FOR PUBLICATION
UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
ECOLOGY CENTER, INC.,
Plaintiff-Appellant,
v.
DEBORAH AUSTIN, in her official
capacity as Forest Supervisor for
the Lolo National Forest; BRADLEY
POWELL, Regional Forester of
Region One of the U.S. Forest No. 03-35995
Service; UNITED STATES FOREST
SERVICE, an agency of the U.S.
Department of Agriculture,
D.C. No.
CV-03-00018-DWM
Defendants-Appellees, OPINION
MINERAL COUNTY; TOWN OF
SUPERIOR; ST. REGIS SCHOOL
DISTRICT NO. 6; SUPERIOR SCHOOL
DISTRICT NO. 3; MONTANA
COALITION OF FOREST COUNTIES;
TRICON TIMBER,
Defendant-Intervenors-
Appellees.
Appeal from the United States District Court
for the District of Montana
Donald W. Molloy, District Judge, Presiding
Argued and Submitted
February 14, 2005—Seattle, Washington
Filed December 8, 2005
Before: Betty B. Fletcher, M. Margaret McKeown, and
Ronald M. Gould, Circuit Judges.
16029
16030 ECOLOGY CENTER v. AUSTIN
Opinion by Judge B. Fletcher;
Dissent by Judge McKeown
ECOLOGY CENTER v. AUSTIN 16033
COUNSEL
Thomas J. Woodbury, Forest Defense, P.C., Missoula, Mon-
tana, for the plaintiff-appellant.
Aaron P. Avila, Ronald M. Spritzer, Department of Justice,
Washington, D.C., for the defendants-appellees.
16034 ECOLOGY CENTER v. AUSTIN
OPINION
B. FLETCHER, Circuit Judge:
Ecology Center, Inc. (“Ecology Center”) challenges the
United States Forest Service’s (“Forest Service” or “Service”)
Lolo National Forest Post Burn Project (“Project”), which
was designed in the aftermath of the 2000 wildfires on the
Lolo National Forest (“LNF”). Ecology Center raises a num-
ber of procedural and substantive claims under the National
Environmental Policy Act (“NEPA”) and the National Forest
Management Act (“NFMA”). We have jurisdiction pursuant
to 28 U.S.C. § 1291. Because we find that the Forest Ser-
vice’s decision to permit logging in critical old-growth forest
and post-fire habitats was arbitrary and capricious, we reverse
the grant of summary judgment to the Service and remand.
I
In 2000, wildfires burned approximately 74,000 acres on
the Lolo National Forest. While the fires caused considerable
damage to the forest, they also created habitat for species that
are dependent upon post-fire habitats, such as the black-
backed woodpecker.
In response to the 2000 fires, the Forest Service began
developing the Lolo National Forest Post Burn Project and
preparing the requisite Environmental Impact Statement
(“EIS”). The Forest Service considered four alternatives in
detail, including a “no action alternative.” In July 2002, the
Forest Service selected a slightly modified version of “Alter-
native Number Five” for the Project. This alternative
involves, inter alia, commercial thinning of small diameter
timber and prescribed burning in old-growth forest stands, as
well as salvage logging of burned and insect killed timber in
various areas of the forest.
On February 7, 2003, Ecology Center filed its complaint,
raising several claims under NEPA and NFMA. Ecology Cen-
ECOLOGY CENTER v. AUSTIN 16035
ter objects to the Forest Service’s decision to permit commer-
cial logging in old-growth forest stands, raising concerns
about the impact of such logging on the viability of species
that are dependent upon old-growth habitat, such as the pile-
ated woodpecker and the northern goshawk. Similarly, Ecol-
ogy Center questions the Service’s impact analysis of salvage
logging in post-fire habitat, particularly with respect to the
black-backed woodpecker, a sensitive species. Ecology Cen-
ter also raises concerns regarding the impact of the Project on
soil conditions and questions the reliability of the Service’s
soil quality analysis.
Ecology Center moved for a temporary restraining order
and preliminary injunction, both of which the district court
denied. The parties then filed cross-motions for summary
judgment. On November 4, 2003, the district court struck
extra-record declarations that Ecology Center had included
with its motion and granted summary judgment in favor of the
Forest Service.1
1
In separate proceedings, the Sierra Club challenged the same project at
issue here. The district court ruled in favor of the Sierra Club’s claim that
the Forest Service was required to determine how much sediment affected
streams could assimilate before implementing the Project, but ruled in
favor of the Forest Service with respect to its analysis of the Project’s
impact on unroaded areas. In a memorandum disposition decided Decem-
ber 2, 2003, we reversed the district court on both claims: we found that
the Forest Service’s assessment of the Project’s effect on water quality sat-
isfied NEPA, but that its analysis of the Project’s impact on unroaded
areas was inadequate. Sierra Club, Inc. v. Austin, 82 Fed. Appx. 570, 2003
WL 22854670 (9th Cir. 2003). The case was remanded to the district
court, and the parties then settled.
The Forest Service moved for this court to take judicial notice of the
Sierra Club settlement agreement, and for leave to file supplemental infor-
mation and declarations explaining the terms of the settlement; we granted
that motion. After taking judicial notice of the settlement, we invited the
parties to advise us as to whether the settlement moots any portion of this
appeal or affects the relief, if any, that the district court should grant.
Although the settlement significantly reduces the scope of the Project, nei-
ther party argues that Ecology Center’s claims are moot.
16036 ECOLOGY CENTER v. AUSTIN
II
“Because this is a record review case, we may direct that
summary judgment be granted to either party based upon our
de novo review of the administrative record.” Lands Council
v. Powell, 379 F.3d 738, 743 (9th Cir. 2004), amended by 395
F.3d 1019 (9th Cir. 2005). Our review of agency actions chal-
lenged under NFMA and NEPA is governed by the judicial
review provisions of the Administrative Procedure Act, 5
U.S.C. § 706(2)(A). Neighbors of Cuddy Mountain v. U.S.
Forest Serv., 137 F.3d 1372, 1376 (9th Cir. 1998); Ctr. for
Biological Diversity v. U.S. Forest Serv., 349 F.3d 1157, 1165
(9th Cir. 2003). Under the APA, we review to determine if the
agency’s actions were “arbitrary, capricious, an abuse of dis-
cretion, or otherwise contrary to law.” Lands Council, 379
F.3d at 743.
NFMA imposes both substantive and procedural require-
ments on the Forest Service. 16 U.S.C. §§ 1600-1687. Proce-
durally, it requires the Forest Service to develop a land and
resource management plan (“forest plan”) for each forest that
it manages. 16 U.S.C. § 1604(a). Subsequent agency actions
must not only comply with NFMA but also be consistent with
the governing forest plan. 16 U.S.C. § 1604(i); Idaho Sporting
Cong., Inc. v. Rittenhouse, 305 F.3d 957, 961-62 (9th Cir.
