Conflict or Coordination? The Spatiotemporal Relationship Between Humans and Nature on the Qinghai-Tibet Plateau

Abstract

As the Earth's Third Pole and the Asian water tower, the Qinghai-Tibet Plateau (QTP) plays a key role in global climate regulation and biodiversity maintenance. Living in harmony with nature is vital for local and global sustainable development. Current research on the conflicted or coordinated relationship between humans and nature on the QTP at a fine spatial scale remains limited. To fill the gap, we developed the human activity intensity index (HAI) and eco-environmental quality index (EQI) at 1-km resolution and proposed a four-quadrant diagram approach to explore the dynamics between them. The results show a coordinated development on the QTP as the HAI and EQI both increased from 2000 to 2020, and the ratio of coordinated areas to conflicted areas was 5:1. High HAI areas were mainly in big cities such as Xining, Lhasa, Haidong, Xigaze, and along traffic lines. The significant conflicted areas were mainly outside the Lhasa metropolitan, south of the Hengduan Mountains, and along some new roads, and reduced by 8% between 2000–2010 and 2010–2020. The area of high HAI but low EQI was the smallest proportion, mainly in southern Qinghai Lake, southern Brahlung Zangbo River, Gobi oases, and western transport lines, but it implies the highest risk of ecosystem degradation. This research expands the fundamental methodology to address complex human-natural relationships and provides implications for the sustainable development of fragile ecosystems.

Introduction

Humans have spread their footprint all over the planet, and the Earth entered the Anthropocene (Kareiva et al., 2007; Steffen et al., 2007). However, the carrying capacity of the Earth is limited. With the deepening understanding of climate change and environmental crisis, the contradiction between humans and nature is considered the root cause hindering sustainable development (Anderson et al., 2019; Dorninger et al., 2017). Human activities have exceeded planetary boundaries regarding biosphere integrity and the cycling of nitrogen and phosphorus, and the challenge of living in a safe and just space remains substantial (O’Neill et al., 2018). Therefore, it is crucial to understand the relationship between socioeconomic development and ecological environment evolution for global sustainable development. Moreover, balancing human activities and the eco-environment is the core of many UN Sustainable Development Goals (SDGs), including Goals 6, 7, 11, 12, 13, 14, and 15 (Cheng et al., 2021; Liu, Hull, et al., 2018), and many global organizations or programs have also explored the coupling mechanisms and collaborative development approach of the human-natural systems (Kates et al., 2001; Liu, Fang, & Fang, 2020). Analyzing contradictions and coordinations between human activities and the eco-environment has great practical significance for humans' fate and well-being.

The Qinghai-Tibet Plateau (QTP), with an average altitude of over 4,000 m and extensive distribution of glaciers and permafrost, is the source of the nine rivers in Asia, providing fresh water, food, and other ecosystem services to more than 1.5 billion people, and is known as the Earth's Third Pole and Asian water tower (Yao et al., 2012) (Figure 1). It plays an important role in global climate regulation and biodiversity maintenance (Li et al., 2020). However, due to its high elevation, cold temperatures, and dry environment, the QTP's ecosystems are extremely sensitive and fragile to human activities (Han et al., 2022; Liu, Milne, et al., 2018). From 1980 to 2020, the population increased by 1.56 times, the urban construction area by nearly three times, and the highway length by more than four times (Zhang et al., 2019). Intense human activities, such as overgrazing, put enormous pressure on local grassland ecosystems, which may be much stronger than the effects of climate change (Wei et al., 2022). The intensification of human activities poses a great threat to the stability of the QTP's ecosystems and the realization of SDGs (Yang et al., 2022). Thus, illustrating the human-nature relationship on the QTP is of great significance to local and global sustainable development.

Figure 1. Physical geography of Qinghai-Tibet Plateau.

The interaction between humans and nature has existed since people first appeared (Costanza et al., 2007). In the context of increasingly severe environmental problems, many scholars from different disciplinary backgrounds focused on this field closely related to sustainable development and generated valuable insights (Harden, 2012). Research topics include the positive and negative effects of human activities on nature (Kareiva et al., 2007), the constraints and carrying capacity of the natural resources and ecosystems on social development (Fan et al., 2017), and the interactions, feedback, tradeoffs, and adaption between them (Comberti et al., 2015; Holling & Gunderson, 2002; Kareiva et al., 2007). Some research indicates that reducing the scale of the human enterprise is necessary to alleviate society crisis (Dirzo et al., 2022); others show that a win-win vision could be achieved if we make appropriate changes in the way of production, energy use, ecological protection, etc. (Cao et al., 2009; Tallis et al., 2018).

