Use of plant hydraulic theory to predict plant controls over mass and energy fluxes in response to changes in soils, elevation and mortality

Background/Question/Methods

A central challenge in predicting terrestrial ecosystem mass and energy fluxes is understanding how plant physiology interacts with environmental drivers. While an enormous amount of work has quantified the feedback between plant physiology and key drivers such as light, vapor pressure deficit and soil moisture, the resulting parameterizations in ecosystems models are often too complicated for appropriate predictive power. The emerging theory of plant hydraulics provides a parsimonious platform from which to generate testable hypotheses and predictions of ecosystem pools and fluxes. Plant hydraulic theory predicts that plant controls over carbon, water, energy and nutrient fluxes can be derived from the limitation of plant water transport from the soil through xylem and out of stomata. In addition, the limit to plant water transport can be predicted by combining plant structure (e.g. xylem diameters or root-to-shoot ratios) and plant function (response of stomatal conductance to vapor pressure deficit or root vulnerability to cavitation). We evaluate the predictions of the plant hydraulic theory by testing it against spatially varying transpiration in multiple ecosystems. These include boreal and northern hardwood forests experiencing spatial variability in soils and changes in species and environmental drivers across an elevation gradient (2700 to 3400 m) encompassing sagebrush steppe through subalpine forests. We further test the theory by predicting the carbon, water and nutrient exchanges from several coniferous trees in the same gradient that are dieing from xylem dysfunction caused by blue-stain fungi carried by bark beetles.

Results/Conclusions

Across all of the data sets which range in scale from individual plants to hillslopes, the data fit the predictions of plant hydraulic theory. Namely, there was a proportional tradeoff between the reference canopy stomatal conductance to water vapor and the sensitivity of that conductance to vapor pressure deficit that quantitatively fits the predictions of plant hydraulic theory. We incorporated these results into a next generation ecosystem–to-regional-scale coupled-biogeochemistry model TREES (Terrestrial Regional Ecosystem Exchange Simulator) using a Bayesian framework. Our model-data fusion approach successfully predicted ecosystem carbon, energy and nitrogen fluxes from a wide range of data including sap flux, leaf gas exchange, eddy covariance and stand biogeochemistry.