The forests in Finland have been under intensive planning for decades. Currently, mathematical programming is widely used in planning of wood production. Today's multi-functional forestry, however, calls for more flexible decision support methods. MCDM tools have been used in responding to fresh planning challenges. For example, the Finnish Forest and Park Service, entrusted with the care of the vast majority of state-owned natural resources in Finland, endeavours to produce large-scale natural resource plans satisfying the needs of both economic, social, and ecological sustainability. Participatory approach is applied in the process.
Several forestry applications of MCDM methods, particularly those making use of the AHP or the HIPRE program, have been presented. Also, the outranking methods ELECTRE and PROMETHEE have been tested. Due to the nature of forestry applications, statistical techniques for analysing uncertainties in pairwise comparisons and for utilizing interval judgement data have been developed to improve the usability of the AHP. Recently, a hybrid method called A'WOT, making use of the AHP and SWOT, was also introduced into strategic forest planning.
This paper summarizes the experiences gained in applying a MAVT and two outranking methods in connection with a participatory natural resource planning process in Finland. In addition, some results of the method development work related to application needs are briefly presented. The details of the planning cases reviewed here have previously been presented in forestry journals. The purpose of this paper is not only to show how MCDM methods have been applied in forestry, but also to discuss the usability and usefulness of MCDM methods from the viewpoint of supporting forestry decision making—and how they might further be improved. Also, some perspectives for the future development work of MCDM applications in the field of natural resource management are focused on. As a conclusion, the use of more than just one MCDM method in a single planning process is seen usually recommendable. In addition, developing hybrid MCDM methods is regarded as a potential direction for future research. Also, closer co-operation between method developers and appliers is called for to produce more useful applications. Copyright © 2002 John Wiley & Sons, Ltd.