Experiments were conducted to study the desorption characteristics and plant-availability of phosphate sorbed by some important variable-charge minerals including kaolinite, goethite and amorphous A1 oxide. Phosphate desorption from the complexes of goethite-P, kaolinite-P and A1 oxide-P by equilibration with 0.02 M KC1, resin or some commonly used chemical extractants was slow compared to desorption from a permanent-charge mineral (montmorillonite). However, rice plants were not observed under P deficiency in a pot trial with a phosphate-mineral complex as the only P source for both the permanent-charge mineral and the variable-charge minerals at either 50% or 100% sorption saturation with the exception of goethite-P at 50% saturation. In the exceptional goethite-P treatment, plant P concentration (1.0 g kg-l) was on the threshold of P deficiency. From 15% to 31% of the applied P was recovered by the plants within a growing period of three months, depending on sorption saturation and mineral type. Both the dry matter yield and P uptake decreased with decreasing sorption saturation for all the tested complexes except for A1 oxide-P100 (100% saturation). In the case of A1 oxide-P10o, A1 toxicity may have occurred, for poor root growth and high A1 concentration in the plants were observed. The effect of sorption saturation on the yield and P uptake of plant was obvious for kaolinite and goethite but not very significant for montmorillonite. Based on the recovery of applied P, the plant-availability decreased in the following order: kaolinite-P10o > goethite-P10o > A1 oxide-Ps0 > montmorillonite-Pl0o > montmorillonite-Ps0 > kaolinite-P50 > goethite-Ps0. Fractionation of the sorbed P before and after plant uptake showed that most of the P uptake originated from the resin-exchangeable P fraction in montmorillonite-P complex, but came mainly from NaOH-extractable fractions in goethite-P complex, whereas all the resin-P, NaHCO3-P and NaOH-P fractions in kaolinite-and amorphous A1 oxide-P complex made a contribution to P uptake.

An isotopic exchange method was used to characterize quantitatively the fixation and plant availability of phosphate previously sorbed by soils. In general, the exchangeability of the sorbed phosphate was much higher than its desorbability for both soils and clay minerals. Isotopic exchangeability of the sorbed phosphate increased with sorption saturation during the initial stage (1 5-60% saturation), but the increase was less with increasing saturation from 6&90% for all soils tested. Therefore a sorption saturation of 60% was recommended as the upper limit of P fertilization in terms of economical efficiency. For clay minerals, with increasing sorption saturation, the isotopic exchangeability of the sorbed P increased significantly for kaolinite and sesquioxides, but decreased for montmorillonite. Most of the phosphate sorbed by montmorillonite and kaolinite was found to be isotopically exchangeable, but only a small amount of the P sorbed by goethite could be exchanged. The P sorbed by Al oxide exhibited isotopic exchangeability between that of kaolinite and Fe oxide. The isotopically exchangeable phosphate pool could readily account for the P uptake of plants and the available P determined by some commonly used chemical methods, such as Olsen-P and Bray-P.

No abstract