Landscape vegetation change, pattern detection, and interpretation in a subalpine fir forest infested with balsam woolly adelgid.

摘要

Invasive insects and pathogens, climate change, and anthropogenic factors are causing broad-scale change in forest ecosystems. Identifying, understanding, and maintaining historic and functioning ecosystems in changing landscapes can be challenging because multiple internal and external factors interact across scales. In this dissertation, I present a hierarchical approach to investigate patterns of health and mortality of subalpine fir (Abies lasiocarpa (Hook) Nutt. var. lasiocarpa) associated with a non-native herbivorous insect, the balsam woolly adelgid (BWA), Adelges piceae Ratzeburg, and landscape disturbance on the Olympic Peninsula, Washington, U.S.A. Assessment across broad and fine spatial and temporal scales allowed the identification of distinct boundaries of change associated with climate, topography, vegetation, and disturbance processes of fire, winter weather events, mass movement, and endemic and introduced insects and pathogens. I used Landsat imagery at broad spatial and temporal scales to differentiate patterns of very-high-, high-, and low-magnitude disturbance in subalpine-fir forest using the normalized burn ratio (NBR). Low-magnitude decline in NBR gradually increased in area over time (1994-2007) and occurred predominantly at high elevations and southern aspects during years of warm PDO index values, suggesting interactive effects of an invasive biotic agent and climate. I conducted field studies that identified BWA as the biotic agent that was consistently present in affected areas and associated with poor health and mortality of subalpine fir. Using dendrochronology, I determined that variability in the growth of subalpine fir was positively associated with the arrival of BWA in a forest along Hurricane Ridge in Olympic National Park. I compared aerial detection surveys, Landsat imagery, and dendrochronology methods and confirmed that all three methods independently detected effects of BWA for the same time period. These methods provided a more complete assessment that would not have been possible with one method alone. This dissertation demonstrates the utility of a hierarchical approach that combines remote sensing and field validation to detect patterns of disturbance and identify potential fine- and broad-scale causal mechanisms of forest change.