Study on Multi-Scale Characteristics and Influencing Factors of Trade-Offs and Synergies between Ecosystem Services in Jiangxi Province

Abstract

The trade-offs and synergies reveal the profit and loss relationship between ecosystem services, which is of great significance to the sustainable development of natural resources. The ecosystem services in Jiangxi Province, such as net primary productivity (NPP), soil conservation (SC) and water yield (WY) during 2000–2020, were estimated in this study. The correlation coefficient was adopted to analyze the trade-offs and synergies between the three ecosystem services by static space correlation and dynamic space correlation from such perspectives as Watershed, county and grid. Moreover, the influence of the three ecosystem services and the relations between them were explored from four aspects: landform, NDVI, accumulated temperature and precipitation. The results showed that the ecological environment quality in Jiangxi Province was improved and that the distribution of ecosystem services had significant regional characteristics. In the static analysis, ecosystem services at all scales were remarkably synergistic, and synergies weakened rapidly and even turned into trade-offs as the scale decreased. In the dynamic analysis, ecosystem services at all scales were mainly synergistic; the proportion of significant samples was much lower than that in the static analysis, the degree of trade-offs/synergies decreased with the decrease in scale, and the decrease was smaller than that in the static analysis. The major constraints for SC were landform and NDVI. The main constraint for WY was precipitation, and that for NPP was NDVI. Affected by various factors, NPP and SC were stably synergistic, NPP and WY were in a stable trade-off relationship, and the relationship between SC and WY was unstable. The trade-offs and synergies changed with factors and zoning.

1. Introduction

Ecosystem services refer to the natural environmental conditions and effects formed by the ecosystem and maintained by humans [1] and are the benefits that humans directly or indirectly obtain from the ecosystem [2]. The quality and quantity of natural ecosystem services are the natural conditions for human survival and development quality [3]. The increasing human demands for the provision of services have come at the cost of reduced regulatory and cultural services and biodiversity for a long time [4]. China has a vast territory with rich and diverse ecosystem types and an absolute quantity of natural resources. The early extensive economic growth has paid some environmental costs. With the implementation of such policies as Ecological Civilization and Beautiful China and the significant improvement in the overall comprehensive strength of China, the ecological and environmental quality of China has also been greatly improved. However, the dual influence of the uneven spatiotemporal distribution of climate resources and profit-oriented human activities is still one of the constraints affecting the coordinated development of the economy and environment [5].

The trade-offs and synergies between ecosystem services refer to the competitive utilization of ecological components and natural resources [6]. Trade-offs and synergies appear under the combined influence of human interference [7,8], which has been a hotspot for academic circles and decision-makers [9]. After long-term exploration, many research results have been achieved in the research methods [10,11,12,13,14] and driver analysis relevant [15,16,17] to trade-offs and synergy analysis. Many studies have found that ecosystem services are affected by regional physico-geographical factors and socio-economic factors [18,19], and the same pair of services in the same region present different trade-offs and synergies relationships at different temporal [20,21] or spatial scales [22,23], and even changing trade-offs and synergies relationships [24]. For instance, the study of Sun Y.J. et al. in the Guanzhong Basin, China, has shown that the correlations between ecosystem services have a certain evolution law with time [25]. In a study in the central part of the dry region in northern China, Sun Z.X. et al. demonstrated significant differences in the trade-off relationship between ecosystem services at different spatial scales [26]. As the basis for understanding the relations between multiple ecosystem services, carrying out the research on the trade-offs between multi-scale and multi-type services, and clarifying the trade-offs or synergies between multiple services and the scale dependence and spatial difference characteristics and the priorities of service trade-offs management [27,28].

In this study, net primary productivity (NPP), soil conservation (SC), water yield (WY) and other ecosystem services in Jiangxi Province during 2000–2020 were evaluated. Additionally, the trade-offs and synergies between multiple types of ecosystem services were explored at such scales as watershed, county and grid by spatial correlation analysis. Finally, the spatial distribution characteristics and influencing factors of ecological environment status were analyzed.

