The United States Forest Service proposed eight timber sales in the Upper Sunday Creek Watershed region of the Kootenai National Forest in northwest Montana. The environmental impact statement it prepared *757 in anticipation of the sales evaluated the project’s impact on a number of “sensitive species” living in that region. Plaintiffs, a number of environmental groups, challenged the sale first in administrative hearings and ultimately in district court, claiming that the Service’s analysis of the sale’s impact on seven species — the lynx, boreal owl, flammulated owl, black-backed woodpecker, fisher, bull charr, and wet-sloped cutthroat trout— was inadequate under both the National Forest Management Act, 16 U.S.C. §§ 1600 , et seq., and the National Environmental Policy Act of 1969, 42 U.S.C. §§ 4321 et seq. The district court concluded that the Service’s analysis was sufficient and thereafter granted summary judgment for the Service and refused to enjoin the sales. In this expedited appeal, Plaintiffs now argue: (1) that the Service failed to comply with 36 C.F.R. § 219.19 , which requires a minimum level of population viability analysis; and (2) that the Service violated the National Environmental Policy Act because the viability analysis it did perform only examined the effect of the timber sales on wildlife populations living within the project boundaries. Plaintiffs also request fees under the Equal Access to Justice Act, 28 U.S.C. § 2412 .

The district court had jurisdiction pursuant to 28 U.S.C. § 1331 , 28 U.S.C. § 2201 , and 28 U.S.C. § 2202 . We have jurisdiction under 28 U.S.C. § 1291 and affirm the decision of the district court.

I. Background

A. The National Forest Management Act

The National Forest Management Act (“NFMA”), 16 U.S.C. §§ 1600 et seq., requires the Secretary of Agriculture to develop land and resource management plans for units of the National Forest System. 16 U.S.C. § 1604 (a). When the Secretary develops these plans, the NFMA requires him to comply with the National Environmental Policy Act of 1969 (“NEPA”), which in turn encompasses the duty to prepare environmental impact statements (“EISs”). 16 U.S.C. § 1604 (g)(1); Idaho Conservation League v. Mumma, 956 F.2d 1508, 1511 (9th Cir.1992). The NFMA imposes substantive requirements as well, which have been promulgated as regulations. See 16 U.S.C. § 1604 (g)(3); 36 C.F.R. §§ 219 et seq.

The NFMA envisions a two-stage approach to forest planning. Mumma, 956 F.2d at 1511 ; Sierra Club v. Espy, 38 F.3d 792, 795 (5th Cir.1994). At the first stage, “a team ... develops a proposed [Land Resource Management Plan (“LRMP”) ] together with a draft and final EIS.” Mumma, 956 F.2d at 1511 (citing 36 C.F.R. § 219.10 (a) & (b)). Once the LRMP is approved, “[djirect implementation of the LRMP occurs at a second stage, when individual site-specific projects are proposed and assessed.” Id. at 1512 . These site-specific projects must be consistent with the stage-one, forest-wide plan. Id.; Sierra Club, 38 F.3d at 795 (“Site specific analysis ... must be consistent with the LRMP.”); 16 U.S.C. § 1604 (i) (“Resource plans and permits, contracts, and other instruments for the use and occupancy of National Forest System lands shall be consistent with the land management plans.”); 36 C.F.R. § 219.10 (e) (“[T]he Forest Supervisor shall ensure that ... all outstanding and future permits, contracts, cooperative agreements, and other instruments for occupancy and use of affected lands are consistent with the plan.”).

The forest and site-specific plans may be incorporated by reference, or “tiered” — so that the site-specific plan need not reiterate issues adequately discussed in the forest plan. See 40 C.F.R. 1508.28 (“Tiering is appropriate ... [f]rom a program, plan, or policy environmental impact statement to a program, plan, or policy statement or analysis of lesser scope or to a site-specific statement or analysis.”); Sierra Club, 38 F.3d at 796 ; Headwaters, Inc. v. Bureau of Land Management, Medford Dist., 914 F.2d 1174, 1178 (9th Cir.1990). Both stages must, nevertheless, fully comply with the NFMA’s regulations. See 16 U.S.C. § 1604 (i) (requiring site-specific plans to be consistent with forest plans, which in turn must be consistent with NFMA’s substantive requirements).

