The objective of this work was to study the effect of four rates and four sources of P on dry matter production, and of phosphorus content in alfalfa and centrosema grown in pots containing a Red-Yellow Latosol (Oxisol), and to evaluate soil phosphorus by three extractants. The treatments were completely randomized in a 4 x 4 x 6 factorial design, with three replicates. Four sources of P were used at the rates of 0, 50, 100 and 200 mg kg-1, supplied as triple superphosphate (TSP), Arad rock phosphate (ARP), North Caroline rock phosphate (NCRP) and Yoorin magnesium thermophosphate (YTP), and six monthly harvests were made. The various rates of phosphorus applied increased both dry matter and P content in dry matter, alfalfa being more responsive. YTP ranked first in the total dry mater production: in six cuttings 50.17 g and 70.89 g were obtained, respectively with alfalfa and centrosema when 200 mg kg-1 was used. The extractants Mehlich 1 and 3 and resin presented each other a high correlation in the evaluation of the available phosphorus. Considering the average of the three rates of application and all cuttings, total P taken up obeyed the following decreasing order: YTP > TSP > NCRP > ARP.
Centrosema pubescens; Medicago sativa; dry matter; phosphates; analytical methods
