The social insects provide some of the most extraordinary and puzzling products of evolution. They live in colonies in which most individuals serve as workers, giving up their own reproduction in order to help one or a few queens. The degree of cooperation and integration that occurs in some species has led to their colonies being likened to organisms. Yet social insect colony-mates, unlike the cells of an organism, are not genetically identical, so it is much more difficult to explain their extreme self-sacrifice and integration.
Kin selection theory has successfully explained why a social insect might help instead of trying to reproduce alone, but a more difficult problem remains unsolved. Given that an individual decides to join a group, it should often be selected to prefer being a reproductive, to be helped rather than to be a helper. But if this is so, how is anarchy avoided, particularly in large colonies where the queen cannot hope to physically dominate all her subordinates? The most promising hypothesis appears to be collective worker control. Even if each worker might prefer to reproduce, under certain conditions they are predicted to suppress each other in favor of the queen. Though this hypothesis is theoretically appealing, it has not been widely tested. We have taken advantage of extensive past work on a species of social wasp, Parachartergus colobopterus, to test collective worker control hypotheses in three distinct contexts. First, do workers choose who lays the haploid male eggs? Second, do workers choose which of the numerous queens will survive to lay the eggs that will become reproductive females? Third, do workers choose the timing and the identity of new queens?
The methods employed are three. Behavioral observations, recorded on videotape, are made for colonies with individually-marked wasps. These focus on behaviors that show who is reproducing (egg laying) and behaviors that might influence who reproduces, (such as egg eating, feeding, and aggression). The observation colonies are collected and brought back to the lab for DNA microsatellite analysis. Microsatellites are genetic loci that are both numerous and highly variable. They are used in two ways. First, they allow precise estimates of relatedness that will be used to predict who workers should prefer to reproduce. Second, are used to assign maternity, that is, to determine who does reproduce, and how much. Together these three types of data determine what workers should prefer, whether their behavior is consistent with that preference, and whether that preference is actually achieved.