Freezing beetles: an improved model for predicting mountain pine beetle spread in western Canada

The mountain pine beetle, Dendroctonus ponderosae Hopk. is a major disturbance agent in western North America with more than 35 million acres of pine forests affected by an outbreak across British Columbia, Canada through 2008. The beetle is currently expanding its range into previously uninfested pine forests of Alberta causing concern of wider spread through jack pine corridors. The landscape-level spread of mountain pine beetle is mainly facilitated by favorable temperatures and availability of susceptible pine. The objective of this study was to refine a statistical model to provide inference on probability of an occurrence of a mountain pine beetle infestation in a given area using temperature variables based on the biology and phenology of the insect, spatial variables including neighborhood structures, and temporal variables including previous years’ infestations. While prolonged periods of -40ºC weather decrease beetle populations, associated cold-temperature signals have not been exhaustively studied in a landscape-level, statistical model despite the important implications to the effects of climate change. We found that the total number of “cold snap”s defined as periods of at least five continuous days of average winter temperatures below -20°C as well as the average duration of a cold snap during December were negatively associated with the response probability of finding an infestation.