Regular play opportunities during the grow-finish phase improve pig trainability and learning in a cognitive bias test

Abstract

Promoting play behaviour could enhance pig welfare in intensive production systems. Play is associated with positive emotions, and regular play opportunities may positively alter pigs' affective states. The affective state of pigs can be indirectly assessed through a behavioural test that determines an optimistic or pessimistic cognitive bias (CB). This study explored whether grow-finish pigs reared with intermittent play promotion over nine weeks exhibited more optimistic CB compared to conventionally-housed pigs. A subsample of 18 pigs was selected from three treatments (6 pigs/treatment): i) Control (CON), ii) Novelty (NOV), and iii) Play pen (PLP). Play was promoted 3x/week with novel objects provided either in a home pen (NOV, 1 m²/pig) or a larger playpen (PLP, 2.9 m²/pig). At 19 weeks of age, pigs were trained to approach a 'hot spot' (GO/correct) with a positive sound cue (POS) reinforced with a positive reward or avoid it (NO-GO/correct) with a negative sound cue (NEG) reinforced with a negative experience. Afterwards, all pigs' CB was assessed using an ambiguous sound cue (AMB) in two tests (CBT) and GO/optimistic and NO-GO/pessimistic responses were recorded. In training, 66 % of NOV and PLP pigs successfully trained for CBT, while 0 % of CON pigs did (p = 0.036, Fisher’s exact). During testing, the proportion of optimistic responses to the AMB cue exhibited by pigs in trained NOV and PLP treatment groups did not differ from pigs in untrained CON group in CBT 1 (p = 0.751, Kruskal-Wallis) and CBT 2 (p = 0.595). For the POS cue, the proportion of correct responses of trained NOV and PLP treatment pigs was higher compared to pigs from untrained CON group in CBT 1 (at a tendency, p = 0.075) and CBT 2 (p = 0.011). Pigs from the three treatment groups did not differ in their proportions of correct responses to the NEG cue in either CBT. The pig-level proportion of fearful/aggressive behavioural reactions to the negative experience tended to be higher in CON compared to trained and untrained play (PLY) pigs (p = 0.072, Mann-Whitney). Four PLY pigs and one CON pig showed a bold reaction to the negative experience. Results suggest that rearing pigs in a more stimulating environment with play opportunities improves trainability and learning, possibly due to reduced fear. Due to collinearity between treatment and successful training, no conclusions can be drawn about the effect of play promotion on the pigs’ affective states. Future studies should refine CBT methods to enhance understanding of the relationship between play and affective states in pigs.