Spatiotemporal dynamics and associations of soil organic carbon and total nitrogen in the Qilian Mountain Grasslands.

Spatiotemporal dynamics and associations of soil organic carbon and total nitrogen in the Qilian Mountain Grasslands - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Carbon Balance Manag . 2026 Mar 7;21:79. doi: 10.1186/s13021-026-00422-8 Spatiotemporal dynamics and associations of soil organic carbon and total nitrogen in the Qilian Mountain Grasslands Jiafang Cai 1 , Meiling Zhang 1, ✉ , Youyi Zhao 1 , Yanjun Gong 1 , Hanying Wang 1 Author information Article notes Copyright and License information 1 Center for Quantitative Biology, College of Science, Gansu Agricultural University, Lanzhou, 730070 China ✉ Corresponding author. Received 2025 Jun 8; Accepted 2026 Feb 18; Collection date 2026 Dec. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13224459  PMID: 41793629 Abstract Understanding the coordinated changes in soil carbon and nitrogen is essential for evaluating ecosystem responses to environmental change, particularly in ecologically fragile alpine regions such as the Qilian Mountains. In this study, the denitrification–decomposition (DNDC) model was used to assess the spatiotemporal dynamics of soil organic carbon density (SOCD) and total nitrogen density (STND) in the 0–30 cm soil layer from 1975 to 2024. The results revealed that SOCD and STND were higher in the northern and east-central grasslands and lower in the southwestern regions. Both stocks exhibited fluctuating but overall increasing trends, with notable increases aligned with major ecological protection policies in China. To better understand the coupling of soil carbon and nitrogen, we constructed a composite indicator called soil carbon and nitrogen density (SCND) using principal component analysis. This indicator captures the synergistic accumulation of organic carbon and total nitrogen driven by shared ecological processes and was further used to explore its associations with environmental factors, enabling an integrated assessment of soil carbon–nitrogen dynamics. The results revealed that elevation and soil bulk density were the main direct drivers of carbon and nitrogen accumulation, both of which exerted negative effects, whereas the other factors acted through indirect pathways. These findings underscore the importance of topography and soil structure in regulating carbon and nitrogen dynamics. It is recommended to plant deep-rooted grass species, limit heavy machinery, and maintain long-term ecological protection to prevent declines after initial gains from interventions. In addition, the carbon-to-nitrogen (C/N) ratio showed increasing spatial heterogeneity over time, with high values in the western and central regions, where nitrogen input can be enhanced by introducing legumes or applying organic fertilizers. In the northern and southeastern areas, grazing exclusion or low-intensity grazing is recommended to promote organic matter accumulation. Vertically, the C/N ratio decreased with soil depth, indicating strong carbon and nitrogen coupling within the soil profile. Overall, this study highlights the coordinated dynamics of soil carbon and nitrogen in the Qilian Mountain grasslands, providing valuable insights for the sustainable management and resilience improvement of grasslands in this region under changing environmental conditions. Supplementary Information The online version contains supplementary material available at 10.1186/s13021-026-00422-8. Keywords: Alpine grasslands, Carbon-to-nitrogen ratio, Soil carbon–nitrogen coupling, DNDC model, Partial least squares structural equation modelling, Ecosystem management Introduction Amid the acceleration of global warming, alpine regions with complex ecosystems and fragile environments have become key areas for studying ecological sensitivity and adaptation to climate change [ 1 ]. Ecosystems in alpine semiarid zones respond strongly to climate variability, making understanding their key ecological processes important for evaluating global ecosystem stability [ 2 , 3 ]. Carbon and nitrogen are essential elements that support ecosystem structure and function [ 4 ]. Their dynamics in soils influence nutrient availability, microbial activity, plant productivity, and broader biogeochemical cycles [ 5 – 8 ]. Grasslands are the dominant ecosystems in alpine semiarid regions, and their soils store large amounts of organic carbon and nitrogen [ 9 , 10 ]. These soils play a central role in linking climate change and ecological processes because changes in soil carbon and nitrogen pools influence greenhouse gas emissions, feed back to atmospheric carbon dioxide levels, and affect nutrient availability for vegetation growth, thereby regulating ecosystem responses to environmental change [ 11 ]. Long-term changes in soil carbon and nitrogen under climate stress and human disturbance reflect the stability and regulatory capacity of these ecosystems [ 12 ]. Understanding the coordinated dynamics of soil carbon and nitrogen is essential for revealing how soils respond to environmental change and for ensuring long-term ecosystem stability. The synergistic accumulation of SOCD and STND reflects the combined influence of climate, soil properties, vegetation, and anthropogenic activities. Although previous studies have examined the effects of these factors on SOCD or STND separately [ 13 , 14 ], most have focused on individual drivers or compared carbon and nitrogen response pathways in isolation, often overlooking their coupling relationship [ 15 ]. This has limited a comprehensive understanding of carbon–nitrogen interactions. In this study, density refers to the amount of organic carbon or total nitrogen stored per unit area of soil, representing the storage capacity and its temporal variation. The C/N ratio, as an important indicator of soil quality, reflects nutrient balance and organic matter decomposition, and is also used to assess vegetation recovery potential [ 16 ]. However, the C/N ratio and the dynamics of carbon–nitrogen coordination are often studied separately. Focusing on only one of these aspects fails to fully capture the dynamic nature and ecological significance of the soil carbon–nitrogen relationship. Integrating them within a unified analytical framework can help elucidate the mechanisms of carbon–nitrogen coupling and assess soil nutrient status and ecosystem resilience from both structural and functional perspectives. Such a comprehensive approach is highly important for evaluating ecosystem responses and improving land management strategies. This study focuses on the Qilian Mountains, a typical alpine grassland region in Northwest China. On the basis of field observations and simulations using the DNDC model, we analysed the spatiotemporal dynamics of SOCD, STND, and the C/N ratio from 1975 to 2024. To characterize the coordinated changes in soil carbon and nitrogen under shared environmental influences, principal component analysis was applied to extract the first principal component of SOCD and STND. This component was subsequently used to construct a composite index, i.e., SCND, which reflects their synergistic accumulation. Partial least squares structural equation modelling (PLS–SEM) was used to evaluate the direct and indirect effects of environmental and anthropogenic factors on SCND. These factors include climate conditions, topographic features, soil physicochemical properties, and human disturbance. To guide the analyses, this study addresses the following testable questions: (1) How have SOCD, STND, and the C/N ratio changed spatially and temporally in the Qilian Mountain grasslands from 1975 to 2024? (2) Do SOCD and STND exhibit coordinated variation, and can this coordination be effectively captured by a composite indicator (SCND) derived from principal component analysis? (3) To what extent do climate conditions, topography, soil physicochemical properties, and human disturbance directly or indirectly influence the coordinated changes in soil carbon and nitrogen? (4) How do the spatial patterns and vertical distribution of the C/N ratio further reflect the mechanisms of carbon–nitrogen coupling in alpine grassland soils? The main innovations of this study are as