The Sierra Club1 appeals the district court's grant of summary judgment to the United States Forest Service ("Forest Service") and intervenor timber companies in connection with the Forest Service's decision to allow seven timber sales in Georgia's Chattahoochee National Forest, which will enable logging (including clearcutting), road building and related activities. On appeal, Sierra Club asserts that the decision to permit the timber sales, which it contends will damage the forest environment, was arbitrary and capricious and thus violated the National Forest Management Act ("NFMA"), 16 U.S.C. § 1600, et seq., and the substantive regulations promulgated under NFMA. See 36 C.F.R. §§ 219.1 et seq. We reverse.

The Chattahoochee and Oconee National Forests ("Forest") encompass 741,000 acres in the Appalachian Mountains of northern Georgia. In 1991, the Forest Service proposed to sell the timber rights to seven tracts within the Forest, totaling approximately 2,000 acres. In addition to the logging itself, the timber projects would require the construction of eighteen miles of roads into wilderness areas of the Forest, leading to a discharge of 155.1 tons of sediment into surrounding rivers and streams.

The Forest Service adopted the Land and Resource Management Plan ("Forest Plan") for the Forest in 1985 and amended it in 1989.2 Before any sales of timber can occur within the Forest, the Plan requires the Forest Service to conduct a site-specific study to determine whether the proposed timber sale would harm the area or its resident species. After conducting a study of the projected impact of the sales in question, the Forest Service determined that there would be no adverse impact and approved the sales.

Sierra Club subsequently filed suit under the Administrative Procedure Act ("APA"), 5 U.S.C. § 706, contending that these timber cutting projects would harm plant and animal species in the Forest. Sierra Club argued that, in conducting its study, the Forest Service did not obtain, and therefore did not consider, population inventory and population trend data for proposed, endangered, threatened, or sensitive species of plants and animals (collectively, "PETS species"), as required by the Forest Plan and the Forest Service's own regulations. Without such data, Sierra Club claimed that the study of the affected area was inadequate, making the decision to sell the timber parcels arbitrary and capricious. Sierra Club also argued that the decision to approve the sales violated 36 C.F.R. §§ 219.12, 219.19 & 219.26 because the Forest Service lacked the population data required by those regulations as well. Finally, Sierra Club challenged the Forest Plan itself, contending that it does not conform with NFMA because the proposed clearcutting will not adequately protect the Forest's soil, watershed, fish, and wildlife as required by the statute. See 16 U.S.C. § 1604(g)(3)(F)(v).

The district court granted summary judgment to the Forest Service and timber intervenors, holding that the Forest Service was not required to obtain the population and population trend data for PETS species before approving the timber sales and therefore that the Forest Service did not act arbitrarily and capriciously. Moreover, the district court found that Sierra Club's challenges to the timber sales under 36 C.F.R. § 219 did not lie because the regulations deal specifically with the formulation of forest plans, rather than site-specific actions initiated under an extant forest plan. Sierra Club now appeals.

We review grants of summary judgment de novo. Northlake Regional Medical Center v.Waffle House System Employee Benefit Plan, 160 F.3d 1301, 1303 (11th Cir.1998). Under the APA, agency actions should be reversed if they are found to be "arbitrary and capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2)(A).

Sierra Club first argues that the Forest Service violated NFMA by failing to comply with the Forest Plan's requirement that population inventory information be gathered and considered before implementing any decision affecting areas within the Forest. For each proposed project within the Forest, the Plan requires that the Forest Service perform a site-specific Environmental Assessment (EA), including a Biological Evaluation (BE)3 of how the area will be affected by the project.4 Specifically, the Forest Plan states in relevant part:

A biological evaluation of how a project may affect any species federally listed as threatened, endangered, or proposed, or identified by the Forest Service as sensitive, is done as part of the site-specific environmental analysis. This evaluation considers all available inventories of threatened, endangered, proposed and sensitive species populations and their habitat for the proposed treatment area. When adequate population inventory information is unavailable, it must be collected when the site has high potential for occupancy by a [PETS] species.

