Wetsel-Oviatt Lumber Co. v. United States

OPINION MARGOLIS, Judge. This is a government contract action and pre-award bid protest challenging the government’s cancellation of the proposed “Bald Mountain” timber sale. A four-day trial was held in Sacramento, California. Based upon the evidence presented at trial, as well as the parties’ pre-and post-trial briefs, the Court holds that the United States Forest Service did not have a rational basis for canceling the proposed Bald Mountain sale. BACKGROUND The National Forest Management Act (“NFMA”), 16 U.S.C. § 1601 et seq., prescribes the statutory framework for management of National Forest System lands. NFMA establishes a scheme of Land and Resource Management Plans (“LRMPs”) for units of the National Forest System, and generally requires that each forest have its — 558 — own LRMP. See 16 U.S.C. § 1604(a); 36 C.F.R. § 219. LRMPs are required to provide for “multiple use and sustained yield of goods and services from the National Forest System in a way that maximizes long term net public benefits in an environmentally sound manner.” 36 C.F.R. § 219.1 In this regard, pursuant to NFMA, LRMPs developed for national forest land are required to “provide for diversity of plant and animal communities based on the suitability and capability of specific land area in order to meet overall multiple-use objectives.” 16 U.S.C. § 1604(g)(3)(B). Further, in conjunction with the NFMA and LRMPs for National Forest lands, the Forest Service has, since the late 1970s, maintained a “sensitive species” program designed to develop and implement management practices to ensure that species do not become listed as threatened or endangered under the Endangered Species Act’ (“ESA”) because of Forest Service actions. Pursuant to the NFMA, and its predecessor statutes, including the Multiple Use Sustained Yield Act of 1960 (“MUSYA”), 16 U.S.C. §§ 528-531, the United States Forest Service has for several decades offered timber sales to the public. Timber may be harvested from National Forest System lands only if the sale would not violate the National Environmental Policy Act (“NEPA”), NFMA, or otherwise irreversibly damage watershed conditions on the forest land. See 42 U.S.C. § 4321 et seq., 16 U.S.C. § 1604 et seq.; see also 16 U.S.C. § 1604(g)(3). Pursuant to NEPA, if a timber sale, or other federal action, “significantly affect[s] the quality of the human environment” a detailed Environmental Impact Statement, among other documents, must be prepared. See 42 U.S.C. § 4332(C); 40 C.F.R. § 1502.3. To evaluate whether a timber sale or other federal action is significant under NEPA, the Forest Service prepares an Environmental Assessment (“EA”). See 40 C.F.R. §§ 1501.4,1508.9; see also Transcript (“Tr.”) at 590. If an action will not have a significant impact on the environment, the Forest Service issues a Finding of No Significant Impact (“FONSI”). A FONSI is “a document by a Federal agency briefly presenting the reasons why an action, not otherwise excluded [ ], will not have a significant effect on the human environment and for which an environmental impact statement therefore will not be prepared.” 40 C.F.R. § 1508.13. FACTS Issued January 6, 1989, the Eldorado National Forest LRMP was designed to direct management of the Eldorado National Forest. The LRMP became effective February 6, 1989. An EA for the proposed Bald Mountain timber sale was completed and a Decision Notice (“DN”) and FONSI were signed on September 22,1983. Thereafter, a Wildlife Biological Evaluation (“Wildlife BE”), sensitive plant BE, and Cumulative Effects Analysis (“CEA”) for the California Spotted Owl were completed on August 11, August 31, and September 20, 1992, respectively. In the 1992 Wildlife BE, Forest Service wildlife biologists analyzed the potential effect of the Bald Mountain timber sale upon various wildlife including the California Spotted Owl and — to a lesser extent — the Pacific Fisher. Further, the Forest Service determined the composition of each harvest unit in the Bald Mountain timber sale by identifying units by “serai stage.” A “serai stage” refers to an age class or period of growth of a stand of trees. Serai stages are identified based on at least two criteria — tree size class, and the percentage of cover provided by the tree canopy. Serai stages on the Eldorado National Forest are identified as — 559 — “timber strata ”2 according to the following code system: Attribute Code Used Identification Tree Size Class 2 less than 12 inches mean diameter at breast height (dbh) 3 12-23.9 inches dbh 4 greater than or equal to 24 inches dbh ■ Canopy Closure S or A sparse — 10-19% cover Por A light — 20-39% cover NorB medium — 40-09% cover GorC heavy — 70-100% cover Tree size class and canopy closure are not generally applied when referring to individual trees, but rather are used to type a relatively homogenous group of trees as “timber strata” composing a “forest stand.” The 1992 Wildlife BE advised that, “all but one of the proposed units [on the Bald Mountain timber sale] have less than 50 percent canopy closure.” Further, the 1992 CEA for the California Spotted Owl provided: The Bald Mountain Timber Sale will tractor log 421 acres of unsuitable habitat within its sale boundary. Harvesting of Unit 18, containing 7 acres of suitable forage habitat, will be deferred. The proposed sale will not directly affect the quantity and quality of California spotted owl habitat for the following reasons. First, harvest of live trees will only occur in unsuitable habitat. Second, activities resulting from implementation of the sale, such as road construction, reconstruction and proposed post-sale treatments will not remove any suitable habitat____ All proposed sale units were field verified for accuracy of the photo typing. Table 4 illustrates the results of the ground verification for each stand. Plaintiffs exhibit (“PE”) at 0701 (table 4 omitted). Advertising and Bidding on the Bald Mountain Timber Sale Located in the Georgetown Ranger District of the Eldorado National Forest in California, which is within Region Five of the National Forest System, the Bald Mountain timber sale was to consist of 17 cutting units numbered 19-32 and 34-36 in four sale subdivisions. The sale consisted of an estimated 4,380 M (thousand) board feet of timber marked for cutting, and was planned and offered to the public pursuant to the NFMA, MUSYA, and the strictures of NEPA. Advertisement for bids on the Bald Mountain timber sale advised prospective bidders that “[t]he right to reject any and all bids is reserved.” The Bald Mountain timber sale prospectus further indicated that the sale would be conducted by oral auction and initial qualifying written bids were to be submitted on prepared forms by 2:00 p.m. on November 2, 1992. Bidders who submitted qualifying written bids were entitled to participate in a subsequent oral auction. Plaintiff Wetsel-Oviatt Lumber Company, Inc. (“Wetsel”) and two other bidders submitted valid written qualifying bids prior to the bid opening time of 2:00 p.m. on November 2,1992. All three bidders participated in the oral auction and Wetsel ultimately prevailed as the high bidder with an overall bid of $2,294,250. Forest Service officials advised Wetsel that it