2. Materials and Methods

2.1. Study Region

Jiangxi Province is located in central China and the middle and lower reaches of the Yangtze River. Its main geomorphic types are mountains and hills. The western and southern parts of Jiangxi are mountainous and hilly, while the plains are around Poyang Lake (Figure 1a). The annual precipitation is 1751–2523 mm [29]. In terms of regional distribution, the precipitation is more in the east than in the west. The average annual precipitation of the Xinjiang River basin in the east is 1855 mm, and the average annual precipitation of the middle reaches of the Ganjiang River in the west is 1562 mm [30]. The rainy season lasts from April to October. Poyang Lake is the largest freshwater lake in China, which receives water from five rivers, Ganjiang River, Fuhe River, Xinjiang River, Raohe River and Xiushui, and flows into the Yangtze River via the lake outlet channel, with a Watershed area of 162,200 km², accounting for 94.0% of the territorial area of Jiangxi Province [31] (Figure 1b). The main soil types are red soil and yellow soil. The former is the most extensively distributed. There are abundant vegetation resources in Jiangxi Province. The forest coverage in 2020 was 61.6% [32]. Therefore, Jiangxi Province is an important ecological barrier in central China.

2.2. Data Sources and Processing

The Land Use and Land Cover (LULC) data of Jiangxi Province during 2000–2020 came from China’s annual land cover data (CLCD) set published by Professor Yang Jie and Huang Xin’s team at Wuhan University [33]. The 250 m resolution normalized difference vegetation index (NDVI) data during 2000–2020 were collected from the MOD13Q1 product, which was retrieved from the United States Geological Survey (USGS). The climate data during 2000–2020 came from the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences. The 30 m resolution DEM data came from the Geospatial Data Cloud of the Chinese Academy of Sciences (http://www.gsclound.cn, accessed on 8 February 2023). Soil thickness, plant available water content and other data came from the World Agroforestry Center (ISRIC) (http://www.isric.org/data/, accessed on 8 February 2023). The calculation method or value of the parameters required for the model came from literature retrieval [34,35,36,37,38,39,40,41,42,43,44,45].

Basic data required in the study were processed in batches with the software of MRT (Version 4.1, Sioux Falls, SD, USA) [46], SPSS (Version 26.0, Chicago, IL, USA), Envi (Version 5.3, Redlands, CA, USA), and Arcgis (Version 10.5, Redlands, CA, USA) involved. Additionally, the final unified space refers to WGS_1984 _UTM_Zone_50N with a resolution of 250 m.

2.3. Methods and Model

Please refer to the Millennium Ecosystem Assessment and the study of Pan Jinghu et al. for the selection of objects of study [47,48,49,50]. Three ecosystem services, NPP, SC and WY, were selected as the objects of study for the evaluation of ecosystem services in Jiangxi Province by the principles of comprehensiveness, importance, sustainability of evaluation and availability of data.

2.3.1. Estimation of NPP

The NPP was mainly estimated by the remote sensing model for estimating the NPP of vegetation. The estimation formula is as follows:

$$\mathrm{NPP}\left(x,t\right)=APAR\left(x,t\right)\times \epsilon \left(x,t\right)$$

where $APAR\left(x,t\right)$ is the photosynthetically active radiation absorbed by grid x in month t (gC/m²), and $\epsilon \left(x,t\right)$ is the actual utilization of light energy of grid x in month t (gC/MJ).

2.3.2. Estimation of SC

SC was estimated by the sediment delivery ratio (SDR) module of Integrated Valuation of Ecosystem Services and Trade-offs model (Invest). The estimation formula is as follows:

$${\mathrm{SC}}_x=\left(RKL{S}_x-USL{E}_x\right)+DE{P}_x$$

where ${\mathrm{SC}}_x$ is soil conservation on grid x (t); $RKL{S}_x$ is potential soil erosion on grid x (t); $USL{E}_x$ is potential soil erosion on grid x in land cover (t); $S{D}_x$ is Sediment volume intercepting the upstream grid on grid x (t).

2.3.3. Estimation of WY

WY was estimated by the SDR module of the Invest mode. The estimation formula is as follows:

$$Y\left(x\right)=\left(1-\frac{AET\left(x\right)}{P\left(x\right)}\right)\cdot P\left(x\right)$$

where $Y\left(x\right)$ is the annual water yield on grid x (m³), $AET\left(x\right)$ is the annual actual evapotranspiration for grid x, and $P\left(x\right)$ is the annual precipitation on grid x (m³).

