The Ecology and Management of Dead Wood in Western ForestsNovember 2-3, 1999

Reno, Nevada

Abstracts

The Charcoal Filter As a Coarse Filter for Coarse Woody DebrisJames K. AgeeFire played an important role in maintaining and creating conditions suitable for native flora and fauna in the forests of western North America. Recent coarse filter conservation strategies have advocated creating future landscapes that incorporate historic or natural ranges of variability, including fire regimes. Fire varied in frequency, intensity, extent, season, and interactions with other disturbance processes. In so doing, it had quite different effects on production/consumption of coarse woody debris (CWD). In low-severity fire regimes, fire limited increases in forest basal area, and consumed logs at 5-15 year intervals. Coarse woody debris was stable was low at a local ! sc!!!!ale. At the other end of the spectrum, in the high-severity fire regime, fires were infrequent but killed most of the trees across large areas, creating a pulse of coarse woody debris that declined over time and later increased. The most complex CWD dynamics were in the moderate-severity fire regimes, where fire underburned, thinned, and patch-killed trees at an intermediate scale. CWD was created and consumed at 25-75 year intervals, maintaining fairly high levels of CWD on a continuous basis. Implications for management are discussed.

Managing Coarse Woody Debris in British Columbia's Forests: A Cultural Shift for Professional Foresters?Andre ArsenaultCoarse woody debris (=cwd) is recognized as an important component of our forest ecosystems linked to biodiversity and ecosystem processes. Cwd are high centres of biological interaction and energy exchange symbolizing in many ways the complexity of forest ecosystems. Our research on lichens and bryophytes demonstrates that many species are either partially or entirely dependent on cwd. Long-term management of this resource is vital to maintain ecosystem integrity. Previous attempts to recommend province-wide cwd management practices were not successful because of potential increased logging costs and conflicts between utilization standards and recommended cwd volume targets. In ! th!!!!is presentation I will demonstrate that small changes to existing forest practices will go a long way in minimizing impacts on the cwd resource without affecting logging costs or timber access. I propose that a cultural shift in how foresters perceive cwd is a necessary ingredient to improve present practices. I will attempt to initiate this cultural shift by explaining key cwd management principles and by providing some operational examples.

The Pileated Woodpecker As a Keystone Habitat Modifier in Coastal Forests of the Pacific

NorthwestKeith B. Aubry and Catherine M. RaleyThe pileated woodpecker is the largest North American woodpecker and the only species capable of creating large cavities in hard snags and decadent live trees. Pileateds excavate multiple cavities or cavity-starts during the breeding season and excavate openings into hollow chambers for roosting throughout the year. A wide array of species, including many that are of management concern in the Pacific Northwest, use old pileated nest and roost cavities. In addition, pileateds excavate deep into both sapwood and heartwood to forage on invertebrate prey, providing access to such prey for other species. Pileated excavations also accelerate decay processes and nutrient cycling by crea! ti!!!!ng openings in decadent live trees, snags, and logs, and breaking apart both fresh and decaying wood; they may also facilitate inoculation of heartwood decay fungi by exposing fresh wood in live trees to airborne spores. Keystone species are species whose effects on ecosystems are disproportionately large relative to their abundance, and include habitat modifiers whose activities create habitat for many other species and strongly influence key ecological processes. We propose that the pileated woodpecker is a keystone habitat modifier in Pacific Northwest coastal forests and argue that their habitat needs should be afforded special attention in forest management plans and monitoring activities.

Ecology and Management of Deadwood in Forested Ecosystems - A Synthesis of Lessons LearnedAnn M. BartuskaPursuing an ecological approach to management of natural resources has resulted in a new way of viewing dead and dying trees. We have learned about the structural (coarse woody debris, snags) and functional (decomposition, carbon release and sequestration) attributes, and the dynamics of organisms that use "deadwood" for food and shelter. All these attributes underscore the role deadwood contributes to a diverse and resilient forested ecosystem. Deadwood integrates aquatic and terrestrial environments, forcing wildlife and fisheries biologists, siliviculturists, and forest managers to look at the interplay of in-stream, riparian and upland processes. Our collective progress o! ve!!!!r the last decades in understanding the role of deadwood and changes in our management built upon this knowledge has been truly remarkable.

Estimating Snag and Large Tree Densities and Distributions on a Landscape for Wildlife Management.Lisa J.Bate, Edward O. Garton, and Michael J. WisdomWe provide efficient and accurate methods for sampling snags and large trees on a landscape to conduct compliance and effectiveness monitoring for wildlife in relation to the habitat standards and guidelines on National Forests. Our USDA Forest Service General Technical Report (PNW-GTR-425) includes spreadsheets, macros, and instructions to conduct all surveys and analyses pertaining to estimation of snag and large tree densities and distributions on a landscape. Methods focus on optimizing sampling effort by choosing a plot size appropriate for the specific forest conditions encountered. Two methods are available for density analysis. Method one

requires sampling until a desired! p!!!!recision level is obtained for a density estimate. Method two is intended for use in areas that have low snag densities compared to management plan targeted densities. After taking a minimum of 60 samples, one may test for a significant difference between the estimated and targeted densities. In addition, data can be used to calculate a distribution index. The value obtained from the distribution index indicates whether the current distribution of target snags and large trees across a subwatershed is adequate to meet the habitat needs of territorial cavity-nesters and other wildlife species. The methods can also be used to evaluate wildlife use of snags in an area.

Estimating the Density, Length, and Percent Cover of Downed Logs on a Landscape for Wildlife Management.Lisa J.Bate, Torolf R. Torgersen, Edward O. Garton, and Michael J. WisdomWe investigated which sampling methods were best for obtaining estimates of the density, total length, and percent cover of logs important to wildlife species on a landscape. We sampled 18 mixed-conifer stands, one to seven hectares in size, in the Blue Mountains of Oregon and the Salish and Mission Mountains of Montana. These stands contained a gradient of log resources based on their density, total length, and percent cover. Stand conditions also varied based on their harvest history, seral stage, shrub cover, and topography. Complete counts of logs were first conducted; subsequently, variations of the strip-plot and the line-intercept method were employed. Each method was then! e!!!!valuated for its precision, accuracy, and efficiency. Results from our analysis were used to create all the necessary spreadsheets, macros, and instructions to conduct log surveys on a landscape. Log sampling methods will be compatible with previously developed sampling methods for snags and large trees. This will allow resource specialists to simultaneously sample for all three habitat components on a landscape: large trees, snags, and logs.

Effects of Wildfire on In-channel Woody Debris in the Eastern Sierra NevadaNeil Berg, David Azuma, and Ann CarlsonManagement of in-channel woody debris after wildfire is controversial. Post-fire increases in stream discharges can cause downstream flooding. The resulting heightened transport and accumulation of debris can wash out bridges and cause other damage. In-channel debris is often removed or cut into smaller pieces to expedite flushing through the system and to avoid debris jam formation. Biotic uses of debris for fish cover, triggering pool formation, storing sediment, and as food sources for invertebrates and micro-organisms are lost or reduced when debris is removed or cut up.Little quantitative information exists on debris dynamics after wildfire, and no known data are available for the Sierra Nevada. We quantified changes in debris frequency, mobility, volume, aggregation and carbon loading after a 1994 wildfire in the eastern Sierra Nevada through before and after comparative measurements on Badenaugh Creek, and by comparing selected attributes to a nearby "control" stream. Fifty-seven percent of wood volume, and 25 percent of the pieces, were consumed by the fire. The fire reduced aquatic carbon loading from about 2 _ to 1 _ times

terrestrial loadings. Although more pieces moved one year after the fire on Badenaugh Creek than on the control ! st!!!!ream, the size and number of debris jams both immediately and one year after the fire were appreciably reduced from pre-fire levels, probably because fewer pieces were available to form aggregates. Decisions on the management of post-fire debris should consider the interaction between fire intensity, channel width and the size of the remaining wood. If few pieces of channel-spanning length remain post-fire they may pose little downstream danger.

Primary Cavity Excavator Birds in Managed and Unmanaged Grand Fir Forests of Eastern WashingtonKenneth R. BevisPopulations of Primary Cavity Excavator (PCE) birds are hypothesized to vary with availability of standing dead trees (snags). This hypothesis was tested in a grand fir (Abies grandis) forest, a dominant east cascades forest type in 1991 and 1992. Characteristics of snags utilized by these birds were also examined. Seven plots in a continuum of forest management intensity were measured for PCE bird abundance, habitat variables and snag selection. PCE birds selected larger diameter snags with advanced decay for feeding and nesting. PCE species composition shifted across the range of plots. Regression models showed total populations as best predicted by densities of large! d!!!!iameter snags. Forest management implications involve long term maintenance of snags in adequate densities and advanced decay states in order to maintain PCE birds.

Dead Wood Dynamics in Aquatic EcosystemsRobert BilbyStream and river ecosystems are intricately interconnected with the terrestrial ecosystems through which they flow. In forested landscapes one of the most obvious indications of this connection is the great abundance of wood deposited in stream channels. Dead wood in streams affects the structural, functional and biological attributes of these systems. Wood influences channel form by creating pools and waterfalls, altering channel width and depth and creating conditions conducive to the deposition of sediment and organic matter. Organic matter accumulated by large woody debris is an important food source for many invertebrates. The pools formed by wood in streams provide habitat ! fo!!!!r fish and the invertebrate production associated with wood increases food for fish. Wood and the sediment it accumulates provides preferred germination sites for many common riparian plants. Wood quantity in streams flowing through many forested regions of the world has decreased over time as a result of practices that removed wood from channels or eliminated the future source of wood by removing trees from riparian areas. Placement of wood in streams has been used a to correct impacts of past practices. However, this approach is expensive and not always effective. Comprehensive restoration of aquatic ecosystems in forested landscapes will require conservation and restoration of riparian areas.

Created Snags and Wildlife, Is There a Connection?Pat Boleyn, Eric Wold, and Ken ByfordStanding dead trees are important resources for vertebrate species worldwide. Despite current understanding of their importance for wildlife, snags had been eliminated from managed landscapes. Within the past decade, however, most land management agencies have devised strategies to maintain snags. This includes creating snags from live trees and leaving them in stands after cutting. Little information exists on the use of these snags by wildlife. This study had the following objectives: 1) to document whether created snags were used by wildlife and 2) to elucidate factors associated with the use of created snags by wildlife (primarily cavity-nesting birds) at the stand and tree l! ev!!!!el. Fieldwork was in 1997 and 1998 in 55 randomly selected stands across the Willamette National Forest in Oregon. We documented sign of woodpecker use, along with snag and stand characteristics. Of the 1267 created snags sampled, approximately half were used by cavity nesting birds. Analyses using logistic regressions demonstrated that foraging use of created snags by cavity nesting birds was more likely in snags that were completely dead (some are left alive after attempted killing)and in those that had old woodpecker sign present. Use of created snags by Pileated Woodpeckers was also analyzed. Implications for management are discussed.

Influence of Snag Angle-of-Fall on Riparian Large Woody Debris RecruitmentDon C. Bragg and Jeffrey L. KershnerRiparian large woody debris (LWD) recruitment models usually assume that snags fall in random directions. However, little empirical work to verify this assumption or the consequences of its violation has been done. To test the impact of snag fall direction, the riparian LWD recruitment model CWD (version 1.4) was used to examine fall patterns including random, trimodal, and 3 types (towards the channel, quartering towards the channel, and parallel to the channel) of unimodal. For streams with both banks forested, there were no statistically significant differences between the 300 yr cumulative LWD recruitment for random (mean - 79.7 m3 100 m reach-1! !!!!) and trimodal (mean - 78.4 m3 100 m reach-1) patterns, as well as the unimodal pattern (mean = 77.6 m3 100 m reach-1) that quarters towards the channel (all P>0.85). The unimodal pattern falling towards the channel had significantly greater cumulative recruitment than any other treatment (mean = 90.7 m3 100 m reach-1, P<<0.001), while the unimodal pattern paralleling the channel provided the least debris (mean = 59.0 m3 100 m reach-1, P<<0.001). The results differed slightly when only one bank of the stream was forested. There was no significant difference between the random and trimodal snag fall patterns (means = 39.9 and 38.4 m3 100 m reach -1, respectively, P>0.80) and the unimodal pattern directed towards the channel contributed the most LWD for this series of runs (mean = 76.6! m!!!!3 100 m reach-1). The unimodal patterns quartering towards the channel delivered more than the random or trimodal patterns (mean = 64.0 m3 100 m reach-1, P<<0.001) while the unimodal pattern parallel to the channel delivered less than all other treatments (mean = 27.5 m3 100 m reach-1, P<<0.001). This work suggests that LWD recruitment models should not assume random snag fall patterns if they are to accurately simulate LWD recruitment, especially in areas strongly

affected by unimodal failure controls (e.g., prevailing wind patterns) or those dominated by partially forested stream banks.

Entrainment and Deposition of Logs in StreamsChristian A. Braudrick and Gordon E. GrantLarge wood is an integral component of forested streams, yet we know little about thresholds of wood movement and deposition. We used physically-based models and flume experiments to investigate when wood moves in streams and a second set of experiments to asses how far wood moves once in transport. Both our model and the experiments indicate that log entrainment is primarily a function of the piece angle relative to flow direction, whether or not the log had a rootwad, the density of the log, and the piece diameter. Stability increased if the pieces had rootwads or were rotated parallel to flow. Although previously reported as the most important factor in assessing wood stabilit! y,!!!! piece length only influenced the entrainment of pieces when logs were oriented parallel to flow, and had no influence on pieces oriented normal to flow. However, piece length does influence the distance logs travel once they are in motion. Travel distance decreased as the ratio of piece length to both channel width and radius of curvature increased. These experiments and physically-based models offer a first-order approach to evaluating the stability and dynamics of both naturally derived and management emplaced logs.

