Alchemy and Uncertainty: What Good Are Models?

Bunnell, Fred L. 1989. Alchemy and uncertainty: what good are models? Gen. Tech. Rep. PNW-GTR-232. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. 27 p. Wildlife-habitat models are increasing in abundance, diversity, and use, but symptoms of failure are evident in their application, including misuse, disuse, failure to test, and litigation. Reasons for failure often relate to the different purposes managers and researchers have for using the models to predict and to aid understanding. This paper examines these two purposes and the nature of problems or failures in model application. A five-step approach toward solution is presented: (1) recognize the problem, (2) nurture modeling teams, (3) match purpose with test, (4) confront basic beliefs, and (5) evade known errors. Nurturing of modeling teams requires recognition of different risks, rewards, and timeframes of researchers and managers. Most wildlife-habitat models have mixed purposes (prediction and understanding). Each purpose should and can be evaluated separately. Some basic beliefs surrounding how we test statements are based more on faith than reason. These can be obstacles to using and testing models. Models of wildlife habitat are efforts to apply research to management; a better approach to using and testing models is a better approach to applying research on wildlife-habitat relations.