The few individuals grouped together in the third cluster move very little, at low frequency and stay hidden more, resulting in round trips of longer duration. The three behavioural clusters were identified using
clustering methods that take into account all behavioural variables under the assumption that each behavioural strategy shows a normal distribution in a population. The retrieved clusters were then cross-validated using an independent clustering method, which showed that the grouping was robust (Fig. 2). Interestingly, the grouping of individuals in either two or three clusters was equally robust with the reassignment of individuals showing the same two individuals being misclassified. The three individuals comprising the third cluster, although being classified as belonging to the second cluster if only two groups are predefined, show a distinct and unique behaviour compared with the entire sample characterized by extremely low levels of exploration. In JAK inhibitor review summary, three groups differing in the amount of exploration and the time to the onset of exploration were detected. van Oers et al. (2004) showed the importance of the latency of the first movement when investigating
avian exploration syndromes as this reflects the willingness of individuals to take risks. Our analysis demonstrated differences among the three clusters in the latency to the first movement. Indeed, animals in cluster one start moving earlier than individuals in group
two, with the maximal latency observed for individuals in cluster two. Generally, frogs moved close to the walls of the tank and did not explore the centre much. This behaviour involving exploration close to a physical structure such as a wall reduces visibility to predators and provides some shelter while exploring. Similar behaviours where animals disperse and explore using landscape elements have been demonstrated for other taxa (Baguette et al., 2013). Among our behavioural clusters, animals in cluster three showed significantly less movement away from the walls of the those cage compared with the animals in the other two clusters. Behavioural syndromes (bold vs. shy) are typically recovered in studies analysing exploration behaviour (Dingemanse & de Goede, 2004; Wilson & Godin, 2009). Bold individuals are defined, in this context, as those that show curiosity and a willingness to explore; they move a lot at high frequency and take risks by moving away from walls or other structures that provide shelter. On the opposite end, shy individuals stay hidden long, explore little and use landscape elements during exploration to avoid open space. When analysing the behavioural clusters discovered in our data, it becomes evident that animals in cluster one can be characterized as bold, those in cluster three as shy and those in cluster two as intermediate. Indeed, our data show a large group of male X.