Spatial Coupling Characteristics and Factors Influencing Soil–Vegetation Relationships in the Lower Part of the Shiyang River Basin

Abstract

The relationship between soil and vegetation is an essential scientific issue in surface environment change. (1) Background: Since the implementation of the Shiyang River Basin governance plan, it has become necessary to quantitatively evaluate the impact of ecological restoration on soil–vegetation spatial coupling. (2) Methods: A coupled model and a coupled coordination model are adopted in order to investigate the spatial coupling characteristics of soil–vegetation systems. Additionally, we explore the influences of climate factors and soil properties on the level of spatial coupling and coordination. (3) Results: From 2015 to 2020, the soil–vegetation spatial coupling coordination in the lower reaches of the Shiyang River Basin was poor, and the average annual proportion of areas with medium and low degrees of uncoordination reached 79.3%. The level of spatial coupling coordination is differed under different vegetation coverage scenarios, and the bare land mainly showed low and moderate imbalances, accounting for 90.3% of the annual average area. The annual average proportions of short coverage and canopy coverage coordinated areas were 53.4% and 49.3%, respectively. In particular, vegetation in the Minqin hinterland is highly sensitive to environmental changes. With the implementation of ecological water conveyance, the spatial coupling coordination between soil and vegetation has improved slightly; however, the effect is not obvious. (4) Conclusions: Precipitation, temperature, and potential evaporation affect the level of coupling coordination between soil and vegetation, with the former having a positive effect and the latter two having negative effects. In addition, soil enriched with sulfate and sand contributed to the disharmony of soil–vegetation relationships in the study area.

1. Introduction

Soil is an important part of the Earth’s surface system, providing the water and nutrients necessary for the growth and development of plants [1], as well as serving as the leading site for many biological, physical, and chemical processes [2]. Soil moisture, soil salinization, and vegetation are critical elements of soil–vegetation systems and are important environmental factors in arid and semi-arid areas, reflecting regional ecological conditions. In arid and semi-arid regions, soil moisture is a crucial factor affecting vegetation growth. Soil salinization poses a serious threat to crop growth and plant survival, resulting in reduced agricultural yields and vegetation cover [3]. The ecological imbalance between soil moisture, soil salinization, and vegetation has become the main reason for long-term environmental deterioration in arid areas. The correlation between soil and vegetation is an important scientific problem regarding the process of surface environment change.

The soil–vegetation system is an open system relevant to a wide range of fields, including geophysics, biology, geology, agronomy, forestry, meteorology, environmental science, physical geography, surveying, and mapping, and there has been a significant amount of research on the soil–vegetation system related to ecosystems, geographic information systems, soil erosion, vegetation restoration, soil moisture, and ecological restoration [4,5,6,7,8]. There are various research topics and certain research foundations within soil–vegetation systems and related disciplines [9,10,11]; however, research on the coupling relationship of soil–vegetation systems in arid and semi-arid areas is still at the stage of exploring the temporal response mechanism of the biophysical elements using experimental data. Li et al. [12] have reviewed the research progress and methods of soil–vegetation hydrology at different scales. The phenomena and laws relating to soil–vegetation hydrology observed at different scales are diverse [13]. Therefore, in this paper, we aim to explore the coupling characteristics of soil–vegetation systems using remote sensing data at the soil surface-cover scale, in order to expound the coupling coordination levels of the soil–vegetation system from the perspective of the vegetation’s demand for soil, water and nutrients, which is of great significance for ecological management in arid and semi-arid areas.

Coupling implies that two or more elements are closely related and influence each other [14,15,16], and it is used as a measure of the interdependence of two or more factors. Coupling coordination refers to the measurement of the level of coordination development between two or more elements with coupling relationships that are undergoing mutually influential processes. The coupling strength and process coordination between elements can be defined by exploring the coupling relationship and the level of coupling coordination between them. Studies on spatial coupling between elements, covering diverse areas, including human economic and social development and ecosystem services, have been carried out extensively [17,18,19,20,21,22]. These studies have demonstrated that spatial coupling relationships can effectively represent the interactions between factors.

