Wildland Fires
Forest Service and BLM Need Better Information and a Systematic Approach for Assessing the Risks of Environmental Effects Gao ID: GAO-04-705 June 24, 2004Decades of fire suppression, as well as changing land management practices, have caused vegetation to accumulate and become altered on federal lands. Concerns about the effects of wildland fires have increased efforts to reduce fuels on federal lands. These efforts also have environmental effects. Congressional requesters asked GAO to (1) describe effects from fires on the environment, (2) assess the information gathered by the Forest Service and Bureau of Land Management (BLM) on such effects, and (3) assess the agencies' approaches to environmental risks associated with reducing fuels.
Wildland fires can have dramatic effects on environmental resources and ecosystems, including production of large amounts of smoke, loss of trees, and erosion of soil into streams and lakes. However, fires can also benefit resources by recycling soil nutrients, renewing vegetation growth, and adding gravel to streams, which improves spawning habitat for fish. The 20 wildland fires that we surveyed burned over 158,000 acres of federal land and had complex, wide-ranging, and sometimes contradictory, effects on both individual resources, such as trees and streams, and ecosystems. For example, the short-term effects of the Missionary Ridge fire in Colorado that burned almost 50,000 acres of trees and other vegetation included increased debris and sediment that affected water quality in some areas. However, in other areas, officials said even dramatic changes to streams would not be detrimental in the long term. The Forest Service and BLM gather specific information on the environmental effects of individual wildland fires, such as soil erosion. The agencies do not, however, gather comprehensive data on the severity of wildland fire effects on broad landscapes and ecosystems--that is, large areas that may involve one or more fires. The agencies recently developed a monitoring framework to gather severity data for fires, but they have not yet implemented it. These data are needed to monitor the progress of the agencies' actions to restore and maintain resilient fire-adapted ecosystems, a goal of the National Fire Plan. The National Fire Plan directs the Forest Service and BLM to target their fuel reduction activities with the purpose of lowering the risk of environmental effects from wildland fires in areas that face the greatest losses. However, the agencies do not systematically assess the risks across landscapes that fires pose to different environmental resources or ecosystems or the risks of taking no action on fuel reduction projects. At the landscape level, the Forest Service and BLM do not have a formal framework for systematically assessing the risk of fire to resources and ecosystems, although some of the forests and BLM field offices have developed risk assessments on their own or in collaboration with regional, state, or local efforts. At the project level, while the agencies recognize the need to better analyze the risk of acting to reduce fuels versus not doing so, neither fire planning guidance nor National Environmental Policy Act guidance specify how to do this. Opportunities exist to clarify how the agencies should analyze the effects of not taking action to reduce fuels. The agencies can clarify interim guidance to implement the Healthy Forests Restoration Act, and the agencies can, in conjunction with Council on Environmental Quality (CEQ), further develop the lessons learned from a CEQ demonstration program carried out in 2003. Without a risk-based approach, these agencies cannot target their fuel reduction projects across landscapes or make fully informed decisions about which effects and project alternatives are more desirable.
RecommendationsOur recommendations from this work are listed below with a Contact for more information. Status will change from "In process" to "Open," "Closed - implemented," or "Closed - not implemented" based on our follow up work.
Director: Team: Phone:GAO-04-705, Wildland Fires: Forest Service and BLM Need Better Information and a Systematic Approach for Assessing the Risks of Environmental Effects
This is the accessible text file for GAO report number GAO-04-705
entitled 'Wildland Fires: Forest Service and BLM Need Better
Information and a Systematic Approach for Assessing the Risks of
Environmental Effects' which was released on July 26, 2004.
This text file was formatted by the U.S. Government Accountability
Office (GAO) to be accessible to users with visual impairments, as part
of a longer term project to improve GAO products' accessibility. Every
attempt has been made to maintain the structural and data integrity of
the original printed product. Accessibility features, such as text
descriptions of tables, consecutively numbered footnotes placed at the
end of the file, and the text of agency comment letters, are provided
but may not exactly duplicate the presentation or format of the printed
version. The portable document format (PDF) file is an exact electronic
replica of the printed version. We welcome your feedback. Please E-mail
your comments regarding the contents or accessibility features of this
document to Webmaster@gao.gov.
This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed
in its entirety without further permission from GAO. Because this work
may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this
material separately.
Report to Congressional Requesters:
June 2004:
Wildland Fires:
Forest Service and BLM Need Better Information and a Systematic
Approach for Assessing the Risks of Environmental Effects:
GAO-04-705:
GAO Highlights:
Highlights of GAO-04-705, a report to congressional requesters:
Why GAO Did This Study:
Decades of fire suppression, as well as changing land management
practices, have caused vegetation to accumulate and become altered on
federal lands. Concerns about the effects of wildland fires have
increased efforts to reduce fuels on federal lands. These efforts also
have environmental effects. The requesters asked GAO to (1) describe
effects from fires on the environment, (2) assess the information
gathered by the Forest Service and Bureau of Land Management (BLM) on
such effects, and (3) assess the agencies‘ approaches to environmental
risks associated with reducing fuels.
What GAO Found:
Wildland fires can have dramatic effects on environmental resources and
ecosystems, including production of large amounts of smoke, loss of
trees, and erosion of soil into streams and lakes. However, fires can
also benefit resources by recycling soil nutrients, renewing
vegetation growth, and adding gravel to streams, which improves
spawning habitat for fish. The 20 wildland fires that we surveyed
burned over 158,000 acres of federal land and had complex,
wide-ranging, and sometimes contradictory, effects on both individual
resources, such as trees and streams, and ecosystems. For example, the
short-term effects of the Missionary Ridge fire in Colorado that
burned almost 50,000 acres of trees and other vegetation included
increased debris and sediment that affected water quality in some
areas. However, in other areas, officials said even dramatic changes
to streams would not be detrimental in the long term.
The Forest Service and BLM gather specific information on the
environmental effects of individual wildland fires, such as soil
erosion. The agencies do not, however, gather comprehensive data on
the severity of wildland fire effects on broad landscapes and
ecosystems”that is, large areas that may involve one or more fires.
The agencies recently developed a monitoring framework to gather
severity data for fires, but they have not yet implemented it. These
data are needed to monitor the progress of the agencies‘ actions to
restore and maintain resilient fire-adapted ecosystems, a goal of the
National Fire Plan.
The National Fire Plan directs the Forest Service and BLM to target
their fuel reduction activities with the purpose of lowering the risk
of environmental effects from wildland fires in areas that face the
greatest losses. However, the agencies do not systematically assess
the risks across landscapes that fires pose to different environmental
resources or ecosystems or the risks of taking no action on fuel
reduction projects. At the landscape level, the Forest Service and BLM
do not have a formal framework for systematically assessing the risk of
fire to resources and ecosystems, although some of the forests and BLM
field offices have developed risk assessments on their own or in
collaboration with regional, state, or local efforts. At the project
level, while the agencies recognize the need to better analyze the risk
of acting to reduce fuels versus not doing so, neither fire planning
guidance nor National Environmental Policy Act guidance specify how to
do this. Opportunities exist to clarify how the agencies should
analyze the effects of not taking action to reduce fuels. The
agencies can clarify interim guidance to implement the Healthy Forests
Restoration Act, and the agencies can, in conjunction with CEQ, further
develop the lessons learned from a CEQ demonstration program carried
out in 2003. Without a risk-based approach, these agencies cannot
target their fuel reduction projects across landscapes or make fully
informed decisions about which effects and project alternatives are
more desirable.
What GAO Recommends:
This report recommends that the Secretaries of Agriculture and the
Interior (1) develop a plan to implement the agencies‘ monitoring
framework, (2) develop guidance that formalizes the assessment of
landscape-level risks to ecosystems, and (3) clarify existing
guidance, working with the Council on Environmental Quality (CEQ), to
assess the risks of environmental effects from reducing fuels.
Commenting on the draft report, Agriculture and Interior agreed that
more data are needed and prioritization of fuels work can be improved,
but had concerns about developing guidance on a risk-based approach.
CEQ commented that its guidance is not intended to address risk
analysis.
www.gao.gov/cgi-bin/getrpt?GAO-04-705.
To view the full product, including the scope and methodology, click
on the link above. For more information, contact Barry T. Hill at
(202) 512-9775 or hillbt@gao.gov.
Contents:
Letter:
Results in Brief:
Background:
Wildland Fires Have Wide-Ranging Effects on Environmental Resources and
Ecosystems, Depending on a Range of Factors:
Forest Service and BLM Do Not Gather Comprehensive Information on the
Varied Effects of Wildland Fires on Ecosystems and Landscapes:
The Forest Service and BLM Do Not Systematically Assess the Risks to
Environmental Resources and Ecosystems to Target and Conduct Fuel
Reduction Activities:
Conclusions:
Recommendations for Executive Action:
Agency Comments and Our Evaluation:
Appendixes:
Appendix I: Objectives, Scope, and Methodology:
Appendix II: Fire Regime Condition Class Analysis:
Appendix III: Definition of Fire/Burn Severity:
Appendix IV: Selected Wildland Fire Survey Results:
Appendix V: Remote Sensing Data and Systems:
Appendix VI: Examples of Models for Assessing Wildland Fires and Fuels:
Appendix VII: Consolidated Comments from the Departments of Agriculture
and the Interior:
Appendix VIII: Comments from the Council on Environmental Quality:
Appendix IX: GAO Contacts and Staff Acknowledgments:
GAO Contacts:
Staff Acknowledgments:
Tables:
Table 1: Assessment of Overall Effects on Individual Resources in the
Short and Long Term:
Table 2: Description of Characteristic and Uncharacteristic Fire
Effects and Vegetation Conditions:
Table 3: Forest Service and BLM Office Assessments and the Risk Elements
Addressed:
Table 4: Results of GAO's Review of 10 Fuel Reduction Project
Environmental Assessments:
Table 5: Classes of Fire Severity for Soils and Vegetation:
Table 6: Acres Burned by Fire, Land Ownership, and Percent of Forest and
BLM Field Office Land Base:
Table 7: Types and Uses of Federal Land within Fire Perimeters:
Table 8: Miles of Perennial Streams and Number of Floods, Debris Flows,
or Landslides within Fire Perimeters:
Table 9: Municipal Watersheds within Severely Burned Areas:
Table 10: Fires Affecting Threatened and Endangered Species Populations
and Habitats:
Table 11: Air Emissions by Fire:
Table 12: Fire Effects on Soils by Burn Severity and Erosion Potential:
Table 13: Fire Effects on Vegetation:
Figures:
Figure 1: Number of Fires and Acres Burned, 1960-2002:
Figure 2: Effects of Wildland Fire in Treated and Untreated Areas Burned
by Wildland Fire:
Figure 3: Conceptual Short-and Long-Term Effects on Vegetation After a
High-Severity Wildland Fire:
Figure 4: Relationship of Ecosystem and Landscape Levels:
Figure 5: Acres of Vegetation Burned Lightly, Moderately, and Severely
in 20 Sample Fires:
Figure 6: Location of Forest Service and BLM Wildland Fires Visited and
Surveyed:
Figure 7: Wildland Fire Hazard:
Figure 8: Remote Sensing Technologies and the Data Produced:
Figure 9: Images and Data Collected Using Aerial and Satellite
Technology Compared with an Actual Site:
Abbreviations:
BAER: Burned Area Emergency Response:
BLM: Bureau of Land Management:
CEQ: Council on Environmental Quality:
ESR: Emergency Stabilization and Rehabilitation:
FRCC: Fire Regime Condition Class:
HFRA: Healthy Forests Restoration Act:
NEPA: National Environmental Policy Act:
NFMA: National Forest Management Act:
NOAA: National Oceanic and Atmospheric Administration:
Letter:
June 24, 2004:
Congressional Requesters:
Decades of fire suppression, in conjunction with land management
activities that have excluded fire from the nation's forests and
rangelands,[Footnote 1] such as roads and trails, grazing, and
development near public land, have caused the accumulation of brush,
small trees, and other vegetation on federal and other lands. Recent
fire seasons have shown that these land management practices have had
unforeseen consequences. The accumulation and alteration of vegetation,
in combination with an extended drought that has covered much of the
country, has caused wildland fires to burn more intensely than they
would under more natural or historical vegetation conditions. In
response to changing views of wildland fire, the Federal Wildland Fire
Management Policy developed by the Departments of Agriculture and the
Interior in 1995 recognized the natural role of wildland fire and the
potential for "catastrophic" fires to occur in areas of accumulated
vegetation. In 2000, federal scientists and land managers estimated
that 182 million acres of land in the United States had accumulations
of vegetation that were highly altered from more historical conditions.
Under historical conditions, many forest and rangeland ecosystems--
which are different ecological units distinguished by physical
characteristics such as mountains, plains, and river basins, as well as
their associated plant and animal communities--have adapted to wildland
fire, surviving and regenerating after fires occur. Under these
conditions, wildland fire can often have beneficial effects for
resources and ecosystems, such as recycling soil nutrients, renewing
vegetation growth, and sustaining ecological functions. However,
federal scientists and land managers believe that the adverse effects
of wildland fires are exacerbated in ecosystems with uncharacteristic
vegetation conditions, that is, in which vegetation has accumulated or
been altered by fire exclusion. Adverse effects of wildland fire on
individual resources include reduced air and water quality, soil loss,
and loss of threatened and endangered species and their habitat. In
addition, wildland fires that cover broad landscapes can adversely
affect all or parts of forest or rangeland ecosystems. In particular,
scientists and land managers are concerned that, after years of fire
exclusion, in dry forest ecosystems the large old trees that used to
survive fires now burn and die, and will not be replaced for over 100
years, eliminating sources of seeds and turning forests into
shrubfields. Scientists and land managers are also concerned that the
natural occurrence of wildland fire in grassland ecosystems has been
altered by invasive species, such as cheatgrass, that have replaced
native vegetation. Furthermore, communities in the interface of
wildland areas that develop into areas where there are uncharacteristic
fuel accumulations may experience exacerbated effects of wildland
fires.
In 2001, in response to one of the worst fire seasons in over 50 years,
the Departments of Agriculture and the Interior reiterated the
principles in the Federal Wildland Fire Management Policy and began
working with state and local agencies and tribal governments to develop
an interagency National Fire Plan to coordinate federal, state, local,
and tribal efforts. Together, the policy and plan offer a new approach
to wildland fires by broadening the emphasis to include reducing
vegetation, or fuels, and reintroducing fire, where possible, to
restore ecosystems to more resilient fire-adapted conditions. The
reintroduction of fire to certain federal lands does not mean, however,
that all fires will be allowed to burn without management. Currently,
there are two approaches to wildland fire management. First, all
unplanned wildland fires are suppressed or are managed--given favorable
weather conditions--to achieve beneficial effects to resources. Second,
wildland fire management activities also include the reduction of fuels
to protect communities and maintain or improve ecological conditions of
the land. Fuel reduction activities include mechanical methods such as
chainsaws, chippers, mulchers, and bulldozers, and prescribed burns.