Pursuant to the Forest Plan, a BE was done as part of the EA for each of the seven timber projects at issue. Thereafter, the Forest Service determined that no further evaluations were necessary and issued Findings of No Significant Impact (FONSIs) for each tract. The locus of this dispute is whether the Forest Service, in conducting its BEs and EAs, adequately researched the potential impact of the proposed timber sales before issuing FONSIs.

There is no disagreement between the parties that numerous plants and animals identified by the Forest Service as sensitive as well as several that are endangered inhabit the proposed timber project areas. In addition, the parties agree that the habitat in sections of the project areas are suitable for other sensitive and endangered species. However, the Forest Service had no population inventory information and little in the way of population data for thirty-two of the thirty-seven vertebrate PETS species that inhabit the Forest. Sierra Club contends that, in light of the acknowledged presence of many PETS species in the areas at issue, the Forest Service was required by the Forest Plan to gather population data before permitting the timber sales to proceed. By failing to collect these data, Sierra Club argues, the Forest Service violated the Forest Plan and the provision of NFMA mandating compliance with the Plan.

The Forest Service, on the other hand, argues that its data are adequate and that population studies are required only if the site has a high potential for occupancy by PETS species. It maintains that its field visits and consultation of compartment maps, CISC5 data and Georgia Natural Heritage Program ("GNHP")6 maps indicate that the sites of the timber sales either do not have high potential for occupancy by PETS species, or suffice to demonstrate the continued viability of those PETS species that do occupy the areas. This habitat information, it asserts, is adequate to satisfy the requirements of the Forest Plan. Moreover, in its view, the Forest Service has the discretion to make determinations of potential impact based on information other than population inventory information, strictly defined.

While the Forest Service's interpretation of its Forest Plan should receive great deference from reviewing courts, "courts must overturn agency actions which do not scrupulously follow the regulations and procedures promulgated by the agency itself." Simmons v. Block, 782 F.2d 1545, 1550 (11th Cir.1986). Moreover, the Forest Service cannot ignore the requirements of the Forest Plan. As NFMA makes plain, "[r]esource plans and permits, contracts, and other instruments for the use and occupancy of National Forest System lands shall be consistent with the land management plans." 16 U.S.C. § 1604(i); see also Thomas Jefferson University v. Shalala, 512 U.S. 504, 512, 114 S. Ct. 2381, 129 L. Ed. 2d 405 (1994) (no deference due to agency interpretation that contradicts the regulation's plain language); Florida Power & Light Co. v. Lorion, 470 U.S. 729, 744, 105 S. Ct. 1598, 84 L. Ed. 2d 643 (1985) (reviewing court may remand a case to the agency "[i]f the record before the agency does not support the agency action [or] if the agency has not considered all relevant factors ...").

The Forest Service admits in numerous places in the record that sensitive species do occur within the project sites and acknowledges that those individuals would be destroyed by the proposed timber sales. It then notes in each case that because the species also exist elsewhere within the Forest, the timber projects would not significantly impact the species' diversity or viability. Yet, the Forest Service reached this conclusion without gathering any inventory or population data on many of the PETS species. Though these species are, by definition, at risk, nothing in the record indicates that the Forest Service possessed baseline population data from which to measure the impact that their destruction in the project areas would have on the overall forest population. We are nevertheless asked to defer to the Forest Service's conclusion that there will be no significant impact upon these species from the proposed timber projects. Absent record support for the Forest Service's assertions, this we cannot do. Agency actions must be reversed as arbitrary and capricious when the agency fails to "examine the relevant data and articulate a satisfactory explanation for its action including a 'rational connection between the facts found and the choice made.' " Motor Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S. Ct. 2856, 77 L. Ed. 2d 443 (1983) (quoting Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168, 83 S. Ct. 239, 9 L. Ed. 2d 207 (1962)).