2002) (“[A]ll management activities undertaken by the Forest
Service must comply with the forest plan, which in turn must
comply with the Forest Act . . . .”). NFMA also requires that
the Forest Service adopt regulations specifying guidelines for
the achievement of NFMA’s substantive mandates. 16 U.S.C.
§ 1604(g)(3).
“In providing for multiple uses, the forest plan must com-
ply with substantive requirements of the Forest Act designed
to ensure continued diversity of plant and animal communities
and the continued viability of wildlife in the forest . . . .”
Idaho Sporting Congress, 305 F.3d at 961; 16 U.S.C.
§ 1604(g)(3)(B). In addition to the mandate to maintain wild-
ECOLOGY CENTER v. AUSTIN 16037
life viability, the Forest Service must maintain soil productiv-
ity. 16 U.S.C. § 1604(g)(3)(C).
Unlike NFMA, NEPA imposes no substantive require-
ments. Instead, it is designed “to force agencies to publicly
consider the environmental impacts of their actions before
going forward.” Idaho Sporting Congress, 305 F.3d at 963.
“Agencies must adequately consider the project’s potential
impacts and the consideration given must amount to a ‘hard
look’ at the environmental effects.” Id. Specifically, NEPA
requires the preparation of a detailed Environmental Impact
Statement. 42 U.S.C. § 4332(2)(C). An EIS must “provide full
and fair discussion of significant environmental impacts and
shall inform decisionmakers and the public of the reasonable
alternatives which would avoid or minimize adverse impacts
or enhance the quality of the human environment.” 40 C.F.R.
§ 1502.1.
A. “Treatment” of Old-Growth Habitat
The Project involves what the Forest Service characterizes
as rehabilitative “treatment” of old-growth (and potential old-
growth) forest stands; this treatment entails the thinning of
old-growth stands via commercial logging and prescribed
burning. The Forest Service cites a number of studies that
indicate such treatment is necessary to correct uncharacteristic
forest development resulting from years of fire suppression.
The Service also points out that the treatment is designed to
leave most of the desirable old-growth trees in place and to
improve their health.
Ecology Center highlights the scientific uncertainty and
debate regarding the necessity, design, and long-term effects
of such old-growth treatment. In particular, Ecology Center
alleges that the treatment of old-growth forest harms species
that are dependent upon old-growth habitat. For example,
Ecology Center claims that, even if treatment leaves most old-
growth trees in place, it removes or alters other essential ele-
16038 ECOLOGY CENTER v. AUSTIN
ments within old-growth habitat and disturbs bird species cur-
rently nesting or foraging within targeted stands.2 Although
treatment may be designed to restore old-growth to “historic
conditions,” Ecology Center points out this can be a mislead-
ing concept: for example, information regarding historic con-
ditions is incomplete; altering particular sections of forest in
order to achieve “historic” conditions may not make sense
when the forest as a whole has already been fundamentally
changed; many variables can affect treatment outcomes; and
the treatment process is qualitatively different from the “natu-
ral” or “historic” processes it is intended to mimic.
1. NFMA
While Ecology Center does not offer proof that the pro-
posed treatment causes the harms it fears, the Service does not
offer proof that the proposed treatment benefits — or at least
does not harm — old-growth dependent species. Ecology
Center argues that because the Forest Service has not assessed
the effects of old-growth treatment on dependent species, the
Service cannot be reasonably certain that treating old-growth
is consistent with NFMA’s substantive mandate to ensure spe-
cies diversity and viability. As a result, especially given the
scientific uncertainty surrounding the treatment of old-growth
stands, the Forest Service’s decision to treat additional old-
growth stands was arbitrary and capricious.
Although the Forest Service points to a report which notes
that two species of woodpecker were observed foraging in
treated old-growth forest, it does not otherwise dispute the
charge that it has not directly monitored the impact of treating
2
For example, the pileated woodpecker is dependent upon old-growth
snags (standing dead trees), but treatment involves timber harvesting,
which “creates the potential for snag loss.” Although the Service acknowl-
edges this danger, it believes that proposed “snag mitigation measures,”
combined with the potential for snag-creation during the prescribed burn-
ing process, will offset most of the snag loss caused by logging.
ECOLOGY CENTER v. AUSTIN 16039
old-growth on dependent species. Instead, the Service main-
tains that it need not do so because (1) it has observed the
short-term effects of thinning old-growth stands via commer-
cial logging and prescribed burning on forest composition, (2)
it has reason to believe that certain old-growth dependent spe-
cies would prefer the post-treatment composition of old-
growth forest stands, and (3) its assumption that treatment
does not harm old-growth dependent species is therefore rea-
sonable. The Service further argues that we must defer to its
methodological choices regarding what to monitor and how to
assess the impact of old-growth treatment.
[1] An agency’s choice of methodology is entitled to defer-
ence. See, e.g., Salmon River Concerned Citizens v. Robert-
son, 32 F.3d 1346, 1359 (9th Cir. 1994). However, there are
circumstances under which an agency’s choice of methodol-
ogy, and any decision predicated on that methodology, are
arbitrary and capricious. For example, we have held that in
order to comply with NFMA, the Forest Service must demon-
strate the reliability of its scientific methodology. Lands
Council, 379 F.3d at 752 (holding NFMA “require[s] that the
hypothesis and prediction of the model be verified with obser-
vation”). Here, as in Lands Council, the Forest Service’s con-
clusion that treating old-growth forest is beneficial to
dependent species is predicated on an unverified hypothesis.
While the Service’s predictions may be correct, the Service
has not yet taken the time to test its theory with any “on the
ground analysis,” id., despite the fact that it has already
treated old-growth forest elsewhere and therefore has had the
opportunity to do so. Just as it would be arbitrary and capri-
cious for a pharmaceutical company to market a drug to the
general population without first conducting a clinical trial to
verify that the drug is safe and effective, it is arbitrary and
capricious for the Forest Service to irreversibly “treat” more
and more old-growth forest without first determining that
such treatment is safe and effective for dependent species.
This is not a case in which the Forest Service is asking for the
opportunity to verify its theory of the benefits of old-growth
16040 ECOLOGY CENTER v. AUSTIN
treatment. Rather, the Service is asking us to grant it the
license to continue treating old-growth forests while excusing
it from ever having to verify that such treatment is not harm-
ful.
