B. Lynn Winmill, Chief U.S. District Court Judge *1139INTRODUCTION
The Court has before it cross-motions for summary judgment, and a motion filed by plaintiffs to consider extra-record material. The Court heard oral argument on May 11, 2018, and took the motions under advisement. For the reasons expressed below, the Court will grant the plaintiffs' motion for summary judgment, deny the defendant's motion for summary judgment, and deem moot the plaintiffs' motion to consider extra-record material as the Court did not review that material in ruling for plaintiffs.
SUMMARY
For years, Wildlife Services has responded to requests from Idaho livestock producers to kill or remove predators like coyotes that threaten their herds. When the agency decided to expand its operations to kill or remove predators to game animals and protected species, it prepared a draft Environmental Assessment (EA) and circulated it to various agencies and the public. That draft prompted numerous critical comments, especially from other agencies with long experience and expertise in managing game animals and protected species: The Bureau of Land Management, Forest Service, and the Idaho Department of Fish and Game, among others.
Instead of studying these concerns in greater depth in an Environmental Impact Statement (EIS), Wildlife Services largely rejected these criticisms, finding that they were invalid for various reasons. Under the National Environmental Policy Act, an agency may use a convincing and objective analysis to reject criticisms and refuse to prepare a full EIS. But that was not done here. While Wildlife Services responded in detail to the criticisms, their reasons for rejecting them were not convincing and objective; the agency failed to take the required "hard look" at the concerns raised by the other agencies. Consequently, the Court finds that Wildlife Services acted in an arbitrary and capricious manner in deciding not to prepare an EIS. The Court will therefore grant the plaintiffs' motion for summary judgment and deny the motion filed by Wildlife Services.
FACTUAL BACKGROUND
Wildlife Services is a federal agency charged with killing or removing predators like mountain lions and coyotes that prey on wild game animals, damage agricultural interests, and pose a danger to humans. See 7 U.S.C. §§ 8351 - 8352. The agency does not conduct operations unilaterally, but instead responds to requests for assistance from individuals or other agencies such as the Bureau of Land Management (BLM), the Forest Service, and the Idaho Department of Fish and Game (IDFG). Wildlife Services has a Memorandum of Understanding with each of these agencies that governs its operations when those agencies request assistance. AR-037168. The agency has similar agreements with the various Tribes in Idaho.
Upon request, Wildlife Services can provide three types of wildlife damage management assistance: technical assistance, direct control assistance, and research assistance. AR-038561. Direct control assistance, *1140which is at issue here, consists of field activities conducted or supervised by Wildlife Services personnel. AR-037260
Most direct control assistance requests in Idaho concern conflicts with coyotes. AR 037303. Coyotes kill more cattle and sheep in Idaho than any other predator, causing a substantial economic loss to ranchers. AR-37171, 37174-75. In response to requests from those ranchers, and various agencies, Wildlife Services killed over 3,860 coyotes in Idaho in 2016. See Answer (Dkt. No.5) at ¶ 55.
There were two past occasions when Wildlife Services responded to agencies' requests to kill or remove predators of protected species and game animals. In the first, conducted from 1997 to 2002, Wildlife Services assisted IDFG with a coyote and mountain lion population reduction effort that was conducted in select game management units in southeastern Idaho to enhance the mule deer population. AR-37184. And during the past 10 years, Wildlife Services and the Forest Service have entered into agreements to manage badger, red fox and coyote predation to protect the northern Idaho ground squirrel, a threatened species under the Endangered Species Act. AR-37185. More recently, IDFG asked Wildlife Services to evaluate a program to remove ravens to protect the sage-grouse. Id. In the future, Wildlife Services anticipates receiving requests to provide predator management to protect sage-grouse eggs and chicks. AR-37186
Wildlife Services has been operating under a programmatic EIS issued first in 1994, and then reissued in 1997 with some corrections. It analyzed Wildlife Services' activities across the country. WS has never prepared an EIS for its Idaho activities, but has instead been operating here under two Environmental Assessments (EAs), one completed in 1996 for the northern and central regions of Idaho, and the other completed in 2002 for the southern region. Because both are outdated, Wildlife Services decided to conduct a new evaluation, this time in a single EA. In the words of Wildlife Services, "[t]he new analysis reviews the impacts of the existing program (environmental baseline), develops new and updated alternatives for PDM, and updates the review of potential environmental impacts of the proposed alternatives." AR-37169.
