FOR PUBLICATION
  UNITED STATES COURT OF APPEALS
       FOR THE NINTH CIRCUIT

WILDWEST INSTITUTE;                    
FRIENDS OF THE BITTERROOT, INC.,
              Plaintiffs-Appellants,
                v.
DAVE BULL; ABIGAIL KIMBELL;
UNITED STATES FOREST SERVICE,                No. 07-35044
             Defendants-Appellees,
BITTER ROOT RESOURCE                          D.C. No.
                                           CV-06-00066-DWM
CONSERVATION AND DEVELOPMENT                   OPINION
AREA INC.; RAVALLI COUNTY; SULA
VOLUNTEER FIRE DEPARTMENT;
ROBERT WETZTSEON; BECKI
LINDERMAN; ROCKY MOUNTAIN LOG
HOMES,
  Defendant-intervenors-Appellees.
                                       
      Appeal from the United States District Court
              for the District of Montana
    Donald W. Molloy, Chief District Judge, Presiding

               Argued December 12, 2007
             Submission Vacated July 7, 2008
              Resubmitted October 29, 2008
                  Pasadena, California

                  Filed November 6, 2008

 Before: Alfred T. Goodwin, Diarmuid F. O’Scannlain, and
            Raymond C. Fisher, Circuit Judges.

               Opinion by Judge O’Scannlain

                            15201
                  WILDWEST INSTITUTE v. BULL             15205


                         COUNSEL

Thomas J. Woodbury, Forest Defense, P.C., Missoula, Mon-
tana, argued the cause for the plaintiffs-appellants and filed
briefs.

Stacey Person, Environmental & Natural Resources Division,
U.S. Dept. of Justice, Washington, DC, argued the cause for
the federal defendant-appellees and filed briefs.

Julie Weis, Hagland Kelley Horngren Jones & Wilder LLP,
Portland, Oregon, argued the cause for the defendant-
interventor-appellees and filed briefs.


                         OPINION

O’SCANNLAIN, Circuit Judge:

  We must decide whether the United States Forest Service
complied with federal environmental law in its management
of the Bitterroot National Forest.

                               I

   In order to effectuate its management role over the national
forests, the National Forest Management Act (“NFMA”), 16
U.S.C. §§ 1600 et seq., requires the United States Forest Ser-
vice (“Forest Service”) to develop and to maintain forest
resource management plans. 

Such plans must,
among other things, “provide for diversity of plant and animal
communities based on the suitability and capability of the
specific land area.” Id.§ 1604(g)(3)(B). Additionally, the
15206             WILDWEST INSTITUTE v. BULL
Healthy Forests Restoration Act (“HFRA”), 16 U.S.C.
§§ 6501 et seq., directs the Forest Service to take action to
“reduce wildfire risk” and “enhance efforts to protect water-
sheds and address threats to forest and rangeland health.” 

(3). Specifically, the Forest Service is required
“[a]s soon as practicable” to implement an “authorized haz-
ardous fuel reduction project[ ]” on federal land where “the
existence of an epidemic of disease or insects, or the presence
of such an epidemic on immediately adjacent land and the
imminent risk it will spread, poses a significant threat to an
ecosystem component, or forest or rangeland resource.” 

   In satisfying its HFRA obligations, the Forest Service must
also comply with the National Environmental Policy Act
(“NEPA”), 42 U.S.C. §§ 4321 et seq. See 16 U.S.C.
§ 6514(a). Thus, the Forest Service must prepare an environ-
mental impact statement (“EIS”) when formulating a hazard-
ous fuels reduction project, which identifies alternatives to its
proposed action. 

42 U.S.C. § 4332(C). Such
requirement forces the agency to take a “hard look” at its pro-
posed action, and it must be “prepared early enough” in the
decision-making process “so that it can serve practically as an
important contribution . . . and will not be used to rationalize
or justify decisions already made.” Metcalf v. Daley, 

, 1141-42 (9th Cir. 2000) (internal quotation marks and
citation omitted); 40 C.F.R. § 1502.5. The Forest Service is
prohibited from “irreversibl[y] and irretrievabl[y]” commit-
ting resources before making its final decision. 40 C.F.R.
§ 1502.2(f); see also 

(internal quo-
tation marks omitted).

   The Forest Service must also give public notice of its deci-
sionmaking process and allow for public collaboration. See 16
U.S.C. § 6514(e)-(g); 42 U.S.C. § 4332; 40 C.F.R.
§ 1500.2(d). For example, HFRA requires that the Forest Ser-
vice hold “a public meeting at an appropriate location proxi-
mate to the administrative unit of the Federal land on which
                    WILDWEST INSTITUTE v. BULL                    15207
the authorized hazardous fuel reduction project will be con-
ducted” and provide advance notice of such. 16 U.S.C.
§ 6514(e)(2). Likewise, the Forest Service must give the pub-
lic an opportunity to comment on its proposed action, and it
must “assess and consider both individually and collectively”
such comments in preparing its final EIS. 40 C.F.R.
§ 1503.4(a). Specifically, the Forest Service must disclose and
respond to “ ‘any responsible opposing view which was not
adequately discussed in the draft statement . . . .’ ” Ctr. for
Biological Diversity v. U.S. Forest Serv., 

, 1167
(9th Cir. 2003) (quoting 40 C.F.R. § 1502.9(b)).

