Decayed Wood Advisor

Print Close

 

Caveats and Cautions- A MUST READ!

Table of Contents

LITERATURE CITED


INTRODUCTION


This document discusses important caveats and cautions associated with the data in DecAID. All data are imperfect and thus have limitations. Even with imperfect data, DecAID is still a compilation of the best available data relating to snags and down wood. The cautions and limitations associated with the underlying data need to be kept in mind by the user when applying DecAID to projects. Used properly, DecAID can help guide management of dead wood to meet management goals. The meta-analysis approach of DecAID, combining data from across multiple studies, and the comparison to forest inventory data, strengthens the evidence over attempting to apply data from studies individually.

CAVEATS AND CAUTIONS IN USING THE WILDIFE SUMMARIES


Interpreting the cumulative species curves

The cumulative species curves cannot be expected to accurately depict specific sites or individual stands. The curves represent a summary of field studies and, in general, what might be expected across broad areas such as subwatersheds or larger areas. Wildlife species richness might not increase as orderly as the curves suggest, on any given site or within any given area, with increasing size or amounts of wildlife trees or down wood as the curves suggest, but this is largely a scale issue. Species occurrence can vary substantially among areas with different spatial patterns of snags and down wood, surrounding landscape conditions, site histories, and site conditions including presence of tree species and specific wood rot patterns. At the site level, tree species-specific rot patterns greatly influence wildlife species occurrence.

The cumulative species curves represent observed patterns of species' use and selection of dead wood size and amounts, insofar as such sizes and amounts were present in individual study areas. Ideally, the best evidence of what to provide for dead wood-using species would come from a large number of replicated, controlled, manipulative experiments that would specifically vary only dead wood sizes and amounts holding all other environmental factors constant, with random allocation of treatment sites and inclusion unaltered control sites, and accurately following the resulting wildlife population size, trend, structure, and persistence and individual fitness (vitality of offspring). Such perfect field experiments have not been conducted, and likely never will -- they are far too difficult, and may be impossible, to implement in the real world, especially across the full array of pertinent species.

Therefore, in the absence of such definitive experiments, it is our fundamental assumption that patterns of species' use and selection of dead wood size and amounts represent behaviors that have adaptive advantage for the species and that serve to bolster individual fitness. That is, organisms will behave in their environments to maximize their individual fitness and, as a result, the persistence and viability of populations. This crucial assumption is essentially based on basic tenets of ethology, sociobiology, population ecology, and evolutionary ecology. It underlies our suggesting the use of the cumulative species curves to help guide dead wood management, even if the curves are based on post hoc observational studies in selected sites and not necessarily on more rigorously controlled manipulative experiments per se.

Also, in the absence of such definitive experiments, we combined data from observational studies. In meta-analyses, the strength of evidence from observational studies increases with greater sample size -- that is, with more studies and a wider array of study areas, environmental situations, site histories, number of species, and specific study plots. This is why we have "combined information" across all the study results.

We encourage future studies to test the underlying assumption that wildlife select dead wood size and amounts based on behaviors that have adaptive advantage and significance. Results of these studies will continue to be incorporated into the DecAID database.

Cautions need to be aimed at interpreting and using data from studies that spanned a variety of vegetation structures, treatments, or seral conditions. In a sense, each point for each species from such a study is a probability cloud itself representing variation among such conditions. In the cumulative species curves, regressions among different points may represent spurious relations. This is a major concern in meta-analysis methods of combining data from different studies. The best thing to do is be aware of the conditions in each reported study and interpret results accordingly. This is why, in part, we designed DecAID to allow the user to “drill down” through the species summary information to get at the underlying study data and citations.

What this means for the user:
  • DecAID is not a simulation model, it is a compilation of the best available, though imperfect, empirical data on wildlife relationships with dead and decaying wood.
  • DecAID is not a population viability analysis model. There is more to viability of populations than the dead wood habitat component. DecAID can help managers decide how much, and what sizes, of dead wood to provide for this part of species’ habitat needs.
  • It is up to the user to understand the underlying data and determine if the data are appropriate to their local situation and conditions. This is why DecAID was designed to allow the user to drill down into increasingly finer details of the data. The user is urged to carefully read the Summary Narratives which discuss and provide red flags, where appropriate, for the basis of underlying studies.

