Science Peer-Review Summary of the

Wenatchee National Forest's Dry Forest Strategy

Paul F. Hessburg and John F. Lehmkuhl. USDA Forest Service, Pacific Northwest Research Station, Wenatchee, WA

June 1, 1999

In response to a request by the Wenatchee National Forest Leadership Team, we directed a blind scientific peer review of the Forest's Dry Forest Strategy and a case study of its implementation in the Sand Creek Ecosystem Restoration Project. We obtained reviews from six scientists with specific and acknowledged expertise in the fields of fire ecology, forest landscape ecology and management, forest entomology, forest soils, forest hydrology, and wildlife ecology. The reviewers also had research experience working in the eastern Washington ecosystems where the Strategy is applicable.

What follows is a synthesis of the reviewers' findings. The detailed comments of the reviewers are found in their appended reviews. The format for the synthesis states the question we gave to the reviewers and our synthesis of their findings. Each reviewer addressed the two general questions as well as specific questions pertaining to their field of expertise. We developed the general and specific questions that were provided to reviewers from a larger set of questions submitted to us by both the Forest and appellants to the Sand Creek Project. We attempted to make the reviewers' task manageable by capturing in our reduced set of questions the primary issues and eliminated the redundancy found in the larger set of questions.

GENERAL QUESTIONS

Question 1: Does the Dry Forest Strategy incorporate and synthesize pertinent science in a reasonable manner?

In general, all reviewers felt that the Strategy incorporated and synthesized pertinent science in a reasonable manner. Reviewer 1 (fire ecology) provided additional input to enhance the scope of the strategy by volunteering that the strategy should look past the initial treatment time frame and describe long-term restoration needs as well. This reviewer also said that the National Park Service (NPS) has been using prescribed fire without prior thinning for several decades. The NPS has political and legal constraints on its use of timber harvest, but the reviewer suggests that the Forest may have occasion to more broadly consider the use of fire alone to reduce stocking. This reviewer also took exception to the statement that fuel levels are so high that thinning must take place before prescribed fire can be used. A fire-only strategy, the reviewer cites, may be involved and expensive, but it can be implemented in a broad range of cases without prior thinning.

Reviewers 2 (landscape ecology) and 3 (entomology) generally agreed that the Strategy is acceptable and reasonably constructed, but both agreed that descriptions of the natural roles

Science Review Summary: Dry Forest Strategy & Sand Creek EIS 2

of insects and pathogens in these ecosystems were less well developed than the related fire and vegetation ecology.

Reviewer 3 suggests three cautions and potential additions to the Strategy. First, the reviewer says that the document portrays dry forest vegetation and disturbance dynamics as simple and homogenous, whereas in reality these forests are more diverse and complex. Additional discussion of this variability and complexity would enhance the credibility of the Strategy. Second, the reviewer described the most obvious shortcoming of the Strategy as the weak discussion of the multi-scale role that insects (and pathogens) play in influencing disturbance and vegetation patterns. Third, while there is ample scientific evidence to support a strategy that uses thinning, pruning, prescribed burning, and fuels treatments to manage tree density and reduce the hazard of stand-replacing fires, the current document inadequately addresses potential negative consequences of these actions and steps the Forest might take to minimize adverse outcomes. For example, thinning and prescribed burning treatments can scar and damage residual growing stock making such trees potentially more susceptible to certain root pathogens and bark beetles. The possibilities for these consequences should be acknowledged in the Strategy and potential mitigation measures described.

Reviewer 4 (soils) generally agreed that the Strategy is sound and for the most part complete, but agreed with reviewer 3 that potential negative impacts of suggested treatments and mitigation measures need more in-depth description. Reviewer 4 suggested that the Strategy in general inadequately considers soils and their protection against compaction and erosion stemming from management treatments, especially erosive, fine-textured soils, and those with varied texture profiles.

Reviewers 5 (wildlife) and 6 (hydrology) state that the DF Strategy outlines an appropriate course of action for restoration of a condition that will not return to a stable condition without some intervention. The reviewers agree that the science framework is adequate and appropriately states the need for management action. These reviewers cite no significant shortcomings. Reviewer 6 applauds the use of adaptive management and monitoring in the Strategy to learn from mistakes and improve future actions.

Question 2: The Sand Ecosystem Restoration Project is one project conducted under the aegis of the Dry Forest Strategy. Are the issues addressed in the Purpose and Need described in the Sand Ecosystem Restoration EIS consistent with those articulated in the Dry Forest Strategy, and does the selected alternative (Alternative B as modified by the Record of Decision) of the Sand EIS adequately address those issues?

