UNITED STATES DISTRICT COURT FOR THE DISTRICT OF COLUMBIA
July 29, 2002
SOUTHWEST CENTER FOR BIOLOGICAL DIVERSITY, ET AL., PLAINTIFFS,
GALE NORTON, SECRETARY, U.S. DEPARTMENT OF INTERIOR, ET AL., DEFENDANTS.
The opinion of the court was delivered by: John M. Facciola United States Magistrate Judge
This matter is before me for a Report and Recommendation pursuant to LCvR 72.3. I herein take up the parties' cross motions for summary judgment. Plaintiffs, a group of environmental organizations, bring this action under the Endangered Species Act ("ESA"), 16 U.S.C.A. § 1531 et seq. (2000), to compel the listing of the Queen Charlotte goshawk as an endangered or threatened species. Because the best scientific data available indicates that this subspecies is not endangered or threatened, the decision to not list this subspecies based on its status in southeast Alaska is neither arbitrary nor capricious, and I shall recommend the denial in part of plaintiffs' motion for summary judgment and the granting in part of defendants' cross-motion. On the other hand, because the Fish and Wildlife Service ("FWS") failed to reach a decision on whether the Queen Charlotte goshawk is endangered or threatened in Vancouver Island, a significant portion of the subspecies' range, I shall recommend that the case be remanded for this limited purpose.
Queen Charlotte Goshawk
The Queen Charlotte goshawk (Accipiter gentilis laingi) ("goshawk") is a subspecies of the Northern goshawk (Accipiter gentilis), a forest-dwelling raptor found throughout the northern hemisphere. *fn1 The subspecies is named for the Queen Charlotte Islands of British Columbia, where it was originally identified. Although information is spotty, its range is believed to extend along the Pacific coast from the United States - Canadian border, north through Vancouver Island and the Queen Charlotte Islands in British Columbia and the Admiralty Islands in southeast Alaska. A.R. *fn2 II.039 at 22; A.R.III.B.011 at 3-4. The goshawk is distinguished morphologically from the Northern goshawk primarily by its smaller size and darker plumage. A.R.III.B.011 at 6. In addition, the goshawk has heavier streaking along its feather shafts and a dark cap over its shoulders. Id. at 6.
The goshawk exhibits a strong preference for productive old-growth forests. *fn3 A.R.II.039 at 37, 50-53. They have been found to nest and forage in forests with particularly dense canopies. Id. at 52. The subspecies' affinity for such forests appears related to its success in predation. Where a large amount of brush and smaller trees predominate, the goshawk is not able to maneuver through the understory. It also appears that productive old-growth forests host an abundance of the goshawk's preferred prey. Id. at 61. In addition to preferring older forests, goshawks were found to avoid areas that had been heavily logged, or clearcut. Id. at 66. The conversion of productive old-growth forest to clearcut and early seral stage forest is therefore considered particularly harmful to the goshawk. Under a traditional 100-year rotation schedule for timber harvest, the forest does not mature enough to support goshawk populations. A.R.III.A.07 at 33-34. *fn4
By all accounts, population size estimates of the goshawk are highly uncertain. The Conservation Assessment cites a handful of studies that estimated a range of anywhere from 100 to 800 pairs in southeast Alaska. A.R.II.039 at 22 (citing Crocker-Bedford 1990, 1994 and Iverson 1990).
In the discussion among the goshawk experts panel in July 1997, panelist Ted Swem noted that by extrapolating from the five known nests on Douglas Island as a typical density, 800-1000 pairs not an unreasonable estimate for southeast Alaska. A.R.III.B.009 at 9. No matter what the precise population numbers, it is well-established that the population density of the subspecies is quite low, especially compared with densities of the Northern goshawk in other parts of North America.
A.R.III.A.007 at 24-25. Finally, although there are no baseline or comparative statistics, scientists are confident in assuming that the goshawk numbers are declining due to the loss of productive old-growth forest in southeast Alaska and in British Columbia. A.R.II.039 at 66.
Endangered Species Act
The ESA defines a species (including subspecies) as "endangered" if it is "currently in danger of extinction throughout a significant portion of its range . . . ." § 1532(6). A species is "threatened," in turn, if it "is likely to become an endangered species within the foreseeable future throughout all or a significant portion of its range." § 1532(20). A species may be listed as threatened or endangered based on five factors enumerated in the statute:
(A) the present or threatened destruction, modification, or curtailment of its habitat or range;
(B) overutilization for commercial, recreational, scientific, or educational purposes;
(C) disease or predation;
(D) the inadequacy of existing regulatory mechanisms; or
(E) other natural or manmade factors affecting its continued existence. § 1533(a)(1). FWS must find the presence of only one of the above factors in order for a species to be listed. In addition, FWS must rely solely on the "best scientific and commercial data available" in arriving at a listing decision. § 1533(b).
There are two avenues by which a species is listed as endangered or threatened. First, FWS itself may initiate the listing. § 1533(a)(1), (b)(2). Alternatively, an "interested person" may petition FWS to add (or remove) species from either the endangered or threatened species lists. § 1533(b)(3)(A).
Once FWS receives such a petition, it has 90 days to decide whether it presents "substantial scientific or commercial information indicating that the petitioned action may be warranted." Id. If so, FWS must "promptly commence a review of the status of the species concerned." Id. Within 12 months after the petition is filed, FWS must determine that either (1) the petitioned action is warranted, in which case it must publish a proposed rule designating the species for protection; (2) the petitioned action is not warranted; or (3) the petitioned action is warranted but immediate promulgation of a rule is precluded by other pending proposals. § 1533(b)(3)(B). Findings that a petitioned action is not warranted or is "warranted but precluded" are subject to judicial review. § 1533(b)(3)(C)(ii).