follows: (1) The DNDC model is applied to simulate long-term soil carbon and nitrogen dynamics in the Qilian Mountain grasslands, providing methodological support for regional ecological assessment; (2) SCND is developed as a composite index to represent the coordinated state of soil carbon and nitrogen, thereby improving the understanding of carbon–nitrogen dynamics in alpine environments. These contributions offer new insights into soil carbon and nitrogen changes in fragile alpine ecosystems and support improved grassland management under climate change. More broadly, this study enhances the understanding of carbon–nitrogen coupling mechanisms in mountain grassland ecosystems and provides a scientific basis for improving ecosystem resilience and sustainable management in alpine semiarid regions. Materials and methods Study area The study area is located in the Qilian Mountains in northwestern China (Fig. 1 a), spanning the border zone between Gansu and Qinghai provinces (94°10′–103°04′ E, 35°50′–39°19′ N) and covering approximately 224,100 km 2 . This region lies at the intersection of the Qinghai–Tibet Plateau, the Mongolian Plateau, and the Loess Plateau, and is jointly shaped by high elevations and a continental climate, resulting in distinctive alpine semiarid environmental conditions. The elevation ranges from approximately 1,000 m to 6,000 m, contributing to substantial spatial heterogeneity in temperature and precipitation. The mean annual temperature is approximately 0 °C, while the annual precipitation varies from 300 to 600 mm and generally decreases from east to west. The dominant soil textures include sandy soil, loam, sandy loam, and calcareous loam. Fig. 1. Open in a new tab Study area and sampling sites in the Qilian Mountains. ( a ) Location, elevation, and meteorological stations in the Qilian Mountains. ( b ) Grassland types and distribution of sites. Dark green indicates alpine steppe, and light green indicates alpine meadow. Brown points: sites from published studies; black points: field sampling sites The Qilian Mountain grasslands also exhibit notable lithological and pedological diversity. Parent materials originate mainly from mixtures of metamorphic rocks, sedimentary strata, and quaternary unconsolidated deposits, which provide various inorganic components that influence soil development. Common soil textures include sandy soil, loam, sandy loam, and calcareous loam. According to the World Reference Base for Soil Resources, the dominant soil groups in the region are Cambisols, Leptosols, and Calcisols, characterized by shallow profiles, weak horizon development, and localized carbonate accumulation typical of alpine semiarid environments. Natural grasslands are widely distributed across the region and are dominated by alpine steppe and alpine meadow ecosystems (Fig. 1 b). Common plant genera include Stipa, Kobresia, Poa, and Festuca, which contribute most of the aboveground biomass and belowground inputs. These vegetation communities play an essential role in regulating organic matter inputs and carbon–nitrogen cycling in regional soil system. Field data collection A total of 90 soil samples were collected from 10 field sites in the Qilian Mountains during 2022, and an additional 90 samples were collected from the same sites in 2023 using identical coordinates. Here, 90 refers to the total number of soil subsamples, while the 10 sites represent the distinct field sampling locations. At each site, three 5 m × 5 m plots were randomly established, and samples were taken from three soil depths (0–10 cm, 10–20 cm, and 20–30 cm) within each plot, yielding 9 samples per site and 30 plots in total. After surface debris was removed, the soil samples were collected using a ring knife and immediately sealed in self-sealing bags to minimize moisture loss. The samples were transported under consistent conditions to the laboratory, where 50 g of each sample was weighed into clean aluminium boxes, oven-dried at 105 °C for 24 h, and analysed for soil moisture content, bulk density, and pH. For the analysis of organic carbon and total nitrogen, the soil samples were air-dried at room temperature, ground, and passed through a 0.15 mm sieve. Afterwards, 500 mg of the sieved soil was weighed and analysed using a fully automated nitrogen-form carbon analyser (Skalar Primacs SNC-100, ho-IC-E) at 1100 °C. In addition, 18 records of organic carbon density (0–20 cm soil depth) from published studies between 2003 and 2013 [ 17 – 28 ] were integrated to enrich the dataset and expand the spatial coverage. Each record corresponds to a sampling point (Additional file 2: Table S1). Although literature data were obtained using the potassium dichromate oxidation method, previous studies have shown strong agreement with our analytical method [ 29 ]. Therefore, the two datasets were considered comparable and were jointly used for model calibration and validation. Basic data In this study, multisource geospatial datasets covering meteorology, soil, vegetation, management, grazing, and topography were integrated. All the data were classified into three categories: field sampling, published literature, and open-access datasets. The open datasets were obtained from national authoritative scientific data centres and specialized databases maintained by leading research institutions in China, and underwent quality checks, resampling, and reprojection as needed. All the datasets were standardized to the GCS_WGS_1984 coordinate system to ensure spatial consistency. The meteorological data were derived from two sources depending on the time period. Soil data included land use, texture, bulk density, pH, and clay content from open datasets; soil organic carbon content (SOC), soil total nitrogen content (STN), bulk density, and pH from field measurements; and SOC from published studies. Land use data were used as inputs to the DNDC model. For this purpose, raster land use grids were extracted using ArcGIS for each of the 28 sampling sites and incorporated into the model. All the sites remained classified as grassland throughout the 50-year period, confirming land use consistency and ensuring the reliability of this input. For other model inputs, measured SOC, bulk density, and pH were used at the field-sampled sites. For the literature-based sites, SOC and STN were taken directly from publications, whereas the other required variables were extracted from the raster datasets at the corresponding locations using ArcGIS. Vegetation and management data were obtained from local statistical yearbooks for Qilian County. Owing to the high spatial heterogeneity in mountainous regions, these parameters were adjusted during model localization. The topographic data were based on 30 m digital elevation models (DEMs) from the National Cryosphere Desert Data Center and were precisely extracted for each sampling site using ArcGIS. Elevation, as a spatial variable, does not change over time but acts as a spatial driver of carbon and nitrogen dynamics. In contrast, the grazing intensity varies over time and space; therefore, data were extracted for each site and sampling year using ArcGIS. The grazing intensity data were likewise extracted by site and year. Both datasets were used as input variables in the structural equation model to assess the effects of terrain and anthropogenic activity on soil carbon–nitrogen dynamics. The detailed data sources are listed in Table 1 . Table 1. Overview of data categories, sources, and temporal coverage in this study Category Content Type Format Source Years Meteorological Daily temperature and precipitation Open-access dataset NetCDF National Cryosphere Desert Data Center ( http://www.ncdc.ac.cn ) 1975–1994 Raster data China Meteorological Data Service Center ( http://data.cma.cn ) 1995–2024 Soil Land use Open-access dataset Raster data Earth Resource Data Cloud ( http://gis5g.com ) 1975–2024 Texture, bulk density, pH, and clay content Open-access dataset Raster data National Earth System Science Data Center ( https://www.geodata.cn ) 1975–2024 SOC, STN, bulk density, and pH Field sampling Tabular 10 sites, 90 samples 2022–2023 SOC Literature Tabular 18 sites, 18 records 2003–2013 Vegetation Grassland type Open-access dataset Vector data Chinese