Dynamics of Soil K Release

It is well known that soil management and potassium (K) fertilizer applications affect soil‐K pools. However, it is not quite clear how K is distributed among K pools in different soil types shortly after its application and after many years of frequent K fertilizer applications. The objective of this work was to determine the size of three soil‐K pools after recent and long‐term K fertilizer applications to identify K distribution patterns. The information generated could be useful for modeling. Soluble and exchangeable K (Ks and Ke) release curves of two Mollisols and an Entisol from Hungary, collected in long‐term experiments (Nagyhörcsök, Kompolt, and Orbottyán), and of Vertisol, Mollisol, and Entisol from Mexico, collected in farmer fields, were determined. The treatments sampled in the Hungarian experiments included checks that had not received K fertilizer or had received very low rates (60–80 kg ha−1 K2O) every second or third year during the last 10–30 yr and plots heavily fertilized with K (1440 up to 8800 kg ha−1 K2O) during the same period. The Ks pool (the sum of several sequential extractions made with CaCl2 [0.01 M]) and Ke pool (the sum of several sequential extractions made with NH4OAc [1N pH7]) were determined in samples without and with added K fertilizer (100 ppm K) before and after a 50‐day incubation (35°C, field capacity) period. Non‐exchangeable K was determined by the HNO3 (1N) procedure. In general, slightly cultivated Mexican soils contained more K than the soil of the long‐term Hungarian experiment either fertilized with K or not. Soil incubation without added K slightly increased the Ks and Ke; however, a relatively large increment was observed in non‐exchangeable K (Kne). When K was added to the soil samples of check plots of Mollisols from Nagyhörcsök and Kompolt before incubation, Ks and Ke increased, but a decrease was observed in the Entisol from Orbottyán. The Ks and Ke released as a function of the sequential extractions followed a negative power function in all cases. When the Ks and Ke values were expressed as a fraction of total Ks or Ke (the sum of all sequential fractions), only one model was necessary to account for all soils and treatments. Hungarian and Mexican soils behave in the same way. These results suggest that the release mechanism is independent of the soil type, clay mineral type, and previous management. The actual Ks and Ke being released in each extraction depends only on the size of their respective pools. This information is useful for modeling purposes.