The fruit fly Drosophila melanogaster has a sophisticated visual system and performs a wide range of visuomotor behaviours such as foraging, mating, and navigation. These behaviours may operate over very different time scales, from tens to thousands of milliseconds. Observation and characterisation of these behaviors is technically difficult based on these short timescales, and methodologically challenging because of the behavioural variability involved.  

Widefield stabilisation is a robust visuomotor behaviour essential to maintain stable flight in response to environmental disturbances. Most of the current understanding of this behaviour comes from tethered experiments, where the subject is rigidly restrained and the normal sensorimotor experience disrupted. We have created a novel virtual reality system and an experimental paradigm that allows studying this behaviour in freely moving subjects, in a naturalistic environment. 

While in free flight we actively manipulate the subject by systematically eliciting the optomotor response. Continued elicitation of this behaviour allows us to control the position of the fly in 3D. To probe the behaviour in greater detail we can also perturb the fly or show challenging visual stimulus. Using control theoretic methods taken from the engineering fields we can build behavioural models for individual flies and individual trials. 

Using this paradigm, and in combination with the powerful genetic tools available for Drosophila, we are uncovering the frequency and temporal characteristics, and the role of certain neurons in the optomotor response in free flight. 

From the success of these experiments we are also investigating other visuomotor behaviours, using a similar virtual reality assay, in freefly swimming zebrafish.