The Condition of Oregon’s Forests and Woodlands:
Implications for the Effective Conservation of Biodiversity
By: Catherine Macdonald, Steven Buttrick, and Michael Schindel
For millennia, fire has played an important role in shaping the composition, structure and
processes of most native ecosystems. Suppression of wildland fire over the past 60 to 100
years, along with widespread livestock grazing and logging, has altered the
characteristics of most ecosystems in Oregon. When fires occur following long-term fire
suppression and other habitat modifying practices, fire behavior can be more intense with
more severe ecosystem effects.
Ecosystem changes resulting from alterations in natural fire regimes affect habitat
conditions for plants and animals. While changes in habitat suitability for an individual
species is certainly nothing out of the ordinary, for native species and ecological systems
that are already in decline due to other land use pressures, changes from fire suppression
or alternatively unnaturally severe fires may add one more nail to the coffin. Eighty-four
percent of places identified by scientists as important for global conservation are
estimated to be at risk from changes that have created too much, too little or the wrong
kind of fire (The Nature Conservancy 2005).
Over the past ten years, federal agencies have treated thousands of acres of forests and
woodlands through application of prescribed fire, thinning, and wildland fire to reduce
fuel loads and/or modify fire behavior (National Fire Plan,
http://126.96.36.199/index.htm). Despite these efforts the general consensus is that
conditions in our fire-prone wildland forests and woodlands are getting worse, not better.
Under-investment in the use of fire and forest restoration, lack of or limitations in
processing infrastructure, market conditions, fragmented ownership patterns, and long-
standing disagreements over how our forests should be managed continue to be barriers
to progress in addressing this problem.
To frame this problem at the statewide scale, we analyzed the recently released draft
LANDFIRE Rapid Assessment fire regime and condition class data for fire-prone forests
and woodlands and lands within the Wildland Urban Interface. We selected forests and
woodlands for analysis based on their fire regime characteristics, current condition, and
proximity to the Wildland Urban Interface. We then compared our estimates of the acres
of forests and woodlands that would need to be treated annually to address
uncharacteristic fuel loads, restore fire as a natural process, or reduce fire risk in the
Wildland Urban Interface to reported federal treatment accomplishments to get a first
approximation of the gap between current and needed restoration efforts.
Recognizing that the LANDFIRE Rapid Assessment data on conditions are coarse and
the predictions greatly simplify the problem and solutions, we propose a roadmap for
developing a Wildland Restoration and Conservation Plan for Oregon. In addition to the
benefits to Oregon’s biodiversity, a blueprint such as this would provide critical and
timely information and context to evaluate and take advantage of any economic
opportunities, such as biomass utilization, associated with restoration.
In Oregon, wildfires maintained extensive grasslands and savanna habitats in the interior
valleys, maintained open park-like stands of ponderosa pine forests in southwestern and
eastern Oregon, and periodically re-set succession in our climax lodgepole pine and
coastal spruce forests. The frequency, intensity, and ultimate size of fires on the
landscape depend on elevation, topography, wind characteristics, relative humidity, and
fuel loads. When considered together, certain patterns in fire frequency, predictability,
seasonality, size, and severity can be grouped into Historic Natural Fire Regimes (Table
Table 1: Fire Regimes from Schmidt et al. (2002) as modified for use by LANDFIRE.
0–35 year frequency, low and mixed severity
35–200 year frequency, low and mixed severity
200+ year frequency, stand-replacement severity
Fire suppression – as well as logging, domestic livestock grazing, and the introduction
and establishment of non-native species – has altered much of the natural vegetation in
the United States. In fire-adapted forests and woodlands, exclusion of fire often results in
increased density of trees and shrubs, proliferation of ladder fuels, accumulation of dead
and down fuels, a shift in composition to less fire-resilient species, and an increase in the
vulnerability of older overstory trees to insects and disease; it also contributes to the
establishment of invasive non-native species. These changes impact plant and animal
populations. For example, acorn woodpeckers have less food and fewer potential nesting
sites in the higher density narrow canopy oak and overtopping Douglas-fir than they had
in the open-grown trophy-form oaks in our western Oregon oak savannas and
woodlands. Sage grouse avoid shrub steppe habitats in southeast and central Oregon
where western juniper has become established.
