Mathematical Biology Seminar

James Powell
Departments of Mathematics & Statistics/Biology
Utah State University, Logan, Utah
Wednesday, Oct. 7, 2009
3:05pm in LCB 225

Abstract: Maintaining an appropriate seasonality is a basic ecological requirement for insects living in seasonal environments. Critical life history events (the timing of which is termed /phenology/) must mesh with seasonal cycles, and it is often selectively advantageous for individuals in the population to synchronize their activities with one another as well. In most terrestrial insects, some explicit physiological mechanism, such as diapause (hibernation which ends with a specific environmental cue, as when day-length exceeds a fixed duration), maintains both aspects of seasonality. However, many ecologically important insects, such as the mountain pine beetle (/Dendroctonus ponderosae/), apparently lack an explicit physiological timing mechanism. Seasonality of such insects is said to be under direct temperature control. How such insects maintain seasonality has been a mystery, since their physiological clocks move at a rate nonlinearly dependent on environmental temperatures.

In this talk, we first discuss the mechanistic basis for direct temperature control of seasonality in the mountain pine beetle, which is responsible for more forest damage across North America than all other disturbances together, including fire. This bark beetle attacks healthy pine trees; successful reproduction is contingent on host mortality. Pines under attack defend themselves strongly using toxic resin, which can repel a fixed number of attacking beetles. This creates strong selective pressure for populations of beetles to mature and emerge simultaneously (synchrony), and at an appropriate time of year (seasonality); synchrony of adult emergence is absolutely necessary for the mass-attack strategy that overcomes tree defenses.

We connect a distributional model describing mountain pine beetle phenology with a model of population success measured using annual growth rates derived from aerially detected counts of infested trees. This model bridges the gap between phenology predictions and population viability/growth rates for mountain pine beetle. The model is parameterized and compared with 10 years of data from a recent outbreak in central Idaho, and is driven using measured tree phloem temperatures from north and south bole aspects and cumulative forest area impacted. A model driven by observed south-side phloem temperatures and that includes a correction for forest area previously infested and killed is most predictive and generates realistic parameter values of mountain pine beetle fecundity and population growth. Extensions of this model to include spatial and demographic structure of host forest (both leading to outbreak ` are discussed.