EVALUATION OF A MECHANISTIC MODEL OF POTASSIUM UPTAKE BY COTTON IN VERMICULITIC SOIL

The Cushman-Barber model was evaluated for K uptake by cotton (Gossypium hirsutum L.) from vermiculitic soils of high K-fixation capacity and low solution-phase K+ concentration. On these soils, cotton exhibits late-season K deficiency while other crop species remain unaffected. Four soil treatments of NH4-N, K, or both were combined factorially in a vertically split-pot system to create eight plant treatments of different uniform and nonuniform soil nutrient environments. Depending on the treatment, initial model output produced both substantial under- and overpredictions of whole-plant K accumulation. Model precision was greatly improved by changing the Michaelis-Menten kinetic parameters for uptake at the root surface to reflect differences in shoot K/N balance, a more accurate measure of plant K demand in the presence of variable soil N supply. Regression revealed a linear relationship between the predicted and the observed K uptake (r2 = 0.87) but, across treatments, the model underpredicted accumulation by 43%. Model predictions were further improved by estimating soil buffer capacity from a Langmuir fit of a K adsorption isotherm rather than from the relationship between exchangeable solid-phase and solution-phase K pools. Sensitivity analyses were performed to identify key determinants of cotton K acquisition from vermiculitic soils of high K-fixation capacity. The analyses demonstrated that both Michaelis-Menten kinetic and soil supply parameters were strong determinants of K uptake on these soils and thus warrant more emphasis than previously suggested by model validations conducted on soils with greater K supply and less K-fixation capacity.