Incomplete species lists

No empirical study has provided data on all snag-associated species (primary and secondary users) in any given vegetation condition. Thus, the data tables and cumulative species curves in the wildlife component of DecAID provide only a partial insight into the full assemblage of wildlife (and fungi, cryptogams, and vascular plants, as well as invertebrates) that are associated with snags and down wood.

What this means for the user:
  • Don't rely just on the cumulative species curves for information on wildlife relationships with dead wood. Not all data were available in a format compatible with the cumulative species curves. Additional data are available in DecAID under the sections Introduction to Available Data and Ancillary Data. Complete lists of wildlife species using dead wood in each vegetation conditions are available through links in the section General Wildlife-Habitat Relations With Wood Decay Elements. Lists of species associated with decayed wood and the role they play in ecosystem functions are available in the section Ecological Functions and Processes of Decayed Wood Elements.

Smoothing the curves

The cumulative species curves jitter and bounce from vagaries in everything from sampling design to differences in specific vegetation conditions at study sites. One could smooth the curves by using some smoothing function. We tested this but chose not to do this, so that the empirical data could be preserved in the curves, showing the conditions for each individual species or species group; smoothed curves would not provide species-specific information. Also, many curves for vegetation conditions and structures not shown here are relatively data poor, and smoothing algorithms would not be appropriate in such cases. Also, as noted above, the cumulative species curves depict relations of species to dead wood and serve to summarize data across studies, but are not intended to depict mathematical functions per se, so curve-smoothing would not be appropriate.

Consider all species' uses and habitat conditions

It is vital that all uses of wildlife trees or down wood, such as for breeding, feeding, and roosting, for given species be considered simultaneously when assessing impacts on species or when devising management guidelines in a particular wildlife habitat and structural condition. Considering only one type of wildlife use, especially a use that correlates with the smallest or fewest snags or down wood pieces, can prove insufficient to meet population needs. DecAID provides some data on these other uses, as available from studies, but such data are not consistently or thoroughly reported from all studies.

Also, DecAID and this wildlife component address terrestrial, upland conditions, and do not specifically address riparian, aquatic, and wetland conditions separately although such conditions might be part of a broader landscape in which some wildlife studies were conducted.

What this means for the user:
  • The user should consider all uses of wildlife trees or down wood by wildlife, including for breeding, feeding, and resting.
  • DecAID does not address dead wood in riparian, aquatic, and wetland environments. If desired, the user should seek information and guidance elsewhere on management of dead wood in such environments.

Consider all decadence elements

Some species select for decadence conditions of hollow trees, hollow down logs, and dead parts of live trees. Such elements should be considered in addition to the traditional focus on (solid) snags and down wood. We have documented available empirical study data and wildlife-habitat relationships information on wildlife use of such decadence elements. In many cases, such substrates may be rare enough to warrant complete protection where found, in coordination with health and safety standards.

In some studies, the reported "snag" densities included both fully dead trees (snags) and live trees with dead parts (e.g., trees with heart rot, dead tops, etc.). It was not possible to separate these out, and in such cases we included all reported densities. When the studies did report these separately, we included only the snag portion. The user may want to drill down to read about the individual studies.

What this means for the user:
  • Do not overlook the wildlife value of wood decay elements other than snags and down wood, such as of hollow trees and logs, and dead parts of live trees. Although studies are needed to quantify wildlife relations with such additional wood decay elements, evidence is clear that such elements provide critical substrates for many species.
  • As needed, the user can drill down into the data to determine which studies may have included live trees with dead parts along with fully dead trees (snags). The user needs to understand the underlying data by drilling down to the information on individual studies to assess if partially dead trees are also included in density estimates.

