The impact of sown flower strips on plant reproductive success in Southern Sweden varies with landscape context

Abstract

In agricultural landscapes, sown flower strips can benefit pollinators and pollination of nearby plants, but their impact on pollination in the wider landscape is poorly studied. We evaluated effects on reproductive success of field bean (Vicia faba) and woodland strawberry (Fragaria vesca) using data from two study systems, both including study sites (1 km radius) with (flower strip sites) or without flower strips (control sites). To assess whether flower strips enhance pollination in the wider landscape, we compared the reproductive success between plants growing in field borders (> 160 m to nearest flower strip) at flower strips sites and control sites. We also tested if flower strips reallocate pollination functions in the landscape. We did this by comparing the reproductive success of plants at flower strip sites, growing adjacent to the flower strips with plants growing in a more distant field border at the same site (> 160 m). Finally, we tested if these potential effects depended on the heterogeneity of the landscape. In field borders without an adjacent flower strip, plant reproductive success was unaffected by the presence of a flower strip at the site, and increased with increasing landscape heterogeneity independently of site type (flower strip vs. control). In contrast, adjacent to the flower strips, reproductive success declined with increasing landscape heterogeneity, resulting in a positive net effect of adjacent flower strips in homogeneous landscapes and a negative effect in heterogeneous landscapes. Our results show that while decreasing landscape heterogeneity may impair pollination in homogeneous landscapes, this can be locally mitigated by sowing flower strips. However, in heterogeneous landscapes, flower strips may instead reduce pollination of adjacent plants.