PHOSPHORUS AND IRON MOBILIZATION IN FLOODED SOILS FROM BRAZIL

The dynamics of phosphorus in flooded soils has much agricultural and environmental significance, One specific feature of such dynamics is t hat it is influenced largely by changes in Fe forms. In this work we i nvestigated, in the laboratory, the effects of continuous flooding and 25-day redox cycles on the P and Fe dynamics in subsoils of seven Oxi sols (Orthoxs, Humoxs, and Udoxs) and surface horizons of 19 lowland s oils (Fluvents and Aquepts) from the State of Minas Gerais, Brazil. Th e amounts of Fe and P extracted by acetate/EDTA (Fe-ae, P-ae), citrate at pH 6 (Fe-c,P-c), and citrate/ascorbate at pH 6 (Fe-ca,P-ca), which provided a measurement of Fe forms with decreasing solubility (from s oluble Fe to the more stable ferrihydrites), were determined before an d after the continuous flooding and redox cycles treatments. In most O xisols, the Fe forms were barely affected by continuous flooding or re dox cycles. In the lowland sons, acetate/EDTA-Fe and citrate-Fe, i.e., the most labile Fe forms, were increased by such treatments, particul arly in the soils containing more organic matter. In contrast, citrate /ascorbate-Fe either did not change or it decreased, suggesting that ( i) no significant amounts of crystalline Fe (hydr)oxides were reduced and transformed to more soluble forms and (ii) the acetate/EDTA-Fe and citrate-Fe increases were largely the result of the transformation of some of the less soluble citrate/ascorbate-Fe forms. Continuous flood ing and redox cycles also induced changes in acetate/EDTA-P, citrate-P and citrate/ascorbate-P that were greater in the lowland soils than i n the Oxisols, but there was no consistent increase or decrease in any of these extractable P forms. However, the citrate/ascorbate-P, on av erage, increased, suggesting that some organic P was mineralized. As a result of the aforementioned changes an Fe and P forms, the P/Fe mole ratio, on average, decreased in the more soluble (acetate/EDTA- and c itrate-extractable) and increased in the less soluble (citrate/ascorba te-extractable) Fe forms. In the continuous flooding treatment, the Fe and P concentrations in solution generally increased until days 20 to 50 and then decreased; the P concentration occasionally showed a seco nd maximum after 50 to 170 days of continuous flooding, probably cause d by organic matter mineralization. After 209 days of continuous flood ing, the P concentration in solution was of the same order of magnitud e as that at the beginning of flooding. The average Fe concentration i n solution during the flooding period was related to acetate/EDTA-Fe a nd citrate/ascorbate-Fe. The variable best suited to predict average P concentration in solution was citrate/ascorbate-P. The variance accou nted for by this variable alone (33%) increased significantly (to 66%) when clay and citrate/bicarbonate/dithionite-extractable Fe were incl uded as variables in the regression. This suggests that P concentratio n in solution is controlled by the amount of P released on reduction o f the citrate/ascorbate-extractable Fe forms and by the P-adsorbing pr operties of the crystalline son minerals (kaolinite, gibbsite, goethit e, and hematite). It has been suggested that the acetate/EDTA and citr ate/ascorbate methods will be useful for evaluating the Fe and P avail ability in waterlogged sons and sediments, if there is analysis of org anic matter and phosphate sorption capacity.