WildEarth Guardians v. National Park Service

TYMKOVICH, Circuit Judge. This appeal concerns WildEarth Guardians’ challenge to the National Park Service’s elk and vegetation management plan for Rocky Mountain National Park. Wild-Earth filed suit in federal district court challenging the plan and the final environmental impact statement the National Park Service (NPS) prepared in conjunction with the plan. WildEarth contends the NPS violated the National Environmental Policy Act (NEPA) by failing to include the reintroduction of a naturally reproducing wolf population as one of the alternatives considered in the environmental impact statement. WildEarth also challenges the agency’s proposal to allow volunteers to assist the agency in reducing the elk population. The district court affirmed the agency action, and WildEarth appealed. We find the record supports the agency’s decision to exclude consideration of a natural wolf alternative from its environmental impact statement. We also find the agency’s interpretation of the National Parks Organic Act and Rocky Mountain National Park Enabling Act persuasive, and that its elk management plan does not violate those statutes. Accordingly, exercising jurisdiction under 28 U.S.C. § 1291, we AFFIRM. I. Background Rocky Mountain National Park (RMNP), located in northern Colorado, was established in 1915. The Rocky Mountain National Park Enabling Act (RMNP Act) bans hunting or killing wild *1181 life within the park, with very limited exceptions. The park has always had a substantial elk population. But most elk predators, especially wolves and grizzly bears, were exterminated in the park area prior to its establishment, and Congress’s decision to ban hunting in RMNP allowed the park’s elk population to grow without constraint. In the 1930s, the National Park Service (NPS) became concerned that the growing number of elk threatened the park’s vegetation through overgrazing. In 1944, the NPS began to control the number of elk by relocating or killing them. This practice was the norm until 1969, when the NPS changed its elk management policy. The agency theorized that increased hunting in the areas around RMNP would sufficiently control the elk population, as elk tend to wander in and out of the park. This policy was not successful, however, as commercial and residential development near RMNP decreased the number of open spaces where hunting was allowed and RMNP’s elk became habituated to residential areas. As a consequence, the number of elk in RMNP has more than tripled since 1969. Several studies conducted in the 1990s found that the park’s elk population is substantially larger, more sedentary, and more concentrated than it would be under natural conditions. As a result, elk overgraze much of the park’s vegetation, eliminating some plant species and making it difficult for others to regenerate. In response, the NPS decided it needed a new elk management policy for the park, both to reduce the overall number of elk and to make the population fluctuate from year to year, as would occur under natural conditions. The NPS expected this would also have a beneficial effect on the park’s vegetation. In August 2002, the NPS assembled an interagency planning team to develop a new elk management plan. The participating agencies included the United States Forest Service, the Colorado Division of Wildlife (CDOW), 1 and several nearby counties and municipalities, with the NPS designated as the lead agency. In May 2003, the NPS published a notice in the Federal Register of its intent to prepare a new elk and vegetation management plan for RMNP and an environmental impact statement (EIS) for the plan. 68 Fed.Reg. 32,084-02 (May 29, 2004). The NPS solicited public comments through a variety of channels, including newsletters, a website, and public meetings. The NPS received around 1,100 public comments on its proposal, which it used to develop a preliminary draft of alternatives for the management plan. In July 2004, the agency publicly released these draft alternatives. One of the proposed alternatives was reintroducing a self-sustaining wolf population to RMNP (the natural wolf alternative). The NPS convened a meeting of biologists and other experts in March 2005 to discuss the feasibility of the natural wolf alternative. And once again, the agency sought public comments on the proposed alternatives. Based on the second round of public comments and feedback from its experts, the NPS selected four alternative plans for analysis in an EIS. In a publicly released August 2005 newsletter discussing these alternatives, the NPS announced it would analyze the introduction of a small number of intensively managed wolves into the *1182 park, in conjunction with the use of sharpshooters, but would not include the natural wolf alternative in its EIS. The agency explained this alternative was infeasible due to lack of support from coordinating agencies, concerns by neighboring communities, the high potential for human-wolf conflicts, and the likelihood that management of wolves in the park would be expensive and time-consuming, distracting from the goal of the NPS’s plan — managing elk. In April 2006, the NPS publicly released a draft EIS that considered five alternative management plans: (1) the current plan (the no-action alternative); (2) rapid reduction of the elk population, which the agency identified as its preferred alternative; (3) gradual reduction of the elk population; (4) a combination of managed killing and elk contraception; and (5) a combination of managed killing and the introduction of a small number of intensively managed gray wolves. 