Reviewers generally felt that the Sand Project was a consistent implementation of the Strategy. Reviewer 1 indicates that the Sand EIS addresses the issues described in the Purpose and Need section, and that the selected alternative is consistent with the DF Strategy. This reviewer's only criticism of the selected alternative is that it treats less of the landscape than could be considered ideal for the purposes of moving the vegetation and fuels toward a more sustainable landscape condition. The Project begins to fragment the continuity of fuels

Science Review Summary: Dry Forest Strategy & Sand Creek EIS 3

but leaves a significant portion the landscape untreated and at risk. The reviewer says that it would be beneficial to expand the discussion of the private lands fuel situation in context of the Project area. The reviewer suggests that longer-term plans for repeated burning should be included in the Plan. Reviewer 2 is in general agreement with reviewer 1.

Reviewer 3 answers both parts of Question 2 in the affirmative, but cites the need for better organization and presentation of information in the Strategy and the EIS using standard formats and terminology in order to facilitate cross referencing the documents. Reviewer 4 found the Sand EIS-Purpose and Need is generally consistent with the DF Strategy, with the exception that the EIS, as with the Strategy, may need to consider soils more carefully. This reviewer felt as did reviewer 1 that the extent of the treatments might not be enough to actually protect adjacent untreated areas or private lands from stand-replacement fires, one of the goals of the Strategy.

Reviewer 5 found the Sand EIS and Purpose and Need are in accord with the DF Strategy, while attempting to meet the concerns of the public for recreation. Because of the long-term for restoration of stand structures after treatment, the retention of large remnant trees and snags, should be reconsidered and emphasized to ensure their survival and availability to wildlife. The reviewer agreed with reviewers 1 and 4 in stating that limiting treatments to a relatively smaller area than in other alternatives substantially increases the probability of stand-replacing fires in those areas.

Reviewer 6 also found the Sand EIS consistent with the DF Strategy. The reviewer felt that the project fit well into an adaptive management mode with the presence of large treated and untreated blocks that could be monitored over the long-term to assess the effects of treatment.

SPECIFIC QUESTIONS

Fire Disturbance

Question 1: Considering native fire regimes and biophysical environment conditions as relevant context:

a) Will the treatments and extent of treatment described in the preferred alternative (AIt. B) of the Sand EIS significantly reduce ground fuels and the likelihood of severe or extreme fire behavior under a wildfire burn scenario?

Reviewer 1 indicated that the proposed treatments and treatment extent will significantly reduce problems in the treated areas. But it is still unknown whether the spatial extent of the treatments is sufficient to significantly lower landscape-scale risk. That is, whole landscape management experiments are in their infancy, and no test of the question of extent of treatment has been made to determine critical thresholds for treatment to dramatically reduce risk of conflagration. This reviewer suggests that severe fire behavior will still occur in

Science Review Summary: Forest Strategy & Sand Creek EIS 4

untreated areas, and intimates that there is some potential, perhaps due to the limited spatial extent of treatments, and treated portions may still be at risk.

b) As a result of implementing Alt. B. will landscape patterns of vegetation composition, structure, density, large snags (>25" d.b.h.) and ground fuels be moved toward or away from native conditions? [Please address patterns of composition, structure, density, snags, and fuels separately in your answer.]

Reviewer 1 speculated that the selected alternative will move the Sand Creek ecosystem towards more natural structure, composition, density, large snag and coarse woody debris abundance. Speculation was in part based on the absence of explicit standards in the plan for snag and down wood leave densities and spatial patterns stratified by biophysical setting, and partly due to the necessary and somewhat subjective conceptual assembly of parts of the plan that the reviewer had to accomplish to derive the post-treatment results for residual stands (he had to "crystal ball" what it would look like when the project was completed).

Question 2: Considering relevant science, and considering existing fuel and fire behavior conditions in the Sand Ecosystem as they are described in the EIS (Chapter 3 -- Affected Environment) and those of comparable systems, which treatment options or mixes hold the most promise for moving landscapes toward native structure and functioning? Consider in your answer a wide spectrum of management options, including a) no active management; b) silvicultural treatments; c) prescribed fuels treatments; d) management using prescribed natural fires; e) no active fire suppression-let burn.

Reviewer 1 was strongly averse to implementing no active management, prescribed natural fire, and by extension, no active fire suppression management scenarios. Considering this reviewers earlier comments to the general questions above, it is clear that this reviewer favors active management treatments for achieving the stated goals of the purpose and need using a diverse combination of silvicultural and prescribed fire treatments whose assignment to particular stands is constrained by specific soil, fish, and hydrologic consequences to be avoided. See also this reviewers answer to Question 1 regarding the use of prescribed fire treatments without silvicultural pretreatment of vegetation.

Reviewer 2 is in general agreement with the comments of reviewer 1 to the Fire Disturbance, Question 2.

