Housing conditions do not alter cognitive bias but affect serum cortisol, qualitative behaviour assessment and wounds on the carcass in pigs

Measures of animal emotions are essential to assess animal welfare. Recently, the cognitive bias technique has been proposed as a measure of animal affective state. This technique is based on the premise that subjects in negative affective states make more negative judgements about ambiguous stimuli than subjects in positive affective states. In the present study, 44 female pigs were divided into two groups of equal size (22 pigs each): one group was allocated in enriched housing conditions (more space allowance, presence of straw and solid floor) and the other in barren housing conditions (lower space allowance and slatted floor) in order to induce differences in the affective state. Three cognitive bias tests (CBT) based on spatial discrimination were performed: 1 week before starting the housing conditions (CBT1) and 1 and 5 weeks (CBT2 and CBT3, respectively) after it. Moreover, 3 and 4 days after each CBT, a qualitative behaviour assessment (QBA) and a serum sampling for the assessment of cortisol concentration were carried out. Finally, the number of wounds was counted on the pig carcass at slaughter. The results showed that the cognitive bias did not differ between treatment groups in any of the two CBT carried out after starting the housing conditions (p > 0.05). However, during the CBT2 and CBT3 when compared with the barren group the enriched group presented a lower concentration of serum cortisol (p = 0.008 and p = 0.011, respectively), a higher QBA score (p = 0.022 and p = 0.027, respectively) and a lower number of wounds on the carcass (p = 0.05). Considering the QBA, serum cortisol and carcass wounds results, the CBT used in this study was not valid or not sensitive enough to assess the variation in the affective state between pigs raised in different housing conditions.