The Role of Modeling in Research on Microfloral and Faunal Interactions in Natural and Agroecosystems

A variety of mathematical tools can be applied to constructing models in biology. The distinction between correlative and explanatory models (Gold 1977) is more important for our purposes than the particular mathematical tool employed. Explanatory models are structured to be analogous to the real system under study and thus to embody hypotheses about how the real system operates. The structure of correlative models--regression models, for example—is not constrained by what is known about mechanisms operating in the real system. Both correlative and explanatory models are expected to yield realistic predictions. Although both types of models are needed in basic research, we restrict our attention largely to explanatory models because they are especially useful as an aid to understanding.