BIOC@PT - Automatic Forest Biodiversity Sensors

Objectives

It meets a triple aim of biovigilance, biodiversity monitoring and the study of spatiotemporal patterns of biodiversity. It involves automating field surveys (using sensors) and species determination (using artificial intelligence or metabarcoding techniques). Our exploratory approach concerns two important and relatively poorly known forest taxonomic groups (insects and bryophytes) and could be transferred to other groups. During the project, we are developing prototypes of insect sensors with automatic photographic traps, and testing their operation in forests. Automatic insect traps are selective (xylophagous Agrilus, flower-visiting insects) or generic (flying insects). During the project, we are also implementing a database of reference pictures (insects of the genus Agrilus, forest bryophyte spores), and we are developing neural networks designed to determine the species sampled by entomological sensors and airborne spore sensors (cyclone sampler). A metabarcoding feasibility test will be carried out on bryophyte spore samples, and the results compared with the visual recognition method using deep learning.

Participants

INRAE units involved

• UR EFNO - Écosystèmes forestiers
• UMR SADAPT - Sciences pour l'action et le développement : activités, produits, territoires 
• UR P2E - Laboratoire de Physiologie, Ecologie et Environnement - Université d'Orléans
• USC Ecodiv-Rouen - Étude et compréhension de la biodiversité 
• UMR Herbivore - VetAgro Sup

Partners

• IRBI - Institut de Recherche sur la Biologie de l'Insecte - Université de Tours / CNRS