Using fuzzy sets to deal with uncertainty in single-species, single-season occupancy models.

Abstract

Occupancy modeling is a valuable tool for managing wildlife populations. Current occupancy models provide estimates of occurrence based on a point estimate for the species detectability and presence-absence. However, detectability can vary based on many variables ranging from weather to personnel. Therefore, I propose the use of fuzzy sets rather than point estimates for detectability and binomial presence-absence data during calculations of occupancy. Fuzzy occupancy estimates are easier to determine, more robust, and generally more informative than traditional point-based occupancy models. Consequently, managers will have better information available for comparing occupancy values among sites. Fuzzy sets are especially useful when parameters of the study violate key data standards.