Modelling the Spatiotemporal Coordination Between Ecosystem Services and Socioeconomic Development to Enhance Their Synergistic Development Based on Water Resource Zoning in the Yellow River Basin, China

Abstract

The synergistic development of ecosystems and socioeconomic systems constitutes a critical foundation for achieving Sustainable Development Goals (SDGs). Large river basins characterized by ecological and socioeconomic spatial heterogeneity frequently present contradictions and conflicts in regional sustainable development, thereby impeding the realization of SDGs. This study employed the Yellow River Basin (YRB), a typical large sediment-laden river system, as a case study. Based on the secondary water resource zones, the spatial variability and temporal evolution of ecosystem service value (ESV), population (POP), GDP, nighttime light (NTL), and Human Development Index (HDI) were analyzed at the water resource partition scale. A consistent mode was applied to quantify the spatiotemporal consistency between ESV and socioeconomic indicators across water resource partitions. The results indicated that from 1980 to 2020, the ESV of the YRB increased from 1079.83 × 10⁹ to 1139.20 × 10⁹ yuan, with no notable spatial pattern variation. From upstream to downstream, the population density, GDP per unit area, and NTL per unit area displayed increasing trends along the river course, whereas the total population, GDP, and NTL initially increased and then declined. Temporally, the population fluctuated with an overall upward tendency, while GDP and NTL experienced significant growth. The spatial distribution and temporal evolution of HDI remained comparatively stable. The coefficients of variation for population, GDP, and NTL were significantly higher than those for ecosystem services and HDI. The study highlighted an overall lack of coordination between ESV and socioeconomic development in the YRB, with relatively stable spatial patterns. These findings could offer a theoretical reference for the formulation of policies to enhance the synergistic development of ecosystems and socioeconomic systems in the YRB.

1. Introduction

Environmental protection during socioeconomic development is crucial for achieving sustainable development [1]. Throughout history, human settlements have consistently been established near water bodies, and river basins have played a vital role in human survival, socioeconomic development, and the preservation of ecological functions [2,3]. As essential components of the Earth’s surface system, river basins serve as carriers of critical natural resources, including water, energy, and food, and represent the geographical units where natural processes and human activities interact most intensely [4,5]. Therefore, the coordinated development of ecosystems and socio-economic systems within river basins is critical for realizing the United Nations 2030 Sustainable Development Goals [6,7]. However, the spatial heterogeneity and regional disparities intensified by current and projected climate change impacts continue to pose substantial obstacles to achieving these goals [8,9].

In recent years, extensive research has been conducted on the interrelationships between river basin ecosystems and socioeconomic development, as well as on the internal interactions among various ecological services within river systems [10,11,12,13]. Nevertheless, major river basins are inherently complex geographic entities that can be distinguished by pronounced spatial differentiation in both natural and anthropogenic elements. From source to estuary, marked heterogeneity exists in ecosystem types and socioeconomic development levels, forming critical constraints on integrated basin management [14,15]. Consequently, effective zoning strategies are essential to advance sustainable development practices in river basins.

An accurate understanding of the spatiotemporal consistency between ecosystem service functions and socioeconomic development levels is essential for implementing effective basin-wide system governance. In recent years, researchers have increasingly focused on the spatial heterogeneity of trade-offs and synergies among ecosystem service functions within river basins, including the dynamic evolution of these interactions in the upper reaches of the Blue Nile River [11]. Within the Yellow River Basin (YRB), the spatial differentiation between ecological systems and human geographical elements is particularly pronounced and highly correlated [16,17]. Investigating the interactions between human activities and natural resources in the YRB presents considerable challenges, primarily due to the basin’s vast spatial extent, long-term and intense anthropogenic disturbances, and significant internal variability [18].

Extensive research has been devoted to understanding ecosystem services and their coupling with socioeconomic drivers in the YRB [19,20,21,22,23]. For instance, studies have examined how ecological patterns, structures, and functions within the YRB respond to drought events [24]. Other investigations have analyzed the synergies and trade-offs among multiple ecosystem services, including carbon storage, water yield, habitat quality, and soil retention [22]. The spatiotemporal heterogeneity of relationships between ecosystem service functions at multiple spatial scales (300 km², 1000 km², 3000 km², and sub-watersheds) has also been explored. The results indicate that the synergistic benefits of ecosystem services, such as habitat quality, carbon storage, soil conservation, food production, water yield, and landscape aesthetics, generally outweigh the associated trade-off effects in the YRB [25]. Furthermore, recent research has investigated the coupling between ecosystem service functions and socioeconomic factors at the sub-watershed scale [23].

However, according to current research, most regional studies have concentrated on specific sub-areas within the YRB, often simplifying the basin into upper, middle, and lower reaches or focusing on the nine provincial-level administrative divisions encompassed by the basin [23,26,27,28]. Several provinces only partially intersect with the YRB, rendering such studies insufficient for fully capturing the basin’s status from the perspective of natural systems. Consequently, the spatiotemporal heterogeneity of the relationship between ecosystem services and socioeconomic development within the YRB remains poorly understood. Further research is required to comprehensively and systematically evaluate the spatiotemporal coordination between ecosystem service functions and socioeconomic development across the YRB by utilizing multiple indicators.

