0721-B4

Forest ecosystem management: the millennium planning approach

Emin Zeki Başkent and H. Ahmet Yolasığmaz ¹

Abstract

Forest management in this century is obliged to promise holistic integration of various forest values, such as recreation, water quality, erosion control and wildlife protection, with the conventional wood production at forest level as they are demanded exhaustively by society. Forest ecosystem (landscape) management provides the right concept, principles, planning approaches and the techniques to design management towards the production of multiple values on a sustainable and shareable basis without jeopardizing the ecosystem health and integrity. This paper explains the conceptual framework of forest ecosystem management, discusses various approaches, and provides the principles and the improvements over the conventional approach focusing on management paradigm, modelling and software engineering techniques. It further looks at the pitfalls that are being discovered in implementing forest ecosystem management guidelines. The paper concludes that composition and spatial configuration of forest values must be quantified before attempting to integrate them within a management model. It is necessary to use both meta-heuristic and mathematical programming techniques with object-oriented modelling design to develop an ecosystem-based management model. Well-rehearsed management design, smarter use of modelling techniques and information technology, understanding forest ecosystems dynamics, and effective public involvement could make the realization of ecosystem management on the ground happen.

Introduction

Forest management of today has already begun to change towards more holistic approach of managing -sustaining and sharing- ecosystem values based on ecosystem sustainability. Traditional management approach, involving in extraction of specific forest resource mix, has not satisfied the global need of the society from the forested ecosystems. Of economic, ecological and socio-cultural values of forest ecosystems, the traditional approach focused on and favored economic value outcomes over other values which are now becoming important goals to achieve in management planning. In addition to timber, recreation, water, erosion protection values, biodiversity values including wildlife habitat conservation are globally accepted planning goals to achieve. It is now indispensable to revise or restructure forest management design and implementation within the planning process.

Forest ecosystem management, an ecologically sustainable management of forest ecosystems, aims to manage full range of forest values such as biodiversity, productivity, soil-water conservation, socio-economic benefits, and cultural heritage values for both present and future generation. Specifically, it is an approach to manage forested landscapes for both commodity and ecological values such as conservation of biological diversity, maintenance of productive capacity of forest ecosystems, and maintenance of forest ecosystem health and vitality by controlling spatial landscape structure and its dynamics. It is a unique concept capable of reflecting economic, ecological, and technologic and the social aspects of management planning. These elements are reflected simultaneously in decision-making process to provide the best stewardship and to sustain ecosystem functions while providing ordinary goods and services for people. The theme of forest ecosystem management is sustainable and sharable use of all forest values without jeopardizing the ecological integrity and the health (Yafee 1998, Grumbine 1994, Baskent and Jordan 1995, Salwasser 1994, Seymour and Hunter 1992).

Started in the first quarter of 1990's in North America particularly, ecosystem management was meant to protect the old growth forest and help advocate to conserve the world forest resources. It went further to cover the health and integrity of forest ecosystems calling for the management plans to include habitat areas, critical ecosystems as well as full participation of community in the plan. As such, the traditional planning approach, planning concept and principles along with the modeling approach have all been revised to respond to multiple needs of the society in managing the forest ecosystems. Therefore, ecosystem management has emerged.

In this paper, we explain forest ecosystem management paradigm, management concept, planning approaches, principles, modeling approaches, solution and software engineering techniques. Furthermore, we lay out some of the potential unfolding pitfalls in implementing forest ecosystem management.

Management Paradigm

Forest management paradigm is meant to establish management objectives and lay out a decision-making process. Currently, we can distinguish at least three different management paradigms - integrated resource management, forest zoning management, and natural disturbance model management. Integrated Resource Management advocates the use of each hectare of a forest simultaneously for different values such as timber, fuel wood and wildlife habitat. In reality, however, this approach has lived more in theory than in practice due to our inability to deal with the complexity of designing management that sustains multiple resource values from the same piece of forest.

Forest zoning management like Triad Approach and the Forestland Allocation Strategy (Seymour and Hunter 1992, Binkley 1997) segregates forest landscape into three zones: intensive timber management areas, ecological reserves, and so-called new forestry areas. Intensive timber management areas are used to maximize commodity production by exploiting the easily accessible and the most productive parts of the forest landscape. The ecological reserves are, on the other hand, permanently set aside and exclude human disturbances. The remaining, new forestry, areas are for management interventions that mimic natural disturbance processes, while producing limited commodity values. The premise of forest zoning is that intensifying silvicultural research and capital investment in the intensive management areas will offset losses in other areas (Binkley 1997).

