Guang Li’s research while affiliated with Gansu Agricultural University and other places

The agro-pastoral ecotone of Gansu Province, a critical component of the ecological security barrier in northern China, is characterized by pronounced ecological fragility and climatic sensitivity. Investigating vegetation dynamics in this region is essential for balancing ecological conservation and sustainable development. This study integrated MODIS/NDVI remote sensing data (2000–2020), climate, land, and anthropogenic factors, employing Sen’s slope analysis, coefficient of variation (Cv), Hurst index, geodetector modeling, and partial correlation analysis to systematically unravel the spatio-temporal evolution and driving mechanisms of vegetation coverage. Key findings revealed the following: (1) Vegetation coverage exhibited a significant increasing trend (0.05 decade−1), peaking in 2018 (NDVI = 0.71), with a distinct north–south spatial gradient (lower values in northern areas vs. higher values in southern regions). Statistically significant greening trends (p < 0.05) were observed in 55.42% of the study area. (2) Interannual vegetation fluctuations were generally mild (Cv = 0.15), yet central regions showed 2–3 times higher variability than southern/northwestern areas. Future projections (H = 0.62) indicated sustained NDVI growth. (3) Climatic factors dominated vegetation dynamics, with sunshine hours and precipitation exhibiting the strongest explanatory power (q = 0.727 and 0.697, respectively), while the elevation–precipitation interaction achieved peak explanatory capacity (q = 0.845). (4) NDVI correlated positively with precipitation in 43.62% of the region (rmean = 0.47), whereas average temperature, maximum temperature, ≥10 °C accumulated temperature, and sunshine hours suppressed vegetation growth (rmean = −0.06 to −0.42), confirming precipitation as the primary driver of regional vegetation recovery. The multi-scale analytical framework developed here provides methodological and empirical support for precision ecological governance in climate-sensitive transitional zones, particularly for optimizing ecological barrier functions in arid and semi-arid regions.

The Global Navigation Satellite System Interferometric Reflectometry (GNSS-IR) has demonstrated significant potential for soil moisture content (SMC) monitoring due to its high spatiotemporal resolution. However, GNSS-IR inversion experiments are notably influenced by vegetation and meteorological factors. To address these challenges, this study proposes a multi-factor SMC inversion method. Six GNSS stations from the Plate Boundary Observatory (PBO) were selected as study sites. A low-order polynomial was applied to separate the reflected signals, extracting parameters such as phase, frequency, amplitude, and effective reflector height. Auxiliary variables, including the Normalized Microwave Reflection Index (NMRI), cumulative rainfall, and daily average evaporation, were used to further improve inversion accuracy. A multi-factor SMC inversion dataset was constructed, and three machine learning models were selected to develop the SMC prediction model: Support Vector Regression (SVR), suitable for small and medium-sized regression tasks; Convolutional Neural Networks (CNN), with robust feature extraction capabilities; and NRBO-XGBoost, which supports automatic optimization. The multi-factor SMC inversion method achieved remarkable results. For instance, at the P038 station, the model attained an R² of 0.98, with an RMSE of 0.0074 and an MAE of 0.0038. Experimental results indicate that the multi-factor inversion model significantly outperformed the traditional univariate model, whose R² (RMSE, MAE) was only 0.88 (0.0179, 0.0136). Further analysis revealed that NRBO-XGBoost surpassed the other models, with its average R² outperforming SVR by 0.11 and CNN by 0.03. Additionally, the analysis of different surface types showed that the method achieved higher accuracy in grassland and open shrubland areas, with all models reaching R² values above 0.9. Therefore, the accuracy of the multi-factor SMC inversion model was validated, supporting the practical application of GNSS-IR technology in SMC inversion.

