Project Title: Landscape-scale enhanced mountain pine beetle and climate change threat assessment
Principal Investigators: Barbara Bentz, USDA Forest Service, Rocky Mountain Research Station, Logan, UT; Jim Powell, Utah State University, Mathematics and Statistics Department & Biology Department, Logan, UT
Collaborators: C. Fettig, and S. Seybold, Western Bark Beetle Research Group and PSW; M. Hansen, A. Lynch, and J. Negron, Western Bark Beetle Research Group and RMRS; J. L. Hayes, R. Kelsey, J. Lundquist, Western Bark Beetle Research Group and PNW; J. Hicke, University of Idaho; J. Powell, Utah State University; J. Regniere, Canadian Forest Service
E-mail Contact: Barbara Bentz, bbentz[at]fs.fed.us
Summary: Mountain pine beetle (MPB) has caused tree mortality across millions of forested acres in western North America in recent years. Increasing temperature associated with climate change is a major factor influencing MPB outbreak dynamics. Because temperature is such an important driver of MPB population dynamics, the USFS, RMRS research project focused on bark beetles has been analyzing MPB response to temperature for 20+ years. Several data-based mechanistic models that describe MPB response to temperature have been developed, and two models are currently considered operational. A phenology model predicts lifestage-specific developmental timing and a cold tolerance model predicts probability of MPB larval mortality due to cold temperature. Both models are driven using hourly phloem temperature.
The cold tolerance model and a variant of the phenology model (the ‘adaptive seasonality model’) have been implemented within BioSim (Régnière and St. Amant 2007) to make landscape scale predictions. The adaptive seasonality model provides conservative predictions of climate suitability (Yes or No) based on two requirements: 1) univoltinism (e.g. a 1-year life cycle) and 2) emergence of a median individual during an appropriate window of time. This model has been very useful for making broad predictions about climatic suitability and MPB success. However, because only the ‘median individual’ is kept track of in the adaptive seasonality model, it does not incorporate important aspects of distributional timing of population emergence. The distributional timing and number of adults emerging at any given time is important to the mass attack strategy of this insect. Moreover, although univoltinsm is the optimum for population success, recent field research suggests that populations with at least some proportion semivoltinism (e.g. a 2-year life cycle) can also outbreak. What is needed is a flexible model that connects the number of adults emerging, and their timing, to population success.
Gilbert et al. (2004) describe a variant of the MPB phenology model that incorporates developmental variability within life stages, and model output includes the temperature-dependent distribution of adult emergence through time. This simulation model is currently coded in Matlab. Using this variant, we have made significant progress on a model that connects MPB phenology and emergence distribution with a demographic model of population growth rate. This model bridges the gap between phenology predictions and population viability/growth rates for MPB, and is based on phloem temperature and aerial over-flight data (e.g., FHP, ADS). The ultimate goal is improved landscape-scale outbreak probability predictions under climate change scenarios. Because this MPB phenology model is mechanistic, it can be directly evaluated using data that describe the inherent processes that are the basis for the model. We have evaluated the MPB phenology model and parameterized the ‘growth model’ using phloem temperature data from MPB-infested trees and aerial over-flight data of a MPB outbreak in central Idaho. A next step is to evaluate model performance in an additional area. With this model we can mechanistically translate a year-long temperature profile to adult emergence timing, and ultimately MPB population success (as identified by ADS information).
Our overall goal is to develop and evaluate an improved tool for predicting climate change effects on mountain pine beetle outbreak potential. This tool can then be applied more broadly to a wide array of projects funded by WWETAC, including ongoing research to retrospectively analyze interactions among climate, fire, and bark beetle outbreaks, and forest vegetation climate change projections. Combining our refined MPB model with downscaled temperature projections developed by the Pacific Northwest Research Station MAPSS team will provide managers with an important tool for making site specific climate change predictions of beetle suitability and associated tree mortality.
1. Collect field measurements of hourly phloem temperature within an MPB outbreak in Washington.
2. Use the field-collected temperature data, phenology model, aerial-overflight data of MPB impacts, and vegetation cover maps to evaluate and re-parameterize the MPB demographic model as needed.
3. Evaluate historic trends in outbreak dynamics using the parameterized model and associated historic temperatures and historic ADS information in Washington (previously compiled by Ager and Preisler).
4. Explore options for connecting our MPB demographic model with Dynamic General Vegetation Models and downscaled gridded temperature projections being developed by the MAPSS Team.
The main product will be updated model of MPB phenology and demographics for improved landscape-scale outbreak probability predictions using historical and climate change scenarios.
Gilbert E., Powell J.A., Logan J.A., & Bentz B.J. (2004) Comparison of three models predicting developmental milestones given environmental and individual variation. Bull Math Biol 66:1821–1850.
Régnière, J. & St-Amant, R. 2008. BioSIM 9 user’s manual. Information Report LAU-X-134. Quebec, Canada, Natural Resources Canada, Canadian Forest Service, Laurentian Forestry Centre.
Project ID: FY09AA64