Carbon dynamics during drought-induced mortality - in search of the missing sink

Background/Question/Methods

Plants carbon reserves are thought to be an energy buffer during periods of environmental extremes. In most studies, carbohydrates are considered the main reserve compounds mirroring the tree carbon balance. However, since the latter is usually unknown this assumption has not been verified.
Here we present results of an intensively monitored experimental manipulation of whole-tree carbon balance using reduced atmospheric [CO₂] (starvation) and drought. Net above-ground assimilation, belowground respiration, carbon storage pool size and allocation to plant compartments and to specific carbon pools (glucose, fructose, sucrose, starch, biomass) were assessed at high temporal resolution. We tested whether observed carbohydrate pools could be estimated from the measured carbon balance.

Results/Conclusions

For trees grown at high [CO₂] the model fairly predicted patterns of carbon storage across tree compartments and storage pools. Surprisingly, carbohydrate pools estimated from the carbon balance showed a positive deviation from observations, indicating the existence of a carbon pool not assessed in our study. When expressed relative to GPP, deviations were greatest in the carbon starvation treatments and were independent of water supply. However, absolute allocation to the unknown sink was relatively constant in needles and branches within a given drought treatment. This suggests that both hydration status and carbon availability control allocation to this sink, potentially involved in osmoregulation or nitrogen storage. Given the large proportion of GPP allocated to this sink during carbon starvation, it represents a non-negligible carbon expenditure and thus may be critical to survival during periods of physiological stress.