Bumble-BEEHAVE: An agent-based population model for bumble bees and its application as a decision tool for farmers.Matthias A. Becher1,2, Grace Twiston-Davies2, Juliet L. Osborne2

The Fields Institute for Research in Mathematical Sciences, 2Environment & Sustainability Institute, University of Exeter, Penryn Campus, Penryn, Cornwall, TR10 9FE UK;

The decline in abundance and species-richness of bumblebees raises serious concerns as they are important pollinators for crops and wild flowers. While honeybees have been studied for a long time, resulting in a plethora of data as well as models on colony dynamics, foraging behaviour, and host-parasite interactions data on bumblebees is comparatively scarce and few models are available.

Nevertheless, a bumblebee population model that is based on individual behaviour, and resulting in colony dynamics could be immensely useful to improve our understanding of the threats and stressors bumblebees face in the wild.

Here I present our new bumblebee population model Bumble-BEEHAVE. The model simulates multiple bumblebee species in a spatially-explicit landscape. Bees, represented as individuals or as cohorts, decide about their activity, using a stimulus-threshold approach. Activities include egg laying, foraging for nectar and pollen, and brood care. Successful colonies will produce new queens and males. Stressors such as parasites, predators and pesticides, can, to some degree, be included. Bumble-BEEHAVE can be used to predict and identify the variables associated with bumblebee colony success. Beesteward, a modified version of Bumble-BEEHAVE, was then designed to be specifially used by farmers as a management tool for the conservation of pollinators in agricultural landscapes.