Modelling the distribution of wetland birds on the Red Sea coast in the Kingdom of Saudi Arabia

Species distribution modelling is useful for addressing knowledge gaps for poorly studied geographical areas by identifying potentially suitable habitats for species across landscapes. This study was undertaken to identify areas containing wetland habitats to support actions aimed at the conservation of wetland bird species on the Red Sea coast of the Kingdom of Saudi Arabia (KSA). We used a maximum entropy approach to build habitat suitability models for 22 wetland bird species (gulls, shorebirds, and terns). Ten variables relating to topography, habitat, latitude, coast complexity, man-made structures, and human settlements were used to produce individual habitat suitability models for each of the bird species. The areas under the curve (AUC) for the final model were 0.9013 and 0.879 for the training and testing data, respectively, and the jackknife analyses suggested that the models generally performed well. Using the modelled distributions of the 22 species along the Red Sea coast, the analysis suggested 17 core areas where the habitat and landscape configuration were suitable for supporting high species richness. Of these 17 sites, one is already protected, and there is a need to protect the remaining sites. The use of these models to inform conservation strategies in the Red Sea region of the KSA is discussed.