Abstract
Optimisation of agricultural water management in arid and semi‐arid regions requires the availability of water–salinity crop production functions. A two‐year experiment was conducted in the northern Golestan province of Iran to assess the water–salinity production function of wheat. The treatments in the experiment consisted of four levels of irrigation water, i.e. 50 (W1), 75 (W2), 100 (W3) and 125 (W4) % of crop water requirement, and four levels of water salinity, respectively 1.5 (S1), 8.5 (S2), 11.5 (S3) and 14.2 (S4) dS m^−1^. The plots were arranged in a randomised complete block design with three replications and water quantity as main plot treatment and water quality as subplot treatment. The data were analysed using linear, quadratic, Cobb–Douglas and transcendental functions, complemented with an economic analysis. The results indicate that for the given climate–soil conditions, transcendental functions best predict wheat yield under both water and salinity stress conditions. Yield reduction caused by a unit increase of matric potential is found to be larger than that caused by a unit increase of osmotic potential. The marginal rate of technical substitution indicates that each one of the two factors studied, namely soil salinity and water supply, can be substituted with the other in a wide range in order to achieve equal amount of yield. Copyright © 2008 John Wiley & Sons, Ltd.