Cognitive and Composite Behavioural Welfare Assessments of Pet Cats between the Ages of 9-22 Months, Living in Single and Multi-Cat Households.

Abstract

Although agonistic interactions between cats are often regarded clinically as a source of stress, there is currently limited research evidence regarding the welfare impact of keeping multiple cats as pets. The aim of this study was to compare welfare indicators between cats living in domestic single and multi-cat households, as well as between multi-cat households where agonistic behaviour was/was not reported by owners. Indicators included a spatial judgment bias task (JBT), where longer latencies to ambiguous probes are interpreted as being related to a more 'pessimistic' mood state, and the cat stress score (CSS), where high scores are indicative of high stress levels. Of 128 focal cats between the ages of 9-22 months, 94 were from multi-cat households, 126 had useable CSS data and 42 had JBT results suitable for analysis. CSSs were significantly lower for cats showing a more 'pessimistic' response in the JBT. It is possible that the cats that appeared to be the most relaxed may have been showing inactivity relating to negative affective states and/or were the least active/food motivated, and therefore slower in the JBT. CSSs were significantly higher in cats from single compared with multi-cat households, and did not vary with reports of agonistic interactions in multi-cat households. JBT results did not vary depending on the presence of, or reports of agonistic behaviours between, cohabiting cats. These data suggest that cats from single-cat households may be more likely to show signs of acute stress than those in multi-cat households. Alternative explanations are possible. For example, lower CSSs in the multi-cat group may reflect 'relief' effects resulting from separating cats for the test period, or inactivity relating to negative affective states. Due to the narrow sample population and broad scope of husbandry conditions, the potential for confounding variables limits the degree by which results can be used to inform causation of the relationships identified. Further research is warranted to replicate this work and explore potential confounders.