Uncharacteristically severe fires damage soils, increase stream temperatures and
sedimentation, cause nutrient loss, and impact vegetation, causing mortality both above
and below ground. Unnaturally frequent fires in shrub steppe habitats invaded by
cheatgrass consume sagebrush, reducing nesting habitat for sage sparrows, killing or
weakening native bunchgrasses and forbs and encouraging cheatgrass expansion.
Quigley et al. (1996) estimated that across the inland Northwest the percentage of forests
predicted to burn with high severity has increased from 20 to 50 percent from historic to
To evaluate the current condition of lands in relation to their historic or “natural reference
condition,” the Forest Service developed a three-level classification of Fire Regime
Condition Classes (Schmidt et al. 2002). Condition Class describes the degree to which
factors like vegetation condition and structure, fire frequency, and severity depart from
natural or historical ecological reference conditions. Condition Class 1 represents no,
minimal, or low departure; Condition Class 2 represents moderate departure; and
Condition Class 3 represents high departure (Table 2). The greater the departure, or the
more highly departed the vegetation conditions, the more likely wildland fires will be
uncharacteristic in relation to the historical fire behavior (severity, intensity and pattern)
and further degrade vegetation structure and condition.
Table 2: Fire Regime Condition Classes from Schmidt, et al., (2002) as interpreted by
Hann and Strohm (2003).
Condition Departure from
Natural Range of
Vegetation composition, structure and fuels are similar to
those of the natural regime and do not predispose the
system to risk of loss of key ecosystem components.
Wildland fires are characteristic of the natural fire regime
behavior, severity, and patterns. Disturbance agents,
native species habitats, and hydrologic functions are
within the natural range of variability.
Vegetation composition, structure, and fuels have
moderate departure from the natural regime and
predispose the system to risk of loss of key ecosystem
components. Wildland fires are moderately
uncharacteristic compared to the natural fire regime
behaviors, severity, and patterns. Disturbance agents,
native species habitats, and hydrologic functions are
outside the natural range of variability.
Vegetation composition, structure, and fuels have high
departure from the natural regime and predispose the
system to high risk of loss of key ecosystem components.
Wildland fires are highly uncharacteristic compared to the
natural fire regime behaviors, severity, and patterns.
Disturbance agents, native species habitats, and
hydrologic functions are substantially outside the natural
range of variability.
* Natural range of variation = the ecological conditions and processes within a specified area, period of
time, and climate, and the variation in these conditions that would occur without substantial influence from
Schmidt et al. (2002) used this classification and remotely sensed vegetation data from
the early 1990’s to map Fire Regime and Condition Class at a 1-km2 resolution for all
federal and non-federal lands in the conterminous United States, excluding agricultural,
barren, and urban/developed lands. They found that 38 percent of the lands assessed were
moderately altered from the natural regime (Condition Class 2) and an additional 15
percent were significantly altered from the natural regime (Condition Class 3).
While Schmidt et al.’s (2002) assessment of conditions was coarse, it identified the
significant challenges and opportunities we face in restoring forest resilience, reducing
risk of unnaturally severe fires, and managing lands to avoid this problem in the future.
To improve upon Schmidt et al.’s (2002) assessment, the US Forest Service, US
Geological Survey, The Nature Conservancy and others are working together on a five-
year mapping effort called LANDFIRE (http://www.landfire.gov/). LANDFIRE is
charged with developing consistent nationwide data to more precisely identify the extent
and location of wildfire risks associated with hazardous fuels, and to better target fuel
LANDFIRE recently released Rapid Assessment geospatial data and models of potential
natural vegetation groups, fire regimes, and fire regime condition classes for Oregon.
Rapid Assessment products are mapped at a 30 meter resolution and use the most recent
vegetation maps available to classify recent satellite imagery. Reference condition
models and descriptions were developed with input from over 250 managers.
The LANDFIRE Rapid Assessment compares the current percent distribution of
vegetation in five structural stages to modeled reference distributions. The structural
stages are defined as:
Early Seral: post replacement disturbance such as stand replacement fire, or clear cut;
Mid Seral Closed: mid successional; mid age; competition stress; fire suppression
Mid Seral Open: mid successional; mid age; disturbance maintained;
Late Seral Open: late successional; mature age; disturbance maintained; and
Late Seral Closed: late successional; mature age; competition stress.
