Fine-scale habitat modelling of wildlife species using spatial information tools

Abstract

Species habitat models (SHM) or species distribution models are numerical tools that combine observations of species occurrence or abundance with environmental estimates. Models that predict distributions of species by combining known occurrence records with digital layers from satellite imagery and Geographic Information System (GIS) inputs together with environmental variables have much potential for conservation applications and scenario modelling. In this research, four analytical techniques for multispecies modelling, i.e. maximum entropy (Maxent), random forest, artificial neural network, and mahalanobis typicality, will be explored to assess the species distribution. A methodological approach will be at the fine (5-10m) spatial scale, where studies at this level are lacking. Spotted-tailquoll, Darling Downs Earless Dragon, Black-Breasted-Button-Quail, and Bulloak Jewel Butterfly, which are endangered and threatened species had been selected as case studies to test this framework.