Identification of degraded forest soils by means of a fuzzy-logic based model

The term "forest soil degradation" is frequently used in forest ecology. It is a highly integrated site property summarizing negative effects over a wide, yet undefined range of variables and a complex range of processes. In forest ecology, different, undefined, and idiosyncratic meanings of the term 'soil degradation' are used. The evaluation for a particular soil is therefore not clear and may be inconsistent among experts. We integrated indicators of forest soil degradation into a fuzzy-logic based model and predicted forest soil degradation for a set of sites by means of standard soil chemical data and easily accessible site characteristics. For validation we used expert judgements on selected sites. We also compared if the predicted soil degradation agrees with the results of a recent assessment of the "naturalness" (hemeroby) of Austrian forests. The predicted results were consistent with our expectations: sites with a long history of nutrient exploitation were found to be degraded by the model. The fuzzy-logic based model is open. Rules can be changed, additional rules can be included and others can be removed, if desired. We want to promote fuzzy-logic based modeling as a means to support experts decisions in complex situations, where clarification can be added by crisply defining the pathway of the decision making process.