EN中文

Journal of Resources and Ecology >

2025 , Vol. 16 >Issue 2: 326 - 339

DOI: https://doi.org/10.5814/j.issn.1674-764x.2025.02.004

Ecosystem Quality and Ecosystem Services

Characterization of Spatial-temporal Evolution of Landscape Ecological Risk in the Three-River Source Region

  • LI Hui , 1 ,
  • ZHOU Bin , 2, * ,
  • WU Xiaoying 1
Expand
  • 1. School of Economics and Management, Minzu Normal University of Xingyi, Xingyi, Guizhou 562400, China
  • 2. Department of tourism, Ningbo University, Ningbo, Zhejiang 315211, China
* ZHOU Bin, E-mail:

LI Hui, E-mail:

Received date: 2024-04-20

  Accepted date: 2024-07-30

  Online published: 2025-03-28

Supported by

The National Natural Science Foundation of China(42171223)

The National Social Science Foundation of China(19BJY205)

The Department of Education Youth Science and Technology Talent Growth Fund of Guizhou([2022]107)

The Mountain Discipline Construction Project of Minzu Normal University of Xingyi(XKJS202332)

Fold

Abstract

The Three-River Source Region is an important ecological security barrier in China. Revealing the spatiotemporal evolution characteristics of its landscape types and ecological risks is of great significance for promoting ecological restoration and landscape pattern optimization in the Three-River Source Region. Selecting the Three-River Source Region for a case study and applying the land-use data from four periods (the 1990, 2000, 2010, and 2020), we constructed a landscape ecological risk assessment model for the region based on the landscape pattern index. We then quantitatively assessed the ecological risks and determined the characteristics of their spatial-temporal evolution. The results showed that: (1) The overall landscape ecological risk in the Three- River Source Region tended to decrease from northwest to southeast, and the distribution of landscape ecological risk was closely related to the natural plateau zones and the changes in land cover. (2) From 1990 to 2020, the areas covered by grasslands, water bodies, croplands, and construction land in the Three-River Source Region increased, while the areas of woodlands and unused land decreased. The spatial-temporal changes in the ecological landscape risk were consistent with the characteristics of the changes in the landscape types. The areas categorized as highest, higher, medium, lower and lowest risk areas, while highest and higher risk areas decreased by 9.76%, medium risk areas increased by 1.03%, lower risk areas increased by 8.99%, and lowest risk areas decreased by 0.26%, respectively. (3) Overall, the Three-River Source Region was dominated by very low to medium ecological risk, the areas of which accounted for more than 70% of the entire study area. Overall ecological risks are decreasing, and there is positive spatial autocorrelation of landscape ecological risks in adjacent evaluation units.

Key words： landscape pattern; ecological risk; spatial-temporal evolution; Three-River Source Region

Cite this article

LI Hui , ZHOU Bin , WU Xiaoying . Characterization of Spatial-temporal Evolution of Landscape Ecological Risk in the Three-River Source Region[J]. Journal of Resources and Ecology, 2025 , 16(2) : 326 -339 . DOI: 10.5814/j.issn.1674-764x.2025.02.004