Consider use vs. selection

Data based on patterns of selection (use compared with availability) should be interpreted differently than data based on occurrence or just use with no comparison to availability. Selection, as for particular sizes or densities of snags or down wood, when demonstrated, provides far greater evidence of how wood decay elements provide for wildlife species. Still, use data can be applied to develop helpful guidelines for what to provide while selection studies can be implemented to test the guidelines. Whether studies reported use or selection is indicated in the underlying data for the graphs in DecAID.

We advocate the use of the terms "use" (an organism is shown to at least associate with a particular decay element such as snags) and "select" or "selection" (an organism is shown to statistically use a decay element in greater frequency than it is available in the environment). We strongly urge to avoid the terms "need," "require" (or "requirement"), and "depend" ("dependent on," etc.), as these are terms and concepts that, at best, would beg for intensive field manipulation experiments on the physiological and perhaps sociobiological characteristics of organisms and populations; such studies are generally lacking for the array of species and vegetation conditions addressed in DecAID, and in many cases may not even be possible to conduct. These terms to be avoided are largely not statistically verifiable.

Some of the 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.

What this means for the user:
  • It is up to the user to understand the underlying data and the strength of the data points. Selection is determined by looking at underlying data 2 levels below the cumulative species curves - a value or comment in the column p-value will help the user assess the significance of the results.
  • Don't rely just on the cumulative species curves for information on wildlife relationships with dead wood. Not all data were available in a format compatible with the cumulative species curves. Additional data are available in DecAID under the sections Introduction to Available Data.

Understand the basis of the underlying studies

Understanding the range over which wildlife tree or down wood sizes or densities were studied is also important. In some cases, there may appear to be no correlation or selection because more than adequate sizes or densities of wildlife trees or down wood were already present in the study area, and the wildlife response had already leveled off.

Even if selection is demonstrated, some species may still be able to persist at lower population levels if their selected sizes or densities of wildlife trees or down wood were not available. However, it is largely impossible to predict this for most species. Thus, providing for snag and down wood sizes or densities according to the empirical studies likely would provide a greater likelihood of providing for associated wildlife.

Data on stand averages of wildlife tree or down wood density may or may not represent unmanaged conditions. Often, we could not determine this from the literature, so great care needs to be exercised when interpreting such data.

What this means for the user:
  • It is up to the user to understand the underlying data and determine which data are appropriate to their local situation and conditions.
  • Do not expect that providing for lower tolerance levels of size and amount of snags and down wood will provide for associated wildlife at the same level of confidence and population sizes or densities than would providing for higher tolerance levels.

Population response

The ultimate, and really the authentic, measure of the effectiveness of snag and down wood management guidelines is how well they provide for fit individuals and viable populations. Fitness is the reproductive vitality of offspring, and viability is the persistence of well-distributed populations over the long term. Few if any studies we reviewed for use in DecAID truly measured fitness and viability, and most studies did not report demographics (including population density and trend) in relation to dead wood.

Under some conditions, populations of snag- and down wood-associated species may be limited by factors other than snag density, size, and condition. For example, based on a simulation model, Raphael (1983) suggested that, in Sierra Nevada mixed conifer forests, secondary cavity-nesting birds may be limited by territoriality rather than cavity abundance when snags are sufficiently numerous to provide nesting habitat for primary cavity-excavating species, at least for a time. His model then suggests that, in a burned forest with no further recruitment of snags, numbers of both primary cavity excavator species and secondary cavity-nesting species are limited by snag abundance as snag numbers decline beyond about year 20.

What this means for the user:
  • DecAID is not a population viability analysis or population demography model. There is more to viability of populations than the dead wood habitat component. DecAID can only help managers decide how much dead wood to provide for this part of a species habitat needs.

Remember hardwoods too

Available studies inadequately represent how live and dead hardwood trees provide for natural or excavated cavities for many species, such as for Downy Woodpecker (Picoides pubescens) and Acorn Woodpecker (Melanerpes formicivorous) in Westside Lowland Conifer/Hardwood Forests. Hardwoods would have to be added to the consideration, particularly in forest types in which aspen, oaks, maples, and other broadleaf or hardwood species naturally occur or dominate.