2.3.4. Analysis and Verification Method of Trade-Offs and Synergies

Types were calculated in this paper at multiple scales by Pearson correlation trade-offs analysis method. The calculation formula is as follows:

$$R=\frac{\sum \left(x-\overline{x}\right)\left(y-\overline{y}\right)}{\sqrt{\sum {\left(x-\overline{x}\right)}^2}\sqrt{\sum {\left(y-\overline{y}\right)}^2}}$$

where R is the correlation coefficient between ecosystem services x and y. A positive correlation coefficient indicates a synergistic relation, while a negative value means a trade-off relationship.

The significance of the correlation was tested by the T test of zero hypothesis. The calculation formula is as follows:

$$T=\frac{R}{\sqrt{\frac{1-R^2}{n-2}}}$$

where R is the corresponding correlation coefficient, and n is the sample size. According to the table of critical values of bilateral T test, when $\left|T\right|>T_{0.05, n-2}$, the correlation results are significant; when $\left|T\right|>T_{0.01, n-2}$, the correlation results are extremely significant.

2.3.5. Analysis of Influencing Factors for Ecosystem Services

In this study, the factors were explored using the GeoDetector model, and their influences on the 3 types of ecosystem services in Jiangxi province were analyzed. The calculation formula is as follows:

$$q=1-\frac{{\sum }_{h=1}^L{N}_h{\sigma }_h^2}{N{\sigma }^2}$$

where L is the number of partitions of factor X; $N$ and $N_h$ represent the number of units in the whole area and layer h, respectively; and ${\sigma }_h^2$ and ${\sigma }^2$ indicate the variance of dependent variable Y in layer h and the whole area, respectively. The range of $q$ is [0, 1]. The higher the $q$, the more obvious the spatial variation in Y, indicating that factor X explains $100\times q\%$ Y [51].

3. Results

3.1. Spatiotemporal Changes in Ecosystem Services

3.1.1. Spatiotemporal changes in NPP

On the time scale, the NPP in Jiangxi Province from 2000 to 2020 showed an increasing trend and increased from 1.05 × 10⁸ t in 2000 to 1.14 × 10⁸ t in 2020 (Figure 2a). The mean of NPP per hectare increased from 6.33 t/hm² to 6.86 t/hm². The NPP of all Watersheds showed an increasing trend. The NPP of the Fuhe River Watershed and Xinjiang River Watershed increased the most significantly, while the Poyang Lake and the main stream of the Yangtze River showed the minimum increase.

On the spatial scale, the distribution of NPP was high in the middle and low in the surrounding areas during 2000–2020 (Figure 3a–e). The low-value area was around the Poyang Lake region, where the landform types mainly are plains and waters. The high-value areas were in the upper reaches of the Ganjiang River, Dongjiang River Watershed and Raohe River Watershed, where the main landforms are mountains and hills.

3.1.2. Spatiotemporal Changes in SC

On the time scale, the SC in Jiangxi Province from 2000 to 2020 showed an increasing trend and increased from 3.06 × 10¹⁰ t in 2000 to 3.78 × 10¹⁰ t in 2020 (Figure 2b). The mean of SC per hectare increased from 1839.44 t/hm² to 2271.89 t/hm². It dropped after it peaked in 2010. Affected by precipitation, the year-to-year fluctuations of the SC of Xinjiang River Watershed and Raohe River Watershed were great. The SC of the Ganjiang River watershed, Raohe River Watershed and Xiushui Watershed increased the most greatly, while that of the Dongjiang River Watershed decreased gently. The SC of the other Watersheds showed a relatively gentle increase.

On the spatial scale, the SC of Jiangxi Province was low in the middle and sporadically high in the surrounding areas (Figure 3f–j). The low-value area was in the area around the Poyang Lake, while the high-value areas were mainly in mountainous and hilly areas and concentrated in the Wuyi Mountains in the Xinjiang River Watershed, the Jiuling Mountain in Xiushui River Watershed and the Huaiyu Mountain at the junction of Xinjiang River and Raohe River Watersheds.

