A crop yield-based pollination index reveals the impacts of land cover and pesticide use on realized pollination at a landscape scale

Abstract

The current pollinator loss raises numerous concerns, because animal pollination maintains the diversity of wild plant communities and is essential for human food. Yet, its consequences for pollination remain unclear, because there is no simple relationship between pollinator abundance or diversity and pollination efficiency. Large scale indicators based on direct measurements of pollination are therefore essential for an accurate assessment of the consequences of pollinator loss for pollination services. Here, we test whether the “realized pollination index” (RPI), a crop yield-based index, is informative on pollination services in a French agricultural landscape. We calculated the index at different spatial resolutions, down to individual fields, compared it with a “potential pollination index” (PPI) based on land cover/land use, and examined its relationship with pesticide use. At the landscape scale, the RPI was positively related to the land cover-based pollination index, and its temporal variations were strongly correlated with those observed at regional or national scales. At the field scale, the RPI was negatively related to pesticide use. These results suggest that the yield-based RPI captures variation in pollination services across landscapes accurately, with large-scale effects of landscape composition, including the spatial arrangement of semi-natural elements and cultivated fields, and more local effects of pesticide use. It could provide an unprecedented, easy-to-use tool for implementing automated pollination monitoring in farmland, from single fields to large areas, and help stakeholders to design biodiversity-friendly territories that efficiently deliver ecosystem services.