Search Results (4,028)

Understanding how prey and predator species partition activity patterns across time and space is essential for elucidating behavioral adaptation and ecological coexistence. In this study, we examined the diel and seasonal activity rhythms of two sympatric rodent species—Rhombomys opimus (Great gerbil) and Meriones meridianus (Midday gerbil)—and their primary predators, Otocolobus manul (Pallas’s cat) and Vulpes vulpes (Red fox), in a desert-steppe ecosystem on the northern slopes of the Qilian Mountains, China. Using over 8000 camera trap days and kernel density estimation, we quantified their activity intensity and spatiotemporal overlap. The two rodent species showed clear temporal niche differentiation but differed in their synchrony with predators. R. opimus exhibited a unimodal diurnal rhythm with spring activity peaks, while M. meridianus showed stable nocturnal activity with a distinct autumn peak. Notably, O. manul adjusted its activity pattern to partially align with that of R. opimus, whereas V. vulpes maintained a crepuscular–nocturnal rhythm overlapping more closely with that of M. meridianus. Despite distinct temporal rhythms, both rodent species shared high spatial overlap with their predators (overlap index OI = 0.64–0.83). These findings suggest that temporal partitioning may reduce predation risk for R. opimus, while M. meridianus co-occurs more extensively with its predators. Our results highlight the ecological role of native carnivores in rodent population dynamics and support their potential use in biodiversity-friendly rodent management strategies under arid grassland conditions. Full article

Earthquakes, as one of the most destructive natural disasters, often cause significant casualties and severe economic losses. Accurate prediction of earthquake fatalities is of great importance for pre-disaster prevention and mitigation planning, as well as post-disaster emergency response deployment. To address the challenges of small sample sizes, high dimensionality, and strong nonlinearity in earthquake fatality prediction, this paper proposes an integrated modeling approach (PCA-IWOA-XGBoost) combining Principal Component Analysis (PCA), the Improved Whale Optimization Algorithm (IWOA), and Extreme Gradient Boosting (XGBoost). The method first employs PCA to reduce the dimensionality of the influencing factor data, eliminating redundant information and improving modeling efficiency. Subsequently, the IWOA is used to intelligently optimize key hyperparameters of the XGBoost model, enhancing the prediction accuracy and stability. Using 42 major earthquake events in China from 1970 to 2025 as a case study, covering regions including the west (e.g., Tonghai in Yunnan, Wenchuan, Jiuzhaigou), central (e.g., Lushan in Sichuan, Ya’an), east (e.g., Tangshan, Yingkou), north (e.g., Baotou in Inner Mongolia, Helinger), northwest (e.g., Jiashi in Xinjiang, Wushi, Yongdeng in Gansu), and southwest (e.g., Lancang in Yunnan, Lijiang, Ludian), the empirical results showed that the PCA-IWOA-XGBoost model achieved an average test set accuracy of 97.0%, a coefficient of determination (R²) of 0.996, a root mean square error (RMSE) and mean absolute error (MAE) reduced to 4.410 and 3.430, respectively, and a residual prediction deviation (RPD) of 21.090. These results significantly outperformed the baseline XGBoost, PCA-XGBoost, and IWOA-XGBoost models, providing improved technical support for earthquake disaster risk assessment and emergency response. Full article

Dietary reviews contain rich emotional and objective information; however, existing knowledge extraction methods struggle with low-resource scenarios due to sparse and imbalanced label distributions. To address these challenges, this paper proposes a few-shot supervised learning approach. First, we develop a professional dietary–emotional schema by integrating domain knowledge with real-time data to ensure the coverage of diverse emotional expressions. Next, we introduce a dataset optimization method based on dual constraints—label frequency and quantity—to mitigate label imbalance and improve model performance. Utilizing the optimized dataset and a tailored prompt template, we fine-tune the DRE-UIE model for multivariate knowledge extraction. The experimental results demonstrate that the DRE-UIE model achieves a 20% higher F1 score than BERT-BiLSTM-CRF and outperforms TENER by 1.1%. Notably, on a 20-shot subset, the model on the Chinese dataset scores 0.841 and attains a 15.16% F1 score improvement over unoptimized data, validating the effectiveness of our few-shot learning framework. Furthermore, the approach also exhibits robust performance across Chinese and English corpora, underscoring its generalization capability. This work offers a practical solution for low-resource dietary–emotional knowledge extraction by leveraging schema design, dataset optimization, and model fine-tuning to achieve high accuracy with minimal annotated data. Full article

