Ray Wills - GIS Research at CALM


Between 1992 and 1995, my program of research with the Department of Conservation and Land Management (CALM) in Western Australia was aimed at establishing a GIS database bringing together relevant spatial biophysical data and remote-sensed imagery to provide information on the spread of Phytophthora and its proximity to populations of susceptible threatened taxa, and includes a study to model the dynamic effects of infection.

The research team included Andrew Conacher, Alex Chapman and Graeme Behn, as well as other collaborators from within CALM, the Geography Department at The University of Western Australia, and CSIRO's Division of Mathematics and Statistics.

The project targeted the Two People's Bay-Mt Manypeaks area in CALM's Albany District. This area is significant because of a long history of Phytophthora impact and a high conservation value due to the large representation of susceptible plant species, the presence of 22 plant taxa on CALM's Priority list, including 7 Declared Rare Flora (DRF) and four of which are highly susceptible to Phytophthora cinnamomi (Adenanthos cunninghammi, Andersonia sp (TPB) G Keighery 8229, Banksia brownii, and Banksia verticillata), vegetation associations of special conservation interest, as well as the presence of a number of endangered animal species including the Noisy Scrub-bird, the Western Bristlebird, the Western Whipbird, and the recently rediscovered Gilbert's Potoroo.

A view of the Two Peoples Bay area, east of Albany, Western Australia using ArcView software.

The research aimed to provide a reliable, graphical decision-support system of use in monitoring and controlling the spread of dieback disease, facilitating conservation effort across landscape units of different tenure, and aiding in the identification of areas in need of rehabilitation or restoration. Development was being carried out on ARC/INFO GIS and the presentation tool will use both ARC/INFO and ArcView 2.


The dieback model provided predictions of the spread of Phytophthora cinnamomi based on hydrogeomorphic characteristics, and the susceptibility of vegetation and soils. The model can be conceptualised as a broad scale model which utilises available data to provide an indication of possible dieback threats. The model relied on LANDSAT imagery to identify sources of infection, from which the algorithm predicted the direction and rate of spread. Hydrological modelling techniques were facilitated through the raster-based GRID module of Arc/Info which provided the ability to combine numerous 'surfaces' within the same spatial context. The factors affecting dieback at higher resolutions require detailed knowledge of micro-hydrogeological processes and cannot be addressed using existing data; investigations to improve the resolution of available data were undertaken. Future developments had aimed to incorporate linear dispersal and Gaussian dispersal algorithms to assist in the prediction of dieback spread along roads and tracks.

Model development was partly based on my earlier research.

This project was funded by the

Australian Nature Conservation Agency


Created November 9, 1995

Last updated: May 6, 1997.

This page has been visited times since 6 January 2006