FRIENDS OF ANIMALS v. FISH AND WILDLIFE SERVICE (2026)

This case involves tension between development of land and preservation of animal species threatened with extinction. Some of these species colonize on land where development is expected. So when development looms, federal law strikes a balance between the interest in development and survival of the species.

This appeal involves two main sets of issues:

1. Comparing habitats: Development can threaten the survival of animal colonies. But habitats vary in their suitability for colonization. For example, habitats vary in the quality and quantity of vegetation suitable for animals. Can regulators adequately protect animals by moving them without comparing the suitability of their habitats? We answer no.

2. Logistics: When deciding whether to allow development, regulators can't foresee every contingency. For example, regulators may rely on funding sources that unexpectedly vanish. But do regulators adequately plan for these contingencies by drawing on regulatory expertise and experience? We answer yes.

1. The Service establishes a General Conservation Plan for Utah prairie dogs.

These issues arise from regulatory efforts to protect Utah prairie dogs, which live in Utah grasslands. During the twentieth century, the species suffered a massive drop in population from disease, poisoning, and destruction of habitats. In response, the Fish and Wildlife Service listed the species as endangered and later as threatened.¹

These listings triggered the Endangered Species Act and accompanying regulations, which ordinarily bar the taking of endangered or threatened species within the United States. See 16 U.S.C. § 1538(a)(1)(B) (prohibiting the taking of endangered species); 16 U.S.C. § 1533(d) (allowing the Secretary of the Interior to prohibit the taking of threatened species); 50 C.F.R. § 17.31 (establishing a default prohibition against the taking of species that are threatened); see also 16 U.S.C. § 1532(19) (defining take as “to harass, harm, pursue, hunt, shoot, wound, kill, trap, capture, or collect, or to attempt to engage in any such conduct”). Despite this general bar, the Service can issue permits allowing takes that are incidental to conduct that is otherwise legal. Id. § 1539(a)(1)(B). For example, a developer may get a permit to build a school or hospital on land occupied by prairie dogs. Joint App'x vol. 3, at 36.

A permit can be issued only if a developer submits a conservation plan and the Service finds that

• the developer would minimize and mitigate the impact of the taking to the maximum extent practicable,

• the developer would ensure enough funding to minimize and mitigate the impact,

• the taking would not appreciably reduce the likelihood of the survival and recovery of the species, and

• the developer would take any other measures that are necessary or appropriate to implement the plan.

16 U.S.C. § 1539(a)(2)(A)–(B).

In 2018, the Service issued a General Conservation Plan that covered land in seven counties in southwest Utah. The Plan included three “recovery units”—areas that were geographically identifiable and “essential to the conservation and recovery of the entire population of Utah prairie dogs.” Joint App'x vol. 5, at 79.

The Plan was designed to streamline the issuance of permits over a ten-year period. Id. vol. 3, at 30–32, 42. In that period, regulators expected construction on habitats containing large colonies of Utah prairie dogs. Id. at 40.

The Service recognized that developers would focus on just a few counties. For those counties, the Plan would relieve developers of the need to get individual permits. Instead, the Plan would allow the Service to issue a “master permit” to a county. Id. at 35–38. The county could then allow developers to use the master permit by issuing “certificates of inclusion.” Id. at 134–37.

A certificate of inclusion would allow developers to take prairie dogs under conditions specified by the Plan and the county. In places without master permits, the Service could authorize the take of prairie dogs by issuing individual permits to developers. Id. at 36–37.

To predict the potential impact on prairie dogs, the Plan designated two types of areas: (1) major development areas and (2) minor development areas.

Major development areas are lands

• that are owned by someone other than the federal government and

• that are “built out or adjacent to built out areas.”

Id. at 40. Minor development areas are lands that are

• owned by someone other than the federal government and

• less likely to experience significant development over the ten-year period.

Id. at 41. Minor development areas sometimes contain low-quality habitats, which are areas lacking the required biological, ecological, or functional requirements to sustain the species’ life cycle. Alan D. Copsey, Symposium: Guidance for Growth, The Protection of Wildlife Under Washington's Growth Management Act, 16 U. Puget Sound L. Rev. 1101, 1126 n.145 (1993).

The Service used historical averages to project the loss of habitat in major development areas and minor development areas. Joint App'x vol. 3, at 84. But the Service recognized that future development could exceed the historic averages. Id. at 85 (acknowledging that “it is possible that future development would be higher than historical averages, at least in some years, with changes in the economy or human population growth”). So the Service also predicted the impact on prairie dogs if the destruction of their habitats were to increase fivefold. Id. at 85–86.

To address the impact of future development, the Plan supplied two primary approaches for mitigation. The first approach would be used in “major development areas” and “small colonies on low-quality habitats in the minor development areas.” Id. at 58. For these areas, the Plan required

• the movement of prairie dogs from development sites and

• the establishment of new colonies on federal or protected lands.

This movement of prairie dogs is called translocation. Id. at 57–60. For minor development areas with medium- or high-quality habitats, the Plan required protection of habitats through conservation banks, land acquisitions, and conservation easements. Id. at 60–63. (Conservation banks are permanently protected and managed lands, and conservation easements are restrictions in deeds.)

The mitigation measures would be funded by Utah and fees paid by developers. Id. at 106–10. To allow adjustments, the Plan required

• the permittees to monitor the success of mitigation and

• the Service to annually review compliance.

Id. at 76–80, 227.

The Service projected that future development (like construction of schools and hospitals) would destroy habitats, leading to the take of up to 7,152 prairie dogs over 10 years. Id. at 88. But the Service concluded that the mitigation measures would fully offset the impact. Id. at 94, 97. For these conclusions, the Service found

• that the Plan and the proposed permits would not likely jeopardize the continued existence of the Utah prairie dog, id. vol. 2, at 86–123, 157–74,

• that the minimization and mitigation measures would fully offset the take of the prairie dogs, id. at 104–13, 160–68,

• that funding was assured, id. at 168–72, and

• that the Plan would not significantly impact the environment, id. at 175–297.

Based on these findings, the Service issued master permits to three Utah counties (Iron, Garfield, and Beaver), authorizing incidental takes through development. Id. at 151–56; id. vol. 3, at 6–19.

2. Friends of Animals unsuccessfully challenges the Plan in district court.

Friends of Animals petitioned for judicial review, claiming violations of the Endangered Species Act and the National Environmental Policy Act. The district court denied Friends of Animals’ petition for review.

3. Friends of Animals has standing.

The Service challenges Friends of Animals’ standing, arguing that a disconnect exists between the claims and the documents allowing the take. The district court rejected this challenge, and we conduct de novo review. Colo. Outfitters Ass'n v. Hickenlooper, 823 F.3d 537, 544 (10th Cir. 2016).

