Climate variability and the distribution of ponderosa pine

Background/Question/Methods

Bioclimatic modeling has been used extensively to understand the climatic factors controlling modern distributions and to predict future distributions of species under changing climates. These models are normally based on climatological values, with few taking into account any parameter related to the year-to-year variations in climate. We present here a study of ponderosa pine (Pinus ponderosa), an ecologically and economically important species in the western USA, for which an existing model (Norris et al., 2006) has been modified to take into account interannual climate variability.

Results/Conclusions

The initial model failed to accurately predict the distribution in the north-eastern part of the range, a region characterized by a highly variable winter temperature regime, with a regular occurrence of severe winter conditions. Including a measure of interannual climate variability in the bioclimatic model improved the simulated modern distribution by approximately 2.5% over the whole range, but by 25% in the north-east. To test the model outside of the modern calibration period, we first performed a hindcast of the location of tree populations during the Last Glacial Maximum, a period with important changes in both climate and variability. Based on output from a general circulation model (PMIP; http://pmip2.lsce.ipsl.fr/), the results show a restricted, southern distribution in general agreement with available paleo-data.

Finally, we explore potential changes in the distribution under IPCC scenarios for future global climate change. Sensitivity tests indicate that incorporation of climate variability has a strong effect on predicted ranges.  Many populations are predicted to remain in place when only mean climatology is applied, but are at risk when climate variability is taken into account. The results allow the identification of areas of potential disappearance and new establishment, as well as an assessment of migrational corridors, and emphasize the importance of accounting for changes in variability, especially for populations at distributional limits.

Norris, J.R., Jackson, S.T. and Betancourt, J.L. 2006. Classification tree and minimum-volume ellipsoid analyses of the distribution of ponderosa pine in the western USA. J. Biogeogr, 33, 342-360.