Potassium (K) release through chemical weathering of soil minerals is an issue that has seen a renaissance by today's discussion on sustainable agriculture with a small intput of mineral fertilizers. Exchangeable K may be depleted if weathering is slower than the annual net output of K from the field, hampering the availability of K to growing crops. In this investigation we assessed rates of K release, and fixation, under field conditions in selected soils from the southern boreal to northern temperate zone in Sweden. The objective was to identify soils, where weathering might be sufficient or insufficient, respectively, to meet the crops' demand for K.

We used five long-term field experiments located on varying parent materials in South and Central Sweden, each having an array of K fertiliser application rates. The soils may be divided into three groups, ranked from high to low clay content in the following order (classification according to Soil Taxonomy within parentheses):

1. One Clay (Typic Haplaquept) and one silty clay (topsoil) on clay (Oxyaquic Haplocryoll)
2. One sandy loam (Oxyaquic Hapludoll) and one loam (Oxyaquic Hapludoll)
3. A loamy sand (Humic Dystrochrept)

According to earlier investigations, the “rich” soils of Group 1, besides their high clay content, have a high concentration of mica/illite within the clay fraction, ca 50% of the clay fraction in the topsoil. The corresponding figures for soils of Group 2 and 3 are in the range 19–25%.

Net weathering rates (weathering – fixation) were calculated based on K inputs and outputs from the soil during a forty year long period, and on changes in topsoil and subsoil exchangeable K (ΔK_ex). This yielded an estimate of net K release by chemical weathering according to Eq. 1 (all terms are amount of K per unit time and land area; e.g., kg ha^–1 yr^–1). In this paper, the term ‘weathering' bears upon all processes that release K from the soil, except for any changes in the exchangeable pool. Net fixation of K is implied by a negative K_weathering term.

K_weathering + K_fertilizer + K_manure + K_deposition = K_harvest + K_leaching + ΔK_ex                                                                      (Eq. 1)

The calculated weathering and fixation rates were highly dependent on the net, positive or negative, K input to the soil. The estimates ranged from 61 kg ha^–1 yr^–1 of K released by weathering, to 84 kg ha^–1 yr^–1 of K trapped in fixation. Minimum use of K fertiliser called forth maximum release of K by weathering. In plots receiving no mineral K fertilizer, the clays (Group 1) and the loams (Group 2) released 40–60 kg ha^–1 yr^–1 by weathering, which was close to the annual off-take of K by the crops. However, the loams of Group 2 had a greater sensitivity in exchangeable K towards outputs and inputs than had the clays of Group 1. It therefore appeared that whereas the clays (Group 1) most probably had a K release capacity greater than the crops' demand, the loams (Group 2) might be close to the limit of their potential performance. The loamy sand (Group 3) delivered less than 10 kg K ha^–1 yr^–1 by weathering, which was less than the annual crop off-take from this soil, 29 kg ha^–1 yr^–1. This soil clearly needs K fertilizer, or else K_ex and thereby K availability to crops decline.

We also investigated the pool of K extracted with 2 M HCl at 100°C (‘K_HCl'; extraction during 2 h). The size of this K pool showed a significant response to weathering and fixation; the average loss or gain in this pool was 56± 44% of the weathering and fixation, respectively. This confirms the conception that K_HCl, which is a routine analyte in the mapping of Swedish agricultural land, represents a soil K reserve that can be replenished or depleted reversibly due to cropping and fertilization.