LAUSR

Facilities housing captive animals are full of staff who, every day, interact with the animals under their care. The expertise and familiarity of staff can be used to monitor animal welfare by means of questionnaires. It was the goal of our study to examine the association between chimpanzee (Pan troglodytes) welfare, happiness, and personality. To these ends we collected two waves of welfare and subjective well-being ratings of 18 chimpanzees housed at the Edinburgh Zoo and one set of ratings of 13 chimpanzees housed at Fundacio Mona. Ratings were made on a welfare questionnaire that included 12 items related to stress, psychological stimulation, and behavioural indicators of negative and positive welfare states, and a 4-item subjective well-being questionnaire. In addition, ratings were made on the 54-item Hominoid Personality Questionnaire and an abbreviated version of this scale consisting of 37 antonym pairs. We used generalizability theory to test whether welfare ratings generalized across items, raters, chimpanzees, and time. We then assessed the validity of the welfare and subjective well-being questionnaires by examining their associations with behaviour. Finally, we tested whether the welfare and subjective well-being ratings were associated with personality. Welfare ratings generalized across items, raters, chimpanzees, and time. Principal components analysis and regularized exploratory factor analysis indicated that ten welfare items and all four subjective well-being items formed a single dimension (welfareSWB). LASSO regression found that lower welfareSWB was associated with regurgitation, coprophagy, urophagy, and decreased proximity to nearest neighbour. A linear model that adjusted for age, sex, and facility, indicated that higher Extraversion and lower Neuroticism were related to higher welfareSWB. Welfare ratings were reliable and associated with subjective well-being and personality, demonstrating that staff ratings are a valid and potentially valuable tool for chimpanzee welfare assessment.