Geospatial modelling for optimum management of fertilizer application and environment protection

Soil as an important source, guaranties the plant growth and supplies more than 97% of food need of world. Knowledge of soil spatial variability is important in natural and environment resource management, interpolation and soil sampling design, but requires a considerable amount of geo-referenced data. Soil cation exchange capacity (CEC) is a vital indicator of soil fertility quality and pollutant sequestration capacity. Plants such as rice need to provide their nutrient elements by using fertilizers for much more production in surface unit. For this purpose, it is essential to recognize macro-elements amount in soils and prepare their ideal maps. 247 soil samples were collected from depth 0–30 cm with distance minimum 250 m and maximum 1500 m using a stratified random sampling scheme on central areas of Guilan province located in north of Iran. CEC, total nitrogen, available potassium and phosphorus maps prepared using kriging geostatistical method. Evaluation criteria values of root mean square error (RMSE) and mean absolute error (MAE) derived for potassium 27.84 and 0.106, phosphorus 8.17 and 4.63, total nitrogen 0.059 and 0.025 and CEC 4.06 and 3.09, respectively. Criteria value of RMSE and MAE showed that accuracy of prepared maps was ideal. The fit of the experimental semivariograms to the theoretical models indicated that kriging could successfully interpolate soil variables. Thus, the kriging geostatistical method used on a large scale could accurately evaluate the spatial variability of soil nutrient properties. With regard to soil nitrogen and phosphorus maps, application of more amounts of nitrate and phosphorus fertilizers than their optimum level cause ground waters pollution and environment damages therefore their application must be carried out with high consideration. Potash fertilizers consumption in land with high CEC results its fixation, too. Precise attention to CEC map and on-time fertilizer application can solve this problem. Therefore, accurate notice to different amounts of these parameters in prediction maps, critical and optimum levels can well manage fertilizers application, prevents additional costs to farmer, pollution of ground waters and environment resources.