The listing of a species is a significant event, for it triggers a series of protective measures designed to restore the species to a healthy population level. Such protections include the designation and acquisition of critical habitat, prohibitions on killing or harming, and mandatory consultations by other federal agencies to ensure that the species is not being harmed by an agency action. §§ 1533(b)(6)(C), 1534, 1538(a), 1536.
FWS is also required to establish a formal recovery plan for every listed species. § 1533(f). In sum, these protections make the ESA a profoundly important (and controversial) statute. In the words of the Supreme Court itself, the ESA is "the most comprehensive legislation for the preservation of endangered species ever enacted by any nation." Tennessee Valley Authority v. Hill, 437 U.S. 153, 180 (1978).
Tongass National Forest and the TLMP
The Tongass National Forest ("TNF") contains roughly 16.9 million acres, covering over 85 percent of southeast Alaska. A.R.II.039 at 1. Of this total acreage, 59 percent is classified as forested land (at least 10 percent tree cover). Id. Nearly 30 percent, or 5.05 million, of the 16.9 million acres is classified as productive old-growth forest, the goshawk's preferred habitat. Id. at 5.
The commercial harvest of productive old-growth forest in southeast Alaska has occurred since the late 1800's and has increased in intensity over the last several decades. Between 1909 and 1995, about 450,000 acres of productive old-growth forest (or 7% of the total) was harvested in TNF, with the vast majority of this amount harvested since 1954, when industrial-scale logging was introduced. Id. at 7. Another 450,000 acres outside TNF have also been harvested, leading to a total loss of 900,000 acres of productive old-growth forests in southeast Alaska. Id. at 66. The loss of this habitat is assumed to be limiting the goshawk's population levels. Id. at 66.
TNF, like all national forests, is subject to the National Forest Management Act, 16 U.S.C.A. § 1600 et seq. (2000), which requires each national forest to promulgate a Land Management Plan. This plan outlines how the various and often competing recreational, natural resource protection, and commercial uses of the forest will be balanced. See 16 U.S.C.A. § 1604(a), (b). The plan must be updated periodically, and at least once every 15 years. § 1604(f)(4), (5). Throughout most of the events leading up to this lawsuit, TNF was governed by the 1979 Tongass Land Management Plan ("TLMP"). In May 1997, the Forest Service issued a revised TLMP.
Overall, the 1997 TLMP allows the continued harvest of over 2.5 million acres of productive old-growth forest in TNF. A.R.II.097(a) at 3, Table 1. Projections of future cuts show that another 8% of TNF's original productive old-growth forest will be harvested by 2055. A.R.III.A.07 at 36. At the same time, when combined with other federal and state protections, the TLMP prohibits logging on 75% of the original productive old-growth forest in southeast Alaska. A.R.III.B.011 at 31.
The 1997 TLMP adds several new mechanisms for protecting a range of species that thrive in productive old-growth habitat, including the goshawk. See A.R.II.096, 097(a). The plan calls for a network of small, medium and large old-growth reserves in which all timber harvesting is prohibited. A.R.II.096 at 2-60. Although the reserves were not established with only the goshawk in mind, they are intended to benefit this subspecies by allowing productive old-growth forests to mature and remain intact. The non-reserve areas, known as the "matrix," are subject to a variety of timber uses and protections. For the most part, the plan retains the 100-year rotation schedule for harvested matrix land. A.R.II.096 at 2-66. The TLMP also establishes a protection zone of 100 acres around each known goshawk nest in the matrix, far below the 300-600-acre recommendation in the Assessment. A.R.II.096 at 4-91. Finally, certain harvest restrictions apply in matrix lands that are adjacent to rivers, marine shorelines, and scenic areas.
At the heart of the present dispute is whether the protections embedded in the 1997 TLMP are adequate to prevent the goshawk from becoming extinct. Plaintiffs contend that even with the reserve and matrix protections, the goshawk is still quite vulnerable. FWS, in contrast, asserts that the added protections are sufficient to ensure the goshawk's persistence.
Timber Harvest and Goshawk Protection Efforts in British Columbia
Roughly half of the goshawk's range is located in the 37.25 million acres of coastal British Columbia. Thus, timber harvesting and goshawk protection programs were considered in the revision of the TLMP and in FWS's listing decision. Unfortunately, information on goshawk ecology in British Columbia is even scarcer than for southeast Alaska. As in southeast Alaska, it is unknown how many goshawks exist in British Columbia, although survey efforts suggest that densities are low. A.R.III.A.07 at 25. Intensive searches on Vancouver Island from 1994 to 1996, for example, turned up only 19 nests. Id. at 25.
In addition, data on timber management in British Columbia is more limited than for southeast Alaska. Instead of the detailed projections of future timber harvests that were available for TNF, FWS was forced to rely on general description of a management strategy. A.R.III.A.07 at 43. This strategy, like TNF's 1997 TLMP, incorporates both a reserve system and a matrix. While information from British Columbia authorities was slow in arriving, FWS eventually concluded that the combination of reserve and matrix lands would protect 64% of the total original productive forest on the Queen Charlotte Islands. A.R.III.B.11 at 25.
Vancouver Island, in turn, has been logged to a much greater extent. Only 36% of the original productive forest is expected to remain protected from timber harvesting. A.R.III.B.11 at 25. This figure is all the more alarming, considering that Vancouver Island contains 79% of the original productive old-growth forest on insular (non-mainland) British Columbia. A.R.III.A.07 at 49. In other words, compared to the Queen Charlotte Islands, Vancouver Island has more than twice as much productive forest, and these forests have been logged at a much greater proportion.
