Soil nutrients play vital roles in vegetation growth and are a key indicator of land degradation. Accurate, rapid, and non-destructive measurement of the soil nutrient content is important for ecological conservation, degradation monitoring, and precision farming. Currently, visible and near-infrared (Vis–NIR) spectroscopy allows for rapid and non-destructive monitoring of soil nutrients. However, the performance of Vis–NIR inversion models is extremely dependent on the number of samples. Limited samples may lead to low prediction accuracy of the models. Therefore, modeling and prediction based on a small sample size remain a challenge. This study proposes a method for the simultaneous augmentation of soil spectral and nutrient data (total nitrogen (TN), soil organic matter (SOM), total potassium oxide (TK₂O), and total phosphorus pentoxide (TP₂O₅)) using a generative adversarial network (GAN). The sample augmentation range and the level of accuracy improvement were also analyzed. First, 42 soil samples were collected from the pika disturbance area on the QTP. The collected soils were measured in the laboratory for Vis–NIR and TN, SOM, TK₂O, and TP₂O₅ data. A GAN was then used to augment the soil spectral and nutrient data simultaneously. Finally, the effect of adding different numbers of generative samples to the training set on the predictive performance of a convolutional neural network (CNN) was analyzed and compared with another data augmentation method (extended multiplicative signal augmentation, EMSA). The results showed that a GAN can generate data very similar to real data and with better diversity. A total of 15, 30, 60, 120, and 240 generative samples (GAN and EMSA) were randomly selected from 300 generative samples to be included in the real data to train the CNN model. The model performance first improved and then deteriorated, and the GAN was more effective than EMSA. Further shortening the interval for adding GAN data revealed that the optimal ranges were 30–40, 50–60, 30–35, and 25–35 for TK₂O, TN, TP₂O₅, and SOM, respectively, and the validation set accuracy was maximized in these ranges. Therefore, the above method can compensate to some extent for insufficient samples in the hyperspectral prediction of soil nutrients, and can quickly and accurately estimate the content of soil TK₂O, TN, TP₂O₅, and SOM.

Plateau zokors (Eospalax baileyi) are an agent actively involved in causing soil erosion and meadow degradation in the Yellow River Source Zone of West China. This study aims to quantify the amount of soil and nutrient loss from zokor mounds in relation to slope gradient and rainfall intensity, and to assess the amount of soil loss in zokor-infested areas compared with healthy meadows in Henan County, Qinghai Province. The results showed that zokor mounds were gradually lowered at a rate of 1.8–3.9 cm h⁻¹. Soil loss occurred two min after the rain began, reaching the maximum level during the first 20 min. The rate of soil loss and nutrient loss increased with the rainfall intensity and slope gradient. When the rainfall intensity rose from 5 to 10 mm h⁻¹, and from 10 to 15 mm h⁻¹, the total soil loss on 10° slopes increased by 2.5 times and 3.9 times, respectively, and soil nutrient loss increased by 1.7 times and 2.7 times, respectively. As the slope gradient steepened to 20°, the corresponding figures were 2.8 times and 4.3 times for total soil loss, and 1.8 times and 2.9 times, respectively, for soil nutrient loss. When the slope rose to 30°, the soil loss increased by 3.0 and 4.5 times, and the soil nutrient loss increased by 1.8 times and 3.1 times, respectively. There was a power function between soil loss and surface runoff (S = 0.2371Q^2.2307, R² = 0.9529). The soil was eroded at a rate of 256.6 g m⁻² h⁻¹ from zokor mounds, 17.7 times higher than in intact meadows, and 1.8 times higher than in partially recovered meadows. Most of the eroded soils had a mean diameter of 0–1.2 mm. It is recommended that artificial control of plateau zokors should be implemented, together with other ecological restoration measures to restrain the soil erosion problem caused by zokor activities.

Land-use change plays an important role in ecological change; knowing the trends in land-use change can quickly help identify problems in regional ecosystems. In 2000 to 2020, the development of a global economy caused increasing extreme weather events worldwide and lead to exacerbating changes in types of land-use. The Yellow River source area is an important water source and a central part of ecological protection efforts in China. The fragile ecosystems make the area sensitive to environmental changes. Therefore, in protecting the ecological security of the basin, simulating changes in the ecosystem service value under different scenarios is a meaningful procedure. A patch-generating land use simulation model was used to simulate different land use scenarios in 2030, including an ecological protection scenario, a production priority scenario, a carbon neutral scenario and a natural development scenario. The analysis shows that significant progress has been made in water conservation but grassland conservation faces enormous challenges. The rate of development, occupation of farmland and land dedicated to construction has increased. Unused land increased dramatically from 2010–2020 and has not been mitigated by existing policies. Based on the unit area value equivalent coefficients, the ecosystem service value rankings for the seven land use types were as follows: Grassland > Wetland > Water Area > Forest > Farmland > Unused Land > Construction Land; the four types of ecosystem service value are ranked as follows: regulating services > supporting services > supply services > cultural services; the four scenarios of ecosystem service value are ranked as follows: ecological protection scenario > production priority scenario > carbon neutral scenario > natural development scenario. The ecosystem service value of the Yellow River source area would increase by CNY 1.641 billion in 2030 with ecological protection goals and decrease by CNY 1.421 billion with the current of development. This study provides valuable insights and implications for land use, ecological protection and sustainable development by shedding light on watershed change issues and assessing and predicting the ecological status of the Yellow River source area.

The plateau pika (Ochotona curzoniae) actively contributes to soil erosion and meadow degradation in western China’s Yellow River source zone. This study aimed to elucidate the effects of the pika mound numbers on the hydrodynamic characteristics and soil water erosion through simulated rainfall experiments. The inhibition effects of restored vegetation growth on meadow degradation and soil erosion were explored using a revegetated pika mound as a control. The results showed that at a rainfall intensity of 30 mm/h, the soil loss per unit time increased and then decreased with rainfall time and that 15–20 min of rainfall duration was the sensitive period for soil loss in different pika mound patch lands. The degradation of meadows due to the activities of plateau pika is an essential factor influencing soil erosion, and the soil erosion rate is positively correlated with both the pika mound numbers and slope. The mean flow velocity can better describe the process of the soil erosion, and its value increased with the number of pika mounds and slope. The Reynolds number ranged from 57.85 to 153.63 (Re <500), and it was preliminarily determined that all slope runoff was laminar flow. The Froude number increased linear function with pika mound numbers (p < 0.01), and both the pika mound numbers and slope were significant factors affecting it (p < 0.05). The Darcy–Weisbach resistance coefficient instead decreased with the pika mound numbers and slope, and the inhibitory effect of vegetation on soil erosion was probably limited when the mound numbers reached a certain level. According to the results of the grey correlation and Pearson correlation analysis, changes in the number of pika mounds led to variability in the soil water erosion properties by altering the landscape scale effect. The number of pika mound patches (NP), edge length index (TE), area index (AREA), and volume index (V) were the key influencing factors on the soil erosion rate and hydrodynamic parameters. We conclude that plateau pika activities intensify meadow degradation, and the continuous increase in pika mounds decreases the vegetation cover and flow resistance and increases soil erodibility. Controlling the intensity of plateau pika activities will effectively prevent soil erosion in degraded areas.