The Sierra Club brought suit against the United States Forest Service ("Forest Service") for alleged violations of the National Forest Management Act, 16 U.S.C. Secs. 1600 et seq. ("NFMA"), in the Forest Service's management of the Bighorn National Forest in Wyoming. The district court subsequently enjoined the Forest Service from offering most Bighorn forest land for timber harvest under its forest management plan. Sierra Club v. Cargill, 732 F. Supp. 1095, 1102 (D.Colo.1990). After altering the plan in response to the district court injunction, the Forest Service moved to dissolve the injunction. The district court denied the motion, and the Forest Service appeals. We exercise jurisdiction over this interlocutory appeal under 28 U.S.C. Sec. 1292(a)(1). We reverse the district court and remand with instructions to dissolve the injunction.

The National Forest Management Act requires that the Forest Service establish a "land and resource management plan" ("plan") for use of national forest land. The requirements of establishing and amending such a plan are contained in 36 C.F.R. Sec. 219. Section 219 requires that the Forest Service consider "a broad range of reasonable alternatives" for forest management. 36 C.F.R. Sec. 219.12(f). Specifically, the Forest Service must consider alternative management plans accounting for different multiple-use objectives so that a plan can be chosen which "comes nearest to maximizing net public benefits." Id.

These principles are evident in the regulations specifically addressing land use for timber harvesting. The regulations outline a three-stage analysis for evaluating the suitability of land for timber harvesting. Stage one involves a land suitability analysis under which land is unsuitable for harvest if: 1) it currently and historically has less than ten percent tree cover; 2) technology is not available to ensure that timber production will not cause irreversible damage to soil or watersheds; 3) it cannot be restocked within five years; and 4) it has been administratively withdrawn from timber production. 36 C.F.R. Sec. 219.14(a). Land that is suitable under stage one is then evaluated under stages two and three. Stage two requires an economic analysis to determine what the management costs and returns are for the different areas remaining after Stage one analysis and what timber production management intensity results in the greatest financial return for each area. 36 C.F.R. Sec. 219.14(b). Stage three then requires a more broad based economic analysis focusing on the value of timber harvest in an area in relation to the value of other "multiple-use" objectives for that area (including a broad range of uses such as recreation, wildlife habitat, watershed, and range land). 36 C.F.R. Sec. 219.14(c).

Section 219 requires a full multiple-use analysis in two situations: when the Forest Service first formulates a plan and when there is an amendment to an existing plan that would result in a "significant" change in the plan. 36 C.F.R. Sec. 219.10(f). "If the change resulting from [an] amendment is determined not to be significant for purposes of the planning process, the Forest Supervisor may implement the amendment following appropriate public notification and satisfactory completion of NEPA procedures." Id.

The Forest Service established a land and resource management plan for the Bighorn National Forest ("the Bighorn") which covered the harvesting of timber. The NFMA and its regulations require that timber harvest plans ensure that any area to be harvested can be regenerated within five years. 16 U.S.C. Sec. 1604(g)(3)(E); 36 C.F.R. Sec. 219.14(a)(3). The plan approved by the Forest Service for the Bighorn included a provision for harvesting lodgepole pine under a seven-year rather than a five-year regeneration standard.

The district court found such a provision invalid under the NFMA and enjoined the Forest Service from harvesting timber in the Bighorn pending a reevaluation of the plan using the five-year standard. Specifically, the injunction prohibited widespread harvesting in the Bighorn until "the Forest Service has first made a determination, based on research and experience, and pursuant to the implementing regulations and [the] three-stage analysis [outlined in Forest Service Regulations], that the land is suitable for harvest." Cargill, 732 F. Supp. at 1102.

In response to the district court's injunction, the Forest Service used the three-stage analysis outlined in 36 C.F.R. Sec. 219.14(a)-(c) in the context of an Environmental Assessment to analyze the effects of the change to the five-year regeneration standard only on the existing timber harvest plan. From this assessment the Forest Service concluded that no significant change in the plan resulted so that the five-year standard could simply be integrated into the existing plan without major amendment. The Forest Service never undertook a full Sec. 219 multiple-use reanalysis of the plan.

