Fungus-growing ants (Attini) feed off a fungus they cultivate in a mutualistic symbiosis in underground chambers by providing it substrate they collect outside the colony. The tribe of Attine ants ranges from small colonies of the paleo- and basal Attine species with a few hundred workers that forage on crude substrates such as insect frass and dry plant material, to large colonies of the leaf-cutting ants with several thousands to several million workers that provide live plant material to their fungus gardens. Leaf-cutting ants are the dominant herbivores of the Neo-tropics, and have a major contribution to cycling of nitrogen and phosphorus in their direct environment and are, furthermore, considered pest species as they have a large impact on human agriculture. These factors make leaf-cutting ants an ideal study subject to better understand the mechanisms that make this mutualistic symbiosis so successful.
To understand the evolutionary development of domestication of the fungus over the phylogeny of the Attine ants, I compared the average number of nuclei per cell for the fungal symbionts, for each of the different groups of fungus-growing ants. I found that the fungal symbionts of the paleo- and basal Attine ants, which have a relative low level of domestication, have two nuclei per cell, the standard for Basidiomycete fungi, but that the average number increased to 7-17 nuclei per cell in the highly domesticated fungi of the higher Attine ants and leaf-cutting ants. Furthermore, I was able to estimate that approximately half of these nuclei were represented by different genomes, giving the fungus a ploidy level of 5n-6n.
In mutualistic symbioses it is important the partners stay true to each other. In fungus-growing ants, new founding queens bring a piece of fungus to build up their new colony. However, in rare occasions fungal symbionts might come into contact with symbionts from other colonies. I showed that in both leaf-cutting ant genera incompatibility reactions between fungal strains can avoid intermixing of different strains, and that these reactions strengthen when genetic distance is increased. This pattern, however, becomes distorted when fungal symbionts are contested across ant genera.
The most important mechanism in the succession of this mutualism of leaf-cutting ants is the controlled degradation of plant material. I show that in the area of Gamboa, Panama, the two leaf-cutting ant genera forage for rather different plant material, with Atta species specializing on tree-leaves and Acromyrmex focusing more on flower material and herbal plant material. This difference is reflected in the overall enzyme activity patterns in the fungus gardens, with Atta specializing more on specific enzyme groups and Acromyrmex having an overall high enzyme activity.
Finally, I show that the fungal symbiont of the leaf-cutting ant Acromyrmex echinatior produces large amounts of biodegrading enzymes in special structures called gongylidia. The ants eat these structures, but enzymes pass the ant gut without being digested, and are excreted by the ants in their fecal fluid which they mix with freshly foraged plant material placed on the top of the fungus garden. The enzymes are still active and have therefore an important role in the biodegradation of the plant material. With this I show that the fungus evolved some incredible adaptations to a mutualistic life with the ants.