Decayed Wood Advisor

Snag DBH Down Wood Diameter Snag Density Down Wood % Cover I&D Species List I&D Inventory Data
Data Caveats Scientific Names Back to Query Page Glossary Literature Cited Tutorial

Summary Narrative:

Advice on Decayed Wood in the Vegetation Condition

 

(WLCH_WCO_L)

TABLE OF CONTENTS


SYNTHESIS AND MANAGEMENT IMPLICATIONS

DecAID is a summary of the current knowledge and best available data on dead wood in Pacific Northwest ecosystems. The primary emphasis is on wildlife relationships to dead wood and summaries of dead wood levels from inventory data, however the data are presented in an ecosystem context. Snag dbh, down wood diameter, snag density, and down wood percent cover are statistically summarized using tolerance levels. For details on the statistical summarization of data see the Statistical Basis paper.

There are multiple possible interpretations of the data summaries, and many options for ways to apply the interpretations to forest management. Below we present one option we feel is a reasonable interpretation of the summarized data.

Considerations of scale: landscape and stand levels

Wildlife and inventory data summarized in the DecAID Advisor can be applied to management and planning decisions at a range of spatial scales and geographic extents. The calculated tolerance levels (80%, 50%, 30 %) for wildlife data can be applied to stand-level management. However, we do not advise that a particular tolerance level be applied to all stands across a landscape. Rather, decisions about how to distribute different levels of dead wood across a landscape can be guided by the distribution information from unharvested plots (see paragraph below).

When using the inventory data, analysis areas (landscapes or watersheds) should be sufficiently large to encompass the range of variation in wildlife habitat types and structural conditions that occur in the area in which the inventory data were collected (i.e. representative of variation at the sub-regional scale). It is impossible to specify a single minimum size analysis area that is most appropriate to all ecoregions and geographic areas. However, as a general rule-of-thumb we suggest that analysis areas be at least 20 square miles (12,800 acres, 5,120 ha) in size. This coincides with the small end of the range of sizes typical of 5th-field hydrologic unit codes (HUCs). The exception would be when managing habitats that are rare at a broad geographic scale, even if they are common at the watershed scale (e.g. recent post-fire habitats). Even at the scale of the 5th-field HUC, wildlife and inventory summaries for multiple vegetation conditions will need to be considered in the analysis and planning process.

On the other end of the scale, the wildlife and inventory summaries are appropriately applied to very broad analysis areas for regional assessments, National Forest planning, and so forth. The maximum size analysis area to which the DecAID summaries should be applied is defined by the geographic distribution of the wildlife studies and inventory plots on which the summaries are based – essentially Oregon and Washington. As at smaller scales, multiple wildlife habitats and vegetation conditions will need to be considered in the analysis and planning process.

If the objective is to manage for natural conditions rather than focusing on wildlife species, mimic the distribution of unharvested acres (unharvested proportion of the landscape) in different snag density classes (Figures WLCH_WCO_L.inv14 and WLCH_WCO_L.inv15) and down wood percent cover classes (Figures WLCH_WCO_L.inv16 and WLCH_WCO_L.inv17) across the landscape. The percentages should be thought of as general guidelines rather than taken literally since the distributions to a substantial degree reflect the plot size and sample design of the inventories (see Caveats and Cautions). The majority of the plot-sized areas on the unharvested landscape support densities of snags > 25.4 cm (10.0 in) dbh of up to 60 snags/ha (24/acre) and densities of snags > 50.0 cm (19.7 in) dbh of up to 30 snags/ha (12/acre). Areas of higher snag densities do occur on a portion of the unharvested landscape. Over a landscape balancing high density areas of dead wood with moderate and low-density areas may be desirable. Note that while DecAID focuses on dead wood-associated wildlife species, there are species that are negatively affected by high levels of dead wood.

Data are not available on the spatial distribution of dead wood within stands. Wildlife data indicate that species often use clumps of snags and down wood. However, the wildlife data offer no guidance on the number or size of clumps that should be managed. In lieu of actual data, managers should take advantage of naturally occurring clumps in the stands they are managing. Areas between clumps should not necessarily be devoid of snags. A mix of clumps and more widely distributed snags should occur in the stand.

Within the Larger Trees Structural Condition Class, canopy closure can range from 10 to 100%. It may not be possible to maintain high levels of dead wood on sites with few trees, especially those sites that are sparse due to inherent site factors such as soil or productivity. Higher density areas of dead wood can be maintained only on sites that currently have high live biomass or have the potential to produce high levels of biomass.

The location of dead wood (particularly snag) clumps may be determined, to a large degree, by operational considerations. See Neitro et al. (1985) and Rose et al. (2001) for discussions of operational considerations when managing snags.

Considerations for developing, creating and retaining decayed wood elements

The following is a summary of information found in the Considerations for Developing, Creating and Retaining Decayed Wood Elements paper that is specific to the Westside Lowland Conifer-Hardwood Forest (WLCH) Wildlife Habitat Type.

Walter (2004) examined use of snags left after harvest activities using different harvest intensities and patterns. He found that clustered and scattered snags received equal use in terms of nesting and foraging. Trees that were topped received more use than trees that were not topped.

Snag density and dbh

Wildlife data on use of snags in this vegetation condition are available for 18 species and 3 groups of species from 20 studies across Oregon, Washington and southern British Columbia. Data are available for snags used for nesting, denning, roosting, and foraging. Data specific to the Washington Coast subregion are available for cavity nesting birds (Zarnowitz 1982, Zarnowitz and Manuwal 1985), northern pygmy owl (Giese et al. 1997), pileated woodpecker (Aubry and Raley 2002, Raley and Aubry 2006). In addition data are available from Vancouver Island, B.C. for pileated woodpecker (Hartwig 1999, Hartwig et al. 2004).

