Powers, Jennifer Sarah. 1995. Spatial and temporal dynamics of the Douglas-fir bark beetle (Dendroctonus pseudotsugae, Hopk.) in the Detroit Ranger District, Oregon: a landscape ecology perspective. Corvallis, OR: Oregon State University. 107 p. M.S. thesis.
I developed a conceptual model of Douglas-fir bark beetle dynamics and associatedhost mortality across spatial and temporal scales. I proposed that a hierarchy offactors influence host resistance to attack at different spatial scales. I then tested thismodel by measuring the association between the occurrence of beetle-kill and differentfactors that might render trees more susceptible to attack at different spatial andtemporal scales. At a stand level, there was not a strong relationship between death bythe beetle and resistance to attack measured by tree growth rate prior to attack.Furthermore, tree growth rates were influenced by stand basal area but not by othersite level factors such as aspect, elevation, total annual solar radiation, or potentialevapotranspiration. At the landscape scale there were strong associations between theoccurrence of beetle-kill and i) portions of the landscape that were potentially drierand received more solar radiation, and ii) portions of the landscape that had moremature and old-growth conifer vegetation. The distribution of patches of beetle-killedtrees with respect to other patches was aggregated at the scales of approximately 1 and4 kilometers. There was no significant relationship between the occurrence of beetle-kill and linear landscape features such as road and stream networks and vegetationboundaries. Statistical analysis of the long term records of beetle-kill showed thatbeetle epidemics were related to the occurrence of windstorms and drought at certainspatial scales. The original conceptual model assumed that the strength of therelationships between the presence of beetle-kill and explanatory variables across all
spatial and temporal scales was similar. These results showed that landscape scalephenomena and temporal patterns were more strongly correlated with beetle-kill eventsthan physiological interactions at the scale of individual trees.