Although the forest zoning paradigm has received a lot of attention and endorsements from environmentalists, it provides little flexibility in designing forest ecosystem management for a number of reasons. First, permanent forest zoning ignores a fundamental reality - most resource values are widely distributed spatially over the whole forest landscape (Franklin 1998). For example, water quality cannot be maintained by setting aside and isolating one part of a watershed. The same can be said for many wildlife habitats, where species require large tracts of forest in large contiguous patches. Second, intensifying management in certain areas will require additional cost and increase uncertainty. It is uncertain, for example, whether areas zoned for intensive management can maintain a high state of productivity over long time horizons.

Another management approach, the natural disturbance model (NDM), has also appeared in places across North America. Examples can be found in the Pacific Northwest of the United States and in Ontario and Alberta in Canada. Unlike forest zoning, the NDM is based on the principle of "ecological sustainability" and is focused on maintaining the full range of forest ecosystems with a continuum of resources of varying intensities interspersed across the landscape. It's believed that the diversity, structure and ecological processes of forests are molded by natural disturbances (Attiwill 1994), and, therefore, the primary objective in the NDM is mimicking the intensity, severity and frequency of natural disturbances, such as fire (Bergeron et al. 1999). Given this, Daishowa-Marubeni International in northern Alberta, for example, has forgone harvest block size limits and green up delay, leaves residual trees in clear cuts, and has extended stand rotation ages. It's widely accepted, however, that human disturbances cannot truly replicate the patterns generated by natural disturbances, even if we had the requisite landscape modeling know-how.

Similar to, yet subtlety different from the natural disturbance model, forest ecosystem management (FEM), holistic approach would use natural disturbance patterns as "a reference", not as processes to mimic. In FEM, natural disturbance patterns would help forest managers identify a forest management target. FEM would identify intervention schedules to explicitly control spatial structure, i.e. composition and configuration of forest conditions. This would require scheduling interventions in strategic geographic locations so as to ensure that a landscape's structural state continued capable of maintaining all ecological processes as well as providing other desired values. Since FEM wouldn't permanently divide a landscape into different forest uses, as in the case of forest zoning management, it wouldn't foreclose any future management options as our understanding of forest ecosystems improved.

These approaches, in fact, show the evolutionary process of forest management concept and the presented FEM is not revolution. As such, the approaches are complementary and not exclusive or fundamentally different from each other.

Principles of Ecosystem Management

Given a holistic approach to forest ecosystem management, a number of significant guiding principles underlying most of these management issues can be listed as following (Brooks and Grant 1992, Baskent and Yolasıgmaz 1999, Forbes G 1998):

