DECISION AND ORDER

Sharon L. Gleason, UNITED STATES DISTRICT JUDGE.

This is an action in which Plaintiffs Southeast Alaska Conservation Council, Alaska Rainforest Defenders, Center for Biological Diversity, Sierra Club, Defenders of Wildlife, Alaska Wilderness League, National Audubon Society, and Natural Resources Defense Council seek the invalidation of portions of the 2018 Environmental Impact Statement and 2019 Record of Decision for the Forest Service's Prince of Wales Landscape Level Analysis Project for the Tongass National Forest. Briefing on the merits concluded on August 23, 2019.1 Oral argument was held on February 7, 2020.2

BACKGROUND

The factual background for this case was set out at some length in the Court's prior order granting Plaintiffs' motion for a preliminary injunction.3 It is repeated here to a certain extent with some additions.

The Tongass National Forest ("Tongass") is a 16.7 million-acre forest in Southeast Alaska.4 The nation's largest national forest,5 the Tongass has seen timber harvesting of varying intensity over the past 100 years.6 In the 1950s, the Forest Service awarded several 50-year timber sale contracts in the forest to "provide a sound economic base in Alaska through establishment of a permanent year-round pulp industry."7 But logging in the Tongass began to slow in the 1980s and 1990s, when several of these long-term contracts were terminated due to market fluctuation, litigation, and other factors.8

Prince of Wales Island, a large island in the Alexander Archipelago, lies within the Tongass.9 Two large pulp mills once operated on the island, where industrial scale logging occurred in the second half of the 20th century, but both mills closed in the 1990s.10 There are 12 communities on the island with a total of approximately 4,300 residents, many of whom are Alaska Native.11 Tourism and sport and commercial fishing are important to the local economy,12 and many residents rely to some degree on subsistence hunting, fishing, and gathering.13

Pursuant to the National Forest Management Act ("NFMA") and its implementing regulations, the Forest Service has developed land and resource management plans, also called forest plans, to govern its management of the Tongass.14 Forest plans "operate like zoning ordinances, defining broadly the uses allowed in various forest regions, setting goals and limits on various uses ..., but do not directly compel specific actions, such as cutting of trees in a particular area or construction of a specific road."15 Any activity occurring within a national forest must comply with the governing forest plan,16 which the Forest Service is required to revise at least every 15 years.17 The current forest plan for the Tongass was issued in 2016, following the completion of an environmental impact statement ("EIS").18 The Forest Plan provides that "[t]imber harvest unit cards will document resource concerns and protection measures," and requires that these "unit cards, including a map with relevant resource features, ... be provided electronically when Draft or Final NEPA documents and decisions are published."19

In late 2016, the Forest Service initiated environmental planning for a proposed project within the Tongass: the Prince of Wales Landscape Level Analysis Project ("Project").20 The agency describes the Project as "a large landscape-scale NEPA analysis that will result in a decision whether or not to authorize integrated resource management activities on Prince of Wales Island over the next 15 years."21 The Forest Service released a final EIS for the Project on October 19, 201822 and issued a Record of Decision ("ROD") selecting the preferred alternative from the EIS on March 16, 2019.23

The Project encompasses all of the land within the national forest system on Prince of Wales Island, consisting of roughly 1.8 million acres.24 It authorizes four categories of activities within this area: vegetation management, including timber harvesting; watershed improvement and restoration; sustainable recreation management; and "associated actions."25 The EIS for the Project does not specify when and where individual activities will occur within the Project Area. Rather, the Project is designed to be a flexible planning framework intended to allow the Forest Service to tailor resource management to changing conditions on the ground over the course of the Project's 15-year term.

The Forest Service appended to the EIS what it terms an Activity Card for each of the 46 activities included in the four activity categories.26 "The Activity Cards describe each potential activity and the related resource considerations," and include "[p]roject-specific design criteria and mitigation measures."27 The Activity Cards were designed using "on-the-ground inventories, computer (GIS) data, and aerial photographs to assess project area conditions and resource-specific concerns."28 The Activity Cards describe and govern activities at the project level, but they do not identify the specific geographic areas within the Project Area where each activity will occur.29 Unlike prior sales, the Project EIS was not accompanied by timber harvest unit cards with maps detailing specific harvest configurations.30

In preparing the Project EIS, the Forest Service also developed a Logging System Transportation Analysis ("LSTA") to "identif[y] potential stands for timber harvest and the associated transportation network that would be needed."31 The LSTA "was developed for National Forest System lands within the project area layer using information from the Forest GIS library, aerial photos, and the Forest Service Activity Tracking System database."32 The LSTA identified 125,529 acres of potential timber harvest in the Project Area: 48,140 old-growth acres and 77,389 young-growth acres.33 The LSTA also identified 643 miles of new roads, 505 of them temporary and 138 permanent.34 The Forest Service represented this information in a Commercial Vegetation Management map,35 which it appended to the ROD, and to which it provided a link in the EIS.36