Prescribed burns may be set to restore or maintain desired vegetation
conditions.
The degree to which fire can be reintroduced to different forest and
rangeland ecosystems depends on the risk fire poses to environmental
resources and ecosystems. Under the National Fire Plan, land managers
are to identify ways to reduce the risk to communities and ecosystems
from wildland fire. Risk, according to the National Academy of
Sciences, involves hazardous events or conditions and the potential
loss of or damage to something of value because of the hazard. In the
case of wildland fires, the hazard involved is not only the fire, but
also the excess vegetation, or fuel, that has accumulated or been
altered on federal lands. A primary way to lower risks involves
reducing the amount, type, or continuity of vegetation available to
burn. The National Fire Plan applies to several federal agencies that
manage public lands and wildland fires, including the Forest Service
within the Department of Agriculture and the Bureau of Land Management,
National Park Service, Bureau of Indian Affairs, and the Fish and
Wildlife Service within the Department of the Interior. These agencies
are all members of the Wildland Fire Leadership Council, formed in 2002
to support and coordinate implementation of the National Fire Plan and
Federal Wildland Fire Management Policy. Two of these agencies--the
Forest Service and the Bureau of Land Management (BLM)--manage
approximately 450 million acres or 60 percent of the nation's federal
land.
In managing the effects of wildland fires, the Forest Service and BLM
face a second type of risk--that the actions they take to reduce fuels
and to restore ecosystems may damage additional resources such as
species, habitat, or water, whereas if they do not take action, the
effects of a future fire may be exacerbated. Consequently, the
agencies' assessment of the potential effects of their activities
involves weighing the risk of action to reduce fuels against the risk
of doing nothing. Under the National Environmental Policy Act (NEPA) of
1969, agencies generally evaluate the likely environmental effects of
projects they are proposing using a relatively brief environmental
assessment, or, if the action would be likely to significantly affect
the environment, a more detailed environmental impact
statement.[Footnote 2] One purpose of this analysis is to ensure that
agencies have available detailed information concerning potentially
significant environmental impacts to inform their decision making. The
Forest Service and BLM typically conduct such analyses at two levels:
the entire national forest or BLM land unit, which can encompass
several broad landscapes, and the more specific project level, which
addresses smaller areas within the landscape.
Concerns about the severity of recent wildland fires and their
environmental effects have led to increased efforts to reduce fuels on
federal lands, which culminated in the enactment of the Healthy Forests
Restoration Act (HFRA) in December 2003.[Footnote 3] There has been
considerable disagreement over the extent of environmental effects of
wildland fire on federal lands, and what, if anything, to do about
them. The timber industry and other groups advocated increasing the use
of mechanical tree thinning and timber sales to reduce the vegetation
accumulating on the nation's forests and rangelands. Critics of this
approach, which included environmental groups, cited its potentially
detrimental effects on environmental resources, particularly large old
trees. Both the advocates and the critics generally agreed on the
actions needed to address risk to communities; however, there is little
agreement as to what steps, if any, should be taken to reduce the risk
posed to ecosystems. In this context, you asked us to (1) describe the
effects wildland fires have on environmental resources and ecosystems,
(2) assess the information the Forest Service and BLM gather on the
extent of environmental effects of wildland fires, and (3) assess the
approaches the Forest Service and BLM take to assess the risk to
environmental resources from wildland fires and the vegetation that has
accumulated or been altered on federal lands.
To describe the effects that wildland fires have on environmental
resources and ecosystems and assess the information the Forest Service
and BLM gather on the extent of these effects, we designed a survey of
local federal land managers who maintain this data. Our survey
contained questions about 20 wildland fires that we randomly selected
from a universe of 614 wildland fires. Some of the questions required
that the land managers provide their opinions of conditions or make
predictions about the future effects of a wildland fire rather than
providing data about effects that had already occurred, and therefore
there is a greater amount of uncertainty regarding the accuracy of
these responses. We identified the 614 fires through Forest Service and
BLM reports completed in fiscal years 2000 through 2002 on the
emergency actions needed to stabilize areas burned in the fires. We
used these reports to identify our sample because the reports are
developed for wildland fires that the agencies determined are likely to
have had considerable environmental effects. We conducted a random
sample of these 614 fires to ensure that we had a range of small,
medium, and large wildland fires.[Footnote 4] To gather information on
the approaches the Forest Service and BLM use to assess the risks to
environmental resources posed by wildland fire, we reviewed federal
wildland fire policies, the National Fire Plan, and agency guidance and
planning and project documents. We also interviewed federal and state
agency officials, scientists from several of the Forest Service's
research stations, and university experts on fire and fire effects. We
attended two national conferences on fire issues and visited national
forests and BLM state offices in eight western states that had
experienced large fires. We conducted our review from April 2003
through April 2004 in accordance with generally accepted government
auditing standards. Appendix I provides further details about the scope
and methodology of our review.
Results in Brief:
The 20 wildland fires that we surveyed burned over 158,000 acres of
federal land and had complex, wide-ranging, and sometimes
contradictory, effects on both individual resources, such as trees and
streams, and ecosystems. For example, the Missionary Ridge fire in
Colorado burned almost 50,000 acres. The loss of trees and vegetation
in some of the burned areas, as well as chemical and physical changes
in the soil, has caused increased flooding and debris flows in local
streams, which has affected water quality in the short term. However,
in the long term, land managers indicated that even dramatic changes to
streams in other burned areas would not be detrimental. Of the 20
wildland fires in our survey, the land managers viewed the effects of
the wildland fires as adverse, neutral, or beneficial, depending on a
number of factors, including the short-and long-term time frames in
which the effects were described, and the type and condition of the
vegetation that had existed in the burned area. The managers also
reported that the 20 fires had effects across broad landscapes and that
these effects varied in severity.
The Forest Service and BLM gather specific information and data on the
effects of some individual wildland fires on environmental resources,
such as soil erosion or acres of trees burned, for the purpose of
stabilizing burned areas. However, they do not gather comprehensive
data on the long-term severity of wildland fire effects on broad
landscapes and ecosystems--that is, on large areas that may involve one
or more burns--because they do not have a monitoring plan to gather
landscape data across fires. Wildland fires can have varying effects
over time and space, and, as a result, it is important that the
agencies have comprehensive data to monitor the progress of the their
actions to restore fire-adapted ecosystems, a goal of the National Fire
Plan. The agencies recently developed a monitoring framework that
includes fire severity, but this plan has not yet been implemented.
Data on severity would help the agencies to assess whether, over time,
fires in forest and rangeland ecosystems are burning with more or less
severe effects and whether the ecosystems are being restored to more
fire-adapted, or resilient, conditions. Without the ability to identify
the broad landscape effects of fire on vegetation conditions as fires
occur, the agencies will have difficulty showing whether they have met
their identified desired conditions and whether different ecosystems
are becoming more or less resilient to fire.
The Forest Service and BLM, when planning fuel reduction activities, do
not have a systematic approach that allows them to assess the risks of
environmental effects from wildland fires at the landscape level or the
specific project level. As a result, the agencies do not systematically
assess the risks that fires pose to different environmental resources
or ecosystems or the risks of taking no action on fuel reduction
projects, although the National Fire Plan directs them to target their
fuel reduction activities in areas that face the greatest losses. At
the landscape level, the Forest Service and BLM do not have a common
framework for assessing the risk of fire to environmental resources and
ecosystems as part of their fuel reduction efforts because they and the
Congress have placed high priority on assessing areas that threaten
communities. The agencies have not focused exclusively on communities,
as several of their field offices have begun efforts to assess the
risks of environmental and ecosystem effects in planning their fuel
reduction activities. Without a systematic approach to assessing risk
to ecosystems at the landscape level, the agencies cannot effectively
target their fuel reduction activities to protect areas that face the
greatest losses, or, conversely, identify areas that can benefit from
the reintroduction of fire. To formalize a common framework for such
risk assessments, the Forest Service and BLM could use the experience
of other agencies that conduct risk assessments, such as the Federal
Emergency Management Agency or the Environmental Protection Agency, as
well as the experience of those field offices that have independently
conducted such assessments. In assessing the risks associated with
individual projects, the Forest Service and BLM have in some instances
assessed, under NEPA, the risk of acting to reduce fuels versus the
risk of not doing so. We reviewed 10 of the agencies' assessments and
determined the agencies did not systematically assess the risks of
taking or not taking action to reduce fuels. Agency guidance is not
specific about how this assessment should be performed and whether
these analyses should be contained in NEPA documentation. The agencies
have opportunities to specify how the risks of not reducing fuels
should be assessed and where this assessment should be documented. The
agencies have developed interim guidance for implementing the Healthy
Forests Restoration Act, but the guidance does not go far enough in
describing the analysis needed for showing the effects of not reducing
fuels. Also, the agencies and the Council on Environmental Quality
(CEQ) are developing model environmental assessments and lessons
learned from a demonstration program developing model environmental
assessments for fuel reduction projects, in which the agencies
participated. Without a risk-based approach at the project level, the
agencies cannot make fully informed decisions about which effects and
project alternatives are more desirable.
We are making recommendations to the Secretaries of Agriculture and the
Interior to help ensure that the Forest Service and BLM develop (1) the
information needed to better understand the full extent of
environmental and ecosystem effects from wildland fires, (2) a
systematic framework for the assessment of risks at the landscape level
to target where fuel reduction activities need to occur, and (3)
specific guidance on a risk-based approach to make trade-offs among the
environmental effects of acting to reduce fuels or doing nothing.
In commenting on a draft of this report, the Departments of Agriculture
and the Interior stated that the report provided a thorough analysis of
a complex set of issues. They agreed that information on the long-term
effects of fire is needed and noted that on May 18, 2004, they approved
a monitoring framework that includes such information. They also agreed
that prioritization of fuel reduction projects can be improved but
expressed a number of concerns about our recommendation that they
develop a systematic risk-based approach to help prioritize projects.
Finally, they did not agree with our recommendation that they provide
specific guidance on the assessment of the effects of not taking action
to reduce fuels. CEQ also provided comments on this recommendation,
stating that we should not imply that CEQ's guidance to help develop
fuel reduction projects was meant to discuss risk analysis and the
risks of not taking action to reduce fuels. While we made modifications
to our report to address these concerns and to clarify our
recommendations, we continue to believe that our recommendations are
warranted.
Background:
Wildland fire is an inevitable natural ecological disturbance that has
helped to shape ecosystems over time. Fires are driven by climate and
weather conditions, topography, and fuels--including trees, brush,
grasses, dead leaves and needles, and other material that will burn.
Thousands of fires are started each year by natural causes, such as
lightning, or human causes, such as arson. These fires burn millions of
acres of state and federally owned land (see fig. 1).
Figure 1: Number of Fires and Acres Burned, 1960-2002:
[See PDF for image]
[End of figure]
Although fire is a natural component of many ecosystems, and although
humans have used fire for land and resource management purposes for
thousands of years--such as creating improved pasture for animals and
improved land for agriculture--fire can be unpredictable and
potentially destructive. The potential destructiveness of fire is a
particular concern for the growing number of communities on the fringe
of wildland areas that are prone to fire. These communities create a
wildland-urban interface, where houses and other infrastructure are in
or near wildland fuels. Because fires can have dramatic social,
economic, and environmental effects, land management agencies and
federal land managers have sought to suppress fires for much of the
twentieth century. In particular, large, intense fires in 1910 focused
federal policy on suppression to prevent damage to ecological
resources. Suppression, in combination with land management activities
such as building roads and trails, grazing, and increasing development
near public lands, has excluded fire from ecosystems and caused the
uncharacteristic accumulation of vegetation in some forest and
grassland ecosystems. In 2000, the Forest Service and BLM completed a
national study of fuel conditions called the Coarse-Scale Analysis,
which estimated that 182 million acres of the nation's land have an
uncharacteristic buildup of fuels.[Footnote 5] The analysis produced
categories of vegetation conditions ranked as low, medium, and high.
The categories, called fire regime condition classes, represent the
increasing accumulation and alteration of vegetation conditions and the
potential for uncharacteristic wildland fire and its effects. (See app.
II for a detailed description of the analysis.) In 2002, the agencies
updated the analysis for the western states, estimating that almost 183
million acres in western states alone have highly altered vegetation.
Based on additional analysis, the agencies estimated that the amount of
highly altered vegetation nationwide could vary from 90 to 200 million
acres. In the 2002 analysis, the agencies estimated that 99 million
acres of Forest Service and BLM lands in western states have highly
altered vegetation. Refinement of the analysis for the nation is
expected to be completed in 2005.
The National Fire Plan[Footnote 6] recognizes the need for restoring
historic, or characteristic, vegetation conditions as an important way
to reduce the risks of wildland fire and its effects. Under historic
conditions, each vegetation type has a characteristic fire "regime," in
which the vegetation and species have adapted to, and benefit from, the
kind of fires that occur there. Fires that occur in a given fire regime
display similar fire behavior, which refers to how frequently fires
burn, how intensely they burn, and how large they grow. Furthermore,
the effects of fires in different fire regimes can be more or less
severe, depending on the types of fires that typically burn there. For
example, ecosystems such as ponderosa pine forests benefit from and are
sustained by the frequent occurrence of less intense fires to remove
brush and small trees, which allows the large trees to survive and
grow. The severity of effects of these fires on resources and the
ecosystem are usually low or moderate. On the other hand, other
ecosystems, such as lodgepole pine forests, rely on less frequent but
more intense fires to remove all the trees and regenerate a new stand
from seeds dropped by fire-adapted cones. These fires are typically
intense, but they are characteristic of the ecosystem and are needed to
sustain it.
In 2001 and 2002, as part of the National Fire Plan, the federal
agencies, states, and others involved in wildland fire management
developed a 10-year strategy and implementation plan to reduce the
risks of wildland fire to communities and ecosystems. The strategy
established four broad goals for wildland fire management: (1)
improving fire prevention and suppression for those areas that need it;
(2) reducing hazardous fuels, using both natural and managed fire or
mechanical means; (3) restoring fire-adapted ecosystems, both by
reducing fuels and rehabilitating burned areas; and (4) promoting
community assistance to help conduct all these fire management
activities. The implementation plan established specific measures for
showing progress toward each of the goals.[Footnote 7]
Reducing hazardous fuels is one of the key tools for reducing the risks
of wildland fires. Evidence from fires, such as that shown in figure 2,
encourages managers and scientists to believe that areas treated to
reduce vegetation can help to slow down the progress of wildland fires
that occur; in addition, the evidence leads managers to believe that
treated areas do not suffer as severe effects from burning as they
would without the treatment.[Footnote 8] In addition, researchers have
conducted modeling that indicates strategically placed fuel reduction
areas can slow the spread of wildland fire across a landscape.
