ECOSYSTEM MANAGEMENT: A LANDSCAPE ECOLOGY PERSPECTWE

WATER RESOURCES BULLETIN VOL. 32, NO. 2

AMERICAN WATER RESOURCES ASSOCIATION

APRIL 1996

ECOSYSTEM MANAGEMENT: A LANDSCAPE ECOLOGY PERSPECTIVE I Mark E. Jensen, Patrick Bourgeron, Richard Everett, and Iris Goodman2

ABSTRACT: Ecosystem management is an evolving philosophy that many government agencies have adopted in the multiple-use, sustained-yield management of federal lands. The primary objective of this philosophy is to sustain the integrity of ecosystems (i.e., their function, composition, and structure) for future generations while providing immediate goods and services to an increasingly diverse public. This objective can be achieved through integrated land evaluation and optimal land use planning that promotes the maintenance or development of landscape patterns and processes that meet societal expectations within the limits of the land's ecological potentials. Landscape ecology and conservation biology principles are critical components of this philosophy. This paper describes how some of these principles can be efficiently used in formulating a framework for ecosystem management on federal lands. The role of landscape ecology in ecosystem characterization and description is stressed, and the appropriateness of integrated ecological assessments to ecosystem management is discussed. (KEY TERMS: ecosystem management; landscape ecology; conservation biology; land use planning.)

Advocates of sustainable resource flows from public lands (e.g., industry) generally acknowledge the merits of these values; however, they also emphasize that the goals and objectives of ecosystem management must include economic and social needs because humans are also part of the ecosystem (Heissenbuttel, 1995; Slocombe, 1993). Wagner (1995) suggests that ecosystem management is the "skillful manipulation of ecosystems to satisfy specified societal values." This definition is useful because it does not imply that optimization of biodiversity, ecosystem health and integrity, and commodity production be included in every ecosystem management effort. Instead, these values are articulated on a case-by-case basis given the planning objectives and the laws and regulations that apply to a given planning area. Management goals, therefore, determine the ecosystem(s) to be managed (i.e., the appropriate planning area), the goods and services, and the desired conditions (e.g., biodiversity, health and integrity, sustainability) of an ecosystem in ecosystem management (Wagner, 1995). Critics of ecosystem management justifiably point to the lack of consistent definitions for ecosystems, the ambiguity of describing their characteristics, and the imprecise nature of delineating their spatial locations [Fitzsimmons, 1996 (this issue); Frissell and Bayles, 1996 (this issue)]; however, concepts of ecosystem management and its application are rapidly evolving, and it would be counter-productive to hasten to an all-inclusive definition that could limit further philosophical development. However, neither should we broadly apply a developing philosophical approach

INTRODUCTION There are about as many definitions of ecosystem management as there are groups that advocate its use (or limitations) as a basis for land use planning. According to Wagner (1995), "Definers commonly couch their definitions in terms of their own values, and it is the differences in individuals and group values that produce the definitonal differences." For example, members of the environmental community commonly emphasize the values of preserving biodiversity, ecosystem health and integrity, and sustainability in their definitions of ecosystem management (e.g., Grumbine, 1994; Wilcove and Blair, 1995).

!Paper No. 95102 of the Water Resources Bulletin. Discussions are open until October 1, 1996. 2Respectively, Landscape Ecologist, USDA, Forest Service, Northern Region, Pacific Northwest Experiment Station, Federal Bldg., Missoula, Montana 59807; Senior Ecologist, Western Conservation Science Department, The Nature Conservancy, 2060 Broadway, Suite 230, Boulder, Colorado 80302; Range Ecologist, USDA Forest Service, Pacific Northwest Experiment Station, Wenatchee, Washington; and Research Hydrologist, U.S. Environmental Protection Agency, Office of Landscape Characterization Research and Development, Las Vegas, Nevada.