Question 3: Appellants of the preferred alternative of the Sand EIS advocate that no active silvicultural or fuels management may be needed or appropriate at the spatial scale of project area (i.e., several thousands of acres), but that natural and management-ignited fires should be allowed to bum. Comment on the likely effects of this approach over a 50-100 year time frame on existing fuel conditions, fire behavior attributes, and on the sustainability of existing Late-Successional Reserves

Science Review Summary: Dry Forest Strategy & Sand Creek EIS EIS 5

(LSRs) and Managed Late-Successional Areas (MLSAs) within the Sand Creek Ecosystem.

Reviewer 1 again states that use of naturally ignited fires to address the specific issues outlined in the purpose and need is not a rationale strategy. This reviewer further suggests that it is physically possible to think about doing the job using only prescribed fire treatments, but one would have to simply dismiss the cost of such treatments, which may be quite high. The reviewer states that the new national fire policy does allow for monies to Forests to address fuel problems. Whether funding will be sufficient is unknown.

This reviewer also highlights the conflict between using prescribed fire to reduce fuels and manage tree densities and goals for late-successional forests, which are generally multi-layered with low crown bases. The more fire, the less multi-layered structures. We suggest that in this light, the forest may wish to evaluate natural spatial patterns of fire regime areas within landscapes that are the focus of planning, to identify specific areas of the aforementioned conflict on a project and Forest-wide basis.

Reviewer 1, in describing aspects of a fire alone strategy, suggests that there are potential pitfalls to be addressed when implementing such a strategy in addition to no economic return and no cost recovery. For example, it will be difficult using prescribed fire only to remove the largest of the small size classes; there are ecological consequences of eventual consumption of all (most) woody debris; damage to residual trees; added smoke from logs consumed by fire that could be utilized; visual effect of leaving many small snags; limited control over residual tree spacing.

As regards effects on LSRs and MLSAs, reviewer 1 generally suggests that a fire alone scenario could be successful but perhaps not as successful and with less precision (as to larger time lag fuel classes) than a thin-burn strategy.

Bark Beetle Disturbance

Question 1: Considering native bark beetle disturbance regimes and biophysical environment conditions as relevant context, will the treatments described in the Strategy and the selected alternative (Alt. B -Modified) of the Sand EIS significantly reduce the likelihood of mountain pine beetle and western pine beetle outbreaks?

Reviewer 3 summarized factors that are associated with increasing likelihood of bark beetle outbreaks and stated that the selected alternative will significantly reduce the susceptibility of treated stands to beetle outbreaks.

Question 2: Drawing on your knowledge of pertinent scientific literature and your experience of dry interior West ecosystems, what historical vegetation conditions (ca. 1850-1900) were normally associated with increased patch and landscape vulnerability to pine bark beetles? How does that compare with current conditions?

Science Review Summary: Dry Forest Strategy & Sand Creek EIS 6

Reviewer 3 summarized the pertinent scientific literature describing vegetation conditions associated with increased vulnerability to tree, patch, and landscape mountain pine beetle and western pine beetle disturbance. Based on this synthesis, the reviewer stated that current high density stands in the Sand Ecosystem are likely at high risk to beetle infestations. Occurrence of other predisposing events such as droughts or windstorms could precipitate a major beetle outbreak in this area. Lacking any ameliorative treatments, the reviewer stated that trees in these high density stands will become increasingly stressed from competition and more susceptible to beetle infestation than they are presently.

Question 3: Allowing any bark beetle and fire mortality that might occur, could the goals of of the Northwest Forest Plan (NWFP) for Late-Successional Reserves and MSLAs (EIS Chapter 1 - Purpose and Need) be met without any active silvicultural or fuels management in the Sand Project Area?

Reviewer 3 states that it is unlikely that the goals of the NWFP for LSRs and MSLAs will be met for the long-term in the Sand Creek Ecosystem without any silvicultural or fuels treatment. The reviewer states that without corrective treatments, stand replacing fires will occur at some time in the future, and susceptibility to bark beetle infestation will only increase. The reviewer closes by stating that "maintaining and enhancing old growth forests in the project area will require active management to move the forests in the direction of more sustainable conditions consistent with those that occurred in the presettlement landscape."

Soils

Question 1: Given the native fire regimes, soils, geology, and geomorphic processes that have shaped the landforms of the Sand Creek Ecosystem as context:

a) Were soil erosion and mass movement events normally associated with upland fire and weather disturbances or were such events somewhat extraordinary?

