REACTIVITY OF PHOSPHORUS IN NORTHEASTERN BRAZILIAN SOILS ASSESSED BY ISOTOPIC DILUTION

To diagnose fertilizer requirements in regions of high soil variability where P-fixing capacities can vary 100-fold, single-extraction P tests must be complemented by an evaluation of P sorption. The isotopic-dilution kinetics of P-32 was tested for its potential to measure available P and to predict the degree of soil-fertilizer P reactivity. Six soils from northeastern Brazil were fertilized with P, incubated (1, 42, and 84 d) and then seeded with pearl millet [Pennisetum glaucum (L.) R. Br.]. Soil samples were analyzed for water-soluble P (P(w)), resin P, P-32 in solution 1 min after addition over P-32 added (R1/R0), rate of P-32 disappearance from solution between 1 and 100 min (n), reversible P (P(rev), determined by double isotopic dilution), mobile P (P(m) = P(w) + P(rev)), and isotopically exchangeable P (E1 and E-value [E] for 1-min and 24-h equilibrium periods, respectively). Use of R1/R0 successfully differentiated soils in terms of net fertilizer effects on P(w) and P(m) with time. We found that P(w) and P(m) were better predictors of plant P uptake than resin P, E1, and E, since the latter quantities overestimated available P in a high-P-fixing Oxisol. Based on this and published data, we propose a critical value of R1/R0 (0.26) below which E values consistently overestimated labile P, while P(w) correlated satisfactorily with available P. A modified technique overcomes limitations of measuring very low concentrations of P(w) in such soils. To diagnose labile P and fertilizer responses in soils with R1/R0 > 0.26, we propose a simplified methodology that avoids the double isotopic dilution.