was the high bidder and Cecil L. Wetsel, Jr., President of Wetsel-Oviatt Lumber Co., Inc., signed a confirmation of the bid. Wetsel’s bid was responsive and Wetsel was, at the time of the bid, a responsible and qualified bidder. Notwithstanding Wetsel’s high bid, the Forest Service neither awarded the Bald Mountain timber sale to Wetsel, nor rejected Wetsel’s bid, on November 2,1992. Pressure to Reconsider the Sale In the Fall of 1992, individuals and environmental groups including Forest Alert, Friends Aware of Wildlife Needs (“FAWN”), Craig Thomas, Erin Noel, John Tecklin, Karen Schamback, and Linda Blum informed Forest Supervisor John Phipps, who was then stationed on the Eldorado National Forest, that they believed the Forest Service was required by federal law to reconsider the environmental documentation for green timber sales within the Eldorado National Forest to determine whether the sales were in compliance with NEPA, NFMA, Forest Service regulations, and the Eldorado forest — 560 — plan. On November 30, 1992, Phipps met with Noel and Tecklin. At that meeting, Phipps indicated that he would have his staff review the Bald Mountain timber sale to determine if there was new information or changed circumstances that had not been considered in the existing environmental analyses. Such environmental reconsideration of timber sales on the Eldorado National Forest produced a “highly charged” atmosphere that resulted in internal conflict among members of the Forest Service staff. According to Phipps, the Forest Service staff had strong views and split into two camps, one of which believed that timber sales should proceed, and the other of which believed that timber sales should be canceled. Environmental Reconsideration In March 1993, Phipps decided to withhold award of the Bald Mountain timber sale to Wetsel while further environmental review of the sale was being completed. Also in March 1993, Phipps convened an interdisciplinary team to review the existing EAs for certain green timber sales, including the proposed Bald Mountain timber sale, and determine whether the EAs should be revised. The team was given 30 days to report back to Phipps. Within the 30 days allotted, the interdisciplinary team presented Phipps with information which the team and Phipps believed indicated that the documentation for 24 awarded timber sales needed to be revised pursuant to NEPA, and that further analysis was necessary on the environmental documentation for the Bald Mountain timber sale. Consequently, Phipps decided to form a longer term interdisciplinary team to evaluate the effects of the timber sales on wildlife, vegetation, and water quality. Phipps directed that the team complete the appropriate analyses and revise the EAs as soon as practicable. The California Spotted Owl On November 24, 1993, the Forest Service issued its Biological Evaluation of the Cumulative Effects of the Eldorado National Forest Timber Sales Upon the California Spotted Owl (“1993 Owl BE”), prepared by wildlife biologist for the Eldorado National Forest Dawn Lipton. The 1993 Owl BE evaluated the effects that harvesting 31 Eldorado National Forest timber sales, including the Bald Mountain sale, would have upon the California Spotted Owl. In completing her assessment, Lipton considered information provided in a document released by the Forest Service in mid-1992 entitled The California Spotted Owl: A Technical Assessment of Its Current Status (“CASPO report”). To analyze the effects of the Eldorado timber sales upon the California Spotted Owl, Lipton and other Forest Service biologists developed and applied the “disturbance index” methodology. Conceived to objectify the effect that removal of timber on the Eldorado National Forest would have upon the California Spotted Owl, Lipton explained in the 1993 Owl BE that: [t]he disturbance index provides a simple measure of the change in timber strata surrounding a spotted owl activity center, in relation to the amount of each timber strata currently available and the value of the timber strata as owl habitat. It is particularly sensitive to the amount of 4G, 4N, and 3G strata affected in relation to the amount available surrounding an owl activity center. The higher the disturbance index the greater the risk of not maintaining habitat or habitat components essential for future management options, and the greater the risk of not maintaining the current distribution of owl sites. There is not, however, a clear threshold that can be related to this index. * * * * Although there is no data to support a threshold, it seems reasonable, assuming protected activity centers are adequately maintained, to conclude that sites with less than five percent disturbance index are not likely to be substantially affected by surrounding harvest activity, nor are future management options likely to be foreclosed. Joint Stipulation of Facts at IN 28-29. In effect, the disturbance index is a fraction. The denominator of that fraction was — 561 — determined by identifying the different types of timber strata within one-mile and one-mile to two-mile circles surrounding an owl “protected activity center” (“PAC”). To determine the timber strata within these circles, the 1993 Owl BE relied upon a 1991 Forest Service vegetation inventory. Typing for the entire circles, including for areas of the Bald Mountain timber sale, was determined based on the 1991 Forest Inventory. Timber strata within the circles were then weighted for both value to, and use by, the California Spotted Owl. By contrast, the numerator of the disturbance index fraction represents timber strata occurring within harvest units within one-mile and one-mile to two-mile circles of a particular owl PAC.3 Unlike the denominator, the 1993 Owl BE relied on aerial photographic timber strata typing conducted by JoAnn Fites as part of the Forest Service’s Landscape Analysis. However, like the denominator, timber strata in the numerator was weighted for value to the owl. Thus, the numerator represents the removal of owl habitat within a two-mile circle, and the two-mile circle data in the denominator is the entire area considered. The disturbance index, therefore, is a ratio of the habitat removed, weighted for the value of that habitat to the owl, to the habitat available, weighted for value. In this regard, the 1993 Owl BE notes that the index: is particularly sensitive to the amount of 4G, 4N, and 3G strata affected in relation to the amount available surrounding an owl activity center. The higher the disturbance index the greater the risk of not maintaining habitat or habitat components essential for future management options, and the greater the risk of not maintaining the current distribution of owl sites. PE 1 at 11. For the area constituting the Bald Mountain timber sale, the 1993 Owl BE concluded that harvest of the timber sale would affect two owl sites, ED098 and ED072. For ED098, the 1993 Owl BE noted a disturbance index value of 11.01%. For ED072, it noted a disturbance index value of 5.22%. Thus, the Bald Mountain timber sale, according to the 1993 Owl BE, “substantially affected” these two owl sites because the disturbance index value was calculated above the five percent threshold established as appropriate by the biologists