Reviewer 4 said that soil creep and mass movement would be normal in the types of soils and topography found in the Sand Creek watershed, most particularly on steep slopes underlain with sloping bedrock. Wildfire, rain on snow, or extreme storms can all exacerbate the situation on steep slopes. Any soil disturbance is likely to increase soil movement, particularly if vegetative cover and the protective O horizon are removed.

b) Will the treatments proposed in Alt. B result in unprecedented erosion and soil mass movements?

Outside of some exceptional areas, reviewer 4 thought that the treatments would not result in unprecedented soil movement. Winter ground-based and helicopter logging should be no problem. Erosion control measures on upslope landings should be sufficient. Skyline logging, however, may be problematic if trees are not fully suspended. Drag lines in these

Science Review Summary: Dry Forest Strategy & Sand Creek EIS 7

Question 2: Given current climatic trends and the range of natural variability in watershed conditions and hydrology in the affected landscapes, will the Dry Forest Strategy and its implementation in the Sand Ecosystem Restoration Project (Alternative B):

a) Result in unprecedented sediment deposition in stream systems such that aquatic health will be jeopardized in the short- or long-term?

Reviewer 6 considered there is always some short-term risk of extraordinary weather events resulting in excessive delivery of water and sediment to streams from roads and landings. However, a greater long-term risk lies in leaving the forest in a condition where stand-replacement fires could result in major sediment and water delivery occurring during ordinary precipitation events.

b) Result in unprecedented water yield, or the timing and severity of peak flow events, particularly as they relate to severity of flooding in downstream populated areas?

The proposed action alone is unlikely to result in unprecedented flooding, according to Reviewer 6. The estimated changes in water yield and peak flow for Alternative B are well within measurement error of most stream gauges and are therefore not of concern. The estimated water yield increases are well below normal annual variation in water yield.

c) Account for anomalous climate events, such as El Nino, that might produce likewise anomalous flood events?

According to reviewer 6, the proposed Alternative B will move the forest into a more stable situation that will better resist anomalous climate events, be they wet or dry. Alternative B recognizes the problems of roads, yarding, and landings and attempts to minimize the amount of roading and ground yarding that may expose soil. A no management prescription would likely increase the disturbance impacts associated with unprecedented hydrologic conditions, such as flooding due to an El Nino cycle, by increasing the likelihood of a catastrophic fire. Even 'normal' hydrologic events following hot fires result in excessive sediment erosion and transport to streams. Indeed, this happened after the Entiat fire.

Wildlife

Question: Considering the current pattern of dry forest stands and landscapes, with their varied habitats for a wide array of species, their biophysical potential for habitat development, and their native disturbance ecology, will the Dry Forest Strategy and its implementation in the Sand Ecosystem Restoration Project (Alternative B):

a) Maintain or restore amount and spatial pattern of habitats for the tong-term maintenance of plant and animal biodiversity especially species associated with late-successional forest, in affected landscapes?

Science Review Summary: Dry Forest Strategy & Sand Creek EIS 9

Reviewer 5 felt that the return to a more open forest structure and the reduction in fuels will reduce the likelihood of stand-replacement fires and will increase the sustainability of the forest. As the landscape moves to a greater mix of forest types, species associated with open forests should increase as the expense of shade-tolerant species. Species associated with large trees or snags should become more common in the long-term as these are created in areas where they were deficient. The reduction in spotted owl habitat, an umbrella late-successional forest species, appears to be an adequate tradeoff for the greater long-term sustainability of habitat with respect to stand-replacement fire. (See the fire ecologists comments on forest sustainability vs. the area treated).

b) Maintain or restore sufficient stand-level habitats, primarily large or defective '"wildlife" trees and standing and down dead wood produced by fire, insect, and pathogen disturbance?

Reviewer 5 found that the Strategy does not explicitly address this issue, but the inference is that by opening stands and increasing tree size that larger more-persistent snags will be created. The Sand EIS appears to less satisfactorily address snags. The discussion is not clear on how large trees will be retained and exactly what are the targets. More explicit direction should be given on maintaining snags in sufficient numbers, to provide short-term habitat, retaining trees for snag recruitment in the medium term, and restoring large diameter trees to provide snags in the long term. Ultimately, snag leave and recruitment densities should represent a range, and ranges should be unique by biophysical environment setting.

This review was approached by addressing specific questions relative to my field of expertise as posed in the cover letter requesting this review. As this review was intended to be a "blind" review I have not identified myself as the author of the review except in the accompanying letter. The questions for reviewers are highlighted in bold and my answers are in regular type after the question.