What this means for the user:
  • Do not overlook the value of live and dead hardwood trees, and partially dead trees of all kinds, which are under-represented in the literature and DecAID.
  • Some studies may have included hardwoods in their data. The user can inspect the ancillary data to determine which studies may have addressed hardwoods.

What snag and down wood use data represent

Most of the data on snag density from the wildlife literature were recorded at nest, roost, or den sites. The size of vegetation sample plots varied among studies, although plots usually were < 0.1 ha (1/4 acre). Snag densities in these “used” plots often were higher than snag densities in random plots in the surrounding stand. This difference might indicate that wildlife use or perhaps select for clumps of snags; this may be clarified by inspecting snag densities from inventory data from unharvested plots. Conversely, where inventory and wildlife data suggest similar values of snag or down wood sizes and amounts, as is often the case for some species, does not necessarily constitute evidence that wildlife do not select for high density clumps.

In the wildlife studies, if the vegetation sample plots were placed among snags occurring in locally dense clumps < 1 ha in size, then extrapolating the snag densities from such sample plots to a per-ha basis may yield very high snag densities that may not be appropriately interpreted as stand-wide averages or management objectives. The user of DecAID can drill down to view information on the underlying studies and carefully read the Summary Narratives to determine if the data suggest stand averages or local clump snag densities.

Studies examining wildlife use of snag densities inconsistently reported size (dbh) classes of snags. We combined data from studies into two general size classes of > 25 cm dbh and > 50 cm dbh because these were the most commonly reported breaks in the literature; these size classes are displayed separately in the cumulative species curves in DecAID. We excluded some data from the curves if the reported snag size (dbh) class was too different from those we used in DecAID. Often, the same study did not report data on the two snag size classes we used in DecAID. As a result, there may be different sets of studies underlying data points, so that a data point on the curve for densities of snags > 50 cm dbh may actually show a higher snag density for the same species than the > 25 cm dbh curve. This is possible, even though > 50 cm should be a subset of > 25 cm, because the data on the 2 different curves may have come from 2 different studies.

Down wood was measured by a variety of methods among studies. The method of measurement will affect the reported diameter of the piece of wood and which down wood is included in estimates of amount of down wood. Diameter of down wood pieces is typically measured at either large end, small end, or point of intersection with a transect. Due to log taper, the point of diameter measurement influences reported down wood diameters and whether or not a log was included in the tally for amount of down wood. Authors did not always report where diameter measurements were taken, however, if available that information is provided in DecAID underlying data.

In general, placement of wildlife study plots – particularly if placed at wildlife-use locations, such as nests -- can tell you more about snag clumps than about stand- or landscape-wide average of snag and down wood sizes and amounts.

What this means for the user:
  • The user needs to read the Summary Narratives carefully. Any potential problems with the data associated with plot size, differences in size of dead wood measured, or other potential problems are mentioned in association with the data points on the cumulative species curves.
  • Again, we also urge the user of DecAID to do their best to understand the underlying data. This is why we provide the means to view all underlying studies and their data.

Sources of variation

As with summarizing results from any empirical wildlife or ecological study, the variation in the data underlying the statistical summaries in DecAID may arise from several sources. Variability in the data may be from variation among individuals in a wildlife population, and we depicted this source of variability with tolerance intervals.

However, some of the variability may also derive from variations in habitat conditions across study sites or across entire research studies, that is, across both geographic space and time periods. The user is encouraged to drill down to the detailed information on each study and decide for themselves which data points may be relevant to their situation. The user can view the basis of the sampling for each study, including sampling plot size and number, as well as a summary of the vegetation conditions in each study.

There may be problems associated with combining wildlife plot data sampled across a wide range of plot sizes used in the various wildlife studies. We extrapolated the reported data to a per-hectare basis, so if the initial plot size was small, this might inflate the per-hectare calculations of snag or down wood density. Moreover, since many of the wildlife studies centered plots on snags or trees used by animals (as with nest trees, roost trees, etc.), this too could serve to inflate estimates of snag or tree density when expanded to a per-hectare basis, particularly if the used snag or tree was part of a clump of trees. However, we did simple linear regressions on plot size and our extrapolated snag density, within wildlife habitat types, and results were not statistically significant (various tests, p>0.05). We also regressed plot size against standard error of snag and down wood density, and found little correlation (p>>0.05). This is encouraging, in that it suggests a relatively low bias from variation in plot size, that is, that plot size did not systematically influence density estimates when viewed across studies, even if plot size might have biased some estimates within individual studies.

