Human behaviour, the flow of information and the spread and control of crop disease

While spread of plant disease clearly depends strongly on biological factors controlling transmission, epidemics in agricultural crops very often also have a human dimension. Farmers make decisions about which crops and/or varieties to grow and when and where to plant them, as well as deciding throughout the growing season whether and how aggressively to control disease (Thompson et al., 2018). Since these decisions are typically based on imperfect knowledge of disease pressures as well as the perceived effectiveness of control, the spread of information is a very important factor, together with individual farmer's responsiveness and level of risk-aversion.

This requires mathematical models to couple human behaviour and disease spread. This is somewhat well-studied for human disease, particularly in the context of vaccination. However, vaccination typically provides long-lasting protection, whereas crop farmers must make decisions both between and within each growing season. For plant disease there is also often a significant spatial aspect to disease spread, complex transmission via a range of pathways (e.g. aerially, on seed, by vectors, via tools, from epidemics in alternate or alternative hosts) and large amount of cryptic (i.e. undetected) infection. There is also often little concrete information about key parameters controlling disease spread and yield responses (Cunniffe et al., 2015).

This project will concentrate on building models coupling human behaviour and crop disease. The initial focus will be on extending an existing simple non-spatial model of maize lethal necrosis (Hilker et al., 2017) - a viral disease complex which is spreading fast and is a significant threat to food security in parts of sub-Saharan Africa - to include stochasticity, spatial spread, the flow of information through networks of growers and the two-way coupling between farmers' decisions and disease spread. However, since at its core this project depends on strategic decision making among self-interested entities, simpler models will also be developed and investigated using tools taken from game theory. Particularly important will be the theory of so-called "iterative games", which focuses on the case in which decisions are repeatedly made.