                                   II

                                   A

   In the summer of 2000, the Bitterroot National Forest in
Montana suffered severe damage caused by wildfires. The
fires affected more than 307,000 acres and the entire Middle
East Fork community1 was evacuated. Although the Middle
East Fork community was spared, many unburned fuels
remain in the area, making it a high-risk area for future wild-
fires. Additionally, the region is in the midst of a Douglas-fir
bark beetle epidemic, which is contributing significantly to
the existing fuel levels. As a result of these conditions, the
Forest Service developed the Middle East Fork Hazardous
Fuel Reduction Project (the “Project”).

   The Project was developed under HFRA. The stated pur-
poses for the Project are to (1) reduce wildland fire threats to
the Middle East Fork community, (2) restore fire-adapted eco-
systems in the Middle East Fork landscape, and (3) restore
stands affected by the Douglas-fir bark beetle epidemic by
treating infested areas and lands at risk. To determine how
best to accomplish these goals, the Forest Service was
  1
    The Middle East Fork area is two miles east of Sula, Montana, in Rav-
alli County.
15208             WILDWEST INSTITUTE v. BULL
required to “study, develop, and describe the proposed agency
action; the alternative of no action; and an additional action
alternative,” if that alternative “is proposed during scoping or
the collaborative process . . . and meets the purpose and need
of the project.” 16 U.S.C. § 6514(c)(1). Thus, here the Forest
Service considered: (1) a no-action alternative (“Alternative
1”); (2) its preferred alternative (“Alternative 2”); and (3) an
alternative proposed by The Ecology Center and Native For-
est Network (now known as the WildWest Institute), Friends
of the Bitterroot, and the National Forest Protection Alliance
(“Alternative 3”).

   The following significant events occurred during the analy-
sis and development of the Project. The Forest Service held
at least twelve public meetings in the aftermath of the 2000
fires. Thereafter, the Forest Service began developing a vege-
tation treatment plan for the Middle East Fork Area. On
March 18, 2004, a meeting was held in Sula, Montana to “dis-
cuss the results of the analysis and the initial recommenda-
tions” of implementing fuel reduction, addressing the bark
beetle problem, and addressing timber harvest. Notice of the
meeting was published in the Ravalli Republic newspaper, as
well as through a press release. Another public meeting was
held in Sula on September 28, 2004. A letter of invitation
was mailed to approximately 120 individuals, organizations,
and other agencies.

   In November 2004, the Forest Service published a Notice
of Intent to prepare an Environmental Impact Statement in the
Federal Register. The Draft EIS analyzed the three alterna-
tives discussed above and was made available to the public
through the Bitterroot website. It was also mailed to members
of the mailing list maintained by the Forest Service. The
notice of availability was published in the Federal Register on
April 22, 2005, thereby starting the 45-day comment period.
Public comments were received and incorporated into the
Final EIS, which was released in September 2005. During the
comment process, the Forest Service responded to seventeen
                      WILDWEST INSTITUTE v. BULL                     15209
Freedom of Information Act requests and provided over
15,500 pages of information.

   When the Final EIS was released, Forest Supervisor Dave
Bull held a press conference to announce the final decision.
Certain vocal opponents of the Project were excluded from
the press conference, including Jim Miller, President of the
Friends of the Bitterroot. Thereafter, Supervisor Bull publicly
apologized for holding a closed press conference.

   In March 2006, the Forest Service published its Record of
Decision (“ROD”), which detailed its decision to implement
a modified version of Alternative 2. Alternative 2—Modified
called for treating2 4,938 acres in the Middle East Fork area,
with approximately 59% of all such treatments (2,893 acres)
involving commercial treatments. The Forest Service made a
number of changes to its original proposed Alternative 2 dur-
ing the time between its publication of the Final EIS and the
ROD in response to objections raised by WildWest. First, the
Forest Service decided not to treat old growth habitat. Second,
the modified plan would not treat units with greater than 15%
detrimental soil reliance or units or portions of units projected
to possibly have greater than 15% disturbance after the treat-
ments. Third, the modified plan did not call for performing
summer ground-based tractor harvest operations because the
units could have greater than 15% disturbance after such
treatments. Finally, the modified plan would “not treat stands
where it has been determined that further field review is
needed to corroborate the determination by the soil scientist
of evidence of past harvest.” Overall, the Forest Service
dropped 1,534 acres from its proposed treatment plans in
response to WildWest’s objections.
  2
    Most acres were to be treated by prescribed fire, a technique that seeks
to reduce dangerous fuels such as brush and dead wood. Other treatments
included removal of small diameter trees in overstocked areas, removal of
infected or infested trees, and removal of larger trees at increased risk for
beetle infestation.
15210             WILDWEST INSTITUTE v. BULL
                                B

