Plant hydraulic controls over the susceptibility of trees to mortality following climate-enhanced disturbances

Background/Question/Methods

Climate-enhanced disturbances to forests, including drought and insect-induced mortality, are increasing globally. The bark beetle epidemic spanning western North America has affected multiple species of conifers, with or without co-occurring drought. Recent modeling efforts implicate both hydraulic failure and carbon starvation processes in tree death, but the role of combined effects of drought and insect-induced mortality modeling remain highly uncertain and understanding of physiological failure points is needed. The mechanism of mortality for bark beetles has been found to be xylem blocking by fungi. In this context, we asked whether incorporating the combined effects of drought and bark beetle attack, and hydraulic and physiological processes, improves predictions of tree mortality. We addressed this question using Terrestrial Regional Ecosystem Exchange Simulator (TREES) applied to pinon pines in New Mexico, and mid-elevation lodgepole pine and higher elevation Engelmann spruce and fir forests in Wyoming. TREES combines dynamic plant hydraulic conductance with canopy biochemical controls over photosynthesis, and the dynamics of structural and non-structural carbon (NSC) through a carbon budget that responds to plant hydraulic status and a simple phloem transport model with both leaf and root NSC pools. Bayesian model-data fusion was used to develop testable hypotheses on a multitude of responses.

Results/Conclusions

Sap flux, needle gas exchange and xylem vulnerability to cavitation all supported a hydraulic failure mechanism, defined on a continuum of percent loss of conductivity, of mortality. This was confirmed by Bayesian model-data fusion testing xylem dysfunction, stomatal and non-stomatal controls on photosynthesis and carbon allocation. Tree transpiration in all conifer species declined after climate-driven disturbance. For lodgepole pine and spruce infestation gas exchange estimates of maximum photosynthesis, day respiration and quantum yield were unchanged. The TREES model predicted a declining critical transpiration value for hydraulic failure that matched empirical measurements. For both lodgepole pine and spruce the model also predicted the contrasting dynamic of NSC increasing over the growing season for living trees and decreasing for dying trees. For pinon pines the combined effects of xylem occlusion by fungi and drought on hydraulic failure improved predictions of tree mortality over models that considered drought effects only. Our results show that models need to integrate hydraulic and physiological responses to drought and beetle-induced mortality. Our analyses also show that uncertainty regarding belowground drought response remains a large driver of variation in hydraulic failure and co-occurring physiological processes.