The district court didn't err. For standing, a plaintiff must show that “(1) it has suffered an ‘injury in fact’ that is (a) concrete and particularized and (b) actual or imminent, not conjectural or hypothetical; (2) the injury is fairly traceable to the challenged action of the defendant; and (3) it is likely, as opposed to merely speculative, that the injury will be redressed by a favorable decision.” Friends of the Earth, Inc. v. Laidlaw Env't Servs. (TOC), Inc., 528 U.S. 167, 180–81, 120 S.Ct. 693, 145 L.Ed.2d 610 (2000) (quoting Lujan v. Defenders of Wildlife, 504 U.S. 555, 560, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992)). An organization has standing only if at least one member has standing. See Summers v. Earth Island Inst., 555 U.S. 488, 498, 129 S.Ct. 1142, 173 L.Ed.2d 1 (2009).

Friends of Animals alleges that the Plan and related permits will impede members’ ability to view Utah prairie dogs. In response, the Service distinguishes between the Plan and the permits authorizing the take, asserting that Friends of Animals has relied on deficiencies in the Plan rather than in the permits themselves. Because permits are required for the take, the Service argues that Friends of Animals can't trace its injury to the Plan itself.

This argument reflects a misunderstanding of the claims: Friends of Animals challenges not only the Plan, but also the permits themselves. See Joint App'x, vol. 1, at 21 (Friends of Animals’ request for the district court to “vacate and remand the decisions approving [the General Conservation Plan] and [incidental take permits]”). Indeed, the Service bases the permits on the Plan; and the terms of the permits incorporate the Plan. Id. vol. 2, at 152 (master permit for Beaver County, incorporating the terms of the Plan); Id. vol. 3, at 8, 15 (master permits for Garfield County and Iron County, incorporating the terms of the Plan). So the alleged injury is traceable to the Plan itself.

The Service disagrees, characterizing the Plan as a programmatic framework that doesn't authorize the take. But the Plan itself allows an incidental take of up to 7,152 prairie dogs. Id. at 88; see also id. at 30 (describing how conservation plans are “required as part of an application for an incidental take permit”). Given the existence of an injury traceable to the Plan, Friends of Animals has standing.

4. We review the Plan under standards established by the Administrative Procedure Act.

The claims are governed by the Administrative Procedure Act. See Defs. of Wildlife v. U.S. Forest Serv., 94 F.4th 1210, 1220 (10th Cir. 2024) (stating that the Administrative Procedure Act governs a claim under the Endangered Species Act); Cure Land, LLC v. U.S. Dep't of Agric., 833 F.3d 1223, 1230 (10th Cir. 2016) (stating that the Administrative Procedure Act governs a claim under the National Environmental Policy Act). Under the Administrative Procedure Act, a court can overturn an agency's decision when it is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A). The decision is arbitrary and capricious if the agency has

• “relied on factors which Congress has not intended it to consider,”

• “entirely failed to consider an important aspect of the problem,”

• “offered an explanation for its decision that runs counter to the evidence before [it],” or

• taken an action “so implausible that it could not be ascribed to a difference in view or the product of agency expertise.”

Motor Vehicle Mfrs. Ass'n of the U.S. v. State Farm Mut. Auto. Ins., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983).

Friends of Animals argues that the Service's issuance of the Plan and related permits was arbitrary and capricious under the Administrative Procedure Act. To address this argument, we conduct de novo review of the agency's decision, applying the same deference to the decision as the district court. W. Watersheds Project v. Haaland, 69 F.4th 689, 700 (10th Cir. 2023). Given this deference, Friends of Animals must show that the Service acted arbitrarily and capriciously. Audubon Soc'y of Greater Denver v. U.S. Army Corps of Eng'rs, 908 F.3d 593, 602–03 (10th Cir. 2018). Our deference is “especially strong” when the decision falls within the agency's expertise in technical and scientific matters. Utah Env't Cong. v. Russell, 518 F.3d 817, 824 (10th Cir. 2008). Despite that deference, we consider only the Service's stated reasons. State Farm, 463 U.S. at 43, 103 S.Ct. 2856. “[P]ost hoc rationalizations” won't suffice. Dep't of Homeland Sec. v. Regents of the Univ. of Cal., 591 U.S. 1, 23, 140 S.Ct. 1891, 207 L.Ed.2d 353 (2020).

5. The Service acted arbitrarily and capriciously in evaluating the Plan's assessment of prairie dog habitats.

For a 10-year period, the Plan allowed a take of up to 7,152 prairie dogs and 1,750 acres of land. The Service could approve this take under the Endangered Species Act only if the Plan had minimized and mitigated the effects “to the maximum extent practicable.” 16 U.S.C. § 1539(a)(2)(B)(ii).

The Service concluded that the Plan had provided adequate mitigation to fully offset these losses of prairie dogs and habitat. Joint App'x vol. 2, at 163. For this conclusion, the Service pointed to its

• use of protocols for translocation and

• consideration of habitat quality when selecting sites for conservation banks, land acquisitions, and conservation easements.

Id.; see also id. vol. 3, at 53 (stating that the Plan was designed to protect “sufficient habitat of appropriate quality to offset impacts from habitat loss”). Friends of Animals challenges this conclusion, arguing that the Service

• should have compared the biological value of lost habitats and new habitats and

• should not have relied on translocation to mitigate the impact of development in major development areas.

We agree with Friends of Animals that the Service arbitrarily and capriciously failed to compare the biological value of habitats. But this failure didn't otherwise prevent the Service from relying on translocation as the preferred mitigation measure in major development areas and low quality habitats in minor development areas.

A. The Service should have compared the biological value of habitats.

For the major development areas and low-quality habitats in minor development areas, the Service regarded translocation as the preferred measure of mitigation to fully offset the losses of prairie dogs. Id. at 58–60. But the Service failed to require the replacement of old habitats with habitats of equivalent biological value.

The Service would translocate the Utah prairie dogs by trapping and moving them to new habitats. Id. at 60, 82; id. vol. 5, at 60–62. And the Service recognized the need to assess the quality of the new habitats for the long-term recovery of the species. Id. vol. 2, at 163; id. vol. 3, at 98. For example, the Service had earlier underscored the importance of open habitats for foraging; well-drained, deep soils; moist herbaceous vegetation; and connectivity of habitats. Id. vol. 5, at 10, 33, 35, 37–38, 53, 71–72.

In comparing the old and new habitats, the Service applied a concept of biological value. Id. at 111–16. This concept recognized the need to fully replace the biological value of habitats through conservation. Id.; see also id. at 111 (the Service's handbook defining fully offset to require replacement of the biological value lost by implementing conservation measures with equivalent biological value).

