Developing habitat suitability models for the endangered plant species Kelussia odoratissima Mozaffarian using management data: Application of Bayesian belief network

Abstract

Conservation of endangered species is very important for maintaining native biodiversity. Management priorities, like climate change, is effective in the distribution of endangered species. To develop effective management and conservation strategies for the future, it is necessary to understand the current species potential distributions. However, for most species, few data are available on their current distributions, let alone on projected future distributions. Kelussia odoratissima Mozaffarian is an endemic perennial and medicinal forb species. The evaluation of K. odoratissima's conservation status places it within the Endangered (EN) category according to the IUCN classification. We demonstrated the benefits of Bayesian Belief Networks (BBNs) for predicting the distribution of endangered and medicinal K. odoratissima species using expert opinion. An influence diagram was developed to recognize the important factors influencing habitat suitability K. odoratissima, and it was populated with probabilities to produce a BBNs model. The behavior of the model was examined using sensitivity analysis. Environmental suitability, management condition, climate suitability, utilization time and levels were identified as the main variables influencing habitat suitability of K. odoratissima. The generated BBNs model had good accuracy because the ROC area under the curve was 0.918. We aim to demonstrate the ability of this approach to integrate field studies with expert knowledge, especially when empirical data are lacking. The BBNs model excels at illustrating species-habitat relationships and rapidly estimating habitat suitability, serving as a valuable tool for conservationists and decision-makers.