Quantifying the effects of phenology on ecosystem evapotranspiration in planted grassland mesocosms using EcoCELL technology

Use of plant phenological variables in models predicting evapotranspiration (ET) has largely relied on relatively simple (e.g., linear) relationships which may not be sufficiently accurate to predict small-yet ecologically significant-changes in plant phenology that are expected to occur in response to global climate change. A dearth of experimental data reflects the difficulties in quantifying these relationships against the background of large environmental variability that occurs in the field. Our main objective was to quantify how plant phenology (leaf area index [LAI] and root length density [RLD]) affect ET and its components during an entire vegetation cycle in large-scale model grassland (Bromus tectorum) ecosystems using the Ecologically Controlled Enclosed Lysimeter Laboratory (EcoCELL)-a unique open flow and mass balance laboratory. We also aimed to compare the three methods employed by the EcoCELL laboratory to measure ecosystem ET (whole-ecosystem gas exchange, weighing lysimetry, and weighing lysimetry combined with time domain reflectometry [TDR]) in order to independently confirm the performance of the unique gas exchange technology. Cumulative ET during the 190 days of the experiment measured with the three different methods compared very well with each other (mean errors <1%). We found that ET reached maximum levels at relatively low LAI (2-3), but as LAI increased beyond this value, small increase in transpiration were more than offset by decreases in soil evaporation, thereby causing declines in ET. A combined rectangular hyperbola (effects on transpiration) and linear (effects on soil evaporation) function between LAI and ET accounted for almost 90% of all variability in measured daily ET. RLD showed relationships to ET similar to those observed for LAI due to high covariance between RLD and LAI, but root length densities did not explain any additional variability in daily ET beyond that explained by LAI under the well-watered conditions of the experiment. Taken together, our results show that: (i) the EcoCELL mesocosm laboratory can precisely and accurately quantify hydrologic processes of large soil-plant monoliths under controlled environmental conditions; (ii) plant canopy phenological changes affect ecosystem ET, and the contribution of transpiration, in non-linear ways; (iii) these non-linear responses must be accounted for when assessing the consequences of changes in plant phenology-e.g., due to global environmental change-on ecosystem hydrology.