Identification of Ecological Management Zoning on Arid Region from the Perspective of Risk Assessment

Abstract

Strengthening ecosystem monitoring and improving the efficiency of ecological risk assessment are of great significance for the sustainable development of ecosystems in an arid area. Using remote sensing monitoring data of land use, the ecological risk status of typical arid areas is assessed, its spatial heterogeneity is analyzed from the perspective of space-time, and the key areas of ecological risk management are finally identified. The results show that (1) the transformation mode of landscape ecological risk was dominated by medium–low risk to medium risk, and low risk to medium–low risk as a secondary level; in addition, the ecological risk level of the region increased. From 2000 to 2020, the transformation mode of landscape ecological risk was dominated by medium risk to medium–low risk, and medium–low risk to low risk as a secondary level; in addition, the regional ecological risk level shows a moderating trend. (2) The spatial difference in the ecological risk level in the Hexi region is obvious, showing a gradual decline from east to west, and the spatial difference in the ecological risk level in the west is significant. (3) A total of 1194 grid units are identified as key areas for ecological risk management in the Hexi region, accounting for 43% of the study area. This study provides important theoretical basis for ecosystem monitoring and risk assessment management in similar areas of arid regions.

1. Introduction

Ecological risk refers to the possibility and loss of the negative impact of natural factors and human activities on the function, structure and stability of the ecosystem. Ecological risk assessment is an effective tool in measuring and evaluating this negative impact [1,2,3]. The expansion of human activities leads to changes in the landscape pattern and structure, and the ecological security state becomes very unstable, thereby threatening human well-being [4,5,6,7]. It is a very frequent situation in arid regions around the world. In ecologically fragile areas, the spatial–temporal heterogeneity of natural and social ecosystems is more significant [8]. Therefore, ecological risk assessment and the identification of key areas for ecological risk management in such areas are of great significance for future ecosystem management and the formulation of risk response strategies.

Ecological zoning not only helps us understand abstract nature, but also serves as an important guiding form for achieving sustainable management of ecosystems [9]. Global ecological zoning and existing ecological zoning in different countries have made absolute contributions to protecting the Earth and regional ecosystems. Researchers have proposed an important concept of global ecological functional zoning at a global macro scale [10]. This concept is different from biogeographical zoning; however, it considers the geographical characteristics and ecosystem services of typical regions. Landscape ecological risk assessment emphasizes the impact of a landscape pattern on specific ecological functions or processes. It is important to pay attention to the overall loss of landscape in providing ecosystem services and ecological functions and evaluate the target towards vulnerability, elasticity, and stability. The research results focus on the comprehensive representation and spatial visualization of multiple risk sources which can provide decision support for sustainable landscape planning and ecosystem management [11,12,13,14].

Landscape risk assessment takes a specific area as a comprehensive risk receptor, which can effectively reflect the impact of the dynamic evolution of the landscape pattern on regional ecological processes and ecosystem health. The assessment methods mainly include the risk “source–sink” method and landscape index method [1,13,15]. The risk “source–sink” method needs to be combined with specific ecological processes or disaster risks in the evaluation process. It is more suitable for directional risk evaluation with clear risk stress factors in a certain area, and it is widely used in non-point-source pollution risk evaluation. For example, Polyakov et al., evaluated the performance of the AnnAGNPS (Annualized Non-Point Source Pollution Model) in simulating runoff and soil erosion in a 48 km² watershed located on the Island of Kauai, Hawaii [16]. Wang et al., identified the source–sink risk of cultivated land non-point-source pollution in the Three Gorges reservoir area based on the minimum cumulative resistance mode [17]; Li et al., linked potential heat sources and sinks with the urban heat island to explore the heterogeneous impact of the landscape pattern on the surface temperature [18]. The landscape index method is based on land use, specifically constructed according to the landscape disturbance index and landscape vulnerability index, and pays attention to the comprehensive ecological risk effects of the landscape spatial pattern and its changes. It is more suitable for the comprehensive risk assessment of watersheds, regions, cities and other scales, and the evaluation results are characterized by the product of landscape loss and risk probability [15,19]. For example, Zhang et al., and Kang et al., used the landscape pattern index method to study the landscape ecological risk of an inland river basin in an arid area based on land use change [20,21]; Zhang et al., analyzed the landscape ecological risk level of national key areas along the Belt and Road initiative [22]; Mo et al., and Li et al., assessed the landscape ecological risk levels of megacities [23,24]. With the application of the landscape index method in research practice, the reliability of the key step of quantifying the vulnerability of different landscape types into constants in the existing methods has been questioned by many scholars [13,25]. A new framework for ecological risk assessment of wave energy converter projects has been established through research, making it a freely accessible online tool for interactive assessment and visualization of stress and ecological risks [26]. In Australia, a new ecological risk assessment framework applied to fisheries, termed Ecological Risk Assessment for the Effects of Fishing (ERAEF), has applicability in a wide range of fisheries [27]. Ran et al., developed a landscape ecological risk assessment framework based on comprehensive landscape pattern characteristics and landscape fragility dynamics and applied this framework to analyze the spatial–temporal changes in landscape ecological risk in the China Yangtze River Economic Belt, and the research results were widely recognized [28].

