Evolution of self-organized division of labor in social insects

PhD ceremony: Ms. A. Leitão Fernandes Duarte, 16.15 uur, Academiegebouw, Broerstraat 5, Groningen

Dissertation: Evolution of self-organized division of labor in social insects

Promotor(s): prof. F.J Weissing, prof. I.R. Pen

Faculty: Mathematics and Natural Sciences

In social insect colonies as found in bees and wasps, thousands of individuals coordinate efficiently to perform the different tasks related to colony growth and survival, with different groups of individuals taking over different tasks. This is achieved in a self-organized manner, without any apparent form of centralized command. Individuals are thought to obey simple behavioral rules and react only to local cues and interactions. In her thesis Ana Fernandes Duarte used computer simulations to study the evolution of behavioral rules that can potentially lead to division of labor.

One of her main findings is that when there is strong pressure for individuals to maximize the amount of work performed, they tend to become task generalists, willing to perform any task available. However, when switching tasks incurs a cost, then selection favors a divergence of behavioral phenotypes, leading to task specialization.

Fernandes Duarte also found that the type of behavioral architecture matters for the outcome of evolution. When behavior is under fully genetic determination, division of labor occurs only in groups in which individuals possess specific combinations of genes. Against expectations, she found that the presence of multiple mating under these conditions did not facilitate the evolution of division of labor. When the behavioral architecture also allows for the influence of experience in task choice, specialization is much more widespread in the population.

The work of Fernandes Duarte is among the first to combine the perspectives of self-organization and evolution, and to follow the evolutionary trajectories of behavioral rules for division of labor. Her work highlights that combining mechanistic and evolutionary explanations of behavior can significantly alter theoretical predictions.