who prepared the 1993 Owl BE. At the same time — in calculating the disturbance index for owl sites within the area of the Bald Mountain timber sale — the 1993 Owl BE recognized that the timber strata data that formed the basis for the disturbance index value ultimately derived was inaccurate. Further, Forest Supervisor Phipps was aware that the use of two different data sets in the numerator and denominator of the disturbance index fraction would create a bias that would weigh in favor of canceling the Bald Mountain sale, although Phipps had no idea as to the degree of bias. The Pacific Fisher Further, on November 8,1993, Forest Service biologist Dawn Lipton and Forest Service ecologist JoAnn Fites4 issued a draft Evaluation of Habitat Requirements for Fisher5 on the Eldorado National Forest (“November 8, 1993 Fisher BE”). A revised draft was subsequently released sixteen days later on November 24, 1993 (“November 24, 1993 Fisher BE”),6 which also included For — 562 — est Service biologist Helen Bombay among its drafters. Both draft 1993 Fisher BEs evaluated the potential effects of sold and unsold timber sales, including the Bald Mountain timber sale, on potential Fisher habitat. The draft 1993 Fisher BEs also utilized A Literature Review for Management of Marten and Fisher on National Forest in California, by Martin Freel, dated July 1991, as a source of parameters. Significant difference, however, existed in the conclusions of the BEs with regard to the effect of the Bald Mountain timber sale upon the Pacific Fisher. In the November 8,1993 Fisher BE, Fites and Lipton conclude that: [u]se area H currently contains little mature forest habitat and is unlikely to provide for resident fisher use. The area may be important as habitat in the future and for future management options but is of low value to the species at present. The effects of the Bald Mountain timber sale on this Use Area are not likely to have much affect on the fisher. PE 28 at BM-00212. By contrast, 16 days later, the November 24, 1993 Fisher BE concluded that “[u]se area H currently contains little mature forest habitat and is unlikely to provide for resident fisher use. Harvest of the Bald Mountain Timber Sale will further reduce the amount of large tree, dense canopied forest available and will increase existing fragmentation.” PE 11 at 19-20. Located within “potential fisher use area H,” 7 the Bald Mountain timber sale occupies 421 acres, or 3.17%, of potential use area H, which includes 13,254 acres of national forest land. None of the studies or summaries of studies relied upon by the Forest Service indicate that the Pacific Fisher has lived on, or actually used, potential use area H in recent history.8 Addressing the potential effects of the proposed Bald Mountain timber sale, the November 24, 1993 Fisher BE stated: The Bald Mountain Timber Sale will affect 421 acres within [] potential fisher use area [¶] ... [and] will remove 149 acres of 4G/N timber strata ... Use area H currently contains little mature forest habitat and is unlikely to provide for resident fisher use. Harvest of the Bald Mountain Timber Sale will further reduce the amount of large tree, dense canopied forest available and will increase existing fragmentation. The area may be important as habitat in the future and for future management options but is of low value to the species at present. If Use Area H is not maintained, the distance between Use area F/G and I exceeds 8 miles, however, potentially precluding effective dispersal and use of habitat based on information in Freel (1991)____ Based on this assessment, harvest of the Bald Mountain Timber Sale is likely to adversely affect the fisher. It may, therefore, contribute to a trend toward Federal listing or loss of viability of the species. PE 11 at 19-20. To determine timber strata within potential use area H, both draft 1993 Fisher BEs relied on the Eldorado National Forest’s 1991 Vegetation Inventory, which was acknowledged to be only 63% accurate. Information contained in the draft 1993 Fisher BEs was used in compiling a final version Fisher BE, issued January 20,1994 by Helen Bombay and Dawn Lipton,9 entitled Biological Evaluation of Habitat Requirements for Fisher on the Eldorado National Forest in Relation to Twenty-Four Sold Timber Sales (“ 1994 Fisher BE”). Unlike the 1993 Fisher BEs, the 1994 Fisher BE did not specifically address the effects of the proposed Bald Mountain timber sale upon the Pacific Fisher. — 563 — The Landscape Analysis and lpG Timber Strata Also pursuant to Phipps’ direction, Forest Service biologists prepared biological evaluations of the potential effects of the sold and unsold timber sales, including the Bald Mountain timber sale, on forest strata.10 The forest biological evaluation, entitled Landscape Analysis for Eldorado National Forest, was prepared by Fites and Lipton, and issued on August 3, 1993 (“1993 Landscape Analysis”). Revision and update of the 1993 Landscape Analysis, entitled Landscape Analysis — Late Serai Forests, was prepared by Fites, Lipton, Bruce Bingham, and Susan Durham, and issued January 21,1994. The Landscape analyses were particularly important to determining the environmental effects of the Bald Mountain timber sale. Not only did the Landscape analyses address the potential effects of the Bald Mountain timber sale upon timber strata, but — as noted — the Landscape analyses’ classification of forest strata also formed the foundation for the findings and conclusions of Forest Service biologists and employees in the Owl BE and the Fisher BEs. Indeed, as indicated previously, assessment of the potential effects of the Bald Mountain timber sale upon the California Spotted Owl and Pacific Fisher depended upon, and were related to, the timber strata in the area of the Bald Mountain timber sale because both the California Spotted Owl and the Pacific Fisher tend to prefer certain timber strata. According to Management Practice 56 of the Forest-Wide Standards and Guidelines set forth in the Eldorado National Forest Land Management Plan: [t]he intent of plant diversity management [would] be the support of populations of wildlife and plant species native to the area of management, maintaining approximately the current distribution and frequency of forest types and enhancing visual quality. Each serai stage shall have at least a five percent representation within the area of management. With regard to the Bald Mountain timber sale, the relevant area of management is Combined Watershed 13. Thus, to determine the effects of the Bald Mountain timber harvest, Fites attempted to ascertain the effect that removal of timber from the cutting units of the proposed Bald Mountain sale would have upon timber strata in Combined Watershed 13. To determine the amount of 4G/N timber strata within the cutting units of the proposed Bald Mountain timber sale, Fites relied on aerial photographic interpretation of timber within each unit. Taking direction from the CASPO report because part of her data was to be used for a Spotted Owl habitat analysis, Fites took a conservative approach, moving borderline timber stands to larger size classes. The 1993 Landscape Analysis acknowledged that, [r]eeent stand exam data by sale unit would provide the