Opinion
Case No. CV-04-367-E-BLW.
September 27, 2005
MEMORANDUM DECISION AND ORDER
INTRODUCTION
The Court has before it motions for summary judgment filed by both parties. The Court held oral argument on the motions and they are now at issue. For the reasons expressed below, the Court will enjoin the McGarry and Big Bend logging projects, but will not direct the Forest Service to modify its Revised Forest Plan and will not grant a blanket injunction against all further logging in the Targhee National Forest.
BACKGROUND
1. Litigation BackgroundNEC seeks to enjoin logging in the Targhee National Forest (TNF) on the ground that the Revised Forest Plan (RFP) for the TNF is flawed. To challenge the RFP, NEC must show that harm from logging is "imminent" and that the RFP "plays a causal role with respect to the imminent harm from the logging." Ohio Forestry Ass'n, Inc. v. Sierra Club, 523 U.S. 726, (1998).
To meet its burden under Ohio Forestry, NEC identifies two logging projects planned in the TNF: (1) the Big Bend Ridge Vegetation Management Project, and (2) the McGarry Salvage Timber Project. The McGarry project proposes to log about 3.5 million-board-feet of dead and dying trees from about 500 acres of the TNF. The Forest Service prepared an Environmental Assessment (EA), and issued a Finding of No Significant Impact on November 19, 2003.
In an earlier decision in this case, the Court enjoined the McGarry sale, finding a link between the sale's approval and an inadequacy in the RFP. The Court found it likely that NEC would prevail on its claim that the RFP is flawed because it authorizes logging (1) without regard for its effect on large snag habitat, and (2) without regard for whether it drops the snag habitat to a level that would not support the 61% biological potential figure set forth in the RFP-EIS. Those flaws, the Court held, could violate the command of the National Forest Management Act (NFMA) to ensure the viability of eight species of woodpeckers in the TNF.
In that decision, the Court did not consider the Big Bend sale because it was not imminent. That sale is now imminent, and NEC seeks to enjoin it from going forward.
The Big Bend project proposes to log about 9.1 million board feet of timber on 2693 acres, promotes the growth of aspen trees on 580 of those acres, proposes to construct 8.8 miles of roads and close 33.4 miles of roads. The Forest Service prepared an Environmental Impact Statement (EIS), and approved the project on December 19, 2003.
While NEC seeks to stop both logging projects, its overall goal is much broader: NEC asks the Court to "invalidate the RFP decision, remand to the [Forest Service] for further study and analysis, provide clear and detailed instruction as to the need to base a decision to revise the 1985 Forest Plan on an honest analysis of the cumulative impacts from implementation of that plan, especially with regard to failed goals and objectives, and base its next proposal on the best science and information currently available." See NEC Reply Brief at p. 39.
This case thus presents both a narrow challenge to two logging projects and a much wider challenge to the RFP itself. The Court will resolve these issues after reviewing the factual background of this case.
2. Factual Background
Both logging projects at issue here were approved by the Forest Service pursuant to the TNF RFP. That plan revised the TNF's original land and resource management plan approved in 1985 ("1985 Forest Plan"). That 1985 Forest Plan "emphasized an extensive salvage and reforestation program of dead lodgepole killed by a massive pine beetle epidemic over the previous 30 years." See RFP EIS at p. I-4.
The "extensive" logging contemplated by the 1985 Forest Plan was a continuation of aggressive logging that had been proceeding since at least 1960. In that year, the Forest Service initiated a clear-cutting program with the largest timber sale in the lower 48 states in agency history. See 42 AR 30163. To accommodate this logging, a sawmill was built in St. Anthony, and for the next 25 years, loggers cut over 23,000 acres of timber. Id.
By the 1970s, mountain pine beetles were killing millions of trees in the TNF. Id. To manage the salvage of these dead trees, the Forest Service issued the 1985 Forest Plan. That Plan projected a timber harvest of 864 million board feet during the first decade of the plan. See 19 AR P011072.
As the logging envisioned by the 1985 Forest Plan neared completion, the Forest Service decided to revise that Plan. The Forest Service recognized that the extensive logging in the past was "a departure from a sustained yield of timber harvest and could not be continued beyond the first decade (1985-1995) in an environmentally sound manner." See RFP EIS at p. I-4. Moreover, the Forest Service was finding it increasingly difficult to "meet the standards and guidelines in the 1985 Forest Plan." Id.