River basin water resource zoning refers to the classification of basins into different functional zones by integrating administrative boundaries with variations in water resources and socioeconomic conditions. Regarding such zoning within the YRB, prior studies have investigated the impact of land-use changes induced by human activity [29] and explored the spatiotemporal patterns of extreme precipitation events [30]. However, these studies are far from sufficient to meet the demands of zoning management for the coordinated development of ecosystem services and socio-economic progress in the YRB. In this context, this study conducted a regional-scale analysis of the YRB, based on secondary water resource zones. A five-year interval was adopted to examine the spatial distribution and temporal evolution of ecosystem service values and socioeconomic systems (including population, GDP, nighttime light data, and HDI) across the basin from 1980 to 2020. A consistency model was introduced to assess the spatiotemporal alignment between ecosystem service values and socioeconomic development across different water resource zones. This study offers a spatially explicit assessment of the relationships between ecosystem service values and socio-economic factors across the YRB, advancing place-based insights for sustainable basin management. Additionally, this study incorporated changes in the waterbody and sandy land area to further investigate the influencing factors affecting ecosystem and socio-economic coordination.

2. Study Area

As illustrated in Figure 1, the YRB extends across the three major geomorphic terraces of eastern, central, and western China. The upper and middle reaches of the basin are subject to severe desertification and function as the primary sediment source regions. The basin spans multiple hydroclimatic zones and encompasses a variety of ecological systems and highly uneven levels of socioeconomic development. Thus, it is considered a typical region exhibiting ecological fragility and high sensitivity to climate change [24,31].

In this study, the water resource zoning of the YRB follows China’s secondary water resource zoning framework. China’s water resource zoning system is delineated based on hydrological characteristics and natural geographical conditions, while also considering the integrity of river basins and administrative boundaries. This system comprises three hierarchical levels: (a) primary zoning is defined by the natural distribution of major rivers; (b) secondary zoning further subdivides areas based on similarities in water resource conditions; (c) tertiary zoning focuses on water quantity calculation and supply–demand balance analysis.

According to the Yellow River Water Resources Bulletin, the YRB is divided into eight secondary water resource zones: upstream of Longyangxia; Longyangxia–Lanzhou; Lanzhou–Hekou Town; Hekou Town–Longmen; Longmen–Sanmenxia; Sanmenxia–Huayuankou; downstream of Huayuankou; and the Endorheic region. As summarized in

Table 1. Water and sediment in the secondary water resource zone of the Yellow River Basin.

Code	Secondary Water Resource Zone	Area /104 km2	Mean Elevation/m	Total Water Resources /108 m3	Water Consumption/108 m3				Sediment Yield /108 t
					Agriculture	Industry	Domestic	Ecological
I	Upstream of Longyangxia	13.20	4064.46	209.3	1.31	0.04	0.36	0.06	0.075
II	Longyangxia–Lanzhou	9.10	3058.02	134.4	16.99	2.39	4.82	3.77	0.172
III	Lanzhou–Hekou Town	15.32	1465.12	40.4	127.88	12.25	10.12	25.59	0.049
IV	Hekou Town- –Longmen	12.24	1217.42	62.8	8.06	6.08	4.58	2.40	1.414
V	Longmen–Sanmenxia	19.08	1266.50	160.3	63.78	12.12	24.34	7.77	0.890
VI	Sanmenxia–Huayuankou	4.17	773.30	63.1	11.89	5.98	6.32	3.38	−1.050
VII	Downstream of Huayuankou	2.24	131.06	37.9	28.98	3.49	5.24	2.44	−0.300
VIII	Endorheic region	4.23	1364.23	11.4	5.90	0.39	0.24	0.44	-

Note: Total precipitation and total water resources were obtained from data of the second National Water Resources Survey and Evaluation in the YRB (2012–2015) (https://www.ncdc.ac.cn/portal/metadata/2f4c5fc1-5a44-4561-9356-f3eb14ccd729, accessed on 2 April 2025). The water consumption data were obtained from the Yellow River Water Resources Bulletin in 2022, released by the Yellow River Conservancy Commission of the Ministry of Water Resources (YRCC, MWR). The sediment yield data were obtained from the Yellow River Sediment Bulletin in 2022, released by the YRCC, MWR.

, these zones exhibit pronounced spatial heterogeneity in land surface features, channel morphology, and water–sediment conditions [32,33,34,35].

3. Methodology

3.1. Data Collection and Processing

Multi-source datasets encompassing the spatial and statistical attributes of the YRB were integrated in this study. When multiple sources were available, datasets explicitly tailored for the YRB or China were prioritized, with the preference given to those offering longer time series. The data employed in this study and their respective sources are outlined as follows:

(i) Population, Gross Domestic Product (GDP), nighttime lighting index (NTL), and Human Development Index (HDI) were adopted to characterize socioeconomic development. Population data (1990–2015) were obtained from Wang and Wang (2022) [36]; GDP data (1985–2019) were sourced from Zhao et al. (2017) [37]; NTL data (1984–2020) were referenced from Lixian Zhang et al. (2024) [38]; HDI data (1990–2015) were cited from Kummu et al. (2018) [39], incorporating metrics such as life expectancy, education level, and quality of life.

(ii) The land use data (1980–2020) were sourced from the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (RESDC) (http://www.resdc.cn, accessed on 12 November 2024).

