Landscape-level honey bee hive density, instead of field-level hive density, enhances honey bee visitation in blueberry

Context To date, managing honey bees and wild bees within crop fields remains challenging. Landscape structure is often overlooked when studying the pollination contribution of honey bees. Increasing our understanding on how to predict honey bee visitation in crops is crucial for sustainable management of agroecosystems. Objectives With this study we investigated which landscape and field-level variables determine honey bee and wild bee visitation, and whether honey bee or wild bee visitation influence crop pollination. Methods Sixteen highbush blueberry fields were surveyed for honey bees, wild bees, and crop pollination in Washington, USA. Additionally, within a radius of 1000 m around each field all honey bee hives were located and the surrounding landscape was characterized. Results Honey bee hive numbers in the landscape positively correlate with the proportion of blueberry in the landscape. Honey bee visitation was best predicted by landscape-level hive density within a radius of 1000 m, whereas semi-natural habitat and field-level hive density did not impact honey bee visitation. The amount of semi-natural habitat and blueberry within a radius of 1000 m had a positive and negative impact, respectively, on wild bee visitation. Honey bee visitation had a positive effect on blueberry seed set. Conclusion We conclude that honey bee visitation is determined by the number of honey bee hives in the surrounding landscape. Hence, field-level hive density recommendations miss contributions from other hives in the landscape. Furthermore, semi-natural habitat did not impact honey bee visitation and contributes to diversifying pollinator diets and provides wild bee habitat.