Print PageWidespread, drought- and infestation-induced tree mortality is emerging as global ecological response to a changing climate. Rapid loss of canopy cover associated with tree drought mortality can rapidly alter ecosystem structure and function, potentially affecting feedbacks to the earth system through changes in carbon, water, and energy budgets. Therefore predicting forest response to a changing climate is a pressing challenge in global change ecology, yet a mechanistic, physiological understanding of drought-induced tree mortality is lacking. Two non-exclusive hypotheses have been proposed, hydraulic failure and carbon starvation. Hydraulic failure would occur when weak transpiration regulation during drought causes complete xylem cavitation that forestalls water transport. Carbon starvation could occur if trees cease stomatal conductance during drought, foregoing photosynthesis and relying on mobile carbohydrates for respiration until these are depleted, or cannot be moved among tissues. A previous experiment demonstrated that for piñon pine, a strong stomatal regulator, carbon exchange dynamics through mortality were consistent with carbon starvation, yet no published study has yet directly quantified mobile carbohydrates through drought-induced death. Here we present an analysis of foliar, non-structural carbohydrates from this glasshouse experiment, where simulated drought killed trees under ambient and warmer (~+4°C) temperatures, to further elucidate the mortality mechanism.
Results/Conclusions
Drought trees had substantially less foliar total non-structural carbohydrates (NSC) than watered controls, consistent with near-zero net photosynthesis reached early in the experiment. When normalized by watered control means, foliar NSC declined during drought, but was not entirely depleted prior to tree mortality. Analysis of NSC components revealed that starch content drove NSC dynamics and soluble sugar content varied little for both drought and control trees. Although trees in the warmer drought treatment cumulatively respired CO2 faster than ambient trees, there was no temperature effect on declines in NSC or component trends. Although our data are limited to foliar tissue, these results seem consistent with carbon starvation occurring via mobilization and/or transportation failure, but not with complete reserve exhaustion. Therefore, the severe hydraulic stress on the xylem observed during the experiment may have inhibited phloem function, thus preventing foliar NSC from meeting respiratory demands of stem or root tissue. Ongoing field and growth chamber drought experiments will provide further insight into whole plant dynamics and NSC trends under realistic conditions. Notably, our results add support that two proposed mechanisms for drought-induced tree mortality are interrelated, indicating that tree death is a complex physiological process.