Condition Class is determined based on the difference in percentages for each vegetation
type within large landscapes (averaging 1 million acres in size). A departure between 0
and 33 percent is placed in Condition Class 1; 34-66 percent in Condition Class 2, and a
departure of greater than 66 percent in Condition Class 3. For example, a difference in
current and modeled reference conditions of 70 percent means that 70 percent of the acres
of a given vegetation type within a subsection is not characteristic of the reference
condition, so the entire unit would be mapped as Condition Class 3.
We summarized the draft LANDFIRE Rapid Assessment data to evaluate the scope of
restoration needed to return reference conditions to Oregon’s fire-prone forests and
woodlands and to reduce fuels in the Wildland Urban Interface. Outside the Wildland
Urban Interface, we targeted only those forests and woodlands where fire regimes were
most likely to have played a significant role in shaping current conditions – forests and
woodlands with low or mixed severity Fire Regimes with moderately or highly altered
species composition and structure.
Low and Mixed Severity Fire Regimes with altered species composition and structure:
We included all but one of the forest and woodland types in Fire Regimes I and III in
Condition Classes 2 and 3. Generally, forests and woodlands in Fire Regimes I and III
have seen more significant departures in both structure and composition due to fire
suppression and are most threatened by uncharacteristically severe wildfires (Agee 1993).
Condition Class 3 lands typically have the greatest fuel loads and are the most likely to
generate unnaturally severe wildfires1. We included lands in Condition Class 2 under the
assumption that acting now to restore conditions and ecological processes in these
landscapes would be more feasible and less costly. We did not include Douglas-fir-
hemlock dry mesic forests common in the West Cascades and parts of the Coast Range.
While these forests are in Fire Regime III, they have a relatively long fire return interval
of 100+ years. We assumed that changes in the structure and condition of these forests
are more likely the result of harvest practices rather than fire suppression.
We did not include forests and woodlands in Fire Regimes II and IV such as lodgepole
pine and subalpine fir forests, where stand-replacing fires were the norm, unless they
occurred within the Wildland Urban Interface. From an ecological perspective, treating
fuels to eliminate severe fires in these systems would produce unnatural habitat
conditions with potential negative impact to wildlife and watersheds (Brown et al. 2004).
Wildland Urban Interface: In addition to the lands identified above, we summarized the
number of acres of forests and woodlands in Fire Regimes II and IV, and one type in Fire
Regime V, in all Condition Classes in the Wildland Urban Interface. We defined the
boundaries of the Wildland Urban Interface using data developed by the SILVIS Lab,
Department of Forest Ecology and Management, and University of Wisconsin-Madison
(Radeloff et al. 2005). The SILVIS Lab used data from the US Census and USGS
National Land Cover Data to create a spatially explicit map of the Wildland Urban
Interface and Intermix (WUI) corresponding with the Federal Register definition of WUI.
In our analysis, we included areas mapped as low-density interface and higher, and low-
density intermix and higher. In Fire Regime V, we only included the acres of western
juniper pumice woodlands in Eastern Oregon. While these woodlands are classified as
having a low frequency of fire, they burn more frequently now, due in part to the
presence of invasive species such as cheatgrass.
We did not include any west-side Fire Regime V forests inside or outside the Wildland
Urban Interface. These forests are characterized by infrequent (200+ years) stand-
replacing fires. Outside of extended drought, Sitka spruce, western hemlock, and mesic
Douglas-fir forests are relatively insulated from wildfires and do not require treatments to
make them fire safe. Altered conditions in these forests are primarily a result of timber
1 Some of the lands mapped in both Condition Class 2 and 3 may have been assigned to these
classes based on factors other than fire suppression and may not have uncharacteristic fuel
accumulations, e.g. Sitka spruce-hemlock and Douglas-fir-hemlock dry-mesic forests.
harvest. While restoration of structural diversity in these second growth west-side forests
may advance land management goals and objectives, the issues with these forests were
beyond the scope of this paper.
