Evaluation of Water Use Efficiency Algorithms for Flux Variance Similarity­-Based Evapotranspiration Partitioning in C3 and C4 Grain Crops

Abstract:

Flux Variance Similarity (FVS)-based evapotranspiration (ET) partitioning method is gaining popularity due to its significant advantages of logistical and spatial representativeness over other ET partitioning methods. The FVS partitioning requires estimating leaf-level water use efficiency (WUE) if WUE measurements are not available. The WUE can be estimated from eddy covariance (EC) data for carbon dioxide and water vapor, plus (usually) a parameterized model for the intercellular carbon dioxide concentration (c_i). Despite the high sensitivity of FVS partitioning to WUE estimations, relatively little work has been done comparing c_i models and WUE algorithms. We tested four c_i parameterizations (Constant ppm, Constant ratio, Linear, and Square root) and one optimized WUE approach. All models were evaluated using high frequency EC data during peak growth for three rainfed C₃ grain crops (wheat - Triticum aestivum L., canola - Brassica napus L., and soybean - Glycine max L.) and two irrigated C₄ grain crops (maize - Zea mays L. and grain sorghum - Sorghum bicolor L. Moench). Although Linear and Sqrt models produced more results for soybean, Constant ppm and Constant ratio models consistently resulted in large numbers of successful half-hour partitioned fluxes for all other crops. Differences in number of successful partitioned outputs occurred mainly due to incompatible WUE estimates. Large ranges in ratios of transpiration to ET (i.e., 0.06 in soybean, 0.10-0.12 in wheat, canola, and maize, and 0.22 in grain sorghum) were observed among the models. This study offers insights into performances of different WUE models for the FVS partitioning of ET in different grain crops.