3.1.3. Spatiotemporal Changes in WY

On the time scale, the WY in Jiangxi Province from 2000 to 2020 showed an increasing trend and increased from 1.69 × 10¹¹ t in 2000 to 1.91 × 10¹¹ t in 2020 (Figure 2c). The mean of WY per hectare increased from 10,147.53 t/hm² to 11492.86 t/hm². It dropped after it peaked in 2015. Only the Dongjiang River Watershed showed a gentle decreasing trend; other Watersheds showed an increasing trend. The increase in the area around Poyang Lake, Raohe River watershed and Yangtze River Watersheds was great, while that in Ganjiang River Watershed and Xiangjiang River Watershed was the smallest.

On the spatial scale, the WY in Jiangxi Province was generally low in the north and south and high in the middle and east from 2000 to 2020 (Figure 3k–o). The low-value areas were mainly in the Ganjiang River Watershed, while the high-value areas were mainly in the Xinjiang River Watershed.

3.2. Analysis of Trade-Offs and Synergies between Ecosystem Services

At present, the correlation between ecosystem services of the object of study is studied by scholars by analyzing the trade-offs and synergies between ecosystem services with an annual or multi-year average value of ecosystem services. On the basis of the static spatial correlation method, it was proposed in this study to perform dynamic spatial correlation analysis with the time series file of ecosystem services, i.e., analyzing the correlation between the multi-period (more than three periods) data of certain ecosystem services of the object of study and the multi-period data of another service. The differences in trade-offs and synergies between ecosystem services in Jiangxi Province on a variety of statistical scales were revealed statically and dynamically. The correspondence between correlation and the level of trade-offs/synergies referred to relevant literature [52] (

Table 1. Table of Correspondence Between R-Value and Level of Trade-off/Synergy.

R-Value	0	$0<\left|\mathit{R}\right|\le 0.3$	$0.3<\left|\mathit{R}\right|\le 0.7$	$0.7<\left|\mathit{R}\right|\le 1$
Level of trade-off/synergy	No	Weak	Moderate	Strong

).

3.2.1. Analysis of Static Spatial Correlation

In this study, for convenient comparison, the grid was used as the evaluation unit to calculate the mean of ecosystem services over the years; and the correlation between services was calculated from the three statistical scales of samples, watershed, county and grid, so as to judge the trade-offs and synergies. On watershed and county-level scales, the ArcGIS zoning statistical function was used to calculate the total value of ecosystem services of grids in each region. All three scales passed the significance test at the 0.01 level (

Table 2. Correlation Coefficient Between Three Ecosystem Services in Jiangxi Province.

Correlation Type	Watershed Scale	County Scale	Grid Scale
NPP-SC	0.942 **	0.766 **	0.301 **
NPP-WY	0.992 **	0.885 **	−0.246 **
SC-WY	0.937 **	0.670 **	0.004 **

Note: ** indicates significant association at the 0.01 level (bilateral) (two-tailed).

).

The study showed that the correlation between services decreases rapidly with the decrease in scales. The relation between the three ecosystem services at the watershed and county scale is strong synergy, while at the grid scale, the relation between NPP and WY is a weak trade-off, and the other two relations are weak synergy or moderate synergy.

In order to further reveal the differences in the correlation between different watersheds at the grid scale, a grid-scale correlation analysis was carried out on the relation between ecosystem services in the nine major watersheds in Jiangxi Province. It was found that the correlation results between the services in different watersheds were more or less different from the results on the grid scale throughout Jiangxi Province (

Table 3. Correlation Coefficient Between Three Ecosystem Services in Nine Watersheds in Jiangxi Province and Grid Scale.

Correlation Type	Gangjiang Watershed	Fuhe Watershed	Xinjiang Watershed	Raohe Watershed	Xiushui Watershed	Other Watersheds	The Area Around Poyan Lake	Dongjiang Watershed	Xiangjiang Watershed
NPP-SC	0.309 **	0.372 **	0.351 **	0.198 **	0.237 **	0.375 **	0.349 **	0.298 **	0.341 **
NPP-WY	−0.408 **	−0.451 **	−0.403 **	−0.234 **	−0.259 **	−0.070 **	0.024 **	−0.463 **	−0.600 **
SC-WY	−0.071 **	−0.176 **	−0.143 **	−0.017 **	−0.072 **	−0.077 **	−0.104 **	−0.045 **	−0.194 **

Note: ** indicates significant association at the 0.01 level (bilateral) (two-tailed).