In addition to snags, defective live trees are important habitat in this habitat type. Aubry and Raley (2002) found that defective live trees (dead tops and/or hollow) were important for pileated woodpecker nests and roosts; density of dead topped trees was higher at nest and roost sites than random sites. Defective trees are also important for spotted owl nesting habitat; Hershey et al. (1998) found density of broken topped live trees was higher at nest sites than random sites and that these defective trees were used for nest trees.

In addition to wildlife data, inventory data from inventory plots should provide information on snag habitat under natural conditions. Use inventory data from all unharvested plots to estimate landscape distribution of snag densities; use inventory data from unharvested plots with measurable snags to provide information on that portion of the landscape where snag habitat is to be provided, this assumes that dead-wood-associated species only utilize those parts of the landscape where snags are present. Most of this habitat type has not yet missed a fire cycle, thus dead wood in unharvested stands should closely mimic historical or natural conditions.

Down wood percent cover and diameter

Wildlife data on use of down wood in this vegetation condition are available for 25 species and 1 group from 9 studies from across western Oregon, Washington, and southern British Columbia. Most of the wildlife data are from the one study in the Oregon Cascades (Umpqua National Forest) (Maguire unpublished data). The study on the Umpqua National Forest also tends to be a drier, less productive area, with more frequent fires than areas in the Washington Coast subregion. Data specific to the Washington Coast subregion are available for: VanDyke’s salamander (Blessing et al. 1999), and foraging pileated woodpeckers (Raley and Aubry 2006). Data are also available from Vancouver Island, B.C. for pileated woodpecker (Hartwig 1999), and deer mouse (Carter 1993).

Additional data not included on the cumulative species curves also indicate a positive relationship between dead wood and a variety of species. Carey and Harrington (2001) found that shrew moles reached a peak abundance in stands with 10% cover of down wood. Deer mouse abundance was positively correlated with percent cover of down wood (Carey and Harrington 2001) and with total number of pieces of coarse woody debris (Carter 1993). However, survival of deer mice peaked at down wood volume of
2.0m3/0.01 ha and declined to zero at >6m3/0.01 ha (Manning and Edge 2004). McDade (2001) found a positive correlation between down wood and crickets, snails and ensatina. Large logs provide important habitat for Plethodontid amphibians, because they are better at regulating temperatures and moisture than smaller logs (Kluber et al. 2009).

In addition to wildlife data, inventory data from inventory plots should provide information on down wood habitat under natural conditions. Use inventory data from all unharvested plots to estimate landscape distribution of down wood amounts; use inventory data from unharvested plots with measurable down wood to provide information on that portion of the landscape where down wood habitat is to be provided, this assumes that dead-wood-associated species only utilize those parts of the landscape where down wood is present. Most of this habitat type has not yet missed a fire cycle, thus dead wood in unharvested stands should closely mimic historical or natural conditions.

When comparing wildlife and inventory data, it should be noted that inventory data on percent cover of down wood may be lower than wildlife data because the majority of wildlife studies include class 5 down wood (often the most numerous class (Spies et al. 1988)), but inventory data do not consistently include class 5 down wood. The minimum diameter of down wood measured on inventory plots was 12.5 cm (4.9 in) while the minimum for most wildlife studies was 10 cm (4 in). As a result slightly less wood was included in the measurement of cover for inventory plots. In addition, it should be noted that several of the data points on the wildlife curves include just decayed wood or a larger minimum diameter for wood included in percent cover calculations.

INTRODUCTION TO VEGETATION CONDITION

Habitat Type Description

The Westside Lowlands Conifer-Hardwood Forest Habitat Type is described in Chappell et al. (2001). The following description is summarized primarily from this publication.

Geographic Distribution: The WLCH_WCO Habitat Type occurs in western Washington at relatively low elevations, specifically the Puget Lowlands, Olympic Peninsula, and the Coast Range.

Figure WLCH_WCO.inv-1 shows locations of forest inventory plots used in the dead wood summaries for Westside Lowland Conifer-Hardwood Forest Wildlife Habitat Type in the Washington Coast subregion. “Evidence of Harvest on Plot” indicates if there was evidence of previous harvest activity recorded for the plot, or if there was a road within the plot area. Data from plots with no recorded harvest or roads were the basis for describing natural conditions for this vegetation condition.

Physical Setting for the Washington Coast subregion:
Climate – moist to wet with relatively mild temperatures.
Mean annual precipitation – 90 to 254 cm (35 to 100 in) mostly as rain; snowfall is rare and transitory.
Elevation – from near sea level to about 610 m (2,000 ft).

Composition: Douglas-fir, western hemlock, western redcedar, and/or big-leaf maple are the dominant tree species in most stands. Sitka spruce is a dominant species along the coast where summer fog is common. Common dominant and co-dominant shrubs include: salal, dwarf Oregongrape, vine maple, salmonberry, trailing blackberry, and several species of huckleberry. Common forbs and ferns include: swordfern, Oregon oxalis, deerfern, bracken fern, vanillaleaf, and foamflower.

Structural Condition Description

Large tree stands have average tree size (quadratic mean diameter) greater than 50.0 cm (19.7 in) and tree stocking or cover of at least 10%. These stands are usually late-successional, including but not limited to old-growth forests. Very large trees usually are scattered throughout the stand. A grass/forb or shrub understory is often present. Stands may or may not have distinct canopy layers.

Vegetation inventory data, based on the entire plot sample, indicate the area within this wildlife habitat type and structural condition class is distributed as follows:

INTRODUCTION TO AVAILABLE DATA

Wildlife data

The wildlife data for the Westside Lowland Conifer-Hardwood Habitat Type are not divided by subregion. The amount of wildlife data available is low and habitat conditions in the subregions are similar. Combining the data allows for a more robust set of cumulative species curves. Where the combined data may be misleading or inappropriate for this subregion it is noted below.

