Modeling of wheat (Triticum aestivum) yield in relation to soil characteristics and management practices

Abstract

A study was carried out to evaluate the production potential of soils in Agro-Ecological Sub-Region (AESR) 4.2 of Udaipur district in Rajasthan for wheat (Triticum aestivum L. emend. Fiori & Paol.) crop. Grain yield was monitored in farmers' fields consecutively for 3 years during winter season of 2004-05, 2005-06 and 2006-07. Multivariate regression model was developed to relate crop production to the set of four independent variables, viz. soil depth, available water content, irrigation and sowing time, and management. By using the regression equation, yield was predicted in 2005-06 and 2006-07, which was validated with actual yield. The available wa- ter capacity strongly affected the crop yield. In shallow sandy-loam soils, due to low available water content, minimum six irrigations were required for obtaining potential yield of 5.5-5.6 t/ha, whereas five irrigations were required in moderately deep to deep sandy clay-loam soils to produce potential yield of 5.7-5.8 t/ha. In deep clay soils four irrigations were sufficient to produce the potential yield. A saving of 15-20% water was achieved through the sprinkler system. The modelling exercise of the response of spatially distributed wheat yield was strongly affected by the soil variability within the study region.