The Forest Service then moved to dissolve the district court's injunction. The district court denied the motion stating that it was "not satisfied that the Forest Service has made an adequate determination, based on research and experience, and pursuant to the implementing regulations and three-stage analysis, sufficient to permit the lifting of the injunction with respect to the offering of land for commercial timber production in the Bighorn National Forest...." Order Regarding Motion for Dissolution of Injunction, at 5 (D.Colo. Aug. 13, 1992). In evaluating a Forest Service motion for a stay pending appeal, this court ordered the district court to clarify the reasons behind its refusal to lift the injunction. In response the district court said further:

[T]he Forest Service has failed to comply with 36 C.F.R. 219.14(c), [stage 3 of the analysis], ... requir[ing] that several alternative plans be developed and evaluated in accordance with the regulations and guidelines set forth.... [T]he Forest Service only evaluated the "preferred alternative in the plan." Furthermore, the Forest Service failed to adequately analyze and consider each of the factors outlined in 36 C.F.R. 219.14(c)(1)-(3) and Stage II costs and benefits are ignored. The Forest Service also fails to include a justification to log areas which can only be harvested at a loss. In addition, the Forest Service uses an unrealistic timber production goal in determining the suitable timber base.

Clarification of Order Regarding Motion for Dissolution of Injunction, at 1-2 (D.Colo. October 23, 1992).II. District Court's Refusal to Dissolve the Injunction

We review a district court's ruling as to whether to dissolve an injunction for an abuse of discretion. Securities & Exch. Comm'n v. Blinder, Robinson & Co., 855 F.2d 677, 679 (10th Cir.1988), cert. denied, 489 U.S. 1033, 109 S. Ct. 1172, 103 L. Ed. 2d 230 (1989). The district court abused its discretion in this case because it applied the wrong standard of review and, as a result, the wrong analytical framework in making its determination as to whether to dissolve the injunction.

An initial matter is what the Forest Service was required to do under the terms of the injunction (which the Forest Service did not appeal). In this case the district court could only order the Forest Service to do what is required under valid Forest Service regulations. See Wilderness Soc'y v. Tyrrel, 918 F.2d 813, 818 (9th Cir.1990) (pointing out that "we are not free to impose our own notions of procedural propriety upon the Forest Service"). The question faced by the district court with regard to the injunction was whether the Forest Service acted properly under 36 C.F.R. Sec. 219.

In its Environmental Assessment, the Forest Service determined that no "significant" change in the Bighorn plan was implicated in changing from the seven-year to the five-year regeneration standard. As outlined above, where no significant change is implicated, a full Sec. 219 reanalysis, including multiple-use considerations, is not required. 36 C.F.R. Sec. 219.10(f). However, in deciding on whether to dissolve the injunction, it is apparent that the district court reviewed the actions of the Forest Service by the standards of the full Sec. 219 reanalysis framework. See Clarification of Order Regarding Motion For Dissolution of Injunction, at 2. (criticizing the Forest Service's actions primarily because it only considered the effect on the "preferred alternative in the plan" and did not do a full multiple use review). As a result, if the Forest Service's determination that no significant change occurred is valid, the framework under which the district court reviewed the matter is largely irrelevant.

The threshold question, then, is whether the Forest Service erred in treating the move from a seven- to a five-year regeneration standard as a "non-significant" change. The regulations provide little guidance as to when a change is significant, directing only that the determination should be "[b]ased on an analysis of the objectives, guidelines, and other contents of the forest plan." 36 C.F.R. Sec. 219.10(f). However, as a general matter we review such an agency determination only as to whether it is arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law. 5 U.S.C. Sec. 706(2)(A); see Wilson v. Hodel, 758 F.2d 1369, 1372 (10th Cir.1985); see also Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 359, 109 S. Ct. 1835, 1850, 104 L. Ed. 2d 351 (1989) (applying a "plainly erroneous or inconsistent with the regulation" standard to an agency construction of its own regulation); Village of Los Ranchos De Albuquerque v. Marsh, 956 F.2d 970, 972-73 (10th Cir.) (applying arbitrary and capricious review to an agency determination that a change in a project results in no "significant" change requiring a full environmental impact statement under NEPA), cert. denied, --- U.S. ----, 113 S. Ct. 59, 121 L. Ed. 2d 27 (1992). Further, in this case the regulation expressly commends the determination of the significance of an amendment to the Forest Supervisor's judgment. 36 C.F.R. Sec. 219.10(f). Applying the deferential standard is especially important where, as here, the agency determination is extremely fact bound. See Los Ranchos, 956 F.2d at 972.