The wildlife data on use of snags summarized in the cumulative species curves for this Wildlife Habitat Type and Structural Condition Class are for 18 species and 3 groups of species from 20 studies from across western Oregon, Washington, and southern British Columbia. Data are available for snags used for nesting, denning, roosting, and foraging. Data specific to the Washington Coast subregion are available for cavity nesting birds (Zarnowitz 1982, Zarnowitz and Manuwal 1985), northern pygmy owl (Giese et al. 1997), pileated woodpecker (Aubry and Raley 2002, Raley and Aubry 2006). In addition data are available from Vancouver Island, B.C. for pileated woodpecker (Hartwig 1999, Hartwig et al. 2004).

The wildlife data on use of down wood summarized in the cumulative species curves for this Wildlife Habitat Type and Structural Condition Class are for 25 species and 1 group from 9 studies from across western Oregon, Washington, and southern British Columbia. Most of the wildlife data are from the one study in the Oregon Cascades (Umpqua National Forest) (Maguire unpublished data). The study on the Umpqua National Forest also tends to be a drier, less productive area, with more frequent fires than areas in the Washington Coast subregion. Data specific to the Washington Coast subregion are available for VanDyke’s salamander (Blessing et al. 1999)
, and foraging pileated woodpeckers (Raley and Aubry 2006). Data are also available from Vancouver Island, B.C. for pileated woodpecker (Hartwig 1999), and deer mouse (Carter 1993).

Several additional studies report information on wildlife and other species' use of dead wood that was not incorporated into the cumulative species curves. Some of these wildlife studies clearly demonstrated selection by some species for specific amounts and sizes of snags and down wood. Some of these studies presented results only in summary form, such as in regressions and statistical summaries, which we could not use in the species-specific presentations of DecAID. Still, these studies have value in corroborating other evidence of selection patterns.

The following additional information is available for the Washington Coast, links are provided to the annotated bibliography:
To view a list of studies on wildlife species use of snags or down wood click one of the following links. Information on wildlife species, including and in addition to those for which specific snag and down wood use data were available, were taken from the Species-Habitat Relations (wildlife-habitat relationships) database of O’Neil et al. (2001). This additional information includes a full listing of all dead wood associated wildlife species expected to occur in each wildlife-habitat type and structural condition class, and the species’ key ecological functions (see the General Wildlife-Habitat Relations With Wood Decay Elements and Ecological Functions and Processes of Decayed Wood Elements).

Inventory data

Snag data were collected on 119 inventory plots (all sampled area in WLCH_WCO_L), including 41 unharvested plots. Down wood data were collected on 42 plots, including 22 unharvested plots (20% of all sampled area in WLCH_WCO_L). Snags were sampled on all ownerships, whereas down wood was sampled only on federal lands in this vegetation condition. Because the forest area represented by each of these plots varies, the dead wood distributions and size summaries are based on sampled area rather than on numbers of plots, with plots given different weights in the calculations.

Tolerance levels for snag density and down wood cover for the total landscape were based on a subsample of all inventory plots, totaling 52 plots for snags and 14 plots for down wood. Tolerance levels in unharvested forests were calculated using plots on federal lands only (not subsampled): snag and down wood data were available for 22 plots.

Tolerance levels in unharvested forests on federal lands also were calculated using just those plots with measurable dead wood (i.e., based just on plots which included at least one piece of dead wood of the specified minumum size). There were 22 plots with snags >25.4 cm (10.0 in) dbh, 21 plots with snags > 50.0 cm (19.7 in) dbh, 22 plots with down wood >12.5 cm (4.9 in) diameter, and 19 plots with down wood > 50.0 cm (19.7 in) diameter.

The inventory numbers presented in the Integrated Summary and Synthesis and Management Implications sections are those based on plots with measurable snags or down wood. Summaries based on this subset of plots are compared to wildlife species data, assuming that dead-wood-associated species would occur on those parts of the landscape where dead wood was present. Much of the wildlife use data were collected in plots associated with some type of use of dead wood (e.g., nesting, roosting, or foraging activities).

The minimum size of snag measured was 25.4 cm (10.0 in) dbh and 2.0 m (6.6 ft) tall. The minimum size of down wood measured was 12.5 cm (4.9 in) diameter and 1.0 m (3.3 m) long.

INTEGRATED SUMMARY OF WILDLIFE DATA AND INVENTORY DATA FROM UNHARVESTED PLOTS

Wildlife and inventory data are statistically summarized using tolerance levels. For details on the statistical summarization see the Statistical Basis paper.

Snag dbh

Down wood diameter


Snag Density

Landscape distribution of snags in natural conditions

Data from unharvested plots can assist managers in setting objectives aimed at mimicking natural conditions (Figures WLCH_WCO_L.inv-14 and WLCH_WCO_L.inv-15). The following data describe distribution of snags on unharvested plots (n=41) in the WLCH_WCO_L vegetation condition.

Down wood percent cover

Landscape distribution of down wood in natural conditions

Data from unharvested plots can assist managers in setting objectives aimed at mimicking natural conditions (Figure WLCH_WCO_L.inv-16 and WLCH_WCO_L.inv-17). The following data describe distribution of down wood > 12.5 cm (4.9 in) in diameter on unharvested plots (n=22) in the WLCH_WCO_L vegetation condition.

ANCILLARY INFORMATION ON WILDLIFE SPECIES USE OF DECAYED WOOD ELEMENTS

See Ancillary Data Methods for caveats and information on how these data are weighted, summarized, and displayed.

Tree height

In Westside Lowlands Conifer-Hardwood Forest, nest tree height in this habitat type ranged from 2 m (7 ft) for the downy woodpecker (DOWO) to 34 m (112 ft) for pileated woodpecker (PIWO) (Figure WLCH.sp-11). The data for DOWO were from just 1 nest. Five of 19 species used nests les than 10 m (33 ft).The six smallest mean nest tree heights were from studies done in clearcuts or, in the case of the DOWO, in an alder forest. Heights of trees used for foraging ranged from 15 m (49 ft) for the pileated woodpecker (PIWO) to 23 m (75 ft) for the northern flicker (NOFL). Den tree height was recorded for the northern flying squirrel (NFSQ) at 21 m (69 ft). Roost tree heights ranged from 24 m (79 ft) for big brown bat (BBBA) to 44 m (144 ft) for fringed myotis (FRMY). Roosting tree heights were recorded primarily for bat species, except for the PIWO, and were among the highest trees for all the species and their uses. Ormsbee (1996a&b) and Ormsbee and McComb (1998) noted that roost snags of long-legged myotis were as tall or taller than the surrounding canopy, and that snag height was the most important selection criteria for roost sites. Waldien (1998) and Waldien et al. (2000) found 45% of long-eared myotis roost snags extended into the upper canopy. Long-eared myotis also used stumps as roosts (Table WLCH.sp-12). Of the stumps that were used, they were taller on the downhill side than the available random stumps (Waldien 1998, Waldien et al. 2000).

