الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 160 |  |  |
| | | from | 160 | | |
Abstract
Sustainable soil management with appropriate understanding of soil characteristics makes a difference in maintaining or improving the fertility level on one side and avoiding soil degradation, on the other sidethis could be a worldwide issue of significance. Therefore, the objectives of undertaking the present study are 1) the ability to characterize the spatial variability across a soil of the selected area using GIS technique and 2) Identification site-specific management (SMZs) using principal components analysis (PCA) and cluster analysis. A total of 120 geo-referenced representative soil samples (from 0 to 0.60 m depth) were collected from Sahl Al-Hussainiyah, El-Sharkia Governorate, Egypt. The collected soil samples were prepared and analyzed for soil physiochemical properties such as soil pH, electrical conductivity (ECe), soluble ions, exchangeable cations, soil cations exchange capacity (CEC), calcium carbonate (CaCO3), soil organic matter (OM), available nitrogen (N), available phosphorus (P), available potassium (K), soil particles analysis, bulk density (BD) and hydraulic conductivity (k). Semi-variogram analysis was used to evaluate the spatial distribution pattern of each soil property. Interpolation mapping was carried out using ordinary kriging method.
Spatial distribution pattern of varies from moderate to strong spatial dependence for most of the soil characteristics using semi-variogram analysis and ordinary kriging. The results showed that the best fit semivariogram model of ECe, S.Na+, S.Mg2+,Cl-, Ex.Na+, Ex.Ca2+, Ex.Mg2+, ESP, Ava. K, clay, silt, sand, BD and porosity was “stable model”. Whereas K-Bessel model was the best fit model ofpH, CEC and OM. Meanwhile, the best fit semivariogram model ofSO42-, Ex.K+, CaCO3, ava. N and ava. P was “Exponential model”. Tetraspherical, spherical and Circular models were the best fit model for S.Ca2+, K+ and HCO3-, respectively.The results showed five zones for each soil attribute.
Some important soil properties (i.e. pH, ECe, CaCO3, organic matter, available nutrients (N, P and K), CEC, clay and bulk density) were selected to delineate site-specific management zones (MZs) of study area using principal components analysis (PCA) and agglomerative hierarchical clustering (AHC). Four principal components (PCs) were Identified and they were able to explain 73.2% of the variability, where the PC1, PC2, PC3, and PC4 are able to explain 26.4, 19.8, 14.4 and 12.7% of the total variance, respectively. These PCs were used for further cluster analysis. Four MZs were identified based on AHC. As follow MZ1, MZ2, MZ3, and MZ4 were 3.01, 36.5, 31.0 and 29.5% of the total study area with statistically significant differences between the studied attributes of the different MZs (p < 0.05).Therefore, the methodology of current study can be used in delineation SMZs and the study area should not be treated as a one homogeneous unit, due to the great variation from one zone to another, beside the benefical effects on lowreine the cost of soil management considerably.