Analysis of the Trade-Off/Synergy Effect and Driving Factors of Ecosystem Services in Hulunbuir City, China

Abstract

An in-depth understanding of the spatiotemporal heterogeneity of ecosystem service (ES) trade-offs and synergies, along with their driving factors, is crucial for formulating key ecological restoration strategies and effectively allocating ecological environmental resources in the Hulunbuir region. This study employed an integrated analytical approach combining the InVEST model, ArcGIS geospatial processing, R software environment, and Optimal Parameter Geographical Detector (OPGD). The spatiotemporal patterns and driving factors of the interaction of four major ES functions in Hulunbuir area from 2000 to 2020 were studied. The research findings are as follows: (1) carbon storage (CS) and soil conservation (SC) services in the Hulunbuir region mainly show a distribution pattern of high values in the central and northeast areas, with low values in the west and southeast. Water yield (WY) exhibits a distribution pattern characterized by high values in the central–western transition zone and southeast and low values in the west. For forage supply (FS), the overall pattern is higher in the west and lower in the east. (2) The trade-off relationships between CS and WY, CS and SC, and SC and WY are primarily concentrated in the western part of Hulunbuir, while the synergistic relationships are mainly observed in the central and eastern regions. In contrast, the trade-off relationships between CS and FS, as well as FS and WY, are predominantly located in the central and eastern parts of Hulunbuir, with the intensity of these trade-offs steadily increasing. The trade-off relationship between SC and FS is almost widespread throughout HulunBuir. (3) Fractional vegetation cover, mean annual precipitation, and land use type were the primary drivers affecting ESs. Among these factors, fractional vegetation cover demonstrates the highest explanatory power, with a q-value between 0.6 and 0.9. The slope and population density exhibit relatively weak explanatory power, with q-values ranging from 0.001 to 0.2. (4) The interactions between factors have a greater impact on the inter-relationships of ESs in the Hulunbuir region than individual factors alone. The research findings have facilitated the optimization and sustainable development of regional ES, providing a foundation for ecological conservation and restoration in Hulunbuir.

1. Introduction

The Hulunbuir region, as a key national ecological function area and an important ecological security barrier in northern China, is a crucial part of the “Three-North” shelterbelt [1]. Its grassland ecosystem’s water and carbon regulation functions are critical to both the country’s dual-carbon goals and ecological security. The “Hulunbuir City National Ecological Civilization Construction Demonstration City Plan (2019–2025)” integrates ecological civilization construction into the “green development” evaluation system for relevant districts, cities, and municipal departments. However, in recent decades, rapid economic development, unsustainable human activities (such as overgrazing and land reclamation), and adverse natural factors like climate change have led to the weakening of the synergy between water and carbon in the Hulunbuir region. More than 66% of grasslands have degraded [2], with over 90% of these areas experiencing moderate to severe degradation [3], accompanied by varying degrees of desertification. This has caused severe damage to the structure and function of the ecosystem, significantly reducing its stability and resilience. In this context, thoroughly investigating inter-relationships among ESs along with their driving mechanisms will provide crucial scientific foundations for formulating regional ecological coordination strategies [4].

