ObjectivesThe study is aimed to reveal the evolution of different soil potassium fractions and the balance of potassium in yellow paddy soils after long-term fertilization, in order to provide scientific basis for improving soil potassium fertility and fertilizer application.

MethodsFour fertilization treatments were designed as control (CK), pure chemical fertilizer (NPK), NPK plus cattle manure (NPKM) and NPK plus straw (NPKS). The potassium contents of different potassium fractions in the topsoil, the organic carbon content were determined in selected years. Plant samples were collected to analyze relationship between soil potassium level, soil K balance, plant nutrient and input of potassium in 4 consecutive years.

ResultsCompared with CK, the average soil total potassium contents in the fertilization treatments were increased by 2.2%–5.0%, the differences among the treatments were higher under double cropping rice system than under single cropping system. The soil readily available potassium contents were affected more than other potassium forms in all the fertilization treatments, the increases were in range of 43.9%–102.4%, and the largest increment was in the NPKS treatment. The slowly available potassium was not significantly different in all treatments. The soil readily available potassium content is significantly and positively correlated with the total potassium. There were most significantly positive correlations among soil total potassium, readily available potassium and organic carbon contents. The potassium contents and potassium uptakes in the plants were increased in the fertilization treatments at the harvest time, with the significant difference found in the NPKS treatment. In the NPK, NPKM treatments, the annual soil potassium was deficit of 89.4 and 18.6 kg/hm², while in NPKS treatment, that (K₂O) was surplus of 101.1 kg/hm². The potassium application rate, readily available potassium content and potassium balance were significantly and positively correlated to each other. The calculated soil potassium exhibited an equilibrium state when the annual application rate of exogenous potassium reached 161.8 kg/hm², and corresponding soil available potassium content was 95.3 mg/kg. The readily available potassium contents, potassium balance and potassium contents in the plants were significantly and positively correlated to plant yields.

ConclusionsThe soil total potassium and readily available potassium contents were improved by long-term fertilization, with the largest increment of readily available potassium in the NPKS treatment. The yield increment is positively correlated with the readily soil available potassium content and potassium balance. The soil readily available potassium is more sensitive in directly reflecting the productivity than slowly available potassium and total potassium. Chemical NPK fertilizer combined with straw treatment produces obviously more soil potassium surplus than the NPK and NPKM, and thus leads to significant increase in soil readily available potassium content.