This research explores the long-term relationship between water resources, irrigated land use change and crop production within a computable general equilibrium modeling framework. The modeling approach is developed on a variant of the MIT Emissions Prediction and Policy Analysis (EPPA) model that describes three agriculture sectors-crops, livestock and managed forestry-five land types-cropland, pasture land, managed forest land, natural grass land and natural forest land-and conversion among these land types. I further develop this framework by describing crop production as the aggregate production of crops grown on irrigated and non-irrigated cropland. Water resources, through the parameterization of regional irrigable land supply curves, limit conversion to irrigated cropland and thus constrain regional crop production.

Land use change, dynamics of irrigated land and regional water demand and crop production are investigated with the new model structure. Non-irrigated cropland is found be expanding faster than irrigated cropland. However, regionally, competition from biofuels for non-irrigated cropland may drive further expansion in irrigated cropland. Regarding water demand, most regions are withdrawing a very small share of their renewable water resource. Crop production levels are compared to results from a model that does not include water constraints. Global crop production declines a small amount with the most significant regional effect observed in the Middle East where regional water constraints have severely restricted the area by which irrigated cropland can expand. This result highlights the importance of considering water resource constraints in regions that experience, or might experience, shortages of water.