The Service illustrated the required analysis with loss of a habitat spanning 100 acres. Id. at 113. The loss might be offset with restoration of land for a new habitat. Id. The comparison of the old and new habitats would involve these “key questions”:

[W]hat value did the habitat lost have to the covered species? What value does the replacement habitat have to covered species (e.g., did the replacement habitat provide for the same life stage of the covered species as that lost)? Does the replacement ratio need to be greater than 1:1 to compensate for the lag time between impacts and full eco-function of the replacement habitat, to allow for restoration uncertainties, or is [it] consistent with previously-defined recovery objectives? Is the identified conservation habitat likely to remain suitable in reasonably anticipated future climate scenarios? Is there more value to the species by replacing the habitat that is lost with a different habitat type (e.g. breeding vs. foraging habitat)?

Id.

The Service acknowledges the “key questions” of replacing lost habitat with habitat of equal value. Appellee's Resp. Br. at 23 (stating that “[the Plan's] framework does consider ‘the quality of habitat and colonies’ ” (cleaned up)). And the Service says that the Plan is designed to protect “habitat of appropriate quality. Joint App'x vol. 3, at 53. After acknowledging this goal, however, the Service adopted the Plan without comparing the biological values of the old and new habitats.

The Service protests that until developers seek permits, no one can know the size of the development or select the lands for mitigation. But the Plan authorizes the take without requiring anyone to assess the biological quality of the existing habitat.

This omission might be unimportant if the existing habitats were inferior sites. But that's not the case; the Service acknowledges that major development areas “retain a high level of conservation value based on the high numbers of prairie dogs in these areas and the connectivity that may remain with” prairie dogs in the minor development areas. Id. at 58. For example, the Service points out that the major development areas contain 46 colonies that are considered medium or large. Id. at 97. Given the sizes of these colonies, the Service regards them as valuable contributors to the resiliency of the species. Id. at 50.

Despite this contribution to the resiliency of the species, the Plan does not require the replacement of lost habitats with habitats of equivalent biological value. The Service defends this omission by arguing that (1) regulators will consider the quality of habitats before issuing individual permits, (2) the Plan analyzes where development is expected, and (3) the Plan includes mechanisms to evaluate the success of mitigation and to allow adjustments.

First, the Service suggests that regulators will address habitat quality when a developer applies for an individual permit. In the application, the developer must provide the results of a survey as to Utah prairie dogs, a map of the proposed development and any existing colonies of Utah prairie dogs on the property, a description of the proposed development and the acreage of habitat that is affected, and the identification of measures to minimize and mitigate the loss of habitat prior to construction. Id. at 204, 214. The developer also needs to provide an assessment of habitat quality “where applicable.” Id. But developers in major development areas don't need to assess habitat quality. See id. at 211, 220. So regulators wouldn't have information on habitat quality when reviewing applications for development in the major development areas.

Second, the Service contends that the Plan includes projections about the particular areas for development. Based on these projections, the Service anticipates complete offset of lost habitat by either

• a new colony on 400 acres (based on historic levels of development) or

• 3 new colonies on 1,200 acres (based on a “stepped-up” scenario).

Id. at 99–100. The Service explains that it can't anticipate the places for future development. Id. at 38.

These arguments provide little reason to expect a full offset of lost habitat. Although the plan relied primarily on translocations to offset the loss of habitat, the Service didn't assess the quality of future sites. In fact, the Service acknowledged that some of the existing sites might have greater biological value than the new sites. For example, the Service planned to focus on establishing new colonies or reestablishing colonies on federal lands. See id. at 58–59. Yet the Service also observed that

• the numbers and densities of prairie dogs on private lands were generally higher than on federal lands and

• the federal lands were more arid than private lands.

Id. at 49. The Service thus provides little reason to expect that the Plan would preserve existing biological values.

The Service reasons that major development areas will be jeopardized by development irrespective of the Plan: “Overall, prairie dog colonies in the major development areas do not contribute to our long-term objectives for recovery of the species in the wild, because the existing and anticipated development, combined with loss and fragmentation of habitat in these areas have already or would result in the loss of habitat connectivity and population viability.” Id. at 40. This reasoning is circular because the Plan is what would allow for future development.

Third, the Service argues that it acted reasonably by monitoring the Plan's success, allowing adjustments if habitats worsened with translocations resulting in inferior habitats. Id. at 72–74. But the Service hasn't explained how monitoring will guarantee the replacement of lost habitats with habitats of equivalent biological value. Without an explanation, the Service assumed

• that development was inevitable and

• that monitoring would ensure equivalent biological content of future habitats.

These assumptions were arbitrary and capricious absent a comparison of biological values.

B. Reliance on translocation wasn't arbitrary or capricious.

Apart from the failure to compare biological values, Friends of Animals challenges the Service's reliance on translocation to mitigate the impact of the take in major development areas. For these challenges, Friends of Animals argues that

• translocation often kills prairie dogs,

• the Service used a flawed measure of success for the translocations, and

• the Plan doesn't require translocation.

But these arguments don't undermine the Service's reliance on translocation as a preferred measure of mitigation.

We typically defer to an agency when it applies its expertise in scientific and technical matters. San Juan Citizens All. v. Stiles, 654 F.3d 1038, 1045 (10th Cir. 2011). This deference is appropriate here because the Service used its scientific judgment and experience when selecting translocation as a preferred mitigating measure. Joint App'x vol. 3, at 40, 57–60; see p. ––––, above. For example, the Service explained that “[t]ranslocation effectiveness ha[d] improved across the years and now plays an important role in Utah prairie dog recovery.” Joint App'x vol. 3, at 57; see also id. at 59 (noting that 80% of post-1985 translocation sites are still occupied by prairie dogs, in contrast to 22% of pre-1986 sites); id. at 59 (explaining that “[t]he improvement in translocation success over time is the result of active study and modification of methods that include vegetation treatments, restrictions on movements of certain age and sex categories, shortened holding times prior to release, food and water supplementation, installation of plastic burrows, use of retention cages and nest boxes, and plague prevention dusting”). The Service pointed to two translocation sites as “recent evidence” of effectiveness in keeping prairie dogs on site and protecting them from predators. Id.

Granted, translocation typically results in a low survival rate. For example, the Service estimated a 10% survival rate. But the Service explained that

• this rate could stem in part from other factors, such as dispersal, predation, plague, and the short life spans of individual animals;

• “prairie dogs have relatively short life spans and many of the translocated prairie dogs would not have survived at their home site even absent translocation efforts;” and

• the survival rate might be higher than 10% given the possibility of dispersal following translocations.

Id. at 87.

