Affective states, which refer to feelings or emotions, are a key component of animal welfare, but these are also difficult to assess. Drawing upon a body of theoretical and applied work, we critically review the scientific literature on the assessment of affective states in animals, drawing examples where possible from research on dairy cattle, and highlighting the strengths and weaknesses of scientific methods used to assess affective states in animals. We adopt the "valence/arousal" framework, describing affect as a 2-dimensional space (with valence referring to whether an experience is positive or negative, and arousal referring to the intensity of the experience). We conclude that spontaneous physiological and behavioral responses typically reflect arousal, whereas learned responses can be valuable when investigating valence. We also conclude that the assessment of affective states can be furthered using mood assessments and that the use of drug treatments with known emotional effects in humans can be helpful in the assessment of specific affective states in animals.

Early cow-calf separation prevents much of cows’ natural maternal behaviour. Early separation is thought to prevent the development of a cow-calf bond. To assess this bond, we measured motivation of dairy cows to reunite with their calf. To vary the degree of bonding, some cows were allowed continued contact with their calf and others were separated from their calf soon after birth, following standard practice on most farms. Among cows allowed continued contact, some were able to suckle their calf and others were prevented from suckling (by covering the cow’s udder with an udder net). Cows were habituated to the weighted-gate apparatus before calving by daily training with the (un-weighted) gate. After calving, cow willingness to use the gate was assessed by determining if she would push open the gate to access to her own calf. Testing occurred once daily, with weight on the gate gradually increased. After passing through the gate, the dam’s calf-directed behaviour was recorded. Suckled cows pushed a greater maximum weight (45.8 ± 7.8 kg) than separated cows (21.6 ± 6.7 kg) and non-suckled cows (24.3 ± 4.5 kg), with no differences between separated and non-suckled cows. Once reunited, latency to make nose contact and duration of licking did not differ between treatments. We conclude that motivation for calf contact is greater for cows that are suckled.

Animals that experience situations likely to induce negative emotions show changes in judgment associated with pessimism. Few studies have focused on whether animals express stable differences in pessimism and whether these differences are related to personality traits. The first aim of this study was to explore if dairy calves are consistent over time in making judgments under ambiguous situations. Our second aim was to determine whether individual differences in judgment bias are related to conventional personality traits assessed using four standardized tests (Open field, Novel object, Human reactivity and Social motivation test). We subjected animals to two sessions of judgment bias and personality trait tests at 25 and 50 d of age. Individual differences in judgment bias were consistent over time with some animals persistently making more pessimistic choices compared to others. Two main dimensions of personality (Fearfulness and Sociability), obtained through principal component analysis, were also highly consistent over time. Pessimism was related to fearfulness, with more fearful calves making more pessimistic judgments. We conclude that dairy calves differ in the way they perceive and react to ambiguity and that this relates to individual differences in fearfulness.

The assessment of individual traits requires that tests are reliable (i.e. consistency over time) and externally valid, meaning that they predict future responses in similar contexts (i.e. convergent validity) but do not predict responses to unrelated situations (i.e. discriminant validity). The aim of this study was to determine if dairy calf personality traits (Fearfulness, Sociability and Pessimism), derived from behaviours expressed in standardized tests, predict individuals’ responses in related situations. The first experiment tested if the trait ‘Sociability’ was related to the expression of social behaviour in the home-pen, with calves assigned individual proximity scores (based on proximity to other calves) while they were in their home-pen at approximately 113 and 118 d of age. The second experiment aimed at exploring whether traits ‘Fearfulness’ and ‘Pessimism’ were related to the calves’ emotional response to transportation. All calves were subjected to two 10-min transportation challenges done on two consecutive days. Emotional response was assessed using the maximum eye temperature (measured using infrared thermography) and the number of vocalizations emitted. Social proximity scores (Experiment 1), vocalizations emitted and maximum eye temperature after loading (Experiment 2) were consistent over time. In addition, the results showed good convergent validity with calves scoring higher in Sociability also having higher proximity scores in the home-pen, and animals scoring higher in Fearfulness and Pessimism showing a more intense emotional response to transportation. The results also showed good discriminant validity, as neither Fearfulness nor Pessimism were associated with the expression of social behaviours (Experiment 1) and Sociability was not associated with the animal’s emotional response to transportation (Experiment 2). We conclude that the methodology used to measure personality traits shows good reliability and external validity.

Judgment bias tests (JBTs) use responses to ambiguous stimuli to infer emotional states in animals. However, with repeated testing, animals can learn to recognize the previously ambiguous stimuli rendering the test less effective. We describe a novel approach to this problem. Calves (n = 9) were trained in a spatial discrimination task to associate five locations with a specific probability of reward/punishment (Positive: 100%/0%; Near-Positive: 75%/25%; Middle: 50%/50%; Near-Negative: 25%/75%; Negative: 0%/100%). As predicted, calves showed increased latencies to touch locations that had higher probabilities of punishment and lower probabilities of reward. To validate our methodology for detecting mood changes, we followed calves in the hours after routine hot-iron disbudding, a time when animals were likely experiencing post-operative inflammatory pain. At 6 h after disbudding, when inflammatory pain was likely to peak, calves expressed increased approach latencies to the Positive, Near-Positive and Middle locations. These results suggest that calves perceived the value of the reward as being lower (i.e., anhedonia) or had lower expectations of positive outcomes (i.e., pessimism). When re-tested at 22 and 70 h after disbudding, we found no evidence of pessimism or anhedonia (i.e., latencies had returned to baseline). We conclude that our probability-based judgment bias task can detect pain-induced mood changes.