The Project EIS addresses four alternatives in detail, including a no-action alternative.37 Each activity under each alternative must be consistent with the applicable Activity Card and certain alternative-specific features, and each activity must also occur within the areas identified for that activity in the LSTA.38 Each action alternative establishes a maximum potential amount of timber harvest and road construction.39 Focusing on timber harvest, Alternative 2, the preferred alternative, allows a maximum of 23,269 acres of old-growth harvest and 19,366 acres of young-growth harvest, or roughly 34 percent of the total potential acreage in the LSTA.40 However, the EIS does not identify where the harvest authorized by each alternative would occur within the potential acreage identified in the LSTA.41

In order to capture the "maximum effects" of the Project, the Project EIS makes several assumptions in addressing each alternative.42 First, in analyzing each alternative, the Forest Service indicates that it assumed that all acres of potential harvest in the LSTA would be harvested and all roads proposed by the alternative would be built.43 Second, the Forest Service assumed that all acres would be harvested using clear-cut methods.44 Third, the Forest Service assumed that each Wildlife Analysis Area—a land division used by the Alaska Department of Fish and Game—would be harvested to the maximum acreage available.45

As noted above, the alternatives do not provide the specific locations or configurations of harvest or roadbuilding within the LSTA. Instead, the Project EIS provides that "site-specific locations and methods" for activities such as timber harvest "will be determined during implementation" over the 15-year lifespan of the Project.46 It explains that siting decisions and the parameters of actual timber sales will be determined pursuant to an Implementation Plan, in a way that is consistent with the alternative selected by the ROD and the Activity Cards developed for the EIS.47 However, the EIS makes clear that these subsequent, site-specific decisions will not be subject to additional NEPA review.48 The Forest Service terms this approach "condition-based analysis."49

The Implementation Plan published with the ROD sets out a nine-step process for making site-specific determinations.50 This process includes checking the action against the relevant Activity Card, the final EIS, and the ROD, as well as engaging in "workshops and other public involvement techniques."51 It is during this process, also, that site-specific unit cards will be developed that describe particular harvest configurations.52 The ROD explains that the Implementation Plan "is integral to the analysis of effects in the [Project EIS] and the Selected Alternative in the [ROD]," and was "developed ... to provide a linkage from the [Project EIS] to the project-specific work without the need for additional NEPA analysis."53 That said, the ROD describes the Implementation Plan as a "living document" that "may need to be adjusted."54

The Forest Service began implementing the Project shortly after issuing the ROD. It held a public workshop on April 6, 201955 and published an "Out-Year Plan" for fiscal year 2019 that included a proposed timber sale of 1,156.34 acres, known as the Twin Mountain Timber Sale.56 The Forest Service also published draft unit cards for the sale, which identify the specific locations and method of timber harvest in graphical and narrative form.57

Plaintiffs initiated this case on May 7, 2019.58 The Complaint is brought pursuant to the Administrative Procedure Act ("APA"), 5 U.S.C. §§ 702-06, and alleges that the Project EIS violates three federal laws: (1) the National Environmental Policy Act ("NEPA"), 42 U.S.C. § 4332; (2) the Alaska National Interest Lands Conservation Act ("ANILCA"), 16 U.S.C. § 3120; and (3) the National Forest Management Act ("NFMA"), 16 U.S.C. § 1604.59 The Complaint seeks declaratory judgment, vacatur of the ROD "or portions of it deemed not in compliance with law," and "preliminary and permanent injunctive relief as needed to prevent irreparable harm from implementation of the [Project]."60 Plaintiffs refine their claim in their merits briefing, requesting vacatur of the "portions of the ROD authorizing vegetation management and road construction."61

On September 23, 2019, the Court entered a preliminary injunction, which prohibited the Forest Service from awarding a contract or authorizing ground-disturbing activities associated with the Twin Mountain Timber Sale during the pendency of this case.62 In the preliminary injunction order, the Court found that Plaintiffs had raised serious questions going to the merits of their NEPA claim.63 The order also informed the parties that the Court intended to issue a final decision on the merits no later than March 31, 2020.64

JURISDICTION

The Court has subject matter jurisdiction pursuant to 28 U.S.C. § 1331, which "confer[s] jurisdiction on federal courts to review agency action, regardless of whether the APA of its own force may serve as a jurisdictional predicate."65

LEGAL STANDARD

Plaintiffs' claims arise under the APA.66 Under that statute, a reviewing court shall not set aside an agency's decision unless it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law."67 Agency action is arbitrary and capricious if it

A court's review of whether an agency action is arbitrary and capricious should be "searching and careful," but "narrow," as a court may not substitute its judgment for that of the administrative agency.69 Courts will generally "uphold agency decisions so long as the agencies have `considered the relevant factors and articulated a rational connection between the factors found and the choices made.'"70

"Agency action is `not in accordance with the law' when it is in conflict with the language of the statute relied upon by the agency."71 "Whether agency action is `not in accordance with law' is a question of statutory interpretation, rather than an assessment of reasonableness in the instant case."72

DISCUSSION

The Court will address Plaintiffs' NEPA, ANILCA, and NFMA claims in turn.