Ecosystem Assessment and Ecological Security Pattern Database ( http://www.ecosystem.csdb.cn ) — Growing season, rotation, and harvest method Tabular Local statistical yearbooks, adjusted as needed during model calibration 1975–2024 Management and grazing Fertilizer type/amount/depth/frequency, irrigation condition, and livestock numbers (beef, sheep, horses) Open-access dataset Tabular Local statistical yearbooks, adjusted as needed during model calibration 1975–2024 Grazing intensity Open-access dataset Raster data National Ecosystem Science Data Center ( http://www.nesdc.org.cn ) 1975–2024 Topography DEMs Open-access dataset Raster data National Cryosphere Desert Data Center ( http://www.ncdc.ac.cn ) — Open in a new tab Research methods DNDC model In this study, the DNDC model was employed to simulate the carbon and nitrogen dynamics of grassland soils. The DNDC model was selected because of its strong ability to capture biogeochemical responses to environmental drivers and land management practices at a daily time step. Compared with other commonly used models, such as CENTURY and RothC, which typically operate at coarser temporal resolutions and provide limited representation of grassland management, the DNDC model simulates key processes, such as decomposition, nitrification, denitrification, and gas fluxes, with greater temporal detail. Moreover, it incorporates site-specific anthropogenic activities including manure application, grazing, and fertilization, making it particularly suitable for the alpine grasslands of the Qilian Mountains, where both environmental and management conditions change seasonally. Model simulations were conducted continuously for the period 1975–2024 at a daily time step using climate, soil, vegetation, and management input data. Key management parameters were calibrated using available field measurements to improve simulation performance. The model simulates matter and energy exchanges among soil, vegetation, and the atmosphere to quantify key processes such as decomposition, nitrogen transformation, gas emissions, and nutrient transport [ 30 ]. Detailed model descriptions, structures, inputs, and outputs are provided in Additional file 1: Fig. S1 and Supplementary Text 1 [ 31 ]. Statistical analysis Model sensitivity was evaluated using the sensitivity index (SI) following [ 32 ], based on Monte Carlo simulations of key DNDC input parameters. The soil organic carbon content was converted to density using the method of [ 33 ], and total carbon and nitrogen stocks were estimated as described by [ 34 ]. The accuracy of the model simulations was assessed using Pearson’s correlation coefficient (r) and the Nash–Sutcliffe efficiency (NSE), with r > 0.80 indicating good agreement and NSE > 0.75 indicating satisfactory performance [ 35 ]. To extract long-term trends in carbon and nitrogen stocks, the 1975–2024 time series data were smoothed using both a three-year moving average and the LOWESS method. The moving average highlights general trends by reducing short-term fluctuations, whereas LOWESS captures local variations, enabling a complementary view of changes across multiple time scales. The optimal smoothing window was selected on the basis of the mean squared error and coefficient of variation (Additional file 2: Table S2). Spearman’s correlation coefficients were used to examine the relationships between key environmental variables and soil carbon and nitrogen. PLS‒SEM was applied to explore direct and indirect regulatory pathways. Statistical analyses were conducted in a stepwise manner, starting with trend and correlation analyses, followed by structural equation modeling to assess direct and indirect driving mechanisms. Geostatistical interpolation was performed using ordinary kriging. Spatial analyses were conducted in ArcGIS and Python, statistical analyses were performed in SPSS and R, and visualizations were generated using Origin. The main packages used include arcpy (Python) for spatial processing, statsmodels (Python) for LOWESS smoothing, and ggplot2, plspm, and stats (R) for visualization and PLS–SEM (Additional file 3). Unless otherwise stated, figures presenting spatial, temporal, or vertical patterns are based on descriptive statistics only. Results Simulation accuracy and parameter settings To enhance the applicability of the DNDC model in the Qilian Mountain grasslands, a sensitivity analysis was conducted to identify key input parameters affecting the simulation results. On the basis of local climate, soil, and management conditions, parameter ranges were defined using measured values, and 1,000 Monte Carlo simulations were performed. The sensitivity index was subsequently used to quantify the influence of each factor on the SOCD and STND outputs. The results (Table 2 ) revealed that most input factors were more sensitive to STND than to SOCD, with soil bulk density and pH being the most influential. With respect to SOCD, the sensitivity to air temperature and precipitation was relatively high, indicating that the model emphasizes the role of climate variability in regulating soil carbon dynamics. Table 2. Sensitivity analysis of the DNDC input parameters based on Monte Carlo simulations, with SI(SOCD) and SI(STND) representing sensitivity indices for soil organic carbon density and total nitrogen density Parameter Baseline value Perturbation range SI (SOCD) SI (STND) Air temperature (°C) 1.965 ± 2 0.4287 0.5572 Precipitation (mm) 316.84 ± 10% 0.1845 -0.1542 Bulk density (g·cm − 3 ) 1.14 ± 20% 0.1475 0.7629 SOC content (g·kg − 1 ) 0.0222 ± 20% 0.1141 0.5596 pH 8.43 ± 20% -0.1789 -0.9269 Manuring date (days) 12 ± 5 0.0477 0.1030 Manure C/N ratio 6 ± 40% 0.0204 -0.1107 Manure amount (kg·hm 2 ) 116.67 ± 30% 0.0889 0.2056 Grazing stock (num) 65 ± 20% 0.0009 0.0007 Grazing days (days) 184 ± 30 -0.0818 -0.1311 Grazing hours per day (h) 6 ± 20% 0.0020 0.0043 Open in a new tab Given the relative stability of climate and soil conditions, this study focused on calibrating highly sensitive and controllable parameters, particularly the amount of manure applied and the number of grazing days, to ensure that the simulated values closely matched the measured data. Field observations and published data supported the entire model calibration and validation process. The localized DNDC model effectively simulated SOCD and STND in the Qilian Mountain grasslands (Fig. 2 ), with correlation coefficients (r) of 0.95 and 0.89 and NSE values of 0.90 and 0.82, respectively, indicating high accuracy and reliability [ 36 ]. The calibrated model was then used to simulate long-term carbon and nitrogen dynamics from 1975 to 2024. The key input parameters are listed in Table 3 . Fig. 2. Open in a new tab Linear regression between simulated and measured values: ( a ) SOCD; ( b ) STND. The solid red line represents the fitted regression line, and the shaded area indicates the 95% confidence and prediction bands. Model performance was evaluated using r and NSE values Table 3. Main input parameters of the DNDC model for alpine grassland in the Qilian Mountains Parameter type Parameter name Value Unit Climate Simulated years 50 years Format for climate file(s) Jday, MaxT, MinT, and Prec °C; mm Soil Land-use Dry/Moist grassland — Nitrate 0.5 mg/kg Ammonium 0.05 mg/kg Crop Number of new crops 1 — Crop type 12 Perennial_grass — Cover crop No — Planting time March 1st date Harvest time August 31st date Transplanting No — Residue return ratio 0.2 fraction Manure Amendment Applications per year 12 times Manure 4 Slurry animal waste — Solid C/N ratio 6 — Graze Applications per year 3 times Grazing hours per day 4 hours Beef/veal 25 heads/ha Sheep 35 heads/ha Horse 5 heads/ha Excreta handle Deposite in field — Open in a new tab Spatiotemporal distribution of soil organic carbon and total nitrogen Spatiotemporal dynamics From 1975 to 2024, the 0–30 cm SOCD of grasslands in the Qilian Mountains exhibited pronounced spatial and temporal variations (Fig. 3 ). Spatially, areas with high SOCD values were concentrated mainly in the northern and east-central regions and tended to decrease towards the periphery, whereas areas with low SOCD values were distributed primarily along the southwestern margin. Temporally, SOCD showed a phased fluctuating pattern: from 1985 to 1994, SOCD increased relative to 1975–1984, with a relative change of 40.91%, indicating an increase in carbon sequestration capacity; from 1995 to 2004, it declined by 21.96%, likely because of climate variability and vegetation degradation; from 2005 to 2014, it rebounded by 9.18%; and from 2015 to 2024, it further increased by 27.51%, reaching the highest level on record, suggesting that the grassland carbon sink function continued strengthening. Fig. 3. Open in a new tab Spatial distribution of grassland SOCD in the Qilian Mountains, showing average values by decade and over the 50-year period The overall STND trend generally aligned with that of SOCD, albeit with some temporal deviations (Fig. 4 ). Notably, STND continued to increase between 1995 and 2004, whereas SOCD decreased during the same period. Over the past five decades, both soil organic carbon and total nitrogen densities in the 0–30 cm layer of the Qilian Mountain grasslands exhibited broadly similar temporal variations and considerable spatial overlap, suggesting a degree of synchrony and potential coupling between carbon and nitrogen cycling processes. Fig. 4. Open in a new tab Spatial distribution of grassland STND in the Qilian Mountains, showing average values by decade and over the 50-year period Stock trends To reduce interannual variability and better identify long-term trends in 0–30 cm soil carbon and nitrogen stocks, both three-year moving average and LOWESS smoothing were applied. The moving average emphasizes overall trends by minimizing short-term fluctuations, whereas LOWESS captures finer-scale variations, and their complementary use enables a more comprehensive understanding of carbon and nitrogen dynamics [ 37 , 38 ]. Over the past 50 years, the 0–30 cm soil organic carbon stocks (SOCS) in the grasslands of the Qilian Mountains has exhibited a staged fluctuating pattern (Fig. 5 a). The LOWESS-smoothed trend revealed four distinct phases: an initial fluctuating increase, a subsequent decline, a period of stabilization, and a final upward trend. Fig. 5. Open in a new tab Long-term trends in 0–30 cm soil organic carbon ( a ) and total nitrogen ( b ) stocks in grasslands of the Qilian Mountains from 1975 to 2024, showing the raw data, 3-year moving average, and LOWESS-smoothed results. The trend lines illustrate temporal patterns only, and no inferential statistical tests were applied Since 1975, the SOCS in the 0–30 cm layer of alpine grasslands in the Qilian Mountains was 1.16 × 10 9 t. From 1975 to 1978, the SOCS increased with fluctuations, mainly because of low human disturbance under prevailing social conditions. This allowed natural recovery of the grassland ecosystem, with enhanced plant productivity increasing carbon inputs. Relatively warm climate conditions during this period also favoured SOC accumulation. Between 1979 and 1987, the SOCS remained stable as grassland management became more institutionalized, leading to a new dynamic balance between carbon input and output. After the establishment of a national nature reserve in 1988, the SOCS increased steadily [ 39 ], reaching a local peak of 3.41 × 10 9 t in 1994. The subsequent decline may be attributed to carbon sequestration saturation, weakened reserve management, and warming-induced drought stress. By 2016, although regional warming continued, sustained ecological restoration projects and consistent protection policies helped stabilize the SOCS. Following the implementation of the national park system in 2017, grassland protection was further strengthened, and the SOCS continued to rise, reaching 5.57 × 10 9 t by 2024 [ 40 ], demonstrating a clear temporal alignment between ecological policies and long-term carbon accumulation. During the same period, the soil total nitrogen stocks (STNS) also fluctuated (Fig. 5 b). In the early years, a peak occurred, which was likely influenced by processes such as biological nitrogen fixation by vegetation. Between 1979 and 1987, the STNS declined but then rebounded, reflecting the heightened sensitivity of the nitrogen cycle to fluctuations in precipitation, mineralization rates, and anthropogenic disturbances. After 1988, the STNS increased gradually, with growth accelerating after 2017. Notably, although the SOCS and STNS experienced a short-term increase following the establishment of the nature reserve in 1988, the increase lasted only four years before declining again. This suggests that ecological protection efforts must be long-term and sustained, as short-term interventions may yield only temporary effects [ 41 ]. Moreover, we also observed consistent 5–8-year oscillation patterns for both variables. These short-term cycles are unlikely to be entirely driven by discrete management events. Rather, they may reflect inherent feedback mechanisms within the DNDC model, such as nonlinear interactions among soil organic matter decomposition, nitrogen mineralization, and plant uptake processes, which can lead to delayed responses and damped oscillations. In addition, natural interannual variability in climate drivers may also contribute. Overall, despite differences in the mechanisms governing carbon and nitrogen stock dynamics, their long-term trends were broadly consistent, implying that both may be driven by similar environmental factors and exhibit synergistic responses. Factor identification and pathway analysis Correlation analysis and indicator construction To identify the key factors influencing SOCD and STND, eight representative variables were selected on the basis of the DNDC model input parameters, data availability, and model limitations, covering climate, topography, soil physicochemical properties, and anthropogenic disturbances. Core drivers, such as temperature, precipitation, pH, bulk density, and clay content, were chosen on the basis of their roles as input variables in the DNDC model and their ecological relevance. To account for topographic and anthropogenic effects not explicitly simulated in the DNDC model, elevation, slope, and grazing intensity were included in the PLS–SEM analysis. In the PLS–SEM framework, variables are classified as either latent or observed. Latent variables are conceptual constructs that cannot be measured directly, whereas observed variables can be directly quantified. In this study, the SCND model was constructed as a latent variable on the basis of both SOCD and STND. All the other variables were treated as observed variables (Fig. 6 ). Spearman’s rank correlation coefficients were calculated on the basis of the results of the Shapiro–Wilk test [ 42 , 43 ], providing a theoretical basis for path specification in the structural equation model. Fig. 6. Open in a new tab Heatmap of correlations between SOCD, STND, and their environmental drivers. Variables: SOCD (soil organic carbon density), STND (soil total nitrogen density), MAT (mean annual temperature), MAP (mean annual precipitation), ASL (altitude above sea level), SG (slope gradient), pH (soil pH), BD (bulk density), CC (clay content), and GI (grazing intensity). Correlations were assessed using Spearman’s correlation coefficients, color intensity indicates the strength and direction of the correlations Given the strong positive correlation between SOCD and STND, along with their highly similar spatial and temporal patterns and high variance inflation factors indicating severe multicollinearity, we combined the two variables into a single composite indicator, namely, the SCND index. Principal component analysis was used for dimensionality reduction, and the first principal component, which explained 98.7% of the original variance (Additional file 1: Fig. S2), was adopted to represent SCND for subsequent structural modelling [ 44 ]. As an integrated indicator of the synergistic accumulation of soil carbon and nitrogen, SCND captures the coupling dynamics of organic matter and nitrogen during long-term ecological processes. Because carbon and nitrogen cycles in soils are often regulated by similar environmental drivers, the synchronized variation in SCND reflects the ecosystem’s response to both climatic and anthropogenic disturbances [ 45 ]. This composite metric provides a comprehensive basis for understanding the ecological functioning of alpine