Plants exert major control over hydrologic budget- and, thus, over their own community stability- by their own transpiration and by their partial control over soil evaporation via soil shading. Furthermore, they exert major control over atmospheric humidity and consequent radiation balance in climate. Predicting further plant status and future climate will require predicting evapotranspiration (ET) over large regions and under varied climates. Models that are useful for diverse sites and species will require understanding of ET control at the process level, physiologically and micrometeorologically. ET prediction also presents significant challenges in quantifying vegetation and its spatial and physiological heterogeneity. We review how process-level understanding can be gained and tested by scaling down from satellite data and scaling up from leaf gas exchange. We have synthesized a model of leaf conductance and fitted its parameters to data on Larrea tridentata and Prosopis glandulosa. To scale up to whole-shrub and whole-stand evapotranspiration, we must sum over all leaves, accounting for (1) their different current microclimates and (2) their physiological states of acclimation to long-term histories of temperature and irradiance. We have developed models for each of these phenomena and are fitting them to data on many individual leaves. Our model can be integrated to predict whole-plant and whole-stand ET. We outline tests that will be performed on a 1000 m x 300 m transect of mixed shrubland on the Jornada Experimental Range.

Ogden, UT