the legacy effects of keystone individuals on collective behaviour

royalsociety |  The ability to execute effective collective behaviour is vital for social groups. The coordinated gliding of fish schools when evading predators or the emergent nest structures of social insects represent collective adaptations that afford groups advantages that are not achievable for solitary individuals [1]. Such collective traits have captured the imagination of scientists including ecologists [2,3], behaviourists [4,5], mathematicians [6] and engineers [7,8], perhaps, more than anything else, because these collective traits are thought to emerge without central control [9]. In classic models of collective behaviour [10,11], individuals are treated as functionally equivalent. Yet, a cursory glance at any group reveals that, even among clones, no two individuals behave in precisely the same way [12]. Only recently have models of collective behaviour begun to predict how such behavioural variation is expected to impact collective outcomes (e.g. [13,14]). We consider here an extreme case of how individual variation can impact collective behaviour, where the behavioural traits of just one or a few highly influential individuals shape the behaviour of entire societies.

We define individuals that exhibit a disproportionate large influence over collective behaviour as keystone individuals (or just ‘keystones’). Though one may reason that keystone individuals might be a relatively rare phenomenon, a recent literature review on the topic identified more than 80 case studies where just one or a few highly influential group members shape group behaviour and success [15]. Movement leaders [16–19], knowledgeable tutors [20–22], hyperaggressive males [23], catalytic individuals [24,25] and disease superspreaders [26] represent just some of the ways in which keystone phenotypes can influence group function.

The reliance of a group on one influential individual may reduce its robustness to perturbation. For instance, groups' ideal collective phenotype could be compromised if their keystone individuals leave or die [27]. Such groups might also be more susceptible to manipulation by predators or pathogens because the keystone may serve as a fulcrum by which an exploitative agent can manipulate the entire group [28,29]. A possible solution to this potential system fragility would be that keystone individuals impose long-lasting changes on the behaviour of other group members which, in turn, could maintain their influence over the group's collective phenotypes even after their departure. Although catalytic effects by keystone individuals have been suggested for some social systems [24,25,30], there are few experimentally verified examples of these effects being long lasting (but see: [31,32]), and even fewer studies have determined what factors control the duration of these effects. Determining what controls the onset and duration of behavioural changes induced by keystone individuals is important for understanding the robustness of collective systems.