Another source of variation may derive from measurement error, including combination of different levels of precision from different studies. We cannot partition out this source of variation, as few if any studies tested and reported measurement error and level of precision of their measurements (number of significant figures).

The upshot is that some of the outlier values, particularly shown on the 30% and 80% tolerance level curves, may result from a combination of these various sources of variability. Given the information presented (or rather, not presented) in most studies, we cannot partition out the relative contribution of each source of variability. Thus, the information presented in the cumulative species curves could be viewed and treated as testable management hypotheses that, if used to guide forest management, might also be empirically tested and updated. In the absence of such tests, however, the data summaries presented in DecAID nonetheless are of value for guiding management.

Another wrinkle to be aware of in the wildlife study data is that plot size may influence variation in the estimates of large snag density (and possibly also large down wood percent cover). Plot size often varied among wildlife studies; the user can view the actual plot size used in each study by drilling down to the data summaries. The studies with the smallest size plots -- especially < 0.1 ha -- may have higher variations in snag density (and possibly down wood percent cover) among those plots. Mostly, this was true only for the larger diameter snags (and down wood), such as snags > 50 cm dbh. This is to be expected given the more scattered and scarce distribution of such larger wood decay elements. This does not invalidate use of these studies, and in fact we display the minimum snag or down wood diameter included in density and percent cover estimates in the data summaries of each study. But the user may wish to evaluate the larger size classes of snags and down wood separately based on this difference in variation.

What this means for the user:
  • There are some potential statistical problems when combining data in a meta-analysis. In DecAID we tested for some of the common problems and they appear to not be significant. The benefit of combining data to increase sample size and strengthen evidence of dead wood associations should outweigh the potential problems. DecAID is still a compilation of the best available, though imperfect, empirical data on wildlife relationships with dead and decaying wood.
  • The user needs to read the Summary Narratives carefully. Any potential problems with the data due to variability associated with differences in plots sizes or other potential problems are mentioned in association with the data points on the cumulative species curves.
  • Again, it is up to the user to understand the underlying data and determine which data are appropriate to their local situation and conditions.

Unknown bias in wildlife data due to uneven sampling

The wildlife studies, on which the wildlife portion of DecAID is based, were conducted in a variety of landscapes and site conditions. Typically, the studies (a) did not report how the general study areas and specific study sites were chosen relative to others, and (b) did not describe how the vegetation conditions within the general study areas and specific study sites differed from conditions within a broader area, especially within the wildlife habitat and vegetation condition classes used in DecAID. Thus, there is no way to know to what degree the study areas and sites varied from conditions generally present, and thus no way to gauge the bias in study area and site selection. In turn, this means there is no way to estimate the degree of bias in the selection of landscapes in the wildlife studies summarized in DecAID.

In general, this unknown bias is likely reduced when a greater number and extent of studies are conducted in a particular wildlife habitat and structural condition. Thus, the DecAID user may wish to drill down to the underlying data and evaluate whether the component studies either pertain to their locations or vegetation conditions of interest, or, alternately, if the number and breadth of studies may adequately capture the range of conditions within a wildlife habitat and structural condition so as to potentially reduce bias to an acceptable level.

Note that although the wildlife studies included in DecAID were not necessarily designed to sample the full range of conditions, the systematically-based inventory data collection (see below) was structured to sample a full or wide range of conditions.

Combining information helps widen the breadth of available conditions addressed. Organisms can use and select the amounts and sizes of snags and down wood, and the other attributes of these and other wood decay elements, only among those that are present in a particular area or landscape. We presumed that combining information across all available, multiple studies and study areas helps to ensure to some degree the range of available wood decay elements and their attributes. Combining information across studies -- particularly across studies conducted among a variety of conditions within a wildlife habitat and structural conditions -- may help ensure that the span of conditions is better represented, although the level of bias is still unknown.