More basically, however, the Forest Service argues that nothing in the regulations requires it to keep data on sensitive species and that it is therefore not necessary for it to do so. While it is true that the regulations make no such demand, the Forest Plan explicitly does so. The Forest Plan states that when adequate population inventory information is unavailable and the site has a high potential for occupancy by PETS species, then the Forest Service must gather that information. Here, the Forest Service admits that the project areas actually contain PETS species. It nonetheless maintains that its data, though devoid of any inventory information as to some PETS species, remain adequate to assess potential impact upon the species, forest-wide. The information which the Forest Service deems "adequate" is in reality no information at all in terms of many of the PETS species. Since the agency's position is contrary to the clear language of the Plan and the statute, it is not entitled to deference. We consequently hold that the Forest Service's failure to gather population inventory data on the PETS species occurring or with a high potential to occur within the project areas is contrary to the Forest Plan and, therefore, that the decision to approve the timber sales without considering this information is arbitrary and capricious.

Sierra Club next claims that the Forest Service's decision to proceed with the timber sales violated 36 C.F.R. §§ 219.19 & 219.26 because it failed to collect population data specifically for Management Indicator Species (MIS)7 (as required by § 219.19), and for all affected species (as required by § 219.26).

Section 219.26 creates a general obligation that the Forest Service gather and keep data to ensure species diversity in the planning area. It states in relevant part:

Forest Planning shall provide for the diversity of plant and animal communities and tree species consistent with the overall multiple use objectives of the planning area. Such diversity shall be considered throughout the planning process. Inventories shall include quantitative data making possible the evaluation of diversity in terms of its prior and present condition.

Section 219.19 specifically requires that the Forest Service monitor the population of Management Indicator Species, stating:

Fish and wildlife habitat shall be managed to maintain viable populations of existing native and desired non-native vertebrate species in the planning area.... (1) In order to estimate the effects of each alternative on fish and wildlife populations, certain vertebrate and/or invertebrate species present in the area shall be identified and selected as management indicator species.... (6) Population trends of the management indicator species will be monitored and relationships to habitat changes determined. [emphasis added]8

Sierra Club contends that, taken together, these two regulations obligate the Forest Service to maintain population data on all affected species in the planning area. Since the Forest Service lacks quantitative inventory data on many MIS, and the data it does have indicate that the populations of some species are inexplicably declining, Sierra Club argues that the Forest Service's approval of the timber projects was arbitrary and capricious. See Motor Vehicle Mfrs. Ass'n, 463 U.S. at 43, 103 S. Ct. 2856 (agency decisions that fail to consider important aspects of the problem or that run counter to the evidence before the agency are arbitrary and capricious).

The Forest Service responds first that neither § 219.19 nor § 219.26 apply at the site-specific level. Rather, they are relevant only during the formation of the Forest Plan and, because the Forest Plan is not a final agency action, the Sierra Club cannot challenge the Forest Plan. Second, the Forest Service contends that even if Sierra Club could bring its § 219 challenge at the site-specific level, it would still not be entitled to relief on the merits of its claims. The Forest Service notes that § 219.19 does not explicitly require the Forest Service to gather data on MIS. The regulation simply states that population trends of the MIS must be monitored and relationships to habitat changes determined. Furthermore, the Forest Service contends that to interpret § 219.26 to require that data be kept on all species makes nonsense out of § 219.19's concept of management indicator species. If the Forest Service must keep data on all species, it argues, then no purpose is served by the MIS.

We agree that the regulations refer to the formulation of Forest Plans rather than to specific projects proposed under already enacted Forest Plans. Section 219 begins by explicitly stating that "[t]he regulations in this subpart set forth a process for developing, adopting, and revising land and resource management plans for the National Forest System," 36 C.F.R. § 219.1, and the regulations make repeated reference to the forest planning process. However, the planning process does not end with the Forest Plan's approval. The obligations of the Forest Service with regard to the Forest Plan continue throughout the Plan's existence. The regulations require that the Forest Service monitor the plan's impact and, when necessary, revise the plan.9 Section 219.10(g) requires that forest plans be revised every ten years and also whenever the Forest Supervisor "determines that conditions or demands in the area covered by the plan have changed significantly or when changes in ... policies, goals, or objectives would have a significant effect on forest level programs." 36 C.F.R. § 219.10(g). One of the purposes of this constant oversight is to establish benchmarks in order to better assess the impact of specific actions upon the forest environment. Sierra Club is therefore entitled to challenge the Forest Service's compliance with the Plan as part of its site-specific challenge to the timber sales. See Wilderness Society v. Alcock, 83 F.3d 386, 390 (11th Cir.1996) (court will not hear challenge to Forest Plan until site-specific action is proposed). A contrary result would effectively make it impossible for a plaintiff to ever seek review of the Forest Service's compliance with a Forest Plan.