Empirical confirmation is needed, although some forests, such as the
forests in the Sierra Nevada, are working to apply these ideas in their
fuel reduction treatments. Debate continues not only over the
effectiveness of treatments, but over the extent and duration of
treatments needed. These are areas that federal and university
researchers continue to pursue through the Joint Fire Science
Program[Footnote 9] and other research programs. Despite uncertainties
related to the effectiveness of fuel reduction treatments, federal and
other wildland fire managers believe that they know enough to proceed
with treatments in particular areas while research is completed.
Figure 2: Effects of Wildland Fire in Treated and Untreated Areas
Burned by Wildland Fire:
[See PDF for image]
[End of figure]
The Federal Wildland Fire Management Policy requires that federal lands
with burnable vegetation have a fire management plan. Of the 750
million acres managed by the Departments of Agriculture and the
Interior, the Forest Service and BLM manage 453 million acres of forest
and rangeland. Although the Forest Service manages most of the federal
forested land in the nation--about 192 million acres--about 55 million
acres of BLM's 261 million acres are forested, while the remainder
contain grass and shrublands.[Footnote 10] A fire management plan
produced by each forest or BLM field office establishes the objectives,
strategies, and resources needed to carry out the fire program for that
office. The plan divides a forest or BLM field office into smaller fire
management units for which fire management strategies, including
suppression, prescribed fire, wildland fire use, and nonfire fuel
treatments, are coordinated. The forests and BLM offices--in
conjunction with other federal agencies--have been directed to complete
updated fire plans in 2004.
The Federal Wildland Fire Management Policy also states that each
forest and BLM field office should base its fire plan on its land
management plan. Both the Forest Service and BLM manage their lands for
multiple uses, including timber production, wildlife, recreation, and
wilderness uses. Under the National Forest Management Act (NFMA), the
law that directs the planning of national forests in the Forest
Service, all of the 155 national forests have land and resource
management plans for the lands they manage.[Footnote 11] Generally,
these plans divide a forest into smaller management units with specific
desired conditions to meet the agency's objectives for the different
resources in that area. Similarly, BLM field offices, which are
organized under state offices in 12 western states, develop resource
management plans under the Federal Land Management Policy Act for the
lands they manage. Similar to the national forests' plans, these plans
identify the specific desired conditions that will meet the agency's
objectives in that area. During the next 8 years, over half of the
forests will be updating their land and resource management plans; BLM
offices are also in the process of updating their resource management
plans. Although many of the existing plans included little or no
discussion of wildland fire and its effects, vegetation and fuel
conditions, or the tools for managing wildland fire, the new plans will
discuss these as appropriate. Currently, each agency's regulations
require an environmental impact statement to accompany a plan
revision.[Footnote 12]
To implement their land and resource management plans, the agencies
carry out specific projects--addressing, for example, fuel reduction,
timber sales, grazing, habitat improvement, and recreation. Because
these projects may cause environmental effects, the agencies generally
carry out either an environmental assessment, which is a less detailed
analysis, or an environmental impact statement for their proposed
projects. These analyses may consider different approaches for carrying
out a project--called action alternatives. The agencies may also
consider an alternative that involves taking no action--called the no-
action alternative. In developing their analyses, the agencies are
required to disclose the potential environmental effects of
alternatives.
Wildland Fires Have Wide-Ranging Effects on Environmental Resources and
Ecosystems, Depending on a Range of Factors:
While they burn and afterward, wildland fires have dramatic effects on
environmental resources and ecosystems, including the production of
large amounts of smoke, the burning of trees and other vegetation, and
the erosion of soil into streams and lakes. However, fires can also
benefit resources by recycling soil nutrients, renewing vegetation
growth, and adding material to streams that improves spawning habitat
for fish. The 20 fires included in our survey highlighted the complex,
wide-ranging--and sometimes contradictory--effects of fire on both
individual resources, such as trees and streams, and ecosystems. For
the 20 wildland fires in our survey, the land managers viewed the
effects of the wildland fires as adverse, neutral, or beneficial,
depending on a number of factors, including when the effects were
described--in the short term or the long term--and the type and
condition of the vegetation in the area that burned. The managers also
reported that the 20 fires had effects across broad landscapes and that
these effects varied in severity. The wildland fires in our survey
burned over 158,000 acres of federal land in 10 states: Arizona,
California, Colorado, Idaho, Louisiana, Montana, Nevada, Oregon, Utah,
and Wyoming, with as few as 243 acres and as many as almost 50,000
acres burning in one fire. (See app. III for the definition of severity
used in our survey and app. IV for a detailed description of our survey
results.):
Fire Effects on Individual Resources Vary in the Short and Long Term:
Fire effects can be considered as adverse, neutral, or beneficial
depending, in part, on which resource is evaluated and the time frame
over which the effects are considered. Fire effects are often described
at three times: (1) immediately after the fire; (2) in the short term,
which lasts from 1 to less than 10 years after the fire; and (3) in the
long term, which lasts 10 years or more after a fire.[Footnote 13]
Unlike fire damage to homes, ecological damages from fire are more
difficult to determine immediately after the fire because burned areas
look devastated, even when these conditions are part of the natural,
fire-adapted cycle. For example, although large, intense fires can kill
vegetation in the burned area and generate substantial smoke, some
vegetation, such as aspen and native grasses, regrows quickly from root
systems. Also, although fires can kill individual animals in the short
term, in the long term, many species are attracted to burned areas
because of increases in food sources from new plant growth, increased
numbers of insects and other prey, or because of increased denning or
nesting habitat that dead trees provide. Figure 3 shows, conceptually,
the effects and recovery of vegetation after a high-severity wildland
fire over the short and long term.
Figure 3: Conceptual Short-and Long-Term Effects on Vegetation After a
High-Severity Wildland Fire:
[See PDF for image]
[End of figure]
When we surveyed Forest Service and BLM officials about the effects of
the 20 fires that occurred on federal lands, land managers consistently
responded that fire effects would be less adverse in the long term than
in the short term for each ecological resource, even though their
responses differed across resources. Officials identified whether the
fire had an adverse, neutral, or beneficial effect on each of several
resources in both the short and long term. As shown in table 1, while
many land managers in our survey indicated these fires would have
adverse effects on individual resources in the short term, fewer
responded that the effects would be adverse in the long term. A
discussion of the effects on each of the individual resources follows
the table.
Table 1: Assessment of Overall Effects on Individual Resources in the
Short and Long Term:
Air: Resource and time period: Short term;
Beneficial: 1;
Neither beneficial nor adverse: 7;
Adverse: 9;
No basis to judge, not applicable, or not answered: 3.
Air: Resource and time period: Long term;
Beneficial: 1;
Neither beneficial nor adverse: 1;
Adverse: 2;
No basis to judge, not applicable, or not answered: 16.
Threatened and endangered species habitat: Resource and time period:
Short term;
Beneficial: 4;
Neither beneficial nor adverse: 3;
Adverse: 3;
No basis to judge, not applicable, or not answered: 10.
Threatened and endangered species habitat: Resource and time period:
Long term;
Beneficial: 3;
Neither beneficial nor adverse: 8;
Adverse: 1;
No basis to judge, not applicable, or not answered: 8.
Other species' habitat: Resource and time period: Short term;
Beneficial: 5;
Neither beneficial nor adverse: 5;
Adverse: 10;
No basis to judge, not applicable, or not answered: 0.
Other species' habitat: Resource and time period: Long term;
Beneficial: 9;
Neither beneficial nor adverse: 6;
Adverse: 4;
No basis to judge, not applicable, or not answered: 1.
Soil: Resource and time period: Short term;
Beneficial: 5;
Neither beneficial nor adverse: 5;
Adverse: 10;
No basis to judge, not applicable, or not answered: 0.
Soil: Resource and time period: Long term;
Beneficial: 4;
Neither beneficial nor adverse: 9;
Adverse: 7;
No basis to judge, not applicable, or not answered: 0.
Vegetation: Resource and time period: Short term;
Beneficial: 10;
Neither beneficial nor adverse: 0;
Adverse: 10;
No basis to judge, not applicable, or not answered: 0.
Vegetation: Resource and time period: Long term;
Beneficial: 8;
Neither beneficial nor adverse: 4;
Adverse: 6;
No basis to judge, not applicable, or not answered: 2.
Water and watersheds: Resource and time period: Short term;
Beneficial: 11;
Neither beneficial nor adverse: 0;
Adverse: 9;
No basis to judge, not applicable, or not answered: 0.
Water and watersheds: Resource and time period: Long term;
Beneficial: 6;
Neither beneficial nor adverse: 9;
Adverse: 5;
No basis to judge, not applicable, or not answered: 0.
Source: GAO survey of Forest Service and BLM land managers.
Notes: Because the officials provided answers about the effects of a
fire on each resource, the columns do not add to 20.
The responses are based on the opinions of land managers.
[End of table]
Effects on air: Although officials reported that nine fires in our
sample had adverse short-term effects on the air, only two expected
long-term adverse effects on air quality, while many did not indicate
what long-term effects these fires had on the air. For example,
although the Pony Express II fire in Nevada released an estimated 54
tons of particulate matter into the air when it burned, BLM officials
did not expect any long-term effects from this fire on air quality
because the burned area is far from homes or towns and there are no
nearby sources of air pollution that might have a cumulative effect.
Effects on threatened and endangered species habitat: Agency officials
reported that 10 of the fires in our sample had no identifiable effect
on threatened and endangered species habitat in the short term.
Similarly, the majority of the fires had either no identifiable long-
term effect on the habitat of these species or had a neutral effect.
Eight threatened and endangered species inhabited the areas covered by
5 fires in our sample, including the Canada lynx and the Northern
spotted owl. (See table 10 in app. IV.) Officials indicated that
although none of the fires in our sample posed a threat to the survival
and recovery of a threatened or endangered species population in the
short term, these 5 fires had at least some local impact on a
threatened or endangered species or its habitat. Fires have complex
effects within and among populations of endangered species because
their effects on habitat can both negatively and positively influence
their chances of survival. For example, a nearly 2,500 acre fire in
Louisiana's Kisatchie National Forest had a negative effect on the red-
cockaded woodpecker's nesting habitat, while improving its foraging
habitat by thinning vegetation--a factor the Forest Service official
reported is likely to aid in its recovery. During site visits, Forest
Service officials in Montana told us that the effect of a wildland fire
on endangered fish, such as the bull trout, depends more on whether the
affected streams are contiguous to other streams than on the fire
itself. Locally, some fish may be killed, but if streams are well
connected, other fish can find refuge by migrating away until the fire
is over and then returning to recolonize burned areas. On the other
hand, isolated fish populations living in an environment without these
critical stream linkages are likely to be very vulnerable to fire. For
example, in Arizona after the Aspen fire in 2003, the Fish and Wildlife
Service removed the endangered fish, the Gila chub, from isolated
reaches of Sabino Creek near Tucson to prevent it from being killed by
potential runoff from burned lands.
Effects on other species' habitats: Agency officials reported that 10
fires had adverse effects on other species' habitat in the short term,
while 5 fires had beneficial effects. In some cases, officials
indicated that the loss of vegetation caused a loss of cover and
habitat for species such as the sage grouse, which is a species that
concerns land managers. However, officials stated that fires had
beneficial effects on grasses by increasing their productivity, in turn
providing forage for grazing animals. In the long term, officials
reported that 9 fires had a beneficial effect on species' habitat,
while 6 had neutral effects. For example, officials stated that
although short-term effects may be adverse, the return to a historic
fire regime increased the diversity of vegetation and would ultimately
help species like the snowshoe hare.
Effects on soil: While officials reported that 10 fires had adverse
effects on soil in the short term, they reported that 9 fires had
neutral effects in the long term. For example, officials indicated that
the short-term loss of vegetation cover after the Horse Creek fire
would cause soil erosion and loss. In the long term, officials reported
that most effects on soil would diminish, although an official reported
that soil erosion after the Pony Express II fire would decrease soil
productivity in intensely burned areas, and another official indicated
that soil productivity would be increased because of increased organic
matter released in the Sheep Mountain fire.
Effects on vegetation: Officials reported that, in the short term, 10
fires had beneficial effects on vegetation, while 10 fires had adverse
effects. For example, BLM officials described the mix of burned and
unburned areas within the perimeter of the Sheep Mountain fire in
Wyoming as beneficial because it created a mosaic of vegetation types
of different ages, with more grasses growing in burned areas. After
another fire, however, officials stated that the fire had removed
native vegetation and allowed the spread of cheat grass. In the long
term, officials viewed 8 fires as having beneficial effects, while 6
had adverse effects. For example, officials described the Missionary
Ridge fire as helping to return the long-term balance of different
vegetation. Officials indicated that other fires would increase the
chance of invasive species to spread.
Effects on watersheds: Nine of the fires in our sample had adverse
effects on water and watersheds in the short term, while 11 had
beneficial effects. In the long term, officials reported that 9 fires
will have neutral effects and 5 fires will likely cause adverse effects
to water and watersheds. Of the 20 fires, 3 severely burned 10
watersheds that supply domestic water to municipalities or towns, and
in two cases, officials said the fires had a negative effect on water
quality that lasted from 3 to 5 years. In areas burned by 8 fires,
floods, debris flows, or landslides occurred within the fire perimeter,
yet the long-term effects of fire on water and watersheds are expected
to be more neutral as these effects subside. For example, although the
Horse Creek fire in Oregon resulted in a short-term increase in the
sediment in stream channels, BLM officials reported that the
sedimentation will decline as vegetation recovers and sediment
deposited into the channels will be moved downstream by natural stream
flows in the long term.
Effects of Wildland Fire Vary Depending on Topography, Climate and
Weather, and Vegetation Conditions:
Researchers and land managers describe fire effects using levels of
severity: low severity, moderate severity, and high severity. (See app.