Wildlife Services began the process of drafting the EA by soliciting public comments on the effects of its predator damage management activities in Idaho. After receiving those comments, Wildlife Services issued a Draft EA on June 19, 2015, and again solicited public comments.
The June 2015 Draft EA discussed five alternatives: (1) continuing Wildlife Services' existing activities in Idaho; (2) ceasing its activities; (3) providing non-lethal assistance only; (4) providing non-lethal assistance before any lethal control; and (5) the "preferred" alternative, expanding its existing activities to encompass killing predators to protect game animals and protected species. AR-32195.
More specifically, the preferred alternative included expanding Wildlife Services' existing predator control actions to including killing native wildlife at the request of IDFG to benefit other desired wildlife species, including a new proposal to kill ravens and other predators for the benefit of greater sage-grouse, and to kill native predators to benefit Columbian sharp-tailed grouse, mule and white-tailed deer, bighorn sheep, and pronghorn antelope, among other animals. AR 32099-100 (listing target predators and wildlife to be "protected"). The Draft EA also claimed Wildlife Services might conduct other, unidentified predator control activities so long as it determined their cumulative impacts *1141would be within the range considered in the EA. AR-32203.
Wildlife Services received hundreds of comments from the public and other agencies, including the BLM, the Forest Service, the Fish and Wildlife Service, and the Idaho Department of Fish and Game (IDFG). Many of the comments were critical of various aspects of the Draft EA.
Comments on Draft EA-Objectivity
For example, three agencies-the BLM, the Forest Service, and the IDFG-all commented that the Draft EA was not an objective analysis of the environmental impacts. The BLM commented that "[t]he document thus far does not read like a real analysis of the potential [Predator Damage Management] outside of lethal methods. Instead, it sounds like a pre-decisional defense of lethal methods, and fails to consider the real benefits of alternative approaches." AR-31350. The IDFG's large carnivore coordinator, Steve Nadeau, echoed that comment:
This is a very complete look at the potential impacts of control actions from one perspective and builds a nice case for conducting PM [Predator Management] in Idaho. It does not however provide an adequate perspective of enormous availability of literature and research that shows the ineffectiveness or neutral benefit of the actions, thus bringing into question the objectiveness of the EA.
AR-32058. The Forest Service commented "[t]here is a weakness in this document in that by portraying only one side of the issue, and cherry-picking papers, it is assuming that there is no controversy." AR-31086.
Comments on the Draft EA-Ineffectiveness of PDM
The BLM commented that "there is no guarantee that predator control will have the intended impact on prey populations. For example, the literature is full of examples of how mammalian predators respond to losses of individual members of populations by adjusting space use or increasing litter sizes." AR-31292. The Forest Service agreed, commenting that "[t]he effectiveness and efficiency of predator control for actually limiting damage is quite controversial." AR-31083. Regarding Wildlife Services' proposal to kill ravens to protect sage grouse, the BLM commented that "[c]ontrolling raven populations without changing the factors that increase their populations is likely to be futile over the long-term." AR-31295.
The Forest Service was also concerned that the Draft EA failed to consider the best science on PDM:
There is a more current and extensive literature base that provides alternative evidence on the effectiveness and efficiency of predator control that was not considered included in the discussions. For example: [listing 9 scientific papers discussing the effectiveness of predator control]. This is not an exhaustive list and a more thorough literature review ought to be done.
AR-31002.
Comments on Draft EA-Environmental Impacts
The agencies also took issue with the Draft EA's failure to evaluate the likely environmental effects of its proposed actions. Some of these comments discussed the issue of "trophic cascade," referring to the impacts that "cascade" down the food chain when a predator is removed. For example, the removal of a predator might allow the predator's herbivore prey to multiply, resulting in more consumption of certain plants. AR-031335. The Forest Service was critical of the Draft EA's discussion of trophic cascades, commenting *1142that it is "written off in too broad and unsubstantiated way[s] thus dismissing the need to seriously consider if these issues warrant being considered in detail." AR-31071.
