Phosphate adsorption of some acid soils in relation to their properties.

Effect of soil properties on phosphate adsorption was examined on 14 surface (0–0.15 m) soil samples of Eutric Dystrochrepts and Mollic Vdorthents from hilly area of Haryana; Typic Haplequolls, Aquic Hapludolls of U.P.; Typic Hapludalfs, Typic Haplustalfs, Typic Paleohumelts of H.P.; and unclassed soils of Assam. The samples of each soil were equilibrated for 24 hrs at 25±2°C with aqueous solution of KH2PO4 containing (0–500 μg P mL−1) in a 1:20::soil:solution ratio. The amount of P adsorbed increased and distribution coefficient (Kd) and percentage of added phosphate adsorbed (Xad) decreased with increasing solution P concentration. Phosphate adsorption was satisfactorily described by the two surface Langmuir (R2 = 0.97–0.99, P = 0.01) and Freundlich equations (R2 = 0.97–0.99, P = 0.01). The values of b2 (adsorption maxima of lower energy sites) were 2.82 to 8.78 folds as compared to b1 (adsorption maxima of higher energy sites); whereas the values of k1 (bonding energy constants of higher energy sites) were 4.58 to 51.99 folds of k2 (bonding energy constants of lower energy sites). The contribution of high and low energy sites to total adsorption maxima ranged from 10.23 to 26.18 and 73.82 to 89.77%, respectively. Inverse relationship was observed between adsorption maxima and bonding energy constants. The pH was correlated negatively and significantly with constants of Langmuir (b1, b2, k1) and Freundlich-K, while non-significantly with k2 and n. The Al was correlated positively and significantly with b1, k1, k2 and Freundlich-K but non-significantly with b2 and n. The Al + Fe were correlated non-significantly with b2, k1, k2, K and n, while significantly with b1. The Fe and clay had non-significant correlations with all the constants of Langmuir and Freundlich equations. However, when the effect of individual property (e.g. Al, Fe) was compared for soils, not varying in other soil properties, it was found to be highly marked. The multiple regression equations revealed that soil properties (pH, EC, OC, CaCO3, CEC, clay, Al, Fe and available P) jointly accounted for 73.80% variation in b1, 64.30% in b2, 73.0% in k1, 43.0% in k2, 85.30% for Freundlich-K and 26.40% for n in these soils.