203

WATER RESOURCES BULLETIN

Jensen, Bourgeron, Everett, and Goodman

Evolution of Ecosystem Management Concepts Within the USDA-Forest Service

without reiterative monitoring, evaluation, and validation. Criticism of the ambiguity of ecosystems, their boundaries, and spatial locations appears to be a rehash of the continuum-classification argument (Brown and Curtis, 1952; Whittaker, 1967), but on a grander scale that now includes socioeconomic as well as biological attributes. Ever since Gleason (1917) published on "unit community" theory, we have known that each patch of vegetation is unique, but we can group vegetation into communities for discussion and management purposes. Ecosystems (classifications} are human constructs that only have value if they assist us in understanding complex biologicalsocioeconomic systems and aid us in conserving our resource base while meeting societal expectations. Although the future success of ecosystem management may now be limited in certain areas due to degraded resource conditions (Frissell and Bayles, 1996), we suggest that ecosystem management is currently our best option for conserving resource.s ~nd management opportunities for the future. In a Similar manner, some have suggested that the ambiguity in ecosystem definitions seriously impacts the value of ecosystem management in framing public policy (Fitzsimmons, 1996); however, the linkage of public expectations and resource conditions in ecosystem management is better suited to developing sound public policy than previous management philosophies. In this paper, we present a landscape-ecology-based description of ecosystem management that addresses many of the concerns currently being raised about its appropriateness as a guiding philosophy for land management. We begin by describing the evolution of ecosystem management concepts within the USDAForest Service to illustrate why a landscape-ecologyb.ased-framework for ecosystem management is appropriate to the goals and objectives of many federal and state land management agencies. Science principles and methods that support this framework are then discussed (e.g., hierarchy theory, pattern recognition, and system dynamics) to demonstrate their utility in ecosystem description. The role of landscape ecology and conservation biology in conservation area planning is described to illustrate the importance of such principles in the development of strategies for biodiversity maintenance. We conclude our paper with a brief description of how integrated ecological assessments that incorporate landscape ecology principles can be efficiently used in the realization of ecosystem management goals.

WATER RESOURCES BULLETIN

The legal precedent for ecosystem management ~y the USDA-Forest Service originated with the Organic Administration Act of 1897, which was more specifically interpreted by Congress through the MultipleUse Sustained-Yield Act (MUSYA) of 1960 (Jensen and Everett, 1994). Refinement of the intent of the MUSYA is evident in the National Environmental Policy Act (NEPA) of 1969, which directed that federal lands be managed to "encourage productive and enjoyable harmony between rna? and his envi~o~­ ment; to promote efforts which will pre~ent or eliminate damage to the environment and biosphere and stimulate the health and welfare of man; (and) to enrich understanding of the ecological systems and natural resources important to the Nation." The National Forest Management Act of 1976 reemphasized the importance of multiple-use, sustained-yield management and directed the Forest Service to develop long-term plans to describe how they would meet the intent of MUSYA and NEPA. Accordingly, the agency entered into an era of forest planning that included plan development, implementation, monitoring, and revision (Grossarth and Nygren, 1994; Morrison, 1994; Shepard, 1994). The discussion above is provided to illustrate that the laws (and presumably the societal values they portray) that govern USDA-Forest Service management of federal lands advocate sustainable flows of goods and services while ensuring for the long-term maintenance of ecosystem health and integrity. The For·est Service is responsible for implementing the intent of this legislative direction, which has required that the agency change its philosophies towards management given advances in ecosystem science and changes in public desires concerning its stewardship of National Forest System lands (Jensen and Everett, 1994; Kennedy and Quigley, 1994). The evolution to ecosystem management by the USDA-Forest Service required that the agency abandon its traditional mechanistic, reductionist world view (Botkin, 1990), which promoted use of agricultural-based production models in wildland management (NRC, 1990), reliance on "old forestry" concepts such as timber primacy (Shepard, 1994), and an interpretation of multiple use as the producti?n of commodities while maintaining other amenity values through identification of optimum yields of desired (often competing) uses. Kessler et al. (1992) suggest that this historical interpretation of multiple use is not appropriate to a society that increasingly demands the maintenance of healthy, diverse, and productive wildlands. These authors also suggest that 204