Accurately monitoring total nitrogen (TN) content in field soils is crucial for precise fertilization management. TN content is one of the core indicators in soil fertility evaluation systems. Rapid and accurate determination of TN in the tillage layer is essential for agricultural production. Although UAV-based multispectral remote sensing technology has shown potential in agricultural monitoring, research on its quantitative assessment of soil TN content remains limited. This study utilized UAV (unmanned aerial vehicle) multispectral imagery and field-measured TN data from four key growth stages of silage corn in 2022 at Huari Ranch, Minle County, Hexi region. The support vector machine–recursive feature elimination (SVM-RFE) algorithm was applied to select vegetation indices as model inputs. A total of 18 models based on machine learning algorithms, including BP neural networks (BPNNs), random forest (RF), and partial least squares regression (PLSR), were constructed to compare the most suitable inversion model for TN in the rhizosphere soil (0–30 cm) of silage corn at different growth stages. The optimal period for TN inversion was determined. The SVM-RFE algorithm outperformed the models built without feature selection in terms of accuracy. Among the nitrogen inversion models based on different machine learning algorithms, the PLSR model showed the best performance, followed by the RF model, while the BPNN model performed the worst. The PLSR model established for the mature growth stage at soil depths demonstrated the highest inversion accuracy, with R and RMSE values of 0.663 and 0.281, respectively. The next best period was the tasseling stage, while the worst inversion accuracy was observed during the seedling stage, indicating that the mature stage is the optimal period for TN inversion in the study area.

Process-based crop growth models, as an important analytical tool in agricultural production, face the problem of calibrating many parameters during the application process, and sensitivity analysis (SA) can quantify the effects of the model input parameters on the model output and provide an important basis for parameter calibration. In this study, we combined the good performance of the Agricultural Production Systems sIMulator Next-Generation (APSIM NG) model in simulating crop growth, soil carbon and nitrogen cycles, and soil N2O emissions with the efficient computational efficiency of the extended Fourier amplitude test (EFAST) method. The sensitivity of the APSIM NG model to the simulation of soil N2O emissions was systematically investigated using the EFAST method in a dry-crop wheat field in the semi-arid region of the Loess Plateau in Longzhong, China, where 28 crop cultivar parameters, 15 soil parameters, 4 meteorological parameters, and 4 field management parameters were selected. The parameters were selected based on the existing literature and the official documents of the model, and the parameter boundaries were determined based on the initial values of the APSIM NG model and the measured data and adjusted upward and downward by the standard normal distribution. In this study, parameters with a first-order sensitivity index (Si) > 0.05 and a total sensitivity index (STi) > 0.10 were identified as having a significant influence on the model outputs. The results of this study demonstrated that soil N2O emission modeling in dry-crop wheat fields showed high sensitivity to the following parameters: (1) Among the crop cultivar parameters, the sensitivity from high to low was the leaf appearance rate, maximum leaf area, maximum nitrogen concentration of the grain, and thermal time from the starting grain-fill stage to end grain-fill stage. (2) Among the soil parameters, the sensitivity from high to low was a lower effective moisture limit, wilting coefficient, and ammonium nitrogen content. (3) Among the meteorological parameters, precipitation and solar radiation showed high sensitivity. (4) Among the field management parameters, the nitrogen application rate exhibited the most significant sensitivity. For this reason, we believe that by prioritizing the calibration of the most sensitive parameters through the results of the sensitivity analysis in this study, the workload of the APSIM NG model in the calibration process can be effectively reduced, which is conducive to the rapid localization and application of the model.

Excessive nitrogen application and low nitrogen use efficiency have been major issues in China’s agricultural development, posing significant challenges for field management. Nitrogen is a critical nutrient for crop growth, playing an indispensable role in crop development, yield formation, and quality enhancement. Therefore, precisely controlling nitrogen application rates can reduce environmental pollution caused by excessive fertilization and improve nitrogen use efficiency. This study employs multispectral remote sensing images, combined with field-measured nitrogen content, to develop canopy nitrogen content inversion models for maize using three algorithms: backpropagation neural network (BP), support vector machine (SVM), and partial least squares regression (PLSR). The results reveal that there is a degree of redundancy in the information contained in various spectral indices. Feature selection effectively eliminates correlated and redundant spectral information, thereby improving modeling efficiency. The spectral indices Green Index (GI) and Nitrogen Reflectance Index (NRI) exhibit strong correlations with nitrogen content in the maize canopy, suggesting that the green and red spectral bands are crucial for retrieving maize’s biophysical and biochemical parameters. In studies on nitrogen content inversion in the maize canopy, the random forest (RF) algorithm, coupled with PLSR, demonstrated superior predictive performance. Compared to the standalone PLSR model, accuracy improved by 3.5%–6.5%, providing a scientific foundation and technical support for precise nitrogen diagnosis and fertilizer management in maize cultivation.

