The translation of readily available soil data into soil hydraulic conductivity parameters, which describe water infiltration at the soil surface, is often limited by complex, non-linear interactions associated with soil structure that are rarely captured by classical predictor sets. We hypothesize that soil taxonomic information, particularly World Reference Base (WRB) soil groups, encodes meaningful pedogenic and structural information that can help bridge this gap. We test this hypothesis using eleven Soil Science-Informed Machine Learning (SoilML) models trained on Europe-wide in-situ observations from the Soil Water Infiltration Global (SWIG) database. Models were developed to predict infiltration-based effective saturated hydraulic conductivity (Ks) using combinations of WRB soil groups, land-use type, and classical soil properties (clay, sand, silt, organic carbon, and bulk density), with additional hydrological state variables included in selected scenarios. Our results showed that ’tree-based’ models, specifically CatBoost, outperform linear regression and kernel-based neural approaches. Importantly, incorporating WRB soil groups and land-use information consistently improves predictive performance, increasing explained variance by up to 20% compared to texture-only models. We demonstrate that WRB soil classification provides useful additional information that reflects pedogenic development and site-specific structural effects not represented by soil texture alone. This highlights the potential of integrating soil taxonomic and land-use information into data-driven pedotransfer approaches for more robust upscaling of continental-scale soil hydraulic properties. However, through the use of spatial cross validation, it is found that although SoilML models perform well in regional interpolation, they fail miserably in terms of generalization across unknown geography.
Biobased fertilizers could be sustainable alternatives to mineral inputs and restore soil functionality, but their effect in calcareous soils with limited phosphorus (P) phytoavailability remains unclear. This study aimed to evaluate the effect of certain biobased fertilizers on soil functionality. A two-year field experiment in a wheat-sunflower rotation on a calcareous Vertisol was developed involving: Commercial Plant Residue (CPR), Composted Digestate (DIG), Composted Olive Mill Pomace (OMP), Composted Urban Organic Waste (UOW) and Vermicompost (VER), which were compared with Diammonium Phosphate (DAP) and no P addition (CON) for their effect on soil [pH, soil organic matter (SOM), nutrients availability and microbial communities (bacteria and fungi)] and plant growth (biomass, yield, and P uptake). Compared with CON, UOW improved soil P availability by 55% and plant P uptake by 26%, and produced higher apparent P use efficiency (PUE) than DAP, while promoting putative SOM decomposers such as Microascus , and Arachnomyces . OMP increased SOM accumulation by 41%, fungal β diversity, and microbial P biomass by 225%, relative to CON, indicating a shift toward biologically mediated P retention and recycling, which may support their residual effect but limited short-term P availability. VER favored rapid nutrient turnover and enhanced PUE, whereas DIG and CPR showed limited effects on soil variables, likely due to a stoichiometric imbalance affecting the SOM mineralization process. UOW was the most promising biobased fertilizer to replace DAP and restore soil functionality in this calcareous soil. Overall, fertilizer composition and C:N:P stoichiometry drove the agronomic and ecological performance of these biobased fertilizers.
SOC in European croplands has declined over recent decades, affecting CO2 emissions and soil functioning. A more detailed understanding of the relative contribution of agricultural management and climate change in these losses is lacking.We analyzed SOC dynamics over 29 years (1994–2023) in a long-term field experiment in Puch, Germany, comparing organic amendments (green manure, three compost types) combined with different mineral N fertilization levels. We further examine how recent climate change may have influenced SOC trajectories during the experimental period.Organic amendments significantly increased SOC content, with fine sized compost achieving the highest levels, 33% above the non-amended control, which showed the lowest SOC. High mineral N fertilization also significantly enhanced SOC content by 7–9%, though its effect (η2 = 9.7%) remained smaller than that of organic amendments (η2 = 68.3%), despite substantially increasing crop yield and associated crop residue C input.SOC content increased until 2012 but substantially declined in all treatments thereafter despite high C inputs from organic amendments and rising C inputs from crop residues. This decline coincided with rising temperatures and more frequent droughts at the site. RothC model configurations that explicitly accounted for climatic controls best reproduced observed SOC trajectories, suggesting that climate-related effects may have contributed to the observed SOC decline.Although compost application mitigated SOC decline on croplands, additional measures are required to fully prevent climate-induced losses. The long–term effectiveness of organic amendments depends on amendment properties and climatic conditions. This pattern raises questions about SOC accumulation under combined influences of organic matter inputs and climatic warming.
