Simulating seasonal soil water balance in contrasting semi-arid vegetation communities

Soil water and nutrient dynamics of a forest ecosystem process model have been modified to simulate productivity and seasonal water use patterns in Artemisia, Agropyron, and Bromus communities in south-central Washington, USA. Measured soil water profiles indicate that each of the three communities exerts substantially different influence on the hydrologic balance of the intermountain sagebrush-steppe, and each requires unique model parameterization to accurately simulate the natural balance between soil water availability and annual carbon accumulation. Yearly simulations predicted soil moisture content for each community for 56 dates in 1992 with R(2) values ranging from 0.93 to 0.98. Model relationships between soil water availability, transpiring leaf area, water use efficiency, and respiration costs of substantial below-ground productivity illustrate key ecophysiological considerations for arid land process modeling.