The BLM had a similar view, commenting that the Draft EA ignored current science on how prey and their surrounding environment are affected when predators are removed:
This section would be greatly improved with a more balanced discussion of the implications of predator control on prey species and ecological processes in general. As it reads now, the document ignores years of research on the nature of complex predator/prey relationships. How does predator control impact the stability of prey species populations relative to their competitors? How does predator control impact carrying capacity? How does predator control impact cyclic vs. non-cyclic prey populations? How does predator control affect the composition and abundance of other predator species or a guild of other prey species?
AR-31292.
Comments on Draft EA-Lack of Site-Specific Studies
The Draft EA did not include site-specific evaluations, and the agency comments were critical of that failure. For example, the BLM commented that "[i]n most cases coyotes are generalist predators, however, there is evidence that they also perform as specialist predators in more simple ecosystems. This points to the need for more site-specific analyses of predator removal." AR-31341.
The Forest Service expressed concern that the lack of information on the effects of local control activities makes it difficult to evaluate the conclusion that those activities will have no significance on a state-wide basis:
There is a lack of specificity of species' biology and habitat requirements (i.e. population structure, home range size, relationship of individuals to populations, etc) which means the general statements on the focus of local control activities and how small those are in relation to the state fail to provide any disclosure of effects at a scale below the state-level which I think is necessary to support the conclusions made of non-significance at the state-scale.
AR-31003. In response to these criticisms of a lack of site-specific information, Wildlife Services stated that its operations will only occur for short periods (90 to 180 days) during the time of year when addressing predators is most beneficial like the spring (during calving and lambing). AR-37381. To this, the BLM commented that "[t]he length of time or season that predator removal occurs during is only one component of intensity. What about predator removal operations that occur during the same season at the same location for several years in a row?" AR-031341. The BLM noted that 180 days is "half the year, so not a 'short period.' " Id.
Final EA
In preparing the final EA, Wildlife Services added over 200 pages of responses to the numerous concerns of the other agencies. see, e.g., AR-37182-202; 37380-89; 37465-520; 37497-99, 37616-19. For the most part, Wildlife Services rejected criticisms by stating that its predator damage management (PDM) activities would be (1) dispersed throughout the state; (2) carried out only during a short period of the year; (3) unlikely to adversely affect the targeted population; and (4) not designed to eliminate predators on a long-term basis. In addition, Wildlife Services found inapplicable the scientific articles raising concerns *1143about the effectiveness of PDM and its effects on the environment.
Because the agency's responses are crucial to a legal analysis of the Final EA and FONSI, the Court will discuss in detail those responses below.
Response to Comments on Draft EA-Objectivity
In response to comments that the Draft EA was not the objective study that NEPA requires, and that the proposed expansion of PDM was "highly controversial" or "highly uncertain," either of which would require an EIS, Wildlife Services stated that,
[t]he failure of any particular special interest group to agree with every act of a federal agency does not create a controversy and the NEPA does not require the courts to resolve disagreements among various scientists as to the methodology used by an agency to carry out its mission ( Marsh vs. Oregon Natural Resources Council , [490] U.S. 360, 378, 109 S.Ct. 1851, 104 L.Ed.2d 377 (1989) ).
AR-37241.1 The agency went on to praise its scientists who provided the environmental analysis as "leaders in the design and implementation of controlled studies to evaluate predation and predator control methods. They collaborate with experts from around the world to conduct these studies and [their] findings are published in peer-reviewed literature." AR-37247.
Response to Comments that PDM is Ineffective
The Final EA responds to comments that lethal PDM methods are ineffective by citing studies showing that they work:
Research has shown that in areas without some level of damage management, losses of adult sheep and lambs to predators can be as high as 8.4% and 29.3% of the total number of sheep, respectively (Henne 1975, Munoz 1977, O'Gara et al. 1983). Additional research has indicated that sheep and lamb losses are generally lower where predator damage management was applied (Nass 1977, Tigner and Larson 1977, Howard and Shaw 1978, Shaw 1987, Howard and Booth 1981). Shwiff and Merrell (2004) reported a 5.4% increase in the numbers of calves brought to market when coyotes were removed by aerial operations. Wagner and Conover (1999) found that total lamb losses declined 25% on grazing allotments in which coyotes were removed by winter aerial operations five to six months ahead of summer sheep grazing.