The district court clearly did not apply this standard because it simply ignored the Forest Service's determination that the change in question is not a significant one. At step one the court reviewed the actions as if the change were significant. See Clarification of Order Regarding Motion for Dissolution of Injunction, at 2; Order Regarding Motion for Dissolution of Injunction, at 5.

Applying the proper deferential standard, this Court cannot say, nor could the district court appropriately say, that the Forest Service acted arbitrarily and capriciously or abused its discretion in treating the move from a seven-year to a five-year regeneration standard as a "non-significant" change. The Forest Service completed an Environmental Assessment reconsidering the Bighorn timber plan. The assessment's conclusion that there was no significant change was based primarily on its finding that a relatively small number of acres (4,377 acres) are affected by the change and that the remaining acreage is largely the same in character and location as that originally included in the timber base. See Environmental Assessment; see generally Forest Service Handbook 1909.12, ch. 5.31 (directing Forest Supervisors to apply several factors in determining significance including: 1) timing; 2) location and size of area involved; 3) the degree to which goals, objectives, and outputs are affected; and 4) whether management prescriptions are modified).

We address specifically several objections by the Sierra Club relevant to this finding by the Forest Service. First, the Sierra Club contends that the Forest Service erred in using an incorrect basic assumption in evaluating the effects of the five-year standard on the plan. The "Allowable Sale Quantity" ("ASQ")1 for an area is an important variable in determining the size of the suitable timber base, and there is no dispute that the ASQ under the original Bighorn plan has proven to be unrealistic. See Environmental Assessment, at 6. The Sierra Club argues, and the district court seems to agree, that failure to adjust the ASQ in the plan reanalysis renders this reanalysis invalid. See Clarification of Order Regarding Motion for Dissolution of Injunction (including as part of its reasoning the fact that "the Forest Service uses an unrealistic timber production goal in determining the suitable timber base"). This proposition is untenable in the context of this case.

The Forest Service has indeed made the determination that a change in the ASQ of the Bighorn plan is necessary. Id. The Forest Service also apparently contemplates that this change in the plan will be "significant." As a result, the Forest Service is in the process of doing a reanalysis of the plan based on a more realistic ASQ. See id. The Forest Service is not required to cease all "non-significant" adjustments in the Bighorn plan, such as an otherwise "non-significant" change in the regeneration standard, based only on the fact that a significant change is on the horizon. Such a rule would improperly tie the hands of the Forest Service in plan management and thereby thwart the purpose of the regulations.

The Sierra Club also suggests that, because the Forest Service used some updated data in arriving at the 4,377 acre number, it should be required to use an updated ASQ as well. The regulations, however, impose no general requirement that the Forest Service adjust a management plan to account for all new data when a subset of new data becomes available. Indeed the regulations specifically give the Forest Service authority to make "non-significant" management decisions which necessarily includes incorporating new data into such decisions where appropriate. See 36 C.F.R. Sec. 219.10(f).

In this case, in determining that use of the five-year regeneration standard resulted in only a 4,377 acre net decrease in acres available for timber harvest, the Forest Service used what it considers more accurate data and review tools than were used in the original plan evaluation. Specifically, there was "(1) a change in data bases, (2) [use of] more precise mapping capability, and (3) [use of] additional regeneration criteria." Environmental Assessment, at 10. This new data turned an apparent decrease of 74,022 acres under the new five-year standard (after Stage one of the Sec. 219.14 analysis) into a decrease of only 4,377 acres (after reanalysis using the new data under stages two and three). In the context of this case, we do not find, nor could the district court appropriately find, that the use of this data to compute acreage without incorporating a new ASQ was arbitrary and capricious or an abuse of discretion.

