Sensor-based soil data libraries have proven valuable for a wide range of applications. Given the increasing adoption of portable X-ray fluorescence (pXRF) spectroscopy for soil-related studies, we aimed to create a library of soils characterized by pXRF and develop prediction models for soil particle size fractions (clay, silt, and sand contents) based on 16,989 samples collected at multiple soil horizons, from all Brazilian biomes (Amazon, Caatinga, Cerrado, Atlantic Forest, Pantanal, and Pampa), under multiple climate zones, land uses and derived from several parent materials. Random Forest (RF), Support Vector Machine (SVM), Gradient Boosting Machine (GBM), and Genetic Programming (GP) were used to train the models. The dataset was split into training (13,593) and independent validation (3,396) sets. The accuracy of models was assessed via root mean square error (RMSE) and coefficient of determination (R2) values. Sand, silt and clay contents ranged from 0 to 100%, 0 to 96%, and 0 to 93%, respectively. Strong correlations were found between particle size fractions and Fe and Si contents, but some strong correlations were biome-dependent, such as those between K and silt (0.77) in Pantanal. Random Forest outperformed the other algorithms, achieving RMSE and R2 values, respectively, of 8.2% and 0.89 for sand, 5.6% and 0.71 for silt, and 6.3% and 0.89 for clay contents. This pXRF library resulted in adequate predictions for sand and clay contents. This national library may contribute to establish a global library of soils characterized via pXRF. This dataset should be continuously increased to support soil chemical characterization and develop prediction models for varied soil properties.
This study evaluates the responses of soil properties to lithology, topography, and LULC in the Galyan watershed using soil data (n = 417) collected from LULC classes overlying different lithological units (Çatak, Hamurkesen, and Kaçkar), together with topographic variables. The study also investigates the combined effects of anthropogenic gradients, including cropland and hazelnut cultivation (Corylus avellana), and natural vegetation gradients (grassland, Alnus glutinosa, Fagus orientalis, and Picea orientalis) across lithology. In addition, LULC changes (2006–2023) were analyzed to evaluate potential risks to watershed sustainability. Soil samples were analyzed for particle size distribution, pH, electrical conductivity, organic matter, total nitrogen, total phosphorus, and exchangeable base cations. MANOVA indicated statistically substantial multivariate effects of lithology, LULC, and their interaction on soil properties. CATPCA revealed structured relationships among soil, topography, and LULC classes, while correlation analysis evaluated the relationships between topography and soil properties. Çatak soils had higher levels of clay, silt, pH, EC, Ca2+, and Mg2+. Kaçkar and Hamurkesen soils had higher sand, organic matter, total nitrogen, and total phosphate. Natural lands exhibited a vegetation gradient in response to topographical conditions. Alnus glutinosa was associated with lower silt, clay, pH, EC, TN, and exchangeable base cations, whereas Fagus orientalis and Picea orientalis were associated with higher values. Topography further structured these patterns: lower, warmer areas were associated with finer textures and higher base cation concentrations, while higher, cooler areas were associated with higher concentrations of organic matter, nitrogen, and phosphorus. Despite being based on a single soil sampling campaign, the spatial differences emerging from the combined effects of lithology, topography, and LULC provide valuable insights into potential soil responses and have important implications for sustainable watershed management. LULC analysis indicated a notable expansion of agricultural and impervious areas, suggesting potential implications for nutrient dynamics.
Climate change and anthropogenic disturbances have complicated vegetation-rainfall relationships, making it crucial to understand their coupling mechanisms, especially in water-limited ecosystems. Within this context, we examined the dynamics and drivers of the vegetation-rainfall relationships in the Yellow River Basin (YRB) since 2000 using multi-source remote sensing data, Mann-Kendall trend detection, and an attribution framework integrating Principal Component Analysis (PCA) and Hierarchical Partitioning (HP). Results demonstrated that Net Primary Productivity (NPP) and Rainfall Use Efficiency (RUE) exhibited a significant increasing trend (80.43% and 42.57% of the YRB respectively), particularly in mid-basin forest areas (95.48% and 67.47% of the midstream respectively). However, Vegetation Sensitivity to Rainfall (VSR) declined across 25.43% of the basin, indicating that vegetation growth is becoming progressively less dependent on natural rainfall, with the most pronounced decline observed in the middle basin (31.9% of its area). Human activity intensity and ecological response index (HEI) accounted for 73.5% of the variance in the drivers of RUE and VSR, while climate-water stress index (CWI) explained 10.9%. Specifically, RUE was predominantly governed by natural factors and vegetation structure (e.g., leaf area index), whereas VSR was primarily driven by human activities (e.g., population density). These divergent drivers reveal an efficiency-resilience trade-off, suggesting that regional ecosystem management should emphasize both vegetation cover and rainfall use efficiency alongside ecosystem resilience to rainfall variability-a dimension essential for ecological security and sustainable development.