B. National Environmental Policy Act

The National Environmental Policy Act of 1969 requires agencies of the federal *758 government to prepare an EIS whenever they propose to undertake any “major Federal action[ ] significantly affecting the quality of the human environment.” 42 U.S.C. § 4332 (2)(C). The goal of NEPA is two-fold: (1) to ensure the agency will have detailed information on significant environmental impacts when it makes its decisions; and (2) to guarantee that this information will be available to a larger audience. Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 349 , 109 S.Ct. 1835, 1845 , 104 L.Ed.2d 351 (1989). NEPA’s goal is satisfied once this information is properly disclosed; thus, NEPA exists to ensure a process, not to ensure any result. Id. at 350 , 109 S.Ct. at 1846 (“[I]t is now well settled that NEPA itself does not mandate particular results, but simply prescribes the necessary process.”); Sierra Club, 38 F.3d at 796 .

C. Facts

The Kootenai National Forest is a 2.2 million acre tract of land nestled against the Salish Range of the Northern Rockies, in northwestern Montana. The Forest Service (“the Service”) completed its stage-one, forest-wide plan for the Forest in 1987 (hereinafter “Kootenai Forest Plan”). Five years later, the Service entertained notions of selling timber from a 28,485 acre area of the Forest known as the Sunday Creek Watershed. By late 1992, the Service refined its plans and proposed eight timber sales from a 12,374 acre tract in the upper portion of the Watershed (hereinafter “Upper Sunday”).

The Service prepared a site-specific EIS for the Upper Sunday area which contained seven alternative proposals for timber sales, and one no-action alternative. After a period of public comment, the Service decided to proceed with Alternative E-Modified. This alternative provided for the harvest of 13.7 million board feet of timber from, among other portions of the Upper Sunday area, 1,237 acres of mature (over 200 years old), interior forest called “old growth habitat.” The Service prepared a supplemental' Biological Assessment which discussed the effects of the chosen alternative on wildlife within the Upper Sunday area. In February 1994, the District Ranger signed the Record of Decision approving the Alternative E-Modified Plan. In April 1994, various groups appealed the District Ranger’s decision. The Appeals Officer affirmed the District Ranger’s approval of the timber sales, but ordered the Service to prepare additional documentation to support its environmental analysis before allowing it to continue with the sales.

Inland Empire Public Lands Council and other environmental groups (hereinafter “Plaintiffs”) filed suit in district court on August 25, 1994. Plaintiffs alleged that the Service’s Upper Sunday EIS was deficient and violated both NFMA and NEPA. Plaintiffs first contended that the EIS did not conduct a proper population viability analysis for the seven “sensitive” species living in the area: the lynx, boreal owl, black-backed woodpecker, flammulated owl, fisher, bull eharr, and the wet-sloped cutthroat trout. 1 Plaintiffs claimed that the Service fell short of what the NFMA required because it never examined the species’ population size, their population trends, or their ability to interact with other groups of the species living in neighboring patches of forest. The district court rejected this argument on summary judgment, reasoning that Plaintiffs were quibbling over the choice of scientific methodologies, a decision to which a reviewing court should defer.

Plaintiffs argued in the alternative that the Forest Service erred in confining its population viability analysis to the Upper Sunday area, rather than going beyond that area to examine the effect of the sales on populations of these species living on land “adjacent to” the project area. They claimed that this error violated the NEPA’s “cumulative impact” requirement. See 40 C.F.R. § 1508.7 . The district court granted the Service summary judgment on this issue as well, ruling that this, too, was a matter of methodology.

When the district court denied Plaintiffs’ motion for a preliminary injunction to enjoin *759 the timber sales, Plaintiffs filed this expedited appeal.

