Introduction: Soil salinization is a critical environmental issue affecting agricultural productivity worldwide, particularly in arid and semi-arid regions. This study focuses on the Xinjiang region of China, specifically the Xiao Haizi and Sha Jingzi irrigation areas, to explore the use of remote sensing technology for surface soil salinity estimation.Methods: Exhaustive and filter-based feature selection methods were employed by integrating soil salinity data measured on the ground with 32 spectral features derived from Landsat 8 OLI remote sensing images. A 5-fold cross-validation method was used to identify feature combinations that resulted in higher R² values. Moreover, the inversion accuracy of soil salinization monitoring models built using different feature combinations was compared across five machine learning algorithms: Support Vector Machine (SVM), XGBoost, Decision Tree (DT), Random Forest (RF), and AdaBoost.Results: The results revealed that: (1) The AdaBoost and DT algorithms demonstrated high efficacy and precision in the prediction of soil salinity, with AdaBoost outperforming other algorithms in the validation set (R² value of 0.892, MAE of 1.558, RMSE of 2.043), and DT showing the best performance in the training set (R² value of 0.917, MAE of 0.838, RMSE of 1.182). (2) Feature combination 3, consisting of Salinity Index 5, Salinity Index 1, and Salinity Index 8, not only effectively extracted soil salinity information but also significantly improved the accuracy and efficiency of model estimations, effectively reflecting the actual situation of soil salinization in the irrigation area.Discussion: This research provides robust methodological support for using remote sensing technology for soil salinity monitoring and management.

In this paper, salinized soils with different degrees of salinity are sampled in Junggar Basin of Xinjiang of China. The X-ray diffraction, transmission electron microscopy, and inductively coupled plasma mass spectrometry are employed to investigate the morphology and distribution characteristics of smectite in salinized soil profiles. In the salinized soil profiles of this region, crystals of smectite are poor where lattice fringes are not parallel. In all soil layers, the content of smectite in the soil increases with the decrease in content of illite, which has demonstrated significant negative correlation (r = 0.79, n = 50, p < 0.01) between illite and smectite. This phenomenon has demonstrated that illite may be transformed into smectite in salinized soils of studied regions. In general, the transformation process of illite to smectite is affected by climate condition. The δ¹⁸O values of secondary carbonate in the 0–10 cm soil layers is higher than that in deep soil layers, which indicates that δ¹⁸O concentrates in surface soil and reflects temperature rise during soil layer formation. The δ¹³C values of secondary carbonate and soil organic matter in 0–10 cm soil layers are higher than that in deep soil layers. It indicates that C₄ plants were the main plants, which reflects that the climate was relatively dry during the formation of the surface soil. Thus, the climate during the surface soil formation is arid, which is not conducive for leaching K⁺ from illite of the 0–10 cm soil to form smectite. As a result, the content of the smectite becomes lowest in the soil surface. In the relative humid condition of deep soil layers, the K⁺ of the illite of the soil would be relative easily leached and more smectite may be formed. Furthermore, the presence of salt in the salinized soil would promote the formation of smectite in Junggar Basin of Xinjiang. A lot of Ca²⁺, Na⁺ and Mg²⁺ in the soil solution of salinized soils would enter into the illite and occupy K⁺ positions. The studied result shows that the amount of smectite would increase with the increase of salt below 10 cm of the soil layer, where the amount of smectite would be significantly correlated with soil electrical conductivity (r = 0.64, n = 39, p < 0.01). In the Junggar Basin in Xinjiang, therefore, the salinized soil below 10 cm would have the necessary water conditions and chemical components for illite transformation to smectite.

Biochar is widely recognized as an effective soil amendment, yet its residual effects on saline soils remain underexplored. This study investigates the residual impact of a single biochar application on soil properties, bacterial community assembly, and cotton growth over three years. A controlled pot experiment was conducted using four biochar treatments − 0%, 1%, 3%, and 5% (w/w) - along with two cotton varieties: salt-sensitive and salt-tolerant. Results showed that biochar reduced soil salinity by 0.80–1.39 g·kg−1 in the second year, with no effects observed in years 1 and 3. By the third year, biochar at 5% increased soil pH by 0.34–0.44 units. Biochar application enhanced soil organic carbon by 135.15%–763.19%, total nitrogen by 12.88%–241.92%, and available potassium by 22.00%–168.84%, in a dose-dependent manner. However, available phosphorus was unaffected by biochar and showed a gradual decline over time. Biochar altered bacterial community composition but had limited impact on alpha diversity. A shift from homogeneous to heterogeneous selection in bacterial community assembly was observed in the third year, correlating with soil nutrient changes. Cotton growth exhibited an “A-shaped” response, with Limited benefits in the first and third years but improvement in the second year. The 1% biochar promoted cotton growth by at least 12.07%, whereas higher concentrations (3% and 5%) had no effect, emphasizing an optimal dose effect. These findings highlight biochar’s potential as a tool for improving saline soils, while underscoring the importance of optimizing application rates and considering its long-term residual effects.