Ecological restoration is widely implemented to recover degraded lands, yet the microbial mechanisms underpinning soil organic carbon (SOC) sequestration during this process remain poorly understood, particularly regarding how nutrient availability shapes microbial community traits and necromass accumulation. In this study, we utilized a space-for-time substitution method across a tropical ecological restoration chronosequence (comprising rubber monoculture, near-natural rainforest, and primary rainforest) and integrated analyses of soil properties, microbial biomarkers, and molecular sequencing to investigate the linkages between microbial nutrient limitation, community structural and functional traits, and soil carbon fractions. Our results showed that ecological restoration significantly improved litter and soil nutrient stoichiometry, alleviating microbial carbon (C) and phosphorus (P) limitation. This variation drove a shift in microbial life-history strategies from high growth yield strategies to resource-acquisition/stress-tolerance strategies, increasing microbial carbon use efficiency (19–60%), richness (27–79%), biomass (10–64%), and network complexity (1.9–3.3-fold) while reducing competitive interactions. Concurrently, stochastic processes became more dominant in community assembly. Metagenomic analysis revealed enhanced expression of genes involved in C fixation, recalcitrant C degradation (e.g., lignin degradation), and P cycling. Consequently, microbial necromass C, particularly fungal-derived C, increased significantly (13–61%) and emerged as a stronger direct driver of SOC accumulation than plant-derived C, despite a decrease in its proportional contribution to the SOC pool. These results suggest that alleviated nutrient limitation is linked to coordinated shifts in microbial metabolism and community characteristics, which in turn promote necromass accumulation and contribute to SOC sequestration. This study underscores the necessity of integrating microbial metabolic traits into soil carbon management strategies for degraded tropical plantations.
Gullies supply and transport sediment, yet their contribution to catchment-scale sediment budgets varies and remains difficult to quantify, particularly in data-limited catchments. We examined the role of gullies in the sediment budget of a sub-humid catchment in the Ethiopian Omo-Gibe using the SIMWE (SIMulated Water Erosion) model, morphometric analysis of field-inventoried gullies, and flow- and sediment-monitoring. A total of 145 gullies were inventoried to analyse their morphology and spatial distribution. Flow and sediment dynamics were simulated using the two-dimensional, process-based SIMWE model. The model was calibrated against seasonal, event-based flow and sediment data. The results indicate that rainfall variability, vegetation cover, and agricultural practices affect both the SIMWE model and the sediment rating curve parameters. Additionally, incorporating seasonally differentiated model parameters improved decadal sediment simulations (2015–2025), highlighting their role in capturing intra-annual catchment-scale erosion processes. However, simulated decadal sediment yield differed by 37% from reservoir inlet bathymetric estimates, reflecting uncertainties arising from temporally constant model infiltration assumptions and limitations in the validation data. Furthermore, the morphometric analysis shows that most gullies in the catchment exhibit V-shaped cross-sections and contribute only 1–1.4% to total sediment yield, suggesting that the gullies primarily act as sediment pathways rather than dominant sources during this period. These findings offer new insights into sediment connectivity and emphasise the need to account for temporal variability when parameterising models to improve sediment budget assessments and inform targeted conservation measures in gully-prone catchments. Future research using coupled hydro-geomorphological models may incorporate spatio-temporal variability in model parameters and infiltration rates.
In South America, pre-Columbian cultures were responsible for the formation of two types of anthropogenic soils: Dark-Earths (DE) and Shellmounds (SM), both characterized by high nutrient content, especially phosphorus (P), in contrast to the surrounding natural soils. Previous work has reported high total P content in DE, while SM have hardly been studied. In this study, 75 samples from different geographical locations were analyzed: 38 tropical, equatorial and sub-polar SM; 30 tropical-equatorial DE; and 7 control soils (CS), to improve our comprehension of P dynamics. Geochemical forms of P were investigated using sequential chemical extraction and 31P NMR. The results revealed significant differences between natural soils and Anthrosols. Control soils showed lower pH (4.3 ± 0.3), lower total organic C (0.5 ± 0.1%) and total N (0.03 ± 0.01%) than DE (pH:5.6 ± 1.6, TOC:1.6 ± 0.6%; TN:0.16 ± 0.18%) and SM (pH:7.2 ± 1.2, TOC:4.3 ± 4.3%; TN:0.24 ± 0.33%). Total P was markedly lower in CS (62 ± 84 mg·kg−1) than in DE (1028 ± 1043 mg·kg−1) and SM (12,812 ± 10,170 mg·kg−1). In control soils, P was mainly associated with recalcitrant organic matter, clays, and Al hydroxides, whereas Anthrosols displayed a more complex and variable distribution of P forms. In the acid DE samples, without carbonates, P distribution was similar to that of control soils. In contrast, carbonate-rich DE and SM were dominated by Ca-bound P, except for SM samples from the subpolar region, where P associated with humic substances was the predominant form. These findings highlight the edaphogeochemical anomaly of pre-Columbian anthropogenic soils, characterized by elevated total P and distinct geochemical forms. Differences in P speciation between CS, DE and SM soils reflect anthropogenic inputs and environmental drivers. Carbonate-rich environments promote Ca-bound P accumulation, whereas colder subpolar conditions favour organic-association P forms, indicating that climate-related factors may also influence P speciation across the different climatic regions studied. Overall, these results emphasize the functional divergence of DE and SM soils.
