Determining water use efficiency of wheat and cotton: A meta-regression analysis

A great challenge for agricultural production is to produce more food with less water, which can be possibly achieved by increasing crop water use efficiency (WUE). This study systematically reviewed 52 cases from 49 empirical studies with field experimental results on wheat and cotton. This research investigated yield-water use relations under both furrow and micro irrigation systems, compared optimal water use to achieve maximum WUE and maximum yield, calculated water saving potentials under various scenarios, and evaluated the effects of influential factors using meta-regression analysis. The results suggest that to achieve maximum WUE rather than maximum yield, water use for wheat can be reduced by 30.4% with a grain yield decrease of 14.8%, and water use for cotton can decrease by 51.4% with a yield reduction of 51.7%. Compared with furrow irrigation, micro irrigation reduces wheat water use by 22.7% and increases yield by 36.7%. While for cotton, micro irrigation reduces water use by 36.8% and decrease yield by 21.4%. Under the scenario of a 10% yield reduction, water use decreases by 25% for wheat and by 20–22% for cotton. Compared with maximum yield, other yield levels reduce water use by 2–15% on average for wheat, and by 15–17% for cotton. Achieving maximum WUE reduces water use by 14–31% compared with other sub-optimal WUE levels. The meta-regression analysis showed adoption of micro irrigation systems, and farm management practices on soil and water significantly improved wheat and cotton WUE. Assessments of the publication selection bias and genuine effects illustrate the application of weighted least squares in conducting meta-regression analysis.