95th ESA Annual Meeting (August 1 -- 6, 2010)

OOS 43-6 - Merging evidence to quantify climate impacts on large competing trees

Thursday, August 5, 2010: 9:50 AM
401-402, David L Lawrence Convention Center
James Clark, Nicholas School of the Environment, Duke University, Durham, NC

Forest responses to climate change will depend on growth, fecundity, and mortality impacts in the context of competition.  Current models used to predict species responses, termed climate-envelope models, are controversial, because i) calibration and prediction are based on correlations between species abundance and climate correlations in space, rather than responses to climate change over time, and ii) they omit competition.  To determine the relative importance of changes in time, variation in space, and competition for light, we applied a longitudinal analysis of 27,000 individual trees over 6 to 18 years subjected to experimental and natural variation in risk factors.  Sensitivities and climate and resource tracking identify which species are vulnerable and in what ways.  A hierarchical Bayes model that includes error in the underlying demographic processes and in data derived from multiple sources provides a unique opportunity to determine the uncertainty in responses of all demographic rates for all species.


Results show that response to changes in time differs from that predicted from correlations in space.  The most important impact is spring temperature on fecundity, and not growth or survival.  Of secondary importance is for growing season moisture.    However, the effect of competition on growth and mortality risk exceeds the effects of climate variation in space or time for most species.  Species vary widely in terms of the interactions between climate and competition.  For many species increased light availability amplifies the positive response to longer growing season and reduced drought.  This effect is most pronounced for juvenile trees.  The interaction between season length and summer drought was also most pronounced for juvenile trees but split between amplification and compensation.    Because sensitivities to changes in time and competition are more important than correlations in space, current models miss the most important effects.   By directly comparing sensitivity to climate in time and space, together with competition, the approach identifies which species are sensitive to climate change and why, including the heretofore overlooked impact on fecundity.