ESV serves as a comprehensive metric for assessing ecosystem functions, typically quantified using land use data. Population and GDP represent fundamental indicators of social and economic development, respectively. NTL data provide a robust proxy for human activity intensity, while the HDI offers a multidimensional measure of societal progress. These well-established indicators have gained broad acceptance in both academic research and management applications, demonstrating strong representativeness and reliable data accessibility.

3.2. ESV Assessment Model

The equivalent factor method, which is a system of standardized coefficients for ESV estimation developed by Xie et al. (2008) [40], was adapted to the actual conditions in China. The net profit of food production per unit area of farmland was taken as the baseline, representing an ESV of one standard equivalent factor. This approach has been widely employed in both basin-scale and global-scale ESV assessments [41,42].

The economic value of the ecosystem service equivalent factor per unit area is generally assumed to equal 1/7 of the economic value of food production per unit area, calculated as follows:

$$E={\displaystyle \frac{1}{7}}{\sum }_{f=1}^n{\displaystyle \frac{d_f{p}_f{q}_f}{D}}(f=1,2,\cdots ,n)$$

(1)

where $E$ is the price of food production per unit area of arable land in each region (yuan/hm⁻²), $p_f$ is the average price of type f crops (yuan/t⁻¹), $q_f$ is the yield per unit area of type f crops (kg/hm⁻²), $d_f$ is the acreage of type f crops (hm²), and $D$ is the total acreage of all crops (hm²).

To account for spatial heterogeneity, Yang et al. (2023) [23] incorporated indicators, such as NPP, rainfall, and soil conservation, to adjust the equivalent factors specific to the YRB. The ESV per hectare for different land-use types in the YRB can be referenced in Reference [23].

The value of ecosystem services was calculated as follows [23,42,43]:

$${ESV}_k={\sum }_{f=1}^m{A}_k\times {VC}_{kf}$$

(2)

$${ESV}_f={\sum }_{f=1}^n{A}_k\times {VC}_{kf}$$

(3)

$$ESV={\sum }_{f=1}^m{\sum }_{k=1}^n{A}_k\times {VC}_{kf}$$

(4)

where ${ESV}_k$ represents the ecosystem service value of the k-th land use, ${ESV}_f$ represents the ecosystem service value of the f-th function, $ESV$ represents the total ecosystem service value, $A_k$ represents the area of the k-th land use, ${VC}_{kf}$ represents the equivalent value coefficient (yuan ha year⁻¹) of the k-th land use and f-th ecosystem service functions, where the specific details of ${VC}_{kf}$ can be seen in Reference [23], m represents the land use type, and n represents the ecosystem service type.

3.3. Consistent Mode of Ecosystem Services and Socioeconomic Development

By calculating the coefficient of variation (CV) of ESV and GDP, the deviation of ecosystem services from the level of economic development was quantified. Accordingly, a consistency model of ecosystem services and economic development was established to evaluate whether coordination existed between ESV and GDP. The method used to assess the consistency between ecosystem services and other socioeconomic indicators (population, NTL, and HDI) followed the same approach as the model for ESV and GDP.

First, the coefficient of variation representing the degree of dispersion and concentration of GDP and ESV was independently computed using the following formula:

$$CV={\displaystyle \frac{1}{Y_0}}\sqrt{{\displaystyle \frac{{\sum }_{i=1}^n{(Y_i-Y_0)}^2}{n}}}$$

(5)

where $Y_0$ is the basin average of ESV or GDP; $n$ is the number of study units; and $Y_i$ is the value of ESV or GDP of the i-th study unit. A lower $CV$ value indicates smaller differences in GDP or ESV across study units. A coefficient of variation greater than 1 indicates a high degree of dispersion in the data, i.e., the standard deviation of the data is greater than the mean.

Second, a consistency model was developed to quantify the degree of coordination between ecosystem services and socioeconomic development (CEE) across watershed regions [43]. The formula used for this calculation is as follows:

$${CEE}_i={\displaystyle \frac{R_{{ESV}_i}}{R_{{GDP}_i}}}={\displaystyle \frac{{ESV}_i/{\sum }_{i=1}^n{ESV}_i}{{GDP}_i/{\sum }_{i=1}^n{GDP}_i}}$$

(6)

where ${CEE}_i$ is the consistency index between ecosystem services and economic development of the i-th unit; $R_{{ESV}_i}$ is the ecosystem service agglomeration level of the i-th unit; and $R_{{GDP}_i}$ is the socioeconomic agglomeration level of the i-th unit.

As shown in

Table 2. Classification of consistency between ecosystem services and socioeconomic development.

Criteria	Major Categories	Subcategories
CEE ≤ 0.50	Socioeconomic agglomeration is higher than ecosystem service agglomeration.	Socioeconomic agglomeration is much higher than ESV agglomeration.
0.50 < CEE < 0.80		Socioeconomic agglomeration is higher than ESV agglomeration.
0.80 ≤ CEE < 1.20, and CEE ≠ 1	Ecosystem services are largely harmonized with socioeconomic agglomeration.	ESVs are basically harmonized with socioeconomic agglomeration.
CEE = 1	Ecosystem services are fully harmonized with socioeconomic agglomeration.	ESVs are fully harmonized with socioeconomic agglomeration.
1.20 ≤ CEE < 2.0	Clustering of ecosystem services over socioeconomic clustering.	ESV agglomeration is higher than socioeconomic agglomeration.
CEE ≥ 2.0		ESV agglomeration is much higher than socioeconomic agglomeration.

