Simulating water use and N response of winter wheat in the irrigated floodplains of Northwest Uzbekistan

The crop-soil simulation model CropSyst was used to simulate growth, water- and N-uptakes of irrigated winter wheat (Triticum aestivum L cv. Kupava) in Khorezm, in the dry lands of northwest Uzbekistan, Central Asia. CropSyst was calibrated using the findings of field experiments of 2005/06 and 2006/07 and validated for the 2007/08 season. A relative root mean squared error of 11% proved the accuracy between simulated and observed aboveground biomass and grain yield in 2007/08. Scenario analyses showed that N-leaching was high and ranged from 63 to 106 kg ha(-1) when irrigated between 749 and 869 mm during the first two cropping seasons. The simulated N-leaching was lowest and ranged from 7 to 15 kg ha(-1) when irrigation was only 148-395 mm during 2007/08. The considerable N losses during leaching and high N-uptakes by wheat together resulted in a negative N-balance even during applications of 180 and 240 kg ha(-1) of N-fertilizer. N scarcity in the N-balance was reduced with increasing N-fertilizer amounts and ranged from 29 to 153 kg N ha(-1) in 2005/06 and 2006/07. Despite a common shallow groundwater table in the region during some time of the year, scenario analysis revealed that only full irrigation water (580 mm) and N supply according to crop demand (180 kg ha(-1)) guaranteed high grain yields, unless the water table is permanently shallow to overcome irrigation deficits. Limited irrigation and N application (40% and 55% of 'optimal', respectively) in combination with a groundwater table below 3 m resulted in a 55% yield decline. The CropSyst wheat model proved a robust tool for assessing the influence of water and N dynamics under conditions of varying irrigation and shallow groundwater tables. It thus has potential as a decision support not only in northwest Uzbekistan, but also in comparable regions of Central Asia. (C) 2010 Elsevier B.V. All rights reserved.