Exploring the Sensitivity of Fuzzy Decision Models to Landscape Information Inputs in a Spatially Explicit Individual-Based Ecological Model

Introduction

This is part of an ongoing exploration of incorporating fuzzy logic into spatially explicit, individual-based ecological models of dispersal. Following the theoretical discussion of Robinson (2002), a prototypical model of small mammal dispersal behavior was used to demonstrate how the fuzzy control of dispersal agents could be implemented (Robinson and Graniero 2005a). The implementation showed how the Extensible Component Objects for Constructing Observable Simulation Models (ECO-COSM) system could be loosely coupled with geographic information system (GIS) database for spatially explicit ecological simulation modeling of individual behavior (Graniero and Robinson 2006). If the problem is viewed from a geocomputational management perspective, we can say that an animal agent must be able to query the state of relevant GIS layers within its local perceptual range and use that information to make decisions regarding its movement behavior. Its movement behavior inturn leads eventually to a change in the state of the agent. Within the ECO-COSM framework, this is handled by the Probe mechanism. By obtaining Probes from relevant Probeable landscape layers , an agent can acquire a perceptual inventory of its world (Graniero and Robinson 2006). Thus, the general approach is consistent with Bian’s (2003) hybrid approach to representing the world in individual-based modeling, which incorporates a traditional grid model of the environment and an object-oriented model of individual organisms.