dennis e. bschora123.g.akamai.net/7/123/11558/abc123/forestservic...decision be affirmed. decision...

United States Department of Agriculture

Forest Service

Alaska Region P.O. Box 21628 Juneau, AK 99802-1628

Caring for the Land and Serving People Printed on Recycled Paper

File Code: 1570 Date: October 27, 2008

Mr. Nathaniel Lawrence Natural Resources Defense Council 3723 Holiday Drive Olympia, WA 98501 Dear Mr. Lawrence:

Pursuant to 36 CFR 215.18, I have reviewed the administrative record for the Scratchings II Timber Sale Final Environmental Impact Statement (FEIS) and Record of Decision (ROD). The Tongass Forest Supervisor signed the ROD. I have also considered the Appeal Reviewing Officer's (ARO) recommendation (enclosed) regarding the disposition of your appeal (Appeal Number 08-10-00-0013 A215). The ARO recommended that the Forest Supervisor's decision be affirmed.

Decision

After careful review of all the information in the record, I concur with the ARO's recommendation and I affirm the Forest Supervisor's decision. Your requested relief is denied.

My decision incorporates, by reference, the entire administrative record, which includes the appeal and project planning records, and constitutes the final administrative decision of the Department of Agriculture (36 CFR 215.18(c)). The ROD may be implemented 15 business days following the date of this decision (36 CFR 215.9(b)).

Sincerely, /s/ Dennis E. Bschor DENNIS E. BSCHOR Regional Forester Enclosure cc: Forrest Cole, Greg Killinger, Karen Iwamoto, Winnie Blesh

08-10-00-0013 A 215 (NRDC)

Reviewing Officer Recommendation

United States Forest Alaska Region P.O. Box 21628 Department of Service Juneau, AK 99802-1628 Agriculture

File Code: 1570 Date: CCT 2 7 2008

Subject: Scratchings I1 Timber Sale Appeal Recommendation

To: Appeal Deciding Officer

This is my recommendation, as Appeal Reviewing Officer, on the action you should take, as Appeal Deciding Officer, on the pending appeal of the Scratchings I1 timber sale project. The following appeal was filed under 36 CFR 21 5:

Appeal No. 08-10-00-0013, filed by the Natural Resources Defense Council, Greenpeace, Cascadia Wildlands Project, Juneau Group of the Sierra Club, and Tongass Conservation Society (NRDC et al).

The decision being appealed is the decision by the Tongass Forest Supervisor, Forrest Cole, to authorize the sale of timber and the construction of roads on Suemez Island, Alaska. The Selected Alternative would allow the harvest of approximately 3 million board feet (MMBF) of timber and the construction of 0.8 miles of National Forest System (NFS) road and 1.4 miles of temporary road.

Background

A Notice of Intent for the Scratchings Environmental Impact Statement (EIS) was published in the FederalRegisteu in July 2005, and a Notice of Availability of the Draft EIS (DEIS) was published in August 2006. The Tongass Forest Supervisor signed the first Record of Decision (ROD) for the Final EIS (FEIS) (referred to as the Scratchings I ROD) on March 12,2007. The Scratchings I decision included only those activities within the roaded portion of the project area, and the Scratchings I ROD was not appealed.

The Scratchings I1 ROD was signed on July 21,2008, and supplemented the Scratchings I ROD. The Scratchings I1 decision authorized the harvest of seven units and the construction of the roads associated with those units. These units were included in the FEIS and the Selected Alternative for the Scratchings I ROD, but were deferred from the March 2007 decision as they were located all or partly within the inventoried roadless areas and thus were subject to the Settlement Agreement reached between the parties in Natural Resources Definse Council er al vs. US. Forest Service, et al. [Case No. 1:03-cv-0029-JKS, CaseNo. 1:04-cv-0010-JKS, Case No. 1:04-cv-0029-JKS, and Case No. 1:06-cv-0005-JKS].

The 2008 Tongass Forest Plan Amendment was completed before the Scratchings I1 ROD was issued, with the ROD for the Amendment being signed on January 23,2008, and becoming effective on March 17,2008. The ROD for the 2008 Forest Plan Amendment adopts the Timber Sale Program Adaptive Management Strategy, under which portions of the suitable land base become available for project-level planning in three phases. The Scratchings project is within

It's Cool to Be Safe ~ r ~ n t e d on Recycled Paper 43

Appeal Deciding Officer 2

the Phase 1 portion o f the suitable land base, which allows planning to continue for this project and implementation o f the project once the planning process is completed.

My review o f this appeal was conducted pursuant to 36 CFR 215.19. The appeal and planning records have been carefully reviewed in my consideration o f the objections raised by the appellants and their requested relief. The Craig Ranger District office prepared the enclosed indices o f the documentation supporting the decision, which are keyed to specific points raised by the appellants. My recommendation hereby incorporates by reference the entire appeal record.

The appellants list many interrelated issues in their appeal o f the Scratchings project. Although I may not have listed each specific issue, I have considered all o f the issues raised in the appeal and believe that they are adequately addressed in the following discussions. While I am addressing all o f the issues raised in the appeal, even those that relate only to the Scratchings I decision, I do so without waiving any legal arguments about the appellants' failure to appeal Scratchings I.

Issue, Whether the Forest Service provides for wildlife viability in the project area.

Appellants assert:

the conservation strategy will not ensure species viability and violates National Forest Management Act (NFMA); and the Forest Service did not disclose or discuss the controversy about the adequacy o f the conservation strategy.

Discussion

The larger issue o f viability raised by the appellants is more appropriately addressed in the Forest Plan and this point was noted in the Scratchings FEIS (Appendix B, p. 1 1 5). The conservation strategy is one o f the foundations o f the Forest Plan and was part o f a long, iterative process that included peer reviews up until the Forest Plan was approved in 1997. Those reviews included Dr. Bruce Marcot's Technical review ofpopulation viability conservation planning strategy for TLMP revision (July 6, 1992). Dr. Mascot concluded:

Overall, I support the viability assessment process that the viability committee has develop.. .It is my professional opinion that the process and basis for the proposed conservation strategy is scientifically sound, given our current knowledge base on these species in southeast Alaska. I stress "current knowledge" because we seem to lack much basic empirical data on wildlife in southeast Alaska.

