Exploring Geographic Parameter Space with a Gis Implementation of Cellular Automata

Many of the recent studies of vegetation spread and the invasion of introduced species incorporate GIS and some form of dynamic modelling in oneor twodimensional space. Most work, however, concentrates on presenting a predetermined model in an abstract computer simulation environment. The focus has been on the system’s ability to display the model (results), as opposed to its ability to develop or compose a model, acting as a creative tool in an applied situation. The research presented in this paper uses a GIS-based cellular automata (CA) framework to study and create an applied and ecologically significant model of spread for the invasive plant species Rhamnus alaternus. Instead of conventionally testing a previously created model, parameters such as generation time, fruiting phenology, fruigivore abundance, reproductive rate and vegetation cover are examined through simulation for their contribution to spread and the resulting spatial pattern. Initial simulation runs focus on the different effects of stochastic versus deterministic parameter values. Further studies examine the effect of different relationships, conditions and interactions between parameters. Upper and lower limits, the interactions between, and the relative importance of each parameter are examined in a number of combinations. Each parameter is then accepted or declined to create a conceptual model outlining the ecologically most significant features of the species. This method effectively examines and tests the ability for models of generic invasion and vegetation dynamics to be transposed into a contextually rich environment through a GIS-CA framework of model creation.