Structure and efficiency of pollination networks in changing environments – POLLINET

Bees are ecologically and economically vital pollinators whose widespread declines associated to human activities and other environmental stressors have serious consequences for our ecosystems and food security. Risks that we may be facing a pollination crisis, in which crop yields begin to fall because of colony collapses and inadequate pollination, have made research on plant-pollinator systems an absolute societal priority. While many studies in pollination ecology have focused on high-level functional associations between plant and pollinator species (within communities and ecosystems) based on field surveys, a comprehensive understanding of these complex phenomena also urgently requires fundamental knowledge on lower-level interactions between individual plants and pollinators (within populations) based on experiments. Specifically, growing evidence shows that many pollinators use predictable foraging routes to exploit spatially scattered floral resources, thus creating appreciable potential for critically influencing pollen flow and plant reproduction patterns. The aim of POLLINET is to develop a multilevel mechanistic analysis of plant-pollinator systems, by dissecting the complex foraging strategies of bees, identifying their impact on plant reproduction, and determining how this varies in response to specific environmental changes. Our approach is based on a comparative analysis of the spatial behaviour of two key pollinators, the wild bumblebee (Bombus terrestris) and the managed honeybee (Apis mellifera), in controlled yet ecologically relevant, environments, where we will selectively vary the spatiotemporal distribution of floral resources, their nutritional composition and the density of pollinators. Comparing naturally co-occurring bee species with different social ecologies will inform about the importance of pollinator diversity in these complex and fragile interactions. We will combine (1) manipulative experiments on bees foraging for nectar and pollen in fully-automated arrays of artificial flowers equipped with tracking systems (radio frequency identification, video cameras, harmonic radars) in the lab and in the field, (2) network analyses of their spatial movements using modern mathematical tools of graph theory, (3) and computer simulations of agent-based models linking bee behaviour to plant reproduction. The project will be implemented by an interdisciplinary team of 10 scientists led by an early career researcher (Lihoreau), in collaboration with high profile academics in international institutions. Our unique integration of concepts from cognitive, behavioural, nutritional, spatial, pollination, plant and population ecology to those of computer biology, physics and mathematics, will shed new light on fundamental mechanisms underpinning a vital yet threatened ecosystem service in the context of global environmental changes.