71 Fed. Reg. 20,414-03 (Apr. 20, 2006). The draft EIS reiterated the NPS’s reasons for excluding the natural wolf alternative. The NPS again sought public comment on its draft EIS and held several public meetings during the comment period. The agency then considered the more than 3,100 comments it had received and prepared a final EIS. The agency released its final EIS in December 2007. 72 Fed.Reg. 70,342-01 (Dec. 11, 2007). Although the agency had identified rapid reduction as its preferred alternative in the draft EIS, the final EIS selected a different alternative, gradual reduction. The final EIS also made a small but important change — expanding the definition of those who could assist the NPS with killing elk to include qualified volunteers. The final EIS also took pains to distinguish killing elk for management purposes, which it called culling, from hunting. The final EIS defined “culling” as a highly controlled method for managing an elk population and “hunting” as a loosely regulated recreational activity. After the final EIS was released, Wild-Earth sought judicial review of the NPS’s decision. WildEarth alleged the NPS acted arbitrarily and capriciously by excluding consideration of the natural wolf alternative from its EIS. WildEarth also alleged the NPS’s decision to allow volunteers to participate in culling activities was tantamount to hunting, and violated the RMNP Act. The district court entered judgment for the NPS, concluding the agency took a hard look at the relevant data and articulated a rational connection between that data and its conclusion that the natural wolf alternative was infeasible. The court also found the agency’s distinction between hunting and culling was reasonable, and that the use of volunteers to assist in culling activities did not violate the RMNP Act. II. Discussion A. Standard of Review We give no deference to a district court’s review of agency action, reviewing its decision de novo. Forest Guardians v. U.S. Fisk & Wildlife Serv., 611 F.3d 692, 710-11 (10th Cir.2010). But our review of the NPS’s actions is considerably more deferential. We review the NPS’s compliance with NEPA under the Administrative Procedure Act (APA), which authorizes us to set aside agency action only when it is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” Prairie Band Pottawatomie Nation v. Fed. Highway Admin., 684 F.3d 1002, 1008 (10th Cir.2012). When reviewing agency action, our task is to ensure the agency examined the relevant data and articulated a rational *1183 connection between that data and its decision. Citizens’ Comm, to Save Our Canyons v. Krueger, 513 F.3d 1169, 1176 (10th Cir.2008); FCC v. Fox TV. Stations, Inc., 556 U.S. 502, 513, 129 S.Ct. 1800, 173 L.Ed.2d 738 (2009). Our deference to the agency is more substantial when the challenged decision involves technical or scientific matters within the agency’s area of expertise. Morris v. U.S. Nuclear Regulatory Comm’n, 598 F.3d 677, 691 (10th Cir.), cert. denied, — U.S. -, 131 S.Ct. 602, 178 L.Ed.2d 436 (2010). Accordingly, we will not set aside an agency’s decision unless: the agency (1) entirely failed to consider an important aspect of the problem, (2) offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise, (3) failed to base its decision on consideration of the relevant factors, or (4) made a clear error of judgment. Forest Guardians, 611 F.3d at 711. “Deficiencies in an EIS that are mere ‘flyspecks’ and do not defeat NEPA’s goals of informed decisionmaking and informed public comment will not lead to reversal.” New Mexico ex rel. Richardson v. BLM, 565 F.3d 683, 704 (10th Cir.2009). And even if an agency violates the APA, this does not require reversal unless the appellant demonstrates prejudice resulting from the error. Prairie Band, 684 F.3d at 1008. As these principles imply, a “presumption of validity attaches to the agency action and the burden of proof rests with the appellants who challenge such action.” New Mexico, 565 F.3d at 704. B. NEPA WildEarth’s sole NEPA claim is that the NPS deviated from NEPA’s required procedure by declining to consider the natural wolf alternative in its environmental impact statement. WildEarth argues the wolf alternative fit the purpose and need of the proposed action, and thus required the NPS to consider it in an EIS. Agencies must consider alternatives to any project that might have a significant effect on the quality of the human environment. 42 U.S.C. § 4332(2)(C)(iii). But agencies need not consider every possible alternative to a proposed action, only “reasonable” alternatives. 40 C.F.R. § 1502.14(a); New Mexico, 565 F.3d at 703. A “rule of reason” applies to an agency’s decision to prepare an EIS, as well as the agency’s choice of alternatives to include in its analysis. DOT v. Public Citizen, 541 U.S. 752, 767, 124 S.Ct. 2204, 159 L.Ed.2d 60 (2004). In other words, agencies are not required to consider alternatives they have “in good faith rejected as too remote, speculative, or ... impractical or ineffective.” Custer County Action Ass’n v. Garvey, 256 F.3d 1024, 1039 (10th Cir.2001). “Alternatives that do not accomplish the purpose of an action are not reasonable, and need not be studied in detail by the agency.” Citizens’ Comm. to Save Our Canyons v. U.S. Forest Serv., 297 F.3d 1012, 1031 (10th Cir.2002) (internal quotation and citation omitted). Agencies must “briefly discuss the reasons” for eliminating unreasonable alternatives from an EIS. 40 C.F.R. § 1502.14(a). WildEarth acknowledges that NEPA does not require an agency to consider impractical alternatives, but it argues the natural wolf alternative was practical. In particular, WildEarth points to studies, emails, and other documents in the record discussing the benefits of this alternative. The evidence