The Service argues that under Inland Empire Public Lands
Council v. U.S. Forest Serv., 88 F.3d 754 (9th Cir. 1996), we
must defer to its decision to monitor only the effect of treat-
ment on forest composition, instead of also monitoring the
effect of treatment on dependent species. However, Inland
Empire is inapposite here. In that case, we permitted the Ser-
vice to determine that it was complying with its duty to main-
tain species viability by using a “proxy-on-proxy” method for
monitoring species population. Id. at 761. But see Idaho
Sporting Congress, 305 F.3d at 972-73 (finding use of proxy-
on-proxy approach arbitrary and capricious where there is evi-
dence that the “methodology does not reasonably ensure via-
ble populations of the species at issue”). Under the proxy-on-
proxy approach, the Service does not ensure that it is main-
taining species viability by directly monitoring species popu-
lations. Instead, the Service designates certain “management
indicator species” as proxies for other species with compara-
ble habitat needs. It then designates certain kinds of habitat as
proxies for the management indicator species. Finally, it
assumes that a species’ viability is maintained so long as the
requisite amount of the species’ habitat is maintained. Here,
the Service is not simply maintaining the amount of old-
growth habitat necessary to support old-growth dependent
species — it is altering the composition of old-growth habitat
through an invasive process.
[2] Although the Service concedes that the opinions of
well-qualified experts vary with respect to the appropriateness
of management activities in old-growth areas, it also argues
that it must have the “discretion to rely on the reasonable
opinions of its own qualified experts even if, as an original
matter, a court might find contrary views more persuasive.”
Marsh v. Oregon Natural Resources Council, 490 U.S. 360,
ECOLOGY CENTER v. AUSTIN 16041
378 (1989). However, this is not a case in which different
experts have studied the effects of commercial thinning and
prescribed burning in old-growth forests and reached different
conclusions. Here, experts have differing hypotheses regard-
ing the effects that treating old-growth has on dependent spe-
cies, yet the Forest Service proposes to continue treating old-
growth without first taking the time to observe what those
effects actually are. In light of its responsibilities under
NFMA, this is arbitrary and capricious.
2. NEPA
[3] “The EIS did not address in any meaningful way the
various uncertainties surrounding the scientific evidence”
upon which the decision to treat the Lolo National Forest old-
growth rests. Seattle Audubon Soc’y v. Espy, 998 F.2d 699,
704 (9th Cir. 1993). Although the EIS identifies the public’s
concerns regarding the impact of treatment on dependent spe-
cies as “key” or “driving” issues,3 the EIS does not actually
explain in any detail the bases of those concerns, much less
address them. Cf. id. (finding NEPA violation where Forest
Service failed “to include a full discussion of the scientific
3
In a section of the EIS entitled “Issue Development,” the Service
describes “[p]otential issues,” which are “defined as actual and perceived
effects, risks, and hazards of the Proposed Action.” Issues were identified
both internally and through “public involvement.” In relevant part, the EIS
states:
There is a concern that the reduction of canopy closure and stand
density caused by fuels reduction activities and commercial thin-
ning may adversely affect habitat for several bird species that are
dependent upon late-successional dry forests with closed cano-
pies and high basal area.
...
There is a concern that the . . . commercial thinning of unburned
trees, and the removal of understory fuels, could adversely affect
Proposed, Threatened, or Endangered Species and listed Sensi-
tive, Threatened, or Endangered Species . . . known to be present
within or to be potential to the analysis area.
16042 ECOLOGY CENTER v. AUSTIN
uncertainty” surrounding its strategy for maintaining spotted
owl viability). The EIS discusses in detail only the Service’s
own reasons for proposing treatment, and it treats the predic-
tion that treatment will benefit old-growth dependent species
as a fact instead of an untested and debated hypothesis. Even
if the Service considered these issues but concluded that it
need not or could not “undertake further scientific study”
regarding the impact of treatment on dependent species, it
should have “explain[ed] in the EIS why such an undertaking
[wa]s not necessary or feasible.” Id. For these reasons, we
also find that the Service’s analysis of the impact of treating
old-growth to be inadequate under NEPA.4
B. Salvage Harvesting of Black-Backed Woodpecker
Habitat
The Forest Service has designated the black-backed wood-
pecker as a “sensitive species,” a species for which viability
is a concern. The black-backed woodpecker’s viability is sen-
sitive because it is particularly dependent upon post-fire land-
scapes, and the combination of fire-suppression efforts and
past salvage logging has drastically reduced the amount of
post-fire habitat. According to the EIS, “[d]ue to this reduc-
tion in habitat, black-backed woodpeckers went from being
relatively abundant to relatively rare.” To make matters
worse, not all fires create black-backed woodpecker habitat,
and the burned stands that do qualify only serve as black-
backed woodpecker habitat for a limited number of years. As
a result, new post-fire habitat must be continuously generated.
Ecology Center and the Forest Service agree that prior to
the 2000 fires, there was a critical shortage of black-backed
4
Ecology Center also raises claims challenging the Service’s analysis of
the Project’s impact on two specific old-growth dependent species, the
pileated woodpecker and the northern goshawk. Because we find that the
Service’s decision to treat old-growth violates both NFMA and NEPA, we
do not reach these more narrow claims.
ECOLOGY CENTER v. AUSTIN 16043
woodpecker habitat. Specifically, the Service estimated that
from 1993 to 1998, fires created a total of 11,045 acres of
post-fire habitat (9160 of those acres were burned in 1998
alone). This total was estimated to be 6% of what would have
been created historically over a comparable six-year period.
At that level, the Service considered the black-backed wood-
pecker and other post-fire habitat dependent species to be “at
extreme risk.”
Writing in November 2000 about the “effect [on] black-
backed woodpeckers,” Forest Service scientists explained:
“We concluded in 1998 that salvage of any post-fire habitat
on the Lolo [National Forest] . . . ‘will impact individuals or
habitat and with a consequence that the action may contribute
to a trend towards federal listing or cause a loss of viability
to the population or species.’ ” They explained that a small
amount (less than 400 acres) of salvage could be permitted,
but only if there was “a commitment to creation of post-fire
habitat of an equal amount” through prescribed fires. Under
those two conditions, they believed that salvaging “would not
likely . . . cause a loss of viability to the population,” even
though it “may impact individuals or habitat.”
From the EIS, it appears that at the time the Project was
developed, there was a total of 19,219 acres of black-backed
woodpecker habitat in the Lolo National Forest: there were
9349 acres that were created by pre-2000 fires and were less
than five years old, plus 9870 acres of habitat that were cre-
ated by the 2000 fires. Of those acres, 9100 are located in the
Project area. The selected alternative originally proposed sal-
vage harvesting 1020 of those acres. However, the Record of
Decision reduced that amount to 815 acres.5
The Forest Service maintains that the Project complies with
NFMA because the 2000 fires created a large amount of
potential habitat, and only a small portion will be salvaged.