Ecosystem services (ESs) are the environmental conditions and important benefits that natural ecosystems provide for human survival and development by virtue of their unique structure and functions [5,6]. The WY and CS serve as fundamental regulating services. They directly support regional water resource security and climate regulation through hydrological cycles and biogeochemical processes. Investigating ESs holds substantial practical importance, both for enhancing human welfare and facilitating sustainable development [7,8]. Scholars globally have extensively investigated the dynamics of trade-offs and synergies among ESs. Their research covers conceptual frameworks, driving factors, spatiotemporal dynamics, methodological approaches, etc. [9,10,11,12,13,14,15]. Huang et al. [16] conducted an analysis of the driving factors behind the trade-offs/synergies of ESs, with the results indicating that socioeconomic factors and land use types have a strong explanatory power. Du et al. [17] utilized integrated spatial overlay and correlation methods to analyze water-related ESs in the Yangtze River Basin from 2001 to 2021. Their study focused on three key services: WY, water purification, and SC. The results revealed substantial spatial variations in trade-off synergy relationships. Zhao et al. [18] investigated spatiotemporal variations inter-relationships among three ESs—CS, water purification, and habitat quality—across China’s Yangtze River Delta. Their results demonstrated spatially heterogeneous patterns in both the trade-off/synergy relationships and their magnitudes. Yang et al. [19] employed sub-watershed-scale correlation analysis to examine trade-off/synergy dynamics among three ESs in the Pearl River Delta, China. Their results revealed distinct variations in these interactions across urbanized regions. Although significant progress has been made in ES trade-off synergy research, several limitations persist. Thematically, current studies have a limited capacity to determine whether ES responses to drivers are linear or nonlinear, and they also lack sufficient exploration of the complex linkages between service relationships and natural, socioeconomic, and climatic drivers. Methodologically, predominant reliance on Geographical Detector analysis faces inherent constraints, particularly regarding suboptimal discretization of continuous variables [20].

This study combines literature analysis and model analysis, focusing on four ESs closely related to the socioeconomic and ecological development of Hulunbuir City: CS, WY, SC, and FS. It systematically explores the trade-offs and synergies between these services from 2000 to 2020 and identifies their key driving factors. The study aims to deepen the understanding of the complex interactions between different ESs, providing a solid theoretical foundation and practical guidance for future ecological restoration in the Hulunbuir region, thereby promoting sustainable environmental development.

2. Materials and Methods

2.1. Overview of the Study Area

The Hulunbuir region is situated in the northeastern region of Inner Mongolia Autonomous Region, China, spanning geographical coordinates from 115°31′ E to 126°04′ E and 47°05′ N to 53°20′ N [21] (Figure 1). With a total area of approximately 253,000 km², it represents the largest prefecture-level administrative division in China. This region serves as a crucial ecological barrier, playing a significant role in maintaining regional ecosystem stability. The study area exhibits a temperate continental monsoon climate, characterized by distinct spatial heterogeneity. Mean annual temperatures range from −5 °C to 5 °C, while annual precipitation varies between 250 mm and 450 mm [22,23,24]. Both temperature and precipitation demonstrate a clear decreasing trend from the southeast to northwest regions. The vegetation is dominated by typical steppe, forming a complete ecological gradient from east to west. This gradient sequentially includes forest steppe, meadow steppe, typical steppe, and desert steppe. The soil types are primarily composed of chernozem, chestnut soil, aeolian sandy soil, and bog soil [25,26]. Among these, chernozem and chestnut soil constitute the dominant zonal soil types in the grassland ecosystem.

2.2. Data Source and Processing

The study utilized 30 m resolution land use data from three years: 2000, 2010, and 2020. All datasets were obtained from the product developed by Professor Huang Xin’s research team at Wuhan University https://zenodo.org/record/4417810 (accessed on 5 March 2024) The land use types were extracted and processed using ArcGIS 10.8, and classified into six categories: arable land, forest land, grassland, water, built-up land, and unused land (Figure 2). The area and proportion of each land use type were subsequently calculated. Meteorological data consisted of mean annual precipitation, potential evapotranspiration, and mean annual temperature (

Table 1. Land use type area and proportion in Hulunbuir region from 2000 to 2020.