most precise means for classifying the stand diameter composition within each sale unit. This data was not available so aerial photo interpretation was used. Sale unit boundaries mapped on 1:12,000 scale color aerial photographs were obtained from Ranger District Timber personnel for the site specific assessment____ It should be noted that there is no way to precisely measure such a determination from aerial photographs, but the most accurate determination was made with an error on the side of being conservative for purposes of this analysis. The term conservative as applied here means that for borderline situations, a larger crown size was assigned____ Crown density was estimated ocularly. The percent of each unit that was comprised of 4G, 4N, or possible old growth was ocularly estimated for each unit, with an estimated accuracy of 5-10%____ A random subset of the photo-typing work was sent to the National Aerial Photography Center in Salt Lake City for an error assessment of — 564 — the timber typing completed with aerial photos. PE 10 at 351. At approximately the same time Fites conducted her assessment, the Forest Service, through Fites, hired an outside contractor— Hammon, Jensen, Wallen & Associates — to do a photo interpretation of the timber sale units, including the Bald Mountain timber sale units. Fites had worked with Hammon, Jensen, Wallen & Associates previously, and they came highly recommended. Results from typing done by the independent contractor varied greatly from Fites’ typing and indicated that the Bald Mountain harvest would have only a minimal impact upon late serai timber strata. Further, as indicated in the 1993 Landscape Analysis, Fites had Forest Service photogrammetry experts at the Salt Lake City Geometronics Center perform a photo interpretation of timber strata within the Bald Mountain units based on a random sample of aerial photography provided by Fites. Fites’ timber strata interpretation and that of the Geometronics Center correlated at 50 percent. Phipps nevertheless directed that Fites’ data be used because it was more conservative. Phipps further discounted the assessment of the Geometronics Center based upon personal conflicts he had with the individual in charge of that center. To determine the timber strata present within Combined Watershed 13, Fites used the 1991 Forest Vegetation Inventory, which was primarily compiled based upon satellite imagery. Questioning the accuracy of the 1991 inventory, Fites did an accuracy assessment of that 1991 inventory, issuing the results on April 30, 1993. Based on that accuracy assessment, Fites determined that the 1991 inventory was likely only 63% accurate. In this regard, Fites indicated that 37% of the trees classified as 3G/N were actually 4G/N.11 In conducting the accuracy assessment, however, Fites used an inappropriate method, although Fites did not realize that her methodology was inappropriate at the time she conducted the accuracy assessment and compiled the 1993 Landscape Analysis. Nevertheless, Fites admitted that she may have realized that her accuracy assessment was faulty as early as January 1994, and certainly knew before the decision was made to cancel the proposed Bald Mountain timber sale. Fites further stated that she could not recall whether such knowledge was ever communicated to either contracting officer Phipps, or his successor, Robert Harris. Based upon Fites analysis of the photo typing, the 1993 Landscape Analysis determined that there were 82 acres of 4G timber strata present in the Bald Mountain cutting units, and that harvest of these units would lower the percentage of 4G timber strata in Combined Watershed 13, from 5.2% to 4.7%. Cancellation of the Sale With environmental documentation at or nearing completion with regard to the Bald Mountain timber sale, Forest Supervisor Phipps — the individual who made the decision not to award the Bald Mountain timber sale in March 1993, and began the process of environmental reconsideration of the Bald Mountain timber sale — left the Eldorado in November 1993. Acting Forest Supervisor Harris arrived on the Eldorado at approximately the same time Phipps left.12 In December 1993, Beth Paulson, Forest Environmental Coordinator, created a two-page document entitled Summary of Effects of the Bald Mountain Timber Sale, which summarized, based on the analysis conducted — 565 — by the Forest Service biologists and staff, the effects of the Bald Mountain timber sale on the California Spotted Owl, the Pacific Fisher, and the late serai forest. Based on Paul-son’s summary, Acting Forest Supervisor Robert Harris informed Wetsel by letter dated February 24, 1994, that all bids for the Bald Mountain timber sale were being rejected and that the Bald Mountain timber sale would not proceed. Harris and the Forest Service’s rejection were premised upon 36 C.F.R. § 223.100(a)13 and Item 8 of the Instruction to bidders. The Forest Service and Harris posited three reasons for the decision to cancel award of the Bald Mountain timber sale: (1) that harvest of the Bald Mountain timber sale would have significant adverse effects on habitat for the California Spotted Owl; (2) that harvest of the Bald Mountain timber sale would have significant adverse effects on habitat for the Pacific Fisher; and (3) that the planned harvest of the Bald Mountain timber sale would violate the late serai forest guideline for 4G timber strata under the Eldorado National Forest Land and Resource Management Plan. This lawsuit by plaintiff Wetsel against the defendant United States, acting through the Forest Service, followed. DISCUSSION This case requires the Court to resolve two disputed issues. First, the Court must determine whether the Forest Service breached its implied contractual duty to fairly and honestly consider Wetsel’s bid on the Bald Mountain timber sale by canceling the sale. Second, if the Forest Service did breach its contract to fairly and honestly consider Wetsel’s bid, the Court must determine what remedy Wetsel is entitled to as a result of that breach. I. The Forest Service Decision to Cancel the Proposed Bald, Mountain Timber Sale An invitation for bids issued by the government for bidding on a government contract carries an implied contractual obligation to fairly and honestly consider all responsive bids. See Parcel 49C Ltd. Partnership v. United States, 31 F.3d 1147,1150-51 (Fed.Cir.1994); Prineville Sawmill Co. v. United States, 859 F.2d 905, 909 (Fed.Cir. 1988); see also Heyer Products Co. v. United States, 147 Ct.Cl. 256, 177 F.Supp. 251, 252 (1959). The government breaches this implied contract if its consideration of the bids is found to be arbitrary, capricious, without rational basis, or an abuse of discretion. See Parcel 49C Ltd. Partnership, 31 F.3d at 1153; Prineville, 859 F.2d at 909; see also Keco Indus. v. United States, 203 Ct.Cl. 566, 574, 492 F.2d 1200 (1974). Government reservation of the right to reject bids on a government contract does not absolve the government from its obligation to fairly and honestly consider a contractor’s bid. See Prineville Sawmill Co., 859 F.2d at 912. If the government breaches its implied duty to fairly and honestly consider a contractor’s responsive bid, the contractor may seek relief, including