Phosphate Sorption in Vertisols of Southwestern Spain

Abstract
Low phosphorus availability is a critical impediment to agricultural use of Vertisols. Several studies have explored the P sorption characteristics of Vertisols, but few have investigated the relationship between soil physicochemical and the P sorption parameters calculated from uniform-surface and two-surface Langmuir equations. Phosphate sorption was investigated in surface-samples of 19 Vertisols in southwestern Spain. The phosphate sorption maxima (PAM) deduced from the uniform-surface Langmuir equation ranged from 150 to 2566 mg P kg-1, with a mean value of 1115 mg P kg-1. The calculated sorption maxima (TPAM) for the two-surface Langmuir model ranged from 410 to 15406 mg P kg-1, and the mean value (4296 mg P kg-1) increased about 3 times compared with the basic Langmuir equation. The two-surface Langmuir model fit the sorption data better than the uniform layer model. Therefore, an underestimation of the P-sorption was observed when the traditional Langmuir equation was used. Total surface area was highly correlated with sorption maxima: n = 0.708*** for PAM and 0.900*** for TPAM. For the two-surface equation, correlation analysis suggested that the “active” CaCO3 (ACCE) was the most active sorbent of P on the high energy sites, whereas citrate-dithionite-bicarbonate extractable Fe (Fed) contributed to P sorption on the low energy sites. The prediction of the PAM and TPAM was improved by combining all or several of following soil properties: Fed, ACCE, clay, and total surface area in a multiple-regression analysis. The equations obtained could offer a rapid estimation of P-sorption in Vertisols of southwestern Spain. Moreover, the results obtained from a greenhouse experiment with Agrostis tennuis and two levels of P (0 and 218 mg P kg-1) showed that the amount of available soil P was a positive function of Fed but a negative function for ACCE.