Type	2000		2010		2020
	Area/ha	Proportion/%	Area/ha	Proportion/%	Area/ha	Proportion/%
Cultivated land	2,304,756	9.14	2,340,260	9.28	2,351,670	9.33
Forest land	14,562,486	57.78	14,860,186	58.96	14,978,113	59.43
Grassland	7,940,193	31.50	7,605,217	30.17	7,446,119	29.54
Water	306,415	1.22	257,002	1.02	274,009	1.09
Built-up land	37,553	0.15	51,481	0.20	42,106	0.17
Unused land	53,373	0.21	90,630	0.36	112,759	0.45

). These datasets were obtained from the National Earth System Science Data Center http://www.geodata.cn/data/ (accessed on 8 March 2024). Topographic data were obtained from the Geospatial Data Cloud https://www.gscloud.cn/ (accessed on 10 March 2024). The dataset included digital elevation model (DEM) data. Slope data were calculated from the DEM. Soil-related parameters, including plant-available water capacity and root-restricting layer depth, were extracted from the World Soil Database https://www.fao.org/ (accessed on 1 April 2024). The socioeconomic dataset comprised two key indicators: gross domestic product (GDP) and population density. Gross domestic product values were acquired from the Resource and Environment Science Data Center http://www.resdc.cn/ (accessed on 28 May 2024). Population density estimates originated from the LandScan database https://landscan.ornl.gov (accessed on 28 May 2024). Population density and GDP data processing was conducted in ArcGIS 10.8. The derived socioeconomic data underwent projection transformation followed by mask extraction. Vegetation coverage (FVC) data were obtained from the Earth Resources Data Cloud platform www.gis5g.com (accessed on 1 June 2024). The coordinate system of the data in this paper has been uniformly converted to WGS_1984_UTM_Zone_50, with a unified resolution of 30 m.

2.3. Research Methods

2.3.1. ES Calculation Methods

(1)

InVEST model part module algorithm

Ecosystem services were quantitatively assessed using the InVEST model, with three primary submodules evaluated: CS, WY, and SC. The computational algorithms underlying these submodules are presented in

Table 2. InVEST model part module algorithm.

InVEST Model	Calculation Method	Remarks
CS	$C_{tot}=C_{soil}+C_{dead}+C_{above}+C_{below}$	Ctot—total carbon storage Csoil—soil carbon density Cdead—dead organic matter carbon density Cabove—aboveground biomass carbon density Cbelow—belowground biomass carbon density
WY	$Y\left(x\right)=\left(1-{\displaystyle \frac{AET(x)}{P(x)}}\right)\times P\left(x\right)$ ${\displaystyle \frac{AET\left(x\right)}{P\left(x\right)}}=1+{\displaystyle \frac{PET\left(x\right)}{P\left(x\right)}}-{\left[1+{\left({\displaystyle \frac{PET\left(x\right)}{P\left(x\right)}}\right)}^{\omega }\right]}^{{\displaystyle \frac{1}{\omega }}}$ $PET\left(x\right)=K_c\left({\mathcal{l}}_x\right)·E{T}_0\left(x\right)$ $\omega \left(x\right)=Z{\displaystyle \frac{AWC\left(x\right)}{P\left(x\right)}}+1.25$	Y(x)—annual water yield (mm) at pixel x AET(x)—annual actual evapotranspiration at pixel x P(x)—annual precipitation (mm) at pixel x PET(x)—potential evapotranspiration of grid x Kclx—plant evapotranspiration coefficient ET0(x)—reference crop evapotranspiration AWC(x)—plant available water content ω—non-physical parameter characterizing soil properties under natural climatic conditions Z—empirical constant
SC	$RKLS=R·K·LS$ $USLE=R·K·LS·C·P$ $SDR=RKLS-USLE$	RKLS—potential soil erosion amount USLE—actual soil erosion amount SDR—sediment delivery ratio K—soil erodibility factor R—rainfall erosivity factor LS—slope length and steepness factor P—support practice factor C—cover management factor

Note: CS—carbon storage, WY—water yield, and SC—soil conservation.

[27].

(2)

Forage supply

The forage supply in Hulunbuir was estimated using vegetation net primary productivity (NPP) data combined with grassland belowground biomass ratios. The formula is as follows:

$$Y_{\mathrm{m}}={\displaystyle \frac{NPP}{t\left(1+r\right)}}$$

(1)

In the formula, Y_m denotes grassland productivity per unit area (kg/m²), where t represents the biomass-to-productivity conversion coefficient (0.45 [28]). The parameter r indicates the belowground-to-aboveground biomass ratio [29].

