Phosphate sorption by reduced and reoxidized rice soils

Abstract

Intact cores of two rice soils of south-eastern Australia were flooded for 146 days to simulate part of a rice-upland crop rotation. The phosphorus sorptivity, and acetate (pH 4.8) and oxalate (pH 3.0) extractable iron levels were monitored for 210 days, before, during and after flooding. The effects of additional organic matter (rice straw) and the growing of rice plants were examined. In non-flooded soils phosphorus sorptivity was mainly influenced by clay content; there were no relationships between extractable iron and phosphorus sorbed. On flooding there were large increases in acetate and oxalate extractable iron and phosphorus sorptivity. Additional organic matter initially increased the rates of formation of extractable iron, and phosphorus sorptivity increased accordingly. However, with prolonged waterlogging, levels of oxalate iron and phosphorus sorptivity reached values dependent on the free iron oxide content of the soils. The growth of rice plants had no significant effects on phosphorus sorptivity or extractable iron at any time of sampling. During oxidation of previously flooded soils, levels of phosphorus sorptivity and oxalate iron decreased rapidly, but did not return to levels occurring before reduction. The results are consistent with domination of phosphorus sorption processes by ferrous hydrous oxides during the flooded (reduced) phase and by poorly crystalline ferric hydrous oxides during the post-flooding reoxidation phase. Agronomic implications of the results are mentioned.