EN中文

Journal of Geographical Sciences >

2023 , Vol. 33 >Issue 9: 1897 - 1920

DOI: https://doi.org/10.1007/s11442-023-2158-y

Research Articles

Exploring the response of ecosystem services to landscape change: A case study from eastern Qinghai province

  • MA Jiahao , 1 ,
  • WANG Xiaofeng , 2, 3, * ,
  • ZHOU Jitao 2 ,
  • JIA Zixu 1 ,
  • FENG Xiaoming 4 ,
  • WANG Xiaoxue 2 ,
  • ZHANG Xinrong 1 ,
  • TU You 1 ,
  • YAO Wenjie 1 ,
  • SUN Zechong 2 ,
  • HUANG Xiao 1
Expand
  • 1. School of Earth Science and Resources, Chang’an University, Xi’an 710054, China
  • 2. School of Land Engineering, Chang’an University, Xi’an 710054, China
  • 3. Key Laboratory of Shaanxi Land Consolidation, Xi’an 710054, China
  • 4. State Key Laboratory of Urban and Regional Ecology, Research Center for Eco-Environmental Science, CAS, Beijing 100085, China
*Wang Xiaofeng (1977-), Professor, specialized in ecological remote sensing and ecosystem services. E-mail:

Ma Jiahao (1998-), PhD Candidate, specialized in ecological remote sensing and ecosystem services. E-mail:

Received date: 2022-10-14

  Accepted date: 2023-04-13

  Online published: 2023-10-08

Supported by

The Second Tibetan Plateau Scientific Expedition and Research Program(2019QZKK0405)

The Chinese Academy of Sciences, Strategic Pilot Science and Technology Project (Class A)(No.XDA2002040201)

The Fundamental Research Funds for the Central Universities, CHD(300102352201)

Fold

Abstract

The degradation of ecosystem structure and function on the Qinghai-Tibet Plateau is the result of a combination of natural and anthropogenic factors, with landscape change driven by global change and human activities being one of the major ecological challenges facing the region. This study analyzed the spatiotemporal characteristics of ecosystem services (ESs) and landscape patterns in eastern Qinghai province (EQHP) from 2000 to 2018 using multisource datasets and landscape indices. Three ecosystem service bundles (ESBs) were identified using the self-organizing map (SOM), and changes in ecosystem structure and function were analyzed through bundle-landscaped spatial combinations. The study also explored the interactions between ESs and natural and human factors using redundancy analysis (RDA). We revealed an increase in total ecosystem service in the EQHP from 1.59 in 2000 to 1.69 in 2018, with a significant change in landscape patterns driven by the conversion of unused land to grassland in the southwest. Forestland, grassland, and unused land were identified as important to the supply of ESs. In comparison to human activities, natural environmental factors were found to have a stronger impact on changes in ESs, with vegetation, meteorology, soil texture, and landscape composition being the main driving factors. However, the role of driving factors within different ESBs varied significantly. Exploring the response of ecosystem services to changes in landscape patterns can provide valuable insights for achieving sustainable ecological management and contribute to ecological restoration efforts.

Key words： eastern Qinghai province; ecosystem service bundles; land use change; landscape pattern; ecological management

Cite this article

MA Jiahao , WANG Xiaofeng , ZHOU Jitao , JIA Zixu , FENG Xiaoming , WANG Xiaoxue , ZHANG Xinrong , TU You , YAO Wenjie , SUN Zechong , HUANG Xiao . Exploring the response of ecosystem services to landscape change: A case study from eastern Qinghai province[J]. Journal of Geographical Sciences, 2023 , 33(9) : 1897 -1920 . DOI: 10.1007/s11442-023-2158-y

