Effect of Gestation Housing System (Individual vs. Group) on the Reproductive Performance of Sows over 6 Parities Under Commercial Conditions.

There is increasing pressure on the US swine industry to change from individual- to group-housing of sows in gestation. A low-cost approach for existing units would be to convert stalls to group pens. There has been limited research on the impact of this approach on sow performance. The objective of this study was to evaluate the effects on sow reproductive performance of housing sows during gestation either in individual stalls or small pens converted from individual stalls. This study was carried out at a commercial breed-to-wean facility using a RCBD with 2 gestation housing treatments: Individual vs. Group. Sows were housed in individual stalls until 35 d of gestation, when they were moved to treatment. Individual treatment: individual sow in gestation stall (0.54 x 2.07 m; floor space 1.12 m²/sow). Group treatment: pens of 8 (2.20 x 4.71 m; floor space 1.30 m²/sow), created by combining 8 individual stalls. A replicate was 1 pen of 8 sows, and 8 sows in individual stalls. A total of 426 replicates and 6802 parity records were collected. Management was according to standard commercial procedures. The experimental unit was individual sow. Data were analyzed using PROC MIXED of SAS; the model included fixed effects of treatment and random effects of block and replicate. There were no effects (P > 0.05) of treatment on conception or farrowing rate, days from weaning to breeding, or the number of piglets born alive. Pre-weaning mortality was greater (P < 0.05) for Group (15.2%) compared to Individual (14.2%) housing. The percentage of sows removed in gestation was greater (P < 0.05) for Group (14.3%) than Individual (12.1%) housing, which was mainly due to greater removals (P < 0.05) for injuries and poor body condition (16.3 vs. 10.0% of total removals, respectively). There were treatment effects (P < 0.05) on sow body condition score and body weight; however, differences were numerically small and not practically important. There were no treatment effects (P > 0.05) on any other reproductive measures. This study found relatively small differences in sow performance between the housing systems evaluated. This suggests that converting existing individual stall facilities to small group pens has minimal impact on sow productivity. Nevertheless, the differences observed, although relatively small, favored individual housing of sows and are of commercial significance. Further research is needed to validate these findings and to determine the underlying cause(s) of these effects.