Hartwig (1999) found that trees with cavities excavated by pileated woodpeckers in southeastern Vancouver Island were taller than the available trees (p<0.01). Pileated woodpeckers on the Olympic Peninsula selected for nest and roost trees that were > 27.5 m (90 ft) tall and against nest and roost trees < 17.5 m (57 ft) tall (p<.001) (Aubry and Raley 2002). Bate (1995) found the abundance of hairy woodpeckers and red-breasted sapsuckers in the central Oregon Cascades increased with increasing density of hard, tall (> 6.1 m (20 ft)) snags across the landscape.

Tree species

In Westside Lowlands Conifer-Hardwood Forest, Douglas-fir had the most species of wildlife using it (18), being mostly of primary use (Figure WLCH.sp-12). Eleven wildlife species used western hemlock, which was a primary tree species of use by about half of those wildlife species. The only studies where Douglas-fir was not used by wildlife species in this habitat type either had a sample size of 1 (Zarnowitz 1982, Zarnowitz and Manuwal 1985), lumped all conifer species together (Carey et al 1997), or was limited to primarily hardwood stands (Chambers et al 1997). However, very few studies provided selection data and this primary use of Douglas-fir may reflect the availability of this tree species rather than a selection for it; for example, Lundquist (1988) found that cavity nesting birds in the southern Washington Cascades preferred nesting in white pine snags. However, white pine snags were rare and only 13% of the nests were found in this species; Douglas-fir contained 34% of the nests and western hemlock contained 28%. Similarly, Aubry & Raley (2002) noted that while pileated woodpeckers on the Olympic Peninsua used western hemlock most frequently for roosting and nesting, based on availability woodpeckers actually selected against western hemlock and selected Pacific silver fir for nest trees and western redcedar for roost trees. Aubry and Raley (2002) believe these tree species were selected because of their susceptibility to infection by heartwood decay organisms, which create suitable cavities for pileated woodpecker use.

Other tree species used less frequently, and mainly of secondary and lesser use, were western redcedar, grand fir, alder, bigleaf maple, and unidentified cedar species. In the southern Washington Cascades, Lundquist (1988) found that the abundance of Vaux’s swifts were positively correlated with large (>100 cm (39 in) dbh) live Douglas fir and western redcedar.

All species for which roosting data were collected used Douglas fir as their primary tree species. Long-eared myotis also used western hemlock as the primary species to roost in, and pileated woodpeckers on the Olympic Peninsula selected western red-cedar as the primary roost tree species ( Aubry and Raley 2002). Roosting pileated woodpecker used seven different tree species, the largest variety among all the wildlife species in this habitat type. This data set comprised two studies, one in the Oregon Coast Range and the other on the Olympic Peninsula, where tree species composition differs between the two areas. Nesting pileated woodpeckers used five different tree species, while the remaining wildlife species used from one to four species.

Tree mortality condition

With the exception of the northern spotted owl, the studies that recorded mortality conditions of trees used by birds indicated that either half or the majority of the trees used were dead (Table WLCH.sp-3). Of these records, all but three pileated woodpecker records (Mellen 1987, Aubry and Raley 2002) indicated the majority of dead trees was equal to or exceeded 67%. Pileated woodpecker records that showed less than 67% dead tree use were either from studies on the Olympic Peninsula (Aubry and Raley 2002) or were for roosting use (Mellen 1987, Aubry and Raley 2002). On the Olympic Peninsula, although half of the nests and roosts were found in decadent live trees, they were much less available than snags (Aubry and Raley 2002); they also noted that the proportion of dead vs. live trees used varied by tree species. For example, 65% of the western hemlock used by nesting pileated woodpecker were live, whereas 86% of the Pacific silver fir used were dead. Conversely, 68% of the western hemlock and 86% of the Pacific silver fir used by roosting pileated woodpeckers were dead, whereas 77% of the western redcedar used were live trees. In southwest Washington, Lundquist and Manuwal (1990) found that hairy woodpeckers foraged primarily in snags, while chickadees foraged in live trees; brown creepers and red-breasted nuthatches foraged in both live and dead trees.Raley and Aubry (2006) found pileated woodpeckers foraging primarily on live trees.

For the two mammals where these data were recorded (American marten and northern flying squirrel), with the exception of two flying squirrel records (Carey et al 1997), live trees were primarily used. Feen (1997) however, noted that denning flying squirrels preferred dead trees, while dens in live trees were associated with deformities.

While snag creation is often used as a form of mitigation for creating snag habitat features in areas devoid of them, Schreiber (1987) noted that “green snags” or snags recently created from green trees will not provide for short-term nest requirements for cavity-nesting birds.

Hollow live trees and snags

The use of hollow trees by wildlife was only recorded in one study (Aubry and Raley 2002), which found pileated woodpecker using hollow trees for roosting; all of the roost cavities that were inspected in this study (n=20) were hollow chambers formed by heartwood decay organisms.

Snag decay

In Westside Lowlands Conifer-Hardwood Forest (Table WLCH.sp-5), Mannan et al. (198l) found that hairy woodpecker, northern flicker, and pileated woodpecker used snags of all decay stages for foraging, while Lundquist (1988) found various species of woodpeckers to forage on moderate and soft snags. Hartwig (1999) found that almost half of the snags used by foraging pileated woodpeckers had extensive decay.