Creating Wildlife Habitat Structures in Snags, Logs, and StumpsTimothy Kent BrownUp to 2/3 of our wildlife species use dead wood structures or woody debris for some portion of their life cycles. Unknown numbers of insects and invertebrates also inhabit dead wood. In order to utilize wood, wildlife must excavate their own cavity or use a cavity created by some other organism or process. By weilding a chainsaw to sculpt cavities in logs we can create or restore den and nest sites for a variety of cavity and shelter dependent wildlife. Techniques for working around snags and logs or modification of these structures to render them safe are presented. I will focus on my experience of maintaining, improving, establishing habitat for an array of cavity dependent wil! dl!!!!ife in a diversity of landscape settings from wilderness areas to backyard urban forests. These structures have been created in forests from Alaska to California as well as other states throughout the United States. Habitat improvement techniques for bats, amphibians, birds, carnivores, bears, the tinyest deer mice and pollinating insects are presented. I will show samples of dens and cavities, bat roost structures and other created structural forms from among my nearly 1900 techniques for working in trees, logs, stumps and planted structures and snags. I will illustrate techniques for creating successful den, cover and feeding sites for wood ducks, bats, bears, marten, squirrels and mice.

Maintenance and Improvement of Snags and Logs for Pollinating and Predaceous InsectsTimothy Kent BrownA myriad of insects can be found inside snags and downed woody debris. Some are primary cavity excavators like beetles, and ants that create cavities for protection and rearing their young.

Some cavities in living trees and snags may be caused by wood rotting fungi. Many insects are opportunistic,...using cavities of all kinds for their own reproduction or for overwintering habitat. Predaceous and pollinating insects like ants, wasps, bees and ladybugs contribute to forest health and productivity. Ants and ladybugs, for instance prey on a number of forest insect pests. Solitary and colonial bees pollinate a variety of flowers and berry-producing shrubs. Special structures and ins! ta!!!!llations for pollinating and predaceous insects in standing and downed wood materials may be created by various mechanical means. Techniques to maintain and create insect habitat structures in trees, snags and downed wood in forests and urban woodlots are presented.

Maintaining, Placing and Creating Suitable Logs for Wildlife in Aquatic HabitatsTimothy Kent BrownA variety of wildlife use wood in aquatic wetland habitats. Migratory and resident aquatic-dependent birds, mammals, amphibians, and insects use aquatic logs. Logs can be added to suitable aquatic habitats to enhance wildlife productivity. A multitude of benefits are derived from these logs. Aquatic logs provide suitable loafing sites for geese and ducks in urban parks. Turtles bask on logs. Salamanders secure eggs to fine branches attached to recently deposited logs. Trees with branches can be secured in estuaries and intertidal areas to provide nest substrates for herring. Logs have been modified in some scenarios for nesting cover. Water level fluctuations are mitigated by anc! ho!!!!ring logs properly in safety zones. This talk will discuss techniques to mitigate water level fluctuations impact on amphibians.

The Value of Dead Wood to Vertebrates in the Pacific NorthwestEvelyn BullMany species of birds, mammals, amphibians, and reptiles utilize standing and downed dead wood. Woodpeckers depend on decayed wood for the excavation of nest and roost cavities in standing trees. Both avian and mammalian secondary cavity nesters then claim the abandoned cavities for their nesting or roosting. Many of the woodpeckers and secondary cavity nesters forage on forest insect pests. Factors that affect the type and extent of vertebrate use of dead wood include the physical orientation, size, decay state, tree species, and overall abundance.Some types of heartwood decay organisms create hollow chambers in living trees which eventually die to become hollow dead trees. Hollow standing trees are used by Vaux's swifts (Chaetura vauxi) for nesting and roosting, pileated woodpeckers (Dryocopus pileatus) for roosting, black bears (Ursus americanus) for overwintering, American martens (Martes americana) for denning and resting, and many other species of small mammals for shelter. Once the trees fall, many of the same species continue to use the hollow structures except the avian species.

Solid logs provide cover or travel lanes for small mammals. Accumulations of logs stacked on top of each other provide important habitat in the open spaces formed under the snow where martens and small mammals spend much of the winter. Large-diameter logs are used extensively

by pileated woodpeckers and black bears for foraging on carpenter ants. Extensively decayed logs provide habitat for amphibians and reptiles.

How Should We Distribute Dying and Dead Wood in Space?Fred L. Bunnell, Mark Boyland, and Elke Wind We consider two aspects of spatial distribution: density and degree of aggregation or clumpiness. Available data suggest that density of all cavity nesters combined responds curvilinearly with density of snags such that there is an upper asymptote of about 2.4 pairs/ha with a half-saturation at about 2.4 large snags/ha. Individual species show different responses, and there is no apparent effect of territoriality among cavity nesting smaller species. Dispersed retention of trees and snags strongly favors secondary cavity nesters and increases their abundance beyond that found in mature or old-growth forests. There are few data on effectiveness of small patches, but large patches ! su!!!!stain all primary cavity nesters within clearcut areas as large as 1,000 ha. Vertebrate responses to various forms of partial cutting are reviewed. Despite good operational and biological reasons for patchwise retention, there are negative effects on some species. Patches create aggregations of downed wood. Several rodent and small carnivore species respond positively to aggregations of downed wood, but dispersed retention of downed wood is important to many species, and especially important to some mosses, liverworts, and possibly fungi.

How Dead Trees Sustain Live Organisms in Western ForestsFred L. Bunnell, Barbara Johnston, and Elke WindFor organisms, the trees of a forest represent a continuum of opportunities to meet life requirements, beginning with live trees, through standing and dead trees, to logs, and finally soil. Most organisms relying on dead wood are not vertebrates, and dead wood contributes to biological richness in four broad ways: substrate, cavity sites, foraging sites, and shelter or cover (logs). Among vertebrates in the Pacific Northwest, 69 species commonly use cavities, while others make opportunistic use of cavities, and 47 species are strongly associated with downed wood. Proportions of the vertebrate fauna using dead wood are related to characteristics of the fire regime. Among weak exca! va!!!!tors, about 70% of reported nest sites are in dead trees; strong excavators make greater use of live trees. Among mammals, most bat roosts are found in dead trees whereas carnivores tend to use living and dead trees about equally. Woodpeckers tend to select similar trees for foraging as for nesting, reflecting the influences of decay. Foraging sites, however, are often of smaller diameter and more advanced decay than are nesting sites. Several vertebrate species appear to prefer large pieces of downed wood. Management implications are discussed.

Diameters and Heights of Cavity Sites: Their Implications to Management.Fred L. Bunnell, Elke Wind, and Mark BoylandAmong primary cavity nesters in the Pacific Northwest, the same bird species select smaller trees when nesting in hardwoods than when nesting in conifers. There is a significant tendency for

larger species to use larger nest trees (p < 0.001), but the trend is more weakly expressed in hardwoods and the slope is more shallow. Each of these findings is expected if rot governs the selection of nest trees. The same bird species select taller nest trees in more productive coastal forests than in inland forests. Actual nest heights are much shorter than nest tree heights, but species-specific averages are rarely below 5 m. The different relations for nest tree height and nest height re! fl!!!!ect dominating influences of diameter and decay. Frequency distributions of nest tree diameters reveal the importance of management targets well above the minimum diameter of nest trees. Diameters near the median diameter at which heart rot typically occurs appear appropriate, but a range of diameters is helpful in sustaining species richness. Because they do not excavate their own cavities, larger mammals require older, larger trees where rot is advanced. Bats also use larger trees, particularly when nesting in conifers. Management implications are discussed.

Dying and Dead Hardwoods and Their Implications to ManagementFred L. Bunnell, Elke Wind, and Kathy MartinAlthough they usually comprise less than 10% of forest cover in western forests, dying and dead hardwoods make important contributions to sustaining vertebrate richness. They are highly preferred as cavity sites, are important foraging sites for several bird species, and provide preferred substrate for many cryptogams and invertebrates. Two cavity-nesting species choose hardwoods for 70% or more of their nest sites, while many prefer hardwoods, even in coastal forests where hardwoods are scarce. Because hardwoods acquire heart rot earlier than do conifers, they offer cavity sites at younger ages and smaller sizes than do conifers. Strong excavators use an increasing proportion of! o!!!!lder or dead trees as their nesting is increasingly restricted to conifers (p < 0.001). Bird species richness is greatest in hardwood:conifer mixtures. The riparian avifauna shows stronger associations with hardwoods and shrubs, than with other habitat elements, nonetheless upland hardwood stands are important. Hardwoods also provide foraging opportunities at younger ages and smaller sizes than do conifers. Many cryptogams and invertebrates show marked preference for, or are limited to, hardwoods. Management implications are discussed.

Results from Sampling Dead Woody Material on the Sierra National ForestKevin CaseyIn 1997-1998, 1400 individual sample points were measured on the Sierra NF as part of the National FIA (Forest Inventory and Analysis) Grid. At each sample point, measurements were made for standing dead trees of all diameter classes, down logs (10 feet by 10 inch minimum), and small woody debris of three size classes. Dead trees and logs were categorized by decay class, providing information relevant to wildlife habitat as well as volume and weight summaries. The data set covers all vegetation types, wilderness & non-wilderness. In total, it represents the most comprehensive survey of dead wood ever collected on a Sierran forest. These data have value for analyzing forest struct! ur!!!!e, wildlife relationships and for modeling fire effects.

Functions of Coarse Woody Debris in Forest EcosystemsEfren Cazares and James M. TrappeCoarse woody debris (CWD) in various stages of decay is characteristic of the forest floor in natural forests. Contributed by breakage, windthrow, fire or standing death of large trees, it carries over into young forests as a legacy from the preceding old forests. In the Douglas-fir region, large fallen trees may persist for centuries, slowly decaying and changing in form until incorporated into the soil. CWD contributes organic matter vital to a healthy soil condition. Nitrogen and minerals accumulate in CWD over time, providing a substrate important in nutrient cycling in the forest system. It can absorb up to 6 times its own weight of water to serve as a reservoir into the sum! me!!!!r dry season, thereby during drought providing a refugium for myriad organisms and a medium for active functioning of roots, biological nitrogen fixation and nutrient cycling. It harbors specialized fungi, arthropods and amphibians, including rare and threatened species, and it provides topographic diversity to the forest floor and essential cover and habitat for small mammals. Although many functions of CWD are well documented, little specific information is available on the fungal and nutrient cycling components, especially in comparison with mineral soil.

Ecological Restoration and the Dead Wood Resource in Ponderosa Pine Forests: Effects on VertebratesCarol L. ChambersTerrestrial vertebrates are ecologically important in food webs (as foragers, predators, and as prey), and as dispersers of seed and mycorrhizal fungi. They also have significant influences on organic matter decomposition and mineral cycling. Vertebrate population densities change in response to change in vegetation structure and composition. Although some wildlife species are highly mobile and can move to more favorable habitat (birds), others (small mammals) are often restricted to relatively small home ranges (2 to 20 ha per population), so forest practices may have differential effects on vertebrate populations. Ponderosa pine forest ecosystems in the Southwest have changed o! ve!!!!r the last 130 years as a result of logging, fire suppression, and livestock grazing. Ecological restoration (thinning and prescribed burning) may help reestablish structure and function of indigenous ecosystems by recreating pre-settlement (ca. 1870) tree species composition and size class distribution and reestablishing a frequent, low intensity fire regime. However, restoration treatments may affect the availability of dead wood to wildlife (e.g., prescribed fires may incinerate snags and logs). I will discuss potential effects of restoration treatments on dead wood and impacts on habitat for terrestrial vertebrates. As an example, I will describe small mammal communities at 3 sites in northern Arizona (south rim of the Grand Canyon National Park, Camp Navajo Army Depot west of Flagstaff, and Mt. Trumbull, near the north rim of the Grand Canyon), relate abundance of small mammals to availability of standing and down wood, and describe how resto! ra!!!!tion may affect dead wood and impact small mammal communities.

CWD Attributes Before and After Harvesting in Old and Second Growth Forests: Four Case Studies in the Coastal Western Hemlock Subzone VariantGerry DavisCoarse woody debris (CWD) is an important structural component in forested ecosystems. While CWD management is required under regulation in British Columbia, Canada, we lack basic information on CWD demography and the influence of harvesting policy.

I will present pre- and post-harvest data on live standing, dead standing and downed wood in three old growth (>250 years) stands and one natural fire origin second growth (100-120 years) stand. The study area is the Coastal Western Hemlock (CWHwh1) biogeoclimatic zone of British Columbia on the Queen Charlotte Islands. At each site, permanent sample plots (1000 m2 trees; 400 m2 CWD) were established at a density of one per two hectares. The stands were harvested the following year using current practices and the permanent sample plots were reassessed. In addition sample plots (50 m2) were established in piles, roadside accumulati! on!!!!s and backspar trails. The results are discussed within the context of: stand age, future rotation ages, potential sources of CWD recruitment, and both historical and current provincial harvesting utilization standards.

