Combining multiple tools to provide realistic potential distributions for the mouflon in Sardinia: species distribution models, spatial pattern analysis and circuit theory

Abstract

Species distribution models (SDMs) are commonly used to predict potential species distribution in ecology, biogeography, and conservation. However, the lack of consideration for many different aspects of the species’ ecology often result in maps of limited practical value. Yet the wealth of information that is often available (especially for larger vertebrates) could help improve these models. Here, considering the mouflon ( Ovis aries ) in Sardinia as a case study, we propose a new integrated approach based on different aspects of the species ecology. In particular, we combined an SDM calibrated with an ensemble modelling approach with a morphological spatial pattern analysis and a model of structural connectivity. With this approach it was possible to map at the same time the distribution of potentially suitable habitat patches (and particularly of those being able to host pregnant females) and of the structural corridors connecting these patches. Overall, while the SDM predicted that 14.5% (3480 km 2 ) of Sardinia was suitable for the presence of the mouflon, with our new integrated approach the percentage of areas suitable dropped to a much lower and more realistic 6.6% (1584 km 2 ). Our results have potentially important consequences for the practical conservation and management of the species in Sardinia.