Ecosystem Management: A Landscape Ecology Perspective

The land evaluation method of Beek and Bennema (1972) provides a conceptual foundation for ecosystem management that incorporates Overbay's suggestions and is based on landscape ecology principles. In this approach, land evaluation includes inventory, classification, and analysis to determine optimal land uses (Zonneveld, 1988). Inventory requires data from relevant land properties that answer the following questions: What is there? Where is it? When is it present? How does it function? Analyses of such data address questions of "what, where, when, and how" in relation to the alternative land uses considered or the management actions to be implemented. Zonneveld (1988) suggests that this process is appropriate to both internal (one individual holding) and external (multiple holdings) integrated land evaluation. In the latter situation (e.g., regional land-use planning), he recommends that land evaluation consider sociological, ecological, technological, and economic factors. Simultaneous synthesis of these factors provides for optimum land-use planning or ecosystem management. Jensen and Everett (1994) provide a modification of Zonneveld's concepts (Figure 1) where ecosystem management is displayed as the optimum integration of societal desires and requirements, ecological potentials of the landscape, and economic plus technological considerations. The following steps (Jensen and Everett, 1994) describe how the land evaluation process may be used to achieve ecosystem management:

a management philosophy is required that recognizes that forest lands (as living systems) have importance beyond traditional commodity and amenity uses [i.e., they are important life-support systems (Dawkins, 1972)]. This philosophy should emphasize development of management objectives that relate to ecological and aesthetic conditions of the land and promote implementation of practices that maintain resource values and yields compatible with those conditions (Kessler et al., 1992). Various professional societies and groups have also emphasized the need for natural resource managers to take a more holistic, ecosystembased approach to land management (Lubchenko et al., 1991,;NRC, 1990; SAF, 1993). Overbay ( 1992) responded to the need by the USDA-Forest Service for a new management direction when he proposed it adopt ecosystem management in its stewardship of federal lands. Overbay (1992) defined ecosystem management as "the maintenance of sustainable ecosystems while providing for a wider array of uses, values, products and services from the land to an increasingly diverse public" and suggested that the following six principles be used to describe the initial components of ecosystem management: (1) multiple-use, sustained-yield management depends on sustaining the diversity and productivity of ecosystems at many geographic scales; (2) the natural dynamics and complexity of ecosystems must be considered; (3) descriptions of desired conditions for ecosystems should integrate ecological, economic, and social values; (4) coordination of goals and plans with affected publics is essential to success; (5) ecological classifications and inventories should be integrated; and (6) monitoring and research should be integrated with management to continually improve the scientific basis of ecosystem management. Many of these points (and some of the landscape ecology principles outlined later in this paper) were incorporated into the new proposed planning regulations for the agency (USDA-Forest Service, 1995).

1. Determine the desires and requirements of people who will be influenced by the planning outcome. 2. Des~ribe the ecological potential of the land for meeting stated societal needs. Such descriptions must include a description of the range of conditions required to maintain long-term system sustainability, a description of current conditions, and a description of desired landscape conditions that achieve societal needs. 3. If desired landscape conditions fall outside the range of conditions required for long-term system sustain ability, inform the people who will be affected. Public awareness of ecosystem potential is critical in developing achievable "desired future condition" strategies for land management. Public desires are refined through this process, based on an understanding of sustainable ecosystem criteria. 4. Once a socially acceptable, sustainable vision of the landscape is achieved, it is then contrasted against available technology to determine if it can be implemented. For example, in many instances, the desired landscape condition may differ from existing conditions. In these situations, available technologies must be considered to determine if the existing landscape can be changed to some desired set of conditions.

A LANDSCAPE ECOLOGY FRAMEWORK FOR ECOSYSTEM MANAGEMENT Landscape ecology (in its broadest sense) has been used in Europe for planning and land management since the term was coined by a German geographer, Carl Troll, in 1939 (Golley, 1994). In this role, landscape ecology was an integrating subject that synthesized information between hydrologists, geomorphologists, vegetation scientists, soil scientists, economists, sociologists, and land use planners (Last et al., 1982).

205

WATER RESOURCES BULLETIN

Jensen, Bourgeron, Everett, and Goodman

5. Determine what parts of the stated human desires can be fulfilled given economic factors. If resources are not available to construct the desired landscape, the public should be notified and alternative strategies developed. In most situations, shortterm economic reasoning and large management impacts contribute to situations that violate land ecological and human values (Zonneveld, 1988). Accordingly, these factors should be avoided in the development of strategies for ecosystem management.

to be clearly defined. The scientific products generally used in conservation and land management planning are descriptions, predictions, and prescriptions. Describing ecological systems of interest is a routine scientific endeavor that leads to the generation of classifications and maps. Predictions concerning future states of ecological systems under various management scenarios require determination of the relations that exist between patterns and hypothesized causal factors, or agents of pattern formation. Once a correlation or a cause-effect relation between pattern and process is determined, predictions are made using (1) the summarization of data and information performed during the classification process and/or the generation of maps, (2) statistical or simulation modeling and/or, (3) a combination of the first two approaches. Rigor is required if ecosystem predictions are to be appropriately translated into conservation or land management actions.

