Assessing the potential for evolution in the insect baculovirus Lymantria dispar nucleopolyhedrovirus

A challenge to understanding infectious disease ecology is that processes occurring at multiple scales likely impact the ecological and evolutionary patterns found in nature. Both pathogen growth within hosts and pathogen transmission between hosts are likely important to pathogen dynamics, but studies typically focus on only one of these scales or the other. An integrated approach to disease modeling may thus provide novel insights. The challenge to using an integrated approach is that within-host dynamics are particularly poorly studied, because of difficulties associated with directly tracking pathogen dynamics within hosts. Here we demonstrate a method by which pathogen growth within hosts can be inferred through fitting mechanistic birth-death models to easily collectable dose-response data, alleviating the need to directly track pathogen dynamics over time. We then use this method to test various models of baculovirus growth in gypsy moth (Lymantria dispar) hosts, to identify the biological mechanisms important to pathogen growth in this system. This analysis yields novel insights into the ecology and evolution of the gypsy moth and its baculovirus with implications for strategies to control gypsy moth population outbreaks. Furthermore, this analysis reveals that host differences alone are insufficient to explain the observed variability in host outcomes, and that demographic stochasticity in pathogen growth is likely important in explaining this variability. The importance of demographic stochasticity, in turn, suggests that genetic drift may occur during pathogen growth within hosts. We next combine our birth-death model of baculovirus growth within hosts with a stochastic SEIR disease model that describes transmission dynamics between hosts for this insect baculovirus, and we use this nested model to explore the importance of genetic drift on pathogen diversity within hosts. Using reasonable parameter values, our model predicts that genetic diversity within hosts will be strongly affected by the drift that occurs both during pathogen population bottlenecks at transmission, and during the subsequent population growth of pathogen within hosts. We next compare the levels of pathogen diversity within hosts predicted by the models to Illumina sequence data from 223 field-collected hosts. This analysis shows that genetic drift is indeed important to explaining the observed patterns of pathogen diversity. We take this as evidence that genetic drift occurring within hosts is important to understanding patterns in disease ecology, suggesting that stochasticities that act both within hosts and between hosts need to be accounted for to understand pathogen evolution in general. We then discuss the implications of these findings for evolution of virulence theory and host-pathogen interactions.