Studies on soil potassium. I. A theoretical model of the Q/I relationship

Abstract

To describe the Q/I isotherm for labile soil potassium, Beckett (1964a) has proposed a theoretical niodel which assumes two types of exchange site and characterizes the isotherm by four parameters. This paper discusses the model and shows how it can be fitted to data by least squares so that tests of hypotheses concerning the parameters can be carried out. The usefulness of this approach was tested on published data and on new experimental data for 10 soils (nine from Queensland, one from South Australia) subjected to a range of cropping and storage treatments. In all cases the model described the data adequately, although this was to be expected since it contains four adjustable parameters. Statistical tests on some of the published data showed that graphical estimates of parameters, and conclusions based on such estimates, are unreliable. Tests of significance suggested that exhaustive cropping changed the form of the Q/I isotherm for only three of the 10 Australian soils. These three soils all contained illite, and yielded more non-exchangeable potassium to plants than the other seven. It is suggested that the changes were due to increases in the interlamellar surface areas accessible to exchange, and that some of the newly exposed exchange sites were highly specific for potassium. If this is so, the simple two-site model is theoretically inadequate for these soils at least, since after cropping it appears to group together, as non-specific, sites with a wide range of specificities.