1 Introduction

Located in the Qinghai-Tibet Plateau, Three-River Source Region is a cold plateau with a fragile ecosystem (Wu et al., 2022). The region is an important part of China’s “two (ecological) barriers and three (protection) zones” strategic layout (Piao et al., 2019), having a fragile ecological environment that is sensitive to global climate change (Yao et al., 2012). The ecological protection of Three-River Source Region is vital to ecological civilization construction in China and in terms global ecological and environmental governance. In recent years, climate warming and humidification, modern agricultural development, as well as increases in recreational activities, production scale, and population, have created the people-land contradiction in the Three-Rivers Source Region has become more tense, and ecological risks are increasingly intensifying (Zhang et al., 2022a). Ecological risk assessment is an important basis for decision-making regarding environmental governance and ecological restoration in this region (Suter et al., 2003). However, due to the fragile ecological environment in this area, which is highly responsive to global changes and human activities, it exhibits strong spatial heterogeneity, high landscape fragmentation, and poor ecological stability and resilience. Traditional ecological risk assessment methods are unable to present a comprehensive representation of its ecological risks from multiple sources and the visualization of their spatiotemporal evolution processes. The ecological risk assessment based on the landscape pattern index can more comprehensively characterize the ecological risk and the spatial and temporal evolution of the ecological risk in the context of human activities, and its indicator system is becoming more and more perfect, and the method is scientific and feasible (Yang et al., 2024). Therefore, it is of great significance to evaluate its ecological risks based on a landscape ecology perspective (Peng et al., 2015).
Ecological risk assessment research emerged in the 1970s as an important field of ecosystem assessment, and it has become a hot topic in disciplines such as geography and ecology (Wei et al., 2023). Early ecological risk research underwent a transition from focusing on single risk factors to considering multiple sources of risk factors. Initially, the focus was on specific risk sources such as chemical pollution agricultural pollution diffusion, and soil erosion (Hope, 2006). For example, Foster et al. (1991) provides a discussion of the mechanisms by which the use of pesticides in agriculture contaminates groundwater pollution and compares this problem to the problem of nitrate contamination in agricultural land use practices. But it later shifted to considering various factors such as climate change (Lazo et al., 2000), natural disasters (Zhang et al., 2009), and human activities (Nunneri et al., 2008). Large-scale, comprehensive, multidimensional ecological risk source assessment has become a research hotspot (Hayes and Landis, 2004). For example, Yu et al. (2016) comprehensively evaluated the land ecological risk in Sheyang County, Jiangsu Province, in terms of the probability of occurrence of multi-source risk, the resistance and resilience of land ecosystems, and the damage that may be caused by exposing land ecosystems to risky environments. In terms of risk receptors, the focus has shifted from solely considering human health (Suter, 1997) to focusing on the overall regional biological community (Critto et al., 2005) and ecosystem (Wallack and Hope, 2002). The most common research methods include the Minimum Cumulative Resistance model (Zhang et al., 2017), Morphological Spatial Pattern Analysis (MSPA) (Rogan et al., 2016), and Principal Component Analysis (Zhang et al., 2016). For example, Zhou et al. used the least cumulative resistance model and gravity model to construct the regional ecological security pattern of Xi’an City (Zhou et al., 2023). However, the traditional ecological risk assessment is not suitable for the regional ecological security pattern under the combined effect of human activities and natural factors. However, the traditional ecological risk assessment pays insufficient attention to the regional ecological risk changes under the combined effects of human activities and natural factors and the spatial heterogeneity of regional risks (Mei et al., 2024). Therefore, some scholars have suggested that landscape ecology should be included in the evaluation system, and human-land interactions should be considered comprehensively (Deng et al., 2024). Focusing on emphasizing the scale effect of the research object and spatial heterogeneity (Liu et al., 2023).
In 1983, German geographer Carl Troll first proposed the concept of landscape when interpreting remote sensing images of East Africa (Fu, 1983). In 1990, Carolyn further introduced landscape ecology concepts such as disturbance scenarios, assessment boundaries, and spatial heterogeneity of landscapes into ecological risk assessment, providing a new perspective for the study of landscape pattern-ecological processes and ecological risk research (Peng et al., 2015). Landscape ecological risk assessment has received widespread attention.
Compared to traditional ecological risk assessment, landscape ecological risk assessment emphasizes the determination and prediction of the degree of impact of human activities or natural disasters on specific landscape components, structures, functions, and processes through the embedding of landscape elements, the evolution of landscape patterns, and landscape ecological processes. It focuses on using land use patterns as a comprehensive risk entity to reveal the interference and impact of human activities on ecosystem functions and structures (Liu et al., 2012). From the perspective of research methods, landscape ecological risk assessment has undergone a transition from being based on the source-sink method to being based on landscape pattern index methods. Early landscape ecological risk assessments mainly used the source-sink analysis method, first identifying risk sources and receptors, then conducting exposure and hazard analysis, and finally dividing ecological risk levels (Yu et al., 2016). However, with further research, the high correlation between the spatial-temporal distribution of landscape ecological risks and changes in land use patterns has been confirmed. The viewpoint that risk receptors come from the ecosystem itself has become the consensus of most scholars, and the comprehensive risk assessment method based on landscape patterns has been widely applied (Mei et al., 2024). In terms of evaluation model construction, there are currently three main types: landscape pattern index models based on land use changes, external pressure-exposure-stability models, and natural-human society-landscape pattern models (Li et al., 2019). Among them, landscape pattern index models based on land use changes focus on characterizing ecological risks through the resistance ability, vulnerability, and area changes of landscape types themselves (Li et al., 2017). The external pressure-exposure-stability model focuses on characterizing ecological risks through the external forcing and internal pressure capacity of the landscape (Li et al., 2008). The natural-human society-landscape pattern model focuses on incorporating various natural and human social disturbance factors into the ecological risk assessment system (Li et al., 2019). Due to the spatial proximity impact relationship of landscape ecological risk and the spatial heterogeneity of landscape ecological risk, landscape pattern index models are the most common in landscape ecological risk assessment.
In terms of research subjects, the current hotspots in landscape ecological risk assessment research are basically the same as those in regional ecological risk assessment, mainly focusing on ecologically sensitive areas such as watersheds (Wang et al., 2010; Pan and Liu, 2016), industrial and mining areas (Bayliss et al., 2012; Wu et al., 2013), and nature reserves (Wang et al., 2021). For example, Chen and Pan (2003) evaluated the ecological risk of the Sangong River Basin in Fukang, Xinjiang, based on the Landscape Loss Index and the Comprehensive Risk Index, and found that its ecological risk exists in three high landscape ecological risk areas, and revealed the direction of evolution of its localized environments, which provides a reference for the analysis of the regional landscape pattern as well as the evaluation of the regional ecological risk. Furthermore, due to the flexibility of landscape pattern index models in terms of evaluation scale, they are better able to reflect the heterogeneity of geographic spatial units within a region. Therefore, the results of small-scale regional landscape ecological risk assessments are often an important supplement to large-scale regional ecological risk assessments. Wang et al. (2022) analyzed the overall ecological risk and influencing factors of the Qinghai-Tibet Plateau, while Sun et al. (2023b) further improved the accuracy of land classification using an optimized PLUS model, confirming most of Wang et al.’s research conclusions. However, there are differences in the results of ecological risk assessments in the western part of the Qinghai-Tibet Plateau, and neither study assessed the landscape ecological risk of each natural geographic region, so the spatial heterogeneity of ecological risk needs to be confirmed.
The Three-River Source Region is one of the ecologically sensitive and fragile areas in the global plateau range, and it is also an important barrier for the ecological and environmental security of our country. Analyzing the temporal and spatial evolution characteristics of its landscape ecological risks scientifically is of great significance for strengthening the ecological security barrier of the Three-River Source Region, maintaining the health of the Qinghai-Tibet Plateau ecosystem, and promoting the sustainable development of our country’s ecology. As vegetation coverage is an important indicator for measuring ecological risk (Zhang and Ning, 2023), scholars in early ecological risk assessments mainly focused on vegetation changes and their driving forces in the Three-Rivers Source Region. Liu et al. (2008) studied vegetation changes in the Three-Rivers Source Region from 1970 to 2004 and found that grassland degradation continued to occur during this period, but at a slow pace and with significant regional differences. Liu et al.’s research showed that, thanks to changes in warm and humid climate and the implementation of ecological conservation plans, the improvement in vegetation area in the Three- Rivers Source Region from 2000 to 2011 exceeded the degradation area, resulting in an overall improvement in the ecological environment (Liu et al., 2014). Zhang and Ning’s research further confirmed this conclusion, showing that vegetation coverage in the Three-Rivers Source Region has significantly increased over the past 20 years, and temperature, precipitation, altitude, slope, and land use systems all have an impact on vegetation recovery (Zhang and Ning, 2023). In addition, scholars have also discussed ecological risks such as soil erosion (Wang et al., 2017), grassland degradation (Zhang et al., 2019), snowstorm disasters (Shao et al., 2019), soil heavy metal pollution (Zhou et al., 2021), and water pollution (Li et al., 2022) in the Three-Rivers Source Region. However, overall, there are relatively few studies based on the perspective of landscape ecology, with only Guo et al introducing Shannon diversity index and Shannon evenness index to study the ecological environment quality of the Three-Rivers Source Region (Guo et al., 2016). Zhang et al used a pressure-state-response model and introduced landscape pattern vulnerability index and landscape functional vulnerability index to evaluate the land ecological security of the Three-Rivers Source National Park (Zhang et al., 2022a). From the existing research literature on landscape ecological risk evaluation, it is not difficult to find that the existing research either focuses on the analysis of single-source ecological risk in the Three- River Source Region, or introduces only part of the landscape pattern indices for the evaluation of its ecological risk, and lacks the long-term and systematic ecological risk assessment based on the perspective of landscape ecology, which can not adequately reflect the spatial heterogeneity of its ecological risk and its spatial-temporal evolutionary characteristics.
In view of this, taking the Three-Rivers Source Region as the research object, combining its geographical development process with the ecological development process, using remote sensing images from 1990, 2000, 2010, and 2020 as the main data sources, introducing indices such as landscape fragmentation, landscape isolation, and landscape dominance, systematically constructing a landscape ecological risk assessment model to evaluate the spatiotemporal distribution and evolution characteristics of landscape ecological risks in the Three Rivers Source Region. It can provide a certain scientific basis for the ecological risk control, land zoning management, ecological restoration, and policy formulation to alleviate the tension between humans and the environment in the Three-Rivers Source Region. On the other hand, it helps to detail the spatial heterogeneity differences of landscape ecological risks in the hinterland of the Qinghai-Tibet Plateau, which is an important supplement to the research on ecological risk assessment in the Qinghai-Tibet Plateau. At the same time, it helps to enrich the theoretical achievements of landscape ecological risk assessment, verify the feasibility and special advantages of landscape pattern theory and spatial heterogeneity theory in regional ecological risk assessment and provide empirical cases for landscape ecological risk assessment in the high-altitude and cold-sensitive areas of the plateau.

2 Materials and methods

2.1 Study area

As is shown in Figure 1, the Three-River Source Region is located in the south of Qinghai Province, with an average altitude of more than 4000 m (Sun et al., 2023a). The headwaters of the Yangtze River, Yellow River, and Lantsang River are located in the region, providing crucial replenishment of China’s freshwater resources (Huang et al., 2020). The western and northern parts of the area are mainly mountains and plains, with many bottomlands and marshes, whereas the southeastern part contains many high mountains and valleys with steep slopes. The climate is classified as a plateau continental climate, transitioning from subtropical to cool-temperate, semiarid, extremely arid, and cold. Influenced by the topography and climate, the vegetation demonstrates horizontal and vertical zonation and rich biodiversity (Huang et al., 2020). The region serves as a value plateau biological germplasm resource and ecological barrier in China, providing special ecological service functions and ecological value (Zhang et al., 2023). In recent years, affected by global warming, urbanization, overgrazing, and other factors (Wang et al., 2004), the ecology of the region has been facing threats such as reduced water resources, desertification, and reduced biodiversity (Shao et al., 2010). The fragile natural environment and human-land conflicts have intensified the decreasing ecological stability of the Three-River Source Region, rendering the ecosystem extremely vulnerable (Wang et al., 2023).
Figure 1 Location of the Three-River Source Region

Note: Henan Mongolian A. C. in the figure represents Henan Mongolian Autonomous County.