grasslands [ 46 ]. Results of partial least squares structural equation modelling The results of the PLS‒SEM pathway analysis (Fig. 7 ) (Additional file 2: Table S3) revealed that SCND was influenced by multiple environmental factors, with altitude above sea level (ASL) and soil bulk density (BD) identified as key drivers. The altitude had a negative effect on SCND by directly suppressing microbial activity and vegetation growth and indirectly modulating carbon and nitrogen accumulation through its effects on air temperature, precipitation, and soil pH. Bulk density, acting as a central mediating variable, served as a crucial bridge across multiple regulatory pathways. Fig. 7. Open in a new tab PLS‒SEM of environmental factors related to SCND in the Qilian Mountain grasslands. Numbers indicate path coefficients; solid lines represent nonsignificant paths; dashed lines represent nonsignificant paths; and line thickness reflects the relative magnitude of the coefficients. Asterisks denote significance levels (* p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001). R 2 values for all endogenous variables are shown in the figure to indicate the explanatory power of the structural model. Path significance was assessed using bootstrapping Specifically, higher elevations lead to lower temperatures, which suppress plant growth and organic matter input while reducing soil porosity, thereby increasing soil bulk density. Temperature fluctuations also influence grazing intensity by regulating the growth cycle of pasture vegetation: rising temperatures can extend the growing season, increase vegetation productivity, and increase the livestock carrying capacity of grasslands, ultimately intensifying grazing pressure. In contrast, increased precipitation creates colder and wetter conditions, delaying snowmelt and green-up, which shorten the effective grazing period and suppress grazing activity. Furthermore, ecological protection measures, such as the establishment of national park systems, impose institutional restrictions on grazing, limiting herders’ ability to adjust grazing intensity in response to precipitation and thereby obscuring its natural regulatory effect. In contrast, the timing of grazing is often based on temperature rise; under fixed management windows, grazing behaviour becomes more directly responsive to temperature variation. This results in an asymmetric regulatory mechanism that deviates from the conventional understanding that “precipitation enhances grazing intensity” [ 47 ] and underscores the distinct response of the Qilian grasslands to hydrothermal variability. In areas with low organic matter content, grazing tends to disrupt soil structure and reduce bulk density [ 48 ], thereby influencing the SCND model through the reverse pathway. Elevation also affects soil acidification processes by altering temperature, which decreases the soil pH and weakens the aggregation capacity of alkaline cations, indirectly modifying soil structure and physicochemical properties [ 49 ]. Soils with lower bulk density, characterized by higher porosity and better aeration, are more conducive to microbial activity and nutrient uptake by plant roots, thereby increasing the efficiency of carbon and nitrogen accumulation and ultimately increasing SCND. All the variables were standardized prior to the PLS–SEM estimation, so the reported path coefficients represent standardized effects and are directly comparable (Table 4 ). Bootstrapping results based on 5,000 resamples indicated that the total effect of elevation on SCND was − 0.224, consisting of a strong negative direct effect and a substantial positive indirect effect, reflecting a dual regulatory mechanism. Soil bulk density had a significant negative effect, confirming its direct inhibitory role in soil carbon and nitrogen accumulation. Among the other variables, MAT and MAP influenced SCND exclusively through indirect pathways. Although soil pH had a positive direct effect, it was offset by a negative indirect effect, resulting in a nonsignificant total effect. Grazing intensity had a positive but nonsignificant indirect effect on SCND, suggesting that human disturbance may have latent regulatory effects on carbon and nitrogen cycling processes. Overall, the PLS-SEM results highlight elevation and soil bulk density as the dominant direct controls on soil carbon–nitrogen coupling, whereas climatic and anthropogenic factors mainly act through indirect pathways. Table 4. Total, direct, and indirect effects of factors on soil carbon and nitrogen density (SCND) estimated using partial least squares structural equation modelling Factor Total effect Direct effect Indirect effect MAT -1.618 -0.718 -0.900 MAP 0.094 0.000 0.094 ASL -0.224 -1.193 0.970 BD -1.329 -1.329 0.000 Ph 0.229 0.834 -0.604 GI 0.615 0.000 0.615 Open in a new tab Total effect = Direct effect + Indirect effect Spatial and vertical variability in the soil C/N ratio Spatial distribution patterns While PLS–SEM identified the main driving mechanisms of SCND, this composite indicator primarily reflects the synergistic density state of soil carbon and nitrogen, which is insufficient to fully capture their coupling relationship and dynamic balance. The C/N ratio, which is derived from SOC and STN, serves as a key indicator of soil organic matter quality and nutrient balance and provides a more intuitive representation of carbon–nitrogen coupling characteristics [ 50 ]. To gain a more comprehensive understanding of its evolution, this study further examined the spatiotemporal distribution of the C/N ratio and its phase transition patterns. In 1975, the C/N ratio was generally high (Fig. 8 ), with notably elevated values in the southeastern region compared with those in the high-elevation northwestern and southwestern margins. This pattern indicated a carbon–nitrogen imbalance, likely driven by ample carbon input and delayed nitrogen accumulation. By 2000, the C/N ratios had generally declined, with the southeastern region becoming a low-value zone, while elevated values appeared along the southwestern margin. This pattern indicates a spatial restructuring of carbon and nitrogen accumulation, potentially influenced by shifts in regional precipitation. By 2024, spatial disparities in C/N ratios had become even more pronounced. Fig. 8. Open in a new tab Spatial distribution of the 0–30 cm soil C/N ratio in grasslands of the Qilian Mountains during different time periods Vertical distribution patterns The C/N ratio decreased with soil depth (Fig. 9 ), with the surface layer showing the highest mean value and variability, reflecting the dynamic nature of carbon and nitrogen inputs and transformations near the surface. In contrast, the lower mean values and reduced variability in the 10–30 cm layers suggest a more stable carbon and nitrogen status in the subsoil, primarily shaped by the gradual decomposition of organic matter and its stabilization through microbial processing and mineral association. These patterns indicate increasing consistency in carbon–nitrogen relationships with depth. Fig. 9. Open in a new tab Vertical distribution of the C/N ratio in grassland soils across different depths (0–10 cm, n = 1400; 10–20 cm, n = 1400; 20–30 cm, n = 1400). The five horizontal lines in each box plot represent the maximum value, upper quartile, median, lower quartile, and minimum value. The points inside the box indicate the mean C/N ratio for each soil layer. The whiskers extend to the maximum and minimum values. The adjacent density plots illustrate the distribution of C/N ratios, with the symmetry axis aligned to the mean. No formal statistical tests were applied; the analysis is descriptive and intended to illustrate vertical distribution patterns The C/N ratios across the soil layers were significantly positively correlated, with the strongest correlation observed between the 10–20 cm and 20–30 cm layers (Fig. 10 ). This pattern reflects strong vertical continuity in carbon and nitrogen cycling, shaped by long-term ecological processes and climatic conditions. Such consistency enhances the stability of the soil carbon–nitrogen balance and indicates that disturbances in surface layers may propagate downwards, highlighting the importance of maintaining