).

The trade-offs/synergies between NPP and SC and NPP and WY were basically consistent with the results on the grid scale of the whole province (Table 1 and Table 2, Figure 4). The synergy between NPP and SC in most watersheds was stronger than the results at the grid scale throughout Jiangxi Province. The synergy between Raohe River Watershed, Xiushui Watershed and Dongjiang River Watershed was weaker than the results at the grid scale in this province.

NPP and WY were in a trade-off relationship in most watersheds, which was consistent with the results on the grid scale throughout Jiangxi Province. The trade-offs in Ganjiang River Watershed, Fuhe River Watershed, Xinjiang River Watershed, Xiushui Watershed, Dongjiang River Watershed and Xiangjiang River Watershed were stronger than the results on the grid scale in the whole province, while the trade-offs in Raohe River Watershed and the main stream watershed of the Yangtze River were weaker than the results on the grid scale throughout Jiangxi Province. The services in the Poyang Lake area were in a weak synergic relationship.

The SC and WY in all watersheds were in a weak trade-off relationship, which was opposite to the synergic relationship, the results on the grid scale in the whole province. The regions where the trade-off relationship was relatively stronger were concentrated in the Fuhe River Watershed, Xinjiang River Watershed and the central Jiangxi and eastern Jiangxi plains where the Poyang Lake is located.

3.2.2. Analysis of Dynamic Spatial Correlation

In the dynamic analysis, the correlation between the ecosystem services in the objects of study over time was analyzed by creating a time series file of each ecosystem service and with watershed, county and grid as objects of study. Additionally, the T significance test was performed.

1. Watershed scale

NPP and SC were mainly in a synergic relationship. The two services in Xiushui Watershed, Raohe River Watershed, Xinjiang River Watershed, the area around Poyang Lake and the main stream watershed of the Yangtze River were in a moderate synergic relationship, while those in the Dongjiang River Watershed were in a weak trade-off relationship. They were in a weak synergic relationship. Trade-offs/synergies did not pass the significance test (Figure 5a). NPP and WY were in a synergic relationship throughout Jiangxi Province. They were in a strong synergic relationship in Xiushui Watershed and Raohe River Watershed, a weak synergic relationship in Dongjiang River Watershed and a moderate synergic relationship in the other watersheds. The synergic relations did not pass the significance test (Figure 5d). WY and SC were in a strong synergic relationship throughout Jiangxi Province. The relations in Raohe River Watershed, Xinjiang River Watershed, Fuhe River Watershed, Dongjiang River Watershed and the main stream watershed of the Yangtze River passed the significance test at the 0.05 level (Figure 5g).

2. County scale

NPP and SC were mainly in a synergic relationship: a strong synergic relationship in 8 of the 100 counties (cities and districts) in the province, a moderate synergic relationship in 39 counties, a weak synergic relationship in 27 counties and a trade-off relationship in 26 counties. They were in a significant synergic relationship in Nankang District (Figure 5b). NPP and WY were mainly in a synergic relationship: a strong synergic relationship in 18 counties (cities and districts), a moderate synergic relationship in 14 counties, a weak synergic relationship in 59 counties and a trade-off relationship in 9 counties. They were in a significant synergic relationship in Fengxin County (Figure 5e). WY and SC were mainly in a synergic relationship: a strong synergic relationship in 87 counties (cities and districts), a moderate synergic relationship in 11 counties, a weak synergic relationship in 1 county and a trade-off relationship in 1 county. They were in an extremely significant synergic relationship in Pengze County and Yudu County, a significant synergic relationship in 28 counties in central and northeastern Jiangxi, 7 counties in western Jiangxi, 2 counties in northern Jiangxi and 4 counties in southern Jiangxi (Figure 5h).

3. Grid scale

The ecosystem services in Jiangxi Province were mainly in a synergic relationship on the grid scale. The area where NPP and SC were in a synergic relationship accounted for 64.74% of the total area of Jiangxi Province (

Table 4. Proportion of Area of Trade-off and Synergy between Ecosystem Services (%).

