Ecologists have relied increasingly on computer simulation to model population dynamics. New models that focus on the behavior and interactions of individuals, rather than aggregate measures of population dynamics, have increased in use. With the increase in computationally intensive models, however, ecologists have been forced to find compromises between the accuracy of the new techniques and computational efficiency of the old techniques. Multiple-resolution modeling is one proposed method to solve this problem. A multiple-resolution model uses the accurate. but computationally intensive, models only when necessary; otherwise, it uses the less accurate, but computationally efficient, models.

This dissertation examines issues related to the development of a multiple-resolution modeling framework. Developing this framework required the formulation of a general, population modeling schema on which to build the multiple-resolution modeling framework and the formulation of a general structure for multiple-resolution modeling. Software engineering techniques were applied in the development of the framework.

A set of multiple resolution models were developed and tested. The results of these tests do not verify the generality of multiple-resolution modeling; rather, they suggest that the framework can produce correct multiple-resolution models.