the integrity of the entire soil profile. Fig. 10. Open in a new tab Pairwise correlations and distribution patterns of C/N ratios across soil depths. The diagonal panels show the distribution of C/N ratios within each soil layer using histograms and density curves. The off-diagonal panels display scatter plots comparing C/N ratios between soil layers, with fitted lines indicating linear trends and dashed ellipses outlining the data spread. The Pearson correlation coefficients and adjusted R 2 values describe the strength of the relationships Discussion In this study, the spatiotemporal dynamics of SOCD and STND in the 0–30 cm soil layer of the Qilian Mountain grasslands, as well as changes in soil carbon and nitrogen stocks over the past 50 years, were systematically evaluated, with a particular focus on their responses to environmental and anthropogenic factors. From 1975 to 2024, the spatial distributions of SOCD and STND increased in the northern and east-central regions but gradually decreased in the southwest. This spatial pattern closely mirrors the regional differentiation reported for shrub community distributions in the alpine zone of the Qilian Mountains [ 51 ], indicating that similar elevation-related environmental gradients consistently shape both vegetation patterns and soil carbon and nitrogen accumulation in this region. Overall, the carbon and nitrogen stocks exhibited a fluctuating trend, tending to increase, decrease, stabilize, and increase again, with strong temporal and spatial synchronization. Traditional studies often analyse carbon and nitrogen separately and overlook their coregulatory mechanisms. To more comprehensively reflect the soil carbon–nitrogen status, this study developed a synthetic indicator, namely SCND, to reveal the coupled carbon–nitrogen dynamics driven by multiple factors. This approach offers a more holistic representation of soil nutrient processes and ecosystem responses. Structural equation modelling based on SCND revealed that elevation and soil bulk density were the dominant direct drivers and were both negatively associated with carbon and nitrogen accumulation, highlighting the importance of topography and soil structure. Similar to path analysis results reported for different land-use types on the southern slope of the Qilian Mountains [ 52 ], our findings indicate that climatic factors, soil pH, and grazing intensity primarily influence soil carbon and nitrogen through indirect pathways. By explicitly integrating carbon and nitrogen into a coupled indicator, the present study further demonstrates the coordinated nature of these indirect effects across regional grassland systems. In light of topographic and soil structural features, the spatial coherence of carbon and nitrogen patterns can be interpreted as arising from the combined influence of multiple environmental controls. The northern and east-central regions, characterized by lower elevations and relatively warmer conditions, are dominated by highly productive alpine meadows with greater vegetation cover, higher organic inputs, and looser soil structure, which together promote carbon and nitrogen accumulation. In contrast, the southwestern region, associated with higher elevation, a colder and drier climate, sparser vegetation, and higher soil bulk density, exhibits reduced capacity for organic input and nutrient retention. This spatial differentiation suggests that topography and soil physical structure jointly shape the spatial distribution of carbon and nitrogen. Therefore, in high-elevation areas, water and nutrient retention measures are recommended to improve soil physical properties and increase carbon and nitrogen accumulation. Although we discuss the roles of topography and soil structure, only elevation and soil bulk density significantly affected the PLS–SEM results. Therefore, interpretations related to topographic and structural influences are limited and should be considered indirect. In addition, we found that grazing activity in the Qilian alpine grasslands was more responsive to temperature than to precipitation because of the region’s dependence on snowmelt-driven regreening. This response pattern differs from the commonly reported precipitation-dominated grazing dynamics in many grassland systems, where grazing activity closely tracks rainfall variability. In the Qilian Mountains, institutional restrictions under national park policies further reduced the adaptive response to precipitation variability [ 53 ]. These results suggest an asymmetric regulatory effect of hydrothermal factors on grazing, which should be considered in future management policies [ 54 , 55 ]. To improve management flexibility under climate variability, monitoring ground and air temperatures is recommended to identify critical phenological thresholds for pasture greening and senescence and to implement a “flexible grazing period” as an adaptive alternative to fixed-date grazing systems. Our analysis of the C/N ratio revealed that high values were detected mainly in the western and central grasslands, indicating that carbon accumulation outpaced nitrogen release, which may have inhibited vegetation recovery [ 56 ]. Conversely, low C/N ratios generally promote nutrient cycling, but overly low values could increase the risk of carbon loss and soil degradation [ 57 ]. To address these contrasting conditions, region-specific management strategies are advised. In areas with high C/N ratios, nitrogen input can be increased by the introduction of leguminous species or the application of organic fertilizers. In areas with low C/N ratios, grazing exclusion or low-intensity rotational grazing can promote organic matter accumulation and improve ecological resilience. Vertically, the carbon and nitrogen indicators were strongly consistent within the 0–30 cm soil layer. This vertical coupling aligns with previous observations on the Tibetan Plateau [ 58 ], indicating that surface disturbances can affect deeper soil dynamics. Hence, protecting the entire soil profile—especially by facilitating deeper carbon and nitrogen inputs—is critical. We recommend promoting the planting of deep-rooted native grasses and restricting heavy machinery to increase subsurface carbon and nitrogen sequestration. Although this study employed a multimodel approach and integrated diverse datasets, several limitations related to data inputs should be acknowledged. First, topography is not explicitly considered as an input variable in the DNDC model. To compensate for this, we incorporated elevation into the PLS–SEM to quantify its direct influence on SCND, thereby addressing this structural limitation and improving our understanding of topographic controls. Second, grazing intensity was not directly used in DNDC simulations, which used livestock numbers (i.e., beef, sheep, and horses) as input variables. For model calibration, a single set of livestock parameters was selected to best approximate regional-scale C and N dynamics. However, in reality, grazing pressure varies across sites. To capture this spatial heterogeneity, grazing intensity data were included as an explanatory variable in the PLS–SEM, allowing for a more refined analysis of the relationship between grazing and soil C and N status at the site level. Third, owing to the high elevation and rugged terrain in the Qilian Mountains, field sampling was logistically challenging, leading to a limited number of observation points. This may influence the robustness of the model validation and the generalizability of the findings. Future research should be grounded in long-term, multiregional observational data, with expanded measurements of key soil, vegetation, and management variables (e.g., soil texture, clay content, grassland type, and detailed grazing intensity records) to improve the accuracy and robustness of statistical and mechanistic analyses. A broader set of relevant factors should be incorporated to fully explore the pathways through which environmental and anthropogenic drivers influence soil carbon and nitrogen dynamics rather than being limited by model input constraints. Integrating detailed grazing management records (e.g., grazing prohibition, rotational grazing, and grass–livestock balance), high-resolution