II. Population Viability Analysis

Plaintiffs first claim that the district court erred in granting summary judgment on their claim that the Forest Service’s Upper Sunday EIS violated the National Forest Management Act. We review de novo the district court’s grant of summary judgment. Nevada Land Action Ass’n v. United States Forest Serv., 8 F.3d 713, 716 (9th Cir.1993).

As noted above, the NFMA imposes substantive duties on the Forest Service, one of which is the duty to “provide for diversity of plant and animal communities.” 16 U.S.C. § 1604 (g)(3)(B). Regulation 219.19, one of the many regulations promulgated to ensure such diversity, states in relevant part that:

36 C.F.R. § 219.19 . 2 This duty to ensure viable, or self-sustaining, populations, applies with special force to “sensitive” species. Oregon Natural Resources Council v. Lowe, 836 F.Supp. 727, 733 (D.Or.1993) (“The Forest Service has interpreted the viable population provision [Regulation 219.19] as requiring additional attention to certain ‘sensitive species.’”); Forest Conservation Council v. Espy, 835 F.Supp. 1202, 1206 (D.Idaho 1993) (“In keeping with [Regulation 219.19], the Forest Service at times designates ‘sensitive species’ ... within a particular planning area.”), aff'd, 42 F.3d 1399 (9th Cir.1994). Because neither party disputes the Service’s ultimate obligation to ensure viable populations, the key to this appeal is deciding what type of population viability analysis the Service must perform in order to comply with Regulation 219.19.

Each party suggests its own answer. The Forest Service proposes that its “habitat visibility analyses” were sufficient. For four of the species (the black-backed woodpecker, lynx, fisher, and boreal owl), the Service did the following: It consulted field studies that disclosed how many acres of territory an individual of each species needed to survive and the percentage of that acreage that was used for nesting, feeding, denning, etc. (e.g. a lynx needs a 200 acre territory, 20 acres — or 10% — of which must be suitable for denning). The Service then assumed that these percentages would hold true regardless of the size of the individual’s territory (e.g. that a lynx would need 10% of whatever acreage of territory it inhabited to be denning habitat). The Service examined each proposed alternative to see how many acres of each type of relevant habitat would remain after the timber was harvested (e.g. Alternative 1 would leave 2,000 acres of denning habitat). It next determined what percentage of the decision area that the remaining types of habitat constituted (e.g. decision area was 10,000 acres so that remaining denning habitat is 20% of the decision area). The Service concluded a species would remain viable as long as the threshold percentage of each type of habitat remaining in the chosen alternative was greater than the percentage required for that species to survive (e.g. the lynx population would remain viable because Alternative 1 left 20% denning habitat and a lynx needs only 10% of its territory to be suitable for denning). See Final EIS at III:38-39, 41-42, IV:76, 80, 83-87; Biological Assessment, Ad *760 dendum 2, at 3-7. 3

The Service’s analysis of the remaining species was not as detailed. As to the flam-mulated owl, the Service noted that the Upper Sunday area contained 366 acres of habitat suitable for nesting and feeding, enough for three potential owl territories; it noted that the timber sales would reduce the size of one of those territories. Final EIS at 111:40. As to the bull charr trout, the Service stated that the trout only marginally used the streams within the decision area, but that the timber sales would not appreciably raise the sediment or carbonate levels in those streams. Final EIS at IV:64-67; Biological Assessment at 21. 4

Plaintiffs contend that the Service’s manifold “habitat viability analyses” are insufficient. They argue that Regulation 219.19 also requires the Service to examine: (1) the population of each species; (2) the population dynamics (trends, etc.) of each species; and (3) whether the species could travel between different patches of forest (“linkages”). Mills Decl. 5 Plaintiffs claim that their form of analysis is the minimum required by law.