The Service also acted reasonably in projecting the number of prairie dogs that would be moved. For this projection, the Service reasonably relied on its experience with other habitat protection plans. Id. at 87. The Service acknowledged that some development may take place outside the time-period for translocation (generally July 1 through August 31). See id. at 22, 57. But any development would occur only after the developer had obtained authorization from the Service or a county. Id. at 37–38. And developers seeking individual permits needed to “commit to conducting translocations of prairie dogs” whenever “feasible.” Id. at 222.

It's true that the Service or the county might allow development when translocation isn't feasible. Id. vol. 2, at 288 (acknowledging that the Plan “does not mandate translocations, because translocations may not always be practical given a developer's timeframe”). From past experience, the Service could reasonably project translocation of 71% of the prairie dogs on developing properties. Id. at 288; id. vol. 3, at 87.

Given the Service's expertise and explanation, reliance on translocation wasn't arbitrary or capricious despite questions involving the low survival rate, the measures of success, and the timing of development projects.

6. The Service didn't act arbitrarily or capriciously in evaluating the Plan's approach in minor development areas.

Friends of Animals also questions the mitigation measures in the minor development areas. For these areas, the Service relied on measures involving conservation banks, land acquisitions, and conservation easements. Id. at 60.

Conservation banks are permanently protected lands that are conserved and managed for endangered and threatened species. Id. at 21; see p. ––––, above. The Service listed three conservation banks with reserves where prairie dogs could be relocated. Id. at 61. The Service predicted that these banks could compensate for the loss of 312 acres of habitat. See id. at 101–02.

But under the stepped-up projections, the lost habitat could increase from 312 acres to 612 acres. Id. at 101 (Plan); see also id. vol. 2, at 54 (incidental take permit). So Friends of Animals questions the adequacy of

• conservation banks to offset the loss of habitat and

• standards to evaluate the new habitats.

But these questions don't show arbitrariness or caprice in the Service's selection of mitigation measures in the minor development areas.

The Service recognized that development might outstrip the mitigation potential of conservation banks. If the conservation banks were exhausted, the Service admittedly would need to acquire land and obtain conservation easements. Id. vol. 3, at 62. Granted, those measures might require greater funding. Id. But the Service noted the flexibility in its fee structure. See id. at 177–80 (discussing the options to structure fees for the minor development areas).

The Service also explained that it would consider habitat quality when picking sites for conservation banks, land acquisitions, and conservation easements. Id. at 102. When considering habitat quality, the Service would rely on the Utah prairie dog recovery plan, a paper on population structure, Five-Year Management Plans, and new scientific information. Id. at 60.

Given its expertise, the Service acted reasonably in selecting mitigation measures for the minor development areas. Those measures could require greater funding, but the Service recognized this possibility and planned accordingly.

7. The Service acted arbitrarily and capriciously in concluding that the Plan wouldn't jeopardize survival of the species.

The Service concluded that the Plan wouldn't reduce the likelihood that Utah prairie dogs would survive and recover. See 16 U.S.C. § 1536(a)(2) (requiring federal agencies to consider whether their actions could “jeopardize the continued existence” of a threatened species). In support, the Service reasoned that

• most of the authorized take would take place in the major development areas,

• these takes would be fully offset by the creation of new colonies on federal or protected lands, and

• the losses in minor development areas would be fully offset by the protection of other colonies.

Joint App'x vol. 2, at 172. This reasoning is unsupported.

A. The Service failed to consider the loss of habitat in the major development areas.

Friends of Animals argues that the Service didn't adequately consider the quality and extent of the habitat being lost in the major development areas.

The Service recognized the importance of the major development areas. For example, the Service had adopted a recovery plan in 2012 that

• recognized the heightened risk to prairie dogs from urban development on non-federal lands and

• found that protecting certain habitats in “recovery units” was necessary for the species to survive and recover.

Id. vol. 5, at 74–76, 79 (designating certain areas as “recovery units”), 86–89 (discussing the protection of habitat in those units). The 2018 Plan

• observed that major development areas were located within the recovery units,2

• incorporated the entirety of the 2012 recovery plan, and

• reaffirmed the earlier plan's biological goals.

Id. vol. 3, at 30–32, 45, 49, 52.

Despite incorporating the 2012 recovery plan, the Service adopted the new plan in 2018 without protecting habitats equivalent to those lost in the major development areas. For example, the Service expressed skepticism about the ability of prairie dogs to remain in the major development areas over the long term based on future development and the resulting separation of colonies. Id. vol. 2, at 172; id. vol. 3, at 40. This reasoning could explain the inability to restore past disruptions to colonies, but wouldn't explain a willingness to allow future destruction of habitats in the major development areas.

While downplaying the long-term importance of major development areas, the 2018 Plan recognized the conservation value of these areas because of their colonies of prairie dogs and the connectivity to colonies in the minor development areas. Id. at 58. Given the concentration of prairie dogs in the major development areas and connectivity to colonies in minor development areas, the Service apparently assumed that it could preserve the species while sacrificing existing habitats in the major development areas. But the Service didn't

• assess the quality or quantity of the new sites or

• ensure that the counties would make these assessments before issuing certificates of inclusion.

The Service thus had no way to assure survival and recovery under the Plan.

The Service noted that the major development areas contain 5,200 acres of private land on three recovery units. Id. at 47. These were areas identified in the 2012 recovery plan as “essential to the conservation and recovery of the species.” Id. at 141. The 2012 plan had characterized the recovery units as “individually necessary to conserve the genetic, demographic, and ecological diversity necessary for the long-term sustainability of Utah prairie dogs.” Id. vol. 5, at 79. After underscoring the importance of these units in 2012, the Service adopted the new plan in 2018 despite the expected loss of 1,138 acres. Id. vol. 2, at 106–07. The loss of that habitat could undermine preservation of the species. For example, a 2000 Princeton study had urgently recommended preservation of the species in one of the recovery units (the West Desert Recovery Unit). Id. vol. 4, at 50. But the new plan said nothing about the colonies in this unit.

By failing to explain the change in its approach, the Service couldn't reasonably square the survival and recovery of the species with the large-scale loss of habitat in the major development areas. See Sierra Club v. Marsh, 816 F.2d 1376, 1388–89 (9th Cir. 1987) (concluding that the Endangered Species Act requires an agency to ensure the creation of a new refuge before allowing modification of an existing habitat), abrogated in part on other grounds as recognized in Cottonwood Env't L. Ctr. v. U.S. Forest Serv., 789 F.3d 1075, 1088–91 (9th Cir. 2015).

B. The Service didn't act arbitrarily or capriciously by relying on scientific studies and various time-periods.

Friends of Animals also questions the reliance on a 2006 study and the use of multiple time-periods for different purposes.