AR-37246-47. But studies such as these were criticized in a 2016 study entitled "Predator Controls Should Not be a Shot in the Dark, referred to as the Treves study, published in the peer-reviewed journal "Frontiers in Ecology and the Environment." AR-37616. Treves criticized the lack of scientific rigor in existing studies on PDM effectiveness, and specifically castigated the Wagner and Conover study (quoted above by Wildlife Services) as containing *1144numerous design flaws that caused biased results. AR-37617. Treves recommended a suspension of PDM until more rigorous studies could prove PDM's efficacy. AR-37247.
The Treves' criticisms hit a nerve-the co-author of the Wagner and Conover study is Dr. Kimberly Wagner, the primary writer and editor of the Final EA. AR-37533. The Final EA strikes back at the Treves study, accusing it of being "purposefully deceptive or sloppy." AR-37618-19. In a four-page analysis, the Final EA concludes that Wildlife Services' own scientists "do not agree with the authors' assessment that existing research is flawed," and attacks the Treves study, claiming it (1) "makes a fundamental error in interpreting the [Wagner and Conover] study design," (2) "selectively disregards studies from Australia," and (3) "misrepresented another study conducted by Dr. Eric Gese [a Wildlife Services scientist]." AR-37618-19.
The IDFG had commented on the Draft EA that it failed to provide an objective study of PDM's effectiveness, in part because it did not discuss an "extensive study [of Wildlife Services] involvement in southern Idaho, Hurley et al. 2011." AR-32058. The Hurley study concluded that coyote removal by Wildlife Services in southeastern Idaho did not benefit mule deer populations. In the Final EA, Wildlife Services does not discuss the Hurley study in any detail, but merely states that "based on information from [Hurley] and published studies ... coyote removals for deer population enhancement are not anticipated." AR-37184.
Response to Comments on Draft EA-Environmental Impacts
In the Final EA, Wildlife Services responded to criticisms that it had ignored concerns about the impacts of PDM on the environment, specifically the potential for trophic cascades and adverse effects on biodiversity. AR-37509. In a section of the Final EA entitled "WS-Idaho Impact on Biodiversity and Ecosystem Resilience," the agency reviewed numerous studies concluding that predator removal can, in certain instances, have an adverse effect on biodiversity, harm ecosystem resilience, and cause unintended consequences down the food chain (trophic cascades). AR-37380-89.
About the biodiversity concerns raised by these studies, Wildlife Services found them inapplicable because they examined ecosystems where a predator had been completely removed for a period of years, while Wildlife Services' operations would be more limited:
WS-Idaho damage management activities would occur in localized areas of Idaho and would not be conducted throughout the year, but would occur for short periods after damage had occurred (i.e., reactive damage management). Activities conducted to reduce threats of damage (i.e., proactive damage management) would likely occur for short periods (90 to 180 days) during the time of year when addressing predators would be the most beneficial to reducing threats of damage (e.g., the period of time immediately preceding and during calving and lambing in the spring). On average, WS-Idaho conducted activities on properties comprising nearly 6.2 million acres during FY 2013, which would represent about 11.7% of the land area per year in Idaho. WS-Idaho generally only conducts activities on a small portion of the land acres allowed under an MOU, AWP, cooperative service agreement or other comparable document. For example, a landowner may allow WS-Idaho to conduct activities on the 1,000 acres they own but WS-Idaho personnel might only conduct activities on 5 *1145acres of the property. In addition, the number of predators addressed annually by WS-Idaho and other entities is likely a small percentage of the actual populations of those species in the State. Therefore, the effects on biodiversity would be of low magnitude.
AR-37381-82.
In that same section, Wildlife Services responded to criticisms that the Draft EA ignored the potential for trophic cascades and mesopredator release. AR-37383. Wildlife Services reviewed numerous studies discussing how tropic cascades result from the removal of an apex predator like wolves, or a mesopredator like coyotes, from the environment:
WS-Idaho has reviewed these studies but many are not applicable to the types of PDM proposed for Idaho because they involve reviews of the complete absence of apex consumers from the system.
