Response of Ecosystem Services to Vegetation Cover Threshold Based on a Multi-coupling Model

Response of Ecosystem Services to Vegetation Cover Threshold Based on a Multi-coupling Model | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 9 July 2025 V1 Latest version Share on Response of Ecosystem Services to Vegetation Cover Threshold Based on a Multi-coupling Model Authors : Mengting Bai , Weiwei Cheng , and Fawen Li 0000-0002-9342-3242 [email protected] Authors Info & Affiliations https://doi.org/10.22541/au.175206448.86915042/v1 276 views 144 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract As a major strategic project for ecological restoration and ecological security in China, the Grain-for-Green Program not only serves as a successful example of global ecological governance, but also holds significant academic value and practical guidance significance for deepening the theory of ecological civilization construction and optimizing the implementation path of ecological projects. Under certain climatic conditions, the extent to which natural vegetation can recover has long been a subject of debate in the field of eco-hydrology. Based on models such as LPJ-GUESS, RUSLE, and InVEST, this paper calculates four ecosystem service indicators (carbon sequestration service, water production service, soil conservation, and habitat quality), and considers factors such as topography, precipitation, temperature, land use, and population comprehensively. The GAM model is used for multi-factor analysis to identify the threshold points of vegetation coverage. The results show that the vegetation coverage in the Ziya River Basin increased significantly from 2001 to 2018. The spatial variations of various indicators under different land use patterns are not the same. The carbon sequestration, habitat quality, and soil conservation index increased with the increase in vegetation coverage, while the water production service decreased with the increase in vegetation coverage. The threshold values of the impact of vegetation coverage on ecosystem services in forest land and grassland were 0.69 and 0.6, respectively. When the vegetation coverage exceeded the above thresholds, the vegetation coverage had an inhibitory effect on the ecosystem. The research results can provide necessary technical support for the ecological restoration of river basins, the effective allocation of resources, and the formulation of management strategies. Response of Ecosystem Services to Vegetation Cover Threshold Based on a Multi-coupling Model Mengting Bai 1 , Weiwei Cheng 1 , Fawen Li 1 State Key Laboratory of Hydraulic Engineering Intelligent Construction and Operation ，Tianjin University，Tianjin 300072，China ) The corresponding author’s contact information: Fawen Li, email: [email protected] Abstract: As a major strategic project for ecological restoration and ecological security in China, the Grain-for-Green Program not only serves as a successful example of global ecological governance, but also holds significant academic value and practical guidance significance for deepening the theory of ecological civilization construction and optimizing the implementation path of ecological projects. Under certain climatic conditions, the extent to which natural vegetation can recover has long been a subject of debate in the field of eco-hydrology. Based on models such as LPJ-GUESS, RUSLE, and InVEST, this paper calculates four ecosystem service indicators (carbon sequestration service, water production service, soil conservation, and habitat quality), and considers factors such as topography, precipitation, temperature, land use, and population comprehensively. The GAM model is used for multi-factor analysis to identify the threshold points of vegetation coverage. The results show that the vegetation coverage in the Ziya River Basin increased significantly from 2001 to 2018. The spatial variations of various indicators under different land use patterns are not the same. The carbon sequestration, habitat quality, and soil conservation index increased with the increase in vegetation coverage, while the water production service decreased with the increase in vegetation coverage. The threshold values of the impact of vegetation coverage on ecosystem services in forest land and grassland were 0.69 and 0.6, respectively. When the vegetation coverage exceeded the above thresholds, the vegetation coverage had an inhibitory effect on the ecosystem. The research results can provide necessary technical support for the ecological restoration of river basins, the effective allocation of resources, and the formulation of management strategies. Key words： ecosystem services, threshold effect, GAM model, vegetation coverage 1 Introduction The ecological environment is the fundamental condition for human survival and the basis for social and economic development. Protecting and improving the ecological environment is a prerequisite for the sustainable supply of ecosystem services, and optimizing the supply of ecosystem services also contributes to the protection and restoration of the ecological environment (Zuo et al.,2022). Vegetation coverage is an important component of the Earth’s surface ecosystem and plays a significant role in maintaining ecological