In this research field, the urbanized areas (Fang et al., 2016), river basins (Yin et al., 2021), forests (Davidson et al., 2012), and coastal zone (Yi et al., 2018) received more attention. For nature system, researchers focus on biodiversity (Cardinale et al., 2012), vegetation (Zhang, Yang, et al., 2022), air pollution (Liu, Cui, & Zhang, 2022), river environment (Leprieur et al., 2008), soil erosion (Wei et al., 2006), etc. Land use and cover change (Cheng et al., 2022), population density (Cropper & Griffiths, 1994), and nighttime light (Cai et al., 2021) are usually used as proxy variables for human activities; also some scholars constructed a comprehensive index named human footprint (Li, Wu, Gong, et al., 2018; Venter et al., 2016). The methods used to analyze the complex relationship between human and nature include network approach (Kluger et al., 2020), system dynamics (Reed et al., 2022; Yoon et al., 2021), coupling coordination analysis (Jiang et al., 2022; Shi, Feng, et al., 2022), complex system theory (Bennett & McGinnis, 2008; Wang & Grant, 2021), environmental Kuznets curve model (Zhou et al., 2015), metacoupling framework (Liu, 2017), supply-demand balance analysis (Liu, Xing, et al., 2022), questionnaires (Allison et al., 2021; Liu, Zhang, et al., 2020), and qualitative analysis (Ives et al., 2017). These methods advance the research of complex human-natural systems from different perspectives, but face substantial constraints at fine grid scales. Certain studies employing raster data: Vačkář et al. (2012) analyzed spatial relationship between population density, land use intensity, and biodiversity in the Czech Republic using 10-km grids. Shi, Feng, et al. (2022) illuminated the coordination between landscape ecological risk and urbanization across 10 km × 10 km in China utilizing a coupling coordination degree. Zhou et al. (2023) measured the conflict and coordination dynamics between urbanization and vegetation cover for Chinese cities with a 250 m vegetation index. Simkin et al. (2022) predict the future impact of urban land expansion on biodiversity across three shared socioeconomic pathway (SSP) scenarios. However, these studies concentrated on certain human activities or natural elements, lacking a holistic consideration of human and natural systems.

Archaic hominins have occupied the QTP since 160,000 years ago (Chen et al., 2019). In the past half-century, human activities have caused limited but rapidly increasing disturbance to the QTP ecosystem, especially in the largest proportion of grassland, valuable water conservation areas, and biodiversity conservation areas (Jie et al., 2015; Li, Zhang, Wang, et al., 2018). Research on the relationship between humans and nature on the QTP mainly focuses on the impacts of urbanization, land use, infrastructure construction, agriculture, and animal husbandry on the eco-environment (Miao et al., 2021; Wang et al., 2015; Zhang, Zhang, et al., 2022). The level of urbanization here is quite low, but in recent years the urban population has grown rapidly, and the construction land has expanded fast (Wang et al., 2020). Although urbanization will lead to more pollutant emissions, it is beneficial to reduce the overall disturbance of the ecosystem caused by human activities by shifting the population from farming and pastoral areas (Tian et al., 2021). From 2000 to 2015, the coordination between urbanization and the eco-environment showed an increasing trend (Feng & Li, 2021). Grassland degradation is mainly contributed by human activities in the short term, which further contributes to biodiversity loss, increased soil erosion, reduced carbon sequestration, and reduced local human well-being (Chen et al., 2020; Dong et al., 2020). The construction and operation of Qinghai-Tibet Railway and Sichuan-Tibet Railway will bring potential risks to the local eco-environment (Cui et al., 2022; Luo et al., 2020; Qin & Zheng, 2010). From a system perspective, the codevelopment of industries, population, and resources and environment was analyzed using the carrying capacity model (Niu et al., 2020), the ecological quality was evaluated by linking human activities and ecosystem services (Sun et al., 2020; Wang, Huang, et al., 2022), and the regional sustainability was measured through integrating ecological capacity, ecological footprint, and resources efficiency (Fan & Fang, 2022). Meanwhile, China's ecological protection and restoration projects on the QTP, including reforestation, grassland restoration, and setting up natural reserves, are improving the eco-environment in many areas (Fu et al., 2021; Jin et al., 2019). In recent years, sewage discharge and exhaust emission from mining and cities, garbage caused by increasing tourists, and light pollution have become new serious environmental problems (Chen, Zhang, et al., 2021; Feng et al., 2020; Wang, Lv, et al., 2022).

