Carlton v. Babbitt

OPINION PAUL L. FRIEDMAN, District Judge. This case is back before the Court on two related matters: (1) the plaintiffs’ motion to supplement the administrative record and (2) the plaintiffs’ challenge to the defendants’ decision on remand not to reclassify the Selkirk population of grizzly bears from “threatened” to “endangered” under the Endangered Species Act (“ESA”), 16 U.S.C. §§ 1531 et seq. In 1995, this Court found that the government’s refusal to reclassify the Selkirk population as “endangered” was arbitrary and capricious and not supported by the evidence in the record. Carlton v. Babbitt, 900 F.Supp. 526, 531-34, 537-38 (D.D.C.1995). It remanded the matter to the United States Fish and Wildlife Service (“FWS”) for further analysis. Id. The FWS has made supplemental findings and has once again found that reclassification of the Selkirk population is not warranted. Supplementary Information for the Court Regarding the Not Warranted Petition Finding for the Selkirk Grizzly Bear Population (Mar. 15, 1996) (“Supplementary Information”) at 11. Upon review of the administrative record and the Supplementary Information, as well as a number of the documents submitted by plaintiffs, the Court again remands this matter to the FWS because it has not established (1) that the Selkirk population can sustain the current rate of human-caused mortality, (2) that the present regulatory mechanisms are adequate, (3) that the Selkirk population is not endangered simply — 105 — by virtue of its size, and (4) that Canadian habitat will continue to be available to the Selkirk population. I. BACKGROUND Once a dominant predator in the Western United States, the grizzly bear has been all but eliminated from its historic range. Between 1800 and 1975, the grizzly bear population in the lower 48 states shrank from an estimated 50,000 bears to fewer than 1000. U.S. Fish And Wildlife Service, Grizzly Bear Recovery Plan (1997) (“Plan”) at 9, Supplemental Administrative Record (“Supp. AR.”) Tab 18. As a result, the Secretary of the Interior found in 1975 that the grizzly bear was likely to become in danger of extinction within the foreseeable future. Under the authority given him by the ESA, he listed the grizzly bear in the lower 48 states as a “threatened” species. 40 Fed.Reg. 31,-734 (1975). In implementing the ESA the FWS has the discretion to define distinct, disconnected subpopulations of a protected species that are “endangered” or “threatened” in their own right. 16 U.S.C. § 1532(16); 50 C.F.R. § 424.11(k) (1997). The FWS recognizes five such subpopulations for the grizzly bears of the lower 48 states, including the populations for the Selkirk ecosystem and the Cabi-nei/Yaak ecosystem. Plan at 11-13. In January and February 1991, plaintiffs requested that the FWS reclassify the Selkirk and Cabinet/Yaak populations from “threatened” to “endangered.” 57 Fed.Reg. 14,372 (1991), Administrative Record (“A.R.”) Tab 15. A “threatened” species is one that is “likely to become an endangered species within the foreseeable future,” while an “endangered” species is one that “is in danger of extinction.” 16 U.S.C. § 1532(20); 16 U.S.C. § 1532(6). An “endangered” species receives more protection under the ESA than a “threatened” species. In April 1992, the FWS responded to plaintiffs’ petitions, finding that there was sufficient information to initiate a formal review of the status of the two populations. 57 Fed.Reg. 14372 (1992). In the subsequent review, however, the FWS chose not to reclassify either population from “threatened” to “endangered.” For the Selkirk population, the FWS determined that reclassification was not warranted because “human-caused mortality is decreasing, reproduction and survivorship rates are adequate ..., and a proactive management program has been implemented.” 58 Fed.Reg. 8250, 8251 (1993) (citations omitted). For the Cabinet/Yaak population, the FWS acknowledged that reclassification was warranted but found that it was precluded “by work on other species having a higher priority for listing.” Id. After providing the government with the proper notice, plaintiffs filed suit in this Court seeking review of the FWS determination regarding both populations. The Court found that both decisions of the FWS had been arbitrary and capricious and remanded the matter to the FWS, which was directed to provide a reasoned analysis for its decisions. Carlton v. Babbitt, 900 F.Supp. at 537-38. In reviewing the Selkirk determination, the Court applied the factors that the FWS was required to review in making listing decisions. Specifically, the factors are: (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. 