Slow-release nitrogen fertilizers enhance crop production and reduce environmental pollution, but their slow nitrogen release may cause insufficient nitrogen supply in the early stages of crop growth. Mixed nitrogen fertilization (MNF), combining slow-release nitrogen fertilizer with urea, is an effective way to increase yield and income and improve nitrogen fertilizer efficiency. This study used urea alone (Urea) and slow-release nitrogen fertilizer alone (C/SRF) as controls and employed meta-analysis and a random forest model to assess MNF effects on crop yield and nitrogen partial factor productivity (PFPN), and to identify key influencing factors. Results showed that compared with urea, MNF increased crop yield by 7.42% and PFPN by 8.20%, with higher improvement rates in Northwest China, regions with an average annual temperature ≤ 20 °C, and elevations of 750–1050 m; in soils with a pH of 5.5–6.5, where 150–240 kg·ha⁻¹ nitrogen with 25–35% content and an 80–100 day release period was applied, and the blending ratio was ≥0.3; and when planting rapeseed, maize, and cotton for 1–2 years. The top three influencing factors were crop type, nitrogen rate, and soil pH. Compared with C/SRF, MNF increased crop yield by 2.44% and had a non-significant increase in PFPN, with higher improvement rates in Northwest China, regions with an average annual temperature ≤ 5 °C, average annual precipitation ≤ 400 mm, and elevations of 300–900 m; in sandy soils with pH > 7.5, where 150–270 kg·ha⁻¹ nitrogen with 25–30% content and a 40–80 day release period was applied, and the blending ratio was 0.4–0.7; and when planting potatoes and rapeseed for 3 years. The top three influencing factors were nitrogen rate, crop type, and average annual precipitation. In conclusion, MNF should comprehensively consider crops, regions, soil, and management. This study provides a scientific basis for optimizing slow-release nitrogen fertilizers and promoting the large-scale application of MNF in farmland. Full article

Knowledge tracing (KT), a core educational data mining task, models students’ evolving knowledge states to predict future learning. In online education systems, the exercises are numerous, but they are typically associated with only a few concepts. However, existing models rarely integrate exercise information with high-order exercise–concept correlations, focusing solely on optimizing models’ final predictive performance. To address these limitations, we propose the Hypergraph-Driven High-Order Knowledge Tracing with a Dual-Gated Dynamic Mechanism (HGKT), a novel framework that (1) captures correlations between exercises and concepts through a two-layer hypergraph convolution; (2) integrates hypergraph-driven exercise embedding and temporal features (answer time and interval time) to characterize learning behavioral dynamics; and (3) designs a learning layer and a forgetting layer, with the dual-gating mechanism dynamically balancing their impacts on the knowledge state. Experiments on three public datasets demonstrate that the proposed HGKT model achieves superior predictive performance compared to all baselines. On the longest interaction sequence dataset, ASSISChall, HGKT improves prediction AUC by least 1.8%. On the biggest interaction records dataset, EdNet-KT1, it maintains a state-of-the-art AUC of 0.78372. Visualization analyses confirm its interpretability in tracing knowledge state evolution. These results validate HGKT’s effectiveness in modeling high-order exercise–concept correlations while ensuring practical adaptability in real-world online education platforms. Full article

