Measuring interlayer potassium release rates from soil materials. I. Critical evaluation on the use of resin and other batch procedures for determining kinetic data

Laboratory procedures are evaluated for their principal suitability to derive unbiased kinetic data of interlayer K release (days to weeks, Luvisol ‘Ohlendorf’, loess, Ap horizon). The final intention is to relate K release rates to their main governing soil environmental variables rather than to conventional procedural standards. Of actual interest are technical aspects of quantifying the influence of solute K (C_K), which is considered to be the most important among those variables. The relevant range is that of a few μM K, similar to the rhizosphere of plants where diffusion-controlled, non-destructive K release is thought to be of importance. Tested non-resin batch techniques, e.g. the time dependent shift of K-Ca exchange isotherms, are considered to be unsuitable because the derived release rates were markedly biased by shaking the suspensions. Generally, the batch technique just showed that additional proton induced release of nonexchangeable K (K_nex) is negligible above pH 4. Ca-resin procedures extracted considerable amounts of K_nex but most of it had to be attributed to the addition of fresh (K-free) exchange solution rather than to the presence of the resins. Further, an active regulation of C_K in the intended range was only hardly possible via the resin/soil ratio but was inexact and only achieved with inacceptable effort. Further modifications, e.g. containing the resins in dialysis tubes, improved the technique but did not eliminate the main problems: Ca-resin does not act as a real sink for K. H-resins did act as a sink but, due to other constraints, are also rejected: C_K is too low (< 1 μM, no effective regulation possible) and the pH of the suspension steadily declines (< 3 after one week). The one promising procedure among those tested was a continuous (non-resin) percolation technique (> 900 h) which permitted a flexible regulation of C_K by the percolation rate.