home > overview > documents > previous abstracts

ForestERA Presentations Abstracts

Oral presentation to the 90th Annual Meeting of the Ecological Society of America , August 7-12, Montreal, Canada.

RESTORING FIRE AS A KEYSTONE PROCESS: INSIGHTS FROM THE PINE FORESTS OF ARID NORTH AMERICA

Authors: Sisk, Thomas D., Donald A. Falk, Melissa Savage and Patrick McCarthy

The exclusion of fire from long-needled pine ecosystems has led to the accumulation of biomass that fuels increasingly large crown fires across much of arid North America. This new fire regime contrasts with the frequent, low-intensity fire that typifies the paleoecological record across much of the region. Novel fire regimes, and the risks they pose to people, are the products of past attempts to engineer forest ecosystems. Throughout most of the 20th century, foresters traded old-growth forests, including the “package” of ecosystem services benefiting people (e.g., the provision of clean water, erosion control, herbaceous productivity, and wildlife habitat) for two products – timber and livestock – that were thought to have overriding value for the increasing human population of the region. While producing short-term benefits to people, this management paradigm led to unanticipated ecosystem change, generating enormous costs associated with fire suppression, as well as the loss of life, property, and environmental quality. As a result, the restoration of natural fire regimes has garnered considerable political will and community support. However, formidable new challenges have emerged during the century of fire exclusion: fuel loads have increased, human communities have expanded into the forest, and nonindigenous species have dispersed widely. While restoration efforts strive to reverse the errors of the past, these new obstacles constrain management options. Furthermore, unfavorable economics associated with large-scale manipulations in forests that retain little marketable timber preclude widespread intervention to advance ecological objectives. Conversely, increasing demands for water, and the costs associated with watershed degradation, provide positive incentives for investment in restoration. We draw on cases studies from the southwestern USA and northern Mexico to develop a framework for assessing the likelihood of success in reestablishing natural fire regimes as a means to ecological restoration. We offer recommendations about how ecologists can respond to the challenge of engineering an intelligent transition to forest conditions that will support natural processes and, eventually, the return of the most-valued ecosystem services.

Poster for the 123rd Annual Meeting of the American Ornithologists' Union, August 23-27, Santa Barbara, California.

MODELING THE INFLUENCE OF FOREST STRUCTURE AND COMPOSITION ON AVIAN COMMUNITIES IN NORTHERN ARIZONA PONDEROSA PINE FORESTS

Authors: Prather, John W., Steven S. Rosenstock, Brett G. Dickson, Kerry Griffis-Kyle, William M. Block, Carol L. Chambers, Thomas D. Sisk, Yaguang Xu and Haydee M. Hampton

We examined the influence of landscape-level patterns of forest structure and composition on avian communities across 800,000 acres of ponderosa pine dominated forest on the Western Mogollon Plateau, Arizona. Between 1996 and 2003, presence/absence data on typical ponderosa pine forest birds were taken at 375 point count locations scattered across the study area. In 2003, the Forest Ecosystem Restoration Analysis (ForestERA) project created 90m resolution spatial data layers representing forest composition and structure (basal area, tree stem density, canopy cover) for the study area using remote-sensing imagery. We used classification and regression tree (CART) methodologies to determine relationships between avian species richness, as well as the distribution of 9 bird species, with forest structure and composition. In general, species were more likely to be present, and species richness was higher, in pine-oak than in pure ponderosa pine. Areas that had recently been exposed to low intensity fires also tended to have higher species richness. Tree density was the most influential forest structural variable in determining species distributions. Basal area and canopy cover generally had less influence, but often interacted with tree density. Individual species were more likely to be present, and species richness was higher, in areas with lower tree densities, but higher basal area and/or canopy cover. Our results suggest that there are coarse, but predictable, patterns in avian distributions and species richness, which are related to forest structure and composition. These relationships can be used to predict changes in avian communities based on forest management practices.

Poster for the 2005 Annual Conference of the American Society for Photogrammetry and Remote Sensing, March 7-11, Baltimore, Maryland.