....
[T]he current program would not result in the elimination of any predator population and impacts on apex predators are only temporary and in relatively small or isolated geographic areas, compared with population levels of target species and would not have significant overall impacts on prey populations or ecosystem function. Consequently, this alternative is not anticipated to result in ecosystem impacts noted in the studies with complete removal or sustained extreme reductions in predator populations.
AR-37387.
Regarding ravens, the Final EA concludes that the preferred alternative "would likely slightly increase target animal take, especially ravens." AR-37457. However, the Final EA concludes that the increased raven take would be inconsequential:
WS-Idaho take under this alternative would only result in an overall estimated reduction of [less than] 600 ravens per year or 1.2% of the estimated population. Assuming that known cumulative human-caused raven mortality (Table 4.16) is additive to all other sources of mortality, WS-Idaho's raven take of 2,291 birds would result in only a little more than the annual population increase from reproduction. However, if raven mortality is in some part compensatory to other forms of mortality (i.e. some of the ravens killed by WS-Idaho would have died anyway from other causes) then the Idaho raven population would continue to increase making trophic cascade inconsequential or discountable because WS-Idaho's influence on the raven population would never result in the total extermination or permanent displacement of ravens in Idaho. Impacts on local areas where raven removals are conducted are also likely to be short-term because of the mobility of ravens and reproduction by individuals remaining in the population. Consequently, indirect ecosystem level impacts of the proposed raven removal are unlikely because of the limited scope and duration of raven population impacts. Additionally, the raven population in Idaho is currently at an artificially high level in the State, supported by human-generated artificial food and habitat. Even short-term local population reductions could help to return raven densities to levels more consistent with available habitat and natural resources. Consequently, a reduction in raven densities is not anticipated to result in adverse indirect impacts on prey populations or trophic cascades.
AR-37458.
Regarding PDM's effects on species protected under the Endangered Species Act (ESA), Wildlife Services states that it will *1146complete "a Section 7 consultation with the [Fish & Wildlife Service] to ensure compliance with the ESA and to ensure that the proposed management actions are not likely to jeopardize the continued existence of any endangered or threatened species." AR-37234.
Preparation of Final EA and FONSI
As these responses make clear, Wildlife Services made almost no changes to either its assumptions or its proposals contained in the Draft EA. In November of 2016, Wildlife Services issued its Final EA, followed by a Decision and Finding of No Significant Impact (FONSI) in November of 2016. The FONSI selected the preferred alternative, which would expand Wildlife Service's PDM program to include lethal and non-lethal removal of predators threatening game animals and protected species.
The plaintiffs responded by filing this lawsuit arguing, among other things, that Wildlife Services should have prepared an Environmental Impact Statement (EIS). The parties have filed cross-motions on this issue, and the Court will resolve it below.
ANALYSIS
Legal Standards for Preparing an Environmental Impact Statement
An Environmental Impact Statement ("EIS") is required for all "major Federal actions significantly affecting the quality of the human environment." See 42 U.S.C. § 4332(2)(c). Whether the action "significantly" affects the quality centers on two factors: context, which requires analysis of an action's impact at multiple levels (e.g.-regional, national, local), and intensity, which is the "the severity of the impact." See 40 C.F.R. § 1508.27. Intensity is evaluated by enumerated factors, which may individually or collectively constitute sufficient intensity to require an EIS. Blue Mtns. Biodiversity Project v. Blackwood , 161 F.3d 1208, 1212 (9th Cir.1998).
Intensity Factor-"Controversial"
One of the intensity factors examines whether the federal action is "controversial"-that is, whether "substantial questions are raised as to whether a project ... may cause significant degradation of some human environmental factor," Northwest Envtl. Def. Ctr. v. Bonneville Power Admin. , 117 F.3d 1520, 1539 (9th Cir.1997), or there is "a substantial dispute [about] the size, nature, or effect of the major Federal action." Blue Mtns. , 161 F.3d at 1212. A substantial dispute exists when evidence, raised prior to the preparation of an EIS or FONSI, casts serious doubt upon the reasonableness of an agency's conclusions. Id. NEPA then places the burden on the agency to come forward with a "well-reasoned explanation" demonstrating why those responses disputing the EA's conclusions "do not suffice to create a public controversy based on potential environmental consequences." National Parks & Conservation Ass'n v. Babbitt , 241 F.3d 722, 736 (9th Cir. 2001), abrogated on other grounds by Monsanto v. Geertson Seed Farms , 561 U.S. 139, 130 S.Ct. 2743, 177 L.Ed.2d 461 (2010). The term "well-reasoned explanation" is simply a less direct way of saying that the explanation must be "convincing." Id.