What this means for the user:
  • There are potential problems of bias associated with any study, no matter how well designed. The benefit of the meta-analysis used in DecAID is that it increases sample sizes, and multiple studies strengthen evidence of dead wood associations.
  • DecAID is still a compilation of the best available, though imperfect, empirical data on wildlife relationships with dead and decaying wood.

CAVEATS AND CAUTIONS IN USING THE INVENTORY SUMMARIES


Under-representation of post-fire conditions

The inventory data in most cases do not represent recent post-fire conditions very well because the plots sample conditions arising from a variety of disturbances, including but not limited to fire. The sample plots of older forests might represent at least some post-fire conditions; however, young stands originating after recent wildfire are not well represented because they are an extremely small proportion of the current landscape and often have been salvaged or otherwise treated. Information about disturbance history and stand origin, especially post-fire conditions, is pertinent for interpreting conditions for wildlife species such as Black-backed Woodpecker that use and select for dense clumps of snags in recent post-fire situations.

Post-fire information, such as wildlife data or current conditions in the analysis area, can only be compared to inventory data if the appropriate scale is used for analysis. The analysis area should have the same proportion of the high snag density classes as the percent of area (y-axis) in the DecAID inventory histograms. Large fires can hugely skew the current conditions, even in a 5th field HUCs, to the point that the analysis area is no longer representative of habitat conditions from which the inventory data were collected. See the Considerations of scale: landscape and stand levels section of the Summary Narratives.

What this means for the user:
  • Inventory data are not summarized for the post-fire structural condition class.
  • Inventory data should not be used to determine "natural" reference conditions for post-fire situations unless the size of the analysis area is large enough to reflect the size and intensity of the disturbance you are assessing.

Effect of scale of the inventory designs

The dead wood estimates must be interpreted in light of the inherent scale imposed by the inventory designs. Each observation that entered our summaries was an individual field plot. Each plot encompassed about a one- hectare area, within which snags were sampled on fixed- or variable-radius subplots and down wood was sampled on line transects. Plot area, subplot sizes, and transect lengths varied somewhat within and among the datasets. Within-plot variability is not represented in this study.

Also, because the inventory plots sampled an area smaller than a typical forest stand, the plot-level observations should not be thought of as representing stand-level conditions. Rather, our summaries describe aggregate properties of variability of dead wood on multiple plots that sample a given wildlife habitat. We believe it is reasonable to apply distributional information about dead wood that is based on many inventory plots in a given vegetation condition to a management “unit” at the scale of a landscape or sub-watershed.

What this means for the user:
  • Use the appropriate scale for analysis. See the Considerations of scale: landscape and stand levels section of the Summary Narratives.

What the inventory data represent


The dead wood summaries cannot be assumed to apply to areas that are not represented in the inventory data. At the time of this data compilation (2002), National, state, and county parks and privately owned reserves were not sampled, and down wood was not sampled on nonfederal lands in Oregon and western Washington. Within National Forests, wilderness areas were sampled at one quarter the intensity as areas outside wilderness. Wetlands, coastal dunes, aspen, and riparian forests were not well represented in the sample-based inventories and were excluded. In addition, because Bureau of Land Management (BLM) lands at the time were sampled less intensively than Forest Service (FS) lands, the inventory tolerance levels are more indicative of dead wood conditions on FS lands. In general, where dead wood on these ownerships differed significantly, there was less dead wood on FS lands than BLM lands. See the Statistical Basis document for more detail.

The distribution and estimated variation of dead wood within each wildlife habitat is the result of the interaction between plot size and the spatial pattern of dead wood. Smaller plot sizes would result in greater variability, since smaller plots are more likely to sample dense clumps of dead wood or fall in gaps where no dead wood exists.