   After the Forest Service’s final decision to implement the
Project, WildWest filed suit in the United States District
Court for the District of Montana, asserting that the Forest
Service’s decisionmaking process, as well as its substantive
decision, violated NEPA, NFMA, and HFRA. Specifically,
WildWest asserts the Forest Service committed procedural
violations by (1) irretrievably committing resources in favor
of its preferred alternative before making its final decision, (2)
failing to engage in adequate public collaboration, and (3)
ignoring competing scientific views. Substantively, WildWest
challenges the Project’s impact on soil productivity, old
growth habitat, species viability, and watershed sedimenta-
tion.

   Along with its complaint, WildWest moved for a temporary
restraining order and preliminary injunction, primarily relying
on its three procedural claims. The district court denied the
motion, concluding in part that WildWest had made an insuf-
ficient showing that it was likely to prevail on the merits of
such claims. We affirmed. WildWest Inst. v. Bull, 

(9th Cir. 2006).

   Thereafter, the parties filed cross-motions for summary
judgment. The district court granted summary judgment in
favor of the Forest Service defendants on all of WildWest’s
claims. The district court ruled against WildWest on the mer-
its of its claims, but it also held that the procedural claims
were waived as WildWest only “summarize[d] those argu-
ments as necessary to preserve them for appeal” because such
claims were previously raised at the preliminary injunction
stage. WildWest Inst. v. Bull, 

, 1241 (D.
Mont. 2006) (internal quotation marks omitted).

   After the district court’s summary judgment ruling, Wild-
West moved for an emergency injunction pending appeal in
the district court, which was denied. Order Denying Motion
                  WILDWEST INSTITUTE v. BULL              15211
to Stay, Jan. 18, 2007. WildWest then made a similar motion
in this court, which we also denied. Order of February 5,
2008. WildWest now appeals the district court’s summary
judgment ruling.

                              III

   [1] Before issuing its final decision, the Forest Service is
prohibited from taking any action that “limit[s its] choice of
reasonable alternatives” identified in the decision-making pro-
cess. 40 C.F.R. § 1506.1(a)(2). Such prohibition extends to
“commit[ting] resources” which would prejudice the Forest
Service’s selection of alternatives. 

Construing
a related regulation, we have held that a premature “ ‘irrevers-
ible and irretrievable commitment of resources’ ” violates
NEPA. 

(quoting Conner v. Burford,

, 1446 (9th Cir. 1988)). As the Fourth Circuit
has explained: “The proper inquiry in a NEPA case is . . . not
whether an agency has focused on its preferred alternative,
but instead whether it has gone too far in doing so, reaching
the point where it actually has ‘[l]imit[ed] the choice of rea-
sonable alternatives.’ ” Nat’l Audubon Soc’y v. Dep’t of the
Navy, 

, 206 (4th Cir. 2005) (alteration in origi-
nal) (quoting 40 C.F.R. § 1506.1(a)(2)).

   [2] Here, WildWest argues the Forest Service prematurely
and irretrievably committed financial resources in favor of its
preferred alternative by marking trees in preparation for log-
ging during the public comment period. Our cases have
focused on the commitment of natural resources, not neces-
sarily the agency’s financial resources. For example, in
Friends of Southeast’s Future v. Morrison, 

(9th Cir. 1998), we determined that there was no NEPA viola-
tion because the Forest Service had not “irreversibly and irre-
trievably committed the resources of Ushk Bay to logging.”

(emphasis added). Although the government had
developed a tentative schedule designating certain forest areas
for harvest, we emphasized that the schedule “ma[de] no com-
15212             WILDWEST INSTITUTE v. BULL
mitment of any part of the national forests because the gov-
ernment retain[ed] absolute authority to decide whether any
such activities will ever take place on the lands.” 

(internal punctuation and citation omitted).

   Conversely, in Conner, we held that the government did
prematurely commit resources in violation of NEPA where it
sold gas and oil leases on national forest land without “reserv-
[ing] . . . the absolute right to prevent all surface-disturbing

; see also 

(holding that irretrievable commitment occurred when
the government agreed to assist an Indian tribe in resuming
whaling without conditioning such agreement “upon a NEPA
determination that the . . . whaling proposal would not signifi-
cantly affect the environment”). We specifically distin-
guished, however, similar leases that “make no commitment
of any part of the national forests to surface-disturbing activi-
ties by the lessees because the government retains absolute
authority to decide whether any such activities will ever take
place on the leased lands.” 

(first
emphasis added).

   WildWest is correct that a financial commitment can, in
some instances, constitute an irretrievable commitment. For
example, if an agency spent most or all of its limited budget
on preparations useful for only one alternative, it may well
have taken action “[l]imit[ing] the choice of reasonable alter-
natives.” 40 C.F.R. § 1506.1(a). In the instant case, the Forest
Service’s expenditure of $208,000 to pre-mark trees was
clearly not so substantial an investment that it limited such
choice.

