The objective of this investigation was to examine whether the silt in loess-derived soils represents a major K+ source for plants. The experiment was carried out with 14 arable loess-derived soils (Alfisols) from which, in one series, the clay was removed and in which, in another series, complete soils were used. In this work, the latter are denoted as whole soils, whereas the soils without clay are labeled silt + sand fractions, of which the silt proportion was about 80%. Potassium was extracted from whole soils and from silt + sand fractions by electroultrafiltration (EUF). Potassium quantities extracted from the silt + sand fractions were about half the K+ quantities extracted from whole soils. Potassium rates extracted and plotted on extraction time yielded curves that fitted the Elovich equation.
Dry matter yield of ryegrass (Lolium perenne) grown in pots on whole soils did not differ significantly from dry matter yields obtained from grass grown in pots on silt + sand fractions. Thus, it is suggested that the K+ present in the silt + sand fraction was readily available to the grass. Close correlations between K+ uptake of the grass and K+ soil parameters were obtained with the EUF extraction at 400 V and 80°C, with which the interlayer K+ was mainly obtained. The b values of the Elovich equation reflect the K+ release rates of interlayer K+. The data provide evidence that in loess-derived soils, the silt fraction, presumably because of its high mica concentration, is of great importance for the supply of K+ to plants.