balance, regulating climate, and conserving water and soil (Cui et al.,2009). It serves as both an indicator of the health of the ecological environment and a core approach for ecological restoration. Through monitoring vegetation coverage and implementing ecological engineering projects, a virtuous cycle of ”vegetation restoration - environmental improvement - sustainable development” can be promoted. The vegetation coverage threshold refers to the point at which the structure and function of an ecosystem undergo significant changes when the vegetation coverage is lower or higher than a specific value (Tang et al.,2015), such as the degradation of a forest ecosystem into a grassland ecosystem or the transformation of a desert ecosystem into a grassland ecosystem. This concept is of great significance in ecological research, as it helps to understand the stability and resilience of ecosystems and provides a scientific basis for ecosystem management and environmental protection. As a turning point from quantitative to qualitative change, the threshold has received considerable attention from many scholars in recent years. Although there are numerous studies on vegetation coverage thresholds, the construction of indicator systems is relatively simple. For instance, Li (Li et al.,2024) in the Mongolian Plateau explained the drought response thresholds and distribution patterns of different vegetation types. Chen (Chen et al.,2024) proposed a sustainable vegetation coverage threshold of 65% for the Loess Plateau based on an analysis of the impact of past, present, and future vegetation changes on water. Xie (Xie et al., 2024) used the Geodetector to analyze the spatiotemporal impact of vegetation coverage on water conservation in the Loess Plateau. Zhao (Zhao et al.,2024) analyzed the spatiotemporal characteristics of vegetation coverage in the Wei River Basin and explored the correlation between vegetation coverage and sediment transport modulus at the basin scale. Su (Su et al., 2022) used the LPJ-GUESS model to analyze the impact and changing trend of vegetation coverage on soil conservation services. In addition, most current studies on the impact of vegetation coverage on ecosystem services use parameter estimation methods (Chen et al., 2023,Wang et al.,2024, Zhang et al.,2020), which assume a functional relationship between the two, such as quadratic or cubic terms. However, if the relationship is more complex, it is difficult to accurately capture the threshold point. Using ecological models to simulate the growth process of vegetation is a more effective analytical method for studying vegetation coverage thresholds. The LPJ-GUESS model, developed by Ben Smith and others at the University of Delft, is a process-based dynamic global vegetation model designed to study the dynamic changes of vegetation and ecosystems at regional or global scales. The model can simulate the interaction between vegetation and the atmosphere and soil. The study of vegetation coverage thresholds can provide scientific basis for ecosystem protection and restoration, climate change adaptation strategies, and natural disaster prevention and early warning. The response of the ecological environment to vegetation coverage thresholds varies in different regions, indicating that scientific and precise management should be implemented in ecological management (Zhang et al.,2020) . Since the implementation of the Grain-for-Green Program in 1999, China’s vegetation ecosystem has undergone significant positive evolution. The average forest coverage rate has increased by more than 4 percentage points, and vegetation restoration has enhanced the stability of soil aggregate structure, leading to an overall improvement in ecosystem service functions. In terms of ecosystem regulation, the sediment load in the Yangtze River Basin has significantly decreased, and the area of desertification has shrunk by an average of 3.45% per year, effectively curbing the expansion of soil erosion and land desertification and promoting the transformation of terrestrial ecosystems towards a stable state. At the same time, vegetation restoration has promoted the realization of ecological product value, driven the development of characteristic industries, and established a coordinated mechanism for ecological protection and improvement of people’s livelihood, providing a systematic solution for global ecological restoration. However, increased vegetation coverage disrupts the original dynamic relationship of the water cycle by elevating vegetation water demand, evapotranspiration, and soil water consumption, thereby further inhibiting vegetation growth. (Wen, 2020) . This will inevitably increase the risk of drought and other hazards. Therefore, vegetation coverage has a certain threshold effect on ecosystem services. Different land uses or vegetation types have obvious differences in their impacts (Guillen-Cruz et al.,2021). The threshold effect of vegetation coverage is a key indicator of the resilience and vulnerability of ecosystems. Its determination requires the integration of data-driven and mechanism models, and the interaction between human activities and climate change should be emphasized. This study is based on the cutting-edge theoretical framework of ecosystem service assessment. By integrating multi-source data and process model simulation methods, it systematically optimizes the construction paradigm of the existing indicator system. Firstly, the dynamic global vegetation model LPJ-GUESS is used to simulate ecological processes. Then, a nonlinear response analysis framework based on the generalized additive model (GAM) is constructed to focus on exploring the complex coupling mechanism between vegetation coverage and the comprehensive index of ecosystem. 