III for the definition of severity used in our survey.)[Footnote 14]
The severity of effects depends on the intensity of the fire--the
amount of heat released in a fire--and its duration in relation to the
historic fire regime. The intensity of a fire depends on its
topography, climate and weather, and vegetation or fuels. First,
topography includes locally unique site properties, such as the slope
of the terrain, the direction in which the ground slopes, and the soil
moisture, each of which affect how intensely a fire burns. For example,
fires burn faster and more intensely on steep slopes, which allow a
fire to move uphill driven by winds, and on south-facing slopes, which
are drier than north-facing slopes. Second, climate and seasonal
weather conditions such as drought cycles and high winds also determine
how a fire will burn and how severe the effects of burning will be.
Climate and weather also determine the extent to which storms occur
after a fire; stronger and more frequent storms can result in increased
erosion and landslides. Finally, the type and condition of vegetation
in an area determines how much "fuel" is available to burn and thus how
intense a fire will be and how severe its effects may be. For example,
rangelands have less vegetation, and therefore lower amounts of fuel to
burn, than forested areas. Furthermore, areas with accumulated
vegetation have more fuels to burn than they would under more natural
conditions.
Whether or not the environmental effects of a wildland fire are
considered as adverse, neutral, or beneficial depends on the degree to
which vegetation conditions have been changed from the historic fire
regime in an area. For example, a fire that burns in a high-elevation
forest filled with spruce and fir trees--a fire regime that
historically has fewer but more intense fires, with more severe
effects--is less likely to have adverse effects to the environment and
that ecosystem than an uncharacteristically intense wildfire that burns
in an ecosystem in which frequent, low-intensity fires occurred
historically, such as ponderosa pine. Forest Service and BLM scientists
and land managers describe areas in which vegetation has accumulated
abnormally or has been altered as having uncharacteristic vegetation
and fuel conditions and areas in which vegetation has accumulated at
normal levels or not been altered as having characteristic vegetation
and fuel conditions. Likewise, they describe fires that are similar to
those that occurred under an area's historic fire regime as
characteristic and those that are not similar to the historic fire
regime as uncharacteristic. Characteristic fires tend to have effects
on the environment and ecosystems that are appropriately severe for
that vegetation type and fire regime, and which are therefore not
considered negative, whereas uncharacteristic fires usually have
unexpectedly severe environmental effects, which are often considered
negative. Of the 20 fires included in our survey, 10 burned with
predominantly characteristic effects, 3 burned with a mix of
characteristic and uncharacteristic effects, and 7 burned with
predominantly uncharacteristic effects. Table 2 shows that of the 10
fires with predominantly characteristic effects, 6 occurred in areas in
which vegetation conditions experienced low levels of alteration or
accumulation, and the remaining 4 occurred in areas with moderate
levels of vegetation alteration or accumulation. Fires that resulted in
both mixed and uncharacteristic effects occurred only in areas in which
vegetation conditions were moderately or highly altered or accumulated
(see table 2).
Table 2: Description of Characteristic and Uncharacteristic Fire
Effects and Vegetation Conditions:
Fire: Sheep Mountain (Wyoming);
Federal acres burned: 21,370;
Vegetation conditions, amount of alteration: Low;
Characteristic.
Fire: Burgdorf Junction (Idaho);
Federal acres burned: 17,207;
Vegetation conditions, amount of alteration: Low;
Characteristic.
Fire: Elko 13/#3 (Nevada);
Federal acres burned: 12,544;
Vegetation conditions, amount of alteration: Medium;
Characteristic.
Fire: Abert (Oregon);
Federal acres burned: 10,100;
Vegetation conditions, amount of alteration: Medium;
Characteristic.
Fire: Stables (California);
Federal acres burned: 4,162;
Vegetation conditions, amount of alteration: Low;
Characteristic.
Fire: Horse Creek (Oregon);
Federal acres burned: 1,839;
Vegetation conditions, amount of alteration: Low;
Characteristic.
Fire: Pony Express II (Nevada);
Federal acres burned: 1,806;
Vegetation conditions, amount of alteration: Low;
Characteristic.
Fire: Crusoe (Nevada);
Federal acres burned: 1,386;
Vegetation conditions, amount of alteration: Low;
Characteristic.
Fire: Elk Mountain (Montana);
Federal acres burned: 667;
Vegetation conditions, amount of alteration: Medium;
Characteristic.
Fire: Y-Mountain (Utah);
Federal acres burned: 437;
Vegetation conditions, amount of alteration: Medium;
Characteristic.
Fire: Missionary Ridge (Colorado);
Federal acres burned: 49,990;
Vegetation conditions, amount of alteration: High;
Mixed.
Fire: Rough Diamonds (Idaho);
Federal acres burned: 7,268;
Vegetation conditions, amount of alteration: Medium;
Mixed.
Fire: Springer (Arizona);
Federal acres burned: 666;
Vegetation conditions, amount of alteration: Medium;
Mixed.
Fire: Crimson Clover (Idaho);
Federal acres burned: 14,466;
Vegetation conditions, amount of alteration: Medium;
Uncharacteristic.
Fire: Boulder Hills (Montana);
Federal acres burned: 5,400;
Vegetation conditions, amount of alteration: High;
Uncharacteristic.
Fire: Cow Hollow (Oregon);
Federal acres burned: 3,022;
Vegetation conditions, amount of alteration: High;
Uncharacteristic.
Fire: Longleaf Vista (Louisiana);
Federal acres burned: 2,497;
Vegetation conditions, amount of alteration: High;
Uncharacteristic.
Fire: Tipton Ranch (Nevada);
Federal acres burned: 2,025;
Vegetation conditions, amount of alteration: High;
Uncharacteristic.
Fire: Hyampom (California);
Federal acres burned: 1,053;
Vegetation conditions, amount of alteration: High;
Uncharacteristic.
Fire: Horse (Idaho);
Federal acres burned: 243;
Vegetation conditions, amount of alteration: Medium;
Uncharacteristic.
Total acres;
Federal acres burned: 158,148.
Source: GAO survey of Forest Service and BLM land managers.
Notes: This sample cannot be projected to all wildland fires.
The responses are based on the opinions of land managers.
[End of table]
As table 2 shows, the third largest fire in our sample, the Burgdorf
Junction fire in Idaho, had characteristic effects. For this reason,
Forest Service officials considered the majority of the effects from
this fire to be beneficial, even though the fire burned more than
17,000 acres of federal land, including areas that provided habitat for
several threatened and endangered species. Overall, the officials
considered the fire effects to support processes for maintaining the
ecosystem, which includes lodgepole pine and Douglas fir forests. For
example, officials stated that the debris flows from the fire
introduced gravel into streams, providing new spawning grounds for
fish. In addition, officials stated that burned areas of the lodgepole
forest were turned into more open stands of brush and grasses,
improving gray wolf and lynx habitat.
Of the three fires that officials identified as having a mix of
characteristic and uncharacteristic effects, one--the Missionary Ridge
fire in Colorado--was the largest fire included in our sample. Forest
Service officials noted that the adverse effects of the fire in the
short term included numerous floods and debris flows, which affected
the water quality of streams supplying water to surrounding
municipalities. They also indicated that in areas where the fire burned
uncharacteristically, long-term adverse effects on streams included
destabilized banks and loss of riparian area. On the other hand, the
officials noted that in areas where the fire burned characteristically,
changes to the streambed and riparian areas would not be adverse over
the long term.
Of the seven fires with uncharacteristic effects, Forest Service
officials identified the smallest fire in our sample--the Horse fire of
243 acres--as having adverse effects on resources. A Forest Service
official reported that this fire in Idaho's Salmon-Challis National
Forest had immediate adverse effects on the vegetation because of the
size of the severely burned area, although he believed that in the long
term, fire-killed trees might benefit the Canada lynx by providing
denning areas. Similarly, BLM officials viewed the 2,025-acre Tipton
Ranch fire in Nevada as having adverse effects on vegetation because
the fire exacerbated the conversion of native plant species such as
grass and sage brush to invasive grasses such as cheat grass. Our
survey showed that the number of acres vulnerable to population by
noxious and invasive plant species such as cheat grass--which competes
with native vegetation and alters the historic fire regime--after the
20 fires increased from about 32,130 acres to about 58,800 acres (83
percent). Several officials raised concerns about the spread of such
invasive species as cheat grass into sagebrush-grass and pinyon-juniper
vegetation types because it is highly flammable, and areas dominated by
it may burn frequently. More frequent fires in such ecosystems may
eliminate the native plants such as sage brush, which is important
habitat for sage grouse.
Fires Have Broad Landscape Effects:
In addition to its effects on individual resources such as soil, water,
and air, fire creates landscape patterns to the extent that it burns
large areas and leaves other areas lightly burned or unburned. As shown
in figure 4, landscapes are geographic areas of varying sizes,
encompassing tens of thousands of acres or more, that may contain
smaller landscapes and interacting and interconnected ecosystems that
are defined by geological, soil, climate, and other physical factors.
Landscapes are separated by natural features, including watersheds,
such as the example in figure 4 from the Interior Columbia Basin, and
encompass different stands of trees and, in some ecosystems, patches of
open areas among the stands of trees. Landscapes may include a mix of
government and private lands and may cross state boundaries.
Figure 4: Relationship of Ecosystem and Landscape Levels:
[See PDF for image]
[End of figure]
The landscape effects of fire include the patterns, or patches, of
vegetation that are burned with varying degrees of severity, including
some that are not burned at all. Under natural conditions, when fires
burn some areas severely and other areas lightly or not at all, they
create irregularly sized openings in vegetation layers, changing the
size, shape, and age of the vegetation in those patches. These
landscape effects, which reflect the degree of environmental change
from burning and affect other environmental resources accordingly, can
be described as being of low, moderate, or high severity. Under natural
conditions, fires with low-severity effects are those that kill the
least amount of vegetation; some burned areas may appear much like the
unburned forest. Fires with moderate-severity effects are those in
which vegetation is killed but some trees remain standing; the
regeneration that occurs after the fire results in stands of trees of
different ages. Fires with high-severity effects are those in which
most of the trees and vegetation are killed over large areas, leaving
open areas in which the tree stands that regenerate will be the same
age. Ecosystems in which vegetation and fire are characteristic, or
fire-adapted, are resilient, and the landscape reflects the functioning
of interdependent plant and animal communities. In ecosystems in which
vegetation and fire are uncharacteristic, fire severity can exceed the
capacity of the ecosystem to regenerate, and the landscape reflects
changes to the plant and animal communities that used to exist there.
Figure 5 shows the range of burn severity patterns attributed to our
sample of 20 fires. Of the 7 fires with uncharacteristic effects in our
survey, the most frequent reason officials cited for a fire to be
considered uncharacteristic was not the size of the fire, but the size
of the patches the fire had burned severely. The fires demonstrate a
wide variety of burn severity patterns within two extremes. While
Wyoming's Sheep Mountain fire severely burned 100 percent of the
federal lands within its perimeter, all of the federal acreage burned
by Oregon's Abert fire burned at low severity.
Figure 5: Acres of Vegetation Burned Lightly, Moderately, and Severely
in 20 Sample Fires:
[See PDF for image]
[A] Percentages may not add to 100 because of rounding.
[B] Percentages do not add to 100 because the method used to measure
burn severity did not always distinguish unburned acres. Acres in the
fire perimeter include only federal acres.
[C] Survey respondent indicated these values are unknown.
[End of figure]
Forest Service and BLM Do Not Gather Comprehensive Information on the
Varied Effects of Wildland Fires on Ecosystems and Landscapes:
Although the National Fire Plan established a goal of restoring forest
and rangeland ecosystems to conditions that are more fire-adapted, and
therefore more resilient to fire, land managers do not have
comprehensive data on the broad landscape effects of wildland fire to
help them monitor these effects over time. While the Forest Service and
BLM gather information on the severity of environmental effects from
individual wildland fires, they do not gather data that captures the
long-term severity of fires across landscapes. Through emergency
stabilization programs, the forests and BLM field offices gather
information on the effects on soils and watersheds to estimate the
likelihood that soil disturbances caused by individual fires will
result in flooding and landslides. However, while the agencies' data
collection efforts include fire histories--that is, the occurrence,
location, and size of fires--they do not have a monitoring plan to
gather landscape data across fires and they do not yet consistently map
the long-term, landscape-level severity of wildland fires. This data
would help the agencies to assess whether, over time, fires in forest
and rangeland ecosystems are burning with more or less severe effects
and whether they are being restored to more resilient, or fire-adapted,
conditions.
Forest Service and BLM Collect Data on the Environmental Effects of
Individual Fires to Help Them Restore and Rehabilitate Burned Lands:
Although the Forest Service and BLM are generally not required to
gather environmental data on the effects of wildland fires, the
agencies' field offices do collect data that indicate the potential for
flooding, erosion, and landslides to occur in the short term after a
fire for the purpose of treating areas that need emergency
stabilization. Both the Forest Service and BLM use multidisciplinary
teams of experts, such as ecologists and soil scientists, to gather and
review data on individual fires that have altered conditions enough to
warrant the emergency stabilization. These teams--called Burned Area
Emergency Response (BAER) teams by the Forest Service and Emergency
Stabilization and Rehabilitation (ESR) teams by BLM--gather data on
vegetation, soils, and stream channels and evaluate burn severity and
erosion hazard potential in areas affected by fire. Both agencies use
this evaluation process to assess the potential emergency and identify
appropriate treatments to stabilize areas and to protect homes and
other values at risk, such as roads. The data are reported and
maintained in fire-specific files in the forest and BLM field offices.
Although the BAER and ESR reports do not address the long-term effects
of fires, the Forest Service and BLM generally collect more extensive
data on the effects of large fires affecting areas where they plan to
conduct rehabilitation or restoration work.[Footnote 15] The Forest
Service uses a variety of funds, including funds raised from salvaging
dead and damaged trees, to pay for rehabilitation work. When the forest
staff identify fire-damaged areas that they wish to rehabilitate, they
gather data on environmental resources for the purpose of developing an
environmental assessment or environmental impact statement for a
salvage sale and associated rehabilitation or restoration projects. For
example, the staff at Bitterroot and Sequoia National Forests
determined that they would conduct rehabilitation projects for very
large fires that occurred on their lands in 2000--the Bitterroot fires
and the McNally fire. In the case of the Bitterroot fires, the forest
staff collected extensive data, including detailed information on soil
and watershed characteristics, vegetation, wildlife and wildlife
habitat, and noxious weeds. During and after the McNally fire in
central California, the Sequoia National Forest staff gathered similar
data on the fire's effects and measured the smoke output from the fire
and its effect on air quality. Because the rehabilitation of rangelands
that have burned often involves seeding the burned sites with fast-
growing grasses to retain soils and forage for wildlife, BLM receives
funding to pay for such work. Field office staff collect data on these
rehabilitation needs as part of their rehabilitation efforts; the data
collected under this program includes data on topography, soils, native
and non-native plants, wildlife habitat, and threatened and endangered
species that inhabit the project area.
