A procedure using 32P isotopic exchange kinetics was suggested to characterize the availability of soil P. The ratio of the radioactivity remaining in the solution over the total radioactivity (r1/R) was closely and inversely proportional to the maximum P-fixation capacity (M) estimated by the Langmuir isotherm (r = 0.82**, significant at the 0.01 probability level) and to the P-buffering capacity (Mb) (r = 0.71**). The 1/r1/R) ratio was also correlated with oxalate-Al (r = 0.69**) and oxalate-Fe (r = 0.35**). The kinetic parameter, n, described as the decrease of radioactivity remaining in the solution with time, was related to the soil pH (r = −0.52**), the prevailing form of soil P (Mehlich ratio Q, r = −0.50**), the P-fixation capacity (r = 0.44**), and organic matter (r = 0.44**). The isotopically exchangeable P (E1 value) was closely related to the plant P uptake on soils having low P-adsorption capacity (Group III, r = 0.90**). However, E1 value may overestimate the available P on high P-fixing soils especially for acid Spodosols (Group I, r = 0.47**). The water soluble P (M1) or the labile P pool (M1 + M2) showed a curvilinear (logarithmic) relationship with plant P uptake and the correlation coefficient was higher than that obtained from E1 when grouping all soils together. The effect of soil properties on the relation between plant P uptake and E1, log M1, or log (M1 + M2) was confirmed by the increase of the coefficient of determination (R2) of the multiple regression including soil parameters as second variable. An exponential relationship was observed between relative yield (%) and E1, M1, (M1 + M2); the R2 value was 50.4, 65.1, and 68.1%, respectively. The three classes of soil P fertility (poor, medium, rich) were established for E1, M1, and (M1 + M2) according to the Cate and Nelson procedure.
Joint contribution from Service de Recherche en Sols, Ministère de l'Agriculture, des Pecheries et de l'Alimentation du Québec, and from Service de Radio-Agronomie, Centre d'Etude Nucléaire de Cadarache, France.