  [3] Thus, on the current record, the Forest Service’s pre-
marking of trees did not irretrievably commit it to a particular
course of action. Although the Forest Service undertook pre-
paratory actions in favor of logging, it clearly retained the
authority to change course or to alter the plan it was consider-
ing implementing. Such is clearly demonstrated by the fact
                  WILDWEST INSTITUTE v. BULL               15213
that it excluded 410 acres of pre-marked timber from the
modified plan it ultimately adopted. See Friends of South-
east’s 

(holding the government did
not irretrievably commit forest resources where it retained the
authority to change the amount of timber ultimately har-
vested); see also Headwaters v. Forsgren, 

, 1129 (D. Or. 2002) (“The record does not show . . . that
by pre-marking . . . trees, the Forest Service made such an
‘irreversible and irretrievable commitment of resources’ . . .
to effectively foreclose other alternative actions.”). The Forest
Service’s actions here easily support this assertion. Thus, we
conclude the Forest Service’s decision to pre-mark trees for
cutting did not prematurely commit it to a specific course of
action in violation of NEPA.

                               IV

   WildWest also argues the Forest Service failed to facilitate
sufficient public collaboration. Specifically, WildWest asserts
the Forest Service (1) provided insufficient notice of the Sula,
Montana, meeting by removing conservation proponents from
the notice mailing list and by failing to announce that a HFRA
project would be discussed; (2) failed to indicate in its Final
EIS that it received over 11,000 public comments opposing its
proposed action; and (3) failed to give notice of when its Final
EIS would be issued and improperly excluding conservation
proponents from the press conference announcing such deci-
sion.

   NEPA dictates that federal agencies “shall to the fullest
extent possible . . . [e]ncourage and facilitate public involve-
ment in decisions which affect the quality of the human envi-
ronment.” 40 C.F.R. § 1500.2(d). As mentioned above, HFRA
also sets forth a specific public participation scheme. See 16
U.S.C. § 6514(e)-(g). However, in its reply brief, WildWest
clearly asserts that it is not alleging the Forest Service vio-
lated HFRA. Thus, we proceed to analyze its public collabo-
ration claims solely under NEPA.
15214               WILDWEST INSTITUTE v. BULL
                                   A

   As we have previously noted, the regulations governing
public involvement adopted pursuant to NEPA “are general in
approach.” Bering Strait Citizens for Responsible Res. Dev. v.
U.S. Army Corps of Eng’rs, 

, 952 (9th Cir.
2008). There are few specific requirements. See 40 C.F.R.
§ 1506.6.

   [4] If an agency holds a public meeting, NEPA requires
that it provide notice of such. 40 C.F.R. § 1506.6(b). The par-
ticular form such notice must take, however, is not specified.3
Envtl. Coal. of Ojai v. Brown, 

, 1415 (9th Cir.
1995). Here, the Forest Service held two public meetings,
and, before doing so, published an announcement in a local
newspaper, issued a news release, and sent individual notices
to interested members of the public identified on a mailing
list. In any event, the record suggests WildWest did in fact
receive notice from the Forest Service’s mailing through its
predecessor, The Ecology Center. We also reject WildWest’s
claim that the Forest Service’s announcement was incomplete
because it did not state that the meeting would address a
“HFRA hazardous fuel reduction project.” The Forest Ser-
vice specifically stated that the meeting would allow the pub-
lic an opportunity to discuss its “recommendations concerning
fuel reduction, thinning, bark beatles, [and] timber harvest.”

                                   B

   The Forest Service is required to “assess and consider . . .
both individually and collectively” the public comments
received during the NEPA process and to respond to such in
  3
   The regulations distinguish between notice requirements for actions of
national concern and those of only local concern. Because WildWest does
not allege that the Forest Service failed to provide adequate national
notice, we assume for purposes of this appeal that it need only provide
adequate local notice.
                    WILDWEST INSTITUTE v. BULL                  15215
its Final EIS. 40 C.F.R. § 1503.4(a). Again, however, the reg-
ulations do not mandate the form such response must take,
instead providing a list of options including: modifying the
agency’s proposed action, developing or evaluating additional
alternatives, supplementing its analyses, making factual cor-
rections, or explaining “why the comments do not warrant
further agency response.”

   [5] WildWest argues that the Forest Service violated NEPA
by failing to identify in the Final EIS the number of opposing
comments it received. We disagree. There is no requirement
that the Forest Service identify the number of public com-
ments it received. See 40 C.F.R. §§ 1502.9(b), 1503.4. Fur-
ther, the Final EIS responds appropriately to the public
comments. The Forest Service summarized the nature of the
comments it had received and the primary concerns that were
raised. It also provided specific substantive responses to such
concerns and included a six and a half page list of the com-
menters.