equitable relief, in the Court of Federal Claims. See 28 U.S.C. § 1491. As noted previously, Contracting Officer Harris gave three reasons for canceling the Bald Mountain timber sale: (1) the harvest of the Bald Mountain timber sale would have significant adverse effects on habitat for the California Spotted Owl; (2) the harvest of the Bald Mountain timber sale would have significant adverse effects on habitat for the Pacific Fisher; and (3) the planned harvest of the Bald Mountain timber sale would violate the late serai forest guideline for 4G timber strata under the Eldorado National Forest Land and Resource Management Plan. Thus, the sole issue before the Court is whether the Forest Service decision to cancel the proposed Bald Mountain timber sale for these three reasons was arbitrary, capricious, or without rational basis. A The Late Serai Forest Guideline for bG Timber One of three reasons Harris posited for canceling the proposed Bald Mountain timber sale was that the timber harvest would — 566 — violate the late serai forest guideline for 4G timber strata under the Eldorado National Forest LRMP. Harris’ conclusion, drawn from the 1993 Landscape Analysis, was based on his determination that the Bald Mountain harvest would reduce the percentage of 4G timber strata in combined Watershed 13 from 5.2 to 4.7%. Such a decrease would violate Forest Service Management Practice 56 in the Eldorado National Forest LRMP, which specified that at least 5% of each timber strata or “serai stage” be present in any management area. 1. Harvest of 4-G/N timber from, the Bald Mountain Gutting Units In producing the 1993 Landscape Analysis and determining that the Bald Mountain harvest would reduce the amount of 4G timber below acceptable levels, Forest Service ecologist Fites relied exclusively on aerial photo interpretation of the harvest units without conducting any on the ground verification. Among experts who utilize timber strata typing, it is generally recognized that typing timber strata without on the ground verification is not good practice. Indeed, Forest Supervisor Phipps erroneously believed Fites had verified her data using on the ground verification and stated that he would not rely on the results of photo interpretation if he knew that on the ground verification had not been conducted. In contrast to the Bald Mountain assessment, when concerns over timber strata typing arose with respect to another sale on the Eldorado National Forest, the Forest Service conducted on the ground verification. No such effort was made with regard to the proposed Bald Mountain timber sale. Nonetheless, Fites justified her typing as “conservative” in accordance with the direction provided by the CASPO report. The CASPO report, however, directs timber typing of sale units to be based upon field verification of each unit to be cut, and that one should use a conservative approach— assigning larger class 4 to borderline size 3/4 strata — only after field verification has been conducted. In this regard, the CASPO report cautions, “[D]o not rely on current timber inventories to determine stand strata. Base this determination on field verification of each cutting unit during stand inventories in preparation for sales.” Tr. at 950. Thus, Fites’ typing of timber strata for units of the proposed Bald Mountain timber sale declined to apply a favored practice of verifying photo interpretation of typing with on the ground verification. Further, Fites’ justification for applying a conservative approach whereby borderline 3/4 size class timber strata were assigned size class 4 without conducting an on the ground verification was flawed in light of the CASPO report’s requirement that on the ground verification precede conservative timber strata typing. If Fites’ failure to conduct on the ground verification of the timber strata within the area of the Bald Mountain timber sale was the sole challenge to the Forest Service’s decision to cancel the Bald Mountain sale, the Court would not find the Forest Service’s decision to be arbitrary or capricious. However, the record reveals more serious flaws. First, Fites’ unverified photo interpretation produced results radically different from Forest Service identification of timber strata within the area of the Bald Mountain sale conducted in 1992 as part of the 1992 Wildlife BE and CEA. The 1992 results had been verified by on the ground examinations and, prior to Fites’ study, Phipps believed the 1992 results to be rehable. According to the 1992 Wildlife BE, there existed only 27 acres of 4G/N timber in the sale units of the Bald Mountain sale. Fites’ typing, by contrast, indicated 149 acres of 4G/N timber within the Bald Mountain units, and that the Bald Mountain harvest would remove 82 acres of 4G timber. Notwithstanding this discrepancy, Fites’ typing went unquestioned by either Phipps or Harris. In addition, the results of the independent contractor Fites hired to conduct photo typing of the area of the Bald Mountain sale varied greatly from Fites’ typing and indicated that the Bald Mountain harvest would have only a minimal impact upon late serai timber strata. Fites, however, rejected the independent contractor’s results, notwithstanding Fites’ instruction to the independent contractor that the contractor’s analysis — 567 — be conservative with regard to larger or denser classes of timber strata. Finally, the Forest Service’s own aerial photography specialist group, the National Geometronics Center, reviewed Fites’ typing and concluded that it was likely only 50% accurate. Further inquiry by the Forest Service as to the degree of error, however, was stifled by the desire of Forest Service employees to reach a conservative result, or otherwise because of personal conflicts between the Forest Supervisor and a staff member at the National Geometronics Center — the branch of the Forest Service relied upon to assess the accuracy of the data compiled. In sum, Phipps and Fites recognized that Fites’ typing of timber strata within the Bald Mountain cutting units conflicted with other recent and contemporaneous attempts to type timber strata in the same areas. In fact, Fites was not confronted just with conflicting data and analysis regarding timber strata in the units of the proposed Bald Mountain timber sale. Rather, Fites knew that her typing of timber strata within the proposed Bald Mountain cutting units was likely erroneous, yet she made no effort to correct for error. Further, Fites knew of procedures that might correct errors in her assessment — including on the ground examination of forest stands — but made no attempt to conduct them herself, or to alert Phipps that there might be a need for such on the ground examination. Indeed, Phipps thought Fites’ data had been verified by on the ground examination. Also, significantly, although Harris apparently had access to the environmental documentation drafted while Phipps was Forest Supervisor, the record indicates that Harris never reviewed that documentation other than to perform cursory editorial work. 2. kG Timber in Combined Watershed IS The flawed process by which the Forest Service ultimately determined to cancel the Bald Mountain timber sale went beyond typing of the timber strata within the cutting