Question 1. Does the Dry Forest Strategy incorporate and synthesize pertinent science in a reasonable manner? YES. It has been well-accepted by the scientific community that the historic ecosystems of the Sand Creek area, as with many other dry forest ecosystems, were composed of clumped but widely-spaced tree cover, mostly of fire-tolerant tree species such as ponderosa pine (see references in Agee 1993). The understory was composed mostly of low shrubs, grasses, and forbs that sprouted after being burned. Fire return intervals were as frequent as 2-3 years but probably averaged 10-20 years. The recording of 20-30 fires on individual trees is evidence that the typical fire was not intense. Crown fires were uncommon. The most recent publication that suggests crown fire did play a role in pine forest (Shinneman and Baker 1997) is from the northern Black Hills of South Dakota This forest type is transitional to boreal forest (in fact, white spruce is a shade-tolerant understory tree in that type), and inferences from these northern Black Hills stands are not relevant to the Wenatchee National Forest. Sufficient relevant literature is cited in the "Dry Forest Strategy".

The one criticism that I have of the strategy is that it does not look much past the initial treatment timeframe to what will be the treatment alternatives once the initial "correction" to the stand has been made. Perhaps this is too far into the future to predict, but any landscape plan should address not just the immediate needs but at least conceptually the long-term plans.

Some specific comments meets on the text of the Dry Forest Strategy, referenced to page numbers:

Page 1. The grand fir plant associations are: A. grandis/Arctostaphylos nevadensis; A. grandis/Spirea betulifolia/Pteridium aquilinum.

Page 4. Second par. from end. "Patch" size often refers to the ecological condition created by a fire, which in these forest types was very small (1/4 acre or less). The patch referred to here is the size of the fire, yet in the next year it would be almost impossible to reconstruct, as the major evidence, blackened grass, would be greened up. Thus, the patch is very small but the fire extent is very large (see Agee 1998, for a description of landscape ecology of western fire regimes). Although it is not clear what actually created these ranges of small patch sizes, my guess is that group-kills by bark beetles of old patches maintained the pattern, and that fire cleaned up the coarse woody debris and

prepared the site for a new tree age class. Bark beetles may have been a keystone set of species in these ecosystems, but not at the scale they now act.

Page 6. Lessons from recent fires. Most of the large recent fires burned in types similar to what is seen in Sand Creek. It is not at all without reason to expect similar fires to strike in the Sand Creek area in the future (although I'm not sure it may be possible assign a probability over the next decade as was done in the EIS). Clearly, the threat is there within a biologically meaningful timeframe into the future, and it does require active management. Under "What Can Be Done", the Arno reference is probably Arno and Ottmar 1993 (?) as no Arno 1993 reference is in the literature cited. This is a short, two-page discussion article and has no data that support the statement that thinning is necessary on most areas. The Agee et al. reference (Journal of Forestry?) is not in the literature cited, but Agee (pers. comm.) notes that this article is coming out in Forest Ecology and Management (see literature cited). Its text deals primarily with fuelbreaks, and does not makes any statements as specific as "...it is necessary to reduce fuels by reducing the density of small trees on most areas." The issue of whether thinning is "usually" necessary before burning is an important point to discuss here.

The National Park Service has been using fire for 30 years in dense forests without prior thinning (see van Wagtendonk 1985), so it is clearly possible to use prescribed fire for fuel reduction without thinning in the Sand Creek ecosystem. The NPS has political and legal issues constraining their ability to use timber harvest (with or without burning). and the more relevant question is "what is really desirable?". The desires may include ecological and economic conditions. Avoiding thinning by saw avoids stumps, but may be useful in some areas (note: in the Sand Creek EIS, almost half of the treated acres will be prescribed fire alone). There are some ecological, visual, and economic issues that make timber harvest an attractive allowable option under some conditions. Thinning can provide more precision and remove the larger of the small size classes more safely than can prescribed fire alone. Fire alone creates a sea of small snags that are visually unattractive until they fall. Harvest can avoid some of these. If no harvest is allowed, all of the dead material remains on site, will fall, and then have to be burned by a second or third fire; enough may be on the ground to create soil heating and root damage to residual trees. Again, harvest will avoid much of this impact. Smoke from fire-killed material that is later burned is another issue that harvest can largely avoid. So harvest can be an important adjunct to burning (see Agee 1993, page 403; Agee 1997) strictly on ecological grounds, but the social and economic effects can also be beneficial (assuming the sale is not below-cost).

Page 12. The statement that fuel levels are so high that thinning must take place first before prescribed fire is used is generally not true. As noted above, the NPS has used fire successfully in heavily fueled stands for a long time. Fire alone is an involved strategy, and usually involved a series of fires; furthermore it is possible that in some stands ecological restoration is difficult using fire alone because some of the "too dense" trees are of sufficient size that fire alone cannot take them out (without taking almost the whole stand). Thinning creates its own fuels, particularly where tops are left in the woods, but

can separate ground fuels from aerial fuels if a low thinning is employed, so there are some tradeoffs with thinning. A combination of thinning and fire can be a quicker, more cost-efficient, and more ecologically-efficient method of restoration than fire alone. However, this is not the same as saying that "prescribed fire cannot be used until other activities occur first." I don't believe that statement is generally true.