Distribution Patterns of Birds Associated with Coarse Woody Debris in Natural and Managed Eastern Boreal ForestsPierre Drapeau, Antoine Nappi, Jean-Francois Giroux, Alain Leduc, and Jean-Pierre SavardIn boreal forests, several bird species use dead trees for feeding or nesting and are depending on them for their survival. Studies on wildlife use of dead trees have shown that the availability of snags is greatly influenced by the age of the forest and the type of perturbations (natural vs anthropogenic). Accordingly, cavity nesting birds seem largely affected by these changes in availability of snags. In North American boreal forests, relationships between birds and dead wood availability have predominantly been documented in western forests. The dynamics of dead wood and the distribution patterns of birds associated with this habitat feature remains largely unknown in eastern! b!!!!lack spruce forests. Distribution patterns of birds associated with dead wood were documented in the eastern black spruce forest of northwestern Quebec, Canada. Study areas were composed of four forest landscapes (50-100 km2) that were naturally disturbed by different fire events (< 2 years, 20 years, 100 years and > 200 years) and two recently logged (20 years, 80 years) landscapes. Birds were surveyed by point counts. Overall, 348 point counts were distributed over the six forest landscapes. Vegetation plots centered at each point count were used to sample standing dead trees and logs. The first objective of this study was to document and explain the distribution patterns of birds associated with dead wood in naturally disturbed forest landscapes. The second objective was to evaluate the effects of forest management on this avian guild by comparing bird patterns and dead wood availability between natural and managed forests landscape! s !!!!of equivalent ages (20 years, 80-100 years). Species richness and abundance of cavity nesting birds reached their peak in the mature forest landscape. Snag users were also abundant in recently burned forests, especially

woodpeckers. Management implications of our findings for this avian guild are discussed.

Response of Wood-boring Beetles (Coleoptera: Buprestidae, Cerambycidae) to Prescribed Burning in Southwestern OregonMaureen V. Duane, Darrell W. Ross, and Chris NiwaWood-boring beetle larvae are important to nutrient cycling, decomposition, and wildlife species as a source of food. Adult wood-boring beetles are thought to be attracted to fire and burned logs by heat and smoke. Because prescribed burning is increasingly being used as a management tool to remove downed wood and other fuels, it is important to understand how these beetles react to this alteration of habitat. We conducted two studies to monitor response of wood-boring beetles to prescribed burning. First, individual logs were placed in a prescribed burn and burned to various degrees of severity. Adult beetles flying to the logs were monitored using flight intercept traps. Bark s! am!!!!ples were taken to determine percent utilization by beetle larvae. Results were compared across severity classes of burning. Second, a retrospective study was conducted by monitoring adult wood-boring beetles in eight sites, spanning one to fifteen years since burning. Overall abundance and diversity was compared over time with adjacent check plots. The goal of these two studies is to determine the effect of prescribed burning on the habitat of wood-boring beetles, and to aid in the development of coarse woody debris and fire management programs.

Mortality as a Source of Coarse Woody Debris in Managed StandsCarl FiedlerMortality was recorded annually over a 15-year period in six stands. A thinning study (110 tpa, 220 tpa, 440 tpa, and an unthinned control) was initiated at the beginning of the period in each of four even-aged stands [one lodgepole pine (LP), one ponderosa pine (PP), and two PP/Douglas-fir (DF)/western larch (WL) stands], while an individual tree selection study (selection cutting and a control) was initiated in two uneven-aged PP stands. Several common trends emerged across treatments and species: the highest mortality in all four even-aged stands occurred in the controls, primarily due to suppression. Weather was the primary (but minor) mortality agent in the three thinned tre! at!!!!ments. Mountain pine beetle (MPB) was the primary killer of PP, and western spruce budworm of DF. In the two uneven-aged PP stands, MPB and suppression were the primary causes of death in the control, while weather was the primary agent in the selection cutting treatment. Fifteen years of mortality records from treated areas in all six stands suggest that one or more large trees per acre may need to be girdled, felled, or killed in some other way to ensure sufficient snags or coarse woody debris in managed stands.

Snag Recruitment in Subalpine Forests in the North Cascades, Washington StatePaul T. Flanagan, Penelope Morgan, and Richard L. EverettInformation on snag locations and densities is useful for managing many species of wildlife. Using a combination of belt transects, fixed plots, and aerial photographs, we recorded snag species, locations, and causal agents of tree mortality in subalpine forests in the Entiat watershed

in Washington State. The overall snag density (all standing dead trees) was 51 per hectare. Subalpine fir (Abies lasiocarpa) and lodgepole pine (Pinus contorta) were the most common species of snags. Weather-related effects created more snags than any other disturbance in the period between stand-replacing fires. The density of dominant and codominant snags did not differ! b!!!!y aspect or slope categories, but the density of intermediate and suppressed snags was highest on steep south-facing slopes. Snag densities were lowest in stand initiation and open stem exclusion structural stages.

Modeling the Delivery of Large Wood to Streams with Light Detection and Ranging Data (LIDAR)Cody FleeceLight Detection and Ranging (LIDAR) devices emit laser pulses in a scanning pattern from aircraft towards the earth. Objects on the earth reflect some pulses back to the airborne source. The differing travel distances of these pulses can be transformed into a model of the earth's surface as well as a three-dimensional characterization of forest structure. LIDAR data have been shown to accurately predict characteristics of forest structure such as canopy height, basal area, total above ground biomass, foliage biomass, and canopy reflection. Airborne LIDAR data acquisition occurred over a 14,000 acre study area at the MacDonald-Dunn Research Forest near Corvallis, Oregon.. A digita! l !!!!elevation model (DEM) with 6 meter resolution was created from the last return laser signals. This DEM was then used to generate a stream network for the forest. Canopy height was generated by subtracting the DEM surface from the elevation surface generated by the first return signals. The canopy height values generated by the resultant grid were used in a model predicting the delivery of large wood and coarse particulate organic matter to streams throughout the forest. The LIDAR data provided a means to predict wood loading to streams over a much wider area than possible in traditional studies.

Dead Wood and Decadence Characteristics of California Black Oak (Quercus kelloggii) in the Central Sierra NevadaBarrett A. Garrison, Robin L. Wachs, Terry A. Giles, and Matthew L. TriggsDead wood in live trees is often overlooked by resource managers when assessing dead wood characteristics of forest habitats. Assessments typically focus on snags and logs while dead branches in live trees are rarely quantified despite literature demonstrating their wildlife values. Dead branches are a "hidden" resource that, if ignored, may bias management efforts correcting deficits in snag abundance. In mid-elevation forests where hardwoods and conifers occur, hardwoods provide important dead wood resources, such as dead branches, sooner than conifers. As part of a wildlife community study in the central Sierra Nevada, we measured decadence components on individually tagged Ca! li!!!!fornia black oak (Quercus kelloggii) 13-152 cm dbh in four 21.1-ha stands. Acorn production, dbh, oak mistletoe (Phoradendron villosum) infestation and dead branches have been measured since 1994. Number of dead branches >= 13 cm basal diameter ranged from 0.0-16.0 per tree (mean = 1.1, SD = 1.8, n = 585). Larger diameter oaks had more dead branches (P < 0.001), while acorn production (P < 0.003) and

number of mistletoe bunches (P < 0.001) was greater in mid-sized trees. Managers should recognize that live hardwoods provide many decadence attributes that have wildlife values.

One-to-Four Year Changes in Seven Fire-Killed Conifer Species in Eastern WashingtonJames S. HadfieldA five-year study was established in three National Forests on the east slopes of the Cascade Mountains in eastern Washington to determine the causes and timing of appearance of changes in seven conifer species killed by wildfires in 1994. Tree species are Douglas-fir, grand fir, subalpine fir, western larch, lodgepole pine, ponderosa pine, and Engelmann spruce. Each of 30 plots initially contained 25 trees, all larger than 6 inches DBH. Burn severity and scorch height were determined for each tree. Five trees are being cut and sampled each year in each plot beginning one year after the trees were killed. Wood changes are being recorded at five positions on the stems from the bas! e !!!!to the 4-inch top. Tree changes being monitored include wood char, wood stain, wood decay, cracks, bark loss, sporophores, woodpecker feeding, insect infestation, branch loss, and stem breaks. Data are being complied annually to reveal the prevalence and locations on the stems of tree changes. Volumes of wood affected by stain, decay, and cracking are being calculated for all tree species. Data for seven tree species are provided that show chronologically causes of tree changes, appearance of changes, and prevalence of changes by stem position varies by tree species.

Relationship of Primary Cavity-nesting Birds to Snag Distribution and Abundance in Dry ForestsJames G. Hallett and Margaret A. O'ConnellThe relationship between the distribution, density, and condition of snags and their use by primary cavity-nesting birds is central to evaluating predictions of current models for snag retention. We are examining this relationship at the Turnbull National Wildlife Refuge, Cheney, WA. We selected 15 stands of ponderosa pine forest, which cover the range of variation in density and distribution of snags at the refuge. Transects with from 6 to 12 point-count stations were established on each stand. All snags within 100 m of each transect were mapped using GPS, tagged, and described. In 1998 and 1999, point-count surveys were conducted seven times per stand from early May until mid-J! un!!!!e. In 1998, about 1,700 of 7,000 observations were of cavity-nesting species including nuthatches (3 spp.), woodpeckers (4 spp.), and chickadees (2 spp.). The distribution and relative abundance of each cavity-nesting species is related to snag distribution and density using traditional stepwise and logistic regression models, and aalso with spatial modeling. We evaluate these patterns to those predicted by models for snag retention.

Woody Detritus Dynamics and Management: A Global PerspectiveMark E. HarmonIn the last decade woody detritus, particularly the coarse fraction, has become an important part of many scientific and management questions. Although the role of this material in providing habitat and carbon cycling is generally understood, perspectives on its role in nutrient cycling are

still evolving. Based on what is know to date, forest managers are moving away from a "blanket" removal of all the woody detritus possible to leaving and even enhancing the amounts in forests. This leaves the question of how much woody detritus is required. Initially this has been solved by the application of static standards based on a set of general objectives, but in the future a more! d!!!!ynamic and specific objective-oriented approach should be developed. The increasing number of studies on tree mortality and decomposition give a global view of how these processes vary with forest type and climate. These data also provide the basis for the dynamic rather than static management of woody detritus. However, to be successful this must be coupled with a detailed understanding of how certain species and ecosystem processes vary with the amount of woody detritus.

Effect of forest age, structural elements and prey density on the relative abundance of pileated woodpecker (Dryocopus pileatus) on south-eastern Vancouver IslandCarol L. Hartwig, Donald S. Eastman, and Alton S. HarestadThe relative abundance of pileated woodpecker (Dryocopus pileatus abieticola) in four 1450-ha landscapes of differing forest ages was correlated with ecological variables using call-playback surveys, habitat surveys, and GIS in dry Douglas-fir (Pseudotsuga menziesii) forests on south-eastern Vancouver Island in 1996-1997. Pileated woodpeckers did not utilise intensively managed forests <80 years old due to structures of young forests. The relative abundance of pileated woodpecker was greatest in a landscape with 51% of >140-year-old stands, although the relative abundance was not significantly different from an immature forest landscape with 49% of >80 ! ye!!!!ar-old forests, or a forest landscape with 70% of >140-year-old stands.

Dead Wood and Insects: Function of Insects in the Death Cycle of ConifersMichael I. Haverty and Patrick J. SheaInvertebrates play a variety of roles in forest ecosystems, but one of the least understood is their role in the death cycle of conifers. The death cycle is the process whereby living trees succumb to the activities of invertebrates, primarily insects, die, then begin to degrade until part or all of the tree is reintroduced into the same forest ecosystem. Insects play many roles as primary mortality agents of conifers. Defoliators have ravaged the fir forests of western North America causing extensive reduction of growth and often mortality. Bark beetles, primarily the genus Dendroctonus, are among the most dramatic agents of tree mortality. Bark beetles mass att! ac!!!!k individual trees and colonize the phloem and inner bark then launch the decomposition process by breaching the protective bark and providing prey base and substrate for the various trophic levels that follow. Bark beetle activity under the bark of trees results in an increase in activity of insect predators and secondary wood-degrading insects on the outer surface of the bark. This sudden increase in activity by bark beetles and associates is mediated by the semiochemicals released by the bark beetles and the host. During this colonization process numerous species of Picidae (woodpeckers) and Sittidae (nuthatches) and other bark-gleaning species visually respond to this activity, feed on external insects, and perforate the bark in search of insect prey. The wood-degrading insects that follow bark beetles continue the process by

penetrating the sapwood and heartwood and begin to reduce the solid wood to a substrate pervaded by gallery systems and! f!!!!rass of wood-boring beetles and termites. Most of the activity by insects is in the early stages of the degradation process, that is, before the bark sloughs and the larger branches begin to break away from the bole. Insects utilize the degrading conifers as food and shelter. The tunneling activity of wood-boring beetles are the primary cause of perforations in wood and provide an avenue of ingress for fungi and other micro-organisms that accelerate the degradation or death cycle of trees. Carpenter ants utilize wood for harborage, creating large, open galleries and chambers, but they do not utilize the wood for food. Relatively few terrestrial animals have the ability to catabolize cellulose, the primary component of wood. Termites harbor micro-organisms that facilitate the digestion and utilization of cellulose; these micro-organisms help termites reduce the volume of wood, convert cellulose to primary biochemical building blocks that termites use for ! nu!!!!trition. Further, the ability of the termite gut microbiota to fix nitrogen, eventually results in secondary productivity of a high protein/high fat prey.