5
The Sierra Club settlement further reduced that amount to 155 acres.
16044 ECOLOGY CENTER v. AUSTIN
The Service further points out that the Project includes miti-
gation measures designed to minimize the adverse impact of
salvaging.6 However, “[r]esearch has also indicated that sal-
vage logging, even when large numbers of snags are left, is
detrimental to this species.” For example, one study “found
that black-backed woodpeckers did not nest in areas that were
considered lightly salvaged.” In addition, the Project’s mitiga-
tion measures do not provide for the creation of an equal
amount of post-fire habitat elsewhere.7
1. NEPA
[4] The EIS fails to adequately explain the basis for the
Forest Service’s conclusion that eliminating a portion of the
newly-created habitat will not adversely affect the black-
backed woodpeckers’ viability. Prior to the 2000 fires, the
amount of post-fire habitat was so low that the black-backed
woodpecker and other post-fire habitat dependent species
were deemed at “extreme risk,” and Forest Service experts
concluded that “salvage of any post-fire habitat . . . would . . .
impact individuals or habitat . . . with a consequence that the
action may contribute to a trend towards federal listing or
cause a loss of viability to the population or species.”8 Yet,
6
For example, the Project provides that no harvesting will occur wher-
ever black-backed woodpecker nests are located; each spring, before har-
vesting actually begins, targeted stands will be re-surveyed for black-
backed woodpecker nests, and loggers will be instructed to preserve any
identified nest trees as well as reserve patches surrounding the nesting
trees.
7
Although other aspects of the Project involve prescribed burning, “the
Lolo prescribed fire program has not demonstrated an ability to recruit
large amounts of post-fire habitat, although there is an improving trend in
1997-98.” Most kinds of prescribed burning do not create the necessary
post-fire habitat; only higher-intensity fires create post-fire habitat. There
is no indication that the kind of prescribed burning that the Project entails
is designed to create black-backed woodpecker habitat.
8
As noted above, these experts concluded that some salvaging could be
tolerated, but only if limited to a small amount and mitigated by the cre-
ation of an equal amount of post-fire habitat elsewhere.
ECOLOGY CENTER v. AUSTIN 16045
the Project EIS states — without meaningful explanation —
that even though salvaging post-fire habitat may negatively
impact individual black-backed woodpeckers, it will “not
likely result in a trend towards federal listing.” Without more,
this general statement regarding the possible impact and risk
involved “do[es] not constitute a ‘hard look’ absent a justifi-
cation regarding why more definitive information could not
be provided.” Neighbors of Cuddy Mountain, 137 F.3d at
1380; 40 C.F.R. § 1502.1 (providing that the EIS “shall be
supported by evidence that the agency has made the necessary
environmental analyses”).
We can discern from the EIS only one possible basis for the
Service’s changed assessment of the impact of salvaging: the
fact that the 2000 fires created additional post-fire habitat,
which raised the total amount of post-fire habitat. However,
we cannot conclude from this fact alone that the habitat level
is no longer critically low and that the previously prescribed
restrictions on salvaging should no longer be enforced.9 To be
reasonably certain that the post-Project habitat levels would
be sufficient to ensure species viability, one must know where
the threshold between “critical” and “sufficient” levels of
burned habitat lies. Because the EIS does not disclose what
this threshold is, much less explain how the threshold was
determined, we cannot evaluate the Service’s decision.
Indeed, we cannot even be certain that the Service determined
and considered this factor when making its assessment.
[5] Similarly, the EIS does not indicate how much further
habitat levels have to drop before the species is downgraded
from sensitive to threatened or how the Service plans to gen-
erate additional black-backed woodpecker habitat to counter-
balance losses from salvaging, the passage of time, and ongo-
9
We note that the Service considered the creation of 11,045 acres of
post-fire habitat over a six-year period to be critically low (6% of the his-
toric rate); thus, the creation of 19,219 acres over a comparable period
does not necessarily represent a sufficient increase.
16046 ECOLOGY CENTER v. AUSTIN
ing fire-suppression efforts. Where, as here, “the information
in the . . . EIS was so incomplete or misleading that the deci-
sionmaker and the public could not make an informed com-
parison of the alternatives, revision of an EIS may be
necessary to provide ‘a reasonable, good faith, and objective
presentation of the subjects required by NEPA.’ ” Animal Def.
Council v. Hodel, 840 F.2d 1432, 1439 (9th Cir. 1988) (quot-
ing Johnston v. Davis, 698 F.2d 1088, 1095 (10th Cir. 1983)),
amended by 867 F.2d 1244 (9th Cir. 1989); see also Idaho
Sporting Cong. v. Thomas, 137 F.3d 1146, 1150 (9th Cir.
1998) (“[A]llowing the Forest Service to rely on expert opin-
ion without hard data either vitiates a plaintiff’s ability to
challenge an agency action or results in the courts second
guessing an agency’s scientific conclusions. As both of these
results are unacceptable, we conclude that NEPA requires that
the public receive the underlying environmental data from
which a Forest Service expert derived her opinion.”).
[6] Thus, we hold that the Forest Service failed to either
adequately explain its impact assessment or provide the infor-
mation that is necessary to understand and evaluate the Forest
Service’s decision to permit salvaging of the black-backed
woodpeckers’ rare habitat, in violation of NEPA.
2. NFMA
[7] The designation of a species as sensitive arises from the
Forest Service’s obligations under NFMA. Pursuant to 16
U.S.C. § 1604(g)(3)(B), the Forest Service is required to “pro-
vide for diversity of plant and animal communities.” The For-
est Service’s duty to maintain viable populations “applies
with special force to ‘sensitive’ species.” Friends of the
Clearwater v. Dombeck, 222 F.3d 552, 556 n.2 (9th Cir.
2000) (citation and quotation omitted). The Lolo National
Forest Plan requires, more specifically, that the Forest Service
“manage to maintain population viability” for plant and ani-
mal species “that are not threatened or endangered, but where
viability is a concern (i.e., sensitive species).”
ECOLOGY CENTER v. AUSTIN 16047
[8] Because the Forest Service failed to provide the factual
basis for its analysis and failed to adequately explain its deci-
sion, we cannot be reasonably certain that the salvaging —
which the Service concedes may harm individual black-
backed woodpeckers — will not jeopardize the black-backed
woodpeckers’ viability. Cf. Native Ecosystems Council v. U.S.
