Abstract

Three popular approaches exist for quantifying personality in animals: behavioral coding in unconstrained and experimental settings and trait assessment. Both behavioral coding in an unconstrained setting and trait assessment aim to identify an overview of personality structure by reducing the behavioral repertoire of a species into broad personality dimensions, whereas experimental assays quantify personality as reactive tendencies to particular stimuli. Criticisms of these methods include that they generate personality dimensions with low levels of cross-study or cross-species comparability (behavioral coding in unconstrained and experimental settings) or that the personality dimensions generated are not ecologically valid, i.e. not reflecting naturally occurring behavior (trait assessment and experimental assays). Which method is best for comparative research is currently debated and there is presently a paucity of personality research conducted in wild subjects. In our study, all three described methods are used to quantify personality in a wild animal subject, the Barbary macaque (Macaca sylvanus). Our results show that the structures generated by unconstrained behavioral coding and trait assessment were not equivalent. Personality dimensions derived from both trait assessments and experimental assays demonstrated low levels of ecological validity, with very limited correlation with behaviors observed in non-manipulated circumstances. Our results reflect the methodological differences between these quantification methods. Based on these findings and the practical considerations of wild animal research, we suggest future comparative studies of quantification methods within similar methodological frameworks to best identify methods viable for future comparisons of personality structures in wild animals.