Optimal drip irrigation management in shallow groundwater areas needs to clarify the effects of flow rate and layout designs on the soil moisture, salt distribution, cotton root length density, plant height, leaf area, and yield. In this study, a one-year field experiment was conducted from April to October 2018 in the fifth company of the 16th Regiment in Alar City, Xinjiang, to investigate the effects of various drip flow rates and layout designs of cotton growth. Two drip flow rates (2.8 and 5.6 L·h⁻¹) and two layout designs (one film, two drip tapes, and six rows; one film, three drip tapes, and six rows) were applied to explore the optimal combination, resulting in a total of four treatments that were irrigated three times in the whole growth period. Soil moisture, salt distribution, cotton root length density, plant height, and leaf area were measured. The main results were as follows: (1) Under the same layout designs, the soil moisture content was higher and the soil salinity was lower when the drip flow rate was 5.6 L·h⁻¹, and the cotton root length density, plant height, leaf area, and yield were significantly higher than that of 2.8 L·h⁻¹. (2) Under the same drip flow rate, the soil desalination rate, cotton growth indexes, and yield under the three-tapes treatment were significantly higher than the values of the two-tapes treatment. The actual yield of treatment D was 21.56%, 19.23%, and 11.71% higher than that of treatments A, B, and C, respectively. (3) The crop evapotranspiration of cotton during the two irrigation cycles showed an increasing trend, and the groundwater contribution showed a smaller and then increasing trend. Overall, the combination of three tapes and a drip flow rate of 5.6 L·h⁻¹ had the highest cotton yield and net income, which were 6211.36 kg·hm⁻² and 4820.21 kg·hm⁻² for the theoretical and actual yields. The results of this study can provide a reference for the management of limited irrigation leaching soil salinity and cotton cultivation in shallow groundwater areas.

This study aimed to investigate the legacy effects of prolonged brackish irrigation on rhizobacterial communities and agricultural productivity in wheat. Here, we conducted pot experiments to investigate the mechanisms through which different irrigation regimes (irrigation using brackish groundwater and normal water) regulate wheat production. We applied four irrigation treatments across different stages of wheat growth (early stages, seedling-to-jointing, and late stages, jointing-to-maturity). This included irrigation exclusively using normal water during both stages (RR), using normal water followed by brackish groundwater (RW), exclusively using brackish groundwater (WW), and using brackish groundwater followed by normal water (WR). Under the premise of retaining 10 seedlings per pot, the average number of effective spikes per 10 plants in the RR, RW, and WR treatments was approximately 1.3, 1.1, and 1.1 times that of WW (19 ± 1), respectively. The spike weight per 10 plants in the RR, RW, and WR treatments was approximately 1.8, 1.5, and 1.3 times that of WW (12.75 ± 1.74 g), respectively. Compared with brackish groundwater irrigation, the use of normal water during the early stages significantly reduced the relative abundance of Pseudomonadota and increased that of Chloroflexota (p < 0.05). The number of effective spikes was positively correlated with the relative abundances of Actinobacteriota, Acidobacteriota, Chloroflexota, and Bacteroidota, but negatively correlated with the abundance of Pseudomonadota (p < 0.05). Irrigation regimes altered the rhizobacterial community structure. However, the legacy effect of long-term irrigation using brackish groundwater resulted in the dominance of stochastic processes in bacterial community assembly and stability of the Shannon diversity across all treatments. The complexity of the rhizobacterial co-occurrence network was lower in the RW treatments than that in the WW treatments (p < 0.05). Structural equation modeling revealed that irrigation using normal water during early stages boosted the number of effective spikes in wheat. This enhancement was achieved by increasing rhizobacterial diversity, reducing rhizosphere sodium, and simplifying the microbial network. This study challenges the “legacy effect” of brackish water irrigation by demonstrating that optimal irrigation timing is key to enhancing crop yield.