THE ATTACK DYNAMICS AND ECOSYSTEM CONSEQUENCES OF STEM BORER HERBIVORY ON SITKA WILLOW AT MOUNT ST. HELENS

Abstract

Variation in plant quality across space and time is considered a driving force behind the heterogeneous distribution of herbivorous insects on their host plants. At the same time, herbivory itself can mediate ecosystem processes that can cause feedbacks directly affecting plant quality. Here I examine both of these processes in a primary successional system to ask how insect herbivory can shape successional outcomes. I performed a three year observational study to determine which host plant factors - stress, vigor, and sex - were associated with insect herbivory by the poplar willow weevil (<italic>Cryptorynchus lapathi</italic>) on Sitka willow (<italic>Salix sitchensis</italic>), a dioecious pioneer shrub recolonizing Mount St. Helens after the 1980 eruption. I found that weevils prefer or perform best on vigorously growing willows that are seasonally water stressed. This result highlights the need to integrate hypotheses regarding insect response to stress and vigor into a single phenologically based framework focusing on nutrient mobilization to early insect herbivore life stages. I performed a field experiment involving leaf litter from stems attacked and not by weevils to determine whether weevils mediate nutrient cycling by altering willow leaf litter quality or resources available in its root environment. I found that although weevils do not consume leaves directly, stem herbivory is associated with a large reduction in leaf phosphorus, which in turn decelerates phosphorus cycling on Mount St. Helens. Lastly, I performed observational and experimental studies to show that the large female bias seen in willow on Mount St. Helens is not caused by weevil herbivory or other late acting ecological factors, but likely result from biased seed sex ratios. Taken together, these results suggest that weevil herbivory is retarding willow colonization in upland areas on Mount St. Helens, possibly allowing for alternative successional trajectories.