RT Journal Article
SR Electronic
T1 Using biological control data to understand host-pathogen dynamics
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 285080
DO 10.1101/285080
A1 Joseph R. Mihaljevic
A1 Karl M. Polivka
A1 Constance J. Mehmel
A1 Chentong Li
A1 Vanja Dukic
A1 Greg Dwyer
YR 2018
UL http://biorxiv.org/content/early/2018/03/19/285080.abstract
AB Mathematical models have provided important insights into infectious disease spread in animal populations, but are only rarely used in environmental management. Part of the problem may be that direct tests of the models have focused on long-term model predictions, whereas short-term epizootics (= epidemics in animals) can sometimes be more informative about mechanisms of disease transmission. To illustrate this point, we tested models of density-dependent disease transmission and host variation in infection risk using multiple, single-host-generation epizootics of a fatal baculovirus of the Douglas-fir tussock moth (Orgyia pseudotsugata). The tus-sock moth baculovirus causes epizootics naturally, but because of its narrow host range, it is used in biological control to mitigate defoliation, an approach known as “microbial control”. Using a combination of experiments, observational data, and nonlinear-fitting algorithms, we fit a set of stochastic epidemiological models to data from tussock moth microbial control programs. We then used formal model comparisons to show that there are strong effects of host and pathogen density and host variation on epizootic severity, and that transmission events at small scales help explain epizootic severity at large scales. When host density is high, even very low initial virus densities can lead to high cumulative infection rates, suggesting that, in some cases, baculovirus sprays may have been applied to populations that would have collapsed from natural epizootics, even without intervention. Our study illustrates the usefulness of linking epidemiological modeling and nonlinear fitting routines in understanding animal diseases, and shows that simple models can provide important guidance for microbial control programs.