Behaviour 2019
Group efficiency: decoupling division of labor and worker expertise
Nicole Dykstra1,2, Mélinda Pozzi3, Simon Garnier1,2. 1Rutgers University, Newark, NJ, United States; 2New Jersey Institute of Technology, Newark, NJ, United States; 3University of Neuchâtel, Neuchâtel, , Switzerland

The success of a social insect colony depends on its members correctly accomplishing a multitude of tasks such as foraging, colony defense and brood care. This is often achieved by dividing labor among workers, with specific individuals performing the same task almost exclusively. However, most studies conflate division of labor – specific workers performing specific tasks – with worker expertise – performing the task better than one’s peers. That confusion limits our ability to determine the evolutionary relationships between division of labor and expertise, and their respective impact on the collective output of the colony. We used a simulation approach to decouple the two concepts and found that, when tasks are assigned according to a worker's expertise, large colonies benefited from expert workers, while smaller colonies benefited from generalist workers, especially as task diversity increased. When tasks are assigned randomly, however, both large and small colonies achieved better performance with generalist workers. Our results suggest that without a mechanism to match task assignments with a worker’s expertise, division of labor is not always beneficial to the colony.