A non-linear model for optimal allocation of irrigation water and land under adequate and limited water supplies. A case study in Southern Italy.

A non-linear mathematical optimization model was used to define optimal water and land allocation and optimal cropping patterns for a southern Italy irrigation district. The objective function was based on crop-water production relations, the parameters of which were derived from an extensive experimental data set including nine arable and vegetable crops. Among production costs, the irrigation binary tariff system adopted in the study region was considered. The study, performed across a wide range of total water supply, identified: (i) optimal water allocation, by keeping crop surface constant; (ii) optimal cropping patterns, when allocation of available land was also investigated. When a fixed surface was allocated to each crop, and total water supply was not limiting, maximum return was achieved by providing the greatest fraction of available water to vegetable crops (pepper, tomato, eggplant, beans), with seasonal volumes corresponding to full restitution of irrigation water requirements (from 6758 to 3571m3 ha-1), while assigning volumes equal to 78, 62, 48 and 41% of water requirements to summer and springmaize, autumnal and spring sugar beet. When the allocation of water and land was investigated simultaneously, optimal cropping patterns and seasonal volumes were identified. The latter varied as a function of crop and total water supply but also, at a given supply, for the same crop.