The effects of dry sow housing conditions on welfare at farrowing
Measures of production, physiology, behaviour and pathology were used to assess the effects of confined and loose dry sow and farrowing housing systems on the welfare of the sow and her litter. Litter size decreased and piglet mortality increased greatly after the sixth parity, regardless of housing system, resulting in a sharp decline in the number of piglets weaned per litter. Stall-housed sows gave birth to the most piglets per sow per year, but also had the highest piglet mortality. Overall, piglet mortality was higher in farrowing pens than in crates. Sows from the large group had a significantly larger number of returns to service after farrowing in crates. Behaviourally, all sows adapted well to the farrowing house. All sows showed an increase in the number of posture changes, reaching a maximum during the 24 hours immediately prior to parturition. However, this increase was greatest in those sows in farrowing crates, which had previously been housed in an open environment. Heart rate was influenced by stage of gestation, posture and behaviour. Stall-housed sows had a higher basal heart rate and heart rate response to feeding than group-housed sows, perhaps indicating decreased cardiovascular fitness and an increased sympathetic nervous response to stimuli such as food. When farrowing in crates, group-housed sows had a higher heart rate response to the suckling event than stall-housed sows. This may be due to general unresponsiveness in stalled sows or to high reactivity to the suckling event in group sows caused by frustration of mother-infant interaction. When involved in agonistic interactions, the change in heart rate was greatest for sows which lost a physical interaction. Stall-housed sows had weaker bones than group-housed sows, and different muscular conformation, probably due to lack of exercise. Bone and muscle weakness may increase the susceptibility of stall-housed sows to lameness. When lying down, stall-housed sows had greater difficulty and took longer than group-housed sows. The times taken for stall sows to lie down and to stand up quickly were positively correlated with body length. For group-housed sows lying down in the open, the time taken was positively correlated with proportional locomotory muscle weight. Spatial restriction when lying resulted in the loss of muscular control. There was a positive correlation between body length and the number of piglets crushed for stallhoused sows and group-housed sows farrowing in crates. There was also a positive correlation between body length and crushing mortality for group-housed sows farrowing in pens. This indicates that sows can have problems controlling movements, even in the presence of piglets. The results presented reveal several welfare problems resulting from stall housing during pregnancy. It would appear unreasonable to confine sows during farrowing, if they have previously been housed in an open environment. However, not enough is known about the causes of piglet mortality and any decision concerning the continued use of farrowing crates must take account of the trade-off between sow welfare and piglet welfare.