ANALYSIS OF BIAS AND VARIANCE IN FOREST STRUCTURAL MAPPING USING ETM IMAGERY AND GROUND PLOT DATA

Authors: Xu, Yaguang, John W. Prather, Haydee M. Hampton, Brett G. Dickson, and Thomas D. Sisk

Use of Enhanced Thematic Mapper (ETM) imagery, coupled with training data from ground plots, has been broadly applied in the creation of forest structural maps across large landscapes. Over the last three years, the ForestERA (Forest Ecosystem Restoration Analysis) project team has used ETM imagery, to map tree stem density and basal area in ponderosa pine dominated forest on the western Mogollon Plateau in northern Arizona. During this process we noted significant bias in predicted values for these forest structural attributes, a result that has also been noted in mapping efforts undertaken by other groups. This poster presents the results of an analysis of the causes of, and relationships between, bias and variance in our predictive models.

We also illustrate the consequences of bias and variance in forest structure mapping efforts. Inaccuracy of predictions results primarily from two sources: 1) image saturation of the ETM sensor under dense forest conditions, and 2) ground plots that do not completely measure forest structure over the extent of an ETM pixel. Due to the existence of this non-image processing bias, forest structural attributes are overestimated when they are low and underestimated when they are high. Additionally, in ponderosa pine, when basal area exceeds 45m²/ha, or tree density exceeds 850 trees/ha, the generally linear relationship between these two attributes breaks down, further affecting the relationship between actual values and those predicted using ETM. Under these conditions the bias dominates the prediction error, and mapping accuracy drops significantly.

Poster for the Joint meeting of the Southwest and Intermountain Region Society of American Foresters, May 11-13, Boulder City, Utah.

SPATIAL TOOLS FOR PREDICTING THE EFFECTS OF FUELS AND RESTORATION TREATMENTS ON FORESTED LANDSCAPES

Authors: Dickson, Brett G., Yaguang Xu, Haydee M. Hampton, John W. Prather, Jean A. Palumbo, John D. Bailey, William H. Romme, and Thomas D. Sisk

Forest fuels reduction and restoration treatments can be important in managing the threat of fire to communities and resources. To minimize this threat, it is presently recognized that efforts to reduce or manipulate forest fuels must occur at large scales and that fuels treatment and restoration objectives should be compatible. However, the effects of these treatments on various ecosystem attributes, including forest structure, fire behavior, and wildlife, are not well understood and should be assessed before resource managers recommend and implement large treatments. Moreover, management policies that address the environmental effects of treatments at the stand level may not consider the unique and cumulative effects of such treatments at the landscape level. For this research the objectives of the Forest Ecosystem Restoration Analysis (ForestERA) project were to 1) broadly characterize a range of treatment alternatives for ponderosa pine (Pinus ponderosa) that could be used to model the effects of these treatments on various forest structure attributes; and 2) develop spatial tools framed within a spatial decision support system for predicting and comparing the impacts of alternative forest management strategies on forest structure, fire hazard, and wildlife. By presenting a range of treatment alternatives within a flexible user interface, ForestERA tools will allow predictive models of forest fuels and restoration treatments to be overlaid with taxonomic distributions, inhabited areas, and other spatial data relevant to forest and fuels management.

Oral presentation abstract for The Nature Conservancy Fire Learning Network, February 16, 2005, Silver City, New Mexico.