Snag use for roosting was recorded for the pileated woodpecker, long-legged and long-eared myotis (LLMY and LEMY). LLMY used all three decay conditions and LEMY used hard and moderately decayed snags. For pileated woodpeckers on the Olympic Peninsula however, only 27% of their roost snags were hard (Aubry and Raley 2002).

The remaining use noted in this habitat type was nesting. There were five records for nesting hairy woodpecker, indicating their use of moderate and hard snags, but no use of soft snags. This was the same pattern for the three records of nesting red-breasted sapsucker. While several studies indicated northern flicker to use all decay stages (Mannan et al. 1980, Schreiber 1987, Schreiber and deCalesta 1992), most of the studies indicated flickers using mainly snags of moderate or soft decay. On the Olympic peninsula, 77% of the snags used by pileated woodpecker for nesting were hard, which contrasts with the condition of the roost trees they used.

In the Oregon Coast Range, Nelson (1988) found the density of chestnut-backed chickadees, brown creepers, northern flickers, and red-breasted sapsuckers to be positively correlated with the density of large class 2 snags (p<0.01 for all four species), the density of hairy woodpeckers, pileated woodpeckers and red-breasted nuthatches to be positively correlated with the density of large class 1 snags (p<0.001, p<0.001, and p<0.01, respectively) and large class 2 snags (p<0.01, p<0.001, and p<0.01, respectively). Bate (1995) found that hairy woodpecker and red-breasted sapsucker abundance increased in the central Oregon Cascade Range with increases is the amount of hard, snags > 6 m (20 ft) across the landscape. Rosenburg and Anthony (1993) found that the density of Townsend’s chipmunks increased linearly with an increasing density of large, hard snags (p=0.002). Erikson and West (2003) found density of snags in early stages of decay were positively associated (P=0.02) with bat activity.

Bark sloughing is an identifying feature in describing snag decay stages and several authors noted wildlife use of sloughing bark. Cross et al. (1996) noted that fringed myotis and long-legged myotis were mainly roosting in the crevices provided by the sloughing bark of snags. Aubry (2000) found a positive correlation (p<=0.002) between density of soft snags and abundance of ensatina and western red-backed salamanders. Sloughed bark that collected at the base of stumps and snags provided habitat for ensatinas (Aubry and Hall 1991).

Feen (1997) noted that northern flying squirrel dens in live trees were associated with deformities.

Tree top condition

Use of broken top trees was noted for twelve bird species and a bat species (Table WLCH.sp-6). All bird species used the broken top trees for nesting, and the pileated woodpecker and long eared myotis used broken top trees for roosting. Of the few studies that indicated the proportion of trees that had broken tops (Brett 1998, Waldien 1998, Waldien et al 2000, Aubry and Raley 2002), well over half of the trees in each of these studies had broken tops. Aubry and Raley (2002) noted that the densities of dead top live trees surrounding pileated woodpecker nest and roost trees was higher than at random sites.

Hershey et al. (1998) found that the density of broken-topped trees > 53 cm (21 in) dbh was significantly higher (p=0.001) at northern spotted owl nest sites than random sites.

Down wood length

For Westside Lowlands Conifer-Hardwood Forests, the mean down wood lengths ranged from 1.1 m (3.6 ft) nesting winter wren (WIWR) to 20 m (66 ft) for pieces where Oregon slender salamander were present (Figure WLCH.sp-13). Pileated woodpeckers used down wood averaging about 10 m (33 ft) long as foraging substrate. Two studies located long-eared myotis (LEMY) roosts in logs averaging 9 m (30 ft) in length.


Doyle (1990) found that the number of Townsend’s chipmunk captures was significantly correlated (p<0.0001) with the total length of decayed logs. Hartwig (1999) found that pileated woodpeckers foraged on logs that were larger in diameter (p=0.02) and length (p=0.03). Corkran et al. (1997) found that an increase in log length increased the odds of Oregon slender salamander occupancy on a plot.

Down wood species

In the Westside Lowlands Conifer-Hardwood Forest, down wood species data were available for four salamander species, pileated woodpeckers (PIWO), and long-eared myotis (LEMY) and (Figure WLCH.sp-14). All species used Douglas-fir as the primary species of down wood. The salamander studies looked at either species presence or nesting use of down wood. PIWO also foraged on western hemlock, western redcedar, and hardwood logs. LEMY also roosted in western hemlock and western redcedar logs.

Hollow down wood

Data are not available for this component in Westside Lowlands Conifer-Hardwood Forest Habitat Type.

Down wood decay

In Westside Lowlands Conifer-Hardwood Forest, down wood decay data were available for five salamander species and one genus of salamander, three small mammal species, and two bird species, one invertebrate, one insect, and bryophytes (Table WLCH.sp-10). Species used a variety of decay stages, and the down wood used by most species had some level of decay, however, woodpeckers foraged on sound logs and crickets were associated with sound logs.

Preference for a particular decay class was shown for the ensatina (Butts 1997, Butts and McComb 2000), which preferred decay classes 3 and 4 (moderate to well decayed). Bury and Corn (1988) found more clouded salamanders (n=76) in class 2 logs than expected (p<0.001), and found more Oregon slender salamanders (n=57) in class 4 and 5 logs than expected (p<0.05). Through logistic regression, Corkran et al (1997) found that a 1 m increase in total length of class 3 logs 55-75 cm (22-30 in) diameter across the landscape would increase the odds of occupancy by Oregon slender salamanders an average of 1.4 times; in addition, a 10% increase in bark cover increased odds of occupancy 1.9 times.

