Dry sows in dynamic groups: An investigation of social behaviour when introducing new sows

In the investigation 22 groups of sows (with a total number of 73 sows, being partially observed repeatedly) were consecutively observed for 96 h with the aid of video technology after the introduction of 5 new sows into a group of 10 resident sows, thus creating a dynamic group of 15 animals. The aim was to analyse the frequency of agonistic interactions (AI) and the lying behaviour of new and resident sows. Additionally, the lying behaviour three weeks after the introduction of new sows was recorded in half of the 22 groups. The number of AI per group was 45.2 on the first day and decreased significantly to 30.0 AI/group on the fourth day after the introduction. Between 82.4 and 87.9% of all AI included the participation of at least one new sow. Including all days of observation 2.3 AI per dyad occurred between resident and new sows (n = 50 dyads), 0.5 AI/dyad within the subgroup of resident sows (n = 45 dyads) and 1 AI/dyad within the subgroup of new sows (n = 10 dyads). Altogether 12,454 lying events were recorded from which 6134 occurred clearly either next to a member of the own subgroup (resident and resident, new and new) or next to a member of the other subgroup (resident and new or vice versa). On the 1st day new sows showed significantly more lying events than resident sows (21.2/24 h, 14.9/24 h, respectively). By the 21st day this number decreased significantly (13.9/24 h) no longer showing significant differences to the number observed for resident sows (14.2/24 h). On the 1st day resident as well as new sows lay down significantly more often next to a member of the own subgroup (new sows 94.3%; resident sows 96.8%). The number of lying events next to a member of the own subgroup decreased significantly within 3 weeks after the introduction of new sows, indicating the advancing integration of new sows into the dynamic group (new: 1st day 94.3%, 21st day 46.0%; resident: 1st day 96.8%, 21st day 74.6%).