The replenishment of labile phosphorus (P) from less labile soil P fractions was evaluated after 10 successive resin extractions and incubation during four months. Soil samples from the A, B and Bw horizons of Latosols (Oxisols) and A and Bt horizons of Luvisols were collected from the upper, middle, and lower position in the ladscape of three toposequences of each soil class. The evaluation was based on sequential fractionation of soil P with resin, NaHCO3 (organic and inorganic fractions), NaOH (organic and inorganic fractions), H2SO4, and by a H2SO4/H2O2 digestion, performed before the successive extractions and after incubation. There were no significant differences in P contents in a comparison of the slope positions. The 10 successive resin extractions removed 2 to 5 times more P than the first extraction in the Latosols and 2 to 9 times more in the Luvisols. The last extractions stabilized around 1 mg kg-1 in all horizons of both soil classes. After incubating soil samples, resin P recovered 20-30 % of its initial P value in the A horizons (1.8 and 3.3 mg kg-1) and 50-90 % in the subsuperficial layers (1.1 and 1.2 mg kg-1) of the Latosols and Luvisols, respectively. Changes occurred in all other fractions, being greater in the Luvisols and A horizons, although most of the P remained in the less labile P forms. The most depleted fractions were NaOH-P in the Latosols and H2SO4-P, followed by NaOH-P in the Luvisols, suggesting that they were the most important in replenishing labile P in these semi-arid soils.
toposequence; horizons; bicarbonate-P; hydroxide-P; acid-P; residual-P