2 Study area and data sources 2.1 Overview of the study area The Ziya River Basin belongs to the southern branch of the Haihe River Basin. It has two major tributaries, the Hutuo River (north) and the Fuyang River (south), which merge at Zangjiaqiao in Xian County and are then called the Ziya River. The basin starts from the Taihang Mountains in the west and borders the Bohai Sea in the east and the Zhangwei River in the south. It spans over Shanxi, Hebei provinces and Tianjin, located between 36°03’ N to 39°35’ N, and 112°20’ E to 117°50’ E. The basin covers an area of approximately 46,800 square kilometers. The overall topography of the Ziya River Basin is west high and east low. The northwest part of the basin is mountainous terrain, located within Shanxi Province, while the eastern and central parts are plains, located within Hebei Province. The landform is characterized by alternating mountains, low hills and plateaus. The mountainous area accounts for about 67% of the total basin area, while the plain area accounts for approximately 33%. The vegetation coverage in the Ziya River Basin shows spatial differences between the north and the south. The vegetation coverage in the south is relatively less, while it is more abundant in the north. This distribution feature is related to the gradient change of water and heat conditions. The south is warmer and has less precipitation, while the north is relatively cooler and has more precipitation. Topographically, the vegetation types and richness in the mountainous river sections are significantly higher than those in the plain areas. There are 199 plant species in the mountainous river sections, belonging to 43 families and 135 genera, with dominant species including reed (Phragmites australis), Mongolian thistle (Artemisia mongolica), and cypress (Vitex negundo), etc. In contrast, the riparian zone of the plain river sections has only 36 plant species, belonging to 22 families and 35 genera, with the dominant species being reed and reverse-angled amaranth (Amaranthus retroflexus). From 2001 to 2018, the land use types in the Ziya River Basin mainly included forest land, grassland, construction land, water bodies and unused land. Among them, cultivated land occupied a dominant position in the land use structure, with its area accounting for more than 45% in both 2001 and 2018, reflecting the important position of agriculture in the economic activities of this basin. The transfer of land use types mainly occurred between cultivated land and construction land. From 2001 to 2018, the transfer area between cultivated land and construction land accounted for 23.7% of the total area, and this proportion even reached 60% between 2000 and 2010. This change trend is closely related to the rapid urbanization process within the basin. With the advancement of urbanization, a large amount of cultivated land has been converted into construction land to meet the needs of urban expansion, infrastructure construction and population growth. Figure 1 shows the land use map of Ziya River Basin in 2018. Figure 1 Location and Land Use of the Ziya River Basin 2.2 Data Sources The vegetation data mainly consist of the Normalized Difference Vegetation Index (NDVI), which is based on the MODIS-NDVI product. Other data such as precipitation and temperature are sourced from the National Qinghai-Tibet Plateau Data Center, with a resolution of 1km. The DEM is from the Geospatial Data Cloud, with a resolution of 90 meters. The land use is from the Wuhan University CLCD land use database, with a resolution of 30m; the population is from the worldpop population raster dataset, with a resolution of 1km. To facilitate calculation, all the data are projected into WGS_1984_Web_Mercator_Auxiliary_Sphere, and the resolution is uniformly resampled to 1km, with a row and column count of 438, 447. 3 Research Methods 3.1 LPJ-GUESS Model Buildings The LPJ-GUESS (Lund-Potsdam-Jena General Ecosystem Simulator) model is a dynamic vegetation model developed by the Department of Natural Geography and Ecological Sciences at Lund University. It evolved from the LPJ-DGVM (Dynamic Global Vegetation Models) model and can simulate the physical and physiological processes of the land surface at different scales, including vegetation growth, phenological changes, carbon and water cycles, and many other ecosystem processes (Sitch et al.,2003). The LPJ-GUESS model is based on the principles of plant physiology and ecology and can distinguish different vegetation functional types, thus enabling more accurate simulation of plant physiological and ecological processes. Moreover, the model has high flexibility in spatial resolution and can be simulated at global, regional, and local scales. The necessary input data for the model include climate