Plaintiffs first petitioned FWS to list the goshawk as an endangered or threatened species in May 1994. 1st A.R.III.A (administrative record from previous case). FWS issued a "not warranted" finding in May 1995. 60 Fed. Reg. 33784 (June 29, 1995). Although FWS found the protections for the goshawk inadequate under the then-applicable 1979 TLMP, FWS based the "not warranted" decision on increased protection measures expected to be implemented under the revised TLMP, that was currently being drafted. 1st A.R.I.A. Plaintiffs challenged the decision and this Court granted summary judgment for plaintiffs on September 25, 1996, holding that FWS could not rely on "possible future actions of the Forest Service to provide sanctuary for the goshawk." Southwest Center for Biological Diversity v. Babbitt, 939 F.Supp. 49, 52 (D.D.C. 1996). The case was remanded to FWS with instructions to make a listing decision based on the then-current 1979 TLMP plan.
In November 1996, the Forest Service, FWS and Alaska Department of Fish and Game ("ADFG") issued a "Conservation Assessment" for the Q.C. goshawk. See A.R. II.039. The Conservation Assessment was crafted to aid the Forest Service in establishing sufficient goshawk protection measures in its revised TLMP. Among other findings, the Conservation Assessment advised that an expanded reserve system and a lengthier timber harvest rotation schedule would provide more goshawk habitat and lessen the risk that the subspecies would become extinct. A.R.II.039 at 83. In addition, the Assessment recommended a nest protection zone of 300-600 acres. Id. at 54.
In May 1997, as FWS was approaching the new listing decision deadline in light of the remand from the District Court, the Forest Service was nearing the end of its years-long TLMP revision. FWS therefore requested and received the permission of the court to extend the listing deadline a few months in order to base the decision on the newly completed 1997 TLMP, rather than the obsolete 1979 TLMP. FWS also re-opened the comment period on whether the Q.C. goshawk should be listed. 62 Fed. Reg. 32070 (June 12, 1997); A.R.III.C.02. Plaintiffs and other groups submitted comments in support of an ESA listing, asserting that the protections in the 1997 TLMP were inadequate to protect the goshawk from extinction. See A.R.I.B.201, 202, 203. In July 1997, FWS convened a panel of five goshawk experts to advise it on whether the goshawk meets the listing criteria for "endangered" or "threatened" status in light of the new TLMP. See A.R.III.B.009. First, the panelists predicted the likelihood that certain threats to the goshawk, such as reduction in foraging success or increase in predation, would occur. A.R.III.B.009 at 31. They then engaged in a sophisticated discussion of what definition of "endangered" to use and arrived at a methodology based on criteria set forth by Georgina Mace and Russell Lande and adopted by the International Union for the Conservation of Nature. Id. at 34-36. They determined that the goshawk would be considered endangered if there was a 20% chance that it would become extinct over the next 30 years and would be considered threatened if at any point in the next 100 years, there was a 20% chance that it would become extinct in the following 30 years. Id. at 36, 40. Using a rigorous scoring system, the panel decisively voted that the goshawk failed to meet the criteria for either endangered or threatened listing based on its status in southeast Alaska. Id. at 40-41; A.R.III.B.007, 008. The panel also voted on the status of the goshawk in British Columbia, although they were frustrated by the lack of reliable information. A.R.III.B.009 at 41-42. The vote scores, when aggregated, indicated nearly an exact toss-up as to whether the goshawk was endangered in British Columbia, with an average of 51% in favor of endangered. Id. at 41-42.
On August 28, 1997, FWS, issued a "not warranted" finding on plaintiffs' listing petition. A.R.III.B.011. In particular, FWS concluded, that because the 1997 TLMP and related measures prohibited logging across 75% of the original productive old-growth forest, there would be adequate habitat "to ensure that goshawks will persist in well-distributed local populations in southeast Alaska." Id. at 31. It concluded that the goshawk does not face the "present or threatened destruction, modification or curtailment of [the species'] habitat or range," one of the criteria for listing under the ESA. § 1553(a)(2).
Plaintiffs initially filed the present action challenging the 1997 "not warranted" decision in April 1998 and moved for summary judgment in January 1999. In July 1999, Judge Sporkin "preliminarily" ruled that plaintiffs had met their burden of showing that defendants "have not fully complied with their statutory duties in determining whether this subspecies of goshawks is endangered or threatened." Southwest Center for Biological Diversity v. Babbitt, Civ. No. 98-934, Order (D.D.C. July 20, 1999).
But instead of reversing outright, Judge Sporkin, frustrated by the lack of scientific information regarding goshawk population trends, remanded the case with instructions for FWS to conduct a population count. Id.
The government appealed, arguing that a district court could not order it to compile additional information. In June 2000, the Court of Appeals agreed, finding that Judge Sporkin exceeded his authority under the ESA and could not order FWS to gather more data. Southwest Center for Biological Diversity, et al. v. Babbitt, 215 F.3d 58, 60 (D.C. Cir. 2000). The case was thus remanded to the District Court for another review of the parties' initial cross-motions for summary judgment.
Upon remand, the case was reassigned from Judge Sporkin to Judge Urbina, who subsequently allowed the parties to submit supplemental memoranda. On February 8, 2002, Judge Urbina referred the cross-motions to me for a Report and Recommendation.
Plaintiffs challenge FWS's "not warranted" decision based on two of the ESA's statutory factors. First, they argue that the goshawk should be listed as either threatened or endangered based on its status in southeast Alaska. Second, they contend that, because the goshawk is threatened in British Columbia as a whole and on Vancouver Island in particular, both "significant portions of its range," the subspecies must be listed. § 1532.
Standard of Review
Plaintiffs bring this suit pursuant to the ESA's citizen suit provision, § 1540(g). The FWS' actions pursuant to the ESA are reviewed under the Administrative Procedures Act ("APA"). See Bennett v. Spear, 520 U.S. 154, 157 (1997). Under the applicable APA provision, a reviewing court must strike down an agency action only if it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C.A. § 706(2)(A) (1996). Although the court must make a "thorough, probing, in-depth review," Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 415-16 (1971), the agency's decision must be upheld if it has "considered the relevant factors and articulated a rational connection between the facts found and the choice made." Baltimore Gas & Elec. Co. v. Natural Resources Defense Council, 462 U.S. 87, 105 (1983); see also Small Refiner Lead Phase-Down Task Force v. EPA, 705 F.2d 506, 521 (D.C. Cir. 1983). Furthermore, courts should defer to the agency's expertise on highly technical matters. Marsh v. Oregon Natural Resource Council, 490 U.S. 360, 378 (1989); New York v. Reilly, 969 F.2d 1147, 1150-51 (D.C. Cir. 1992).