                                  C

   [6] Finally, WildWest contends that the Forest Service vio-
lated NEPA by failing to announce when it would make its
final decision and by excluding individuals who opposed its
decision from the press conference announcing such. Again,
NEPA does not require that federal agencies give notice of
when it will announce its final decision or that it make such
announcement a public event. Thus, even though the Forest
Service’s decision to hold a closed press conference was per-
haps unfortunate, it was not a violation of NEPA.

                                  V

  WildWest contends that the Forest Service violated
NFMA’s substantive soil productivity requirement.4 Relat-
  4
   As we turn to WildWest’s substantive claims, we are reminded by this
court’s recent en banc decision in The Lands Council v. McNair, No. 07-
15216                 WILDWEST INSTITUTE v. BULL
edly, it argues that the Forest Service violated NEPA by disre-
garding the data and opinions of its own soils expert, Ken
McBride (“McBride”).

                                      A

   [7] In addition to their general obligation to respond to pub-
lic comments under 40 C.F.R. § 1503.4(a), federal agencies
must specifically “ ‘discuss at appropriate points in the final
[EIS] any responsible opposing view which was not ade-
quately discussed in the draft [EIS] and . . . indicate the agen-
cy’s response to the issues raised.’ ” Ctr. for Biological

(quoting 40 C.F.R. § 1502.9(b)).
A failure to do so is itself a NEPA violation. 

An
agency must also “insure the professional integrity, including
scientific integrity, of the discussions and analyses” included
in its EIS. 40 C.F.R. § 1502.24.

  The Forest Service’s Draft EIS incorporated McBride’s soil
analysis report, but McBride indicated that his report had been
edited, resulting in a “deliberate removal of information that

35000, 

(9th Cir. July 2, 2008) (en banc), that we do
not “act as a panel of scientists that instructs the Forest Service how to val-
idate its hypotheses regarding wildlife viability, chooses among scientific
studies in determining whether the Forest Service has complied with the
underlying Forest Plan, and orders the agency to explain every possible
scientific uncertainty.” 

Rather, we only require “that the Forest
Service . . . support its conclusions that a project meets the requirements
of the NFMA and relevant Forest Plan with studies that the agency, in its
expertise, deems reliable. The Forest Service must explain the conclusions
it has drawn from its chosen methodology, and the reasons it considers the
underlying evidence to be reliable. We will conclude that the Forest Ser-
vice acts arbitrarily and capriciously only when the record plainly demon-
strates that the Forest Service made a clear error in judgment in
concluding that a project meets the requirements of the NFMA and rele-
vant Forest Plan.” 

Although the issuance of McNair between
oral argument and the filing of this opinion does not change our views on
any substantive topic, it does strengthen them.
                  WILDWEST INSTITUTE v. BULL              15217
accurately portrayed the conditions of the soils and the pre-
scriptions and mitigations needed to address those degraded
soil conditions.” Thereafter, the Forest Service formed a peer
review group to “check for consistency and repeatability and
compare [its] results . . . to [McBride’s].”

   The peer review group performed a sample analysis of
some of the units previously analyzed by McBride. The group
used a modified Howes methodology, which allows soil to be
classified under six different categories but does not assume
that all soil disturbance is detrimental. McBride, on the other
hand, used the “ ‘rapid’ method for determining soil distur-
bance,” under which “only areas that are unchanged from nat-
ural conditions are recorded as non detrimental disturbance.”
Though the peer group identified a comparable amount of dis-
turbance, it found a lesser degree of detrimental disturbance
and it concluded all of the sample areas satisfied soil quality
guidelines. The group further indicated that its chosen meth-
odology was more appropriate in “assessing the threshold for
an activity unit,” whereas McBride’s methodology “should be
used to flag project areas that may depart from the soil quality
guidelines for more accurate and intensive assessment at the
project level.”

   [8] In its Final EIS, the Forest Service adopted the peer
review group’s findings, concluding that McBride’s method-
ology “overestimate[d] the amount of detrimental soil dam-
age.” (emphasis added). It further discounted McBride’s
survey as “primarily ocular . . . with any measurements being
used to verify what was being observed.” Contrary to Wild-
West’s assertion, however, the Forest Service did not com-
pletely disregard McBride’s findings. Rather, the Final EIS
states that even though McBride’s methodology “illustrate[s]
a conservative approach to estimating soil condition,” his
findings regarding the level of disturbance are “used in the
direct, indirect, and cumulative effects discussions.” In fact,
the Forest Service modified the Project in its ROD, providing
that it would “not treat any units with 15% or greater soil dis-
15218             WILDWEST INSTITUTE v. BULL
turbance or units or portions of units projected to possibly
have 15% or greater detrimental disturbance after proposed
treatments,” resulting in postponed action on numerous units.