WildEarth points to falls into three broad categories: (1) evidence of the biological benefits of wolf reintroduction, such as studies concluding that wolves not only reduce the number of elk but also *1184 prompt elk to congregate in smaller groups and spend less time grazing any particular area, further reducing their impact on vegetation; (2) information about successful wolf reintroduction in Yellowstone National Park and Banff National Park in Canada; and (3) evidence discussing the feasibility of wolf reintroduction, including a 1994 survey showing that 70.8% of Coloradans support wolf reintroduction and a 2004 report by CDOW’s Wolf Management Working Group that discussed the potential benefits of wolf tourism and an offer by an environmental group to compensate livestock owners for wolf predation. While the record supports some benefits to a natural wolf option, that is not what guides us. What guides us is a rule of reason, where the agency explains its decision to take certain proposed options off the table because of a lack of practicality. The NPS did that here. The agency found the natural wolf alternative would be impractical despite some marginal upside, and the record supports that decision. For example, wolf reintroduction may have been successful in Yellowstone and Banff, but the record reflects that those parks are not a good comparator for RMNP. RMNP is many times smaller than Banff and Yellowstone, and also much closer to residential and commercial developments at the park entrances, plus it is near a heavily populated urban area, Colorado’s Front Range Urban Corridor. The NPS determined RMNP has relatively little suitable wolf habitat due to its small size and abundance of steep, high-altitude terrain, which wolves dislike. And as a consequence of the lack of habitat and wolves’ natural tendency to disperse, NPS experts predicted that any wolves in RMNP would be very likely to leave the park boundaries, prompting conflicts with neighboring communities. Such conflicts would likely include predation on livestock and pets. All this would require intensive, costly management of wolves by RMNP personnel, diverting the park’s resources and attention from the very problem the NPS is trying to address — elk overpopulation and degraded vegetation. And given RMNP’s relatively small size and the near certainty that wolves would leave park boundaries, the NPS would need the cooperation of Colorado wildlife agencies to manage wolves outside the park, where the NPS has no jurisdiction. Yet CDOW was unwilling and unable to do so. To add to the complexity of the proposal was the fact that the gray wolf species is endangered, lending a level of state management not required of other species. See Colo.Rev. Stat. § 33-2-105.5 (prohibiting any state or local agency from reintroducing threatened or endangered species into Colorado without authorization from the legislature). WildEarth argues the NPS should not have considered CDOW’s lack of support when determining the feasibility of the natural wolf alternative. WildEarth reads federal regulations to require NPS to ignore CDOW’s opposition when determining which alternatives to include in its EIS. 40 C.F.R. § 1502.14(c) (requiring agencies to consider alternatives otherwise outside their jurisdiction in an EIS). We disagree with WildEarth’s broad reading of § 1502.14(c). This regulation is intended to prompt agencies to consider otherwise appropriate alternatives that the agency lacks jurisdiction to authorize. See Sierra Club v. Lynn, 502 F.2d 43, 62 (5th Cir.1974). But it is not meant to force an agency to consider alternatives rendered infeasible by the actions of another agency. If the NPS concluded the natural wolf alternative was infeasible because it could not bring wolves to RMNP without CDOW’s permission, then 40 C.F.R. § 1502.14(c) might require the NPS to include that alternative in its EIS if the *1185 alternative were otherwise feasible. But that is not the case. The NPS could bring wolves to RMNP with or without CDOW’s approval. But without CDOW’s cooperation in managing wolves outside RMNP’s boundaries, the NPS estimated wolf reintroduction was unlikely to succeed. This is a question of feasibility, not jurisdiction. Consequently, the NPS did not violate § 1502.14(c) by excluding the natural wolf alternative from its EIS. 2 In addition, WildEarth fails to support its claim that the NPS’s March 2005 meeting on the feasibility of the natural wolf alternative was somehow improper, a “shadow process” that excluded the public from participating in the consideration of this alternative in an EIS. Op. Br. at 28. WildEarth suggests representatives from CDOW’s Wolf Working Group attended the meeting and influenced the NPS to drop the natural wolf alternative from its EIS, but the record does not support this assertion. Records from the meeting show only one CDOW representative in attendance, who gave a presentation on chronic wasting disease in elk. WildEarth cites nothing in the record establishing that any other CDOW representatives were at the meeting, let alone that the CDOW attempted to improperly influence the agency or any other attendees. 