Young sows are subordinate and vulnerable in group-housing systems because they usually lose most fi ghts and suffer more injuries than mature sows at mixing. This study was conducted to evaluate effects of sorting by parity on reducing aggression and associated stress with the aim to improve welfare and performance of fi rst-parity sows in a group-housed system. Sows and gilts (n = 180) from 6 breeding groups were used. Within each group, 2 groups of 15 females were mixed in each of 2 treatment pens after weaning and remained there throughout the entire gestation period. The control pen consisted of 11 multiparous and 4 fi rst-parity sows, and the treatment pen consisted of 11 gilts and 4 fi rst-parity sows. Before mixing and at the end of the gestation period, sows and gilts were weighed individually, assessed for BCS, and measured for backfat thickness. Injury scores were assessed before and 48 h after mixing and wean-to-mating intervals, farrowing rate, and litter performance at the subsequent farrowing were recorded for all females. Aggressive interactions involving fi rst-parity sows were video recorded for 72 h immediately after mixing in each pen. Data were analyzed using the Glimmix procedure of SAS with a Poisson regression model for count data and a Gaussian model for continuous data. All females in treatment pens sustained fewer scratches (P = 0.01) after mixing than females in control pens. First-parity sows in treatment pens fought more frequently (P = 0.01), tended to fi ght for longer periods (P = 0.08), and won more fi ghts (P = 0.04) of parallel pressing but had fewer injures (P = 0.03) after mixing, gained more BW (P = 0.01) during gestation, and had greater farrowing rates (P = 0.03) compared with fi rst-parity sows in control pens. The results suggest that sorting by parity shielded fi rst-parity sows from severe injuries caused by mixing-induced aggression so that their welfare and performance can be improved in group housing systems.
- Gestation housing