2.2 Study methodology

2.2.1 Division of landscape risk assessment units

With the fishnet tool in ArcGIS 10.8 software, a 1.5 km×1.5 km grid was used to divide the Three-River Source Region into 1887 landscape risk assessment units. Using Fragstats 4.2 software, the number of patches and the patch areas of each landscape type within each unit were determined.

2.2.2 Construction of landscape ecological risk assessment index (ERI)

Based on the empirical relationship between land-use patterns and the ecological risks established in prior studies (Zhou et al., 2020), the following landscape indices were selected to construct a landscape ecological risk assessment model for the Three-River Source Region: fragmentation, segmentation, dominance, disturbance, vulnerability, and loss. The landscape types in the Three-River Source Region were transformed into spatialized ecological risk variables using sampling methods, and the landscape ERI was calculated for 1887 landscape risk assessment units using the formulas shown in Table 1.
Table 1 Landscape pattern indices and their ecological definitions
Landscape pattern index Formula Description
Landscape fragmentation index (Ci) ${{C}_{i}}={{n}_{i}}/{{A}_{i}}$ Ci denotes the fragmentation index of landscape type i, with larger values representing less internal landscape stability; ni indicates the number of patches of landscape type i; Ai denotes the area of landscape type i (Zhang et al., 2018)
Landscape segmentation index (Ni) ${{N}_{i}}=\sqrt{\frac{{{n}_{i}}}{A}}\times \frac{A}{2{{A}_{i}}}$ Ni indicates the degree of spatial segmentation of patches of landscape type i, with larger values representing a more complex spatial distribution of various patches of the landscape; ni indicates the number of patches for the type of landscape i; A denotes the total area of the entire landscape; Ai denotes the area of landscape type i (Qu et al., 2022)
Landscape dominance index (Fi) ${{F}_{i}}=\frac{{{M}_{i}}+{{L}_{i}}+{{P}_{i}}}{3}$ Fi denotes the degree of landscape dominance, with larger values representing higher importance of the patch in the landscape; Mi denotes the patch frequency; Li denotes the patch density; Pi denotes the proportion of patches (Xie, 2008);
Landscape disturbance index (Ei) ${{E}_{i}}=a{{C}_{i}}$+$b{{N}_{i}}$+$c{{F}_{i}}$ Ei denotes the degree of disturbance of regional ecosystems; a, b, c are the weights of the landscape fragmentation index, landscape segmentation index, and landscape dominance index, respectively,$a+b+c=1$
The values of 0.5, 0.3, and 0.2 were assigned based on the results of prior studies and the actual situation in the study area (Xie, 2013);
Landscape vulnerability (Di) Unused lands = 6; water bodies = 5;
croplands= 4; grasslands = 3; woodlands = 2; construction lands = 1; normalized		 Di denotes the resistance of different landscapes to external interference;
Landscape vulnerability indicates the ability of different landscape types to resist external risks and disturbances. Larger values represent increased sensitivity and less resistance to disturbances (Qu et al., 2022)
Landscape loss index (Ri) ${{R}_{i}}=\sqrt{{{E}_{i}}/{{D}_{i}}}$ Ri indicates the landscape loss index (Zhang et al., 2018)
Landscape ecological risk index (ERIk) $ER{{I}_{k}}=\underset{i=1}{\overset{n}{\mathop \sum }}\,\frac{{{A}_{ki}}}{{{A}_{k}}}\times {{R}_{i}}$ ERIk denotes the landscape ERI of the k-th assessment unit; n is the number of landscape types; Aki is the area of the landscape type i in the risk
communities of risk k; and Ak is the total area of the risk communities (Zhang et al., 2018)

2.3 Data sources

The land cover data of the Three-River Source Region were obtained from the Institute of Geographic Sciences and Natural Resources Research, CAS, including four periods of data in the 1990, 2000, 2010, and 2020, with a spatial resolution of 30 m×30 m. In this study, the land cover of the whole study area was reclassified into six types: croplands, woodlands, grassland, water bodies, construction land, and unused lands referring to the Level-1 Land Use and Cover Change (LUCC) Standard for Land-use Classification of the Resource and Environment Science and Data Center of the Chinese Academy of Sciences. The basic data of administrative divisions are sourced from the National Catalogue Service for Geographic Information. The boundary vector data of the Three-River Source Region were obtained from the National Tibetan Plateau Data Center.

3 Results and analysis

3.1 Landscape type changes in the Three-River Source Region

The land use in the Three-River Source Region from 1990 to 2020 was classified into six types: grasslands, croplands, construction lands, woodlands, water bodies, and unused lands. Grasslands and unused lands were the main landscape types, accounting for more than 89% of the entire study area. These were followed by woodlands and water bodies, accounting for approximately 10% of the entire study area. Cropland and construction land areas were small, accounting for less than 1% of the entire study area. In terms of the temporal changes in landscape types, the intensity of land-use changes during the study period followed the sequence of construction lands > unused lands > croplands > grasslands > water bodies > woodlands. The proportions of grasslands, water bodies, and construction lands showed an overall upward trend, with increases of 5.7%, 5.1%, and 70.9%, respectively; the proportions of woodlands and unused lands trended downward overall, with decreases of 0.47% and 19.3% respectively; and cropland area was characterized by an increase and then a decrease, with an overall increase of 14.7% during the study period (Figure 2). These changes in landscape types reflect the expansion of construction land in the Three-River Source Region. Tourism and urban development drove such changes; however, owing to the establishment of the Three-River Source Nature Reserve and the Three-River Source National Park and the state’s emphasis on ecological protection, desertification in the Three-River Source Region has been controlled, the area of unused land has decreased, and grassland and water body areas have substantially increased, highlighting the impacts of human resource development activities and policy interventions on the ecological environment in Three-River Source Region (Zhang et al., 2018).
Figure 2 Land use types of the Three-River Source Region from 1990 to 2020

