Mycorrhizae are the most common symbiosis on Earth. Around 90% of all terrestrial plants form associations with mycorrhizal fungi living in the soil. There they form intimate bonds, with fungal hyphae living within the plant roots themselves. Both parties benefit from the partnership through resource exchange, defence against pathogens, and drought tolerance. However, although the existence of trade between plant and fungal partners is well established, we do not yet understand mycorrhizal associations well enough to be able to predict the diverse outcomes observed. Controversy remains around how much control each party has over the exchange, with some pointing to the existence of parasitism by both partners as evidence of imbalance.
My DPhil project builds on traditions from behavioural ecology and modern multi-agent reinforcement learning techniques to explore what constitutes good strategies for mycorrhizal partners when facing their complex and unpredictable world. By building functions of the symbiosis into a model of mycorrhizae, we can explore how and why both fungus and plant make decisions. In doing so, we might better explain outcomes under various environmental conditions. Such knowledge could allow us to formulate targeted interventions across applications, from forestry and agriculture to nature restoration, to produce desirable outcomes.