Correlation Type	NPP-SC	NPP-WY	SC-WY
Strong trade-off	3.87 (0.97)	3.35 (1.00)	0.03 (0.01)
Moderate trade-off	13.13	8.97	0.34
Weak trade-off	18.26	12.15	1.47
Weak synergy	23.58	20.19	1.35
Moderate synergy	29.59	38.28	12.88
Strong synergy	11.57 (2.87)	17.06 (4.47)	83.97 (35.22)

Note: The values in brackets are the proportion of grids that passed the 0.05 significance test in the whole province.

). The total proportion of weak and moderate synergies was 49.73%, while the proportion of strong synergies was 11.57%. The area where NPP and WY were in a synergic relationship accounted for 75.53% of the total area of Jiangxi Province. The total proportion of weak and moderate synergies was 58.47%, while the proportion of strong synergies was 17.06%. The area where WY and SC were in a synergic relationship accounted for 98.16% of the total area of Jiangxi Province. The proportion of strong synergies was 83.97%. The trade-offs/synergies between services were insignificant as a whole; the area where NPP and SC, NPP and WY, and WY and SC were significantly synergic accounted for 2.87%, 4.47% and 35.22% of the total area of Jiangxi Province, respectively (Table 4, Figure 5c,f,i).

3.3. Analysis of Influencing Factors for Trade-Offs/Synergies

The study determined the acting direction and intensity of influencing factors using SPSS and the GeoDetector model, respectively, so as to interpret the impacts of different influencing factors on the trade-offs/synergies in ecosystem services. Four factors, precipitation, ≥10°C active accumulated temperature, vegetation and landform, were selected in this study by literature retrieval [53,54,55] and comprehensive comparison. The factor vegetation was replaced by NDVI. Jiangxi Province was divided into four to six regions by natural breakpoint method or references [56], and all influencing factors passed the significance test (

Table 5. Correlation between Ecosystem Services and Influencing Factors.

Type of Service	Landform Area	≥10 °C Active Accumulated Temperature Area	Precipitation Area	NDVI Area
SC	0.378 **	−0.039 **	0.059 **	0.342 **
WY	−0.277 **	−0.439 **	0.890 **	−0.217 **
NPP	0.679 *	0.230 **	−0.114 **	0.893 **

Note: The asterisks indicate degree of significant association (* for p<0.05, ** for p<0.01) (bilateral) (two-tailed).

, Figure 6). The factor detection results all passed the significance test ($p\le 0.05$) (

Table 6. Q statistics of spatially divergent impact factors of Ecosystem Services.

Type of Service	Landform	≥10 °C Active Accumulated Temperature	Precipitation	NDVI
sc_$q$	0.184	0.012	0.009	0.147
wy_$q$	0.073	0.531	0.761	0.092
npp_$q$	0.480	0.078	0.032	0.747
p value	0.000	0.000	0.000	0.000

). Landform and NDVI had significant effects on SC and NPP, while accumulated temperature and precipitation had significant effects on WY. Limited by the not high significance of the dynamic analysis results, only change rules of trade-offs/synergies were explored on the static analysis grid scale.

3.3.1. Analysis of Trade-Offs/Synergies under Landform Conditions

The landforms in Jiangxi Province were grouped into four types according to the Geomorphology of China (1:1,000,000) and the geomorphology of Zhou Chenghu: plains, platforms, hills and mountains [56,57]. NPP and SC were weakly synergic, and the correlation decreased with the increase in relief. NPP and WY were uncorrelated in plains, and were in a weak trade-off relationship in platforms, hills and mountains, and the correlation decreased with the increase in relief. SC and WY were in a weak trade-off relationship in plains and platforms with smaller relief but were weak synergic in hills and mountains with larger relief (Figure 7a).

3.3.2. Analysis of Trade-Offs/Synergies under Accumulated Temperature Conditions

The 5-year mean ≥10 °C active accumulated temperature during 2000–2020 was 5708–6751 °C. Jiangxi Province was divided into six accumulated temperature areas (such as I, II...) from low to high. NPP and SC were less influenced by accumulated temperature and were in a moderate synergic relationship. NPP and WY were in a moderate trade-off relationship; there was a relatively higher correlation between Areas II and VI. SC and WY were in a weak synergic relationship in Areas III and V and a weak trade-off relationship in Areas I, II, IV and VI (Figure 7b).