Forest Serv., No. 04-35375, 2005 WL 1906996, at *6 (9th
Cir. Aug. 11, 2005) (finding NFMA violation where “we can-
not reasonably determine that the Forest Service has complied
with the [Forest] Plan”). We therefore hold that the decision
to salvage harvest black-backed woodpecker habitat is also
arbitrary and capricious under NFMA.
C. Soil-Quality Analysis
[9] Ecology Center next challenges the Forest Service’s soil
quality analysis. Under the applicable Regional Soil Quality
Standard, the Forest Service cannot allow an activity that
would create detrimental soil conditions in fifteen percent of
the activity area.10 Ecology Center claims that the methodol-
ogy the Forest Service used to determine what percentage of
soil is in a detrimental state in each activity area was insuffi-
ciently reliable because the Service estimated soil conditions
on the basis of maps, samples from throughout the Forest,
aerial reconnaissance, and computer modeling, but did not
verify those estimates by directly observing soil conditions in
the activity areas.11
10
An “activity area” is a unit of forest in which an activity, in this case
salvage harvesting, is to take place; this Project proposes harvesting in
approximately 128 different activity areas or “cutting units.” The Regional
Soil Quality Standard further provides that if fifteen percent or more of the
activity area already has detrimental soil conditions, then the Project will
not be permitted to make it worse. Rather, the Project should then aim to
improve the soil conditions.
11
Specifically, the Service looked at maps to determine where damaging
past activities (e.g., timber harvesting), jammer roads, firelines, and skid
trails, as well as areas of high soil burn severity overlap with proposed
activity areas. In addition, the Service input data from the Land Systems
16048 ECOLOGY CENTER v. AUSTIN
[10] We addressed a nearly identical claim, involving the
same Regional Soil Quality Standard, in Lands Council.12 See
Lands Council, 379 F.3d at 752-53. In that case, “based on
samples from throughout the Forest, and aerial photographs,
the Forest Service estimated the quality of the soils in the
Project area using a spreadsheet model.” Id. at 752. We held
that the “Forest Service’s reliance on the spreadsheet models,
unaccompanied by on-site spot verification of the model’s
predictions, violated NFMA.” Id. at 752-53. In so holding, we
relied on the analysis of an earlier district court case, Kettle
Range Conservation Group v. U.S. Forest Serv., 148 F. Supp.
2d 1107 (E.D. Wash. 2001), where the Forest Service, “based
on assumptions [from general data from the Forest soils], geo-
logical maps, and aerial photographs, . . . estimated the condi-
tion of each unit, tried to determine which units might exceed
established standards, and projected potassium levels.” Id. at
1127 (emphasis in original).
Although the methodologies in this case, Lands Council,
and Kettle Range all appear to suffer from the same critical
defect, the Forest Service argues that Lands Council is not
controlling here. Specifically, the Service argues that the
Regional Soil Quality Standard is not binding with respect to
projects in the Lolo National Forest. In addition, the Forest
Service argues that this case is factually distinguishable from
Lands Council because, notwithstanding the non-binding
Inventory (“LSI”) and the Burned Area Emergency Rehabilitation Report
(“BAER Report”) into models to generate estimates of the project’s possi-
ble effects. The BAER Report was based on field reviews and helicopter
flyovers, as well as “transect surveys” (an approved method of field obser-
vations). The LSI and BAER Report data was not collected for the pur-
pose of determining compliance with the Regional Soil Quality Standards.
Because the Project was developed after the BAER transect surveys were
conducted, the transects did not cover the vast majority of the activity
areas; only a few were crossed by coincidence.
12
Because Lands Council was decided after briefing, we ordered supple-
mental briefing.
ECOLOGY CENTER v. AUSTIN 16049
nature of the Standard, the Forest Service completed suffi-
cient on-the-ground verification of its soil quality estimates.
We address each argument in turn.13
1. The Regional Soil Quality Standard
The Forest Service argues that the Standard is not binding
because it was never incorporated into the Lolo National For-
est Plan, and although it was incorporated into the Forest Ser-
vice Manual, this court has held that the Forest Service
Manual does not have the independent force and effect of law.
Western Radio Servs. Co. v. Espy, 79 F.3d 896, 901 (9th Cir.
1996).
Even assuming, arguendo, that the Standard does not have
the independent force and effect of law with respect to the
Lolo National Forest, it would nonetheless be arbitrary and
capricious for the Forest Service to ignore it because both the
draft EIS and final EIS discuss the Standard as if it is binding
and claim that the Service developed the Project in compli-
ance with its provisions. Thus, even if we were to agree that
the Standard is merely advisory, we would then be compelled
to find that the draft EIS and final EIS are misleading in viola-
tion of NEPA. See Resources Ltd. v. Robertson, 35 F.3d 1300,
1304 n.3 (9th Cir. 1994) (rejecting the argument that the For-
est Service could treat the Interagency Grizzly Bear Guide-
13
The Forest Service also reiterates its argument that this court must
defer to its choice of methodology and reliance on its own experts. This
argument fails here for the same reason it did in Lands Council: although
this court is required to defer to agency expertise, the agency is not permit-
ted to adopt and rely upon a methodology without reasonably verifying its
reliability. Lands Council, 379 F.3d at 752 (“The Forest Service, granted
appropriate deference, still does not demonstrate the required reliability of
the spread sheet model.”) Moreover, in this case, one of the Forest Ser-
vice’s own experts, Region Scientist John Nesser, called the Forest Ser-
vice’s soil analysis into question because it failed to assess soil conditions
by field testing the actual activity areas — the exact reason that Lands
Council deemed the Forest Service’s methodology inadequate.
16050 ECOLOGY CENTER v. AUSTIN
lines as optional where the Fish and Wildlife Service made its
“no jeopardy” conclusion contingent on adherence to the
Guidelines); Animal Def. Council, 840 F.2d at 1439 (noting
that a misleading EIS violates NEPA). Moreover, the purpose
of the Standard is to ensure compliance with the substantive
mandates of NFMA. The Forest Service does not explain how
it can be certain that the Project complies with NFMA if the
Project was not developed in accordance with the Standard.
2. On-Site Verification of Soil Conditions
First, the Service points to the transects it conducted during
the BAER assessment of the 2000 fires’ aftermath. However,
during the Project development, one of the Service’s own
experts pointed out that the transects targeted burned areas,
not proposed harvest units; as a result, most of the activity
areas were not transected.14 Thus, the BAER transects do not
distinguish this case from Lands Council, where the Service
similarly “did not test much of the activity area.” Lands
Council, 379 F.3d at 752, as amended by 395 F.3d at 1034.