The management and risk assessment of ecosystems in arid areas are of great significance for the production and living within the region, as well as for the protection of the global ecological environment. The Hexi region of China is not only a typical fragile ecological environment area in northwest China, but it also has global representativeness due to its cultural and economic development. As the golden zone of the Silk Road Economic Belt, the Hexi region carries China’s transportation, trade, and other functions that connect the international world. In 2000, the Chinese government has implemented the western development strategy, and the urbanization of Hexi has accelerated. In 2013, as the Chinese government put forward the “the Belt and Road” initiative [29], the Hexi region was developed and constructed as a key area of the Belt and Road, and the land use and development pattern has changed significantly. With the rapid development of the social economy, the contradiction between regional development and the ecological environment has become more and more obvious, and a series of ecological environment problems, such as desertification and vegetation degradation, have become increasingly prominent [30,31], which has seriously hindered regional ecological environment protection and ecosystem management.

Incorporating an improved landscape ecological risk assessment framework, this study discussed the temporal and spatial characteristics, dynamic transfer patterns and dominant structure of land use and landscape ecological risks in the Hexi region in the 20 years around 2000 from the perspective of time and space, and identified the key areas of ecological risk management in the Hexi region. Targeted measures have been proposed to address the key issues faced by ecosystem management in different types of regions, providing important theoretical basis for ecological risk management decision-making in the Hexi region of China and even in arid regions around the world.

2. Materials and Methods

Taking the typical ecologically fragile areas in arid areas as the research object, based on the remote sensing monitoring data of land use, and integrating the net primary productivity and normalized vegetation index, this study discussed the spatio-temporal evolution and spatial correlation of land use and landscape ecology risks in Hexi from 1980 to 2020 and finally identified the key areas of ecosystem risk management (Figure 1).

2.1. Study Area

The Hexi region is located in the northwest of Gansu Province, China, and is a typical arid area in China and even globally. It borders the Tibetan Plateau in the south, the Mongolian Plateau in the north, the Tarim Basin in the west and the Loess Plateau in the east (37°10′–42°50′ N, 93°20′–104°00′ E). It is approximately 1000 km long from east to west, and 30–120 km wide from north to south. It is located in the west of the Yellow River. Spatially, there are three inland river basins, namely Shule River, Heihe River and Shiyanghe River, respectively, from west to east. The administrative division consists of 5 cities and 20 districts and counties, with a total area of 247,800 km², accounting for 58.15% of the total area of Gansu Province (Figure 2). By 2020, the Hexi district’s GDP was CNY 229.137 billion, accounting for 25.41% of Gansu’s GDP, with 4,395,600 residents, accounting for 17.58% of Gansu’s resident population [32]. On the one hand, the Hexi region is an important channel for China to open to the west and successfully implement the grand blueprint of the “One Belt and One Road” concept, as well as a golden location on the Silk Road economic belt and a key channel of national strategy. On the other hand, the resources, industries and culture of the Hexi region play an extremely important role in the whole country [33].

2.2. Data Sources

The data used in this study and their sources are shown in

Table 1. Data information and sources.

Data	Details	Resolution	Sources
Land use remote sensing monitoring data	1980, 2000, and 2020	30 m	Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (http://www.resdc.cn, accessed on 5 January 2022)
Basic geographic data	Vector data of administrative boundaries, cities, and rivers	Line/point data
Digital elevation mode	For extracting the slope and elevation	30 m
Socioeconomic raster data	Population density raster data for 1980, 2000, and 2020	1 km	WorldPop data platform (https://www.worldpop.org, accessed on 10 January 2022)
	GDP density raster data for 1980, 2000, and 2020	1 km	National Earth System Science Data Center (http://www.geodata.cn), accessed on 10 January 2022
Meteorological data	Precipitation and temperature for 1980, 2000, and 2020	1 km	Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (http://www.resdc.cn, accessed on 3 February 2022)
Net primary productivity	1980, 2000, and 2020	1 km	US National Aeronautics and Space Administration (http://modis.gsfc.nasa.gov, accessed on 3 February 2022)
Normalized difference vegetation index	1980, 2000, and 2020	1 km	US Geological Survey (USGS) (https://www.usgs.gov, accessed on 5 February 2022)

. The land use remote sensing monitoring data are based on the National Resources and Environment database of the Chinese Academy of Sciences and the Landsat remote sensing image data of the United States as the main information source. It is one of the most accurate remote sensing monitoring products available in China at present. The overall accuracy of type interpretation is above 90% [34,35].

2.3. Methods

2.3.1. Land Use Transfer Matrix

The method can reflect the dynamic process of mutual transformation between land types. It not only includes static data of a regional area at a certain point in time, but also contains more abundant information of regional area transfer and transfer at the beginning and end of the research period. It is helpful to further understand the structural characteristics before and after the transformation of different terrestrial species [36,37]. Its calculation formula is as follows:

$$S_{ij}=\left[\begin{array}{cc}\begin{array}{cc}{S}_{11}& S_{12}\\ S_{21}& S_{22}\end{array}& \begin{array}{cc}\cdots & S_{1n}\\ \cdots & S_{2n}\end{array}\\ \begin{array}{cc}⋮& ⋮\\ S_{n1}& S_{n2}\end{array}& \begin{array}{cc}⋮& ⋮\\ \cdots & S_{nn}\end{array}\end{array}\right],$$

(1)

where $S$ represents the area, $n$ represents the number of land types, $i$ represents the initial area, and $j$ represents the final area.

2.3.2. Construct Landscape Ecological Risk Assessment Mode

The disturbance intensity of human activities and natural changes, as well as the ability of the landscape system itself to maintain a stable ecological structure and function, are of decisive significance to the generation of regional ecological risks. Meanwhile, different landscape types have different abilities to resist external disturbances and maintain the stability of their own systems [23]. The risk assessment system is constructed from the two dimensions of landscape disturbance and vulnerability [19,23,38]. Ran et al., developed a landscape ecological risk assessment framework based on integrated landscape pattern characteristics and landscape vulnerability dynamics, and they applied this framework to analyze temporal and spatial changes in landscape ecological risks in the Yangtze River Economic Belt in China. The research results were widely recognized [28].