The Shiyang River basin is a typical arid inland river basin with a unique geographical location and a perennially arid climate. Severe soil water scarcity and desertification in the lower reaches pose a major threat to ecological security. Therefore, the purpose of this study is to: (1) investigate the coupling characteristics of the soil–vegetation system using a coupling model and a coupling coordination model; (2) analyze the characteristics of soil–vegetation coupling coordination under different vegetation cover conditions; and (3) explore the effects of climate factors and soil properties on the coupling coordination level. This study provides a reference for environmental improvement and evaluation in arid and semi-arid areas.

2. Research Data and Methods

2.1. Study Area

This study was conducted in Minqin Oasis in the lower part of the Shiyang River Basin (Figure 1), a typical inland arid area [23], with an altitude of 1300–1500 m and annual precipitation of less than 150 mm [24]. It is surrounded by the Tengger Desert to the northeast and the Badain Jaran Desert to the northwest [24]. In the northern part, near the edge of the Tengger Desert, the yearly precipitation is 50 mm, the potential annual evaporation is 2000–2600 mm, and the drought index is 15–25 [25]. Arid climatic conditions and frequent human activities have resulted in a decrease in the natural groundwater supply, characterized by a decline in regional groundwater levels [24], increased salinity, the wilting and death of large areas of natural vegetation, and the acceleration of land desertification and salinization [25]. The highly deteriorating ecological environment in the Shiyang River Basin has attracted significant attention from the Chinese government [26,27]. As the approval period of the planning revision aimed to contain the ecological deterioration of the basin as soon as possible, the critical management emergency project of the Shiyang River Basin was initiated in advance from 2006 to 2007 [28]. It has been more than ten years since a series of critical governance measures were implemented in the Shiyang River Basin. Among them, ecological water conveyance and agricultural water-saving irrigation reconstruction projects in the lower part of the Shiyang River Basin have achieved specific results. Therefore, exploring the spatial coupling characteristics of the soil–vegetation system, based on soil moisture, soil salinization, and vegetation cover, could provide a basis for evaluation of the effectiveness in Shiyang River governance.

2.2. Data Sources

Landsat TM/OLI data with a spatial resolution of 30 m and a time resolution of 16 days were acquired from NASA (https://search.earthdata.nasa.gov/search, accessed on 23 October 2021) and were used in this paper. First, the images were pre-processed by performing atmospheric and geometric corrections. Then, the soil moisture, normalized difference vegetation index, and soil salinization index from 2015 to 2020 (image cloud coverage in 2019 was high, which was not selected) were calculated using bands 1–7 of Landsat satellite images. The image data had high spatial resolution, making them suitable for small watershed-scale research.

Land-use data were obtained from the Resources and Environmental Science and Data Center, Chinese Academy of Sciences (https://www.resdc.cn/, accessed on 3 December 2021), with a spatial resolution of 30 m. According to the vegetation cover field theory [29] (vegetation below 5 m is defined as short vegetation, vegetation above 5 m is defined as canopy vegetation, and other types are defined as bare land) proposed by Song et al. (2019) [29], we divided land use into tree vegetation (TV), short vegetation (SV), and bare ground (BG); see

Table 1. Land use classification criteria based on vegetation field theory.

Primary Classification	Secondary Classification	Vegetation Field Classification
Arable land	Dryland	Short Vegetation (SV)
Grassland	High-coverage grassland
	Medium-coverage grass
	Low-cover grassland
Forest	Woodland	Tree Vegetation (TV)
	Shrub wood
	Sparse woods
	Other land use
Unused land	Sandy	Bare Ground (BG)
	Gobi
	Marshland
	Bare land
	Bare rock
	Saline–alkali soil

for the classification criteria.

The soil property data adopted in this paper were obtained from the Harmonized World Soil Database, comprising grid data with a spatial resolution of 1 km, providing the soil types, soil phases, and soil physicochemical properties of each grid. Moreover, we acquired meteorological data from the National Climatic Data Center (ftp://ftp.ncdc.noaa.gov/pub/data/noaa/isd-lite/FTP server, accessed on 22 February 2022). The monthly and annual mean temperature and precipitation were calculated using the data of meteorological stations in Shiyang River Basin and its surrounding areas. Spatial interpolation was carried out on monthly and annual scale data, in order to determine the relationships between the temperature, precipitation, evapotranspiration, and coupling coordination index under different vegetation cover scenarios.

