Do I sound anxious? Emotional arousal is linked to changes in vocalisations in domestic chicks (Gallus gallus dom.)

A major goal in animal welfare science is the development of methods that quantify or ‘read-out’ current affective states in freely behaving animals. In mammalian models, changes in acoustic parameters within vocalisations have been linked to differences in emotional arousal, although there are very few studies showing changes in vocalisations that act as indicators of valence (positive or negative affect). Currently, there are very few studies on vocal indicators of emotion in birds. To determine the link between acoustic parameters within vocalisations and emotion in domestic chicks (Gallus gallus dom.), we used a well-validated paradigm that purports to elicit two distinct, negative emotional states: anxiety and depression. In this paradigm, chicks in social isolation initially show high rates of distress calls (anxiety-like phase: high arousal/negative affect) that decrease over the next 20–30 minutes to a depression-like phase (low arousal/negative affect). We analysed acoustic parameters of distress vocalisations from Legbar chicks (4–7 days old) that were placed in social isolation for 30 minutes, either with or without a mirror (Isolated: n=18; Mirror: n=16), (a mirror is known to reduce the negative behavioural and physiological indicators of isolation). Chicks in the Isolated condition produced louder calls, of a higher frequency, and calls were individually longer, and more ‘degraded’ (higher entropy and lower harmonics-to-noise ratio) compared to the Mirror condition. However, the call rate did not drop below 50 % of the initial rate in either condition, indicating the chicks in this study did not reach the criteria for the depression-like phase of the paradigm. To investigate further we analysed vocalisations at times during a trial when calls were at a high rate per minute (indicating a higher level of emotional arousal at that point in time) and we found those vocalisations were louder, more ‘degraded’, longer and higher in frequency across both conditions. These results are consistent with findings across several species of mammal, indicating that changes in emotional arousal in negatively valanced situations give rise to consistent changes in vocal parameters across a wide range of species. Indeed, further work could look to see if the same links are found in groups such as chelonians, crocodilians and geckos, as their vocal production mechanisms share some similarities. There could also be scope to use these changes to automatically detect shifts in emotion in captive animals.