Measuring personality traits in Eurasian red squirrels: A critical comparison of different methods

Abstract

AbstractAnimal personality, behavioural differences among individuals which are consistent through time and contexts, is generally described by one or more traits. Different methods are used to measure these traits, such as behavioural observations and trapping indices. Comparing several methods allows to validate different tests and to better identify which aspect of an animal's personality is being measured. Here, we measured activity, exploration and immobility of Eurasian red squirrels (Sciurus vulgaris) using observations from the open field test (OFT), and compared them with trappability and trap‐diversity indices calculated from capture–mark–recapture data in six independent study sites. Trappability measures the willingness to enter a baited trap (boldness), while trap‐diversity can be a proxy for exploration tendency. Our first aim was to test their repeatability, thus their appropriateness as candidate measures of personality traits. Next, we explored the predictions that (i) trappability, measuring boldness, does not correlate with any of the OFT personality traits, since risk‐taking cannot be expressed in our arena test; and (ii) trap‐diversity correlates with exploration obtained from the OFT. Considering the species ecology and study design, we used multivariate Bayesian models based on different time intervals (session/year) and habitat types (alpine/plain). We found significant repeatabilities for trappability, for trap‐diversity and for most OFT personality traits, except exploration in the plain sites. Independently from habitat type or time interval, trappability did not correlate with either activity, exploration or immobility from OFT, thus measuring a different personality axis, which we considered a proxy for boldness. Trap‐diversity did not correlate with exploration from OFT, but seems related to a different aspect of red squirrels' exploration tendency. Our study emphasizes that caution is needed when using behaviours from OFT to measure multiple personality traits and that applying extra methods based on animals’ responses to trapping can provide reliable proxies for boldness and exploration tendency.