Spatial terrain analysis for matching native tree species to sites: A methodology

This methodology has been developed to predict the distribution of preferable biophysical sites for three of the favored plantation species for private land rehabilitation in the Atherton Tablelands of tropical north Queensland: Araucaria cunninghamii, Eucalyptus cloeziana, and Flindersia brayleyana. The site prediction is expressed in a spatial and quantitative form of site classification. The sites (termed 'toposites') are characterized by soil and terrain qualities using spatial analytical techniques afforded by the predictive capacity of DTMs (digital terrain models) and conceptual regolith-terrain relationships in an explicit approach to forest site analysis through a geographic information system (GIS). Knowledge-based relationships and recorded soil-topographic data are used in this methodology to create conceptual models of climatic, terrain and parent material interaction. These guide the algorithms for digital terrain analysis through use of transfer functions and topographic DTM derivatives. Spatial regolith-terrain attribute predictions for potential toposites from DTMs are then combined with modeled climatic surfaces. The spatial arrangement of potentially favorable toposites for each of the selected species can be portrayed in raster format involving fuzzy classification. This form of classification provides multiple membership values for given regolith-terrain attributes for each grid cell. These can then be grouped as desired in an a posteriori classification to produce 'toposite preference' in raster or vector map form.[PUBLICATION ABSTRACT]