Economic and production reasons dictate that the breeding stock of domestic pigs are normally fed a food ration less than their ad libitum intake (Whittemore 1987), and commercial levels of food restriction have been shown using operant conditioning to result in sustained feeding motivation (Lawrence et al 1988). However, operant conditioning procedures can be criticised on the basis that they impose artificial conditions on the animal.

The practice of food restriction with dry sows has been directly linked to the performance of stereotypic behaviour (Appleby and Lawrence 1987) which maybe interpreted as an indicator of poor welfare (Mason 1991). Stereotypies in sows have often been associated with physically restrictive housing such as stalls and tethers. However, stereotypies can also develop in loose housed sows if they are also food restricted (Terlouw et al 1991). The performance of stereotypies may then reflect underlying feeding motivation and the subsequent potentiation foraging behaviour (Hughes and Duncan 1988). Stereotypies may arise where restrictive housing in some way interferes with the expression of that foraging behaviour perhaps by modifying or channelling complex and variable foraging behaviour into more simple and often repeated forms (Lawrence and Terlouw, in press.