Loess, kaolin and bentonite are widely distributed in cold regions, and their geotechnical mechanical properties deteriorate significantly under repeated freeze–thaw cycles (FTCs) and low-temperature weathering, which seriously threatens the long-term stability and service life of infrastructure projects. This study investigates and compares the long-term strength and microstructural evolution of loess, kaolin, and bentonite subjected to FTCs. The applicability of existing long-term strength prediction models was evaluated, and microstructural parameters were characterized using mercury intrusion porosimetry (MIP) and scanning electron microscopy (SEM). Correlation analyses between strength and microstructural parameters were performed, and empirical models describing their relationships were developed. The results show that the long-term strength of loess and kaolin initially increases but then declines sharply with increasing FTCs, whereas the strength of bentonite exhibits continuous deterioration. The Vyalov equation is more suitable for predicting the long-term strength of loess, while the Roman equation performs better for kaolin and bentonite. The time analogy method demonstrates strong predictive capability for all three special soils. SEM imaging indicates a transition from face-to-face to point contacts between soil particles. SEM and MIP results reveal that after FTCs, loess with coarser particles forms an overhead structure dominated by macropores and mesopores. Kaolin presents a dense layered structure with a high proportion of mesopores, whereas bentonite contains considerable amounts of both micropores and macropores, with both structures present. Redundancy analysis (RDA), Pearson correlation analysis, and grey relational analysis reveal that pore morphology exerts the greatest influence on the long-term strength of loess, whereas pore arrangement is the dominant factor for kaolin and bentonite. The grey relational analysis shows that the long-term strength of loess has a higher correlation with macropores, while the long-term strengths of kaolin and bentonite are more closely correlated with mesopores and micropores, respectively. The differences in mechanical behavior among the three soils under FTCs are fundamentally controlled by their mineral-specific structural reorganization mechanisms, including cementation breakdown, particle reorientation, and interlayer delamination. These findings establish a linkage between macroscopic long-term strength and microscopic parameters, providing theoretical support and practical guidance for the long-term performance and hazard mitigation of frozen soil engineering in cold regions.
Actively heated fiber-optic (AHFO)-based soil water content measurement technology has been widely applied in geological, agricultural, and environmental monitoring due to its distributed and real-time sensing capability. However, variations in heating power among AHFO sensors are often neglected in large-scale deployments, reducing the accuracy and applicability of soil water content measurements. The primary novelty of this study is the development of a Tt–w calibration framework that explicitly considers heating power and enables the calibration relationship at target heating powers to be derived from a single reference-power calibration. The proposed framework was verified through laboratory experiments using loess samples and further applied to in situ monitoring of soil water content in loess over a distance of up to 218 m, demonstrating its applicability to large-scale soil water content monitoring. Laboratory results demonstrated an approximately linear relationship between Tt and Q. Using 10 W/m for 20 min as the reference, the predicted Tt values at 20 W/m and 35 W/m yielded RMSEs of 1.37°C and 2.02°C, showing good agreement with the experimental results. Further in situ tests indicated that the soil water content values determined by the proposed model showed good agreement with neutron probe measurements (RMSE = 0.0197 g/g), enabling the spatiotemporal dynamics of shallow soil water content and temperature in the loess region to be captured and demonstrating the effectiveness and potential of the model under complex field conditions.