, ${CEE}_i>1$ indicates that the concentration of ecosystem services exceeds that of economic development, whereas ${CEE}_i<1$ suggests an ecosystem service concentration lower than the economic concentration. When ${CEE}_i$ approaches 1, this reflects a balanced coordination between ecosystem services and economic development within the region. Additionally, the coordination level was classified into four primary categories and six subcategories [43].

4. Results

4.1. Dynamic Changes in Ecosystem Service Value

As shown in Figure 2 and

Table 3. ESV changes of eight secondary water resource zones in the Yellow River Basin from 1980 to 2020.

Code	ESV (109 Yuan)					Change Rate (%)
	1980	1990	2000	2010	2020	1980–1990	1990–2000	2000–2010	2010–2020	1980–2020
I	161.81	162.71	162.77	171.05	171.77	0.55 ↑	0.04 ↑	5.09 ↑	0.42 ↑	6.15 ↑
II	134.11	134.11	134.18	136.15	137.30	−0.01 ↓	0.05 ↑	1.47 ↑	0.84 ↑	2.37 ↑
III	201.92	199.24	199.32	200.21	213.20	−1.33 ↓	0.04 ↑	0.44 ↑	6.49 ↑	5.59 ↑
IV	141.36	141.40	143.01	148.77	152.06	0.03 ↑	1.13 ↑	4.03 ↑	2.21 ↑	7.57 ↑
V	280.28	280.05	281.50	292.14	296.51	−0.08 ↓	0.52 ↑	3.78 ↑	1.50 ↑	5.79 ↑
VI	77.28	77.27	77.21	81.13	82.46	−0.01 ↓	−0.08 ↓	5.08 ↑	1.63 ↑	6.70 ↑
VII	40.72	38.85	37.89	41.41	42.95	−4.59 ↓	−2.49 ↓	9.29 ↑	3.72 ↑	5.47 ↑
VIII	42.35	42.21	42.25	42.22	42.96	−0.32 ↓	0.09 ↑	−0.07 ↓	1.75 ↑	1.44 ↑
Total	1079.83	1075.84	1078.12	1113.08	1139.20	−0.37 ↓	0.21 ↑	3.24 ↑	2.35 ↑	5.50 ↑

Note: “↑” indicates an increase, while “↓” indicates a decrease.

, the overall spatial pattern of ESV in the YRB exhibited no significant variation from 1980 to 2020. The ESV in all water resource zones upstream of Sanmenxia consistently exceeded 100 × 10⁹ yuan, significantly surpassing that below Sanmenxia. During this period, the total ESV of the basin increased from 1079.83 × 10⁹ to 1139.20 × 10⁹ yuan. All eight water resource zones showed an upward trend, with growth rates of 6.15% (I), 2.37% (II), 5.59% (III), 7.57% (IV), 5.79% (V), 6.70% (VI), 5.47% (VII), and 1.44% (VIII), respectively.

A comparative analysis across partitions indicated that the Longmen–Sanmenxia section (V) exhibited the highest average ESV from 1980 to 2020, reaching 286.10 × 10⁹ yuan. In contrast, the lowest ESV was observed in the downstream Huayuankou zone (VII), with an average of only 40.36 × 10⁹ yuan. The remaining six partitions exhibited descending ESV magnitudes in the following order: Lanzhou–Hekou Town section (III) with 202.78 × 10⁹ yuan, above-Longyangxia zone (I) with 166.02 × 10⁹ yuan, Hekou Town–Longmen section (IV) with 145.32 × 10⁹ yuan, Longyangxia–Lanzhou zone (II) with 135.17 × 10⁹ yuan, Sanmenxia–Huayuankou section (VI) with 79.07 × 10⁹ yuan, and the Endorheic region (VIII) with 42.40 × 10⁹ yuan.

Between 1980 and 1990, the YRB experienced an overall decline in ESV, with the most notable reduction (4.59%) observed in downstream of Huayuankou (VIII). Only two zones recorded slight increases during this period: the above-Longyangxia zone (I) with +0.55% and the Hekou Town–Longmen section (IV) with +0.03%. After 1990, all eight partitions generally demonstrated increasing ESV trends, although exceptions were identified in some decades. Specifically, the Sanmenxia–Huayuankou section (VI) declined by 0.08% between 1990 and 2000, the downstream Huayuankou section (VII) decreased by 2.49% during 1990–2000, and the Endorheic region (VIII) declined by 0.07% from 2000 to 2010.

4.2. Changes in Socioeconomic Factors in the Yellow River Basin

4.2.1. Population

• Spatial distribution pattern

With respect to spatial distribution, the population density across water resource partitions exhibited a progressively increasing trend from upstream to downstream areas (Figure 3). In terms of temporal dynamics, the population density from 1990 to 2015 in the above-Longyangxia zone (I) increased by a factor of 1.66, whereas the below-Huayuankou zone (VII) experienced a 1.12-fold increase. The average population density in zone VII remained substantially higher than that in zone I, with the ratio decreasing from 194:1 in 1990 to 131:1 in 2015.

In terms of total population distribution, a first-increasing-then-decreasing trend was observed from upstream to downstream. The Longmen–Sanmenxia section (V) had the largest total population. During 1990–2015, the average total population in section (V) was 91 times that in the above-Longyangxia zone (I) and 3 times that in the below-Huayuankou zone (VII). Regarding the variation in spatial differences, the population multiple in the Longmen–Sanmenxia (V) section compared to the above-Longyangxia zone (I) decreased from 106 to 82 times over the same period, whereas the multiple relative to the below-Huayuankou zone (VII) remained relatively stable, at approximately three times.