Glycyrrhiza uralensis (G. uralensis), a leguminous plant, is an important medicinal and economic plant in saline–alkaline soils of arid regions in China. Its main bioactive components include liquiritin, glycyrrhizic acid, and flavonoids, which play significant roles in maintaining human health and preventing and adjuvantly treating related diseases. However, the cultivation of G. uralensis is easily restricted by adverse soil conditions in these regions, characterized by high salinity, high alkalinity, and nutrient deficiency. This study investigated the impacts of four multistrain microbial inoculants (Pa, Pb, Pc, Pd) on the growth performance and bioactive compound accumulation of G. uralensis in moderately saline–sodic soil. The aim was to screen the most beneficial inoculant from these strains, which were isolated from the rhizosphere of plants in moderately saline–alkaline soils of the Hexi Corridor and possess native advantages with excellent adaptability to arid environments. The results showed that inoculant Pc, comprising Pseudomonas silesiensis, Arthrobacter sp. GCG3, and Rhizobium sp. DG1, exhibited superior performance: it induced a 0.86-unit reduction in lateral root number relative to the control, while promoting significant increases in single-plant dry weight (101.70%), single-plant liquiritin (177.93%), single-plant glycyrrhizic acid (106.10%), and single-plant total flavonoids (107.64%). Application of the composite microbial inoculant Pc induced no significant changes in the pH and soluble salt content of G. uralensis rhizospheric soils. However, it promoted root utilization of soil organic matter and nitrate, while significantly increasing the contents of available potassium and available phosphorus in the rhizosphere. High-throughput sequencing revealed that Pc reorganized the rhizospheric microbial communities of G. uralensis, inducing pronounced shifts in the relative abundances of rhizospheric bacteria and fungi, leading to significant enrichment of target bacterial genera (Arthrobacter, Pseudomonas, Rhizobium), concomitant suppression of pathogenic fungi, and proliferation of beneficial fungi (Mortierella, Cladosporium). Correlation analyses showed that these microbial shifts were linked to improved plant nutrition and secondary metabolite biosynthesis. This study highlights Pc as a sustainable strategy to enhance G. uralensis yield and medicinal quality in saline–alkali ecosystems by mediating microbe–plant–nutrient interactions. Full article

The surface of the Great Wall harbors a large number of non-vascular plants dominated by cyanobacteria, lichens and mosses as well as microorganisms, and form biocrusts by cementing with the soils and greatly alters the pore structure of the soil and the ecohydrological processes associated with the soil pore space, and thus influences the soil resistance to erosion. However, the microscopic role of the biocrusts in influencing the pore structure of the surface of the Great Wall is not clear. This study chose the Warring States Qin Great Wall in Weiyuan, Gansu Province, China, as research site to quantify thepore structure characteristics of the three-dimensional of bare soil, cyanobacterial-lichen crusts, and moss crusts at the depth of 0–50 mm, by using optical microscopy, scanning electron microscopy, and X-ray computed tomography and image analysis, and the precipitation infiltration process. The results showed that the moss crust layer was dominated by large pores with long extension and good connectivity, which provided preferential seepage channels for precipitation infiltration, while the connectivity between the cyanobacterial-lichen crust voids was poor; The porosity of the cyanobacterial-lichen crust and the moss crust was 500% and 903.27% higher than that of the bare soil, respectively. The porosity of the subsurface layer of cyanobacterial-lichen crust and moss crust was significantly lower than that of the biocrusts layer by 92.54% and 97.96%, respectively, and the porosity of the moss crust was significantly higher than that of the cyanobacterial-lichen crust in the same layer; Cyanobacterial-lichen crusts increased the degree of anisotropy, mean tortuosity, moss crust reduced the degree of anisotropy, mean tortuosity. Biocrusts increased the fractal dimension and Euler number of pores. Compared with bare soil, moss crust and cyanobacterial-lichen crust increased the isolated porosity by 2555% and 4085%, respectively; Biocrusts increased the complexity of the pore network models; The initial infiltration rate, stable infiltration rate, average infiltration rate, and the total amount of infiltration of moss crusted soil was 2.26 and 3.12 times, 1.07 and 1.63 times, respectively, higher than that of the cyanobacterial-lichen crusts and the bare soil, by 1.53 and 2.33 times, and 1.13 and 2.08 times, respectively; CT porosity and clay content are significantly positively correlated with initial soil infiltration rate (|r| ≥ 0.85), while soil type and organic matter content are negatively correlated with initial soil infiltration rate. The soil type and bulk density are directly positively and negatively correlated with CT porosity, respectively (|r| ≥ 0.52). There is a significant negative correlation between soil clay content and porosity (|r| = 0.15, p < 0.001). Biocrusts alter the erosion resistance of rammed earth walls by affecting the soil microstructure of the earth’s great wall, altering precipitation infiltration, and promoting vascular plant colonisation, which in turn alters the erosion resistance of the wall. The research results have important reference for the development of disposal plans for biocrusts on the surface of archaeological sites. Full article

