More than Moran: coupling statistical and simulation models to understand how defoliation spread and weather variation drive insect outbreak dynamics.

Understanding the processes that underlie species fluctuations is crucial to the development of efficient management strategies for outbreaks of destructive forest pests. Yet, the role of biotic and abiotic factors as well as their interactions in synchronizing outbreaks is not understood, despite many empirical and theoretical studies of species fluctuations. Here, we use a combined statistical-simulation model to investigate how defoliation spread and autocorrelated weather affect outbreaks of a major defoliator of North American boreal forest, the spruce budworm ( Choristoneura fumiferana (Clemens)). We modelled the regional dynamics of spruce budworm and based the model on data collected from spatiotemporal aerial surveys of defoliation from 1968-2015 in Quebec, Canada. The effects of weather on local forest stand defoliation and mortality transitions, along with defoliation spread probability and distance, were estimated statistically. Simulations were run with these estimates to identify the effects of spatiotemporal weather autocorrelation on synchronicity of outbreaks. Defoliation spread together with weather variables was found to best fit the observed outbreak size. Simulation models suggest that positive temporal autocorrelation in weather promotes outbreaks, indicating that a series of suitable years could encourage outbreaks. [ABSTRACT FROM AUTHOR]