In terms of standard deviation, the population distribution exhibited a spatial zoning pattern that initially increased, then decreased, and increased again along the river. The highest standard deviations were observed in the Longmen–Sanmenxia (V) and Sanmenxia–Huayuankou (VI) sections, with values consistently exceeding 460 between 1990 and 2015. Moderate standard deviations, averaging approximately 260 were recorded in the Longyangxia–Lanzhou (II), Lanzhou–Hekou Town (III), and downstream of Huayuankou (VII) sections. In contrast, lower values (below 100) were recorded in the upstream Longyangxia zone (I), Hekou Town–Longmen section (IV), and Endorheic region (VIII).

2.

Temporal variation trends

From 1990 to 2015, the population density and total population in the water resource zones exhibited a fluctuating, yet overall upward trend (Figure 3). The population growth rates for the eight zones during this period were 1.66, 1.18, 1.49, 1.22, 1.29, 1.20, 1.12, and 1.15 times, respectively. The standard deviations of population across all eight zones increased significantly from 1990 to 2015, indicating a notable intensification in the spatial concentration of population distribution. This trend reflects the inevitable consequences of socioeconomic development and urbanization.

4.2.2. GDP

• Spatial distribution pattern

The spatial distribution pattern of GDP per unit area in the water resource zones revealed an overall increasing trend from upstream to downstream (Figure 4). Between 1985 and 2019, the average GDP per unit area in downstream of Huayuankou (VII) was approximately 14 times higher than that upstream of Longyangxia (I). In terms of spatial disparities, GDP per unit area in the upstream zone (I) increased by 10.58 times during this period, whereas in the downstream zone (VII), it increased by 16.49 times, rising from 10 times in 1985 to 16 times in 2019.

Regarding total GDP, the spatial pattern along the basin exhibited a trend of first increasing and then decreasing. The GDP in the Longmen–Sanmenxia section (V) was the highest. From 1985 to 2019, the total GDP in section V was 14 times that in section I and 6 times that in section VII. Regarding the spatial distribution pattern of total GDP, from 1985 to 2019, the ratio of the Longmen–Sanmenxia (V) section to the upstream of Longyangxia (I) increased from 10 to 16. The ratio of total GDP in section V to that in the downstream of Huayuankou (VII) remained stable at approximately 5, and the ratio of section VII to section I increased from 2 to 3.

A spatially increasing trend was observed along the river with respect to the standard deviation of GDP. The lowest standard deviation occurred in the upstream region (I), with an average value of 0.294 during 1985–2019, whereas the highest was recorded in the downstream region (VII), with an average of 6.278 over the same period. The standard deviation in section VII was nearly 20 times greater than that in section I, indicating a gradual intensification of spatial agglomeration of economic development from the source to the mouth of the river.

2.

Temporal variation trends

The GDP per unit area and the total GDP of water resource regions exhibited a pronounced upward trend (Figure 4). From 1985 to 2019, the growth multiples of GDP for the eight water resource zones were 10.58, 14.83, 16.01, 19.38, 16.01, 14.98, 16.49, and 15.05, respectively. During the same period, the standard deviation of GDP across these zones increased by more than 15 times, demonstrating a substantial intensification in the spatial concentration of GDP distribution, which was consistent with the general principles of economic development.

4.2.3. NTL

• Spatial distribution pattern

Regarding the spatial distribution pattern of NTL per unit area, an overall increasing trend from upstream to downstream was observed across water resource zones (Figure 5). From 1985 to 2020, the average NTL per unit area in the downstream of Huayuankou (VII) was 143 times greater than that in the upstream of Longyangxia (I). During this period, the NTL per unit area increased by 13.32 times in the upstream section (I) and by 2.87 times in the downstream section (VII). The gap between the two sections narrowed considerably, decreasing from 328 times in 1985 to 70 times in 2020.

In terms of the spatial distribution of the total NTL, a trend of initial increase followed by a decline was identified from upstream to downstream. The highest total NTL was recorded in the Longmen–Sanmenxia section (V). Between 1985 and 2019, the average total NTL in this section was 101 times that of the upstream Longyangxia section (I) and four times that of the downstream Huayuankou section (VII). From 1985 to 2020, the regional gaps in the total NTL volume narrowed significantly. Specifically, the ratio of the total NTL in the Longmen–Sanmenxia section (V) to that in the upstream Longyangxia section (I) decreased from 232 to 51 times. The ratio between the Longmen–Sanmenxia (V) and downstream Huayuankou (VII) sections remained approximately four times. Meanwhile, the ratio between downstream Huayuankou (VII) and upstream Longyangxia (I) decreased from 58 times to 13 times.

Regarding the spatial distribution of the standard deviation of the NTL, a continuous upward trend was observed along the river. The lowest standard deviations were recorded upstream of Longyangxia (I), averaging 57 from 1985 to 2020. The highest standard deviation occurred between Sanmenxia and Huayuankou (VI), with an average of 781. From 1985 to 2020, the average standard deviation of the NTL of downstream of Huayuankou (VII) was approximately 13 times that of the upstream section at Longyangxia (I). These results reflect a gradual enhancement in the spatial concentration of socioeconomic development from the river’s source to its mouth.

