Dynamic classification: conceptual development and applications in wildlife management

Abstract

Information is a prerequisite for effective management of wildlife habitat. In response to the need for management-oriented information regarding the suitability of an area as wildlife habitat, a new methodology was developed and demonstrated. This methodology involved the conceptual development of a dynamic classification approach.

The proposed methodology sought to avoid many of the problems inherent in conventional. classification due to the inflexibility of the latter technique. Using dynamic classification methodology, an entit;y is described with respect to 1 or m.ore attribute axes relevant to the objectives of the specific cl.assificat:ion effort. Attribute axes may represent naturally occurring (i.e., physical and theoretically empirical) attributes or synthetic attributes such as suitability for some purpose.

To demonstrate conceptual utility, an original computerized cellular mapping system was developed to display information graphically. Maps of habitat suitability and other habitat-related information were produced for a total of 5 wildlife species on 3 study areas in Virginia. Within this demonstration, levels of habitat attributes subject to change over time were estimated from forest stand. data predicted with a modified Markov chain algorithm. Performance of the prediction program was determined from a 40 year hindcast procedure. Conceptual validity of a dynamic classification approach was examined using epistemological arguments.

The computer mapping package was found to be an effective vehicle for displaying information derived from dynamic classification. The vegetation prediction system appeared to be a feasible technique for predicting certain wildlife habitat attributes which are dynamic over time. It was concluded that dynamic classification was conceptually valid and is an effective methodology for producing information specific to the objectives of a given classification effort. Although conclusions: were based on an application of dynamic classification in a wildlife management context, it was speculated that the overall concept of dynamic classification may have additional utility in other fields within the natural sciences.