data (temperature, precipitation, radiation, etc.), soil data (soil type, texture, nutrient content, etc.), and land use data. The output data of the model include leaf area index (LAI), annual net primary productivity of the ecosystem (NPP), and evapotranspiration (ET), etc. Photosynthesis by plants absorbs carbon dioxide and converts it into organic carbon. The absorption amount varies depending on the physiological characteristics of the plant types. Autotrophic respiration of plants and heterotrophic respiration of soil microorganisms decompose the organic carbon and release it back into the atmosphere in the form of carbon dioxide, with the rate being affected by factors such as plant biomass, temperature, and humidity. The decomposition of plant litter, part of the carbon becomes carbon dioxide, and part becomes soil organic carbon, which is influenced by the amount of litter, decomposition rate, and environmental conditions. The carbon absorbed by the vegetation is distributed to various organs for growth and other purposes, and the distribution method is related to the plant type, which affects the carbon cycle. The core formula of the LPJ-GUESS model is as follows: In the formula, R m represents the amount of carbon consumed by autotrophic respiration. C tissue is the carbon content of the tissue. C∶N represents the carbon-nitrogen ratio of the plant tissue. r is a constant, representing the respiratory coefficient. g(T) is the modified Arrhenius equation. In the formula, T represents temperature (in degrees Celsius), referring to either air temperature or soil temperature. In the formula, GPP represents the total amount of carbon fixed by vegetation through photosynthesis, measured in gC/m 2 . NPP is the net primary productivity, which is the difference between photosynthesis and respiration, also measured in gC/m 2 . The vegetation water consumption module in the model calculates by integrating the plant and soil water dynamics. Evapotranspiration is the sum of evaporation (E) and transpiration (T), and in the LPJ-GUESS model, it is usually estimated using the Penman-Monteith formula. This formula can comprehensively consider the influence of meteorological conditions, plant physiological characteristics, and environmental factors on evapotranspiration. Specifically, the formula involves variables such as net radiation, air temperature, wind speed, and saturation and actual vapor pressure difference. In the formula: ET 0 represents the water requirement of the reference crop, in millimeters per day. R n represents the net radiation at the crop surface, in MJ/m 2 ·d. G represents the soil heat flux, in MJ/m 2 ·d. T represents the average temperature, in degrees Celsius. u 2 represents the average wind speed at a height of 2 meters, in meters per second. e s represents the saturated vapor pressure, in kPa. e a represents the actual vapor pressure, in kPa. Δ represents the slope of the curve of saturated vapor pressure versus temperature, in kPa/°C. γ represents the dry-wet constant, in kPa/°C. 3.2 Model calibration and evaluation indicators This study used the carbon fixation and evapotranspiration indicators simulated by the LPJ-GUESS model for calibration and validation. The data were the average values of 18 years from 2001 to 2018. 600 pixel points on the Zaya River Basin were randomly selected. The NPP indicators simulated by the model were verified against the GLASS-MODIS data, and the ET indicators were verified against the PML-V2 data. This study used the root-mean-square error ( RMSE ) and coefficient of determination ( R 2 ) for calibration and validation. In the formula: M i represents the simulated value, which is the average of the simulated values. O i represents the measured value, which is the average of the measured values. RMSE and R 2 can directly reflect the proportion of the prediction error relative to the range of data variation, and are one of the commonly used indicators for evaluating the performance of regression models. The smaller the RMSE , the better the prediction ability of the model usually is. The closer R 2 is to 1, the better the fitting effect of the model is. 3.3 Method for calculating ecological service indicators (1)Fractional Vegetation Cover The fractional vegetation cover(FVC) is obtained through pixel binning, which is calculated based on the NDVI values of bare soil and high vegetation coverage: In the formula: NDVI represents the normalized vegetation index. NDVI soil and NDVI vge are the NDVI values corresponding to bare soil and pure vegetation respectively. (2) Ecosystem carbon sequestration service The carbon sequestration services of the ecosystem are characterized by the NPP data simulated by the LPJ-GUESS model. (3)Ecosystem water supply service In the formula: W represents water supply, mm. PRE represents precipitation, mm. ET represents evapotranspiration, mm, which is characterized by the ET data simulated by the LPJ-GUESS model. (4)Ecosystem soil conservation service Soil conservation services refer to a series of activities and management measures that aim to reduce soil erosion, improve soil quality, maintain soil productivity and ecological balance. The ecosystem soil conservation services are calculated using the RUSLE model: In the formula: SC represents