The steps outlined above refine human desires based on land ecology, technology, and economic considerations. Such refinement requires that the public be informed of land evaluation findings and that public opinion be solicited throughout the planning process. Basic principles of landscape ecology are critical to an understanding of the land's ecological potentials and provide the foundation on which the above ecosystem management framework is based. These principles are briefly reviewed in the following discussion.

General Principles of Hierarchy Theory

The concepts of scale and pattern are interwoven (Hutchinson, 1953; Levin, 1992). Complex ecosystem patterns, landscapes, and the multitude of processes that form them exist within a hierarchical framework (Allen and Starr, 1982; Allen et al., 1984; O'Neill et al., 1986). Scale dependency is very significant because the relationship between ecological processes (and the patterns they create) change with spatial scale (Turner, 1990; Davis et al., 1991). In recent years, considerable attention has been directed to describing the formal hierarchical organization of ecological systems. As applied to landscape ecology, hierarchy theory provides a needed framework for the description of the components of an ecosystem and their scaled relations. Four tenets of hierarchy theory are required for an understanding of landscape patterns and their dynamics (Allen et al., 1984; O'Neill et al., 1986):

ECOsYSTEM MflftfiGEMErtT

Figure 1. A Conceptual Landscape Ecology Based Framework for Ecosystem Management.

DESCRIPTION OF ECOLOGICAL SYSTEMS Ecological systems are groups of interacting, interdependent parts (e.g., species, resources) linked to each other by the exchange of energy, matter, and information. Ecological systems are considered complex because they are characterized by strong interactions between components, feedback loops, significant time and space lags, discontinuities, thresholds, and limits (Costanza et al., 1993). To describe ecological systems, pattern recognition techniques are used. These techniques have been used for many years in the medical and social sciences, which also deal with complex systems. Ecological systems can be described at many different scales (Levin, 1992). Hence, the spatial and temporal relations of specific ecological systems (or any of their components) need WATER RESOURCES BULLETIN

1. The whole/part duality of systems states that every component of a system, ecological or otherwise, is a whole and a part at the same time. For example, a forest (a whole) is made up of trees (the parts). However, at larger spatial scales, the forest is part of a regional landscape. In that case, the regional landscape is the whole and the forest becomes a part. The notion of whole/part duality is very important to the characterization of ecological systems. 2. Patterns, processes, and their interactions can be defined at multiple spatial and temporal scales. These scales need to be clearly identified.

206

Ecosystem Management: A Landscape Ecology Perspective

3. There is no single scale of ecological organization that is correct for all purposes. This is an important consideration because scientists often provide information/interpretations on ecological systems at a single or limited number of scales. 4. The definition of an ecological hierarchy (component patterns and processes) is dictated by the objectives of a study or planning endeavor.

Ecosystem Characterization Ecosystem characterization is ( 1) the process of appropriately relating pattern and processes at all scales of interest, and (2) the mapping of the entities thus determined. To match patterns and processes, ecological hierarchies such as the four templates described in the preceding section are superimposed. Figure 3 illustrates this point with a terrestrial ecosystem example of subalpine fir forests in the northern Rocky Mountains. In this example, the relationship of each pattern to specific biotic processes, disturbances, and environmental constraints emerge. For example, at the level of an individual tree in the canopy, the corresponding biotic process is tree replacement (e.g., from lodgepole pine to subalpine fir), which is driven by tree fall (disturbance), which is influenced by slope and aspect (environmental constraints). At a finer scale, the pattern is the seedling, the biotic process is germination, and the environmental constraint is the snowbank on which the seed lands. Above the individual tree level, the stand becomes the whole of which the individual tree is a part. The biotic process at the stand level is succession, which integrates individual tree replacement at larger spatial and temporal scales. Indeed, stand level succession from lodgepole to subalpine fir may take over 200 years. Disturbances at this scale may include windthrow and stand-consuming fires. Environmental constraints include slope, aspect, and landform. The proper match of patterns to their agents of formation is very important to ecosystem characterization because any incorrect coupling limits the predictiveness of future ecological system states which, in turn, inhibits development of realistic land management plans.

