Geographic variation in the repeatability of a personality trait

Animal personalities are interindividual behavioral differences that are consistent across time or contexts. Increasing research is revealing the adaptive significance of personalities, although the mechanisms driving this variation remain largely unknown. A possible source of variation in personality traits is interpopulational differences in the strength of selection acting upon them. The response to selection can be measured indirectly via the behavior's repeatability, as repeatability generally sets an upper limit to trait heritability. However, no information currently exists on geographic variation in personality repeatability. We therefore quantified repeatability in exploratory behavior, a common personality trait, over multiple trials for 3 populations of rufous-collared sparrow (Zonotrichia capensis), focusing on 3 specific measures (exploration speed, diversity of perches visited, and number of hops). We also asked how differences in repeatability of these 3 measures affect other aspects of exploration, such as the temporal consistency of intercorrelations between the measures. Exploration speed was highly repeatable across all populations, whereas diversity was only repeatable in 2 of 3 populations and hopping behavior not at all. These differences in repeatability lead to temporal variation in the correlation matrices of the 3 exploration measures. Finally, only trial number influenced interindividual variability in exploration, whereas population identity, experimental conditions (i.e., conducting the novel environment assay under laboratory or field conditions), and time since capture all had no effect. Our findings highlight the complexity of using measures of behavioral consistency as a definition of personalities and emphasize the value of quantifying interpopulational patterns of trait repeatability.