Comparing Deterioration and Ecosystem Value of Decay-resistant and Decay-susceptible Species of Dead TreesPaul E. HennonThe pattern and rate of deterioration of dead trees vary by species. The heartwood of some trees, such as yellow-cedar (Chamaecyparis nootkatensis), contains compounds that inhibit decay. Bark sloughing and decay of sapwood in early snag classes proceed in a manner similar to other tree species. Limited heartwood decay occurs above ground, however, resulting in persistent snags with intact tops. Soil-borne fungi cause decay at the ground line where cedar snags finally break and fall to the ground, often in a single piece of nearly inert heartwood, some 80 to 100 years after death. Yellow-cedar snags generally do not yield habitat for cavity nesting animals unless! t!!!!hey contained heart rot before death. Dead trees of species lacking specialized heartwood compounds, such as western hemlock (Tsuga heterophylla), exhibit less distinction between sapwood and heartwood decay. Also, modes of tree death, heart rot levels, and type of saprophytic decay may differ considerably by stand age. Hemlocks that die standing deteriorate more rapidly than cedars and gradually are reduced in height as they break into pieces. Woody debris produced by dead standing hemlocks becomes partially decomposed before reaching the forest floor or streams and thus is less persistent and, in streams, more mobile.

Characteristics of Ponderosa Pine Snags Selected as Roosts by Long-Legged Myotis Bats, Myotis volansMichael Herder and Jennifer G. JacksonThis study is the result of a three year radio telemetry project to locate and characterize ponderosa pine snags used as bat roosts at Mount Trumbull, in northwestern Arizona. Study objectives included characterizing bat roost site preferences in pine and oak snags to document the effects of ecosystem restoration work to restore the forest to pre-settlement conditions. Over 900 bats of 13 different species were captured over open. Myotis volans, M. thysanodes, and

Eptesicus fuscus were the most abundant species, comprising 40.1%, 22.8%, and 12.5% respectively of the total. More than 60 net nights of capture effort was conducted in ponde! ro!!!!sa pine habitat (66.7%), pinyon-juniper habitat (30.3%), and sagebrush-grasslands habitat (3%). A total of 45 roost sites were located using radio telemetry tracking methods. Of these, 40 day roosts of Myotis volans were located in ponderosa pine snags. Roost snags used by Myotis volans shared several characteristics including a d.b.h. measurement 28 inches, height 70 feet, elevation 6,480 feet above mean sea level, presence of large fissures and/or exfoliating bark, and distance to water < 1.5 miles. Other roost tree characteristics, such as percent slope, position on slope, aspect, canopy and basal area, and distance to grazing, treatment areas, or foraging sites showed a high degree of variability.

Conservation Plans and the Protection of Snags and Coarse Woody Debris on Industrial ForestlandsLorin L. Hicks and Henning C. StabinsForest practices on private industrial timberlands have steadily progressed to incorporate many of the issues and approaches currently in use on public forests. One of the most significant advances for protecting wildlife habitat on private lands has been the development of Habitat Conservation Plans. Originally developed as a planning process for landowners to mitigate for the incidental "take" of threatened and endangered species, HCPs have expanded become long-term, multi-species landscape plans designed to address many aspects of wildlife and fish management, including the conservation of standing dead trees and downed woody debris. Plum Creek completed a 50-year HCP fo! r !!!!169,000 acres of company land in Washington’s Central Cascades in June 1996. The HCP addresses the biological needs of 285 vertebrate species, including anadromous fish and cavity-dependent wildlife. This paper describes how the HCP was designed and presents examples and information on implementation of practices to address key riparian and upland management strategies for retention of dead and dying trees.

The Dynamic Relationship of Woodpecker Habitat Selection and the Successional Decay of Ponderosa PineKerry L. Hughes, E.O. Garton, Steve Zack, and Patrick J. SheaWoodpecker foraging and nesting activity is closely tied to tree decomposition stages, yet the distinct relationships between woodpecker use and the various phases of snag decay are poorly documented. It is clear, however that particular species are associated with recently killed trees, whereas others primarily use snags in more advanced stages of decay. In addition, snag decomposition may be accelerated by foraging woodpeckers which puncture the bark, creating vectors for the colonization of sapwood decay fungi. In 1998 we initiated a study to examine the relationships between woodpecker foraging and snag decay processes in an Eastside pine forest community located in northeast! er!!!!n California. Our objectives were to (1) quantify relative foraging intensity across a chronosequence of known snag ages, and (2) determine the relationship between foraging intensity and sapwood decay. To address the first objective, a foraging intensity index based on the cumulative number of foraging excavations per square

meter was used to compare foraging activity across known snag ages. Preliminary results revealed no significant difference in cumulative foraging abundance between snags of different ages, suggesting that woodpecker foraging activity occurred primarily in the younger age classes. To address the second objective, we are currently recording relative measures of sapwood decay and woodpecker foraging intensity in ponderosa pine snags experimentally baited with bark beetle pheromone in 1996 and 1998. These results will be synthesized into a conceptual model incorporating species-specific woodpecker foraging data, entomological dat! a,!!!! and snag structural changes. This model is intended to provide further insight into the relationship between woodpecker habitat selection and successional patterns of decay organisms in ponderosa pine.

The Efficacy of Inoculating Fungi into Conifer Trees to Promote Cavity Excavation by Woodpeckers in Managed Forests in Western WashingtonMartin J. Huss and James C. BednarzCavity-using birds and mammals provide ecological services that benefit and enhance the forest ecosystem, such as the control of forest-damaging insects and the dispersal of seeds. To develop management prescriptions to promote the development of a cavity-using community in forests, we have investigated the relationship that wood-decay fungi play in facilitating nest excavation by primary-cavity nesting species (i.e., woodpeckers). We located 90 woodpecker nests in western Washington and examined trees for the presence of basidiocarps or conks (fungal reproductive structures) and collected wood samples to identify the decay organisms associated with these trees. The f! !!!!21 ungus most often found to produce basidiocarps on woodpecker trees was Fomitopsis pinicola (the red-belted conk). Basidiocarps of this fungal species were found on 41.1% of all woodpecker nest cavities (i.e., 61 hairy woodpecker [Picoides villosus], 26 northern flicker [Colaptes auratus], and 3 red-breasted sapsuckers [Sphyrapicus ruber] nest trees). This fungus was also recovered from wood samples of 11.5% of all woodpecker nest trees that did not possess conks. We also identified 32 other fungal types, representing 20 genera from woodpecker trees, but many probably do not represent primary causes of wood decay but consist of secondary invaders or colonizers. On the basis of (1) the ability to grow the fungus in culture, (2) the type of decay caused, and (3) the degree of pathogenicity, we selected F. pinicola as the organism to inoculate into trees experimentally to promote colon! iz!!!!ation by woodpeckers. During the summers of 1997 and 1998, we inoculated 65 clusters of 10 trees (total = 650 trees) in 34 separate managed forest stands in western Washington. The experimental design of inoculations that we followed will allow us to examine the effect of blank (control) vs. fungal inoculations, the influence of tree species (i.e., Tsuga heterophylla, Pseudotsuga menziesii), the effect of available snags, and the effect of size/age classes on trees on future colonization by woodpeckers. Of 20 randomly selected trees that we inoculated in 1997, we retrieved F. pinicola from 14 (70%) of these in 1998. As we collected only one wood sample from each tree, it is likely that the fungus was present and not detected from a least some of six trees that yielded negative results. Vegetative compatibility analyses confirmed that, in every case, the F. pinicola isolated was the same genetic type inoculated ! th!!!!e year previously. Thus, our results indicated that at least 70% of the inoculations performed in 1997 were successful in establishing

the fungus in the intended tree. Follow-up monitoring of experimental inoculation sites will continue at 3-yr intervals to determine if woodpeckers begin using these trees.

Macroarthropod Fauna Utilizing Coarse Woody Debris of Pinus ponderosa and Abies concolor in Northeastern California.Ethan Koenigs, Patrick J. Shea, Michael I. Haverty, and Robert BorysThis study was initiated to determine the diversity and community structure of macroarthropods found in the coarse woody debris (CWD) of a pine/fir dominated forest. With the changing goals of the Forest Service, emphasis has been placed upon understanding ecosystem processes that support long-term sustainability and biological diversity. Most studies on CWD to date have been associated with mesic Douglas-fir forest of the Pacific Northwest. There is a paucity of information on the role of CWD in other forest types. Our study site is located northeast of Mt. Shasta in the Klamath National Forest, Goosenest Adaptive Management Area. For the purposes of this study CWD consists of f! al!!!!len logs and stumps >6" in diameter. Three types of traps were constructed to survey the arthropod fauna entering and emerging from CWD: An acrylic pane trap with a water filled collection basin, a collar trap that wraps around fallen logs, and a solid plastic cylinder placed over stumps. Equal numbers of traps were placed on Abies concolor and Pinus ponderosa. Three different stages of decomposition, from recently dead (Class I), to soft but structurally intact (Class III), were sampled from each tree species. Collections were made on a biweekly basis in 1997, 1998, and 1999. The insects from all traps were sorted by order, except for the Coleoptera, which were sorted to the family or species level. The most numerous groups in almost all cases were the Diptera, followed by the Coleoptera and Hymenoptera. Of the Coleoptera, the Staphylinidae were the most numerous in all cases excluding the earliest decay stage in bot! h !!!!species, where Scolytidae are dominant. Adult termites have been mysteriously lacking from all traps in all years.

Snags and Down Wood in the Interior Columbia Basin ProjectJerome J. Korol, Miles A. Hemstrom, Wendel J. Hann, and Rebecca A. GravenmierSnag and Down Wood density tables were developed at a one square kilometer resolution for the Interior Columbia Basin Ecosystem Management Project (ICBEMP) analysis area. The table attributes include three broad Potential Vegetation Groups (Dry Forest, Moist Forest, and Cold Forest), two time periods (Historic, approximately 1850; and Current, approximately 1995), two fire regimes (frequent, low intensity stand-maintaining fires and infrequent, high intensity stand-replacing fires), and two management characterization classes (Managed or Unmanaged). Snags and Down Wood are divided into two size classes (Large and Small, with a 21 inch break-point) and five density classes (None, ! Ra!!!!re, Uncommon, Common, and Abundant). Each category, such as Large Snag, Uncommon; Small Down Wood, Rare; etc. has its own set of density category definitions (e.g the Large Snag, Uncommon density is 1.0-2.5 per acre). Computer simulations with the Columbia River Basin Succession Model (CRBSUM) will be used to estimate future Snag and Down Wood densities for the periods 10 years and 100 years from Current for the three proposed Alternatives in the ICBEMP Supplemental Draft Environmental

Impact Statement.

The Distribution and Role of Large Woody Debris in Upper Prairie Creek, a Pristine Northern California Redwood WatershedSharon Kramer and Randy KleinWe inventoried large woody debris (LWD) in approximately 7 km of the mainstem of upper Prairie Creek (drainage area = 8 to 21 km2), an important coho salmon stream in northern California, in 1997 and again in 1999. Information on LWD characteristics and loading in small coastal streams in redwood ecosystems is scant, especially in pristine redwood watersheds. We inventoried all pieces of LWD exceeding 10 cm in diameter and 2 m in length, consistent with most other inventories from the Pacific Northwest.Initial surveys were conducted in 1997, and repeated in 1999 following a 10-year flow event in Nov. 1998, thus providing information on the LWD redistribution from a moderate flood. Results from the 1997 survey indicated that redwood comprised over 60% of the total number of pieces of LWD in Prairie Creek, and over 85% of the total volume, with maple, alder, other conifers contributing the remainder. Nearly 25% of the total volume of redwood LWD was contributed by a small fraction of the extremely large pieces (in excess of 2 m in diameter). The importance of these large, stable pieces to the geomorphology and ecology of Prairie Creek, as well as flood-induced redistribution, will be ! di!!!!scussed.

Demography of Snags in Eastside Pine Forests of CaliforniaMichael Landram, William F. Laudenslayer, Jr., Thomas Atzet, and Diane WhiteThe importance of standing dead trees has been well documented, however, relatively little is known about the spatial and temporal patterns of snag creation, decay, and loss. In 1988, we initiated a study on snag abundance and bird use in eastside pine forests of northeastern California dominated by yellow pines (Pinus ponderosa or P. jeffreyi). Dramatic changes in snag abundance that 3 year study argued for further inquiry into snag demographics and we continued the study through 1999. We randomly selected 24-5 ha plots in six snag density classes ranging from 0 snags to approximately 10 per hectare; classes were based on counts of all snags > 3! 8 !!!!cm in diameter and >6.5 m tall within each plot. Plots chosen had not experienced recent fire but fire affected two plots in the course of the study. We have detected approximately 3,000 snags to date and have collected data on characteristics of each snag yearly. Most snags were created in pulses associated with droughts and bark beetle activity; pulses are separated by years of low snag creation. Losses of snags was more gradual. The spatial and temporal pattern of change in snag numbers was not consistent across plots. Diameter greatly influenced longevity of snags, but other site specific factors also influenced life spans of individual snags.