2.3.2. Ecosystem Service Trade-Off/Synergy Assessment Methodology

The coupling coordination degree model is a quantitative analysis tool. It is primarily used to study the interactions between multiple systems. Additionally, it analyzes the relationships of coordinated development among these systems. Coupling refers to a dynamic linkage mechanism formed between different motion forms or system structures through material cycles, energy conversion, and information transfer. Ecosystem services themselves exhibit typical systemic characteristics. There exist complex material–energy–information networks among the internal components. Based on this theoretical framework, the coupling coordination degree model can precisely characterize the interaction intensity between ES. The calculation formula is as follows:

$$C=2{\displaystyle \frac{\sqrt{F(x)-F(y)}}{F(x)+F(y)}}$$

(2)

$$D=\sqrt{C\cdot T}$$

(3)

$$T=a\cdot F(x)+b\cdot F(y)$$

(4)

where C is the coupling degree and C ∈ [0, 1]; F(x) and F(y) represent two different ESs; T is the integrated harmonization index for both ESs; a and b are the parameters, because the two ESs participate equally in the coupled system, so a = b = 0.5; and D represents the degree of coupling coordination of the system, and the larger the value of D, the more coordinated the two.

The trade-offs/coordination relationships among ESs are divided into extremely strong trade-off, strong trade-off, relatively strong trade-off, relatively weak trade-off, extremely weak trade-off, extremely weak synergy, relatively weak synergy, relatively strong synergy, strong synergy, and extremely strong synergy according to the degree of coupling synergy. The specific classification criteria are shown in

Table 3. Coupling coordination degree types and their classification criteria.

Coordination Degree D	0 < D < 0.1	0.1 ≤ D < 0.2	0.2 ≤ D < 0.3	0.3 ≤ D < 0.4	0.4 ≤ D < 0.5	0.5 ≤ D < 0.6	0.6 ≤ D < 0.7	0.7 ≤ D < 0.8	0.8 ≤ D < 0.9	0.9 ≤ D < 1
relation	extremely strong trade-off	strong trade-off	relatively strong trade-off	relatively weak trade-off	extremely weak trade-off	extremely weak synergy	relatively weak synergy	relatively strong synergy	strong synergy	extremely strong synergy

.

2.3.3. Optimal Parameter Geographic Detector

The conventional Geographical Detector method is susceptible to subjective bias when classifying discrete–continuous mixed data due to its reliance on manually predefined parameters, which may compromise the objectivity of data discretization. In contrast, the OPGD employs intelligent optimization algorithms to adaptively determine optimal segmentation thresholds. This approach enables more precise extraction of latent information embedded in explanatory variables within geospatial feature spaces. Four classification methods are used: natural breaks, quantile, equal interval, and geometric. Considering the natural and socioeconomic characteristics of the Hulunbuir region, this study selected eight representative indicators: mean annual temperature (X₁), mean annual precipitation (X₂), fractional vegetation cover (X₃-FVC), digital elevation model (X₄-DEM), slope (X₅), land use type (X₆), gross domestic product (X₇-GDP), and population density (X₈). By discretizing the data and comparing the q-statistics of different parameter combinations, the combination with the highest q-value is selected. This approach systematically evaluates and quantifies the driving factors of the interactions between ESs in the Hulunbuir region.

3. Results

3.1. Spatiotemporal Distribution Characteristics of Ecosystem Services

From 2000 to 2020, ESs in Hulunbuir City exhibited significant spatiotemporal differentiation patterns. The spatial distribution of CS and SC showed a characteristic of “high in the northeast and central regions, low in the west and southeast.” Carbon storage displayed an overall stable trend, while SC showed an increasing trend. Water yield exhibited a distribution pattern of “high in the central and western transition zone and southeast, low in the west.” It showed an increasing trend year by year. Forage supply primarily displayed a “high in the west, low in the east” spatial gradient. It showed a decreasing trend year by year (Figure 3).