Second, the Forest Service argues that its soil analysis is
nonetheless sufficiently reliable because it utilized data from
areas with ecological characteristics similar to the proposed
harvest units. Lands Council expressly rejected this same
argument. Id. (“The Forest Service concedes that it did not
test much of the activity area, but argues that because it tested
similar soils within the Forest, and similar soils act the same
way, then the methodology is sound.”). We note that com-
ments on the Project by the Service’s own expert support this
holding: Region Scientist Nesser argued that the Service’s
conclusions regarding soil conditions in the activity areas
14
This point is corroborated by the draft EIS, which states that “not all
proposed harvest units were visited with line transects” and that “much of
the [soil quality] determination was based on information from the trans-
portation and timber shop with respect to past activities and regeneration
level of the jamer [sic] roads.”
ECOLOGY CENTER v. AUSTIN 16051
were “not credible” because the Service did not test the activ-
ity areas themselves, as the Standard requires.15
Third, the Service points to some informal field reports
contained in the record, some of which indicate that a small
percentage of the activity areas were observed directly.16 The
record provides little information that enables us to assess the
reliability or significance of these reports; for example, we do
not know the qualifications of the persons conducting the
field review, the methodology utilized, or whether the field
observations confirmed or contradicted the Service’s esti-
mates. Moreover, the final EIS repeatedly explains that the
Service will conduct field testing to verify its estimates prior
to any ground-breaking activities,17 and there is no indication
in the draft EIS or final EIS that the Forest Service actually
15
Specifically, Region Scientist John Nesser commented:
The statement on page 3-35 [of the draft EIS] that ‘All proposed
activity units are at 3 percent (detrimental disturbance) or less
prior to any new activities’ is not credible, given fire effects and
the effects of past timber harvest and road building activities.
This is because transects were done in the various burned areas,
rather than in proposed harvest or other activity units. The burned
units are not the correct activity unit to use for post fire activities.
. . . statements about meeting the 15 percent limits for detrimental
disturbance are both irrelevant and wrong. Without using the cor-
rect activity areas for analysis, the cumulative effects assessment
is not correct.
16
Some of the field reports cited by the Service actually say very little
about soil conditions.
17
The final EIS explains that a “[w]alk through of the proposed harvest
and burn units will be completed prior to the implementation of any
ground disturbing activities in the activity units. The walk through will
identify the occurrence and percent of the activity area with displacement,
rutting, compaction, high burn severity, depth of organic matter, pedestal-
ing, and slumping.” (emphasis added). Similarly, throughout its discussion
of the Project’s impacts, the EIS repeatedly notes that it “will verify” the
conclusions it reached regarding soil conditions by having “qualified soil
specialists” conduct “field reviews” of the proposed activity units “to eval-
uate the percent detrimental condition prior to any ground disturbing activ-
ities.”
16052 ECOLOGY CENTER v. AUSTIN
consulted and relied upon any of the field reports it now cites
when it was making its selection amongst the Project alterna-
tives. Thus, we cannot conclude that these reports distinguish
this case from Lands Council.
[11] Finally, we consider whether the Service’s plan to ver-
ify soil conditions in the activity areas after authorizing the
Project, but before actually commencing harvesting activities,
satisfies NFMA and NEPA. The fact that the Service plans to
conduct on-site verification prior to any harvesting implies
that even the Service recognizes that its soil-quality estimates
need to be verified. Thus, we conclude that the Forest Ser-
vice’s decision to authorize the Project first and verify later
does not distinguish this case from Lands Council, and that
the Project violates NFMA. We also conclude that the Ser-
vice’s plan to verify its estimates post-decision fails to satisfy
NEPA, because “NEPA requires consideration of the potential
impact of an action before the action takes place.” Neighbors
of Cuddy Mountain, 137 F.3d at 1380 (quoting City of
Tenakee Springs v. Clough, 915 F.2d 1308, 1313 (9th Cir.
1990) (emphasis in original)); Kettle Range, 148 F. Supp. 2d
at 1127 (“They [the Forest Service] apparently planned to
[field test] sometime in the future. They should have done it
before they completed the FEIS. The failure to do so is a vio-
lation of NEPA because it shows that USFS did not give a
‘hard look’ at the effects of the Project on the soils in the
analysis area.”).
III
The Forest Service’s selection of Modified Alternative
Number 5 for the Lolo National Forest Post-Burn Project vio-
lates both NEPA and NFMA. We reverse the district court’s
summary judgment in favor of the Forest Service, and direct
the district court to enter summary judgment on behalf of
ECOLOGY CENTER v. AUSTIN 16053
Ecology Center. We remand this case to the Forest Service for
further proceedings consistent with this opinion.18
REVERSED and REMANDED.
McKEOWN, Circuit Judge, dissenting:
The Ninth Circuit, like the other circuits, repeats frequently
the legal mantras of administrative review in the context of
environmental decisions: “arbitrary and capricious,” “hard
look,” and “no second guessing.”1 These standards are easy to
articulate, but it is more difficult to know when we have
crossed the line from reviewer to decisionmaker. In this case,
we have gone too far.
As summarized in Neighbors of Cuddy Mountain v. United
States Forest Service, 137 F.3d 1372, 1376 (9th Cir. 1998),
we review agency actions “to determine if they were arbitrary
or capricious, an abuse of discretion, or not in accordance
with law.” The limited nature of this inquiry underscores the
latitude in implementation and interpretation that Congress
intended for its agents. This latitude does not relieve us of our
duty to insure that the agency took a “hard look,” as required
18
We do not reach the evidentiary issue: whether the district court
abused its discretion in excluding the extra-record expert declarations. We
note that, in the course of the preparation of a new EIS on remand, the par-
ties may submit the declarations and any other relevant information to be
included as part of the administrative record.
1
See, e.g., Neighbors of Cuddy Mountain v. U.S. Forest Serv., 137 F.3d
1372, 1376 (9th Cir. 1998) (noting that we review Forest Service actions
under the APA and the National Forest Management Act to determine if
they were “arbitrary and capricious”); Idaho Sporting Cong., Inc. v. Rit-
tenhouse, 305 F.3d 957, 963 (9th Cir. 2002) (requiring a “hard look” by
agencies at environmental effects of projects); Idaho Sporting Cong., Inc.
v. Thomas, 137 F.3d 1146, 1150 (9th Cir. 1998) (noting that courts should
not “second guess[ ] an agency’s scientific conclusions” and, for that rea-
son, requiring hard data to discourage second guessing).