THE FOREST ECOSYSTEM RESTORATIN ANALYSIS PROJECT: METHODS FOR THE ASSESSMENT OF FIRE THREAT IN SOUTHWESTERN CONIFEROUS FORESTS

Authors: Prather, John W., B. G. Dickson, Y. Xu, T. D. Sisk, and H. M. Hampton

The Forest Ecosystem Restoration Analysis (ForestERA) Project has developed spatial data and a GIS-based toolset to help support collaborative efforts to prioritize and plan forest management in southwestern coniferous forests. The project encourages the active participation of diverse stakeholders by engaging them in science-based planning processes that include the development, assessment, and comparison of alternative plans and their expected impacts. In support of these efforts, ForestERA is developing layers related to forest structure, fire behavior, wildlife habitat, and watershed attributes. Fire-related data and model outputs lie at the heart of many landscape assessments and have received commensurate attention in our work. To assess fire risk, we used a spatially-weighted Bayesian approach. Our output layer shows the predicted probability of a large fire (> 20 ha) starting within a given 1km² of the landscape. To assess fire hazard we developed input layers for the FlamMap fire modeling program, and used that program to produce outputs of predicted fire behavior. FlamMap is related to the widely used FARSITE fire modeling program, but predicts the behavior of fire at all points across an entire landscape, rather than predicting the temporal dynamics of spread. Thus, FlamMap can quickly and easily be used to compare relative fire behavior across large landscapes under a fixed set of weather conditions. We believe that this approach to modeling fire hazard is robust because it draws on existing local conditions while introducing a minimum of assumptions about historical factors and stochastic events, and because the output layer can be easily incorporated into multi-factor overlay analysis in the GIS environment. Alternative approaches to assessing fire hazard, such as FRCC, may also be useful within the ForestERA framework. However, the difficulty of reconstructing historical conditions in a manner that satisfies a diverse group of stakeholders presents multiple scientific and policy challenges. ForestERA has created or collected foundational data that may be used as surrogates for reference condition, and are more easily explained and ground-truthed. ForestERA tools and data are being used by an increasing number of stakeholder groups to plan and assess forest management actions at multiple scales on the western Mogollon Rim. Currently, we are expanding our project to two new areas, the eastern Mogollon Rim, and in north-central New Mexico. As these efforts evolve, we continue to explore new approaches that may lead to more efficient and effective means of providing useful information for landscape-level planning.

Oral presentation abstract for the 19th Annual Symposium of the International Association of Landscape Ecology (I.A.L.E.) U.S. Regional Association, March 31 – April 2, 2004, Las Vegas, NV

SPATIAL TOOLS SUPPORTING PUBLIC PARTICIPATION IN PLANNING FOREST RESTORATION

Authors: Hampton, Haydee M., John W. Prather, Yaguang Xu, Ethan N. Aumack, Brett G. Dickson, Thomas D. Sisk

The Forest Ecosystem Restoration Analysis project has completed a 2.5-year effort to develop a spatial decision support system for assisting public participation in forest management decisions. Our system is being used in a collaborative planning effort across two million acres of forests in northern Arizona. Collaborators include federal, state, and local government agencies, environmental organizations, community groups, and academics. In recent years, progress towards developing agreement on forest restoration plans has been impeded, in part, by the challenge of incorporating diverse stakeholder values for multiple criteria, such as watershed and community protection, and of predicting the impacts of forest treatments across hundreds of thousands to millions of acres – the scale at which wildfire and many other ecological processes operate. Our transdisciplinary GIS-based tools are comprised of an integrated set of spatial data and models, including predicted vegetation, fire risk and hazard, post-fire erosion, sedimentation, and flooding, wildlife habitat quality, and invasive plant occurrence and spread. Our decision support system allows alternative landscape-scale plans to be developed according to user preferences, and it can be used to predict their cumulative effects on wildlife and fire hazard, greatly facilitating the comparison of alternative plans. Applications are open and inclusive of all interested parties, and transparent and clearly documented, complete with uncertainty analyses. Our results demonstrate a new capability for bringing scientific understanding and science-based tools to a broad community involved in formulating goals and desired outcomes for the restoration of degraded forest ecosystems.