We afford substantial deference to the Service's expertise in technical and scientific matters. See p. ––––, above. Exercising that expertise, the Service pointed to a 2006 study involving the habitat and population of prairie dogs. Joint App'x vol. 2, at 116. This study had concluded that regulators should take a percentage of a species’ population (rather than a fixed number of animals) to maintain the species and prevent extinction. Id.

Friends of Animals gives three criticisms of the Service's reliance on the study:

1. The study addressed take from the shooting of prairie dogs rather than the destruction of habitat.

2. The study involved a different species of prairie dogs that wasn't threatened with extinction by the loss and fragmentation of habitats.

3. The Plan used a quota even though the study had preferred the use of percentages.

We reject these criticisms because Friends of Animals misunderstands the context for the agency's reliance on the study.

The Service referred to the 2006 study to explain the overall impact of activities on the population of prairie dogs. Id. This study had addressed the take of prairie dogs based on a percentage of the known population, concluding that use of a percentage could “help ensure maintenance of a sustainable population, with no risk of extinction.” Id. Pointing to this study, the Service reasoned that the allowable take would not jeopardize the survival of Utah prairie dogs. Id.

The Service recognized that the 2006 study had involved a different species of prairie dog. But the Service noted that the 2006 study had shown the ability to preserve this species with takes exceeding those under the Plan. Id. (anticipating an overall take of 17% under the Plan, compared to a sustainable take of 20–25% of the black-tailed prairie dog population in the 2006 study). So the Service didn't act arbitrarily or capriciously by using the 2006 study.³

Friends of Animals challenges not only the Service's reliance on this study, but also the use of different time-periods to quantify the impact of development. For the impact on habitat, the Service considered data for Iron County from 2014 to 2017. Id. vol. 3, at 84–86. And for the impact on populations of prairie dogs, the Service relied on data from 2010 to 2014. Id. at 50. For the percentage of prairie dogs being translocated, the Service used data from three other years (1998, 2001, and 2002). Id. at 87.

The Service explained its use of different time-periods. For the impact on habitat, the Service explained that

• development had increased from 2014 to 2017 and

• the agency had not “want[ed] to underestimate” the loss of occupied habitat.

Id. at 84. For translocations, the Service referred to 1998, 2001, and 2002 because translocations had peaked in these years and the Plan had put a premium on translocation. Id. at 87.

For population density, the Service used the 2010–14 period without explanation. But Friends of Animals doesn't say why use of this time-period would undermine the Service's conclusion. See Rocky Mountain Wild v. Dallas, 98 F.4th 1263, 1290 (10th Cir. 2024) (“[A]n agency's violation of the [Administrative Procedure Act] ‘does not require reversal unless the appellant demonstrates prejudice resulting from the error.’ ” (quoting WildEarth Guardians v. Nat'l Park Serv., 703 F.3d 1178, 1183 (10th Cir. 2013))).

Friends of Animals argues that the Service chose a time-period that would artificially reduce the estimated take. But Friends of Animals doesn't explain

• how a different time-period would affect the estimated take or

• why the choice of data would affect a projection about the ultimate survival and recovery of the species.

Moreover, Friends of Animals didn't preserve this argument in the administrative proceedings. See W. Watersheds Project v. U.S. Bureau of Land Mgmt., 76 F.4th 1286, 1293 (10th Cir. 2023) (“If a party fails to raise its objection, it has failed to exhaust its administrative remedies and has consequently forfeited its argument.”).

So we reject Friends of Animals’ challenges to the Service's reliance on the 2006 study and the data from different time-periods.

8. The Service didn't act arbitrarily or capriciously in finding adequate funding.

The Service found adequate funding for the Plan, relying on the assessment of mitigation costs. Joint App'x vol. 2, at 168–72; id. vol. 3, at 106–11.

Mitigation in major development areas and low-quality habitats in minor development areas would consist of translocation of prairie dogs, habitat treatments, plague management, and establishment of new sites. Id. To fund these mitigation measures, the Service entered into an implementation agreement with the Utah Department of Natural Resources, the Utah Division of Wildlife Resources, three Utah counties, the U.S. Forest Service, and the U.S. Bureau of Land Management. Id. vol. 2, at 124–50. Under this agreement, the state agencies would “[p]rovide funding, as available to assist with establishment of new translocation sites on Federal or protected lands, the translocation of animals to translocation sites, and plague management.” Id. at 130. For mitigation costs elsewhere in the minor development areas, the Service would rely largely on fees paid by developers. Id. at 169. To determine the amount of fees, the Service considered the actual take of prairie dogs in the minor development areas over the past five years. Based on the existing fee for available conservation banks ($7,000 per acre), the Service projected a need to raise $77,420. Id. at 169–70; id. vol. 3, at 108. With that projection, the Service adopted a fee schedule expected to generate $93,419. Id.

Friends of Animals makes four challenges to the Service's projection as to future funding:

1. The Service assumed that development would occur primarily in major development areas, but the Plan didn't apportion permits based on the type of area.

2. Utah agreed to fund mitigation in certain areas only if funding were available.

3. The Service failed to consider Utah's cost for translocations.

4. The Service couldn't guarantee funding for habitat protection in minor development areas.

We reject these challenges.

First, the Service could rely on its experience in managing habitats without apportioning permits between the major and minor development areas. Given that experience, we defer to the Service's conclusion that fees from developers would cover the mitigation costs in minor development areas, with Utah to cover the mitigation costs elsewhere.

Second, the Utah agencies committed to fund mitigation in major development areas and in minor development areas with small colonies and low-quality habitat. Though the commitment was qualified by the availability of funds, the Service could again rely on its experience. For example, Utah had contributed an average of $340,000 every year from 2005 to 2018 for the conservation of Utah prairie dogs. Id. vol. 2, at 291–92.

The Service recognized that Utah law prevented agencies from incurring obligations based on an expectation of future appropriations. Id. vol. 3, at 107 (citing Utah Code Ann. § 63G–6a–1204). But Utah had

• spent decades using agencies to recover losses in the species and

• promised that future funding would depend “on the cost of offsetting the annual loss of occupied Utah prairie dog habitat,” Joint App'x vol. 2, at 130.

In these circumstances, the Service reasonably determined that Utah agencies would continue to provide enough funding for translocations.

Third, Friends of Animals contends that the Service

• underestimated the cost of required translocations under the Plan and

• didn't account for the cost of other translocations associated with the recovery of prairie dogs.

For these contentions, Friends of Animals highlights

• a possibility that translocation cost might surpass Utah's funding and

• a potential need for more funds to acquire land.

Despite these concerns, Friends of Animals hasn't undermined the Service projections.

