ANALYSIS OF SPRUCE BUDWORM OUTBREAK CYCLES IN NEW BRUNSWICK, CANADA, SINCE 1952

Spruce budworm populations in New Brunswick have been surveyed an- nually since 1952 by sampling egg masses (later, overwintering larvae) as part of the insecticide application program. Although not designed for an ecological investigation, we extracted as much information from the survey data as we could with respect to several ecological issues. (1) All populations across the province tended to cycle in unison, although three major regions were distinguished by dissimilarity in peak and trough levels. We found that these regional distinctions were a result of random variation in the egg recruitment rate, rather than due to factors associated with comparatively fixed ecoregional (e.g., to- pographic, climatic, or forest type) characteristics. (2) We found, among all regions, a significant correlation in the rate at which eggs were recruited to each generation, thus providing evidence for the Moran effect as the mechanism underlying population synchrony that caused the province-wide outbreaks. (3) We discuss, with the aid of simulations, the nature and significance of random variations in the egg recruitment rate to explain observed differences in the spatial and temporal patterns of population cycles. Finally, we remark on problems in forecasting.