Intensity Factor-"Highly Uncertain"
Another intensity factor examines whether "uncertainty may be resolved by further collection of data, or where the collection of such data may prevent speculation on potential ... effects." Native Ecosystems Council v. U.S. , 428 F.3d 1233, 1240 (9th Cir. 2005). This factor was a key to a court holding that Wildlife Services must prepare an EIS to evaluate wolf removal *1147designed to reduce livestock predation. In Wildlands v. Woodruff , 151 F.Supp.3d 1153, 1165 (W.D. Wash. 2015), plaintiffs challenged Wildlife Services' EA and FONSI approving a project to remove wolves by trapping and lethal methods. The court held that Wildlife Services had to prepare an EIS in part because "it is highly uncertain whether lethal wolf removal actually reduces livestock depredation," noting that the comments revealed a "dispute within the scientific community." Id. at 1166.
Intensity Factor-"Unique Lands"
An action may also require an EIS when it takes place in lands with unique geographic characteristics, "such as proximity to historic or cultural resources, park lands, prime farmlands, wetlands, wild and scenic rivers, or ecologically critical areas." See 40 C.F.R. § 1508.27(b)(3).
Analysis of Intensity Factors-Controversy and Uncertainty
Here, several federal agencies with long experience and expertise in managing game animals and protected species-precisely the type of animals that Wildlife Services was proposing to protect in its expanded PDM program-expressed serious concerns about that proposed expansion. Wildlife Services rejected most of those concerns, raising the central issue in this case: Did Wildlife Services give "well-reasoned" and "convincing" explanations for that rejection?
Throughout its responses to criticisms, Wildlife Services often repeated four points, alleging that its PDM operations were (1) dispersed throughout the state; (2) carried out only during a short period of the year; (3) unlikely to adversely affect the targeted population; and (4) not designed to eliminate predators on a long-term basis.
But those explanations are not convincing when applied to the predator most often targeted by Wildlife Service for lethal or nonlethal removal-the coyote. For example, as the BLM pointed out, these removals could occur for a period of 180 days, or half the year, and nowhere does Wildlife Services rebut or allay the BLM's concern that predator removal operations could "occur during the same season at the same location for several years in a row." AR-31341. Indeed, Wildlife Services noted in the Final EA that "[i]n some local areas of Idaho, coyote removals may be conducted annually over a period of several years." AR-37305. Thus, the agency's "short period" argument is not convincing.
The explanation that PDM operations are dispersed throughout the State is contradicted in another part of the Final EA where Wildlife Services states that impacts from its PDM will occur "in relatively small or isolated geographic areas ...." AR-37389. The local nature of impacts is seen clearly with coyote removals, which are conducted on about 11.6% of Idaho land, AR-37306. The dispersed impacts explanation is therefore not convincing.
The lack of dispersed impacts for coyote removal, in combination with the removals being conducted repeatedly for years, raises a concern that local populations could be depleted below sustainable levels, a concern expressed by both the Forest Service and BLM raised, as discussed above. Lacking any site-specific information, and relying entirely on State-wide information, Wildlife Services could easily have missed the adverse impacts on local populations. The Final EA contained no convincing explanation as to why Wildlife Services refused to collect site-specific information to cure this blind spot.
The lack of crucial data is exacerbated by the unreliability of the data that Wildlife Services does possess. For example, *1148Wildlife Services' Assistant Regional Director concedes that the agency's data on coyote killings on BLM land is inaccurate. See Yeary Declaration (Dkt. No. 22-3 ). He alleges that some of the coyote killings listed in that data base should not have been counted because they occurred off BLM land or were double-counted. Id. at ¶ 8. For example, he notes that in the Twin Falls District for 2011, the coyote take was between 430 and 839, rather than the 1,345 that were originally reported. Id. at ¶ 9. And for 2013, the coyote take "was between 155 and 629, rather than the 629 that was originally reported." Id.
