Mountain pine beetle (Dendroctonus ponderosae) populations alternate between epidemic and endemic phases. In their epidemic phase, they are one of the primary insect pests of pine forests in western North America, causing high host mortality over millions of hectares. The transitions between low-density endemic populations, which kill modest numbers of weakened, small trees to high-density epidemics causing up to 90% host mortality are influenced by many factors including climate, weather, host population characteristics, and other disturbances.
During the winter months, beetle larvae in host tree phloem are exposed to fluctuating temperature conditions. Low minimum overwinter temperatures, as well as rapid changes in temperature during critical periods contribute to high overwinter beetle mortality in many years. On the other hand, mild winters are favorable for larval survival and may help incipient epidemics overcome population thresholds needed for the development of epidemics.
We used spatial records of historical epidemics, daily weather records, and our spatial adaptation of the physiological overwinter survival model of Régnière and Bentz to describe the relationship between overwinter survival and population phase. We created an agent-based model, incorporating the physiological model, to explore differing scenarios in which overwinter survival predicts or fails to predict eruptive dynamics.
Results/Conclusions
Preliminary results of the spatial analysis indicate that in many locations, high overwinter survival was associated with the new and ongoing epidemics. For some winters, notably the winter of 2000/2001, the positive association persisted for time lags of up to several years. For some other winters, particularly winter 2005/2006, the relationship was not strong, and in some cases appeared negative for lag times of more than one year. Although in most winters, high survival was associated with increased beetle kill in many locations, in the majority of sampling locations there was no obvious relationship.
The lack of a universally positive relationship between high estimated winter survival and beetle epidemics is consistent with the hypothesis that weather is one of many factors shaping eruptive dynamics, and in isolation does not consistently predict epidemics. Results from the agent-based model support previous field and modeling studies indicating that other forest conditions must be favorable to beetles for mild winters to trigger a population eruption. In the agent-based model, high host density was particularly important for initiation of epidemics. This is consistent with field observations that thinning can buffer against beetle epidemic mortality.