3.2. Spatial Heterogeneity Analysis of Interactions Between Ecosystem Services

From 2000 to 2020, the relationships among various ESs in the Hulunbuir region exhibited distinct spatiotemporal heterogeneity. The distribution of ESs in different regions exhibited spatial unevenness due to factors such as human activities, land use changes, and variations in the natural environment (Figure 4). The spatiotemporal heterogeneity characteristics of the three pairs of ESs—CS and WY, CS and SC, and SC and WY—are relatively similar. The trade-off relationships are primarily concentrated in the western region of Hulunbuir, while the synergistic relationships are predominantly observed in the central and eastern regions. Over time, the level of synergy between CS-WY and SC-WY has consistently increased. Similarly, the spatiotemporal heterogeneity characteristics of the CS-FS and FS-WY pairs are relatively consistent. The trade-off relationships are primarily concentrated in the central and eastern regions of Hulunbuir, with an increasing trend over time. The trade-off/synergy relationships of FS-SC exhibit relatively stable spatiotemporal heterogeneity. The trade-offs are widely distributed across the entire Hulunbuir region, with their intensity steadily rising.

3.3. Drivers of Ecosystem Service Trade-Off and Synergy: A Systematic Analysis

3.3.1. Driving Factors and Multicollinearity Test Results

The trade-off/synergy of ESs are influenced by various factors, including climate, topography, economics, etc. This study selects eight factors for analysis: average temperature, annual precipitation, FVC, DEM, slope, land use type, GDP, and population density. The correlation among influencing factors may interfere with model judgment and reduce computational accuracy; therefore, multicollinearity tests need to be performed first. This study uses two indicators, tolerance and variance inflation factor (vif), to assess the multicollinearity of influencing factors on ES trade-offs/synergies. For the trade-offs/synergies of CS-WY in 2000, eight influencing factors were overlaid for analysis. As shown in

Table 4. The results of the impact factor multicollinearity test.

Driving Factors	Tolerance	VIF
average temperature	0.344	2.905
annual precipitation	0.215	4.647
fractional vegetation cover	0.338	2.958
digital elevation model	0.509	1.966
slope	0.864	1.158
land use type	0.549	1.822
gross domestic product	0.769	1.300
population density	0.767	1.304

, the tolerance values for all factors are greater than 0.2. Additionally, the VIF values are all below 5. This indicates that the collinearity test was successfully passed, and no redundant factors are present.

3.3.2. Optimal Discretization Processing

Taking the average annual temperature (X₁) as an example, it is shown that when the average annual temperature is divided into 10 categories by natural breakpoint method, the average annual temperature has the greatest explanatory power to the trade-off/synergy relationship among ESs in Hulunbuir region. The results are shown in Figure 5 as follows:

3.3.3. Single-Factor Detection

According to the factor detection results (

Table 5. Probing force values for different explanatory variables.

Explanatory Variable	Detection Force Value	Mean Temperature	Precipitation	FVC	DEM	Slope	Land Use Type	GDP	Population Density
CS-WY	Year 2000 Year 2010 Year 2020	0.255 0.157 0.231	0.559 0.653 0.650	0.698 0.821 0.732	0.219 0.405 0.262	0.004 0.005 0.008	0.151 0.283 0.242	0.239 0.192 0.236	0.130 0.169 0.127
CS-SC	Year 2000 Year 2010 Year 2020	0.283 0.173 0.222	0.651 0.662 0.655	0.745 0.830 0.838	0.284 0.247 0.288	0.155 0.156 0.162	0.563 0.602 0.570	0.150 0.182 0.103	0.032 0.054 0.050
CS-FS	Year 2000 Year 2010 Year 2020	0.289 0.270 0.287	0.633 0.694 0.618	0.733 0.842 0.879	0.150 0.175 0.176	0.086 0.068 0.069	0.546 0.587 0.593	0.149 0.237 0.231	0.005 0.033 0.032
WY-SC	Year 2000 Year 2010 Year 2020	0.354 0.115 0.230	0.781 0.844 0.809	0.788 0.855 0.819	0.249 0.291 0.306	0.063 0.072 0.107	0.340 0.479 0.462	0.162 0.177 0.112	0.032 0.038 0.017
WY-FS	Year 2000 Year 2010 Year 2020	0.237 0.261 0.244	0.498 0.508 0.487	0.670 0.791 0.848	0.104 0.124 0.113	0.068 0.055 0.055	0.363 0.428 0.439	0.188 0.249 0.283	0.033 0.085 0.069
SC-FS	Year 2000 Year 2010 Year 2020	0.241 0.247 0.235	0.385 0.438 0.350	0.702 0.814 0.879	0.055 0.079 0.080	0.024 0.019 0.019	0.234 0.294 0.293	0.117 0.230 0.268	0.005 0.021 0.022

