Conflict in insect societies
The Paradigm Shift
Social insect colonies may attain a remarkable level of integration. With respect to foraging behaviour for example, ant and bee colonies may often seem to act like real superorganisms. Since the 1960s however, people started to become increasingly aware that social insect colonies are in fact often full of conflicts and conspiracies. This stems from the fact that insect colonies are made up of genetically non-identical individuals, each transmitting genes in different ways, and hence each having distinct reproductive interests. In social insect research this culminated into a paradigm shift from cooperation to conflict as a major topic of study. In this laboratory, study of conflict over queen replacement in Diacamma ants was one of the factors that paved the way to much current research on the evolution of conflict in social groups. Among the first follow-up studies was one by B. Gobin on the ponerine ant Gnamptogenys menadensis (a collaboration with C. Peeters from Jussieu). Rather peculiar for this species was that most reproduction is performed by mated workers ('gamergates'), and not by queens as usual. One result of his work was that when a colony loses it's gamergate, dominance interactions among the colony members determine who will mate and succeed in heading the new colony (see fig. below).
Hierarchies of hopeful reproductives. In Gnamptogenys menadensis, a queenless ant from Indonesia, dominance rank determines who will mate and reproduce and who will not. All ants are colour marked so that their aggressive behaviour can be recorded individually.
Conflict in Cells and Colonies
Parallel to the discovery of conflict in social insects was an increased notion of conflict within individual organisms. Such 'intragenomic conflict', it turned out, could take on various forms, ranging from sterilisation of all male offspring, diabetes during pregancy or elimination of half of all offspring. The reason for the presence of these conflicts was obvious : individual organisms are composed of genomic subunits that are transmitted in different ways, hence their conflicting reproductive interests. Mitochondria, for example, are transmitted through the maternal line only, whereas nuclear genes are transmitted through male and female offspring equally. This is why mitochondrial genes in many plant species can get fitness benefits from sterilising the male function - it frees resources that can then be reinvested into the female function, resulting in a transmission advantage for the maternally inherited gene. Similarly, the nuclear genome is composed of a mix of genes derived from either the mother or the father, and each type of gene may have distinct interests. For example, in polyandrous species male-derived genes would favour higher investment in embryonic growth than that which would benefit maternally derived genes. Male-derived (paternally imprinted) genes therefore often cause diabetes during pregnancy. In the germ-line finally, alleles were found that override the normal Mendelian inheritance and achieve a higher than 50% chance of ending up in the gametes. Such meiotic drive alleles usually kill all the gametes that do not contain the allele, resulting in the production of just half as much sperm as usual.
Conflicts in social insect colonies occur for very similar reasons. Workers and queens for example transmit their genes in different ways. While a queen transmits her genes equally efficiently through both sexes, the workers transmit their genes, as a result of the haplodiploid sex determining system, three times as efficiently through sister queens than through brothers. Hence, there exists a queen-worker conflict over optimal sex allocation, with the workers favouring a 3:1 female biased sex allocation, in contrast to the equal investment favoured by the queen.
The research of T. Wenseleers aims to compare the evolution of conflict within organisms and social insect colonies, so as to reveal common principles. Special attention goes to mechanisms that favour collective group interests over selfish individual interests (e.g. social policing in the honey bee suppressing selfish worker laying and recombination suppressing meiotic drive). At a less fundamental level, he is also trying to document the occurrence of sex ratio distorting pathogens in ants. For this he has been looking mainly at the bacterium Wolbachia (see photograph below), a close relative of our own mitochondria, but one that has taken a manipulative rather than a mutualistic route. It is a maternally transmitted bacterium that occurs in the ovaries of various arthropods where it causes several types of reproductive alterations. In parasitoid wasps for example, the bacterium makes its host reproduce asexually, resulting in all-female broods. Effects of sex ratio distortion also occur through other mechanisms like feminisation of genetic males or selective killing of male offspring, and in most insect species the bacteria spread by sterilising competing uninfected females (cytoplasmic incompatibility). Using a PCR-based assay he has recently shown that Wolbachia occurs abundantly in ants, with up to 50% of all species being infected (Wenseleers et al. 1998).
The enemies within. Wolbachia bacteria (B) and mitochondria (m) in the egg cytoplasm of the ant Gnamptogenys menadensis. Being maternally transmitted, Wolbachia bacteria have developed several strategies to make their host produce broods of mostly female offspring. The specific effect of Wolbachia on ant reproduction is currently being investigated.
Conflict over Male Production
If parent-offspring conflict over sex allocation is one major focus in the study of conflict in insect societies, queen-worker conflict over male production is another. This type of conflict originates from the fact that every individual worker is always more highly related to its own offspring than to queen produced males, whereas the queen in turn is always more highly related to her own offspring than to the worker's offspring. Not only can this create a conflict between the queens and the workers; when the queen is multiply mated or when several queens co-exist in the nest, there will also be a conflict among the workers themselves over who will reproduce. Theory predicts that workers are always expected to try to lay their own eggs, but that they will be counterselected to raise other worker's eggs in colonies where worker-worker relatedness is low (low relatedness can result from multiple maternity or paternity). Such inhibition of each other's reproduction can act as a social policing mechanism, maintaining social cohesion even when relatedness among colony members is low. So far, it has only been extensively documented in honey bees. With the advent of new DNA techniques (see fig. below) theories on the evolution of worker reproduction have just recently become amenable to empirical tests.
Microsatellite genotyping. Amplification of tandemly repeated direpeats allows one to determine the kin structure of Formica ant colonies and maternity of field collected male eggs or adults.
Copyright 1999, Katholieke Universiteit Leuven
Information provider: Katholieke Universiteit Leuven
Comments for the authors: Diane Allard
Page design: Tom Wenseleers. Last update: 20/12/04