Dr. Mascot goes on to identify other information needs and makes suggestions on how to improve the conservation strategy. Continuing reviews o f the evolving conservation strategy were included in the April 26, 1994, interagency meeting convened to cover wildlife viability peer review results. This meeting included discussion about the peer reviewers comments, such as "[tlhe reviewers were troubled that they were unable to predict what the 'null hypothesis' or 'default behavior' would look like, since we have

Appeal Deciding Officer

very limited data." As part o f the "give-and-take" o f the review process, an interagency committee developed a Response to thepeer review ofi apropose strategy,for maintaining well-distributed, viable populations of wildlife associated with oldgrowth ,forests in southeast Alaska (May 1994). Ultimately, the conservation strategy was improved and adopted, but not before it received a wide range o f comments from agency staff, other scientists, and interagency biologists.

The reviews o f the conservation strategy did not end with the adoption o f the 1997 Forest Plan. As part o f the 2008 Forest Plan Amendment process, a 2006 conservation strategy workshop identified key findings, some o f included:

the Conservation Strategy is still sound; there is a low risk o f species viability problems related to Forest Plan implementation; there is a need to better understand the role and management o f the matrix component - o f the Strategy, including the role o f non-NFS lands; and the wildlife monitoring program needs to be updated.

In addition, the Review of conservation science produced since 1997 and its relationship to the Tongass National Forest Land and Resource Management Plan (Haufler 2006) discussed coarse and fine filter strategies and their relevancy to the Tongass National Forest. Haufler concluded:

[tlhe review o f the recent literature supports the appropriateness o f the science used in the Tongass Forest Plan. The coarse filter strategy linked with a species assessment remains a supported conservation approach ... While no new information undermines the basic structure o f the conservation strategy o f the plan, a finer scale analysis o f viability for selected species would allow a better assessment o f the status o f populations that might be o f concern as important endemics (page 24).

In my opinion, the 2008 Forest Plan Amendment record demonstrates that differing perspectives were considered and factored into the conservation strategy. The record also shows that the strategy is not meant to be a static device, but is intended to be dynamic as new concepts develop and additional species-specific information is obtained.

With regard to the Scratchings project specifically, the Scratchings I1 ROD is appropriately tiered to the comprehensive landscape old-growth habitat reserve strategy designed for the Tongass Forest Plan. This strategy was developed with consideration o f the cumulative effects o f a full 100-year implementation o f the Forest Plan, including the effects o f the Scratchings project and all other potential sales within the Tongass National Forest. As described in the 2008 Forest Plan Amendment ROD (page 16), the analysis in the 2008 Final EIS for the Forest Plan concluded that "...the amended Forest Plan will nrovide an amount and distribution o f habitat adequate to maintain viable populations o f vertebrate species in the planning area and will maintain the diversity o f plant and animal communities." Any project that tiers to and fully implements the landallocations and standards and guidelinesif the Forest Plan will maintain viable populations o f old-growth associatedspecies.


The Scratchings FEIS states that the Forest Plan anticipated some of the small old-growth reserves (OGR's) would be adjusted during project-level planning to better meet Tongass National Forest criteria and objectives. The FEIS (pages 3-55 through 3-64) describes the interagency review process and the recommendations to adjust the small OGR's in the project area. The 2007 Scratchings ROD (referred to as Scratchings I ROD, page 2) approved those adjustments through a non-significant amendment to the 1997 Forest Plan. These adjustments were carried forward in the adoption of the 2008 Forest Plan Amendment.

Conclusion

In my opinion, the Scratchings project is consistent with the conservation strategy for the Tongass Forest Plan; therefore, species viability has been fully and appropriately addressed.

Issue. Whether the Forest Service adequately analyzed and disclosed the potential effects of the project on deer.

Appellants assert:

the wrong deer multiplier was used in light of the 2008 TLMP amendment; the model should have used the size-density dataset instead of the volume-strata dataset and the Forest Service failed to disclose the controversy about this dataset; the model erroneously assumes a linear habitatlpopulation relationship; the model overestimates carrying capacity because of its assumption of linearity, its disregard for the juxtaposition of habitat patches, and the effects of severe winters which are not accounted for in the model; the wrong deer multiplier was used because of the lack of summer range in the project area; private land was improperly excluded from the deer habitat capability analysis; the wrong geographic scales were used for the deer habitat capability analyses; the quartile analysis of deer habitat capability is flawed; the Forest Service's claim of completing a "worst case" analysis is incorrect.

Discussion

Many of appellants' issues have been addressed in previous appeal responses. Appellants continue to assert that the Forest Service has failed to adequately analyze the effects of Tongass timber sale projects on deer because the analysis relies on the use of the deer model and coefficients and datasets that they believe are inappropriate. I disagree. As stated in the Scratchings FEIS, in numerous other timber sale project EISs, and in the Forest Plan EIS, the numbers derived from the use of the deer model are "useful for comparing alternatives, but [are] not intended to estimate actual deer populations" (FEIS, p. 3-33). These same documents, along with other documentation in their project records, clearly disclose the limitations and cautions regarding the use of habitat capability models, and explain why the Forest Service continues to use the deer model to compare the differences in effects on deer habitat capability (not population numbers) between project alternatives.


With regard to the Scratchings project specifically, the EIS and project record indicate that the deer model was run using "100 deer per square mile for habitat scores of 1.0," and using the volume strata dataset (FEIS, pp. 3-32 to 3-33). This is consistent with current Tongass National Forest direction regarding the use of the deer model, as outlined in the May 25,2005, memo from the Forest Supervisor and the Annual Monitoring and Evaluation Report for Fiscal Year 2000 (FEIS, p. 3-32; also Scratchings WildlifdBotany/Subsistence Resource Report, Document 498).