The Yellow River Basin is one of the most ecologically fragile zones in China. Landscape ecological risk assessment and ecological security pattern construction are effective ways to find out and solve the problems of the ecological environment in this region. In this study, based on LUCC and driving factor data, we evaluated the spatial distribution characteristics of landscape ecological risk from the grid scale using the ecological risk index (ERI) and minimum cumulative resistance (MCR) models, especially proposing a new method for constructing the ecological security pattern from the perspective of landscape ecological risk assessment and finally forming an ecological security pattern network composed of points, lines, and surfaces. The results indicated that the ecological risk of the Yellow River basin showed a spatial distribution pattern of “High in southwest and northwest, low in other regions.” The middle–low level ecological risk accounted for more than 86% of the total area. The study has formed an ecological security pattern network consisting of 12 core ecological sources, 92 ecological nodes, 12 key corridors, 16 auxiliary corridors, and 4 different controlled areas. According to these study findings, we provide ecological protection strategies that strengthen the ecological source conservation, attaching importance to the construction of ecological corridor, increasing the restoration of ecological nodes, and carrying out the zoning regulation. This study will provide a new insight for the construction of the ecological security pattern network based on the results of landscape ecological risk assessment.

Accurately monitoring soil moisture content (SMC) in the field is crucial for achieving precision irrigation management. Currently, the development of UAV platforms provides a cost-effective method for large-scale SMC monitoring. This study investigates silage corn by employing UAV remote sensing technology to obtain multispectral imagery during the seedling, jointing, and tasseling stages. Field experimental data were integrated, and supervised classification was used to remove soil background and image shadows. Canopy reflectance was extracted using masking techniques, while Pearson correlation analysis was conducted to assess the linear relationship strength between spectral indices and SMC. Subsequently, convolutional neural networks (CNNs), back-propagation neural networks (BPNNs), and partial least squares regression (PLSR) models were constructed to evaluate the applicability of these models in monitoring SMC before and after removing the soil background and image shadows. The results indicated that: (1) After removing the soil background and image shadows, the inversion accuracy of SMC for CNN, BPNN, and PLSR models improved at all growth stages. (2) Among the different inversion models, the accuracy from high to low was CNN, PLSR, BPNN. (3) From the perspective of different growth stages, the inversion accuracy from high to low was seedling stage, tasseling stage, jointing stage. The findings provide theoretical and technical support for UAV multispectral remote sensing inversion of SMC in silage corn root zones and offer validation for large-scale soil moisture monitoring using remote sensing.

Progressively higher atmospheric nitrogen (N) deposition increasingly affects soil ecosystems' elemental cycling and stability. Biochar (BC) amendment has emerged as a possible means of preserving soil system stability. Nevertheless, the pattern of soil–microbial nutrient cycling and system stability in response to BC after high N deposition in ecologically sensitive regions remains uncertain. Therefore, we investigated the effects of high N (9 g N·m⁻²·a⁻¹), BC (0, 20, 40 t·ha⁻¹), and combinations of the treatments on soil organic carbon (SOC), total nitrogen (TN), total phosphorus (TP), microbial biomass carbon (MBC), nitrogen (MBN), phosphorus (MBP), microbial entropy (qMB), and stoichiometric imbalance (Cimb:Nimb:Pimb). We found that high N addition decreased topsoil (0–20 cm) TP, C:N, qMBN, and Cimb:Nimb values and increased TN, C:P, N:P, qMBP, Cimb:Pimb, and Nimb:Pimb values. However, BC addition increased 0–40 cm soil qMBC and Nimb:Pimb values and decreased qMBN, Cimb:Nimb, and Cimb:Pimb values. Meanwhile, high BC additions attenuated BC's promotion of soil–microbial nutrients. We observed that a mixture of high N and BC increased the 0–40 cm SOC and TP content, promoted the accumulation of MBN and MBP in the subsoil (20–40 cm), and decreased the topsoil Cimb:Pimb and Nimb:Pimb values compared to high N additions. The impact of high N and BC additions on N and P elements varied significantly between the different soil depths. In addition, redundancy analysis identified C:N, MBC, MBN, and C:P as pivotal factors affecting alterations in soil qMB and stoichiometric imbalance. Overall, adding BC reduced the negative impacts of high N deposition on the stability of soil–microbial systems in the Loess Plateau, suggesting a new approach for managing ecologically fragile areas.

