Modelling distribution data to predict subterranean fauna habit in the Pilbara

Starting in 2022, this PhD project is part of the ARC Training Centre in Data Analytics for Resources and Environments (DARE) at the University of Western Australia. It studies where subterranean fauna live in the Hamersley Province of Western Australia. The researcher used a machine learning method called Random Forest to build species distribution models for key groups of cave and groundwater animals. The models work at different levels of classification and in different parts of the Pilbara, using geology, soil, topography, climate and time-related data.

The project delivered three main findings. First, it compared habitat suitability for troglofauna (animals living in underground air-filled spaces) across two areas, showing how environmental drivers differ from place to place. Second, it modelled stygofauna (groundwater animals) at order level (Amphipoda) and family level (Paramelitidae) to test how classification detail affects prediction accuracy. Third, it built a model for the family Eriopisidae to compare its habitat preferences with those of Paramelitidae.