A process-based approach to estimate lodgepole pine (Pinus contorta Dougl.) distribution in the Pacific Northwest under climate change

Abstract

Listen

Translate article

Lodgepole pine (Pinus contorta Dougl.) is a widely distributed species in the Pacific Northwest of North America. The extent that the current distribution of this species may be altered under a changing climate is an important question for managers of wood supply as well as those interested in conservation of subalpine ecosystems. In this paper, we address the question, how much might the current range of the species shift under a changing climate? We first assessed the extent that suboptimal temperature, frost, drought, and humidity deficits affect photosynthesis and growth of the species across the Pacific Northwest with a process-based model (3-PG). We then entered the same set of climatic variables into a decision-tree model, which creates a suite of rules that differentially rank the variables, to provide a basis for predicting presence or absence of the species under current climatic conditions. The derived decision-tree model successfully predicted weighted presence and absence recorded on 12,660 field survey plots with an accuracy of ~70%. The analysis indicated that sites with significant spring frost, summer temperatures averaging <15°C and soils that fully recharged from snowmelt were most likely to support lodgepole pine. Based on these criteria, we projected climatic conditions through the twenty-first century as they might develop without additional efforts to reduce carbon emissions using the Canadian Climate Centre model (CGCM2). In the 30-year period centered around 2020, the area suitable for lodgepole pine in the Pacific Northwest was projected to be reduced only slightly (8%). Thereafter, however, the projected climatic conditions appear to progressively favor other species, so that by the last 30 years of twenty-first century, lodgepole pine could be nearly absent from much of its current range. We conclude that process-based models, because they are highly sensitive to seasonal variation in solar radiation, are well adapted to identify the importance of different climatic variables on photosynthesis and growth. These same variables, once indentified, and run through a decision-tree model, provide a reasonable approach to predict current and future patterns in a species’ distribution.