3.2 Characteristics of spatial-temporal shifts in landscape types in the Three-River Source Region

Regarding the shifts in the land-use types in Three-River Source Region, the total amount of shifts in the different land-use types from 1990 to 2000 was low; only the conversion from grasslands to croplands and unused lands was relatively considerable. A total of 120.09 km² of grassland was reclaimed as cropland, and 246.84 km² of grassland degraded into unused land such as bare, sandy, and saline-alkali land. From 2000 to 2010, owing to the establishment of Three-River Source Nature Reserve and the implementation of the first-stage ecological conservation and restoration project in the Three-River Source Region, the grassland and water body areas of the Three-River Source Region increased by 20570.23 km² and 1229.67 km², respectively (Table 2). The expansion of grasslands mainly occurred from the conversion of cropland, woodland, and unused land; the expansion of lakes and wetlands mainly occurred from the conversion of unused land and grassland. This finding indirectly indicates that, during this period, the Three-River Source Region has achieved remarkable results in converting farmland back to grassland, desertification management, and restoration of wetland and water ecosystems. The growth in construction land was fastest in the Three-River Source Region from 2010 to 2020, and the most notable feature of the land-use change was the transformation of grassland to unused and construction land. This finding demonstrates the increased need for construction land in Three-River Source Region due to the need for socioeconomic development. Additionally, this indicates that the problem of grassland desertification in the region has not yet been fundamentally resolved. Overall, the land-use changes in the Three-River Source Region from 1990 to 2020 reflect both the impact of policy interventions on environmental protection (Zhang et al., 2018) and the evolution of the basic geographical conditions, socioeconomic factors, and ecological protection policies that have jointly acted on the landscape types in the Three-River Source Region (Niu et al., 2022).
Table 2 Transfer of land use types in the Three-River Source Region from 1990 to 2020 (unit: km2)
Transfer of land use types 1990-2000 2000-2010 2010-2020
Croplands→Woodlands 0.03 0.92 2.97
Croplands→Grasslands 1.15 118.96 54.19
Croplands→Water bodies 12.52 18.97 10.71
Croplands→Construction lands 4.80 5.99 9.49
Croplands→Unused lands 5.92 0.66 3.65
Woodlands→Croplands 7.30 0.91 3.69
Woodlands→Grasslands 38.21 271.76 768.03
Woodlands→Water bodies 5.81 7.83 10.06
Woodlands→Construction lands 0.04 0.62 1.51
Woodlands→Unused lands 0.25 9.16 15.37
Grasslands→Croplands 120.09 348.34 53.41
Grasslands→Woodlands 50.94 203.14 771.93
Grasslands→Water bodies 109.40 276.82 539.95
Grasslands→Construction lands 5.47 24.35 71.47
Grasslands→Unused lands 246.84 4487.29 1183.75
Water bodies→Croplands 6.69 4.09 8.75
Water bodies→Woodlands 0.46 2.10 7.51
Water bodies→Grasslands 133.72 101.05 591.62
Water bodies→Construction lands 0.10 0.08 2.32
Water bodies→Unused lands 114.13 80.35 239.33
Construction lands→Croplands 0.04 1.50 7.01
Construction lands→Woodlands 0.00 0.08 0.42
Construction lands→Grasslands 0.02 1.14 15.49
Construction lands→Water bodies 0.00 1.52 0.74
Construction lands→Unused lands 0.00 0.05 0.38
Unused lands→Woodlands 0.50 24.89 17.22
Unused lands→Grasslands 146.85 20077.32 1165.47
Unused lands→Water bodies 49.55 924.53 332.27
Unused lands→Construction lands 0.02 3.34 6.88
Unused lands→Croplands 6.72 7.04 4.38

3.3 Characteristics of landscape ecological risk distribution in the Three-River Source Region

The landscape fragmentation, segmentation, dominance, disturbance, vulnerability, and loss indices of each type of land use in the Three-River Source Region from 1990 to 2020 were measured using the landscape ERI model. The ERI of the 1887 evaluation units in the Three-River Source Region were derived. The spatial interpolation of the landscape ERI was symbolically presented using the ordinary kriging interpolation method. The 1990 landscape ecological risk results were categorized into five classes based on the natural breaks (Jenks) classification as lowest risk (ERI≤0.12), lower risk (0.12<ERI≤0.135), medium risk (0.135<ERI≤0.15), higher risk (0.15<ERI≤0.165), and highest risk (0.165<ERI). To facilitate comparison, this division was used as a reference to obtain the 1990-2020 4-phase landscape ecological risk spatial distribution map of the Three-River Source Region (Figure 3). The statistics on the areas and proportions occupied by different risk levels are shown in Table 3.
Figure 3 ERI Distribution in the Three-River Source Region
Table 3 Landscape ecological risk area and proportion in the Three-River Source Region from 1990 to 2020
Year Lowest risk Lower risk Medium risk Higher risk Highest risk
Area
(103 km2)		 Proportion
(%)		 Area
(103 km2)		 Proportion
(%)		 Area
(103 km2)		 Proportion
(%)		 Area
(103 km2)		 Proportion
(%)		 Area
(103 km2)		 Proportion
(%)
1990 1.58 0.41% 104.57 27.35 174.93 45.74 71.22 18.62 30.12 7.88
2000 1.77 0.46% 106.93 27.96 172.76 45.18 71.29 18.64 29.67 7.76
2010 0.23 0.06% 137.38 35.92 180.12 47.10 55.46 14.50 9.23 2.41
2020 0.57 0.15% 138.97 36.34 178.86 46.77 55.71 14.57 8.31 2.17
Overall, the landscape ecological risk in the Three-River Source Region during the study period was dominated by low- and medium-risk areas, which accounted for more than 70% of the entire study area. This was followed by higher- and highest risk areas, accounting for more than 15% of the entire study area; lowest risk areas accounted for a small proportion of the entire study area. Owing to the establishment of multifaceted ecological management systems such as national nature reserves, national parks, and compensation programs for ecological protection, the average values of the landscape ecological risk in 1990, 2000, 2010, and 2020 were 0.1438, 0.1437, 0.1406, and 0.1398, respectively, showing an overall decline in ecological risk and an overall improvement in the ecological environment.
The landscape ecological risk of the Three-River Source Region demonstrated a general pattern of high-risk values in the northwest and relatively low values in the southeast. The high-risk areas mainly concentrated in the northwestern counties of Zeku and Qumalai, distributed along the Kunlun and Hoh Xil Mountains. The region is characterized by high altitude, low temperature, extensive bare land, and poor conditions for plant growth and survival, mainly consisting of alpine meadows and grasslands, which are highly ecologically vulnerable (Xia et al., 2021). In addition, the region is characterized by many highland lakes, swamps, snow-capped mountains, and glaciers; the levels of landscape fragmentation, segmentation, and disturbance are high, leading to higher ecological risks. The low-risk areas mainly concentrated in Zaduo, Nangqian, and Yushu counties in the south and in Banma, Jiuzhi, Gande, and Maqin counties and Henan Mengonlian A. C. in the southeast. The region hosts many rivers and valleys, with a low elevation, with warm and humid airflow entering along the valleys. The climate is warm and humid, and vegetation coverage is high. The degree of landscape dominance is high, resistance to disturbances is strong, and the ecosystem vulnerability is low (Huang et al., 2022).

3.4 Characteristics of spatio-temporal changes in landscape ecological risk in the Three-River Source Region

In terms of the temporal changes in the landscape ecological risk in the Three-River Source Region, the average value of ecological risk decreased from 0.1438 to 0.1398 from 1990 to 2020, which is a decrease of 2.8%. The distribution of landscape ecological risk is consistent with the changes and shifts in landscape types. The total amount of shifts in different landscape types was low in 1990-2000, and the changes in the area and proportion of area of the different classes of ecological risks were minimal. The shift of grassland to unused land and the degradation of grassland led to an increase in the proportion of the high-risk areas. The total amount of shifts for the different landscape types was larger in 2000 vulnerability 2010, and the changes in ecological risk were also the largest. The implementations of ecological protection and construction projects have curbed the trend in ecological degradation (Zhang et al., 2022b), and the substantial increases in the grassland and water body areas in the northern part of Qumalai County and the northern part of Maduo County have led to a combined decrease of 9.49% in the proportions of highest risk and high-risk areas, thereby substantially reducing the ecological risk. The proportions of lowest risk and low-risk areas in the Three-River Source Region increased by 0.51% from 2010 to 2020, which were mainly transformed from the surrounding medium- and highest risk areas. More transformation occurred from adjacent risk classes than from across different classes. However, the shift of grassland to construction and unused land during this phase resulted in an increase in the proportion of high-risk areas.

3.5 Spatial autocorrelation analysis of landscape ecological risks in the Three-River Source Region

The results of spatial autocorrelation analysis revealed the interactional relationships between the reference spatial units and the adjacent units by determining the existence of a significant correlation between the attribute values of a certain parameter and the attribute values of spatially adjacent units. This analysis included both the global and local Moran’s I indices. A positive index value indicates a positive correlation between the two, and the correlation strengthens as the value increases (Fan et al., 2016).