Cavity Nesting Bird Use of Snags in Eastside Pine Forests of Northeastern CaliforniaWilliam F. Laudenslayer, Jr.Relations between snags (standing dead trees) and cavity nesting birds in eastside pine habitats were examined in the breeding seasons of 1989, 1990, 1991 on 24-5 ha strip transects on the

Modoc and Lassen National Forests and Lassen Volcanic National Park, California. Transects were placed in randomly selected plots of eastside pine habitat dominated by either ponderosa (Pinus ponderosa) or Jeffrey pine (P. jeffreyi) with varying densities of existing snags. Cavity nesting birds generally placed their nests in snags larger in diameter and taller than randomly selected alternative snags found in the forests studied. Snags with active nests had greater d! ia!!!!meters and heights than alternative snags; both differences were significant. Snags with historical nest cavities generally were of larger diameter than snags without historical nest cavities. Despite the heavier nesting use of larger snags, many large snags, with similar visual deterioration characteristics, showed no indication of historical nesting use.

Assemblage of Vascular Plants on Deadwood Microsites within Aspen-Dominated Boreal ForestsPhilip Lee and Kelly SturgessThis study examines the role of downed logs, stumps, and root throws on understory composition of aspen-dominated boreal forests. Measures of microsite availability, and vascular plant composition and abundance were taken from a matched pair of 28 year old wildfire and harvest stands. Only large (>20 cm dia.) downed logs in advanced decay stages were suitable for colonization by vascular plants. Suitable downed logs covered more than 5x the area of stumps and root throws combined. In comparison to the forest floor of either the wildfire or harvest stands, detrended correspondence analysis revealed that different assemblages of vascular plants colonize logs and stumps. Surprisingl! y,!!!! assemblages on root throw pits and mounds were not different from the assemblages on the forest floor of either wildfires or harvest stands. Logs and stumps in wildfire and harvest stands start with similar assemblages of plants. However, as both logs and stumps decayed, the assemblages they supported became more similar to their respective wildfire or harvest forest floor. Ordination plots of species scores suggest that regenerating trees and low herbs disproportionally colonized logs and stumps over the forest floor. Examination of seed and veg banks from deadwood suggests that dispersal limitations may plays a major role in the types of communities developing on deadwood. Ordination with assemblages from old seral stages strongly suggests that colonization of deadwood in younger stands facilitates understory changes to forest floor communities found at older seral stages.

Ant Abundance, Diversity, and Dependence on Dead Wood in Central Interior British ColumbiaB. Staffan Lindgren and A.M. MacIsaacAnts are known to be ecologically significant invertebrates in many types of ecosystems. They affect physical and chemical soil properties, plant distribution, and forest health etc. Ants utilize dead wood substrates for nesting, starting with species that nest sub-cortically in stumps within a year of tree death, to species that live in stumps and other wood with advanced decay. Virtually no information is available on the species diversity and abundance of the ant fauna of the north-central interior of British Columbia. We surveyed the area around Prince George, B.C., to determine the species diversity, abundance and dependence on dead wood as nest sites. This

paper gives an ov! er!!!!view of how ant diversity and abundance relates to dead wood availability and condition in forests of different age and species composition. Nest habitat-specificity with reference to dead wood utilization is discussed.

The Use of Residual Downed Wood by Female Short-tailed WeaselsKimberly A. Lisgo and Fred L. BunnellDowned wood provides structure which is important in the habitats of mustelids and small mammals. It can provide secure travel corridors and in winter, subnivian access and thermal cover. Timber harvesting practices provide residual downed wood, which could be managed to suit the needs of wildlife and mitigate some of the effects of timber harvesting on habitat for wildlife. In the mixedwood boreal forest of Alberta, we examined the use of residual downed wood by female short-tailed weasels (Mustela erminea) (n=4) (August to December 1996). We found that female weasels preferentially used areas of residual downed wood in 3-year-old regenerating aspen (Populus ! tr!!!!emuloides) cutblocks during summer and fall (P < 0.001). There are 3 reasons why female weasels may selectively use these areas: 1) the amount of food; 2) the number of rest sites; and 3) availability of travel corridors. Our data suggest that food abundance, specifically the southern red-backed vole (Clethrionomys gapperi), and number and type of rest sites are 2 factors that may explain the preferential use of residual downed wood by female weasels. Recommendations for the management of residual downed wood are discussed.

How Much Dead Wood in Stream Channels is Enough?Thomas E. LislePrivate forest managers often ask for guidelines on how much dead wood should be retained in streams to fulfill ecosystem functions of key organisms while allowing timberland owners to profit from riparian stands. There are three approaches to answering this question for a particular reach of channel: (1) Ecological: Determine a relation between dead wood volumes in channels and habitat condition or populations. This approach fails because of the complexities of sizes, shapes, and arrangements of dead wood in a variety of stream channels harboring a variety of communities. (2) Reference: Compare dead wood volumes with those in pristine channels. A precise threshold cannot be defi! ne!!!!d using this approach because frequency distributions of dead wood volumes are wide, even within pristine channels in similar settings, and distributions for managed and pristine channels invariably overlap. (3) Budgetary: Put existing wood volumes in context with past and present rates of input of wood and the condition of ripariian stands, using regional frequency distributions as a starting point to suggest significant departures from average conditions. This approach can be successful but demands more analysis and the solution takes the form of defining departures from desirable trends in wood volumes, no precise threshold values.

Methods for the Quantification of Coarse Woody Debris and the Identification of its Spatial PatterningDuncan C. Lutes

Methods for the quantification of coarse woody debris volume and the description of spatial patterning were studied at the Tenderfoot Creek Experimental Forest in central Montana. The line transect method was found to be an accurate, unbiased estimator of large down debris (>10 cm diameter) volume on 1/4 hectare fixed-area plots, when perpendicular lines were used. The Fischer Photo Guide for woody fuels did not quantify large down debris as precisely as the line transect method on 1/4 hectare plots. Chi-square tests found that down debris had a clumped spatial distribution in 23% of the tests. Pieces were found randomly distributed in the remaining 77% of the tests. Tests of log! o!!!!rientation showed evidence of directional clumping in 8 of 13 tests. The Paired Quadrat Variance method found no consistent scale of clumping in 24 plots of variance. Snags were found to have a clumped distribution in 37% of the tests, a random distribution in 62% of the tests and one test indicated a uniform distribution. The Variable Area Transect method accurately described snag density, however it was not successful at determining the density of down pieces.

Effects of Ecosystem Restoration Treatments on Cavity-nesting Birds, Their Habitat and Their Insectivorous Prey in Dry Fire-maintained Forests of Southeastern British ColumbiaMarlene M. MachmerLong term fire suppression has resulted in excessive forest ingrowth and associated structural and functional changes to dry fire-maintained forests. Ecosystem restoration treatments are currently being tested for future broad scale operational use in the Nelson Forest Region of southeastern British Columbia. This experimental study investigates effects of restoration treatments on cavity-nesting birds, their habitat and their insectivorous prey under several treatment combinations: (1) prescribed fire, (2) partial harvesting, (3) a combination of fire and harvesting and (4) untreated controls. Cavity-nesting birds require standing dead or diseased trees (wildlife trees) for nest! in!!!!g and roosting that are often targeted for removal during harvesting. Many of these species and woodpeckers in particular forage on trees scorched and killed by fire which attract bark and wood boring insects. Proposed treatments are therefore expected to have measurable effects on this avian guild. Pre- and post-treatment data on (i) the nesting density and relative abundance of cavity-nesters, (ii) the availability of wildlife trees, (iii) the foraging intensity of woodpeckers and (iii) the species diversity and relative abundance of insects have been gathered each breeding season, beginning in 1996. A final year of post-treatment monitoring will be completed in 1999 and comprehensive management recommendations will be put forward based on comparison of the quantitative effects of various restoration treatment options on cavity nesting birds.

Deadwood Management in British ColumbiaAndy MacKinnon, John Parminter, Jeff Stone, Nancy Densmore, Gerry Davis, AndrŽ Arsenault, and Todd ManningIn this paper, we highlight the historical development, current legislative, and operational practices of dead tree management and research in British Columbia. The BC Ministry of Forests is responsible for managing more than 82 million hectares of Crown (public) land in British

Columbia. BC's Forest Practices Code, which regulates forest management, guides forest managers to leave wildlife tree patches in harvested areas as required in higher level plans, and to include general objectives respecting target levels of retention for CWD in forest development plans. At the same time, occupational health and safety regulations govern the retention of potentially dangerous trees in fo! re!!!!stry operations and current policy with respect to CWD only enables the management of the CWD resource within current harvesting utilization standards. Managing for coarse woody debris and snags without impacting timber supply is the challenge facing BC's forest managers today.

The Relationship of Dead Wood and Terrestrial Vertebrate Diversity in Southwestern OregonChris C. MaguireTwenty-four mature forest stands in southwestern Oregon were used to test the hypothesis that the diversity (= richness) of small terrestrial vertebrates is positively correlated with dead wood volume. Dead wood was separated into two components: (1) coarse woody debris (CWD), defined as down wood at least 10 cm in diameter, and (2) snags, defined as standing dead wood at least 20 cm in diameter and 0.5 m in height. The volume of CWD in stands ranged from 50 to 860 m3/ha and snag volumes ranged from 10 to 240 m3/ha. Small mammals and amphibians both numbered between 1 and 7 species per stand based on a pitfall sampling effort of approximately 3,600 ! tr!!!!ap nights per stand over two years. Regression analysis revealed that the diversity of all terrestrial vertebrates taken as a single group increased with increasing volumes of CWD. Viewed as separate taxonomic groups, the richness of small mammals, insectivores, and amphibians all correlated positively with CWD volume; rodent richness showed no significant relationship with CWD. None of the vertebrate groups disclosed significant correlations between species richness and snag volume. The results of this study suggest that CWD volume is a better predictor of stand level diversity of small terrestrial vertebrates than either snag or total dead wood volumes in southwestern Oregon.

British Columbia's Dangerous Tree Assessment Process - A SummaryTodd ManningNew dangerous tree assessment guidelines have recently been developed by the Wildlife Tree Committee of British Columbia. They are intended to provide information and technical procedures for assessing and safely retaining trees in various work situations. These range from park operations, to tree planting and stand tending, to various forest harvesting methods and silvicultural systems. The decision to select any given tree as a leave tree must consider worker safety, the type of work activity and level of ground/tree disturbance, tree condition and site factors, and wildlife habitat value. These guidelines will be of interest to persons involved in any activity where the manage! me!!!!nt of trees for wildlife habitat or other purposes is desirable.

An Ecological Functional Basis for Managing Decaying Wood for WildlifeBruce G. Marcot

The traditional approach to managing decaying wood for wildlife has been to list species associated with snags, down wood, and other wood decay elements, and providing sizes and amounts to meet their assumed needs. An expanded approach more consistent with the spirit of ecosystem management would also describe how the ecological roles of species in turn affect wood decay elements, and how providing for wood decay elements provides for a suite of species whose key ecological functions (KEFs) influence the ecosystem far beyond those decay elements per se. This expanded approach describes the "functional web" of how managing for wood decay elements helps support a surprisingly wide ! ar!!!!ray of ecological functions within ecosystems, such as many trophic relations, species interactions, soil aeration, primary cavity and burrow excavation, and dispersal of fungi, lichens, seeds, fruits, plants, and invertebrates. An example displays this functional web for wildlife species associated with various wood decay elements (snags, down wood, litter, duff, mistletoe brooms, dead parts of live trees, hollow living trees, natural tree cavities, bark crevices, and live remnant or legacy trees) in Washington and Oregon. The challenge is posed for management is to think functionally and beyond simple species-habitat relations.

DecAID: Wildlife ComponentBruce G. Marcot, Kim Mellen, Susan A. Livingston, and Cay OgdenThe wildlife component of DecAID is based on a thorough review, analysis, and synthesis of the empirical literature on wildlife-dead wood relations. The wildlife analysis consists of compiling data on snag and log size, snag density, and amounts of down wood related to individual species or groups of wildlife species as presented in the literature, for various habitats and types of wildlife use (breeding, feeding, roosting). The wildlife use data are arranged in 3 cumulative species richness curves representing means and plus or minus one standard error (or equivalent variant), which are interpreted as levels of confidence of providing for individual species or groups along the c! ur!!!!ves. One can consult the curves to determine (1) for a given snag or down wood level which species or groups are provided, at three levels of confidence, and (2) for a target objective of providing for a specified percent of species, the levels of snags or down wood needed to achieve that objective. Given the outcome, the other components of the DecAID model can then be consulted to determine hazards or mitigation for risks of fire and insects and disease, and to provide for fungi and non-pest invertebrates.

Large Wood Redistribution and Recruitment from Debris Flows in the Central Oregon Coast RangeChristine L. MayDebris flows in the Pacific Northwest play a major role in routing wood and sediment stored on hillslopes and in first- through third-order channels to higher order channels and valley floors. Field surveys conducted in 1997 documented characteristics of the initiation site, runout zone, and deposit of 53 debris flows triggered during the winter of 1996 in the Siuslaw Basin of the central Oregon Coast Range. The volume of large wood in debris flows that traveled through forested channels was only slightly greater than the volume delivered by debris flows through

clearcuts. This finding reflects the legacy of large wood stored in low-order channels. Furthermore, size-class distrib! ut!!!!ions of wood in deposits and trees on the hillslope were not well correlated. The average diameter of wood in the deposit was greater than the diameter of trees present on the surrounding hillslopes, and the diameter distributions in deposits were similar among forest age classes. Valley width was the primary control on the type of deposit that formed, and the degree of interaction with the receiving channel. By increasing the frequency, magnitude, and altering the composition of debris flows, forest practices can affect the long-term potential for developing complex channel morphology and aquatic habitat.