As appellants correctly point out, the deer analysis conducted for the 2008 Forest Plan Amendment "standardized" the deer model scores by dividing the output values by 1.3. It also used the new forest-wide size-density vegetation model rather than the volume-strata dataset, although it did have to reclassify the 7 size-density categories into the 3 vegetation categories used by the deer model (2008 Forest Plan Amendment FEIS, pp. 3-265 to 3-266). Appellants believe that in making these adjustments, the Forest Service has "recognized and corrected its long-standing erroneous use of the deer multiplier" and has "[settled] . . . the question as to which dataset should be used." They assert that the Forest Service should have made these adjustments in its use of the deer model for the Scratchings project. I disagree. While the planning team for the 2008 TLMP Amendment did make adjustments to the information used in the deer model, I do not believe this means that the Forest Service has "recognized" any errors in how previous model runs were completed for the Forest Plan or other timber sale projects, and I believe that it is inappropriate to compare the two separate analyses and conclude that one is "wrong" just because the outputs are different. While the 2008 Forest Plan Amendment and the Scratchings project used the same model, the Forest standardized the output values and used a different vegetation layer when running the model for the Plan Amendment. Therefore, the numbers are not comparable and the Forest Plan EIS acknowledged these differences (p. 3-266). However, both the model runs used for the 2008 Amendment and those used for the Scratchings project are useful tools when used as intended - to compare and rank the alternatives in a relative sense.

The Scratchings FEIS and supporting documentation (including the Wildlife/Botany/Subsistence Resource Report) were completed many months before the Forest Plan Amendment, and there is no requirement for the Forest Service to halt project planning and conduct new analyses every time new information becomes available during forest planning. As stated in the transition language of the 2008 Amendment ROD (p. 70), the Scratchings project falls within Category 2. For projects in this category, the ROD directed the Forest Supervisor to "incorporate the new direction in the amended Forest Plan to the extent this can be done without causing major disruptions in the implementation of these projects." The text goes on to state that the changes to be considered include the legacy standards and guidelines for goshawk and marten, and new direction regarding goshawk nests. While some of the information used in the deer model was adjusted for the 2008 Forest Plan Amendment analysis, neither the ROD or the 2008 Forest Plan contain any new direction with regard to the use of the deer model at the project level.

I have considered whether the Scratchings project would still meet the Forest Plan objective to "[plrovide, where possible.. . 18 deer per square mile.. . in biogeographic provinces where deer are the primary prey of wolves" (2008 Forest Plan, p. 4-95) and maintain an adequate supply of deer to meet estimated human demand from the project area if the Forest Supervisor had made the same adjustments in the deer model runs for the project. In doing so, I considered the


estimated deer per square mile numbers by Wildlife Analysis Area (WAA) provided by the appellants and included in the 2008 Amendment Planning record. I also considered the estimated hunter demand described in the Scratchings EIS and the minimum number of deer needed to meet that demand (pp. 3-193 to 3-194). Based on this review, I believe that WAA 901 (the project area) and the Southern Outer Island Biogeographic Province (WAAs 901,902, 1003, 1524, and 501 5) would still provide sufficient habitat capability to meet these objectives. Furthermore, the relative ranking of the alternatives did not change as a result of this analysis; therefore, I have no reason to believe the Forest Supervisor would have selected a different alternative if he had this "new" information before him when making his decision.

I should add one further point about the size-density dataset used (with modifications) for the 2008 Amendment and advocated by appellants. The 2008 Plan Amendment record includes the publication Predictive mapping for tree sizes and densities in southeast Alaska (2005), which describes the size-density dataset. This document includes multiple caveats:

We caution that there is a substantial amount of variation and noise in any forest mapping exercise.. .With so much variation, error, and noise in the data, it is highly unlikely that any mapping model will lead to conclusive, definitive, and totally acceptable results.. .

We need more ground-sampled data and closer work with biologists, ecologists, silviculturists, planners, and managers to see if our model can help with their particular needs, interests, or applications.

Therefore, while the size-density dataset was used for the 2008 Plan Amendment, I believe it is reasonable for the Forest Supervisor to hold off on the use of this dataset for site-specific project planning until he believes it is an appropriate tool for that use.

Appellants' assert that the model overestimates carrying capacity because of its assumption of linearity, its disregard for the juxtaposition of habitat patches, and the effects of severe winters which are not accounted for in the model. The Scratchings FEIS, the Wildlife/Botany/ Subsistence Resource Report completed for the project, and numerous documents in the Scratchings project and 2008 Plan Amendment records discuss many of these issues raised by the appellants. The model used for the 1997 Forest Plan, the 2008 &rest Plan Amendment, A d the Scratchings project analyses is simple, based on professional judgment, uses limited data and few variables, and is best used for comparison among alternatives. It was not designed to consider all of the variables appellants believe it should account for.

Appellants imply that because the model does not account for variables such as habitat fragmentation or severe winters, the Forest Service has not considered them in its analysis of the effects of the Scratchings project. This is not true. The Scratchings FEIS and project record clearly indicate that these variables have been considered. For example, the FEIS discusses habitat fragmentation in the project area, including patch size changes (p. 3-69),, changes in acres of interior productive old-growth habitat (p. 3-70), a map of productive old-growth (p. 3-71),


and travel corridors between old-growth habitat patches (pp. 3-73 through 3-75). The effects of the alternatives on old growth habitat patches and wildlife travel corridors are disclosed on pages 3-76 to 3-79.

With regard to appellants' assertions regarding severe winters, it is not appropriate to use the current deer model to attempt to quantitatively display the effects of stochastic events, such as a higher than average snowfall winter. The deer model uses average winter snow depth as one of the parameters to estimate winter habitat capability. The model is not designed and should not be used to reflect a one-time event, even though that event may happen with greater frequency (2008 Forest Plan Amendment FEIS, Volume 1, p. 3-296). The WiIdlife/Botany/Subsistence Resource Report includes a detailed discussion of the effects of severe winter weather on deer populations (p. 69).