2.

Temporal variation trends

Both the unit-area NTL and total NTL in the water resource zones exhibited a significant upward trend (Figure 5). From 1985 to 2020, the growth multiples of NTL across the eight water resource zones were 13.32, 6.43, 4.30, 7.51, 2.93, 2.48, 2.87, and 23.71, respectively. During this period, the total NTL in upstream of Longmen (exceeding fourfold) increased at a notably faster rate than those downstream of Longmen (less than threefold). Furthermore, the standard deviation of NTL across all eight water resource zones increased by more than 1 times from 1985 to 2020, indicating a substantial enhancement in the spatial concentration of socioeconomic activity, in line with the general patterns of economic development.

4.2.4. HDI

From 1990 to 2015, no significant changes were observed in the spatial distribution pattern or temporal trends of HDI (Figure 6). The average HDIs per unit area across the eight water resource regions during this period were 0.572, 0.568, 0.620, 0.625, 0.602, 0.611, 0.628, and 0.640, respectively. Over this period, the HDI in the Longyangxia upstream (I) and Longyangxia–Lanzhou (II) regions transitioned from low to medium levels of human development. In contrast, the HDIs in the Lanzhou–Hekou Town (III), Hekou Town–Longmen (IV), Longmen–Sanmenxia (V), Sanmenxia–Huayuankou (VI), and Huayuankou downstream (VII) regions underwent shifts from low to high levels of human development.

4.3. Spatiotemporal Coordination Between Ecosystem Services and Socioeconomic Development

As shown in the

Table 4. Variation coefficient of ESV, POP, GDP, NTL, and HDI in the Yellow River Basin.

Variation Coefficient	1980	1990	1995	2000	2005	2010	2015	2020
ESV	0.6080	0.6103	0.6202	0.6132	0.6149	0.6109	0.6105	0.6117
POP		1.1039	1.1037	1.1263	1.1188	1.1086	1.1264
GDP	0.9668	0.9668	0.9884	0.9861	0.9849	0.9887	0.9826	0.9816
NTL	1.1621	1.1420	1.1292	1.0857	1.0562	1.0381	0.9620	0.9424
HDI		0.6134	0.6134	0.6134	0.6134	0.6134	0.6134

, the coefficients of variation for population, GDP, and NTL were substantially higher than those for ecosystem services and HDI. Owing to the influence of regional location and market mechanisms, the variation coefficients of socio-economic indicators such as population, GDP, and nighttime light index remained relatively high, reflecting the pronounced spatial distribution differentiation across water resource zones. Given China’s generally high standards of basic healthcare and education, the coefficients of variation for HDI among the water resource zones were relatively low.

From the perspective of coordination between ESVs and population, GDP, and NTL, the YRB displayed a marked imbalance between ESVs and socioeconomic development, whereas the overall spatial pattern remained relatively stable (Figure 7). In the upper reaches (I and II), the dominant pattern was characterized by ecosystem service aggregation being significantly higher than socio-economic (GDP, NTL, and POP) aggregation. In the middle reaches (V and VI), socioeconomic (GDP, POP, and NTL) aggregation exceeded that of ecosystem services. In the downstream region (VII), socio-economic (POP and NTL) aggregation was markedly higher than ecosystem service aggregation. Water resource zones that exhibited coordination between ecosystem services and economic aggregation were primarily concentrated in the central section of the basin (III and IV).

5. Discussion

5.1. Potential Impact of Changes in Water Body and Sandy Land Area on the Relationship Between ESVs and Socio-Economic Development

There are complex and close interrelationships among socioeconomic factors, ecological environments, and water–sediment dynamics in the YRB [44,45,46]. Ecological restoration measures have been shown to improve environmental conditions, while simultaneously affecting synergistic and trade-off dynamics between water and sediment [47]. Water scarcity remains a critical factor that contributes to the ecological vulnerability of the YRB [48]. Land use and land cover changes are profoundly influenced by socioeconomic development, and ecosystem conditions are closely associated with land cover types, with ecological fragility largely shaped by land transformation [49]. Previous studies have examined land-use changes across the upper, middle, and lower reaches of the YRB [50]. Among the various land use categories, water bodies exhibited the highest ecosystem service value, whereas sandy areas demonstrated the lowest. The spatial distribution and temporal variation in surface water and sandy areas exert a significant influence on the coordination between ecosystem services and socioeconomic development within the basin’s water resource zones [43]. Accordingly, this study investigated the spatiotemporal characteristics of surface water bodies and sandy areas across eight water resource zones in the YRB from 1980 to 2020. By integrating the spatial and temporal distributions of water and sandy areas, coordination patterns between ESVs and socioeconomic development were elucidated at the scale of water resource zoning.