16 U.S.C. § 1533(a)(1); 50 C.F.R. § 424.11(c) (1997). The Court found that (1) the FWS conclusion that human-caused mortality is decreasing was not supported by the record, (2) the FWS determination that the regulatory mechanisms were adequate because mortality had decreased was similarly flawed, and (3) the FWS inadequately considered the susceptibility of the Selkirk population to certain threats by virtue of its small size. Carlton v. Babbitt, 900 F.Supp. at 532-34. The Court also requested the FWS to give further consideration to its reliance on the — 106 — contiguity of Canadian grizzly bear populations to ensure the Selkirk population’s survival. Id. at 584. For the Cabinet-Yaak population, the Court acknowledged that the FWS had the discretion to make certain listing determinations a priority over others but found that it had not proffered sufficient evidence that it was actively working on other listings. Carlton v. Babbitt, 900 F.Supp. at 536. In reviewing a “warranted but precluded” finding, the Court was required to determine whether the FWS had shown that it was working on other pending proposals and that it was making expeditious progress. Id. Because the FWS had not produced any evidence in this regard, the Court remanded the matter to the FWS to address the insufficiency. In its supplemental findings, the FWS has retained its finding that the reclassification of the Selkirk population was unwarranted but has revised its justification for that finding. Instead of maintaining that human-caused mortality was decreasing, the FWS determined that the population was stable and could sustain the current level of human-caused mortality. Supplementary Information at 3-4. In particular, the FWS responded to the Court’s concerns by finding that (1) the Selkirk population could sustain the current annual human-caused mortality rate of four percent, (2) the number of regulatory initiatives in the United States and Canada are adequate to forestall the population’s decline, (3) the Selkirk population’s size at 26 to 36 bears is sustainable, and (4) Canadian policies and cooperation justify reliance by the FWS on the Canadian portion of the Selkirk population in evaluating its health. Id. at 2-11. For the Cabineb-Yaak population, the FWS merely restated its earlier findings. Plaintiffs now challenge the supplemental findings made by the FWS with respect to the Selkirk population and request the Court to find that the supplemental findings are arbitrary and capricious and in violation of the ESA. To support their arguments, plaintiffs also have moved to supplement the administrative record with a number of additional documents. Plaintiffs do not pursue the Cabinet-Yaak issue. See Plaintiffs’ Response to Defendant’s Notice of Filing at 1 n. 1. II. DISCUSSION A Standard of Review Agency decisions under the ESA are reviewed under the Administrative Procedure Act. See Las Vegas v. Lujan, 891 F.2d 927, 932 (D.C.Cir.1989); Cabinet Mountains Wilderness v. Peterson, 685 F.2d 678, 685 (D.C.Cir.1982). The reviewing court may set aside such agency actions, findings or conclusions when they are arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law. 5 U.S.C. §§ 706(2)(A). There is a strong presumption in favor of upholding decisions of the FWS in view of its expertise in the area of wildlife conservation and management and the deferential standard of review. Carlton v. Babbitt, 900 F.Supp. at 530; see Marsh v. Oregon Natural Resources Council, 490 U.S. 360, 375-78, 109 S.Ct. 1851, 104 L.Ed.2d 377 (1989). Nonetheless, a reviewing court must undertake a “thorough, probing, in-depth review” of the agency’s decision and then decide whether it was “based on a consideration of the relevant factors and whether there has been a clear error of judgment.” Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 415-16, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971). If the agency has “considered the relevant factors and articulated a rational connection between the facts found and the choice made,” its decision must be upheld. Baltimore Gas & Elec. Co. v. Natural Resources Defense Council, Inc., 462 U.S. 87, 105, 103 S.Ct. 2246, 76 L.Ed.2d 437 (1983). B. Supplementation of the Administrative Record Review of agency action “is to be based on the full administrative record that was before the [agency] at the time [it] made [its] decision.” Citizens to Preserve Overton Park v. Volpe, 401 U.S. at 420. In certain limited circumstances, however, a court may consider evidence not included in the administrative record that more fully explicates an agency’s decision. See Beach Communications, Inc. v. FCC, 959 F.2d 975, 987 — 107 — (D.C.Cir.1992); Esch v. Yeutter, 876 F.2d 976, 991 (D.C.Cir.1989); Nat’l Trust for Historic Preservation v. Blanch, 938 F.Supp. 908, 916 n. 10 (D.D.C.1996). Supplementation of the administrative record has been found to be appropriate, inter alia, when “the agency failed to consider factors which are relevant to its final decision” or when “evidence arising after the agency action shows whether the decision was correct or not.” Esch v. Yeutter, 876 F.2d at 991; see also Beach Communications, Inc. v. FCC, 959 F.2d at 987 (quoting Esch v. Yeutter, 876 F.2d at 991) (“Where [the] court ‘needs more evidence to enable it to understand the issues clearly,’ ... we have discretion to supplement an administrative