Transmission dynamics of a virus in a stage-structured insect population

Despite the blossoming interest in host-microparasite epidemiology, and in use of viruses in the biological control of insect pests, few empirical studies have attempted to quantify transmission and mortality rates under field conditions. We report a laboratory and field study in which the transmission parameter (v) and mortality rate (Ux) due to nuclear polyhedrosis virus (NPV) are measured in different larval instars of the cabbage moth, Mamestra brassicae (Lepidoptera: Noctuidae). Laboratory studies of food consumption and virus susceptibility were used to produce crude estimates of relative transmission rates in successive instars. Increases in the rate of feeding outstrip increases in virus resistance with instar, so we predict that transmission rates should increase in older larvae (assuming rate of intake of virus to be proportional to rate of feeding). This prediction was tested in a field experiment in which a constant initial density of susceptible and infected (moribund) larvae were reared together on cabbage plants for 2-8 d. Estimates of the linear transmission parameter (v) did not differ between instars and gave a mean value of 2.16 X 10-12 for all instars and time points. Causes for the discrepancy between predictions based on laboratory data and field measurements are discussed. Differences were found between instars in the time from infection to death (T) (equivalent to 1/cV, where ax is the rate of mortality due to viral infection). Second-instar larvae died more rapidly once infected than third instars, which, in turn, died more rapidly than fourth instars. The effect of host population stage structure on patterns of viral infection can be pronounced and if recognized, may significantly increase the accuracy and predictive value of models of host pathogen systems.