In this study, the fishing net creation tool in ArcGIS10.4 was used to generate 2724 grid cells (10 km × 10 km) as assessment units, covering the entire study area. The formula for calculating the landscape ecological risk index is as follows:

$$E_k={{\displaystyle \sum }}_{i=1}^n\sqrt{D_{ki}\times V_{ki}}\times \frac{A_{ki}}{A_k} ,$$

(2)

where $E_k$ represents the landscape ecological risk index of grid unit $k$, $i$ represents the land use type, $D_{ki}$ represents the disturbance index, and $V_{ki}$ represent the vulnerability index. $A_{ki}$ represents the area of the land type $i$ in grid unit $k$.

Landscape disturbance index ($D_{ki}$) mainly consists of three components of fragmentation, separation and dominance [22,28]. The calculation formula is as follows:

$$D_{ki}=aL{C}_{ki}+bL{S}_{ki}+cL{D}_{ki},$$

(3)

$$L{C}_{ki}=\frac{N_{ki}}{A_{ki}},$$

(4)

$$L{S}_{Ki}=\frac{\sqrt{S_{Ki}}}{2P_{ki}} \left(S_{ki}=\frac{N_{ki}}{A_k},P_{ki}=\frac{A_{Ki}}{A_k}\right),$$

(5)

$$L{D}_{ki}=\frac{Q_i+M_{ki}}{4}+\frac{P_{ki}}{2} ,$$

(6)

where $N_{ki}$ represents the number of patches of type $i$ land in grid unit $k$, and the meanings of $A_k$ and $A_{ki}$ are the same as those in Formula (2). $L{C}_{ki}$ represents the fragmentation, $L{S}_{Ki}$ represents the separation and $L{D}_{ki}$ represents the dominance. Variables $a$, $b$ and $c$ represent the weights of $L{C}_{ki}$, $L{S}_{Ki}$ and $L{D}_{ki}$, respectively. They are 0.5, 0.3 and 0.2 in reference to the relevant literature. $S_{ki}$ represents the ratio of the area of type $i$ land to the area of grid unit $k$; $P_{ki}$ represents the ratio of the area of type $i$ land to the area of grid unit $k$; $Q_i$ represents the ratio of the number of grids appearing in type $i$ land to the total number of grid units; $M_{ki}$ represents the ratio of the number of patches of type $i$ land to the total number of patches in grid unit $k$.

Landscape vulnerability index ($V_{ki}$) represents the vulnerability of an ecosystem’s internal structures represented by different landscapes which can reflect the resistance capacity of different landscape types to disturbances caused by human activities and natural changes. In previous studies [25,39], the experts divided the vulnerability of landscape types into six levels, from high to low: unused land, water area, cultivated land, grassland, forest land and construction land. They then obtained the vulnerability index of each landscape type through normalization processing. On the basis of expert rating and in combination with the latest research results [28], this study takes the vulnerability level of each type of landscape as the empirical value ($E{V}_i$) and screens 8 indicators from the three dimensions of exposure, sensitivity and adaptability to calculate the complex correction factor ($C{F}_k$). Finally, the modified landscape vulnerability index ($V_{ki}$) is obtained. The calculation formula is as follows:

$$V_{ki}=E{V}_i\times C{F}_k,$$

(7)

$$C{F}_k=\frac{E{F}_k}{\overline{E{F}_k}},$$

(8)

$$E{F}_k={{\displaystyle \sum }}_{j=1}^8{w}_j\times m_{jk},$$

(9)

where $V_{ki}$ represents the landscape vulnerability index of type $i$ land in the modified grid unit $k$, $E{F}_k$ is the weighted sum of the modified indicators in grid unit $k$, $w_j$ represents the weight of index $j$, and $m_{jk}$ represents the standardized value of index $j$. The selected 8 indicators are shown in

Table 2. Landscape vulnerability index correction factors.

Dimensions	Indicators (Abbreviation)	Unit	Weight	Orientation
Exposure	GDP density (GDP)	104 yuan/km2	0.1042	Positive
	Population density (Pop)	Population/km2	0.1074	Positive
Sensitivity	Elevation (Ele)	m	0.1299	Positive
	Slope (Slp)	%	0.1248	Positive
	Temperature (Temp)	°C	0.1291	Negative
	Precipitation (Prep)	mm	0.1341	Negative
Adaptive capacity	Net primary productivity (NPP)	g·C/m2	0.1356	Negative
	NDVI values (NDVI)	/	0.1345	Negative

Note: The index values used for the calculation are the average of each unit and were standardized. The weights of the indicators were obtained based on the entropy method.

.

2.3.3. Change Rate of Landscape Ecological Risk

The change rate of the unit landscape ecological risk could effectively identify the spatial–temporal heterogeneity of risk changes [22,40]. The calculation formula is as follows:

$$RE{C}_k=\frac{\left(E_k^b-E_k^a\right)}{E_k^a}\times \frac{1}{t}\times 100\%,$$

(10)

where $RE{C}_k$ represents the change rate of the landscape ecological risk of grid unit $k$; $E_k^a$ and $E_k^b$ represent the landscape ecological risk index of grid unit k at moment $a$ at the beginning of the study period and moment $b$ at the end of the study period, respectively; $t$ represents the time interval.