   WildWest also argues that the Forest Service removed from
its analysis all of McBride’s findings regarding cumulative
soil impacts. However, the Final EIS contains a detailed anal-
ysis of the cumulative impact caused by logging and other
historical activities, and specifically acknowledges McBride’s
concern that “[t]ime periods of 20-40 years may be required
to improve the surface layer.” The Final EIS also includes an
analysis of the soil disturbance in each subwatershed. That the
Forest Service reordered the format of its cumulative impact
analysis from the Draft EIS to the Final EIS does not support
WildWest’s contention that it failed to address such issue. See
Nat’l Parks & Conservation Ass’n v. U.S. Dep’t of Transp.,

, 682 (9th Cir. 2000) (explaining that “[w]e
review the EIS as a whole”).

                               B

   [9] Under NFMA, the Forest Service may harvest timber
from national forests only where “soil, slope, or other water-
shed conditions will not be irreversibly damaged.” 16 U.S.C.
§ 1604(g)(3)(E)(i). Further, NFMA directs the Forest Service
to evaluate the effects of its management plans on the produc-
tivity of the land “based on continuous monitoring and assess-
ment in the field.” 

The Bitterroot
National Forest Plan does not provide specific numeric soil
quality standards. Therefore, the Forest Service applied its
Region One Soil Quality Standards (“R1-SQS”), which pro-
vide that activities should not result in a cumulative detrimen-
tal impact of more than 15% of the activity area. WildWest
makes two arguments related to this substantive requirement.
First, it argues that the R1-SQS are facially unreliable. Sec-
ond, it argues that the Forest Service failed to properly con-
sider cumulative impacts in determining whether the project
satisfied such standard.
                    WILDWEST INSTITUTE v. BULL                   15219
                                   1

   WildWest failed to raise its facial challenge in the district
court. Generally, an issue raised for the first time on appeal
is deemed waived. Cold Mountain v. Garber, 

,
891 (9th Cir. 2004). We have discretion, however, to consider
a newly raised issue (1) in the “exceptional” case where “re-
view is necessary to prevent a miscarriage of justice or to pre-
serve the integrity of the judicial process,” (2) when the issue
arises while on appeal “because of a change in the law,” or (3)
when the issue is “purely one of law and either does not
depend on the factual record developed below, or the perti-
nent record has been fully developed.” 

   WildWest did not address these discretionary factors in its
briefing, instead baldly asserting that “this case presents an
appropriate fact situation for this Court to pass on the ade-
quacy of the R1-SQS.” However, on our own review of such
factors, we conclude it is not appropriate to consider Wild-
West’s newly raised argument in the first instance. The first
two factors do not apply here, and we conclude that the cur-
rent record is not sufficiently developed as to the issue Wild-
West seeks to raise.

                                   2

  [10] WildWest also contends the Forest Service violated
NFMA by failing properly to consider cumulative impacts on
soil productivity.5 WildWest analogizes this case to Lands
Council v. Powell, 

(9th Cir. 2005), where we
held that the Final EIS violated NEPA because it contained
“no discussion of the connection between individual harvests
and the prior environmental harms from those harvests that
  5
   Much of WildWest’s argument on this point appears to conflate
NFMA’s substantive requirement with NEPA’s procedural requirement
that agencies conduct cumulative impact analyses. 40 C.F.R. § 1508.7. We
have previously rejected WildWest’s procedural challenges on this point.
15220             WILDWEST INSTITUTE v. BULL
the Forest Service now acknowledges.” 

As dis-
cussed above, however, the record shows that the Forest Ser-
vice took a “hard look” at cumulative impacts from past
timber harvests, as well as other historical activities and
events, and concluded that these past activities did not reduce
soil or site productivity in violation of the soil quality stan-
dards.

  WildWest also argues the Forest Service erred by narrow-
ing its soil conditions analysis to particular proposed harvest-
ing units instead of taking a broader landscape view. Agencies
have “discretion to determine the physical scope used for
measuring environmental impacts” so long as they do not act
arbitrarily and their “choice of analysis scale . . . represent[s]
a reasoned decision.” Idaho Sporting Cong., Inc. v. Ritten-
house, 

, 973 (9th Cir. 2002).

   [11] Here, although the Final EIS included broader sub-
watershed data as background, the Forest Service explained
that such landscape-level analysis is unreliable because it can-
not be correlated to site-specific impacts under the R1-SQS.
Specifically, it determined that such correlation is impossible
“because of the variability in soil texture, the amount of
organic matter and ground cover, soil response to past proj-
ects, and the intensity of past projects.” On this record, we
cannot conclude that such reasoning is arbitrary.

                               VI

   [12] WildWest makes several arguments with respect to the
Forest Service’s management of old growth habitats. NFMA
requires that the Forest Service’s management of national for-
ests “provide for diversity of plant and animal communities.”
16 U.S.C. § 1604(g)(3)(B). NFMA does not, however, man-
date a specific old growth standard. Rather, such standards
arise under the relevant Forest Plan. “[A]ll management activ-
ities undertaken by the Forest Service must comply with the
forest plan, which in turn must comply with the Forest Act
                  WILDWEST INSTITUTE v. BULL               15221
. . . .” 

. Also applicable here is
HFRA’s requirement that the Forest Service “maintain, or
contribute toward the restoration of, the structure and compo-
sition of old growth stands” when undertaking fuel reduction
projects. 16 U.S.C. § 6512(e)(2).