2.3. Research Methods

2.3.1. Calculation of Indices

The rationality and scientific nature of index system construction are directly related to the accuracy of target-level evaluation [30]. In arid and semi-arid regions, soil moisture and salinization are closely related to vegetation growth. Therefore, for this study, we selected the normalized difference water index (NDWI), the salinity index (SI), and the vegetation coverage (FVC), which are more accurate than others, in order to extract soil moisture, soil salinization, and vegetation information. Among these indices, FVC can be calculated using the normalized difference vegetation index (NDVI). The calculation formulas are as follows [31]:

NDWI = (G − NIR)/(G + NIR),

(1)

$$SI=\sqrt{B\times R}$$

(2)

NDVI= (NIR − R)/(NIR + R),

(3)

$$FVC=\frac{\left({NDVI-NDVI}_{soil}\right)}{\left({NDVI}_{max}-{NDVI}_{soil}\right)}\times 100\%$$

(4)

where G, NIR, R, and B represent the reflectance of the green band, near-infrared band, red band, and blue band in the Landsat images, respectively. The value range of SI is from 0 to 1, and the value range of NDWI and NDVI is from −1 to 1. The closer the NDWI is to 1, the higher the soil moisture content is; the closer the NDWI value is to −1, the lower the soil moisture content. The closer the NDVI is to 1, the higher the vegetation coverage. NDVI_soil and NDVI_max represent the bare ground vegetation index and maximum vegetation index, respectively.

2.3.2. Coupling and Coupling Coordination Models

A coupling model is a method used to measure the strength of the correlation between multiple elements or systems [31]. It can describe the dynamic processes of mutual stress, interdependence, and restriction between components or systems [32]. A coupling model can only simply represent the strength and magnitude of the interaction between components or systems; it cannot distinguish the advantages and disadvantages of interactions. A coupling coordination model, derived from a coupling model, can represent the benign coupling level of interaction and judge the coordination level of elements or systems in the development process [33]. In particular, a coupling coordination model can comprehensively and objectively reflect the coordination level of the interaction between soil and vegetation in the ecological restoration process. The calculation formulas [34,35,36] are as follows:

$$h_i=\alpha s_i+\beta w_i,$$

(5)

$$C_i=2\times \sqrt{\frac{h_i\times fv{c}_i}{{\left(h_i+fv{c}_i\right)}^2}},$$

(6)

$$t=0.5h_i+0.5fv{c}_i,$$

(7)

$$D_i=\sqrt{t\times C_i},$$

(8)

where C_i is the coupling coefficient, which ranges from 0 to 1 and indicates the magnitude of coupling relationship between elements; h_i is the comprehensive soil index; fvc_i is the vegetation coverage; α and β are the weights of soil salinization s_i and soil moisture w_i, respectively; and D_i is the coupling coordination level (

Table 2. Classification criteria of coordination level and coupling relationship.

Coupling Coordination Level		hi > fvci		hi < fvci
		si > wi	si < wi	si > wi	si < wi
0 ≤ D < 0.1	Moderate imbalance	Saline land	Flood land	Salt Marsh	Marsh
0.1 ≤ D < 0.4	Low imbalance
0.4 ≤ D < 0.7	Low coordination
0.7 ≤ D < 0.9	Moderate coordination
0.9 ≤ D < 1	Ecological stability

), which ranges from 0 to 1.

3. Results and Analysis

3.1. Characteristics of Soil–Vegetation System

3.1.1. Comprehensive Indices of Soil Moisture and Soil Salinization

When calculating the comprehensive soil index, the weight of each sub-factor was determined according to the theoretical basis established by previous studies [37,38], the importance of the soil index, and experimental verification. In particular, the weight of the soil moisture was determined to be 0.7, while the weight of soil salinization was 0.3 [39]. Using the standard deviation classification method, the data were classified as indicated by the histogram in the figure, which shows the maximum value, minimum value, standard deviation, and average of pixels from 2015 to 2020. According to the comprehensive soil index, the area was divided into high water content area, water–salt balance area, and high salt content areas (Figure 2). The areas with high water content were mainly in the hinterland of Minqin Oasis, covering a small area. The area with water–salt balance occupied the largest proportion, followed by the area with high salt content. The soil moisture content of Minqin Oasis was generally low.