These projections included the annual amount needed to translocate 1,500 prairie dogs ($408,596), Id. vol. 3, at 182, and Friends of Animals contends that this amount dwarfs what the Service had been able to get in the past. But the Service's projection of $408,596 included translocations conducted separately from the Plan. Id. at 89–90 (discussing a separate program involving translocations from properties that were already developed or adjacent to developed properties). And the Service didn't need to ensure funding for the translocations from properties that were already developed or adjacent to properties already developed because

• these translocations were optional as mitigation measures, see id. at 111, and

• the Plan provided that programs for those translocations would be developed separately by Utah or the counties.

Id. at 111 (stating that “[t]ranslocations independent of development are an optional program under [the Plan],” that “this portion of the [Plan] [did] not need to meet the assured funding criterion,” and that “[p]rograms for translocations independent of development would be developed by the State or counties”).⁴

The Service also didn't skirt the potential need for funding to obtain land to translocate the prairie dogs from major development areas. The Service

• projected that translocation sites would “typically be located on Federal lands,” id. vol. 2, at 169; see also id. vol. 3, at 58 (stating that translocations would occur “primarily on Federal lands”); id. at 59 (stating that “[t]he focus of this mitigation effort would be on Federal lands”), and

• acknowledged federal agencies’ “long track record of over 30 years ․ of managing prairie dogs and establishing translocations sites for the species on [federal lands],” id. vol. 2, at 169.

So the Service could reasonably assume that it wouldn't need to buy land for most of the translocations.⁵

To be sure, the Service indicated that

• it could establish “new colonies” on protected land and

• the land would come from conservation banks, conservation easements, or land acquisitions.

Id. at 59. So establishing translocation sites on protected lands could require funding. But the Utah agencies committed to provide funds for new translocation sites, and the Plan allowed adjustments of future contributions in light of “changed circumstances.” Id. at 68–69, 71–72.

Friends of Animals also asserts that the Service

• needed four translocation sites and

• estimated the cost for only two.

But Friends of Animals didn't make that assertion in district court or request plain-error review. So we don't consider this assertion. See United States v. Leffler, 942 F.3d 1192, 1196 (10th Cir. 2019).

In its reply brief, Friends of Animals also questions the Service's reliance on the expense to translocate 1,500 prairie dogs, arguing that this estimate rests on a miscalculation of the number of prairie dogs that could be translocated to each site. Joint App'x vol. 3, at 88. We don't consider that argument because it came too late. See United States v. Leffler, 942 F.3d 1192, 1197 (10th Cir. 2019) (“In this Circuit, we generally do not consider arguments made for the first time on appeal in an appellant's reply brief and deem those arguments waived.”).

Fourth, for the minor development areas, the Service could reasonably rely on conservation banks, conservation easements, and land acquisitions. Friends of Animals questions this reliance based on the inability of existing conservation banks to fully offset the loss of 612 acres of occupied habitat in minor development areas. See Joint App'x vol. 3, at 101 (projecting that between “35 to 612 [acres] of occupied Utah prairie dog habitat may be impacted in the minor development areas over the 10-year term of this [Plan]”).

But the Service acted reasonably by

• projecting the amount needed to acquire new habitats and

• providing flexibility in the fee schedule to adjust for unexpected costs.

The Service considered the annual acreage of occupied habitat lost over the prior five years (2.2 acres per year). Id. at 108. If that rate were to continue, the Service could reasonably project a full offset of the lost habitat.

But the Service designed a fee structure to allow mitigation even if development were to exceed projections. The structure provided funding that exceeded the amount needed for mitigation for 2012–2016. Id. at 108. In addition, the Service preserved flexibility if conservation banks couldn't offset the losses of habitats. Id. at 71–72 (identifying a “[l]ack of available conservation banks and therefore the need to acquire conservation properties” as a changed circumstance that could result in a fee increase).

Friends of Animals argues that fee increases wouldn't provide a solution absent coverage of past costs. But the Plan provided flexibility by allowing for periodic fee adjustments to address unexpected costs. See id. at 71. In addition, Utah agencies could adjust fees and contributions to cover shortfalls. Id. at 72.

We thus reject Friends of Animals’ challenges to the Plan's provisions for funding.⁶

9. The Service's environmental analysis was flawed under the National Environmental Policy Act.

Finally, Friends of Animals argues that the Service violated the National Environmental Policy Act by

• failing to consider the impact on habitats,

• making unsupported assumptions about future development, and

• neglecting to consider reasonable alternatives.

For the reasons discussed above, we agree that the Service failed to consider the impact on habitats. But we reject Friends of Animals’ two other arguments.

A. The Service should have considered the impact on habitats.

The National Environmental Policy Act requires federal agencies to take a “hard look” at environmental impacts when making decisions. Wyoming v. U.S. Dep't of Agric., 661 F.3d 1209, 1263 (10th Cir. 2011) (quoting Citizens’ Comm. to Save Our Canyons v. Krueger, 513 F.3d 1169, 1179 (10th Cir. 2008)); see 42 U.S.C. § 4332(2)(C). The Act “ensures that the agency, in reaching its decision, will have available, and will carefully consider, detailed information concerning significant environmental impacts.” Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 349, 109 S.Ct. 1835, 104 L.Ed.2d 351 (1989).

To gauge the environmental impact, an agency must prepare an environmental impact statement when proposing a major federal action that would significantly impact the environment. 42 U.S.C. § 4332(2)(C). If the impact is uncertain, the agency must prepare an environmental assessment. See 40 C.F.R. §§ 1501.3, 1501.4, 1508.9 (2018).⁷ If the environmental assessment shows no significant impact, the agency can proceed without an environmental impact statement. See 40 C.F.R. § 1501.4(e) (2018); 40 C.F.R § 1508.13 (2018).

So the Service needed to assess the significance of the Plan's impact on the environment. This assessment required consideration of the “context and intensity” of actions under the Plan. 40 C.F.R. § 1508.27 (2018). For this assessment, the Service needed to consider factors such as

• “impacts that may be both beneficial and adverse,”

• the extent of controversy over the effects,

• the degree of certainty regarding the effects, and

• the potential effect on endangered and threatened species.

40 C.F.R. § 1508.27(b)(1), (4)–(5), (9) (2018).

These factors could trigger a need for an environmental impact statement. “[B]ut the simple existence of an effect does not trigger that obligation—the ‘relevant analysis is the degree to which the proposed action affects’ a listed factor.” WildEarth Guardians v. Conner, 920 F.3d 1245, 1261 (10th Cir. 2019) (quoting Hillsdale Env't Loss Prevent'n, Inc. v. U.S. Army Corps of Eng'rs, 702 F.3d 1156, 1180 (10th Cir. 2012) (emphasis in original)).

