Remote Sensing and GIS in Ecosystem Management

By Al Sample

Recent advances in remote-sensing technology and the processing of remote-sensing data through geographic information systems (GIS) present ecologists and resource managers with a tremendously valuable tool -- but only if they are able to understand its capabilities and capture its potential.

Remote Sensing and GIS in Ecosystem Management identifies and articulates current and emerging information needs of those involved with the management of forest ecosystems. It explores the potential of remote-sensing/GIS technologies to address those needs, examining:

• the need for landscape-scale analysis to support forest ecosystem research and management
• current challenges in the development of remote-sensing/GIS applications
• case studies of different forest regions in the United States
• the potential for further development or declassification of military and aerospace remote-sensing/GIS technologies

As well as providing important information for ecologists and resource managers, the book will serve as a valuable resource for legislative and judicial policymakers who do not have a technical background in either remote sensing or resource management but who are nonetheless called upon to make decisions regarding the protection and management of forest ecosystems.

• Language: English
• Publisher: Island Press
• Release date: Apr 24, 2013
• ISBN: 9781610913096

Book preview

1994

Introduction

V. Alaric Sample

Ecologists and resource managers alike are being challenged by the need to greatly expand the scale on which they analyze and manage forest ecosystems. This need arises partly from the growing public concern over the loss of sensitive plant and animal species, in the United States and worldwide. The rate of loss is unprecedented in human history and is especially pronounced in species associated with late-successional forest ecosystems, which continue to dwindle as human population expands. At the same time, the current single-species approach to bringing sensitive species back from the brink of extinction is raising broad concerns due to its unpredictable, and sometimes extensive, economic and social impacts. But it is also of concern to ecologists who fear that it simply won’t work, and that a more comprehensive approach is needed—one based on protecting entire plant and animal communities by using an ecosystem management approach at a landscape scale.

As ecologists grapple with the incipient science of ecosystem management, so resource managers are beginning to grapple with what amounts to a fundamental shift in the way they conceive of and manage forest resources. After more than a century as the guiding principal of forestry, sustained-yield management of a few commercial tree species is evolving and gradually being replaced by another principle: the protection and sustainable management of forest ecosystems, taking into consideration the full array of plant and animal species that occupy these complex natural communities.

Recent advances in remote sensing technology, and the processing of remotely sensed data through geographic information systems (GIS) to directly address specific analysis or management queries, present ecologists and resource managers with a tool of tremendous potential value—but only if they understand its capabilities sufficiently to capture that potential. Application of remote sensing/GIS technology has helped ecologists and resource managers recognize landscape-level ecological issues. However, use of this relatively new technology to support ecosystem-based landscape management is still in its early developmental stages. Remote sensing/GIS technology has advanced rapidly in recent years, but it is not clear that ecologists and resource managers are fully aware of its expanding capability to address their needs for landscape-scale spatial information. Nor is it clear that ecologists and resource managers have adequately identified and articulated their information needs to specialists in remote sensing and GIS applications who may be able to match those needs with new or emerging capabilities.

The central purpose of this volume, then, is to identify and articulate current and emerging information needs of ecologists and resource managers for use in policy development and decision-making regarding the management of forest ecosystems, and to explore the current and potential applications of remote sensing/GIS technology to address those information needs.

A second purpose is to provide a basic information resource to legislative and judicial policymakers who do not have a technical background in either remote sensing or resource management, but who are nonetheless called upon to make decisions and provide direction regarding the protection and sustainable management of forest ecosystems. Policymakers confronted with issues over land and resource allocation are often presented with visual representations developed with remote sensing and GIS technology, but many have little or no understanding of how the information is developed, what its limitations are, and how different representations can be derived from the same basic data.

