INCORPORATING RULE-BASED REASONING IN THE SPATIAL MODELING OF SUCCESSION IN A SAVANNA LANDSCAPE

A Spatial modeling technique incorporating rule-based reasoning with conventional models is presented. The approach utilizes geographic information systems (GIS) and rule-based expert systems to spatially model secondary succession on a savanna woodland site, in southern Texas, that had mesquite clusters distributed on a grassy matrix. Cluster distribution found in an aerial photograph was used to recreate the initial landscape for the study site. Relative growth rates for clusters in various size classes were determined from data obtained from later photographs. A rule base containing factors believed to have regulated landscape development was linked to the growth model. Landscape patterns were then generated, analyzed, and displayed through manipulation of a raster-based GIS software. The simulation result was compared to a recent photograph to evaluate model performance. The simulation result verified the roles of annual rainfall and soil characteristics. It also suggested that, with proper control, the landscape could be maintained at its current state. On the other hand, if cluster development continues at its present rate, succession from open savanna to closed-canopy woodland would occur.