Gilbert and Allwine (1991) found that Pacific jumping mouse captures in the western Oregon Cascades were positively correlated with log decay classes 3 (p=0.01) and 4 (p=0.02). Abundance of the Pacific water shrew and of the Trowbridge's shrew was positively correlated with the number of decayed logs (class 4-5) and negatively correlated with class 1 and 2 logs (Corn et al 1988). Doyle (1990) found that the number of Townsend's chipmunk captures were positively correlated with the total length of decayed logs and number of snags (p<0.0001). Doyle (1987) found that total length of decayed logs (class 3-5) was significantly higher (p< 0.001) at sites were western red-backed voles were captured as compared to where they were not captured. Conversely, Garman and Cole (1999) found the capture density of western red-backed voles were positively associated with soft logs 10-30 cm (4-12 in) dbh (p<0.0001), hard logs > 30 cm (12 in), and hard logs> 50 cm (20 in) (p<0.0001).

Hershey et al. (1998) found that volume and basal area of decayed (class 4&5) down logs was significantly (p<0.001) higher at northern spotted owl nest sites (n=30) than at random sites.

McDade (2001) found a positive association between well-decayed wood and ensatina and snails. Conversely, crickets were positively correlated with sound down wood.

Rambo (2001) found that decayed logs had the richest bryophyte flora and supported the greatest percentage of bryophyte diversity of any substrate.


GENERAL WILDLIFE-HABITAT RELATIONS WITH WOOD DECAY ELEMENTS

The following tables provide lists of wildlife species that use specific wood elements in this vegetation condition. The lists are derived from the Wildlife-Habitat Relationships database of O’Neil et al. (2001). (These are likely to be more complete lists of species than those lists above, for which quantitative field data on snags and down wood were available.) This fuller set of wood elements provides for a wide array of wildlife species (Rose et al. 2001). Additionally, lists of wildlife species with wood decay elements in all forest or woodland wildlife habitats in Washington and Oregon can be found on the DecAID Repository Web Site.

LANDSCAPE-LEVEL DISTRIBUTION OF DECAYED WOOD ELEMENTS: Comparison of Current and Natural Conditions Based on Forest Inventory Data

Comparisons of natural conditions to the broader landscape condition provide insight into the cumulative effects of forest management and other activities on the amount and distribution of dead wood. The following data compare the 50% tolerance levels of dead wood on unharvested plots (natural conditions) and all plots (current landscape conditions) in the WLCH_WCO_L Vegetation Condition.

Snags > 25.4 cm (10.0 in) dbh are 25% more dense on unharvested plots (34.0 tph (13.7/acre) than in the entire landscape (27.2 tph (11.0/acre)) (Tables inv.3b and inv.7b). Snags are present on all of the unharvested area, whereas 10% of the total area lacks snags. 62% of the unharvested area supports > 30 snags/ha (12.1/acre), whereas 37% of the landscape as a whole supports snags at this density (Figures WLCH_WCO_L.inv-14 and WLCH_WCO_L.inv-18).

Large snags (> 50.0 cm (19.7 in) dbh) are 57% more dense on unharvested plots (26.2 tph (10.6/acre)) than in the overall landscape (16.7 tph (6.7/acre)) (Tables inv.4b and inv.8b). 5% of the unharvested area does not support any large snags, whereas 19% of the total landscape lacks large snags. 57% of the unharvested area supports >20 snags/ha (8.1/acre), whereas 28% of the total landscape supports this density of large snags (Figure WLCH_WCO_L.inv-15 and WLCH_WCO_L.inv-19).

Cover of down wood > 12.5 cm (4.9 in) diameter is similar on unharvested plots (8.7%) and in the entire landscape (9.2%) (Tables inv.5b and inv.9b). Down wood is present on all area of this vegetation condition. 58% of the unharvested area and 57% of the landscape as a whole contains >8% cover of down wood (Figures WLCH_WCO_L.inv-16 and WLCH_WCO_L.inv-20).

Cover of down wood > 50.0 cm (19.7 in) diameter on unharvested plots (5.3%) is slightly less than on the entire landscape (5.6%) (Tables inv.6b and inv.10b). 14% of the unharvested area does not support any large down wood, and 11% of the total landscape lacks large down wood. 66% of the unharvested area supports > 4% cover of large down wood, and 60% of the landscape as a whole supports this much cover (Figures WLCH_WCO_L.inv-17 and WLCH_WCO_L.inv-21).

RELATIONSHIPS OF FUNGI TO DECAYED WOOD ELEMENTS

Fungi are important to proper ecosystem functioning by providing ectomycorrhizal associations and nutrient cycling as well as being useful as food sources and medicinal sources. Therefore it is desirable to maintain the most fungal diversity possible. Since fungal species vary in resource requirements, the following guidelines for maintaining down wood will assist in providing habitat for a diversity of fungi:
For additional information see the DecAID Fungi Components Document.

CONSIDERATIONS FOR STAND DYNAMICS

General trends of stand dynamics in this habitat type

A discussion of stand dynamics for the Westside Lowlands Conifer-Hardwood Forest Habitat Type is found in Chappell et al. (2001). The following discussion is summarized from this publication.

While fire is the most common natural disturbance in this Habitat Type, wind is also a major disturbance in some areas. Minor, small-scale windthrow events are common, whereas major, large-scale events occur about once ever few decades. In some areas, landslides are another cause of natural disturbance.

After a natural, stand-replacing disturbance, grasses and forbs dominate the site for a few years before shrubs and young trees begin to dominate the site. Shrubs may dominate a site for long periods of time if trees are not able to establish soon after disturbance. This may especially be the case for shrubs that resprout. Douglas-fir is usually dominant after fire in all but the wettest areas. Due to its fire resistance, Douglas-fir often survives moderate-severity fires. After the tree canopy closes the understory becomes sparse. At about 60 to 100 years suppression mortality occurs and the stand opens up enough to allow understory development. Shade-tolerant trees such as western hemlock or western redcedar become established in the understory, gradually creating a multi-layered stand by age 200 to 400 years. Western hemlock increases in the canopy as the stand continues to develop. Douglas-fir would continue to dominate the canopy for 800 to 1,000 years in the absence of disturbance. In all but the wettest areas, however, a disturbance usually occurs well before the stand reaches this age.

