Just Accepted

This article has been peer reviewed and accepted for publication. It is in production and has not been edited, so may differ from the final published form.

Developing a regional species distribution model and validating with independent survey data: a case study of an avian apex predator, the greater sooty owl (Tyto tenebricosa)

Dylan Westaway , Yoko Shimizu, Chris Jolly, Scott Burnett

Abstract

Context. Comprehensive understanding of the distribution and habitat requirements of wildlife species is crucial for the development of effective conservation strategies. For cryptic predators, such as owls, obtaining accurate population metrics can be challenging. The rise of citizen science has created large amounts of data which are increasingly being used for conservation purposes. Aims. To create, and externally validate, a species distribution model (SDM) for the greater sooty owl (Tyto tenebricosa) throughout south-east Queensland (SEQ), Australia. Methods. A Maxent model was developed by combining citizen science records and environmental variables relevant to greater sooty owl ecology. The resulting model was then validated by incorporating Maxent-derived habitat suitability values as a variable in occupancy modelling performed on an independent dataset collected in the field across agricultural, suburban and remnant forest landscapes. Key results. The Maxent model showed good discriminatory ability (AUC = 0.95), with vegetation type (36.4%), elevation (26.8%) and annual precipitation (18.4%) contributing most to the model. Estimated detection probability (p = 0.19) and occupancy (ψ = 0.31) were relatively low across field surveys. Maxent-derived habitat suitability values had a significant positive relationship with occupancy and performed best in predicting greater sooty owl occupancy compared to other covariates. Conclusions. The species distribution model showed good discriminatory ability and was validated externally highlighting the potential value of citizen science data. The model suggests that rainforest and wet eucalypt open forest vegetation types, high rainfall and elevation provide optimal greater sooty owl habitat in SEQ. Implications. Our study represents a baseline which can be used to identify current greater sooty owl habitat, monitor habitat into the future and guide conservation actions and further research. We recommend further surveys into areas of identified high potential habitat and advocate for increased protection of important owl resources such as roosting and nesting sites.

WR25062  Accepted 02 September 2025

© CSIRO 2025