land-use data, and extended field surveys will help capture spatial heterogeneity and human impacts. Future work should also address the effects of freeze–thaw cycles and extreme climate events, which may trigger delayed or nonlinear responses. Additionally, extending the sampling depth beyond 30 cm—to 50 cm or more—will support a more complete understanding of subsurface carbon–nitrogen interactions and long-term sequestration. Furthermore, controlled scenario testing or comparative model analysis may help elucidate the mechanisms behind the observed 5–8-year oscillation patterns in the DNDC simulation outputs, providing deeper insight into model behaviour and soil biogeochemical feedbacks. Collectively, these efforts will improve the precision, completeness, and ecological relevance of research on alpine grassland carbon and nitrogen cycling. Conclusion This study revealed the coordinated spatiotemporal dynamics of soil organic carbon density (SOCD) and soil total nitrogen density (STND) in the Qilian Mountain grasslands from 1975 to 2024 based on DNDC model simulations. Both SOCD and STND exhibited fluctuating but overall increasing trends, with higher values concentrated in the northern and east-central regions and marked enhancement following major ecological protection policies. To characterize soil carbon–nitrogen coupling, a composite indicator (SCND) was developed. Structural equation modelling identified elevation and soil bulk density as the dominant direct controls, while temperature, precipitation, soil pH, and grazing intensity mainly acted through indirect pathways. The C/N ratio showed increasing spatial heterogeneity and strong vertical coupling within the 0–30 cm soil layer. Future research should prioritize long-term, multiregional field observations to support model validation and improve understanding of soil carbon and nitrogen dynamics. Permanent monitoring plots should be established across representative elevation gradients and management regimes, with expanded measurements of soil properties, vegetation characteristics, and grazing intensity. From a management and policy perspective, several recommendations can be made. Conservation and restoration efforts should focus on low- and mid-elevation grasslands. Measures to reduce soil compaction should be implemented, including restrictions on heavy machinery use and appropriate regulation of grazing pressure. Grazing management should incorporate flexible grazing periods aligned with seasonal vegetation dynamics. Spatial variation in soil C/N ratios should be explicitly considered in management decisions, with nitrogen inputs adjusted in high C/N areas and grazing intensity controlled in low C/N areas. Overall, this study provides a scientific foundation for integrating modelling approaches with long-term field monitoring to support sustainable management of mountain grassland ecosystems under ongoing environmental change. Supplementary Information Supplementary Material 1 (27.4KB, docx) Supplementary Material 2 (22.9KB, docx) Supplementary Material 3 (211KB, docx) Author contributions All the authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by [Jiafang Cai], [Meiling Zhang], [Youyi Zhao], [Yanjun Gong] and [Hanying Wang]. The first draft of the manuscript was written by [Jiafang Cai], and all the authors commented on previous versions of the manuscript. All the authors read and approved the final manuscript. Funding This work was supported by the National Natural Science Foundation of China (No. 32260353) and the High-End Foreign Expert Introduction Program of the Ministry of Science and Technology of China (No. G2022042009L). Data availability The data sources are indicated in the text, and all the data are available upon request from the authors. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Consent was given by all the authors. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References 1. Pepin NC, Arnone E, Gobiet A, Haslinger K, Kotlarski S, Notarnicola C et al. Climate Changes and Their Elevational Patterns in the Mountains of the World. Rev Geophys. 2022;60: e2020RG000730. 2. Singh PK, Chudasama H. Pathways for climate change adaptations in arid and semi-arid regions. J Clean Prod. 2021;284:124744. [ Google Scholar ] 3. Farooq TH, Yan W, Chen X, Shakoor A, Rashid MHU, Gilani MM, et al. Dynamics of canopy development of Cunninghamia lanceolata mid-age plantation in relation to foliar nitrogen and soil quality influenced by stand density. Global Ecol Conserv. 2020;24:e01209. [ Google Scholar ] 4. Prasad S, Malav LC, Choudhary J, Kannojiya S, Kundu M, Kumar S et al. Soil Microbiomes for Healthy Nutrient Recycling. 2021. pp. 1–21. 5. Elbasiouny H, El-Ramady H, Elbehiry F, Rajput VD, Minkina T, Mandzhieva S. Plant Nutrition under Climate Change and Soil Carbon Sequestration. Sustainability. 2022;14:914. [ Google Scholar ] 6. Jílková V, Jandová K, Kukla J. Responses of microbial activity to carbon, nitrogen, and phosphorus additions in forest mineral soils differing in organic carbon content. Biol Fertil Soils. 2021;57:513–21. [ Google Scholar ] 7. Lawlor DW. Carbon and nitrogen assimilation in relation to yield: mechanisms are the key to understanding production systems. J Exp Bot. 2002;53:773–87. [ PubMed ] [ Google Scholar ] 8. Geng Y, Ding Y, Zhou P, Wang Z, Peng C, Li D. Soil microbe-mediated carbon and nitrogen cycling during primary succession of biological soil crusts in tailings ponds. Sci Total Environ. 2023;894:164969. [ DOI ] [ PubMed ] [ Google Scholar ] 9. Zhang C, Yan R, Liang Q, Na R, Li T, Yang X, et al. Study on soil physical and chemical properties and carbon and nitrogen sequestration of grassland under different utilization modes. Acta Prataculturae Sinica. 2021;30:90–8. [ Google Scholar ] 10. Mason ARG, Salomon MJ, Lowe AJ, Cavagnaro TR. Microbial solutions to soil carbon sequestration. J Clean Prod. 2023;417:137993. [ Google Scholar ] 11. WANI OA, KUMAR SS, HUSSAIN N, WANI AIA, BABU S, ALAM P, et al. Multi-scale processes influencing global carbon storage and land-carbon-climate nexus: A critical review. Pedosphere. 2023;33:250–67. [ Google Scholar ] 12. Chen P, Xie Y, Ren X, Cheng C, Wei X. Spatial variation of soil organic carbon density in the black soil region of Northeast China under the influence of erosion and deposition. J Clean Prod. 2024;475:143616. [ Google Scholar ] 13. Zhang G, Fu G. Changes in soil organic carbon, total nitrogen and total phosphorus in 2000–2020 and their driving mechanisms in Tibetan alpine grasslands. Glob Planet Change. 2024;239:104484. [ Google Scholar ] 14. Dai L, Ge J, Wang L, Zhang Q, Liang T, Bolan N, et al. Influence of soil properties, topography, and land cover on soil organic carbon and total nitrogen concentration: A case study in Qinghai-Tibet plateau based on random forest regression and structural equation modeling. Sci Total Environ. 2022;821:153440. [ DOI ] [ PubMed ] [ Google Scholar ] 15. Niu S, Song L, Wang J, Luo Y, Yu G. Dynamic carbon-nitrogen coupling under global change. Sci China Life Sci. 2023;66:771–82. [ DOI ] [ PubMed ] [ Google Scholar ] 16. Pan Y, Fang F, Tang H. Patterns and Internal Stability of Carbon, Nitrogen, and Phosphorus in Soils and Soil Microbial Biomass in Terrestrial Ecosystems in China: A Data Synthesis. Forests. 2021;12:1544. [ Google Scholar ] 17. Cao G, Tang Y, Mo W, Wang Y, Li Y, Zhao X. Grazing intensity alters soil respiration in an alpine meadow on the Tibetan plateau. Soil Biol Biochem. 2004;36:237–43. [ Google Scholar ] 18. Cao S, Chen K, Cao G, Zhu J, Lu B, Wang J. Effects of grassland degradation on soil carbon density of Kobresia meadow in the Qinghai Lake Basin. Res Soil Water Conserv. 2014;21:71–5. [ Google Scholar ] 19. Li X-G, Li F-M, Zed R, Zhan Z-Y, Bhupinderpal-Singh. Soil physical properties and their relations to organic carbon pools as affected by land use in an alpine pastureland. Geoderma. 2007;139:98–105. [ Google Scholar ] 20. Cao S, Chen K, Cao G, Zhu J, Lu B, Zhang T, et al. Distribution characteristics of soil organic carbon density of Kobresia humilis meadow in the Qinghai Lake Basin. Acta Ecol Sin. 2014;34:482–90. [ Google Scholar ] 21. Li Y. Response of alpine grassland to land use pattern changes on the Qinghai–Tibet Plateau. Gansu Agricultural University; 2013. 