Abstract
CONTEXT: An increasing concern in the face of human expansion throughout natural habitats is whether animal populations can respond adaptively when confronted with challenges like environmental change and novelty. Behavioural flexibility is an important factor in estimating the adaptive potential of both individuals and populations, and predicting the degree to which they can cope with change. STUDY DESIGN: This study on the three-spined stickleback (Gasterosteus aculeatus) is an empiric illustration of the degree of behavioural variation that can emerge between semi-natural systems within only a single generation. Wild-caught adult sticklebacks (P, N = 400) were randomly distributed in equal densities over 20 standardized semi-natural environments (ponds), and one year later offspring (F1, N = 652) were presented with repeated behavioural assays. Individuals were challenged to reach a food source through a novel transparent obstacle, during which exploration, activity, foraging, sociability and wall-biting behaviours were recorded through video observation. We found that coping responses of individuals from the first generation to this unfamiliar foraging challenge were related to even relatively small, naturally diversified variation in developmental environment. All measured behaviours were correlated with each other. Especially exploration, sociability and wall-biting were found to differ significantly between ponds. These differences could not be explained by stickleback density or the turbidity of the water. FINDINGS: Our findings show that a) differences in early-life environment appear to affect stickleback feeding behaviour later in life; b) this is the case even when the environmental differences are only small, within natural parameters and diversified gradually; and c) effects are present despite semi-natural conditions that fluctuate during the year. Therefore, in behaviourally plastic animals like the stickleback, the adaptive response to human-induced habitat disturbance may occur rapidly (within one generation) and vary strongly based on the system's (starting) conditions. This has important implications for the variability in animal behaviour, which may be much larger than expected from studying laboratory systems, as well as for the validity of predictions of population responses to change.