Corn Irrigation Scheduled by Tensiometer and the Lambert Model in the Humid Southeast

A need for a grower acceptable irrigation scheduling technique was brought about by rapid expansion in irrigated land in the Southeast. A modified Jensen-Haise evapotranspiration formula was adapted by Lambert et al. into an irrigation scheduling and forecasting model for this region. The purpose of this study was to determine if the Lambert model was as effective as tensiometer scheduling in meeting irrigation needs of field corn (Zea mays L.). Corn was grown in seven replicated field experiments and two nonreplicated production scale experiments. Yield and water were compared after irrigation was applied either when soil water pressure in the 150 to 300 mm zone fell below −30 kPa as indicated by tensiometers or when called for by the model. The comparisons were conducted for 3 years on two soils, Tifton loamy sand (fine-loamy, siliceous, thermic Plinthic Paleudult) and Bonifay sand (loamy, siliceous, thermic, grossarenic Plinthic Paleudult). In five of the comparisons a nonirrigated check was included. The treatment irrigated by the Lambert model yielded 11.1 Mg ha⁻¹ over all seven experiments; the tensiometer irrigated treatment yielded 11.0 Mg ha⁻¹, and the nonirrigated treatment yielded 4.3 Mg ha⁻¹. There were no significant differences between the two irrigated treatments within any year on either soil. The mean yearly water required by the model was 285 mm, compared to 237 mm for the tensiometer. In four of the seven experiments the model called for more water than the tensiometer treatment. Model predicted soil water content was generally greater than measured soil water content for the Bonifay soil but was not significantly different for the Tifton soil. Experience with the Lambert model indicated that it was difficult to determine some of the required model inputs but with correct inputs the model could accurately predict soil water content and corn water requirements.