ResearchBlogging.orgIN THE PAST week or so I’ve been writing about the attine ants, which have a complicated mutualistic network combining cultivated fungi and actinomycete bacteria, and are parasitized by Escovopsis fungi and perhaps black yeasts as well! Today I’m writing about the attine ants again, but along a very different angle. In this case this paper examines the influence of evolution upon reproductive behavior. I actually ran across the paper by accident, having previously planned to write about a paper on evolution and reproductive behavior in humans. This paper nicely transitions between these two themes.

Among organisms in general it is a bad idea evolutionarily to abandon breeding in favor of helping another individual raise its offspring. There are examples of social cheaters among groups as different as myxobacteria, slime molds, vertebrates, and insects. The myxobacteria and slime molds are bacteria and eukaryotes respectively that have converged upon a similar lifecycle. These are organisms capable of lone existence, but which mass together during unfavorable conditions to produce a stalk that launches spores. Ideally every strain in the group will be represented equally in the spores produced, but some cheater strains are able to produce more than their share of spores. Among vertebrates, there are examples in many groups of species that parasitize the nests of others, foisting off their young upon an unsuspecting individual. Cowbirds and cuckoos are well known among the birds, and cuckoos gave their name to the cuckoo catfish, which parasitizes cichlids. A similar parasitism is seen among insects where some wasp species will infiltrate another species nest and lay their eggs. The ants take this type of parasitism even further, with some species going to war against others to capture their larvae, which are raised in their captor’s colony and tend their brood. But the type of reproductive cheating occurring in attine ants is different from all of these!

The colonial insects have an unusual setup because a colony contains one breeding female, and her daughters do not breed, but devote their lives to caring for their mother’s eggs and larvae. In the attine ants the daughters are not clones. Ants are haplodiploid, that is, males are haploid and females are diploid. Among haplodiploid species sisters are 3/4 related instead of the usual 1/2 relationship. The father passes on an identical chromosome set to each of his daughters (since he only has one set, that’s his only option!), while the chromosome set inherited from the queen is only 1/2 shared. This may help to explain the evolution of social structures, since workers would be more closely related to their sisters (3/4) than they would be to their daughters (1/2)! However, the increase in relatedness towards sisters is balanced by a decrease in relatedness to brothers, and the end result appears to be no net benefit. Then according to the theory of evolution we should see the appearance of cheaters, ants attempting to contravene the normal reproductive process to increase the inheritance of their genes.

Attine ant castesThis has been discovered in the attine ant Acryomyrmex echinatior. These ants are haplodiploid, but the queen mates with more than one male, so the workers in the colony are 3/4 related to daughters from the same father, but only 1/4 related to daughters from a different father. Typically eggs develop in approximately equal proportions into small workers, large workers, or queens (all three seen in the picture to the right). This differentiation appears to be environmentally controlled.

The decrease in relatedness of lineages from different fathers occurring in the same nest and the environmental control of queen development present a situation ripe for exploitation by a cheater male. If a male inherits a mutation causing his daughters to develop preferentially as queens, he will achieve greater reproductive success.

Patrilineal worker caste skewA study of five colonies discovered in three colonies six paternal gene lines out of 30 that produced queens preferentially. Mapping the phenotype distribution of the offspring (small worker, large worker, or queen) for each genotype produced not a straight line, but a concave curve. This indicates that there are two mechanisms operating in different paternal lineages. In one, small workers and queens predominate, in the other large workers and queens predominate. Since development is environmentally controlled, mutations causing sensitivity to stimuli triggering larger size will produce large workers and queens preferentially, while mutations triggering early transition from large worker to queen will produce small workers and queens preferentially.

Mechanisms of royalty biasing

Since a colony is dependent upon its workers, a colony with too many cheaters will suffer a decrease in fitness. This applies selective force upon ants to detect cheaters, most likely due to an odd size distribution in developing larvae. The cheating patrilines found in the study were sufficiently rare that their size variance was likely lost in the background of normally maturing larvae. The large-worker biased patrilines had a size distribution very similar to non-cheating genotypes and produced queens at only a moderate excess. The small-worker biased patrilines produced a significant excess of queens, but also were rare in a colony, probably due to a much more abnormal size distribution of larvae. Among insects producing daughters that are not clones cheating is predicted to occur, but this will be matched by an evolution of control measures. There is probably widespread low-frequency cheating among the social insects.

Hughes, W.O., Boomsma, J.J. (2008). Genetic royal cheats in leaf-cutting ant societies. Proceedings of the National Academy of Sciences, 105(13), 5150-5153. DOI: 10.1073/pnas.0710262105