Significant spatial differences in water area and dominant types of water bodies were observed across the water resource zones in the YRB (Figure 8). From 1980 to 2020, the average water area ranked from largest to smallest was as follows: Lanzhou–Hekou Town section (38.69 × 10⁴ hm²), upstream of Longyangxia (29.78 × 10⁴ hm²), Longmen–Sanmenxia section (18.56 × 10⁴ hm²), downstream of Huayuankou (13.68 × 10⁴ hm²), Hekou Town–Longmen (11.51 × 10⁴ hm²), Longyangxia–Lanzhou section (10.42 × 10⁴ hm²), Sanmenxia–Huayuankou section (8.17 × 10⁴ hm²), and the Endorheic region (7.01 × 10⁴ hm²). With regard to water body composition, the Lanzhou–Hekou Town, Longmen–Sanmenxia, Longyangxia–Lanzhou, and Sanmenxia–Huayuankou sections exhibited the highest proportions of floodplains, followed by rivers and canals. The sections downstream of Huayuankou and Hekou Town–Longmen had the highest share of rivers and canals, followed by floodplains. The area upstream of Longyangxia had the highest proportion of lakes, followed by floodplains, whereas the Endorheic region was dominated by floodplains and lakes.

The average sandy land area from 1980 to 2020 in descending order was as follows: Lanzhou–Hekou Town section (125.75 × 10⁴ hm²), Endorheic region (110.12 × 10⁴ hm²), Hekou Town–Longmen section (73.28 × 10⁴ hm²), upstream of Longyangxia (30.66 × 10⁴ hm²), Longmen–Sanmenxia section (0.77 × 10⁴ hm²), Longyangxia–Lanzhou section (0.46 × 10⁴ hm²), downstream of Huayuankou (0.24 × 10⁴ hm²), and Sanmenxia–Huayuankou section (0.01 × 10⁴ hm²).

From 1980 to 2020, the spatial distribution pattern of water areas in the YRB remained relatively stable across water resource subregions (

Table 5. Water area changes in the Yellow River Basin (Unit: ×10⁴ hm²).

Code	Annual Time Series								Mean
	1980	1990	1995	2000	2005	2010	2015	2020	1980–2020	1980–2000	2005–2020
I	26.79	28.71	27.00	28.93	29.07	32.44	32.48	32.79	29.78	27.86	31.69
II	9.92	10.04	10.59	10.19	10.43	10.50	10.67	11.01	10.42	10.19	10.65
III	44.20	39.07	31.91	39.02	39.08	37.33	37.89	40.98	38.69	38.55	38.82
IV	11.88	11.80	12.12	11.78	11.53	10.87	10.79	11.29	11.51	11.89	11.12
V	20.57	18.87	20.19	18.19	18.91	16.62	16.87	18.22	18.56	19.46	17.66
VI	8.66	8.41	6.54	6.74	8.73	8.45	8.74	9.11	8.17	7.59	8.75
VII	19.07	14.38	11.68	10.84	12.01	13.33	13.66	14.50	13.68	13.99	13.38
VIII	8.12	7.87	5.69	7.72	7.06	6.25	6.09	7.25	7.01	7.35	6.66
Total	149.21	139.14	125.72	133.41	136.82	135.79	137.19	145.16	137.81	136.87	138.74

). The water areas were predominantly concentrated in the subregions upstream of Hekou Town (above Longyangxia, Longyangxia–Lanzhou, and Lanzhou–Hekou Town), collectively accounting for 58.40% of the basin’s total water area.

During the same period, the total water area of the basin first decreased and then increased, resulting in a net decrease of 2.71%. The spatial trends in change in the water area varied across regions. Notable increases were observed in the two upstream subregions above Lanzhou, with the Longyangxia region increasing by 22.40% and the Longyangxia–Lanzhou region by 10.99%. In contrast, the downstream regions experienced initial decreases followed by slight increases, whereas the overall trend remained decreasing. The subregion downstream of Huayuankou recorded the most significant decline, with a reduction of 23.96%.

Similarly, the spatial distribution of sandy land remained stable and was mainly concentrated in the Jiziwan region (encompassing Lanzhou–Hekou Town, Hekou Town–Longmen, and the Endorheic region), which accounted for more than 90% of the basin’s total sandy area (

Table 6. Sand area changes in the Yellow River Basin (Unit: ×10⁴ hm²).

Code	Annual Time Series								Mean
	1980	1990	1995	2000	2005	2010	2015	2020	1980–2020	1980–2000	2005–2020
I	31.44	30.26	29.95	30.65	37.67	28.50	28.50	28.34	30.66	30.58	30.75
II	0.48	0.47	0.48	0.48	0.47	0.43	0.43	0.43	0.46	0.48	0.44
III	128.50	129.99	140.39	128.24	127.94	119.11	116.89	114.98	125.75	131.78	119.73
IV	82.61	82.47	64.91	72.35	75.48	69.15	67.97	71.33	73.28	75.59	70.98
V	1.13	1.12	0.90	0.80	0.75	0.63	0.34	0.48	0.77	0.99	0.55
VI	0.00	0.00	0.00	0.02	0.02	0.01	0.01	0.01	0.01	0.01	0.01
VII	0.54	0.60	0.14	0.38	0.16	0.02	0.02	0.02	0.24	0.42	0.06
VIII	110.79	110.88	108.61	110.02	114.92	108.71	107.95	109.09	110.12	110.08	110.17
Total	355.50	355.79	345.39	342.95	357.41	326.56	322.11	324.67	341.30	349.91	332.69

).