Literature review shows that most of the previous research analyzed the relationship between a specific type of human activities and a kind of eco-environment element, with few studies taking human activities and eco-environment as two single systems to explore the macro relationship between them. Current methods to quantify the interactions between humans and the environment have significant limitations, especially in the fine spatial dimension. Raster-scale data are rarely used to study the spatiotemporal evolution and relationship between people and nature on the QTP, causing a lack of understanding of the coordinated development of ecological environment and socioeconomic development at a fine spatial and temporal scale. To fill these research gaps, we constructed two comprehensive indexes, named human activity intensity index (HAI) and eco-environmental quality index (EQI), and proposed a four-quadrant diagram approach to identify dynamic interactions between human activities and eco-environment. A set of 1 km raster data were used to reveal the spatiotemporal evolution of the human-natural systems on the QTP from 2000 to 2020. This work could provide a methodology for other studies and support for the sustainable development of the Earth's Third Pole.

Methodology and Data Basic Concept and Research Approach

Human activities, including cultivation, grazing, fishing, mining, transportation, power generation, and buildings, inevitably impact the eco-environment system (Figure 2a) (Govorushko, 2016). Eco-environment elements like water, soil, atmosphere, animals, vegetation, and microorganism, support and carry socioeconomic development and human well-being (Goudie, 2018). These eco-environment and socioeconomic elements embedded in the coupled human and natural systems are connected and interacted via exchanges of energy and matter (Figure 2b). Thus, humans and nature have been interacting, influencing, and transforming each other over time and space (Liu, Fang, & Fang, 2020; Wu, 2013). In the process, new structures, functions, and mechanisms emerge to form a dynamic and complex world (Wang et al., 2018). Chinese traditional philosophy considers humans as an integral part of nature (Bourdeau, 2004), and the integrated system of the two is called the Coupled Human and Natural Systems (CHANS) (Kramer et al., 2017). In recent years, greater access to diverse gridded data has allowed us to characterize human activity and ecological environmental information at a highly granular scale and examine the interconnections between people and nature in each grid (Figures 2c and 2d). This has vastly broadened the purview of traditional CHANS research.

Figure 2. Conceptualization and abstraction of coupled human and natural systems. (a) Actual world; (b) coupled human and natural systems in a region; (c) rasterization of human activities and eco-environment; (d) coupled human and natural systems at the grid scale.

Figure 3 is the flowchart of this research. To explore the interaction between humans and nature on the QTP, we collected and sorted out various human activity and ecological environment data from 2000 to 2020, and transformed these data into a 1-km × 1-km grid scale. Given the complex interactions, we constructed the HAI and EQI indexes based on a set of indicators (Tables 1 and 2) to measure human activity and eco-environment. On this basis, we analyzed the spatial and temporal patterns of the two separately as well as the spatial matching between them. Furthermore, we classified the human-nature interactions on the QTP into four groups through a four-quadrant diagram from a dynamic perspective (Figure 4), and identified potential risk areas over the past 20 years. In particular, the densely populated metropolitan area and national nature reserve are emphatically analyzed.

Figure 3. Research methodology flowchart.

Table 1 Data Description, Source, and Processing Methods of Human Activity Indicators

Data set		Year and format		Data source	Data processing
Population density		2000, 2005, 2010, 2015, 2020; Raster, 1 km		WorldPop	For all locations with more than 1,000 people km2, we assigned a score of 100. For more sparsely populated areas, we logarithmically scaled the score using 33.333 × log (population density + 1) (Williams et al., 2020)
GDP density		2000, 2005, 2010, 2015, 2019; Raster, 1 km		Chen et al. (2022)	Scale data into the 0–100 range using Min-Max Normalization, and the maximum value is the average of the top 50 values in the 5 years
Nighttime light		2000, 2005, 2010, 2015, 2020; Raster, 1 km		Zhang, Ren, et al. (2021)	Assigned from 0 to 100 according to intervals determined by 10 equal quantiles (Mu et al., 2022)
Grazing intensity		2000, 2005, 2010, 2015, 2019; Raster, 250m		Liu (2021)	The data in 2005 used the average of 2000 and 2010, and the data in 2015 used the average of 2010 and 2019. Resampling to 1-km resolution. Scale data into the 0–100 range using Min-Max Normalization, and the maximum value is the average of the top 50 values in the 5 years
Land use/cover		2000, 2005, 2010, 2015, 2020; Raster, 1 km		www.resdc.cn/data.aspx?DATAID=335	Assign score: 100 for urban land, 50 in 1 km buffer; 80 for construction land, 40 in 1 km buffer; 60 for cropland, 20 in 1 km buffer; 0 for other lands
Road network		2000, 2005, 2010, 2015, 2018; Shapefile		Digitization by authors	Assign a score of 60 for 0.5 km out for either side of a highway, national road, and provincial road. Assign 80 and 40 for 0.5 and 1.5 km out for either side of a railway (Venter et al., 2016). Convert to raster data

Table 2 Data Description, Source, and Processing Method of Eco-Environmental Indicators

[Table omitted. See PDF]

Figure 4. Theoretical analysis and the four-quadrant diagram approach. (a) Macroscopic evolution of temporal dimension. (b) Types of CHANS in the spatial dimension.

