Christina Grozinger

Populations of pollinating bees are showing declines across the world. These declines are the result of land use and climate change. Agricultural intensification reduces the availability of flowering plants that bees depend on for food, and increases the risk of pesticide exposure. Climate change can alter the phenology of flowering trees and herbs and their interacting bee pollinators, reducing the ability of the bees to obtain adequate nutrition from different habitats including forest and the open landscape. These factors interact to cause bees to be more susceptible to pathogens and parasites. While these underlying factors are well-understood, it is currently not possible to predict the effects of these factors on specific bee populations. If we can establish site-specific predictions of local land use and climate effects on bee populations, we can predict which populations will be most threatened. Furthermore, we can identify the main risks to those populations, which can be mitigated by adjusted management practices, of the bees and surrounding landscape to make bee pollinators more resilient to environmental changes. Modelling the effect of land use and climate on individual bee species and bee communities requires large data sets of entire landscapes, which have both great temporal and spatial range and resolution. Traditional approaches to monitoring bee populations and their interactions with the surrounding plant community are labour intensive. We propose to harness advances in remote sensing, genomics, and machine learning to assess landscape characteristics, bee behavior and health, and bee forage availability and preference to contribute with novel models and technology to a current political debate of how to conserve biodiversity without compromising food security.