Applications of modeling to vegetation management

Abstract

A review of modeling applied to vegetation management shows that these models range in resolution from simple yield equations to complex representations of processes affecting growth and competition for light, water, and nutrients. The latter models inform scientists and managers about mechanisms involved, while the former are more likely to be applied by managers to estimate effects of competition. Six generic model forms were identified. Models may be further categorized by whether they focus on weed population dynamics or on processes directly affecting growth of crop plants. There is scope for these aspects of vegetation modeling to be combined. Extremely complex models are scientifically satisfying as repositories of knowledge, but they tend to be excessively parametarized and recursive. Models with many parameters are difficult to fit to specific situations because they are ambiguous; the same overall estimates of growth can be achieved in a variety of ways. In addition, their recursive nature leads to compounded errors. Simple growth and yield models, by contrast, are usually very efficient and accurate at estimating local estimates of growth, but they are inadequately sensitive to the variety of site and site management practices that vegetation managers wish to represent. A new kind of hybrid model is proposed that combines the efficiency of forest mensurational techniques with sufficient complexity to represent the results of scientific studies associated with vegetation management for use by managers.