3 On the contrary, the record reflects that the attendees at the March 2005 meeting were all scientists from the NPS and other institutions and agencies, including the U.S. Fish and Wildlife Service (USFWS), Colorado State University, and Banff National Park. The workshop notes reveal that the assembled experts expressed many doubts about the feasibility of the natural wolf alternative. For example, the experts concluded: “there would be no control over where the wolves would go once they left the park,” “[mjanagement, wolf control and compensation expense would be higher than other wolf options,” “[m]ore time spent managing external issues rather than managing to meet the objectives,” and “[potential to harm wolf restoration efforts in other areas if an attempt in [RMNP] failed.” Aple. Supp. App. at 1088. For these and other reasons, the experts concluded that “this option is not considered feasible or likely to be successful.” Id. at 1032. Agencies are entitled to rely on the opinions of their experts so long as these conclusions are not arbitrary and capricious. Wyoming Farm Bureau Fed’n v. Babbitt, 199 F.Sd 1224, 1241 (10th Cir.2000). WildEarth cites nothing establishing that this conclusion was arbitrary and capricious. Instead, WildEarth attacks the credentials of the assembled experts, pointing out that although the experts were mostly biologists, their conclusions addressed the social implications of wolf reintroduction. We are not sure the con- *1186 elusions drawn at the meeting are so easily categorized, but even if we agree with WildEarth that they were social, there is 'no indication the experts were unqualified to draw these conclusions. Nine of the eleven presentations given at the workshop were on biological topics, such as wolf reproduction and the effects of wolf reintroduction on elk and willow in Yellowstone. But two experts gave presentations on social issues, and WildEarth presents no evidence the experts’ conclusions were outside their areas of expertise. Further, WildEarth argues that, while agencies are entitled to rely on their experts, they cannot exclusively rely on expert opinion without allowing for public comment. Relying on Ctr. for Biological Diversity v. Morgenweck, 351 F.Supp.2d 1137 (D.Colo.2004), WildEarth argues the NPS should not have made a decision about the natural wolf alternative until its EIS was complete, as NEPA requires agencies to gather information about an alternative through the EIS process. Morgenweclc dealt with a petition to list an endangered species. Id. at 1143. Morgenweclc held that when the USFWS receives such a petition, the Endangered Species Act and associated regulations require the agency first to determine whether the petition is meritorious on its face, and then to gather information about the status of the species, in part by soliciting public comments. Id. Morgenweclc found that the USFWS improperly solicited input from select state and federal agencies when determining whether the petition was meritorious, which was proper only during the information-gathering phase, and only by seeking public comments, not by privately contacting select parties. Id. Morgenweclc is not applicable here. NEPA does not prohibit an agency from gathering information from outside sources, as well as its own experts, to determine whether an alternative is feasible and thus a candidate for analysis in an EIS. See, e.g., WildEarth Guardians v. U.S. Forest Serv., 828 F.Supp.2d 1223, 1237-38 (D.Colo.2011) (holding the Forest Service did not act arbitrarily and capriciously by relying on the expertise of an outside agency to exclude from its NEPA analysis a proposal submitted by WildEarth). And unlike the USFWS in Morgenweclc, which did not solicit public comments and instead privately contacted select parties, the NPS here solicited two rounds of public comments on its proposal before eliminating the natural wolf alternative from further consideration. Again, NEPA requires agencies to analyze only reasonable alternatives in an EIS. 40 C.F.R. § 1502.14(a). Agencies are not required to analyze infeasible alternatives so long as they articulate the reasons supporting their decisions. Id. The NPS followed NEPA by articulating the reasons for its decision to exclude the natural wolf alternative from its EIS, and those reasons find support in the record. Thus the agency complied with NEPA’s required procedure. WildEarth concludes by pointing to several emails and other internal communications from NPS employees expressing the opinion that the natural wolf alternative would be feasible, or at least should be included in the EIS. But the fact that some individual NPS employees believed the natural wolf alternative should be included in the EIS does not demonstrate that the agency ignored its own experts or inexplicably changed its mind. WildEarth cites no evidence showing this was a consensus view, rather than a recommendation from a few individual employees. Even if some NPS employees held this view, a diversity of opinion by local or lower-level agency representatives will not preclude the agency from reaching a contrary decision, so long as the decision is not arbitrary and capricious and is other *1187 wise supported by the record. See Nat’l Ass’n of Home Builders v. Defenders of Wildlife, 551 U.S. 644, 659, 127 S.Ct. 2518, 168 L.Ed.2d 467 (2007). The NPS cites ample evidence supporting its decision, including the consensus opinion of the experts at its March 2005 meeting. Even the evidence WildEarth cites identifies factors both supporting and undermining the feasibility of the natural wolf alternative. Accordingly, WildEarth has not demonstrated that the NPS ignored its own experts when it decided to exclude the natural wolf alternative from its EIS. Finally, it is worth reiterating that the NPS continued to consider wolves as a management tool at RMNP even after it eliminated the natural wolf alternative from its EIS. The agency included a modified wolf alternative in its EIS, combining a more limited role for wolves with the use of sharpshooters to cull elk. WildEarth suggests this alternative is not a substitute for the natural wolf alternative at least in part because these wolves would be sterilized, but this is not