Factors Influencing the Dynamics of Large Wood in Headwater Streams of the Oregon Coast RangeChristine L. MayFactors influencing large wood recruitment and redistribution were investigated in an unharvested watershed in the southern Oregon Coast Range. Stream size and topographic setting strongly influenced processes by which wood entered the stream and was subsequently redistributed. In low-order streams draining steep hillslopes, processes associated with slope instability dominated large wood recruitment. In the low gradient valley floors of mid-order streams, windthrow was the dominant recruitment process, but debris flows played an important, yet temporally variable role in wood redistribution from the headwaters. Headwater streams received wood from further upslope than mid-order ! st!!!!reams, suggesting that the adjacent forest exerts a wider zone of influence on the channel in headwater basins. Input and redistribution processes also influenced piece location relative to the direction of flow and thus, affected the functional role of wood. Wood recruited directly from local hillslopes and riparian areas was more abundant throughout the network, was typically positioned along side or spanned the full width of the channel, and provided a unique framework that functioned to trap sediment and wood in transport. In contrast, fluvially redistributed wood had a patchy spatial distribution, pieces were commonly located in mid-channel positions, and were closely associated with scouring of the stream bed and banks rather than deposition.

Use of a Large-scale Experiment to Examine the Role of Coarse Woody Debris in Southeastern Pine ForestsTimothy S. McCay, James L. Hanula, Susan C. Loeb, Steven M. Lohr, James W. McMinn, and Brett D. Wright-MileyA long-term experiment into the ecological consequences of woody-debris manipulation was initiated at the Savannah River National Environmental Research Park in the Upper Coastal Plain of South Carolina. Each of four 9-ha plots in each of four blocks was subject to one of the following treatments: removal of all snags and fallen logs, removal of fallen logs only, a catastrophic pulse of debris (to be implemented in year 2000), and control. The monitoring of invertebrate, herptile, avian, and mammalian assemblages -- along with abundance, quality, and dynamics of woody debris -- began with the onset of woody debris removal in 1997 and continues. Treatments have resulted in a 5- to! 1!!!!0-fold difference in woody-debris abundance.

To date, significant differences among treatments have only been detected for bird assemblages. Preliminary results underscore the benefits and limitations of using large-scale experiments. This experimental design allows tests of hypotheses regarding the effect of volume, type, and decay condition of woody debris on faunal assemblages. Coupled with mechanistic studies of habitat use and trophic interactions, this experimental approach can result in stronger inferences regarding the function of woody debris than those obtained through natural history observation or correlative studies.

Simulation of Large Wood Dynamics in Small Streams of the Pacific NorthwestMark A. Meleason and Stanley V. GregoryLarge wood in streams can profoundly influence stream morphology, hydrology, habitat complexity, and nutrient dynamics. Intra-basin variation in the distribution and abundance of stream wood is due in part to the natural history and management of the basin as well as local influences within the basin. We developed a computer simulation model to investigate causes of variation of wood in streams. The model, STREAMWOOD, is an individual based stochastic model that operates on an annual time step at the reach scale. Stream systems that can be simulated range from a single reach to a small basin. Stream wood dynamics considered are tree entry, breakage, movement, and decomposition. R! ip!!!!arian forest input are from either a simplified forest gap model built within STREAMWOOD or from a user specified input file. The model is run under a Monte Carlo procedure and the results are reported as average conditions per reach. The current version of STREAMWOOD was developed for coniferous-forested streams of the Pacific Northwest fifth-order and smaller. Species considered include Douglas fir, western hemlock, western red cedar, and red alder. Verification, validation, and sensitivity analyses are in progress. Test runs simulating wood dynamics in small streams have been conducted.

Coarse Wood Dynamics ModelKim Mellen and Alan AgerThe Coarse Wood Dynamics Model (CWDM) analyzes the dynamics of snags and down logs in forested ecosystems. CWDM models Douglas-fir and western hemlock snags and logs. The model assesses snag fall, height loss and decay, and log decay at 5 year intervals for a period of 300 years. Snags either fall whole or partially break off, creating down logs. Snags also decay from hard to soft. Logs decay more slowly than snags, from sound to decayed conditions, and eventually disappear into the forest floor as duff. Snag fall and height loss rates are from Forest Inventory and Assessment (FIA) remeasurement data on snags on private lands in western Washington. Decay rates are from Graham (19! 82!!!!) based on work at the H.J. Andrews Experiment Station, Willamette National Forest. The model can track remnant snags and logs (those existing on site at the beginning of the assessment time) and new snags and logs created from green trees. The Forest Vegetation Simulator (FVS) can be used to determine number and sizes of snags created by suppression mortality.

DecAID: A Decaying Wood Advisory Model for Oregon and WashingtonKim Mellen and Bruce G. MarcotDecAID is a knowledge-base advisory model provides guidance to managers in determining the size, amount, and distribution of dead and decaying wood (dead and partially dead trees and down wood) necessary to maintain wildlife habitat and ecosystem functions. The model is intended to update and replace the snag-wildlife models presented in Thomas (1979) and Brown (1985). While the relationship of dead and decaying wood to wildlife habitat is a major component of the model, DecAID also provides expert advice on prescribing and interpreting existing conditions of dead wood with regards to: wild and prescribed fire and fuels, pest and non-pest insects, disease, fungi, and use of inven! to!!!!ry data to describe current managed and natural levels of dead wood. The model also produces a list of ecosystem functions performed by those species associated with dead and decaying wood. Advice on the distribution of dead wood at the stand and landscape scale is provided in the model.

A Method for Monitoring the Effects of Prescribed Fire on Key Wildlife Habitat ComponentsRichard Miller and Tammy Randall-ParkerA method is needed to gather information and to monitor the effects of prescribed fire on key wildlife habitat components. After trying and failing with a point based monitoring method, a grid based mapping method of monitoring evolved. This method is effectual for monitoring live ponderosa pine, snags, down logs, and live oaks. This project has an eventual target of ten plots monitored. The presentation will explain the method and something about the lessons learned along the way.

Stag Retention and Use by Arboreal Mammals in Eucalypt Forests in Queensland, AustraliaDamien Moloney and Stephen DeStefanoDead standing trees ("stags" in Australia) are critical for wildlife in forests throughout the world. Dynamics of stag retention are an interaction of tree mortality, size, and deterioration rate. In North America, cavity formation often involves primary excavators, notably woodpeckers. In Australia, however, primary excavators do not exist, and hollow development in Eucalypt forests is an interaction of mechanical and biological processes (e.g., limb breakage, fire scars, fungal infections, and termite activity). Historically, timber management involved removal of stags, and stags are lost to fire and wind at a greater rate. The current status of stags in south-east Queensland w! as!!!! determined from permanent inventory plots. Hollow development was correlated with rates of wood decay and time since death. Density of hollow bearing trees was 10.2 ± 8.9 stems ha-1; ~50% of the hollow trees were stags. Relative importance of live versus dead trees to two species of arboreal mammal was investigated. Proportion of standing dead trees versus live was approximately 1:30; and dead trees used by arboreals versus live varied between species and sites. Longevity of all stags was estimated to be 50 years. As the retention of stags may provide a significant hollow resource in managed stands, adequate densities and distribution

is extremely important to forest wildlife communities on both continents.

Water, Wildlife, Recreation, Timber ...........Coarse Woody Debris?John MountCoarse woody debris (cwd) is a valuable natural resource that plays a key role in the ecosystem. As a private land owner, Edison Forestry has been tracking cwd for six years. The tracking that's done in conjunction with the timber inventory has been given equal time and importance. The inventory process tracks the age and size class of downed wood, the species and size of snags and the species and size of green snags. All data is then input into a Geographic Information System, where maps showing densities of cwd of any form can be displayed. The combining of these two data collection processes has provided a cost effective way to acquire some knowledge about the amounts and d! is!!!!tributions of cwd on Edison lands. Through this process we hope to gain base line data and valuable clues about the role that cwd plays in various habitat types that will assist in management decisions concerning fire, nutrient recycling, fungi and wildlife needs.

Our burn program and timber harvesting program are the two main cost effective ways we have to manipulate cwd. We utilize low intensity burns that allow us to rid the forest of flashy and small fuels, while retaining the larger logs for wildlife. This method speeds up the nutrient recycling process and creates a safer situation with respect to wildfire. Our timber harvest program is an uneven aged single tree selection process, which allows us to retain all snags and any green snags we choose. We have tremendous flexibility with our logging operation, which is beneficial to both us and our logging contractors. We have had the ability for many years to let the loggers leave any un! me!!!!rchantable log species in the forest for wildlife. In return their quick to do us "favors", like fall trees into meadows for structure, drag leftover logs back into the forest, or place cull logs in riparian areas. All of these seemingly small practices add up to a significant contribution to the effect cwd has on the habitat.

Example of Biomass Removal CostsJack NaylorThe presentation offers examples of costs associated with Biomass removal for three differing site conditions within the Plumas National Forest. Contract requirements and operating costs are reviewed as well as a brief discussion concerning fire load reduction, future silviculture plans, and the potential Biomass/BTU harvest at each site.

The three example areas (composed of 12 contract areas totaling approximately 226 acres) differed in both vegetation cover and future silviculture. They consisted of:1. Dense stands of 24-40 year old (8-19"dbh) white fir to be thinned to a 20'X20' spacing, with all residual "older growth" trees and dead snag retained.2. White Thorn/Manzanita/Black Oak complex to be cleared for replanting, with all Black Oak, alive or dead retained.

3. Pine plantations (8-15 years) to be released from competition with White Thorn/Manzanita/Tan Oak brush.The costs are broken down into five categories;1. Labor - faller/cutter, equipment operator, oiler/mechanic.2. Equipment - excavator/grapple, lowboy, fuel/maintenance truck.3. Water truck/fire equipment.4. Administration5. Total cost/acre for each unitIt was the U.S. Forest Service's decision to require low ground pressure tracked excavators with brush-rake grapples to insure low ground disturbance (lower compaction, decreased soil erosion, and keep soil water retention). The piling and windrowing of materials allowed more selectivity of tree, slash, and brush removal, less ground disturbance during brush removal and log and brush piling, better protection of the plant species left, and less soil and rocks in the burn piles.

Regional Patterns of Dead Wood in Forested Habitats of Oregon and WashingtonJanet L. Ohmann and Karen L. WaddellDead trees are an important element of productive and biologically diverse forests, yet little is known about how dead wood varies across broad regions. Such information is needed for assessing habitat suitability and ecosystem health at the regional scale, and for guiding management of stands and landscapes. We compiled data on snags and fallen trees from 15,000 field plots across 49 million acres of forest in Oregon and Washington. Dead wood abundance and characteristics were summarized for vegetation types that provide distinct wildlife habitats. Current patterns of snags and logs are highly variable and associated with environmental gradients, forest composition, and successi! on!!!!al stage. Moist maritime forests contain more dead wood than drier interior environments. Large wood is most abundant in young natural forests and in late-successional forests on federal lands. In young managed forests on private lands, most large snags and logs are remnants from previously harvested forests, which will not be replaced. The information on regional variation of dead wood has been used in developing the DecAID model, which guides management of the amounts, characteristics, and spatial distribution of dead wood within stands and across landscapes.

Decay Processes -- Why Simply Managing for Snag and Log Quotas is not EnoughCatherine G. ParksWood decay is a significant ecological process affecting forest systems. Pathogenic and saprophytic fungi that cause disease, death, and decay of trees promote biological and structural diversity, create critical habitat for wildlife, and act as recycling agents to convert trees to nutrient-rich organic matter. Forest ecologists have largely ignored that how a tree dies markedly influences several ecological processes. Trees killed or weakened by fire, drought, defoliation, beetle attack, or disease, each have a distinct decay trajectory that influences their structure and potential use by wildlife. Experimental management techniques that promote specific decay conditions in snag! s !!!!and logs are presented.

The Nutritional Significance of Coarse Woody Debris in Three Rocky Mountain ForestsCindy Prescott and Raija LaihoThe contribution of CWD relative to other aboveground litter types to nitrogen and phosphorus cycles was assessed in three Rocky Mountain forests: a self-thinning lodgepole pine forest, a mature white spruce forest, and an old-growth forest of Engelmann spruce and subalpine fir. Weight loss and changes in N and P concentrations in decomposing log segments were measured over a 14-year period. Input rates were measured during ten years for CWD, one year for ground vegetation, and five years for other aboveground litter types. N and P release from each type of decomposing litter were simulated for a period of 40 years. The role of CWD in nutrient cycling in these forests was assesse! d !!!!by determining 1) the proportion of the annual aboveground inputs of C, N and P that is derived from CWD, and 2) the proportion of the C, N and P released from aboveground litter during the simulated 40-year period that is derived from CWD. During the 14 years pine logs lost 71% of their original weight, spruce lost 38% and fir lost 40%. If current trends continued, pine logs would disappear completely in about 35 years, spruce logs in 45 years, and fir logs in 35 years. Nutrient release during log decomposition differed between the two elements and among the three species of logs. The pine logs gained N, spruce logs released some N, and the fir logs released almost 30% of their original N content. This pattern was related to the initial N concentrations in the logs, which were lowest in the pine logs and highest in the fir logs. A similar pattern was even more striking for P: the fir logs with very low initial P concentration immobilized addition! al!!!! P equivalent to almost four times their initial content. These findings suggest that wood decay organisms may be competing with vegetation for limiting nutrients. The proportion of CWD in aboveground litter input was 19% at the pine site, 3% at the spruce site, and 24% at the fir site. The contribution of CWD to N and P release during the 40-year simulation was 2% or less except at the fir site where CWD released 5% of the N. Our findings indicate that CWD is not a significant source of available N and P in these forests; it may actually compete with vegetation for limiting nutrients. The amounts of woody debris to be maintained should be based on management objectives related to other functions of CWD, rather than nutritional considerations.

