The mix matters: behavioural types and group dynamics in water striders

The effect of an individual's behavioral type (e.g., its boldness or aggressiveness) on fitness likely depends on the environmental context. In many species, an important component of an individual's environment is its social environment - the mix of individuals in its social group. Accordingly, much of game theory assumes that the mix of behavioral types (BTs) in a social group influences individual fitness and group dynamics. Few experimental studies, however, have directly investigated how the group's mix of BTs influences individual and group outcomes. Here we manipulated social group composition in the water strider Aquarius remigis and evaluated the effects of our manipulation on individual behavior and various group outcomes (overall group activity, aggression, mating success). We formed 12 groups that differed substantially in average male BT (activity and aggression level), each with a low variance in BT. That is, one group had only the most active and aggressive males, a second group was made up of the next most active and aggressive males, and on down to a group of all very inactive and unaggressive males. All groups also had females. We found that, on average, groups made up of more active-aggressive males continued to be more active than other groups, but that contrary to predictions, these groups did not tend to enjoy higher mating success. Instead, a major factor affecting group mating activity was the presence of hyper-aggressive males. Hyper-aggressive males drove females out of the group and thereby decreased the group's overall mating activity. We discuss these findings in terms of their importance to the study of behavioral plasticity in social groups and the potential role of keystone individuals in determining group dynamics.