entirely accurate. Only the male wolves would be sterilized, and only at first. If wolves proved useful as a management tool and successfully established themselves in the park, they would later be allowed to breed. Under this plan, RMNP’s wolf population would be more tightly regulated than under the natural wolf alternative, but this conclusion is appropriate given that the NPS’s goal was not to reintroduce wolves but to manage elk. In sum, we find that the NPS met NEPA’s requirements when it excluded the natural wolf alternative from its EIS. The agency discussed the reasons for its decision in a newsletter it released prior to its release of the draft EIS, in the draft EIS, and in the final EIS, as required by 40 C.F.R. § 1502.14(a). The agency drew a rational connection between these reasons and its conclusion by examining the data in the record, consulting experts at its March 2005 meeting on wolf reintroduction, and repeatedly explaining why it excluded the natural wolf alternative from its EIS. Krueger, 513 F.3d at 1176. This is all NEPA and the APA require. C RMNP Act In its environmental impact statement, the NPS selected gradual reduction in the elk population as its preferred alternative for RMNP’s elk management plan. The EIS also specified that qualified volunteers could assist the NPS in culling elk. The term “qualified volunteers” was defined to include members of the public who received special training in wildlife culling and firearms use and passed a proficiency test. WildEarth contends the use of volunteers transforms the NPS’s culling into hunting, which is prohibited by the RMNP Act. In making its determination, the NPS concluded the use of volunteers would not violate the RMNP Act because the volunteers would be culling elk, and not hunting. The agency found that culling differed from hunting, which it defined as a loosely structured “recreational activity” including elements “of fair chase and personal take of the meat.” Aple. SuppApp., Vol. Ill, at 676. In contrast, culling is a highly regulated and tightly controlled activity whose primary purpose is the “efficient and humane [reduction of] herds of animals that are habituated to the presence of humans.” 4 Id. The agency took the posi *1188 tion that its use of qualified volunteers did not transmute its culls into impermissible hunts because the purpose and structure of the culls distinguished them from hunting, regardless of who participated. 5 The RMNP Act provides in relevant part: “All hunting or the killing, wounding, or capturing at any time of any wild bird or animal, except dangerous animals when it is necessary to prevent them from destroying human lives or inflicting personal injury, is prohibited within the limits of [RMNP]....” 16 U.S.C. § 198c. But the National Park Service Organic Act (Organic Act) states: “The Secretary of the Interior ... may also provide in his discretion for the destruction of such animals and of such plant life as may be detrimental to the use of any said park....” 16 U.S.C. § 3. The NPS argues its plan is permissible under § 3 and does not constitute hunting, so § 198c does not apply. WildEarth contends the NPS’s plan involves hunting and violates § 198c. As' a preliminary matter, the parties and the district court evaluated the NPS’s action under the deferential standard announced in Chevron U.S.A., Inc. v. Natural Res. Defense Council, Inc., 467 U.S. 837, 842-42, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), which we can apply when an agency interprets an ambiguous statute. But we do not automatically apply Chevron every time an agency advances an interpretation of a statute it administers. Whether Chevron applies “depends in significant part upon the interpretive method used and the nature of the question at issue.” Barnhart v. Walton, 535 U.S. 212, 222, 122 S.Ct. 1265, 152 L.Ed.2d 330 (2002) (citing United States v. Mead Corp., 533 U.S. 218, 229-31, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001)). We also consider the agency’s expertise, the importance of the question to the agency’s administration of the statute, and the degree of consideration the agency has given the question. Id. at 222, 122 S.Ct. 1265. Although the presence or absence of notice-and-comment rulemaking or formal agency adjudication is also relevant, these factors are not dispositive. Id. When Chevron does not apply, we still defer to the agency, but only to the extent its reasoning is persuasive. S. Utah Wilderness *1189 Alliance v. BLM, 425 F.3d 735, 759 (10th Cir.2005) (internal citation omitted). In this case, the NPS’s interpretation of the RMNP Act and Organic Act was not adopted as part of a formal rulemaking procedure or adjudication, nor is there any indication the agency considered its interpretation for any great length of time — it was announced only as a part of the final EIS. We do not know whether the NPS considers its interpretation binding for other purposes, despite the fact that this interpretation was announced in an EIS, after a formal notice-and-comment procedure. But as we explain below, whatever deference we give to the agency, the NPS did not violate the RMNP Act. 1. The RMNP Act and Organic Act The conflict is this: The Organic Act, 16 U.S.C. § 3, permits the killing of “detrimental animals” in national parks, but the RMNP Act, 16 U.S.C. § 198c, prohibits all hunting and killing of animals in RMNP, except animals that are dangerous to humans. The NPS argues that because it has not authorized hunting in the park, § 198c’s hunting ban is irrelevant. Wild-Earth counters that the distinction between hunting and culling is illusory and § 198c prohibits implementation of the NPS’s management plan. The problem with