Oral presentation abstract for the International Association of Wildland Fire (IAWF) Wildland Fire 2004 Conference, March 3-4, 2004, Reno NV. Session: Landscape Level Fire Management Planning, Wednesday, March 3

LANDSCAPE-LEVEL ASSESSMENT OF FIRE, BIODIVERSITY, AND WATERSHEDS TO GUIDE ECOLOGICAL RESTORATION OF PONDEROSA PINE FORESTS

Author: Thomas D. Sisk, Center for Environmental Sciences and Education, Northern Arizona University

Stand-replacing crown fires have increased in size and destructiveness across the western USA during the past five decades. In many forests, the suppression of frequent, low-intensity ground fires, combined with livestock grazing and timber harvest, has resulted in large areas covered by smaller, fire-prone trees. Efforts to restore forest structure to conditions that would allow a return to historical fire regimes are hindered by the inability of managers and the public to assess the effects of alternative forest management practices on fire threat, biodiversity, and watersheds. Our research team, motivated by the need to examine the cumulative effects of multiple independent management decisions over large planning areas, provides spatial data and modeling tools to guide landscape-level assessments and planning. Data layers describing forest composition and structure allow modeling of fire threat, wildlife habitat, and watershed condition over hundreds of thousands of hectares. Current efforts to apply this modeling toolbox engage diverse stakeholders in assessment efforts, developing, analyzing, and comparing alternative management scenarios and their anticipated effects. I will present modeling results that emerged from the first formal application of this approach, the 800,000-ha Western Mogollon Plateau Assessment in Arizona. Additional efforts are planned for using these landscape-level spatial models to prioritize treatment areas and allocate resources for forest management in a manner that minimizes negative impacts on wildlife while maximizing restoration objectives, namely the return of frequent, low-intensity fire that sustains native forest structure and decreases the likelihood of catastrophic crown fire.

Poster Presentation from the 7th Biennial Conference of Research on the Colorado Plateau, U.S.G.S., Flagstaff, AZ, Nov. 3-6, 2003

MAPPING FOREST STRUCTURE USING MULTI-SOURCE SPATIAL DATA FOR STRATEGIC FOREST MANAGEMENT IN PONDEROSA PINE ECOSYSTEMS

Authors: YAGUANG XU(1), John W. Prather(1), Haydee M. Hampton(1), Ethan N. Aumack(1), Brett G. Dickson(2), and Thomas D. Sisk(1)

(1) Center for Environmental Science and Education, Northern Arizona University, Flagstaff, AZ 86011: yaguang.xu@nau.edu
(2) Department of Fishery and Wildlife Biology, Colorado State University, Fort Collins, CO 80521 and USDA Forest Service, Rocky Mountain Research Station, Flagstaff, AZ 86001,dickson@cnr.colostate.edu

Maps of forest structural parameters at multiple scales are spatial data foundations for landscape analysis aimed to aid the forest management decision-making process. Remote sensing imagery and a variety of ground measurements are the major data sources available for developing forest structure maps. Using existing ground datasets from multiple sources and remotely-sensed imagery with multiple spatial resolutions, the Forest Ecosystem Restoration Analysis (ForestERA) research team has mapped tree composition, total basal area, tree density, and canopy cover over four million acres of ponderosa-dominated forests in northern Arizona (at 10 to 90 meters spatial resolutions). This poster presents the techniques used to build these structural layers and the resultant maps. In this poster, we focus on two practical data analysis techniques for constructing these data layers and estimating the accuracy of them: 1) using ground plot measurements in varying-sized groups, and 2) deriving regional forest structural layers from low-cost aerial photographs such as USGS digital orthophoto quadrangles. The two methods were developed in this research project for constructing forest structural maps to fit the requirement of landscape ecological analysis when the number of ground samples is insufficient, or the distribution of ground data are not ideal, for spatial modeling.