One of those difficult-to-meet guidelines concerned quantities of old growth timber. The 1985 Forest Plan defined the term "old growth" and then set a minimum habitat level of 3% old growth in average optimum blocks of 300 acres, plus another 3% replacement habitat. See Plaintiff's Exhibit 2.
Maintaining old growth standards is important because "[m]any of the species present on the [TNF] utilize old growth and other `older' forest habitat." See Forest Service Brief at p. 41. For example, the Great Gray Owl, designated by the TNF as a "sensitive species," uses old-growth forests with large-diameter trees or snags for nesting. See 49 AR 36171-71. Another TNF sensitive species, the Northern Goshawk, uses a variety of habitats, but nests in mature or old growth forest stands. See Process Paper D at p. 230. Eight species of woodpeckers use cavities in dead and decaying wood for nesting. Id.
The past history of clear-cutting raised a question whether the standards for old growth set in the 1985 Forest Plan were being met. In 1994, during the process of revising the 1985 Forest Plan, TNF Wildlife Biologist Mark Orme surveyed TNF data in an attempt to estimate the amount of old growth. To determine whether the data met old growth requirements, Orme used the most recent definition of old growth, known as the R-4 definition, developed in 1993.
Orme gleaned data from two sources. First, he used timber stand data. Timber stands are polygons of similar vegetation ranging in size from several acres to over 100 acres. Analyzing this data covering 621,800 acres of the TNF (about 1/3 of the TNF), he concluded that only "8,045 acres were identified as old growth." See AR P049623. He observed that this was "1.3 percent of the 621,800 forested acres covered by stand exam information." Id. Obviously, this was less than half of the 3% standard set in the 1985 Forest Plan.
Second, Orme examined data on the "permanent forest inventory plots." These plots were created as part of the Forest Service's inventory of timber state-wide conducted in 1991, and were set forth in a report entitled "Idaho's Forests, 1991." Orme concluded that 19 plots out of 412 in the TNF were identified as old growth. Id. He noted that "we have not identified acres of old growth associated with the 19 plots." Id.
After reviewing the data, Orme concluding that "[t]here does not appear to be a lot of old growth acres." Id. He explained that "[t]he number one parameter that disqualified most stands as old growth was the age parameter. We simply do not have trees that are old enough to meet the old growth definition." See AR 49637. Orme sent the data to Dee Sessions, his colleague at the TNF, who commented that "[y]ou're right, there doesn't seem to be a lot of old growth. In fact, I doubt there are 8,045 acres." See AR P049631.
Two years later, in 1996, the Forest Service took another look at old growth, and found that 36 of the 412 plots, or 8.7% of the plots, met "all of the old growth characteristics that could be determined from the permanent forest inventory plots." See Process Paper D at p. 11. This study did not discuss the forest stand data that had caused such concern within the Forest Service two years earlier. In fact, the 1996 study did not even reference the 1994 study.
A study that completely ignores troubling data is suspect. The 1996 study therefore does not begin on strong footing, and a closer review reveals further flaws. First, the plot data on which the 1996 study was based was collected prior to the establishment of the R-4 old growth definition, and hence the data did not measure some criteria important to the R-4 definition. Specifically, the plot data did not measure (1) downed dead trees, (2) live tree decadence, or (3) dead tree height. In reviewing the 1996 study, the Forest Service conceded that "[i]f some of the plots were deficient in downed dead trees, then our calculation pertaining to the quantity of old growth will be high." See RFP-EIS at p. III-7.
A second area of concern was created when the Forest Service adjusted the plot data by adding "1-inch to all of the diameter at breast height (dbh) measurements to allow for growth." Id. at III-8. The Forest Service recognized that "[a]dding 1-inch dbh is probably optimistic for old trees," but justified it on the ground that "we did not want to eliminate plots which were close to the minimum required dbh." Id. In other words, the Forest Service manipulated the data to reach a desired result.
The review to this point has revealed flaws in the 1996 study that raise substantial questions about its methodology and conclusions. But an even more basic weakness strikes at the very core of the study.
