Defining nutrient management zones (MZs) is crucial for the implementation of site-specific management. The determination of MZs is based on several factors, including crop, soil, climate, and terrain characteristics. This study aims to delineate MZs by means of geostatistical and fuzzy clustering algorithms considering remotely sensed and laboratory data and, subsequently, to compare the zone maps in the north-eastern Himalayan region of India. For this study, 896 grid-wise representative soil samples (0–25 cm depth) were collected from the study area (1615 km²). The soils were analysed for soil reaction (pH), soil organic carbon and available macro (N, P and K) and micronutrients (Fe, Mn, Zn and Cu). The predicted soil maps were developed using regression kriging, where 28 digital elevation model-derived terrain attributes and two vegetation derivatives were used as environmental covariates. The coefficient of determination (R²) and root mean square error were used to evaluate the model’s performance. The predicted soil parameters were accurate, and regression kriging identified the highest variability for the majority of the soil variables. Further, to define the management zones, the geographically weighted principal component analysis and possibilistic fuzzy c-means clustering method were employed, based on which the optimum clusters were identified by employing fuzzy performance index and normalized classification entropy. The management zones were constructed considering the total pixel points of 30 m spatial resolution (17, 86,985 data points). The area was divided into four distinct zones, which could be differently managed. MZ 1 covers the maximum (43.3%), followed by MZ 2 (29.4%), MZ 3 (27.0%) and MZ 4 (0.3%). The MZs map thus would not only serve as a guide for judicious location-specific nutrient management, but would also help the policymakers to bring sustainable changes in the north-eastern Himalayan region of India.

Soil Organic Carbon (SOC) is a crucial indicator of ecosystem health and soil quality. Machine learning (ML) models that predict soil quality based on environmental parameters are becoming more prevalent. However, studies have yet to examine how well each ML technique performs when predicting and mapping SOC, particularly at high spatial resolutions. Model predictors include topographic variables generated from SRTM DEM; vegetation and soil indices derived from Landsat satellite images predict SOC for the Lakhimpur district of the upper Brahmaputra Valley of Assam, India. Four ML models, Random Forest (RF), Cubist, Extreme Gradient Boosting (XGBoost), and Support Vector Machine (SVM), were utilized to predict SOC for the top layer of soil (0–15 cm) at a 30 m resolution. The results showed that the descriptive statistics of the calibration and validation sets were close enough to the total set data and calibration dataset, representing the complete samples. The measured SOC content varied from 0.10 to 1.85%. The RF model’s performance was optimal in the calibration and validation sets (R²c = 0.966, RMSE_c = 0.159%, R²v = 0.418, RMSE_v = 0.377%). The SVM model, on the other hand, had the next-lowest accuracy, explaining 47% of the variation (R²c = 0.471, RMSEc = 0.293, R²v = 0.081, RMSEv = 0.452), while the Cubist model fared the poorest in both the calibration and validation sets. The most-critical variable in the RF model for predicting SOC was elevation, followed by MAT and MRVBF. The essential variables for the Cubist model were slope, TRI, MAT, and Band4. AP and LS were the most-essential factors in the XGBoost and SVM models. The predicted OC ranged from 0.44 to 1.35%, 0.031 to 1.61%, 0.035 to 1.71%, and 0.47 to 1.36% in the RF, Cubist, XGBoost, and SVM models, respectively. Compared with different ML models, RF was optimal (high accuracy and low uncertainty) for predicting SOC in the investigated region. According to the present modeling results, SOC may be determined simply and accurately. In general, the high-resolution maps might be helpful for decision-makers, stakeholders, and applicants in sericultural management practices towards precision sericulture.

Changes in land use have several impacts on soil organic carbon (C) and nitrogen (N) cycling, both of which are important for soil stability and fertility. Initially, the study area was barren uncultivated desert land. During the late 1960s, the introduction of a canal in the arid region converted the barren deserts into cultivated land. The objectives of the present study were to evaluate the effects of various land use systems on temporal changes in soil organic C and N pools, and to evaluate the usefulness of different C and N management indices for suitable and sustainable land use systems under arid conditions. We quantified soil organic C and N pools in five different land uses of the Ghaggar flood plains, in hot, arid Rajasthan, India. The study focused on five land use systems: uncultivated, agroforestry, citrus orchard, rice–wheat, and forage crop. These land use systems are ≥20 years old. Our results showed that total organic carbon (TOC) was highest (7.20 g kg⁻¹) in the forage crop and lowest in uncultivated land (3.10 g kg⁻¹), and it decreased with depth. Across different land uses, the very labile carbon (VLC) fraction varied from 36.11 to 42.74% of TOC. In comparison to the uncultivated system, forage cropping, rice–wheat, citrus orchard, and agroforestry systems increased active carbon by 103%, 68.3%, 42.5%, and 30.6%, respectively. Changes in management and land use are more likely to affect the VLC. In soil under the forage crop, there was a considerable improvement in total N, labile N, and mineral N. Lability index of C (LIC), carbon management index (CMI), and TOC/clay indices were more sensitive to distinguishing land uses. The highest value of CMI was observed in the forage crop system followed by rice–wheat and agroforestry. In the long term, adoption of the forage crop increased soil quality in the hot, arid desert environment by enhancing CMI and VLC, which are the useful parameters for assessing the capacity of land use systems to promote soil quality.

Numerous applications in agriculture, climate, ecology, hydrology, and the environment are severely constrained by the lack of detailed information on soil texture. The purpose of this study was to predict soil particle-size fractions (PSF) in the Ri-Bhoi district of Meghalaya state, India, using a random forest model (RF). For the modeling of soil particle-size fractions, we employed 95 soil profiles (456 depth-wise layers) gathered from a recent national land resource inventory as well as currently accessible environmental variables. Sand, silt, and clay content were predicted using the Random Forest model at varied depths of 0–5, 5–15, 30–60, 60–100, and 100–200 cm. Our results showed the R² for sand was found to be 0.30 (0–5 cm), 0.28 (5–15 cm), and 0.21 (15–30 cm). For the sand, silt, and clay fractions, respectively, the concordance correlation coefficient (CCC) was found to be greater in the 0–30 cm, 0–60 cm, and 0–15 cm depths. When there is a reasonably close monitoring of the coverage probability with a confidence level along the 1:1 line, prediction interval coverage probability (PICP) gives a decent indicator of what to anticipate. The most crucial variables for the prediction of sand and silt were channel network base level (CNBL) and LS-Factor, whereas Min Temperature of Coldest Month (°C) (BIO6) was discovered for clay prediction. For all three soil texture fractions, the range between the 5% lower and 95% higher prediction bounds was large, indicating that the existing spatial predictions may be improved. The maps of soil texture were significantly more precise, and they accurately depicted the spatial variations of particle-size fractions. Additionally, there is still a need to investigate novel methodologies for extensive digital soil mapping, which will be very advantageous for many international initiatives.