Appellants' assert that the model overestimates deer habitat capability. I disagree. In fact, I believe the deer model may underestimate habitat capability, for multiple reasons. As pointed out in the 2008 Plan Amendment FEIS (Volume 1, p. 3-266), "[llands under non-federal ownership have an assumed habitat capability of zero," which is generally not the case. Additionally, "forest management on the Tongass has produced more forage than assumed by the model, through the management of second-growth stands" (Id., p. 3-268), which indicates that many young growth stands should be assigned greater habitat suitability indices. The 2008 Plan Amendment EIS also suggests that "changes to the 1997 HSI model may be warranted such that higher suitability values are assigned to habitats that provide important summer forage, such as recent clearcuts, unproductive forest, and low volume old growth" (Id., p. 3-232).

In my opinion, the Scratchings project analysis appropriately relied on the approved deer habitat capability model, as well as other information, to estimate the effects of the alternatives on deer habitat capability. The project record clearly discloses the limitations of the model, as well as the fact that it was never intended to estimate actual deer populations. The 2008 Forest Plan Amendment FEIS (p. 3-232) describes a "new tool for evaluating deer habitat appropriate for analysis at finer scales, called the . . .FRESH-Deer model" that is "currently being developed by the Forest Service in cooperation with the University of Alaska." This model is not available for use at the current time, but the Forest continues to review it for possible application at the project level in the future.

With regard to appellants' assertions regarding the lack of summer range (and consequent overuse of winter range) in the Scratchiugs project area, there is no indication in the Scratchings FEIS or in the Wildlife/Botany/Subsistence Resource Report (Document 498), other than the comment submitted by the Sitka Conservation Society, et al. (Appendix B, p. 96) and noted by appellants, that there is a concern regarding the overuse of winter range. The project record includes documentation of two consultations with the Alaska Department of Fish and Game (ADF&G) (Documents 335 and 495) regarding the project, and there are no concerns noted on these records about the overuse of winter range or suggestions regarding an adjustment of the deer multiplier at the project level to compensate for any such overuse. Therefore, any adjustment to the deer multiplier at the project level would have been inappropriate unless there was other project-specific data to suggest otherwise. I did not find any such data in the record.


Appellants also assert that private land in the project area was not included in the deer model analysis. The statements that "the acreage of private land was not figured into the calculations in 2007," found in Appendix B of the FEIS (p. 122), page 65 of the WildlifelBotanylSubsistence Resource Report, and cited by appellants, appear to be in error. This statement is not repeated in the FEIS itself (other than in the appendix as noted). The Scratchings ROD indicates that "Suemez Island is 37,127 acres in size, 35,960 acres of which are [NFS] lands. In the northwest corner of the island, the remaining 1,167-acre parcel is owned by the University of Alaska" (ROD, p. 1). The project boundary, as shown in the FEIS (Figure 2-I), follows the shoreline of the Island, excluding small adjacent islands. Document 769 in the project record, titled "FEIS deer model numbers," clearly shows at multiple locations that the acreages used for model runs totaled 37,244 acres. This indicates that the deer model analysis included all of the acres in WAA 901 (this is also indicated by the titles ofTable 3W-1: Deer Density in the Scratchings Project Area (WAA 901), and Table 3W-2: Deer Density in the Scratchings Project Area (WAA 901) by Alternative). The slight discrepancy in total acres (37, 127 verses 37,244 acres) is likely due to the fact that small adjacent islands are included in the WAA but are not within the project boundary. In addition, an email dated December 29, 2006, states "that private lands are given a zero in the model," and the data file of the Scratchings deer model run clearly indicates that 1163 acres of private land were assigned an HSI value of zero. While I could not locate this email or the data file in the appeal record forwarded by the Craig Ranger District, they are clearly dated before the Scratchings decisions and should be added to the project record if they aren't already included in the record.

Appellants also assert that the geographic scales used for the Scratchings analysis were inappropriate. Deer habitat capability was calculated at three different scales for the Scratchings FEIS: by Value Comparison Unit (VCU) (WildlifelBotany/Subsistence Resource Report, p. 28), by WAA (FEIS, p. 3-34), and by biogeographic province (FEIS, p. 3-38). As appellants note, and as stated in the FEIS (p. 3-38), the "Tongass National Forest Implementation Clarification Policy (TPIT) states that since wolves tend to have home ranges that cross several WAA's, the appropriate scale for analysis should be multiple WAA's or the biogeographic province."

As discussed above, the Scratchings project record indicates documentation of two consultations with ADF&G personnel (Documents 335 and 495) regarding the project, and these records do not note any concerns regarding the scales used for the deer habitat capability analysis (and related wolf analysis). While there are certainly an infinite number of ways to group habitat units (VCU's or WAA's) and conduct an infinite number of analyses, the geographic scales considered in the Scratchings deer habitat capability analysis (and related wolf analysis) are reasonable and consistent with current Tongass National Forest direction.

Appellants also question another component of the deer habitat capability analysis conducted at the project level, and also at the Forest Plan level, which includes comparing acres by quartile between alternatives. The rationale for this method of analysis is stated in the 2008 Forest Plan Amendment FEIS, (Volume I, p. 3-267): "[tlo take into account impacts to deer across the Tongass inhabiting areas that vary naturally in their habitat quality, high quality habitat was defined as the quartile of the land base with the highest HSI scores within.. .each WAA." This direction is further outlined in the Memorandum of Understanding (MOU) between the State of


Alaska and the Forest Service, Alaslia Region on Coastal Zone Management Act/Alaska Coastal Management Program Consistency Reviews (Attachment I ) , that states that the Forest Service will provide information on ''[all1 deer winter range in project area (that scores above zero in most recent interagency approved version of deer HSI model) distinguished by quartile (i.e. by 25 percent of acres)."

There are an infinite number of ways to apportion, compare, and analyze data. In fact, the appellants offer an interesting alternative for developing quartiles based on habitat capability, with the objective of identifying the "most productive winter habitat." For the Scratchings analysis, the WildlifelBotanylSubsistence Resource Report displays productive winter habitat using past and current acres by quartile (p. 70). Table 3W-4 in the FEIS (p. 3-35) displays anticipated changes in the fourth quartile (when compared to both 1954 and 2007), by alternative. The fourth quartile is defined as "high value deer winter habitat in the Scratchings Project Area" (footnote to Table 3W-4). This information is briefly discussed by alternative in the FEIS (pp. 3-34 and 3-35). In my opinion, this development and comparison of habitat capability by quartiles followed Forest Plan and MOU direction.