2.3.4. Variation Coefficient of Landscape Ecological Risk Index

The variation coefficient can effectively measure the variation degree of the array. By calculating the variation coefficient ($CV{E}_k$) of the landscape ecological risk index, the stability of the ecological risk of each grid unit can be effectively monitored. The larger the coefficient of variation value, the worse the stability of risk change; otherwise, the more stable it is [41,42]. The calculation formula is as follows:

$$CV{E}_k=S{D}_k/M{N}_k,$$

(11)

where $CV{E}_k$ represents the variation coefficient of the landscape ecological risk index in unit $k$; $S{D}_k$ and $M{N}_k$ represent the standard deviation and average value of the landscape ecological risk index in grid unit $k$, respectively.

3. Results

3.1. Spatiotemporal Dynamics of Land Use from 1980 to 2020

3.1.1. Spatiotemporal Distribution Characteristics of Land Use Change

The land use types in the Hexi region showed obvious regional differentiation from southwest to northeast, displaying a spatial distribution pattern from woodland to grassland to cultivated land and construction land, and then to unused land (Figure 3). Unused land is the largest land use type in the Hexi region, mainly distributed in the west, north and east. The second is grassland, mainly distributed in the south and southeast. The sum of the two land use types accounts for approximately 90% of the total area of the Hexi region (Figure 4).

During the research period, the cultivated land, water area, and construction land area in the Hexi region showed an increasing trend, with the highest annual average growth rate of construction land. From the perspective of the severity of land use change, the dynamics of land use change have become more significant since 2000 (

Table 3. Land use change rate in the Hexi region from 1980 to 2020.

Rate of Change (%)	Cultivated Land	Forest Land	Grassland	Water	Construction Land	Unused Land
1980–2000	3.28	0.04	–0.47	–0.95	6.21	0.13
2000–2020	14.22	–0.29	–0.60	19.75	62.39	–1.61
1980–2020	17.97	–0.26	–1.06	18.62	72.48	–1.48

). In particular, the rate of land use change in construction land increased from 6.21% to 62.39%, and that of cultivated land rose from 3.28% to 14.22%. At the same time, the unused land area changed from an increasing trend to a decreasing trend, which was mainly because, after 2000, the Hexi region entered the stage of rapid urbanization, and population growth and urban expansion caused the demand for construction land and food production to increase significantly; thus, urban construction land and cultivated land gradually expanded to the unused marginal land.

3.1.2. Transformation Modes of Land Use

In the Hexi region, different land use types were converted into each other, and the land use transformation modes were different in different periods (Figure 5).

The land use transformation mode was mainly from grassland to unused land and from unused land to grassland in the period from 1980 to 2000. From the perspective of net transfer out, the area of grassland transfer was the largest, and the area converted to unused land was 774.57 km², accounting for 60.24% of the total area of grassland transfer, and the area converted to cultivated land accounted for 28.14% of the total area of grassland transfer. From the perspective of net transfer in, the converted area of grassland was the largest, followed by unused land and cultivated land. At the same time, grassland and unused land were the largest contributors to each other; the two contributed almost the same amount of area to cultivated land, and the sum of the area was 87.38% of the total converted area of cultivated land.

The land use transformation mode was mainly from unused land to cultivated land and from unused land to grassland in the period from 2000 to 2020. From the perspective of net transfer out, the unused land transfer area was the largest, which was 3591.47 km², and the area converted to cultivated land and grassland was 1470.54 km² and 1033.02 km², accounting for 40.95% and 28.76% of the total unused land transfer area, respectively. From the perspective of net transfer in, the maximum converted area of cultivated land was 2421.37 km², and the contribution rates of unused land and grassland were 60.73% and 32.11%, respectively. Grassland was the next largest area, and the main source was unused land.

Comparing the intensity of land use change in the two periods before and after 2000, it can be seen that the intensity of land use change was the most intense from 2000 to 2020, with a change area of 6072.54 km², and that of 1980 to 2020 had a change area of 3048.03 km².

During the study period, with the passage of time, the transformation rate of land use types in different periods was different and presented significant characteristics in terms of spatial distribution (Figure 6).

In the whole study period, the area of cultivated land, construction land and water areas increased significantly, the area of unused land and grassland decreased slightly, and the area of forest land remained essentially stable. During the period from 1980 to 2000, the increased cultivated land was mainly distributed in Gaotai County, Linze County, Ganzhou District and Yongchang County in the central part of the Hexi region, and the increased construction land was mainly distributed in Jiayuguan City, Ganzhou District, Jinchuan District and Liangzhou District in the central part. The reduced grassland area was mainly distributed in Suzhou District, Linze County, Ganzhou District, Sunan County, Shandan County and Tianzhu Tibetan Autonomous County in the centra area, and Dunhuang City and Yumen City in the west. The reduced unused land is consistent with the increased cultivated land. From 2000 to 2020, land use changes were more frequent, and the increase and decrease in land use area were especially significant in the spatial distribution. The construction land area increased greatly, mainly distributed in Dunhuang City, Jiayuguan City, Suzhou District, Ganzhou District and Liangzhou District, etc. At the same time, there was a relative spatiotemporal coincidence between the area of unused land and grassland and the area of cultivated land and construction land. In general, the increase in cultivated land and construction land was mainly distributed in the central districts and counties, with relatively dense population activities. The unused land and a small part of the grassland in these districts and counties will eventually be converted into land for production and living due to the spread of production and the increased living space of residents.

3.2. Spatiotemporal Dynamics of Landscape Ecological Risks from 1980 to 2020

3.2.1. Spatiotemporal Distribution Characteristics of Landscape Ecological Risk

In this study, by using the ArcGIS10.4 software, the calculated ecological risk index was divided into five levels: high-risk area, medium-risk area, medium–low-risk area and low-risk area. We used the natural break point method to assess the landscape ecological risk level in the Hexi region from 1980 to 2020. During the study period, the risk change characteristics of different risk levels and the change rate of ecological risks in different periods were different in the Hexi region, showing different spatial distribution characteristics.