22. Liu W. Study on soil organic carbon storage in natural grassland ecosystem in the Loess Plateau. Graduate University of Chinese Academy of Sciences, Ministry of Education Key Laboratory of Soil and Water Conservation and Ecological Environment; 2011. 23. Qi B. Study on the soil carbon storage of alpine grassland under different degrees of degradation in Qinghai Lake region. Gansu Agricultural University; 2005. 24. Qin Y, Yi S, Ren S, Li N, Chen J. Responses of typical grasslands in a semi-arid basin on the Qinghai-Tibetan Plateau to climate change and disturbances. Environ Earth Sci. 2014;71:1421–31. [ Google Scholar ] 25. Yang C, Long R, Chen X, Xu C, Wang J. Characteristics of topsoil carbon, nitrogen and phosphorus density under different alpine grassland types in eastern Qilian Mountains. Chin J Grassland. 2008;1:1–5. 26. Yang X, Huang M, Wang J, Liu H. Relationship between belowground biomass and environmental factors in alpine grasslands of the Qinghai–Tibet Plateau. Acta Ecol Sin. 2013;33:2032–42. [ Google Scholar ] 27. Zhang P, Zhang T, Chen N. Characteristics and influencing factors of soil carbon and nitrogen distribution along vertical zones in the northern slope of the Qilian Mountains. Chin J Appl Ecol. 2009;20:518–24. [ PubMed ] [ Google Scholar ] 28. Zhang X, Sheng Y, Li J, Wu J, Chen J, Cao Y. Changes of alpine ecosystem along the ground temperature of permafrost in the source region of Datong River in the northeastern Qinghai-Tibet Plateau. J Food Agric Environ. 2012;10(1 Part 2):970–6. [ Google Scholar ] 29. Bao S. Soil and Agricultural Chemistry Analysis. 3rd ed. Beijing: China Agriculture; 2000. [ Google Scholar ] 30. Li C, Frolking S, Harriss R. Modeling carbon biogeochemistry in agricultural soils. Global Biogeochem Cycles. 1994;8:237–54. [ Google Scholar ] 31. Li C. User’s Guide for the DNDC Model (Version 95). Institute for the Study of Earth, Oceans, and Space, University of New Hampshire. 2000. https://www.dndc.sr.unh.edu/model/GuideDNDC95.pdf . Accessed 5 Jun 2025. 32. Lembaid I, Moussadek R, Mrabet R, Bouhaouss A. Modeling Soil Organic Carbon Changes under Alternative Climatic Scenarios and Soil Properties Using DNDC Model at a Semi-Arid Mediterranean Environment. Climate. 2022;10:23. [ Google Scholar ] 33. Li J, Yu X, Xia S, Zhao W, Wang S, Xu C. Spatial distribution characteristics of soil organic carbon density and storage in the Baiyangdian wetland area. Acta Ecol Sin. 2020;40:8928–35. [ Google Scholar ] 34. Zhang Y, Wang H, Gao Z, Qiu H. Estimation of soil organic carbon and total nitrogen density and storage in forest ecosystems of South China. Soils Fertilizers Sci China. 2024;12:17–26. 35. Nash JE, Sutcliffe JV. River flow forecasting through conceptual models part I — A discussion of principles. J Hydrol (Amst). 1970;10:282–90. [ Google Scholar ] 36. Moriasi DN, Arnold JG, Van Liew MW, Bingner RL, Harmel RD. Veith. Model Evaluation Guidelines for Systematic Quantification of Accuracy in Watershed Simulations. Trans ASABE. 2007;50:885–900. [ Google Scholar ] 37. Lovric M, editor. International Encyclopedia of Statistical Science. Berlin, Heidelberg: Springer Berlin Heidelberg; 2011.pp.866–9. 38. Cleveland WS. Robust Locally Weighted Regression and Smoothing Scatterplots. J Am Stat Assoc. 1979;74:829. 10.2307/2286407. [ Google Scholar ] 39. State Council of China. Establishment of Qilian Mountain National Nature Reserve. Beijing; 1988. 40. General Office of the CPC Central Committee; General Office of the State Council. Overall Plan for Establishing the National Park System. Beijing; 2017. 41. Bennett AF, Haslem A, Cheal DC, Clarke MF, Jones RN, Koehn JD, et al. Ecological processes: A key element in strategies for nature conservation. Ecol Manage Restor. 2009;10:192–9. [ Google Scholar ] 42. SHAPIRO SS, WILK MB. An analysis of variance test for normality (complete samples). Biometrika. 1965;52:591–611. [ Google Scholar ] 43. Spearman C. The Proof and Measurement of Association between Two Things. Am J Psychol. 1904;15:72. [ Google Scholar ] 44. Abdi H, Williams LJ. Principal component analysis. WIRE Comput Stat. 2010;2:433–59. [ Google Scholar ] 45. Han Y, Jia Y, Wang G, Tan Q, Liu X, Chen C. Coupling of soil carbon and nitrogen dynamics in drylands under climate change. Catena (Amst). 2023;221:106735. [ Google Scholar ] 46. Xiang Z, Xue Y, Guo W, Hartman MD, Liu Y, Parton WJ. Development of a plant carbon–nitrogen interface coupling framework in a coupled biophysical-ecosystem–biogeochemical model (SSiB5/TRIFFID/DayCent-SOM v1.0). Geosci Model Dev. 2024;17:6437–64. [ Google Scholar ] 47. Zhu Y, Shen H, Akinyemi DS, Zhang P, Feng Y, Zhao M, et al. Increased precipitation attenuates shrub encroachment by facilitating herbaceous growth in a Mongolian grassland. Funct Ecol. 2022;36:2356–66. [ Google Scholar ] 48. Gao Y, Han X, Wang S. Effects of grazing on grassland soil. Acta Ecol Sin. 2004;4:790–7. 49. Ratier Backes A, Frey L, Arévalo JR, Haider S. Effects of Soil Properties, Temperature and Disturbance on Diversity and Functional Composition of Plant Communities Along a Steep Elevational Gradient on Tenerife. Front Ecol Evol. 2021;9:758160. 50. Tong R, Wu T, Jiang B, Wang Z, Xie B, Zhou B. Soil carbon, nitrogen, and phosphorus stoichiometry and its influencing factors in Chinese fir plantations across subtropical China. Front Forests Global Change. 2023;5:1086328. 51. Huo H, Feng Q, Su Y. Shrub communities and environmental variables responsible for species distribution patterns in an alpine zone of the Qilian Mountains, northwest China. J Mt Sci. 2015;12:166–76. [ Google Scholar ] 52. Diao E, Cao G, Cao S, Yuan J, Yu M, Chen Z, et al. Path analysis and variation of soil carbon and nitrogen contents under different land use types on the southern slope of the Qilian Mountains. Arid Zone Res. 2021;38:1346–54. [ Google Scholar ] 53. Feng X, Qiu H, Pan J, Tang J. The impact of climate change on livestock production in pastoral areas of China. Sci Total Environ. 2021;770:144838. [ DOI ] [ PubMed ] [ Google Scholar ] 54. Fu YH, Piao S, Zhao H, Jeong S, Wang X, Vitasse Y, et al. Unexpected role of winter precipitation in determining heat requirement for spring vegetation green-up at northern middle and high latitudes. Glob Chang Biol. 2014;20:3743–55. [ DOI ] [ PubMed ] [ Google Scholar ] 55. Jonas T, Rixen C, Sturm M, Stoeckli V. How alpine plant growth is linked to snow cover and climate variability. J Geophys Res Biogeosci. 2008;113:G3. 56. Jiang W, Gong L, Yang L, He S, Liu X. Dynamics in C, N, and P stoichiometry and microbial biomass following soil depth and vegetation types in low mountain and hill region of China. Sci Rep. 2021;11:19631. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 57. Tao F, Huang Y, Hungate BA, Manzoni S, Frey SD, Schmidt MWI, et al. Microbial carbon use efficiency promotes global soil carbon storage. Nature. 2023;618:981–5. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 58. Wang J, Zhong Z, Wang Z, Chen B, Zhang X, Shen Z, et al. Distribution characteristics of soil nitrogen to phosphorus ratio in alpine grassland ecosystems on the Qinghai–Tibet Plateau. Chin J Appl Ecol. 2013;24:3399–406. [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Material 1 (27.4KB, docx) Supplementary Material 2 (22.9KB, docx) Supplementary Material 3 (211KB, docx) Data Availability Statement The data sources are indicated in the text, and all the data are available upon request from the authors. Articles from Carbon Balance and Management are provided here courtesy of BMC ACTIONS View on publisher site PDF (5.5 MB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Impacts of organic matter amendments on carbon and nitrogen dynamics in grassland soils#624391
• Soil microbial community structure and carbon dynamics in response to compost and livestock management in grassland soils#156403
• Modeling the distribution of soil organic carbon in salt marshes dominated by various plant species along Egypt's Delta coast.#295236
• Carbon and nitrogen pools in Padanian soils (Italy): Origin and dynamics of soil organic matter#388266
• Seasonal dynamics of soil CO<sub>2</sub> efflux across land use systems and implications for mitigation under a changing climate#642507