We summarized the acreage data by ownership (private, tribal, FS, BLM, and Other
public), and within public lands by management category and accessibility,
distinguishing Wilderness and Roadless areas from all other forest designations. The
existing ownership and management category layer was derived from a geospatially-
explicit statewide coverage developed by the Oregon Natural Heritage Information
Center (2005). To bracket our estimates of the total acres in Condition Class 2 and 3 in
these forests and woodlands, we further evaluated the Rapid Assessment data to include
fire regime departure values.
We calculated the number of forest and woodland acres that would need to be treated
annually through the application of wildland fire use, prescribed burning, and thinning to
restore reference conditions and to reduce fuels in the Wildland Urban Interface,
assuming treatment periods of 20 or 25 years. Current treatment levels for all habitats in
Oregon were taken from accomplishments reported on the National Fire Plan website
(http://www.fireplan.gov/NFP_HFT_YTD.cfm?StateName=Oregon) for fiscal years
2003-2005. We estimated the number of acres of forests and woodlands treated in 2005
based on a more detailed report of accomplishments made available by the Forest
Service/Bureau of Land Management’s Fire and Aviation Program (2006) and coarse
estimates of the percentage of the selected forest and woodland habitats described above
in each of the reporting Forests and Districts.
Fires resulting from both lightning and anthropogenic ignitions have played a major but
varied role in shaping vegetation structure, composition, and dynamics in Oregon. The
most common fire regimes in Oregon are Fire Regimes III, I, and IV, which collectively
make up 79 percent of the state (Table 3; Figure 1). Fifty-five percent of Oregon’s
vegetation historically experienced low or mixed severity fire in Fire Regime I and III.
Forest and woodland habitats are found in each of the Fire Regimes (Table 3; Appendix).
Low elevation forests, woodlands, and grasslands across the state are classified in Fire
Regime I. The majority of mid-elevation forests and woodlands are classified as Fire
Regime III, except in southwestern Oregon where they are also classified in Fire Regime
I based on the higher fire frequencies. Forest types dominated by trees with low
tolerance to fire such as Sitka spruce dominated forests on the Oregon Coast, subalpine
forests at higher elevations in the Cascades, and lodgepole pine forests in Central Oregon
are in Fire Regimes IV and V.
Table 3: Distribution of Fire Regimes in Oregon by percent area and major associated
Generally low elevation types including grasslands, oak and pine
savannas and woodlands as well as mesic ponderosa pine and drier
mixed conifer forests
Higher elevation and moister grasslands (Idaho fescue) and
mountain big sage habitats
Drier ponderosa pine forests, more mesic Eastern Oregon mixed
conifer forests, and dry to mesic Douglas-fir - hemlock forests in
western Oregon, as well as juniper steppe woodlands and low
Wyoming big sagebrush and lodgepole pine forests
Moist forests such as: Douglas-fir - hemlock forests, mountain
hemlock, and Sitka spruce – Hemlock; High elevation systems
such as Pacific silver fir high elevation; and sparse understory
vegetation types such as salt desert shrub.
There are 34,100,000 acres of forest and woodland habitats in Oregon (Figure 2;
Appendix). LANDFIRE Rapid Assessment data categorized 31 million acres (91 percent)
of forests and woodlands in Condition Class 2 or 3 (Figure 3). Of the 31 million acres of
forests and woodlands in Condition Class 2 and 3, 20,970,000 (Figures 4 and 5) acres
were classified in Fire Regimes I and III. An additional 55,000 acres of forests and
woodlands in Fire Regime II, IV, and V occur within the Wildland Urban Interface2.
Sixty percent of these forest and woodland acres are managed by public agencies,
primarily the Forest Service and Bureau of Land Management (Table 4; Figure 6). The
majority (84%) of these public lands (12,976,000 acres) are outside of Wilderness and
Table 4: Forest Service, Bureau of Land Management and other state and federal
agencies forest and woodland acres (and percentage of each agencies’ wooded acres) in
select Fire Regime and Condition Classes in Oregon.
Bureau of Land
FR I CC2
FR I CC3
2 WUI acres were not calculated by ownership.
Bureau of Land
FR III CC2
FR III CC3
An average of 64 percent of the lands identified in Condition Class 2 and 3 were
classified as outside of reference conditions.
