An experiment was carried out under greenhouse conditions at the Soil Science Department of the Universidade Federal de Lavras, Minas Gerais, Brazil, to determine critical levels of different phophorus fractions and acid phosphatase activity in bean plants (Phaseolus vulgaris L. cv. Carioca-MG); the relationship between these fractions and the acid phosphatase activity and grain production, and the relationship between acid phosphatase activity determined "in vivo" and "in vitro". The plants were cultivated on four lowland soils of southern Minas Gerais, Brazil and subjected to five phosphorus doses, with and without lime application. The experimental design was completely randomized, in a factorial scheme 4 x 5 x 2 with four replications, as follows: four soils [Humic Gley (GH), Bog Soil (O), Alluvial (A) and Low Humic Gley (GP)]; five P doses (75, 150, 300, 500 and 800 mg dm-3); and two levels of liming application (with and without lime). Each plot consisted of a three dm-3 pot where three bean plants were grown. One of the plants was harvested at flowering with its leaf concentrations of total P (Pt), total P soluble in acid (Pts) and organic (Po) and inorganic (Pi) P soluble in acid being evaluated. The other two plants were harvested at the end of the cycle, with their grain production evaluated. The critical levels of P fractions in the bean plants cultivated, in the GH, O, A and GP soils without liming, were, respectively: Pt = 2.3, 2.2, 2.0 and 1.3 mg g-1; Pts = 110, 67, 78 and 63; Po = 32, 19, 20 and 21 and Pi = 77, 47, 58 and 43 µg g-1. With liming, the critical P fraction levels were: Pt = 2.8, 3.0, 3.5, and 4.2 mg g-1; Pts = 105, 96, 130 and 154; Po = 47, 38, 42 and 52 and Pi = 59, 58, 88 and 103 µg g-1. Acid phosphatase activity was not a good characteristic of plant nutrition status under conditions of severe deficiency of P.
critical levels; plant nutrition; enzyme activity