3.5.1 Global autocorrelation analysis of ecological risks

GeoDa software was used to calculate the global Moran’s I index and draw the scatter plot of the landscape ecological risk in the Three-River Source Region from 1990 to 2020 (Figure 4). The results showed that Moran’s I index was 0.632, 0.636, 0.457, and 0.576 in 1990, 2000, 2010, and 2020, respectively, in the Three-River Source Region. A spatially positive correlation was found between the landscape ERIs in the Three-River Source Region, and adjacent regions showed a high degree of spatial correlation and a high degree of similarity in landscape ecological risk, showing an aggregation effect.
Figure 4 Moran’s I of ERI in the Three-River Source Region

3.5.2 Local autocorrelation analysis of ecological risks

To further portray the specific spatial correlations of landscape ecological risks in the Three-River Source Region, LISA aggregation maps showing local autocorrelation were produced through local autocorrelation analysis (Figure 5). The landscape ERI displayed strong “High-High” and “Low-Low” types of aggregation in the Three-River Source Region from 1990 to 2020. The “High-High” type of aggregation accounted for approximately 18.12% of all study units in 2020. Distributed in Zeku and Qumalai counties in the northern part of the study area, these areas of high landscape ecological risk also have higher levels of landscape ecological risk in their adjacent areas. The “Low-Low” type of aggregation accounted for approximately 20.61% of all study units, being mainly distributed in Nangqian and Yushu counties in the southern part and Banma, Jiuzhi, Gande, and Maqin counties in the southeastern part of the study area. For these areas with low landscape ecological risk, the adjacent areas were also characterized by low levels of ecological risk.
Figure 5 LISA distribution of ERI in the Three-River Source Region
In terms of temporal changes, the “Low-Low” and “High-High” types of aggregation showed minimal changes from 1990 to 2000. From 2000 to 2010, the “Low-Low” and “High-High” types of aggregation of landscape ecological risk tended to be concentrated; however, both types of aggregation of landscape ecological risk in 2010-2020 experienced a slight spread. Comparing the LISA aggregation map showing the local autocorrelation (Figure 5) and the spatial distribution map of landscape ecological risk (Figure 3), we found that the “High-High” type, with spatially localized auto-correlated aggregated areas, corresponded well to the high-risk and highest risk areas in the landscape ecological risk distribution map; the main landscape types were unused and construction land. The internal landscape stability of these areas as lower due to their more dispersed distribution and the higher level of human activities. The area of the “Low-Low” type of aggregation was consistent with the areas with very low and lower risk in the landscape ecological risk spatial distribution map. The main landscape types were grasslands and woodlands, which are featured by high vegetation coverage and strong resistance to disturbances.

4 Discussion and conclusions

4.1 Discussion

(1) The pattern of the landscape ecological risk distribution was strongly related to the natural zones of the plateau and land-cover changes. At higher elevations, as heat and moisture decrease, vegetation cover often transitions from dense mountain forests to alpine shrub meadows, alpine grasslands, and alpine deserts. Landscape fragmentation gradually increases, connectivity deteriorates, and interactions between different landscape types are inhibited, leading to increased spatial heterogeneity and landscape vulnerability, and consequently to higher landscape ecological risk. In contrast, the landscape ecological risk in lower elevation areas was relatively lower due to the favorable climatic conditions, high-density vegetation cover, and strong resistance to disturbances. This conclusion is consistent with those of Gao et al. (2021) for the Sichuan-Yunnan ecological barrier area but differs from those of Zheng et al. (2022) on the landscape ecological risk in the mountainous areas of the Yunnan borders. The probable reason for this difference is that the areas of cropland and construction land in the Three-River Source Region are relatively small, and natural factors are the main factors influencing its ecological risks. The Gaoligong Mountains, located in the border area of Yunnan, are relatively low in altitude; after becoming a national nature reserve, human activities have been reduced in the area. Compared with construction land, cropland, and rocky desertification areas in the Three-River Source Region, the ecosystems in the mountainous areas are, conversely, relatively stable, and the ecological risks are relatively low. Overall, landscape ecological risk assessment is a complex systematic project, and regional landscape ecological risk is influenced by both natural and human factors (Liang and Song, 2022). Because human activities are intertwined with social perceptions, cultural integration, and other unknown factors, they more strongly interfere with the environment than changes in natural landscape types. Therefore, using the landscape type as the main indicator of ecological impact inevitably introduces bias. In the future, a more accurate system of evaluation indices needs to be established.
(2) Landscape ecological risk is an important indicator of changes in landscape use types. The 30 m×30 m data used in this study provide a value supplement to the data from large-scale studies as they can portray the characteristics of spatial-temporal shifts in landscape types and ecological risk distribution of localized areas in more detail. The landscape types in the Three-River Source Region most widely changed from 2000 to 2010, with grassland and water bodies increasing by 5.9% and 5.4%, respectively; the maximum change in its ecological risk was reached in this period, with a cumulative decrease of 9.49% compared with the high-risk and highest risk areas. The spatial-temporal changes in the landscape ecological risk aligned with the changes in landscape types. Specifically, the landscape loss indices of construction land, unused land, and croplands were higher, and the ecological risk they produced was also relatively high. The landscape loss indices of woodlands, grasslands, and water bodies were lower; the ecological risk in the regions was also relatively low. This conclusion is consistent with those from a study on the ecological risk of the Yellow River Basin by Du et al. (2022). However, this finding is different from that of a lower landscape loss index of construction land in a study of Daiyue District, Tai’an City, by Shi et al. (2013). A possible reason for this is that the construction land in the mountainous areas of the plateau showed patchy fragmentation and increased segmentation compared with the plains, which led to a higher landscape loss index. At the same time, the conclusion differs from that the ecological risk of urban construction land landscape is low in coastal cities proposed by Li et al. (2023). The possible reason is that urban construction land in coastal areas has the characteristics of concentrated patches, uniform and regular shapes, and relatively stable conditions, thus resulting in low ecological risks. Overall, the sample in this study on landscape ecological risk assessment was small. In the future, case sites of different geographic regions and types should be compared and generalized to further clarify the inherent relationships between landscape type and landscape ecological risk.
(3) A positive spatial correlation was found between the landscape ecological risks in the Three-River Source Region, and the spatial distribution of the aggregation of risk was essentially consistent with the spatial distribution of the level of ecological risk. This conclusion is consistent with the findings of studies on the coastal zones of Guangdong (Cheng et al., 2022) and Tianjin (Xiao and Tian, 2014), but the causes of ecological risk level transformation differ. Human activities in Three-River Source Region are limited, and the altitude widely differs in the area, so the area is affected by multiple factors such as topography, water, heat, and soil. Its ecosystem displays vertical zonation, and the transformation between landscape ecological risk levels was triggered more by the vertical gradient differences caused by the changes in hydrothermal conditions. In contrast, the transformation between landscape ecological risk levels at sites such as Guangdong and Tianjin was more often triggered by human activities such as port, city, and town construction, as well as industrial and agricultural production. In the future, a comprehensive ecological risk prediction system considering topographical, climatic, and socio-economic conditions should be established based on the correlation and transformation of landscape ecological risks to provide a more scientifically robust theoretical basis for regional ecological risk prevention and control.
Overall, the fragmentation of landscapes in the Three- River Source Region is high, with strong spatial heterogeneity and fragility, and poor resistance to disturbance. Therefore, it is necessary to strengthen the overall protection of river sources, enhance the construction of important ecological corridors, and improve the connectivity and disturbance resistance of landscapes. Compared to low-altitude areas, although human activities are less intense in the Three-River Source Region, their impact is significant. The landscape loss index of construction land, unused land, and cultivated land is relatively high, resulting in higher ecological risks. It is important to coordinate economic and social development with ecological environmental protection, control the disorderly expansion of construction land, promote returning cultivated land to forests and grasslands, returning pastures to grasslands, strengthen wetland protection and restoration, and achieve harmonious coexistence between humans and nature.