A NOTABLE LIMITATION ON MAPPING LARGE-AREA FOREST STRUCTURE USING TM IMAGERY AND GROUND PLOT DATA

Authors: Yaguang Xu(1), John W. Prather(1), Haydee M. Hampton(1), Brett G. Dickson(2), Ethan N. Aumack(1), and Thomas D. Sisk(1)

(1) Center for Environmental Science and Education, Northern Arizona University, Flagstaff, AZ 86001, yaguang.xu@nau.edu
(2) Department of Fishery and Wildlife Biology, Colorado State University, Fort Collins, CO 80521, dickson@cnr.colostate.edu

Maps of forest structural attributes such as tree density and basal area form the spatial data foundation required for fire behavior and wildlife habitat analysis (or modeling) at landscape and regional scales. TM imagery coupled with training data from ground plot measurements is a major source of spatial data used to build such structural attribute maps. Over two years we mapped forest structural attributes across two million acres of the ponderosa pine dominated forest in northern Arizona using regression tree analysis, with and without committee models. We encountered a considerable limitation in using TM imagery for mapping forest structure, although the technical approach using the two types of data is usually considered an efficient way to map structure over large forested areas. Our results demonstrate that where forest was very dense (> 800 trees / ha) or extremely sparse (< 100 trees / ha), accuracy of the structural maps decreased significantly. We were unable to erase this problem by changing predictive layers, ground plot training datasets, or analytical methods. Thus, we conclude that this problem decrease in accuracy was caused by insensitivity of TM imagery to the areas with very low tree cover, and weak penetrating capacity of TM imagery to very high tree cover areas. A canopy cover map derived from USGS Digital Orthophoto Quads (DOQs) was used to illustrate this conclusion. The consequences of the genetic drawback of TM imagery may result in profound effects on some fire behavior modeling and wildlife habitat analysis in which the structural maps are used as primary input layers.

Oral presentation abstract for the 73^rd Annual Meeting of the Cooper Ornithological Society. 30 April–3 May 2003. Flagstaff, Arizona.

USING REMOTELY-SENSED FOREST STRUCTURE DATA TO HELP PREDICT THE EFFECTS OF FOREST MANAGEMENT ON PONDEROSA PINE FOREST BIRDS IN NORTHERN ARIZONA.

Authors: Prather, John W., Yaguang Xu, Thomas D. Sisk, Haydee M. Hampton, and Ethan N. Aumack.

The Forest Ecosystem Restoration Analysis (ForestERA) project is a landscape-scale assessment of forest structure across northern Arizona. The goal of this project is to link forest structural characteristics to wildlife distributions and fire behavior so that managers may better prioritize forest treatments and assess cumulative effects of those treatments. As part of this effort we are using remote-sensing to create highly accurate GIS layers of forest composition (vegetation type) and structure (canopy cover, basal area, and tree density) across northern Arizona. We are collaborating with individuals and groups that have field data related to avian presence and abundance in this region to model the distributions of many avian taxa. Initial modeling efforts suggest that this effort will be successful in better defining avian responses to forest management. For example classification trees reveal that many passerine birds are absent in areas of high basal area and tree density. This suggests that treatments designed to reduce fire threat will be beneficial to many of these species. In addition, models of potential habitat for Mexican Spotted Owl and Northern Goshawk, based on literature and expert opinion, can be used to determine the impacts of proposed treatments on habitat availability for these species. By addressing these issues before treatment implementation begins, planners can better design treatments to address concerns about biodiversity and can assess the cumulative effects of multiple treatments on species of concern.

Oral presentation abstract for the Ecological Society of America Meeting: August 1 - 8, 2004, Portland, Oregon.

AN ASSESSMENT OF VARIOUS DEFINITIONS OF THE WILDLAND-URBAN INTERFACE (WUI): IMPLICATIONS FOR MANAGEMENT IN A SOUTHWESTERN LANDSCAPE.

Authors: Prather, John W., Ethan N. Aumack, Haydee M. Hampton, Yaguang Xu, Brett G. Dickson, and Thomas D. Sisk.