To restore conditions to all acres, public and private, in the categories identified above
within a 20-25 year treatment period, between 840,000 to 1 million acres would need to
be treated through the use of wildland fire use, prescribed fire, and thinning per year in
Oregon. Looking just at public lands, annual treatment levels would need to range from
670,000 acres a year over 20 years including wilderness and roadless areas to 447,000 a
year over 25 years and not including wilderness and roadless areas (Table 5).
Table 5: Annual acres of Forests and Woodlands needing treatment by category in 20, 25
year restoration timeframes.
All public lands, including
All non wilderness and non
roadless public lands,
* All forest/woodland acres in FR I CC 2+3, FR III CC 2+3, all FR II and IV in WUI.
Over the past three years, the federal agencies have treated between 208,000 and 368,000
acres a year on public lands across all habitat types in Oregon – not just forests and
woodlands (National Fire Plan 2006). The Bureau of Land Management treated 30 to 48
percent of the total treatment acres over the last three years; the Forest Service treated 48
to 60 percent of the total treatment acres. For the Federal Fiscal Year 2005, of the
215,656 acres treated by the Forest Service and Bureau of Land Management in Oregon,
we estimated that approximately 73 percent or 156,475 acres of forest or woodlands were
treated (Table 6).
Table 6: Estimate of acres of select forest and woodland habitats treated by the Forest Service and
Bureau of Land Management to reduce Hazardous Fuels in fiscal year 2005. (Overall treatment
data as reported to by the USFS and BLM (as of 1/3/2006). Percent of Forest Treatments in FR I &
III outside the WUI and all forest types inside the WUI coarsely approximated based on percent
area in forest or woodland habitats.)
Forest Treatments Treatment Acres
Columbia River Gorge
Total FS and BLM
Reported treatment of acreage includes thinning, prescribed fire, and wildland fires.
Agencies report acres each time they are treated. Many acres may take multiple
treatments to restore reference conditions over a landscape. Current data are not available
on actual acres restored annually to reference conditions.
Scope of the Problem: Concern over fuel build-up in our forests has grown in the past 10
years. Understandably, much of the effort to address the problem has focused on reducing
fire risks in our Wildland Urban Interface rather than restoring ecosystem processes in
our wildland forests and woodlands. While we agree with the importance of improving
fire safety around our human communities, our estimates call attention to the much
greater restoration needs outside the WUI and the need to take a much more
comprehensive look at the problem.
Given the past 60-100 years of fire suppression, our estimated annual treatment needs
make sense. Agee (1990) estimated that historically an average of 794,000 acres of
Oregon’s forests burned each year based on studies of the fire histories of our forests and
woodlands. The bulk of the acres, some 650,000 acres, were of the low and mixed
severity fire regime Ponderosa pine and mixed conifer forests.
Here we have estimated that the annual rate of treatment needed to restore forest and
woodland conditions on public lands and increase safety in the Wildland Urban Interface
over the next 20-25 years is 3.3 to 4.6 times current agency treatment rates. The gap
between current and needed treatment levels appears to be greater on Forest Service
managed forest lands than on Bureau of Land Management holdings.
While our numbers call for a substantial increase in treatment efforts, consideration of
several additional factors could further increase the actual annual treatment needs. First,
we based our estimate on the assumption that only one treatment is needed per acre. In
reality, the Condition Class 3 lands (6,257,000 acres) will, in most cases, need more than
one treatment over the course of several years – one or more manual or mechanical
treatments to reduce fuel loads and help restore composition and structure, slash disposal
or biomass removal, and prescribed fire to reintroduce natural processes. Using the same
approach to counting treatment acres currently used by the Forest Service and Bureau of
Land Management, and assuming that the Condition Class 3 forests and woodlands we
selected would need at least two treatments, our estimates of the annual treatment levels
would increase by at least 33 percent.
Added to this, active management is needed to maintain stands in Fire Regime I and III
that are currently in Condition Class 1; and for some habitats, re-treatment may be
necessary within the 20-25 year treatment period. The Rapid Assessment data identified
780,000 acres of forests and woodlands on public land in Fire Regimes I and III,
Condition Class 1. Prescribed fire should be the preferred management tool for initial
treatment or re-treatment in these circumstances both based on cost and the added value