Modelling of K/Ca exchange in agricultural soils
Résumé
The exchange selectivity of soil toward K relative to Ca increases in the depletion profile near roots during plant uptake. The modelling of the soil K/Ca exchange is thus necessary for accurate soil–plant transfer modelling. The aims of this work were i) to study the change in the K affinity of 45 agricultural soil samples that came from 15 French K trials in the range of small K saturation ratios of the soil CEC (< 25%); ii) to test two multi-site models for their ability to represent the K/Ca exchange data, i.e., a low and infinitely high K-selective sites model (M2) with two adjusted parameters and a low and high K-selective sites model (M3) with three adjusted parameters; and iii) to propose a model of the exchange parameters. Compared to model M3, model M2 predicted the exchange data well despite its lower number of adjusted parameters. The two parameters of model M2 could be modelled, i.e. the Gapon selectivity coefficient for the low K-selective sites (Gn) and the cation exchange capacity of the infinitely high K-selective sites (Si). For the Podzol Gn was lower (0.75) than Gn for the other studied soils, which had a higher mineral content, and Gn for these other soils linearly increased with an increase in the clay content of the soils. For all of the soils, Si linearly increased with the clay content and with the initial exchangeable K content of the soil. Using this description of the two parameters of model M2, more than 97% of the variance of the exchange data could be explained. Thus, when exchange sites' heterogeneity is taken into account, the K/Ca exchange selectivity coefficient can be adequately and simply described. This conclusion can be drawn regardless of the formalism used to model the exchange data, including the Vanselow and the Gaines and Thomas formalisms.