                                A

   The Record of Decision states that the Project “will not
treat old growth habitat.” WildWest challenges this assertion,
arguing the Forest Service erred in its classification of old
growth habitat.

   The Forest Service analyzed old growth habitat in the Proj-
ect area based on “stand-wide structure and characteristics,”
as dictated in the Forest Plan. As old growth characteristics
vary by region, it further relied upon criteria specifically “de-
scrib[ing] old growth forest types of the northern region . . .
[which] was intended to provide local definitions of old
growth to be used in the implementation of Forest Plans.”
Such criteria require a minimum number of live trees. Due to
the Douglas-fir bark beetle infestation, the Forest Service con-
ducted field surveys for purposes of classifying old growth
habitat in both 2004 and 2005. Over 4,421 acres were evalu-
ated on the ground.

   [13] WildWest contends the Forest Service’s classification
methodology was flawed because it based its analysis on “im-
minently dead” trees. It further argues, without citation to any
authority, that while the regional criteria utilized by the Forest
Service is proper for determining whether a stand is old
growth as an initial matter, it is not proper for determining
that a stand is no longer old growth. As the Forest Service
points out, the “imminently dead” standard applied to tree
marking, not to classifying old growth status. And in any
event, WildWest’s arguments on this point are not convinc-
ing. The Forest Service properly applied its selected method-
ology, and it disclosed such methodology, as well as its
15222                 WILDWEST INSTITUTE v. BULL
findings, to the public.6 It further addressed objections to its
methodology raised during the comment period.

   WildWest further argues that the Project violates substan-
tive old growth standards. Accepting the Forest Service’s con-
clusion that the Project, as finally implemented, does not treat
old growth, however, such arguments are unavailing.7 See
Lands 

(“[B]ecause no old growth
forest is to be harvested under the selected alternative, we
reject the contention that the Project will be impermissible
[because] . . . the ‘allocated old growth’ within the Forest is
less than the Forest Plan requirement.”).

  [14] We also reject WildWest’s argument that the Forest
Service is required to designate replacement habitat.8 Assum-
ing (but not deciding) that any such requirement exists, we
conclude the Forest Service has satisfied it by demonstrating
  6
     WildWest argues that the Forest Service violated NEPA by failing to
include the results of its surveys in the administrative record. NEPA
requires that the Forest Service disclose the hard data supporting its expert
opinions to facilitate the public’s ability to challenge agency action. See
Idaho Sporting Cong. v. Thomas, 

, 1150 (9th Cir. 1998),
overruled on other grounds by McNair, 

. Contrary to
WildWest’s assertion, however, the data supporting the Forest Service’s
assessment of units in the Project area, including silvicultural diagnoses,
silvicultural prescriptions, common stand exams, and walk through exams,
are contained in the record.
   7
     To the extent WildWest challenges the substantive standards them-
selves, we also conclude such arguments are not properly before us. See
Neighbors of Cuddy Mountain v. Alexander, 

, 1067 (9th Cir.
2002)(“[N]ot all forest-wide practices may be challenged on the coattails
of a site-specific action; there must be a relationship between the lawful-
ness of the site-specific action and the practice challenged.”).
   8
     WildWest relies on Lands Council v. Vaught, 

(E.D. Wash. 2002), in which the district court held that to comply with
NFMA, the Forest Service must “demonstrate either that adequate old
growth acreage exists in the [forest] to satisfy the [Forest Plan’s] old
growth standards or that the timber slated to be harvested under the Project
is not needed to fulfill old growth standards.” 

Vaught is not
binding on this Court, and we decline to address its merits at this time.
                  WILDWEST INSTITUTE v. BULL               15223
that the Project retains “the largest, healthiest and dominant
residual trees” that may some day become old growth. Addi-
tionally, the Final EIS indicates that the proposed treatments
will contribute to the health of the remaining trees, “mak[ing
them] more resistant to insect and disease and future fires,
meaning that in the future they have a higher probability of
contributing toward the minimum criteria for live larger trees
per acre for their respective old growth type group.”

                               B

   WildWest also argues the Forest Service failed adequately
to consider the impact the Project, particularly its logging
component, will have on population trends of management
indicator species (“MIS”) such as the pileated woodpecker,
northern goshawk, and black-backed woodpecker. The Bitter-
root Forest Plan provides that “[t]he amount and distribution
of old growth will be used to ensure sufficient habitat for the
maintenance of viable populations of existing native and
desirable non-native vertebrate species, including two indica-
tor species, the pine marten and pileated woodpecker.” The
Final EIS includes assessments of the Project’s effect on the
pine marten and pileated woodpecker, and explains that the
Project would not likely contribute to loss of their viability.
WildWest does not contest the Forest Service’s conclusions
with respect to the marten, but it does object to the Forest Ser-
vice’s conclusions regarding the pileated woodpecker and two
other species, the northern goshawk and black-backed wood-
pecker.

