Linking pollinator abundance in field margins to crop pollination service

Crop pollination services are an often-used argument for supporting pollinators in agricultural landscapes. However, the link between pollinator occurrence in the proximity of fields and realized crop pollination service remains poorly studied. We examined how the abundance of different pollinator taxa in field margins predicts pollination services in adjacent insect-pollinated crop in boreal agricultural landscapes. Additionally, we studied how pollinators in field margins are affected by pesticide use and landscape structure. We conducted a field experiment in 34 spring-sown turnip rape (Brassica rapa ssp. oleifera) fields and their permanent margins in Southern Finland in summer 2017. The fields situated in two landscape types, differing in the cover of arable fields within 500 m distance. We counted pollinators in the field margins from early June until end-July along 50-m-long transects, and monitored pollinator visits on turnip rape flowers three times during flowering. Through a farmer questionnaire, we collected data on farming activities in the fields. Total pollinator abundance and species richness in field margins were poor indicators of pollination service in adjacent turnip rape fields. However, pollinator groups differed in this respect: High abundance of bumblebees and syrphid flies in field margins indicated high crop visitation rate by these taxa. Syrphid flies preferred to visit crop near field edges, while bumblebees used more field interior. Butterflies were abundant pollinators in field margins but rarely visited the adjacent crop. Honeybees, by contrast, were dominant pollinators in the crop regardless of their abundance in field margins. In total, one third of pollinator species or groups present in field margins were observed to visit turnip rape. Pesticide use in turnip rape fields reduced pollinator abundance in field margins. The negative effect tended to be more pronounced in landscapes dominated by arable fields than in landscapes with low field cover. Pesticide use affected more bees and butterflies than syrphid flies. Syrphid flies preferred landscapes with low field cover, whereas bees benefitted from high cover of perennial grasslands in the surrounding landscape. Our study showed that the value of field margins for crop pollination cannot be reliably assessed based on total pollinator abundance, but the assessment should focus on the few pollinator groups and species with demonstrated importance as crop pollinators. These results highlight the need for greater arguments than crop pollination for pollinator conservation in agricultural landscapes. While the effects of pollinator abundance and diversity on crop yield may be weak, the impacts of agricultural activities on pollinator conservation are obvious. Reduced pesticide use and increased land use heterogeneity would benefit pollinators in boreal agricultural landscapes.peerReviewe