16054 ECOLOGY CENTER v. AUSTIN
under the National Environmental Policy Act (“NEPA”), or to
insure compliance with forest plans under the National Forest
Management Act (“NFMA”). Nor does record review mean
that we abdicate our obligation to scrutinize the agency’s
underlying data, assumptions, and conclusions. As a review-
ing court, we play a critical role in the review process and we
have not been shy in letting the U.S. Forest Service know
where it has fallen short as a matter of law and process.2 How-
ever, the majority’s extension of Lands Council v. Powell,
379 F.3d 738 (9th Cir. 2004), amended by 395 F.3d 1019,
1024 (9th Cir. 2005), represents an unprecedented incursion
into the administrative process and ratchets up the scrutiny we
apply to the scientific and administrative judgments of the
Forest Service. Because the majority has, in effect, displaced
“arbitrary and capricious” review for a more demanding stan-
dard, I respectfully dissent.
The project under review is designed to address the after-
math of the Lolo National Forest wildfires in Montana in
2000 through selected thinning of smaller timber, salvage of
timber, prescribed burning and regeneration of thousands of
acres. The administrative record in this case is huge—a 1900+
page Final Environmental Impact Statement (FEIS), 150
detailed maps and 20,000 pages of background information.
The majority highlights only parts of this record, criticizes the
qualifications of the Forest Service’s personnel, and questions
various scientific judgments. In faulting the Forest Service’s
soil quality analysis and concluding that old-growth forest
will not be impaired, the majority changes our posture of
review to one where we sit at the table with Forest Service
2
See e.g. Natural Res. Def. Council v. U.S. Forest Serv., 421 F.3d 797,
816 (2005) (holding that the Forest Service’s timber sale plan was arbi-
trary and capricious, and violated NEPA, because it was based on a clear
misinterpretation of data on market demand for the timber); Idaho Sport-
ing Cong. v. Thomas, 137 F.3d at 1154 (holding that a Forest Service plan
violated NEPA because it had failed to take a “hard look” at the environ-
mental effects of the proposed timber sale).
ECOLOGY CENTER v. AUSTIN 16055
scientists and second-guess the minutiae of the decisionmak-
ing process.
My concern is perhaps best illustrated in the majority’s
application of Lands Council to the soil quality analysis issue.
In Lands Council, we addressed the Forest Service’s analysis
of disturbed soil conditions in a watershed restoration project
in the Idaho Panhandle National Forest. The Forest Service
had taken soil samples from throughout the forest but not in
the activity area. Id. at 1035. Based on these samples and
aerial photographs, it calculated the impact of the watershed
restoration project on soil quality in the activity area. Id. at
1034.
In Lands Council, we rejected the Forest Service’s choice
of scientific methodology because it did not walk or test soil
in the activity area. Id. We concluded that the “Forest Ser-
vice’s reliance on the spreadsheet models, unaccompanied by
on-site spot verification of the model’s predictions, violated
NFMA.” Id. at 1035. Given the specific record and circum-
stances in Lands Council, this resolution may have made per-
fect sense. But in the abstract, without the ability to compare
the record in Lands Council with the record here, there is no
legal basis to conclude that the NFMA requires an on-site
analysis where there is a reasonable scientific basis to uphold
the legitimacy of modeling. NFMA does not impose this sub-
stantive requirement, and it cannot be derived from the proce-
dural parameters of NEPA.
Application of Lands Council makes even less sense in this
specific case. Here, the Forest Service conducted exactly the
kind of on-site analyses found lacking in Lands Council—soil
samples in activity areas. The record lists five activity areas
surveyed in soil sample transects and details the walkthrough
of a number of activity areas. The majority concedes that
these on-site analyses exist but criticizes them for being too
few and of poor quality.3 Resting on this characterization, the
3
See Op. at 16048 n.11 (finding the Forest Service’s soil transects “did
not cover the vast majority of the activity areas; only a few were crossed
16056 ECOLOGY CENTER v. AUSTIN
majority rejects the Forest Service’s soil analysis on the
ground that it was “nearly identical” to the one we found
lacking in Lands Council.
Lands Council makes compliance with NFMA and NEPA
a moving target. If one reads Lands Council to require on-site
analysis in every case, which I respectfully suggest is an over-
reading of the case, the circumstances here comply with that
dictate. Yet the majority now says, without support in the
record and absent a challenge by Ecology Center at the
administrative level, that the on-site analysis was insufficient
and that we do not possess enough information to know if the
inspectors were qualified. From this judgment, we are left to
conclude that not only does the court of appeals set bright-line
rules, such as requiring an on-site, walk the territory inspec-
tion, but it also assesses the detail and quality of that analysis
—even in the absence of contrary scientific evidence in the
record. Page after page of the record contains details on field
analysis, such as the following representative sample of an
evaluation form:
by coincidence”); see, e.g., id. at 16051 (“The record provides little infor-
mation that enables us to assess the reliability or significance of these
reports; . . . we do not know the qualifications of the persons conducting
the field review, the methodology utilized . . . .” ). To paraphrase an old
joke, the majority found the food terrible and the portions small.
ECOLOGY CENTER v. AUSTIN 16057
In my view, it is not our role to criticize this evaluation
because, for example, the form does not list the evaluator’s
qualifications. Nor are we in a position to unscramble and
pass judgment on the investigator’s field notes, as in the fol-
lowing excerpt, which are summarized and integrated into the
FEIS:
16058 ECOLOGY CENTER v. AUSTIN
Had Ecology Center offered a credible scientific critique of
this methodology or provided some other basis for us to deter-
mine whether the Forest Service’s judgment was arbitrary and
capricious, we would be in a different position. But just as we
are not supposed to substitute our judgment for that of the
Forest Service, neither can we rely on—in the absence of
evidence—Ecology Center’s arguments in its briefs or in its
expert declaration filed after the administrative review pro-
cess.
ECOLOGY CENTER v. AUSTIN 16059
Lands Council does not direct us to assess the sufficiency
of the Forest Service’s on-site soil quality analyses beyond
the traditional arbitrary and capricious standard; it only asks
us to verify that there is such an on-site sampling. That opin-
ion stated four times that the Forest Service had failed to con-
duct any on-site analysis—it did not say that the Service
failed to conduct enough of them.4 The majority now extends
the “on-site soil analysis” doctrine, initiated by Lands Coun-
cil, while abandoning its reasoning. Although on-site analysis
could plausibly be conducted in such a manner as to be arbi-
trary and capricious, the majority’s justification for finding
this project to be arbitrary and capricious imports a bright-line
rule where individualized analysis is appropriate.
More importantly, in Lands Council we did not purport to
create a general rule requiring on-site verification for all sci-
entific hypotheses adopted by the Forest Service regardless of
context. We did not even adopt a rule requiring all soil analy-
ses to be verified by on-site sampling. Our decision was spe-
cific to the facts of that case. See Lands Council, 395 F.3d at
1035 (“Under the circumstances of this case, the Forest Ser-
vice’s . . . methodology, to be reliable, required that the
hypothesis and prediction of the model be verified with obser-
vation.”) (emphasis added).
