Exploration of positive social relationships in group-housed gestating sows
Résumé
Social behaviour constitutes a crucial aspect of positive animal welfare in farm animals. While preferential relationships for positive social interactions have been reported in ruminants, like cows and sheep, evidence for their existence in pigs is limited, despite potentially significant implications for their welfare. We aimed to assess whether group-housed sows had preferential partners for positive social interactions and whether individual attributes influenced these preferences. We used four groups of gestating sows comprising 17 to 19 sows from various parities and dominance ranks. Two groups were housed in a barren environment and two groups were housed in a larger environment with straw bedding. We continuously recorded emitted and received socio-positive interactions, including body exploration (i.e. nosing and grooming body) and snout-to-snout contacts, on video recordings for three hours/day on five consecutive days at the end of gestation. Social interactions were averaged over the five days and expressed as counts/hour. A total of 2831 social interactions (averaging 40 ± 3.7 interactions/sow) were scored. Sows engaged in an average of 2.76 ± 0.25 socio-positive interactions/hour, comprising 2.07 ± 0.20 body exploration and 0.69 ± 0.06 snout-to-snout contacts/hour. Low-parity sows displayed more body exploration than primiparous and high-parity sows (p < 0.001), and engaged in more snout-to-snout contacts than high-parity sows (p = 0.04). Sows in a barren environment displayed more snout-to-snout contacts than sows in an enriched environment (p = 0.02), with no difference observed in body exploration. Using social network analysis methods and permutation tests, we identified non-random dyadic associations for both body exploration and snout-to-snout contacts in all four groups of sows. While dominance did not affect social preferences, sows preferred to explore sows from the same parity class in most groups. Networks for body exploration and snout-to-snout contacts were highly correlated. In summary, our study highlights the presence of preferential relationships for socio-positive contacts within groups of gestating sows, influenced in part by parity. Incorporating these relationships into management practices could be pivotal in enhancing the welfare of sows on farms.