units. Rather, flaws extended to Forest Service typing of 4G timber strata within Combined Watershed 13. In fact, in determining that the Bald Mountain harvest would violate Eldorado National Forest LRMP Management Practice 56 by lowering timber strata in Combined Watershed 13 from 5.2% to 4.7%, the 1993 Landscape Analysis used the 1991 Forest Inventory to type timber strata within Combined Watershed 13. The 1993 Landscape Analysis therefore ignored the Forest Service and Fites’ findings that the total amount of 4G timber in Combined Watershed 13 had been significantly understated. A determination that the 1991 Forest Inventory, which relied on satellite imagery to classify timber strata, was inaccurate did not become apparent until Fites issued an accuracy assessment of the 1991 inventory on April 30, 1993. Fites’ accuracy assessment revealed that 37% of the timber strata classified as size class 3 should have been classified as size class 4. Fites acknowledged the 37% variability in both the 1993 and 1994 Landscape Analyses. Nevertheless, the 1993 Landscape Analysis omitted revealing that the 37% variability demonstrated that 37% of timber strata in size class 3 should have been typed size class 4. The 1994 Landscape Analysis which, although released prior to the decision to cancel the Bald Mountain timber sale, omitted any analysis of the effects of the Bald Mountain harvest. A gross error of 37% would, and did, significantly affect Fites and the Forest Service’s determination that harvest of the Bald Mountain sale would violate Management Practice 56 of the Eldorado National Forest LRMP.14 Indeed, Forest Service policy dictates accuracy of at least 75%. The 1993 Landscape Analysis, which specifically addressed the effect of the proposed Bald Mountain timber sale on the late serai forest, — 568 — made no effort to adjust its analysis to account for the 37% error in timber strata typing. By contrast, the 1994 Landscape Analysis applied a 37% correction factor to timber strata typed size class 3 which acknowledged that such timber strata was actually size class 4. Adjustment included timber strata in Combined Watershed 13. Nevertheless, the 1994 Landscape Analysis did not specifically address the effect of the Bald Mountain timber sale upon 4G timber strata and the late serai forest, despite issuance of the 1994 Landscape Analysis on January 21, 1994, over a month before cancellation of the proposed Bald Mountain timber sale.15 Thus, despite the Forest Service’s knowledge that at least 37% of size class 3 timber strata was misclassified, and despite Fites and the Forest Service’s attempts to correct for the 37% error in other environmental analyses of timber strata on Combined Watershed 13, no attempt was made to correct the assessment regarding the effects of the proposed Bald Mountain timber sale. Consequently, Fites and the Forest Service erroneously calculated that there existed 5.2% 4G timber strata within Combined Watershed 13 when the Forest Service determined the effect of the proposed Bald Mountain timber sale upon the late serai forest. Further compounding the flaws already present in the Forest Service’s reconsideration of environmental circumstances surrounding the proposed Bald Mountain timber sale, Fites acknowledged that, prior to the decision to cancel the Bald Mountain timber sale, she was aware that her accuracy assessment itself was likely flawed. Notwithstanding Fites’ acknowledgment that her own accuracy assessment was likely flawed, Fites made no attempt to amend either the 1993 or 1994 Landscape Analyses. Lack of valid accuracy assessment runs contrary to Fites own belief that — in utilizing the 1991 Forest Inventory or any map layer — a valid accuracy assessment of timber strata for a particular area ought to be conducted. A valid accuracy assessment, Fites testified, is “standard procedure.” In sum, at the time Acting Forest Service Supervisor Harris made his decision to cancel the proposed Bald Mountain sale, significant errors existed in the environmental analyses upon which he based his decision to cancel the sale. These errors were known to Harris and Forest Service personnel. No effort was made to account for inaccuracy in the timber strata typing of Combined Watershed 13 when analyzing the potential environmental effects of the proposed Bald Mountain harvest. No effort was made to correct for inaccuracies. The Court has doubts as to whether Harris even understood the impact of the inaccuracies upon environmental review of the Bald Mountain timber sale — as he should have — when he made the decision to cancel that sale. Further, Forest Service personnel apparently had doubts even as to the accuracy of their own error assessments. Such information was never communicated to Harris. Thus, Harris could not have considered this vital information in determining whether to cancel the proposed Bald Mountain timber sale. Consequently, at the time Harris decided to cancel the proposed Bald Mountain timber sale, the Court finds that the Forest Service had no idea as to what timber strata was present in Combined Watershed 13. As such, the Court finds that the Forest Service could not have made a reasoned decision to cancel the proposed Bald Mountain timber sale because the sale violated the late serai forest guideline for 4G timber under the El Dorado National Forest LRMP. B. The Pacific Fisher The November 24, 1993 Fisher BE, upon which the Forest Service’s decision to cancel the Bald Mountain timber sale was based, relied on Fites’ erroneous timber strata data compiled in the 1993 Landscape Analysis. To determine the total acreage of 4G/N timber strata within area H, the Forest Service relied on the erroneous 1991 Timber Inventory, even though the estimate of the amount — 569 — of 4G timber strata was known to provide only 63% accuracy. No adjustment to the 1993 Fisher BE was made to account for the 37% error acknowledged in the 1993 Landscape Analysis. The November 24, 1993 Fisher BE determined that the Bald Mountain timber sale would remove 149 acres of 4G timber strata within potential Fisher Use area H, thereby increasing the proportion of unsuitable fisher habitat. Because the 1991 Inventory was inaccurate and understated the amount of 4G timber in Potential Use Area H, use of the 1991 data magnified the effect of the Bald Mountain harvest upon Potential Use Area H. The Forest Service knew of the error, as acknowledged in the Fisher BE, as early as April 30, 1993, the date of Fites’ accuracy assessment. Further, the Forest Service recommendation in the November 24, 1993 Fisher BE to cancel the sale substantially relied on Fites’ determination that the Bald Mountain harvest would remove large patches of 4G timber. Although the November 24, 1993 Fisher BE was issued 16 days after the November 8, 1993 Fisher BE — which concluded that the Bald Mountain harvest would not have much effect on the Pacific Fisher habitat — the November 24, 1993 Fisher BE concluded that the Bald Mountain timber sale would have a significant effect on Pacific Fisher habitat. Lipton, at trial, commented for the first time that the change in the conclusion with regard to the effect of the proposed Bald Mountain harvest on the Pacific Fisher was largely due to information, obtained