In deference to an agency’s expertise, we review its interpretation of its own regulations solely to see whether that interpretation is arbitrary and capricious. 5 U.S.C. § 706 (2)(A); Oregon Natural Resources Council v. Marsh, 52 F.3d 1485, 1488 (9th Cir.1995). This is especially true when questions of scientific methodology are involved. Inland Empire Public Lands Council v. Schultz, 992 F.2d 977, 981 (9th Cir.1993) (“We defer to agency expertise on questions of methodology unless the agency has completely failed to address some factor, ‘consideration of which was essential to a truly informed decision whether or not to prepare an EIS.’”) (citation omitted); see Sierra Club v. Marita (“Marita II”), 46 F.3d 606, 619-20 (7th Cir.1995) (holding that Forest Service’s failure to employ “conservation biology” methodology when conducting population viability analysis was not arbitrary or capricious). Thus, we will uphold the Forest Service’s interpretation “unless it is plainly erroneous or inconsistent with the regulation.” Nevada Land Action Ass’n, 8 F.3d at 717 (citations and internal quotations omitted). 6

*761 We start, as we must, with the plain language of the Regulation. Idaho First Nat’l Bank v. Commissioner of Internal Revenue, 997 F.2d 1285, 1289 (9th Cir.1993) (“If the [statutory] language ... is unambiguous, and its literal application does not conflict with the intentions of its drafters, the plain meaning should prevail.”). The Regulation specifically provides that the Forest Service may discharge its duties though habitat management as long as “habitat [is] provided to support, at least, a minimum number of reproductive individuals and that habitat [is] well distributed so that those individuals can interact with others in the planning area.” 36 C.F.R. § 219.19 (emphasis added). 7

We do not believe that the habitat management analysis conducted in this case for the black-backed woodpecker, lynx, fisher, and boreal owl was in any way “plainly erroneous” or “inconsistent” with this regulatory duty. Regulation 219.19 ultimately requires the Forest Service to maintain viable populations. In this case, the Service’s methodology reasonably ensures such populations by requiring that the decision area contain enough of the types of habitat essential for survival. In applying this methodology, the Service recognizes that decision areas are artificial boundaries that change depending on the project at issue, and that the species inhabiting these areas pay no attention to such boundaries.

We recognize that the Service’s methodology necessarily assumes that maintaining the acreage of habitat necessary for survival would in fact assure a species’ survival. The Service is entitled to rely on reasonable assumptions in its environmental analyses. See, e.g., Sierra Club v. Marita (“Marita I”), 845 F.Supp. 1317, 1331 (E.D.Wis.1994) (finding it permissible to assume that population trends affecting one species in a particular habitat will similarly affect other species in the same habitat), aff'd, 46 F.3d 606 (7th Cir.1995); Greenpeace Action v. Franklin, 14 F.3d 1324, 1335-36 (9th Cir.1992) (finding it permissible for Service to assume that declines in the Stellar sea lion population would be the same for the harbor seal population, given their similarities). We find the above-stated assumption eminently reasonable and therefore do not find that the Forest Service’s habitat analyses for the black-backed woodpecker, lynx, fisher, and boreal owl were arbitrary or capricious. 8

Nor do we believe that the less rigorous analysis performed for the remaining three species — the flammulated owl, the bull charr trout, and the wet-sloped cutthroat trout— was arbitrary and capricious. See Final EIS at 111:40, IV:81-82; Biological Assessment at 21. The Service’s failure to engage in a more intensive analysis for the bull charr trout and the wet-sloped cutthroat trout is understandable, as neither species would be *762 affected by the timber sales. See Final EIS at IV:64-67 (noting that the bull eharr trout’s habitat within the decision area would not be affected by the timber sales); Biological Assessment at 21 (same); Final EIS at 111:37 (listing the 12 sensitive species that “possibly occur[]” in the decision area and omitting the wet-sloped cutthroat trout from that list).