Finally, appellants question the use of the term "worse case scenario" used in reference to deer model outcomes by alternative (p. 3-35). As discussed previously, non-NFS lands were not excluded from the deer model runs, but were included and given an HSI score of zero. Additionally, as noted in the WildlifeiBotanyISubsistence Resource Report (p. 69), the tern "worse case scenario" is also "intended to describe the amount of deer habitat that would be harvested if all harvest units were clearcut. In some partial harvest units, the effects on deer may be less than that predicted by the model." The current deer model treats all harvest units as clearcuts. However, not all units will receive this treatment. The FEIS indicates that, for the Selected Alternative, 1101 acres will be clearcut, while 234 acres will be harvested under a two- aged management system (with 41 acres deferred from harvest). Under a two-aged system, a majority of the unit will be clearcut, but a minimum of 15 percent of the setting pretreatment basal area, based on standing live trees, will be retained (pp. 3-106 to 3-107). In my opinion, the use of the term "worse case scenario" is acceptable in the context of this analysis.

Conclusion

In my opinion, the Forest's use of the deer model to estimate the potential effects of the project on deer habitat capability is consistent with current Tongass National Forest direction, and the Forest's continued use of the deer model, in conjunction with other analyses, is reasonable and supported by the project record.

Issue. Whether the Forest Service adequately assessed the status of endemic species in the project area.

Appellants assert:

the analysis relied on old, inadequate survey information; the FEIS failed to discuss information about the possible presence of and risks to endemics on Suemez Island that was available even without additional surveys.


Discussion

The 2008 Forest Plan Amendment (p. 4-97) includes the following direction:

Use existing information on the distribution of endemic mammals to assess project-level effect. If existing information is lacking, surveys for endemic mammals may be necessary prior to any project that proposes to substantially alter vegetative cover.. .Surveys are necessary only where information is not adequate to assess project-level effects.

The text goes on to state that the "extent and rigor of surveys will be commensurate with the degree of existing proposed forest fragmentation, and the potential risk to endemic mammals that may be present" (Id.).

The WildlifelBotanylSubsistence Resource Report indicates that University of Alaska-Museum personnel conducted 412 trap nights on Suemez Island during 1993. District personnel also conducted 34 trap nights during the summer of 2000. For those 446 trap nights, only "[tlhree species of small mammals were trapped off Suemez, sorex, peromyscus and microtis" (p. 19). Additional trapping conducted during 1998 and 1999 documented the presence of flying squirrels, wolves, and at least one ermine (FEIS, Appendix B, p. 30). The number of trap nights for the 1998 and 1999 surveys are not reported in the FEIS. The project record does not show any survey efforts more recent than 2000; however, nothing indicates that additional trapping would be any more successful or provide different information on the distribution of endemic mammals in the project area.

The Scratchings FEIS contains a table, constructed from multiple data sources, that details the number of small mammals that have been trapped from Suemez and other nearby islands (Table 3W-11, p. 3-42). Contrary to appellants' assertion that an outside expert, Dr. Cook, suggested 30 individuals per island as a reasonable minimum sample size, the April 2006 Tongass Forest Plan Interagency Conservation Strategy Review reports that Dr. Cook stated "sample sizes of 30,for each distinctpopulation [italics added] would be adequate" (p. 56). A population may be found on multiple adjacent islands. Table 3W-11 indicates that, except for ermine, the level of trapping effort for genetic testing has been met or exceeded.

The WildlifelBotanylSubsistence Resource Report describes and assesses the effects on small mammal species that are no longer considered endemics, including the long-tailed vole (p. 51) and the dusky shrew (pp. 51 -52). This is likely because a draft of the Report was completed in July 2005, prior to the April 2006 Forest Plan Interagency Conservation Strategy Review. The FEIS discusses endemics in general (pp. 3-41 and 3-42), and by the following selected species: Alexander Archipelago wolf (pp. 3-37 through 3-39), Suemez ermine (p. 3-43), marten (pp. 3-43 and 3-44), Prince of Wales flying squirrel (p. 3-44), and Keen's mouse (pp. 3-44 and 3-45). The potential effects of the Scratchings project on endemics in general are disclosed in the FEIS (pp. 3-45 through 3-51).

The Tongass Forest Plan includes a comprehensive conservation strategy to assure viable and well-distributed wildlife populations (1997 Forest Plan FEIS, Volume 4, Appendix N). The strategy is based on a system of Small, Medium, Large, and Very Large Old-Growth Habitat


Reserves (OGR's) and other non-development land use designations. The Scratchings project area contains four small OGR's and one Special Interest Area, which functions as a medium OGR (FEIS, pp. 3-55 and 3-56). All of the small old-growth reserves, as adjusted with the Selected Alternative for the Scratchings ROD I, exceed Forest Plan criteria for amount of productive old growth. As stated in the FEIS (Appendix B, p. 32), "sufficient habitat for endemics species will be maintained due to the Forest Plan Conservation Strategy."

The FEIS discusses additional aspects of the conservation strategy, including productive old- growth acres and percent change by alternative (Table 3W-12, p. 3-45), landscape connectivity (pp. 3-67 and 3-68), fragmentation and patch size (pp. 3-68 and 3-69), acres in "interior productive old-growth habitat patches" (pp. 3-69 through 3-73, including Figure 30G-I), and wildlife travel corridors between old-growth habitat patches (pp. 3-73 through 3-75). The potential effects of the project on these factors, by alternative, are discussed on pages 3-76 through 3-79 of the FEIS.

