Grassy–herbaceous land moderates regional climate effects on honey bee colonies in the Northcentral US

Abstract

Abstract The lack of seasonally sustained floral resources (i.e. pollen and nectar) is considered a primary global threat to pollinator health. However, the ability to predict the abundance of flowering resources for pollinators based upon climate, weather, and land cover is difficult due to insufficient monitoring over adequate spatial and temporal scales. Here we use spatiotemporally distributed honey bee hive scales that continuously measure hive weights as a standardized method to assess nectar intake. We analyze late summer colony weight gain as the response variable in a random forest regression model to determine the importance of climate, weather, and land cover on honey bee colony productivity. Our random forest model predicted resource acquisition by honey bee colonies with 71% accuracy, highlighting the detrimental effects of warm, wet regions in the Northcentral United States on nectar intake, as well as the detrimental effect of years with high growing degree day accumulation. Our model also predicted that grassy–herbaceous natural land had a positive effect on the summer nectar flow and that large areas of natural grassy–herbaceous land around apiaries can moderate the detrimental effects of warm, wet climates. These patterns characterize multi-scale ecological processes that constrain the quantity and quality of pollinator nutritional resources. That is, broad climate conditions constrain regional floral communities, while land use and weather act to further modify the quantity and quality of pollinator nutritional resources. Observing such broad-scale trends demonstrates the potential for utilizing hive scales to monitor the effects of climate change on landscape-level floral resources for pollinators. The interaction of climate and land use also present an opportunity to manage for climate-resilient landscapes that support pollinators through abundant floral resources under climate change.