Delineation of soil management zones for a rice cultivated area in eastern India using fuzzy clustering

Efficient, cost effective and easy to use tools are needed for site specific management of soils for increasing the crop productivity. Currently nutrient management recommendations for rice in the study are typically uniform without consideration of spatial heterogeneity of nutrient content in soil. This study has emphasized the use of management zones (MZs) as an efficient method for variable application of fertilizers. Rajnagar Block of Kendrapada District in India was selected as the study area for the present study and 225 soil samples were collected on 1.5 kilometer grids. Soil samples were analyzed for pH, electrical conductivity, soil organic carbon, available nitrogen, available phosphorous, available potassium DTPA extractable micronutrients i.e., iron, zinc, copper, and manganese. Spatial variability of these soil properties was analyzed and spatial distribution maps were generated using geostatistics and ordinary kriging technique. Further, principal component analysis and fuzzy c-means clustering algorithm were performed to delineate the management zones based on optimum clusters identified using fuzzy performance index (FPI) and normalized classification entropy (NCE). The results revealed that the optimum number of MZs for this study area was three and there was heterogeneity in soil nutrients in three MZs. The delineated MZs provide a basis of information for site-specific fertilizer management in the rice cultivated fields in the study area. (C) 2015 Elsevier B.V. All rights reserved.