Due to timber harvest, grazing, and fire suppression, many dry coniferous forests of the western United States have a structure that has made them prone to high-intensity crown fires. Coupled with a sustained drought, these conditions have recently resulted in a number of extremely large and damaging fires. Because of this, forest management in the southwest is increasingly focused on thinning treatments to reduce hazardous fuels and to help restore fire-adapted ecosystems to their historical range of variation. Congress and the federal land management agencies are committed to reducing the accumulation of wildland fuels that lead to catastrophic wildfires and unacceptable risk to forest ecosystems and human life and property. Much of this effort is focused on protecting communities and human infrastructure, especially through intensive fuels reduction treatments in the wildland-urban interface (WUI). Through the use of a Geographic Information System and predictive models of wildlife habitat we have examined how various definitions of the WUI might change the pattern of, and area covered by, intensive fuels reduction treatments on the Western Mogollon Rim, Arizona. In this region, populated areas are scattered throughout an ecosystem dominated by fire-prone ponderosa pine forests. Definitions ranging from a ½ mile buffer around “communities at risk” to guidelines in the Healthy Forest Restoration Act result in classification of 10% to 20% of the region as WUI. Addition of municipal watersheds, private property outside of “communities at risk”, and buffer zones to protect critical human infrastructure quickly raise the total area covered. Many of the areas identified as WUI are also important habitat for various wildlife taxa, including threatened and sensitive species such as the Mexican Spotted Owl and Northern Goshawk. Through the use of fire behavior models and wildlife habitat models we assess how intensive fuels reduction treatments within WUI areas might reduce fire hazard and impact critical habitat for various wildlife taxa. Our analysis quantifies the tradeoffs inherent in multiple management scenarios, and we evaluate the impacts of WUI policy decisions on multiple wildlife taxa.

Poster presentation abstract for the Ecological Society of America Meeting: August 4-9, 2002
Tucson, Arizona.

USING EXISTING DATASETS TO ADDRESS CONCERNS ABOUT THE EFFECTS OF RESTORATION TREATMENTS ON WILDFIRE.

Authors: Prather, John W., Haydee M. Hampton, Yaguang Xu, Thomas D. Sisk, and Ethan N. Aumack.

Assessing the effects of alternative land management practices on biodiversity is an important aspect of many conservation efforts. However, changes in management are driven by many factors, and in some cases the research needed to determine the effects of management practices on wildlife cannot be carried out in time to influence decisions. In such cases, use of existing datasets may be the only avenue available to assess potential effects of the proposed management practices. In the case of southwestern ponderosa pine, there is great interest in restoring forest structure to presettlement conditions. This effort is driven mainly by concerns about potential for catastrophic fires, and many treatments are slated to take place before data on their effects on wildlife will be available. By combining existing data on wildlife responses to forest management, expert advice on wildlife habitat requirements, and the spatial analysis capabilities of a geographic information system, estimates of the effects of treatments on wildlife can be assessed at the landscape scale. These results can then be used to adjust restoration efforts in ways that reduce negative impacts on species “at risk” due to restoration treatments. We provide here examples of how using these techniques can help merge management practices designed to restore ponderosa pine forests with those designed to protect sensitive species occupying those forests.

Oral presentation abstract for the 7^th Biennial Conference of Research on the Colorado Plateau.  Flagstaff, Arizona. U.S.G.S., Flagstaff, AZ, Nov. 3-6, 2003.

MODELING THE EFFECTS OF FOREST RESTORATION TREATMENTS ON SENSITIVE WILDLIFE TAXA: A GIS-BASED APPROACH.

Authors: Prather, John W., Brett G. Dickson, Yaguang Xu, Haydee M. Hampton , Etha, N. Aumack, and Thomas D. Sisk

