Zhiqiang Yin

Soil thickness is an important indicator for evaluating soil production and ecosystem support functions. However, traditional field surveys cannot characterise the spatial distribution of soil thickness with high resolution. Although machine learning models that integrate multiple environmental variables offer a promising approach to predict the high‐resolution spatial distribution of soil thickness, their predictive accuracy is often constrained in areas with heterogeneous soil parent materials. In this study, the Zhangbei area with heterogeneous soil parent materials was selected as a case to optimise the model. In detail, soil parent materials were used as one of environmental variables incorporated into the machine‐learning model for soil thickness prediction, and their contributions were evaluated. Based on 97 soil thicknesses of surveyed profiles in the Zhangbei area, four machine learning models were developed and compared, including support vector machine (SVM), artificial neural network (ANN), random forest (RF) and extreme gradient boosting (XGB). Results showed that all models performed better than the simple linear baseline model, indicating the advantage of nonlinear modelling for soil thickness prediction. Especially, the XGB model incorporating soil parent material explained 81% of the spatial variation in soil thickness and outperformed models of SVM, ANN and RF. Based on high‐resolution data of soil parent materials, the soil thickness with high‐resolution spatial distribution across the study area was precisely predicted. Overall, our study provides a high‐precision framework for predicting soil thickness in areas with heterogeneous soil parent materials, and its potential applications are expected for ecological restoration and land management.

The Loess Hilly Region of northern Shaanxi is one of the most erosion-prone areas in the world due to its porous, erodible loess, steep slopes, and seasonal rainfall. To address this, conversion of sloping farmland to terraces has been extensively conducted across China’s loess regions, as terracing can reduce soil and water loss and enhance soil fertility. However, disturbance of soil layers during terracing can also lead to short-term decline in farmland productivity. This study investigates the effects of terracing operations at two sites of different ground substrate configurations in the Loess Hilly Region. Utilizing geochemical and molecular biological analysis methods, we examined the changes in the physicochemical and microbial properties of the ground substrate after terracing, using adjacent sloping farmlands as control sites. The results show that when the ground substrate configuration remained intact, terracing increased the average water content (from 8.44% to 14.34%) and soil organic carbon (from 2.74 g/kg to 5.76 g/kg) by 70% and 110%, respectively, and increased soil microbial α-diversity by 90%. The microbial community structure was also enhanced with an increase in relative abundance of soil- and plant-benefiting genera such as Streptomyces and Nocardioides, thereby promoting plant growth. Conversely, when the ground substrate configuration was altered, terracing led to a decrease in soil nutrient and moisture content, which was detrimental to crop growth. Therefore, maintaining the integrity of the ground substrate configuration is crucial during the terracing process to achieve optimal soil and water conservation outcomes.

Heterogeneity of ground substrates is a key controlling factor for vegetation differentiation in mountainous regions. However, the cascading regulatory mechanism involving lithological differences, substrate structure, water retention, and vegetation response remains unclear. This study focuses on three typical ground substrate areas in the Yanshan Mountains of northern China—granite, gneiss, and dolomite regions—located within the ecological barrier zone of the Beijing–Tianjin–Hebei region. By integrating multi‑source data from field trenching, shallow well excavation, in‑situ electrical detection, vegetation quadrat surveys, and remote sensing inversion, the regulatory mechanism of lithological differences on vegetation coverage was revealed. The results indicate that: (1) The development of bedrock fractures and the appropriate grain size composition of the substrate directly affect the water‑holding capacity of the ground substrate, influencing root distribution and plant water‑use strategies, thereby controlling the spatial distribution characteristics of vegetation. (2) Bedrock fracture density shows a significant positive correlation with vegetation coverage (r2= 0.79), with the order being granite > gneiss > dolomite. Ground substrate moisture content is positively correlated with the proportion of silt‑clay particles ( gneiss > granite. (3) The thickness of the bedrock weathering layer (granite > gneiss > dolomite) and fracture density jointly control vegetation root distribution, providing a scientific basis for differentiated vegetation restoration strategies based on lithological characteristics in mountainous regions of northern China.Keywords: lithological differences; vegetation type; vegetation coverage; constraining mechanism; Yanshan MountainFunding: This study was supported by the National Natural Science Foundation of China (Grant No. U2344227)

Systematic assessment of the ecological and human health risks induced by soil metals and metalloids (MMs) from various sources is critical for pollution control and risk prevention. This study takes Hancheng, a typical coal-industrial city in Weibei, China, as an example and employs the geological accumulation index, potential ecological risk assessment, positive matrix factorization (PMF), and Monte Carlo probabilistic health risk assessment to comprehensively quantify and evaluate regional risks. The results show that among the eight soil metals and metalloids, only the average Cr concentration does not exceed the background value of Shaanxi Province, with all elements influenced by anthropogenic activities. Cd and Hg are identified as the most polluting elements, exhibiting pronounced accumulation and ecological risks. The PMF receptor model reveals four sources: mixed transportation and specific industry sources (28.6%), coal-fired sources (31.2%), natural sources (29.1%), and industrial emissions (11.1%), with human-induced sources collectively contributing 70.9%. The geological background of the study area and coal industrial activities exert a synergistic enhancing effect on soil MMs pollution. By integrating the Monte Carlo method with the PMF model, the source-contribution-based probabilistic health risk assessment indicates that children face higher health risks than adults, and the combined effect of multiple MMs significantly increases health risks compared to individual elements. Transportation-related and industrial sources are major contributors to human health risks, with As, Ni, and Cd prioritized as pollutants due to their elevated concentrations. These findings offer a scientific basis for targeted pollution prevention and prioritized management in coal-industrial cities, supporting efforts to mitigate regional ecological and health risks.