3.1.2. Vegetation Coverage

The vegetation coverage of Minqin Oasis has always been a concern, and vegetation coverage change is considered an important indicator of ecological change in this region. We calculated the vegetation coverage of Minqin Oasis from 2015 to 2020 (Figure 3); the line chart in the figure shows the proportion of the temporal variation characteristics for different vegetation coverages from 2015 to 2020. The area with vegetation cover was mainly in the hinterland of Minqin, which is also a region with high soil moisture. The land-use types—which belong to the short cover and canopy cover area—are mainly cultivated land, grassland, forest land. Although these areas are the main areas covered by vegetation in Minqin, the maximum vegetation coverage was less than 70%, and most of them had less than 50%. Overall, the area with 0–20% vegetation coverage accounted for the most significant proportion, the average area of which was more than over 83% during 2015–2020.

3.1.3. Spatial Coupling Characteristics of Soil–Vegetation Systems

The soil–vegetation spatial coupling relationship is shown in Figure 4 (the histogram shows maximum, minimum, and standard deviation information for 2015 to 2020). From Figure 4, no significant difference in the inter-annual coupling relationship of the soil–vegetation system from 2015 to 2020 can be observed, and the variation range of the soil–vegetation coupling coefficient was small. The soil–vegetation coupling coefficient for unused land was small, while the coupling coefficient for vegetation-covered land in the Minqin hinterland was significant. This phenomenon is because the unused land in Minqin County has been in a state of desertification all year long, and the environment is relatively stable. As such, the temporal and spatial variation characteristics of the environmental elements are not prominent. Moreover, the vegetation coverage of the unused land is deficient: basically, no vegetation grows. Therefore, the coupling coefficient between soil and vegetation in unused land tends to be 0, and the relationship is fragile.

Figure 5 shows the classification results for the soil–vegetation spatial coupling coordination level based on the classification criteria in Table 2. The areas of moderate and low imbalance were mostly outside the hinterland of Minqin. Low-level coordination accounted for a small proportion, distributed primarily in the vegetation-covered area in the hinterland of the Minqin and Qingtu Lake areas (i.e., canopy cover and short cover areas in the vegetation-covered situation). In these regions, the vegetation and soil showed a low level of coordination, indicating that the soil conditions satisfied the needs of plant growth, but the region has not developed a stable ecological environment. Figure 6 shows the statistical results of the inter-annual changes in the degree of soil–vegetation coupling coordination at different levels. The results indicated that, in the study area from 2015 to 2020, the proportion of low-level and medium-level uncoordinated soil–vegetation spatial coupling coordination areas was the largest, accounting for 79.3% annually. There was a slight change in the inter-annual coupling coordination levels. Moderate discoordination decreased from 16.5% in 2015 to 6.6% in 2020, mainly changing to low discoordination. Moderate and highly coordinated areas occupied a small proportion, mainly concentrated around the short vegetation cover and canopy cover areas in the hinterland of Minqin, where the soil provides a relatively stable ecological environment for vegetation growth.

3.2. Soil–Vegetation Coupling Characteristics under Different Vegetation Cover Scenarios

Figure 7 shows the spatial distribution characteristics of soil–vegetation coupling coordination levels under different vegetation coverage scenarios from 2015 to 2020, while Figure 8 shows the land use area ratio for varying degrees of coupling coordination. Uncoordinated coupling relationships between soil and vegetation accounted for a large proportion of the bare land. In contrast, the largest proportion of coordinated coupling relationships was observed within the canopy cover and short cover areas. This observation was attributed to the large area of unused land in the study area, including sandy land, dry land, Gobi bare rock, rocky land, and saline–alkali land, with little vegetation growth. The soil moisture in the short cover and canopy cover areas was high, and the vegetation grew well. In addition, the moderate discoordination in some regions with tree (canopy) cover and short cover changed to low discoordination or low coordination from 2015 to 2020.

