The shape of the contact–density function matters when modelling parasite transmission in fluctuating populations

Models of disease transmission in a population with changing densities must assume a relation between infectious contacts and density. Typically, a choice is made between a constant (frequency-dependence) and a linear (density-dependence) contact–density function, but it is becoming increasingly clear that intermediate, nonlinear functions are more realistic. It is currently not clear, however, what the exact consequences would be of different contact–density functions in fluctuating populations. By combining field data on rodent host (Mastomys natalensis) demography, experimentally derived contact–density data, and laboratory and field data on Morogoro virus infection dynamics, we explored the effects of different contact–density function shapes on transmission dynamics and invasion/persistence. While invasion and persistence were clearly affected by the shape of the function, the effects on outbreak characteristics such as infection prevalence and seroprevalence were less obvious. This means that it may be difficult to distinguish between the different shapes based on how well models fit to real data. The shape of the transmission–density function should therefore be chosen with care, and is ideally based on existing information such as a previously quantified contact– or transmission–density relationship or the underlying biology of the host species in relation to the infectious agent.