Many studies have shown that zero tillage (ZT) in combination with a surface crop residue layer - two components of conservation agriculture (CA) practice - can improve the agronomic water balance by increasing the amount of water that is readily plant available. However, no account has yet been published in which this effect had been fully quantified under rainfed semiarid Mediterranean conditions. To tackle the issue, in the 2009/2010 cropping season we studied the soil water dynamics of wheat grown after barley in northern Syria under two contrasting tillage regimes (zero tillage vs. conventional tillage, CT), two levels of surface residue retention (partial and full) and early and late planting. For a comprehensive quantification of the water balance, we applied the crop-soil simulation model CropSyst for the season under study and for the period 1980-2010 (30 years). Results showed that planting date had a notable impact on crop performance and yield (30-year average, early: 2.68 Mg/ha; late: 2.30 Mg/ha). Simulations indicated that planting wheat immediately after the first sufficient rainfall in autumn bears little risk of crop failure due to early season droughts, and more should be done to encourage farmers to do so. ZT and residue management changed yields only very little, even though in 25 out of 30 years, ZT yields were higher than CT yields. About 55% of the seasonal precipitation (similar to 150 mm) was lost by unproductive soil evaporation, whereas ZT and residue retention had only a minor mitigating impact; too little to be clearly distinguishable by field observations. A potential obstacle for meticulous simulation of CA with CropSyst is the model's inability to simulating the dynamic nature of tillage, i.e. its decreasing impact over time, and the beneficial effect of ZT and residue retention on soil water infiltration. However we argue that such impact may be limited on soils with self-mulching characteristics that are common in the region of this study. (C) 2012 Elsevier B.V. All rights reserved.
