Previous work has shown that feeding regime and not physical restraint is a major cause in the development of stereotypies in closely confined sows, (Teriouw et al, 1991). However, this and a previous report on the relationship between feeding regime and stereotypies in sows (Appleby and Lawrence, 1987) have only studied gilts in their first pregnancy. The objective of this present work was to examine longer-term effects of feeding regime and housing on the development of stereotypies in sows.

Thirty-two sows (Cotswold Pig Development Co. Ltd, UK) were allocated to either loose (Lo) or tether (Te) housing and to Low (L; 2.5 kg/day) or high (H; 4.0 kg/day) food levels in a balanced design. Apart from social contact and freedom of movement the housing systems were similar including the provision of chains to the loose housed animals. The behavioural response to these treatments was observed over parities 2,3 and 4 using a time sampling technique, and the data analysed by analysis of variance and linear correlation.