Furthermore, some forest and BLM staff, after years in which numerous
fires have occurred, have developed special reports on the effects of
the different fires across the region. For example, the Intermountain
and Northern Regions of the Forest Service[Footnote 16] assessed the
extent and effects of the large wildland fires these areas experienced
during 2000. They collected data on the (1) number of air quality
advisories to communities affected by the many fires involved; (2)
flooding and sediment in streams with native fish species, such as
cutthroat and bull trout; and (3) adverse effects to sage grouse
habitat. In addition, after a number of fires in 1999 burned about 1.7
million acres in portions of four states that comprise the Great Basin-
-Idaho, Nevada, Oregon, and Utah--BLM issued[Footnote 17] a general
report on the fires' combined effects on this large geographic area. As
part of this effort, BLM assessed the role wildland fire has played
over time in the Great Basin--changing some healthy rangeland
ecosystems populated by native plants into systems dominated by annual
weeds, such as cheatgrass. According to the report, because this annual
grass provides little or no cover or food for wildlife, sage grouse
populations had decreased by more than one-third, a factor in the
grouse's possible consideration for listing as a threatened and
endangered species.
Agencies' Data Collection Does Not Address Long-Term Landscape Severity
of Wildland Fire Effects:
Over the long term, as the National Fire Plan and its activities are
implemented, the Forest Service and BLM will need landscape data on
wildland fire effects, including fire severity, to monitor whether they
are restoring and maintaining fire-adapted ecosystems. Because wildland
fires and the severity of their effects vary across different
landscapes and ecosystems, land managers and scientists need data on
the severity of effects to understand whether the severity of fires is
changing, and therefore whether vegetation conditions need to be
managed differently. A monitoring plan would provide the means for
gathering consistent and comprehensive data over the long term on fire
trends and severity. The agencies have started to develop systems and
methods to gather the needed data including: (1) vegetation data, (2)
historical fire and severity data, and (3) current severity data.
However, they do not yet have the full capability to gather and use
these data. When these data are integrated and assessed, they will
provide the Forest Service and BLM with information on the historical
fire regime that occurred in a given area and the expected fire
severity. With such data, the agencies will have a baseline to
determine whether fire severity is changing because of vegetation and
fuel conditions.
While the Forest Service and BLM use several different methods to
gather information on vegetation conditions, they are working to
develop a system of protocols and procedures to gather consistent
nationwide data. Satellite images are currently used to provide data
for individual forests and BLM offices to use in assessing their
vegetation conditions. However, the Forest Service and BLM, as part of
an interagency effort to gather consistent national data on vegetation
conditions and related fuel conditions, are developing a database and
related modeling tools called LANDFIRE to gather satellite data,
interpret it, and compare and validate the data with data from actual
sites on the ground. Satellite images capture data on thousands of
acres or more, providing a landscape-level view of the resources for
land managers. (See app. V for a discussion of the recording and use of
satellite data.) When it is completed, LANDFIRE is expected to provide
land management agencies with maps of their vegetation that, when
combined with data on the physical conditions of the same areas, will
show the natural vegetation cover type that should exist on the areas.
The Departments of Agriculture and the Interior funded the
implementation of the system in 2004 and expect it to be completed for
western states in 2006, for eastern states in 2008, and for the entire
nation in 2009. In the interim, the agencies will use data that are
already available.
In addition to vegetation data, historical fire and severity data are
important for the agencies to understand the landscape severity of
current fires. While the national forest and BLM field office staff
collect historical fire data, including the occurrence, location, and
size of wildland fires that have burned across a landscape, they do not
consistently collect this data or data on the severity of effects for
individual fires. Of the 13 forests and BLM offices we visited during
our review, most were collecting severity data for large fires, but
only one office had these data in a geographic information system
database. A national database of fire history and severity data would
help the agencies identify and monitor the actual effects of fire on
vegetation and ecosystem conditions. While the Wildland Fire Leadership
Council adopted a monitoring framework that includes fire severity in
May 2004, it has not yet been implemented. The agencies are currently
developing cost estimates and specific plans for implementing the
framework.
Finally, while the forest and BLM field offices gather some data on
severity of fires that burn on their lands, the agencies do not
consistently collect data on burn severity that reflect the long-term
effects of fires. The agencies use a mix of ground, aerial, and
satellite data to measure the burn severity of different fires;
however, the satellite data that they typically gather are better
suited to show the short-term effects of wildland fire. The Forest
Service and BLM generally compile satellite images taken a year before
the fire and immediately after it to help them estimate the emergency
stabilization needs of the area. In contrast, another approach to
measuring severity involves comparing one satellite image taken
immediately before a fire and another about a year after to more
accurately estimate the long-term ecological effects of the fire. This
approach, developed by the National Park Service in an effort to
measure the long-term effects of fire, includes factors that affect the
recovery of an area over the long term, such as the number of seeds
that remain in the soil, the proximity of seed sources from unburned
areas, slope, soil moisture, and the amount of erosion that may occur.
Although this approach has been used successfully on Park Service lands
and by some forests and BLM offices, its use in evaluating Forest
Service and BLM lands has not been widely tested. One of the
applications that is still being developed and tested is the ability to
update vegetation maps from the severity maps. This application could
be useful in updating vegetation maps, including those that will be
created by the LANDFIRE system once it is completed. If it is
implemented as planned, LANDFIRE's vegetation maps will be available to
the forests and BLM offices about every 5 years. According to the
National Park Service, this approach has been used to update some of
the parks' vegetation maps, thereby maintaining their usefulness in
tracking progress in managing vegetation conditions and planning
management activities.
The Forest Service and BLM Do Not Systematically Assess the Risks to
Environmental Resources and Ecosystems to Target and Conduct Fuel
Reduction Activities:
Although the National Fire Plan identifies the need to reduce the risk
of environmental and ecosystem effects from wildland fires by targeting
fuel reduction activities to the areas that face the greatest potential
losses, the Forest Service and BLM do not systematically assess the
risks to resources and ecosystems for the purpose of targeting fuel
reduction projects. Because wildland fires affect both large landscapes
and individual resources at specific sites, it is important to assess
risk at both the broad landscape level and the more specific project-
planning level. At the landscape level, the Forest Service and BLM do
not have a common framework that includes the three elements for
assessing risks: hazard, risk, and values. At the project level,
although the agencies have recognized the need to better analyze the
relative risks of undertaking fuel reduction activities versus not
doing so, they do not have a systematic approach to assess these risks.
Because they lack a systematic risk-based approach for targeting their
fuel reduction projects across a landscape and within a project area,
the agencies cannot ensure that they are reducing fuels in areas of
highest risk to environmental resources and ecosystems.
A Risk Assessment Framework Would Systematically Analyze Hazard, Risk,
and Values:
In general, risk assessment is a process for evaluating a natural
hazard, such as wildland fire, as well as the probability of the
hazardous event occurring and the consequences or potential losses that
would result if the event did occur. According to the National Academy
of Public Administration, a systematic approach to risk assessment
involves three elements:
Hazard: A hazard is the potential event, such as a wildland fire,
hurricane, or earthquake, and the conditions that cause it. In the case
of wildland fire, both the fuel conditions that exist and the fire
itself are the hazard. According to scientists and land managers, the
increased vegetation in different ecosystems around the country has
become more continuous and dense, resulting in larger fires that burn
more intensely. For example, the Tyee fire burned 140,000 acres in
Washington in 1994, in part because excess vegetation had grown into
areas that, under natural conditions, would have less vegetation to act
as fuel for the fire.
Risk: Risk is the probability that an event such as a wildland fire
will occur. By mapping the number and location of fires, scientists
have discovered that some areas are more prone to fires than others.
For example, Florida and the western states are more likely to
experience wildland fire than the states along the East Coast and in
the Midwest. Wildland fire ignites either because of lightning strikes,
which occur along storm paths and prominent landscape features, such as
mountain ridges, or because of human activities, such as camping,
logging, agricultural burning, and careless smoking.
Values: In general, values are the things that might be lost or damaged
because of a hazard. In the case of wildland fire, social values that
might be lost include the lives of both civilians and firefighters,
cultural and historical resources, and artifacts and sacred sites.
Economic values that might be lost include property and other
infrastructure, resources such as timber and water, and recreation and
tourism opportunities. In the case of environmental values, however,
wildland fire can both damage and improve different environmental
resources and ecosystems. The values that might be affected include
ecosystems, species and their habitat, air and water quality, and soil
and vegetation.
To prioritize areas needing fuel reduction, all three elements--hazard,
risk, and value--need to be considered and ranked because, for example,
an area with high vegetation hazard may or may not be in an area where
fires are likely to occur, making it a lower priority for treatment.
Furthermore, a high hazard area may not be close to something of value
that would be lost or damaged in a fire, also making it a lower
priority for treatment. The National Fire Plan calls for the Forest
Service and BLM to collaborate with state, local, and tribal entities
in making decisions about what specific areas are in need of fuel
reduction treatment. An assessment of hazard, risk, and values can form
the basis for informing this collaborative approach.
According to the National Academy of Sciences' National Research
Council, a risk assessment supports decisions that need to be
made.[Footnote 18] For this reason, a framework for assessing the risks
of effects from wildland fires would also identify the types of
decisions that need to be made at different organizational levels--the
national, landscape, and project levels--and the risk information
required to make the decisions. For example, in the case of fuel
reduction activities, the types of decisions that need to be made at
the national level include how many resources to allocate per year to
reduce risk. Landscape-level decisions include determining which parts
of the landscape are at the greatest risk of wildland fire and its
potential effects. Project-level decisions include making tradeoffs
among alternatives and their different effects.
Forest Service and BLM Lack a Framework for Assessing Risks to
Ecosystems Across Landscapes to Target Fuel Reduction Activities:
The Forest Service and BLM have not adopted a framework to
systematically assess the risks of environmental effects of fires to
support their fuel reduction efforts. Without addressing the three
elements of a risk assessment--hazard, risk, and value--the agencies do
not have a systematic way to target their fuel reduction activities
across a landscape. The Forest Service and BLM nationwide assessment of
vegetation, or fuel, conditions, conducted in 2000 and updated in 2002,
addressed only one element of a risk assessment--hazard. Because of the
need to prioritize their fuel reduction efforts, some Forest Service
and BLM field offices have conducted assessments that include one or
more elements of a risk assessment. However, these efforts are
informal, incomplete or uncoordinated and therefore do not
systematically address the need to identify and reduce ecosystem risks.
Nationwide Assessment of Vegetation Conditions Considered Only Hazard
Information, Not Risks and Values:
In 2002, the Forest Service and BLM updated the national assessment of
fuel conditions that estimated that about 99 million acres of the
agencies' land in 11 western states were highly altered from historic
vegetation conditions.[Footnote 19] While the assessment also
identified and collected data on fire occurrence, a factor in
determining the probability that a fire will occur, the researchers did
not include these risk data in the assessment because they were
incomplete or inconsistent. In addition, although the assessment
indicated which ecosystems might burn uncharacteristically and
experience uncharacteristically severe effects from fire, it did not
consider other values at risk, such as threatened and endangered
species habitat that might be damaged or lost. A complete risk
assessment--one that includes risk and values--could give national
leaders a better idea of (1) the amount of fuel reduction that needs to
be done per year to reduce the risks of wildland fires at the national
level and (2) the amount of funding that needs to be allocated to
reduce risks. For example, in 2002, a group of researchers involved in
developing the national assessment conducted an independent study of
options for reducing risks of effects from wildland fire through fuel
reduction projects across landscapes and nationally.[Footnote 20] The
study, which included information on the probability of fire occurring
and the values at risk, concluded that reducing risks would require
more work and funding than was allocated in 2002. While the study
results have not been officially confirmed by the multiple agencies and
organizations involved in conducting fuel reduction activities, it is
an example of the risk-based approach needed to target fuel reduction
activities and funding.
In late 2003, recognizing the need for some direction in how to conduct
risk assessments, the Forest Service and BLM issued guidance in
conjunction with the National Association of State Foresters. The
guidance states that there are a number of valid assessment processes
available for the agencies to use, including one approach that involves
mapping data on four factors: (1) fire occurrence, (2) hazard, (3)
values to be protected, and (4) protection capabilities. This guidance,
however, focuses on communities and does not discuss how the risks to
the environmental resources and ecosystems that the agencies are
responsible for managing are to be assessed. It is important that the
risks to environmental resources and ecosystems be assessed in
considering fuel reduction across landscapes because different
approaches are needed to manage environmental resources and ecosystems
in different fire regimes. For example, it can be more difficult to
treat forests and rangelands that burn with low frequency and high
intensity, such as lodgepole pine forests, as opposed to areas that
burn with high frequency and low intensity, such as ponderosa pine
forests.
The Forest Service's Rocky Mountain Research Station has developed a
modeling protocol to assess the risks, as well as the benefits, of fire
by considering three factors: (1) the probability of fire occurrence;
(2) the expected severity of a fire; and (3) the ecological, social,
and economic value ascribed to an area.[Footnote 21] Other agencies
have developed risk assessment frameworks, tailored to their particular
needs, that include hazard, risk, and values and identify the
organizational levels that should conduct the risk assessment. For
example, the Federal Emergency Management Agency provides funding to
state and local governments for hazard mitigation on the basis of their
assessment of (1) the natural hazard and the probability of future
hazardous events, (2) state and local vulnerability to the hazards, and
(3) the potential losses from the hazardous event. In addition, the
Environmental Protection Agency has developed an ecological risk
assessment framework that defines the values at risk on the basis of
desired environmental conditions and can be applied at different
organizational levels depending on the risk problem.
Some Forest Service and BLM Offices Have Conducted Independent Risk
Assessments:
To identify the areas with the highest levels of risk at the landscape
level, some forests and BLM field offices, in conjunction with state
and local governments, have applied their own approaches to assessing
fuel conditions and risks. We reviewed fire planning processes and
documents at 13 forest and BLM offices and found that several offices
had applied risk assessment frameworks that they had either developed
themselves or contributed to developing. Table 3 shows the forest and
BLM offices that we visited and describes the type of assessment they
conducted, including the elements of a risk assessment that were
addressed.
Table 3: Forest Service and BLM Office Assessments and the Risk
Elements Addressed:
National forest: Arapaho-Roosevelt (Region 2, Colorado);
Assessment conducted and elements addressed: Joint effort with state
and other federal agencies included hazard, risk, and communities at
risk. Subsequent land management plan included environmental values,
such as threatened and endangered species and old-growth vegetation.