both parties’ arguments is that § 198c is broadly written, purporting to ban all hunting and killing of animals within RMNP. WildEarth does not argue that § 198c prohibits the NPS from killing animals in RMNP for management purposes; instead WildEarth argues that § 198c allows only the NPS to use its employees to trim herds and not volunteers. But the logic of WildEarth’s interpretation would bar any plan by the NPS that involves “hunting, or the killing, wounding, or capturing” of animals in RMNP, § 198c, even if it does not involve the use of volunteers. Therefore we must determine whether WildEarth’s interpretation of these statutes is correct. “The starting point in every case involving construction of a statute is the language itself.” Blue Chip Stamps v. Manor Drug Stores, 421 U.S. 723, 756, 95 S.Ct. 1917, 44 L.Ed.2d 539 (1975) (Powell, J., concurring). Statutes must be read as a whole and in relation to one another. United States ex rel. Sikkenga v. Regence Bluecross Blueshield of Utah, 472 F.3d 702, 711 n. 2 (10th Cir.2006). When two related statutes appear to conflict, courts have a duty to construe them harmoniously and give each effect. Morton v. Mancari, 417 U.S. 535, 551, 94 S.Ct. 2474, 41 L.Ed.2d 290 (1974). “The intention of the legislature to repeal must be clear and manifest.” Id. (citing United States v. Borden Co., 308 U.S. 188, 198, 60 S.Ct. 182, 84 L.Ed. 181 (1939)). Normally when two statutes conflict, we interpret the more specific statute as an exception to the more general statute. United States v. Shewmaker, 936 F.2d 1124, 1128 (10th Cir.1991). The Organic Act states that it applies to specific parks to the extent its provisions are not in conflict with any statute made specifically applicable to that park. ' 16 U.S.C. § lc(b). WildEarth suggests this provision requires us to read § 3 as the more general statute and § 198c as the exception to that statute. Under this interpretation, § 3 applies in RMNP except to the extent it conflicts with the specific provisions of § 198c. The problem with WildEarth’s reading of § le(b) is that the RMNP Act incorporates § 3 and makes it specifically applicable to RMNP. 16 U.S.C. §§ 195 & 197. This suggests that § lc(b) does not control our interpretation of §§ 3 and 198c because both provisions are specifically applicable to RMNP. So we could plausibly interpret either statute as the more general provision, with *1190 the other as the exception. Interpreting § 3 as the general provision and § 198c as the exception, as WildEarth suggests, would allow the NPS to destroy detrimental animals in the park except through means prohibited by § 198c, namely “[a]ll hunting or the killing, wounding, or capturing at any time of any wild bird or animal, except dangerous animals when it is necessary to prevent them from destroying human lives or inflicting personal injury.” This interpretation is implausible because of § 198c’s breadth; it bans both hunting and killing and does not exempt the NPS. Read literally, it would allow even the NPS to kill only “dangerous” animals. But animals can be detrimental to a park without being dangerous to humans, as RMNP’s elk demonstrate. If § 198c prohibited killing detrimental but non-dangerous animals, it would effectively repeal § 3. Although this interpretation would still permit the killing of dangerous animals under § 3, it would make § 3 redundant, at least in RMNP, because § 198c also permits killing dangerous animals. We avoid interpretations that render a statute redundant. Enfield ex rel. Enfield v. A.B. Chance Co., 228 F.3d 1245, 1250 (10th Cir.2000). WildEarth’s interpretation would also make it almost impossible for the NPS to manage RMNP’s wildlife. Because interpreting § 198c as an exception to § 3 would do just that, and because we presume Congress enacted § 198c with a knowledge of § 3, 6 see United States ex rel. Boothe v. Sun Healthcare Group, Inc., 496 F.3d 1169, 1176 (10th Cir.2007), we conclude § 198c does not create an exception to § 3. Instead, § 3 is better interpreted as the more specific provision. Under this reading, hunting and killing wildlife is generally prohibited in RMNP, except in the case of dangerous animals (§ 198c) or detrimental animals (§ 3). This interpretation gives effect to both § 198e and § 3, allowing the destruction of both dangerous and detrimental animals, but otherwise prohibiting hunting and killing wildlife in RMNP. But this does not end our inquiry. Section 198c contains two prongs: (1) hunting, and (2) killing, wounding, or capturing. WildEarth argues that any exceptions to § 198c apply only to the second prong, and not to the first. Again, the text of § 198c guides us: “All hunting or the killing, wounding, or capturing ... of any wild bird or animal, except dangerous animals ..., is prohibited within the limits of [RMNP]” (emphasis added). WildEarth’s argument is that the use of the disjunctive “or” and the placement of the dangerous animals exception after “killing, wounding, or capturing” suggests that Congress intended to allow the dangerous animals exception to apply only to the second prong but not the first. This, in turn, suggests that any exception we read into § 198c, namely § 3’s detrimental animals exception, should also apply only to the “killing, wounding, or capturing” of animals in the park because it shows Congress did not intend to create an exception to the hunting ban. We agree that § 3 does not create an exception to the ban on hunting. To harmonize the statutes, as we did above, we read the dangerous and detrimental animals exceptions to § 198c to apply only to its prohibition on killing, capturing, or wounding animals, not to its prohibition on hunting. With these statutory