I. National Environmental Policy Act

Pursuant to NEPA, agencies must prepare an EIS before taking an action "significantly affecting the quality of the human environment."73 Regulations issued by the Council on Environmental Quality require an EIS to include discussion of the direct and indirect effects of the action, as well as "[t]he environmental effects of alternatives."74 After completing an EIS, "the agency must select a course of action within the range of alternatives analyzed and issue an ROD," which "explains why the agency chose a particular alternative, whether all practical means for avoiding or minimizing environmental harm have been adopted, and, if not, why not."75

"An EIS must `reasonably set forth sufficient information to enable the decisionmaker to consider the environmental factors and make a reasoned decision.'"76 The agency meets this obligation if its EIS "contains a reasonably thorough discussion of the significant aspects of the probable environmental consequences."77 The NEPA process "serves two fundamental objectives": "First, it `ensures that the agency, in reaching its decision, will have available, and will carefully consider, detailed information concerning significant environmental impacts," and "second, it requires `that the relevant information will be made available to the larger audience that may also play a role in both the decisionmaking process and the implementation of that decision.'"78

"NEPA requires ... procedural steps but does not require an agency to reach any particular result."79 Indeed, agencies retain significant discretion over their "methodology and planning strategy" when engaging in environmental review, and "NEPA's `requisite "hard look" does not require adherence to a particular analytic protocol.'"80 The act "merely prohibits uninformed—rather than unwise— agency action."81 To determine whether an agency has complied with NEPA's requirements, courts apply a rule of reason, which involves "a pragmatic judgment whether the EIS's form, content and preparation foster both informed decision-making and informed public participation."82

One method available for agencies engaging in long-term planning is to first prepare a programmatic EIS that considers the broad, program-level effects of a coordinated series of actions, and then conduct subsequent site-specific NEPA analysis for each action as it occurs, tiering back to the programmatic EIS.83 Regarding this approach, the Ninth Circuit has explained that "[t]he detail that NEPA requires in an EIS depends on the nature and scope of the proposed action," and that "[t]he critical inquiry in considering the adequacy of an EIS prepared for a large scale, multi-step project is not whether the project's site-specific impact should be evaluated in detail, but when such detailed evaluation should occur."84

The Forest Service chose not to take that approach here. Instead of preparing a programmatic EIS to be followed by site-specific NEPA analyses for individual timber sales as they occur, the agency compressed its NEPA review for the entire 15-year Project into a single document. The Forest Service maintains that its "landscape-scale NEPA analysis" enables informed decision-making about integrated resource management at the programmatic level and contains sufficient site-specific information and analysis to proceed with individual timber sales over the 15-year Project period without additional NEPA review.85

The Court will not here decide whether the nature of the Project required the Forest Service to complete a programmatic EIS to which later planning documents would tier.86 Nor do Plaintiffs request such a decision.87 Instead, the Court will evaluate the analytical method that the agency employed and determine whether the Project EIS, a landscape-level long-term planning document, adequately evaluated the site-specific impacts of as-yet undefined timber sales that could potentially occur on certain acreage on Prince of Wales Island over the Project's term.

Plaintiffs contend that the Project EIS does not provide sufficient site-specific information or analysis to comply with NEPA.88 They argue that this case is governed by the Ninth Circuit's decision in City of Tenakee Springs v. Block.89 In that case, the Circuit reversed a district court's decision not to enjoin "construction of an 11-mile road through the Kadashan watershed" in the Tongass.90 The plaintiffs had challenged the adequacy of an EIS for a five-year operating plan that "specified no timber harvesting in the Kadashan Watershed in the years 1981 through 1986," but did authorize the construction of the road for future harvest activity.91 The Circuit ordered the entry of a preliminary injunction, in part due to its conclusion that the plaintiffs had raised serious questions about the merits of their NEPA claim.92 It explained that the challenged EIS did not "g[ive] any indication of its overall plan for timber harvesting" in the designated area and that "it [was] impossible to determine where and when harvesting will occur on the 750,000 acres of land."93 The Circuit held that the EIS was inadequate, reasoning that the location and timing of logging would affect "the locating, routing, construction techniques, and other aspects of the road, or even the need for its construction."94

In City of Tenakee Springs, the Ninth Circuit separately rejected the trial court's conclusion that the Forest Service had discretion to determine the specificity of its environmental review.95 Instead, it held that "[a]lthough the agency does have discretion to define the scope of its actions, such discretion does not allow the agency to determine the specificity required by NEPA."96 The Circuit explained that "[w]here there are large-scale plans for regional development, NEPA requires both a programmatic and a site-specific EIS."97

