This study comprehensively considers soil formation factors such as land use types, soil types, depths, and geographical conditions in Lanxi City, China. Using multi-source public data, three environmental variable screening methods, the Boruta algorithm, Recursive Feature Elimination (RFE), and Particle Swarm Optimization (PSO), were used to optimize and combine 47 environmental variables for the modeling of soil pH based on the data collected from farmland in the study area in 2022, and their effects were evaluated. A Random Forest (RF) model was used to predict soil pH in the study area. At the same time, Pearson correlation analysis, an environmental variable importance assessment based on the RF model, and SHAP explanatory model were used to explore the main controlling factors of soil pH and reveal its spatial differentiation mechanism. The results showed that in the presence of a large number of environmental variables, the model with covariates selected by PSO before the application of the Random Forest algorithm had higher prediction accuracy than that of Boruta–RF, RFE–RF, and all variable prediction RF models (MAE = 0.496, RMSE = 0.641, R² = 0.413, LCCC = 0.508). This indicates that PSO, as a covariate selection method, effectively optimized the input variables for the RF model, enhancing its performance. In addition, the results of the Pearson correlation analysis, RF-model-based environmental variable importance assessment, and SHAP explanatory model consistently indicate that Channel Network Base Level (CNBL), Elevation (DEM), Temperature mean (T_m), Evaporation (E_m), Land surface temperature mean (LST_m), and Humidity mean (H_m) are key factors affecting the spatial differentiation of soil pH. In summary, the approach of using PSO for covariate selection before applying the RF model exhibits high prediction accuracy and can serve as an effective method for predicting the spatial distribution of soil pH, providing important references for accurately simulating the spatial mapping of soil attributes in hilly and basin areas.

Landscape connectivity is important for all organisms as it directly affects population dynamics. Yet, rapid urbanization has caused serious landscape fragmentation, which is the primary contributor of species extinctions worldwide. Previous studies have mostly used spatial snap-shots to evaluate the impact of urban expansion on landscape connectivity. However, the interactions among habitats over time in dynamic landscapes have been largely ignored. Here, we demonstrated that overlooking temporal connectivity can lead to the overestimation of the impact of urban expansion. How much greater the overestimation is depends on the amount of net habitat loss. Moreover, we showed that landscape connectivity may have a delayed response to urban expansion. Our analysis shifts the way to understand the ecological consequences of urban expansion. Our framework can guide sustainable urban development and can be inspiring to conservation practices under other contexts (e.g., climate change).

Studying the spatial variation patterns and influencing factors of soil organic matter (SOM) in hilly and basin areas is of great significance for guiding agricultural production practices. This study takes Lanxi City as an example and comprehensively considers soil formation factors such as climate, vegetation, and terrain. Based on the genetic algorithm, 47 environmental variables are combined and optimized to construct a random forest (RF) model and an improved version—a random forest model based on genetic algorithm variable combination optimization (RF-GA). At the same time, the SHAP interpretation method is used to quantitatively analyze the spatial distribution characteristics of the SOM content and further identify the main driving factors. Compared with the ordinary Kriging (OK) and random forest (RF) methods, the random forest model based on genetic algorithm variable combination optimization (RF-GA) demonstrates a significantly improved prediction accuracy (R² = 0.49; RMSE = 3.49 g·kg⁻¹), with an MAE = 3.019 and LCCC = 0.67. Among the three models, the R² of the RF-GA model increases by 87.84% and 56.29%. The model prediction results indicate that the SOM content in the study area ranges from 12.11 to 31.38 g·kg⁻¹, showing spatial distribution characteristics of a higher content in mountainous areas and a lower content in plains. A further SHAP analysis shows that terrain, climate, and biological factors are key environmental factors affecting the spatial differentiation of the SOM, with the channel network base level (CNBL), which contributes 20.68% to the model, and DEM, which has a contribution rate of 5.57%, playing particularly significant roles. By regulating moisture, erosion deposition, vegetation distribution, and microclimate conditions, they significantly affect the spatial distribution of the SOM. In summary, the RF-GA and its interpretable prediction model constructed in this study not only effectively reveal the spatial and driving mechanisms of SOM in hilly and basin areas but also provide a solid theoretical basis and practical guidance for accurate mapping, the formulation of sustainable utilization strategies for soil resources, and ensuring national food security.

Accurately mapping paddy rice is crucial for food security, sustainable agricultural management and environmental protection. Recently, Sentinel-2 optical images with a spatial resolution of 10 m and a repeat cycle of five days have demonstrated enormous potential for mapping paddy fields. However, the influence of the temporal selection of Sentinel-2 optical images on mapping paddy rice is still unclear. In this study, the optimal temporal windows were detected by considering all possible temporal combinations during the growing stages from the constructed cloud-free 10-day time series and assessing the classification performances of all combination schemes on paddy rice mapping by F1_score. The results indicated that the combination of two or three phases is necessary for mapping early-cropping paddy rice (EP) and late-cropping paddy rice (LP), achieving the F1_score aim of 0.96. The detection of single-cropping paddy rice (SP) requires a combination of three to five phases and can obtain the F1_score aim of 0.94. Additionally, an automatic workflow for paddy rice mapping has been developed, which does not require any cloud removal but provides complete spatial coverage, suitable for regions with frequent rain and clouds. Through verification in the study area of Yiwu, China, the discrepancies between mapping results and agricultural statistics were within 5%, demonstrating the rationality and efficiency of the proposed framework.

Landscape connectivity can largely affect the level of biodiversity and it is a key concern in conservation planning. Considering that protected areas (PAs) may become functionally isolated “islands” under rapid land-use change, there is an urgent need to expand and connect protected areas to prevent further biodiversity loss and improve PAs effectiveness. The Grain-to-Green Project (GTGP) is the largest reforestation program worldwide with the aim of controlling water and soil loss; however, the opportunities for connectivity gains through GTGP have been widely ignored. Here we provided a three-stage hierarchical framework incorporating soil erosion analysis, cropland suitability analysis and network analysis to prioritize agricultural patches for reforestation under the scheme of GTGP. The potential reforestation patches were identified in the first two stages. Then, four different GTGP strategies were designed, and a set of network metrics were used to determine the best strategy and prioritize patches that significantly enhance PAs connectivity. A typical GTGP region, Wanzhou district (Southwest China), was taken as the study area. We found that: (1) the agricultural patches with high reforestation suitability cover an area of c. 40 km² (1% of the study area); and (2) the efficiency of GTGP strategies varies by species, species with intermediate and high mobility benefit more from a strategy that continuously adds reforestation patches close to PAs, while for species with low dispersal ability, the amount of patches added should also be taken into account to decide the appropriate strategy. We conclude that our framework can provide guidance to restore PAs connectivity with limited land resources in the context of GTGP.