Information on stumps also is lacking from the forest inventory data. Stumps can serve as wildlife habitat as well as an indicator of the below-ground system. Neither do the summaries include data on partially dead trees

Although the estimates of amounts of dead wood are from plots measured at a single point in time, the current conditions express events that have occurred over the past decades to centuries.

The structural conditions were defined by current vegetation structure on the plots. They should be strongly related to stand age and length of time since the most recent stand-replacing disturbance, but data to verify this assumption are not available. Furthermore, the structural conditions are less useful for describing uneven-aged stands that are common in southwest Oregon and east of the crest of the Cascades, and do not reflect the effects of selective timber harvesting or other factors that influence tree density.

The summaries could be improved by better information on the occurrence of contrasting forest conditions within the CVS inventory plots. There currently is no easy way to identify different conditions such as successional stages within the inventory plots. Consequently, some of the inventory plot-level estimates of dead wood and classifications of wildlife habitats and structural conditions represent averages across contrasting conditions. This introduces an unknown level of error into our regional-level characterizations, but we do not think this error is sufficient to compromise our overall findings.

The estimates represent average conditions within a vegetation condition at the regional level, rather than conditions around specific wildlife nest, roost, den, or resting sites.

What this means for the user:
  • Because estimates from inventory data are for a single point in time, incorporate snag and down wood dynamics in to the analysis of dead wood habitat to provide a continuous supply of dead wood.
  • The inventory data do not include information on wood decay elements other than snags and down wood, and reflect site and landscape disturbance histories.
  • The inventory data represent average conditions across the landscape, whereas much of the wildlife data were collected at nest, roost, den, or resting sites. Note, at a minimum, at least on snag or piece of down wood occurs at these wildlife use site, whereas a percentage of inventory plots contained no measurable snags or down wood.

Spatial and historic distributions of snags and down wood in the inventory data

The summaries of sample-based inventory data provide guidance on the (statistical) distributions of dead wood amounts within landscapes, but do not provide the basis for where to spatially locate the different levels of dead wood.

The range of variability in dead wood abundance that is present among plots in the region can help guide distribution of dead wood within a large landscape or watershed being managed. However, caution must be exercised in using the regional plot data, which sample current conditions, to describe the historic range of conditions in dead wood. Data on disturbance history of the plots is incomplete. Plots in unharvested forest may have been altered by fire exclusion and other human influences. On the eastside in particular, current levels of dead wood in some areas may be elevated above historical conditions due to fire suppression and increased mortality, and may be depleted below historical levels in local areas burned by intense fire or subjected to repeated salvage and firewood cutting. Plot data from unharvested forests on the westside, where fire return intervals are longer, may provide a reasonable approximation of historical conditions.

What this means for the user:
  • The distribution histograms of the inventory data are more appropriately used to represent distributions of snag or down wood densities across landscapes rather than among individual sites or stands.
  • When using the inventory summaries of unharvested forest conditions as representations of historic range of conditions in vegetation and structural conditions east of the Cascades, the user should also become familiar with other published estimates of historic conditions. These additional estimates are described and discussed in the Summary Narratives.

CAVEATS AND CAUTIONS IN USING THE INSECT AND DISEASE SUMMARIES


Root Disease Tables
Data used
Only CVS (Forest Service lands) and BLM (Bureau of Land Management lands) inventory data, and not FIA (State and private lands) data, were used in developing the tables which show the percentages of plots with root disease for the various vegetation conditions. CVS and BLM inventories have “damage” codes for the five major root diseases found in the Pacific Northwest (laminated root rot, armillaria root disease, annosus root disease, black stain root disease, and Port-Orford-cedar root disease) as well as a code for “other”, which means the crew thinks that root disease is present but they could not identify the causal agent. In addition, there are three fields for each tree in which one can enter damage codes. Thus, more than one of these primary diseases may be recorded for an individual tree. In addition, these diseases will be recorded if they are detected on a plot tree, or within 30 ft of a plot tree.

FIA data used were collected using different guidelines, these data could not readily be summarized with CVS and BLM data in a meaningful way. There is only one field for recording the factor (called a growth impactor) which most impacts the tree’s growth and survival, and, for many of the codes, the impactor must be expected to kill the tree within 10 years or cause “serious damage” in order to be recorded.

