MEMORANDUM **
Plaintiffs Ecology Center, Inc. and Native Ecosystem Council (collectively “Ecology Center”) appeal summary judgment entered in favor of the Forest Service. Ecology Center argues that the Forest Service’s approval of the Dry Fork Vegetation Restoration Project for the Lewis and Clark National Forest violated the National Environmental Policy Act, 42 U.S.C. § 4321 et seq. (“NEPA”) and the National Forest Management Act, 16 U.S.C. § 1601 et seq. (“Forest Act”).
We review summary judgments de novo. “Judicial review of agency decisions under [the Forest Act] and NEPA is governed by the Administrative Procedure Act (‘APA’), which specifies that an agency action may be overturned only where it is found to be *795‘arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.’ ” Neighbors of Cuddy Mountain v. Alexander, 303 F.3d 1059, 1065 (9th Cir.2002) (quoting 5 U.S.C. § 706(2)(A)).
We conclude that the Forest Service acted arbitrarily and capriciously in approving the Dry Fork Vegetation Restoration Project for the Lewis and Clark National Forest. The Forest Service has failed to demonstrate that the project is consistent with the forest plan’s old growth preservation standards and goshawk monitoring requirements.
I
“Pursuant to the [Forest Act], the Forest Service must demonstrate that a site-specific project would be consistent with the land resource management plan of the entire forest.” Neighbors of Cuddy Mountain v. United States Forest Serv., 137 F.3d 1372, 1377 (9th Cir.1998) (citing. 16 U.S.C. § 1604(D).
Ecology Center argues that the project plan is contrary to the forest plan for old growth timber. The forest plan requires the Forest Service to maintain 5 percent of the “commercial forest land” in each timber compartment in old growth forest condition.
The Forest Service relies on the declaration of Thomas Whitford to show that it has met the requirement for old growth preservation standards. Whit-ford’s declaration was prepared during litigation. The declaration is not part of the administrative record that the Forest Service compiled in support of the project. “[Jjudicial review of an agency decision typically focuses on the administrative record in existence at the time of the decision and does not encompass any part of the record that is made initially in the reviewing court.” Lands Council v. Powell, 395 F.3d 1019, 1029 (9th Cir.2005) (quoting Southwest Ctr. for Biological Diversity v. United States Forest Serv., 100 F.3d 1443, 1450 (9th Cir.1996)). The Forest Service cannot meet any of the exceptions to this requirement. Id. Indeed, the calculations and analysis in the Whitford Declaration highlight the fact that the Forest Service failed to make a determination whether the project was consistent with the forest plan before approving the project. The Forest Service may not use supplemental materials such as the Whitford Declaration “to present information and analysis that it was required[ ] but ... failed to include in its original NEPA documents.” Idaho Sporting Congress v. Alexander, 222 F.3d 562, 567 (9th Cir.2000).
The environmental impact statement does not contain data about the amount of old growth forest on commercial forest lands in each timber compartment, as the standard requires. The Forest Service failed to demonstrate that the project is consistent with the forest plan’s old growth forest standards, and thus it failed to comply with the Forest Act.
II
Ecology Center contends that the Forest Service’s method of population sampling for the northern goshawk satisfies neither its procedural duty to ensure that the project is consistent with the forest plan nor its substantive duty under the Forest Act to “provide for diversity of plant and animal communities.” 16 U.S.C. § 1604(g)(3)(B). The Forest Service has complied with its substantive duty under the Forest Act, but has failed to ensure that the project is consistent with the forest plan.1
*796A
The Forest Service has met its substantive duties under the Forest Act by showing that the project will not affect the viability of the goshawks. The Forest Service surveyed the 362 acres of potential goshawk habitat that the project will affect, which is 4.7 percent of the available potential goshawk habitat in the timber compartments affected by the project, and determined that there are no nesting territories in this area. Contrary to Ecology Center’s assertions, the absence of goshawks in this area is not cause for alarm, but instead indicates that the project will not affect any goshawks. Furthermore, the absence of goshawks on those 362 acres is not surprising because a goshawk’s nesting home range consists of about 6,000 acres.
B
The Forest Act encourages but does not require the Forest Service to “assess population viability in terms of actual population size[ ][or] population trends.” Inland Empire Pub. Lands Council v. United States Forest Serv., 88 F.3d 754, 761 n. 8 (9th Cit.1996) (emphasis added). The forest plan in this case is more stringent than the minimum required by the Forest Act, as it requires the Forest Service to annually monitor all active nesting territories of the goshawk and to take further action if there is a 10 percent or more decrease in active nesting territories.
This standard leaves little room for error. We conclude that the Forest Service has failed to meet it. The Forest Service monitored most but not all of the goshawk nesting territories from 1997 to 2001 in the Jefferson Division of the forest, and from 1999 to 2001 in the Rocky Mountain Division. The Forest Service acknowledged in its 2000-2001 Monitoring and Evaluation Report that “[t]he only trend data that can be examined is from 1997 to 2001 for the Jefferson Division.” However, it failed to analyze the trend data and failed to evaluate whether its requirement to take further action was indicated.2 The forest plan utilizes the requirements of monitoring active nesting territories and further evaluation if a 10 percent decrease occurs in order “to insure that the Forest Act’s requirement of maintaining viable populations of native species, including old growth dependent species, is met.” Idaho Sporting Congress, Inc. v. Rittenhouse, 305 F.3d 957, 966 (9th Cir.2002). The failure of the Forest Service to comply with the governing forest plan does not permit project approval.
Ill
Ecology Center contends that the Forest Service does not use the term “cona*797mercial forest land” consistently. Although the forest plan supports the Forest Service’s interpretation that the term “commercial forest land” applies to 50 percent of the total forest land, Ecology Center is correct that in certain instances, including one in the environmental impact statement, the Forest Service appeared to indicate that the old growth standard applied to suitable forest land instead of commercial forest land. However, the Forest Service is correct that only 20 percent of the total forest land may be developed even though 50 percent is commercial; the remaining 30 percent is designated as “forest land not appropriate for timber production.” The district court, in the first instance, is the proper forum in which to consider this issue.
REVERSED AND REMANDED.
This disposition is not appropriate for publication and may not be cited to or by the courts of this circuit except as provided by Ninth Circuit Rule 36-3.
. The Forest Service is not obligated to adopt a landscape-scale strategy for managing goshawks because this approach was not set forth in a "site specific management recommendation.”
. It is not clear from the data whether or not a 10 percent decrease occurred in active nesting territories. In 2001, the number of active territories in the Jefferson Division dropped to 9 from 11 in 2000 and from 10 in 1998 and 1999. These numbers alone suggest that in 2001 the number of territories in the Jefferson Division was down 18.19 percent from the year before, and down 10 percent from the two years before that. Similarly, between 1999 and 2000, the combined active territories from both divisions dropped from 17 to 13, a 23.53 percent decrease. These basic calculations fail to consider other factors that might be important to a calculation of population trends, such as the increasing total number of nesting territories found each year, the number of known nesting territories not monitored in a given year, the average number of active territories found over a certain number of years, the combination of data from the two divisions, and the expected rate of error. This court is ill-equipped to determine the exact formula that should be used to calculate the population trend; nevertheless, it is apparent that the Forest Service did not perform the required calculation, because a simple calculation of the numbers shows that the 10 percent threshold was triggered, yet apparently no further action was taken.