Here, Plaintiffs argue that the Project EIS, with its condition-based analysis, is similarly deficient, and that the Forest Service impermissibly limited the specificity of its environmental review.98 The Forest Service maintains that it has complied with NEPA by creating a project-level map that "provide[s] information on where timber harvest and road construction activities may take place."99

The Forest Service used the LSTA to identify "potential stands for timber harvest as well as the transportation network needed to access those stands" within the roughly 1.8-million-acre Project Area over a 15-year period.100 Through this process, the agency identified 125,529 acres of timber for potential harvest: 48,140 acres of old growth and 77,389 acres of young growth.101 The Project EIS links to a Commercial Vegetation Management map, which portrays this potential acreage graphically.102 However, the EIS expressly leaves site-specific determinations about the actual location of timber harvest within this potential acreage for future determination.103 For example, Alternative 2—the selected alternative—allows 23,269 acres of old-growth harvest, but does not specify where this harvest will be located within the 48,140 acres of old growth identified as suitable for harvest in the Project Area.104 The Activity Cards likewise do not identify with specificity where harvesting will occur, although they do contain mitigation measures that could constrain future siting decisions.105

The Forest Service maintains that it properly exercised its discretion to determine the scope of the Project while at the same time providing the specificity required by NEPA.106 The agency argues that through Project EIS, it "has provided information on where timber harvest and road construction may take place and is not `attempt[ing] to justify' any lack of information, or to opt-out of any of NEPA's requirements."107 However, similar to the EIS found inadequate by the Ninth Circuit in City of Tenakee Springs, the Project EIS does not include a determination —or even an estimate—of when and where the harvest activities or road construction authorized by each alternative will actually occur.108 Rather, it reserves actual siting decisions for the future, as individual timber sales are offered.

The Forest Service contends that the phrase from City of Tenakee Springs on which Plaintiffs rely—that it "is impossible to determine where and when harvesting will occur on the 750,000 acres of land"— was factually inaccurate, citing the district court's decision on remand.109 But regardless of that statement's factual accuracy, the Circuit's reasoning is still binding precedent: NEPA requires that environmental analysis be specific enough to ensure informed decision-making and meaningful public participation.110 The Project EIS's omission of the actual location of proposed timber harvest and road construction within the Project Area falls short of that mandate.

The Forest Service maintains that Plaintiffs are demanding more detail than is required by NEPA. The agency relies on Stein v. Barton, which concerned a challenge to the EIS for a five-year operating plan for an area of the Tongass that was then being logged pursuant to a long-term contract held by the Ketchikan Pulp Company.111 Among other issues, the plaintiffs in that case argued the "harvesting plans for each area" were not sufficiently site-specific, "object[ing] that they cannot determine from the FEIS `where, when, and how logging and roading activities will occur on the 812,477 acres of land.'"112 The district court rejected this argument, concluding that the EIS "contain[ed] comprehensive, detailed quantitative and qualitative descriptions of the logging and roading plans for each harvest unit."113 The court noted that "the only details that the FEIS does not disclose are exact timetables and locations on the ground for planned harvesting activities in each harvest unit," but that this was not a fatal omission since "the Forest Service does not develop these details at the pre-implementation stage."114

The Forest Service contends that Plaintiffs' arguments are similar here and that NEPA does not require disclosure of the exact location of actual timber harvest in the Project Area.115 But the EIS that survived review in Stein contained significantly greater site-specificity than the Project EIS at issue here. In Stein, the district court described a "nine-volume FEIS [that] employ[ed] a combination of annotated topographic maps, textual, and tabular data to describe the project alternatives and their impacts on cognizable values within the affected areas" and contained "comprehensive, detailed quantitative and qualitative descriptions of the logging and roading plans for each harvest unit."116 Similarly, in Alliance for the Wild Rockies v. Weber, another case the Forest Service cites,117 the district court upheld an environmental analysis for a timber sale that "identif[ied] the project boundaries down to the township and range level" and contained maps that would "allow the Plaintiffs to identify where those activities will take place in relation to bull trout critical habitat," a resource value the plaintiffs had claimed was inadequately addressed.118

The Project EIS at issue here does not approach this level of specificity; it does not delineate harvest units, let alone identify planned activities within them and describe their impacts on localized cognizable values. Nor does the Project EIS allow the public to identify where specific harvest activities will occur in relation to various cognizable values on Prince of Wales Island.119 Far from "unwarranted `fly-specking,'"120 Plaintiffs' objections identify serious shortcomings in the sufficiency of the Project EIS's environmental analysis.

