Experimental and statistical study of saturated hydraulic conductivity and relations with other soil properties of a desert soil

Abstract

Agricultural expansion in the Kharga Oasis, in the western desert of Egypt, depends strongly on irrigation. Soil hydraulic conductivity is therefore a key property for reclaiming desert land and planning irrigation schemes. Soil samples collected at 10‐m intervals in a 120 m by 120 m plot were analysed for hydraulic conductivity together with 12 other basic physical and chemical soil properties. The resulting data were analysed statistically using Pearson correlation, principal component analysis and linear regression. The hydraulic conductivity values varied over four orders of magnitude and, because of the saline–sodic nature of the soil, were about two orders of magnitude smaller than what is generally reported in the literature for similar soil textures. Results showed that the hydraulic conductivity was correlated significantly with soil variables that relate to soil structure, such as wilting point, field capacity and SAR, and less to variables that relate to soil texture, such as silt and clay fractions. Pedotransfer functions for hydraulic conductivity were derived by stepwise multiple linear regression and fitted by residual maximum likelihood. The first model was based directly on soil properties, whereas the second model was based on principal components. Both models showed that part of the variation in hydraulic conductivity encountered in the field could be explained, but with large uncertainty probably resulting from sampling and measurement errors, randomness and soil heterogeneity, or other soil properties that were not observed. The most significant predictors for the first model were wilting point, SAR and silt fraction. The second model used the first two principal components of the soil variables as predictors, the first one related to soil structure and the second to soil texture.HighlightsWe analysed relations between saturated hydraulic conductivity and other basic soil properties. Principal component analysis showed that the variation in hydraulic conductivity is related to soil structure. A pedotransfer function predicted hydraulic conductivity from wilting point, sodium adsorption ratio and silt fraction. A pedotransfer function was derived from principal components one and two related to structure and texture, respectively.