Du suivi des pollinisateurs et de leurs interactions au service de pollinisation en milieu agricole : influence des champs cultivés

Résumé fourni par la source

Animal pollination is essential for the production of 70% of the world's crops. This contribution to crop yields is often associated with the diversity and abundance of pollinators within crops. Pollinator diversity and abundance are often impacted by the landscape, which, depending on its composition, provides spatially and temporally varying amounts of floral resources. Despite the predominance of cultivated fields in agricultural systems, the impact of landscape factors related to cultivated fields remains relatively undocumented, unlike semi-natural habitats. In addition, the life cycle of pollinators suggests that landscape effects can be carried over from year to year. The overall objective of this thesis is to investigate the links between cultivated field effects, pollinator monitoring and crop pollination services. This work is based on the analysis of plant-pollinator interaction networks and the development of an acoustic pollinator monitoring method to study the pollination service and the underlying mechanisms.First, we quantified the contribution of different floral resources from the previous year's cultivated fields to the pollination services of oilseed rape. We show that oilseed rape depends mostly on other floral resources rather than on itself for the maintenance of pollinator species delivering the pollination services. The contribution of crop weeds is at least as important as that from mass-flowering crops. Most of this contribution comes from weeds within cereal fields, which are usually neglected in pollination services studies. In a second step, we investigated whether acoustic pollinator monitoring could be a method to study the pollination services of sunflower. We show that deep learning appears to be the most effective machine learning method for recognizing the sound emitted by pollinators and that the acoustic pollinator monitoring method provides information consistent with the literature on pollinator biology. We also demonstrate that this method of monitoring pollinators, because of its temporal continuity, allows to get rid of some biases inherent to classical methods and thus allows to highlight subtle landscape effects.In conclusion, our work highlights the importance of cultivated fields in pollination services studies. We demonstrate that changes in cropping patterns and practices within cultivated fields, including fields where crops are not pollinator-dependent, can have a significant impact on the pollination services of crops. Although many technical and methodological developments remain to be developed, we show that acoustic monitoring is a promising method for monitoring pollinators in pollination services studies. Finally, we discuss the links between pollination proxies from pollinator monitoring and pollination services.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.