Moreover, the district court in Stein rejected the plaintiffs' site-specificity claims because they had not asserted or "show[n] why disclosure of more details regarding site-specific impacts [was] necessary in order to `foster both informed decision-making and informed public participation.'"121 Here, Plaintiffs maintain that more detailed information about the specific location of timber harvest under the Project is necessary to properly assess its ecological and subsistence impacts.122

The Forest Service contends, however, that the EIS satisfies NEPA because it analyzes the Project's maximum potential impacts.123 In its briefing before this Court, the agency describes the Project EIS as assuming that "the entire ... Project Area would be harvested by clear-cut methods and that every mile of road would be constructed up to the maximum number of harvest acres and miles of road authorized under each alternative."124 The Forest Service maintains that as a result of this worst-case-scenario analysis, "whatever units [it] ultimately selects within the constraints outlined in the alternatives, Activity Cards, and Implementation Plan, the Project will produce environmental effects that fall within those already disclosed and analyzed in the EIS."125

It is not entirely clear from the record, but for at least some of the impacts analyzed, the Project EIS assumes that considerably more timber would be harvested than is actually authorized under each alternative. For example, to determine the effects of each alternative on wildlife and subsistence, the Project EIS focused on the 32 Wildlife Analysis Areas ("WAA")— "land divisions used by [the Alaska Department of Fish and Game] for wildlife analysis and regulating wildlife populations" —situated in the Project Area.126 In its briefing, the Forest Service explained that "for each WAA, the Service assumed [timber] harvest would be concentrated in that WAA and would occur at the maximum level."127 At oral argument, the agency confirmed that this approach could cause the EIS to assume that more harvest would occur under each alternative than actually allowed by that alternative.128 And language in the Project EIS itself indicates that this approach assumed that all potential harvest stands identified in the LSTA were cut, regardless of the acreage authorized by the alternative in question.129 The Project EIS makes these assumptions despite its recognition that "[t]he specific location and amount of harvest in each WAA would be determined during implementation and vary by alternative."130

The Forest Service cites WildEarth Guardians v. Conner to defend its approach.131 There, the Tenth Circuit upheld an Environmental Assessment ("EA") for a tree-thinning project designed to address a beetle infestation in two national forests.132 The EA "evaluat[ed] the Project's effects on lynx in a worst-case scenario in which all the mapped lynx habitat in the Project area is treated."133 Due to the listing of the Canada lynx as a threatened species, the Forest Service had previously amended the forest plans for the two forests to prohibit the clearcutting of more than 15% of lynx habitat or precommercial thinning of more than 1% of lynx habitat in a given analysis area after analyzing the impact of these changes in an EIS.134 In the subsequent EA for the tree-thinning project, the Forest Service took "the conservative approach of assuming that all lynx habitat in the Project area w[ould] be treated."135 Using that approach, the Forest Service found that only 6% of lynx habitat would be subject to clearcutting, and no more than 0.2% subject to precommercial thinning—well below the percentages prohibited by the governing forest plans.136 The Tenth Circuit held that because the project's impacts to lynx habitat were below the caps established in the amended forest plans, which had been adopted after a full analysis of the lynx-habitat impacts in the EIS, the Forest Service "could reasonably assess the maximum impact that the Project could have on the lynx and conclude it was unlikely to adversely affect them."137

Although in the instant case the Forest Service applied an analytical framework similar to the one it used in Wild-Earth Guardians v. Conner, the difference between an EA and an EIS renders that case inapplicable. An EA is meant to determine whether a proposed action will have a significant impact on the environment, such that an EIS is necessary.138 In contrast, an EIS must compare the environmental impacts of different alternatives, not just determine whether environmental impacts will occur.139 While an agency's analysis of a proposed action's maximum potential impacts may be appropriate for an EA, the Forest Service's analytical framework in this case is not sufficient to meet the requirements for an EIS.140

The Forest Service candidly acknowledges in the EIS that its analytical framework overestimates the Project's impacts and is unlikely to reflect the actual extent and nature of activities under each of the proposed alternatives within the Project Area.141 By focusing on the Project's maximum potential impacts for all alternatives rather than its actual or foreseeable impacts for each alternative, the EIS falls short of NEPA's directive to "contain[] a reasonably thorough discussion of the significant aspects of the probable environmental consequences" for each alternative.142 This approach, coupled with the lack of site-specific information in the Project EIS, detracts from a decisionmaker's or public participant's ability to conduct a meaningful comparison of the probable environmental impacts among the various alternatives. For example, in the introduction to the section discussing the Project's impacts to wildlife habitat, the EIS states that "[t]he effects ... are similar between all alternatives because all alternatives assume that all acres proposed for timber harvest will be harvested. The analysis also assumes that all acres will be harvested by even-aged harvest methods."143 Due to these identical assumptions for each alternative, it appears that at least with respect to wildlife impacts, the Project EIS only meaningfully analyzed the different mitigation measures contained in each alternative, not the harvest limits.144 And where the Project EIS does differentiate between alternatives, it does so in partly conditional terms due to its lack of site-specific information. For example, when discussing impacts to high productive old growth habitat, the EIS concludes that Alternative 2 "may result in two [Wildlife Analysis Areas] dropping below 50% habitat remaining."145

The Project EIS identified a total acreage of potential timber harvest, but not the distribution of the specific acreage authorized by each alternative within these areas. This omission is meaningful given the duration and scale of the project.146 Despite "additional parameters that limit the ultimate selection of units and activities,"147 such as mitigation measures contained in the Activity Cards,148 the Project EIS's structure creates ambiguity about the actual location, concentration, and timing of timber harvest and road construction on Prince of Wales Island.149 By doing so, the Project EIS fails to provide a meaningful comparison of alternatives.

