Potassium Response of Cotton on Some Inceptisols and Oxisols of Nothern Nigeria

Consequent upon the shift from the traditional bush fallowing system of agriculture to more intensive continuous cultivation, plant-available soil K which was hitherto considered adequate in the West African savannah zone is now becoming a major yield limiting factor. Information is urgently needed on the K fertilization requirements of the promising crop varieties and on the suitable diagnostic parameters for evaluating these requirements. The objectives of the present study were to: 1) approximate the optimal K fertilization level for cotton (Gossypium hirsutum L.), 2) evaluate the yield predictive abilities of four K availability indices, and 3) estimate the critical values of these indices under the tropical savannah conditions. Four rates of K (0, 25, 50, and 75 kg ha⁻¹) were tested in a 3-year field study conducted on three Oxisols (Typic or Tropeptic Haplustox, fine, gaolinite, isohyperthermic) and two Inceptisols (Oxic and Typic Ustropept, fine, mixed, isohyperthermic). Seedcotton yields were measured all 3 years while soil and index leaf samples were taken from the third year's trials for laboratory chemical analysis. The relationships between seedcotton yield and soil exchangeable K, (Ca + Mg)/K ratio, rates of applied K, and leaf percent K were measured using multiple regression techniques involving quadratic polynomials. Significant yield responses (only at first level of applied K) were obtained on the Oxisols (all of which are derived from sandstones and had topsoil exchangeable K values of ≤ 0.10 meq/100g). Marginal responses obtained on the Inceptisols are, however, indicative of imminent deficiency. Percent K in index leaf best predicted yield (R² = 0.86***; yield = 3099.2 + 6031.6 % K — 1543.3 % K²) and so provides the best measure of K availability for cotton followed by rates of applied K (R² = 0.812*; Yield = 1302.3 + 56.96 app. K — 0.54 (app. K²). Their estimated critical values (values associated with the highest predicted seedcotton yields) were 1.84% K and 50 kg/ha K, respectively for a highest yield range of 2,400 to 2,700 kg ha⁻ˡ. These results may be limited by the relatively few observations used but they are sufficiently close to published values for crops of similar nutrient requirements to justify proposing them as a guide for the savannah areas of Nigeria and similar soils elsewhere.