Mangrove forests, as a quintessential blue carbon ecosystem, play a vital role in the global carbon cycle and climate regulation. However, studies on the spatial distribution patterns of mangrove carbon stocks, inter-community variability, and tidal influence remain limited in typical regions of China. In this study, we conducted a systematic field investigation in the Shankou Mangrove National Nature Reserve, Guangxi, based on 40 vegetation sampling plots and 22 Soil coring transects, to comprehensively evaluate the carbon stock characteristics and their spatial heterogeneity within the mangrove ecosystem. The results revealed that the total carbon stock density in the study area was 120.41MgC·ha⁻¹, with biomass carbon accounting for 55.64MgC·ha⁻¹- lower than the reported national average for Chinese mangroves (84.61 MgC·ha⁻¹)- which is consistent with the characteristics of subtropical mangrove ecosystems. Notably, Soil carbon accounted for 53.77% of the total stock, representing the dominant carbon pool. Spatially, carbon stocks exhibited a clear “high in core zones, low in peripheral zones” pattern. Carbon sequestration capacity varied significantly among plant assemblages and tidal zones, with arboreal communities dominated by Bruguiera gymnorhiza and Rhizophora stylosa demonstrating the highest accumulation potential. Additionally, Soil carbon density decreased progressively from the mid intertidal zone toward the high- and low-intertidal zones, indicating that tidal inundation frequency exerts a crucial control on carbon burial and stabilization. These findings provide a valuable scientific basis and empirical dataset for mangrove carbon accounting, ecosystem restoration, and blue carbon management strategies.

In recent years, China has invested substantial funds in ecological restoration, achieving significant accomplishments. The forest coverage rate in the Chengde Bashang area, located in the transitional zone between the monsoon and non-monsoon regions, has now reached 82%. However, the area has also encountered a series of environmental issues, including lake shrinkage, soil salinization, and large-scale die-offs of planted forests. Whether the forests in this region can achieve sustainable development in the future, and whether ecological restoration should prioritize tree planting or grass cultivation, are critical questions that require attention. By studying the historical vegetation dynamics in afforested areas, we can better understand the relationship between climatic environmental changes and vegetation, providing baseline data for future ecological restoration. This study utilized AMS ¹⁴C dates to establish a chronological framework for the core and employed pollen to investigate vegetation dynamics over the past 5000 years in the artificial Larix Mill. forest area. The vegetation and environmental history of this core can be divided into three zones: Zone 1 (5100–4100 a B.P.): vegetation was dominated by pine and spores, with low herbaceous pollen content. Zone 2 (4100–1400 a B.P.): vegetation was primarily herbaceous. Zone 3 (1400 a B.P.–present): arboreal pollen content increased slightly, but herbaceous plants remained dominant. This period included the warm–dry Medieval Warm Period (1400–900 a B.P.), the cold–humid Little Ice Age (900–300 a B.P.), and the recent 300 years of anthropogenic disturbance. Notably, the large-scale afforestation efforts in recent decades are clearly reflected in the profile. A comparative analysis of records from the monsoon–non-monsoon transition zone reveals that, except for Angulinao Lake, other records were dominated by herbaceous vegetation over the past 2000 years. Additionally, the Mu Us Sandy Land, Hunshandake Sandy Land, Hulunbuir Sandy Land, and Horqin Sandy Land in China have experienced aeolian sand accumulation over the same period. Given the anticipated warming–desiccation trend, phytoremediation strategies should favor xerophytic shrubs and herbaceous over monospecific forest plantations.

Monitoring the distribution of nutrient and beneficial elements (NBEs) in topsoils is important for assessing agricultural and ecological risks and for promoting human health. In order to understand their distribution effectively, it is critical to clarify their sources and main influence factors. In this study, a total of 272 topsoils and 46 soil parent material samples and some samples along 4 vertical profiles were collected in the Mentougou area of Beijing. Then, 12 NBEs were determined for each sample, including elements of N, P, S, Se, Zn, Mn, V, Ge, K, F, Ca, Mg. The mean content of 12 NBEs was statistically calculated and their spatial distributions were mapped. NBEs in topsoils were mainly influenced by the background of parent materials and the subsequent surficial processes. To address the weathered background, the study area was divided into 4 units, including igneous rocks, clastic rocks, carbonate, and loess units. For each unit, cluster analysis was performed to identify the relationship between 12 NBEs in topsoils. Furthermore, the changes in characteristics of 12 NBEs along vertical soil profiles of 4 units were individually analyzed by comparison with the weathered background. Therefore, the main factors influencing the distribution of 12 NBEs in topsoils were identified respectively. According to each element’s main influencing factors, 12 NBEs were divided into 5 groups. The main influencing factors were same for elements in each group but different each other for 5 groups. Besides the weathered background, the distribution of NBEs in topsoils was significantly influenced by subsequent surficial processes, such as organic enrichment, leaching, and calcification. In addition to these natural surficial processes, the influence of human activities, a special surficial process, was also observed on the distribution of N, P, S, and Zn in topsoils in farming areas by interpreting normalized difference vegetation index (NDVI). This study provides a semi-quantitative method to determine the role of weathered background and subsequent surficial processes in the distribution of NBEs in topsoils, which helps to understand the current situation and evolution trend.