Extrapolating from samples requires a statistical analysis to ensure accuracy. Here, 412 plots were used to estimate conditions on the TNF at large. To ensure the accuracy of this estimate, a statistical analysis must demonstrate that the 412 plots represent a statistically valid sample size on which to base predictions about conditions in the entire TNF. The Administrative Record contains no such statistical analysis.
The Forest Service, in its briefing, refers to a report it drafted entitled "Idaho's Forests, 1991." That report "presents the results of the 1990-1991 inventory of Idaho's forest lands." See Report Title Page. This Report, however, says nothing about whether data from the 412 plots may be used to estimate the conditions in the TNF.
The Report does contain a general explanation of the method used to make the broad inventory of timber throughout the state. See Report at pp. 23-25. Those paragraphs, however, contain neither a statistical analysis nor information from which a statistical analysis could be done.
The Report does contain a table showing the "standard error for sampling area." See Report, Table 2 at p. 32. That table identifies standard error for the 3,167 plots spread over all National Forests in Idaho, but does not address the statistical validity of using the 412 plots to estimate conditions in the TNF.
In fact, the Report warns that statistical error will increase if the data is used in units smaller than the state-wide National Forests as a whole: "Standard errors for data other than the totals (Table 2) are not given in this report, but as the data are subset into smaller categories the errors should be expected to increase." See Report at p. 25.
The RFP-EIS did just this — it used the data to estimate something far short of state-wide conditions, i.e., conditions in the TNF. That analysis is subject to more error, although how much is unknown. But at an even more basic level, there is nothing in the Administrative Record discussing whether the 412 plots provide a statistically valid sample of conditions in the TNF. This is especially problematic given that just two years earlier, in 1994, the Forest Service was concerned that the TNF contained very little old growth.
The 1996 study remains the most recent attempt in the record to estimate R-4 old growth in the TNF. The RFP-EIS recognized the incomplete nature of the 1996 study. Introducing its discussion of that study, the RFP-EIS issues a caveat that "[t]he Forest does not have a complete old growth inventory." RFP-EIS at III6-. The RFP-EIS then goes on to summarize the 1996 study, and refers the reader to Process Paper D, which includes the entire 1996 study, for more details.
The RFP set a Guideline for R-4 old growth as follows:
In each principal watershed, the combination of old growth and late seral forest stage acres will be 20 percent or more of the forested acres. Where it exists, at least half of this (ten percent of the forested acres) should meet old growth characteristics.
This provision, the Forest Service concluded, would not be violated by the McGarry and Big Bend logging projects because neither project would log old growth timber. For this and other reasons, the Forest Service approved both projects.
NEC argues that these approvals violate NFMA because the record does not reveal any comprehensible inventory of R-4 old growth in the TNF, and further does not reveal any study of the distribution of R-4 old growth throughout the TNF. Neither amounts nor distribution can be extracted with any confidence from this record. The effects of these failings will be discussed below.
LEGAL ANALYSIS
3. Challenge to McGarry Big Bend Projects
The National Forest Management Act (NFMA) requires that for each National Forest, the Forest Service prepare a Forest Plan that "provide[s] for diversity of plant and animal species." 16 U.S.C. § 1604(g)(3)(B). The Forest Service has a legal obligation to ensure that any specific logging project is consistent with the applicable Forest Plan. Native Ecosystems Council v. U.S. Forest Service, 418 F.3d 953, 963 (9th Cir. 2005). When the Forest Service has "no inventory of R4 old growth" and is "without any knowledge of the . . . distribution of R4 throughout the Forest," the Forest Service "is not acting according to a forest wide plan as required by [NFMA]." Idaho Sporting Congress, Inc. v. Rittenhouse, 305 F.3d 957, 971 (9th Cir. 2002).
That is the case here. The record reveals neither an inventory of R-4 old growth or information on how R-4 old growth is distributed throughout the TNF. The 1996 study, relied upon by the RFP-EIS, is flawed, as discussed above, because (1) it ignored tree stand data from the 1994 study showing significant shortages of old growth; (2) was based on incomplete data that did not measure some R-4 criteria; (3) was based on data adjusted by the Forest Service to achieve a desired result; and (4) is not supported by any statistical analysis showing that it represents conditions forest-wide.