record”) (citations omitted). Here, plaintiffs have asked that the administrative record be supplemented with (1) documents concerning uncounted mortalities among the Selkirk population, (2) documents on the sustainable rate of mortality for grizzly bears, (3) documents concerning the health of the Canadian habitat and grizzly bear populations, and (4) documents concerning the size and stability of the Selkirk population. Sufficient grounds exist for the inclusion of the majority of these materials in the record. The first group of documents consists of public materials that indicate the occurrence of grizzly bear deaths for which the FWS has not accounted. Plaintiffs Motion to Supplement the Administrative Record, Exhs. 2-9 (“Supp.Exhs.”).1 All of these documents are relevant because the number of known grizzly bear mortalities is essential to the FWS findings of sustainability. Because all of these documents were publicly available at the time the FWS compiled its mortality statistics, and all but two were official records from court proceedings, they should have been considered by the agency. Particularly in view of the small size of the Selkirk populations, they should be included in the record. Cf. Mack v. South Bay Beer Distributors, Inc., 798 F.2d 1279, 1282 (9th Cir.1986), abrogated on other grounds by Astoria Federal Sav. and Loan Ass’n v. Solimino, 501 U.S. 104, 111 S.Ct. 2166, 115 L.Ed.2d 96 (1991) (“[A] court may take judicial notice of ‘records and reports of administrative bodies’ ”) (citations omitted). The second group of documents consists of three academic writings: an article by Richard B. Harris on the sustainable rate of human-caused mortality for grizzly bears, Mr. Harris’ 1984 master’s thesis that formed the basis of the article, and a 1995 article by Professor Daniel F. Doak on the “lag time” between critical levels of habitat degradation and any detectable changes in population sizes. Supp. Exhs. 10-12. Since the FWS asserts that Mr. Harris’ article is the basis for its finding that the Selkirk population could sustain a four percent known human-caused mortality rate, the FWS concedes that it should be included in the record. Defendant’s Opposition at 20 n. 13. Because Mr. Harris’ master’s thesis, and not his article, was the justification for the use of the four percent figure in the FWS plan for the management of the grizzly bears, see Plan at 125, it should be included in the record as well. See Esch v. Yeutter, 876 F.2d at 991 (supplementation is appropriate when “an agency considered evidence which it failed to include in the record”). Professor Doak’s article was submitted to the Court to contradict the assertion of the FWS that it can quickly identify any change in the Selkirk population. Plaintiffs have not established that the FWS failed to consider the issue of “lag time,” however, and the article therefore will not be included. See id. The third group of documents concerns the status of the Canadian portion of the Selkirk population. See Supp. Exhs. 13108 19.2 The FWS found that the Canadian habitat was a necessary part of the Selkirk population’s range and relied on this habitat in its assessment of the population’s viability, but it did not analyze possible threats to Canadian habitat. Supplementary Information at 9-11. The documents offered by the plaintiffs are publicly available documents that would have allowed the FWS to conduct this relevant analysis. Since there will be a remand on this issue, the Court need not supplement the record with these documents now, but it assumes that the agency will consider them. Finally, the documents concerning the size and stability of the Selkirk population address an issue that the FWS has explicitly stated that it has investigated. In its supplemental findings, the FWS noted that it explored the concept of minimum viable population and found a range of projections. Supplementary Information at 7. In support of its contention that scientific authority exists for a small minimum viable population, the FWS cites a 1978 study by Dr. Mark L. Shaffer. Id. In response, plaintiffs wish to supplement the record with several documents purportedly demonstrating that a scientific consensus exists that the Selkirk population is too small to be viable. Supp. Exhs. 21-25. The first of the documents in this final group is a declaration by Dr. Shaffer that explicitly disclaims the FWS’s optimistic reading of his article. Since Dr. Shaffer’s article was the only support provided by the FWS to demonstrate that scientific authority exists for a small minimum viable population, his understanding of his own article is particulary relevant and should have been considered in connection with this matter. See Esch v. Yeutter, 876 F.2d at 991. Plaintiffs have not established that the remainder of the articles in the final group should be included, however, because the FWS admits that support exists in the scientific literature for the proposition that the Selkirk population is too small. The FWS therefore