3.3.3. Analysis of Trade-Offs/Synergies under Precipitation Conditions

The 5-year mean precipitation during 2000–2020 was 1537–2228 mm. Jiangxi Province was divided into six precipitation areas from low to high. NPP and SC were in a moderate synergic relationship. The correlation in Areas II–V increased with the increase in precipitation. The NPP and WY were in a moderate trade-off relationship. The curve of the correlation coefficient rose first and then dropped. SC and WY were in a weak trade-off relationship. The overall curve of the correlation coefficient also rose first and then dropped, with Area IV as a turning area (Figure 7c).

3.3.4. Analysis of Trade-Offs/Synergies under NDVI Value Conditions

NDVI was often used to characterize the physiological status of vegetation, green biomass and vegetation productivity in the study area. Jiangxi Province was divided into six NDVI areas from low to high, according to the data collected during 2000–2020. NPP and SC were in a weak synergic relationship in Areas I–V and a weak trade-off relationship in Area VI. NPP and WY were in a moderate synergic relationship in Area I, a weak synergic relationship in Area II, and a trade-off relationship in Areas III–VI. The correlation coefficient between SC and WY decreased first and then increased with the NDVI value. Area III was the turning area (Figure 7d).

4. Discussion

Jiangxi Province has a narrow and long terrain, with a north-south trend, and a relatively independent geographical unit, surrounded by mountains on three sides. There are the Mufu Mountains, Jiuling Mountains and Luoxiao Mountains lining up in the west, the Huaiyu Mountains and Wuyi Mountains in the east, the Jiulian Mountains in the south, and rolling hills in the middle, forming a huge U-shaped basin that overall tilts towards Poyang Lake and opens to the north. The hilly and mountainous areas in the region, which account for about two-thirds of the province’s total area, are undulating and have complex and diverse vertical climates, thus resulting in the complicated and changeable distribution of climate resources, presenting a climate pattern of more heat resources in the south and less in the north, and more water resources in the east and less in the west. Because of the location of Jiangxi province in the humid subtropical region, the vegetation types are complex and mainly distributed with temperate vegetation in the north and tropical-subtropical vegetation in the south. The complex terrain, landform, climate, vegetation and land use will have different impacts on the ecosystem and promote the heterogeneity of ecosystem service relationships in Jiangxi province.

4.1. Spatiotemporal Changes in Ecosystem Services

The NPP, SC and WY of Jiangxi Province during 2000–2020 showed an increasing trend as a whole, indicating that the overall ecological environmental quality of Jiangxi Province was continuously improved. However, since ecosystem services are significantly susceptible to precipitation and other climate factors (Figure 2d), the values of ecosystem services in some years fluctuated (Figure 2a–c). In terms of spatial distribution, the distribution of ecosystem services showed obvious regional features. NPP and SC were low in the middle and high in surrounding areas, while WY was low in the south and north and high in the middle and east. The conclusions are similar to that of other scholars in Jiangxi Province or the Poyang Lake watershed [58,59].

4.2. Analysis of Trade-Offs and Synergies

Because of the dependence of ecosystem service trade-offs and synergies on the spatio-temporal scale, the trade-offs and synergies relationships between ecosystem services at the global scale cannot represent the ecosystem service relationships at a more micro scale [60]. The existing research mostly analyzes the trade-offs and synergies relationships between ecosystems using static spatial correlations [61]. In this study, it is believed that the overall characteristics of ecosystem service relationships can be quickly obtained, and regional differences can be revealed in the static analysis at the basin and county scales. The static analysis at a grid scale can demonstrate the overall correlations of each evaluated unit after integration.

It has been shown that the relationships between ecosystem services at all scales are extremely significant. Services were mainly strongly synergic on large scales. The synergic degree decreased rapidly with the decrease in scales. Services were mainly strongly synergic on large scales. The synergic degree decreased rapidly with the decrease in scales. The ecosystem service values of each sample on the grid scale were the results of the comprehensive influence of all influencing factors. The differences among samples resulted in relatively smaller or even inverted results of the correlation test. The conclusions are similar to that of other scholars [62,63].