4.2 Conclusions

(1) In the studied 30-year period, the landscape ecological risk in the Three-River Source Region was dominated by low- and medium-risk areas, accounting for more than 70% of the entire region. The landscape ecological risk continuously trended downward overall, and the change was relatively drastic more recently. The spatial-temporal changes in landscape ecological risk have been consistent with the spatial-temporal changes and shifts in the landscape types.
(2) The overall landscape ecological risk of the Three- River Source Region trended downward from northwest to southeast, which was closely related to the natural plateau and changes in land cover. The high-risk areas were mainly distributed along the Kunlun and Hoh Xil Mountains, which are featured by high altitudes, poor plant growth environments, and higher levels of landscape fragmentation, segmentation, and disturbance. Low-risk areas were mainly located along river valleys which are characterized by lower elevations, warm and humid climates, dense vegetation cover, and a high degree of landscape dominance.
(3) The landscape ecological risk in the Three-River Source Region showed strong positive autocorrelation, and the levels of landscape ecological risk in spatially adjacent regions were similar. An aggregation of landscape ecological risk was found, with high-risk aggregations mainly located in high-altitude mountainous areas in the northwest and low-risk clusters mainly located in river valleys in the southeast.

4.3 Practical implications

Based on the above landscape ecological risk assessment results, tailored ecological risk prevention strategies should be formulated according to the characteristics of different risk zones in the Three-Rivers Source Region, taking into account local environmental conditions and social development.
(1) The high-risk and moderate-risk areas in the northwest of the Three-River Source Region have few residential areas, and their ecological risks mainly stem from their fragile natural ecological background. This region has a high altitude, a large amount of bare land, and a wide distribution of plateau lakes. The degree of landscape fragmentation is high, the fragility is strong, and the connectivity is poor, making it greatly affected by global climate change. On one hand, efforts should be made to promote carbon neutrality, actively adapt to and mitigate climate change, minimize the ecological disturbances caused by climate warming, maintain the function of the “solid water reservoir” of snow-capped mountains and frozen soil, and enhance the adaptability of ecosystems to climate change (Tang et al., 2022). At the same time, further research on the mechanisms of climate change and risk assessment of extreme meteorological disasters should be conducted, with a focus on preventing and controlling disasters such as hydraulic erosion, wind erosion, and freeze-thaw erosion. On the other hand, active ecological restoration should be carried out, strengthening the management of sandy areas, bare land, and bare rock gravel areas, and controlling soil erosion and desertification.
(2) The central high-risk areas of the Three-Rivers Source are greatly affected by human activities, and the conversion of grassland into unused land and construction land is particularly evident. Population growth, overgrazing, and grassland degradation are the main problems. It is necessary to further deepen the national park protection system, continue to implement the Three-Rivers Source Ecological Protection and Construction Project, strengthen ecological management measures such as returning grazing land to grassland, returning cultivated land to forests, treating degraded grassland, managing black soil beaches, controlling grassland rodent pests, and regulating construction land, based on the priority designation of arable land, permanent basic farmland, and ecological protection red lines.
(3) The southeastern part of the Three-River Source Region, which has low and moderate ecological risks, is densely populated due to its lower altitude, gentle slopes, abundant water resources, and dense vegetation cover. The expansion of urban land use is the main challenge for its ecological environment. It is necessary to further conform to the natural geographical pattern, establish effective land spatial planning, define urban development boundaries, control the total amount of urban construction land, and optimize the urban-rural spatial pattern (Yu et al., 2021). Overall, the Three-River Source region should develop differentiated ecological optimization and governance strategies based on the contradictions between humans and the environment in different regions and stages of development, in order to promote the stability and sustainable development of the Three-River Source Region ecosystem.

5 Limitation and future research

(1) Landscape ecological risk assessment is mainly based on the evaluation of ecological risks caused by land cover change, which is greatly influenced by the resolution. In the future, it is necessary to further improve the resolution of land cover remote sensing data in order to more accurately capture the geographical characteristics of landscape ecological risk distribution and the changes in ecological processes.
(2) The landscape ecological risk assessment system has a certain subjectivity, mainly reflected in the selection of influencing factors and evaluation methods, the setting of indicators, and the interpretation of data, which may lead to differences in evaluation results between different researchers or the same researcher at different time points. Especially in existing studies, the determination of influencing factors and indicator weights is mainly based on the experience and judgment of the researchers, and its applicability and reliability still need to be tested. In the future, it is necessary to establish reliable methods to effectively validate the accuracy of landscape ecological risk assessment results.
(3) Due to the different research scales, different scholars have different ranges for dividing risk assessment units. Overemphasizing risk assessment units may to some extent sever the natural geographical connections of the original surface factors, and may interfere with the overall understanding and comprehensive analysis of landscape patterns. Moreover, due to the lack of unified standards, the evaluation results of landscape ecological risks cannot be horizontally compared between different regions, and can only discuss the relative risks and temporal evolution of ecological risks within the study area. In the future, it is necessary to solve this problem by establishing relatively unified standards or by expanding the scale.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Bayliss P, Van Dam R A, Bartolo R E. 2012. Quantitative ecological risk assessment of the Magela Creek floodplain in Kakadu National Park, Australia: Comparing point source risks from the ranger uranium mine to diffuse landscape-scale risks. Human and Ecological Risk Assessment, 18(1): 115-151.

[2]
Chen P, Pan X L. 2003. Landscape ecological risk analysis of inland watershed in arid areas—Taking the Sangong River Basin in Fukang as an example. Chinese Journal of Ecology, (4): 116-120. (in Chinese)

[3]
Cheng Y, Li Y L, Chang Z B, et al. 2022. Landscape ecological risk assessment based on land use change—Taking the coastal zone of Guangdong Province as an example. Environmental Ecology, 4(11): 23-33. (in Chinese)

[4]
Critto A, Carlon C, Marcomini A. 2005. Screening ecological risk assessment for the benthic community in the Venice lagoon (Italy). Environment International, 31(7): 1094-1100.

PMID

[5]
Deng X H, Wang L, Ou C H, et al. 2024. Dynamic analysis of landscape ecological risks in the Changsha-Zhuzhou-Xiangtan metropolitan area based on the PLUS model. Geography and Geographic Information Science, 40(1): 47-54, 98. (in Chinese)

[6]
Du W T, Li X P, Song J W, et al. 2022. Analysis and prediction of landscape ecological risks in the Yellow River Basin. Bulletin of Soil and Water Conservation, 42(5): 105-113. (in Chinese)

[7]
Fan J H, Wang Y, Zhou Z, et al. 2016. Dynamic ecological risk assessment and management of land use in the middle reaches of the Heihe River based on landscape patterns and spatial statistics. Sustainability, 8(6): 536. DOI: 10.3390/su8060536.

[8]
Foster S S D, Chilton P J, Stuart M. 1991. Mechanisms of groundwater pollution by pesticides. Water and Environment Journal, 5(2): 186-193.