has considered this factor. See id. at 991-92. C. The Supplemental Findings In the Court’s 1995 ruling, it ordered the FWS to reconsider its finding that the reclassification of the Selkirk population was not warranted. Specifically, the FWS was required to respond to the Court’s conclusions that (1) the FWS finding that human-caused mortality is decreasing was not supported by the record, (2) the FWS determination that the regulatory mechanisms were adequate because mortality had decreased was similarly flawed, and (S) the FWS inadequately considered the susceptibility of the Selkirk population to certain threats by virtue of its small size. Carlton v. Babbitt, 900 F.Supp. at 532-34. The Court also expressed concern about the reliance by the FWS on contiguous Canadian populations and whether there was a sufficient basis in the record to support the proposition that “grizzly bears in the Selkirk ecosystem actually move between the United States and Canada and thus share a larger ecosystem.” Id. at 534. The Court is not persuaded that the FWS has adequately responded to any of these issues. 1. Human-Caused Mortality In contrast to its earlier contentions, the FWS does not maintain that human-caused mortality is declining. Instead, the FWS has now found that the Selkirk ecosystem had a known human-caused mortality rate of 1.29 bears per year and that “no discernable statistical trend” existed in the rate of this mortality. Supplementary Information at 3. In other words, the FWS found that the mortality rate of 1.29 bears per year did not appear to be rising or falling. On the — 109 — basis of an article by Richard B. Harris, the FWS concluded that the Selkirk population could sustain a four percent annual known human-caused mortality rate. Id. As a four percent rate for the 26 to 36 bear Selkirk population would allow 1.04 to 1.44 annual deaths, the FWS concluded that the 1.29 annual rate was sustainable. Id. This analysis is flawed in at least two respects. First, the FWS ignored a number of deaths within the Selkirk ecosystem of which the government was well aware. Second, the application of the four percent rate to a population as small as the Selkirk population is without any scientific support in the record, including even Mr. Harris’ article. Plaintiffs point out that the FWS had to have known about several mortalities not noted in its supplemental findings. Because the mortality rate for the small Selkirk population will swing significantly with the inclusion of one or two additional deaths, the FWS conclusion might be completely undermined by taking account of these uncounted mortalities.3 The FWS counted the shooting of “Sy,” a female bear with two young cubs, by Robert Wenger in 1993, but did not count the deaths of the two cubs. Mr. Wenger entered into a plea agreement with the federal government to avoid further prosecution in which he agreed to pay $15,000 in restitution for the loss of the two cubs to the ecosystem. See Supp. Exh.8 (“$15,000 Restitution ($7,500 per cub)”); Idaho Department of Fish and Game Grizzly Bear Mortality File, Supp. Exh. 2 at 2 (Mr. Wenger was “ordered to reimburse Selkirk ecosystem $15,000 for loss of cubs to ecosystem”). Since the federal government was prepared to prove their deaths at a criminal trial beyond a reasonable doubt, the FWS should have counted the deaths of the two cubs in its figures. In addition, the FWS admits that it knows that another male grizzly was killed in October 1995 by Carl Pitts, but it excluded this information from the supplemental findings filed with the Court because this death was not confirmed until a few weeks after the filing was made. Leaving aside the issue of whether the FWS should have amended or supplemented its filing when it became aware of this mortality, it certainly should incorporate this mortality into any future analysis, particularly since the mortality has been established by Mr. Pitts’ guilty plea in federal court. See June 28, 1996 Order Accepting Guilty Plea, Supp. Exh. 4. As for the reliance on Mr. Harris’ article for the use of the four percent mortality figure, it is hard to see how it can be justified. Mr. Harris’ work was based on a population of 448 grizzly bears, over ten times the size of the Selkirk population. Supp. Exh. 10 at 270-71. In the very article on which the FWS relied for its assertion, Mr. Harris explicitly stated that “the concept of sustainable yield is not independent of the size of the population in question.” Id. at 268-69. Furthermore, Mr. Harris explained in more detail in his master’s thesis the reasons why his four percent figure should not be used for smaller populations.4 The FWS unsuccessfully tries to parse Mr. Harris’ words to avoid the explicit limitations of his work. It argues that Mr. Harris’ article merely acknowledges that sustainable yield is “not independent” of the size of the population and does not state that its use is — 110 — inappropriate. Defendants’ Opposition at 21. This attempt to make a semantic distinction is