Changes in the structure and composition of western forests due to fire suppression, grazing, and timber harvest, have resulted in increased threat of catastrophic wildfires and a general reduction in the health of these forests. Large-scale treatments designed to reduce the threat of catastrophic wildfire and restore ecosystem health are now being proposed in many locations across the Colorado Plateau. However, the effects of restoration treatments on biodiversity, have not been well-studied, leading to concern over implementation of treatments across large landscapes. Thus, there is a need for new methodologies linking management to the relationship between biodiversity and forest structure. Moreover, a comprehensive approach to assessing the potential effects of large-scale treatments on wildlife is needed to help monitor and protect species of concern. The Forest Ecosystem Restoration Analysis (ForestERA) Project is creating remotely-sensed data layers of forest structure and composition, researching the effects of different treatments on forest structure, and collaborating with researchers studying the relationships between forest structure and the distribution and abundance of selected wildlife taxa. Using a Geographic Information System (GIS) and specialized coverages, we are able to make general predictions about the effects of different types and placements of treatments on a variety of species in ponderosa pine forests in northern Arizona, including Mexican Spotted Owl, Northern Goshawk, Tassel-eared Squirrel, and many others. The response of these species to associated changes in forest structure can be characterized across large landscapes. By considering a range of treatment options, ForestERA tools will better allow managers to produce predictive models of forest change that may be overlaid with taxonomic distributions, inhabited areas, and other spatial data. Furthermore, ForestERA products will allow managers to assess larger patterns of biodiversity when undertaking the prioritization and placement of forest restoration treatments in the region.

IALE Abstract, July03, ver 1.doc

LANDSCAPE-SCALE MODELING OF FIRE AND BIODIVERSITY TO SUPPORT ECOLOGICAL RESTORATION OF PONDEROSA PINE FORESTS IN THE AMERICAN SOUTHWEST

Authors: Sisk, Thomas D., Haydee M. Hampton, John Prather, Yaguang Xu, and Ethan N. Aumack

Stand-replacing crown fires in the forested portion of the American Southwest have been increasing in size and destructiveness over the past five decades. The suppression of the historical fire regime of frequent, low-intensity ground fires has resulted in atypically dense stands of fire-prone ponderosa pine forest over large areas. Forest management intended restore the structural characteristics that would permit a return to the fire regime that has predominated over the past several thousand years is complicated by the inability of managers and the concerned public to explore alternative forest management practices and their likely outcomes, in terms of reductions of fire threat and the conservation of biodiversity. Currently, forest management issues are mired in public controversy and litigation, endangering forest ecosystem health, biodiversity, public safety, municipal water supplies, and other values. Our research program, which provides new data and tools for landscape-scale assessment of management options, was motivated by the need for new tools to guide forest management in these arid forests. Our GIS-based models of forest composition and structure allow spatial modeling of fire threat and wildlife habitat quality over areas of 3-500,000 ha. New modeling tools simulate alternative forest management treatments and, in turn, their likely impacts on fire behavior and habitat quality. These tools allow managers to prioritize treatment areas and allocate limited resources in a manner that maximizes the benefits of forest ecosystem restoration, while minimizing negative impacts on biodiversity during transition from degraded and fire-prone current conditions to a desired condition where frequent, low-intensity fire can be returned to the landscape, sustaining native forest structure and the biological diversity that depends on it.

Oral presentation abstract for the International Association for Landscape Ecology 6^th World Congress, 13-17 July 2003, Darwin, Australia.

MODELING FIRE AND BIODIVERSITY TO GUIDE ECOLOGICAL RESTORATION OF PINE FORESTS IN ARID NORTH AMERICA

Authors: Sisk, Thomas D., Haydee M. Hampton, John Prather, Yaguang Xu, and Ethan N. Aumack

Stand-replacing crown fires have increased in size and destructiveness in the American Southwest over the past five decades. Suppression of frequent, low-intensity ground fires has resulted in dense stands of fire-prone forest over large areas. Efforts to restore structural characteristics that would allow a return to historical fire regimes are hindered by the inability of managers and the public to compare the effects of alternative forest management practices on fire threat and biodiversity. Currently, forest management issues are mired in controversy, endangering ecosystem function, biodiversity, public safety, and municipal watersheds. Our research program, motivated by the need to examine the cumulative effects of many independent management decisions over large planning areas, focuses on the development of spatial data and modeling tools to guide landscape-scale planning. Data layers describing forest composition and structure allow modeling of fire threat and wildlife habitat over areas of several hundred thousand hectares. Modeling alternative forest management treatments and their effects allows managers to prioritize treatment areas and allocate limited resources to minimize negative impacts on biodiversity while maximizing restoration objectives, namely the return of frequent, low-intensity fire that sustains native forest structure and the biological diversity dependent upon it.

