Probabilistic graphical models for species richness prediction: Are current protected areas effective to face climate emergency?

Abstract

Climate change has been related to the current loss of global biodiversity. In this paper, the effects of different scenarios of climate change on the distribution of the four classes of terrestrial vertebrate species in Andalusia (Spain) are explored. The goal is to obtain potential climatically suitable areas for each group (amphibians, reptiles, mammals and birds) under each proposed scenario and examine the usefulness of the current static design of protected areas. We propose a methodology to construct habitat suitability models, which are used to predict the expected species richness given each projected scenario of climate change. The relative change of the species richness within National and Natural Parks, remainder of Natura (2000) network and unprotected areas is compared. The results of the study show a broad effect of climate change on the species richness distribution. In general, there is a loss of specific richness and a restricted availability of suitable areas. The protected areas located in higher altitudes maintain the best conditions for the survival of the taxa considered in the proposed climate change scenarios.

Keywords

• Bayesian networks
• Terrestrial vertebrates
• Effectiveness of protected areas
• Conservation policy