Group housing of sows: Effects of season and day of grouping on sow welfare and reproductive performance

A decrease in reproductive performance was observed by some farmers during summer, following the entry into force of Directive 2008/120/EC requiring sows and gilts to be group housed during pregnancy. Mixing sows at 28 days of pregnancy might increase both embryo resorption and abortions due to the adverse effects of cortisol on corpus luteum, therefore some farmers asked the Public Veterinary Services the authorization to group sows after 35 days from service. Our aim was to assess the effects of mixing at different gestational ages on sows’ hormonal profiles and reproductive performances. Two batches of sows (one during winter and one during summer) were followed during a reproductive cycle (PIC Camborough-22, N=30 sows per batch, average parity: 4.4±2.8). Each batch was split into two groups: one was kept in single gestation stalls until day 28 of pregnancy and then moved to group housing (MIX28), the other was kept in single stalls until day 42 of gestation and then moved to group housing (MIX42). Group housing was on partially-slatted floors with ad libitum fibrous feed. Bristles were sampled from each sow at the end of pregnancy (when they were moved to farrowing rooms) and bristle cortisol was assayed as an indicator of cortisol levels during the entire pregnancy. Blood samples were drawn from 12 sows per group, at 32 (B1) and 46 (B2) days after service to assess progesterone plasmatic concentration. Productive and reproductive parameters of all sows were recorded. Plasmatic progesterone and hair cortisol values were submitted to two-way ANOVA, using season and day of mixing as the main effects. Hair cortisol at the end of pregnancy did not differ between MIX28 and MIX42 sows (5.53 vs 5.98pg/mg; P<0.60). Progesterone level as well did not show any significant difference between MIX28 and MIX42 sows at 46 days after service (B2) (23.18 and 21.76ng/ml respectively; P<0.41). However, at 32 days after service MIX28 sows had lower progesterone levels if compared to MIX42 (25.15 vs 30.14ng/ml; P<0.01), probably due to a transitory effect of mixing (cortisol) on corpus luteum. A significant (P<0.001) effect of season on plasmatic progesterone was observed, with lower concentration during summer than during winter at both blood samplings, regardless of the day of mixing (26.54 vs. 29.50ng/ml at B1; 18.71 vs. 25.70ng/ml at B2). All progesterone concentrations fell within the physiologic range for pregnant sows. No differences were observed between seasons and between groups in the main productive and reproductive parameters (gestation and lactation duration, newborns, stillborns, mummified and weaned piglets, abortions, etc). Our results show that a prolonged stay (42 days) in single stalls does not increase stress levels of sows (no differences in cortisol from bristles). On the other hand, such a prolongation does not improve productive or reproductive performances. Our findings therefore do not confirm the necessity (as expressed by farmers) to keep sows in single stalls until at least 35 days after service.