Using sensitivity analysis to simplify ecosystem models: a case study

Publisher Website
Google Scholar
Abstract
Theoretical ecology requires simulation because real istic ecosystem models are too complex for mathemati cal analysis. In this paper we propose a way of moving from a complex simulation model based on the "mechanics" of an ecosystem to a much simpler model that exhibits behaviour similar to that of the origi nal model in its major features. The tool used to isolate the key characteristics of the original model is sensitivity analysis. It enables the investigator to identify the variables and parameters that deter mine the essential behaviour of the original model and to formulate a clear picture of that behaviour. With this information in hand, the investigator can devise a simplified model that exhibits the same essential behaviour. While the simplified model behaves like the original model, its inner workings only generally parallel those of the original. Its simplicity allows the investigator to test his intuitive ideas directly, clarifying understanding of the ecosystem originally modelled. Further, the investigator may be able to adapt the simplified model to other similar ecosystems. The methods outlined in this paper are illustrated by a case study of a forest infested by tent caterpillars.