Pre-weaning social behaviours and peripheral serotonin levels are associated with behavioural and physiological responses to weaning and social mixing in pigs

In pig production systems, weaning is a major challenge that is usually paired with social mixing and may greatly affect health and welfare of piglets. Research efforts have been devoted to characterising early predictors of weaning adaptation, but have focused mainly on aggressive and harmful behaviours, whereas socio-positive behaviours have been poorly studied. Furthermore, serotonin (5-HT), a neurotransmitter regulating social behaviours, may also be a pertinent predictor of piglets' adaptation to challenging situations. This study aimed to assess whether social behaviours and blood 5-HT concentration before weaning were associated with behavioural and physiological responses of piglets to weaning. Social interactions (social exploration, aggression, playfight, locomotor play) of 72 focal piglets from 12 litters were scored continuously for 8 h at 42 days of age. At weaning (d48), focal piglets were allocated to four pens of 33 piglets from six litters. During the two days following weaning (d49-50), social interactions were scored continuously for 6 h per day, and behavioural activities were scored with 6-min instantaneous scan sampling. Blood was sampled one week before (d41) and 24 h after (d49) weaning to measure 5-HT concentrations and health-related variables. Exploration of pen mates represented 55% and 79% of all scored social interactions before and after weaning, respectively, and play was not observed after weaning. Using a multivariate analysis paired with clustering analysis on post-weaning behavioural and physiological responses, we identified three clusters of piglets with distinct profiles of adaption to weaning: unhealthy inactive animals, healthy inactive aggressors and healthy active affiliative animals. Compared to other clusters, unhealthy inactive animals at weaning were characterised by lower levels of social exploration and aggression before weaning (p < 0.05 for both). Furthermore, piglets that explored their pen mates more before weaning were more active (p = 0.03) after weaning, while piglets that were involved in greater number of locomotor play episodes (p = 0.009) or that were less aggressive (p = 0.04) before weaning walked more after weaning. Piglets with higher blood 5-HT concentrations before weaning were less aggressive (p = 0.01) and had greater growth (p = 0.009) after weaning. Pre-weaning aggression was also positively associated with post-weaning lymphocyte count (p = 0.04), and pre-weaning locomotor play with post-weaning hydroperoxide concentration (p = 0.05), a marker of oxidative stress. Our findings suggest that pre-weaning social behaviours and blood 5-HT concentration may be relevant predictors of piglets' adaptive responses to social mixing at weaning and deserve more research attention.