By authorizing an integrated resource management plan but deferring siting decisions to the future with no additional NEPA review,150 the Project EIS violates NEPA. The Forest Service has not yet taken the requisite hard look at the environmental impact of site-specific timber sales on Prince of Wales over the next 15 years. The Forest Service's plan for condition-based analysis may very well streamline management of the Tongass and decrease the amount of falldown acreage associated with each timber sale;151 however, it does not comply with the procedural requirements of NEPA, which are binding on the agency.152 "NEPA favors `coherent and comprehensive up-front environmental analysis to ensure ... that the agency will not act on incomplete information, only to regret its decision after it is too late to correct.'"153 Plaintiffs have therefore established Count I of the Complaint; the Project EIS violates NEPA and is therefore not in accordance with law.154

II. Alaska National Interest Lands Conservation Act

Congress enacted ANILCA to "cause the least adverse impact possible on rural residents who depend upon subsistence uses of the resources of [the public lands in Alaska]."155 To achieve this purpose, § 810 of ANILCA imposes procedural requirements upon federal decisionmakers; pursuant to its terms, an agency proposing an action resulting in the "use, occupancy, or disposition of public lands" must evaluate that action's effects on "subsistence uses and needs," the availability of other lands for the same purpose, and "other alternatives" that would reduce the impacts to subsistence uses.156 Upon determining that the action "would significantly restrict subsistence uses," the agency must provide notice to the affected communities, hold public hearings, and make three findings:

When the proposed action requires completion of an EIS, the agency "shall provide the notice and hearing and include the findings required by [ANILCA § 810(a)] as part of such [EIS]."158

The Project EIS contains a section discussing the impacts to subsistence activities.159 This section recognizes that "[s]ubsistence hunting, fishing, trapping, and gathering activities are a major focus of life for many residents on Prince of Wales Island."160 The EIS notes that commenters were particularly concerned about the Project's impacts to Sitka black-tailed deer, which the EIS "considered an `indicator' for potential subsistence resource effects concerning the resources associated with old-growth forest habitat."161 Using the maximum-impact methodology described in the previous section of this order,162 the Forest Service determined that the Project presents "a significant possibility of a significant restriction for the use of deer," largely "due to the loss of deep snow habitat in some WAAs [Wildlife Analysis Areas]."163 The Project EIS identified "five WAAs of concern ... [:] 1214, 1315, 1317, 1318, and 1420," which are "near the communities of Thorne Bay, Coffman Cove, Hollis and Klawock."164

In light of this determination, the Forest Service held seven subsistence hearings in Prince of Wales Island communities after issuing the draft EIS.165 In response to concerns raised at two of these hearings, the agency decided "not to authorize commercial harvest of old-growth stands in the area `North of the 20 Road, and in VCU 5280."166 In the final Project EIS, the Forest Service made the three findings required by ANILCA § 810(a)(3).167

Plaintiffs contend that "[t]he lack of site-specific information in the [Project EIS] violates not only NEPA, but also section 810 of ANILCA."168 They maintain that "analysis of impacts on subsistence uses under section 810 should be at least as site-specific as that for the environmental impacts under NEPA."169 As such, Plaintiffs also argue that the Forest Service's § 810(a)(3) findings were premature since they were made "before deciding the specific location or extent of logging or road construction over the next 15 years."170

The Forest Service argues that Plaintiffs have simply repackaged their NEPA claim.171 However, due to the similarities between the procedural requirements of NEPA and ANILCA, courts have evaluated the two statutes under similar standards.172 Although there is little case law on the issue, at least one court has looked to NEPA decisions to determine the site-specificity required by ANILCA § 810. In City of Tenakee Springs v. Clough, the district court rejected the plaintiffs' claims that supplemental EISs were "deficient because they discuss site-specific impacts of proposed harvesting on subsistence resources but do not correlate those impacts with the specific subsistence needs of each affected community."173 The court cited Stein v. Barton's description of NEPA site-specificity as the relevant standard, stating that "[p]roposed activities must be sufficiently correlated with environmental factors in each affected area to facilitate public discussion of the project."174 But the court found that no additional site-specificity was necessary to comply with ANILCA because the "EIS's identify site-specific impacts on subsistence resources, incorporate maps identifying which sites are important for subsistence use generally, and catalog how and to what extent each community utilizes subsistence resources, including data on per capita consumption."175