The Sierra Club also contends that the Forest Service acted improperly in allowing for the harvesting of some timber at a monetary loss under the plan. The district court in part relied on this factor in refusing to dissolve the injunction. See Clarification of Order Regarding Motion for Dissolution of Injunction, at 2 ("Stage II costs and benefits are ignored. The Forest Service also fails to include a justification to log areas which can only be harvested at a loss"). However, the inclusion of some "below cost" timber in the timber base does not by itself constitute an abuse of discretion. The Sierra Club concedes that the Forest Service may allow for the cutting of below cost timber under the regulations. Brief for Appellee, at 31-32 (citing Big Hole Ranchers Ass'n v. United States Forest Serv., 686 F. Supp. 256 (D.Mont.1988)); see also Thomas v. Peterson, 753 F.2d 754 (9th Cir.1985).

In this case the original Bighorn plan itself provided for the inclusion of largely the same below-cost land in the timber base as is included under the reanalysis. It is important to remember that we are reviewing whether there is a significant change in the plan and not the entirety of the original plan. Further, the Forest Service obviously considered pure cost to some extent in the reanalysis as it excluded from the timber base very steep timber land where very high cost "cable" logging is required. See Environmental Assessment, at 87. As a result, as matter of law we cannot say that the Forest Service acted arbitrarily and capriciously or abused its discretion where it allowed for below-cost harvesting under its Bighorn plan reanalysis.

The district court abused its discretion in not lifting the injunction on timber harvesting in the Bighorn National Forest because it did not apply the requisite deferential standard of review to the Forest Service's decision to treat as "non-significant" the change in the Bighorn plan from a seven-year to a five-year regeneration standard. As a result, the district court applied the wrong analytical framework in the case. Applying the correct standard of review, we find that the decision of the Forest Service to treat the plan change as "non-significant" is not arbitrary and capricious and is within its discretion and that therefore the Forest Service's actions satisfy the requirements of the NFMA and the Sec. 219 regulations. We REVERSE the district court and REMAND the case with instructions to dissolve the injunction.

SEYMOUR, Circuit Judge, dissenting.

I regret that I am unable to join the majority opinion. The majority hinges its decision to reverse upon its assessment of the legality of the original injunction. See Maj. op. at 1548. However, the Forest Service did not appeal that injunction and its terms are therefore not subject to review by this court. We are instead called upon to ascertain only whether the district court abused its discretion in concluding that the Forest Service did not properly perform its obligations under the injunction. The majority upholds the Forest Service's actions by engaging in judicial bootstrapping that I view as a disturbing abdication of the court's obligation to conduct a meaningful review of agency action. I therefore respectfully dissent.

In attempting to comply with the injunction, the Forest Service made changes by way of a plan amendment, "using updated data and procedures to evaluate land suitability." Aplt.Br. at 13. In Stage I, "[t]he suitability reanalysis used a new resource database (RIS) that has been updated since the 1985 Forest Plan was approved and contains site specific data. This analysis also used information from a new timber inventory completed in 1985 which is reflected in the new data base." Aplt.App. at 89. Thus, the reanalysis contained "more precise information about land characteristics, and more precise and scientifically based criteria," and was based on "scientific knowledge and research, professional experience and ... existing, readily available technology." Id. at 70. The new suitability analysis revealed that "74,022 more acres were removed from the tentatively suitable acreage base in the Amended Plan than had been removed in the 1985 Plan." Aplt.Br. at 18. This change was the result of the use of the new database, a change in methodology for computing industrial wood production, and new regeneration criteria. Id. at 19-20.

Although the Stage I analysis resulted in 74,022 acres being removed from the tentatively suitable land base, use of new data in Stages II and III resulted in a difference from the original three-stage analysis of only 4,377 acres. In Stage II, "[a]ll costs and timber prices were updated based on the latest information available." Aplt.App. at 83.

The Stage II analysis revealed that overestimates of planting needs in the original Plan made revision of the cost estimates for regeneration under the strict 5-year standard unnecessary. No changes were made in the analysis for Stage III, except that use of the new database showed that only 8,832 acres subject to a primitive recreation management prescription in the Plan were actually located on lands included within the tentatively suitable timber base, in contrast to 62,100 acres excluded by this prescription in 1985.