Conclusion

In my opinion, Scratchings project is consistent with the conservation strategy for the Tongass Forest Plan, and the FEIS and Wildlife/Botany/Subsistence Resource Report for the Scratchings FEIS adequately discuss and disclose the potential effects of the project on endemic mammals in con~pliance with NEPA. The project record indicates that there was a considerable prior effort directed toward small mammal surveys; no additional surveys were necessary in order to assess project-level effects on endemic species.

Issue. Whether the FEIS complies with NEPA in its discussion and estimate of public costs incurred by the Scratchings timber sale.

Appellants assert:

the Scratchings Timber Sale FEIS lacks an accurate and complete discussion of public costs, including Forest Service administrative costs; the analysis underestimates timber sale costs and overstates demand; the public costs of road construction, decommissioning, and other road costs should have been included in the timber harvest economics analysis.

Discussion

NEPA requires the disclosure of effects on the human environment, not the administrative costs of preparing and managing timber sale projects. The task for the agency is to weigh the economic and other benefits of the project against its environmental costs. The Forest Service Handbook (FSH) 2409.18 provides guidance on how to conduct a financial efficiency analysis of a proposed timber sale during NEPA analysis. As indicated in the handbook, public investment costs do not have to be calculated to meet Handbook direction.

Table 3 1-10 on page 3-1 8 of the FEIS illustrates the ~ d r e s t Service costs and net revenue by alternative, as suggested by FSH 2409.18, including Forest Service administrative costs. In


doing so, the FEIS uses the average costs across the Alaska Region for administering timber sales. These costs are based on calculations outlined in the Declaration of Forrest Cole (lodged with the District Court in NRDC v. Fovest Service, Case Number 504-010CV-JKS (July 2004)). These administrative costs, cited on page 3-12 of the Scratchings FEIS, are based on an average cost per thousand board feet for timber sale planning and support. These costs are only estimates, as the actual costs and revenues will not be finally determined until the sales are sold, but they are useful for comparing the relative differences between alternatives.

In deciding how much timber to offer for sale in any given year, the agency uses the Morse methodology (Morse 2000). That methodology has the advantage of being self-correcting in that when actual harvest falls below demand projections, offerings for future years are reduced. The methodology also adjusts for changes in mill capacity due to openings and permanent closures of facilities.

The effect of underestimating timber demand is much more serious than overestimating demand. When the agency underestimates timber demand, mills can close for lack of adequate timber supply. Conversely, if the agency prepares more timber than is demanded, the excess timber will not be sold and less environmental impacts will occur. Timber demand on the Tongass has always been volatile, and can differ significantly from actual harvest in any given year or series of years. The spreadsheet in the planning record entitled "copy-of vol-under-contract-2007fy," indicates that the portion of the Scratchings ROD I that has alreadybeen offered sold six months after the decision was made, indicating a demand for timber in the area. Table 3 in "Timber Demand Final 2005" (Document 648) indicates that the average time from award date to first harvest for timber sales on the Tongass is 6.9 months. In the most recent two years (2003 and 2004 illustrated in Table 3), the majority of timber sales sold on the Tongass had a shorter lead time. This indicates that buyers needed the wood from the sales they purchased right away, and demand for timber is high. In a market with plenty of supply, mills could be expected to hold timber sales for as long as two years, keeping an inventory available to meet demand. Table 1 in the same document indicates that the current rate of total mill capacity utilization in southeast Alaska is 11 percent; that is far too low to sustain the industry for long. Sawmills in southeast Alaska need timber to sustain operations.

FSH 2409.19 provides additional direction regarding the type of information to include in a financial efficiency analysis. Costs for construction, reconstruction, and maintenance of roads are included in the analysis (FSH 2409.18, Chapter 30, p. 19). The transportation section of the FEIS, beginning on page 3-178, includes an assessment of road costs in Table 3TR-2 (p. 3-186). In addition, "Estimated Road Costs Scratchings FEIS" (Document 867) identifies the estimated costs for road maintenance. The calculation of other road costs is separate from the timber sale financial efficiency analysis because it has elements of long-term transportation planning unrelated to individual timber sales.

A primary reason for transportation planning is to prevent additional maintenance costs of roads by storing or decommissioning them (Document 796, Suemez Island Roads Analysis - Access and Travel Management Plan, p. 19). Once NFS roads are stored or decommissioned, there should not be any post-sale costs associated with them unless severe weather events cause damage. In addition, since temporary roads are not part of the~forest transportation system and


are not necessary for long-term resource management, they will be decommissioned upon completion of timber sale activities. Constructing and decommissioning temporary roads are typically paid for by the timber purchaser (FSH 2409.18, Chapter 30, p. 16) and do not require Forest Service maintenance activities or expenditures. As a result, there is no need for further analysis of costs that occur in a "post-sale" scenario.

Conclusion

The Scratchings FEIS has an accurate and complete discussion of estimated public costs, including Forest Service administrative costs, and complies with NEPA. The analysis used the best available information to assess timber sale costs and timber-demand, in compliance with FSH and other agency direction. In addition, the financial efficiency analysis considered, as required, the key elements related to the timber sale.

Issue. Whether the effects of the project on recreation, scenery, and watersheds will affect natural resource related employment.

Appellants assert:

there will be a diminishment of salmon runs, fishing, and tourism because the ROD stated that the impacts to recreation, scenery and watershed will be moderate or less; that by arbitrarily knocking these anticipated impacts [to recreation scenery, and watershed] back to zero, the ROD is arbitrary and not factually grounded.

Discussion

The regulations implementing NEPA at 40 CFR 1502.16 require agencies to consider the potential direct and indirect effects of proposed activities, and their significance, on the environment. The regulations go on to define these effects as:

Direct effects, which are caused by the action and occur at the same time and place (1 508.8(a));

Indirect effects, which are caused by the action are later in time or farther removed in distance, but are still reasonably foreseeable. Indirect effects my include growth inducing effects and other effects related to induced changes in the pattern of land use, population density or growth rate.. .(1508.8(b)).

