Do local landscape features affect wild pollinator abundance, diversity and community composition on Canterbury farms?

Pollination is an essential ecosystem service that can be affected by habitat features in the immediate environment, termed here ‘local landscape features’. This study tested how five local landscape features (bare ground, native biodiversity plantings, homestead gardens, shelterbelts, and control areas of pasture) affect local pollinator communities on Canterbury farms. We also compared two sampling methods (flower visitation to native potted plants vs sticky traps) to determine if the sampling method affects the results of landscape-feature comparisons. We recorded 928 pollinators of 17 taxa on the potted plants and 791 pollinators of 16 taxa on the sticky traps. There were significant differences in pollinator abundances between the landscape features and the control areas. Both sampling methods recorded fewer pollinators overall at shelterbelts and bare ground sites than control sites, although morning-evening fly was more common at bare ground sites. However, the methods gave contrasting results for biodiversity plantings and gardens: the flower method recorded significantly more pollinators in biodiversity plantings and gardens than it did at control sites, whereas the sticky trap method recorded significantly fewer. As the ower method showed higher pollinator abundances near biodiversity plantings and gardens, planting native insect-pollinated plants on farms could boost populations of wild pollinators, which may improve crop pollination.