Page 15. The intent to "maintain structure, composition, and processes" is admirable, but says very little about what structure or composition or processes are maintained. Fire can be maintained, but at 1994-levels of severity is probably not a desirable process. Adding a phrase such as "within the historic range of variability for that ecosystem" or something similar may be a solution.

Page 16. It's important to note that the management objectives highlighted under #1 are based on desirable changes from current composition. Once restoration was complete in an area, those objectives would have to be altered to be relevant.

Page 17. The "treatments" at the end of the flow diagram are the "options" listed under #2 on page 16 (not clear)? In the lower right of the diagram, following the box labeled Mgmt. Obj. 2, why wouldn't PCT or commercial thinning (option [a]) also be relevant?

Page 21. Under (a), line 6 begins 250 tpa and ends with an open "or" - is there something missing from this line? Both PCT and CT should note that uniform spacing is not desirable. Under b. fuelbreaks, the 16 foot pruning would have no effect on a 12 foot flame length. One would need at least a 30 ft pruning to avoid torching from a 12 ft flame length (and most of the trees would be killed anyway).

Page 22. Fuelwood collection is an end product of fuel treatment, not a structurally-based criterion for fuel treatment. Most mechanical fuel treatments alter structure of the fuel, not the load per se, so a simple descriptor like "25%" of the fuel level isn't very descriptive of the criterion to meet after treatment. Load may be the same but arranged to as to produce less actual fire hazard.

Question 2. The Sand Ecosystem Restoration Project is one project conducted under the aegis of the Dry Forest Strategy. Are the issues addressed in the Purpose and Need described in the Sand Ecosystem Restoration EIS consistent with those articulated in the Dry Forest Strategy, and does the preferred alternative (Alternative B) of the Sand EIS adequately address those issues? YES, and "YES in the proper direction". The Sand EIS is consistent with the Dry Forest Strategy. The preferred alternative treats less of the landscape than what I would consider ideal for the purposes of moving towards more sustainable and natural forest structure. Of the >20,000 acres in the planning area, only 1/3 (about 7,100 acres) of the national forest area is proposed for any treatment, and slightly less than half of those acres will be thinned. This begins to fragment the fuel structure, but leaves a significant portion of this landscape with no treatment at all. One large block (Devils Gulch) is not entered at all

for thinning or prescribed burning. While the proposed treatment does reduce fire hazard somewhat, it leaves >2/3 of the planning area (including the private land) at risk. It would have been helpful for this reader to have a bit more discussion of the private land fuels situation. In summary the project is consistent with the dry forest strategy, but if there is a criticism it is that the project may not go far enough in terms of treated acres. Longer-term plans for repeat burns, etc. would be helpful to include. Is single burn, or harvest/single burn, the only treatment likely for the next 25-30 years? I realize that disturbing the entire area in a short period of time is not desirable, but a long-term strategy would include essentially a maintenance rotation for treatment that would eventually cover most of the acres in the project area. Further, leaving large blocks such as Devils Gulch with no treatment at all provides a large block of land with contiguous fuel, encouraging larger wildfires to occur when and if fire does enter this area.

Fire Disturbance Specific Issue #1: Considering native fire regimes and the biophysical environment as relevant context:

a) Will the treatments and extent of treatment described in the preferred alternative (Alternative B) of the Sand EIS significantly reduce ground fuels and the likelihood of severe or extreme fire behavior under a wildfire burn scenario?

YES, the project will significantly reduce problems in the treated areas. We do not currently have a metric to define precisely how much of the landscape has to be treated for the landscape risk to significantly decline. On the Federal portion of the project area, it appears that the proportions (~50% of federal land within areas that will be treated) are beginning to enter what most specialists would call sufficient fuel fragmentation (at least in the short term), but these are largely seat-of-the-pants estimates (40-70%) of the proportion of landscape necessary to treat. Obviously, in untreated areas such as Devils Gulch, severe behavior will still occur, but the behavior will be altered when the fire enters treated areas, and 1994-type fires are less likely there.

b) As a result of implementing alternative B. will landscape patterns of vegetation composition, structure, density, large trees, and snags (>25" dbh), and ground fuels, be moved towards or away from native conditions? [Please address patterns of composition, structure, density, snags, and fuels separately in your answer].