National forest: Pike San Isabel (Region 2, Colorado);
Assessment conducted and elements addressed: Joint effort with state
and other federal agencies included hazard, risk, and communities at
risk. Land management plan is being revised starting in 2004 and will
include environmental values such as threatened and endangered species
and watersheds.
National forest: Bitterroot (Region 1, Montana);
Assessment conducted and elements addressed: Region conducted an
assessment of hazard, risk, and values, including environmental values
such as threatened and endangered species, water quality, air quality,
and soil condition. The forest's land management plan is being revised,
and the forest will add some of this information.
National forest: Sequoia (Region 5, California);
Assessment conducted and elements addressed: Joint effort with other
forests and federal agencies included hazard, risk, and communities at
risk. Data on environmental values have been collected.
National forest: Wenatchee (Region 6, Washington);
Assessment conducted and elements addressed: An assessment of fire
regime and vegetation conditions began in 2004.
National forest: Payette (Region 4, Idaho);
Assessment conducted and elements addressed: Joint effort with two
other forests included hazard and risk. Another assessment identified
watersheds of concern.
National forest: Coronado (Region 3, Arizona);
Assessment conducted and elements addressed: Assessment of fire regime
and vegetation conditions began in 2004.
BLM office: Colorado;
Assessment conducted and elements addressed: Joint effort with the
state of Colorado included hazard, risk, and values at risk, including
environmental values such as threatened and endangered species.
BLM office: California;
Assessment conducted and elements addressed: Assessment of fire regime
and vegetation conditions began in 2004.
BLM office: Nevada;
Assessment conducted and elements addressed: Joint effort with other
federal and state agencies will assess hazard, risk, and communities at
risk. BLM began its assessment of fire regime and vegetation conditions
in 2004.
BLM office: Oregon/Washington;
Assessment conducted and elements addressed: Assessment of fire regime
and vegetation conditions began in 2004.
BLM office: Idaho;
Assessment conducted and elements addressed: Joint effort with state
and other federal agencies to assess hazard, risk, and values,
including environmental values such as threatened and endangered
species, water quality, air quality, and soil.
BLM office: Arizona;
Assessment conducted and elements addressed: Assessment of fire regime
and vegetation conditions completed in 2003.
Source: GAO.
[End of table]
Although the forests and BLM offices have undertaken independent
efforts, some similarities exist in their approaches. Specifically,
several of the units have attempted to include environment and
ecosystem values in their mapping efforts. For example, Region 1 of the
Forest Service assembled data on species habitat, water quality, soils,
erosion potential, airsheds, and vegetation. The data are mapped, which
provides managers with the location of important resources as they plan
their fuel reduction activities. This team developed a computer program
called the Multi-Resource Information Tool that uses the hazard, risk,
and values information to rank different watersheds by different risk
categories. In a similar way, some of the BLM offices identified the
values present in different management areas, including environmental
resource values. Other BLM offices use a computer program called the
Risk Assessment and Mitigation System to model the protection afforded
these values from their chosen fuel reduction programs.
The purpose of conducting risk assessments at the forest and BLM field
office level is to support decisions about where to target fuel
reduction efforts and where naturally caused fires can be managed and
controlled to achieve resource benefits. The forests and BLM field
offices are required to identify in their fire management plans the
areas in which they will continue suppressing fires, those in which
they will conduct fuel reduction projects, and those in which they can
allow wildland fires to burn to reduce fuels and provide resource
benefits. A risk assessment would help the offices identify these
areas. For example, officials at some of the forests we visited stated
that they wanted to increase their use of wildland fires to achieve
resource benefits. Using wildland fire in this way not only provides
resource benefits, but may also help to limit the overall costs of
suppressing fires. For example, if the Forest Service had been able to
let the Burgdorf Junction fire in Idaho burn, the fire would have
benefited most resources and the agency would have spent much less than
the $26 million used to suppress the fire. The fire, which occurred in
an area identified as having low accumulations of fuel that could have
benefited from a fire to maintain its conditions, was suppressed to
protect a nearby community--a community that would have been protected
by a $1 million fuel reduction project that was considered lower
priority, given funding and the assessment of vegetation conditions.
Although a framework for conducting risk assessments to support the
agencies' landscape-level fuel reduction activities would identify the
organizational level at which the assessment should be conducted,
agency guidance issued in 2002 and 2003 does not direct the forests or
BLM field offices to conduct risk assessments to support their
landscape-level fuel reduction plans. Specifically, guidance issued in
April 2002 directs the Forest Service and BLM to collaborate with other
agencies, including state, local, and tribal agencies responsible for
managing wildland fire and fuels, in planning treatments across
landscapes. This guidance recommends that landscape plans be developed;
however, it does not require a formal risk assessment to support a
plan. In addition, guidance issued in February and March 2003 directs
the forests and BLM field offices to use the methods from the national
assessment of fuel conditions to classify their local lands and fuel
reduction projects according to their alteration from historic fire
regimes. The purpose for classifying projects, in the short term, is to
determine if the agencies are making progress toward reducing the
number of acres of land they manage that have high accumulations of
vegetation. In the long term, the guidance directs that the methods be
used as part of land management planning.
Forest Service and BLM Also Lack a Systematic Approach to Assess the
Risks of Environmental Effects Associated with Fuel Reduction Projects:
The Forest Service and BLM also do not have a systematic approach for
assessing the risks of environmental effects associated with fuel
reduction projects. Because the agencies generally assess the
environmental effects of fuel reduction projects in NEPA documents, we
reviewed 10 environmental assessments for such projects. Although the
agencies' assessments were for projects that would mitigate or avoid
the effects of wildland fire, the assessments did not systematically
assess the hazard, risks, or values associated with the projects. This
is because the agencies lack clear guidance on how to assess the risks
of environmental effects of projects and where to document the
assessment and effects--in environmental assessments or in other
documents. Interim guidance issued for the Healthy Forests Restoration
Act partially addresses the need for documentation of the effects of
not taking action to reduce fuels, but is not yet complete or final.
CEQ guidance on developing model environmental assessments for fuel
reduction projects, issued in 2002, does not address the risks of
environmental effects from reducing fuels or not because the purpose of
the guidance was to facilitate the development of concise environmental
assessment documentation. Guidance on how to do such assessments and
where to document them is important if the agencies are to effectively
use fuel reduction projects to address the risks of environmental
effects from wildland fires.
Some Environmental Assessments Did Not Systematically Assess the Risks
of Not Reducing Fuels:
Our review of 10 environmental assessments of fuel reduction projects-
-6 prepared by the Forest Service and 4 by BLM--revealed that some of
these assessments did not systematically assess the risks of the likely
environmental effects from not implementing fuel reduction projects.
Forest Service and BLM officials recognize that taking no action to
reduce fuels while continuing to suppress fires will contribute to
continued accumulation and alteration of vegetation, perpetuating
hazardous conditions that can create severe fire effects. At the same
time, fuel reduction projects themselves can pose risks to
environmental resources. For these reasons, it is important that the
agencies, in developing fuel reduction projects, discuss the hazard,
risk, and values at risk associated with project alternatives. It is
also important that the agencies analyze a project's environmental
effects. We reviewed the environmental assessments prepared by the
Forest Service and BLM to determine the extent to which they (1)
included a discussion of, or referred to, other documents that
discussed hazard, risk, and value in assessing project alternatives and
(2) presented information and data on the environmental effects of the
no-action alternative. We did not determine whether the assessments
complied with NEPA.
Generally, the agencies analyze the environmental effects of a fuel
reduction project in an environmental assessment, which is an analysis
conducted under NEPA regulations and guidance to discusses such
effects. In general, fuel reduction projects can range in size from a
few hundred acres to several thousand acres. In developing a project,
forest and BLM staff determine the purpose and need for the project and
design one or more alternatives. Although NEPA regulations do not
require the agencies to consider a no-action alternative and its
effects in an environmental assessment, the agencies often do develop a
no-action alternative and consider its effects. The agencies can use
several different models of fire effects and fire behavior to show how
different vegetation will burn in different conditions, such as
weather. (See app. VI for a discussion of models.) Through this
modeling, agency staff can show that reducing vegetation can reduce the
risk of a fire becoming uncharacteristically large and intense and the
risk of associated adverse effects.
Furthermore, in analyzing the effects of a fuel reduction project, the
agencies are also to consider whether the project meets the
requirements of several environmental laws, including the Clean Water
Act, the Clean Air Act, and the Endangered Species Act. In analyzing
water quality effects, for example, forest and BLM staff estimate what
effects fuel reduction activities such as mechanical thinning--which
can involve heavy equipment that compacts soils--will have on soils and
sedimentation of local streams. In analyzing air quality effects,
forest and BLM staff estimate the amount of smoke that a prescribed
burn may produce. Finally, in analyzing the effects of a project on
threatened and endangered species--such as the Northern spotted owl or
the Canada lynx--the agencies may be required to develop biological
assessments of the effects on species population and habitat. In
consulting with the Fish and Wildlife Service and National Oceanic and
Atmospheric Administration (NOAA) Fisheries, the agencies use these
biological assessments to determine what risks the alternatives pose
for species and their continued existence and survival.[Footnote 22]
Table 4 shows the results of our review of the 10 environmental
assessments.
Table 4: Results of GAO's Review of 10 Fuel Reduction Project
Environmental Assessments:
Project name: Rogue;
Hazard, risk, and values included? Yes;
Discussion of environmental effects specifically for not reducing
fuels? Yes.
Project name: Horsethief;
Hazard, risk, and values included? No;
Discussion of environmental effects specifically for not reducing
fuels? Yes.
Project name: Weaver Mountain;
Hazard, risk, and values included? No;
Discussion of environmental effects specifically for not reducing
fuels? Yes.
Project name: Pine Valley;
Hazard, risk, and values included? No;
Discussion of environmental effects specifically for not reducing
fuels? Yes.
Project name: Last Chance;
Hazard, risk, and values included? Yes;
Discussion of environmental effects specifically for not reducing
fuels? Yes.
Project name: Sheafman;
Hazard, risk, and values included? Yes;
Discussion of environmental effects specifically for not reducing
fuels? Yes.
Project name: Deer Point;
Hazard, risk, and values included? Yes;
Discussion of environmental effects specifically for not reducing
fuels? Yes.
Project name: Pinaleno;
Hazard, risk, and values included? No;
Discussion of environmental effects specifically for not reducing
fuels? Yes.
Project name: Cache la Poudre;
Hazard, risk, and values included? No;
Discussion of environmental effects specifically for not reducing
fuels? Yes.
Project name: Ely;
Hazard, risk, and values included? No;
Discussion of environmental effects specifically for not reducing
fuels? Yes.
Sources: Forest Service and BLM (documents); GAO (analysis).
[End of table]
Of the 10 assessments we reviewed, 4--3 from the Forest Service and 1
from BLM--used the hazard, risk, and value framework for a risk
assessment to discuss the action and no-action alternatives. In one
case, for example, the BLM's Rogue assessment estimated the number of
days per year that intense fires would reach the crowns of trees--known
as crown fire--and kill whole trees. The assessment also estimated how
much of the project area contained environmental resources at risk. The
Forest Service's Sheafman project assessment described the risk of fire
on the basis of a recent trend toward large fires in the project area.
However, 6 of the 10 assessments did not use a hazard, risk, and value
framework to discuss the risk of environmental effects. Existing
guidance does not specifically require this or describe how the
assessment should be done.
Each of the 10 assessments included some level of discussion of the
environmental effects of the no-action alternative (see table 4). The
level of detail about the effects varied from a minimal description of
some effects to a detailed discussion of effects that was comparable to
the discussion provided in other alternatives. For example, one of the
assessments only described the general effects of no action as
"increasing risk of damage to water resources." On the other hand,
another of the 10 assessments used extensive tables comparing the
specific effects on a wide range of environmental resources that could
result from taking or not taking action to reduce fuels.
Although each of the 10 assessments we reviewed contained some effects
of not reducing fuels, the effects were not described systematically in
all of the assessments. According to forest and BLM staff, as well as
biologists from the Fish and Wildlife Service and NOAA Fisheries,
predicting the environmental effects of the no-action alternative is
difficult. This difficulty stems from having to predict the occurrence
of wildland fires, which are random events. The staff noted that such
an analysis would need to be based on the assumption that a particular
fire would occur and on estimates of the environmental effects of that
assumed fire--a process they said might be considered "speculative."
For example, to project the potential effects of wildland fire in a
project area on a threatened or endangered species--such as the
Northern spotted owl or Canada lynx--a wildlife specialist would take
data produced by fire models to determine what habitat and food sources
would be affected under anticipated fire conditions. While some say
that this cannot be done, according to agency and CEQ officials, NEPA
involves making reasonable forecasts of effects, and models can provide
a reasonable basis on which to make such forecasts.
Guidance Is Unclear About How to Assess and Document the Risks of
Environmental Effects of Fuel Reduction Projects:
The varied use of risk assessment at the project level is a result of
the fact that the agencies do not have clear guidance about the
systematic assessment of the risks of environmental effects from
wildland fire from fuel reduction projects. Guidance would describe how
relevant hazards, risks, and values would be assessed for fuel
reduction projects and where such information would be documented. Such
guidance could be expected to appear in the agencies' fire planning or
NEPA guidance. However, Forest Service and BLM fire planning guidance
does not provide clear direction on how to conduct and document an
assessment of risk of environmental effects at the project level,
although the agencies recognize the importance of reducing fuels to
mitigate or avoid the environmental effects of wildland fires. As
described above, the agencies' fire planning guidance refers to
assessing the risks to communities at the landscape level and to
assessing the condition of vegetation at the project level. Just as the
guidance does not address the assessment of hazard, risk, and values at
risk at the landscape level, it does not discuss the assessment of
hazard, risk, and value at the project level.
Similarly, NEPA guidance does not describe how to conduct or document
an assessment of the risks of environmental effects associated with
fuel reduction project alternatives. In terms of describing how to
conduct such an assessment, although NEPA guidance requires the
description of an alternative's environmental effects, it does not
discuss an explicit risk assessment approach involving the assessment
of hazard, risk, and value. In terms of documenting the environmental
effects of a project, NEPA does not explicitly require the discussion
of the effects of the no-action alternative in an environmental
assessment. Analysis and documentation of the effects of a no-action
alternative in relation to the effects of action alternatives
facilitates the comparison of the relative risks of taking action or
not to reduce fuels at the project level. While NEPA requires
environmental impact statements to discuss a no-action alternative for
a proposed project, it does not require this of environmental
assessments. In 2002, CEQ issued NEPA guidance for a demonstration
program to develop examples of model environmental assessments for fuel
reduction projects. The demonstration program sought to make
environmental documents more concise by removing unnecessary
information. While the CEQ guidance stated that the agencies may
compare the impact of the proposed action and alternatives with the
current condition and expected future condition in the absence of the
project, it did not address the no-action alternative or its effects in
environmental assessments of fuel reduction projects.