The statistics of the soil–vegetation spatial coupling relationship under different vegetation coverage scenarios are shown in Figure 8. The areas with vegetation coverage higher than 20% in Minqin County were mainly canopy coverage areas and short vegetation coverage areas. Figure 8 shows that the proportion of moderate imbalance in canopy cover decreased from 40.1% in 2015 to 0.7% in 2020, while the proportions of low incoordination, low coordination, and moderate coordination all increased to varying levels. There was also a similar change pattern in the short-coverage scenario. These results indicate that, in the areas where the vegetation coverage is higher than 20%, the spatial coupling coordination level of soil vegetation tends to increase; this is mainly because cultivated land occupies a large proportion in the canopy cover and short cover area, and the measures (e.g., canal irrigation) adopted in the Minqin County in recent years have improved the growing environment for crops, thus also improving the coupling coordination between soil and vegetation.

The results demonstrated that, although the ecological environment of Minqin Oasis has been improved since the implementation of the Shiyang River watershed treatment plan, the soil area that could provide appropriate water and nutrients for plant growth is still small. The soil area of vegetation growth in Minqin Oasis is inadequate, and vegetation coverage in the growing area is low. Creating a suitable environment for vegetation growth, expanding the vegetation area, and improving the vegetation coverage are still the main ecological problems to be solved in this area. The ecological environment in the hinterland of Minqin Oasis is fragile. Affected by the surrounding desertification, vegetation growth is highly sensitive to environmental changes, especially in areas with vegetation coverage higher than 20%. Notably, the soil–vegetation coupling coordination degree significantly differed under different vegetation coverage scenarios.

4. Discussion

4.1. The Influence of Climate Factors on Soil–Vegetation Spatial Coupling Coordination Level

Drought is the leading factor affecting desertification and ecological fragility in the lower part of the Shiyang River Basin [23,24,25]. Therefore, it is necessary to discuss the influence of climate factors on the coupling coordination level. In this study, we considered the influences of annual mean temperature, annual precipitation, and annual evaporation on soil–vegetation spatial coupling coordination level under three different vegetation cover scenarios. The correlations between the annual average and monthly average precipitation, temperature, and evapotranspiration and the coupling degree under different vegetation cover scenarios were calculated separately, as shown in Figure 9. Over the years, precipitation showed a significant positive correlation with the soil–vegetation coupling coordination level under the three vegetation cover scenarios (Figure 9), indicating that the coupling coordination level increases, to some extent, with an increase in precipitation. Both evapotranspiration and temperature had slight adverse effects on the soil–vegetation coupling coordination level; that is, when temperature and evapotranspiration increase, the soil–vegetation coupling coordination level may decrease. However, the relevant coefficients were small, indicating that the degree of influence was also small. The relationships between precipitation, temperature, evapotranspiration, and soil–vegetation coupling coordination level under different vegetation cover scenarios, on a monthly scale for 2015, were analyzed (Figure 9). Precipitation was still the most significant climate factor influencing soil–vegetation coupling coordination level on a monthly scale, having a positive influence. Again, the effects of temperature and evaporation were relatively weak and negative.

The above analysis indicated that the soil–vegetation spatial coupling coordination level is particularly sensitive to changes in climate factors; specifically, the coupling coordination level was most highly sensitive to a change in precipitation, followed by a change in evaporation, and least sensitive to temperature change, consistent with previous arguments [15]. Precipitation can increase the soil–vegetation coupling coordination level, to a certain extent [40], whereas temperature and potential evaporation reduce the level of coupling coordination. Thus, the results indicate that low precipitation, high temperature, and high potential evaporation in the lower part of the Shiyang River Basin are crucial reasons for the low level of spatial coupling coordination between soil and vegetation.

4.2. The Influence of Soil Properties on Soil–Vegetation Spatial Coupling Relationships

The level of soil–vegetation spatial coupling coordination is affected not only by climate factors but also by soil properties [13]. Soil provides the water and nutrients necessary for plant growth, and plants can differentially obtain extra water and nutrients under differing soil property scenarios, resulting in significant differences in vegetation [14,15]. Exploring the influence of soil properties on the soil–vegetation spatial coupling relationship can allow for a better understanding of the physical and chemical properties and soil–vegetation coupling characteristics, providing the necessary theoretical support for the study of soil–vegetation spatial coupling relationships.

