Program
| MSC2c | |
|---|---|
| Greg Dwyer | |
| Department of Ecology & Evolution, University of Chicago | |
| Title | Models of pathogen-driven outbreaks that are motivated by experiments |
| Abstract | Many insects undergo outbreaks, in which their densities rise from levels that are undetectable, to levels at which massive defoliation occurs. Mathematical models have played a key role in identifying the mechanisms that drive outbreak cycles, but most models are constructed only to explain observational data. Testing models with experimental data as well can often reveal the importance of mechanisms that are not immediately apparent in observational data. Work in my lab therefore attempts to use experimental data to motivate the construction of models, which are then tested with observational data. We have used this approach to show that population structure has a strong effect on interactions between the gypsy moth (Lymantria dispar) and its nucleopolyhedrovirus. The nucleopolyhedrovirus is a fatal pathogen that is transmitted as host larvae feed on leaves, and it often causes intense epidemics in outbreaking populations. By allowing larvae to feed on virus-contaminated leaves in the field, we are able to directly test for factors that affect viral transmission rates. Our preliminary experiments showed that heterogeneity in susceptibility among individuals plays a key role in transmission, but inserting our estimates of this heterogeneity into models of pathogen-driven outbreaks leads to a stable point equilibrium in the models. This type of behavior is inconsistent with the cyclic outbreaks observed in gypsy moth populations. Further experiments, however, showed that susceptibility is heritable and appears to evolve, whereas standard models assume that average susceptibility is constant over time. We therefore extended existing models to allow natural selection to drive fluctuations in average susceptibility. In these evolutionary models, susceptibility declines immediately after outbreaks because of selection for resistance, but it rises during inter-outbreak periods because of a fecundity cost of resistance. The resulting models allow for realistic outbreaks even if heterogeneity in susceptibility is constant, and accurately predicts changes in susceptibility observed in the field. In ongoing work, we are further extending our models to allow for mechanisms that underlie susceptibility. In our current models, susceptibility is expressed as a per-capita risk of infection, but relating infection risk to variability among individuals is difficult. We are therefore developing models that relate individual host behaviors to a host's risk of infection, and thus to disease dynamics. In addition, we are developing models that incorporate the evolution of viral virulence, by keeping track of individual virus strains. Our ultimate goal is to combine these separate models into a combined model that relates coevolution at the level of individual hosts and pathogen particles to the long-term dynamics of insect outbreaks. |
| Location | Friedman 153 |