Yeary's "corrections" do not instill confidence in Wildlife Services' own data. After Yeary corrects the numbers, he still has a substantial "margin of error." For example, Yeary apparently cannot be any more precise than to say that the coyote takings for Twin Falls in 2013 are somewhere between 155 and 629. He does not explain that massive gap between the high and low numbers. The high number in his range is identical to the supposedly inaccurate number he contests as being too high. And if some coyotes were in fact taken off BLM land but still within the region, shouldn't they be counted for purposes of determining whether the local population is being depleted below sustainable numbers? The only conclusion that can be drawn from Yeary's "corrections" is that the Wildlife Services' data is an inadequate basis for the agency's evaluations of environmental impacts. Thus, when Wildlife Services explains that its PDM program is "not likely to result in depleted coyote populations," that explanation is not convincing. AR-37305. Similarly, Wildlife Services' confidence that trophic cascades will not result from its PDM appears misplaced. The lack of reliable data infects all the agency's conclusions.
A similar case where Wildlife Services failed to collect crucial data to support its predator removal program is Sierra Club v. U.S. Fish & Wildlife Service , 235 F.Supp.2d 1109, 1134 (D. Or. 2002). There, the court held that where Wildlife Services failed to convincingly address comments to its Draft EA that it lacked critical information on cougar density in a project area in which cougars were slated for removal, there was sufficient controversy and uncertainty to require an EIS. Id. That result applies with equal strength to this case.
Similarly, unconvincing are the agency's attempts to explain away scientific challenges to the effectiveness of predator removal. For example, the IDFG's comments expressed concern that the Draft EA failed to discuss a study-Hurley (2011)-that found no benefit from a Wildlife Services' program to remove coyotes in southeastern Idaho to benefit mule deer populations. The IDFG urged Wildlife Services to analyze Hurley because it "provides a complete look at impacts of control on wildlife populations [and] does an excellent job of reviewing current literature and providing an objective view of [predator management]." AR-32058. But Wildlife Services did not provide any analysis of Hurley in the Final EA. Instead, Wildlife Services merely states that "based on information from [Hurley] and published studies ... coyote removals for deer population enhancement are not anticipated."AR-37184. Wildlife Services missed the point of the IDFG's comment-the Hurley study was an excellent source of analysis not just for that particular program but for the broader question of the effectiveness of predator removal.
Wildlife Services reacted somewhat similarly to comments critical of the agency's plan to remove ravens to protect the sage grouse. These comments cited a study by *1149Christian Hagen in 2011, and argued that predator removal for avian species like the sage grouse was ineffective. AR-37250.
Hagen's study reviewed the literature and concluded that "[p]redator management studies have not provided sufficient evidence to support implementation over broad geographic or temporal scales, but limited information suggests predator management may provide short-term relief for a population sink." AR-49814. The study also concluded that "[m]anagement of predator populations to benefit avian species can include both nonlethal and lethal techniques; the latter is often highly controversial. " AR-49816 (emphasis added).
This study would appear to be a direct challenge to Wildlife Service's raven PDM program, which is not limited to short-term relief for a population sink. At the very least, the study is some support-although not determinative on its own-for finding that the raven removal proposal is "controversial" and should be studied in an EIS.
Wildlife Services never gets to that level of analysis, stopping instead after summarizing Hagan's results and observing that "[t]he question remains as to whether or not predator management can be an effective conservation tools, and if so, under what conditions it may be appropriate to use it (Hagen 2011)." AR-37252.
Wildlife Services' response to the Treves study, discussed above, is more detailed than its response to either Hurley or Hagen. Yet because Treves is so critical of a study co-authored by the primary writer and editor of the Final EA, Dr. Kimberly Wagner, the agency's response (whether written or edited by her) feels like equal parts personal attack and scientific rebuttal-it does not so much convince the reader of the agency's position as cry out for a more objective review, the kind that could be done in an EIS.
