Mutualisms underpin much of the complexity of the natural world, in interactions as diverse as coral bleaching, nitrogen fixation in root nodules, and the human gut microbiomes. Understanding how interspecific mutualisms (cooperation between different species) can be stable over many generations is a key unresolved subject within evolutionary biology.Fungus-farming attine ants are an excellent model system to address this question experimentally as they form complex associations with antibiotic-producing actinobacteria (from the genus Pseudonocardia) that they grow on their cuticle. The extent to which these associations represent stable adaptive remedies against a single fungal pathogen, or ongoing evolutionary ‘arms races’ between bacterial biofilms and multiple pathogens of fungal crops and/or ant farmers has been subject of recent debate.Capitalising on recent discoveries (and a technological advance) in the Copenhagen Centre for Social Evolution, this proposal aims to make a significant contribution towards generally resolving this controversy. We will test the hypothesis that newly hatched ants always start their adult lives with a “native” vertically acquired actinomycete strain but obtain more complex biofilm cultures after they begin foraging outside the nest. I will use an experimental approach to manipulate the nature of the association between the ants and bacteria, to test specific questions about coadaptation, partner fidelity, and specificity in pathogen defence – something that is relatively uniquely possible with this system. In doing so, the proposal will also bring together the best possible expert mentors in Europe, both in the host group and as external collaborators. If this idea were confirmed, it may be the first example of a mutualism that changes function during host development. These results will also contribute to a broader understanding of how mutualistic associations more generally can be maintained over time.