Short- or long-term affect? Disentangling the effects of emotion and mood on pigs’ decisions on a Judgement Bias Task

Abstract

The Judgement Bias Task (JBT) is a promising proxy measure of animals’ affective states, a growing field in animal welfare science and beyond. However, it remains unclear if judgement biases are predominantly affected by long-term mood, short-term emotions or a combination of both. This study aimed to systematically disentangle the effect of mood- and emotion-inducing treatments on pigs’ decisions on a JBT using positive human-pig interactions as affect manipulation. Across three batches, thirty-six pigs were trained on a spatial Go/No-go task with active trial initiation. During the three weeks of training and during testing, 12 pigs were administered five minutes of positive interactions daily with a familiar human (MoodPlus), whereas the remaining 24 pigs did not receive such treatment (MoodMinus). When pigs were tested on the JBT, all 12 MoodPlus and 12 of the MoodMinus pigs received three test sessions that were preceded by five minutes of positive interactions (EmoPlus) and three test sessions without positive interaction before (EmoMinus). The remaining 12 MoodMinus pigs were allocated to a control group, which did not receive any specific positive interaction treatment. Thirty pigs (83 %) reached the learning criterion and were tested across six test sessions. Pigs from all treatment groups showed highly “optimistic” responses to all ambiguous cues, leading to a ceiling effect, which may have overshadowed potential treatment differences (all effects apart from cue type: ${\chi }_1^2$ ≤ 1.75, p ≥ 0.19). Results from Human Approach and Avoidance Tests at the end of the study and the analysis of the behaviour of pigs during EmoPlus sessions indicated that all pigs, independent of treatment, had developed a positive relationship with the experimenter over the course of the study. These results suggest that positive human-animal interactions may not be an ideal treatment in the context of JBT when training requires regular contact with humans. MoodMinus pigs tended to respond more “optimistically” in EmoMinus sessions than Control pigs (χ²₁ = 3.58, p = 0.06), which may indicate a carry-over effect from EmoPlus to EmoMinus sessions. We call for future studies with a different treatment than human-animal interactions that implements both long- and short-term interventions in a systematic way to disentangle the effect of mood and/or emotion on judgement biases.