Pattern Analysis Implementation of hierarchy theory in the description of ecological systems is achieved through explicit characterization of the scaled relations that exist between the patterns of interest (e.g., species, habitats) and the ecological factors that determine such patterns [i.e., the agents of pattern formation (Urban et al., 1987)]. This type of ecosystem characterization is commonly called pattern analysis (Bourgeron and Jensen, 1994) and can be illustrated using fish distributions as the ecological pattern of interest. Fish species may be visualized as exhibiting different patterns of organization (e.g., individuals to metapopulations) that follow different spatial and temporal scales (Figure 2a). The formal definition of the hierarchical arrangement of fish distribution patterns is important because it requires explicit statements about (1) the spatial and temporal bounds of each pattern and (2) the order in which these patterns are nested. Such an objective-specific exercise provides the basis for identification of pattern formation agents (Bourgeron et al., 1994; Urban et al., 1987). The agents of pattern formation can be organized into different hierarchies of biotic processes (Figure 2b), disturbance processes (Figure 2c), and environmental constraints (Figure 2d). Biotic processes important to an understanding of fish distribution patterns may include behavior or physiologic adjustment at the channel unit or stream reach level, dispersal or genetic exchange at the watershed level, and speciation or extinction at the river basin (or broader scales) in this example (Figure 2b). The specific spatial and temporal relations that exist between fish distribution patterns and biotic processes are efficiently described through this type of characterization. In a similar manner, the relation between fish distribution patterns and disturbances (Figure 2c) and environmental constraints (Figure 2d) can be described if each agent is clearly specified and its spatial and temporal bounds clearly identified. At the watershed level, for example, population or guild distributions may be viewed as responding to mass wasting or flooding disturbance processes which in turn are a function of local climate, geology, and landform environmental constraints (Figure 2).

System Dynamics Most ecological assessments or land management plans focus on existing ecosystem patterns which represent static structures. Such assessments or plans, however, cannot ignore the dynamic nature of these patterns. Ecological systems exhibit temporal changes along various developmental pathways, resulting in different types of organization. Furthermore, when system dynamics are considered in planning, most of the conceptual models used assume that all ecosystems reach an equilibrium or quasi-equilibrium (e.g., succession leads to climax) despite the fact that ecosystems are rarely in equilibrium. Although it may be convenient for scientific exercises (modeling) and planning to assume equilibrium at a given scale, it is irresponsible to ignore complex ecological system

207

WATER RESOURCES BULLETIN

Jensen, Bourgeron, Everett, and Goodman

(a)

(b)

Fish Distribution Patterns

Biotic Processes

' i i

I

4

en 10 j~

1

iC'G

~~ 103

~

. 1

10

0

Meters 2 Channel Unit

I

6

1rl

Reach

10

0

3

10

Meters 2

~

!

Watershed

I

I

River

IBasin

Channel Unit

I

Reach

10

6

Watershed

River

I Basin I

............................. - ................ _ ............................................................... ______ ,.................._ ....................................1....................................................................................................................................................................................... .

!

(c)

(d)

I

Disturbance Processes

l

Environmental Constraints

Ii i I

I

0 1

10

3

10

Meters 2 Channel Reach Unit 1

0

6

10

~ ~

Watershed I

3

10 River I Basin

Figure 2. Agents of Pattern Formation.

WATER RESOURCES BULLETIN

208

10

6

Meters 2 Channel Unit

I

Reach

Watershed

I

River

IBasin

Ecosystem Management: A Landscape Ecology Perspective

10 10 10

•a: 10 ~

10 10 10

EXAMPLE: ABLANASC

7 6

6 4 3 2

1

0

10

2

10

4

10

6

METERS

ABlA m SUBALPlNE FIR

10

8

10

10

12

10

2

P1C0 c LODGEPOLE PINE VASe a GROUSEWHORTLEBERRY

PLOT/STAND

LANDSCAPE

I

ECOREGION

CONnNENT

I

Figure 3. An Ecosystem Characterization Example of Subalpine Fir Forests in the Northern Rocky Mountains.

