Avoiding danger and approaching food or mates are one of the most essential and conserved set of behaviours, observed in most species from crabs to primates. Nevertheless, in an ever-changing environment, the type and kinetics of these innate behaviours need to be flexible and context-specific to optimize an animal’s survival. However, it remains unclear to which degree innate behaviours can be influenced by the environment of an animal; as well as how brain circuits are organized to allow reliable and fast, yet flexible and context-specific behaviours. We are addressing these questions by studying the effect of ambient light on innate reactions of rodents. In nature, animals will encounter predators and prey under a variety of different light conditions, which have intrinsically distinct levels of danger and opportunity. However, most studies are performed under daylight conditions, and it is unknown how different light conditions affect predator avoidance and prey capture behaviour.

To induce innate avoidance and approach in freely moving animals, we presented standardized visual stimuli mimicking predators and prey on the top and side monitors of a behavioural box with a Plexiglas shelter in one corner. Rodents were filmed during those experiments and behaviour type (skittish darts, escape, running, stopping, freezing) as well as their kinetics were analysed offline.  While similar paradigms are used by many different research groups, it remains unclear what parameters enable fast and reliable data collection while preventing habituation of the animals to the behaviour-eliciting stimuli. We hence set out to first test various combinations of session intervals and shelter locations, and their effect on visually induced behaviour in Mus musculus. We found that shelter location impacted session duration since mice spent more time foraging if the shelter was placed opposite of the setup entrance, providing more opportunities for stimulus presentations. Furthermore, while the time between early session (3 or 7 days) had little effect on behaviour, short time intervals led to switches between escaping and stopping.

We then used the same experimental paradigm with a 7-day session-interval to test the effect of ambient light on innate behaviour of American (Peromyscus) and Eurasian (Mus) rodent species. We found general and species-specific adaptation to different ambient light levels. For example, a looming stimulus mimicking an attacking predator induced a faster onset and more vigorous reaction in all species under moonlight than under brighter light conditions. For cricket-mimicking stimuli and non-threat stimuli, on the other hand, different species changed their behaviour differently depending on the light level.  In addition, pre-stimulus behaviour could predict some reaction to threat, but only in a light-specific manner. Taken together, we found a strong and species-specific effect of ambient light on various innate reactions, suggesting that successful context-specific avoidance and hunting strategies are passed on to offspring through inheritance rather than learning. Finally, ongoing Neuropixels recordings of the underlying behaviour-driving circuits through the superior colliculus suggest ambient light specific encoding of ethologically relevant visual stimuli.