The Forest Service is required to insure the scientific integrity of its analysis and to "make explicit reference by footnote to the scientific and other sources relied upon for conclusions in the statement." 40 C.F.R. § 1502.24 (2005). "[T]he Forest Service calculations need not be perfect." Native Ecosystems Council v. U.S. Forest Service, 418 F.3d 953, 963 (9th Cir. 2005). Nevertheless, the Court "must still be able reasonably to ascertain from the record that the Forest Service is in compliance with the [Forest Plan]." Id. "If the administrative action is to be tested by the basis upon which it purports to rest, that basis must be set forth with such clarity as to be understandable." Id. (quoting SEC v. Chenery Corp., 332 U.S. 194, 196-97 (1947)).
The Forest Service argues here that any shortcomings in its R-4 old growth analysis are irrelevant because the McGarry and Big Bend projects will not log R-4 old growth. That was essentially the same argument considered and rejected by the Circuit in Neighbors of Cuddy Mountain v. Alexander, 303 F.3d 1059 (9th Cir. 2002), and by this Court in its decision on remand from Neighbors. See Neighbors of Cuddy Mountain v. Madrid, Civil No. 00-247 (Memorandum Decision filed February 4, 2005).
In Neighbors, the Circuit held that even if the particular logging project did not violate old growth standards in the logged area, the project may still violate NFMA if other areas of the forest were not in compliance with old growth standards. Neighbors, 303 F.3d at 1069. The Circuit went on to state that the Forest Service could not "don blinders to the overall condition of a national forest each time it approved a sale, quite literally losing sight of the forest for the trees." Id.
Here, the Forest Service had no statistically sound inventory of R-4 old growth in the TNF — or study of the distribution of R-4 old growth throughout the TNF — at the time it approved the McGarry and Big Bend projects. Thus, the Forest Service approved the projects without knowing if it was in compliance Forest-wide with the RFP requirements for R-4 old growth.
This lack of knowledge is important. If the R-4 old growth falls below the RFP Guidelines, timber stands that are approaching R-4 status may be considered for increased protection from logging. Even younger stands of timber may be set aside in some areas because nothing older is anywhere nearby. In other words, knowledge of R-4 quantity and distribution could change management practices. That is why Neighbors held that forest-wide compliance with the Forest Plan is relevant to the approval of a particular logging project.
Until the Forest Service can say with confidence that it is in compliance with the RFP old growth Guidelines, the McGarry and Big Bend projects must be enjoined. NEC asks the Court to go further and ban all logging in the TNF. As the Circuit held in Rittenhouse when faced with a similar request, "[s]uch a sweeping remedy is not warranted." Rittenhouse, 305 F.3d at 974. In that case, the Circuit noted that if the Forest Plan was not revised or amended, future logging projects may be subject to "the same infirmities present here." Id. But the Circuit refused to issue a blanket injunction, holding that "we prefer to consider such issues in the context of site specific actions, if and when they actually arise." Id.
This Court agrees. The Court will therefore not issue a blanket injunction against all logging in the TNF, but will enjoin the McGarry and Big Bend sales for the reasons expressed above. In addition, the Court reaffirms its prior decision enjoining the McGarry sale on the grounds stated in that decision.
NEC also asks this Court to require the Forest Service to revise the RFP, and to provide guidelines for that revision process. The Court declines on the ground that it is barred from doing so by Ohio Forestry Ass'n, Inc. v. Sierra Club, 523 U.S. 726, (1998). That case held that the Court had no jurisdiction to consider general challenges to a Forest Plan, but was restricted to hearing disputes over site specific projects that had been approved under the Forest Plan. That decision precludes NEC from making a broad-based challenge to the RFP that is not tied to a site specific project.
NEC responds that the Forest Service is failing to protect the viability of a species in violation of NFMA, and that the Administrative Procedures Act (APA) authorizes suits to correct "agency action unlawfully withheld or unreasonably delayed." See 5 U.S.C. § 706(1). However, that provision of the APA "empowers a court only to compel an agency to perform a ministerial or non-discretionary act or to take action upon a matter without directing how it shall act." Norton v. Southern Utah Wilderness Alliance, 124 S.Ct. 2373, 2379 (2004) (internal quotations omitted). NEC has not identified any such ministerial or non-discretionary act that the Forest Service has "unlawfully withheld." For that reason, the Court declines to require the Forest Service to revise the RFP.
The Court will enter a separate Judgment as required by Federal Rule of Civil Procedure 58.