HAI

There exist numerous metrics quantifying the intensity of human activities (human footprint, human modification), which are frequently utilized to characterize human pressures on ecosystems (Venter et al., 2016; Williams et al., 2020). The HAI constructed in this research largely follows the evaluation index systems established by previous studies, encompassing population density, GDP, nighttime light, land cover, and road networks (Huang et al., 2022; Mu et al., 2022; Xu et al., 2016). However, given that grasslands occupy over 60% of the QTP (Li et al., 2019; Zeng et al., 2022), we employed grazing intensity (i.e., the number of grazing livestock per square kilometer) in lieu of previously used pasture area to improve the accuracy of quantifying human activities in grassland ecosystems. Additionally, unlike previous studies that typically assigned equal weights to all indicators, we adopted a different approach based on expert judgment. Specifically, we assigned a weight of 1/12 to population density, GDP, and nighttime light in densely populated urban and mining areas. This decision was made to account for the potential overlap of these three indicators. On the other hand, we assigned a 1/4 weight to the remaining indicators (Shi, Wu, et al., 2022).

The six indicators were transformed into 1-km resolution raster data with a range of 0–100 after an outlier check, projection transformation, raster reassignment, and standardization of the original data. The HAI index was composed by using the following formula: [Image Omitted. See PDF]where α is one of the six indicators and w is the corresponding weight. Land use data gauge human pressure through construction land and cropland, as grassland and grazing intensity have overlapping meanings. Both sides of the road are generally affected by human activities within a certain range. Referring to previous studies, we assign values within the buffer of 0.5 and 1.5 km. Due to the unavailability of data for certain indicators in 2020, data from the previous year substituted as a proxy. Besides, the 2005 grazing intensity employed the average of 2000 and 2010, and the 2015 grazing intensity used the average of 2010 and 2019.

EQI

Eco-environmental quality is a macro concept that reflects the state of ecosystems and the natural environment closely linked to human activities (Goudie, 2018). Many studies have employed pollutant emissions and environmental quality monitoring data to characterize the eco-environmental quality of a given region (Chen, Lu, et al., 2021; Fang et al., 2019). However, such data are difficult to downscale to the grid level. At the grid scale, some scholars have attempted to use NDVI, NDBSI, LST, and WET to reflect greenness, dryness, heat, and humidity, respectively, and further construct an EQI (Boori et al., 2021; Xu et al., 2021; Yue et al., 2019). Nonetheless, these remote sensing indicators primarily reflect the state of ecosystems and lack consideration of environmental pollution.

In 2015, the Ministry of Ecology and Environment of China issued the Technical Criteria for Eco-Environmental Status Evaluation in China (HJ 192-2015), which comprehensively evaluates eco-environmental quality from six dimensions: biological richness, vegetation coverage, network density of water systems, land stress, pollution loads, and pollution event (China, 2015). Our study referred to this criterion but substituted corresponding alternative indicators for those unobtainable at the grid scale: the water conservation index replaced the network density of water systems, and pollution loads were represented solely by PM_2.5 concentrations, to which China currently devotes the most attention. Pollution event was not presently considered. Thus, seven indicators that are influenced by human activities at the interannual scale were selected for the EQI, including vegetation coverage, biological richness, NPP, soil salinization, soil erosion, water conservation, and PM_2.5 concentration. The data source, processing, and weight of each indicator are shown in Table 2.

To reduce the disturbance of the interannual variation of FVC, NPP, and PM_2.5 by climate factors, we take a 3-year time interval average of each indicator. After processing and standardizing the original data, all seven indicators are transformed into 1-km resolution raster data with a range of 0–100. We then invited 15 experts (i.e., six ecologists, six geographers, and three environmentalists) to determine the weight of each indicator based on the AHP method. Soil salinization, soil erosion, and PM_2.5 are negative indicators, and the rest are positive indicators. The formula of EQI is [Image Omitted. See PDF]where β is one of the seven indicators and w is the corresponding weight.

