From individuals to groups and back: interaction between individual variation in behaviour and group performance

Stark differences in individual behavioural responses are a well-known feature of animal diversity. Even within a social group many distinct strategies coexist, and this variation has been recently found out to play a significant role in resource exploitation, social learning and various collective behaviours. How the entire group performs can therefore depend on various characteristics, all linked to its members’ behaviour and the relationships that connect them. While there are theorical analyses focusing on the consequences of systems where individual variation and group environment influence each other by interacting through feedbacks, most of the assumptions and the effects hypothesized by these models have rarely been experimentally studied in controlled conditions. My aim was thus to test if the interactions between variation in behavioural strategies and the social environment might have an effect on the performance of single individuals within the group and of the group itself. I approached the complex issue by performing a series of experiments on a captive population of House Sparrow (Passer domesticus). I started by examining the effect of predation on a potential proxy for life-history traits, i.e. relative telomere length, and the connection of the latter to various behavioural traits. In the second experiment I investigated if social connections between individuals within a group might have an influence on the measurable benefits obtained by its members. Expanding on this topic, I questioned if previous familiarity with a companion might be a factor strong enough to affect exploration of a novel environment, or if the presence of any conspecific would allow social facilitation. This investigation was also a necessary step to take before testing any group-related effect, as an attachment to one own’s group was a necessary prerequisite for the next experiment. In fact, I then assessed the performances of two flocks facing each other over limited resources. While there have been studies comparing groups’ performances, it has rarely been taken into account how two groups would interact together, even if in the natural environment groups sharing resources are quite common. For my last experiment I focused on one of the most well-studied dichotomous behavioural strategy, i.e. the leader/follower dynamic. I decided to investigate this variable strategy not only during exploration but also in a different situation, one of the most crucial in the life of any animal: the attack of a predator and the split-second reactions to it. The first experiment showed no influence of predation on telomere dynamics: relative telomere length however changed with successive samples. In the second experiment results showed that social connections affected the rate of discovery of a novel food source, with individuals more closely connected to the first feeder foraging before the others. In the following experiment I discovered that averaging over familiarity and sex the presence of a companion strongly increased exploratory behaviour. Familiarity with the companion however had an influence on the social exploration of female sparrows: they explored faster and consumed more resources only when accompanied by a familiar individual. In the experiment where two groups of sparrows faced each other we found out a that group membership affected the outcome of the confrontation, as the group that foraged first ended up almost always consuming more of the limited resource. This meant that whoever shared the group with a risk-averse individual, one that foraged first at a novel food source, gained benefits regardless of their own behavioural traits. Finally, individuals that led movements during exploration were followers during a simulated attack and vice versa, showing that social positions in this species are context dependent. In conclusion, these experiments shed light on interactions between variation in behavioural strategies and the social environment. Our results underscored how various assumptions made by theoretical models on the potential role of complex feedbacks between individual traits and the performance of the entire group were indeed correct and testable in a controlled setting. In the future, in order to keep investigating how social animals trade off costs and benefits in crucial contexts such as novel environment exploration and predator attacks it will be imperative to account for the role of diversity within the social environment.