We must give “substantial deference” to the Service's consideration of the need for an environmental impact statement. Seven Cnty. Infrastructure Coal. v. Eagle Cnty., 605 U.S. 168, 180–81, 145 S.Ct. 1497, 221 L.Ed.2d 820 (2025). Given that deference, we consider only whether the Service's assessment “fall[s] within a broad zone of reasonableness.” Id. at 183, 145 S.Ct. 1497; see also id. at 182, 145 S.Ct. 1497 (requiring courts to defer to agencies’ decisions about “how far to go in considering indirect environmental effects”). And even if the Service's analysis had been flawed, we need not vacate the decision absent evidence that a correction might have led to disapproval of the project. Id. at 185, 145 S.Ct. 1497.

The Service concluded that it didn't need an environmental impact statement because the environmental effect was insignificant. Joint App'x vol. 2, at 178 (citing the existing version of 40 C.F.R. § 1508.27). For this conclusion, the Service assumed that the expected take wouldn't reduce the population of Utah prairie dogs. See, e.g., id. at 179.

That assumption informed the Service's assessment of the environmental impact. For instance, the Service needed to consider whether the Plan would imperil a threatened species. 40 C.F.R. § 1508.27(b)(9) (2018). For this inquiry the Service acknowledged the possible impact on Utah prairie dogs, but concluded that the Plan wouldn't significantly affect the population. Joint App'x vol. 2, at 179, 182–83. Similarly, the Service concluded that “there was no data presented to support claims that the [Plan] would jeopardize the species or would otherwise have significant impacts for the species.” Id. at 182; see 40 C.F.R. § 1508.27(b)(4) (2018) (requiring consideration of whether anticipated effects are likely to be “highly controversial”).

For this conclusion, the Service assumed that the impacts of the take would be fully offset through mitigation measures. Joint App'x vol. 2, at 179. Despite that assumption, the Plan authorized the take without requiring the replacement of habitats lost with habitats of equal biological value. See pp. –––– – ––––, –––– – ––––, above. So the Plan didn't provide a meaningful reason to expect a full offset of the anticipated losses of prairie dogs and habitat.⁸

Similarly, the Service assumed that the population of Utah prairie dogs wouldn't be affected by the loss of critical habitats. This assumption is irrational because the Service didn't require anyone to assess the quality or quantity of new sites for the prairie dogs. See id., above. Absent such an assessment at any stage, the Service couldn't reasonably evaluate the significance of the environmental impact.

When an agency relies on an irrational assumption, the analysis is arbitrary and capricious. See WildEarth Guardians v. U.S. Bureau of Land Mgmt., 870 F.3d 1222, 1236 (10th Cir. 2017). Relevant factors include

• whether the assumption was “key to the ultimate decision,”

• whether the agency's analysis underestimated the environmental effects, and

• whether the agency's decision involved “ ‘simple findings of fact’ ” or instead implicated “ ‘the frontiers of science.’ ”

Id. (quoting Baltimore Gas & Elec. Co. v. Nat. Res. Def. Council, Inc., 462 U.S. 87, 103, 103 S.Ct. 2246, 76 L.Ed.2d 437 (1983)).

Together, these factors show that the Service's analysis was arbitrary and capricious. First, the Service's assumption (that the mitigation measures would fully offset the impacts without comparing the biological value of old and new habitats) was critical to the ultimate assessment of the environmental impact as insignificant. Second, the Service underestimated the effects of the Plan on the environment, and this assumption may have affected the assessment of environmental impact. Third, the Service's analysis didn't involve a “scientific frontier.” See id. at 1237 (understanding that term to refer to “barely emergent knowledge and technology”).

So the Service's environmental assessment did not satisfy the National Environmental Policy Act.

B. The Service didn't err in making assumptions about future development.

Friends of Animals also faults the Service for making uncertain assumptions to forecast the loss of habitat and prairie dogs. But the Service's assumptions were reasonable.

The Service assumed that over a 10-year period, development wouldn't destroy more than 1,750 acres of occupied habitat or 7,152 prairie dogs. For this assumption, the Service considered

• data from Iron County, which was “the area within the range of the species that historically has experienced the most development,” Joint App'x vol. 3, at 83, and

• data from the prior four years (2014–2017) because that period “represent[ed] an increase in development from previous years,” Id. at 84.

To extrapolate from the Iron County data to the species’ entire range, the Service relied on a 2011 projection of habitat loss in three areas. Based on that projection, the Service estimated that the habitat loss in Iron County would account for roughly 73% of the total loss of habitat. Because the annual loss in Iron County alone was 25.5 acres, the Service expected future development to result in the annual destruction of 34.9 acres of habitat.

To calculate the affected population in the areas with lost habitat, the Service considered three different parts of the species’ range. These projections were further broken down to consider the amount of habitat loss expected in major and minor development areas. Id. (noting that 90% of the expected loss would occur in major development areas and 10% would occur in minor development areas). The Service also considered the average density of prairie dogs in the three different parts, again broken down based on major and minor development areas. Using the data on habitat loss and density, the Service

• estimated the total take of prairie dogs from development based on historical averages, id. at 85, and

• adjusted this estimate to account for mitigation involving translocation, id. at 87–88 (predicting incidental take after adjusting for minimization measures of 1,594 prairie dogs over ten years).

The Service explained that it was reasonable to rely on Iron County's historical rates because of a projected decrease in the rate of population growth. Id. at 83. Still, the Service included a “stepped-up estimate” to project the impact if development were to exceed past rates. The Service obtained this estimate by multiplying the past loss of habitat by five. Id. at 85–86 (explaining that the stepped-up estimate was 175 acres of habitat lost annually, or 1,750 total); see p. ––––, above. The resulting estimate was the loss of 7,152 prairie dogs.⁹ Friends of Animals challenges this approach for two reasons.

The first criticism involves the Service's reliance on Iron County's data from 2014 to 2017. Friends of Animals argues that in this time-period, Iron County had stricter limits on take than those in the Plan. As an example, Friends of Animals argues that

• Iron County had capped the annual take of prairie dogs at 15 and

• the Plan allowed the take of 7,152 prairie dogs over 10 years.

Regardless of the difference in takes allowed, the Service could

• reasonably rely on data from Iron County to estimate the loss of habitats and

• extrapolate the expected loss of prairie dogs based on this data

Moreover, the Plan requires translocations and they had earlier been voluntary in Iron County. Id. at 87 (noting that an average 20% of prairie dogs on developing properties were translocated in Iron County and projecting the translocation of 71% under the Plan). In addition, the Plan estimated a take of 7,152 prairie dogs only if development were to exceed expectations.

