Inferring plant community phenology via bee-collected pollen

Abstract

Global climate change is producing novel biospheric conditions, presenting a threat to the stability of ecological systems and the health of the organisms that reside within them. Variation in climatic conditions is expected to facilitate phenological reshuffling within plant communities, impacting the plant-pollinator interface, and the release of allergenic pollen into the atmosphere. Impacts on plant, invertebrate, and human health remain unclear largely due to the variable nature of phenological reshuffling and insufficient monitoring of these trends. Large-scale temporal surveillance of plant community flowering has been difficult in the past due to logistical constraints. To address this, we set out to test if metabarcoding of honey bee collected pollen could be used to infer the phenology of plant communities via comparison to in situ field monitoring. We found that honey bees can accurately indicate the onset of anthesis, but not its duration, in the plant species they selectively forage on. Increasing the number of colonies used to monitor, and employing a multi-locus approach for metabarcoding of pollen, substantially increased the species detection power of our approach. Here, we demonstrate that metabarcoding of honey bee collected pollen can substantively streamline the establishment of long-term phenological monitoring programs to document the on-going consequences of global climate change and its impact on the temporal aspects of plant-pollinator relationships.