Aplt.Reply Br. at 6. In this way, 53,268 acres of the 74,022 acres lost in stage I were replaced. Another 16,376 acre difference came under the category of "economic efficiency (logging methods)." Aplt.App. at 89. "The lack of a site specific data base and recreation opportunities map led to a miscalculation in the original analysis." Id. at 90. These differences combined to replace 69,644 acres of the 74,021 acres that had been removed in stage I. Thus, the new analysis differed from the original by only 4,377 acres.

Independent of this litigation, "the Forest Service began monitoring harvest levels under the Bighorn Plan." Aplt.Br. at 9 (footnote omitted). Based on the monitoring data, the Forest Service determined that the forest will actually fall far short of the projected allowable sale quantity. "[M]onitoring reports showed production significantly below the average annual allowable sale quantity for several consecutive years, signalling an imbalance in the Plan's projections." Id. "The fact that the forest has produced substantially less timber than anticipated in the Forest Plan shows that the anticipated balance between 'standards and guidelines' and timber production cannot be maintained without amending the plan." Id. at 29. In fact, the imbalance was so severe that the Forest Service referred to the ASQ and the plan objectives as "incompatible." Aplee. Supp.App. at S-1 (Draft Environmental Impact Statement). "The Forest Supervisor determined that the incompatibility was significant and that it would be necessary to amend the Forest Plan." Id. The Forest Service concedes that changing the ASQ will result in a significant amendment to the plan. Aplt.Br. at 9-10. As the majority recognizes, when an amendment is significant the Forest Service is required to conduct the same complex planning process applicable to promulgation of the original plan, see U.S.C. Sec. 1604(f)(4), including an entirely new three-stage analysis and the assessment of multiple alternatives.

The decision to alter the ASQ was made contemporaneously with the reformulation of the three-stage analysis performed to satisfy the injunction in this case. Nevertheless, the new data regarding actual timber production was not incorporated in the three-stage analysis. The Forest Service offered the following explanation in its Environmental Assessment:

The scope of this proposed action does not include reanalysis of the Forest's 'allowable sale quantity' or timber management program portion of the Forest Plan. Thus the tentatively suitable land base is identical for all forest plan alternatives. The tentatively suitable land base is independent of the multiple-use management objectives that may be changed as a result of the ASQ reanalysis. While the stage III suitability analysis is based on multiple use objectives, the court did not order the Forest Service to change the multiple-use objectives in the preferred alternative. Therefore, the stage III suitability analysis can be completed for the existing preferred alternative now, and the stage III analysis can be modified later as new alternatives are developed as part of the ASQ reanalysis.

Aplt.App. at 61 (emphasis added). The ASQ reanalysis was projected to be completed in August 1993. Aplt.Br. at 10.

As the Forest Service and the majority concede, when adopting a Forest Plan the Forest Service is required to consider a number of alternatives. Each alternative plan contains different management objectives, which determine how much land to set aside for each of the multiple uses: wildlife protection, timber production, recreation, etc. Each alternative is then reviewed to determine which alternative "comes nearest to maximizing net public benefits." 36 C.F.R. Sec. 219.12(f).

When the Forest Service conducts an initial suitability analysis in the formulation of a Forest Plan, the regulations require consideration of a "broad range of reasonable alternatives" (36 C.F.R. 219.14(c)). Alternatives must "reflect a range of resource outputs and expenditures" (36 C.F.R. 219.12(f)). Each alternative therefore sets out a unique combination of products and services. In the management prescriptions for each alternative, the agency defines management prescriptions for the various multiple uses, including outdoor recreation, timber, watershed, range, wildlife and fish, and wilderness. The timber production objective of each alternative determines how many of the "tentatively suited acres" found in the Stage I analysis will become "suited acres" in stage III.

