Die invloed van verskillende grondeienskappe op die bepaling van 'n fosfaatsorpsiemaksimum

Abstract

Phosphate sorption processes in soils are influenced by soil properties that are so closely interrelated that they cannot be studied in isolation or even by eliminating some of them. The graph of the quantity of extractable phosphate in soil versus time, exhibits a characteristic twophase pattern for the soils of the area studied. This pattern initially indicates a rapidly declining phase over the first seven to ten days, followed by a sharp bend in the curve which, thereafter shows a relatively gradual decline for at least 70 days. The behaviour is best explained by fitting a split line regression function. Three parameters are obtained, B1, B2 and B0 respectively, which represent the rapid and gradual slopes as well as the intercept of the spline. In this study such a spline function was fitted to the sorption curve of extractable phosphate versus days of contact of a soluble P-source with the soil particles. Soils with very little or no residual phosphate were used to simplify the analysis of quantification techniques. The use of the spline technique was succesful in the quantification of the sorption characteristics of a soil and the B0 coefficient allowed the prediction of the quantity of extractable phosphate seven days after addition of a known quantity of a soluble phosphate (P) source (r = 0.82). A further attempt was then made to utilize the same methodology in studying the complicating influence of residual P on sorption characteristics. The study was repeated for a set of soil samples from the same sampling area and the same soil forms which contained different levels of residual phosphate from previous fertilization practices. This yielded a simplified fonnula describing the relationship between the extractable phosphate, the clay content, the KCIextractable Al and the residual P-content of the soil calculation of B0 which resulted in an improved r2 = 0.93. The application of the above results in the optimization of Phosphate fertilization practices was then attempted. The possible number of combinations of clay content, KCl-extractable P and the extractable P-Ievel which could be chosen in a geographical area as optimum for crop production, is unlimited. In order to overcome this problem, it was necessary to stratify the sample area with regards to the above-mentioned soil properties. An approach was evaluated in which the B0 property of a specific soil was calculated thereby establishing the actual amount of phosphate required in addition to the residual P already in the soil. The actual amount of a specific fertilizer required to satisfy the calculated demand of the soil could then bedeterrnined by selecting an appropriate linear equation best suited for the specific characteristics of the soil. The success of this approach was evaluated by perfonning a validation experiment in which calculated amounts of the applied phosphate and the actual amounts measured in a variety of soils were compared seven days after application. In this study a high degree of correspondence between these two figures was found illustrating he value of the suggested approach. The results of this study are not only supported by the latest relevant subject literature, but supply a more accurate methodology for optimizing P fertilization applications.