Soil Potassium Balance and Cumulative Cotton Response to Annual Potassium Additions on a Vermiculitic Soil

Cumulative effects of annual K additions on cotton (Gossypium hirsutum L.) growth and soil K balance were examined in a field study on an irrigated, vermiculitic soil. In each of 3 consecutive yr, cotton ‘Acala SJ2’ was grown with 0, 120, 240, or 480 kg K ha⁻¹ applied to the same treatment plots in a randomized complete block design with 10 replications. The relationship between seed cotton yield and petiole K concentration was similar across years (r² = 0.61, P < 0.001), but slopes of the yield response to applied K increased each year; 2.6-fold more seed cotton was obtained per unit K input in 1987 than in 1985. This cumulative response reflected a yield decline without K addition and increasingly higher yield with each additional input of 480 kg K ha⁻¹. Two processes appeared to control this response. First, without K input, NH₄-extractable K⁺ in surface soil decreased 20% from August 1985 to August 1987. This decrease greatly exceeded the K removal by cotton, and we surmise that an 11% decrease in soil organic C content contributed to this phenomenon. Loss of cation exchange capacity associated with the loss of soil humus may shift K equilibria towards fixation at interlayer sites in vermiculite at the expense of external K pools extracted by NH⁺₄. Second, with high levels of K input, apparent K uptake efficiency from applied K was 50% greater in 1987 than in 1985. Increased efficiency was attributed to the partial saturation of K fixation sites from previous K additions and thus a larger proportion of newly added K remained in plant-available K pools. At the highest K addition level, 86% of the 1440 kg K ha⁻¹ applied in the 3-yr period was fixed beyond extraction by NH⁺₄, and plants remained marginally K deficient. The existence of a cumulative response to applied K on vermiculitic soils would have a marked impact on the economics of K fertilization on cotton.