The fractionation of inorganic phosphorus (Pi) is important to evaluate in which forms phosphorus is sorbed to soil, because the efficiency of an extractant of available P in soils depends on its capacity to extract each fraction of soil Pi. This study evaluated the contents of Pi fractions in representative soils of the State of Paraiba and their relationship with the chemical and physical characteristics of these soils, as well as the correlation between the Pi fractions and P extracted from soils by maize plants and Mehlich-1, Mehlich-3, Bray-1, and by ion exchange resin extractors. The experiment was conducted in a greenhouse in a randomized block design and tested 12 soils, with and without P application, in three replications. After the application of P doses (75, 88 and 103 mg dm-3) in 3.0 dm-3, soil samples were incubated in polyethylene pots for three weeks with distilled water (50% of total porosity of each soil). Plant-available P was determined by four chemical extractants and the fractionation of Pi performed in sub samples of 0.2 dm-3 per pot. The fractions P-Al and P-Fe were predominant in highly weathered soils and in less weathered soils with lower pH and Ca2+, while P-Ca was preferentially found in alkaline soils with very high Ca2+ contents. Phosphorus extracted by maize plants as well as Mehlich-1, ion exchange resin and Bray-1 were found in P-H2O and P-Al fractions, while P extracted by Mehlich-3 was more correlated with P-Ca fraction. P extracted by Mehlich-1 and Bray-1 was more correlated with P absorbed by plants, while P extracted by Bray-1 was low in soils with high Ca2+ contents and high pH. In general, Mehlich-1 quantified P similar to ion exchange resin, even in Ca2+ -rich soils with high pH values.
P fractionation; phosphorus availability; semi-arid soils