This book represents an unusually high level of coordination and collaboration among leading landscape ecologists, resource managers, and remote sensing/GIS applications specialists, and lays a foundation for greater cooperation in the future. The book is organized into three major sections. Part I is a general description of the need for landscape-scale analysis to support forest ecosystem research and management, and current challenges in the development of remote sensing and GIS applications for this purpose. From an ecologist’s perspective, Jerry Franklin discusses the need for a multiscale spatial and temporal approach to understanding and maintaining late-successional ecosystems, including old growth. An effective strategy for protecting biological diversity must look beyond forest reserves and consider the entire landscape matrix, including lands that will continue to be managed primarily for economic purposes. Bill Gregg takes a resource manager’s viewpoint and sees GIS technology as a catalyst for building constituencies for ecosystem management. He sees the increased use of remote sensing/GIS and communications technology as fostering a shared understanding of the complex tradeoffs involved in managing ecosystems to meet local and regional needs while furthering the goal of a sustainable biosphere. As a remote sensing /GIS applications specialist, Roger Hoffer asserts that the recent explosion of data available through these technologies poses new challenges for effectively utilizing the data to produce the information needed to facilitate ecosystem management. An important role of GIS applications specialists is to assist researchers and managers in understanding the choices they face in terms of the type and characteristics of the information desired, the accuracy and reliability of the data needed, and the economic tradeoffs of various sources of data and processing techniques.

Part II consists of case studies of four major forest regions of the United States—the Pacific Northwest, the Southwest, the southern Appalachians, and the northern Lake States—each with a different set of ecological and resource management concerns in a different social, economic, and cultural context. For each regional case, an ecologist and a resource manager each describe their specific information needs, and a remote sensing/GIS applications specialist describes the current and emerging technological capabilities for addressing those needs.

As home to some of the nation’s largest remaining areas of native old-growth forest, the Pacific Northwest was the focus for several early efforts at ecosystem classification using remote sensing and GIS technology. James Rochelle, Peter Morrison, and Warren Cohen each examine aspects of the evolution and future of techniques used to compare historical and current extent, location, and condition of late-successional forests in western Oregon and western Washington. To Rochelle, the key to the development of an ecosystem management approach is the examination and interpretation of spatial patterns of forests as they exist now and as they could exist in the future. Historical trends toward greater fragmentation of these ecosystems, adds Morrison, could be redirected through the application of landscape approaches to forest management. Cohen looks ahead to emerging techniques for correlating the particular spatial and spectral properties of particular sources of remotely sensed data to specific biological attributes to improve the mapping of Pacific Northwest forests. Tom Spies articulates the fundamental change to a multiscale spatial and temporal perspective—in both ecological analysis and resource management—needed to understand and maintain late-successional ecosystems, including old growth. Improvements are needed in our scientific understanding of ecological characteristics such as structure, composition, dynamics, and function. But to achieve ecosystem management on the ground, it is also important to identify and incorporate components in management activities such as objective setting, inventory, planning, implementation, and monitoring. Remote sensing and GIS techniques will continue to evolve and will be important tools for understanding the composition, dynamics, and interactions of landscape components in these unique and highly valued forest ecosystems.

Using Pacific Northwest forests as a case study, John Sessions, Sarah Crim, and Norm Johnson describe an important breakthrough in the development of spatial computer models to guide resource management decision-making. Most of the resource management decision models used in the past, such as the FORPLAN model developed by the USDA Forest Service to guide national forest planning, have used a linear-programming approach. Complex ecological interactions must be approximated and expressed as linear relationships, and then incorporated into the model as a series of constraints on the maximization of a particular objective, such as timber harvests or present net value (PNV). Such models can describe the optimal allocation of lands among a variety of land uses, but they offer little guidance as to exactly where on the landscape those uses will take place, or how different land uses on adjacent areas will relate to one another when viewed at the landscape scale. Sessions, Crim, and Johnson unveil what may become a model for planning and decision models of the future, overcoming the limitations of nonspatial models. Their prototype decision model is designed to allow scientists and resource managers to test a variety of ways to meet resource management objectives with a much clearer idea of the landscape-scale cumulative effects on aesthetics, water quality, biological diversity, and other important values.

