EVALUATION OF SOIL EXTRACTANTS FOR THE PREDICTION OF PLANT-AVAILABLE POTASSIUM IN ONTARIO SOILS

The effectiveness of five extractants, NaCl, AB-DTPA, NH4OAc, Mehlich 3 and HNO3 was evaluated on 72 Ontario soils for prediction of plant-available K to alfalfa (Medicago sativa L.) grown in a greenhouse. Total K uptake by four cuts of alfalfa was used to measure plant available soil K. The average amounts of K extracted were in ascending order: AB-DTPA < NaCl < = NH4OAc < Mehlich 3 < HNO3. The highest simple correlation with K uptake by alfalfa was obtained with HNO3 (r2 = 0.76) and the lowest with NaCl (r2 = 0.61). The K uptake by alfalfa was used as the dependent variable to develop predictive models using stepwise regression, with extractable soil K, exchangeable Mg and Ca, soil pH, organic matter, cation exchange capacity (CEC), and contents of silt and clay as independent variables. The best prediction of K uptake with extracted soil K and its square was AB-DTPA (R2 = 0.84), followed by HNO3 (0.81), NH4OAc (0.81), NaCl (0.78) and Mehlich 3 (0.77). Inclusion of soil properties improved the effectiveness of prediction. The models developed with the different extractants differ in the number of variables which contributed significantly at 5% probability. Based on the final R2 and ease of measurement of variables, the NH4OAc model is one of the best as a predictor of availability of K in Ontario soils when used along with other soil properties such as plant-available nonexchangeable K and soil organic matter.