dynamics that display change along multiple pathways, discontinuities, surprises, and changing environmental conditions (Holling, 1986; Kay, 1991; Costanza et al., 1993). Of great importance to decision makers and land managers is the fact that in the course of undergoing change, ecological systems may follow different pathways. The limit of ecosystem development along each pathway is defined by a point at which the processes that result in changes may balance the processes that lead to ecological organization (Kay, 1991). The fact that ecosystems may have multiple limits (Holling, 1986; Bak and Chen, 1991; Kauffman and Johnson, 1991; Kay, 1991; Costanza et al.,1993), and that some are more stable than others should be considered in land management planning. For example, the climax community is the end-point of ecological succession and as such is an example of a limit (Kay, 1991). However, after a disturbance, succession may proceed along multiple pathways, and a given pathway may stop before the end-point of succession is reached, which leads to several successional limits for the same system. Furthermore, if there is a large scale change in the climate or in the regional pool of

species, new species may enter the system, resulting in new vegetation types (i.e., new limits). Trend analysis of ecosystem pattern and process relations is important to ecosystem management if the temporal dynamics of systems are to be understood. Trend analysis facilitates an understanding of ( 1) the nature of the ecosystem dynamics (stochasticity) and (2) the periodicities, limits, and trends of system dynamics. Assessments of historic or "natural" variability are currently being conducted by many land management agencies in an attempt to improve this understanding (Swanson et al., 1994). Such assessments are commonly conducted to meet three primary purposes: (1) to understand the formation of contemporary patterns; (2) to determine whether there is any long-term change in the ecological system; and (3) to assess whether there is any short-term change in pattern that can be correlated with global climate change, landscape fragmentation, or any other management scenario. A typical historical variability assessment is illustrated in Figure 4, which displays the relative percentage of ground fires versus crown fires over time in selected dry Douglas-fir forests of the Northern

209

WATER RESOURCES BULLETIN

Jensen, Bourgeron, Everett, and Goodman

CONSERVATION BIOLOGY

Rockies. In this example, both types of fire regime displayed characteristic ranges until fire suppression activities became effective in the mid-1900s. Following fire suppression, the ratio of ground to crown fires changed abruptly. Compared to the past 300 years, these forests are now outside of their historic range of variability with respect to fire dynamics. The results of this type of assessment do not necessarily imply that past patterns have to be recreated. Instead, this type of assessment provides a larger context for an understanding of the temporal dynamics that have influenced the system (Swanson et al., 1994). Such assessments are commonly conducted to determine if present conditions have been experienced historically or not. In the latter case (i.e., being outside of the historic range of variability) the ecological system of interest is in a state for which we have no information; therefore, our present knowledge and predictive abilities may not be useful for understanding system behavior, and present technology may be useless in preventing rapid large-scale change to the system (Hann et al., 1994).

The coarse and fine filter concepts of conservation biology (Hunter, 1991) are consequences of the hierarchical structure of ecological systems and of their long-term dynamics. In simple terms, the coarse filter concept states that if an entire ecological system (e.g., landscape) is managed, the parts (e.g., species) would be managed as well. The coarse filter has been advocated as a more cost-efficient way to maintain common species for which little detailed data might exist (Scott et al., 1990). The converse of the coarse filter concept is the fine filter concept, which states that parts need to be managed individually. In practice, the fine filter should be used to manage rare species or communities which otherwise would have fallen through the cracks of the coarse filter approach (Jenkins, 1976). It has become fashionable to oppose both management strategies; however, it is most appropriate to consider the two strategies not only as valid for different purpose but also as complementary for a given purpose. 1b understand this concept better, consider the two approaches in the light of ecological hierarchies developed for specific ecological systems and purposes. For example, in a northern Rockies subalpine forest landscape (Figure 3), consider the goal of maintaining the landscape within its historic range of variability (i.e, maintaining all communities and species found within that landscape). The plant community (e.g, subalpine fir/grouse whortleberry) is the appropriate level of ecological organization (the coarse filter) for managing the vegetation at the landscape level. The individual plants (species) become the parts (fine filter), as components of the higher level scale (plant community). The appropriate biotic processes, disturbances, and environmental constraints that need to be considered in conservation management are defined for both the coarse filter (in this case the plant community) and the fine filter (the species) by the ecosystem characterization example displayed in Figure 3. If the management goal, by contrast, is vegetation management at the level of an ecoregion, an appropriate scale of ecological organization is the cover type (e.g., subalpine fir forests), and plant communities become the relevant parts (the fine filters). Management objectives define the ecological system of interest and the appropriate scales of ecological organization that need to be addressed (i.e., the coarse and fine filters). Accordingly, management defined analysis scales determine the relevant patterns and agents of pattern formation that need to be addressed in conservation planning.