Many factors, such as various soil indices, can be used to characterize the physical and chemical properties of soil [41]. Other than soil moisture, four soil properties, including soil conductivity, sulfate content, soil carbonate content, and soil sand content, have significant impacts on the ecological security of the lower part of the Shiyang River Basin. Therefore, we considered the above four indicators, with regard to their effects on the soil–vegetation coupling coordination level. The spatial distribution characteristics of the four soil elements are shown in Figure 10. The soil sand content in Minqin was very high, especially in the bare land area outside the hinterland of Minqin, where the soil sand content was more than 40%. In addition, the soil electrical conductivity and soil sulfate content in the northwest and northeast of Minqin were high. The soil vegetation in these regions is uncoordinated (Figure 5). It is difficult for these areas to provide plants with suitable water and nutrients, due to the high salt and high sand content, making it difficult for vegetation to grow and, thus, causing soil–vegetation disharmony. The results indicate that the sediment concentration, soil electrical conductivity, and soil sulfate content decrease the soil–vegetation coupling coordination level; that is, high soil sediment concentration, soil conductivity, and soil sulfate content led to a decrease in the level of coupling coordination between soil and vegetation, which is reflected in the deterioration of the ecological environment and soil desertification.

The carbonate content in the lower part of the Shiyang River Basin was higher in the northwest, the Qingtu Lake area, and the Minqin Oasis hinterland. The soil–vegetation coupling characteristics are different in these areas. The northwest and Qingtu Lake regions showed more discoordination, whereas the hinterland of Minqin Oasis showed low or moderate coordination. This indicates that soil carbonate content has different influence characteristics on soil–vegetation coupling coordination under different vegetation cover scenarios. In the area with high vegetation cover, carbonate is mainly absorbed by plants to provide the required nutrients [42], and the coupling coordination level is high. In areas with low vegetation coverage and bare land, carbonate mainly accumulates on the surface, aggravating soil salinization and resulting in low coupling coordination.

There exists a certain coupling relationship between the soil–vegetation system in the lower part of the Shiyang River Basin. Climate factors, soil properties, and land use all have varying influences on the coupling coordination level. Dry weather, large unused bare landcover, high sulfate content of the soil, and sediment created mismatched conditions for soil–vegetation spatial coupling in the study area. However, there was an increase in low levels of coupling coordination and a shift from low- to medium-level coupling coordination under different vegetation situations. The results showed that the degree of spatial coupling coordination between soil and vegetation has increased, to a certain extent, during the process of ecological environment restoration in the lower part of the Shiyang River Basin. Although the ecological environment is still fragile, it has been improved, to some extent, under different vegetation cover situations.

In this paper, three typical remote sensing indices were selected to construct a soil–vegetation system that is representative and feasible. However, soil–vegetation systems are complex open systems, and studies at different scales require different indicator systems [43]. The indicators selected in this paper are suitable only for exploring the coordination between vegetation growth and changes in soil water/salt content, and can well illustrate the effects on vegetation of soil water/salt content at the pixel scale.

In addition, statistical data on groundwater and soil properties need to be fully considered when constructing a soil–vegetation system. However, only the spatial data that could be directly monitored via by remote sensing images were collected in this paper, making the research results somewhat one-sided. Some scholars have also investigated the mechanisms of soil–vegetation–water influence and have summarized research methods used at different scales [17,18], thus laying a good foundation for research into soil–vegetation–water coupling. In future research, an integrated approach that combines multi-system and multi-indicator elements should be used to systematically study the coupling characteristics between soil and vegetation at different scales, which is of great significance for understanding the vegetation–soil response mechanisms.

5. Conclusions

(1)

The soil system was characterized by soil moisture and salinization, and the spatial coupling characteristics of soil and vegetation in the lower part of the Shiyang River Basin were explored. The results demonstrated that there exists a strong coupling relationship between soil and vegetation in the study area. Vegetation coverage was highly sensitive to soil moisture and salinization, which contributes to our understanding of the response mechanisms between vegetation and soil.

(2)

Under different vegetation cover scenarios, the level of soil–vegetation spatial coupling coordination differed. In particular, bare land showed moderate and low imbalance, and the short cover and canopy areas showed low and moderate coordination. The area of coordination within the study area was small, indicating ecological imbalance and suggesting that soil–vegetation systems have not achieved stability.

(3)

In the lower part of the Shiyang River Basin, arid climatic conditions, high sediment content, and high sulfate content were found to greatly influence the degree of soil–vegetation coupling coordination. These factors have significant impacts on ecological and environmental governance in the study area.