The dynamic analysis takes into account both spatial and temporal differences. the results output are the individual correlations of ecosystem services in the same unit under the continuous influence of the influencing factors. It has been shown that the relations were dominated by synergies on all scales and gradually decreased with the decrease in scales. The changes in the strength of trade-offs/synergies were smaller than that in the static analysis. Limited by the number of observation years, the proportion of significant relations was much lower than that in static analysis. The significant regions on all scales were mainly located in the strong trade-off/strong synergy regions. The extremely significant and significant regions on the grid scale were basically consistent with the extremely significant and significant regions on the basin or county scale, and the proportion of significance increased with the decrease in scales.

4.3. Analysis of Influencing Factors

Most current studies believe that if the differences in various factors such as land use patterns [64], topographic features [65] and climate factors [66] are ignored in the analysis of ecosystem service trade-offs and synergies, the results can not accurately indicate the processes of and interactions between ecosystems [67]. It has been demonstrated that when the external disturbance intensity exceeds the threshold of the ecosystem, the ecosystem services will change dramatically, leading to changes in the relationships between ecosystem services [68].

In this study, SC was greatly influenced by landform and NDVI. Its value benefited from landform, precipitation and NDVI and was constrained by landform. WY was greatly affected by precipitation. Its value benefited from precipitation and was constrained by landform, NDVI and accumulated temperature. NPP was significantly influenced by NDVI and then by landform. Its value benefited from landform, accumulated temperature and NDVI and was constrained by precipitation.

Affected by all factors, NPP and SC were in a synergic relationship. NPP and WY were in a trade-off relationship in most grids. SC and WY were in a synergic relationship in most grids when influenced by landform and NDVI and were in a trade-off relationship in grids when affected by accumulated temperature and precipitation.

4.4. Limitations and Outlook

Based on the research from the static and dynamic aspects, the trade-offs and synergies in ecosystem services in Jiangxi, China, feature obvious variations in regions, spaces and time due to the impact of natural endowments. However, some limitations and uncertainties can be observed in this research. First, it failed to study the trade-offs and synergies in the entire ecosystem, with only three ecosystem services selected for the assessment. Second, regular interval years were selected as time points. The credibility of the results may be improved in the study with continuous time series [69]. Third, influencing factors are selected as natural factors without any consideration of human factors. For this, the relationship between ecosystem services will be further explored through the improved evaluation system in the upcoming research, which can provide decision-makers with more accurate ecological measures [70].

5. Conclusions

The ecosystem services in Jiangxi Province such as NPP, SC and WY during 2000–2020 were estimated in this study with Jiangxi Province as the study area. The research was analyzed the trade-offs and synergies between the three ecosystem services from multi-scale perspectives. Additionally, the influencing factors affecting ecosystem services and the relationship between services are discussed. The main conclusions are as follows:

1. The ecological environment quality in Jiangxi Province was continuously improved during 2000–2020. Affected by climate factors, the ecosystem service value in some years fluctuated but increased as a whole. The distribution of ecosystem services showed obvious regional features. NPP and SC were low in the middle and high in surrounding areas, while WY was low in the south and north and high in the middle and east.

2. In the static analysis, the relations between ecosystem services on all scales were extremely significantly synergic. Services were mainly strongly synergic on large scales. The synergic degree decreased rapidly or even turned into trade-offs with the decrease in scales. In dynamic analysis, the relations were dominated by synergies on scales and gradually decreased with the decrease in scales. The changes in the strength of trade-offs/synergies were smaller than that in the static analysis. The proportion of significance was much lower than that in the static analysis.

3. Landform, accumulated temperature, precipitation and NDVI can significantly affect the spatial distribution and trade-offs/synergies in ecosystem services. Of which, the influence on SC is $\mathrm{Landform} 0.018>\mathrm{Ndvi} 0.147>\mathrm{Accumulated} \mathrm{Temperature} 0.012>\mathrm{Precipitation} 0.009$, the influence on WY is $\mathrm{Precipitation} \left(0.761\right)>\mathrm{Accumulated} \mathrm{Temperature} \left(0.531\right)>\mathrm{Ndvi} \left(0.092\right)>\mathrm{Landform} \left(0.073\right)$, and the influence on NPP is $\mathrm{Ndvi} \left(0.747\right)>\mathrm{Landform} \left(0.480\right)>\mathrm{Accumulated} \mathrm{Temperature} \left(0.078\right)>\mathrm{Precipitation} \left(0.032\right)$.