Studying animal innovation at the individual level: A ratings-based assessment in capuchin monkeys (Sapajus [Cebus] sp.)

Abstract

Large-scale studies of individual differences in animal innovation are rare firstly because discovery behaviour itself is often rare, and secondly because of logistical difficulties associated with obtaining observational data on a large number of innovative individuals across multiple groups and locations. Here we take a different approach, using observer ratings to study innovative behaviour in 127 brown capuchin monkeys (Sapajus [Cebus] sp.) from 15 social groups and 7 facilities. Capuchins were reliably rated by 1 to 7 raters (mean 3.2 ± 1.6 raters/monkey) on a 7-point Likert scale for levels of innovative behaviour, task motivation, sociality, and dominance. In a subsample, we demonstrate these ratings are valid: rated innovation predicted performance on a learning task, rated motivation predicted participation in the task, rated dominance predicted social rank based on win/loss aggressive outcomes, and rated sociality predicted the time that monkeys spent in close proximity to others. Across all 127 capuchins, individuals that were rated as being more innovated were significantly younger, more social, and more motivated to engage in tasks. Sociality, task motivation, and age all had independent effects on innovativeness, whereas sex, dominance and group size were non-significant. Our findings are consistent with long-term behavioural observations of innovation in wild white-faced capuchins. Observer ratings may therefore be a valid tool for studies of animal innovation, and our findings highlight in particular several possible scenarios through which innovative behaviour might be selected for among capuchins.