This study aims to explore the impacts of climate change on potato planting in Gansu Province so as to be able to adjust potato planting pattern scientifically and rationally. (1) Air precipitation and temperature time-series datasets were obtained from 87 meteorological stations in the study area over the past 50 years. The backpropagation neural network was employed to interpolate irregular and missing data in the time-series data. The altitude, the precipitation from June to July, the average temperature in July and the accumulated temperature above 10 °C were selected as the agricultural zoning indicators for the regionalization of potato planting. (2) The linear propensity rate method, cumulative anomaly method, ArcGIS technology and the Mann–Kendall mutation test were employed to examine the spatial–temporal variation in and mutation testing of the three zoning indicators. (3) The experimental results demonstrated that the amount of precipitation from June to July was registered at 139.94 mm, indicating a slight humidifying trend characterized by an annual increase rate of approximately 1.81 mm/10 a. Furthermore, a significant abrupt change was observed in 1998. The average temperature in July was registered at 20.53 °C, which showed an increasing trend at a rate of 0.55 °C/10 a, marked by a sudden shift in 1998. Lastly, the accumulated temperature above 10 °C was registered at 2917.05 °C, manifesting a significant warming trend at a rate of 161.96 °C/10 a, without any abrupt changes. For spatial distribution, the precipitation from June to July showed a decreasing spatial distribution pattern from south to north and from east to west, while its tendency rate showed a gradually decreasing trend from north to south and from east to west. The average temperature in July showed a decreasing spatial pattern from northeast to southwest, while its tendency rate showed a decreasing trend from west to east and from north to south. The accumulated temperature above 10 °C showed a spatial pattern of high accumulated temperatures in the northwestern and southeastern regions and low accumulated temperatures in the remaining regions, while its tendency rate showed a decreasing trend from west to east and from north to south. (4) The impacts of climate change on potato planting in Gansu Province were mainly manifested as a decrease of 0.30 × 10⁶ hm² in the cultivated land area in the most suitable region for potato planting post-1998, while the suitable area diminished by 0.96 × 10⁶ hm², the sub-suitable area expanded by 0.47 × 10⁶ hm², and the plantable area increased by 0.79 × 10⁶ hm². However, the unsuitable area experienced a reduction of 0.30 × 10⁴ hm². The findings of this study can provide a scientific foundation for optimizing and adjusting the potato planting structure, considering the backdrop of climate change. Moreover, they contribute to regional decision-making, thereby promoting sustainable agricultural development as well as enhancing both the yield and quality of potato in Gansu Province.

The farming-pastoral ecotone in Gansu Province, a crucial green barrier in northwestern China, faces a conflict between economic development and environmental fragility. Scientifically assessing the landscape ecological risks in this region is essential for developing an early warning system for ecological security and ensuring the sustainable management of regional ecosystems. This study systematically analyzes the dynamic evolution patterns and clustering characteristics of landscape ecological risk along the topographic gradient in the area, using Land-Use and Land-Cover Change (LUCC) data from five periods (each period representing 10 years) between 1980 and 2020, as well as Digital Elevation Model (DEM) data. Methods such as CA-Markov and the Topographic Position Index (TPI) were employed to assess these risks. Additionally, the study simulates the spatial distribution patterns of ecological risks in future years. The results reveal that from 1980 to 2020, the area of built-up land in the farming-pastoral ecotone of Gansu Province increased by approximately 82%, with roughly 21% of the expansion occurring at the expense of grasslands and 53% from arable land. Ecological risk in the study area exhibits a spatial distribution pattern characterized by lower risk in the southwest and higher risk in the northeast. Overall, ecological risk initially increased and then decreased. The area of high-risk zones has steadily decreased by approximately 50% since 2010, primarily in regions with low topographic gradients and high human activity. In other areas, high-risk levels are directly proportional to the topographic gradient, showing a high degree of spatial clustering. Consequently, land-use planning should be tailored to local conditions, and a long-term monitoring mechanism should be established. This study provides novel insights into the spatiotemporal changes in landscape ecological risk in ecologically vulnerable areas, focusing on topographic gradients.