A most complete answer would be possible if a long-term plan over the next 50 years or so was possible to review. The answer that can be given is that it appears that the Sand EIS alternative B is moving the Sand Creek ecosystem more towards a natural structure, species composition, density, with more natural levels (eventually) of larger snags, and lower (more natural) fuel loads in these ecosystems, in the areas of the project that are treated. In areas where fire alone is used, it tends to thin stands from below, and thus the residual stands will be reduced in density but have large average tree size (immediately, even before the effects of thinning on subsequent tree growth are realized). The lowered density will allow faster growth on remaining trees. The more fire-tolerant trees of any diameter class will be favored. Thinning by fire will typically be a clumped process, with

variable fire intensity leaving few residuals in some areas and too many in others, but overall a good reduction. First using mechanical thinning by saw provides a better control on spacing (leave open grown trees, thin very dense stands, but maintain an overall clumpiness). Thinning can provide more precision and remove the larger of the small size classes more safely than can prescribed fire alone. The standards for residual trees are similar to what fire would leave, favoring larch and pine over fir. Production of snags >25 inches will first require production of live trees of that size and this project is a first step in that direction. Large snags will probably be underrepresented for a long time because most trees are currently small and growth rates on these sites are moderate at best.

Log densities were not clearly identified in the project proposal. In these dry forests, with frequent fire, they were probably clustered around snag patches, and would burn up as the 10-15 year interval fires passed through the forests. While snags were periodically created (usually be group beetle kills), and more created as those already present were being decomposed by fire, logs were probably consumed readily by these fires and represented a spatially and temporally transient resource compared to large live trees or snags.

Fire Disturbance Specific Issue #2: Considering relevant science and considering existing fuel and fire behavior conditions in the Sand Ecosystem as they are described in the EIS (Chapter 3 - Affected Environment) and those of comparable systems, which treatment options or mixes hold the most promise for moving landscapes towards native structure and functioning? Consider in your answer a wide spectrum of management options, including (a) no active management; (b) silvicultural treatments; (c) prescribed fuels treatments; (d) management using prescribed natural fires; and (e) no active fire suppression (let burn).

(a) No active management is a recipe for disaster. We have known this for decades (see Weaver 1943!) and 1994 was a local example. Let's not belabor the obvious.

(b) Silvicultural treatments are a mixed bag. I have seen silvicultural treatments in dry forests in Region 6 in the last 5 years that were absolutely disastrous for fire issues, even though they were planned as forest restoration treatments. On the other hand, properly planned silvicultural treatments can be a real blessing for ecological restoration. The work that has been put into defining these (see Harrod et al. 1999) represents a real contribution, and the local forest is blessed to have this type of analysis available. In general, low thinning treatments with subsequent fuel and small (nonmerchantable tree) reduction by fire is needed. Where low impact yarding can be done, these treatments are even more justified.

(c) I presume this is meant to be prescribed fire treatment. Absolutely essential over much of the landscape, but it can be tied in with silvicultural treatments as desired (see comments under question 1, page 6 above).

(d) I see no rationale at present for the use of prescribed natural fires (pnf) in the Mission Creek watershed. These are naturally-ignited fires that are allowed to burn under monitoring. First of all, they make sense only in large areas of wilderness where constraints on smoke, damage to private lands, etc. are low. Such is not the case here. Furthermore, pnf, or whatever it is called under the new fire policies, applies most to ecological systems where change due to anthropogenic influences is low: fire can "do its thing" in a natural environment. In Mission Creek, the lack of larger natural structure (big trees), abundance of smaller ladder fuels, and increase in fuels generally over natural levels means the average pnf will not be within the range of natural variability for that process (fire). It will be too intense and be natural only in the sense of ignition, not in terms of effects. Prescribed natural fire is NOT a justifiable strategy in this area, and is not even a reasonable option to consider in any depth for this area.

Fire Disturbance Specific Issue #3: Appellants of the preferred alternative of the Sand EIS advocate that no active silvicultural or fuels management may be needed or appropriate at the spatial scale of the project area (i.e., several thousands of acres) but that natural and management-ignited fires should be allowed to burn. Comment on the likely effect of this approach over a 50-100 year timeframe on existing fuel conditions, fire behavior attributes, and on the sustainability of existing Late-Successional Reserves (LSRs) and Managed Late-Successional Areas (MLSAs) within the Sand Creek Ecosystem.