Identification of the Dynamic Relationships Between Human Activities and Eco-Environment

Human activities have both negative and positive effects on nature (Chi et al., 2020; Huimin, 2013). With the spatial agglomeration of population and various industries, humans have brought pressure on the eco-environment, including the expansion of impervious land and cultivated land, deforestation, infrastructure construction, and the discharge of air and water pollutants. Since the 1980s, the concept of sustainable development has gradually spread worldwide, and people have begun to act to protect the eco-environment. More environment-friendly materials and low-carbon energy-saving technologies were created, and more environmental policies were implemented. Human pressure on nature will gradually decrease. At some point, when it reaches a turning point, the influence of human activities on the eco-environment may change from negative to positive, with the CHANS from conflict to coordination (Figure 4a). Some scholars have concluded that this evolutionary trend conforms to the “U-shaped” curve (Fang et al., 2021). Therefore, the relationship between humans and nature is not zero-sum. Under certain conditions, reciprocal relationships can be realized (Comberti et al., 2015).

Where does the relationship between humans and nature in a particular area lie on this U-shaped curve? Which plays a dominant role, conflict or coordination? To solve this issue, from the dynamic perspective of the CHANS evolution, we draw a four-quadrant diagram consisting of changes in HAI and EQI, and the evolutionary relationship between the two is subdivided into four categories (Figure 4b) [Image Omitted. See PDF]

The first quadrant, “coordination of CHANS,” is a hopeful state in which both people and nature thrive. The second quadrant is good for nature with decreased HAI and increased EQI, where the ecosystem has been restored and developed. In the third quadrant, the HAI and EQI tend to decline, likely related to ecosystem and socioeconomic degradations. The fourth quadrant is a state of conflict between people and nature. In this state, human activities and socioeconomic development continue to increase while the eco-environmental system has been degraded. Here, quadrants one and two are considered sustainable development, and quadrants three and four are considered unsustainable.

In addition, we take 10% of the standard deviation of the change of HAI and EQI in 10 years as the threshold. That is, grids fluctuating within 0.1 standard deviations are considered to have insignificant change (gray space in Figure 4b); beyond this threshold, they are considered significant. Areas with significant changes in human activities and the eco-environment need our special attention. The formula for calculating the threshold is as follows: [Image Omitted. See PDF]where TV denotes the threshold value, σ represents the standard deviation, x_j represents the change in the jth grid over 10 years, μ represents the average of change in all grids, and n represents the total number of grids.

Results Spatiotemporal Evolution of Human Activities and Eco-Environment

Figure 5 depicts the mean values of HAI and EQI on the QTP from 2000 to 2020, demonstrating an overarching panorama. In the past 20 years, the HAI has increased significantly, with the increase rate in 2010–2020 being about three times that in 2000–2010. The EQI also showed an increasing trend, but its growth rate was lower than that of HAI, and its increase slowed down in the last decade.

Figure 5. Mean value of human activity intensity index and eco-environmental quality index on the Qinghai-Tibet Plateau from 2000 to 2020.

The spatial distribution pattern of human activities has not changed significantly over the 20 years. The areas of high intensity were mainly distributed near big cities such as Xining, Lhasa, Haidong, and Xigaze, as well as some hub cities along the transportation lines, such as Golmud, Delingha, and Qamdo (Figures 6 and S1 in Supporting Information S1). The HAI increased most significantly in Xining metropolitan region, Lhasa metropolitan region, and South Tibet (Figure 6d). In addition, the HAI rose dramatically in some areas of Hainan and Qamdo. Changes in Hainan came from the construction of the Gonghe Photovoltaic Power Station, which began in 2012 and covers an area of 609.6 km². The change of Qamdo was mainly due to the extensive expansion of cultivated land.

Figure 6. Spatiotemporal evolution of human activity intensity index.

Figure 7 shows the spatiotemporal evolution of EQI. From 2000 to 2020, the overall spatial pattern was stable, with a significant decrease in the EQI from southeast to northwest due to topography and climate (see Figure S2 in Supporting Information S1 for 2005 and 2015). The high EQI was found in the Hengduan Mountains and South Tibet in the Southeast, while the low EQI was mainly in the Qaidam Basin and North Tibet Plateau in the Northwest. We find that most cities on the QTP are primarily located in areas with relatively favorable eco-environmental quality—even Golmud, Delingha, and Mangya further north reside within oases amid the Gobi. EQI increased in Qinghai Province during the last 20 years, whereas Tibet, except Nyingchi region, experienced a decline. The north Kunlun Mountains had the most significant decrease in EQI.

Figure 7. Spatiotemporal evolution of eco-environmental quality index.