Motion to Strike Extra-Record Documents
Defendants filed a motion to strike a certain document that did not appear in the administrative record. *fn5 Plaintiffs seek to supplement the administrative record with a declaration by Russell Lande ("Lande"), a biologist and co-author of an influential paper on listing criteria for species in peril. Lande and Georgina Mace published a paper in 1991 that sets forth an objective means of classifying an endangered species. See A.R.II.055. Their methodology was adopted, with modifications, by the World Conservation Union in its 1996 "Red Book," an internationally recognized authority on listing criteria. Essentially, Mace and Lande proposed that, in order to eliminate subjectivity from a listing decision, objective and quantitative data be plugged into a formula, which would then arrive at an objective probability of extinction. If a species had a greater than 20% probability of becoming extinct within 20 years or 10 generations (whichever is longer), it should be listed as endangered. A.R.II.055. Mace and Lande also included the caveat that this formula would only be applicable where the listing entity had enough input data on the species' population demographics. In Lande's declaration that plaintiffs seek to submit, he accuses FWS of misapplying his methodology because there was not enough quantitative data on the goshawk.
In general, a review of agency action under the APA is limited to the administrative record before the agency. See Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 420 (1971). However, this Circuit has recognized, albeit in dicta, several exceptions to this principle. In Esch v. Yeutter, 876 F.2d 976, 991 (D.C. Cir. 1991), the court listed eight exceptions where supplementation to the record would be permitted. In particular, the court noted that when "a case is so complex that a court needs more evidence to enable it to understand the issues clearly," extra-record information would be permitted. Id.; see also Beach Communications, Inc. v. F.C.C., 959 F.2d 975, 987 (D.C. Cir. 1992). Another exception would apply where "when the agency failed to consider factors which are relevant to its final decision." Esch, 876 F.2d at 991.
Although these exceptions inevitably lend themselves to varying interpretations, it would be difficult to pretend that Lande's declaration does not exist, for it purports to undermine key evidence supporting FWS' "not warranted" decision. In a case directly on point, the court permitted the supplementation of a declaration by the author of an article that explicitly disclaims the FWS's optimistic reading of his article. Carlton v. Babbitt, 26 F.Supp.2d 102, 108 (D.D.C. 1998). Moreover, there is nothing to suggest that plaintiffs were acting in bad faith by delaying Lande's declaration until after the close of the administrative record, for the first hint of the use of the 20% standard is found in FWS' final action, the not warranted decision itself. The administrative record closed upon the issuance of this final action.
In the end, the addition of Lande's declaration is a nonstarter, for it cannot be said to taint FWS' final decision. Lande accuses the agency of misusing the 20% standard in a situation where objective quantitative data on goshawk demography was lacking. Lande Declaration ¶¶ 10-13. In truth, it was not FWS that adopted the 20% standard, but the panel of experts who participated in the July 1997 discussion. Indeed, notes from this discussion reflect that the panelists spent several minutes deliberating what standard to use in defining endangered and threatened. A.R.III.B.009 at 34-36.
Lande also accuses FWS of ignoring the other criteria from the Red Book, but a closer look suggests that the panelists considered and rejected using these criteria. Panelist Mark Fuller, for example, stated (as paraphrased in the notes), "Red Book criteria are not as likely to put an organism in the more critical categories as we are likely to be." Id. at 35. FWS can hardly be accused of misapplying Lande's methodology when an independent panel of experts made a well-informed and sophisticated decision as to what definition should be used. Similarly, this court is hardly in the position to second- guess those experts, especially in light of Marsh v. Oregon Natural Resource Council, 490 U.S. at 378.
Best Scientific Data Available
Before turning to an analysis of whether the Q.C. goshawk should be listed under the ESA, it behooves me to say a word or two on the meaning of the statute's "best scientific . . . . data available" requirement. § 1533(b)(1)(A). This requirement has already played prominently in this lawsuit, and it appears the parties are still not entirely in accord as to its implications.
To reiterate, the ESA requires that "The Secretary shall make [listing] determinations . . . . solely on the basis of the best scientific and commercial data available to him . . . . " § 1533(b)(1)(A). Case law interpreting the provision suggests that the language prevents FWS from manipulating its analysis by unreasonably relying on certain sources to the exclusion of others. In Las Vegas v. Lujan, 891 F.2d 927 (D.C. Cir. 1989), this Circuit held that the agency may not disregard "scientifically superior evidence." Id. at 933. See also Southwest Center for Biological Diversity v. Babbitt, 926 F.Supp. 920, 927 (D. Ariz. 1996) (agency's unexplained reliance on earlier data while ignoring more recent data violated § 1533(b)(1)(A)). On the other hand, the requirement does not mean that relatively minor flaws in scientific data render that information unreliable. Building Industry Ass'n of Superior California v. Norton, 247 F.3d 1241, 1246-47 (D.C. Cir. 2001). In recognizing that scientific studies are often incomplete and open to challenge, the D.C. Circuit emphasized that § 1533(b)(1) requires FWS to utilize the "best scientific ... data available," not the best scientific data possible. Id. at 1246 (emphasis in original).
Another implication of "best scientific data available" requirement is that FWS must rely on even inconclusive or uncertain information if that is the best available at the time of the listing decision. This point was made abundantly clear by the Court of Appeals' opinion reversing Judge Sporkin's order requiring FWS to conduct another goshawk population count. Southwest Center for Biological Diversity, et al. v. Babbitt, 215 F.3d 58 (D.C. Cir. 2000). The Court, relying on Las Vegas, 891 F.2d at 933, noted that the § 1533(b)(1)(A) requirement prevented a court from ordering FWS to compile new information, "even if the available scientific and commercial data were quite inconclusive." Southwest Center for Biological Diversity, 215 F.3d at 60.