Note: CS—carbon storage, WY—water yield, SC—soil conservation, FS—forage supply, FVC—fractional vegetation cover, DEM—digital elevation model, and GDP—gross domestic product.

), mean annual precipitation, FVC, and land use type exhibited the strongest explanatory power for the inter-relationships among ES. Among these factors, FVC emerged as the dominant driver, with all q-values ranging from 0.670 to 0.879. Notably, the influence of FVC showed a progressive strengthening trend over time. Slope and population density demonstrated the lowest explanatory power (q-values) regarding the inter-relationships among ESs. Specifically, slope showed minimal influence on the trade-off and synergy between CS and WY, with all q-values remaining below 0.01. Despite a slight temporal increase, slope consistently showed negligible explanatory power. The significance was adjusted using three methods—Bonferroni, Holm, and Benjamini–Hochberg (BH-FDR)—with a uniform significance level set at 0.05 (BH-FDR also considered 0.10 as a reference). The adjusted p-values for all factors were much smaller than 0.05 (Appendix A), indicating that their q-values are statistically highly significant, with no risk of false positives. All three multiple correction methods consistently passed the significance tests, demonstrating that the results are robust and reliable. The Bonferroni method is the most conservative, while BH-FDR retains higher statistical power while controlling the false discovery rate (FDR).

3.3.4. Interaction Detector

The interaction detector analysis was performed using an OPGD to quantify the joint effects of factor combinations on ES relationships in Hulunbuir. These results are illustrated in Figure 6. The interaction detector results were categorized into three distinct types: nonlinear enhancement: the interaction between two variables strengthens their combined effect on the target in a non-linear way; bivariate enhancement: two variables together explain the target better than individually; and univariate weakening: one variable’s effect on the target is weakened by interaction with another variable. Compared to individual factors’ effects on ES inter-relationships in Hulunbuir, most factor interactions exhibited enhanced explanatory power. The observed spatiotemporal heterogeneity of ES relationships primarily resulted from the integrated effects of multiple driving factors. The interaction X₁∩X₂ consistently dominated the trade-off/synergy relationships among CS-WY(Y₁), CS-FS(Y₃), WY-SC(Y₄), WY-FS(Y₅), and SC-FS(Y₆) in 2000 and 2010, with q-values >0.5. Although X₁ has a relatively small explanatory power in the single-factor detection, the interaction factor X₁∩X₂ exhibits the strongest explanatory power among all the interaction factors. All interaction q-values exceeded 0.5, indicating that this factor pair explained over 50% of the spatiotemporal heterogeneity in ES inter-relationships. In 2020, X₂∩X₇ dominated the interactions for Y₁, Y₃, Y₅, and Y₆ (q > 0.58), while X₂∩X₄ was most influential for Y₂, Y₄. Across 2000–2020, X₂∩X₄ continued to be the dominant interaction factor influencing the CS–SC relationship.