The agencies have opportunities to clarify the analysis and
documentation of the risks of environmental effects from not taking
action to reduce fuels. In February 2004, the agencies issued interim
guidance on the Healthy Forests Restoration Act (HFRA).[Footnote 23]
This guidance, which does not clearly state when a no-action
alternative should be developed, states that the effects of not
reducing fuels should be documented in a project's files. It also
identifies some effects that may be analyzed as part of the
documentation. However, the guidance does not include the range of
environmental resources that may be affected by wildland fire nor does
it state clearly what analyses should be done. In addition, the
agencies and CEQ met in March 2004 to evaluate the lessons learned from
the demonstration program, which CEQ used to develop examples. The
agencies are compiling lessons learned from the demonstration program
that they intend to use in future environmental assessments. Through
these evaluations, CEQ and the agencies have another opportunity to
clarify how the agencies will address the risks of environmental
effects associated with fuel reduction projects and whether or not the
discussion of the effects of taking no action should be included in
environmental documents for projects.
Conclusions:
The Forest Service and BLM, in conjunction with the other federal,
state, local, and tribal land management agencies, face important and
complex challenges in managing wildland fires, and, in particular,
minimizing risks from future fires, including the effects of fires on
environmental resources and ecosystems. The primary challenge that
relates to these effects is, over the long term, balancing the risks of
adverse effects caused by wildland fires with the benefits that such
fires offer to restoring resources and ecosystems. The agencies will
not be able to find such a balance without understanding the range of
environmental and ecosystem effects created by wildland fire, including
the adverse, neutral, and beneficial effects, and without gathering
data and information to monitor and support their decision making about
fire management. The range and variety of effects complicates their
data-gathering efforts, which makes it important that the agencies
select the right effects to monitor and that the agencies test and
adopt new tools that can help them capture the range of effects of
wildland fire. Without collecting landscape data on wildland fire
effects as the National Fire Plan and its activities are implemented,
the Forest Service and BLM--as well as the other fire management
agencies--will not have information and data to give them a better
understanding of how fires are affecting environmental resources and
ecosystems over the long term. With such data, the agencies would be in
a better position to answer the question whether fires are burning with
more than normal severity because of vegetation and fuel conditions or
as a result of other factors. For these reasons, it is important that
the agencies, and more broadly the Wildland Fire Leadership Council,
follow through on their commitment to implement the monitoring
framework passed in May 2004.
A second challenge is that the agencies will not be able to stop all
fires, nor will they be able to reduce fuels on all lands. Without a
risk-based approach, the agencies cannot adequately determine where
they will take deliberate action to reduce fuels and the risks of
potential fire effects and where they will take the risk of fire
occurring. Successful risk management is grounded in risk
identification and assessment. Until the agencies formalize a framework
for assessing the elements of risk at appropriate organizational
levels, they will not have a systematic process for assessing the risks
to environmental resources and ecosystems posed by wildland fires and
will not be able to coordinate their efforts to ensure that they
succeed in reducing these risks. A framework should incorporate and
build on the data that are to be generated by LANDFIRE, as well as fire
and other models. At the landscape level, a framework would clarify
what elements will be included in agency risk assessments, what
decisions will be supported by risk assessments, who will do the
assessments, and when they will be done. This is particularly important
as state and local agencies are moving ahead with risk assessments and
the federal agencies will need to coordinate and collaborate with them.
In carrying out fuel reduction projects, it is important that the
agencies consider the risk of not taking action to reduce fuels and the
long-term risk of wildland fire effects. In fact, HFRA now specifically
requires the agencies to develop a no-action alternative for fuels
reduction projects covered by the act. However, the agencies must also
assess the risks of effects from their actions to reduce fuels to
ensure that they will not exacerbate existing resource problems or
create new problems. Such decisions require information on the
potential environmental effects of all project alternatives and the
ability to make clear comparisons among project alternatives and their
potential effects. Without clear guidance on the complicated analysis
and comparisons that seem warranted to adequately address the effects
of taking action or not to reduce fuels, the agencies lack the ability
to make informed decisions to conduct fuel reduction projects, and as a
result, these projects could face challenges and delay rather than
proceeding more quickly in the face of increased risks. The agencies'
guidance needs to be more specific about (1) how the assessment of
hazard, risk, and values at risk should be done at the project level
and (2) where the assessment of such risks should be documented. Both
the HFRA guidance and the lessons learned from the CEQ demonstration
program provide an opportunity for the agencies and CEQ to clarify
appropriate guidance.
Recommendations for Executive Action:
To improve the agencies' ability to identify and manage the actual and
potential effects of wildland fires on the environment, we recommend
that the Secretaries of Agriculture and the Interior, after consulting
with the Wildland Fire Leadership Council, direct the Forest Service
and BLM to:
* develop a monitoring plan to implement the agencies' framework
approved in May 2004 and include a pilot program for testing on Forest
Service and BLM lands the applicability of, and resource needs
associated with, the burn severity mapping and data tool developed by
the National Park Service;
* develop and issue guidance, in consultation with experts inside and
outside the agencies, that formalizes a framework for systematically
assessing landscape-level risks to ecosystems from wildland fires; and:
* clarify existing guidance, working with CEQ and taking into account
any lessons learned from the CEQ demonstration program, on the
assessment and documentation of the risks of environmental effects
associated with not conducting fuel reduction projects.
Agency Comments and Our Evaluation:
We provided a draft of this report to the Secretaries of Agriculture
and the Interior and the Chairman of CEQ for review and comment. The
departments provided a consolidated written response to our draft,
which is included in appendix VII of this report. CEQ provided written
comments, which are included in appendix VIII.
The departments stated that the draft was well prepared and provided a
thorough analysis of a complex set of issues. In commenting on our
recommendation to develop a monitoring plan and pilot project to test a
tool for mapping the long-term effects of wildand fires, the
departments said they presently have several methods for assessing such
effects. Specifically, the departments said that the Forest Service and
BLM work with the U.S. Geological Survey to assess the burn severity of
large or severe fires and that, at the field level, the agencies
collect site-specific data on wildland fires to support their land
management plans and postfire rehabilitation plans. Furthermore, the
departments mentioned that in May 2004, the Wildland Fire Leadership
Council approved a nationwide monitoring framework for wildland fire
data, including fire severity data. We agree that the agencies have
various methods for assessing the effects of wildland fire, and we
identified several of these in the report. We are also encouraged by
the recent development of a nationwide monitoring framework. However,
this framework does not address the part of our recommendation that
calls for a pilot program for testing a different tool for measuring
burn severity. The pilot program would allow the agencies to determine
the extent to which the National Park Service tool is applicable for
assessing the environmental and ecosystem effects from wildland fires
that start on the agencies' lands. Furthermore, since the framework has
just been approved, no plan for its implementation has been developed.
Accordingly, we are recommending that such a plan be developed. We
modified the report text and recommendation to reflect the recent
approval of a monitoring framework.
In commenting on our second recommendation that the agencies develop
and adopt a framework that supports systematic assessment of the
landscape-level risks to ecosystems from wildland fire and issue
guidance implementing that framework, the departments had several
concerns. First, the departments stated that the Forest Service and BLM
are designing and refining a number of analytical tools to assess
project and landscape-level risk. Specifically, the departments noted
that the Fireshed Assessment process is a promising approach for
evaluating fuel treatment effectiveness across landscapes and that
LANDFIRE will provide nationally consistent data to be used in
landscape-level risk assessments. Second, the departments stated that
the agencies already include a significant amount of risk assessment in
their fuels programs. The departments said that national fuel reduction
priorities represent a judgment about risk; the classification of
wildland-urban interface areas into high, medium, and low priorities
represents another judgment about risk; and the emphasis on treating
lands in different fire regimes and condition classes represents yet
another judgment about risk. The departments also stated that the
relative value of resources is decided through the collaborative
project selection and prioritization process developed with states for
the implementation of the 10-Year Comprehensive Strategy, as directed
by the Congress, and that community wildfire protection plans called
for in HFRA include an assessment of risks associated with high-
priority areas and values identified by the communities. Third, while
the departments agree that prioritization of fuel reduction treatments
can be improved and in their technical comments they agree that they do
not have a single, consistent risk-based approach, they do not believe
that a model or methodology that assesses the risks associated with
fuel reduction treatments across time and at the landscape and project
scales can be developed. Finally, the departments said that it is
misleading to assert that the Forest Service and BLM have focused on
the wildland-urban interface, preventing them from developing a
systematic approach to ecosystem risk assessment.
Our recommendation is based on our belief that to systematically
identify fuel reduction projects, the agencies should support a single
common approach or framework for assessing risk. We recognize that the
agencies are developing a number of analytical tools useful for
assessing landscape-level risk. We note several of these in our report
and believe that they can serve as good examples for the agencies in
developing an overall risk assessment framework. However, a common
approach will allow the agencies to prioritize projects systematically.
In addition, we recommended that the agencies seek the assistance of
external and internal experts, which would allow them to identify the
best tools for conducting risk assessments and would leverage the
experience of the field offices in developing and using different risk
tools, such as the Fireshed Assessment process the departments mention.
In addition to Fireshed, several other worthwhile efforts that are
being implemented in the field could serve as examples.
Regarding the departments' views that they already carry out a
significant amount of risk assessment, it is our view that the results
of these assessments are too broad to target fuel reduction projects at
the landscape or project levels. Given the level of fuel reduction
needs (90-200 million acres) identified in the Coarse-Scale Analysis
and the emphasis of the 10-Year Comprehensive Strategy and
Implementation Plan on reducing risks to the wildland-urban interface
and environment, the agencies need a systematic way to further target
their projects to areas of high risk and to demonstrate to decision
makers and to the public how the decisions have been made and what
results have been achieved with the use of public funds. The
departments state that the collaborative process will identify relative
values and priorities and that community wildfire protection plans will
include assessment of the risks to community-defined values.
Recognizing that the 10-Year Comprehensive Implementation Plan does not
change existing agency statutory and regulatory responsibilities, we do
not disagree. However, we believe that a formalized, common risk
assessment framework would better inform the collaborative efforts
under way, as well as future community efforts, by providing the groups
involved with consistent hazard and risk information and allowing them
to identify values at risk. We did not, as the departments' comments
state, suggest they develop a single, all-encompassing model or
methodology to assess risks associated with fuel reduction treatments
across multiple time frames and geographic scales. Rather, our
recommendation is directed at ensuring that the agencies develop a
common approach to assessing landscape-level risks to environmental
resources and ecosystems from wildland fires. We modified our report
text and recommendation language to clarify that we mean a formalized,
common approach to assessing risks.
Furthermore, we did not criticize the agencies' emphasis on addressing
threats within the wildland-urban interface. We merely observed that
the agencies have not made the same degree of progress in developing an
approach for assessing the risks from wildland fire to ecosystems and
environmental resources as they have for areas within the wildland-
urban interface. We concur with the agencies' statements in the 10-Year
Comprehensive Strategy and Implementation Plan that threats outside the
wildland-urban interface also need to be addressed, and our
recommendation is directed at that need. We did not mean to imply that
communities should not be a high priority for fuel reduction, and we
modified our report text accordingly.
Lastly, the departments disagree with our recommendation that the
agencies develop and issue guidance to clarify the assessment and
documentation of risks of environmental effects associated with taking
or not taking action to reduce fuels. They said that current NEPA
guidance is sufficient and that additional guidance on conducting risk
assessments would be inconsistent with new CEQ guidance, which calls
for short concise environmental assessments. Furthermore, the
departments said that existing direction is generally adequate for
implementing the lessons learned from the CEQ demonstration program. We
did not state that CEQ's guidance fails to meet the intent of NEPA, nor
did we recommend any amendments to that guidance. The purpose of our
recommendation is to help ensure that the agencies have a sound
scientific basis for the prioritization decisions that they will need
to make to implement fuels reduction projects as effectively as
possible, not to enhance agency compliance with NEPA. In fact, interim
guidance issued by the agencies for HFRA partially addresses the need
to document the risks associated with not taking action to reduce
fuels. Specifically, the interim guidance says that "it is important
that the specialists' report retained in the project files document the
anticipated short-and long-term effects of proposed HFRA treatments."
We believe, however, that if the agencies agree that the documentation
of risks and effects is important, this guidance should specifically
require the need for such documentation and should clarify how the
environmental assessment should refer to the effects of the no-action
alternative documented in the project files. Accordingly, we continue
to believe that clear guidance for helping agency personnel determine
the appropriate form and content of the risk assessment associated with
not taking action would improve the agencies' fuel reduction efforts.
We modified our report text and our recommendation language to reflect
the fact that the agencies' interim guidance provides another
opportunity to clarify the analysis of risks of not reducing fuels.
CEQ provided us written comments on the aspects of the report that
reflect CEQ policies. CEQ stated that our draft should not imply that
CEQ's guidance for its demonstration program could, or was intended to,
describe "how to" assess the risks of environmental effects of taking
action to reduce fuels against those of not taking action. CEQ stated
that the purpose of this guidance was to provide a framework to support
the development of environmental assessments that are "concise" and
"public" documents that better serve their core function: briefly
describing sufficient evidence and analysis to support a decision about
whether or not to prepare an environmental impact statement and to
assist agency NEPA compliance. We did not intend to imply that CEQ's
guidance should discuss risk assessment methods, and we directed our
recommendation to the agencies to clarify their guidance and
methodology. We referred to the CEQ guidance in our discussion of risk
assessments for conducting or not conducting fuel reduction projects
only to demonstrate that such guidance does not exist.
CEQ also said that our draft incorrectly states that CEQ and the
agencies plan to finalize a report on the "lessons learned" from the
demonstration program. Rather, CEQ said that it has issued examples of
completed environmental assessments on the Internet and that the
agencies have drafted a document on these lessons. We deleted the
reference to CEQ being a participant with the Forest Service and BLM in
documenting the lessons learned from the program.
The Departments of Agriculture and the Interior and CEQ made other
technical comments, which we addressed as appropriate in the report.
As arranged with your offices, unless you publicly announce the
contents earlier, we plan no further distribution of this report until
30 days after the date of this letter. At that time, we will send
copies of this report to other interested congressional committees. We
will also send copies of this report to the Secretaries of Agriculture
and the Interior, the Chief of the Forest Service, the Director of BLM,
and the Chairman, CEQ. We will make copies available to others upon
request. In addition, the report will be available at no charge on the
GAO Web site at [Hyperlink, http://www.gao.gov.].