The native aspen/pine forests of montane Arizona and New Mexico once formed an unbroken band along the mountains that ranged from the border of Colorado to the Chihuahua Mountains of northern Mexico. Today, the remaining forests provide important habitat for the northern goshawk, the Mexican spotted owl, and many other plant and animal species that characterize these upper-elevation late-successional forest ecosystems. Ecosystem classification using satellite imagery has helped to determine the spatial distribution of the remaining areas of old-growth forests, as well as the composition, structure, and condition of these forests. Margaret Moore describes how the determination of key ecological parameters using remote sensing and GIS technology can serve as an aid in the entire inventory, monitoring, and decision-making process. The controversy over the amount of late-successional habitat necessary to maintain viable populations of regional threatened and endangered species has increased as the amount of habitat has declined. Several old-growth inventories have facilitated protection efforts by using GIS technology to quickly identify probable old-growth using a combination of crown closure, cover type, size classes, and topography. Craig Allen and Jessica Gonzales provide more specific descriptions of efforts by the National Park Service and U.S. Forest Service, respectively, to determine the distribution and abundance of probable old-growth areas in the Jemez Mountains in northern New Mexico. The usefulness of the analysis forecosystem management and planning is discussed, along with some of the limitations and problems encountered in conducting the analysis. Doug Schleusner, who is also involved in the northern New Mexico project, takes the evaluation a step further, exploring the spatial information needs in national forest-level planning and the role that remote sensing and GIS technology might play in decisions made at that level.

The highlands of the southern Appalachian Mountains in Georgia, South Carolina, North Carolina, Tennessee, and Virginia contain some of the last vestiges of the vast native hardwood forests that once covered much of the eastern United States. The loss and fragmentation of these forests has been linked to declining populations of many species of songbirds that summer there and winter in the Amazon and other areas of the neotropics that are also under development pressures. Most of the eastern hardwood forests are in private ownership; these lands have been extensively logged, with the result that the majority of these forests are second or even third growth. The region’s forests are under increasing pressure from rapid population growth, resulting in the development of many forest areas for retirement and seasonal second homes.

A large and fairly contiguous area of the southern Appalachians is under federal supervision as national parks or national forests. In 1989, the United Nations recognized the tremendous natural value of this area by designating it a World Biosphere Reserve. The Reserve is focused on the Great Smokies National Park. A larger zone of cooperation surrounds the park but the management of these lands, particularly national forest land, has been a source of continuing concern. Timber cutting, particularly clearcutting and the new roads that come along with it, are seen by many as threatening the ecological integrity of this especially valuable forest region.

In addition to the possible threat to songbird populations, there is considerable worry over what effects current forest management might have on the local black bear population. While much of the denning, dietary, and isolation requirements of the roughly 3000 bears in the region remain unclear, there is general agreement on some factors. These include large elevated tree cavities (associated largely with old-growth forests), an ample supply of acorns and other hard mast, and low road density. This last factor has become more important with the rise of illegal poaching of the black bear.

Ecosystem management in the South requires effective ecological land classification but, notes Charles Van Sickle, it must also recognize the social and economic context of these ecosystems and that this context varies widely across major physiographic regions. He describes in general terms the differences in approaches being used to analyze ecosystems in coastal plain, piedmont, mountains, and interior highlands. He provides a more specific description of a separate plan developed for an area with a special set of geographic and social features, but which spanned jurisdictional boundaries of three states and three federal administrative areas. David Cawrse discusses what can be done on federal lands in the area and describes the Forest Service’s efforts to evaluate the quantity, quality, diversity, and connectedness of old-growth stands in advance of timber sale planning. Broader resource management plans are developed with a recognition of desired future conditions at three levels: stand, watershed, and region.

Scott Pearson tackles the problem of fragmentation at the landscape scale, recognizing the natural patterns of heterogeneity in southern Appalachian forests and the importance of this heterogeneity to ecological processes. Information on ecological processes, GIS-based maps of natural communities, and projections about activities that modify forests are needed in order to implement management strategies that will minimize forest fragmentation. Richard Flamm and Monica Turner have developed a GIS-based spatial model that will permit resource managers and planners to simulate the effects of alternative management scenarios on landscape structure. The model simulates not only the amount of land cover change that occurs on the landscape, but the spatial expression of this change, allowing issues such as biodiversity conservation, the importance of specific landscape elements to conservation goals, and long-term landscape integrity to be addressed. Of particular value in these regions will be the model’s potential usefulness in enhancing coordination among different landowners in managing a regional landscape.

With the exception of a few limited areas in northern Minnesota and Wisconsin, the native boreal and northern hardwood forests that once clothed the upper Lake States were almost completely cleared away during the wave of European settlement and forest exploitation that crossed the area in the nineteenth century. However, some ecologists believe that, given time, the existing second-growth forest could continue its successional development and eventually recreate forests very similar to the old-growth forests found by the first European explorers and settlers. Several studies to improve the scientific basis for restoring old-growth characteristics are underway. Many have advocated that extensive areas of publicly owned forest land in this region be reserved from commercial timber harvesting in order to foster the development of these late-successional forest ecosystems. The goal is more than historical; it is an attempt to reestablish the species composition and diversity that existed in these relatively stable, highly developed forest ecosystems.

Thomas Crow provides a broad overview of the social and economic as well as ecological considerations that will guide the management of the northern Lake States forests. As the forests regenerated after initial development are reaching maturity, they are expected to play an important role in regional economic development. Nonetheless, there is a growing acceptance of more comprehensive ecosystem management strategies intended to protect a diversity of forest values. This approach requires different information for planning, analysis, and decision-making, as compared to the more narrow, traditional approaches to forest management. New applications of remote sensing/GIS technology are seen as critical to providing this information. David Mladenoff and George Host describe the special challenges the Lake States forests pose for ecologists and GIS specialists. These landscapes lack pronounced topography but have complex variations in soils and landforms that influence ecosystems. The region has a large number of tree species with a wide range of ecosystem characteristics, but with few differentiable spectral signatures. Among the specific information needs identified are cover mapping at various resolutions and temporal change detection, including succession and management effects and land use changes.

David Cleland, Tom Crow, J. Hart, and Joyce Thompson add that ecosystem management requires working definitions and inventories of ecosystems, as well as an understanding of ecological patterns and processes that vary at different spatial and temporal scales. There is a need to further develop GIS capabilities for arraying numerous elements of landscapes and evaluating features across a variety of scales. This kind of integration is crucial to resource managers in conducting spatial analyses of existing and desired future conditions, and analyzing management alternatives and effects. Remote sensing and GIS technology has been used to address the needs of both ecologists and resource managers, and Mark MacKenzie sees this technology as having provided a mechanism for intellectual exchange between these two groups. This integration is providing a more complete understanding of the system and effective ways to manage the resource. GIS analysis of presettlement spatial information with current land cover derived from remote sensing has been used to quantify the nature of past disturbance-induced land cover change. This may provide useful insights into ecosystem processes important to the restoration of late-successional forest ecosystems in the Lake States. When used with current and future spatial data sets, it may also be useful in detecting ecological changes due to succession, management, and disturbance.

Part III is a general description of the potential for further development (or declassification) of military and aerospace remote sensing/GIS technologies and their application to address the heretofore unmet needs for ecological information. Forrest Hall describes NASA’s efforts to assist other agencies, universities, and the private sector in adapting NASA-DEVELOPED technologies for broader scientific and commercial application. Hall also illustrates NASA’s commitment to new environmental applications of space-based remote sensing technologies, using as an example the BOREAS project, an international scientific effort to analyze boreal forest ecosystems around the globe and assess their role in global climate change. The Department of Defense, notes Don Artis, is also beginning to explore opportunities to improve remote sensing and GIS capabilities to support ecosystem management by declassifying data obtained through a wide variety of surveillance satellites. Military remote sensing experts are still largely unaware of the particular needs and potential usefulness of their remotely sensed data for environmental and natural resource management purposes, and there is a new receptiveness to declassifying certain data that are no longer deemed sensitive for national security purposes. If these experts are made aware of these needs, there may be a significant potential for providing remote sensing information of types not currently available from commercial or civilian satellites.

Kass Green provides a summary and overview of the capabilities and limitations of remote sensing/GIS technology to address current ecological information needs, and the potential for broader application and use of GIS technology by landscape ecologists, resource managers, and conservationists. Green notes that, while the technology presents the tools for analysis, it does not guarantee the wisdom to use them effectively and wisely. She draws from several case examples to illustrate both the promise and the potential pitfalls of remote sensing/GIS technology, examining some of the fundamental assumptions of the technology and the sensitivity of analysis results to those assumptions.

Finally, Zane Cornett offers new insights into a somewhat less apparent value of remote sensing and GIS technology to ecosystem management—its use as a device for more effective public involvement in resource management decision-making by federal and state agencies and major private land managers. Many public concerns expressed during resource management planning are motivated by a sense of place, a concern over what the consequences will be for a particular location of special value or interest. As noted above, planning models used in the past have generally been nonspatial. Linear-programming models can specify the optimal number of acres to allocate among various land uses to maximize timber production, wildlife habitat, or present net value, but they are very limited when it comes to answering the question, What changes can I expect to take place on this particular acre of land? Remote sensing and GIS technology offers the first real opportunity to display detailed spatial information for a range of planning alternatives, and then make changes in land allocations in real time based on location-specific public concerns that may not have been expressed in earlier, more abstract stages of the planning process. More informed public choices regarding the protection and sustainable use of forest ecosystems are more likely to result in a consensus with both the breadth and depth necessary to survive the daily challenges of plan implementation.

Part I

Emerging Information Needs for the Protection and Sustainable Management of Forest Ecosystems

1

Developing Landscape-Scale Information to Meet Ecological, Economic, and Social Needs

William P. Gregg, Jr.

There is a growing consensus among managers and the public that natural resources should be managed according to principles of sustainability. Management practices must be ecologically sound, economically feasible, and socially and politically acceptable (Salwasser et al. 1992). Ecosystems are increasingly seen as fundamental units of interaction between people and Nature. Management goals must be holistic. They must reflect clear vision of the desired condition of the land to maintain aesthetics, soil processes, water quality and availability, atmospheric composition, native biological diversity, and areas of special sensitivity to disturbance. They must also reflect a clear vision for maintaining the health, economic prosperity, cultural diversity, and social well-being of the human communities of today and tomorrow. In addition, management practices to implement these goals must take into account the many uncertainties that affect natural and human systems, and the interacting effects of local decisions and regional and global conditions and trends. They must be based on an ethic of cooperation, in which people are seen as partners in planning and decision-making.

Ecosystem management is a process for implementing principles of sustainability. The process can apply to any geographic area where integrated approaches to achieve desired ecological conditions are practicable. However, ecosystem management most often applies to large geographic areas, which have been referred to as greater ecosystems (Grumbine 1990, Clark and Zaunbrecher 1987), regional ecosystems (Keystone Center 1991), regional landscapes (Noss 1983), and biogeocultural regions (U.S. MAB 1989). Although there is substantial recent literature on ecosystem management, a detailed definition has not yet become widely accepted. However, the following characteristics are frequently mentioned (Clark and Harvey 1988, Grumbine 1990, Noss 1990, 1992, Salwasser et al. 1992). Ecosystem management areas are typically open-ended. They are large enough to maintain viable populations of native species, including wide-ranging mammals. They accommodate natural disturbance regimes—that is, they can be expected to enable the long-term survival of native species under expected frequencies and intensities of fire, drought, temperature extremes, outbreaks of pests and diseases, and other stochastic events. Human use and occupancy occur at levels that do not cause ecological degradation. Management has a long time horizon, on the order of decades to centuries, that facilitates natural evolution of ecosystems and species. All organisms benefit from the management process. Charismatic, commercial, and other special-status species are considered in the context of the total management program. People are considered an integral part of the ecosystem. Management reflects understanding of the interactions of natural and human systems at many spatial and temporal scales.

Ecosystem management reflects and fosters an ethic of sustainability, and requires continuing consultation and cooperation among landowners and other stakeholders. As the number of stakeholders increases, the process becomes more effective in facilitating the complex tradeoffs that enable use of ecosystems to meet human needs in ways that sustain natural ecosystem functions and components in a changing