Oral presentation abstract for the 19^th Annual Symposium for the International Association of Landscape Ecology  – US Chapter, Mar 30 – Apr 04, 2004, Las Vegas, Nevada.

Title: EFFECTS OF RESTORATION TREATMENTS ON FIRE BEHAVIOR AND BIODIVERSITY: CAN FIRE HAZARD BE REDUCED WHILE MINIMIZING IMPACTS TO WILDLIFE?

Authors: Prather, John W., Brett G. Dickson, Thomas D. Sisk, Haydee M. Hampton, Yaguang Xu, and Ethan N. Aumack.

The Forest Ecosystem Restoration (ForestERA) Project is an effort to help land managers prioritize areas for restoration treatment and determine the effects of alternative management scenarios on attributes such as fire hazard and biodiversity. Using remotely-sensed imagery, we created Geographic Information System (GIS) data layers representing vegetation structure (basal area, tree density, and canopy cover) and composition across two-million acres of forested land in Northern Arizona. We also derived physiographic layers for the area (slope, aspect, and elevation) from a Digital Elevation Model (DEM). Using these basic data layers we were able to create models of fire behavior using the FlamMap fire-modeling program, and predictive models of habitat for sensitive wildlife taxa, such as Mexican Spotted Owl, Northern Goshawk, and Tassel-eared Squirrel. Areas of high fire hazard often overlap with critical habitats for these species, creating the perception that fuels reduction treatments in southwestern forests will have negatively impacts. While our models confirm that conflicts do arise, we were able to identify a variety of ways to mitigate those impacts while still significantly reducing the chance of catastrophic wildfire. Mitigation processes include choosing lower intensity treatments in critical wildlife habitat, restricting treatments from critical wildlife habitat, or distributing treatments spatially, such that they limit impacts on habitat in any one location. In addition, it may be possible to improve habitat for some species by identifying and treating areas where forest structure diverges from that preferred by the species in question. The ForestERA toolbox provides these and other approaches that can help managers prioritize areas and design plans to reduce fire hazard while minimizing impacts to sensitive species.

Oral presentation abstract for the Southwest Fire Initiatives Conference, 29 April 2003, Flagstaff, Arizona.

USING FLAMMAP AND REMOTELY SENSED FOREST STRUCTURAL DATA LAYERS TO PREDICT FIRE BEHAVIOR ACROSS LARGE LANDSCAPES

Authors: Prather, John W., Yaguang Xu, Haydee M. Hampton, Ethan N. Aumack, and Thomas D. Sisk.

The Forest Ecosystem Restoration Analysis (ForestERA) project is using the fire-modeling program FlamMap and remotely sensed forest structural layers to predict fire behavior in Ponderosa Pine forests across Northern Arizona. Initial modeling attempts suggest that the combination of remotely sensed data and FlamMap can be used to successfully model some aspects of crown fire behavior on large landscapes. This information could be valuable in locating areas of high fire hazard and prioritizing those areas for treatments designed to reduce fire danger. However, at present the conditions under which a crown fire would occur on any given part of the landscape are not predictable. This is due to inadequacies in current fuel models and the lack of a remote sensing platform that can accurately measure tree height and crown base height.

FlamMap is a simplified version of the widely used fire-modeling program FARSITE. It is more useful than FARSITE for predicting fire behavior across large landscapes because it assumes the entire landscape is on fire. However, it cannot be used to predict fire spread or changes in fire behavior due to changes in weather.

Most of the input layers used by FlamMap can be derived from Digital Elevation Models (DEMs), or from remote sensing imagery. This should prove useful as remotely sensed data become more widely available. New remote sensing technologies such as LiDAR should improve input layers to fire modeling programs and help make fire behavior predictions more accurate.

Page last updated August 16, 2005