The description of sorption curves

In this paper I discuss sorption curves: choosing them, fitting them, and interpreting them. The Langmuir equation is not appropriate for reaction of ions with soil. Despite some disadvantages, the Freundlich equation is the best simple equation. Fitting sorption equations causes statistical difficulties because sorption is usually calculated from the change in concentration. The two variables are therefore not independent. If this is ignored, the errors in the coefficients of the fitted curves will be small, but there will be larger errors in the measures of goodness of fit and tests of significance will not be valid. If mild but nevertheless adequate mixing of soil and solution are chosen, sorption and desorption will be the opposites of each other. We can therefore learn much about the rates of desorption by studying the rates of sorption. If a sink is used to capture desorbed nutrients, the potential rate of desorption will be underestimated, as no sink can instantly and completely remove all the nutrients from solution. Especially with phosphate, continuing reaction makes the surface potential more negative. This means that re-application is made to a soil of smaller buffering capacity. This effect has been insufficiently studied. It influences both the response to further applications, and the ability of soil to store nutrients.