Multi-criteria natural resource management with preferentially dependent decision criteria.

The traditional assumption with multi-criteria natural resource management models is that the values of the decision alternatives with respect to one criterion can be assessed independently of the values of the decision alternatives with respect to other criteria. In practice, however, the assumption of independent decision criteria is not always realistic. For example, the importance of the amount of old forests to biodiversity may depend on the amount of dead wood. This paper shows how multi-criteria natural resource management problems can be analysed in a case involving dependent decision criteria. The proposed models are based on pair-wise comparisons of items in the ratio scale and statistical regression analysis of the pair-wise comparisons data. The basic idea behind the dependent models is to ask the decision maker (e.g. forest owner) to assess the decision alternatives or their characteristics simultaneously with respect to dependent decision criteria. The separate models for discrete and continuous cases are given, which enable the number of potential management alternatives to vary from a few to infinite. The models proposed allow a more realistic description of the decision problem in multi-criteria natural resource management.