considerations in mind, we turn to the NPS’s culling option. *1191 2. Culling Versus Hunting Having determined that § 198c continues to prohibit all hunting in the park and allows only for management killing — that is, the killing of dangerous or detrimental animals at the Secretary’s discretion — we must discern what differentiates prohibited hunting from permissible management killing. WildEarth claims it is the identity of the party pulling the trigger. If an NPS employee shoots an animal, it is permissible management killing. It is not clear whether WildEarth considers an authorized contractor a permissible actor, but it strongly condemns the use of volunteer agents, regardless of the shooter’s motives or whether he or she has the NPS’s permission. The NPS, again, argues that the nature and purpose of the act determines whether an animal’s destruction is hunting or management killing, not the identity of those who carry it out. There is no support in the text of either § 3 or § 198c for WildEarth’s position. Neither § 198c’s hunting ban nor the exceptions to that ban are based on the identity of the party destroying the animal. Section 198c’s hunting ban is universal, and the exceptions to the statute’s killing, capturing, or wounding ban focus on the reason the animal is killed, rather than who kills it. Section 198c permits animals to be killed if they are dangerous to humans, and § 3 permits animals to be killed if they are detrimental to the park. And § 3 gives the Secretary discretion to determine how detrimental animals are to be destroyed. Nor does WildEarth satisfactorily explain why, if NPS personnel can shoot an elk without it being considered hunting, the NPS’s agents cannot do so, or can do so only if they are being paid by the NPS. Generally a principal can authorize an agent to perform any lawful act the principal can perform himself. See Phillips Petroleum Co. v. Peterson, 218 F.2d 926 (10th Cir.1954); 2A C.J.S. Agency § 129 (2012). WildEarth never explains why an authorized and carefully supervised volunteer is not subject to this rule. The more plausible distinction between hunting and management killing is the one advanced by the NPS: namely, that the difference between permissible management killing, or culling, and impermissible hunting is that the latter is the recreational pursuit of game for meat and sport, with incidental management effects on game populations, while the former is the closely supervised killing of game to control its population. This interpretation rests on the reason the animal is being killed and is consistent with the text of §§ 3 and 198c. The identity or subjective feelings of the person pulling the trigger do not matter. WildEarth argues this distinction is illusory, pointing out that some dictionaries define “hunt” broadly to include the general pursuit of wild animals or game. Oxford English Dictionary (2d ed.1989) (defining “hunt” as “to go in pursuit of wild animals or game.”). WildEarth also notes that the NPS’s own regulations define “hunting” as “taking or attempting to take wildlife,” and “taking” as “to pursue, hunt, harass, harm, shoot, trap, net, capture, collect, kill, wound, or attempt to do any of the above.” 36 C.F.R. § 1.4(a). Wild-Earth argues that because NPS regulations define “hunting” very broadly, the agency should not be allowed to assert a narrower definition here. While these definitions are broad, we do not see them as restricting the NPS’s park management under § 198c. We agree with the NPS that these definitions were not intended to apply to § 198c. Section 1.4(a) states that its definitions “shall apply to this chapter [of the Code of Federal Regulations] ...” (emphasis added), which *1192 chapter does not include § 198c. In fact, the regulations make no reference to § 198c, nor any other similar hunting restriction. In addition, the regulatory definition of hunting in 36 C.F.R. § 1.4(a) is inconsistent with the text of § 198c and the interpretation WildEarth itself advances. As discussed previously, § 198c contains two prongs: one banning hunting, with no exceptions, and one banning killing, wounding, or capturing, with two exceptions. The problem with the regulatory definition is that it defines “taking” as, among other things, “killing, wounding, or capturing,” as well as “hunting.” In other words, if we were to seriously apply the regulatory definition of “hunting” to § 198c, § 198c would have only' one prong because “hunting” and “killing, wounding, or capturing” would be redundant. This in turn suggests that the § 198c exceptions would apply to all of § 198c, letting the NPS authorize hunting in RMNP to remove detrimental or dangerous animals. We suspect WildEarth would not agree with this interpretation, nor is it the better reading of the statutory text. See Gustaf-son v. Alloyd Co., 513 U.S. 561, 574, 115 S.Ct. 1061, 131 L.Ed.2d 1 (1995) (reminding courts to avoid interpretations of statutes that render some words redundant). In any event, the regulations themselves state they are not to be construed to prohibit the implementation of “approved general management and resource management plans.” 