If you or your staff have questions about this report, please contact
me at (202) 512-3841. Key contributors to this report are listed in
appendix IX.
Signed by:
Barry T. Hill:
Director, Natural Resources and Environment:
List of Requesters:
The Honorable Pete V. Domenici:
Chairman:
Committee on Energy and Natural Resources:
United States Senate:
The Honorable Larry E. Craig
Chairman:
Subcommittee on Public Lands and Forests:
Committee on Energy and Natural Resources:
United States Senate:
The Honorable Mike Crapo:
Chairman:
Subcommittee on Fisheries, Wildlife, and Water:
Committee on Environment and Public Works:
United States Senate:
The Honorable Kit Bond:
United States Senate:
The Honorable Conrad Burns:
United States Senate:
The Honorable Ben Nighthorse Campbell:
United States Senate:
The Honorable John Ensign:
United States Senate:
The Honorable Jon Kyl:
United States Senate:
The Honorable Gordon H. Smith:
United States Senate:
[End of section]
Appendixes:
Appendix I: Objectives, Scope, and Methodology:
This appendix presents the scope and methodology we used to gather
information on the environmental effects of wildland fire and to
understand the data collected on these effects by the Forest Service
and the Bureau of Land Management. It also addresses how we identified
the approaches that these agencies use to assess the risk posed to
environmental resources by wildland fire and the relative risks of
undertaking and not undertaking fuel reduction activities.
To describe the environmental effects of wildland fire and assess the
information the Forest Service and BLM gather on these effects, we
randomly selected a sample of 20 fires from a universe of 614 wildland
fires to include in a survey. The fires we selected ranged in size from
hundreds to tens of thousands of acres in diverse geographic locations.
To identify the universe of wildland fires, we created a database from
two separate lists of Forest Service Burned Area Emergency Response
(BAER) reports and BLM Emergency Stabilization and Rehabilitation (ESR)
reports completed in 2000, 2001, and 2002. We requested these lists of
BAER and ESR reports because the agencies' official lists of wildland
fires did not contain information on the environmental effects of these
fires. Both Forest Service and BLM officials indicated that BAER and
ESR reports contain data on the environmental effects of fire because
they are prepared for the purpose of requesting emergency stabilization
funds to aid in the recovery of the burned areas. We assessed the lists
of these reports for completeness and accuracy and found them to be
reliable for the purpose of selecting a sample of wildland fires. We
also visited the locations of 6 western fires and discussed the data
collected on their environmental effects with local Forest Service and
BLM officials. Figure 6 shows the fires we included in our survey
sample, as well as the fire locations we visited in eight western
states.
Figure 6: Location of Forest Service and BLM Wildland Fires Visited and
Surveyed:
[See PDF for image]
[End of figure]
To develop survey questions for the sample of 20 wildland fires and
better understand the data that the Forest Service and BLM collect on
the environmental effects of these fires, we (1) conducted a literature
review and read reports and scientific studies on this topic; (2)
reviewed federal fire policy documents, handbooks, BAER and ESR
reports, and other reports; and (3) interviewed Forest Service and BLM
officials in national, regional, and state offices, as well as local
land units. Information gathered from the agencies on the environmental
effects of wildland fire allowed us to develop a series of questions
for our survey about the size and extent of a wildland fire and how it
affected individual resources, such as soil, vegetation, air,
watersheds, and threatened and endangered species over the short and
long term. Our survey also contained questions on short-term (defined
as less than 10 years), long-term (defined as greater than 10 years),
and cumulative effects of the sample fire on the landscape or
ecosystem. Some of these questions requested that the land managers
make predictions about the future effects of a wildland fire rather
than providing data about effects that had already occurred. Therefore,
there is a greater degree of uncertainty regarding the accuracy of
responses to these questions than to questions about observed effects.
Once we developed the survey questions, we pretested the content and
format of the survey with BAER and ESR team leaders and other agency
staff, as well as with two scientific and technical experts at the
University of Arizona and the University of Washington. We conducted
pretests with agency officials located in Colorado, Montana, Nevada,
Washington, and Washington, D.C. During pretesting, we simulated the
actual survey experience by asking the individual to complete the
survey, and then we conducted a semistructured interview to determine
whether (1) the questions were clear, (2) the terms used were precise,
and (3) how long it took individuals to answer the questions. In mid-
November 2003, we electronically mailed the survey to staff the
agencies identified as knowledgeable about each fire and instructed
them to return the survey by electronic mail. We obtained a 100 percent
response rate from staff at all 20 field locations.
To identify the approaches the Forest Service and BLM take to assess
the risk to environmental resources from wildland fire and the risks
associated with undertaking and not undertaking fuel reduction
projects, we reviewed National Fire Plan policies, National
Environmental Policy Act regulations, and the December 2001 report
[Footnote 24] prepared by the National Academy of Public
Administration on risk assessment and its application to the National
Fire Plan. We also obtained agency guidance and selected planning and
project documents to see how they addressed the potential
environmental effects of future fires. To obtain specific information
on the agencies' risk assessment practices, we interviewed Forest
Service and BLM officials representing seven national forests and six
BLM field offices and reviewed their land and resource management
plans, in addition to their fire management plans. We also examined
existing and proposed risk assessment approaches discussed in research
papers and governmental and nongovernmental publications, and we
interviewed some of the authors. During 2003, we attended two national
conferences on wildland fire and risk assessment decisions held in
Denver, Colorado, and Portland, Oregon. We also contacted the National
Park Service, professors from five universities, and scientists from
the Forest Service's research stations in California, Montana, and
Colorado, including those who participated in developing the national
assessment of fuel conditions and LANDFIRE.
To identify how the Forest Service and BLM evaluate the relative risks
of undertaking or not undertaking projects to reduce fuels, we reviewed
10 environmental assessments of fuel reduction projects in eight
western states: Arizona, California, Colorado, Montana, Nevada, Oregon,
Utah, and Washington. We selected 6 environmental assessments prepared
by the Forest Service and 4 prepared by BLM for analysis. Of the
assessments we selected, 5 were part of a Council on Environmental
Quality (CEQ) demonstration program for developing model environmental
assessments and 5 were not. We also interviewed officials at Forest
Service and BLM headquarters and staff members at CEQ to obtain
information on agency guidance to field offices and CEQ program
guidance to agencies. Finally, we interviewed headquarters and regional
staff from NOAA Fisheries and the Fish and Wildlife Service to obtain
information on how agency biologists could compare the short-term
effects of fuel reduction projects on threatened and endangered species
with the long-term effects of a future wildland fire in their
environmental assessments.
We conducted our work from April 2003 through April 2004 in accordance
with generally accepted government auditing standards.
[End of section]
Appendix II: Fire Regime Condition Class Analysis:
Wildland fire managers have recognized for several years the need to
reduce excessive vegetation levels that have accumulated and been
altered after decades of excluding fire from different ecosystems.
Until 2000, managers did not have national-level data that could be
used to distinguish and prioritize the different ecosystems that are
more affected than others by fire exclusion. To accomplish such a
prioritization, land managers need data on the current conditions of
vegetation and fuels. In 2000, the staff of the Forest Service's Fire
Sciences Laboratory at the Rocky Mountain Research Station produced an
assessment of spatial data for wildland fire and fuel management called
the Coarse-Scale Analysis.[Footnote 25] The staff gathered and analyzed
national-level data from several sources to create vegetation maps and
from that assessed the condition of this vegetation to act as fuel for
wildland fire. The assessment resulted in a national map of vegetation
and fuel conditions. The assessment was based on "coarse-scale" or
coarse resolution data that are not, and were never intended to be,
applicable for site-level use.
The assessment identified three categories of fuel conditions, called
fire regime condition classes (FRCC), using estimates of the historical
fire regimes--the frequency and intensity of fires typical for
different vegetation types--and estimates of the alteration in current
vegetation. Figure 7 shows the three categories of hazard and the
levels of alteration from historic conditions (hazard) that they
represent.
Figure 7: Wildland Fire Hazard:
[See PDF for image]
[End of figure]
The original assessment estimated that wildland fire could cause severe
effects on about 182 million acres of land in the United States. Of
that land, an estimated 75 million acres of federal land, including
land managed by the Department of the Interior and the Forest Service
within the Department of Agriculture, could suffer from severe effects
as a result of wildland fire. In 2002, the original assessment was
updated to more thoroughly examine grassland vegetation and to make
corrections from more specific data gathered through samples on the
ground. This update focused on 11 western states, reporting that 183
million acres across the west were highly altered from historical
conditions and would experience severe effects from wildland fire.
Based on additional analysis, the agencies estimated that the amount of
highly altered vegetation nationwide could vary from 90 to 200 million
acres.
Federal land management agencies and the National Fire Plan have
adopted the FRCC categories as an indicator of whether forest and
grassland ecosystems are in good condition. The National Fire Plan,
under its goal for achieving fire-adapted ecosystems, tracks the acres
of land that are in the second and third FRCC categories for which
fuels are reduced and conditions changed to a lower category--either
the first or second category. The agencies' field offices, which plan
and implement all projects to reduce fuels, are responsible for
monitoring their progress through a new national reporting system,
which tracks National Fire Plan goals and performance measures.
Because the national-scale data is based on coarse resolution data, the
data cannot be used to identify the areas on the ground that need to
have fuel reduction activities--it can only be used at a Forest Service
regional level or summarized to several western states. To help the
field offices identify which acres need fuel reduction treatment
projects, the Forest Service and the Department of the Interior are
developing a project-level analysis tool. This tool, which the Forest
Service and the Department of the Interior began to implement in spring
2004, requires the field office staff to conduct a field visit to
examine vegetation conditions, to consider them in context of past and
current fire regimes, and to estimate the alteration of fire regimes
and fire intensity if a fire were to burn in the current conditions.
The FRCC concept has been incorporated into LANDFIRE, which is a data
system that, when completed, will provide periodic updates of national
vegetation maps at a higher resolution than the original assessment
provided in 2000. LANDFIRE is also expected to provide FRCC maps and
will contain computer models to help fire planners and land managers
estimate how wildland fire might behave in the estimated fuel
conditions. Full implementation of LANDFIRE was approved in 2003 and
funded by the Forest Service and the Department of the Interior in
2004. The system is expected to be completed for application across the
nation in 2009. In the interim, the agencies will use the FRCC
assessment and project tool to track progress in reducing high levels
of fuels across the nation.
[End of section]
Appendix III: Definition of Fire/Burn Severity:
Table 5 presents the definitions of fire/burn severity as it relates to
effects on soils and vegetation. We asked Forest Service and BLM
officials to use this table in responding to questions about the
specific fires and their effects on soil and vegetation for the 20
wildland fires in our sample.
Table 5: Classes of Fire Severity for Soils and Vegetation:
Class: Unburned;
Soil substrate--litter/duff: Not burned;
Vegetation-- understory/brush/herbs: Not burned.
Class: Scorched;
Soil substrate--litter/duff: Litter partially blackened;
duff nearly unchanged; wood/leaf structures unchanged;
Vegetation--understory/brush/herbs: Foliage scorched and attached to
supporting twigs.
Class: Lightly burned;
Soil substrate--litter/duff: Litter charred to partially consumed;
upper duff layer burned; wood/leaf structures charred, but
recognizable;
Vegetation--understory/brush/herbs: Foliage and smaller twigs partially
to completely consumed.
Class: Moderately burned;
Soil substrate--litter/duff: Litter mostly to entirely consumed,
leaving coarse, light-colored ash; duff deeply burned; wood/leaf
structures unrecognizable;
Vegetation--understory/ brush/herbs: Foliage, twigs, and small stems
consumed.
Class: Heavily burned;
Soil substrate--litter/duff: Litter and duff consumed, leaving fine,
white ash;
mineral soil visibly altered, often reddish;
Vegetation--understory/brush/herbs: All plant parts consumed, leaving
some or no major stems/trunks.
Class: Not applicable;
Soil substrate--litter/duff: Inorganic;
Vegetation--understory/brush/herbs: Not present.
Source: U.S. Fish and Wildlife Service.
Note: The definition appears in the National Wildlife Coordinating
Group Fire Use Working Team, National Interagency Fire Center, Fire
Effects Guide (NFES 2394), (Boise, Idaho: June 2001), citing Fuel and
Fire Effects Monitoring Guide, U.S. Fish and Wildlife Service, June
2001.
[End of table]
[End of section]
Appendix IV: Selected Wildland Fire Survey Results:
This appendix provides selected results from our survey of Forest
Service and BLM officials on the environmental effects of 20 sample
wildland fires (app. I). Tables 6 and 7 present general information
about each fire and the fire location. Tables 8 through 13 provide
information on the generally short-term effects that these fires had
on individual environmental resources such as streams and watersheds,
threatened and endangered species and their habitats, air, soils, and
vegetation.
Table 6 lists each wildland fire by size according to the total number
of federal and nonfederal acres of land burned. It also presents
information on federal and nonfederal acreage within the fire perimeter
but not burned by fire.
Table 6: Acres Burned by Fire, Land Ownership, and Percent of Forest
and BLM Field Office Land Base:
Fire name: Missionary Ridge (Colorado);
Total acres burned: 57,935;
Federal acres burned: 49,990;
Nonfederal acres burned: 7,945;
Federal acres unburned: 11,929;
Nonfederal acres unburned: 615;
Percent of forest or BLM office land base affected by fire: 3.80%.
Fire name: Sheep Mountain (Wyoming);
Total acres burned: 27,574;
Federal acres burned: 21,370;
Nonfederal acres burned: 6,204;
Federal acres unburned: 5,102;
Nonfederal acres unburned: 1,253;
Percent of forest or BLM office land base affected by fire: 0.59%.
Fire name: Burgdorf Junction (Idaho);
Total acres burned: 17,207[A];
Federal acres burned: 17,207;
Nonfederal acres burned: D/K;
Federal acres unburned: 47,000;
Nonfederal acres unburned: D/K;
Percent of forest or BLM office land base affected by fire: 2.90%.
Fire name: Crimson Clover (Idaho);
Total acres burned: 16,172;
Federal acres burned: 14,466;
Nonfederal acres burned: 1,706;
Federal acres unburned: 330;
Nonfederal acres unburned: 0;
Percent of forest or BLM office land base affected by fire: 1.00%.
Fire name: Elko 13/#3 (Nevada);
Total acres burned: 13,104;
Federal acres burned: 12,544;
Nonfederal acres burned: 560;
Federal acres unburned: 256;
Nonfederal acres unburned: 0;
Percent of forest or BLM office land base affected by fire:
