EMDS: Using a logic framework to assess forest ecosystem sustainability

June 24, 2017 | Autor: Keith Reynolds | Categoria: Forestry, Forest ecosystem
Share Embed


Descrição do Produto

EMDS

Using a Logic Framework to Assess Forest Ecosystem Sustainability

ABSTRACT

Keith M. Reynolds More and more, bioregional assessments are being viewed as essential components of ecosystem management. But forestry professionals and others have identified several challenges posed by this new brand of regional-scale analysis. This article summarizes key challenges facing assessments in general, describes use of a logic-based modeling framework with an example of evaluating forest ecosystem sustainability in particular, and discusses ways in which logic-based modeling can help address the challenges of bioregional assessment.

Keywords: assessment; logic-based modeling; sustainability

B

ioregional assessments collect, organize, and evaluate information about a bioregion as a prelude to regional planning and implementation, and are increasingly viewed by natural resource managers, scientists, and the public as essential components of ecosystem management (Johnson et al. 1999b). Since 1990, several major assessments have been conducted in the United States, including assessments for the Northwest Forest Plan (Forest Ecosystem Management Assessment Team [FEMAT] 1993), the Columbia Basin (USDA Forest Service 1996), and the Sierra Nevada (Sierra Nevada Ecosystem Project [SNEP] 1996). A hallmark of these assessments is their attempt to integrate a broad range of biophysical, social, and economic considerations into an evaluation of a bioregion. Although the specific questions driving them and the way they have been conducted have varied over time and 26

Journal of Forestry



June 2001

with geographic context, an underlying objective of assessments has been to evaluate forest ecosystem sustainability. A major accomplishment of the 1992 Earth Summit in Rio de Janeiro was the enunciation of principles for sustainable development of the world’s forest resources (United Nations 1992). Subsequently, the 11 signatory nations to the 1995 Santiago Declaration, representing about 90 percent of the world’s boreal and temperate forest cover, affirmed the recommendations of the Montreal Process that prescribed a set of seven criteria and 67 indicators for evaluating forest ecosystem sustainability (Anonymous 1995). The specifications of the Montreal Process are notable in two respects. First, the specifications provide relatively clear definitions of ecosystem attributes requiring evaluation. Second, the Montreal Process does not prescribe how criteria and indicators are to be interpreted to draw conclusions about the state of

forest ecosystem sustainability. Clearly needed are additional specifications that enable consistent interpretations of monitoring data on sustainability. This article summarizes key challenges facing assessments in general (Johnson et al. 1999b), describes the logic-modeling framework of the Ecosystem Management Decision Support (EMDS) system with an example of evaluating forest ecosystem sustainability in particular, and discusses ways in which the EMDS approach addresses the challenges of bioregional assessment. Assessment Challenges

Bioregional assessments are large, complex endeavors, and we are still searching for a practical balance among concerns for scientific rigor, efficiency, thoroughness, accommodation of values, and application to management. Assessments are driven by policy questions (Gordon 1999) related to perceived ecosystem crises that traditional science is not equipped to handle (Holling 1978; Walters 1986). Precise knowledge about relevant issues often is not available, forcing scientists to rely substantially on professional judgment (Swanson and Greene 1999). The integrative nature required of assessments has been characterized by Thomas (1999), but achieving high levels of information integration in as-

Input spatial data Assessment system

NetWeaver logic engine

ArcView ArcView extension

GUI

Input logic specifications

sessments has proven problematic (Johnson et al. 1999a; Thomas 1999). Gunderson (1999) suggests that assessments require a new form of integrative science. Assessments often emphasize quantitative, technological solutions, but the issues often involve values that are inherently social and political (Cortner et al. 1999). The interests and perspectives of participants who are not quantitatively or technologically oriented may be forced to take a back seat to science, but science alone cannot resolve public policy issues (Nelson 1999). Over-reliance on the quantitative methods of “hard” (biophysical) sciences may preclude balanced treatment of the “soft” (socioeconomic) sciences. Assessments are complex because they deal with numerous interdependent issues that are themselves influenced by myriad interacting factors. Communicating effectively about such complexity takes time and effort, but assessments also need to be performed in a reasonable time frame to maximize their policy relevance (Cortner et al. 1999). Effective public involvement also is important, and has significant implications for the success of subsequent planning and implementation activities. But it comes at a price: Assessments with effective levels of public involvement take longer to complete (Johnson and Herring 1999).

GUI Knowledge base

Output spatial

Output nonspatial

Figure 1. Ecosystem Management Decision Support (EMDS) integrates the knowledge-based reasoning technology of NetWeaver (Reynolds 1999b) into ArcView GIS.

Forest ecosystem sustainability

and

How EMDS Can Help

The EMDS system (Reynolds 1999a) integrates the NetWeaver knowledge-base engine (Reynolds 1999b) into ArcView GIS to provide knowledge-based reasoning for landscape-level ecological analyses (fig. 1). EMDS is a Microsoft Windows application and an extension to ArcView 3.2. The application can be downloaded or ordered on CD at www. fsl.orst.edu/emds. A knowledge base is a formal logic specification for organizing and interpreting information; in the strict sense, it is a form of meta database that represents a problem in terms of propositions about topics of interest and their logical interrelations. In knowledge-

Biophysical integrity

Economic feasibility

Social acceptability

Figure 2. Logic specification for evaluating sustainability of forest ecosystems in a bioregion. Ovals represent logic networks that evaluate propositions. Each premise has its own logic specification that may extend many more levels. NetWeaver knowledge bases are graphically built from modular components like this, simplifying incremental development of complex models.

base design, a policy question is translated into a testable propositional statement. For example, if the policy question is “Are forest ecosystems in the bioregion sustainable?,” the associated proposition might be as simple as “Forest ecosystems in the bioregion are sustainable.” The statement of the proposition by itself is inherently ambiguous

because sustainability is an abstract concept. However, the formal specification of a proposition makes the semantic content of the proposition clear and precise (fig. 2). In NetWeaver, strength of support for propositions is evaluated with fuzzy math (Reynolds 1999b), a branch of applied mathematics that implements qualitative reaJune 2001



Journal of Forestry

27

Table 1. Outline view of logic specification for evaluating biophysical integrity of forest ecosystems in a bioregion, by knowledge-base topic. Biodiversity Ecosystem diversity • Forest land area • Successional stage balance • Reserve area • Protected area • Fragmentation Species diversity • Forest species • Threatened species Genetic diversity Production capacity Timber Plantations Harvest volume Nontimber products Ecosystem health Historic range of variation • Insect infestation • Disease • Exotic plant invasion • Fire damage • Storm damage • Flood damage • Livestock impact Air pollution Ecological processes

Conservation Soil conservation • Erosion • Organic matter • Compaction • Soil toxins Water conservation • Protected water bodies Streams • Flow • Timing Water body biodiversity • Stream biodiversity • Lake biodiversity Water body physical condition • Stream condition… • Lake condition… Carbon cycle Total forest biomass Forest carbon budget Wood product carbon budget

NOTES: Details of the logic specification of propositions have been omitted to conserve space (see Reynolds 2000 for full documentation). Logic outlines for evaluating trends and for evaluating current conditions in relation to long-term goals (desired future conditions) are essentially the same in structure, although actual topic names for trends and desired future conditions are different within the knowledge base.

soning (Zadeh 1976). We might assert, for example, that forest ecosystems in a bioregion are sustainable to the degree that (1) integrity of biophysical environments is maintained, (2) maintaining the biophysical states is economically feasible, and (3) the consequences are socially acceptable (fig. 2). In the sense of logical discourse, requirements for testing the assertion of sustainability are both premises and propositions in their own right. The logical discourse on forest ecosystem sustainability is extended by specifying logic structures for each of the premises (table 1). Each iteration of discourse extends the logic structure another level deeper by logical decomposition of the level above. The pattern of discourse generally proceeds from abstract to concrete propositions, with a tendency for premises of a particular proposition to be less abstract than that proposition. 28

Journal of Forestry



June 2001

Eventually, each logic pathway terminates in a premise that can be evaluated by reference to data. Logic pathways in a knowledge base can thus be construed as a cognitive map of the problem that provides a formal data specification. The specification not only describes what data are to be evaluated, but how the data are to be interpreted to arrive at conclusions. A prototype knowledge base for comprehensive evaluation of the Montreal criteria and indicators has been proposed (Reynolds 2000). The knowledge base contains two primary logic networks that complement each other; the first evaluates current conditions in relation to long-term desired future conditions, and the second evaluates trends by comparing conditions between consecutive assessments. All evaluative functions in the Montreal prototype knowledge base are dynamically defined, so the logic specification for

evaluating forest ecosystem sustainability is quite general and could be applied to any country or major bioregion. Discussion

Knowledge representation is a critical challenge in assessment. What we think we know about ecosystems is often problematic and, although scientific frameworks are valuable organizing tools (Johnson et al. 1999a), a basic difficulty with scientific frameworks is that the concept itself is ill-defined. What constitutes a valid framework? Too often, the term connotes a conceptual model with no well-defined, underlying syntax, so the description is semantically vague at best and unintelligible at worst. However, a knowledge base is a particular form of scientific framework that has a well-defined syntax and set of semantics. Interpretation of data by a knowledge-base engine provides an assessment of system states and processes represented in the knowledge base as topics. Thomas (1999), discussing the difficulties of integrated assessment in FEMAT, foresaw increased reliance on “scientists trained or experienced in synthesis and integration,” but these so-called knowledge engineers are in short supply. One of the goals of EMDS development was to provide resource managers and scientists with a knowledge-based system for landscape analysis that could be used with little or no support from specialists in knowledge integration. Acquiring skill in the art and science of knowledgebase design with EMDS is not trivial, but it is also not difficult. Several EMDS users have now successfully developed significant applications with little or no help. Effectively communicating results, conclusions, and the derivation of conclusions can be formidable and timeconsuming. An advantage of a formal logic framework as a problem specification for assessment is that it provides a clear, concise, and intuitive explanation of the approach to problem solving and the derivation of conclusions. A critical element of the EMDS system is its knowledge-base browser that lets users graphically navigate the structure of an evaluated knowledge base.

Availability of data for assessments or evaluation monitoring is a significant and persistent issue, and is likely to remain so for the foreseeable future. Given this reality, another major design consideration in EMDS was the ability to robustly handle missing data. This consideration was a primary motivation for selecting NetWeaver as the knowledge-base design component and logic engine. EMDS can provide meaningful analyses even with incomplete information, and its analysis of data influence provides a simple basis for optimizing data collection, possibly realizing substantial savings in time, effort, and funds devoted to data collection. Considering the broader context of adaptive management, logic-based specifications for assessment hold promise for simplifying implementation of adaptive management, making it conceptually more coherent. Various comments in Johnson et al. (1999b) imply that an assessment is performed one time to assess the current state of a bioregion, and then an adaptive ecosystem management process is implemented. In other words, assessments are viewed as preludes to, rather than integral components of, adaptive management (fig. 3). The distinction commonly made between assessment and evaluation is artificial (“assess” and “evaluate” are synonyms), and can unnecessarily complicate implementation of adaptive management. Johnson et al. (1999a), for example, note significant problems with developing evaluation monitoring programs for implementation of the Northwest Forest Plan. The same problem has occurred in moving from assessment to planning and implementation in the Columbia Basin, Sierra Nevada, and other assessments. Data requirements for monitoring in the context of adaptive management should not even be an issue. Treating assessment as evaluation in iteration 0 of the cycle, rather than as a preliminary step external to the cycle (fig. 3), both clarifies the role of assessment and simplifies implementation of the adaptive management process. Assessment of a bioregion initially is motivated by concerns about the state of the system. Results of assessment provide the informational context for sub-

Goals

Assess

Knowledge

Technology

Inventory

Plan

Revised goals New knowledge

Evaluate Inventory

Adaptive management cycle

Act

New technology

Monitor

Figure 3. The adaptive ecosystem management process, modified with an inaugural assessment (adapted from Maser et al. 1994). Assessment often has been conceived as a step preparatory to adaptive management, rather than a special case of evaluation.

sequent planning and implementation of actions designed to improve the state of the system. The logic specification of a knowledge base designed for assessment also specifies the monitoring requirements for evaluating effectiveness of management (the action component of the adaptive management cycle). The entire adaptive management process becomes more coherent when criteria for evaluation at the end of a cycle are consistent, or at least consonant, with criteria that initiated the cycle. Networks of knowledge bases also hold promise for clarifying how information moves across scales in multiscale analyses. Integrated analysis can be extended via knowledge-based reasoning to multiple spatial scales (fig. 4, p. 30). Data from fine-scale landscape features such as watersheds is first processed by a knowledge base designed for that scale. Knowledge-base output, shown as evaluated states in the middle of the figure, then goes through an intermediate filter (for example, a spreadsheet or database application) to synthesize information for input to the next coarser scale. A second knowledge base processes the synthesized information to provide an assessment of landscape attributes at the top of the figure. Finally, knowledgebase outputs at the broader landscape

scale can feed back to the fine scale as context information influencing evaluations at that scale. This simple model (fig. 4) provides the basis for a formal logical specification for integrated analyses across scales. Hierarchies or networks of knowledge-based analyses would be consonant with ecosystem theories concerning the hierarchical organization of ecosystems (Allen and Starr 1982). The knowledge-based approach of EMDS is not a replacement for statistical, simulation, optimization, and other models typically used in assessments, nor does it replace the democratic process for adjudicating policy debate. Instead, EMDS complements these other tools by providing a logicbased framework for integration of diverse model results. Knowledge-based problem specification cannot guarantee parsimonious design of assessments, but logical formalism at least tends to strongly encourage an efficient mapping from concerns to data and analytical requirements (Cortner et al. 1999). The adaptive management process is a learning process by design (Walters 1986; Maser et al. 1994). We learn by doing, checking outcomes against expectations, and adjusting management actions accordingly. At any given time, knowledge about ecosystem function is June 2001



Journal of Forestry

29

CONTEXT (biophysical)

CONTEXT (socioeconomic)

KB (sustainability)

SYNTHESIS (socioeconomic)

SYNTHESIS (biophysical)

SYNTHESIS (biophysical/socioeconomic)

KB (biophysical)

DATA (biophysical)

KB (socioeconomic)

DATA (socioeconomic)

Figure 4. Knowledge-based integration across scales. Shaded features have been evaluated by an associated knowledge base. Synthesis is an intermediate step in which knowledge-base outputs from a fine scale are transformed for input to the next coarser scale of analysis. Context pathways indicate how knowledge-base outputs from a coarse scale feed back as data inputs to a fine scale.

incomplete, fragmentary, and perhaps erroneous in some respects. It may occasionally be the case that current concepts about ecosystem function are catastrophically wrong, requiring a major overhaul of logic specifications used to evaluate the state of a system. More typically, however, concepts undergo incremental, evolutionary transformation. The knowledge-based approach in EMDS has two useful features with respect to the learning process. First, any given knowledge base provides detailed, permanent documentation about how we conceive that a system should be evaluated, given the current state of knowledge. Second, the modular architecture of knowledge bases is well suited to incremental, evolutionary adaptation as knowledge changes. Literature Cited ALLEN, T.F.H., and T.B. STARR. 1982. Hierarchy: Perspectives for ecological complexity. Chicago: University of Chicago Press. ANONYMOUS. 1995. Sustaining the world’s forests: The Santiago Agreement. Journal of Forestry 93(4):18–21. CORTNER, H.J., M.G. WALLACE, and M.A. MOORE. 1999. A political context model for bioregional assessments. In Bioregional assessments: Science at the crossroads of management and policy, eds. K.N. John-

30

Journal of Forestry



June 2001

son, F.J. Swanson, M. Herring, and S. Greene. Washington, DC: Island Press. FOREST ECOSYSTEM MANAGEMENT ASSESSMENT TEAM (FEMAT). 1993. Forest ecosystem management: An ecological, economic, and social assessment. Portland, OR: USDA Forest Service and other agencies. GORDON, J.C. 1999. History and assessments: Punctuated equilibrium. In Bioregional assessments: Science at the crossroads of management and policy, eds. K.N. Johnson, F.J. Swanson, M. Herring, and S. Greene. Washington, DC: Island Press. GUNDERSON, L.H. 1999. Stepping back: Assessing for understanding in complex regional systems. In Bioregional assessments: Science at the crossroads of management and policy, eds. K.N. Johnson, F.J. Swanson, M. Herring, and S. Greene. Washington, DC: Island Press. HOLLING, C.S. 1978. Adaptive environmental assessment and management. London: John Wiley & Sons. JOHNSON, K.N., and M. HERRING. 1999. Understanding bioregional assessments. In Bioregional assessments: Science at the crossroads of management and policy, eds. K.N. Johnson, F.J. Swanson, M. Herring, and S. Greene. Washington, DC: Island Press. JOHNSON, K.N., R. HOLTHAUSEN, M.A. SHANNON, and J. SEDELL. 1999a. Forest ecosystem management assessment team assessments case study. In Bioregional assessments: Science at the crossroads of management and policy, eds. K.N. Johnson, F.J. Swanson, M. Herring, and S. Greene. Washington, DC: Island Press. JOHNSON, K.N., F. SWANSON, M. HERRING, and S. GREENE. 1999b. Bioregional assessments: Science at the crossroads of management and policy. Washington, DC: Island Press. MASER, C., B.T. BORMANN, M.H. BROOKES, A.R. KIESTER, and J.F. WEIGAND. 1994. Sustainable forestry through adaptive ecosystem management is an open-

ended experiment. In Sustainable forestry: Philosophy, science, and economics, ed. C. Maser. Delray Beach, FL: St. Lucie Press. NELSON, J.E. 1999. Forest ecosystem management assessment team assessments: Management review. In Bioregional assessments: Science at the crossroads of management and policy, eds. K.N. Johnson, F.J. Swanson, M. Herring, and S. Greene. Washington, DC: Island Press. REYNOLDS, K.M. 1999a. EMDS users guide (version 2.0): Knowledge-based decision support for ecological assessment. General Technical Report PNW-470. Portland, OR: USDA Forest Service, Pacific Northwest Research Station. ———. 1999b. NetWeaver for EMDS version 2.0 user guide: A knowledge base development system. General Technical Report PNW-471. Portland, OR: USDA Forest Service, Pacific Northwest Research Station. ———. 2000. NetWeaver knowledge base specifications for evaluation of forest ecosystem sustainability as prescribed by the Montreal Process. Available online at www. fsl.orst.edu/emds/sustain/sustain2s.htm. SIERRA NEVADA ECOSYSTEM PROJECT (SNEP). 1996. Final report to Congress, volumes I-III. Davis, CA: University of California, Centers for Water and Wildland Resources. SWANSON, F.J., and S. GREENE. 1999. Perspectives on scientists and science in bioregional assessments. In Bioregional assessments: Science at the crossroads of management and policy, eds. K.N. Johnson, F.J. Swanson, M. Herring, and S. Greene. Washington, DC: Island Press. THOMAS, J.W. 1999. Learning from the past and moving to the future. In Bioregional assessments: Science at the crossroads of management and policy, eds. K.N. Johnson, F.J. Swanson, M. Herring, and S. Greene. Washington, DC: Island Press. UNITED NATIONS. 1992. Forest principles: Report of the United Nations Conference on Environment and Development. New York. USDA FOREST SERVICE. 1996. Status of the Interior Columbia Basin: Summary of scientific findings. General Technical Report PNW-385. Portland, OR: USDA Forest Service, Pacific Northwest Research Station. WALTERS, C.J. 1986. Adaptive management of renewable resources. New York: Macmillan. ZADEH, L.A. 1976. The concept of a linguistic variable and its application to approximate reasoning, part III. Information Science 9:43–80.

Keith M. Reynolds (e-mail: kreynolds@ fs.fed.us) is research forester, USDA Forest Service, Pacific Northwest Research Station, 3200 SW Jefferson Way, Corvallis, OR 97331.

Lihat lebih banyak...

Comentários

Copyright © 2017 DADOSPDF Inc.