The Effects of Microsite (Logs Versus Ground Surface) on the Presence of Forest Floor Biota in a Second Growth Hardwood ForestCharlotte Pyle and Michelle M. BrownA study was conducted in the University of Connecticut Forest to determine the significance of forest floor microsite (logs versus ground surface), log decay class, and log surface substrate for the presence/absence of representatives of seven broadly-defined groups of biota. A total of 277 logs was sampled along randomly located transects. For each log, an equivalent area of ground surface was sampled. Chi-square analysis indicated presence/absence of the biotic groups was not independent of forest floor microsite. Seedlings and (non-seedling) vascular plants occurred more frequently on the ground surface than on logs while evidence of small mammals and birds, and algae, fungi, ! an!!!!d bryophytes were present only on logs. Log decay class was significant

only for seedlings. The occurrence of algae, fungi, bryophytes, seedlings, and vascular plants on each of five log surface substrates was not proportional to the occurrence of those substrates. Differences between observed and expected values were greatest for the powdery wood substrate with very high observed frequencies of seedlings and vascular plants on powdery wood but very low frequencies of algae, fungi, and bryophytes. Reasons for the overall lack of importance of logs as seedling microsites in this oak-dominated forest are discussed.

Preliminary Results of Monitoring Effects of Five Prescribed Fires on Key Wildlife Habitat ComponentsTammy Randall-Parker and Richard MillerFive prescribed burns were monitored to assess gains and losses on key wildlife habitat components. The prescribed burs occurred on the Coconino National Forest, Kaibab National Forest, and Walnut Canyon National Monument in Northern Arizona. Live ponderosa pine, snags, down logs, and live oaks were monitored. More than half of the down logs were consumed by the fires we monitored. Snags were moderately impacted by the burns. Few live oaks and ponderosa pine were killed or consumed.

Importance of Dead Wood to Bark BeetlesDarrell W. RossBark beetle population dynamics are closely tied to the availability of breeding sites. At low population densities, beetles breed primarily in scattered trees that have recently died from other causes or in living trees with below average defensive capacities. Most of the time these preferred breeding sites are in limited supply resulting in intense competition both within and between species. When natural or human-caused disturbances increase the abundance of suitable breeding sites, beetle populations can rapidly increase to high densities. At high population densities, some bark beetle species are capable of attacking and killing large numbers of healthy trees in short period! s !!!!of time. The tree mortality resulting from a bark beetle epidemic produces a pulse of dead wood that is sometimes much greater than that produced by the disturbance that initiated the outbreak. A diverse community of arthropods follows bark beetle infestations. Some associated insects such as predators and parasites are attracted to bark beetle infested trees by the pheromones that the beetles use to communicate with one another. Others such as scavengers and detritovores are attracted by volatiles released from the decomposing trees. This diverse arthropod community is involved in many important ecological processes. Natural resource management activities can greatly influence bark beetle populations and, thus, dead wood dynamics.

Deadwood and the Regulatory ProcessChristopher P. RowneyIt is no secret that the manner in which deadwood is viewed has changed considerably over the years. A review of Board of Forestry regulatory mechanisms relating to the presence, or absence, over the last 100 years addressing standing and down wood provides some insight into the

priorities of policymakers. Those priorities have changed significantly, as the role of deadwood is better understood. California has evolved from policies in which deadwood enhances wildfire risks and had no recognized ecological value, to policies that demonstrate at least a nascent understanding of deadwood's role in effective resource management.

The policy decisions that must be made relating to such issues rely on science, but are also driven by societal considerations. Regulatory policy is shifting from species-specific to landscape based approaches to wildlife and watershed management. To be effective, such policy requires stakeholder understanding and support. This is best achieved through education and incentive-based approaches, although some level of prescriptive regulation is unavoidable.

Importance of Snag Clumps to Cavity-nesting Birds in Burned LandscapesVictoria Saab, Ree Brannon, Jonathan Dudley, Larry Donohoo, Vicky Johnson, Dave Vanderzanden, and Henry LachowskiForests affected by wildfire, and subsequent salvage logging are increasingly prevalent across much of the inland Northwest. Many cavity-nesting birds are associated with burned forests but little is known about their responses to postfire salvage logging. We monitored 1,288 nests of 9 cavity nesters from 1994-1998 in unlogged and salvage-logged units of crown-burned ponderosa pine/Douglas-fir forests. Lewis' woodpecker was the most abundant (356 nests) cavity nester on the recent burns (wildfires in 1992 and 1994), while black-backed and white-headed woodpeckers were relatively rare (35 and 14 nests, respectively). Snag densities for trees >23 cm dbh averaged 38.3 ! +!!!! 5.0 per ha in unlogged units and averaged 31.4 + 1.9 per ha in salvage-logged units. Snag densities were significantly higher at nest sites of all species compared to average conditions. In the salvage-logging prescriptions, snags were retained in evenly-spaced, uniform distributions, while within those treatments, cavity-nesting birds selected clumps of snags for their nest sites. This suggests that management can improve the prescriptions to favor cavity-nesting birds by changing the distribution of trees retained (from uniform to clumped), even when the same number of trees are harvested. Black-backed Woodpeckers selected nest sites with the highest tree densities (126.9 + 24.6 snags/ha in unlogged; 106.3 + 15.7 snags/ha in logged), whereas Lewis' Woodpecker selected relatively open nest sites (91.7 + 26.4 snags/ha in unlogged; 62.1 + 3.7 snags/ha in logged). ! Le!!!!wis' Woodpecker nested almost exclusively in salvaged units, whereas black-backed woodpeckers favored the unlogged areas. Partially logged units with clumps of relatively large, soft snags characterized Lewis' woodpecker nest sites, whereas unlogged areas with high tree densities of relatively small, hard snags were typical of black-backed woodpecker nest sites. A range of stand conditions characteristic of black-backed and Lewis' woodpeckers would most likely incorporate local habitat features necessary for successful nesting of other members in the bird community.

Girdled vs. Bark Beetle Created Snags: Utilization by Cavity Dependent Species and Differences in Decay Rate and Insect DiversityPatrick J. Shea, William F. Laudenslayer, Jr., and George Ferrell

Snags are a natural component of forest ecosystems that have a variety of values. Snags provide important feeding substrates, nesting sites, and habitat for a variety of vertebrates and invertebrates and these functions continue after the snag eventually falls and becomes a log. Logs are also important in nutrient cycling as they decompose and slowly return nutrients to the soil. Some forest landscapes are "snag deficient" because of direct action by forest managers or because of natural events such as winter storms or fire. Therefore it is not unusual for forest managers to create snags by various means. In this study we compared snags created by mechanical girdling (chain! sa!!!!w) and bark beetle (western pine beetle aggregation pheromone) attack as to acceptability and suitability to cavity dependent species and insect diversity. Imbedded within the experimental design were two other factors: size of snag and season of mortality. A randomized complete block experimental design was used with 6 replications of each treatment resulting in 48 trees. Six years into the study we have found differences in rate of degradation (tree fading, needle, twig, branchlet retention, top breakage, bark sloughing), woodpecker feeding activity, cavity building, and insect diversity. In summary our data indicate that bark beetle killed trees provide an more biological rich snag that is both suitable and acceptable to cavity dependent species.

Some Thoughts on the Dynamics of Fire and Dead Woody MaterialCarl N. SkinnerThe frequent occurrence of fire in most forested areas of California before this century has been well established. Likewise, the importance of dead woody material to various wildlife species as snags and downed logs has also be well established. Those that would reintroduce fire as an ecological process often appear to be at odds with those that would protect available large, dead woody material. Yet, for wildlife species to have developed a need for this material suggests that there was always at least some available in many landscapes over time. Observations of fire effects on dead woody material, as well as some recent limited data, indicate that fire often consumes most mate! ri!!!!al that is in advanced stages of decay. It appears unlikely that many pieces of large woody material survived fire long enough to fully decompose in California's pre-fire suppression era fire regimes. However, though many appear to be consumed as well, hard snags and logs may, at least in part, survive low severity fires. Fires also help to create snags and, ultimately, downed logs. The frequent low-moderate-severity fires that characterized much of the forested landscapes of California burned with mosaics of severity related to topography and weather conditions. The result was likely a landscape with many of the snags and logs clustered both in time and in space and very sparsely distributed in the intervening time and space. The patterns of severity related to landscape topographic structure and its implications for accumulations of dead wood will be discussed.

Characteristics and Dynamics of Cavity Nest Trees in Forests of Southern British ColumbiaChristoph Steeger and Jakob A. DulisseStanding dead and diseased trees (wildlife trees) are essential habitat components for cavity-

nesting vertebrates but are often targeted for removal in managed forests. To assist forest managers in selecting ecologically-valuable trees for retention during forestry operations, we investigate the characteristics and dynamics of cavity nest trees. Annually since 1994, we have located and assessed nest trees in mature forests of southern British Columbia. We determine nesting species and nest tree characteristics (species, size, condition, and disturbance agents) at the time of first detection and during subsequent years, to document changes in wildlife use and tree condition. To da! te!!!!, we located a total of 505 nests of 15 cavity nesters in 378 trees. Populus tremuloides, Pseudotsuga menziesii, Betula papyrifera and Larix occidentalis accounted for 91% of all nest trees. Of those trees 20%, 98%, x% and y% were dead when first detected, respectively, and 54% had broken tops. For all nest tree cohorts (nest trees found within one particular year) combined, 3.3% of nests trees were lost annually due to uprooting or total breakage and 3.9% experienced partial breakage each year. Complex patterns of use and reuse of nest trees and cavities exist, within and among wildlife species and years.

Snag Densities and Fuel Loads in Jeffrey Pine-Mixed Conifer Forests Under an Unmanaged Fire Regime in the Sierra San Pedro Martir Mountains, MexicoScott L. Stephens and Samantha J. GillFire suppression has increased fuel loads and may have changed the spatial distribution of snags and large woody debris in western coniferous forests. Determining snag and fuel management standards that can be applied to western coniferous forests is difficult because there is very little prehistoric data in this area. Analysis of data taken from the mixed conifer forests of the Sierra San Pedro Martir Mountains (SSPM) Mountains of northern Baja California, Mexico, are presented as a forest ecosystem that can provide information in this area. The SSPM is unique within the California floristic province in that its forests have never been harvested and are still influenced by a lig! ht!!!!ning ignited fires. The mixed conifer forests of the SSPM are composed of Jeffrey pine (Pinus jeffreyi), white fir (Abies concolor), sugar pine (Pinus lambertiana), and lodgepole pine (Pinus contorta). A systematic inventory in a Jeffrey-pine mixed conifer forest using 49 plots was conducted. Surface and ground fuels were inventoried using 3 Brown transects at each plot and snags density was determined using 0.4 ha circular plots centered at each of the 49 plots. Snag density averaged 4.3/ha (range 0 - 23/ha). Surface and ground fuels averaged 16.47 metric tonnes/ha (range 1-166 metric tonnes/ha). Tremendous variability was found in forest structure characteristics in a relatively small area (1.44 km2) with no changes in forest type, elevation, aspect, and soils. Average values were almost never found in any individual plot and this challenges many current US Forest Service g! ui!!!!delines that require average densities of snags, large woody debris, and fuels in many habitats.

Evaluating the Effects of Partial Cutting on Wildlife Trees and Coarse Woody DebrisSusan K. Stevenson and Dagmar G. KeiskerIn the northern interior wet-belt of British Columbia, where stand-destroying events occur infrequently, uneven-aged management is expected to produce stands that are more natural in structure than those that result from clearcutting. The contention that biodiversity can be

maintained through the use of ecologically-appropriate silvicultural systems requires testing. We have begun a long-term replicated study of the effects of several partial cutting prescriptions on a variety of structural attributes and ecosystem processes, including the loss and creation of wildlife trees and coarse woody debris. That component of our study focuses on comparing the occurrence of damage agents ! an!!!!d of structural attributes that are functionally important to wildlife in undisturbed and post-harvest stands. We employ a new classification of functional types of wildlife trees and coarse woody debris. A Type is a configuration of habitat features required by one or more wildlife species for specific functions. For example, Type 1 wildlife trees, which have hard outer wood surrounding decay-softened inner wood, appear to be required by strong primary cavity excavators for nesting. We present preliminary data on the occurrence of disturbance agents, Wildlife Tree Types, and Coarse Woody Debris Types in unharvested cedar-hemlock and Engelmann spruce-subalpine fir stands in the interior wet-belt of British Columbia.

Standing Dead Tree Dynamics Extracted from Growth and Yield Permanent Sample Plots in British ColumbiaJeff Stone and John ParminterPermanent sample plots (PSPs) of the growth and yield (GY) of live trees can provide valuable information on standing dead tree dynamics. GY PSPs traditionally documented the death of a live tree but failed to follow trees beyond death. Fortunately, the repeated measures nature of PSPs enable information on the past dynamics of dead trees to be extracted following an assessment of the standing dead trees. This cohort based information is readily analyzed (e.g., by logistic regression) and can be incorporated into GY models for the prediction of snag dynamics.