4. Discussion

4.1. Causal Analysis of the Spatial Pattern of Ecosystem Services Driven by Land Use

The spatial distribution patterns of ESs are not only constrained by natural geographical conditions but are also closely related to land use types and the intensity of human activities [30]. The western region of Hulunbuir is primarily dominated by grassland ecosystem (Figure 2), which are heavily influenced by grazing activities [31,32]. This area is characterized by low CS and high FS potential (Figure 3). The central and northeastern regions belong to the Greater Hinggan Mountains, where the main land use type is forest land. These areas are minimally affected by human activities, making them high-value zones for CS and SC. Studies by Chao et al. [33], Pang et al. [34], and Wang et al. [35] also indicate that high-value CS and SC areas are located in the northeastern forest land of Inner Mongolia. High WY areas are primarily distributed in the central and western agricultural–pastoral transition zone and the southeastern region. This area features complex terrain and is influenced by warm, moist air currents, resulting in relatively high precipitation. The main land use type is cultivated land, which is subject to strong human activity interference. Hu et al. [36] have also shown that WY in the Ruoergai Plateau is influenced by vegetation type, precipitation, topography, and evapotranspiration.

4.2. Trade-Offs/Synergies of Ecosystem Services

Ecosystem services exhibit significant trade-offs and synergies, with these inter-relationships demonstrating pronounced spatiotemporal heterogeneity [37]. Our study reveals that trade-offs among ESs in Hulunbuir have gradually weakened over time. Conversely, synergistic relationships have shown an overall strengthening trend. CS-WY, CS-SC, and WY-SC exhibited similar spatial distribution patterns. Trade-offs primarily occurred in western Hulunbuir, while synergies dominated the central and eastern region. These spatial patterns were mainly influenced by land use change, climate, and topographic conditions. The western region is predominantly characterized by grasslands and residential areas, with an arid climate and strong evapotranspiration. The reduction in grasslands and the increase in construction land (Figure 2) have led to ecological degradation, resulting in trade-off relationships between CS-WY and CS-SC. The eastern region is primarily composed of high-altitude forested areas, and the forest area has continuously increased between 2000 and 2020, with good vegetation coverage [38]. This region receives higher precipitation, and the low-temperature conditions at elevated altitudes reduce evapotranspiration water loss from vegetation. The relationships between CS-WY and CS-SC were predominantly synergistic, with the degree of synergy showing a temporal increase. A substantial portion of carbon storage originates from soil carbon [39], highlighting the critical linkage between soil health and ecosystem functioning. Effective soil conservation measures, including fractional vegetation cover and windbreak–sand fixation systems, significantly reduce soil erosion and maintain soil structural integrity. These practices consequently enhance soil carbon storage capacity. Carbon storage–forage supply exhibited overall trade-off relationships across the region (Figure 3). However, in the western region of Hulunbuir, the relationship is characterized by synergy, primarily related to land use types. The western region is dominated by grassland (Figure 2), where deep root systems of grasses promote soil organic matter accumulation and consequently enhance CS [40]. Mashizi et al. [41] identified overgrazing as a contributing factor to CS reduction. This conclusion indirectly supports the existence of a synergistic relationship between FS and CS. Overgrazing and improper grassland utilization may also lead to soil degradation and erosion [42]. The well-developed root systems of grasslands contribute to soil stabilization and reduce soil erosion. They also enhance the soil’s ability to maintain its structure. Additionally, proper land use and grazing management promote healthy grass growth. This approach helps prevent soil degradation caused by overgrazing.