Plaintiffs here have criticized FWS for rejecting evidence of the goshawk's decline as "inconclusive." Indeed, FWS may not insist upon conclusive scientific evidence in order to list a species. Defenders of Wildlife v. Babbitt, 958 F.Supp. 670, 680 (D.D.C.,1997). At the same time, this does not mean that whenever evidence is less than fully conclusive, a listing is automatically warranted. Evidence can be inconclusive and yet lean in favor of an endangered status. Conversely, evidence might strongly suggest that a species is not endangered or threatened, yet still be considered inconclusive or uncertain from a scientist's perspective. The underlying scientific evidence regarding goshawk ecology and population numbers is by all accounts riddled with uncertainty. And yet a panel of scientists, in the face of this uncertainty, nevertheless was able to make very certain conclusions that the goshawk is neither threatened nor endangered in southeast Alaska. See infra, discussion of July 1997 experts' panel.
Plaintiffs next contend that FWS improperly relied on a speculative regulatory mechanism in basing its listing decision on the anticipated effects of the 1997 TLMP on the goshawk. The ESA provides that FWS may rely only upon "existing regulatory mechanisms" in its listing determinations. § 1533(a)(1)(D). In contrast, future and uncertain actions cannot justify a negative listing decision. Biodiversity Legal Found. v. Babbitt, 943 F.Supp. 23, 26 (D.D.C. 1996); Oregon Natural Resources Council v. Daley, 6 F.Supp.2d 1139, 1153-54 (D. Or. 1998). Indeed, this was the very basis of the District Court's initial decision in this case to remand the May 1995 "not warranted" finding based on the still-incomplete 1997 TLMP. See Southwest Center for Biological Diversity v. Babbitt, 939 F.Supp. at 52.
Plaintiffs now protest that even though the 1997 TLMP is complete and in place, it is still speculative because its regulatory mechanisms may not be effective in protecting the goshawk from extinction. But no regulatory mechanism would appear to meet plaintiffs' high bar, for there is always the possibility that objectives will not be achieved. The operative question, rather, is whether a set of regulations is concrete and specific enough to ensure that it will in fact be implemented. The 1997 TLMP, despite its built-in flexibility, clearly meets this standard. Furthermore, plaintiffs' comparison to Federation of Fly Fishers, et al. v. Daley, No. C 99-0981 (N.D. Cal. Oct. 24, 2000), is inapposite because the regulatory mechanisms at issue in that case were to be implemented in the future. Here, however, the 1997 TLMP took effect immediately upon its passage.
Finally, plaintiffs' complaint that defendants' failure to consider the ADFG's 1997 Field Study was a material omission is not convincing. The material at issue was an annual progress report on the status of the goshawk in the TNF. Memorandum in Support of Plaintiffs' Motion for Summary Judgment, Exh. B ("ADFG's 1997 Field Study"). Plaintiffs' contention that the report presents new and alarming data confirming the dire state of the goshawk is simply not borne out by the record. The authors of the study reported finding 10 active goshawk nests in 1996, which was well below the numbers reported from two years earlier. Id. at 7. But the report begins by explaining that for a variety of reasons, fewer research hours were spent in the field attempting to locate goshawk nests. Id. at 2. It would be surprising indeed if the researchers had not found fewer nests, given the sharply reduced search effort.
Listing Based on Status in Southeast Alaska
Plaintiffs argue that the dire state of the goshawk in southeast Alaska due to timber harvesting in TNF merits immediate listing. As is common in listing decisions, FWS had to wade through a large quantity of scientific information gathered over several decades in order to determine whether the goshawk merited listing. The administrative record fills three large boxes and includes 110 references, most of which are scientific studies. Amidst these mounds of paper, two sources were particularly useful in arriving at the listing determination: a 1996 Conservation Assessment authored by a group of Forest Service, FWS and ADFG biologists and the conclusions of five goshawk experts specifically convened to advise FWS in its listing determination. A.R.II.039; A.R.III.B.009. *fn6
In conjunction with its revision of the 1979 TLMP, the Forest Service established an interagency Viable Population Committee to determine whether certain old-growth-forest-dependent species required heightened protections in the Forest Plan in order to remain viable in the Tongass. The Committee's 1993 report concluded that several species, including the goshawk, were in danger of extinction unless large tracts of habitat were preserved. 1st A.R.III.B.021. The Committee also recommended a nest protection zone of 1600 acres for goshawk nests outside of reserves. Id.
The Conservation Assessment is a 101-page report that synthesized all available scientific research to arrive at an overview of the subspecies' demographic trends. The report cites no less than 179 references, including many papers written by the various authors of the Assessment. After reviewing studies of the goshawk's ecology in North America as a whole and in southeast Alaska in particular, as well as the state of timber harvesting operations in TNF under the 1979 TLMP, the Conservation Assessment reached three main conclusions regarding the goshawk's status in southeast Alaska:
The first is that the probability of persistence for goshawks throughout southeast Alaska has declined since the middle of the 20th century. Secondly, although persistence may be in immediate peril in specific areas with highly modified landscapes (see "Management Considerations," "Risk Assessment," below) goshawks in most ecological provinces with limited or no habitat modification are likely not in immediate peril. Thirdly, we concluded that a sound habitat management strategy is important to maintain long-term, well-distributed populations. A.R.II.039 at 67(emphasis in original).
These conclusions, read together, suggest that the goshawk is neither threatened or endangered.
It is important to keep in mind that the Conservation Assessment was completed prior to the publication of the 1997 TLMP. Thus, it cannot be read to settle the precise question at issue before us: whether, in light of the 1997 TLMP protection measures, the goshawk is endangered or threatened.
