Forest carbon cycling along an elevation gradient: The influence of species and climate.

摘要

Global climate change will affect both ecosystem structure and function, which, in turn, may generate carbon cycle feedbacks to further climate change. Over the long-term, warming may result in a redistribution of species on the landscape, such that climate and species composition together affect ecosystem carbon dynamics. Over the short-term, changes in temperature and moisture availability may affect carbon uptake and release, although the climate sensitivity of these processes may depend on forest species composition.; I used an elevation gradient to investigate how climate and species composition interact to affect long-term, equilibrium carbon stocks and cycling in a Rocky Mountain subalpine forest. Total ecosystem carbon stocks decreased by 20 +/- 7 Mg ha-1°C-1 along the gradient, indicating that, with warming and an associated shift from Engelmann spruce and subalpine fir to predominantly lodgepole pine, these forests will yield a positive long-term feedback to climate change. These losses would occur because of declines in coarse woody debris and in organic and mineral-soil carbon. Relative aboveground net primary productivity increased with temperature, largely due to the shift from spruce and fir to pine. Further, tree species responded differentially to climate, with pine productivity increasing under warmer, drier conditions while spruce showed no response.; I used manipulations to investigate the short-term sensitivity of decomposition to climate and species. The short-term climate sensitivity of leaf litter turnover (6--11 years) was species-specific: spruce and fir needle litter decomposed more rapidly in winter, and generally under cooler, wetter conditions, while pine litter was not sensitive to seasonal or manipulated climate. The size of the labile soil carbon fraction was well predicted by a combined function of temperature and moisture, while the turnover time of this fraction was not. The microclimate sensitivity of the size of the labile carbon fraction was variable, depending on the species composition of the forest from which the soil was derived. Overall, my results indicate that climate and species interact to control most forest carbon processes. By including these interactions in climate-ecosystem models, scientists can improve predictions for terrestrial carbon cycle feedbacks and future climate change.