Spatial Matching Between HAI and EQI

To identify the spatial matching between HAI and EQI each year, we divided the human-nature relationship on the QTP into four categories based on the average HAI and EQI as thresholds, namely, high HAI, high EQI; high HAI, low EQI; low HAI, high EQI; low HAI, low EQI. Since the overall pattern of spatial matching does not change much, only the 2000 and 2020 maps are presented in Figure 8. The category of high HAI and high EQI means a match between socioeconomic development and the eco-environment. Most cities and towns belong to it, and the area of this category decreased in 2020. The group of high HAI and low EQI is the smallest one and shrunk significantly from 2000 to 2020. It is mainly in southern Qinghai Lake, southern Brahlung Zangbo River, Gobi oases, and western transport lines, where humans are more likely to clash with nature. Low HAI and high EQI are mainly in nonurban areas in the southeast. The ecosystem in that area is less under human pressure, and the area expanded from 2000 to 2020. Low HAI and low EQI are mainly in the northwestern, where the eco-environment conditions are harsh and not suitable for human habitation. Many areas belong to the depopulated zone.

Figure 8. Spatial matching between human activity intensity index and eco-environmental quality index.

Dynamic Relationships Between Human Activities and Eco-Environment

To measure the conflict or coordination between HAI and EQI, based on the four-quadrant diagram approach, we identify the evolution types of the human-nature relationship in three phases, 2000–2010, 2010–2020, and 2000–2020 (Figure 9). By calculating the standard deviation of HAI and EQI changes every period using Equation 3, we derive threshold values to detect statistically significant variations in these indices. HAI changed significantly in 7% of the regions, and EQI fluctuated significantly for 88% between 2000 and 2020. We focus on areas with significant changes, with the other area painted in gray on the map (Figure 9). From 2000 to 2010, the area of conflict accounted for 28.4%, mainly distributed in Lhasa metropolitan region, southern Tibet, and Mangya; the coordination area accounted for 69.8%, mainly distributed in Qinghai Province; other types of areas accounted for small proportion. From 2010 to 2020, the area of conflict accounted for 20.4%, mainly distributed around some towns and on the sides of many new roads in the west; the coordination area accounted for 39.2%, and most of the core areas in big cities belong to this type. The proportion of good for nature type increased to 25.8%, mainly distributed in the Hengduan Mountains, part of the Qilian Mountains, and the southern area of Xigaze. Degradation types accounted for 14.6%, mainly distributed in eastern Qinghai Province and around Yushu.

Figure 9. Dynamic relationships between human activities and eco-environment.

Viewing the overall change from 2000 to 2020, the proportion of the coordinated evolution area to the conflicting evolution area was 5:1 (Figure 9d). The conflict areas were mainly distributed in the outer Lhasa metropolitan, the south of Hengduan Mountain District, and on the side of some main roads. The development of these areas is not sustainable and deserves high attention from the local government. The good for nature type accounted for 16.5%, and both decrease types accounted for <3%. Overall, the human-natural system on the QTP tends to be coordinated in these 20 years.

More than 50% of the population on the QTP is distributed in the Xining and Lhasa metropolitan areas. Much of the living and production supplies for inhabitants in other regions of the QTP must traverse these two cities, and the modernization progression of highland populations is largely predicated upon them. Hence, a detailed analysis centered on these two densely peopled localities is requisite. As shown in Figure 10, the HAI of the Xining metropolitan region from 2000 to 2020 first increased and then decreased, while the EQI was always on the rise. The decline of HAI in the recent 10 years is probably due to the implementation of the conversion of farmland to forest and continuous optimization of grazing intensity (Wang et al., 2023). Despite the constant expansion of Xining's population and construction area, the forest coverage rate increased from 28.6% in 2010 to 36.3% in 2020 (see Figure S3 in Supporting Information S1 for spatial presentation). For the Xining metropolitan region, the coordinated evolution area accounted for 72%, while the conflicted area accounted for only 3% during 2000–2020; in the 25% area, HAI decreased and EQI improved. The HAI in the Lhasa metropolitan region is lower than that in Xining, and the average is about half of that in Xining, but it has kept rising in the past 20 years. The EQI in Lhasa is also lower than in Xining, showing a trend of first decline and then a rise. From 2000 to 2020, 48% of the areas with significant human activity changes in the Lhasa metropolitan region showed the harmonious evolution of the human-nature relationship, 16% showed a decrease in human activities and a better eco-environment, and 26% was in a conflicted state. This indicates that the contradiction between humans and nature in Lhasa metropolitan region is more significant than that in Xining metropolitan area.

Figure 10. Relationship between human activities and eco-environment in key areas.