[9]
Fu B J. 1983. A new field of geography-landscape ecology. Chinese Journal of Ecology, (4): 60-67. (in Chinese)

[10]
Gao B P, Li C, Wu Y M, et al. 2021. Landscape ecological risk assessment and influencing factors in the Sichuan Yunnan ecological barrier area. Chinese Journal of Applied Ecology, 32 (5): 1603-1613. (in Chinese)

[11]
Guo B, Zhou Y, Zhu J F, et al. 2016. Spatial patterns of ecosystem vulnerability changes during 2001-2011 in the Three-River Source Region of the Qinghai-Tibetan Plateau, China. Journal of Arid Land, 8(1): 23-35.

[12]
Hayes E H, Landis W G. 2004. Regional ecological risk assessment of a near shore marine environment: Cherry Point, WA. Human and Ecological Risk Assessment, 10(2): 299-325.

[13]
Hope B K. 2006. An examination of ecological risk assessment and management practices. Environment International, 32(8): 983-995.

PMID

[14]
Huang X, Wang X F, Zhang X R, et al. 2022. Ecological risk assessment and identification of risk control priority areas based on degradation of ecosystem services: A case study in the Tibetan Plateau. Ecological Indicators, 141: 109078. DOI: 10.1016/j.ecolind.2022.109078.

[15]
Huang X T, Ma L, Chen C B, et al. 2020. Predicting the suitable geographical distribution of Sinadoxa corydalifolia under different climate change scenarios in the Three-River Region using the MaxEnt model. Plants, 9(8): 1015. DOI: 10.3390/plants9081015.

[16]
Lazo J K, Kinnell J C, Fisher A. 2000. Expert and layperson perceptions of ecosystem risk. Risk Analysis, 20(2): 179-194.

PMID

[17]
Li H Q, Cheng F F, Song H Y, et al. 2023. Landscape ecological risk assessment of coastal cities in eastern China based on land use change. Journal of Economics of Water Resources, 41 (5): 29-37, 98. (in Chinese)

[18]
Li J G, He C Y, Li X B. 2008. Research on ecological risk assessment of natural/semi natural landscape spaces in rapidly urbanizing areas—Taking Beijing as an example. Journal of Natural Resources, (1): 33-47. (in Chinese)

[19]
Li J L, Pu R, Gong H B, et al. 2017. Evolution characteristics of landscape ecological risk patterns in coastal zones in Zhejiang Province, China. Sustainability, 9(4): 584. DOI: 10.3390/su9040584.

[20]
Li Q P, Zhang Z D, Wan L W, et al. 2019. Landscape pattern optimization in Ningjiang River Basin based on landscape ecological risk assessment. Acta Geographica Sinica, 74(7): 1420-1437. (in Chinese)

DOI

[21]
Li Y, Li B L, Yuan Y C, et al. 2022. Trends in total nitrogen concentrations in the Three Rivers Headwater Region. Science of the Total Environment, 852: 158462. DOI: 10.1016/j.scitotenv.2022.158462.

[22]
Liang Y, Song W. 2022. Integrating potential ecosystem services losses into ecological risk assessment of land use changes: A case study on the Qinghai-Tibet Plateau. Journal of Environmental Management, 318: 115607. DOI: 10.1016/j.jenvman.2022.115607.

[23]
Liu D D, Qu R J, Zhao C H, et al. 2012. Landscape ecological risk assessment in Yellow River Delta. Journal of Food Agriculture & Environment, 10(2): 970-972.

[24]
Liu F L, Yang L, Wang S. 2023. The spatiotemporal evolution and correlation analysis of landscape ecological risk and ecosystem service value in the Jinsha River Basin. Journal of Resources and Ecology, 14(5): 914-927.

[25]
Liu J Y, Xu X L, Shao Q Q. 2008. Grassland degradation in the “Three-River Headwaters” Region, Qinghai Province. Journal of Geographical Sciences, 18(3): 259-273.

[26]
Liu X F, Zhang J S, Zhu X F, et al. 2014. Spatiotemporal changes in vegetation coverage and its driving factors in the Three-River Headwaters Region during 2000-2011. Journal of Geographical Sciences, 24(2): 288-302.

DOI

[27]
Mei Z Y, Zhang Y R, Huang X Y, et al. 2024. Ecological risk assessment of Qinghai Lake Basin based on ecosystem services and analysis of spatial heterogeneity influencing factors. Acta Ecologica Sinica, 44(12): 4973-4986. (in Chinese)

[28]
Niu L N, Shao Q Q, Ning J, et al. 2022. Ecological changes and the tradeoff and synergy of ecosystem services in Western China. Journal of Geographical Sciences, 32(6): 1059-1075.

[29]
Nunneri C, Lenhart H J, Burkhard B, et al. 2008. The use of ecological risk for assessing effects of human activities: An example including eutrophication and offshore wind farm construction in the North Sea. Landscape Online, (5): 1-20.

[30]
Pan J H, Liu X. 2016. Landscape ecological risk assessment and landscape security pattern optimization in Shule River Basin. Chinese Journal of Ecology, 35(3): 791-799. (in Chinese)

[31]
Peng J, Dang W X, Liu Y X, et al. 2015. Review on landscape ecological risk assessment. Acta Geographica Sinica, 70(4): 664-677. (in Chinese)

DOI

[32]
Piao S L, Zhang X Z, Wang T, et al. 2019. Responses and feedback of the Tibetan Plateau’s alpine ecosystem to climate change. Chinese Science Bulletin, 64(27): 2842-2855. (in Chinese)

[33]
Qu Z, Zhao Y H, Luo M Y, et al. 2022. The effect of the human footprint and climate change on landscape ecological risks: A case study of the Loess Plateau, China. Land, 11(2): 217. DOI: 10.3390/land11020217.

[34]
Rogan J, Wright T M, Cardille J, et al. 2016. Forest fragmentation in Massachusetts, USA: A town-level assessment using Morphological spatial pattern analysis and affinity propagation. GIScience & Remote Sensing, 53: 506-519.

[35]
Shao Q Q, Liu G B, Li X D, et al. 2019. Assessing the snow disaster and disaster resistance capability for spring 2019 in China’s Three-River Headwaters Region. Sustainability, 11(22): 6423. DOI: 10.3390/su11226423.

[36]
Shao Q Q, Zhao Z P, Liu J Y, et al. 2010. The characteristics of land cover and macroscopical ecology changes in the Source Region of Three Rivers in Qinghai-Tibet Plateau during last 30 years. 2010 IEEE International Geoscience and Remote Sensing Symposium. Honolulu, USA: IEEE.

[37]
Shi H P, Yu K Q, Feng Y J. 2013. Ecological risk assessment of rural-urban ecotone based on landscape pattern: A case study in Daiyue District of Tai’an City, Shandong Province of East China. Chinese Journal of Applied Ecology, 24(3): 705-712. (in Chinese)

[38]
Sun F H, Zhang Z C, Jiang P G, et al. 2023a. Structure and stability characteristics of zonal soil aggregates in the Three Rivers Source of the Qinghai-Tibetan Plateau. Soil Science Society of America Journal, 87(5): 1042-1055.

[39]
Sun N S, Chen Q, Liu F G, et al. 2023b. Land use simulation and landscape ecological risk assessment on the Qinghai-Tibet Plateau. Land, 12(4): 923. DOI: 10.3390/land12040923.

[40]
Suter G W. 1997. Integration of human health and ecological risk assessment. Environmental Health Perspectives, 105(12): 1282-1283.

PMID

[41]
Suter G W, Norton S B, Barnthouse L W. 2003. The evolution of frameworks for ecological risk assessment from the red book ancestor. Human and Ecological Risk Assessment, 9(5): 1349-1360.