The Court has already determined that the Project EIS does not contain the level of site-specificity required by NEPA. For the same reasons, the Court finds that despite the public hearings it held and the findings it made, the Forest Service has failed to comply with ANILCA § 810. The purpose of that section is to promote informed decision-making, such that the impacts of an action to subsistence activities are considered; those actions "which would significantly restrict subsistence uses can only be undertaken if they are necessary and if the adverse effects are minimized."176 This purpose can only be fulfilled if specific information about the actual, not the potential, proposed action is available. This is made clear by the statute's command to consider site-specific aspects of a proposed action, such as its effect on local "subsistence uses and needs."177

The Forest Service contends that it has fulfilled its duty under ANILCA by "identifying where Project activities would occur, evaluating the maximum impacts of the Project on subsistence uses, and by taking actions to benefit subsistence uses and reduce the adverse effects of the project on ... those uses."178 However, as discussed above, the Commercial Vegetation Management map only identifies the potential, and not the actual, locations of timber harvest and road building within the 1.8-million acre Project Area. The Forest Service's evaluation of the Project's maximum impacts does not evaluate the Project's actual expected impacts.179 By not developing actual site-specific information, the Forest Service limited its ability to make informed decisions regarding impacts to subsistence uses and presented local communities with vague, hypothetical, and over-inclusive representations of the Project's effects over a 15-year period.180

The Implementation Plan does envision additional information-gathering and public participation before site-specific decisions are made.181 However, as Plaintiffs note, this process does not require the Forest Service to make additional findings under ANILCA § 810(a)(3) or provide the public with a right of appeal.182 Moreover, the Implementation Plan is subject to change by the Forest Service, and there is no certainty that its public participation provisions will last for the Project's 15-year duration.183 Without either an up-front discussion of actual site-specific impacts or future ANILCA analysis when siting decisions are made, the Project EIS and ROD are inconsistent with ANILCA § 810. Plaintiffs have therefore established Count II of the Complaint; the Project EIS violates ANILCA and is therefore not in accordance with law.184

III. National Forest Management Act

Pursuant to NFMA, the Forest Service must prepare a land and resource management plan, also called a "forest plan," for each forest it manages.185 Projects occurring in a national forest must comply with that forest's management plan.186 Standard and Guideline TIM3.I.C in the 2016 Forest Plan that currently governs management of the Tongass provides:

Tongass National Forest Supplement 1909.15-2015-1, which "[e]stablishe[d] procedures for producing and distributing unit and road cards associated with NEPA documents,"188 was rescinded by the Forest Service in October 2018.189

The Project EIS does not contain unit cards, and the ROD explains in the section describing Activity Cards that "[u]nit cards would be developed for any timber sales when site-specific locations are determined" through the Implementation Plan.190 Plaintiffs contend that the "Forest Service violated the forest plan requirement to include unit cards with the draft or final EISs for the Prince of Wales Project."191

Responding to comments arguing that the Project EIS had failed to comply with the Tongass Forest Plan, the Forest Service explained that it understood the rescission of Tongass National Forest Supplement 1909.15-2015-1 to have rendered "the timing for when electronic unit cards are provided ... no longer applicable."192 The Forest Service added that the Implementation Plan's "opportunity for public comment on the maps, and unit and road cards meets the intent of Forest Plan TIM3.I.C."193

In its briefing to this Court, the Forest Service advances a different argument. It now contends that TIM3.I.C is ambiguous regarding "the contents and format of unit cards, except that the cards `will document resource concerns and protection measures' and `include[e] [sic] a map with relevant resource features.'"194 The Forest Service maintains that the Activity Cards and Commercial Vegetation Management map provided with the Project EIS comply with this requirement.195

The Forest Service argues that its interpretation of TIM3.I.C is entitled to Auer deference, which accords "defer[ence] to agencies' reasonable readings of genuinely ambiguous regulations."196 But Auer deference is not automatic; the Supreme Court has explained that "before concluding that a rule is genuinely ambiguous, a court must exhaust all the `traditional tools' of construction," meaning it "must `carefully consider[]' the text, structure, history, and purpose of a regulation."197

Addressing first the Forest Service's position in the administrative record, the Court finds no ambiguity in TIM3.I.C about when the agency is required to provide unit cards, notwithstanding the rescission of the supplement cited therein. The Standard and Guideline clearly states that "[t]he unit cards ... will be provided electronically when Draft or Final NEPA documents and decisions are published."198 Under the plain language of the Forest Plan, the Forest Service must provide unit cards when the relevant NEPA document is published. Regardless of the provision's intent,199 the Forest Service departed from the unambiguous directive of TIM3.I.C. The agency's decision to delay the publication of unit cards until the Implementation Plan, after NEPA review was completed for the Project, is therefore inconsistent with the Forest Plan.