From 1980 to 2020, the sandy land area declined substantially. Specifically, the area above Longyangxia decreased from 31.44 × 10⁴ to 28.34 × 10⁴ hm² (a 9.86% reduction); Lanzhou–Hekou Town from 128.50 × 10⁴ to 114.98 × 10⁴ hm² (a 10.52% reduction); Hekou Town–Longmen from 82.61 × 10⁴ to 71.33 × 10⁴ hm² (a 13.65% reduction); and the Endorheic region from 110.79 × 10⁴ to 109.09 × 10⁴ hm² (a 1.53% reduction). These reductions were largely attributed to ecological conservation efforts in the source regions and desertification control measures implemented in the Jiziwan region (Lanzhou–Hekou Town, Hekou Town–Longmen, and the Endorheic region) [51,52].

The total sandy land area across the entire basin peaked in 2005, although the timing of this peak varied across different water resource zones. In the upstream Longyangxia region, the peak occurred in 2005, coinciding with the launch of the Sanjiangyuan Nature Reserve Conservation Project. In the Lanzhou–Hekou Town section, the peak was reached in 1995, whereas it occurred in 1980 in the Hekou Town–Longmen section. The Endorheic Region peaked in 2005. Notably, 2000 marked the initiation of several major ecological and policy programs, including the Beijing–Tianjin Sandstorm Source Control Project, Natural Forest Conservation Project, Western Development Strategy, and promulgation of the China Wetland Conservation Action Plan.

Based on the above analysis, the region upstream of Lanzhou accounted for 24.60% of the total basin water area, with 8.98% of the sandy land. This region exhibited low levels of socioeconomic development, while the ecosystem service aggregation significantly exceeded the aggregation levels of socioeconomic indicators (GDP, NTL, and POP), which primarily clustered within the same water resource zone.

The water resource sub-basin between Lanzhou and Huayuankou comprised approximately 62.61% of the total basin area, with a sandy land proportion of 90.87%. Socioeconomic development in this region is at an intermediate level and is characterized by high industrial and agricultural productivity. This section exhibits the most complex relationship between ecosystem service aggregation and economic aggregation, which can be characterized by a transition from the dominance of ecosystem service aggregation over socioeconomic aggregation to the reverse. A fundamental coordination between the two types of aggregation was observed in this region.

The region downstream of Huayuankou represents approximately 12.78% of the total basin area, with only 0.15% sandy land. It had the highest level of socioeconomic development in the basin, marked by the highest average values of population, GDP, NTL, and HDI. This region also has the densest road network. The social and economic aggregation significantly exceeded the ecological and environmental aggregation levels in this area.

5.2. Policy Recommendations

For the upstream water resource zone, the primary objective in the YRB is ecological conservation. Therefore, while ensuring ecosystem health and functionality, and preventing excessive human encroachment, efforts should focus on boosting regional economic development and residents’ livelihoods through measures such as realizing the value of ecosystem services and establishing cross-regional ecological compensation mechanisms.

For the midstream water resource zone—a water-scarce arid region with fragile ecosystems and intermediate socio-economic development levels across the basin—the focus should be on restructuring the ecosystem landscape and adopting innovative industrial models, such as the integrated “solar-forest-medicinal-pastoral” sand control model and the agrophotovoltaic model (“solar power generation on panels, crop cultivation below panels, and livestock farming between panels”).

For the downstream water resource zone, where land availability is limited and human activity intensity is high, significantly expanding ecosystem space is challenging. Priority should be given to restoring degraded natural ecosystems and enhancing ecological quality. Within human-dominated areas, maximizing the ecological service value of artificial systems (e.g., plantation forests and farmland) is essential.

5.3. Limitations and Future Perspectives

Although we have made every effort to accurately conduct the research in this paper, there are still some limitations due to the temporal and spatial matching and availability of the data. The time series data we used are not all within the period of 1980–2020, such as population and HDI. For ecosystems, we only used a comprehensive indicator of ecosystem service value and did not use multiple classification indicators to characterize the status of ecosystems. Furthermore, this study primarily focused on exploring the relationship between ecosystem service value and socio-economic development at the water resource zoning scale in the YRB. However, we did not perform a quantitative analysis of the underlying causes. Based on the results of this study, we speculate that natural factors (such as terrain, precipitation, temperature, and land cover type) are the fundamental drivers of the present spatiotemporal distribution pattern, while human activities (such as land and energy resource allocation and national policy adjustments) play a moderating role. In future research, we will quantify the relative contributions of natural and human factors to this spatial distribution. Additionally, we will conduct predictive studies using ecosystem and socio-economic scenario data under different climate scenarios.

6. Conclusions

This study investigated the spatiotemporal patterns of ecosystem services and socio-economic development at the secondary water resource zone scale in the YRB by integrating multi-source data. We systematically assessed the distribution characteristics of ESVs, population, GDP, NTL, and HDI, revealing their coordinated relationships across different water resource zones. In terms of ecosystem service values, from 1980 to 2020, the spatial pattern of ecosystem service value in the YRB remained relatively stable. The highest value occurred in the Longmen–Sanmenxia section (V), whereas the lowest was found in the region downstream of Huayuankou (VII). In terms of socioeconomic development factors, from upstream to downstream, the population density, GDP per unit area, and NTL per unit area exhibited increasing trends, whereas the total population, total GDP, and total NTL showed patterns of initial increase followed by decline. Regarding the spatiotemporal consistency between ecosystem service value and socio-economic development, the coefficients of variation for population, GDP, and NTL were substantially higher than those for ecosystem services and HDI. This indicated that the ecosystem service values in the YRB exhibited an inconsistent relationship with socioeconomic development, although the overall spatial pattern remained relatively stable.