Structural limitations of water and sugar transport in trees during drought: Model characterization of the relative importance of xylem embolism, carbon starvation, and lack of phloem transport

Background/Question/Methods

Increased tree mortality due to drought has been reported for several locations worldwide, but the actual physiological mechanisms for the mortality remain under debate, with different studies reaching strikingly different conclusions. The most likely candidates include hydraulic failure in the xylem transport system (i.e., run-away cavitation), carbon starvation (caused by an imbalance between photosynthesis rate and whole plant respiration rate), and failure of phloem transport (failure to transport carbohydrates to the locations where they are needed). In turn, carbon starvation can be caused by a range of processes including stomatal and non-stomatal restrictions to photosynthesis and reduced acclimation of respiration, compared to photosynthesis, to drought.

While several groups are actively engaged in the empirical determination of the relative contribution of all these processes (either with field observations or via experimental manipulations), it is also widely acknowledged that several of these processes are likely to interact with each other, confounding simple interpretations.

Here we present a modeling framework where all the major relevant processes are represented and in which one can ask the question of how properties of the transport systems (xylem and phloem), leaf gas exchange (photosynthesis and stomatal conductance) and external driving variables interact to affect the likelihood for each type of failure.

 Results/Conclusions

We identify five major forms of drought-related failure, i.e., a) xylem hydraulic failure, b) phloem transport failure caused by excessive viscosity, carbon starvation related to c) stomatal limitations or d) non-stomatal limitations to photosynthesis, d) imbalance between photosynthesis and respiration or e) a combination of reduced phloem transport and a lack of locally available carbohydrate for respiration.

Different failure modes are associated with particular set of species-specific functional traits (e.g., isohydricity vs anisohydricity) and particular sets of driving environmental variables.

Our steady state approach does not allow simulate the time required for tree mortality. However, it provides a reasonable estimate of the magnitudes of time involved. For hydraulic failure, the time required for the plant to run out of water is the total xylem volume divided by the transpiration rate (see Hölttä et al. 2009b). For carbon starvation, the time required is the amount of NSC before the drought divided by the respiration minus the photosynthesis rate.