The Court finds the Forest Service's position in its briefing to be no more convincing. While TIM3.I.C does not fully explain what a unit card should contain, it is clear from the Standard and Guideline's language that each card must relate to a discrete geographic area, or "[t]imber harvest unit," within the Project Area.200 The Commercial Vegetation Management map does not identify specific timber harvest units,201 and the Activity Cards "document resource concerns and protection measures at the Project level."202 They are not unit cards within the meaning of TIM3.I.C.203

Moreover, even if TIM3.I.C were ambiguous, the current interpretation advanced in the Forest Service's briefing would not be entitled to Auer deference. Throughout the administrative record, the Forest Service's position is consistent: provision of unit cards after the completion of NEPA review complies with TIM3.I.C due to the rescission of Tongass National Forest Supplement 1909.15-2015-1. It is only in its briefing to this Court that the agency asserts that the Activity Cards and the Commercial Vegetation Management map themselves constituted the unit cards required by TIM.3.I.C.204 "[A] court should decline to defer to a merely `convenient litigating position' or `post hoc rationalization[n] advanced' to `defend past agency action against attack.'"205 And the administrative record itself belies the Forest Service's litigation position; in the ROD, the agency differentiated between Activity Cards and unit cards and explained that the latter would be provided through the Implementation Plan in an effort to comply with TIM3.I.C.206

Due to their failure to include timber harvest unit cards corresponding to discrete geographic locations, the Project EIS and ROD are inconsistent with the 2016 Tongass Forest Plan. Plaintiffs have therefore established Count III of the Complaint; by not complying with the applicable forest plan, the Project violated NFMA and is therefore not in accordance with law.207

Nevertheless, the Forest Service asserts that this error was not prejudicial.208 "Relief is available under the APA only for `prejudicial error.'"209 Plaintiffs challenging agency action bear the burden of showing prejudice, but this is not "a particularly onerous requirement."210 Plaintiffs have met their burden here. They argue that they "were prejudiced by the lack of unit cards that would have given them the opportunity to provide meaningful input on logging locations, impacts, and alternatives."211 The Ninth Circuit has consistently held that an error that affects the public's ability to meaningfully participate in the NEPA review process is prejudicial.212

The Forest Service contends that Plaintiffs had the opportunity to comment on the Activity Cards and Commercial Vegetation Management map, which contain "proposed locations of timber harvest."213 However, as discussed above, these documents do not contain the level of site-specificity required by TIM3.I.C—or by NEPA for that matter—and did not allow Plaintiffs to meaningfully comment on the specific harvest activities that would have been identified on timber harvest unit cards had they been published with the draft EIS.214 The Court therefore finds that the NFMA violation was prejudicial and that relief is warranted under the APA.

IV. Proper Remedy

Having determined that Plaintiffs prevail on all three counts in their Complaint, the Court turns to the question of the proper remedy. Plaintiffs request a judgment declaring that the Project EIS "violates NEPA, section 810 of ANILCA, and NFMA," and "vacating those portions of the ROD authorizing vegetation management and new road construction."215 The Forest Service contends that remand without vacatur may be appropriate, and requests the Court to allow supplemental briefing to address the proper remedy.216

Vacatur is the default remedy under the APA, which directs reviewing courts to "hold unlawful and set aside" unlawful agency action.217 The Ninth Circuit has explained that a court should "order remand without vacatur only in `limited circumstances,'" and "leave an invalid rule in place only `when equity demands.'"218 To determine whether to remand an action without vacatur, a court is to "weigh the seriousness of the agency's errors against `the disruptive consequences of an interim change that may itself be changed.'"219

The Forest Service maintains that "[b]oth considerations are informed by the Court's decision on the merits," and the agency "believes the consequences of vacatur would be extremely disruptive due to the recent low and uncertain supply of timber in Southeast Alaska, which threatens businesses in the region."220 The Forest Service maintains that "[t]he economic need for Project timber and the harm that would be caused by delaying timber harvesting activities authorized by the Project cannot be adequately addressed within the page limits for [its] merits brief."221 In response, Plaintiffs note that the Forest Service "made a strategic choice to devote more pages to the merits of the claims, with no basis to assume it was entitled to supplemental briefing on the remedy."222 Since Plaintiffs did choose to devote a section of their briefing to the remedy issue, they maintain that allowing supplemental briefing would unfairly prejudice them.223

The Court finds that supplemental briefing on the appropriate remedy could be helpful. Any prejudice to Plaintiffs as a result of this delay will be eliminated because the Court will keep the preliminary injunction in effect until the appropriate remedy is determined.

CONCLUSION

In light of the foregoing, IT IS HEREBY ORDERED that:

Plaintiffs' request at Docket 10 for declaratory relief is GRANTED. The Project EIS violates NEPA, 42 U.S.C. § 4332(2)(C), ANILCA, 16 U.S.C. § 3120(a), and NFMA, 16 U.S.C. § 1604(i). It is therefore "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law."224

Each party may file a supplemental brief addressing the proper remedy in this case within 21 days of the date of this order; such brief shall not exceed 15 pages. Each party will then have an additional 14 days to file a response, not to exceed 8 pages.

The preliminary injunction at Docket 27 shall remain in place until the Court enters a final judgment.

Attachment