A MODEL FOR TESTING HYPOTHESES OF GYPSY-MOTH, LYMANTRIA-DISPAR L, POPULATION-DYNAMICS

A model for simulating long-term gypsy moth (Lymantria dispar L.) popu lation dynamics in North America has been developed. Simulated ecologi cal processes include larval dispersal, foliage consumption on differe nt host trees, reproduction, and mortality due to natural enemies (vir us, 4 guilds of parasitoids, and 2 guilds of predators). Population dy namics in several forest stands can be simulated taking into account t he migration of gypsy moth and its natural enemies. The model fits wel l to the following observations and data: (1) quasi-periodic outbreaks at approximately 10-year intervals, (2) life-tables in different outb reak phases, (3) increased parasitism in artificially augmented gypsy moth populations, and (4) phase plots of population dynamics. Simulate d survival curves are intermediate between those reported by J.S. Elki nton et al. and by J.R. Gould et al. It is shown that low-density gyps y moth populations can be stabilized by immigration from high-density areas without any density-dependent local processes. The model support s the following hypotheses about gypsy moth population dynamics: (1) d ensity fluctuations of small mammal predators is the most probable syn chronization factor in gypsy moth populations; (2) the outbreak freque ncy depends on the proportion of susceptible tree species and the dens ity of small mammal predators; (3) bacterial insecticide can be applie d less frequently for gypsy moth control than can a chemical insectici de of the same killing power because the bacterial insecticide spares invertebrate natural enemies; (4) the success of gypsy moth eradicatio n programs depends on the initial gypsy moth population density and on the type of functional response of predators. An alternative form of population bimodality hypothesis is suggested, according to which ther e are two types of gypsy moth population dynamics: eruptive in suscept ible stands and stable in resistant stands.