An ant colony can be considered from multiple, equally valid perspectives: it is a coherent individual in itself, but is in turn composed of potentially millions of individual ants, which in turn are the collaborative efforts of thousands of genes. Selection acts simultaneously at each of these levels, and it is this hierarchical view of a colony that this thesis elaborates on, exploring the ant colony at successively lower levels of biological organisation.
The first chapter considers a colony-level adaptation: how an initially homogenous group of workers within a colony self-organize into a complex, organized workforce. We consider how simple threshold rules obeyed by each individual can lead to an organised division of labour among different castes, and show that the neuropeptide Tachykinin may play a role in organising colony defence by controlling a threshold for aggression in the sterile worker caste. Tachykinin did not appear to control changes in aggressive behaviour in the reproductive castes, however, suggesting that colony defence by workers and by queens may have separate evolutionary origins.
Chapter 2 considers the separate evolutionary interests of the attine ants and their symbiotic fungal cultivar, and tests the role of vertical inheritance in the maintenance of cooperation between the attine ants and their symbiotic fungus. Vertical inheritance is often heralded as the main driver of cooperation between the ants and their fungus, but by conducting a large scale field study on Acromyrmex leaf-cutting ants and their fungus, we demonstrate that this symbiosis is in fact characterized by horizontal transmission. We show that queens regularly steal replacement fungal cultivars from neighbouring nests when they have lost their own, implying that other factors than vertical transmission, namely lifelong commitment through fungal incompatibility mechanisms, must be driving cooperation.
Chapter 3 considers reproductive conflict among the unmated workers within the colony, testing for a role of relatedness among workers in the extent of selfish worker reproduction. In a typical colony, policing prevents worker egg-laying, however following queen death, policing ceases and potential conflict among workers over reproduction emerges. Through a combination of genetic analysis and dissections, we show that relatedness is an important factor in determining the degree of selfish worker reproduction within a colony—and that workers may in fact be able to assess their own individual relatedness to their nestmates. RNA-sequencing also showed that those mechanisms controlling altruistic worker sterility appear to be highly conserved across the social Hymenoptera.
Finally, the fourth chapter considers the potential for conflict among individual genes within each ant. The kinship theory of genomic imprinting highlights that maternally- and paternally-derived genes within each female individual in a social insect colony are often differentially related to colony-mates, and that these genes may therefore be in conflict over how altruistic or selfish they should be towards such relatives. Although imprinting conflict has been predicted to affect many aspects of social insect biology, testing for its presence has proven difficult because of the genetic complexity of a social insect colony, and because of the difficulties in creating reciprocal crosses. This final chapter develops a graphical model that can be used to search for genomic imprinting without the need to create reciprocal crosses, and tests this model in the leaf-cutting ants. Although none of the individual genes we considered appeared to be imprinted, this technique holds promise in testing imprinting predictions across the social insects.