A new method for predicting vegetation distributions using decision tree analysis in a geographic information system

Abstract

/ Decision tree analysis was used to predict the distribution of forest communities in an area on the south coast of New South Wales. Australia, The analysis was carried out using a geographical information system environmental data base of those topographic and geological variables thought to influence the distribution of vegetation and derived from cartographic sources. The resulting maps of forest communities are of a resolution sufficient to delimit individual forest stands and contain much ecological information.A major priority of environmental ntanagement in Australia is the developntent of cost-effective methods for mapping native forests. The aim is to produce maps of these forests that provide information not only about the location and extent of timber resources, but also ecological information to assist forest HI~:lltagers in developing and implententing muhiuse management policies. For forest managentent purposes, vegetation maps should be of sufficient resolution to delimit forest stands, which are contiguous regions occupied by a single community. Such maps would show the extent of forest communities within ntanagement compartments. Survey of forest communities at such fine resolutions is a labor-intensive and costly exercise because of the size and highly variable nature of many of the forests.Most vegetation surveys in these torests employ some combination of field work and phutointerpretation, sometimes in conjunction with satellite image analysis. These are not alternative ntapping ntethods. Manual survey can be assisted greatly by photointerpretation, and the production of reliable vegetation maps from aerial photographs or satellite data requires a significant investment of field work for interpretation and verification (Zonneveld 1988). htitially it appeared that multispectral satellite dam could be digitally analyzed to produce vegetation maps with little investment of labor. Such dam would appear to have great potential as a cheap, readily updaud)le

How a river course influences the species richness and ecological requirements on two opposite riverbanks in a forest area

“…This secondary topographic attribute is highly variable from place to place due to changing slope and aspect [17,18,44]. In the present study, the decisive condition shaping the energy amounts delivered for both riverbanks was the almost parallel course of the valley in a greater part of the analyzed fragment.…”

Spatial modelling of biodiversity at the community level

“…In this strategy, also known as 'predict first, classify later' and 'classification-then-modelling' Davis & Goetz (1990), Franklin & Wilson (1991), Lees & Ritman (1991), Moore et al (1991), Fitzgerald & Lees (1992), Brzeziecki et al (1993), Brown (1994), Brzeziecki et al (1995), Lewis (1998), Zimmermann & Kienast (1999), Keith & Bedward (1999), Hilbert & Ostendorf (2001), Ferrier et al (2002) 1b. Modelling of pre-derived species groups McKenzie et al (1989), Bojórquez-Tapia et al (1995), Ferrier et al (2002) 1c.…”

A methodological approach to potential vegetation modeling using GIS techniques and phytosociological expert-knowledge: application to mainland Portugal

“…A priori probabilities of a forest-type occurring on a topographic position were chosen by expert foresters, rather than induced from sampling O n the other hand, CAIRNS (2001) prefers an approach in which the vegetation model is strictly induced from a training set: general linear models (GLM), artificial neural networks (ANN) and induced classification trees are compared as predictors of vegetation type along the alpine timberline. MOORE et al (1991) also used machine-learning approaches (CART: classification and regression trees) to vegetation mapping in GIS. -FRANKLIN (2002) used classification-trees obtained from a trainine set of 906 plots, kheré pedictive variables were drawn from a DEM a l d from climate information.…”