Comparison of phosphorus soil test extractants for plant availability and environmental assessment

A number of soil tests have been proposed to predict crop response to added P or to assess potential for soil P loss to runoff waters. A series of four separate experiments were conducted over a 10-yr period to evaluate soil test methods on a total of 163 Vermont and New York field soils. The experiments included the following: (i) a pot study with alfalfa grown in the greenhouse with 31 soils either unfertilized or fertilized with 18 mg P kg(-1); (ii) routine chemical analysis on 54 soils; (iii) a 360-d incubation study with 24 soils receiving either 0, 20, or 40 mg P kg(-1) as CaH(2)PO(4), in which soils were analyzed for desorption and adsorption and the equilibrium P concentration (EPC(O); and (iv) another set of 54 agricultural soils incubated with 0 or 40 mg P kg(-1) and analyzed for CaCl(2), distilled water, and ammonium acetate (Vermont 1)-extractable P (VT1P) and EPC(0). Although P extracted by VT1 was significantly correlated with P removed by F extractants, it was better correlated with the ratio of F-extractable P/Al extracted by either acetate or F. Phosphorus additions increased VT1P, as well as P extracted by acetate + F (Vermont 2 [VT2]), and they decreased reactive soil Al (VT1Al) and P adsorption. The amount of P needed to increase VT1P by a certain amount was directly related to the amount of Al in the VT1 extract. Phosphorus availability to plants, CaCl(2)-extractable P, and the EPC(0) were all more closely related to VT1P than P extracted by solutions containing F, such as Mehlich 3 (M3), Bray and Kurtz 1 (BK1), and VT2. In a number of instances the ratio VT2P/VT1Al had a better relationship with CaCl(2)P and EPC(0) than did VT1P. Thus, the fraction of reactive Al that has reacted with P (as estimated by VT1P or the ratio of VT2P/VT1Al) appears to be a better indicator of P availability and potential P desorption to runoff water than is P extracted with F.