Aplt.Br. at 23 (emphasis added). In the reanalysis done here, however, the Forest Service analyzed the land under one alternative only--the existing alternative with the timber production goal of 149.5 million board feet. The Forest Service's justification for this procedure was that elimination of the seven-year regeneration standard resulted in only a small change in acreage suitable for timber production, therefore resulting in an insignificant amendment to the Bighorn Plan. Id. The Forest Service was able to reach this conclusion only by ignoring the significant amendment that it knew was required in order to redetermine a more accurate ASQ.

In the original injunction, the district court set aside the regeneration provisions and remanded the Bighorn Forest Plan for revision. It is apparent from the district court's injunction order that it was concerned about the plan as a whole and the Forest Service's blatant refusal to follow the statutory mandate regarding regeneration. The court specifically enjoined the Forest Service from permitting any timber production in the Bighorn National Forest until the Forest Service first determined, "based on research and experience, and pursuant to the implementing regulations and three-stage analysis discussed above, that the land is suitable for harvest." 732 F. Supp. at 1103 (emphasis added). I emphasize again that the Forest Service did not appeal this injunction. It was therefore required to use its research and experience and the three-stage analysis to reassess the plan.

The Forest Service necessarily used some new data acquired through its research and experience in order to ensure that the amount of land found suitable for harvest was not significantly changed even though the restocking requirement was changed from seven years to five years. Since the resulting change was insignificant, the Forest Service asserts that it was not required to consider multiple alternatives and to do a completely new Stage III reanalysis. At the same time, however, the Forest Service chose to ignore additional new data regarding the ASQ, acquired through its research and experience in the same time frame, where such data admittedly mandated a significant amendment to the plan.1 As the Forest Service itself recognizes, "[c]alculation of the allowable sale quantity is an important, concurrent issue associated with management of the Bighorn National Forest at this time." Aplt.App. at 61 (emphasis added). Furthermore, the new ASQ analysis admittedly will alter the Stage III suitability analysis. Id. Significantly, neither of the preferred alternatives listed in the Draft Environmental Impact Statement prepared in conjunction with the ASQ reanalysis maintains the level of timber production previously authorized. Of four alternatives, the Forest Service prefers Alternatives A and C. Aplee. Supp.App. at S-4. Alternative A decreases the amount of timber available for harvest from the current level of 149.5 million board feet to 58.0 million board feet. Id. Alternative C decreases the amount of available timber to 114.5 million board feet. Id.

The Forest Service thus chose to incorporate "more precise information about land characteristics ... based on scientific knowledge and research," Aplt.App. at 70, where such information supported the original plan objectives and enabled the Forest Service to avoid a significant amendment to the plan, and to ignore such information where it did not. This was done in the face of the district court's injunction requiring the Forest Service to revise the plan and to use its "research and experience" in determining whether the land is suitable for timber production. 732 F. Supp. at 1102. Under these circumstances, I cannot agree that the district court was "arbitrary, capricious, or whimsical," when it refused to let the Forest Service pick and choose which new data to use in doing the reanalysis mandated by the injunction.

The majority recognizes that we do not uphold agency action that is arbitrary, capricious, or an abuse of discretion. Nonetheless, the majority approves the Forest Service's selective incorporation of new data by using a circular argument that begs the relevant question. The majority thus concludes that the Forest Service need not use the new ASQ data because the regeneration amendment is "otherwise 'non-significant.' " Op. at 1549. However, the only reason the amendment is otherwise non-significant is because the Forest Service did not incorporate the new information it had regarding the totally inaccurate ASQ resulting from the prior Stage III analysis.

The majority reasons in a circle rather than addressing the dispositive issue: was the Forest Service arbitrary and capricious in its refusal to incorporate new information regarding significantly decreased production of timber when amending the Plan to comply with the injunction. As the majority states, the Forest Service "used what it considers more accurate data and review tools than were used in the original plan evaluation." Op. at 1549. At the same time, the Forest Service did not incorporate the new information about reduced timber production even though it recognized that harvesting timber at the present ASQ would result in an inability to meet the plan objectives regarding wildlife protection, recreation, and other goals of a proper forest plan. Nothing in the rationale supporting the majority's result convinces me that the Forest Service did not act arbitrarily and capriciously in choosing selectively to use some new information and to exclude other information of great import. Accordingly, I would affirm the district court.