On the other hand, given that the 1979 TLMP goshawk protections were much weaker than those in the 1997 TLMP, the Assessment's conclusion that the goshawk is not in immediate peril under the 1979 TLMP can only be strengthened in the wake of the added 1997 TLMP protections.
The other significant scientific data relied upon by FWS were the opinions of the expert panel that convened for two full days in July 1997. Five biologists with direct experience working with the Q.C. goshawk were asked to vote on the precise questions of whether, in light of the revised 1997 TLMP, the subspecies was endangered or threatened in southeast Alaska. The vote on the endangered status (whether there was a 20% chance that the species would become extinct in southeast Alaska within 30 years) was nearly unanimous, with four experts voting that there was a 0% chance of extinction and one voting that there was a 10% chance. A.R.III.B.007. The votes on whether the goshawk is threatened (if at any point in the next 100 years, there is a 20% chance that the species would become extinct in 30 years) were also decisive, with the five panelists turning in respective votes of between 0%-30% and 9%-37% using two slightly different scoring methods.
A.R.III.B.008. These votes reflect the panelists views, expressed explicitly in the notes of the meeting, that even if all of the productive old-growth forest in British Columbia were harvested, the goshawk would still persist in southeast Alaska. A.R.III.B.009 at 40, 42.
A central assumption underlying FWS' not warranted decision is that the persistence of the goshawk is related to the amount of productive old growth forest that remains intact. *fn7 Thus, according to FWS, the most important aspect of the 1997 TLMP is the amount of productive old growth forest that is protected from timber harvesting. Because the range of protective mechanisms in the reserves and the matrix would preserve 75% of the original productive old growth forest in southeast Alaska, FWS concludes that the goshawk is not likely to become extinct in southeast Alaska. FWS A.R.III.B.011 at 31.
The protection-persistence assumption does not appear to be the result of unfounded optimism or willful ignorance. Rather, it appears that the panel of experts relied on the very same assumption in reaching its nearly unanimous conclusions that the goshawk is neither endangered or threatened in southeast Alaska. For example, one panelist noted that in southeast Alaska, "there are enough remote spots where it just isn't feasible to go in and harvest trees, that it would be impossible to drive goshawks to extinction." A.R.B.III.009 at 42. Moreover, the Conservation Assessment, the most definitive summary of goshawk ecology to date, expressly concluded that "goshawks in most ecological provinces with limited or no habitat modification are likely not in immediate peril." A.R.II.039 at 67.
Plaintiffs challenge the protection-persistence assumption as overly simplistic, noting that because goshawk densities in wilderness and roadless areas of TNF (areas that generally have not been logged and are therefore analogous to the incipient reserve network) are very low. In particular plaintiffs point to the introductory section of the 1997 ADFG Field Report, which cites to a 1995 study by P.F. Schempf finding a goshawk nest density of 1.49 nests per 100 square kilometers in TNF roadless and wilderness areas, much lower than densities of the Northern goshawk in the continental United States. 1997 Field Report at 1. The Field Report summarizes this study to conclude that there is "no evidence to support speculation that wilderness and roadless areas of the Tongass National Forest support a significant reservoir of undocumented goshawk breeding territories that could buffer loss of nesting habitat elsewhere in Southeast Alaska." Id. at 1.
Plaintiffs' assertion that the Schempf study contradicts the protection-persistence assumption is not cogent. It is entirely consistent to say that, even though goshawk densities are quite low in productive old growth forests, the preservation of enough such forest will be sufficient to ensure the persistence of the goshawk. A species may exist in very low levels and yet still remain safe from extinction. In fact, several species occur in naturally low population densities that have nothing to do with human interference. Such species, merely by virtue of their rarity, do not merit listing under the ESA. See Cook Inlet Beluga Whale v. Daley, 156 F.Supp.2d 16 (D.D.C. 2001).
Nevertheless, plaintiffs argue that the range of protections set forth in the 1997 TLMP are inadequate to prevent the goshawk from becoming extinct. Among other things, they point to the insufficiency of the 100-acre nesting zone and the large amount of productive old-growth forest that will remain open to harvest. No doubt, each of plaintiffs' criticisms are valid, insofar as larger nest protection zones and reserves would certainly help the goshawk. But there remains a large gap between asserting weaknesses in the 1997 TLMP and finding that the goshawk is thereby doomed to extinction.
Plaintiffs focus much of their argument on the goshawk's allegedly low population levels. In particular, they point to a 1993 study indicating that 610 interbreeding pairs of goshawks are necessary to assure long-term genetic viability of the subspecies. 1st A.R.III.B.21 at 127. Emphasizing that goshawk populations may be as low as 100 pairs in southeast Alaska, plaintiffs attempt to portray the subspecies' vulnerability as a "no-brainer." A.R.II.039 at 22 (Conservation Assessment's recognition of various studies estimating population levels anywhere from 100 to 800 pairs in southeast Alaska). There are several flaws with plaintiffs' numbers. First, the 610 figure has never been accepted as a definitive number. In fact, the Conservation Assessment concluded that "We do not know how many goshawks are necessary, or in what spatial distribution they need to occur, to ensure their long-term persistence in southeast Alaska." A.R.II.039 at 80. I will not substitute plaintiffs' judgment over that of the eight scientists who authored the Conservation Assessment on this point. Finally, plaintiffs' numbers compare apples and oranges. The 610 figure was not geographically limited to southeast Alaska, whereas the 100-800 pairs figure was.
Another theme of plaintiffs' argument is that the continued timber harvesting of productive old-growth forest in southeast Alaska has devastated the goshawk. No doubt, there is widespread scientific agreement that the goshawk's numbers are declining in southeast Alaska. But, a declining population alone does not indicate that a species is threatened or endangered. The operative question, rather, is whether a population decline is so extensive or precipitous that the species may no longer survive. None of the evidence, as interpreted by those most qualified to pronounce on the subject, supports this conclusion.