4.3. Drivers of Ecosystem Service Trade-Offs/Synergies

The Hulunbuir region faces typical ecological issues such as grassland degradation, soil erosion, and desertification [43]. The study finds that there are clear trade-offs and synergies between different ESs. The underlying mechanisms are not caused by a single factor but instead is the result of the long-term coupling of natural and human factors. This study identified significant influences of natural and anthropogenic factors on ES inter-relationships, particularly FVC, precipitation, and land use types. These factors exhibit strong explanatory power, and their explanatory power shows an overall increasing trend over time. Liao et al. [44] concluded that rainfall and vegetation cover have a significant explanatory power for ecosystem services. However, their study only analyzed single-factor exploration. In contrast, this study also investigates the interactions between factors. In the interaction analysis, the driving explanatory power influencing the trade-offs and synergies of the six pairs of ESs is mainly characterized by two-factor enhancement and nonlinear enhancement. Statistical analyses revealed that the average annual temperature–mean annual precipitation, mean annual precipitation–DEM, and mean annual precipitation–GDP interactions exert significant effects on ES trade-off/synergy relationships in the Hulunbuir ecosystem. On one hand, climate factors play a fundamental regulatory role in the interactions between ESs. Precipitation and temperature, as key variables, have a significant impact on services such as FS, WY, and CS. Precipitation directly regulates grassland biomass and hydrological processes, serving as a key factor for FS and WY [45,46]. Temperature, on the other hand, influences plant phenology and evapotranspiration, thereby affecting the effectiveness of water resources and soil erosion resistance in ecosystems [47]. Both factors jointly influence vegetation physiological activities and community structure processes. They regulate the interactions between multiple ecological functions. On the other hand, land use is a significant human factor influencing the trade-offs and synergies of ESs. The western part of Hulunbuir is primarily a grassland ecosystem. High-intensity human disturbances, particularly overgrazing and the expansion of built-up areas, lead to the decline in carbon storage and soil conservation functions.

Different stakeholders hold different views on ecosystem services. Herders are more focused on increasing the supply of forage to provide adequate food for livestock, which leads them to favor measures that can bring short-term economic returns. However, environmental organizations emphasize the importance of ecological protection, arguing that maintaining the health of ecosystem services can provide long-term economic benefits for society. Therefore, balancing livelihood economics and ecological protection is key to promoting the collaborative management of ecosystem services. The management and optimization of regional ecosystem services in the future should be based on scientific planning, taking into account regional characteristics and socioeconomic needs. Effective ecological protection policies should be developed to maximize the ecosystem service functions and promote sustainable development.

4.4. Insufficient

This study only analyzed ecosystem services for the years 2000, 2010, and 2020, which provides limited temporal resolution and does not explore changes in more recent periods. However, the current study still reflects changes in ecosystem services in the Hulunbuir region, as well as the associated trade-offs/synergies, offering a reference for future research. Additionally, the InVEST assessment model has its limitations. For example, in the WY module, although the focus is on the water conservation benefits of natural precipitation, it does not fully integrate groundwater flow, surface runoff, and the water supply capacity of the watershed. Further improvements and refinements are needed in future studies.

5. Conclusions

To enhance the ecosystem service functions in the Hulunbuir region, reconcile the conflicts between ecological conservation and regional economic development, and promote the collaborative and sustainable development of the ecological-economic system, it is of significant theoretical and practical importance to investigate the interrelationships among the ecosystem services in this region. Such development requires comprehensive understanding of ES inter-relationships specific to this region. Systematic investigation of these relationships carries both theoretical significance and practical management implications.

(1) From 2000 to 2020, the spatial distribution of ecosystem services in the Hulunbuir region was primarily influenced by land use types, exhibiting an overall “high in the east and low in the west” pattern. However, FS followed a “high in the west and low in the east” distribution trend.

(2) Over time, the synergy between CS-WY and SC-WY continues to increase, while the trade-off between CS-FS, FS-WY, and FS-SC strengthens. The trade-off/synergy relationships of ESs are unevenly distributed spatially.

(3) Fractional vegetation cover exerted the most significant influence on trade-offs and synergies among ESs. This influence exhibited a progressively intensifying trend over time. Slope gradient and population density exhibited relatively minor effects on the inter-relationships among ESs. Most interaction factors demonstrated enhanced effects on the inter-relationships among various ESs in the Hulunbuir region compared to individual factors. Among them, the interaction between XI and X2 explains a significant portion of the trade-off/synergy relationships between ESs.