[42]
Tang C C, Wang Y F, Yan K N, et al. 2022. Application of green and low-carbon technologies in the Beijing Winter Olympics and its implications for low-carbon tourism. Journal of Chinese Ecotourism, 12(4): 690-703. (in Chinese)

[43]
Wallack R N, Hope B K. 2002. Quantitative consideration of ecosystem characteristics in an ecological risk assessment: A case study. Human and Ecological Risk Assessment, 8(7): 1805-1814.

[44]
Wang H, Liu X M, Zhao C Y, et al. 2021. Spatial-temporal pattern analysis of landscape ecological risk assessment based on land use/land cover change in Baishuijiang National Nature Reserve in Gansu Province, China. Ecological Indicators, 124: 107454. DOI: 10.1016/j.ecolind.2021.107454.

[45]
Wang H M, Zhang Y G, Li D Q, et al. 2004. Protection and sustainable development for resources in three rivers source nature reservation. The Proceedings of the China Association for Science and Technology. Beijing, China: China Association for Science and Technology.

[46]
Wang R H, Xue Y, Ning H S, et al. 2010. A watershed ecological compensation strategy based on ecological risk assessment. Journal of Arid Land Resources and Environment, 24(8): 1-5. (in Chinese)

[47]
Wang S R, Song Q, Zhao J Y, et al. 2023. Identification of key areas and early-warning points for ecological protection and restoration in the Yellow River source area based on ecological security pattern. Land, 12(8): 1643. DOI: 10.3390/land12081643.

[48]
Wang S S, Tan X, Fan F L. 2022. Landscape ecological risk assessment and impact factor analysis of the Qinghai-Tibetan Plateau. Remote Sensing, 14(19): 4726. DOI: 10.3390/rs14194726.

[49]
Wang Y S, Cheng C C, Xie Y, et al. 2017. Increasing trends in rainfall- runoff erosivity in the Source Region of the Three Rivers, 1961-2012. Science of the Total Environment, 592: 639-648.

[50]
Wei Q Q, He W, Wang J Y, et al. 2023. Spatial and temporal evolutionary characteristics of landscape pattern of a typical Karst watershed based on GEE platform. Journal of Resources and Ecology, 14(5): 928-939.

DOI

[51]
Wu J H, Wang G Z, Chen W X, et al. 2022. Terrain gradient variations in the ecosystem services value of the Qinghai-Tibet Plateau, China. Global Ecology and Conservation, 34: e02008. DOI:10.1016/j.gecco.2022.e02008.

[52]
Wu J S, Qiao N, Peng J, et al. 2013. Spatial differentiation of landscape ecological risks in open-pit mining areas. Acta Ecologica Sinica, 33 (12): 3816-3824. (in Chinese)

[53]
Xia M, Jia K, Zhao W W, et al. 2021. Spatio-temporal changes of ecological vulnerability across the Qinghai-Tibetan Plateau. Ecological Indicators, 123: 107274. DOI: 10.1016/j.ecolind.2020.107274.

[54]
Xiao L, Tian G J. 2014. Ecological risk assessment of land use in Tianjin City. Chinese Journal of Ecology, 33(2): 469-476. (in Chinese)

[55]
Xie H L. 2008. Regional eco-risk analysis of based on landscape structure and spatial statistics. Acta Ecologica Sinica, 28(10): 5020-5026. (in Chinese)

[56]
Xie H L, Wang P, Huang H S. 2013. Ecological risk assessment of land use change in the Poyang Lake Eco-economic Zone, China. International Journal of Environmental Research and Public Health, 10(1): 328-346.

DOI PMID

[57]
Yang J, Duan H X, Song F Q, et al. 2024. Analysis of spatiotemporal evolution of landscape ecological risks in Gaoligong Mountain from 2000 to 2020. Journal of Southwest Forestry University (Natural Science), 44(4): 54-63. (in Chinese)

[58]
Yao T, Thompson L G, Mosbrugger V, et al. 2012. Third pole environment (TPE). Environmental Development, (3): 52-64.

[59]
Yu H, Xu L L, Liu Q Q, et al. 2021. Research on the management of recreational spaces in Canadian national parks. Journal of Chinese Ecotourism, 11(2): 256-265. (in Chinese)

[60]
Yu Y N, Zhu J, Wu S H, et al. 2016. Assessment of land ecological risks driven by multi-sources: A case study of Sheyang County, Jiangsu Province. Journal of Natural Resources, 31(8): 1264-1274. (in Chinese)

[61]
Zhang D L, Liu J Q, Jiang X J, et al. 2016. Distribution, sources and ecological risk assessment of PAHs in surface sediments from the Luan River Estuary, China. Marine Pollution Bulletin, 102(1): 223-229.

DOI PMID

[62]
Zhang F, Yushanjiang A, Wang D F. 2018. Ecological risk assessment due to land use/cover changes (LUCC) in Jinghe County, Xinjiang, China from 1990 to 2014 based on landscape patterns and spatial statistics. Environmental Earth Sciences, 77(13): 491. DOI: 10.1007/s12665-018-7676-z.

[63]
Zhang J Q, Liang J D, Liu X P, et al. 2009. GIS-based risk assessment of ecological disasters in Jilin Province, Northeast China. Human and Ecological Risk Assessment, 15(4): 727-745.

[64]
Zhang L N, Zhang H Q, Xu E Q. 2022a. Information entropy and elasticity analysis of the land use structure change influencing eco-environmental quality in Qinghai-Tibet Plateau from 1990 to 2015. Environmental Science and Pollution Research International, 29(13): 18348-18364.

[65]
Zhang L Q, Peng J, Liu Y X, et al. 2017. Coupling ecosystem services supply and human ecological demand to identify landscape ecological security pattern: A case study in Beijing-Tianjin-Hebei Region, China. Urban Ecosystems, 20(3): 701-714.

[66]
Zhang X J, Zhong L S, Yu H. 2023. Research on the sustainability assessment method of lake-type tourism destinations in the Qinghai-Tibet Plateau from a relationship perspective. Journal of Chinese Ecotourism, 13(4): 621-634. (in Chinese)

[67]
Zhang X Y, Ning J. 2023. Patterns, trends, and causes of vegetation change in the Three Rivers Headwaters Region. Land, 12(6): 1127. DOI: 10.3390/land12061127.

[68]
Zhang X Y, Yu H, Zhang X, et al. 2022b. Comprehensive evaluation of land ecological security in Sanjiangyuan National Park based on multi- source data. Acta Ecologica Sinica, 42(14): 5665-5676. (in Chinese)

[69]
Zhang Y, Zhang C B, Wang Z Q, et al. 2019. Comprehensive research on remote sensing monitoring of grassland degradation: A case study in the Three-River Source Region, China. Sustainability, 11(7): 1845. DOI: 10.3390/su11071845.

[70]
Zheng K J, Li C, Wu Y M, et al. 2022. The spatiotemporal evolution and influencing factors of landscape ecological risks in Yunnan Border Mountainous Areas. Acta Ecologica Sinica, 42(18): 7458-7469. (in Chinese)

[71]
Zhou L H, Wang P T, Bai Y X. 2023. Optimization of the ecological security pattern in Xi’an City based on a minimum cumulative resistance model. Journal of Resources and Ecology, 14(6): 1127-1137.

DOI

[72]
Zhou S Y, Chang J, Hu T H, et al. 2020. Spatiotemporal variations of land use and landscape ecological risk in a resource-based city, from rapid development to recession. Polish Journal of Environmental Studies, 29(1): 475-490.

[73]
Zhou W, Yang H, Xie L J, et al. 2021. Hyperspectral inversion of soil heavy metals in Three-River Source Region based on random forest model. CATENA, 202: 105222. DOI: 10.1016/j.catena.2021.105222.

Options
Outlines

/

〈 〉