Effects of (a switch in) enriched vs barren housing on the response to reward loss in pigs in a negative contrast test

Several studies suggest that animals in a negative emotional state are more sensitive to reward losses as shown by behavioural and neurophysiological responses. In a successive negative contrast (SNC) test, reward losses are induced by decreasing the size of the reward for a task for which animals have been trained. is SNC paradigm has not been widely used in pigs. It is well known that environmental enrichment positively inuences the welfare of pigs, and may induce a more optimistic emotional state, which could reduce their sensitivity to reward losses. We studied pigs in barren (B) or enriched (E) housing, experiencing either a switch in housing conditions at 7 weeks of age or not (4 treatment groups: EE, EB, BE, BB, n=8 pens per group) in an SNC runway task. We hypothesized that B housed pigs, particularly those that changed from E to B housing, would show an enhanced sensitivity to reward losses. One pig per pen was trained to run a 24.6 m U-shaped runway for 6 pieces and one for 1 piece of apple. Each pig received 3 trials per day, with a maximum of 120 sec/trial. Latency to start eating the reward was recorded, and the average was calculated per day. Aer 11 days, all pigs received 1 piece of apple only for another 11 days (reward shi: 6-1 vs 1-1 reward group), i.e. the group originally receiving 6 pieces of apple experienced a reward loss. Eects of pre-housing, post-housing, (original) reward size, day and interactions were analysed using mixed models with a random eect of animal. Fiy-one pigs were successfully trained. Before the reward shi, over the rst 11 days, pre-housing × post-housing × reward size aected the average run-time (P<0.05). All BB pigs ran slower than other pigs (BB: 59.3±2.8; BE: 35.9±1.7; EB: 39.6±2.2; EE: 40.9±2.2, P<0.05), without any other signicant pairwise dierences. Analysis per treatment revealed, however, that EB 6-reward pigs were faster than the 1-reward pigs. Overall latency was higher on the last days (P<0.001). Aer the reward size shied to 1 on day 12, pre-housing × post-housing aected the latency (P<0.001). Post hoc analysis showed that again, BB pigs were slower than other pigs (BB: 88.2±2.7; BE 62.3±2.3; EB: 57.3±2.3; EE: 70.4±2.6, P<0.001), and EB pigs were faster than EE pigs (P<0.05). Pigs ran slower aer than before the reward shi (P<0.001). Nevertheless, pigs in the 6-1 group ran slower than pigs in the 1-1 group aer the reward shi (6-1: 73.9±2.0; 1-1: 66.4±1.8, P<0.05), suggesting that pigs are sensitive to a loss in reward size. is was, however, irrespective of housing given the lack of interactions with reward size. We conclude that housing aected the latency to run down a runway for a reward in pigs, which can indicate a lower motivation in the BB pigs, an eect that was absent in the B pigs that switched to enriched housing (BE pigs). We found, however, no evidence that housing or a switch in housing conditions aected the sensitivity to reward loss.