Detection of low levels of baculovirus for outbreak-terminating epizootics in defoliating insects

Recent empirical and theoretical studies have indicated that epizootics of baculoviruses in defoliating insects may result in the termination of outbreaks starting from lower initial infection rates than previously believed. This suggests that natural epizootics can preempt the need for costly and labor-intensive pest management measures because natural epizootics reduce or end outbreaks before substantial defoliation occurs. Such cases, however, require the ability to detect small amounts of infection at the beginning of the population’s life cycle. At hatching, Douglas-fir tussock moth (Orgyia pseudotsugata) larvae become infected by baculoviruses on the surface of their eggs. Prior to the larval season in an outbreak area, egg masses are collected and assayed for the presence of viruses; however, a large number of eggs may need to be sampled to detect low infection rates in which biocontrol measures are not needed. Here we used simulated sampling to detect infection rates ranging from 10^-3 to 10^-4 and considered whether multiple potential probability distributions of the actual infection rate affected detection. We showed that the level of sampling is considerably higher than in previously published protocols, but that increased effort via either more egg masses sampled (with a fixed number of eggs per egg mass), or more eggs per egg mass sampled without collecting more egg masses, equally improved the accuracy of detecting the virus.