Cautions in interpreting data

From knowledge of these areas in the field, it appears that armillaria root disease is greatly over-counted in some of the vegetation conditions. For example, in the WLCH_OCA vegetation condition, inventory data indicated that armillaria root disease was present in 34% of the plots and unidentified root disease was present on 16% of the plots. In addition, it indicated that 48% of the plots had at least one root disease, including those unidentified ones recorded as “other”. This implies that over the landscape on Forest Service and BLM lands in this vegetation condition, one could detect root disease on 48% of the area. This does not seem reasonable, and points to potential problems in identification and recording of actual root diseases (i.e., fungi acting as pathogens) as opposed to recording fungi which are acting as saprophytes. This seems to be a common mistake made in coding armillaria in particular, and it points to a need to enhance the abilities of the crews in identification, or to change the data collection methods to improve accuracy.

What this means for the user:
  • These data do seem to provide reasonable estimates in for the percentages of plots with annosus root disease and laminated root rot in the different vegetation conditions.
  • Root disease data derive from studies on federal lands. If the user is concerned about root disease on non-federal lands, they will need to determine if the data are pertinent to their lands.

Heart Rot (Stem Decay) Tables
Data used
Only CVS (Forest Service lands) and BLM (Bureau of Land Management lands) inventory data, and not FIA (State and private lands) were used in developing the tables which show the percentages of trees >50-cm dbh by species with visible signs of heart rot (stem decay). CVS and BLM inventories have “damage” codes for three of the common heart rots (red ring rot [Phellinus pini], rust red stringy rot or Indian paint fungus [Echinodontuim tinctorium], and brown cubical rot or Schweinitzii root and butt rot [Phaeolus schweinitzii] as well as a code for “other”, which means the crew detected signs of decay, but could not identify the causal agent. There are three data fields for each tree in which one can enter damage codes, so it is possible for more than one species of decay fungi to be recorded for a single tree.

FIA plots did not have a field for recording presence of stem decays. The Only field which made reference to stem decay was the CULLROT field where crews estimated the tree volume lost to rot. Thus, these data could not be summarized with the CVS and BLM data in a meaningful way.

Cautions in interpreting data

Some of the percentages of trees per hectare presented in the tables are based on a rather small sample size (i.e., there weren’t many trees >50-cm dbh of that species within that vegetation condition), so use the numbers, especially those of some of the less dominant tree species, with some caution.

By far, the stem decays most commonly recorded in trees >50 cm dbh were placed in the unidentified category.

What this means for the user:
  • This indicates that when wildlife trees or potential snags in the field, one may need assistance from forest pathologists in determining which heart rot fungi are present and the effects a particular species has on the tree. This will help in determining the usefulness of a particular decay organism to the wildlife species of interest.
  • Stem decay data derive from studies on federal lands. If the user is concerned about stem decay on non-federal lands, they will need to determine if the data are pertinent to their lands.

Dwarf Mistletoe Tables
Data used
Unlike the root disease and stem decay tables, CVS (Forest Service lands), BLM (Bureau of Land Management lands) inventory data, and FIA (State and private lands) were used in developing the tables which show the percentages of plots with dwarf mistletoe present on several tree species. These data were collected similarly for all inventories.

Cautions in interpreting data

These tables show only whether dwarf mistletoe was present on a plot for a particular tree species, and say nothing about the severity of the mistletoe on the plot.

What this means for the user:
  • When selecting wildlife trees and potential snags in the field, the severity of the mistletoe in the stand is important. If live, heavily infected trees are left in the stand, this must be taken into consideration when predicting whether the stand will meet the long term management objectives of the area.

LITERATURE CITED

Raphael, M. G. 1983. Cavity-nesting bird response to declining snags on a burned forest: a simulation model. Pp. 211-215 in: J. W. Davis, G. A. Goodwin, and R. A. Ockenfels, eds. Snag habitat management: proceedings of the symposium. USDA Forest Service General Technical Report RM-99. Flagstaff AZ.