With regard to the potential effects of the project on recreation,the FEIS discusses the direct and indirect effects of the project and concluded that "[r]ecreation opportunities would not be negatively affected by any alternatives" ((FEIS, p. 3-219 to 3-221). As the FEIS indicates, much of Suemez Island recreation is primarily linked to the road system, and the use of off-highway vehicles (OHV) for hunting is one of the main recreation activities (p. 3-215). The FEIS discloses that there will be a decrease in roads open to OHV travel following the closure and decommissioning of roads under several alternatives, but the effects on OHV users were still considered to be minimal due to the small number of current OHV users and their continued


ability to use the Port Refugio area (p. 219). The FEIS also indicates that there are no outfitterlguides current using Suemez Island; recreation use was estimated to be less than 200 people in the last five years, with less than 20 people observed in 2006; and displacement of visitors was not likely due to the sinall nuinber of users (p. 3-220). The EIS also determined that cumulative effects on recreation opportunities were not likely (p. 3-221). Based on the very low level of recreation use by the public and no recorded ontfitterlguide use, I see nothing that would lead me to conclude that natural resource employment will suffer as a result of the Scratchings project.

The EIS discloses the effects of the project on scenery (FEIS, pp. 3-210 to 3- 212). The harvest units are located in the Timber Production and Modified Landscape LUD's, and the FEIS describes the visual disturbance for those LUD's. As indicated in the FEIS, "[all] action alternatives will either meet or exceed a higher level of visual quality than the Adopted Visual Quality Objective for those areas within the project boundary as designated in the Tongass Land and Resource Management Plan [1997]." While there are effects on scenery, the analysis does not indicate that there will be a noticeable effect on tourism.

Appellants also assert that salmon runs will suffer as a result of this decision. The FEIS indicates that there could be direct and indirect effects on watershed and fisheries resources as a result of an increase in low-intensity, short term sedimentation (FEIS, p. 3-130). These effects could be mitigated by taking several measures to minimize the risk, include maintaining appropriate buffers, avoiding steep and unstable slopes, and utilizing harvest methods and road construction techniques that minimize risk of slope failure. As stated in the FEIS, "[tlhese guidelines have been incorporated into each of action alternative. None of the alternatives are expected to have a major effect on fish habitat or water quality" (p. 3-131). The cumulative effects section of the FEIS for these resources disclosed that the action alternatives "would result in between four and seven watersheds exceeding 20 percent harvest in the last 30 years," with varying effects on water yield (p. 3-137). Despite this, the FEIS concluded that "[all1 action alternatives would result in reduced open road lengths and hence reduced cumulative effects to fish habitat and passage. The risk to fish habitat and passage is reduced by removing culverts and eliminating the potential for plugged pipes" (Id.). In addition, the effects on marine and freshwater essential fish habitat would be avoided or minimized by following the Forest Plan standards and guidelines and implementing BMP's (pp. 3-145 and 3-146). In my opinion, the potential effects of the project have been considered and disclosed and the analysis does not support the appellants' assertion that salmon runs will decrease and cause a decline in fishing and other natural resource employment.

In its consideration of the effects of the project on natural resources and its potential effects on associated employment, the Forest was not required to include an economic analysis of all of those effects. The NEPA regulations at 40 CFR 1502.23 state that "...the weighing of the merits and drawbacks of the various alternatives need not be displayed in a monetary cost-benefit analysis and should not be when there are important qualitative considerations." Furthermore, the 2006 court decision in Forest Conservation Council vs. USFS (DC No.CV-02-01293-JCC) affirmed the Forest Service position that it is not required, nor should it attempt to monetize the value of non-timber resources in its site-specific timber sale analyses.


Conclusion

The FEIS clearly demonstrates that the direct, indirect, and cumulative effects of the project on the various resources were considered. In most cases, these effects were projected to be minimal so there is no reason to believe, as the appellants claim, that the effects have been arbitrarily "knocked back" to zero. The analysis supports the Forest Supervisor's conclusion that the potential effects of the project on recreation, scenery, and watersheds are not likely to affect natural resource related employment. In my opinion, this decision is consistent with NEPA and Forest Plan standards and guidelines.

Issue 6. Whether the FEIS and ROD overestimated the demand for yellow cedar. - Appellants assert:

the Scratchings FEIS and ROD overestimate local demand for yellow cedar; that Alaska yellow cedar and western red cedar probably will be surplus to local needs and will be shipped out of Alaska; and the economics of yellow cedar export was not considered in the Scratchings FEIS.

Discussion

The Scratchings FEIS indicates that the amount of cedar proposed for harvest is commensurate with the amount of cedar occurring in the units (p. 3-9, Tables 31-3 and 31-4). While western red cedar and Alaska yellow cedar are important components of a timber sale, they make up a very small percentage of the volume in these units. Alaska yellow cedar comprises three percent of the proposed harvest under Alternative 3 (which is the basis of the Selected Alternative in both ROD I and 11), and western red cedar comprises 13 percent.

The estimated direct income and annualized jobs associated with the Scratchings alternatives are listed on page 3-15 of the FEIS (Table 31-9). These calculations are generated by the NEAT-R model, and are based on the assulnptions embedded in how timber is appraised. Alaska yellow cedar is appraised for export because the purchaser has the right to export it. They do not export all of it, but that is how it is appraised and priced. Western red cedar is appraised for local manufacture. If a buyer wants to ship volume of any species other than Alaska yellow cedar out of Alaska, they must apply for a permit from the Regional Forester after the sale is purchased. The export of Alaska yellow cedar is accounted for in the estimated appraisal price and estimated job calculations generated by the NEAT-R model through the assumption that yellow cedar will generate logging jobs but no sawmilling jobs.

The annual demand for Tongass timber is calculated with the Morse model (see Document 648, Timber Demand Final 2005). Table 2 of this document indicates that local demand for wood was calculated with the assumptions that some Alaska yellow cedar will be locally sawn and some western red cedar will be shipped to the Lower 48. These assumptions are based on recent trends. In order to stay viable, local sawmills need to be able to maximize revenues by selling a variety of products selling into the best markets they can.