In the Westside Lowland Conifer-Hardwood Forest Habitat Type, abundance of snags and down wood generally peaks in the first 50 years after a fire or other disturbance and is least abundant at about 150 years post disturbance, and increases again after about 200 years (Spies et al. 1988).

List of Insects and Pathogens

For a list of insects and pathogens important in affecting habitat for dead wood-dependent wildlife species in this vegetation condition, see Tables WLCH_WCO.fid-1-6. The functional roles of each insect and pathogen are also displayed. Although other insects and pathogens not shown in these tables also occur in this vegetation condition, their influence upon wildlife habitat generally is of minor significance compared to the listed species. The included insects and pathogens were selected because they are known to create or assist formation of dead wood or other structural habitat elements, or to influence the abundance or long-term sustainability of wildlife habitat. Many exert very significant effects, or should be considered when making management decisions.

Pathogens

In this section, inventory data from unharvested plots are compared to data from harvested plots, not to all plots as in other sections in DecAID. This comparison is displayed so that apparent differences in amount of infection by various pathogens as a result of stand manipulation can be seen. The "unharvested" plots include only those where there was no evidence of tree cutting of any kind and there was no evidence of roads. The "harvested" plots included the remainder of the plots. For more details on how "harvest" was determined on a plot see the Harvested v.s. Unharvested document.

The primary disease causing small-scale disturbances in Westside Lowland Conifer-Hardwood Forest, Washington Coast, is laminated root rot (Phellinus weirii) in Douglas-fir. Inventory data on federal lands indicate 5% and 8% of the plots have infected trees in unharvested and harvested areas, respectively (Table WLCH_WCO.fid-7). This is a disease “of the site” and the causal fungus can survive in large stumps and roots for as long as 50 years (Childs 1963, Hansen 1976, 1979). Trees of any age or size can become infected when they make root contact with a disease-infected root. These “root disease pockets” expand at a rate of about one foot a year, so the root disease pocket increases in size through time as long as there are host trees available. Additional host trees will began to grow within the root disease pocket but will be short-lived. Immune (e.g., hardwood species) or resistant (e.g., western redcedar) tree species will survive and grow within the root disease pockets. As the roots decay, trees will go down, some while still maintaining green foliage. Trees that die from the root disease, or from Douglas-fir beetles which are attracted to stressed, diseased trees, will not remain standing for long (i.e., selecting host trees as wildlife trees within a root disease pocket is not a good idea). These root disease pockets are a source of down wood if in older stands, but would not provide down wood of any appreciable size in younger plantations where host trees were planted within the root disease pocket. For additional information on laminated root rot see report by Thies and Sturrock (1995).

Mortality caused by armillaria root disease (Armillaria ostoyae) is most common in Douglas-fir plantations between the ages of 10 and 25 (Hadfield et al. 1986). Tree killing after the age of 25 is uncommon unless the trees are stressed (Hadfield et al. 1986). Although armillaria root disease can be readily found in this habitat, it is uncommon to find significant mortality of older trees as a result of this pathogen. Inventory data indicate 11 to 19% of the inventory plots had armillaria present (Table WLCH_WCO.fid-7). This is probably a significant overestimate due to the tendency for crews to mark presence of saprophytic forms of armillaria which are commonly seen and which do not behave as a root disease.

There are numerous fungal species which cause decay in conifers and hardwoods in WLCH habitat, and it is common to find heart rot in older trees. The inventory data on federal lands only has specific codes for three of the species, which the remainder being lumped into an “other” category.

Red ring rot (Phellinus pini) can be found in most any species of tree, and is especially common in larger, older trees. Advanced decay caused by this fungus is beneficial to some cavity nesting species, but does not cause trees to become hollow.

Annosus butt rot is quite common in western hemlocks. The amount of butt decay is small in trees younger than 120 years, but increases as trees age beyond that, with significant decay in trees over 150 years old. The probability of breakage along the decayed part of the stem increases through time. Decay caused by this fungus, Heterobasidion annosum, will result in a hollow butt and could be very useful to some cavity nesters or animals which make use of hollow logs. Inventory data (Table WLCH_WCO.fid-8) indicate levels of the pathogen as lower than what is most likely present due to the fact that annosus is difficult to detect in live western hemlocks, and, even in broken trees, it is difficult to find the positive indicator (fruiting body) of this fungus. When decay is detected and the positive indicator cannot be found, the tree would be coded as unidentified root disease or unidentified heart rot.

Indian paint fungus, Echinodontium tinctorium, causes heartrot primarily in true firs and hemlocks. The result is hollow stems which are very useful to cavity nesters and other organisms which make use of hollow logs. No plots were coded as having Indian paint fungus in this habitat (Table WLCH_WCO.fid-8, although it certainly does occur in this habitat. It is possible that some of the sporophores (conks or fruiting bodies) of this fungus were coded as unidentified decay fungi.

There are many species of decay fungi, and there are no specific codes for major ones commonly found on western redcedar. The inventory data indicate the 16% of the trees > 50 cm (20 in)dbh had heart rot (Table WLCH_WCO.fid-8).

Dwarf mistletoes can be important in providing nesting or roosting structures for some species of birds and small mammals. Western hemlock dwarf mistletoe (Arceuthobium tsugense subsp. tsugense) is the most important mistletoe for wildlife use in this habitat. Inventory data indicate 54% and 13% of unharvested and harvested plots, respectively, have some level of hemlock dwarf mistletoe infection (Table WLCH_WCO.fid-9). Douglas-fir dwarf mistletoe does not occur in the WLCH_WCO habitat.

Insects

Overview: Insects most commonly create small-scale disturbances characterized by dead tops in larger trees, individual dead trees of any size, and small to medium patches of dead medium-size and larger trees. Bark beetles cause most of the insect-related mortality, interacting very closely with other natural disturbances, especially wind, root disease, and drought.