the annual average soil retention amount, measured in tons per square kilometer. R is the rainfall erosion force factor. K is the soil erodibility factor. LS is the length slope factor. C is the coverage and management factor. P is the soil and water conservation measure factor. E represents the sediment interception amount. The values of C and P were determined based on previous studies and adjusted according to the current situation of the Ziya River Basin. (5)Ecosystem habitat quality The habitat quality service refers to various functions and services provided by ecosystems that can maintain and improve the quality of biological habitats. Habitat quality services are of vital importance for the protection of biodiversity, the health of ecosystems, and human well-being. Habitat quality is established through the InVEST habitat quality module, which links ecosystem types with threat sources and sets sensitivity parameters: In the formula: HQ xi represents the habitat quality of pixel x in land use type i. H i represents the habitat suitability. Z is the scale constant, usually 2.5. D xi is the weighted average of threat sources of pixel x in land use type i. k is the half-saturation constant, which is selected as 0.5 in this case. (6)Integrated Index of Ecosystem Services To explore the comprehensive influence of vegetation coverage on the above four indicators, it is necessary to normalize these four indicators: In the formula: ST b represents the normalized result of ecosystem services. ST is the initial value. ST min is the minimum value of ecosystem services. ST max is the maximum value of ecosystem services. TES is the comprehensive index of services. ST bi is the normalized result of the i-th type of ecosystem services, i = 4. is the weight, and the weights of the four indicators in this case are all 0.25. 3.4 Ecological Threshold Identification Method The Generalized Additive Model (GAM) is a semi-parametric regression model that combines the characteristics of additive models and non-parametric regression. This model is particularly suitable for dealing with situations where there is a complex and shapeless nonlinear relationship between the independent variables and the dependent variable. One significant advantage of the GAM model is that it does not require prior assumptions about the distribution form of the data, and it can reveal and explain the basic structure of the model, thereby providing valuable insights for data analysis. Therefore, GAM has important application value in handling complex data relationships and improving prediction accuracy. The model assumes that the response variable is the sum of nonlinear functions of multiple predictor variables, and its mathematical form is: In the formula, g() is the link function, represents the parameter terms, and f() is the non-parametric smoothing term. The GAM model does not specify the form of the explanatory variables, fully demonstrating its flexibility. The calculation of the GAM model is implemented through the mgcv software package in R language. The effective degrees of freedom (Edf) in the GAM model is a core indicator for measuring the complexity of the smoothing terms in the model. The higher the Edf, the more complex the relationship between the explanatory variables and the dependent variable. If Edf = 1, it indicates a linear relationship between the explanatory variables and the dependent variable, if Edf = 3, it is similar to a cubic polynomial, indicating a moderately complex curve, if Edf > 3, it indicates a more complex nonlinear relationship. 4 Results and analysis 4.1 Model calibration and validation results The LPJ-GUESS model is an ecosystem simulation tool based on data-driven plant functional type (PFT) information. During the model operation, PFT data serves as the key calibration parameters, providing detailed information about the biological physical and biochemical characteristics of different plant types, such as maximum light energy utilization, optimal nitrogen concentration in leaves, root distribution characteristics, etc. These parameters play a crucial role in the model debugging process and directly affect the accuracy and reliability of the simulation results. The core parameters obtained after model debugging (Table 1) have been optimized and verified, and can better reflect the characteristics of the actual ecosystem. Table 1 Core Parameters of the LPJ-GUESS Calibration Model PFT Tcmin-serv（℃） Tcmin-est（℃） Tcmax-est（℃） Twmin-est（℃） Gdd5min-est（℃） Bine -31 -30 -1 5 500 Bns -1000 -1000 -2 -1000 350 Tene -7 -7 10 5 1150 Tebs -11 -10 2 19 1100 Tebsh -14 -13 2 22 1100 The figure 2 shows the verification results of ET and NPP. Figure 2 Calibration and validation of the LPJ-GUESS model (The left picture is ET, and the right picture is NPP.) After calculation, the R 2 in the ET validation is 0.61, and the RMSE is 27.71 mm; the R 2 in the NPP validation is 0.66, and the RMSE is 0.02 kgC/m 2 . The model simulation accuracy is relatively high. The constructed LPJ-GUESS model has good applicability in the Ziya River Basin. 