Plaintiffs compare this case to Defenders of Wildlife v. Babbitt, 958 F.Supp. 670 (D.D.C. 1997), wherein Judge Kessler held that FWS acted arbitrarily and capriciously in refusing to issue a proposed rule listing the Canada lynx as endangered or threatened. In that case, however, evidence of the lynx's declining population levels was much more convincing than in the instant case. The administrative record showed a "dramatic" drop in lynx numbers, as well as their complete disappearance from 17 states. Defenders of Wildlife, 958 F.Supp. at 674. In addition, there was near unanimous agreement among FWS biologists that the lynx should be listed. Id. at 676. The agency's decision, moreover, flatly contradicted the conclusions of these biologists and relied on "inaccurate" facts. Id. at 676, 682. None of these facts are at play in this case, making any comparison misguided.
At bottom, plaintiffs' argument is based entirely on a disagreement over the science, namely the soundness of the protection-persistence assumption. They point to no material information that FWS failed to consider. They identify no independent biologist who flatly disagrees with the expert panels' conclusions. For me to agree with plaintiffs' arguments would be to accept their interpretation of the data on this highly technical matter over the unanimous opinion of five goshawk experts. This would be flatly inconsistent with the instruction in Marsh v. Oregon Natural Resource Council, 490 U.S. at 378, to show deference to the agency on technical and scientific conclusions.
Listing Based on Status in British Columbia
Plaintiffs next assert that listing is warranted by virtue of the fact that the Q.C. goshawk is in danger of extinction in British Columbia as a whole, which makes up roughly half of the species' range, and Vancouver Island in particular, which represents one-third of its range. Plaintiffs invoke the ESA's requirement that a species be listed if it is in danger of extinction "throughout all or a significant portion of its range." 16 U.S.C.A. § 1532 (emphasis added). Under plaintiffs' theory, this listing provision is not qualified by any political boundaries. That is, the location of the significant portion in which the species is in peril is not a consideration. According to plaintiffs, the fact that the significant portion of the goshawk's range where it is said to be threatened or endangered is in British Columbia, and thus entirely outside of the United States, is irrelevant for the purposes of the listing determination. *fn8
The jurisdictional issue typically arises when a species is present in both the United States and a neighboring nation (either Mexico or Canada), but is threatened or endangered only in the United States. Such a situation exists with many, if not most, of the large mammals listed in the western United States, including the grizzly bear, the gray wolf, and the Canada lynx. It is well established that even if healthy populations of species are present across the border, a species may still be listed in its historical range in the United States. The instant case appears to be unique in that, while the goshawk inhabits both the United States and Canada, plaintiffs seek a listing based solely on its conservation status in Canada. *fn9
There is little case law that discusses the listing of species found exclusively or partially in foreign nations. *fn10 The closest case would be Defenders of Wildlife v. Norton, 258 F.3d 1136 (9th Cir. 2001), wherein the Ninth Circuit held that "a species can be extinct 'throughout . . . . a significant portion of its range' if there are major geographical area in which it is no longer viable but once was. Those areas need not coincide with national or state political boundaries, although they can." Id. at 1145.
The Ninth Circuit's decision is compatible with certain provisions of the ESA that support its application beyond the geographic borders of the United States. For instance, in the findings and purposes section, the United States pledges itself "as a sovereign state in the international community to conserve to the extent practicable the various species of fish or wildlife and plants facing extinction." § 1531(a)(4). Furthermore, the ESA requires the Secretary of the Interior to consider "those efforts, if any, being made by any State or foreign nation ... to protect species" when making listing decisions. § 1533(b)(1)(A)(emphasis added). Finally and most significantly, the statute requires FWS to give actual notice and invite comment from foreign nations in which species proposed to be listed as endangered are found. § 1533(b)(5)(B).
In addition to the statute and case law, FWS itself has interpreted the ESA to allow listing of species that exist exclusively in foreign nations. In fact, FWS in the past has listed several foreign species that do not inhabit the United States at all. A 1984 listing, for example, added several foreign species to the list, including the giant panda. 49 Fed. Reg. 2779 (Jan. 23, 1984).
As a policy matter, there appear to be several practical reasons for listing a foreign species. These rationales were articulated quite clearly by FWS itself in the 1984 listing:
Conservation measures available to foreign species listed as endangered or threatened include the following:
(1) worldwide attention is called to their problems which may result in international efforts to prevent their further decline.
(2) U.S. expertise could be made available (if requested by resident country) to assist in development of management or conservation programs.
(3) limited U.S. funds could be made available (if requested by resident country) for development of management or conservation programs.
(4) the U.S. would strictly regulate import and export, and commercial U.S. trade in these species, thus assuring that any of these activities by persons subject to the jurisdiction of the U.S. do not jeopardize these [animals]. 49 Fed. Reg. at 2782.
Although the 1984 listing dealt with species that occur only in foreign nations, and not in the United States, the same policy reasons apply equally, if not more so, to species that occur both within the United States and foreign nations, but are only threatened or endangered in those foreign nations.
In its "not warranted" decision, FWS considered the goshawk's respective statuses in the Queen Charlotte Islands and Vancouver Island. A.R.III.B.011 at 31. FWS concluded that the species would persist in the Queen Charlotte Islands because 64% of the original productive old-growth forest would be protected from logging. *fn11 Even in light of the increased uncertainty of the British Columbia timber management information, I cannot conclude that FWS' decision that the goshawk is not threatened or endangered on the Queen Charlotte Islands is unreasonable.