Furthermore, the Forest Service and intervenors' substantive argument--that 36 C.F.R. §§ 219.19 & 219.26 do not require the Forest Service to collect any population data--is inconsistent with the language of the regulations. Section 219.19(a)(6) states that "[p]opulation trends of the management indicator species will be monitored and relationships to habitat changes determined." It is implicit that population data must be collected before it can be monitored and its relationships determined. Likewise, § 219.26 requires that inventories of quantitative data be used when evaluating the effect of management alternatives on forest diversity. Before inventories can be evaluated, they have to be collected. Thus we find no merit to the Forest Service's contention that they have no obligation under § 219 to collect population data.

We do agree with the Forest Service that the combination of §§ 219.26 and 219.19 require it only to collect inventory data on MIS rather than on all species in the Forest. To read § 219.26 to require inventory data on all species obviates the need for MIS and reduces § 219.19 to nonsense. On the other hand, the Forest Service and Timber Intervenors' interpretation of § 219.26--that they need not collect data on MIS either--would consign that regulation to a similar fate. By their reading, § 219.26 would have no meaning despite its explicit requirement that quantitative inventory data be used to measure forest diversity. Interpreting a regulation in a manner that robs it of all meaning is unacceptable. Cf. Scott v. City of Hammond, Ind., 741 F.2d 992, 998 (7th Cir.1984) (strong presumption against agency interpretation that renders statute "wholly ineffective").

We believe that the regulations are harmonious when read together. MIS are proxies used to measure the effects of management strategies on Forest diversity; Section 219.19 requires that the Forest Service monitor their relationship to habitat changes. Section 219.26 requires the Forest Service to use quantitative inventory data to assess the Forest Plan's effects on diversity. If § 219.19 mandates that MIS serve as the means through which to measure the Forest Plan's impact on diversity and § 219.26 dictates that quantitative data be used to measure the Plan's impact on diversity, then, taken together, the two regulations require the Forest Service to gather quantitative data on MIS and use it to measure the impact of habitat changes on the Forest's diversity. To read the regulations otherwise would be to render one or the other meaningless as well as to disregard the regulations' directive that population trends of the MIS be monitored and that inventory data be gathered in order to monitor the effects of the Forest Plan. See Sierra Club v. Glickman, 974 F. Supp. 905, 936 (E.D.Tex.1997) ("The unambiguous language of the MIS regulations requires collection of population data.").10

Turning now to the instant case, it becomes clear that the Forest Service's approval of the timber sales without gathering and considering data on the MIS is arbitrary and capricious. The regulations require that MIS be monitored to determine the effects of habitat changes. The timber projects proposed for the Chattahoochee and Oconee National Forests amount to 2,000 acres of habitat change.11 Yet, despite this extensive habitat change and the fact that the some MIS populations in the Forest are actually declining,12 the Forest Service has no population data for half of the MIS in the Forest and thus cannot reliably gauge the impact of the timber projects on these species.

For the foregoing reasons, we reverse the district court's grant of summary judgment to the Forest Service and timber intervenors on the claims that the Forest Service acted arbitrarily and capriciously and violated NFMA, 16 U.S.C. § 1604, by approving the timber sales without gathering data on PETS species despite the directive of the Forest Plan. We also find that the agency's failure to gather inventory data on management indicator species violated 36 C.F.R. §§ 219.19 & 219.26 and we reverse the district court's grant of summary judgement on that claim as well.13 REVERSED AND REMANDED for further proceedings consistent herewith.