Honeybees constitute an important model for studying sensory processing, decision making and learning in the olfactory system, as they can perform complex tasks by detecting and processing olfactory stimuli with a very limited number of neurons.


In a rich environment with many volatile compounds around, honeybees need to find their path to the correct food source. They learn and keep in memory the cues of a flower with nectar, and they tend to look for the same type of flower while foraging.

This behaviour indicates that honeybees need to be good at odour discrimination and in forming memories related to a previously visited food source. Indeed, there is an impressive literature on how the honeybees can successfully achieve complex discrimination and learning tasks.

The objective for modelling olfaction in the Green Brain Project extended previous attempts to model the antennal lobes and their constituent glomeruli (which encode olfactory cues), the projection neurons and the mushroom bodies.

Odours are known to have a distributed representation in the antennal lobe, encoded as differential activation levels of glomerular populations. Odour mixtures are represented as a non-trivial combination of constituent odours’ representations which can be revealed by calcium imaging. Projection neurons in the antennal lobes then make connections to the Kenyon cells in the mushroom bodies, where single odours and mixtures are encoded in sparse firing patterns.

Our models were based on earlier work showing that the anatomy and known electrophysiological properties of the olfactory pathway of insects in combination with spike-timing-dependent plasticity (STDP) and lateral inhibition lend themselves to unsupervised self-organisation of synaptic connections for the recognition of odours.

In the Green Brain Project, we extended this model by adding mechanisms of reinforcement learning and realistic spiking patterns in the antennal lobes with mapping information obtained by calcium imaging.

The following animations, made by Dr. Nowotny, show spiking and rate activity in the bee olfactory system model. They show the response to 200ms control, followed by 200ms Hexanol, followed by 200 ms Hexanol+Nonanol, and finally followed by 200 ms Nonanol.


Rate dynamics