A prototype graphic landscape analysis system: Part 1. Predicting spatial patterns of grazing pressure using GIS
Abstract. A prototype spatial graphic landscape system for use in PC-based micro-computers was developed. Spatial characteristics and attributes of a landscape were entered into a geographical information system to create a data base. A grazing pressure index was modelled using an algorithm which sequentially adjusts grazing capacity of response units for percentage lost to brush density, slope, and distance from water. The grazing pressure model was verified on a poorly watered homogeneous pasture (1302ha) and well watered complex pasture (2959ha). Response units were highly correlated to observed grazing ( r 0.92-0.94) in the poorly watered, homogeneous pasture. However, correlations were poor in the complex, well-watered pasture. Circular barrier effects of slopes, intermittent watering sources and relative differences in ecological condition of adjacent response units were identified as the primary factors reducing model performance.