From the perspective of spatial distribution, the spatial difference in ecological risk levels in the Hexi region is large, showing a gradual decline from east to west, and the level in the west is significant (Figure 7). High-risk areas are distributed in Gaotai, Linze, etc. Medium-high risk areas re distributed in Dunhuang, Subei and Guazhou, while the rest are scattered at the edges of high-risk areas. Medium-risk areas are distributed in large areas of Subei, Guazhou, Dunhuang, Aksai and Minqin. Medium–low-risk areas are distributed on the cross-edge between the medium–high-risk area and another medium–high-risk area. The low-risk area is the smallest, its distribution is roughly consistent with the Qilian Mountains, and it is mainly located in the low-altitude area.

The medium–high risk areas in the Hexi region account for the largest proportion, all above 56%, followed by the area of medium–high risk, accounting for approximately 32%, and the area of low risk accounts for the smallest proportion, forming a risk structure dominated by medium–high risk areas (

Table 4. Area and change in different ecological risk regions in the Hexi Corridor from 1980 to 2020.

Land Type	1980			2000			2020			1980–2000	2000–2020	1980–2020
	Number of Units	Area (km2)	Area Proportion (%)	Number of Units	Area (km2)	Area Proportion (%)	Number of Units	Area (km2)	Area Proportion (%)	Rate of Change (%)	Rate of Change (%)	Rate of Change (%)
High	137	10,301.07	4.16	137	10,129.23	4.09	137	10,055.84	4.06%	–1.67%	–0.72%	–2.38%
Medium–high	1695	13,9351.3	56.28	1701	14,0238.4	56.64	1701	140,645.6	56.80%	0.64%	0.29%	0.93%
Medium	1016	79,909.72	32.27	1022	80,883.24	32.66	988	78,218.12	31.59%	1.22%	–3.30%	–2.12%
Medium–low	237	14,867.53	6.00	229	13,476.24	5.44	251	15,623.2	6.31%	–9.36%	15.93%	5.08%
Low	128	3184.94	1.29	124	2887.388	1.17	136	3071.789	1.24%	–9.34%	6.39%	–3.55%

). From 1980 to 2000, the areas of medium–low risk and low risk decreased significantly, with a change rate of −9.36% and −9.34%, respectively, while the areas of medium risk and medium–high risk increased slightly, indicating that the regional ecological risk was upgraded during this period. During the period from 2000 to 2020, the area of high risk decreased slightly, and the area of medium risk decreased greatly, with a change rate of −3.30%. The high and low risk levels complement each other, and the overall situation is relieved. It shows that the high-risk area, medium-risk area and low-risk area decreased slightly, the medium–low-risk area increased greatly, and the medium–high-risk area showed a small increase, indicating that the regional ecological risk was alleviated as a whole, but some areas were under greater risk stress.

3.2.2. Transformation Mode of Landscape Ecological Risk Level

The transformation characteristics of the landscape ecological risk level in the Hexi region were obvious in the period from 1980 to 2020 (Figure 8). Specifically, from 1980 to 2000, the mode of ecological risk transformation was mainly from low risk to medium risk, and from low risk to low risk to medium risk. From the perspective of net transfer out, the transfer-out area of the medium-low risk area was the largest, which was 7215.86 km², and it was mainly converted into a medium-risk area, followed by the medium-risk area with a larger transfer-out area. From the perspective of net transfer in, the transfer area of the medium-risk area is the largest, and the main sources are the medium–low-risk area and the medium–high-risk area. During the period from 2000 to 2020, the mode of ecological risk transformation is mainly medium risk to low risk and medium–low risk to low risk. From the perspective of net transfer out, the transfer-out area of the medium risk area is the largest, which is 12,790.89 km², followed by the medium-low risk area, which is mainly converted into low risk and medium-low risk areas. In terms of net transfer in, the transfer-in area of the medium-risk area was the largest, followed by the transfer-in area of medium-low risk area, which were 12,432.67 km² and 10,470.63 km², respectively.

It can be found that the ecological risk change area increased from 20,892.82 km² to 37,059.78 km² by comparing the changes in the two periods, with a change rate of 77.38%. The landscape ecological risk level in the Hexi region changed dramatically from 2000 to 2020. Considering the transformation mode of the risk level in this period, it can be found that the overall landscape ecological risk level in the Hexi region showed a downward trend after 2000, and the ecological risk stress was generally alleviated.

3.3. Key Areas of Ecological Risk Management

Ecological risk management needs to consider various factors, such as the natural environment and social economy, from a holistic perspective. Further combined with the results of ecological risk assessment, preventive measures and management countermeasures to avoid, mitigate, inhibit and transfer risks are taken. In order to improve the effectiveness of ecological risk management, save management resources and realize the organic integration of society, economy and ecology, it is necessary to identify the key areas of risk management so as to guide the formulation of preventive measures and management policies more efficiently. This study combined the state of risk grade, the rate of risk change and the stability of risk change referring to relevant studies [28], combined with the actual situation of the study area, took 20% as the selection threshold and determined three types of key areas.

Type I are stable high-risk areas (Figure 9a). This type of area is always at a high-risk level during the whole study period, which creates a huge threat and places pressure on the ecological security of the landscape system in the region, and this is important to consider in ecological risk management. Type I contains 569 grids, and mainly focuses on the middle reaches of the Heihe River and Gulang Basin; the rest are mainly distributed in the interlacing zone between grassland and unused land in the western area of the Hexi region.

