Quantifying floral resources and habitat quality of pollinators in agricultural landscapes using field data and UAV tele-detection

Plant-pollinator interactions are crucial for the ecosystem, but the robustness of their interactions highly depends on the quantity and the quality of floral resources available. In agricultural landscapes, where diversity and quality of floral resources are often in degraded conditions for pollinator populations, the need for quantitative assessment of habitat quality is significant. Most of the studies often only focus on quantity but not on quality, and often at small scales because of the difficulty of collecting enough data on the field. To help understand the distribution of floral resources in agricultural habitats, we evaluated pollen spatial heterogeneity in Belgian agricultural crops at the landscape scale. We assessed floral nutritional quality by measuring chemical components (carbohydrates, lipids and proteins) of the pollen of various entomophilous plant species. We then combined multi-temporal field flower data and UAV flight tele-detection to train a deep learning network capable of counting the precise number of flowers in each type of habitat. This method significantly simplified the quantification of floral resources in ecosystems, enhancing both the efficiency and accuracy of measurements. Using these results and the phenology of each species, we can directly quantify the nutritional potential each landscape offers to pollinators through time, a critical factor in designing effective conservation strategies.

Affiliations

Citations

APA
Chicago
FWB

Buron, M., Radoux, J., Jeannerod, L., Agnan, Y., Defourny, P., & Jacquemart, A.-L. (2024). Quantifying floral resources and habitat quality of pollinators in agricultural landscapes using field data and UAV tele-detection. Eurobee10, Tallinn, Estonia. https://hdl.handle.net/2078.5/240520