Housing during early pregnancy affects fertility and behaviour of sows

From 1.1.2013 EU legislation will permit stall-housing of dry sows for only 4 weeks post-service. Stalls are associated with compromised animal welfare, but group housing comes with risks from social confrontations. Studies on housing effects on fertility have produced contradictory results as systems are inconsistently defined. Effects of housing on pregnancy rate 28 days post-service (P28), early disruption of pregnancy (EDP) and behaviour were investigated in 12 replicates of 40 sows. Half of the animals were stalled (treatment S, mean parity 2.4) and half group-housed adjacent on 5.1 m(2) deep litter per sow (treatment G, mean parity 2.3). Pregnancy was detected using real-time ultrasound. An EDP-diagnosis required a drop of salivary progesterone concentration from >15.9 to <15.9 mmol/l in a series of samples taken on days 17, 20, 24 and 27 post-service. Behaviour was scan sampled on day 27 for 120 min with 10-min intervals. Data was analysed using mixed regression models in MLwiN. Treatment S increased the odds for P28 with 2.3 and the pregnancy rate of a treatment group with 0.1 compared with G. G increased the odds for EDP with 4.6. Treatment did not affect total exploratory or total passive behaviour. S sows explored floor and fixtures (not bedding) more and performed more passive sitting than G sows. Social stress may have caused the impaired reproductive performance in G sows. S sows showed behavioural signs of decreased welfare, but no corresponding reproductive effects. An interdisciplinary approach is needed when evaluating the appropriateness of housing systems.