Sensitivity of a Fuzzy-Constrained Cellular Automata Model of Forest Insect Infestation

Cellular automata (CA) are discrete systems used for modelling complex spatial dynamic phenomena. The discrete nature of CA enables integration with rasterbased geospatial datasets in geographic information systems (GIS), and also can be beneficial when modelling complex ecological processes that evolve over time. However, when modelling forest insect infestations it is difficult to use discrete cell states to represent, for example, the concept of susceptibility of a tree to insect attack. The use of binary or probabilistic approaches for cell states definition is not appropriate because insect disturbances are driven by numerous components of insect-tree relationships that are difficult to understand. Furthermore, uncertain transition zones exist between forest stands of different sizes and different species where a discrete definition of a cell cannot be provided. The objective of this study was to integrate fuzzy set theory with GISbased CA modelling to model tree mortality patterns caused by insect infestation, and to explore the sensitivity of the model to different spatial properties. This study focused on a case study of lodgepole pine, Pinus contorta, mortality patterns caused by infestations of mountain pine beetle (MPB), Dendroctonus ponderosae Hopkins. The use of fuzzy set theory addresses the issue of inherent uncertainty of the geospatial data used for studies of forest infestations, while a test of model sensitivity exp lores the influence of the spatial properties of a fuzzy-constrained CA.