GEOSTATISTICAL EVALUATION, PHOSPHORUS SORPTIONDESORPTION, AND RESPONSE OF BREAD WHEAT (Triticum aestivum) TO LIMING AND PHOSPHORUS IN ACID SOILS OF ARSI ZONE, ETHIOPIA

Abstract

Acid soils research has been one of the main priorities of the government of Ethiopia. However, the pace of research and development efforts to improve the productivity of acid soils was not commensurate with the extent and severity of the problem. This is, especially, true in Arsi-Bale wheat-belt, where soil acidity is a real yield-limiting factor. Cognizant of this fact, the research was conducted with the main objective of generating quantified information and agricultural technologies for improving the management of acid soils and productivity of bread wheat in the highlands of Arsi zone. The research consisted of 4 components, viz., geostatistical analysis of acid soils; phosphorus (P) adsorption-desorption; lime requirement (LR) estimation using models; and evaluation of acid soil management options for bread wheat productivity. In an attempt to geostatistical analysis of acid soils distribution, 165 georeferenced point samples were collected using multistage random sampling method; and analyzed for soil pHw, pH-K, percent organic carbon (% OC), cation exchange capacity (CEC), exchangeable acidity, exchangeable aluminum, and % clay content. Continuous surface maps of the soil physicochemical properties were generated using 4 kriging and 3 deterministic interpolation methods. The soil P adsorptiondesorption properties of Vertisols and Luvisols were assessed using the Langmuir, Freundlich, Freundlich-Langmuir, modified Freundlich and Elovich models. Direct titration of soils using Ca(OH)2 and field experiments were used to develop site specific LR estimation models. The agronomic and economic viability of liming acid soils was evaluated using a field experiment consisting of 5 rates of row applied lime and 3 levels of P2O5 in a factorial design. The results of the first research component revealed that soils with pHw: 4.49 -5.3 covered about 82 % of the total areas, the CEC varied from 2.96 to 37.1 Cmol(+)/kg soils, and exchangeable acidity from 0.1–4.28 Cmol(+)/kg soils. The superior predictor models were Ordinary Kriging for pHw and pH-K; and Universal Kriging for CEC, % OC, exchangeable acidity and exchangeable aluminum. The results of the second research component showed that Luvisols displayed to have higher P sorption maxima, stronger affinity and very low desorption rate than Vertisols. For modeling LRs of soils, pHw was found to be a good predictor variable; and validation of model predictions using asymptotic maximum pH and relative grain yield, correlation analysis and comparison of model coefficients indicated that the models can be used to estimate the LRS of soils for different target pHw. The results of the fourth research component showed that yields and calculated efficiency indices from the highest rate of row applied lime were comparable to the bulk applied one. Combined interpretation of the first 2 components indicated that liming is necessary to improve the availability of applied nutrients and crop productivity on Luvisols. In conclusion, estimation of LRs using models and maps showed that about 65 % of the total area required aglime within the range of 800 – 2000 and 1660 – 3300 kg/ha to bring the soils’ pHw to a target pHw of 5.6 and 5.9, respectively. For resource poor farmers, row application at 600 kg/ha aglime + 46 kg P2O5 is recommended as the economic optimum for areas with initial pHw < 5.1