• Forest management decisions must be based on an ecosystem perspective. This principle sees a forest as composed of organisms hierarchically organized into functional groups and linked through complex processes to their physical environment and each other. Thus ecosystem management recognizes the need to balance the sensitive management practices among the various components.
• The effect of forest management activities is evaluated over a range of spatial scales. Ecological processes occur at different spatial scales (micro sites, stands, watersheds, landscapes and regions) and the effects become quite different from each other. At a landscape scale, management must be applied to ecologically-based units, such as watersheds and ecological land classification divisions, and not human induced administrative units.
• Ecosystem management embodies the spatial effects of both human induced and natural disturbances on the management of forests. Native species, for example, developed adaptive strategies to maintain in the context of disturbances, which have created forest patches of various sizes and configurations. FEM reflects natural disturbance regimes that allow for the survival of local populations in minimum size patches of habitat, as well as metapopulations in functionally-connected patches on the landscape. Conservation of biodiversity requires permanent networks of protected areas that are connected by corridors acting as functional linkages between populations.
• The effect of forest management decisions is also evaluated over longer time horizon. Like spatial scale, expanding the time scale causes new issues to emerge. Long term site productivity, ecosystem resilience, ecosystem health and the long-term viability of biodiversity necessitates thinking across a range of ecosystems time scale. There are groups of native species, for example, that strongly associated with older seral stages of native forest types where short-term rotation, fibre-based forest management has the potential to eliminate them. FEM strategies focus on these species rather than on ubiquitous or less vulnerable species associated with young seral stages.
• Management strategies must leave future options open. Long and unresolved discussion about the uses of forestry, uncertainty about the future natural events coupled with lack of understanding basic principles of ecosystems functions forces management design process to avoid foreclosing on future management options. In short, ecosystem sustainability measured by a number of criteria and indicators becomes a central issue of planning.
• Ecosystem management ensures the sharable uses, rights and responsibilities of forest ecosystem values among the people in the globe. Since the effects of rehabilitation, deforestation or economic utilization of forest values are beyond jurisdictional borders, global view is indispensable, for example, keeping responsible for the people who misuse forest values anywhere in the world.
• Equal participation is necessary among the full range of stakeholders including NGOs in decision-making process. Public and private owners, environmental groups, naturalists, and the scientists should proceed a constructive and sincere partnership for better management decision.
• Forest ecosystem management necessitates better integration of ecological, economic and social relationships. No longer should timber production always have the priority over other forest uses though the commodity production is still necessary for the livelihood.
• Forest ecosystem management is compatible with its forecasting mechanism where key processes determining forest's internal structure, rules, or behaviors are dynamically accounted for. This principle requires the use of process based forest models, not empirical yield tables for example, to understand the long term forest ecosystem dynamics.
• Ecosystem management has three dimensions; the physical structure defined by topographic and geological structure, biologic composition represented by the fauna and the flora, and human dimension represented by the social, economic, spiritual, historical and the cultural values. The ecosystem management endeavor tries to balance these three dimensions for the well being of the future generation by providing future management options.
• FEM combines coarse-filter and fine-filter approach. The coarse-filter approach allows for planning of larger scale arrangements of communities, including their composition size, patch size, adjacency; patch size distribution and seral stage distribution. Fine-filter approach ensures that no important species are missed by the coarse-filter approach. Protecting critical ecosystems such as sensitive and unique sites or standing dead and fallen woody material are necessary to sustain elements of biological diversity.
• Ecosystem management by its nature utilizes the full advantages of adaptive management where understanding and learning the forest ecosystems dynamics happen by doing things in a stepwise manner (Haney and Power 1996).
• Since ecosystem management is inherently data intensive and particularly requires spatial data, information technology is essential. Information systems such as GIS and remote sensing provide the information infrastructure necessary to conduct ecosystem management.

These principles relate to the ideal design and implementation of FEM. However, how they should be implemented for a successful FEM is still under investigation.

Management Modeling Approaches and Solution Techniques

Conventionally, forest management modeling involves four elements (Baskent et. al. 2000):

• A forest description, e.g. areas of stand types and age classes;
• Management objectives and constraints, e.g. maximizing sustainable harvest flow;
• Forest performance indicators, e.g. measurements of forest growing stock and structure;
• Management strategy development, e.g. application rules, amounts, timings and locations of interventions.

Ecosystem management modeling involves all the above elements yet, unlike contemporary modeling approaches, incorporates spatial structure and processes (Baskent and Jordan 1995). First, a forest is described by its composition and configuration. Second, management objectives, constraints and performance indicators include spatial characteristics. FEM includes structural forest objectives, such as spatial and temporal arrangement and distribution of forest patches over a landscape. Last, spatial management strategies would specify the geographical format of interventions; for example, harvest opening sizes, shapes, and proximity to other forest conditions. These new elements are related to the spatial dimension of forest management planning and reflect the fact that forest management objectives should be functionally linked to forest condition and configuration as well as their dynamic process over time.

Conventional modeling techniques such as simulation and mathematical optimization fails to capture all principles and modeling components of forest ecosystem management design. The inevitable FEM modeling elements such as multiple, conflicting, non-linear spatial objectives such as adjacency delay, maintenance of a desired patch size distribution can not efficiently be presented in both techniques. While simulation approaches are easily confounded where structural forest objectives exist, pure optimization approaches are easily swamped with spatial resolution (Baskent et. al 2000, Murray 1999).

FEM requires a spatial forest modeling approach. Here, management interventions and their timings are identified with absolute geographic detail at the smallest forest management units, i.e. stands, so that spatio-temporal characteristics of forest landscape (size, shape, distribution, proximity, dispersion of forest patches) can be predicted and measured with respect to objectives. By all means, the scale, spatial gradient and extend, is part of spatial modeling. The fundamental emphasis lies on the development of management plan that is practical and easily applicable on the ground. Spatial modeling present information about spatial forest dynamics which lead to the understanding of ecosystem functioning. Spatial forest modeling, however, causes uncertainties when spatially allocating silviculture and assigning stand transitions following intervention. For example, in aspatial modeling, we could transition a spruce-balsam fir stratum following a clear cut to different outcomes using percentages, say 50% poor regeneration response and 50% perfect regeneration response. Such proportions, however, cannot be assigned reliably at stand level, unless we develop stand-level transition models. That is not easily done.

