EVALUATING A MODEL FOR PREDICTING NUTRIENT-UPTAKE BY RICE DURING VEGETATIVE GROWTH

Since submerged soils create a unique chemical environment from which roots take nutrients, models that estimate nutrient uptake for upland crops may differ from those for rice (Oryza sativa L.). A mathematical model that had been used to predict nutrient uptake by corn (Zea mays L.), soybean [Glycine mar. (L.) Merr.], and sorghum (Sorghum bicolor L.) was evaluated for predicting uptake by rice under flooded soil conditions. Our objectives were to evaluate such a model for rice and to conduct greenhouse and growth chamber studies to measure the soil and plant parameters required by the model. Sensitivity analyses were also conducted to determine the parameters most influencing N, P, and K uptake. Two soils [Crowley silt loam (Typic Albaqualfs) and Calhoun silt loam (Typic Glossaqualfs)] and three rice cultivars (Lemont, Katy, and Mars) were selected for this study. Linear regression of predicted vs. observed uptake for N, P, and K gave satisfactory fits to the data with R2 values of 0.48, 0.99, and 0.88, respectively. The slopes of the lines for predicted vs. observed uptake for N and K were 1.01 and 0.94, respectively, and the relationships did not depend upon either soil or cultivar. However, the prediction of P did depend on cultivar. The slopes of the lines for predicted vs. observed P uptake by Mars, Lemont, and Katy were 1.81, 1.08, and 1.02, respectively. The parameters most affecting N uptake were root growth rate, N concentration in the soil solution, and the half-distance between root axes. However, parameters most influencing P and K uptake were root growth rate, root radius, and maximal influx rate.