Wildlife Services attempts to avoid this problem by pointing out that it only responds to the requests for predator control by other agencies who are in the best position to determine the effectiveness of any specific PDM program. But that is not a sufficient response under NEPA. E.g., Kern v. BLM , 284 F.3d 1062, 1073-74 (9th Cir. 2002) (holding that agency could not rely on a "promise of a later site-specific analysis" to substitute for an adequate effects analysis).
The agency's unconvincing discussions as to why Treves, Hurley, and Hagen do not apply are set forth in the Final EA among other discussions-much more convincing-either demonstrating predator removal works, or distinguishing studies showing it does not work. There is no doubt that the Final EA undertook a much broader assessment of the science than the Draft EA. But if the Final EA demonstrates anything, it is that Wildlife Services has serious disagreements with leading experts, and has not given their studies the full attention they deserve. And that demonstrates the controversy necessary under NEPA to require an EIS. See Woodruff , 151 F.Supp.3d at 1165 (holding that "significant disagreement among experts regarding the effectiveness of removal to address depredation, a question summarily dismissed by Wildlife Services," warranted an EIS due to controversy).
In conclusion, the lack of reliable data and the unconvincing responses to the serious concerns of agencies with long experience and expertise in the very area Wildlife Services sought to expand its operations into, demonstrates that the expanded PDM program is controversial, and its environmental impacts highly uncertain, so that an EIS is required under NEPA. This case is much like *1150National Parks. There, the Park Service drafted an EA evaluating a plan to increase cruise ships in Glacier Bay National Park. There, as here, the Draft EA was criticized for its incomplete treatment of environmental impacts. The Circuit found that the comments "cast substantial doubt on the adequacy of the Parks Service's methodology and data." National Parks , 241 F.3d at 736. The Circuit concluded that an EIS had to be prepared because "the agency's conclusions were not reached by reasoned extrapolation from the data, rather the data were simply insufficient." Id. at 737. The same result should occur here.
Analysis of Unique Lands Factor
Finally, the "unique lands" factor also favors preparing an EIS. Wildlife Services predicts that there is an "extremely high" probability (meaning there is a 95-100% chance of it occurring) that it will conduct operations in the Boulder-White Clouds Complex in the next five years. AR-37586. It also predicts that there is a "high" probability (66-95% confidence) it will conduct operations in a Wilderness Study Area and an Area of Critical Concern. AR-37587. This is yet another reason for requiring an EIS in this case.
Conclusion
The criticisms of the Forest Service, BLM, Fish and Wildlife Service, and the IDFG make this a unique case; it is rare for the Court to encounter such an unanimity of critical comments from other agencies. Their comments, the lack of reliable data, and the unconvincing responses from Wildlife Services, trigger three intensity factors that combine to require Wildlife Services to prepare an EIS. The factors of controversy, uncertainty, and unique lands, all weigh in favor of an EIS in this case.
If an agency decides not to prepare an EIS, it must supply a "convincing statement of reasons" to explain why a project's impacts are insignificant. Blue Mtns. , 161 F.3d at 1212. The statement of reasons is crucial to determining whether the agency took a "hard look" at the potential environmental impact of a project." Id. As explained above, Wildlife Services failed to take the required hard look in its EA and FONSI. Consequently, the Court will grant plaintiffs' motion for summary judgment and deny the motion filed by Wilderness Services.
Typically, the Court, having granted summary judgment on the ground that an EIS should have been prepared, would remand the matter to the agency to prepare an EIS. In this case, however, the parties have indicated they desire further input on remedies. The Court will therefore direct the Clerk to set up a status conference to discuss the next step in this litigation. The Court will therefore delay entering a separate Judgment, and await counsels' input at the status conference.
The plaintiffs filed a motion to consider extra-record material. The Court will deem it moot on the ground that the Court ruled in plaintiffs' favor without considering the extra-record material.
ORDER
In accordance with the Memorandum Decision set forth above,
NOW THEREFORE IT IS HEREBY ORDERED, that the plaintiffs' motion for summary judgment (docket no. 18) is GRANTED.
IT IS FURTHER ORDERED, that the defendant's motion for summary judgment (docket no. 24) is DENIED.
*1151IT IS FURTHER ORDERED, that the motion to supplement administrative record (docket no. 19) is DEEMED MOOT.