As stands constitute basic units in spatial forest modeling, with each having potentially multiple treatment regimes, FEM design is a combinatorial problem (Murray 1999, Nurullah et. al 2000). It means that the number of decision choices exponentially or astronomically increases when the size and management options linearly increase. As such, common modeling techniques such as linear/goal/integer programming can not overcome the problem. A particular class of algorithms, commonly labeled meta-heuristics, such as simulated annealing and taboo search, have been able to provide "good enough" solutions in reasonable computational time, however (Lockwood and Moore 1993, Boston and Bettinger 1999, Baskent and Jordan 2002, Nurullah et. al 2000, Öhman and Eriksson 2002). These heuristics have the ability to formulate a problem using discretionary rules that would be difficult to formulate mathematically. In meta-heuristic parlance, for example, an FEM design problem would be represented as an objective function which typically involves several components; each expressed as a summation of numerical penalty function values. The objective function thereby accommodates different objectives measured in different units, e.g. timber in cubic meters and patch size distribution in hectares.

The meta-heuristic solution technique provides immense opportunity to solve FEM problems, since it is relatively easy to tailor and customize. However, the technique requires careful programming such as object oriented (OO) software engineering environment. Model built using OO techniques includes modularity (object identification), abstraction (object interaction), and encapsulation (object representation). These features make OO-based forest ecosystem modeling components (a bundle of related objects such as a tree, age cohorts, stands, harvest blocks, or habitat patches) less interdependent on each other and easily maintained, extended and updated. With ever increasing and varying forest management model components, complex relationships, detailed spatial information and better understanding of landscape dynamics, the OO analysis and design approach becomes an essential model building environment for FEM.

Pitfalls in Implementing FEM

While FEM is a decade old concept its implementation has just been realized. There exist few specific applications and thus it is hard to provide a success story involving the comprehensive implementation of FEM. However, some apparent applications and potential pitfalls may be listed. For instance, Quigley et. al (2001) documented a successful FEM implementation using ecological integrity as a criterion of FEM in The Interior Columbia River.

• Providing public consensus in designing and implementing FEM interventions becomes dull when the awareness and stewardship do not exist. Providing a common currency among various stakeholders through a forum is a must.
• Peer-pressured ad hoc management planning without a holistic restructuring of existing planning process keeps the stakeholders busy and thus retards FEM implementation.
• Applying a light-hand silvicultural interventions such as reserve clear cutting do not necessarily mean FEM -it is just one aspect. However, the focus must be to create a target landscape structure where ecosystem health and integrity are maintained.
• FEM does not mean to overlook timber production - it means rational utilization subject to ecological sustainability. Objectives will overlap necessitating conflict resolution.
• Quantitative numbers for non-quantifiable such as biodiversity are generally superficies and in fact proxy measures thus should be interpreted cautiously.
• Short or one rotation may cause a spurious result in FEM model and can not warrant ecosystem sustainability.
• Use of empirical yield tables in modeling forest dynamics will not effectively forecast water production and biodiversity objectives for example as they are process oriented..
• Roads may result in habitat fragmentation, surface water flow, degraded stands attractive for insect infestation and could facilitate dispersal of species which would be indicator or keystone species that have the potential to adversely affect sensitive component of the landscape.
• Comprehensive spatial data base design and implementation is essential. The data base is to be prepared with a multidisciplinary team ensuring that data origin, content, definition, data base design, storage and exchange format -metadata- are all in place.

Conclusions

Ecosystem management is now a reality focusing on spatially re-structuring forest landscape to produce both commodities and ecological values. The ultimate challenge in forest management endeavor, however, is to design creative and innovative management designs so that acceptable balance of managing forest values can be realized in perpetuity. Specifically, composition and spatial configuration of forest values must be quantified before attempting to integrate them within a management model. It is necessary to use both meta-heuristic and mathematical programming techniques with object-oriented modeling design to develop an ecosystem based management model. Well rehearsed management design, smarter use of modeling techniques and information technology, understanding forest ecosystems dynamics, and effective public involvement could make the realization of ecosystem management on the ground happen. Professionals have a major role to usher in its success by being open and proactive to the emerging issues of forest ecosystem science. They should take stewardship in designing management interventions that are functionally linked to forest conditions and values they generate rather than just implementing regulatory statutes.