contrary to Mr. Harris’ statement in his master’s thesis—the document supporting the use of the four percent figure in the management plan—that his research could not be applied to smaller populations. In particular, Mr. Harris was careful to say that “[s]mall populations were unable to sustain the same proportional harvests as large populations. Thus, the common notion of a maximum sustainable rate of harvest was inappropriate when applied to small populations.” Supp. Exh. 11 at 178 (citations omitted) (emphasis added). The FWS also argues that its conservative use of a four percent human mortality rate instead of the six percent mortality rate recommended by Mr. Harris should account for any additional risk to small populations. Defendants’ Opposition at 21. This argument ignores the assertion of the FWS itself in its management plan that the adjustment from six percent to four percent was intended to account for “unknown, unreported, human-caused mortality.” Plan at 20. The FWS cannot ascribe its downward adjustment of the acceptable mortality rate to two different justifications. In fact, the FWS itself has acknowledged the problems with the use of the four percent figure in its recovery plan for the Selkirk population. In the plan, the FWS stated that the use of the Harris figure “does not account for demographic, genetic, or other problems that can be dramatically amplified in such small populations.” Plan at 103. The FWS therefore was fully aware of the limitations on the use of the Harris figure with a population of 26 to 36 bears, making its use arbitrary and capricious.5 2. The Adequacy of Regulatory Mechanisms In its supplemental findings, the FWS detailed a long list of recently initiated regulatory efforts intended to protect the Selkirk population. These initiatives include an agreement with the U.S. Forest Service to decrease road densities in grizzly bear habitat, the elimination of sanctioned grizzly bear hunting by Canadian officials, increased educational and enforcement efforts and increased penalties for grizzly bear shootings. Supplementary Information at 5-6. The FWS concluded that the promise of these efforts was adequate and did not warrant the listing of the Selkirk population. Unfortunately, the conclusion of the FWS runs up against the same logical obstacle that it did previously. See Carlton v. Babbitt, 900 F.Supp. at 533. The FWS finding that the current mortality level is sustainable is not supported because it ignored known mortalities and used an inappropriate benchmark for sustainability. See supra Part II, C, 1. Therefore, to the extent that the FWS concluded that the current regulatory mechanisms are sufficient to keep mortality in the Selkirk population at that level, its conclusion also cannot be supported. The current regulatory approach cannot be adequate so long as it is intended to maintain an unsustainable level of mortality. 3. The Size of the Selkirk Population Where the FWS originally disagreed with the plaintiffs that the Selkirk population consisted of only 26 to 36 bears, see Carlton v. Babbitt, 900 F.Supp. at 533 n. 6, it now accepts this as a proper estimate. Supplementary Information at 7. The FWS, however, now contends that such a small population can be sustained in the Selkirk ecosystem. In taking this position, the FWS falls back on its argument that the Selkirk population can sustain a four percent known human-caused mortality rate. Id. It also asserts that the Selkirk population is not particularly susceptible to threats from environmental and demographic stoehasticity and that there is no scientific consensus on the concept of a minimum viable population. Id. at 7-8. — 111 — As a preliminary matter, the Court notes that the FWS position on this issue is inconsistent with the position it took earlier regarding the Cabinet-Yaak population. The Cabinet-Yaak population is only marginally smaller than the Selkirk population, totaling approximately 15 to 20 bears. Plan at 84. In addition, the Cabinet-Yaak population is located adjacent to the Selkirk population, see Plan at 11, and faces the same human-caused stressors on its habitat as does the Selkirk population. See 57 Fed.Reg. 14,372, 14,374 (1992). Nevertheless, the FWS determined that the Cabinet-Yaak population warranted listing as an endangered species while the Selkirk population does not. 58 Fed. Reg. at 8251; see also Plan at 85 (“[T]he mortality goal for [the Cabinet-Yaak ecosystem] is zero until the three key parameters monitored indicate a population of approximately 100 bears”) (emphasis added). The FWS has not adequately explained its disparate treatment of the two populations. The assertion by the FWS that the Selkirk population is not in danger simply by virtue of its size is also troubling. Its invocation of Mi’. Harris’ four percent mortality rate is flawed for the reasons discussed above. See supra Part II, C, 1. Moreover, the FWS has not demonstrated that genetic interchange occurs between the Selkirk population and other populations. The