| This research sought to predict the potential short- and long-term effects of insect defoliation and disturbance interactions on above-ground forest carbon dynamics in eastern North American forests. I integrated forest inventory data with remote sensing analyses to help determine the spatial processes that drive outbreaks of hardwood defoliators, and implemented these processes in a landscape model to simulate long-term changes in aboveground carbon that accompany invasion of persistent defoliator populations. My empirical analyses with MODIS data and an insect phenology model demonstrated that synchrony between the leaf-out phenology of host tree populations and the timing of insect emergence is a significant driver of defoliation disturbance. Spatial linear models to confirmed that phenological mismatch was a significant predictor of defoliation, in addition to measures of host abundance, topography, suppression, and spatial adjacency. Synthesis of the literature on tree growth and mortality responses to defoliation showed that tree species responses depend on wood growth anatomy, phenology, and nonstructural carbohydrate storage. Landscape simulations in the LANDIS-II model revealed that disturbance scenarios that include both native and introduced defoliators produce the most pronounced changes in aboveground forest carbon, particularly when periodic outbreaks occasionally synchronize or interact with forest management. |