100,-------~==================~----~ CROWN FIRE *GROUND FIRE

1-

I

~ 80 ~ ~----~~----------------~~-r-1 0

Q

j

60 1-~:..____________---*~------;t-----'

u. 0

!Z

w a:

40

(,)

~

20 0~--~~--~--~------.-~--~--.-~

1000 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 YEARS

Figure 4. Historic Variability Characterization of Ground and Crown Fire Regimes Within Selected Dry Douglas-Fir Forest Environments of the Northern Rockies.

In the previous example (Figure 4), it is clear that these ecological systems display conditions that have never been experienced before. Therefore, there is limited ability to make predictions concerning the future of these forests under any scenario of conservation and management, and existing fire suppression technology is likely to be obsolete due to the increase in closed forest structures and associated fuel-loading conditions. The historic range of variability can be established for many patterns or processes of interest. It is a simple tool that, if used wisely, assists determination of the risks, consequences, and costs of a particular management decision. WATER RESOURCES BULLETIN

210

Ecosystem Management: A Landscape Ecology Perspective

suited for the conservation of patterns with both low spatial and temporal variability. Conservation areas protect bio-physical environments that are found within their boundaries. Therefore, conservation areas are well suited for rare elements of biodiversity (species, communities, or ecosystems) because they incorporate within discrete boundaries specific patterns which are not found commonly and the biotic processes, disturbances, and environmental features that ensure their persistence. Management of such conservation areas aims solely at the local persistence of the pattern over time. Patterns and processes that exhibit high spatial and temporal variability are less likely to be confined within the finite boundaries of most conservation areas. Such patterns and processes are likely, therefore, to change over time within conservation areas boundaries. Accordingly, while the representativeness of a conservation area will not change in terms of the biophysical environments (environmental constraints) that it includes, its representativeness in terms of the dynamic patterns and processes that it contains will likely change. Because these changes are part of the historic variability at the regional scale, such changes are important to the maintenance of regional scale patterns of biodiversity. For example, in a large. conservation area (which is a subset of a larger regiOnal fire-dominated landscape in northeastern Minnesota), Baker (1989) found that there was no spatial scale at which the environment within the study area would be in a temporally stable patch-mosaic due to temporal fluctuations in the fire regime because the landscape structure fluctuates significantly over time. A consequence of this observation is that there will likely be patterns that will become extinct within the conservation area. Regionally, these patterns would be perpetuated under natural fire regimes because no two areas of the same size will be in synchrony. However, the local restoration of natural processes does not ensure that regional patterns of biodiversity might persist. In the case of the northeastern Minnesota conservation area, Baker (1989) further noted that current natural fire regimes could result in local landscape structures that are not adequate for maintaining viable moose populations. If maintaining regional moose populations in areas with natural fire regimes is a conservation goal, there is a need for having several asynchronous fire landscapes so that the moose populations will move from one area to the next in this example. The hierarchical structure of ecosystems and their dynamics have four important implications for the design of conservation areas that promote maintenance of regional patterns of biodiversity rather than rare elements only: (1) regardless of their size, no conservation area will be large enough in itself to allow

Representativeness Assessment The evaluation of conservation areas requires both a rationale and information concerning the biotic and abiotic components of a study area. The rationale (or set of conservation values) commonly used in such evaluations include rarity, diversity, and representativeness. Rarity and diversity (which focus solely on the biotic elements of an area) are relatively straightforward criteria that are routinely addressed in land management planning. Representativeness, however, implies that a preserve (or system of preserves) should contain both the biotic and abiotic (e.g., landforms, geology) features that represent the range of natural variation found within some land class or region (Austin and Margules, 1986). Accordingly, assessments of conservation area representativeness require coarser scale descriptions of the regional patterns of environmental and biological variability, the environmental relations of biota, and actual biotic distribution patterns (Austin and Margules, 1986; Margules et al., 1987; Mackey et al., 1988, 1989; Margules and Stein, 1989; McKenzie et al., 1989). Austin and Margules (1986) have proposed that four steps are required as precursors to representativeness assessment: (1) a hierarchical land classification of ecological units which considers both biotic and abiotic components (Austin and Margules, 1986; Mackey et al., 1988; Belbin, 1993); (2) a defi?ition of the relevant ecological properties of these units; (3) a method of allocating areas to the ecological units; and (4) a process for evaluating the representativeness of such areas.