Type II is risk rise rapid area (Figure 9b). The ecological risk was rapidly upgraded with a large change rate, and the ecological security of the landscape system in the region was seriously threatened by continuous upgrading. Rapidly rising risk areas contain 371 grids, of which approximately 21% (78 grid units) overlap with Type I areas. This type of areas is mainly distributed in semi-agglomeration in Jinta, Jiayuguan and Suzhou in the central part of the Hexi region, as well as in the interlacing zone of cultivated land and grassland in the middle and lower reaches of the Shiyanghe River.

Type III is risk-unstable area (Figure 9c). The ecological risk was in an unstable state, the risk levels varied but failed to remain stable, and the regional environmental conditions may increase the probability of a future risk increase. There were 615 grids in the unstable risk areas, of which 43% (283 grid units) coincided with Type II, and the rest were scattered in the marginal interlace zones of cultivated land, grassland, construction land and unused land in various districts and counties.

Combined with the above three types of areas, the key areas account for 43% of the study area, which include 1194 grids (Figure 5d). The results showed that Gulang County and Ganzhou District in the Hexi region were the most critical areas for ecological risk management, accounting for more than 70% of the areas, followed by Suzhou District, Jiayuguan City, Linze County, Gaotai County and Yongchang County, accounting for more than 60% of the areas.

4. Discussions

4.1. Attribution Analysis of Land Use and Landscape Ecological Risk Change

The interaction of various factors such as ecological environment and economic development has finally formed the change characteristics of land use and landscape ecological risk in the Hexi region. During the period from 1980 to 2000, the area of cultivated land increased obviously, the area of construction land increased less, and the area of grassland and water area decreased. The conversion of grassland to unused land leads to a significant increase in ecological risk. Since the beginning of the reform and opening up, in order to solve the problem of grain shortage in Gansu Province, the Gansu Provincial Party Committee formulated the implementation plan for the construction of a commodity grain base in the Hexi region from 1978 to 1981. From 1982 to 1995, the Chinese government put forward the policy of promoting the benefits of Hexi and the poverty of the central part, and transferred 128,000 people from the poverty-stricken areas in the central and southern parts of Gansu Province to Hexi. The implementation of these policies and programs is the main reason for the growth in the cultivated land area and construction land area in the Hexi region. The ecological environment in the Hexi region was poor, and the unreasonable human activities that took place for a long time led to the deterioration of the regional ecological environment, the degradation of grassland vegetation and the contradiction between the production and living demands of a large number of people; the regional supply capacity was prominent, the relationship between man and land was increasingly tense, and the regional ecological risk increased significantly during this period.

Land use change was severe from 2000 to 2020 in the Hexi region. Specifically, the construction land, cultivated land and water area increased significantly, and the unused land area decreased. The land use transformation mode was mainly from unused land to cultivated land and from unused land to grassland. In this period, the landscape ecological risk in the Hexi region showed a moderate trend, which was mainly manifested as the expansion of the medium–low-risk area, and the minor shrinkage of the medium- and high-risk areas. Since 2000, the state has successively issued the Short-Term Governance Plan for the Heihe River Basin (2001), the Key Governance Plan for the Shiyanghe River Basin (2007) and the Comprehensive Plan for Rational Utilization of Water Resources and Ecological Protection in Dunhuang (2011–2020) (2011). The whole Hexi region has become an important area for the promotion of ecological protection and the coordinated development of the economy and society. After 20 years of governance, the Hexi region has initially reversed the trend of ecological deterioration and achieved social stability and rapid economic development. In 2000, the state began to implement the western development strategy. With the support of national planning guidance, major project construction, capital investment, policies and measures, Gansu Province achieved a historic leap in infrastructure construction. In the strategy of developing the western region, a number of important ecological projects have been carried out, such as returning farmland to forest, returning grazing land to grassland, building shelterbelts in the three northern regions and comprehensively improving the Heihe River and Shiyanghe River basins. In November 2013, the Chinese government adopted the Decision of the Central Committee of the Communist Party of China on Some Major Issues concerning Comprehensively Deepening Reform (2015), which clearly proposed to “promote the construction of the Silk Road Economic Belt and the Maritime Silk Road and form an all-round pattern of opening-up”, thus making the construction of the Silk Road Economic Belt a national strategy. Since ancient times, Gansu Province, especially the Hexi region, has been a must-pass and golden section of the Silk Road. Under the Belt and Road initiative, the regional advantages of the Hexi region have been further highlighted. In 2014, the Provincial Party Committee and Government of Gansu Province formulated the Overall Plan for Gansu Section of the Silk Road Economic Belt in accordance with the strategic orientation of opening up to the outside world. In 2015, the Implementation Plan of Gansu Province’s Participation in the Construction of the Silk Road Economic Belt and the 21st Century Maritime Silk Road (2015) was formulated, and the Ecological Protection and Construction Plan of Gansu Province (2014–2020) was issued in the same year, which plans to build a comprehensive ecological security barrier of “three screens and four zones” within seven years. It aims to protect the ecological construction of the Hexi region and the ecological security of the whole province, the northwest and even the whole country.

4.2. Spatial Correlation between Land Use Structure and Landscape Ecological Risk Structure

Based on the county scale, this study analyzed the area proportions of land use types and the area proportion changes in different risk levels in each district and county during the study period. During the whole study period, the land use type and landscape ecological risk level of each district and county had obvious differentiation characteristics. The area proportions of different land use types and areas of different risk levels in all districts and counties had small changes, indicating that the land use structure and landscape ecological risk structure were relatively stable in the study period (Figure 10 and Figure 11).

