This study quantified spatial variability in soil fertility attributes to delineate management zones (MZs) for site-specific nutrient management (SSNM) in a 4-ha maize field in northern Telangana, India. A total of 200 geo-referenced surface (0–15 cm) soil samples were analyzed for pH, electrical conductivity, organic carbon, and available nutrients (e.g., P, K, S, Fe, Mn, Zn, and Cu). Geostatistical analysis using ordinary kriging revealed that spherical models best were the best fit for describing the spatial structure of most parameters, with strong spatial dependence (nugget/sill < 0.25). Principal Component Analysis (PCA) reduced dimensionality, and fuzzy C-means clustering of the principal components delineated three distinct MZs, which were validated by ANOVA. Integration of MZs with targeted yield-based fertilizer recommendation equations enabled differential NPK application, resulting nutrient use efficiency gain equivalent to savings of up to 36 kg N, 39 kg P₂O₅ and 31 kg K₂O ha⁻ ¹ in MZ -3. The maize yield increased from 7.27 t ha ⁻¹ under conventional farmer practices to 7.79 t ha ⁻¹ in MZ -1, 7.93 t ha ⁻¹ in MZ-2 and 8.02 t ha ⁻¹ in MZ -3 with corresponding benefit–cost ratio of 2.54, 2.60 and 2.65. MZ-3 consistently outperformed other zones in yield and economic return, demonstrating the agronomic and economic efficiency of site-specific nutrient management. This work demonstrates the potential of combining geostatistics and fuzzy clustering for optimal nutrient use efficiency and profitability in smallholder maize-based agroecosystems.

Identification and demarcation of management zones (MZs) are required to maximize profit, reduce environmental damage, and enhance soil and crop health. The management zone with uniform spatial homogeneity and production potential may solve the problem of sustainable soil nutrient management. Hence, this current investigation was carried out to evaluate variations in soil parameters in order to demarcate the soil fertility zone for site-specific nutrient management. Overall, 200 soil samples (0-15 cm depth) with geographical coordinate were collected with a grid size of 14.2 m × 14.2 m from 4 ha maize cultivated area of Bogumpadu village ofEllanthakunthamandal, Karimnagar district, Telangana, India. The collected samples were tested with different reagents to know the soil reaction and available nutrients in soil. The geostatistical technique was implemented to assess nutrient variability and preparation of variability map. The spatial variability of soil properties was explained by different models whereas spherical, exponential, and Gaussian models identified as the best-fitted models. Furthermore, the management zone was delineated by principal component analysis and fuzzy C-means clustering algorithm. Five PCs with eigenvalue >1, explaining 99.98 % of variation in overall variation were selected for the next statistics. Three management zone were identified by using the fuzzy performance index (FPI) and normalized classification entropy (NCE). The management zone significantly differs from each other. MZ-1 covers an area of 45.5 %, followed by MZ-2 (29.5 %) and MZ-3 (25%). To evaluate the management zone productivity, site-specific nutrient management experiment was conducted in the maize field. The different fertilizer doses were calculated for three management zones by the Soil test crop response model and compared with farmer fertilizer practices. The result showed the highest maize grain yield in MZ-3 (80. 18 q ha^-1) followed by MZ -2 (79.25 q ha^-1) and MZ-1 (77.89 q ha^-1) and the lowest grain yield with farmer fertilizer practice (72.69 q ha^-1). The highest N, P₂O₅and K₂O fertilizer saving was observed in MZ-3 followed by MZ -2 and MZ-1 compared to farmer fertilizer practices. This study concluded that the management zone concept reduced the application of fertilizer, reduced environmental pollution and increased the maize grain yield and profit.

Spatially explicit soil fertility diagnosis is still difficult in data sparse paddy cultivated areas because single soil test indicators cannot separate soil chemical condition, primary nutrient supply, micronutrient balance and yield relevance. This study developed a soil-test driven, domain-balanced Soil Fertility Index (SFI) framework using 64 georeferenced long-term paddy-field samples from Hanumakonda district, Telangana, India. Ten laboratory indicators (pH, electrical conductivity, organic carbon, available nitrogen, phosphorus, potassium, zinc, iron, copper, and manganese) were transformed into agronomic scores and aggregated into Chemical Condition, Macronutrient, Micronutrient, and Overall Soil Fertility indices. Regularized regression models were used to approximate the derived fertility-index targets from the original soil-test variables and to identify the conditional contribution of individual indicators within each index formulation. Ridge, Lasso and ElasticNet regression models were evaluated using five-fold cross-validation within the training dataset ( n = 51) and independent test-set evaluation ( n = 13). Among the evaluated models, Lasso provided the best prediction of Overall SFI (R ² = 0.594, RMSE = 0.033, RPD = 1.63), whereas ElasticNet performed best for the Micronutrient Index (R ² = 0.683, RPD = 1.85) and showed stable prediction of the Macronutrient Index (R ² = 0.577, RPD = 1.60) Relative zone mapping showed that low to moderate priority classes dominated the predicted overall SFI surface, while the Micronutrient Index produced the sharpest local management contrast. The predicted fertility surfaces were mapped across approximately 140,387 ha within the empirical point-covered sampling domain using a 30 m spatial analysis grid. Independent yield validation supported the agronomic relevance of the framework. Available nitrogen and organic carbon were strongly correlated with yield. Yield validation showed that predicted MNI relative zones significantly separated productivity classes, with mean yield increasing from 22.52 to 26.81 q acre ⁻¹ (Kruskal-Wallis H = 17.52, p < 0.001), whereas Overall SFI, MiNI and SFI 5core zones did not show significant yield separation. The proposed framework provides an interpretable and reproducible decision-support approach for prioritizing soil testing and site-specific nutrient management in data-sparse paddy landscapes.