Evaluating Spatial Variation of Nutrients Content in Soils and Leaves in Vineyards of Khodabande Using Geostatistical Methods

Proper exploitation of available resources in agriculture is possible by examining relationships between plants, soil, and environmental factors. Effective strategy for the development of sustainable agriculture requires maps of the spatial variability of soil nutrients and plant distribution. In this context, the present study is to evaluate the spatial structure with Semi-variogram models in nutrient: P, K, Zn, Cu, Mn, and B in the soil and leaves as well as leaf nitrogen and organic carbon in the soil in vineyards of Khodabande in Zanjan Province. After preparing the Best Semi-variogram model, zoning maps were prepared by using geostatistical methods such as Kriging and Co-Kriging also IDW (powers: 1 to 5) and Kernel model of the interpolation methods. Results showed most of parameters in soil and the leaves in vineyards had high coefficients of variation. Soil content of Mn and P in leaves samples had the highest correlation radius. The nutrients in soil samples had higher average correlation radii than those of leaves. The results of the evaluations by geostatistical and interpolation methods with root mean square error (RMSE) values, mean absolute error (MAE) and efficiency coefficient showed that the Co-Kriging had the best performance for estimating hydraulic properties and Co-Kriging for soils and vineyards leaves nutrients. The best estimates were obtained with Co-Kriging in soil Zn and Kernel method showed similar results for leaves Zn. The results of the spatial variation of soil and leaf nutrients in the area showed that most of the vineyards were deficient in soil organic carbon, phosphorus, iron, and zinc.