Analysing subterranean fauna distribution in Western Australia’s Hamersley Province: Applications in hydrogeology, geology, and biodiversity

Subterranean fauna are specialised and poorly understood, especially in arid regions like Western Australia. This thesis uses Random Forest classification to model the distribution of troglofauna and stygofauna in the Pilbara. It assesses the presence probability areas and environmental drivers for troglofauna (Chapter 2), evaluates the effects of taxonomic resolution on stygofauna models(Chapter 3), and compares habitat preferences between the Eriopisidae and Paramelitidae families (Chapter 4). Chapters 5 and 6 discuss and conclude the thesis, respectively. Despite challenges such as class imbalance, taxonomic resolution, and sampling bias, robust validation ensured the reliability of the outputs.