SCALING PLANT WATER USE FROM ORGANS TO ECOSYSTEMS IN SEMIARID SHRUB AND FOREST ECOSYSTEMS RESPONDING TO DROUGHT AND BARK BEETLES

Predictive understanding of semiarid ecosystem water use from remote sensing models will be greatly improved by cross-scale validation applying mechanistic measurements at the leaf to watershed levels. To address this validation, our questions are 1) what plant structure scalars work best for moving from organs to stand transpiration and 2) is the spatial patterning of transpiration predictable from plant structure and/or temporal drivers. To answer these questions we have focused on an elevation gradient from seasonally drought prone sagebrush dominated basins through bark beetle impacted montane and subalpine forests of Wyoming. In each of these ecosystems we have employed a combination of plant leaf gas exchange, sap flux and eddy covariance approaches to quantifying transpiration and evapotranspiration. In sagebrush ecosystems we have found that plant leaf area from shrub allometric relationships scales water use well in stands greater than 40 years old whereas younger stands require additional herbaceous plant estimates. In contrast, sapwood area is a better scalar than leaf area in the forested areas. This scalar becomes even more prominent with bark beetle infestation because sapwood changes from blue stain fungi occlusion of xylem (causing a 50% reduction in transpiration) are detectable months before leaf area changes. These scalars were further evaluated by showing that spatial autocorrelation in leaf area in the sagebrush stands was ca. 2 m and sapwood area in the forests was ca. 30 m. In the forest, spatial autocorrelation in transpiration linearly declined from 60 to 20 m as vapor pressure deficit increased from 0.5 to 2.0 kPa. This finding can be predicted from hydraulic transport mechanisms. In sagebrush ecosystems, soil moisture below 45 cm, driven by spring moisture, was the dominant driver of evapotranspiration and has been further verified with watering experiments. These spatial and temporal drivers of transpiration are being tested in the Terrestrial Regional Ecosystem Exchange Simulator (TREES) model. Scaling within stands is tested through stable isotopes of soil, plant tissue and water vapor. Scaling beyond the stand level is being evaluated through Remote Sensing platforms including Aerocam, Landsat and MODIS.