To control their movement trajectories, animals rely on sensory feedback arising from self-motion.
One prominent example for this are panoramic image shifts on the retina, termed optic flow.
Among other functions, the associated neural signals are thought to stabilize locomotor performance against internal and external perturbations.
We aim to understand the underlying working principles and functional organization of such a feedback system at the level of individual neurons.

First, I will discuss some of the mechanisms and neuronal substrates, which endow wide-field motion-sensitive neurons with flow field selectivity.
I will then report on the causal roles of optic flow detectors in tethered steering behaviour using new optogenetic tools.
Tethering animals affords precise experimental control over stimuli.
However, it is also unnatural. The animal is forced to behave in a certain way and sensory feedback from self-motion is either lacking or untypical.
Therefore, we also perform selective neuronal activity perturbations in freely roaming flies tracked by computer vision.
From behavioural phenotypes, we hope to obtain new mechanistic insights into visually guided course control in a naturalistic context.