Programmed and flexible: long‐term Zugunruhe data highlight the many axes of variation in avian migratory behaviour

Studies of Zugunruhe – the ‘migratory restlessness’ behaviour of captive birds – have been integral to our understanding of animal migration, revealing an inherited propensity to migrate and an endogenous timing and navigation system. However, differences between Zugunruhe in captivity and migration in the wild call for more data, in particular on variation within and among taxa with diverse migration strategies. Here, we characterise Zugunruhe in a long‐term dataset of activity profiles from stonechats (genus Saxicola) with diverse migratory phenotypes (976 migration periods from 414 birds), using a flexible and consistent quantitative approach based on changepoint analysis. For east African, Austrian, Irish, and Siberian stonechats and hybrids, we report key inter‐population differences in the occurrence, timing, and intensity of Zugunruhe. In line with expectations, we found the highest Zugunruhe intensity in the longest‐distance migrants, more variable patterns in short‐distance migrants, and intermediate characteristics of hybrids relative to their parental groups. Inter‐population differences imply high evolutionary lability of Zugunruhe timing within a robustly structured annual cycle. However, counter to theory, Irish partial migrants showed no segregation between migrant and resident individuals, and previously reported nocturnal restlessness was confirmed for resident African stonechats. Further features of nocturnal restlessness that did not align with migratory behaviour of stonechats were juvenile nocturnal restlessness even prior to postjuvenile moult, and protandry in spring, although stonechats winter in heterosexual pairs. Importantly, Zugunruhe of all populations declined with age, and the intensity of an individual bird's Zugunruhe was correlated with activity levels during other parts of the annual cycle. Our results confirm endogenous, population‐specific migration programmes but also reveal apparent discrepancies between Zugunruhe and migration in the wild. We thus highlight both the continued potential of Zugunruhe study and the need for circumspect interpretation when using migratory restlessness to make inferences about migration in the wild.