Sanjiangyuan National Park is the birthplace of the Yangtze, Yellow, and Lancang rivers, an important source of freshwater resources in Asia, the most concentrated area of biodiversity on the QTP, and a sensitive area for climate change in Asia and even the world. Therefore, it is essential to scan its human-nature relationship. The average HAI in the Sanjiangyuan area during 2000–2020 is very small, about 0.3, but the value tends to rise (Figure 10c). EQI showed an increase from 2000 to 2010 and almost no change from 2010 to 2020. Around 2% of the area shows significant enhancement of human activity, mainly on the sides of the Qinghai-Tibet Railway and other highways, but the good news is that 95% of these areas belong to the coordination type. Although the road construction changed the land cover in some areas, the project adopted nature-based solutions (e.g., 675 bridges for wildlife migration), and the overall eco-environmental quality along the road increased in the years after the railway was built (Bu et al., 2013).

Discussion Has the Human-Natural Relationship on the Third Pole Reached a Turning Point?

In ancient times, human existence on the QTP has been relatively harmonious with nature. Local people believed that the three most essential things in the world are water, pasture, and livestock, which they consider gifts from nature. Both the historical Bonpo and the current Tibetan Buddhism believed by most Tibetan people advocate reverence for nature and protection of nature (Shen & Tan, 2012). However, beginning in the 1960s, the grazing intensity on the plateau gradually increased, and serious grassland degradation processes started. Human activities, especially changing land use practices, were identified as the primary force for eco-environment deterioration (Li et al., 2013; Wang et al., 2015).

Since the 1980s, with China's reform and opening-up policy, industrialization and urbanization on the QTP have accelerated. From 1980 to 2020, the population increased rapidly, urban and rural construction areas expanded continuously, and the proportion of the urban population increased from 15% to 47% (Figure 11). Our results show that, from 2000 to 2020, despite the increased intensity of human activity in urban areas, people living on the QTP trended in harmony with nature except in some population concentration areas and places close to roads. This is consistent with other research (Feng & Li, 2021; Li, Zhang, Liu, et al., 2018; Zhang, Zhang, et al., 2022). We believe this change is closely related to China's Ecological Civilization strategy in recent years. The QTP has been designated as the National Ecological Security Shelter Zone of China, and top-down ecological protection policies have promoted the eco-environment in the Third Pole region (Ouyang et al., 2016; Sun et al., 2012). Some key indicators give a glimpse, e.g., the livestock on the plateau had been on the rise from 1990 to 2006 and began to decline after reaching the peak in 2006, and the NDVI began to rise after reaching its lowest value in 2008 (Figure 11). However, whether the transition attained the hypothetical tipping point (Figure 4a) over the past 10–20 years needs further verification and a longer time series investigation. In addition, our study also reveals the spatial heterogeneity of this phenomenon; e.g., the conflict between people and nature is still significant in the suburbs of Lhasa. Figures 8 and 9 provide more spatial information to understand the human-nature relationship over time, which is crucial for local government's targeted ecological governance, population, and industrial structure adjustment.

Figure 11. Time series trends of some indicators on the Qinghai-Tibet Plateau.

Policy Recommendations

Although our results show that the human-nature relationship on the QTP is becoming more harmonious, the plateau's fragile ecosystem still faces great challenges with population growth, urbanization, infrastructure construction, and the threat of global climate change. Previous research revealed that 28 nature reserves on the QTP showed over 10% increase in human footprint from 2000 to 2020 (Hua et al., 2022). The sustainable development of QTP in the future should put forward a systematic solution to keep a balance between ecological conservation and socioeconomic development (Li et al., 2022). The following approaches may help achieve a win-win situation.

First, it is necessary to speed up the construction of the Third Pole national park group, including the Qilian Mountains, Pudacuo, Qinghai Lake, and Qomolangma, and delineating priority regions for conserving biodiversity, water conservation, and ecosystem carbon sinks (Zhao et al., 2020). Help those living in these core protected areas gradually move out. Second, we need to improve the ecological compensation mechanism and increase financial support for the protection of the QTP ecosystem. It is suggested to divide the supply and use of ecological products by the upper, middle, and lower reaches of the river basin, establish an interregional ecological compensation approach, and explore the way of value transformation of ecological resources. Third, use regional rotational grazing, pasture planting, and winter semihouse feeding to optimize grazing management and establish sustainable grassland use patterns. Fourth, using nature-based solutions in facility construction and operations such as roads, mining, wind, and photovoltaic power, with minimal changes to local ecosystems, any infrastructure projects in their full life cycle must be environmentally sustainable (Li et al., 2016). Fifth, per capita household energy consumption on the QTP was roughly two to three times the national average, while agriculture and livestock generated substantial carbon emissions (Ping et al., 2011). Small photovoltaic power stations and hydroelectric plants proved more cost-effective here. We will encourage agricultural regions to transition from traditional energy sources to clean and renewable ones, and optimize fertilizer use and technical management (Yu et al., 2022; Zhuang et al., 2021). Sixth, the cities' scale on the plateau should be rationally planned to remain within the thresholds of resource environmental bearing capacity, and ecological tourism ought to be developed to mitigate summer tourist detriment to the environment (Hu et al., 2021).