                               1

   [15] WildWest contends the Final EIS does not acknowl-
edge the pileated woodpecker’s preference for larger trees and
snags for nesting, despite the fact that such trees and snags
will be removed under the Project. It further argues other fac-
tors such as patch size and canopy cover were ignored. On the
contrary, however, the Final EIS does consider the pileated
15224                WILDWEST INSTITUTE v. BULL
woodpecker’s habitat needs. For example, it notes that the
pileated woodpecker prefers snags “greater than 21 inches in
diameter at breast height,” “areas with 10% cover,” and nests
“more than 40 feet off the ground.” Although the Forest Ser-
vice acknowledges the possibility that “some snags, which
could be used for nesting, would be removed or burned,” it
notes that “the trees which are potentially important to pile-
ated woodpeckers for nesting will generally be retained
through project design and mitigation.” The Final EIS also
discusses historical impacts on pileated woodpeckers,
explaining that “[p]rior timber harvests, fires, and fire sup-
pression do not appear to have affected species viability
because monitoring indicates they are present on all monitor-
ing transects and detections have increased slightly each year
between 2000 and 2003.” On this record, we conclude the
Forest Service took the requisite “hard look” at the Project’s
potential impact on the pileated woodpecker.

                                     2

   Next, WildWest argues the Forest Service inadequately
analyzed the impact on the northern goshawk, which it deems
a “sensitive species.”9 Acknowledging that the Project avoids
known nesting habitats, WildWest complains the Forest Ser-
vice failed to conduct surveys to identify nest stands. How-
ever, the Final EIS indicates the Forest Service did conduct
such surveys within the Project area that identified existing
and potential nest territories,10 and the Project avoids treat-
ment within thirty acres of these areas.

   According to WildWest, the Final EIS “ignored the best
science available at the time of analysis,” which suggests that
“it is essential to the viability of goshawks that 20-60% older
  9
    Although the goshawk was once deemed a sensitive species, it was
removed from the Forest Service’s sensitive species list in 2004.
   10
      Since goshawks “can be notoriously difficult to find,” the Forest Ser-
vice was unable to locate them in the potential nest stands.
                   WILDWEST INSTITUTE v. BULL               15225
forest levels be maintained in goshawk territories.” To sup-
port its contention, WildWest moved to supplement the record
with a declaration from a former Forest Service employee, Dr.
Sara Johnson. The district court denied such motion, however,
because it was filed after the completion of summary judg-
ment briefing.

   We may consider extra-record materials (1) when neces-
sary to determine whether the agency considered all relevant
factors in making its decision; (2) when the agency has relied
on extra-record materials; (3) when necessary to explain tech-
nical terms or complex subject matter; or (4) when the agency
has acted in bad faith. Inland Empire Pub. Lands Council v.
Glickman, 

, 703-04 (9th Cir. 1996). WildWest
argues the first scenario applies because the Johnson declara-
tion responds to an “unsigned memorandum in the record
dated March 8, 2006—seven months after the F[inal] EIS, and
only a few weeks prior to the final decision.” We disagree.
Although WildWest attempts to characterize the memoran-
dum as “new information not provided to the public in the
F[inal] EIS,” as the district court noted, the memorandum
responds to objections raised during the administrative pro-
cess and explains information previously included in the
NEPA documents.

   [16] We conclude the Forest Service satisfied NEPA by
properly considering the northern goshawk’s habitat needs.
The Final EIS explicitly recognizes that the goshawk prefers
a “nesting habitat with a dense overstory of large trees and an
open understory of grass and shrubs.” It also explains that the
Douglas-fir bark beetle epidemic has had a significant impact
on goshawk habitat in the Project area because it eliminated
the “high live crown canopy.” Although the Project may
affect goshawk habitat, the Forest Service determined that
treatment would retain live overstory, and that “the incremen-
tal small reduction of canopy closure through treatment would
likely be a negligible reduction, if a reduction at all, in habitat
suitability.” Additionally, considering the impact on goshawk
15226             WILDWEST INSTITUTE v. BULL
prey, the Forest Service determined that treatments “may
actually improve or create foraging habitat for goshawks.”

                               3

   [17] Finally, WildWest asserts the Forest Service did not
properly consider the Project’s impact on the black-backed
woodpecker. Although the Forest Service acknowledges that
“[f]ire suppression and extensive salvaging of burned trees
has probably reduced habitat for black-backed woodpeckers,”
the record supports its conclusion that the Project “would
have no impact on black-backed woodpeckers or their habi-
tat.” First, as noted in the Final EIS, the Forest contains
ample habitat for black-backed woodpeckers adjacent to the
Project area. Second, the Project area contains only minimal
black-backed woodpecker habitat because of the lack of mod-
erate to high intensity fires in the past. Thus, we reject Wild-
West’s argument on this point as well.

                              VII

  For the foregoing reasons, we conclude the Forest Service
was properly granted summary judgment on all of the claims
asserted against it by WildWest.

  AFFIRMED.