Diversification of crop and non-crop vegetation in agricultural landscapes: a benefit for pollinator communities

National audience; Vegetation diversification is considered a promising management strategy to promote biodiversity, improve the functioning of agro-ecosystems and increase crop yields in various farming systems. Vegetation diversification can be set up at different spatial scales – field, field border, landscape - and targets both crop and non-crop vegetation. Management strategies based upon the diversification of crop vegetation rely on the combination of different varieties or crop species and on their organization in space and time. Conversely, strategies based upon the diversification of non-crop vegetation rely on the maintenance of semi-natural elements, such grassy field margins, hedgerows, or the introduction of new landscape elements such as flower strips. Previous works have investigated the effects non-crop vegetation diversification. They have shown that non-crop elements fulfill resources (nectar, pollen) and shelters for pollinators, and that increasing the diversity of non-crop vegetation at field borders increases the diversity and abundance of pollinators in crops. At the landscape scale, the diversity, extent and spatial arrangement of crops and non-crop elements, i.e. landscape heterogeneity, play an important role for pollinators and pollination. However, the relative influence of crop and non-crop vegetation diversity in the field and its adjacent border on pollinator communities and pollination in different landscape contexts still remain unknown. It is thus unclear whether crop and non-crop vegetation diversity can have synergistic or antagonistic effects on pollinator communities and pollination. Here, we aim at i) disentangle the effects of crop and non-crop vegetation diversity on pollinators in crop fields and field borders, and ii) evaluate whether these relationships are mitigated by the diversification of crop and non-crop vegetation at the landscape scale. Pollinator surveys were carried out in 10 pairs of monoculture and mixed-crop fields along a gradient of landscape heterogeneity, in the Zone Atelier Armorique, France.