Species distribution modelling to identify and validate areas of occupancy for the magnificent broodfrog (Pseudophryne covacevichae)

Context The wet sclerophyll forests of the Australian Wet Tropics face increasing pressure from emerging threats like climate change and industrial development for wind farms. The magnificent broodfrog (Pseudophryne covacevichae) is a threatened species of the wet sclerophyll forest. A limitation for conservation is poor knowledge of its ecology and distribution. Aims We aimed to better estimate the distribution of P. covacevichae using species distribution modelling (SDM) and evaluate model performance through targeted field surveys. Methods We developed an initial MaxEnt model to guide field surveys and locate new populations. These new records were then incorporated to produce an updated model. Model performance was evaluated using area under the curve and true skill statistic (TSS). Key results The SDM identified previously unrecognised areas of potential habitat beyond the species’ currently known distribution and identified many high-probability areas without records within the currently known distribution. The model resulted in the discovery of 12 new populations (all within the currently recognised distribution). Including the new records in the updated model led to improved model performance (TSS) and increased predicted probabilities in marginal or under-sampled areas. Modelling suggests that the patchy distribution of P. covacevichae is due to natural isolation between areas of suitable habitat, with suitability driven primarily by dry season precipitation and elevation. The findings highlight priority areas for future surveys, particularly in the wet sclerophyll forest on the Windsor Tableland. Further, increased sampling effort at elevations of 650–800 m may refine the species’ lower elevational limits. Conclusion This research provides insight into the potential distribution of P. covacevichae, identifies key areas that require surveys, and underscores the importance of field validation in SDM studies. Accurate distribution data are essential for conservation planning, particularly in mitigating threats from potential habitat loss due to climate change and wind farms. Implications Substantial areas of the species’ distribution occur on unprotected land and the SDM identifies key habitat that could be prioritised for future protection, particularly in the Atherton Tablelands region. The predictions also provide a foundation for genetic studies to assess population structure and gene flow among these isolated populations, which can aid in identifing vulnerable populations and inform conservation strategies to reduce extinction risk.