FWS acknowledges that genetic interchange is needed to sustain small populations of grizzly bears, see Plan at 27, but offers no evidence of such interchange in its supplemental findings. Finally, the FWS fails to present a defensible case that a scientific consensus does not exist about the viability of a population of 26 to 36 bears. In its supplemental findings, the FWS described a range of opinion on the minimum viable population size for grizzly bears. To demonstrate the optimistic and pessimistic sides of this range, the FWS cites the 1978 study by Dr. Mark Shaffer and a 1991 study by B.P. Dennis. Supplementary Information at 7. At the time it presented this finding, however, the FWS had before it a declaration submitted by Dr. Shaffer in Fund for Animals v. Babbitt, 903 F.Supp. 96 (D.D.C.1995), in which Dr. Shaffer disclaimed the optimistic reading given to his work by the FWS. See March 13, 1994 Declaration of Dr. Mark L. Shaffer, Supp. Exh. 21. In his declaration, Dr. Shaffer stated that the FWS “misinterpreted [his] work and has used this misinterpretation to arrive at highly arbitrary recovery goals.” Id. at 2. Because his study did not account for demographic factors such as “genetic factors or the impact of natural catastrophes,” Dr. Shaffer believed that it simply could not be used to justify a low minimum viable population. Id. Instead, he argued that the “best available scientific studies in the field of conservation biology suggest that a population of several thousands of individual grizzly bears is necessary to ensure persistence over the long term.” Id. at 3. Given this Court’s concern about the size of the Selkirk population, as well as the plaintiffs arguments that there is indeed a consensus that the minimum viable population for grizzly bears is much higher than the number of bears in the Selkirk population, the FWS must support its position that the Selkirk population is viable with more than a 1978 study that has been disavowed by its author.6 4. Canadian Habitat Plaintiffs’ final argument is that the FWS did not consider the status of the grizzly bear habitat in the Canadian portion of the Selkirk ecosystem when it relied on the presence of Canadian habitat in its analysis. The FWS countered that this Court already determined that the FWS had adequately considered this factor when it found that “the FWS sufficiently addressed whether there is ‘present or threatened destruction, modification, or curtailment of [the grizzly’s] habitat or range.’ ” Carlton v. Babbitt, 900 F.Supp. at 531-32 (quoting 16 U.S.C. § 1533(a)(1)(A)). Although the Court determined that the FWS had satisfied that requirement for the lands over which the FWS has jurisdiction, it — 112 — explicitly stated in a footnote that “the agency’s reliance on the contiguity of the Canadian portion of the ecosystem may require further explanation.” Id. at 532 n. 3. The FWS therefore was expected to consider this issue. See id. at 534. The FWS decision to look to Canadian habitat makes sense. Grizzly bears have no respect for political boundaries in their movement and will survive in Canada as easily as in the United States. Nonetheless, if the FWS is going to rely on habitat north of the border to sustain the Selkirk population, it must evaluate whether those areas will continue to exist as grizzly bear habitat. While the FWS noted several initiatives by the government of British Columbia to protect the Selkirk population, it did not make findings regarding the present or threatened destruction of the Canadian habitat. The FWS must make such findings before it can rely on the continued viability and availability of the Canadian habitat. III. CONCLUSION For the reasons stated above, the Court finds that the supplemental findings of the FWS fail to consider several relevant factors and draw conclusions that cannot be supported by the evidence in the record. Accordingly, the decision not to reclassify the Selkirk population as “endangered” was arbitrary and capricious and must be remanded to the agency for further consideration consistent with this Opinion. While the agency remains the expert in this field, it appears on the basis of the scientific evidence presented that the Selkirk population is simply too small to justify the agency’s decision not to reclassify it as endangered. While the FWS may once again reassess the status of the Selkirk population on the “basis of the best scientific and commercial data available,” 16 U.S.C. § 1533(b)(1)(A), a population of 26 to 36 grizzly bears seems to be endangered almost by definition. An Order consistent with this Opinion is entered this same day. SO ORDERED. ORDER Upon careful consideration of defendants’ Notice of Filing, plaintiffs’ response and motion to supplement the administrative record, defendants’ opposition, plaintiffs’ reply and the administrative record as supplemented, and for the reasons stated in the Opinion issued this same day, it is hereby ORDERED that plaintiffs motion to supplement the administrative record