Tagore, G.S.Sethy, Sumit Kumar2017-02-272017-02-272016http://krishikosh.egranth.ac.in/handle/1/5810002659ABSTRACT Soil fertility fluctuates throughout the growing season each year due to alteration in the quantity and availability of mineral nutrients. The fertility of soil depends on types of soils, nature of irrigation facilities, pH, organic matter content and addition of fertilizer. The evaluation of soil fertility includes the measurement of available macro and micronutrients are essential to maintain ecosystems and high crop yields. Hence, evaluation of fertility status of the soils of an area or a region is an important aspect in the context of sustainable agriculture. Soils are inherently heterogeneous in nature because of the many factors that, contribute to their formation and the complex interactions of these factors. Soil heterogeneities may arise from management activities and can occur from land use and management strategies. Small-scale farmers, who cultivate the lands, are often unable to afford very costly fertilizers and instead, apply animal manure as per their availability. The differential allocation of manure or other source of nutrients creates a nutrient gradient across the cultivated lands. The resulting spatial variability of soil fertility poses great challenge to land management and reflects in variable yields over farmlands. Traditionally, researchers have attempted to remove spatial variability by blocking and/or statistical averaging procedures. However, these attempts have often resulted in the failure to understand the spatial interdependence of the soil properties. Geo-statistics has been used extensively to characterize the spatial variability of soil attributes due to its ability of quantifying and reducing sampling uncertainties and minimizing investigation costs. Multivariate classification by cluster analysis enables the identification of sub- region in the fields that internally have similar characteristics. Currently, use of zone management technique emerged as the most popular approach to manage spatial variability within agricultural fields. In this technique, the field is subdivided into different zones that have relatively homogeneous attributes in landscape and soil conditions and can be used for direct variable rate fertilizer application. Keeping above in view, the present investigation entitled “Appraisal of Spatial Variability of Soil Properties for Identification of Management Zones in Cropped Areas of Anuppur district using Geo- spatial Techniques” was carried out under AICRP on MSN at Department of Soil Science and Agricultural Chemistry, College of Agriculture, JNKVV, Jabalpur (M.P.) during 2015-16. Geographically, Anuppur district lies in between 220 70’-230 25’ North latitude and 810 10’-820 10’ East longitude with an area of 3701 km2. Administratively, the district divided into four blocks, Pushparajgarh, Kotma, Anuppur and Jaithari. The sites decided randomly distributed over agricultural areas by considering land use and heterogeneity of the soil types. From the area of interest, a total of 283 surface soil samples were collected using GPS. These soil samples were analyzed for physico-chemical properties, available N, P, K, S and micronutrients. After data arrangement, spatial variability maps of soil properties were generated using geo-statistical tool in Arc GIS 10.2 software. Correlations between variables, one-way ANOVA test for comparison mean of groups. PCA to reduce the dimensionality of a data set carried out using SPSS 16.0 software and FuzME software was used for delineation of management zones. Fuzzy k-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). Result obtained from present study using appropriate methodology summarized below: In the district as a whole, the mean bulk density and CEC of soil were found 1.58 Mg m-3 and 19.33 meq/100 g soil, respectively and the soils were found slightly acidic in reaction, safe in electrical conductivity, low to medium in organic carbon content and non-calcareous in nature. The deficiency of macronutrients i.e. N, P, K and S were recorded by 65.02%, 32.16%, 33.57% and 70.67% soil samples, respectively. Micronutrients mainly DTPA-Zn (54.06%) and HWS B (9.89%) were found deficient while DTPA-Fe, Cu and Mn content sufficient in whole district. The exponential model was best fitted for pH, CaCO3, AN, ln(AP), ln(Zn) and ln(B) and spherical model for ln(EC), OC, ln(Fe), yield, sand, ln(silt) and clay. However, the circular model was fitted for ln(AK), Cu, Mn, bulk density and CEC while Gaussian model fitted only for ln(S). The NS ratios of variogram models for AK, B, clay and CEC falls between 9 to 16%, which exhibit strong spatial dependency. The pH of soil had significant positive relation with N and K, while significant negative relationship with micronutrients i.e. Zn, Cu, Mn & Fe and B in soil. In addition, the EC had positive and significant relationship with N, P, K, S and B. The OC of soil showed significant and positive relation with available N and Mn. The colour parameter, hue had significant positive relationship with B content. The value had significant positively related with Fe and Mn while negatively related with N, P and Zn. The chroma of soil positively related with Fe and B, while significant negatively related with N. Result showed the yield of wheat showed significant positive relationships with available N in soil, HWS B in soil, P in wheat straw, K and S in wheat grain. The Zn content in wheat grain showed significant negative relationship with the pH of soil and Zn in wheat straw showed significant negative relationship with EC and available P in soil. Multivariate analysis results, showed the most important factors governing variation in soil composition were BD, sand, silt, clay & CEC, Zn, Cu, Fe and Mn, OC and N, B, S and EC, pH and CaCO3, P and K, respectively. Result showed that seven management zones were identified by PCA & Fuzzy k-means algorithm in whole district. The ANOVA indicated that significant statistical difference existed among the seven MZs for all properties. Overall, it was concluded that the differentiation into seven management zones may be very useful for farmers to adopt site-specific nutrient management, which satisfies the criteria of management zones to be simple, functional, easy to understand and economically feasible. The mean values of soil nutrients in each zone can be used as a reference for variable- rate fertilization.ennullThesis