Understanding the carbon emission efficiency of apple production (APCEE) is critical for promoting green and low-carbon agricultural development. However, the spatiotemporal dynamics and driving factors of APCEE in China remain inadequately explored. This study employs life cycle assessment, super-efficiency slacks-based measures, and a panel Tobit model to evaluate the carbon footprint, APCEE, and its determinants in China’s two major production regions from 2003 to 2022. The results reveal that: (1) Producing one ton of apples in China results in 0.842 t CO₂e emissions. Land carbon intensity and total carbon emissions peaked in 2010 (28.69 t CO₂e/ha) and 2014 (6.52 × 10⁷ t CO₂e), respectively, exhibiting inverted U-shaped trends. Carbon emissions from various production areas show significant differences, with higher pressure on carbon emission reduction in the Loess Plateau region, especially in Gansu Province. (2) The APCEE in China exhibits a W-shaped trend (mean: 0.645), with overall low efficiency loss. The Bohai Bay region outperforms the Loess Plateau and national averages. (3) The structure of the apple industry, degree of agricultural mechanization, and green innovation positively influence APCEE, while the structure of apple cultivation, education level, and agricultural subsidies negatively impact it. Notably, green innovation and agricultural subsidies display lagged effects. Moreover, the drivers of APCEE differ significantly between the two major production regions. These findings provide actionable pathways for the green and low-carbon transformation of China’s apple industry, emphasizing the importance of spatially tailored green policies and technology-driven decarbonization strategies. Full article

As a core device in low-voltage distribution networks, the distribution zone transformer (DZT) is influenced by short circuits, overloads, and unbalanced loads, which cause thermal aging, mechanical stress, and eventually deformation of the winding, resulting in parameter deviations from nameplate values and impairing system operation. However, existing identification methods typically require synchronized high- and low-voltage data and are limited to symmetric three-phase conditions, which limits their application in practical distribution systems. To address these challenges, this paper proposes a parameter identification method for DZTs under three-phase unbalanced conditions. Firstly, based on the transformer’s T-equivalent circuit considering the load, the power flow equations are derived without involving the synchronization issue of high-voltage and low-voltage side data, and the sum of the impedances on both sides is treated as an independent parameter. Then, a novel power flow equation under three-phase unbalanced conditions is established, and an adaptive recursive least squares (ARLS) solution method is constructed using the measurement data sequence provided by the smart meter of the intelligent transformer terminal unit (TTU) to achieve online identification of the transformer winding parameters. The effectiveness and robustness of the method are verified through practical case studies. Full article

As the world’s largest loess deposit region, the Loess Plateau’s vegetation dynamics are crucial for its regional water–heat balance and ecosystem functioning. Leaf Area Index (LAI) serves as a key indicator bridging canopy architecture and plant physiological activities. Existing studies have made significant advancements in simulating LAI, yet accurate LAI simulation remains challenging. To address this challenge and gain deeper insights into the environmental controls of LAI, this study aims to accurately simulate LAI in the Loess Plateau using deep learning models and to elucidate the spatiotemporal influence of soil moisture and temperature on LAI dynamics. For this purpose, we used three deep learning models, namely Artificial Neural Network (ANN), Long Short-Term Memory (LSTM), and Interpretable Multivariable (IMV)-LSTM, to simulate LAI in the Loess Plateau, only using soil moisture and temperature as inputs. Results indicated that our approach outperformed traditional models and effectively captured LAI variations across different vegetation types. The attention analysis revealed that soil moisture mainly influenced LAI in the arid northwest and temperature was the predominant effect in the humid southeast. Seasonally, soil moisture was crucial in spring and summer, notably in grasslands and croplands, whereas temperature dominated in autumn and winter. Notably, forests had the longest temperature-sensitive periods. As LAI increased, soil moisture became more influential, and at peak LAI, both factors exerted varying controls on different vegetation types. These findings demonstrated the strength of deep learning for simulating vegetation–climate interactions and provided insights into hydrothermal regulation mechanisms in semiarid regions. Full article