4.2 Identification of the vegetation coverage threshold 4.2.1 Change in the index space As shown in the figure 3, the five indicators in the Ziya River Basin exhibit significant spatial heterogeneity. Figure 3 Spatial average changes of various indicators in the Ziya River Basin from 2001 to 2018 The average vegetation coverage is 0.61, with a relatively scattered spatial distribution. The average carbon fixation is 482.2 gC/m 2 , higher in cultivated areas and lower in mountainous areas. The average habitat quality is 0.46, lower in plain areas and higher in mountainous areas. The average soil retention is 6636.01 t/km 2 , higher in cultivated areas and lower in mountainous areas. The average evaporation is 434.96 mm, and the average rainfall is 485.2 mm, showing a pattern of lower in plain areas and higher in mountainous areas. 4.2.2 The Adaptation of Vegetation Coverage to GAM Model The vegetation coverage shows complex nonlinear correlation characteristics with the ecosystem service indicators. Through model calculations, the fitting degrees of different land uses such as forest land and grassland in the GAM model are 61.9 % and 74.2 % respectively. The fitting degrees are relatively high, indicating that the GAM model has a strong explanatory ability for the complex relationships between FVC and various ecosystem service indicators. The fitting degrees of each indicator with the GAM model are shown in Table 2. Table 2 Analysis results of the GAM model Indicators Degree of fitting( R 2 ) Forest land Grass land W 81.5 82.6 NPP 46.9 64.6 HQ 47.7 47.2 SC 17.5 19.1 TES 61.9 74.2 4.2.3 The relationship between vegetation coverage and ecosystem service index. Due to spatial heterogeneity, different land use types have different vegetation coverage thresholds. Based on the fitted GAM model, the relationship between vegetation coverage and the comprehensive ecosystem service index is plotted to determine the threshold points. The table 3 shows the smoothed term statistical parameters of each index for different land uses. Table 3 Statistical Parameters of Smoothing Terms for Different Land Uses Indicators Forest Land Grassland Edf p value Edf p value W 8.551 0.000 8.692 0.000 NPP 8.899 0.000 8.867 0.000 HQ 7.888 0.000 8.224 0.000 SC 7.764 0.000 7.030 0.000 TES 8.729 0.000 8.822 0.000 From the above table, it can be seen that the Edf values for both forest land and grassland are at a relatively high level (forest land Edf = 8.729, grassland Edf = 8.822). This further indicates that the relationship between vegetation coverage and water yield services, carbon sequestration, habitat quality, soil retention, and the comprehensive index is rather complex and cannot be fitted using simple linear or quadratic, cubic functions. The p-values for both forest land and grassland are less than 0.001, supporting the validity of the model and indicating that the relationship has a strong statistical significance. 4.2. 4 Identification of the vegetation coverage threshold Based on the above results and analysis of the GAM model, the following conclusions can be drawn: The effect of forest land FVC on ecosystem services The effect of forest land FVC on various indicators of the ecosystem is shown in the following figure 4. The effect of FVC on W The effect of FVC on NPP The effect of FVC on HQ The effect of FVC on SC The effect of FVC on TES Figure 4 The effect of forest land FVC on ecosystem services As can be seen from the above graph, in the forest land, the water production service gradually decreases as the FVC increases, while carbon fixation first increases and then decreases, habitat quality gradually increases, soil retention first decreases and then increases, and the ecological comprehensive index first increases and then decreases. The forest reaches its peak when the FVC is equal to 0.69. Excessive vegetation density can lead to enhanced transpiration, reducing the moisture content in the soil. Additionally, dense vegetation may increase surface runoff, reducing groundwater recharge, thereby reducing the water production service capacity. At moderate FVC, photosynthesis of the vegetation reaches its peak, with the maximum carbon absorption. However, when the FVC is too high, due to insufficient light and poor ventilation, the efficiency of photosynthesis decreases, resulting in reduced carbon fixation. As the FVC increases, the vegetation coverage becomes more dense, providing more habitats and food sources for more organisms, enhancing the biodiversity of the ecosystem, and improving the overall habitat quality. At low FVC, the vegetation is insufficient to prevent soil erosion, and the soil retention capacity is poor. As the FVC increases, the root systems of the vegetation strengthen, and the soil retention capacity improves. However, when the FVC is too high, the overly dense root systems may lead to changes in soil structure, affecting the soil retention capacity, and thus the soil retention capacity increases again at a certain point. The differences in the response curves of each service function to the FVC reflect the non-linear characteristics of the ecosystem. The effect of grassland FVC on ecosystem services The effect of grassland FVC on various indicators of the ecosystem is shown in the following figure 5. The effect of FVC on W The effect of FVC on NPP The effect of FVC on HQ The effect of FVC on SC The effect of FVC on TES Figure 5 The effect of grassland FVC on ecosystem services From the above graph, it can