Conclusion

The annual demand for cedar, along with all other species of commercial trees, is calculated with the Morse annual demand model. These short-term demand calculations adjust for the recent history of shipments of various species and grades outside Alaska. The estimated employment numbers presented in the FEIS is calculated by the NEAT-R model, which gives estimates of costs and revenues in addition to potential employment. These calculations are based on averages, given the volume and species of timber under each alternative and appraisal rules and laws. The purpose of the information generated by the NEAT-R model is to allow the decision- maker to compare alternatives, using the best available information

Issue. Whether the temporary Tongass National Forest exemption from the Roadless Area Conservation Rule (RACR) is consistent with NEPA.

Appellants assert:

the current temporary Tongass National Forest exemption from the RACR's exemption on road building and timber harvest in inventoried roadless areas is illegal because there was no environmental analysis as required by NEPA.

Discussion

The December 30, 2003, final rule to temporarily exclude the TongassNational Forest from the RACR has a specific section entitled "Need for a Supplemental Environmental Impact Statement (SEIS)" (68 Federal Register 75141). This section discusses the supplemental information report (SIR) the Forest Service completed to determine if there were significant new circumstances or information since the RACR FEIS was completed in November of 2000. As stated in the December 30, 2003, Federal Register notice:

The conclusion in the supplemental information report is that the identified new information and changed circumstances do not result in significantly different environmental effects from those described in the roadless rule FEIS. Such differences as may exist are not of a scale or intensity to be relevant to the adoption of this final rule or to support selection of another alternative from the roadless ruling FEIS. Consequently, the overall decision-making picture is not substantially different from what it was in November 2000, when the roadless rule FEIS was completed. The effects of adopting the proposed rule as final have been displayed to the public and thoroughly considered. For all these reasons, no additional environmental analysis is required.

Conclusion

In my opinion, the need for additional NEPA analysis was carefully examined in the SIR for the final-rule to temporarily exempt the Tongass from the roadless rule. The conclusion of that SIR was that no additional NEPA analysis was needed because previous NEPA documents were sufficient. I concur with that conclusion.


Issue. Whether the decision to exclude the Tongass from the RACR considered the correct acres of roadless areas in development land use designations (LUD's).

Appellants assert:

that approximately 234,000 acres in the 1999 Forest Plan ROD decision that were in mostly natural LUD's were "rolled back" to development LUD's in the 1997 Forest Plan decision, and that the effects of this change were not analyzed.

Discussion

The effects of implementing the 1997 Forest Plan were fblly analyzed and displayed to the public during the development of the EIS for that Plan. This analysis was updated in the Supplemental EIS completed in 2003 (2003 SEIS). The 2000 RACR FEIS assumed that the current forest plan at the time, the 1999 ROD, would be the operative plan under the Tongass Exempt Alternative in that FEIS. However, the 1999 ROD was subsequently vacated by the courts in March of 2001.

The 234,000 acres that changed back to development LUD's (as in the 1997 Plan) when the 1999 ROD was vacated represent 1.4 percent of the Tongass and 2.5 percent of the inventoried roadless acres on the Forest (SIR, p. 14). Re-allocating these areas back to development LUD's as they were under the 1997 Forest Plan reduced the total area allocated to the old-growth conservation reserve system in the 1999 ROD by 1.7 percent. As stated in the SIR completed for the final rule, implementation of the 1997 Forest Plan was determined to not have effects that were significantly different from those of the 1999 ROD (SIR, p. 14, 68 Federal Register 75141, December 30,2003).

The effects on inventoried roadless areas, including the acreage changes, have been analyzed and considered in the 1997 FEIS; the 1999 ROD, the 2000 Roadless FEIS, and the 2003 SEIS.

Conclusion

The fact that the 1999 ROD was subsequently vacated by the courts does not invalidate the analysis contained in the 1997 Forest Plan EIS and the 2003 SEIS, nor does it result in effects that have not already been studied and fully considered in the agency's decision-making process. The 2000 RACR FEIS and the 2003 final rule to temporarily exempt the Tongass rightfully relied on the effects analyses in those documents. In my opinion, the record indicates that the effects have been fully evaluated and no further analysis is necessary.

Issue. Whether the Forest Service provided adequate rationale for the temporary roadless exemption for the Tongass National Forest.

Appellants assert:

the current temporary Tongass National Forest exemption from road building and timber harvest in inventoried roadless areas did not provide any rationale for the exemption.


Discussion

As discussed above in response to Issue 7, the SIR the Forest Service completed for the final rule examined whether there were new significant circumstances or information that would require additional NEPA analysis. The SIR reviewed Economics, Subsistence, Transportation, Recreation, Biodiversity, Social, Minerals, Forest Health, Karst, and Water Quality resources, and the conclusion reached for all of those resources was that the analyses completed for the RACR FEIS is still applicable.

The SIR'S conclusions were included in the final rule that established the temporary exemption. In addition, the preamble to the final rule explained that:

Exempting the Tongass from the prohibitions of the roadless rule returns management of the Tongass to the direction contained in a forest plan that has undergone thorough scientific review, which found the Tongass Forest Plan to be consistent with the available science.

The Department [of Agriculture] has concluded that the social and economic hardships to Southeast Alaska [of implementation of the roadless rule] outweigh the potential long-term ecological benefits because the Tongass Forest Plan adequately provides for the ecological sustainability of the Tongass.

As discussed in the roadless rule FEIS (Volume. 1, pages 3-202,3-326 to 3-350, 3-371 to 3-392), substantial negative economic effects are anticipated if the roadless rule is applied to the Tongass, which include the potential loss of approximately 900 jobs in Southeast Alaska. With the adoption of this final rule, the potential negative economic effects should not occur in Southeast Alaska.

Conclusion

The record indicates there is ample rationale for the temporary roadless exemption on the Tongass National Forest.

Recommendation

In my opinion, the analysis in the FEIS and project record is sufficient to support the Forest Supervisor's decision with respect to the issues raised in the appeal. Therefore, I recommend that you affirm the Forest Supervisor's decision.

If they are not already in the record, the December 29,2006, email stating the private land should be assigned an HSI value of zero and the data file of the deer model runs should be added to the project record.

Appeal Reviewing Officer