The Service’s treatment of the flammulated owl is also reasonable. In its EIS, the Service determined that the Upper Sunday decision area contained habitat to support three potential flammulated owl territories and concluded that Alternative E-Modified would shrink the size of the smallest of these territories from 40 to 35 acres. Biological Assessment at 29-30. The Service did not engage in a more extended analysis of the owl’s nesting and feeding habitat requirements because such data were unavailable. See Richard T. Reynolds & Brian D. Linkhart, “The Nesting Biology of Flammulated Owls in Colorado,” Biology & Conservation of Northern Forest Owls 259 (1987) (“It spite of its wide distribution, little is known of the flammulated owl’s nesting biology and population status.”). We believe that an analysis that uses all the scientific data currently available is a sound one. 9 See Seattle Audubon Soc’y v. Moseley, 80 F.3d 1401 , 1404 (9th Cir.1996) (upholding a viability analysis that was “based on the current state of scientific knowledge”). We therefore find no fault with the Service’s analysis of the flammulated owl. 10

Plaintiffs contend that we must still reverse because the Service did not comply with its duties regarding the “management indicator species.” Regulation 219.19 provides that the Service may select “certain vertebrate and/or invertebrate species present in the area” to be “management indicator species” when those species’ “population changes are believed to indicate the effects of management activities.” 36 C.F.R. § 219.19 . 11 Once an indicator species is selected, the Service is obligated to evaluate planning alternatives for projects that affect that species “in terms of both amount and quality of habitat and of animal population trends of the management indicator species.” 36 C.F.R. § 219.19 (a)(2).

We believe that the Service has satisfied this obligation. In this case, the Service selected the pileated woodpecker as the indicator species for the old growth habitat. See Final EIS at 111:42. The old growth areas provide special feeding and nesting conditions, upon which several species are dependent. The Service evaluated the various planning alternatives for the Upper Sunday timber sales in terms of how they would affect the old growth forest and the number of pileated woodpecker nesting and feeding territories therein. See Final EIS at IV:88-89. Specifically, the Service found that the chosen alternative would reduce the pileated woodpecker’s old growth nesting habitat by 11-12% and the feeding habitat by 11-15%. Final EIS at 11:33. Such reduction would “eliminate one nesting block ... and reduce the number of home ranges that support good amount and distribution of nesting and feeding habitat from 10 to 9.” Final EIS at IV:88. Because the number of remaining *763 nesting and feeding territories has a direct impact on the population of the species, the EIS effectively predicts a slight downward population trend in pileated woodpeckers as a result of the timber sales. We therefore do not believe that the Service acted arbitrarily or capriciously when it estimated the effects of the alternatives on the population of the management indicator species by analyzing the amount of the species’ habitat that would be reduced by each alternative. In short, the Service complied -with Section 219.19(a)(2). 12

We therefore affirm the district court’s conclusion that the Service’s population viability analysis was not “arbitrary and capricious.”

III. Cumulative Impacts Analysis

Plaintiffs next argue that the Service had a duty under NEPA to discuss how the Upper Sunday project would affect the populations of sensitive species living both within and “adjacent to” the project area. 13 As discussed above, NEPA is a “process”-oriented statute that does not mandate a specific result. Robertson, 490 U.S. at 350 , 109 S.Ct. at 1846 . Consequently, we may only examine whether “the agency has taken a ‘hard look’ at environmental consequences.” Kleppe v. Sierra Club, 427 U.S. 390 , 410 n. 21, 96 S.Ct. 2718 , 2730 n. 21, 49 L.Ed.2d 576 (1976); Marsh, 52 F.3d at 1488 . Only if the agency’s analysis of environmental consequences is “arbitrary and capricious” or “contrary to the procedures required by law” can we conclude that the agency did not take a “hard look.” Marsh, 52 F.3d at 1488 .

Regulations promulgated under NEPA state that an EIS must consider “[i]mpaets, which may be (1) [d]ireet; (2) indirect; [or] (3) cumulative.” 40 C.F.R. § 1508.25 (c). A “direct effect” is an effect “caused by the action and occur[ring] at the same time and place.” 40 C.F.R. § 1508.8 (a). An “indirect effect” is an effect which is both:

40 C.F.R. § 1508.8 (b) (emphasis added). A “cumulative impact” is defined as:

40 C.F.R. § 1508.7 .

Plaintiffs argue that the timber sales’ impact on animal populations “adjacent” to but outside the Upper Sunday area are “indirect” and “cumulative” effects of the sales because those populations are part of the same “ecosystem” as the populations within the area. They contend that the Service should not be able to define the scope of its population analysis on the basis of artificial project boundaries that ignore how species’ populations are part of an ecosystem.