be seen that in the grassland area, the water supply service gradually decreases as the FVC increases, while carbon fixation first increases and then decreases, habitat quality gradually increases, soil retention gradually increases, the ecological comprehensive index first increases and then decreases, the ecosystem service comprehensive index first increases and then decreases. The ecosystem service comprehensive index reaches its peak when FVC is 0.6. When FVC is low, the root systems of herbaceous plants and the surface litter layer can effectively intercept precipitation and promote water infiltration. However, as FVC exceeds 0.6, dense vegetation can induce excessive soil water depletion through heightened transpiration, while the dense canopy of herbaceous plants may impede surface runoff efficiency in recharging groundwater. At lower FVC levels (< 0.6), vegetation photosynthetic efficiency increases with coverage. Beyond this threshold, however, excessive density in herbaceous plants triggers intensified competition for light and nutrients, elevates pest and disease risks, and ultimately reduces carbon fixation. As FVC increases, it promotes the vertical structure and species diversity of grassland vegetation, providing habitats and food resources for insects and small mammals. Soil retention increases exponentially with the increase in vegetation coverage. The vegetation canopy can intercept rainwater, reducing the direct impact of rainwater on the soil. The higher the vegetation coverage, the less rain erosion of the surface occurs, and in addition, in the grassland area, there may be residual litter on the surface, and the vegetation coverage makes the litter layer thicker, with a stronger soil retention capacity (Xiong et al., 2019). The initial coordinated improvement of carbon fixation and habitat quality dominates the increase in TES. However, beyond the threshold, the decrease in water supply service capacity offsets the increase in habitat quality, resulting in a decrease in the comprehensive index. This phenomenon is closely related to the resource allocation contradiction of the grassland ecosystem. 4 Discussion This study employed a multi-model coupling analysis method to construct an ecosystem service comprehensive evaluation system. Based on the LPJ-GUESS dynamic vegetation model, the regional carbon sequestration function (including biomass carbon pool and soil organic carbon storage) and hydrological regulation services (spatial patterns of evapotranspiration) were simulated; the RUSLE equation and the InVEST module were used to quantify soil retention amount and habitat quality index. On this basis, the generalized additive model (GAM) was used to analyze the nonlinear response relationship between vegetation coverage and each individual service and the comprehensive index, and the critical threshold points that significantly affect ecosystem services were identified. This multi-scale assessment framework achieved a systematic study from ecosystem process simulation to service function quantification and to the analysis of driving mechanisms. The study shows that ecosystem management needs to comprehensively consider multiple factors, find a balance point, and pursue high FVC, but it does not necessarily lead to the best ecological benefits. Different from that, it is necessary to reasonably regulate FVC according to the characteristics and goals of different regions to achieve sustainable development of the ecosystem. The threshold effect of vegetation coverage is mainly applied in the regulation of actual production processes. For example, in forest areas, maintaining FVC = 0.69 through thinning or planting mixed forests can balance carbon fixation and water service demands. In the peak area of grassland TES (FVC = 0.6), implementing rotational grazing or seasonal grazing can balance production and ecological functions. The limitations of this study are as follows: ① In practice, each indicator is also affected by factors such as aspect, soil texture, and resource management methods. Although this study selected DEM, precipitation, and temperature as explanatory variables, it did not fully cover all influencing factors. ② The current research mainly focuses on terrestrial natural ecosystems and lacks exploration of the threshold values for aquatic vegetation and urban green space coverage. 5 Conclusion This paper analyzes the temporal and spatial changes of vegetation coverage, precipitation, evapotranspiration, terrain, carbon fixation, habitat quality, and soil retention in the Ziya River Basin from 2001 to 2018. It explains the nonlinear relationship between vegetation coverage and ecosystem services and quantitatively identifies the threshold of vegetation coverage. The main conclusions are as follows: (1) The LPJ-GUESS model demonstrates excellent performance in simulating the vegetation ecosystem in the Ziya River Basin, being able to accurately simulate the carbon fixation and evapotranspiration quantities generated by the vegetation ecosystem process in this basin. In the verification of evapotranspiration simulation results, the R²reaches 0.61, and the RMSE is 27.71 mm; in the verification of carbon fixation simulation results, R 2 is 0.66, and RMSE is 0.01 kgC/m². (2) From 2001 to 