is GRANTED in part and DENIED in part; it is FURTHER ORDERED that the administrative record is supplemented by Exhibits 2-11 and 21 from plaintiffs’ motion to supplement the administrative record; it is DECLARED that the defendants’ decision on remand not to classify the Selkirk population of grizzly bears as “endangered” under the Endangered Species Act, 16 U.S.C. §§ 1531 et seq., was arbitrary and capricious because the defendants failed to explain sufficiently how they exercised their discretion with regard to the statutory listing' factors and drew conclusions that cannot be supported by the record; it is FURTHER ORDERED that this matter is remanded to the U.S. Fish and Wildlife Service, which shall provide the Court with a reasoned analysis of whether to classify the Selkirk population of grizzly bears as endangered on or before January 21, 1999; and it is FURTHER ORDERED that plaintiffs’ motion for attorneys’ fees is no longer held in abeyance. See Order of May 7, 1996. Defendants shall file an opposition to the motion on or before November 17,1998. SO ORDERED. Footnotes [1] Specifically, Exhibits 2 to 9 are: (1) a 1996 copy of the Idaho Department of Fish and Game’s grizzly bear mortality file, (2) the Information from the criminal proceeding against Carl G. Pitts for killing a grizzly bear, (3) the Order accepting Mr. Pitts’ guilty plea, (4) the Information from the criminal proceeding against Robert Christopher Wenger for killing a grizzly bear and its cubs, (5) the record of the proceeding in which the District Court accepted Mr. Wenger’s guilty plea, (6) the Judgment entering Mr. Wenger’s guilty plea, (7) the record of the sentencing of Mr. Wenger, and (8) a newspaper article noting the death of “Sy," a female grizzly bear, and her two cubs. [2] Exhibits 13 to 19 are: (1) a 1995 report on conservation of grizzly bears by British Columbia Ministry of Environment, Lands and Parks, (2) a 1994 paper by Vivian Banci, Dennis Demarchi and W. Ralph Archibald titled "Evaluation of the Population Status of Grizzly Bears in Canada,” (3) a 1993 paper by Michael L. Gibeau titled "Grizzly Bear Habitat Effectiveness Model for Banff, Yoho and Kootenay National Parks,” (4) a 1994 article from the Christian Science Monitor on the status of grizzly bears in Canada, (5) a 1995 report by the World Wildlife Fund entitled "Large Carnivore Conservation in the Rocky Mountains,” (6) the minutes of the January 19, 1996 Grizzly Bear Scientific Committee in British Columbia, and (7) a 1993 report by the Conservation, Recreation and Cultural Heritage Subgroups of the Kootenay Regional Protected Areas Team entitled "A Protected Areas Strategy for British Columbia.” [3] In fact, even accepting the mortality numbers as reported, the FWS is unjustifiably secure in the Selkirk population's viability. With an annual mortality rate of 1.29, the Selkirk population would have to contain at least 33 bears in order to have a mortality rate as low as four percent. This is near the high end of the range of 26 to 36 grizzly bears that the FWS believes is in the Selkirk population. [4] As Mr. Harris stated in his master’s thesis: Sustainable yield, expressed as a proportion of the standing population, was inversely related to the magnitude of demographic stochasticily. Because small populations were more influenced by demographic stochasticity than large populations, they had lower proportional sustained yields than larger populations. Thus, the concept of a constant proportion of the population as being harvestable was incorrect. Supp. Exh. 11 at 186-87. In other words, Mr. Harris found that because smaller populations are more influenced by random or stochastic events that may undermine the sustainability of a population, the sustainable yield for smaller populations is lower than it is for larger populations. Mr. Harris' concentration on "demographic sto-chasticity" refers to those random events, such as infertility, that affect reproduction and surviv-orship. [5] The FWS also does not discuss the second requirement for a sustainable mortality rate discussed in Mr. Harris' article: that no more than 30 percent of the annual mortalities could be breeding females. Plan at 103. For the Selkirk population, this means that a death of a breeding female could not occur more than once every 2.3 years. From the figures provided in the supplemental findings, it appears that three females were killed between 1992 to 1995. Supplementary Information at 3. While the Court cannot determine if these females were breeding, the FWS should have addressed this factor if it was going to base its findings on Mr. Harris' article. [6] The FWS noted that this Court has already found in another context "that there is disagreement among experts regarding the correct population size that is necessary for viability.” Fund for Animals v. Babbitt, 903 F.Supp. at 115. Nothing in the record of that case, however, suggests that any disagreement exists as to the viability of a population as small as 26 to 36 grizzly bears.