We reject Plaintiffs’ reading of these regulations. As an initial matter, we do not believe that Plaintiffs base their argument on the proper regulations. When environmental plaintiffs challenge an EIS’s “cumulative” *764 and “indirect” effects analysis, they usually contend that the EIS failed to consider the environmental effects , of the project at issue in conjunction with the environmental effects of other past, present, and reasonably foreseeable future projects. See, e.g., Marsh, 52 F.3d 1485 (finding an EIS. for dam project inadequate because it did not properly consider the cumulative effects of the dam project and two earlier dam projects on the fish populations in the river); Thomas v. Peterson, 753 F.2d 754 (9th Cir.1985) (remanding for cumulative impact analysis of timber sales and the construction of a road necessary to facilitate those sales); Natural Resources Defense Council, Inc. v. Callaway, 524 F.2d 79 (2d Cir.1975); Natural Resources Council, Inc. v. Hodel, 865 F.2d 288 (D.C.Cir.1988). In this appeal, Plaintiffs argue that the geographic scope of the Service’s EIS for this project is too small. This does not appear to be a cumulative impact challenge.

We furthermore believe that adopting Plaintiffs’ position as a rule of law would be impractical. Under such a rule, an agency would have to analyze separately each species to determine the area covered by its particular ecosystem and then analyze its population viability in that area; this task could become particularly burdensome if there are a number of different species to examine, each with a different population ecosystem area to analyze. See Seattle Audubon Soc’y v. Lyons, 871 F.Supp. 1291, 1312 (W.D.Wash.1994) (“[T]o plan based on different geographic boundaries for every species in the same ecosystem would be impractical.”). 14 . NEPA does not require the government to do the impractical. Kleppe, 427 U.S. at 414 , 96 S.Ct. at 2732 (noting that “practical considerations of feasibility might well necessitate restricting the scope of comprehensive statements”); Krichbaum v. Kelley, 844 F.Supp. 1107, 1118 (W.D.Va.1994) (“This claim ... would require a level of analysis sufficient to stop all action in the Forest while every conceivable effect is catalogued-. [Plaintiffs insistence [on this action] ... is unavailing on an arbitrary and capricious standard.”), aff'd, 61 F.3d 900 (4th Cir.1995). We therefore hold that the Forest Service did all it was obligated to do. 15

Even if we were to assume that the Service could not confine its analyses to the project boundaries, the Forest Service’s EIS is nevertheless valid. The Service never limited its analysis of cumulative effects to the Upper Sunday area. For the. black-backed woodpecker, boreal owl, flammulated owl, lynx, and fisher, the Forest Service extended its analysis beyond the 12,345 acre Upper Sunday area to include the entire 28,485 acre Watershed. See Final EIS at IV:74, 77, 82, 83 (black-backed woodpecker, boreal owl, and lynx analyses cover Upper Sunday Watershed area); Biological Assessment, Addendum 2, at 5, 8 (same for flammulated owl and fisher analyses). To the extent that they challenge that the Service’s decision not to extend its analysis beyond the Watershed, Plaintiffs have advanced no proof why this decision is arbitrary and capricious, as is their burden.

*765 We therefore affirm the district court on this issue.

IV. Attorneys’ Fees

Plaintiffs have requested attorneys’ fees under the Equal Access to Justice Act (hereinafter “EAJA”). The EAJA awards fees to a party who prevails against the government if the government cannot show that its position in the litigation was “substantially justified.” 28 U.S.C. § 2412 (d)(1)(A),(B); Kali v. Bowen, 854 F.2d 329, 332 (9th Cir.1988). The EAJA applies in environmental cases. Marsh, 52 F.3d at 1491 . Because Plaintiffs did not prevail in this litigation, we decline to award fees.

CONCLUSION

We therefore AFFIRM the decision of the district court and uphold the Forest Service’s population viability and cumulative effects analyses. We DENY Plaintiffs’ request for attorney’s fees.