Friends of Animals also challenges the Service's decision to reach a stepped-up estimate by using a multiplier of five. But Friends of Animals cites nothing to suggest an underestimation of lost habitat under the stepped-up estimate. And the Plan required periodic reevaluations of development and the related loss of habitat. Id. vol. 2, at 118; see also id. vol. 3, at 72 (noting that the Service would evaluate mitigation success every 3 years and amend mitigation measures as necessary).

In sum, the Service hasn't acted arbitrarily and capriciously in relying on Iron County's historical data.¹⁰ See New Mexico ex rel. Richardson v. Bureau of Land Mgmt., 565 F.3d 683, 704 (10th Cir. 2009) (“Deficiencies in an [environmental analysis] that are mere ‘flyspecks’ and do not defeat [the National Environmental Policy Act's] goals of informed decisionmaking and informed public comment will not lead to reversal.”).

C. The Service didn't fail to consider possible alternatives.

Friends of Animals also contends that the Service should have considered alternatives such as

• seasonal restrictions on take during the breeding and young-rearing season,

• “practicable measures identified in previous [habitat conservation plans],”

• limits on the areas where take could occur and the amount of take, and

• preservation of “high-quality habitat, important colonies, and connective corridors.”

Joint App'x vol. 2, at 84.

The Service needed to include a “brief discussion” of “alternatives.” 40 C.F.R. § 1508.9 (2018). But the Service didn't need to consider

• the impact of alternatives rejected in good faith “ ‘as too remote, speculative, or impractical or ineffective,’ ” or

• an alternative “unless it [was] significantly distinguishable from the alternatives already considered.”

New Mexico ex rel. Richardson v. Bureau of Land Mgmt., 565 F.3d 683, 708–09 (10th Cir. 2009) (quoting Colo. Env't Coal. v. Dombeck, 185 F.3d 1162, 1174 (10th Cir. 1999)).

In preparing the Plan, the Service tried to “meet the economic development needs of the local communities ․ while ensuring a strategic conservation approach for the species.” Joint App'x vol. 2, at 198. Friends of Animals hasn't shown how its proposed alternatives would accomplish these objectives.¹¹ Absent such a showing, the Service didn't act arbitrarily and capriciously by failing to consider these alternatives.

10. The district court should determine the appropriate remedy in the first instance.

Because the Service violated the Endangered Species Act and the National Environmental Policy Act, we must decide the appropriate remedy: vacatur or remand. Friends of Animals requests that we vacate the Plan and the related permits. The Service responds that if we decline to uphold the Plan, we should direct the district court to decide the appropriate remedy.

The decision between vacatur and remand turns on

• the seriousness of the error and the likelihood that the Service can justify its decision on remand and

• the disruptive consequences of vacatur of the Plan and the permits.

See Ctr. for Biological Diversity v. U.S. Env'l Prot. Agency, 149 F.4th 1142, 1152 (10th Cir. 2025). The parties haven't briefed these issues, and they involve a “fact-intensive inquiry” typically left to the discretion of the district court. Diné Citizens Against Ruining Our Env't v. Haaland, 59 F.4th 1016, 1049 (10th Cir. 2023). So we leave the remedy for the district court to decide in the first instance. See Devon Energy Prod. Co. v. U.S. Dep't of the Interior, 174 F.4th 734, 744 (10th Cir. 2026) (leaving the decision between vacatur and remand to the district court given the fact-intensive nature of the inquiry).

* * *

We reverse the district court's decision on Friends of Animals’ claims under the Endangered Species Act and the National Environmental Policy Act. With this reversal, we remand for further proceedings consistent with this opinion.

FOOTNOTES

1.   The Endangered Species Act defines an endangered species as a species, other than insects considered as pests, “which is in danger of extinction throughout all or a significant portion of its range.” 16 U.S.C. § 1532(6). A species is threatened when it “is likely to become an endangered species within the foreseeable future throughout all or a significant portion of its range.” 16 U.S.C. § 1532(20).

2.   A study noted the importance of preserving existing colonies on private land in the West Desert Recovery Unit. Joint App'x vol. 4, at 50.

3.   The Service also relied on its experience and the conservation measures associated with the Plan. Joint App'x vol. 2, at 116–17 (explaining that population of the species had “remained stable to increasing” under past plans authorizing takes for development).

4.   Separate funding arrangements existed for these translocations. See Joint App'x vol. 2, at 131 (committing the Utah Division of Wildlife Resources to develop a translocation program with funding assurances); id. vol. 3, at 66 (similar). And the Plan explained that the Service could make changes if a shortfall existed for translocations on land already developed or on adjacent properties. Id. at 67; see also id. vol. 2, at 131 (committing the Utah agency to “monitor the success of the translocations independent of development program” and “make changes to the program if needed to ensure success”).

5.   Friends of Animals questions whether federal lands could provide suitable sites for translocation because they• are often more arid and may be inferior to private land, see Joint App'x vol. 3, at 49 and• can be subject to uses that threaten prairie dogs and their habitats.Given the Service's expertise and experience, we defer to its reliance on federal lands as the primary sites for translocation. See p. ––––, above.

6.   Friends of Animals also argues that the Service should have considered other alternatives. The Service responds that consideration of alternatives is not required if the Plan meets the statutory requirements. See 16 U.S.C. § 1539(a)(2)(B). We need not address this disagreement given our remand. The district court should address this issue in the first instance.

7.   The federal government has rescinded its regulations implementing the National Environmental Policy Act. See Removal of Nat'l Envtl. Policy Act Implementing Reguls., 90 Fed. Reg. 10610 (Feb. 25, 2025). But the cited regulation was in effect when the Service approved the Plan. See Audubon Soc'y of Portland v. Haaland, 40 F.4th 967, 980 n.3 (9th Cir. 2022) (explaining that the court applies the regulation in effect when the agency made the decision under review).

8.   In emphasizing the uncertainty surrounding the Plan's mitigation measures, Friends of Animals refers to the Plan's complexity, including the need to consult private landowners. But Friends of Animals doesn't base the need for an environmental impact statement on the complexity of the Plan.

9.   For the stepped-up estimate, the Service also assumed that the total take possible in minor development areas would increase from 10 percent to 35 percent. Joint App'x vol. 3, at 86. That assumption informed the Service's projection of the take from the loss of habitat. Id.

10.   Friends of Animals hasn't challenged the Service's assumptions that (1) 71% of prairie dogs from occupied habitats lost to development would be translocated and (2) 10% of translocated prairie dogs would survive at translocation sites. See Joint App'x vol. 3, at 87–88. So we need not consider whether those assumptions tainted the Service's environmental analysis.

11.   In a reply brief, Friends of Animals argues that the Service dismissed the alternatives by overestimating the value of mitigation measures. But this argument came too late. See United States v. Leffler, 942 F.3d 1192, 1197 (10th Cir. 2019)

BACHARACH, Circuit Judge.