Nevertheless, the majority generalizes the “unverified
hypothesis” principle articulated in Lands Council beyond the
soil analysis context to other scientific findings made by the
Forest Service. In so doing, the majority demonstrates the
dangers of extending a reference—abstracted from a single,
technically detailed, fact-specific decision—to unrelated fac-
4
See id. at 1034 (“The Forest Service did not walk, much less test, the
land in the activity area.”); id. at 1035 (“The predictions of the model . . .
were not verified with on the ground analysis.”); id. (“[T]he spreadsheet
models, unaccompanied by on-site spot verification . . . , violated
NFMA.”); id. (The “soils analysis was based entirely on the model with
no on-site inspection . . . .”) (emphasis added).
16060 ECOLOGY CENTER v. AUSTIN
tual contexts. For example, regarding the old-growth claim,
the majority reverses the Forest Service’s judgment because
“[h]ere, as in Lands Council, the Forest Service’s conclusion
that treating old-growth forest is beneficial to dependent spe-
cies is predicated on an unverified hypothesis.” Op. at 16039.
The Forest Service hypothesis is, however, supported by
observational data. There is record evidence, based on direct
observation, that treatment produces and preserves habitat for
old-growth dependent species.5
The majority applies Lands Council to prohibit the Forest
Service from inferring simply that because treating old-
growth preserves or creates habitat for dependent species does
not mean that such treatment will not harm the species. Op.
at 16040 (claiming that the Forest Service just “assumes that
a species’ viability is maintained so long as the requisite
amount of the species habitat is maintained”). In reaching this
conclusion, the majority acknowledges and then dismisses
without explanation the very record evidence it says the For-
est Service failed to provide—here, direct observation of cer-
tain species seen foraging in old-growth areas after treatment.6
Applying Lands Council to the old-growth issue is inappro-
priate for yet another reason. The panel assumes that Alterna-
5
For example, a report by Forest Service scientists utilized historic and
stand structure research to plan for treatment that would enhance habitat
for plants and species. The scientists found that treated areas have nesting
and foraging opportunities for cavity nesting species such as pileated
woodpeckers.
6
The majority fails to acknowledge that we “have, in appropriate cases,
allowed the Forest Service to avoid studying the population trends of [spe-
cies] by using . . . habitat as a proxy for . . . population trends.” Lands
Council, 395 F.3d at 1036. In Lands Council, we rejected a proxy-on-
proxy approach for measuring population trends in old-growth forest
because the habitat data the Forest Service offered was “fifteen years old,
with inaccurate canopy closure estimates, and insufficient data on snags.”
Id. The majority does not identify any such problems in using the proxy-
on-proxy approach in this case. See Op. at 16040.
ECOLOGY CENTER v. AUSTIN 16061
tive 5, which involves thinning, salvage, and regeneration,
disrupts a stable status quo “through an invasive process.” Op.
at 16040. Yet the status quo is anything but stable. The Forest
Service presents uncontested evidence that the failure to treat
old-growth areas risks the very harms feared by Ecology Cen-
ter, even though it has provided no evidence to support such
a claim. Op. at 16038 (“Ecology Center does not offer proof
that the proposed treatment causes the harms it fears . . . .”).
In fact, the record reveals that the failure to treat old-growth
areas could result in “considerable loss of old growth trees
from bark beetle predation,” which will put “at risk . . . spe-
cific habitat niches for many wildlife species that are adapted
to the more open growth old-forest character.” Old-growth
areas “are now at risk for major disturbances such as disease
and insect epidemics and high-severity stand replacing fires.”
Inaction or delay threatens the species Ecology Center seeks
to protect.
No one contests the Forest Service’s conclusion that treat-
ment will “provide direct reduction of bark beetle infestation
or risk of future infestation.” Faced with uncontroverted evi-
dence that inaction may harm old-growth areas, the majority
still requires the Forest Service to produce more evidence that
treating old-growth will not harm dependent species before it
will allow the Service to save old-growth areas from other
dangers. Under these circumstances, I cannot agree that the
Forest Service’s decision to treat old-growth areas was arbi-
trary and capricious. Indeed, had the Forest Service taken the
majority’s approach, its decision may well have been arbitrary
and capricious for “fail[ing] to consider an important aspect
of [the] problem.” See Lands Council, 395 F.3d at 1026.
Nor do I think it appropriate to analogize the Forest Service
process to the separate and wholly inapposite regime of the
Food and Drug Administration (“FDA”). The majority writes
that “it would be arbitrary and capricious for a pharmaceutical
company to market a drug to the general population without
first conducting a clinical trial” just as the Forest Service can-
16062 ECOLOGY CENTER v. AUSTIN
not be permitted to “treat more and more old-growth forest
without first determining that such treatment is safe and effec-
tive for dependent species.” See Op. at 16039. Without com-
menting on the obvious differences between humans and trees
(and in fact acknowledging the importance of both to our
environment), this analogy underscores the degree to which
the majority inserts itself into the internal judgments of the
Forest Service. The FDA process dictates a substantive and
specific administrative course of action in terms of clinical tri-
als and other requirements as a prelude to the approval of
drugs and medical devices. Neither NEPA nor NFMA serve
that function in the environmental context. To import the
notion of clinical trials from the FDA context to soil sampling
in federal forests is a leap too far.
Apparently we no longer simply determine whether the
Forest Service’s methodology involves a “hard look” through
the use of “hard data,” but now are called upon to make fine-
grained judgments of its worth. In reaching this conclusion,
the majority takes aim at two firmly established lines of pre-
cedent in administrative law. First, this view is contrary to the
basic principle that we reverse agency decisions only if they
are arbitrary and capricious. This standard of review does not
direct us to literally dig in the dirt (or soil, as it were), get our
fingernails dirty and flyspeck the agency’s analysis. Yet the
majority does exactly that by rejecting the Forest Service’s
soil analysis field checks and its observations and historical
data in treated old-growth forests. The majority’s rationale
cannot be reconciled with our case law requiring “[d]eference
to an agency’s technical expertise and experience,” particu-
larly “with respect to questions involving engineering and sci-
entific matters.” United States v. Alpine Land & Reservoir
Co., 887 F.2d 207, 213 (9th Cir. 1989). See also Westlands
Water Dist. v. U.S. Dept. of Interior, 376 F.3d 853, 871 (9th
Cir. 2004) (reversing a district court decision that certain
alternatives violated NEPA because of inadequate deference
to the agency’s expertise).
ECOLOGY CENTER v. AUSTIN 16063
Because the majority’s analysis cannot be squared with the
deferential review required of us, I respectfully dissent.