Yields of dryland crops in semiarid and arid zones are limited by precipitation, and so water content and placement are very important at each stage of development. Spring wheat (Triticum aestivum L.) grown in a wheat-fallow (WF) rotation system (1 crop in 2 years) generally occupies the greatest area in the Israeli dryland region, more than the continuous wheat (CW) rotation system. To identify the optimal crop management for dryland farming where annual precipitation is <250 mm, we compared the effects of no-tillage (NT) and conventional tillage (CT) on wheat growth and water use efficiency (WUE) in both the WF and the CW rotation systems, and on water storage in fallow (F) plots. During the 4-year period from 1994 to 1997, experiments Here conducted at Gilat Experimental Station, located in the south of Israel (average annual precipitation, 237 mm; soil type, sandy loam loess-Calcic Xerosol). In the fallow year, F-NT increased water infiltration and soil water content in comparison with F-CT. However, most of the water evaporated during the summer, especially from the upper soil layer (0-120 cm). During growth, uncultivated soil with straw mulch increased water content in the upper soil layer and also encouraged the development of a longer root system capable of utilizing deeper water. During 1995, similar grain yields were obtained with both NT and CT treatments, an average of 3.45 t ha(-1) for WF and 2.9 t ha(-1) for CW. In the last 2 drought gears (1996 and 1997), NT management increased yields by 62 to 67% for WF and by 18 to 75% for CW, relative to CT management. During the 2 years when water deficiency occurred during the grain-filling stage (1994 and 1997), NT management increased grain weight by 20% and test weight by 5 to 7%, in addition to the 70 to 200% increase in the total grain yield, relative to CT management. Crop yield and WUE can be increased in arid zones with annual precipitation of < 200 mm, through use of a wheat-fallow rotation system that is managed by NT.