from Fites, that the timber sale would reduce three large patches of 4G/N timber in the cutting units of the Bald Mountain timber sale.16 As the Court has noted previously, the accuracy of Fites’ assessment of timber strata within the Bald Mountain units was dubious, and known to be so by Fites and the Forest Service.17 Because the impact on the Pacific Fisher was based upon data known to be erroneous, the analysis was without a rational basis. Further, the Forest Service was aware that the error was in favor of canceling the Bald Mountain timber sale, yet the Forest Service made no attempt to adjust for this bias. Conclusions drawn from the Fisher BEs, therefore, could not be supported other than by ignoring or disregarding glaring deficiencies in the evaluation itself The Court finds, therefore, that the Forest Service did not have a rational basis for canceling the Bald Mountain sale based upon the effect of that sale on the Pacific Fisher or the fisher’s potential habitat. C. The California Spotted Owl Finally, deciding to cancel the proposed Bald Mountain timber sale because the harvest would foreclose future management options for the California Spotted Owl is flawed due to use of erroneous timber strata data from the 1993 Landscape Analysis. To calculate the effect of the Bald Mountain timber sale upon the California Spotted Owl, the Forest Service used the “disturbance index,” which attempted to calculate the amount of preferred owl habitat — mainly 4G/N timber strata — that would be removed by a timber harvest, in relation to the amount of preferred habitat existing in the area surrounding a protected activity center. However, in calculating the amount of 4G/N timber strata in the areas surrounding owl sites, the 1993 Owl BE relied on the erroneous 1991 timber inventory, even though the 1993 Owl BE acknowledged that such data was known to be erroneous. No attempt was made to correct for data known to be erroneous. Further, different data sets were used in the numerator and denominator of the disturbance index fraction in calculating the dis — 570 — turbance index ratio. At the time the disturbance index was used to calculate the effect of the Bald Mountain timber sale, Phipps was aware that using different data sets would create a biased disturbance index ratio. Phipps justified reliance upon known biased results because he believed that any bias would err in favor of maintaining owl habitat and canceling the proposed timber sale. Nevertheless, Phipps had no idea as to the degree of bias that resulted from the use of different data sets. Thus, Phipps elected to disregard the 5% threshold for the disturbance index — a threshold the Forest Service had chosen — in favor of arbitrary results, so long as he believed those results erred in favor of conserving owl habitat. In sum, Phipps knowingly disregarded the very analytic procedure originally designed to measure the effect of the Bald Mountain harvest upon the California Spotted Owl. Thus, the Court finds that the Forest Service did not have a reasonable basis for determining that harvest of the proposed Bald Mountain timber sale would have a significant adverse impact upon the California Spotted Owl because the analysis upon which that determination was based was founded upon data the Forest Service knew to be erroneous. As with the Pacific Fisher analyses, the Court finds that Harris relied primarily on the summary of environmental analyses provided to him, and never sought to investigate inconsistencies and errors in the environmental analyses themselves. Thus, the Forest Service’s decision to cancel the sale, based upon its potential effect upon the California Spotted Owl, did not have a rational basis. II. Breach of the Contract and the Relief Available In sum, the Forest Service did not have a rational basis for canceling the Bald Mountain timber sale based upon any of the three reasons articulated by Harris. Erroneous data infected analysis of the proposed Bald Mountain timber sale’s effect upon the Late Serai Forest, the Pacific Fisher, and the California Spotted Owl. Although the Forest Service was aware that the data they were using to evaluate the environmental effects of the sale was erroneous, they incorporated the data into their analyses anyway. For some data, the Forest Service attempted to ascertain the degree of error; other times it did not. Whether or not the degree of error was known to the Forest Service, the Forest Service ignored the error in conducting their environmental analyses of the effects of the Bald Mountain timber sale, so long as Forest Service personnel believed that the results of the analyses would lead to a conclusion that overstated the effects of the Bald Mountain timber sale upon the environment. Finally, with flawed environmental documentation complete, or nearly complete, the Forest Supervisor overseeing the environmental review departed. The new contracting officer, Harris, having virtually no background in the environmental review taking place, canceled the proposed Bald Mountain timber sale within three months. In deciding to cancel the sale, Harris relied substantially on summaries of the new environmental analyses. Although he had access to the analyses themselves and knew generally that inconsistencies and errors existed in those documents, he limited his role to minor editorial comments on the analyses. At the same time, there is no evidence that Forest Service personnel, including those preparing the analyses, ever alerted Harris to the fundamental problems that existed in those analyses. When Wetsel submitted its responsive bid on the Bald Mountain timber sale, and the government accepted that bid, the government owed Wetsel a contractual duty to fairly and honestly consider that bid. The government breached its contractual obligation because it canceled the sale, and rejected Wetsel’s bid, without a rational basis for so doing. The Court assumes that the government’s stated reasons to protect the environment would, if supported, be valid reasons for canceling the sale. However, the Forest Service — at best — had no idea whether these reasons were supportable and — at worst — was predetermined to find support for their stated reasons to cancel the sale. “The law of procurement does not tolerate ‘actions reflecting personal predilections of administrative officials, whether ascribable to — 571 — whim, misplaced zeal, or impermissible influence.’” Parcel 49C Ltd. Partnership, 31 F.3d at 1153 (quoting M. Steinthal & Co. v. Seamans, 455 F.2d 1289, 1305-06 (D.C.Cir. 1971.)). To be dear, the Court does not find that Wetsel bargained for any particular environmental standard or methodology to be applied by the Forest Service in meeting its obligations under various environmental statutes and regulations. Rather, Wetsel bargained that the Forest Service would not knowingly use inaccurate data that tended to support a finding that the proposed timber sale would violate controlling environmental standards. Wetsel bargained for fair and honest consideration of its bid. The Court finds that, in this case, Wetsel is entitled to such consideration. The Forest Service is, therefore, directed to reinstate the award process to the point where it left off before the illegality