Modelling the within-host growth of viral infections in insects.

Insects are infected by a variety of pathogens, including bacteria, fungi and viruses, which have been studied largely for their potential as biocontrol agents, but are also important in insect conservation (biodiversity) and as model systems for other diseases. Whilst the dynamics of host-pathogen interactions are well-studied at the population level, less attention has been paid to the critical within-host infection stage. Here, the reproductive rate of the pathogen is largely determined by how it exploits the host; the resources supplied by the host in terms of size and condition; competition with other pathogens; and the speed with which it kills the host (death being an inevitable outcome for obligate-killing pathogens). In this paper we aim to build upon recent developments in the literature by conducting single infection bioassays to obtain data on growth and fitness parameters for phenotypically different and similar strains of nucleopolyhedroviruses in the Lepdipoteran host Spodoptera exigua. Using these data, a simple mechanistic mathematical model (a coupled system of differential equations) is derived, fitted and parameter sensitivity predictions are made which support empirical findings. We unexpectedly found that initial growth of virus within the host occurs at a double-exponential rate, which contrasts with empirical findings for vertebrate host-pathogen systems. Moreover, these infection rates differ between strains, which has significant implications for the evolution of virulence and strain coexistence in the field, which are still relative unknowns. Furthermore, our model predicts that, counter to intuition, increased viral doses may lead to a decrease in viral yield, which is supported by other studies. We explain the mechanism for this phenomenon and discuss its implications for insect host-pathogen ecology.