According to the area proportions of different land use types, the land use structure of each district and county can be divided into two types: a single land type and a composite land type (

Table 5. Division of dominant types of county land use structure.

Dominant Structure	Dominant Land Type	District/County Name (Abbreviation)	Number	Proportion
Single land type	Unused land	AKS, DH, GT, GUZ, JT, LNZ, MQ, SB, YM	9	45%
Composite land type	Unused land—Cultivated land	GAZ, LIZ, SZ, YC	4	20%
	Grassland—Cultivated land	GL, ML, SD	3	15%
	Unused land—Grassland	JC, JYG, SN	3	15%
	Grassland—Forest land	TZ	1	5%

). It should be noted that when the area of a certain type of land use in each district or county accounts for more than 60% of the total land use area, the land use structure of the district or county is defined as a single-land-type dominant structure. When no single land use occupies more than 60% of the total land use, and the sum of two types of land use occupies more than 60% of the total land use, the land use structure of the district and county is defined as the composite-land-type dominant structure. This method of dividing the land structure is suitable for this study area. It was found that the land use in Hexi county is dominated by a composite dominant structure and secondarily characterized by a single dominant structure. Specifically, there were nine districts and counties dominated by unused land, accounting for 45% of the total area. The secondary area was dominated by unused land and cultivated land, and the number of districts and counties dominated by grassland and forest land was the smallest.

According to the regional area proportions of different risk levels, the landscape ecological risk structure of each district and county can also be divided into two types: a single risk type and a composite risk type. The specific method of division is the same as the land use structure division above. Based on the area proportions of different risk levels, the area proportion threshold divided by the risk structure was adjusted to 50% (

Table 6. Division of dominant types of county landscape ecological risk structure.

Dominant Structure	Dominant Risk Type	District/County Name (Abbreviation)	Number	Proportion
Single risk type	High	GAZ, GL	2	10%
	Medium	GUZ, MQ, SB, YM	4	20%
Composite risk type	High–Medium high	GT, JYG, LNZ	3	15%
	High–Medium	ML, SD, TZ, YC	4	20%
	Medium high–Medium	AKS, DH, JC, JT, LIZ, SZ	6	30%
	Medium–Medium low	SN	1	5%

). County landscape ecological risk in the Hexi region is dominated by a composite-risk-dominant structure and secondarily by a single-risk dominant structure. Specifically, in the single-risk-dominated structure, there are four districts and counties dominated by medium risk and two districts and counties dominated by high risk, accounting for 30% in total. In the composite-risk-dominated structure, the number of districts and counties dominated by medium–high and high risk was the largest, followed by high and high–medium risk, and the number of districts and counties dominated by medium and medium–low risk was the lowest.

The land use structure of each district and county in the study area presents the dominant characteristics of a single to composite land type from west to east, as shown in Figure 12a, and the landscape ecological risk structure presents the dominant characteristics of a composite to single to composite risk type, as shown in Figure 12b.

In districts and counties where land use is dominated by a single land type, the ecological risk structure is dominated by a single risk and secondly by a composite risk. In districts and counties where land use is dominated by composite land types, the ecological risk structure is dominated by a composite risk and secondly by a single risk. In this study, the districts and counties were dominated by grassland and cultivated land composite types, and the ecological risks were dominated by high risk and medium–high risk, including Shandan and Minle. Other districts and counties were dominated by an unused land and cultivated land composite type, and the ecological risk showed a single structure dominated by high risk, mainly including Ganzhou, Gulang, etc. The land use structure of Tianzhu Tibetan Autonomous County is dominated by grassland and forest land, and its ecological risks are dominated by high risk and medium–high risk. Tianzhu Tibetan Autonomous County mainly develops the economy of pastoral areas, which is highly dependent on grassland, and the transformation of traditional animal husbandry has not been completed. The degradation of natural grassland, small scale of artificial herbage cultivation, prominent contradiction between the increase in livestock feedlots and insufficient forage supply [43], overgrazing and overloading even occur in some areas, and the grassland vegetation is damaged by both human and natural factors, leading to a high level of regional ecological risk.

4.3. Ecological Risk Management Measures for Key Areas

Type I areas are mainly distributed in the middle reaches of Heihe River and the Gulang Basin. The population and economic development are concentrated in this region, and the agricultural and industrial activities exert great pressure on the regional ecosystem, leading to the long-term high regional ecological risk. In the ecological risk management of this type of key region, attention should be paid to the coordinated development of urbanization and ecosystem protection. Urbanization and ecosystem protection both restrict and promote each other, forming a contradictory unity. Urbanization causes the destruction and waste of ecological resources due to the unrestrained demand for the ecological environment, and the protection of the ecological environment requires social and economic input, which is the mutual restriction of the two. On the other hand, ecological system’s protection refers essentially to the material elements needed to protect the development of urbanization, and a good ecological environment depends on the material and technical support in the process of urban economic development, which reflects the mutual promotion relationship between the two. In the middle reaches of Heihe River and the Gulang Basin, the starting point of urbanization construction is low and the urbanization development is slow; in addition, the urbanization level is relatively backward on the whole. At the same time, the urbanization development relies heavily on the ecological environment, and the mutual restriction between the two is obvious [44] (Tang et al., 2020). According to the actual situation of this type of region, on the one hand, we should speed up the economic construction of urbanization, promote the continuous improvement of the regional urbanization level and promote the construction of an ecological city with strong economic basic conditions. On the other hand, we need to strengthen ecological and environmental protection. In particular, we need to strengthen the supervision and management of environmental pollution caused by urbanization, control the discharge of pollutants and promote the coordinated development of ecological and environmental protection and urbanization.

