Transmission of the gypsy moth nuclear polyhedrosis virus: Theory and experiment

We used the nuclear polyhedrosis virus (LdNPV) of the gypsy moth, Lymantria dispar (Lepidoptera: Lymantriidae), to test a basic assumption of most models of disease dynamics; that the rate of horizontal transmission is directly proportional to the product of the densities of healthy larvae and virus. We made measurements of virus transmission, using small-scale experiments in bags on red oak (Quercus rubra) and black oak (Q. velutina) and observed a decline in the transmission constant as the densities of both healthy larvae and pathogen increased. We hypothesized two possibilities for the non-linearity observed in this system: (1) pathogen inhibition via the effects of induced foliage chemistry and (2) the effect of spatial heterogeneity of LdNPV (pathogen clumping effects). Previous work has linked larval damage on oak foliage to subsequently higher levels of tannins in damaged leaves; and laboratory bioassays have linked higher levels of tannins to a decrease in mortality caused by LdNPV. We damaged LdNPV-contaminated oak foliage using larvae, then bioassayed foliage with test larvae and measured tannins in damaged leaves. None of the experiments showed significant effects of foliage damage on mortality of test larvae or tannin content of damaged leaves, and we concluded that induced foliage responses cannot explain non-linearity of transmission in our previous work, or transmission dynamics in the field. To evaluate the role of pathogen clumping in transmission dynamics, we redefined parameters of the host-parasitoid model of Nicholson and Bailey (1935) for use in the gypsy moth-LdNPV system: a relationship between the assumptions of this model and the continuous time Anderson-May model was described. Using a version of this discrete-time model incorporating spatial heterogeneity, we observed transmission dynamics similar to those in field experiments. To evaluate safety issues in use of genetically altered baculoviruses as insecticides, we field-tested LdNPV that had been genetically-engineered for non- persistence. The EV was released in a forest setting, and did not persist or spread past the first year of release. Concerns regarding movement of baculoviruses were also addressed in experiments exploring the action of rainfall on the translocation of LdNPV