See my comments on naturally-ignited fires above (specific issue fire #2) - they are not a rational strategy in this area. This leaves the question of management-ignited, or prescribed fire, to be addressed as a sole strategy in this watershed. The new national fire policy does allow some monies to be spent upfront to address fuels problems, and it may develop that sufficient money would become available to have a very large prescribed fire project in Mission Creek as envisioned by the "appellants" described above. Costs aside, is it physically possible to think about a project like this in Sand Creek? I would answer yes, but it may not be the best option for reasons described above in the answer to Question 1 (comments related to page 6 of Dry Forest Strategy). Right now, the Sand Creek area is proposed to have about 4,000 acres treated only with prescribed fire, so a significant investment in this strategy by the Forest Service is already evident. Could it be expanded? Physically, yes, with appropriate constraints on fire intensity being met. Whether this is ecologically "best" for the remaining area is not clear. Fire can be used to reduce hazard in the more mesic forest, by reducing ground fuels and raising average height to live crown. However, this may be to some extent in conflict with goals for late-successional forest patches where a more multi-layered structure is desired. From a fire perspective, these tradeoffs are complex. The more fire, the less multi-layered structure but the more sustainable the resulting forest structure is. We (the fire community) don't have good landscape-level answers to these tradeoffs, although new technology such as the GIS-based fire spread model FARSITE (Finney 1998) may help in this regard.

If the prescribed fire alone strategy were employed over a 50 year timeframe (such that >1/2 of the area were burned 2-3 times during that interval), the result during ages 50-100 would be similar to that using a combination of thinning and prescribed burning, with the caveats mentioned in earlier portions of this review (no economic return, so no cost recovery; difficulty of removing with fire the larger of the small size classes; ecological effect of all debris being eventually consumed, and some logs smoldering and damaging residual trees; additional smoke from logs being consumed that would have been utilized; the visual effects of the small snags; and generally less control on spacing of residual trees). Would prescribed fire alone be a disaster? I believe not. But it may not be a preferred means of restoration for the reasons mentioned.

The effect on LSR and MLSA areas depends on how these areas are treated, either with the thin-burn option or the fire-only option. Clearly, the current plan (Alternative B) includes thousands of acres of fire-only treatment already, so the recognition of fire as a sole tool is already in place for some portions of the landscape. So if fire alone were used in these areas, it could be used with success on at least part of the LSR-MLSA landscape. For the Sand Creek MSLA landscape, with 78% of the land in the "dry forest" category, some sort of active management is necessary. Fire alone could be successful (but perhaps not AS successful as combining it with thinning) in treating this area, subject to the caveats mentioned in the previous paragraph. Obviously, the more of the surrounding lands that are treated, the higher the probability that LSR lands can be protected from wildfires that might otherwise enter from those areas. Both the LSR and MLSA assessment appropriately introduce the concepts of thinning and burning, with priorities to the drier sites in both areas. Again, the thinning can provide more precision and remove the larger of the small size classes more safely than can prescribed fire alone, but prescribed fire alone could be used (perhaps tied into to fuelbreaks where thinning would be done [see the Agee et al. paper (in press)].

Untreated areas will remain, largely in the more mesic types, where higher intensity fire was historically more probable, and when or if wildfire occurs there, some intensity levels above historic may occur (if, for example, a mesic forest historically burned with 1/3 of land each in low, moderate, and high severity, it may now burn with 10% low, 30% moderate, and 60% high severity. This may be within the range of some given fire of the past, but is higher than the overall landscape percentages of the past). But until we have a better handle on the drier forest problem, the mesic forest will likely remain a lower priority.

Agee, J.K. 1993. Fire ecology of Pacific Northwest forests. Island Press. Washington, D.C.

Agee, J.K. 1997. The severe weather wildfire -- too hot to handle? Northwest Science 71: 153-156.

Agee, J.K. 1998. The landscape ecology of western forest fire regimes. Northwest Science 72 (special issue): 24-34.

Agee, J.K., B. Bahro, M.A. Finney, P.N. Omi, D.B. Sapsis, C.N. Skinner, J.W. van Wagtendonk, and C.P. Weatherspoon. (in press). The use of shaded fuelbreaks in landscape fire management. Forest Ecology and Management.

Finney, M. 1998. FARS1TE: Fire area simulator. Model development and evaluation. USDA Forest Service Research Paper RMRS-RP-4.

Harrod, R.J., B.H. McRae, and W.E. Hartl. 1999. Historical stand reconstruction in ponderosa pine forests to guide silvicultural prescriptions. Forest Ecology and Management 114: 433446.

Shinneman, D.J, and W.L. Baker. 1997. Nonequilibrium dynamics between catastrophic disturbances and old-growth forests in ponderosa pine landscapes of the Black Hills. Conservation Biology 11: 1276-1288.

van Wagtendonk, J.W. 1985. Fire suppression effects on fuels and succession in short-fire-interval wilderness ecosystems. pp. 119-126 In: Lotan, J.E. et al. (eds) Proceedings Symposium and workshop on wilderness fire. USDA Forest Service Gen. Tech. Rep. INT-1 82.

Weaver, H. 1943. Fire as an ecological and silvicultural factor in the ponderosa pine region of the Pacific slope. Journal of Forestry 41 (1): 7-15.