Phosphorus adsorption and its relation with soil properties in acid soils of Western Kenya

Abstract

Low available phosphorus (P) is one of the major hindrances to crop production in acid soils of western Kenya. Although considerable work has been done to establish P levels in the region, there is paucity of information on which to base fertilizer recommendations due to potential crop production differences caused by different soil types and climate. Phosphorus adsorption capacity and its relationship with some soil properties were evaluated in acid soils from nine locations of western Kenya. Adsorption data was obtained by equilibrating the nine soil samples with 30ml of KH2PO4 in 0.01 M CaCl2, containing 0, 80, 150 and 300 µg ml-1 for 48 hours with shaking for 30 minutes at intervals of 8 hours. Langmuir, Freundlich and Tempkin adsorption models were fitted to the test results and relationship between P adsorption and soil properties determined by correlations. The result of this study showed that the soils were strongly to extremely acidic (pH 4.83 - 3.76), had high exchangeable Al3+ (>2 cmol Al kg-1), Al saturation of (> 20% Al) and calculated maximal phosphorous adsorption varied from 770.83 to 1795.83 mg kg-1 soil. Comparing the models, Freundlich linear model showed a better fit to the tested soils compared to Langmuir and Tempkin models. The regression coefficients (R2) for the fitted Freundlich P adsorption isotherms was highly significant ranging from (0.995- 1.000) for all tested soils. Analysis of relationship between adsorption maximum and soil attributes revealed that adsorption maximum positively correlated with clay content, exchangeable P, exchangeable acidity and Aluminium saturation and negatively correlation with organic matter and electrical conductivity. Due to differences in maximal P sorption capacities within the tested area, blanket P fertilizer recommendation may affect crop productivity in some sites. There is therefore need for further research to determine optimal Phosphorus requirements for soils in each research site.