In British Columbia, the Ministry of Forests administers over 5,500 GY PSPs. These plots which are distributed throughout the province were established for research and inventory purposes. We measured a subset of the plots for standing dead tree breakdown and incorporated this information into the Ministry's GY models TASS and TIPSY. To determine the feasibility of measuring the full PSP program, a pilot study was undertaken to evaluate the time and cost requirements. In this paper, we discuss the dead tree information obtainable from GY PSPs, the time and costs to obtain that information, findings on standing dead tree dynamics, and the development of snag dynamic models in Brit! is!!!!h Columbia.

Coarse Woody Debris in Wyoming Lodgepole Pine Forests after Fire and ClearcuttingDaniel B. Tinker and Dennis H. KnightCWD was measured in burned, clearcut, and unburned/uncut lodgepole pine stands in the Medicine Bow National Forest (MBNF) and in Yellowstone National Park (YNP), Wyoming. Estimates of the amount of CWD consumed or converted to charcoal by fire were made from a recently-burned stand in YNP, and a spatially-explicit simulation model, DEADWOOD, was created to simulate various clearcutting and fire regimes over a 1000-year period. Measurements following a single clearcut or fire revealed that downed CWD > 7.5 cm in diameter did not differ

between the two, but fires created more potential CWD as standing-dead trees. Notably, 35% of the CWD in clearcut stands was inherited wood from th! e !!!!previous stand-replacing fire; this source of CWD will decrease with successive harvests. Only 8% of downed CWD was consumed during a single fire; an additional 8% was converted to charcoal. Following a single disturbance event, clearcutting resulted in an average net loss of 83 Mg/ha, while natural fires created an average net gain in CWD of 95 Mg/ha. After 1000 years of simulation, 100-year fire-return intervals produced CWD that occupied more of the forest floor than did 200- or 300-year intervals. The time required for 100% occupancy of the forest floor by CWD was 1125, 1350, and 1300 years for 100-, 200-, and 300-year fire-return intervals, respectively. Simulations revealed that current harvest regimes will require at least four centuries longer for 100% of the forest floor to be occupied by wood (1800-3600 yr) than under fire regimes. Only when post-harvest CWD slash was simulated at double the present amounts did clearcutting produce suffi! ci!!!!ent CWD to maintain forest floor CWD within the range of variability for naturally-developing stands.

Characteristics of Log Resources in the Interior Columbia River BasinTorolf R. Torgersen and Lisa J.BateLogs have complex physical, biological, and functional attributes. The body of literature on detailed sampling of log resources in diverse habitats in the Interior Columbia River Basin is sparse. Sampling schemes developed for assessing logs as woody fuel usually have translated amounts of logs into weight or volume (tons or cubic feet) per acre. To depict amounts of logs in a more descriptive way,---to assess wildlife habitat, for example---we describe log resources in terms of number, mean length, linear feet, volume, and percent ground cover per acre. These measurements are further categorized by large-end diameter classes, and species composition. We present sampling data on ! lo!!!!g resources in the Blue Mountains of northeastern Oregon, and in the Salish and Mission Mountains of western Montana. Data representing mid- and late-structure mixed conifers and ponderosa pine stands dominate these data. Studies dealing with changes in log resources following salvage harvest to reduce fuels or remove dead and dying volume, and after prescribed burns are presented.

Forest Structure and Tree-use by Woodpeckers in Church Wood Nature ReserveRafael Villegas-PatracaThe general aim of this work was to evaluate the potential of woodlands trees as a nesting habitat for birds. This study focused on an estimation of forest structure, dead wood volume and use of trees (live and dead) by woodpeckers. It was carried out in Church Wood Nature Reserve, in Kent England. Woodpeckers specially were selected for investigation, because their presence is easy to detect by the holes in the trees. Three different species of woodpeckers (green Spotted Picus viridis, Great Spotted Dendrocopus major and Lesser Spotted Dendrocopus minor) use live and dead trees as habitats in Church Wood. This reserve is part of one of! t!!!!he most ancient Oak Woods in Southern England. Some types of disturbances, like storms, cause the accumulation of dead wood in the forest. in this reserve, for example there are many fallen trees

as a result of the 1987 storms. The objective of this study were to 1) estimate the structure (basal area and density) of live oak trees 2) estimate the process of regeneration of oak using the relation between canopy and number of juveniles; 3) estimate the volume of dead tree wood and 4) determine whether dead trees are used by woodpeckers as habitats in preference to live trees. The methodology used consisted of: a survey of all species of trees present in the reserve; measurements of diameter, estimation of canopy cover, regeneration and health of the Oak trees; a survey and identification of all dead trees with measurements of diameter, length (fallen dead trees) and height (standing dead trees) and a survey of woodpecker holes in live Oak trees and ! in!!!! all species of dead trees. The total density in the reserve was 1551.08 ind / ha. According to the Shannon Wiener Biodiversity Index some compartment had higher biodiversity than others. For Oaks the basal area was 38.14 (m2/ha) and 88 % of the individuals have good level of canopy density. According to the regeneration survey, less than 50 % (fewer than half of the trees surveyed)exhibited a low regeneration potential. The total density of dead trees was 97.38 ind/ha (almost 90% of all dead trees had fallen and only a small number standing). The volume of dead wood was 6500.78m3. the number of woodpeckers nests was variable. Some compartments had no recorded nests. The presence or absence of woodpeckers nets in live oak trees was positively correlated with the health of the tree. In some trees more than one woodpeckers nest was recorded. In the case of dead trees the number of woodpecker nest by tree was not correlated wi! th!!!! the DBH or height of the same tree. Oak is the most common tree species used by woodpecker for nesting in Church Wood. Other tree specie used by woodpeckers were Beech and Birch. To conclude, in general the presence of dead trees is very important because they provide habitats and foraging sites for hole nesting birds (woodpeckers) and many other species.

Overview of Agents and Patterns of Coarse Woody Debris Recruitment in Western ForestsMichael R. Wagner and Brytten E. NelsonThis paper will present an overview of agents of tree mortality in western coniferous forests that are responsible for coarse woody debris recruitment under natural conditions. The patterns and scales of occurrence of these mortality agents will be reviewed. Recognition ofthese factors and how they can be used to assess historic disturbance patterns and improve management of coarse woody debris will be discussed. Finally risks associated with management activities designed to create coarse woody debris and their mitigation are summarized.

Composition of Snags in Primeval Spruce-Pine Stands of the Bialowieza National Park, PolandWieslaw WalankiewiczThe Bialowieza Forest complex is situated on the Polish-Belarussan border. Its western part (580 km, c. 45% of the area) belongs to Poland. The forest represents a remnant of the vast lowland forests that once covered great parts of temperate Europe. A block c. 50 km2 of the best preserved primeval forest has been strictly protected within the Bialowieza National Park (hereafter BNP). In the coniferous stands (c. 25% of the BNP area) only two tree species coexist

(with an admixture of birches and oaks.). These are Norway spruce (Picea abies) and Scotch pine (Pinus silvestris). These stands are in deep transforming processes. Th! is!!!! is because the Norway spruce regenerates while the Scotch pine does not since 80 years or so. It means that the spruce is going to replace the pine. Although pine trees live much longer than the spruce this species requires much more light for regeneration. Such situation (open areas with good sun illumination) seems to occur in Bialowieza Forest once per 100 - 200 years. This happens only when fire or wind destroys forest creating vast open areas. Within our study plot there are no younger pine trees than c. 60 years old while the spruce is represented by all age classes. This specific succession affects snag composition. Pine snags are greater in size than spruce snags and they are very durable. Some of them stand even 50 years after a tree dies. Spruce snags usually fall dawn within a few years after a tree dies. Birds use pine snags mostly for excavating cavities. Spruce snags are important feeding places for woodpeckers although in some spruce snag! !!!!s are excavated cavities as well.

Snag and Down Wood Relationships with Fire in Western Hemlock Forests of OregonDiane E. White, Tom Atzet, Lisa A. McCrimmon, and Patricia A. MartinezThis paper describes the relationships between fire and density of snags and down wood (51 cm + DBH) at a landscape level in the Western Hemlock (Tsuga heterophylla) Plant Series in southwestern Oregon. Data were collected from 14 plant associations on the Rogue River, Siskiyou, and Umpqua National Forests. A wide range of environments was covered, with mean annual temperatures between 6oC and 12oC and mean annual precipitation between 127 cm and 292 cm. The variables collected were live tree, snag, and down wood densities, fire disturbance information, and elevation, aspect, slope and slope position. Large, live trees were plentiful, supp! ly!!!!ing a pool from which snags and down wood could be created. The median number of snags per hectare ranged from 2.1 to 15.3. The median number of pieces of down wood per hectare ranged from 0 to 14.9. In all plant associations, the variation in both snags and down wood was large. Snag density was negatively correlated with average fire severity and positively correlated with the length of the last fire interval. Multiple regression showed over half the variation in snag density was explained by these two variables. Down wood density was not significantly correlated with either variable.

DecAID: Insect and Disease ComponentElizabeth A. Willhite, Bruce B. Hostetler, and Roger E. SandquistThe insect and disease component of DecAID provides information about disturbance agents affecting recruitment of decadent wood. Occurrence frequencies of plots or trees affected by selected insect and/or disease species were obtained from vegetation inventory plots in terrestrial upland forests. The examined inventory data was stratified by habitat type and structural class, and included information on species composition and structure of living trees, dead standing trees, and down wood. This portion of the model provides a list of the insects and diseases which are most likely to generate decadent wood in these strata, and describes the relationship of these disturbance agents ! to!!!! decadent wood dynamics. Feedback loops are described where

applicable. The model also describes ways to mitigate or intensify the activity of these agents to decrease or increase decadent wood recruitment.

The Effect of Fire Regime on Coarse Woody Debris in the West, Central Cascades, OregonPamela WrightThis study examined the effect of fire regime on coarse woody debris (CWD) mass using a combination of field data and modeling. The objectives were to use field sampling to determine how CWD differs between two areas that have had different fire regimes (120 to 300 year mean fire return intervals), and investigate how fire frequency and severity, stand growth and development, decomposition, and mortality rates affect the amount of CWD for sites experiencing two different fire regimes. A mass-budget model was developed to compare field data and model results, and better understand the dynamics of CWD. While fire frequency and severity established the pattern of CWD succession, the! i!!!!nteraction of fire regime with the other controlling factors is responsible for the differences in CWD mass. CWD mass was almost twice as high in stands having an infrequent, stand-replacing fire regime (173 Mg/ha) compared with stands having a moderately frequent, mixed-severity fire regime (95 Mg/ha). Factors that appear to have the greatest influence are decomposition rates, fire severity, fire frequency, and fuel consumption, suggesting that environment and stand structure control CWD mass more than fire regime per se. CWD levels have greater temporal variability in the infrequent, stand-replacing fire regime, and mortality occurs during long periods between fire events; whereas, in the mixed-severity regime mortality is more associated with events than with stand mortality over time.

Forest Insect Defoliation and Tree Mortality: The Role of Disturbance Factors in the Creation of Dead WoodAndrew Youngblood and Boyd E. WickmanKnowledge of natural disturbance factors that influence ecosystems over large temporal and spatial scales is essential for understanding the interaction between various ecosystem processes. Disturbance often results in long-term changes that are not readily apparent until long after the disturbance event. Assessing the long-term changes from past disturbance events may aid in identifying ecosystem processes and components most resilient. In northeastern Oregon, one such disturbance that occurs over large temporal and spatial scales is defoliation by phytophagous insects. Concern over fuel accumulation and failed stand regeneration after tree mortality from insect defoliation ofte! n !!!!led to management decisions resulting in stand and ecosystem "restoration" prescriptions for harvesting and planting in mixed conifer stands. These practices were initiated without fully understanding the long-term impacts and implications of defoliation, mortality, snag creation, and dead wood. We describe studies to assess stand structure in long-term study plots originally established after Douglas-fir tussock moth and spruce budworm outbreaks in northeastern Oregon. After Douglas-fir tussock moth defoliation in the Wenaha-Tucannon Wilderness in 1972, and spruce budworm defoliation beginning in 1984 in Starkey Experimental Forest, all trees in 250 m2 plots were tagged and their survival or death

and transition to snag or down log monitored. We report on the current stand composition and structure including snags, the density of down logs, and the change in associated vegetation. We believe this work provides valuable in! si!!!!ghts into the role of large insect outbreaks and the dynamics of dead wood in these mixed conifer stands.

Are There Snags in the System? Comparing Cavity Use among Nesting Birds in "Snag-Rich" and "Snag-Poor" Eastside Pine ForestsSteve Zack, T. Luke George, and William F. Laudenslayer, Jr.The habitat structure and ecological processes for our western coniferous forests have changed dramatically in the past century primarily because of extensive logging and fire suppression; the forests of today often are denser, have a larger component of shade tolerant species, and have fewer large, older trees. The density of large snag trees and their generation has been greatly altered. Snags with cavities are needed as nesting sites for many species of birds, and are used by other wildlife. We have been studying wildlife-forest relationships at two sites in Northern California. The eastside pine forests at the Goosenest Adaptive Management Area in the Klamath NF are "typicial!!!!" in that they have been logged and fire suppressed this past century resulting in a dense forests with few large tree and few snags. In contrast, the eastside pine forests of the Blacks Mountain Experimental Forest in the Lassen NF, particularly those in the Research Natural Areas, have not been logged (but have been fire suppressed), and so contain many large pine trees and snags. Here, we compare the abundance and diversity of cavity-nesting birds and snag cavity use at these sites from 1999 and discuss the implications of our results for eastside pine forest management and wildlife.