Common associations and effects: Douglas-fir beetle is the predominant insect affecting stand dynamics in the Westside Lowland Conifer-Hardwood Forest wildlife habitat type. It typically creates scattered canopy gaps ranging from less than 1/4 to 1/2 acres in size, as well as scattered patches of forest with significantly reduced crown cover where upwards of 500 trees over areas up to 30 acres in size are killed in mature and old growth Douglas-fir dominated forests, resulting in the release of understory trees and shrubs. Regional outbreaks of the Douglas-fir beetle in western Oregon and Washington usually are linked to large-scale wind or winter storm events that produce large quantities of fresh windthrow or breakage of medium-size and larger Douglas-fir trees over fairly broad geographical areas. More localized increases in Douglas-fir beetle activity are sometimes associated with local events of wind or heavy snow and silvicultural activities that place large quantities of fresh suitable host material on the ground during periods of beetle flight and development. Fire-injured Douglas-fir are susceptible to attack by Douglas-fir beetle and wood borers. Fir engraver frequently kills true fir trees infected with root pathogens, during droughty periods, or following disturbances that cause tree injury or stress. Pines may be occasionally killed by mountain pine beetle (pole-size and larger trees) or pine engraver beetles (sapling/pole-size trees), and spruce beetle may occasionally attack and kill mature spruce trees, but activity levels of these beetles in this wildlife habitat type is usually very low. Cedars, western hemlock, Pacific yew, and hardwoods are not commonly killed by insects.

A variety of small bark beetles, including the fir engraver, Douglas-fir engraver, and Douglas-fir pole beetle, sometimes kill the tops of mature trees stressed by drought, root disease infection, or injury.

Balsam woolly adelgid, a non-native sucking insect that infests the boles and branches of true firs, often causes reduced tree growth and mortality of sapling-size and larger trees, and diminished reproduction of reproductive-age trees. Permanent infestations occur in the Pacific Northwest throughout much of the range of its true fir hosts, which display varying degrees of susceptibility to the adelgid (Mitchell and Buffam 2001). Susceptibility varies not only among species of true fir, but also among the range of growing sites occupied by a single true fir species.

Prior to the 1960’s, sporadic outbreaks of the western hemlock looper sometimes caused extensive mortality in old growth hemlock forests growing in coastal areas. These outbreaks tended to be medium to large-scale disturbances, covering up to 50,000 acres and resulting in significant amounts of tree mortality (Furniss and Carolin 1977). The absence of similar outbreaks in more recent times may be due to the cumulative effects of logging, which has significantly altered the abundance and extent of habitat conducive to hemlock looper outbreaks. Hemlock looper defoliation may cause tree mortality, growth reduction, and topkill of western hemlock and associated tree species such as western redcedar, Sitka spruce, Pacific silver fir, and Douglas-fir, although western hemlock usually is the most severely affected species.

Important long-term dynamics: Balsam woolly adelgid is slowly eliminating grand fir from low-elevation (< 330 m (1,080 ft)) areas of the Puget Sound trough and along coastal streams (Mitchell and Buffam 2001).

Fire

The following discussion on fire is from Chappell et al. (2001):

Fire is the major natural disturbance. Natural fire-return levels generally range from about 100 years or less in the driest areas to several hundred years (Agee 1993). Mean fire-return level for the western hemlock zone as a whole is 250 years, but may vary greatly. Most natural fires are associated with occasional extreme weather conditions (Agee 1993). Fires are typically high-severity, with few trees surviving. However, low- and moderate-severity fires that leave partial to complete live canopies are not uncommon, especially in drier climatic areas.

ECOLOGICAL FUNCTIONS AND PROCESSES OF DECAYED WOOD ELEMENTS

In this section information is presented on the array of ecological functions of wildlife species that in turn are supported by wood decay elements in this vegetation condition. That is, wildlife species play specific ecological roles called “key ecological functions” (KEFs) in their ecosystems (also see the Glossary). The extent to which wildlife species are associated with wood decay elements, and the degree to which such elements are provided, helps support an array of KEFs performed by those species. In general, a community that is more functionally diverse and functionally redundant (more species playing each ecological role or that pertain to each KEF category), will likely be more resilient and resistant to disturbance perturbations than will a community that is less functionally diverse or redundant (Marcot and Vander Heyden 2001). Information presented here on KEFs of wildlife is taken from the WHR database of O’Neil et al. (2001).

List of Key Ecological Functions (KEFs) of Wildlife Species Associated With Wood Decay Elements

The following tables list species’ ecological functions that are afforded by wood decay elements. The manager can use this information to determine which ecological functions might be present in a broad area if such wood decay elements were to be provided. Functional redundancy (number of wildlife species) by KEF category of species found in this vegetation condition can be viewed in Table WLCH_L.kef-1.

Lists of KEF categories for wildlife species found in the vegetation condition that use the following specific wood elements can be viewed by clicking the links below:

Additionally, lists of wildlife species with key ecological functions affecting wood decay elements in all forest or woodland wildlife habitats in Washington and Oregon can be found on the DecAID Repository Web Site.

Ecosystem Processes Related to Wood Decay

Wood decay elements provide resources and substrates for many organisms and biological processes that perform vital ecological roles that contribute to overall ecosystem health, soil productivity, and growth of desired plant species. Such processes are part of natural, healthy ecosystems but few studies have quantified them.

Down wood serves as moisture reservoirs, sources of beneficial mychorrizal fungi, nurse logs for many tree and shrub species, and sources of soil organic matter, and also provides other beneficial ecosystem processes. Disturbance events such as fire and blowdown begin a process of releasing useful nutrients stored in woody structures as wood decay proceeds. Decaying tree roots also can contribute to soil nitrogen and soil water-holding capacity, and reduce soil compaction.

Many forest-dwelling mammals disperse spores of beneficial mychorrizal.

Removing woody structures can have short-term benefits to planted seedling growth but longer-term adverse effects on overall forest productivity.

Wildfire can greatly increase the net amount of down wood in a stand, whereas clear-cutting tends to decrease it. Human safety can be a major concern with wildfire or prescribed fires and may override the need to retain wood decay elements in fire-prone forests near human habitations.

For more details see the Ecosystem Processes Related to Wood Decay Document.