occurred. In determining whether to award or cancel the Bald Mountain sale, the Forest Service retains the power to terminate the award process for any legal reason. Cf. Parcel 49C Ltd. Partnership, 31 F.3d at 1154 (awarding similar relief upon finding the government breached its contractual obligation to fairly and honestly consider a contractor’s bid by canceling a government procurement). In this regard, the Court does not mandate that the Forest Service adopt any particular environmental standards or methodology, other than that required by law. Rather, the Court merely holds that the Forest Service cannot arbitrarily and capriciously make its determinations without a rational basis, and in favor of a desired result. The Court’s decision constitutes the full measure of Wetsel’s expectation. CONCLUSION For the reasons stated, the Court holds that Forest Service’s cancellation of the Bald Mountain timber sale was arbitrary, capricious and without rational basis, and that the Forest Service, therefore, breached its contractual obligation to fairly and honestly consider Wetsel’s bid on that sale. The Forest Service should proceed to determine whether to award the Bald Mountain timber sale, or whether that sale should be canceled for any legal reason. The Forest Service shall make such a determination within 60 days from the date of this opinion and so inform the Court. This opinion has been issued under seal. Within 60 days from the date of this opinion, the parties shall file with the Court and designate by ellipses those portions of the opinion, if any, that should remain sealed. A redacted version of the opinion shall be issued thereafter. Footnotes [1] The Endangered Species Act, 15 U.S.C. § 1531 et seq., was enacted to "provide a means whereby the ecosystems upon which endangered species and threatened species depend may be conserved, [and] to provide a program for the conservation of such endangered species and threatened species ..." 16 U.S.C. § 1531(b). Thus, ESA protects endangered and threatened species. See 16 U.S.C. § 1531(c). An “endangered species,” with certain exceptions not relevant to this action, is "any species which is in danger of extinction throughout all or a significant part of its range.” 16 U.S.C. § 1532(6). "Threatened species” are 'any species which [are] likely to become an endangered species within the foreseeable future throughout all or a significant portion of its range.' 16 U.S.C. § 1532(20). [2] "Timber strata” refers to an area of relatively similar vegetation. [3] An owl "protected activity center” is the "best” 300 acres surrounding an owl activity site. According to the 1993 Owl BE, the protected activity center would, ideally, be comprised of 4G/N timber strata. [4] JoAnn Fites is also known as JoAnn FitesKaufman. [5] The Pacific Fisher is a mammal found mainly in the western United States. With an appearance resembling a weasel, the Pacific Fisher generally weighs from two to eight pounds and preys upon other animals both on the ground and in trees. [6] A revision and update of the 1993 Fisher BEs, entitled Biological Evaluation of the Habitat Requirements for Fisher on the Eldorado National Forest in Relation to Twenty-Four Sold Timber Sales, signed by Bombay and Mike Foster for Lipton, was issued on January 20, 1994. Some analyses contained in the 1993 Fisher BEs, specific to the Bald Mountain timber sale, were omitted from the 1994 Fisher BE, which dealt only with 24 sold sales on the Eldorado National Forest. [7] Potential use areas, of which potential use area "H" is an example, are areas designated by the Forest Service as having the potential, though not necessarily the current capability, to support populations of Pacific Fisher. [8] Although neither the California Spotted Owl, nor the Pacific Fisher are listed as threatened or endangered under the Endangered Species Act, both are designated as "sensitive species” on the Region 5 "Sensitive Plant and Animal Species List.” The Owl was so designated in October 1980, while the Pacific Fisher was designated a sensitive species in December 1984. [9] Acting for Dawn Lipton, Mike Foster signed the 1994 Fisher BE. [10] The Bald Mountain timber sale falls within Combined Watershed 13 on the Eldorado National Forest. Combined Watershed 13 contains 17,200 acres of Eldorado National Forest land and 29,987 acres of privately owned land. Thus, the Bald Mountain timber sale, which consists of 421 acres of national forest land, represents 2.4% of the national forest land within Combined Watershed 13, and less than .99% of the entire acreage of Combined Watershed 13. [11] Further evidence that Fites and the Forest Service knew that 37% of trees classified as 3G/N were actually 4G/N is found in a document drafted by Fites, Durham, Lipton and Foster, dated November 8, 1993, and entitled Landscape Analysis-Salvage and Drought. In that document, the Forest Service undertook to estimate the recruitment pool for large snags and logs in a watershed by summing the acres of 4G, 4N, 4P, and 4S timber strata. In the assessment, the Forest Service recognized that "detailed analysis of Forest inventory data has shown that 37% of the size class 3 timber strata is likely to have a substantial number of larger trees as well, particularly in the 3G and 3N strata." PE 115 at 02. Consequently, the Landscape Analysis lumped together the 4G and 37% of the 3G/N timber strata in calculating one assessment of the recruitment pool for large snags and logs. [12] Phipps returned to the Eldorado National Forest, replacing Harris, in June 1994. [13] 36 C.F.R. § 223.100 provides, in relevant part, that "[t]he sale of advertised timber shall be awarded to the responsible bidder submitting the highest bid that conforms to the conditions of the sale as stated in the prospectus unless: (a) Determination is made to reject all bids.” [14] Additionally, the 1993 Landscape Analysis acknowledged that the accuracy assessment demonstrating 37% error in the 1991 inventory was conducted solely for crown class size. Therefore, as the 1993 Landscape Analysis stated, if there were "any level of inaccuracy for other components of the strata label (species or density), then the total accuracy for the inventory maps could be lower.” PE 10 at 356. No accuracy assessment was, however, conducted with regard to species or density. [15] Likewise, the Forest Service's Landscape Analysis-Salvage and Drought, authored by Fites, Lipton, Durham, and Foster, and issued November 8, 1993, considered that 37% of timber strata classified as size class 3 was likely size class 4. [16] Lipton also stated that she changed her conclusion based upon comments from Forest Service employee Dr. William Zielinski, an expert in mammalian ecology and the ecology of the Pacific Fisher, during a 15-20 minute telephone conversation with Zielinski in November 1993. Zielinski, however, denied that he ever substantively reviewed the draft Fisher BE or that he provided substantive comments to Lipton about the Fisher BE. [17] Though the final version of the Fisher BE was issued January 20, 1994 — a month before cancellation of the Bald Mountain timber sale — the 1994 Fisher BE provided no analysis of the potential effects of the Bald Mountain harvest upon the Pacific Fisher.