Potassium nutrition of irrigated potatoes in South Australia. 2. Effect on chemical composition and the prediction of tuber yield by plant analysis

Abstract

I compared the effects of potassium sulfate and potassium chloride fertiliser, at rates up to 1280 kg/ ha K, on the concentrations in petioles of the youngest fully expanded leaves and in harvested tubers of potassium, calcium (tubers only), magnesium, phosphorus, chloride and sulfur (petioles only) and on potassium: magnesium ratios in petioles. For the sites classified as deficient and marginal according to soil test potassium concentrations, increased application rates of potassium sulfate increased the potassium and sulfur concentrations and potassium: magnesium ratios in petioles and decreased magnesium, phosphorus and chloride concentrations. In tubers both potassium and magnesium concentrations increased as potassium sulfate rates increased, while calcium and phosphorus concentrations showed negative trends. Compared with potassium sulfate, potassium chloride resulted in higher potassium, calcium and chloride concentrations and lower sulfur concentrations. Chloride concentrations in petioles from many sites were much higher than those reported in other studies and were related to high chloride concentrations in irrigation waters. Significant correlations were obtained between bicarbonate- extractable potassium concentrations in the surface (0-1 5-cm) soils and potassium concentrations in petioles (r2 = 0.75***) and tubers (r2 = 0.84***). As bicarbonate-extractable potassium concentrations in the surface soils increased from 40 to 600 mg/kg, the potassium concentrations in tubers increased from 1.69 to 2.69% compared with an increase from 5.72 to 1376% in petioles. Significant correlations were obtained between relative yields ((mean treatment yields/maximum mean treatment yie1d)x 100) and potassium concentrations (dry weight basis) in petioles and tubers and potassium:magnesium ratios in petioles. The percentage of variance in relative yield accounted for by each regression was 67.9, 31.0 and 56.9% respectively. The potassium concentration in petioles was the most effective plant index to predict yield response to potassium. For petioles and tubers prognostic critical potassium concentrations of 11.30 k 0.26% and 2.19 ¦ 0.11% respectively were established. Alternatively, a prognostic critical potassium:magnesium ratio of 14.8 ¦ 0.8% was determined. The choice of the mathematical model used influenced the critical values obtained.