2018, vegetation coverage, carbon fixation, evapotranspiration, soil retention, and precipitation all showed a continuous upward trend, but habitat quality showed a continuous decline. Spatially, the distribution of vegetation coverage was relatively scattered, while carbon fixation and soil retention were higher in the plains and lower in the mountains, and habitat quality, evapotranspiration, and precipitation showed a pattern of lower in the plains and higher in the mountains. (3) The nonlinear relationship between vegetation coverage and the comprehensive indicators of ecosystem services reveals its significant threshold effect. Through the calculation of the GAM model, the threshold of vegetation coverage is 0.69 for forest land and 0.6 for grassland. Once this threshold is exceeded, the system may shift from a ”resource-efficient” state to an ”overcompetition” state, and its anti-interference ability will decline. 6 Acknowledgements The authors would like to acknowledge the financial support for this work provided by the National Key Research and Development Program Project of China (Grant no. 2023YFC3006601）. Compliance with Ethical Standards Conflict of Interest The authors declare that they have no conflicts of interest. Data Availability The datasets generated during the current study are available from the corresponding author on reasonable request. References 1. Zuo L.Y., Jiang Y., Gao J.B. Quantitative Separation of Multiple Driving Forces for Ecosystem Services in Ecological Protection Redline Areas. Acta Geographica Sinica. 2022;77(9):2174-88. 2. Cui L.L., Shi J., Yang Y.M., et al. Seasonal Responses of NDVI of Vegetation in Eastern China to Temperature and Precipitation [J]. Acta Geographica Sinica. 2009，64（7）：850-860. 3. Tang H.P., Chen J., Xue H.L. Ecological Threshold: Concepts, Methods and Research Prospects. Journal of Plant Ecology. 2015;39(9):932. 4. Li H, Hu Y, Ao Z. Identification of critical drought thresholds affecting vegetation on the Mongolian Plateau[J]. Ecological Indicators, 2024, 166112507-112507. 5. Chen, Y. P., Wang, K. B., Fu, B. J.et al. 65% cover is the sustainable vegetation threshold on the Loess Plateau. Environmental Science and Ecotechnology, 2024, 100442. 6. Xie Z., Li P., Liu W.Q., Jiang Y.N. The Impact of Climate Change and Vegetation Restoration on the Spatiotemporal Changes of Water Conservation Function in the Loess Plateau. Acta Ecologica Sinica, 2024，44（23）：10915-10935 7. Zhao Y.B., Zhao G.J., Mu X.M., et al. Threshold of Vegetation Coverage for Regulating Sediment Transport in Wei River Basin [J]. Research on Soil and Water Conservation, 2024,31(6):96-102,108. 8. Su Y.L., Chen X.T., Liu S.Y., et al. Spatial-temporal Evolution of Soil Conservation Function of Vegetation Changes in the Loess Plateau from 2001 to 2100 [J]. Journal of Soil and Water Conservation, 2022,36(06):55-62+81. 9. Chen T.T., Wang Y.X., Zeng X.L., et al. Characteristics of ecosystem service relationships in the southwest region and their constraints with vegetation coverage. Acta Ecologica Sinica, 2023，43( 6) : 2253-2270． [10] Wang X.F., Bai J., Feng X.M., et al. Response Threshold of Ecosystem Services in Qinba Mountain Area to Changes in Vegetation Coverage. Acta Ecologica Sinica2024，44（15）：6811-6827． [11] Zhang K, Lu Y.H., Fu B.J., et al. Impact of Vegetation Coverage Changes in the Loess Plateau on Ecosystem Services and Its Thresholds [J]. Acta Geographica Sinica, 2020,75(05):949-960. [12] Wen X, Theau J. Spatiotemporal anyalysis of water-related ecosystem services under ecological restoration scenarios: a case study in northern Shannxi, China. Science of the Total Environment, 2020, 720: 137477. [13] Guillen-Cruz, G., et al. ”Influence of vegetation type on the ecosystem services provided by urban green areas in an arid zone of northern Mexico.” Urban Forestry & Urban Greening 62 (2021): 127135. [14] Sitch S，Smith B，Prentice IC，et al．Evaluation of ecosystem dynamics，plant geography and terrestrial carbon cycling in the LPJ dynamic global vegetation model. Global Change Biology，2003，9: 161－185. [15] Xiong Y.W., Yu B., Bai M.T., Zhang X.Y., Huang G.H, & Furman, A. (2019). Soil properties and plant growth response to litter in a prolonged enclosed grassland of Loess Plateau, China. Journal of Earth Science, 2019,30, 1041-1048. Information & Authors Information Version history V1 Version 1 09 July 2025 Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords ：ecosystem services gam model threshold effect vegetation coverage Authors Affiliations Mengting Bai Tianjin University View all articles by this author Weiwei Cheng Tianjin University View all articles by this author Fawen Li 0000-0002-9342-3242 [email protected] Tianjin University View all articles by this author Metrics & Citations Metrics Article Usage 276 views 144 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Mengting Bai, Weiwei Cheng, Fawen Li. Response of Ecosystem Services to Vegetation Cover Threshold Based on a Multi-coupling Model. Authorea . 09 July 2025. DOI: https://doi.org/10.22541/au.175206448.86915042/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.175206448.86915042/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9ffd08a6afb6593a',t:'MTc3OTQ2NTYyNg=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();