Type II areas are mainly distributed in the transition area from mountain to oasis in the upper and middle reaches of Heihe River and the middle and lower reaches of Shiyanghe River. The mountain area is an important water source in the oasis region, which has an important water conservation function. The vertical zone differentiation of vegetation and the characteristics of the natural ecosystem are obvious. As a gathering place for the population and for production activities, the oasis area rises and develops continuously, with strong humanistic and social factors. Therefore, during the transition from mountain to oasis, the dominant factors of the regional system also transition from natural factors to human factors, leading to a rapid increase in the level of regional ecological risk. In the ecological risk management of this type of key area, attention should be paid to the improvement of the water conservation function and the protection and utilization of water resources in the transition region from mountain to oasis. As an important water conservation area, the Qilian Mountain area is of great significance to ecosystem protection in the Hexi region. In the transition area from mountain to oasis, the main land use type change from woodland to grassland and then to cultivated land, and the function and effect of water conservation weakene. In the middle and lower reaches of Shiyanghe River, the human demand for cultivated land, water and other resources, as well as development and utilization, are constantly increasing; coupled with regional drought and low rainfall, evaporation is extremely widespread, leading to a prominent contradiction in regional water resource utilization and ecological system imbalance. According to the actual situation of this type of region, on the one hand, we should strengthen the natural forest protection in the Qilian Mountain area, carry out the project of returning farmland to forest and grassland, accelerate the restoration of vegetation ecology in the transition area from mountain to oasis and improve the water conservation function. On the other hand, it is necessary to ensure the rational use of water resources and improve the utilization efficiency of water resources. Specifically, it is necessary to ensure overall consideration of the upper, middle and lower reaches of the basin; promote water-saving agriculture throughout the basin; introduce water-saving irrigation equipment; and promote a regional water ecological balance and the sustainable utilization of water resources.

With the exception of the overlapping parts of Type I and Type II areas, the remaining key areas of Type III areas are dispersed in the marginal interleaved zones of cultivated land, grassland, construction land and unused land in various districts and counties. The types of land use in this area are adjacent and cross-distributed. Cultivated land and construction land are the most direct embodiments of human surface transformation, and grassland and unused land retain the original landscape state of the surface relatively. Human activities and natural factors are superimposed, and conflicts and mutual energy transformation in the boundary region are intense. In addition, the region has always been in a disorderly unstable state, leading to the increased instability of regional ecological risks. In this type of key area, the ecological risk management should focus on the surface fragmentation and landscape separation caused by encroachment among different land types. The vulnerability of different land use types has different effects on the regional ecological risk level. It is generally considered that the land from high to low risk is unused land, water area, cultivated land, grassland, forest land and construction land. With population growth and urbanization development, on the one hand, the grassland and unused land around cultivated land and construction land are gradually annexed and degraded. On the other hand, land desertification and salinization are caused by overdevelopment. These two aspects always impact regional development, leading to the regional ecological risk level being extremely unstable. According to the actual situation of this type of region, first, it is necessary to optimize the regional land use structure and not blindly expand it, carry out the scientific and reasonable planning of the regional land use structure, allocate land resources reasonably and prevent resource waste and more serious land use problems from causing ecological risks. Second, we should use land intensively and economically, make good use of the existing construction land to improve land use efficiency, strengthen the research and development of agricultural technology to enhance the potential of cultivated land production and protect the existing ecological land while strengthening ecological protection and restoration.

The following global ecological management policies in arid areas are proposed: (1) Strengthen environmental awareness. Human beings must fully realize the importance of ecological protection in arid areas, achieve resource conservation, green environment, and reduce further damage to the ecology of arid areas. (2) Establish a diversified management mechanism. Strengthen the intensification of regional production and living, innovate economic models and technological development, and promote the formation of a diversified management mechanism that integrates economy, society, and ecology. (3) Improve the environment. Scientifically increase vegetation coverage, increase soil moisture content, reduce air pollution, and provide favorable conditions for regional coordinated development. (4) Strengthen scientific research. Different research methods should be adopted for arid areas with different characteristics, and targeted protection measures should be proposed based on the actual situation. In a word, the management and protection of ecosystems in arid areas require joint efforts on a global scale.

5. Conclusions

This study aims to depict the key areas of ecological risk management in the Hexi region. At the same time, the study confirms that the landscape ecological risk in the Hexi region has declined in the recent years; the spatial pattern features of high in the east and low in the west were described.

In the period from 1980 to 2000, the transformation mode of landscape ecological risk was mainly from low risk to medium risk, and secondly from low risk to medium risk. In the period from 2000 to 2020, the landscape ecological transformation mode from medium risk to medium–low risk was most notable, and that from medium–low risk to low risk was secondary. The transformation mode of the landscape ecological risk at different time periods is closely related to land use change and policymaking in the study area. Combined with the state, rate of change and stability of the landscape ecological risk, the key areas of ecological risk management were identified as mainly concentrated in the middle reaches of Heihe River, the Gulang Basin and the middle and lower reaches of Shiyanghe River, and the key areas accounted for 43% of the total area. In the process of ecological risk management, attention should be paid to the coordinated development of urbanization and ecosystem protection, the protection and utilization of key transition zone resources and the landscape fragmentation in the cross-zones of landscape types. The method adopted in this study improves the scientific nature of the research process, and the research results provide an important theoretical basis for future ecosystem management and decision-making in the Hexi region and similar regions.