For successful FEM:

• As everyone plays a role in ecosystem management, a common understanding and learning capacities must arise
• Spatial management objectives and strategies must be formulated in relation to forest composition and configuration.
• Cross boundary thinking across ecological continuities should be realized
• Managers must have better decision making tools based on process based forecasting and comprehensive spatial database
• Ecologic integrity and forest values must be quantified and managed on the basis of ecosystem sustainability.
• Management actions must be planned and assessed across a wide range of both spatial and temporal scales
• Alternative future management options should be evaluated for better understanding of spatial forest dynamics with performance indicators of forest response

Undoubtedly, forest management philosophy will evolve with changing societal aspiration as well as with our enhanced understanding and wisdom of forest ecosystems. With increasing awareness of scientific and technological development, the opportunity to design better management process as a means of achieving best stewardship will likely emerge.

Literature Cited

Attiwill, P.M. 1994. The disturbance of forest ecosystems: the ecological basis for conservation management. Forest Ecology and Management, 63:247-300.

Baskent, E.Z. and J.A. Jordan.1995. Designing forest management to control spatial structure of landscapes. Landscape and Urban Planning, 34:55-74.

Baskent, E. Z. ve H.A.Yolasığmaz. 1999. Forest Landscape Management Revisited, Environmental Management, Vol.24, No. 4, pp. 437-448.

Baskent, E.Z. and J.A. Jordan. 2002. Forest Landscape Management with Simulated Annealing, Forest Ecology and Management, 165 (2002), 29-45.

Baskent, E.Z., J.A. Jordan and Nurullah, A.M.M. 2002. Designing Forest Management, The Forestry Chronicle, 76(5): 739-742.

Bergeron, Y., B. Harvey, A. Leduc and S. Gauhier. 1999. Forest management guidelines based on natural disturbance dynamics. The For. Chron. 75(1): 49-54.

Binkley, C. 1997. Preserving nature through intensive plantation management. The Forestry Chronicle, 73(5):553-559.

Boston, K. and P. Bettinger. 1999. An analysis of MCIP, Simulated Annealing, and Tabu Search heuristics for solving spatial harvest scheduling problems. Forest Science, 45(2): 292-301.

Brooks, D.J., and Grant, G.E., 1992. New perspectives in forest management: background, science issues, and research agenda, USDA FS. PNW-RP-456. 17p.

Forbes, G. 1998. Forest biodiversity guidelines. UNB Forestry Focus, Volume 23, Number1.

Franklin, J.F. 1998. The natural, the clear-cut, and the future. Northwest Science, (72)2:134-138.

Grumbine, R. E., 1994. What is Ecosystem Management?, Conservation Biology, March, Volume 8, No. 1, pp. 27-38

Haney, A. and Power, R.L. 1996. Adaptive Management for Sound Ecosystem Management, Environmental Management, Nov-Dec 31, 20(6), 879-886.

Lockwood, C.G. and T.G.E. Moore. 1993. Harvest scheduling with spatial constraints: simulated annealing approach. Canadian Journal of Forest Research, 23: 468-478

Murray, A.T. 1999. Spatial restrictions in harvest scheduling. For. Sci., 45(1): 45-52.

Nurullah, A.M.M., G.A. Jordan and E.Z. Baskent. 2000. Spatial stratification. The Forestry Chronicle,76(2): 311-317.

Öhman, K. and L.O. Eriksson 2002, Allowing for Spatial Consideration in Long-Term Forest Planning by Linking LP with Simulated Annealing, Forest Eco. and Mgmt. 161 (2002), 221-230.

Salwasser. H., 1994. Ecosystem Management, Journal of Forestry, August, pp. 6-10.

Seymour, R.R. and M.L. Hunter. 1992. New forestry in Eastern Spruce-Fir forests. April 1992, Maine Agricultural Experiment Station, University of Main, 716pp.

Yaffee, S.L. 1998. Three faces of ecosystem management. Cons. Biol.13(4):713-725.

¹ Professor, Faculty of Forestry, Karadeniz Technical University, Orman Fakültesi, 61080 Trabzon, Turkey. [email protected]