Evolution Characteristic And Its Mechanism Of Potassium In Typical Agriculture Soils Under Long-Term Fertilization

Field and laboratory experiments were conducted to characterize the evolution characteristic of potassium (K) and its mechanism as affected by different climates, cropping systems, and long-term fertilization, and to investigate the effect of long-term fertilization on K fixation capacity using laboratory simulating methods and X-ray diffraction, and the nonexchangeable K release in 0.01 mol/L CaCl2 and 0.01 mol/L oxalic acid extrations, and to study the Quantity-Intensity (Q/I) relationship in seven typical agriculture soils (red soil, paddy soil, grey desert soil, loess soil, fluvo-aquic soil, black soil, and purple soil) in China.The decrease rates of the contents of available K and slow-available K in NP treatment in seven soils were observed in the following order: type A soil (red soil) (5.72 mg/kg/a and 38.7%) > type C soil (black soil and purple soil) (2.01～4.84 mg/kg/a and 11.7%～18.2%) > type B soil (paddy soil, loess soil, grey desert soil, and fluvo-aquic soil), except for available K content in grey desert soil. The decrease of the contents of available K and slow-available K in fluvo-aquic soil was larger than that in other type B soil. The contents of available K and slow-available K in NPK treatment in type A soil and type C soil either remained constant or decreased, whereas that in type B soil in the same treatment either remained constant or increased significantly. The contents of available K and slow-available K in NPKM treatment in type A soil increased slightly, while increased significantly in type B soil except for grey desert soil, and remained constant in type C soil.From the data of X-ray diffraction study, the application of K fertilizer induced no significant changes in the K-bearing mineralogy at the peak of d001 in NP treatment from in NPK treatment in type B soil, whereas that in type A soil and type C soil had slightly difference at the peak of hydromica, chlorite and vermiculite, and mixed-layered mica-smectite component after 15-yr fertilization.When added K ranging from 0.4 to 4.0 g/L, the K fixation rate in seven soils after 15-yr fertilization were observed in the following order: type C soil (19.4%～91.4%)>type B soil(9.8%～76.4%)>type A soil (0.8%～13.7%). The K fixation rate in black soil (27.9%～91.4%) was larger than that in purple soil (19.4%～84.0%) in type C soil, while that in paddy soil and grey desert soil (9.8%～51.8%) was smaller than that in loess soil and fluvo-aquic soil (17.9%～76.4%) in type B soil. Potassium fixation capacity of NPKM treatment remained constent, whereas that of NP treatment had larger increase (12.4%～25.3%) in type C soil, and had little changes in type A soil due to its lower K fixation capacity compared with NPK treatment.Release kinetics of nonexchangeable K confirmed fairly well to parabolic and first-order reaction kinetic models. The nonexchangeable K release rate constant (b) of first-order reaction model in 0.01 mol/L oxalic acid extraction were observed in the following order: type C soil (0.0170～0.0195/h) > type B soil (0.0170～0.0195/h) > type A soil (0.0170～0.0195/h). The b value was greater in black soil (0.0185～0.0195/h) than that in purple soil (0.0136～0.0149/h), and were greater in loess soil and fluvo-aquic soil (0.0144～0.0168/h) than that in paddy soil and grey-desert soil (0.0136～0.0149/h). The b value was different in different treatments following the sequence: NPKM>NPK>CK,N,NP.The value of K+ equilibrium activity ratio (AR0) was greater in type B soil (0.00217～0.00637 (mol/L)1/2) than that in type A soil (0.00041～0.00063 (mol/L)1/2) and type C soil (0.00033～0.00099 (mol/L)1/2), and was greater in paddy soil, loess soil and grey desert soil (0.00360～0.00637 (mol/L)1/2) than that in fluvo-aquic soil (0.00217～0.00340 (mol/L)1/2) under long-term fertilization which K fertilizer was absent.Whatever K fertilizer was applied, the value of potassium buffering capacity (PBC) was observed in the following order: type C soil (48.2～143.8) > type B soil (19.6～48.8) > type A soil (12.7～20.3), and was greater in loess soil and grey desert soil (34.8～48.8) than that in fluvo-aquic soil and paddy soil (19.6～29.1), and had larger increase in type C soil (44%～50%) and little change in other soils compared with the treatments which K fertilizer was applied.The content of total K had little change, whereas that of the slow-available K and available K decreased, especially for type A soil under long-term fertilization which K fertilizer was absent. The K depletion of type A soil and type C soil resulted in the degradation of hydromica and formation of minerals with swelling structures. The K fixation capacity increased due to the changes of the K⁺ saturation, CEC and < 0.002 mm clay content in soils as well as the changes of the component of K-bearing minerals. The release rate constant of nonexchangeable K and K supplying intensity decreased, and free energies of exchange (-â–³G) for the replacement of Ca+Mg by K increased due to the changes of boiling 1 mol/L HNO₃ extractable K and CEC content in soils. Therefore, the soil K contribution index decreased, and the agronomic efficiency of K fertilizer increased. Long-term K fertilization or NPK fertilizer in combination with manure could arreste or decrease the changes mentioned above. The content of slow-available K and available K decreased slightly, and other changes mentioned above were smaller in type B soil. The auther put forward his suggestions for the further study based on different long-term experiments.