Faba bean (Vicia faba L.), a nutrient-rich legume beneficial to human health, is valued for its high L-3,4-dihydroxyphenylalanine (L-DOPA) and phenolic content. This study investigated phytochemical diversity and bioactivity across 29 Chinese faba bean varieties. Phenolics were profiled using ultrahigh-performance liquid chromatography with quadrupole time-of-flight mass spectrometry (UPLC-QTOF-MS) and quantified via high-performance liquid chromatography (HPLC). Antioxidant capacity was evaluated, including DPPH (2,2-diphenyl-1-picrylhydrazyl), ABTS (2,2-azinobis (3-ethylbenzothiazoline-6-sulfonic acid)) radical scavenging activity, and ferric reducing antioxidant power (FRAP), along with α-amylase/α-glucosidase inhibitory effects. Twenty-five phenolics were identified, including L-DOPA (11.96–17.93 mg/g, >70% of total content), seven phenolic acids, and seventeen flavonoids. L-DOPA showed potent enzyme inhibition (IC₅₀ values of 22.45 μM for α-amylase and 16.66 μM for α-glucosidase) but demonstrated limited antioxidant effects. Lincan 13 (Gansu) exhibited the strongest antioxidant activity (DPPH, 16.32 μmol trolox/g; ABTS, 5.85 μmol trolox/g; FRAP, 21.38 mmol Fe²⁺/g), which correlated with it having the highest flavonoid content (40.51 mg rutin/g), while Yican 4 (Yunnan) showed the strongest α-amylase inhibition (43.33%). Correlation analysis confirmed flavonoids as the primary antioxidants, and principal component analysis (PCA) revealed geographical trends (e.g., Jiangsu varieties were particularly phenolic-rich). These findings highlight faba beans’ potential as functional foods and guide genotype selection in targeted breeding programs aimed at enhancing health benefits. Full article

Wool has distinctive biological, physical, and chemical properties that contribute to its value both for the sheep and in global fibre and textile markets. Its fibres are primarily composed of proteins, principally keratin and keratin-associated proteins (KAPs). To better comprehend the genes that underpin key wool traits, this study examined the keratin-associated protein 36-1 gene (KRTAP36-1) in Chinese Tan lambs. We identified three previously reported alleles of the gene (named A, B and C) that were present in the lambs studied, with genotype frequencies as follows: 2.0% (n = 5; AA), 6.9% (n = 17; AB), 13.8% (n = 34; AC), 8.9% (n = 22; BB), 33.4% (n = 82; BC) and 35.0% (n = 86; CC). The frequencies of the individual alleles in the Chinese Tan lambs were 12.4%, 29.1% and 58.5% for alleles A, B and C, respectively. The three alleles were in Hardy–Weinberg Equilibrium. In an association analysis, it was revealed that allele C was associated with variation in the mean fibre curvature of the fine wool of the Chinese Tan lambs, but this association was not observed in their heterotypic hair fibres. This finding suggests that KRTAP36-1 might be differentially expressed in the wool follicles that produce the two fibre types, and that along with other KRTAP genes, it may be involved in determining fibre curvature and the distinctive curly coat of the lambs. Full article

