Knowledge gaps regarding distribution, habitat associations, and population size for rare and threatened range-restricted taxa lead to uncertainty in directing conservation action. Quantifying range metrics and species–habitat associations using Species Distribution Models (SDMs) with remote-sensing habitat data can overcome these setbacks by establishing baseline estimates for biological parameters critical for conservation assessments. Area of Habitat (AOH) is a new range metric recently developed by the International Union for Conservation of Nature (IUCN) Red List. AOH seeks to quantify inferred habitat within a species’ range to inform extinction risk assessments. Here, we used SDMs correlating occurrences with remote-sensing covariates to calculate a first estimate of AOH for the Endangered Madagascar Serpent-eagle Eutriorchis astur, and then updated additional IUCN range metrics and the current global population estimate. From these baselines we then conducted a gap analysis assessing protected area coverage. Our continuous SDM had robust predictive performance (Continuous Boyce Index = 0.835) and when reclassified to a binary model estimated an AOH = 30,121 km2, 13% less than the current IUCN range map. We estimated a global population of 533 mature individuals derived from the Madagascar Serpent-eagle AOH metric, which was within current IUCN population estimates. The current protected area network covered 95% of AOH, with the binary model identifying three additional key habitat areas as new protected area designations to fully protect Madagascar Serpent-eagle habitat. Our results demonstrated that correlating presence-only occurrences with remote-sensing habitat covariates can fill knowledge gaps useful for informing conservation action. Applying this spatial information to conservation planning would ensure almost full protected area coverage for this endangered raptor. For tropical forest habitat specialists, we recommend that potential predictors derived from remote sensing, such as vegetation indices and biophysical measures, are considered as covariates, along with other variables including climate and topography.