THE ROLE AND LIMITS OF NATURE PRESERVES The role of nature preserves (or conservation areas) depends largely on three main attributes: (1) the rationale or value(s) used in their establishment (e.g., rarity, diversity, or representativeness); (2) the relationships among individual conservation areas; and (3) the actual management status of the conservation area (e.g., wilderness, private preserve, RNA, managed area, etc.). Strategies concerning the geographic distribution, size, shape, and management of conservation areas are consequences of these three attributes. Assessments of the role and limits of conservation areas require description of the effects that dedicated land uses have on the conservation of selected patterns and processes. Consideration of the hierarchical structure of ecosystems and their system dynamics suggests that most conservation areas are 211

WATER RESOURCES BULLETIN

Jensen, Bourgeron, Everett, and Goodman

types of policy questions or resource issues being addressed (Table 1). Climate change, for example, is a process that is meaningfully addressed at broader assessment scales. Grizzly bear viability assessments, by contrast, commonly assess conditions within the horne range area of the species (i.e., regional or subregional scale). Soil erosion assessments are commonly related to patterns of soil types as they occur on landforms, which are usually displayed at landscape (watershed) or land unit levels of assessment. Broad-scale ecological assessments facilitate improved land management decisions by placing more localized ecosystem management strategies into proper context. Regional or subregional assessments of mature forest patterns, for example, can greatly assist development of management plans for "old growth" forest networks, which in turn can be used to help direct community development and timber sale activities at the landscape (watershed) or land unit level (Table 1). By placing allocation and regulation strategies into proper context, multi-scale ecological assessments (using scale-appropriate information) facilitate improved management decisions. Descriptions of the historic, current, and probable future states of different ecosystem components (e.g., vegetation, species habitat) are usually required in most ecological assessments. Three basic categories of maps are commonly used in such characterizations: (1) biophysical environment maps, (2) existing or historical ecosystem component maps, and (3) coordinated planning unit maps. Biophysical environment maps are used to describe land or aquatic units that behave in a similar manner given their potential ecosystem component composition, structure, and function (Bailey et al., 1994). Such maps commonly delineate areas with similar response potential and resource production capabilities and are constructed based on landscape components that do not display high temporal variability at a given scale of mapping (e.g., regional climate,

for dynamic equilibrium unless the system is self-contained; (2) no conservation area will be fully representative of all the spatio-ternporal variables that contribute to biodiversity; (3) experience gained from one area has to be extrapolated to other areas carefully; and (4) conservation areas need to be integrated in a matrix of appropriately managed areas if the persistence of desired patterns and processes (biodiversity) at the regional or continental scale is to be achieved. These points suggest that ecosystem management is required if broad-scale objectives for biodiversity are to be achieved.

INTEGRATED ECOLOGICAL ASSESSMENTS Ecological assessments are an important component of any strategy for implementing ecosystem management (Slocornbe, 1993; Jensen and Bourgeron, 1994) and are commonly conducted when major land management or regulatory decisions need to be reevaluated. The need for an ecological assessment should be identified through ongoing monitoring efforts based on one or more of the following items: (1) initial assumptions concerning ecosystem pattern and process relations change due to new knowledge; (2) social values and human needs change resulting in a situation in which new issues are identified that require different levels of assessment for their resolution; or (3) significant change occurs in the ecosystem patterns or processes of a planning area. Hierarchy theory suggests that ecosystems may be described at different spatial scales and that levels of ecosystem organization at coarser scales bound the range of ecological properties that emerge at finer scales (Allen and Starr, 1982; Allen et al., 1984; Bourgeron and Jensen, 1994; O'Neill et al., 1986). Accordingly, ecological assessments should be conducted at multiple spatial scales dependent on the

TABLE 1. Relation Between Ecological Assessment Scales and Selected Resource Issues Assessment Scale

General Size Range (km2)

Climate Change

Global

>106

X

Continental

105-106

X

Regional

104-105

X

Subregional

102-104

Landscape (Watershed)

101-1o2

Land Unit

.1-10

Site