. "Sensitive species" are those species that are not endangered, but for which there is concern about the viability of their populations.

. 36 C.F.R. § 219.27 (a)(6) also requires: "All management prescriptions shall ... [plrovide for adequate fish and wildlife habitat to maintain viable populations of existing native vertebrate species and provide that habitat for species chosen under § 219.19 is maintained and improved to the degree consistent with multiple-use objectives established in the plan.” Because this section refers to § 219.19, our Regulation 219.19 analysis applies with equal force to this section.

. This example is meant to be illustrative only. For each species, the particular habitat analyzed and the effects of the various alternatives on those habitats varied.

. The Service conducted no analysis at all for the wet-sloped cutthroat trout, presumably because it did not believe that the trout inhabited the decision area. See Final EIS at 111:37.

. The Service argues that we cannot consider Mills' declaration, because it was submitted with Plaintiffs' district court motion for summary judgment and was never part of the administrative record. Although the general rule prevents consideration of evidence outside the administrative record, "the court may consider, particularly in highly technical areas, substantive evidence going to the merits of the agency’s action where such evidence is necessary as background to determine the sufficiency of the agency’s consideration." Love v. Thomas, 858 F.2d 1347, 1356 (9th Cir.1988), cert. denied sub nom. AFL-CIO v. Love, 490 U.S. 1035 , 109 S.Ct. 1932 , 104 L.Ed.2d 403 (1989) (citation omitted). This is because “[i]t will often be impossible, especially when highly technical matters are involved,. for the court to determine whether the agency took into consideration all relevant factors unless it looks outside the record to determine what matters the agency should have considered but did not.” Asarco, Inc. v. EPA, 616 F.2d 1153, 1160 (9th Cir.1980). To the extent Mills' declaration is submitted to show that the Service overlooked factors relevant to a proper population viability analysis, we will consider it.

.As a preliminary matter, the Service argues that Plaintiffs' claim is procedurally barred. The Service cites Environment Now! v. Espy, 877 F.Supp. 1397, 1420 (E.D.Cal.1994), for the proposition that Regulation 219.19 "applfies] solely to the promulgation and management of forest plans, not individual timber sales” so as to ban a challenge to the site-specific Upper Sunday plan. Environment Now! adopted the rationale of an Eighth Circuit case, Sharps v. United States Forest Serv., 28 F.3d 851 , 855 (8th Cir.1994), and held that “only challenges to regional guides or forest plans may be brought under [these] provisions.” Environment Now!, 877 F.Supp. at 1420 . The Sharps court relied on the analysis of its district court, which had noted that Regulation 219.19 limited the management of habitat to the "planning area” and that Regulation 219.3 defined "planning -area” as "the area of the Forest System covered by a regional guide or forest plan "&emdash;which, according to the district court, would exclude district-level, site-specific plans. Sharps v. United States Forest Serv., 823 F.Supp. 668, 679 (D.S.D.1993), aff'd, 28 F.3d 851 (8th Cir.1994). (8th Cir.1994).

*761 We are not persuaded. We believe that the Sharps court's reading of Regulations 219.19 and 219.3 is incorrect. Regulation 219.3 defines the "planning area” as "the area ... covered by a regional guide or forest plan.” 36 C.F.R. § 219.3 (emphasis added). Because any district contained within the boundaries of a forest having a plan would be an "area ... covered by a ... forest plan,” it would be also be.a planning area governed by Regulation 219.19.

. The Forest Service’s manual provides no clear guidance on whether a population viability analysis presupposes data on actual populations. Compare Forest Service Manual, § 2621.2 (noting that the “biological assessment” documents required for sensitive species must “[(Identify and consider ... factors that may affect the downward trend of the population ”) (emphasis added) and id. at § 2670.5(22) (defining “viable population” as “[a] population that has the estimated numbers and distribution of reproductive individuals to ensure the continued existence of the species ... ”) (emphasis added) and id. at § 2672.1 ("There must be no impacts to sensitive species without an analysis of the significance of adverse effects on the populations, its habitats, and on the viability of the species as a whole.”) (emphasis added) with id. at § 2670.32(4) ("If impacts [on sensitive species] cannot be avoided, analyze the significance of potential adverse effects on the population or its habitat within the area of concern ... ”) (emphasis added) and id. at § 2670.22(3) (requiring Service to "[d]evelop and implement management objectives for populations and/or habitat of sensitive species”) (emphasis added).