Addressing the technical demands for the rapid, precise detection of earthquake-triggered landslides in loess tablelands, this study proposes and validates an innovative methodology integrating enhanced deep learning architectures with large-tile processing strategies, featuring two core advances: (1) a critical enhancement of YOLOv8’s shallow layers via a higher-resolution P2 detection head to boost small-target capture capabilities, and (2) the development of a large-tile segmentation–tile mosaicking workflow to overcome the technical bottlenecks in large-scale high-resolution image processing, ensuring both timeliness and accuracy in loess landslide detection. This study utilized 20 km² of high-precision UAV imagery acquired after the 2023 Gansu Jishishan Ms 6.2 earthquake as foundational data, applying our methodology to achieve the rapid detection and precise segmentation of landslides in the study area. Validation was conducted through a comparative analysis of high-accuracy 3D models and field investigations. (1) The model achieved simultaneous convergence of all four loss functions within a 500-epoch progressive training strategy, with mAP50(M) = 0.747 and mAP50-95(M) = 0.46, thus validating the superior detection and segmentation capabilities for the Jishishan earthquake-triggered loess landslides. (2) The enhanced algorithm detected 417 landslides with 94.1% recognition accuracy. Landslide areas ranged from 7 × 10⁻⁴ km² to 0.217 km² (aggregate area: 1.3 km²), indicating small-scale landslide dominance. (3) Morphological characterization and the spatial distribution analysis revealed near-vertical scarps, diverse morphological configurations, and high spatial density clustering in loess tableland landslides. Full article

Soil respiration (Rt), consisting of heterotrophic (Rh) and autotrophic respiration (Ra), plays a vital role in terrestrial carbon cycling and is sensitive to soil temperature and moisture. In dryland agriculture, plastic film mulching (PM) is widely used to regulate soil hydrothermal conditions, but its effects on Rt components and their temperature sensitivity (Q₁₀) across regions remain unclear. A two-year field study was conducted at two rain-fed maize sites: Anding (warmer, semi-arid) and Yuzhong (colder, drier). PM significantly increased Rt, Rh, and Ra, especially Ra, due to enhanced root biomass and improved microclimate. Yield increased by 33.6–165%. Peak respiration occurred earlier in Anding, aligned with maize growth and soil temperature. PM reduced Q₁₀ of Rt and Ra in Anding, but only Ra in Yuzhong. Rh Q₁₀ remained stable, indicating microbial respiration was less sensitive to temperature changes. Structural equation modeling revealed that Rt and Ra were mainly driven by soil temperature and root biomass, while Rh was more influenced by microbial biomass carbon (MBC) and dissolved organic carbon (DOC). Despite increased CO₂ emissions, PM improved carbon emission efficiency (CEE), particularly in Yuzhong (+67%). The application of PM is recommended to enhance yield while optimizing carbon efficiency in dryland farming systems. Full article

Enhancing farmers’ digital literacy is both an inevitable requirement for adapting to the digital age and an important measure for promoting the sustainable development of livestock and poultry manure resource utilization. This study surveyed and obtained data from 1047 farm households in Ningxia and Gansu, two provinces in China that have long implemented livestock manure resource utilization policies, from December 2023 to January 2024, and employed the binary probit model to analyze how digital literacy influences farmers’ willingness to adopt two livestock manure resource utilization modes, as well as to analyze the moderating role of three policy regulations. This paper also explores the heterogeneous results in different village forms and income groups. The results are as follows: (1) Digital literacy significantly and positively impacts farmers’ willingness to adopt both the “household collection” mode and the “livestock community” mode. For every one-unit increase in a farmer’s digital literacy, the probability of farmers’ willingness to adopt the “household collection” mode rises by 22 percentage points, and the probability of farmers’ willingness to adopt the “livestock community” mode rises by 19.8 percentage points. After endogeneity tests and robustness checks, the conclusion still holds. (2) Mechanism analysis results indicate that guiding policy and incentive policy have a positive moderation effect on the link between digital literacy and the willingness to adopt the “household collection” mode. Meanwhile, incentive policy also positively moderates the relationship between digital literacy and the willingness to adopt the “livestock community” mode. (3) Heterogeneity analysis results show that the positive effect of digital literacy on farmers’ willingness to adopt two livestock manure resource utilization modes is stronger in “tight-knit society” rural areas and in low-income households. (4) In further discussion, we find that digital literacy removes the information barriers for farmers, facilitating the conversion of willingness into behavior. The value of this study is as follows: this paper provides new insights for the promotion of livestock and poultry manure resource utilization policies in countries and regions similar to the development process of northwest China. Therefore, enhancing farmers’ digital literacy in a targeted way, strengthening the promotion of grassroots policies on livestock manure resource utilization, formulating diversified ecological compensation schemes, and establishing limited supervision and penalty rules can boost farmers’ willingness to adopt manure resource utilization models. Full article