Research Uncertainty and Prospect

The index weight of EQI is based on scores from expert experience, which is subjective to some extent. Any weighting structure is likely to be criticized by some because assigning weights is a value-dependent process (Carpenter et al., 2009). Thus, we adjusted the weights of some indicators of EQI by 25% up or down randomly and found that the spatiotemporal pattern of the results did not change significantly. On a long time scale, eco-environmental evolution is influenced by both human activities and climate change, and EQI variation on the QTP cannot be attributed entirely to human activities (Zhang, Wang, et al., 2021). However, the analysis period herein spans only two decades, and the research shows that human stresses far outstrip climatic influence upon the QTP in the near term (Wei et al., 2022). Our research aim is not to analyze the attribution of eco-environment change, but to explore the risk areas of human-nature contradiction and the human-nature interaction over time. Moreover, we neglected the random error of time selection. We merely selected one year every five years from 2000 to 2020, which could affect the robustness of the results.

We did not account for human-nature telecoupling. For instance, the QTP receives many tourists in summer. In 2019, the QTP hosted 126.43 million visitors from around the globe. The pressure they exert upon the Third Pole cannot be ignored. Studies also show that global trade has remotely impacted the QTP's glaciers (Yi et al., 2019). Prospectively, big data may quantify human activities at superior temporal resolution (Yi et al., 2020). We shall further explore the human-nature interaction mechanisms and both the short and long-range impacts of human activities.

Conclusions

Potential contributions and significance compared to existing literature include: (a) From a systems evolution view, we proposed a four-quadrant diagram method to identify dynamic relationships between human activities and the eco-environment. (b) The human activity intensity and eco-environmental quality were measured using a 1-km grid scale on the Third Pole. (c) We identified areas of human-nature conflict on the QTP from 2000 to 2020, which will benefit more precise ecological restoration and environmental stewardship by local governments.

Results indicate HAI and EQI on the QTP both increased from 2000 to 2020. The areas with high HAI are mainly distributed in big cities like Xining, Lhasa, Haidong, and Xigaze and along transit corridors. Human activities in Xining and Lhasa metropolitans have changed significantly. EQI gradually decreases from southeast to northwest spatially. Most of the cities and towns are located in an area with high HAI and high EQI. The region with high HAI and low EQI is smallest in size, but it has the highest risk of human-nature conflict, mainly located in southern Qinghai Lake, southern Brahlung Zangbo River, Gobi oases, and western transport lines.

Socioeconomic development and eco-environment on the QTP tend to be coordinated during 2000–2020. However, it has spatial heterogeneity, and the area ratio of the coordinated evolution to the conflict evolution is 5:1. The conflicted areas are mainly located outside the Lhasa metropolitan, south of the Hengduan Mountains, and along some new roads. Significant conflict areas in the latter decade comprise 8% less than in the former. Lhasa metropolitan displays a larger portion of conflict areas than Xining. Despite escalating human activity within Sanjiangyuan National Park over the past 20 years, environmental quality there has also risen, demonstrating harmonious coexistence between humanity and nature.

Acknowledgments

This study was supported by the Second Tibetan Plateau Scientific Expedition and Research Program (Grant 2019QZKK1005) and National Natural Science Foundation of China (Grant 42171210).

Conflict of Interest

The authors declare no conflicts of interest relevant to this study.

Data Availability Statement

Population density are obtained from WorldPop (https://hub.worldpop.org/project/categories?id=18); GDP density are available at Chen et al. (2022); nighttime light data are available at Zhang, Ren, et al. (2021); grazing intensity are available at Liu (2021); land use/cover data are obtained from the Resource and Environment Science and Data Center (https://www.resdc.cn/DOI/DOI.aspx?DOIID=54, registration required); NDVI data are available at Juntao (2022); net primary productivity are available at USGS (https://lpdaac.usgs.gov/products/mod17a3hgfv061/); MOD09A1 images are available at USGS (https://lpdaac.usgs.gov/products/mod09a1v006/); soil erosion are obtained from the Resource and Environment Science and Data Center (https://www.resdc.cn/data.aspx?DATAID=259, registration required); water conservation data are available at Wang (2022); PM_2.5 concentration are available at https://sites.wustl.edu/acag/datasets/surface-pm2-5/#V5.GL.02; urbanization rate and livestock number are available at statistical yearbooks of China's provinces (http://www.stats.gov.cn/sj/ndsj/, http://tjj.xizang.gov.cn/xxgk/tjxx/tjgb/).