. We therefore reject Plaintiffs’ argument that the Service must assess population viability in terms of actual population size, population trends, or the population dynamics of other species. We do not mean to suggest, however, that Plaintiffs' suggestions are in any way improper. Indeed, we would encourage such analyses and hold only that they are not required.

. Plaintiffs' experts seem to agree with us on this point. See Mills’ Decl. at 8 (“For these situations [where population-specific demography and life history information is not available], there are a number of alternative approaches and rules of thumb that can be used with whatever data are available.").

. We also doubt that the flammulated owl will be greatly affected by the timber sales. These owls require only 20 acres for a territory, see Biological Assessment at 29, and the timber sales will still leave 35 acres in the smallest of the three potential flammulated owl territories.

. A species chosen as a management indicator species is used as a bellwether — a class representative, if you will — for the other species that have the same special habitat needs or population characteristics. The Service must choose management indicator species "where appropriate” to represent certain classes of species in order to estimate the effects of proposed management activities on fish and wildlife populations. In this case, the Service chose the pileated woodpecker as the management indicator species for old growth dependent species. Old growth forest comprises 20% of the Upper Sunday decision area, and provides special feeding and nesting conditions for several species. The use of management indicator species is intended to allow the Service to thoroughly evaluate the effects of the alternatives on fish and wildlife populations by using a "class representative,” without having to evaluate each species individually.

. For the reasons stated above, we believe that the Service satisfied its obligations under § 219.19(a)(6). See 36 C.F.R. § 219.19 (a)(6) ("Population trends of the management indicator species will be monitored and relationships to habitat determined.”). The Service specifically found that for the smaller, more reclusive species, such as the pileated woodpecker, there is no technically reliable and cost-effective method of counting individual members of the species. In light of the Service’s alternative method of population trend analysis, its failure to monitor the actual population of the pileated woodpecker is not dispositive or unreasonable. See Forest Monitoring Report for Fiscal Year 1992, at 27.

. To the extent that Plaintiffs are simply putting a new face on their Regulation 219.19 argument that the Service's population viability analysis was inadequate, Regulation 219.19's language limiting such analysis "to the planning area" forecloses that argument.

. It would also be difficult to determine when land stops being “adjacent.”

. Marsh does not dictate otherwise. There, we were asked to decide whether a revised EIS covering a dam project properly analyzed the cumulative effects of the instant project in light of two previously constructed dams. In concluding that the EIS was deficient, we were particularly disturbed by the revised EIS’ discussion of the dam’s effects on wild coho and summer steelhead populations and urged that the Army Coups examine the effect of the dam on these fish living in areas both inside and outside the project area. 52 F.3d at 1490 . We do not believe that 'our offhand suggestion stands for the proposition Plaintiffs now advance — especially since the weight of precedent favors no such rule. See Callaway, 524 F.2d at 90 (holding that the Navy, in its project to deepen the Long Island Sound, needed only to consider the effects of past projects on the areas its project covered — not the "whole Long Island Sound”); Kleppe, 427 U.S. at 414 , 96 S.Ct. at 2732 (holding that the Department of the Interior did not have to complete one comprehensive EIS for all mining projects in the ‘’Northern Great Plains Region,” reasoning that "identification of the geographic area within which [cumulative impact factors] may occur, is a task assigned to the special competency of the appropriate agencies”); cf. Hodel, 865 F.2d at 297-300 (requiring the Secretary of Interior to analyze the cumulative effects of offshore drilling near California and Alaska together because whales and salmon would pass through both project drilling areas in the normal course of migration).