University of Khartoum

Physico-chemical Empirical Models to Predict some Soil Properties of Salt-affected Soils in Two Sites in Khartoum State, Sudan

Physico-chemical Empirical Models to Predict some Soil Properties of Salt-affected Soils in Two Sites in Khartoum State, Sudan

Show full item record

Title: Physico-chemical Empirical Models to Predict some Soil Properties of Salt-affected Soils in Two Sites in Khartoum State, Sudan
Author: Ahmed, Sayda Bushra Omer
Abstract: Many attempts have been made in the last five decades to predict some complex soil properties from some easily measured soil parameters. Pedotransfer functions were used for developing empirical predictive models to estimate the needed properties, and prediction performances were then analyzed. This study was carried out to develop prediction models for some soil parameters, e.g. clay percentage, cation exchange capacity (CEC) and sodium adsorption ratio (SAR). Linear and multiple regression models were used to develop the appropriate pedotransfer functions. In order to evaluate the models, root mean square error and determination coefficient (R2) were used. The associations between the soil hydraulic conductivity and different soil properties were also considered. Soil samples were collected from surface horizons (0-30 cm) of two sites in Khartoum State (west Omdurman and Soba). They were treated with solutions of known sodium adsorption ratios (SARs) and electrical conductivities. The equilibrated soils were dried and classified according to their salinity and sodicity into three classes (non-saline non sodic, salinenon sodic and saline sodic). Independent data sets were used to validate the models; the soil samples were selected from different semi-detailed soil survey reports in Sudan. Regarding the evaluation criteria, the results revealed that in the three salinity- sodicity classes there is a good possibility to predict their clay contents from their saturation percentages (SP). Linear regression models, based on the soil clay percentage, were used to predict soil CEC. Validation results indicated that the models, in aggregate, provide a reasonable estimate of CEC for most sandy clay loam textured soils. The results showed that the use of electrical conductivity of the saturation extract (ECe) to predict SAR is most reliable for non saline non sodic soils of both Omdurman and Soba and can provide a short, simple and economical method to estimate soil SAR. Yet, the model predictions for saline-non sodic soils and saline-sodic soils were poor. The multiple regression models were developed to predict clay percentage with different combination of inputs: (i) SAR, ECe and SP, (ii) SAR, ECe, CaCO3% and SP, and (iii) SAR, ECe, CaCO3%, pH and SP. The models were acceptable for prediction of clay content of soils with exclusion of pH parameter except for non saline non sodic soils of Omdurman, where the pH was efficient in predicting the clay content. Low negative correlation (R2 = 0.43) was obtained between hydraulic conductivity and sodium adsorption ratio in non-saline non-sodic soils of Omdurman, whereas for Soba soils the correlation was high (R2=0.78). For saline non-sodic soils the linear correlation was high for both Omdurman and Soba soils (R2= 0.77 and 0.83, respectively). Saline- sodic soils of Omdurman showed a high correlation (R2=0.89).
URI: http://khartoumspace.uofk.edu/handle/123456789/17420
Date: 2015-12-03


Files in this item

Files Size Format View

This item appears in the following Collection(s)

Show full item record

Share

Search DSpace


Browse

My Account