Project title: Characterising pollinator energetics and foraging strategies using machine learning
PI: Geraldine Wright
Pollinators such as bees play a critical role in plant reproduction and contribute to the yields of over 70% of global crops. Despite extensive research into how floral traits such as colour or scent shape bee foraging behaviour, we have poor knowledge about the energetic factors influencing their floral choices. A major reason for this knowledge gap is the difficulty in characterising the energetic expenditure of bees in the wild. Using the common and commercially-important buff-tailed bumblebee (Bombus terrestris), my research will combine on-bee microsensors - ‘bee backpacks’ - with machine learning approaches to classify behavioural state and estimate the energetic expenditure of bees in the field. This system will give unparalleled detail on bumblebee foraging trips, facilitating research into longstanding questions in pollination energetics. These data also have multiple wider applications such as quantifying the floral resources required to support bee populations and in optimising the nectar/pollen rewards of crops to increase pollination.