Plaintiffs' contention that FWS ignored the findings of the experts panel with respect to the goshawk's status in British Columbia is unconvincing. Plaintiffs assert that the panel found that it is more likely than not that the Q.C. goshawk is threatened by extinction in British Columbia. In fact, the five experts each arrived at a percentage score for a likelihood that the goshawk was threatened with extinction in British Columbia. Two panelists voted that it was more likely than not (voting 66% and 56% likelihood), a third believed it was as likely as not (voting 50%) and two more voted that is was less likely (voting 44% and 40%). A.R.III.B.008 at 7. The average of the five scores was 51%.
Plaintiffs point to this number as evidence of the fact that the expert panel determined that it was more likely than not that the goshawk is threatened because of species' status in British Columbia. In truth, these scores, especially in light of the panelists' overt frustration regarding the lack of reliable information from British Columbia forestry officials, are hardly conclusive enough to render FWS' not warranted decision unreasonable and to mandate a proposed listing.
In contrast, the analysis with respect to Vancouver Island is a bit more complicated. Two relevant issues are present. First, does the island constitute a "significant portion" of the goshawk's range, given that it represents one-third of its geographic range? Second, is the goshawk actually threatened or endangered on Vancouver Island?
In its litigation briefs, FWS contends that the goshawk cannot be listed based on its status in Vancouver Island because that area represents only one-third of the subspecies' geographic range, which is not a "significant portion" of its range. I reject this argument based on an analysis of the term "significant." There is no simple formula or hard and fast case law indicating what constitutes a significant portion of a species' range. On the other hand, FWS has in the past given some hint as to what may qualify. In listing the marbled murrelet, another old-growth raptor, FWS noted without comment that one-third of the species' range was a significant portion. 57 Fed. Reg. 45328, 45330, col. 1 (October 1, 1992); cf. 63 Fed. Reg. 49065, 49074 (seeming to indicate in another instance that 30% of a species' range was not significant).
A more thorough analysis, however, suggests that a flat percentage of geographic area is not the sole determinant of significance. As a general rule, species are not evenly distributed across their ranges, but rather tend to concentrate in certain areas where habitat is particularly suitable. Thus, the percentage of geographic area would not linearly correlate to the percentage of a species' population. One-third of a species' geographic range may be found to contain a disproportionately greater or lesser percentage of the total number of individuals. It does not seem fair or sensible, then, to point to some arbitrary geographic percentage as constituting a "significant" portion of a species' range. In fact, FWS argues as much when it insists that the marbled murrelet decision is irrelevant because it concerned a "different species . . . ., in a different geographic area, under different biological and scientific circumstances . . . ." Def.'s Opp. at 29.
I conclude that Vancouver Island is a significant portion of the goshawk's range because not only does it represent one-third of the species' geographic range, but it also appears to contains a relatively large amount of suitable goshawk habitat, i.e., productive old-growth forest. The island contains 79% of the original productive old-growth forest on insular (non-mainland) British Columbia. A.R.III.A.07 at 49. Furthermore, search efforts on the island during three years in the mid-1990s turned up 19 goshawk nesting areas, which is a significant number, considering that similar searches over six years in TNF have discovered only 43 active nests. A.R.III.A.07 at 25; ADFG 1997 Field Study at 7. Given that Vancouver Island represents a relatively rich one-third of the goshawk's geographic range, this portion is certainly significant within the meaning of the ESA.
It appears that FWS never actually reached a decision on the substantive issue of whether the goshawk is threatened or endangered on Vancouver Island. FWS did find that only 36% of the original productive forest will be protected, and noted the high degree of uncertainty regarding the goshawk's persistence. A.R.III.B.011 at 31. In the end, however, FWS apparently decided that it did not need to make a threatened or endangered conclusion with respect to Vancouver Island. Rather, it simply noted that the goshawk's "range-wide persistence is not expected to be significantly influenced by the population status of the birds on Vancouver Island." Id. at 32. FWS stated that even if the goshawk became extinct on Vancouver Island, it would still remain in sufficient numbers on the Queen Charlotte Islands and in southeast Alaska. *fn12 While this may indeed be the case, it does not answer the question of whether the goshawk is likely to become extinct on Vancouver Island, which remains an independent grounds for listing under the ESA. Although in certain cases, courts may "uphold a decision of less than ideal clarity if the agency's path can be discerned," Motor Vehicles Mfctr. Ass'n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983)(citing Bowman v. Transp. Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 285 (1974)), this principle does not apply here because it is not even clear that the agency has made a decision.
FWS' failure to make a finding on the conservation status of the goshawk on Vancouver Island constitutes a material omission in its "not warranted" decision. I recommend that Judge Urbina remand this case to allow FWS to make a specific finding on whether the goshawk should be listed based on its conservation status on Vancouver Island. *fn13
Because I recommend only a remand to FWS, and not an outright reversal, the questions surrounding the implications of listing based solely on the goshawk's threatened or endangered status on Vancouver Island are not yet ripe. Suffice to say, the legal and policy implications of such a listing are not obvious. It is not clear which provisions of the ESA could or should apply to the species on either Vancouver Island or in southeast Alaska. As a fundamental principle of sovereignty, the United States would have no authority to regulate the species within British Columbia proper. Conversely, the import and export of the species could be regulated in the United States, as indicated by the above-quoted FWS listing notice. 49 Fed. Reg. at 2782. Aside from these relatively incontrovertible propositions, there lie a whole host of ESA provisions that may or may not apply to the goshawk in southeast Alaska, even though it is not considered threatened or endangered in that locale. For example, amicus curiae Ketchikan Pulp Company adamantly opposes the designation of any critical habitat in southeast Alaska if the goshawk is only endangered or threatened in British Columbia. Supplemental Brief of Amicus Curiae Ketchikan Pulp Company, June 4, 1999, at 7-8. In any event, it appears that such questions are not before the court at this time, and will only become ripe if and when FWS does proceed with a listing of the goshawk.
Failure to file timely objections to the findings and recommendations set forth in this report may waive your right of appeal from an order of the District Court adopting such findings and recommendations. See Thomas v. Arn, 474 U.S. 140 (1985).