Unifying reserve design strategies with graph theory and constraint programming

Justeau-Allaire Dimitri, Birnbaum Philippe, Lorca Xavier. 2018. Unifying reserve design strategies with graph theory and constraint programming. In : Principles and practice of constraint programming: 24th International Conference, CP 2018, Lille, France, August 27-31, 2018, Proceedings. Hooker John (ed.). Cham : Springer, 507-523. (Lecture Notes in Computer Science, 11008) ISBN 978-3-319-98333-2 International Conference on Principles and Practice of Constraint Programming. 24, Lille, France, 27 Août 2018/31 Août 2018.

Résumé : The delineation of areas of high ecological or biodiversity value is a priority of any conservation program. However, the selection of optimal areas to be preserved necessarily results from a compromise between the complexity of ecological processes and managers' constraints. Current reserve design models usually focus on few criteria, which often leads to an oversimplification of the underlying conservation issues. This paper shows that Constraint Programming (CP) can be the basis of a more unified, flexible and extensible framework. First, the reserve design problem is formalized. Secondly, the problem is modeled from two different angles by using two graph-based models. Then CP is used to aggregate those models through a unique Constraint Satisfaction Problem. Our model is finally evaluated on a real use case addressing the problem of rainforest fragmentation in New Caledonia, a biodiversity hotspot. Results are promising and highlight challenging perspectives to overtake in future work.

Auteurs et affiliations

• Justeau-Allaire Dimitri, CIRAD-BIOS-UMR AMAP (FRA)
• Birnbaum Philippe, CIRAD-BIOS-UMR AMAP (NCL)
• Lorca Xavier, Ecole des mines d'Albi (FRA)

Source : Cirad-Agritrop (https://agritrop.cirad.fr/596408/)

Métriques

[ Page générée et mise en cache le 2024-10-23 ]