A Pilot Study on Behavioural and Physiological Indicators of Emotions in Donkeys

SIMPLE SUMMARY: Animal welfare science is facing the challenge of measuring the subjective feelings and emotions of animals. In the case of donkeys, this could lead to a better understanding of their specific needs, improved training methods, and a healthier relationship with humans and the environment from a holistic perspective that considers the interconnection between the human, animal, and environmental health and welfare (One Health-One Welfare). In this study, we tried to overcome the difficulty of reading animal emotional states by means of some of the non-invasive parameters cited in the scientific literature. These behavioural and physiological parameters were used to compare the emotional response of donkeys in two situations of different emotional content in order to find possible indicators of positive or negative emotions in this species. The results showed significant differences in behavioural (in terms of ears orientation, postural changes, and behavioural complexity) and physiological (heart rate variability) responses when comparing a rewarding vs. a frustrating context. Despite the small sample considered for the analysis, which limits the possibility of generalization of the results, this preliminary investigation highlights promising outcomes and indicates that further investigation of these non-invasive parameters could lead to the identification of valid emotional state indicators in donkeys. ABSTRACT: Recognizing animal emotions is critical to their welfare and can lead to a better relationship with humans and the environment, especially in a widespread species like the donkey, which is often prone to welfare issues. This study aims to assess the emotional response of donkeys through an operant conditioning task with two presumed different emotional contents. Specifically, a within-subject design including positive and negative conditions was conducted, collecting behavioural and physiological (heart rate variability and HRV) parameters. Facial expressions, postures, and movements were analysed by principal component analysis and behavioural diversity indexes (frequencies, activity budgets, richness, Shannon and Gini-Simpson). During the positive condition, both ears were held high and sideways (left: r = −0.793, p < 0.0001; right: r = −0.585, p = 0.011), while the ears were frontally erected (left: r = 0.924, p < 0.0001; right: r = 0.946, p < 0.0001) during the negative one. The latter was also associated with an increased tendency to walk (r = 0.709, p = 0.001), walk away (r = 0.578, p = 0.012), more frequent changes in the body position (V(Body position) = 0, p = 0.022), and greater behavioural complexity (V(Gini-Simpson Index) = 4, p = 0.027). As for HRV analysis, the root mean square of successive beat-to-beat differences (rMSSD) was significantly lower after the negative condition. These non-invasive parameters could be considered as possible indicators of donkeys’ emotional state.