36 C.F.R. § 1.2(d). For these reasons, the definition of hunting in 36 C.F.R. § 1.4(a) does not control our interpretation of § 198e. WildEarth’s last argument is that the NPS’s position is contradicted by its own personnel, who stated in various emails that allowing volunteers to participate in managed kills would constitute hunting. Specifically, WildEarth points to an email sent by an NPS biologist, who reported that an unnamed Interior Department attorney said during a meeting that using volunteers in a culling program could be de facto hunting. This is not dispositive. At best, this is an informal, preliminary opinion by an individual employee, not a formal position adopted by the agency. Again, the fact that the NPS later adopted a different position does not mean that the agency acted arbitrarily and capriciously. Nat’l Ass’n of Home Builders, 551 U.S. at 659, 127 S.Ct. 2518. The record reflects that the NPS’s elk management plan is meant to control the number of elk in RMNP. Elk will not be killed when they are within the target range. Culls will be closely supervised by NPS employees. Some cullers may enjoy the experience, but this is irrelevant so long as they kill elk for management purposes pursuant to the procedures and supervision of the NPS. The primary purpose of hunting is not for controlling a population of detrimental animals but for food and sport. Because the purpose of the NPS’s plan is to control the population of the park’s elk and their effect on vegetation, it is distinguishable from hunting, regardless of whether members of the public volunteer to participate in culls. 7 *1193 III. Conclusion For the foregoing reasons, we AFFIRM the order of the district court. Footnotes [1] In 2011, the Colorado Division of Wildlife merged with the Division of Parks and Outdoor Recreation to become the Division of Parks and Wildlife. We refer to the agency by its old name, as all events significant to this appeal occurred prior to the merger. [2] The NPS’s consideration of another alternative, the limited use of wolves in conjunction with sharpshooters, demonstrates the proper application of 40 C.F.R. § 1502.14(c). The NPS would need the permission of the United States Fish and Wildlife Service (USFWS) to reintroduce wolves into RMNP because gray wolves are a federally listed endangered species. See 16 U.S.C. § 1539(j) (permitting reintroduction of an experimental population of an endangered species only with the Secretary of the Interior’s permission and only where "such release will further conservation of such species”). The NPS included consideration of a limited wolf reintroduction in its EIS, in compliance with § 1502.14(c), despite the uncertainty of obtaining this permission. [3] The record evidence WildEarth cites to support its claim does not actually do so. Wild-Earth cites an email from a USFWS biologist to an NPS biologist expressing his opinion that wolf reintroduction in RMNP would not succeed and would create more problems than it would solve. This email only briefly mentions CDOW’s opposition, and does not mention the March 2005 meeting. [4] The full definitions from the final EIS are as follows: Hunting is a recreational activity that includes the elements of fair chase and personal take of the meat, as well as being a conservation tool. Hunting is administered by the state fish and game agency, which licenses hunters. If areas of the park were to be opened to hunting those areas would need to be closed to visitor use while hunting was taking place. The NPS would need *1188 to absorb the costs of managing hunters, visitors and media during a hunt. Culling is used as a conservation tool to reduce populations that have exceeded the carrying capacity of their habitat. As opposed to hunting, culling is done under very controlled circumstances in order to minimize impacts on park operations, visitors, private inholdings and neighbors. Culling is also an efficient and humane way to reduce herds of animals that are habituated to the presence of humans. Culling is not recreational and does not incorporate the concept of fair chase. Culling would be administered by the NPS and carried out by NPS personnel and their authorized agents, and would not require licensing by the state. The personnel doing the shooting would be responsible for killing and processing several animals in any session. Carcasses from culling operations would be tested for chronic wasting disease and to the extent possible carcasses and/or meat would be donated through an organized program to eligible recipients, including members of tribes, based on informed consent and pursuant to applicable public health guidelines. Short-term road closures (a few hours most likely early in the morning) could be needed while culling activity is ongoing. Aple. Supp.App., Vol. Ill, at 676. [5] - In addition to the characteristics in the EIS definitions, the NPS indicated that its culls would focus on culling female elk, and the number and type of animals killed each year would vary depending on annual population surveys. Culls will not take place when the elk population is within a specified range. Cullers will operate under the supervision of an NPS team leader, will attempt to cull multiple elk in any given event, and will attempt to ensure the "humane dispatch” of targeted animals. Id. at 677. [6] The RMNP Act was first enacted in 1915, but § 198c was added in 1929. See Pub.L. No. 70-1009, § 4, 45 Stat. 1536, 1537(1929). The Organic Act, including § 3, was enacted in 1916. See Pub.L. No. 64-235, § 3, 39 Stat. 535 (1916). [7] In reaching this holding, we do not mean to suggest that the NPS may authorize hunting in RMNP simply by covering it with the fig leaf of population control. As discussed, § 198c unambiguously bars anyone from hunting in RMNP. The culls contemplated by the NPS in its management plan will be tightly controlled and are sufficiently distinct from hunting that they do not run afoul of § 198c’s restriction. But§ 198c would clearly prevent the NPS from creating a scheme similar to a state-regulated hunt by, for example, selling tags to licensed hunters and allowing them to freely roam the park every autumn to shoot elk, even if the agency’s reason for doing so was to reduce RMNP’s elk population.