1 Introduction

Global climate change and human activities are imposing significant pressure on ecosystems (Hu et al., 2022). As a result of rapid economic and social development, humans have had a profound impact on ecosystems, leading to ecological degradation, loss of biodiversity, and a decline in vital ecosystem services (ESs) (Bullock et al., 2011; Xu et al., 2017; Liu et al., 2022b). The Millennium Ecosystem Assessment indicated that 60% of global ESs were degraded (MEA, 2005). As a bridge between human society and natural ecosystems, the conservation and improvement of ESs is essential for achieving sustainable development (Xu et al., 2017; Wang et al., 2023). However, ESs do not exist in isolation, they interact with each other (Dou et al., 2020). These interactions arise when multiple services respond to driving factors or when a change in one service results in a corresponding change in other services (Bennett et al., 2009). How to effectively manage multiple ESs across landscapes is a key challenge in ecosystem management (Raudsepp-Hearne et al., 2010) and one of the most pressing areas of sustainability research (Renard et al., 2015; Zhang et al., 2022b).
Landscape patterns are an important manifestation of land use (Li et al., 2021), and related studies have shown that both landscape composition and configuration have important impacts on ESs (Liu et al., 2020; Ma et al., 2022). Landscape patterns are determined from the composition and configuration of landscape elements, which in turn influence the supply of ESs through ecosystem structure, function, and ecological processes (Turner, 2005). Changes in land use impact ecosystem structure and function and play a pivotal role in ESs provision (Wang et al., 2017; Liu et al., 2021). Therefore, a more realistic assessment of ESs provision must consider the additional benefits offered by diverse ecosystems or land use types at the landscape scale (Frank et al., 2012; Bastian et al., 2014). To obtain a deeper comprehension of how landscape composition and configuration impact ES provision, scholars have introduced the concept of “landscape services” (LSs), which are defined as ESs that are influenced by landscape patterns and determined by the contribution of landscapes and landscape elements to human well-being (Bastian et al., 2014; Duarte et al., 2018).
Changes in landscape patterns can have significant impacts on ecosystem structure and function, leading to corresponding changes in ecological processes, and ultimately resulting in a range of responses in landscape patterns themselves (Duarte et al., 2018). By exploring the relationships between ESs and changes in landscape patterns, we can improve the sustainable management of landscape ecology in the Qinghai-Tibet Plateau region. Many scholars have studied the evolution of landscape patterns and the changing functions of ESs in this region. For instance, Hou et al. (2020) quantitatively assessed the relationships and influencing factors between multiple landscape services to construct sustainable landscape models. Li et al. (2021) analyzed the landscape pattern dynamics of the Yarlung Tsangpo River basin using the landscape index method. Duan et al. (2023) explored the effects of land use change on landscape patterns and the value of ESs in the Sanjiangyuan area. Liu et al. (2022a) investigated the links among land use, ESs, and human well-being in the Yushu region of Qinghai province, providing support for regional ecological management. However, most of these studies have focused on changes in only one aspect of landscape structure or ecosystem functions, or have only briefly analyzed the two without effectively linking them.
Ecosystem service bundles (ESBs) are defined as a recurring combination of multiple types of ecosystem services that occur spatially and temporally (Raudsepp-Hearne et al., 2010). By exploring the interaction between ecosystem function and landscape structure, ESBs provide a novel approach for spatially mapping the linkage between specific ESs and human-dominated land use and associated ESs (Dick et al., 2011; Li et al., 2022). The provision of ESBs is contingent upon ecosystem function and landscape structure, which are in turn influenced by various socioecological factors (Gonzalez-Ollauri and Mickovski, 2017; Karimi et al., 2021). Multiple ESs tend to aggregate in homogeneous or heterogeneous landscapes to form multiple types of ESBs (Mouchet et al., 2014; Shen et al., 2020). Research on ESBs can assist in elucidating the relationships among various ESs, which can ultimately improve multifunctional landscape management. Therefore, comprehending the intricate relationships among multiple ESs and the underlying mechanisms is crucial for promoting sustainable landscape management (Bennett et al., 2009; Lyu et al., 2019; CAO et al., 2020).
The eastern Qinghai province (EQHP), located at the intersection of the Yellow River Basin and the Qinghai-Tibet Plateau, plays a vital role in maintaining regional ecological security. Despite its ecological significance, the EQHP is one of the most densely populated areas of the Qinghai-Tibet Plateau, encompassing important ecological and socioeconomic zones, such as Qinghai Lake and the Huangshui Valley. The region faces ecological challenges due to extensive human activities that result in land desertification, soil erosion, grassland degradation, and biodiversity reduction (Qi et al., 2020; Li et al., 2022). To mitigate these challenges, the government has implemented ecological restoration protection projects such as the Returning Grazing to Grassland Project (Zhao et al., 2022a) and the Natural Forest Protection Project (Yan et al., 2022). However, the effects of these projects on the regional ecosystem remain unclear. To address this gap in knowledge, this study focused on five key ESs: net primary productivity (NPP), water yield (WY), soil conservation (SC), sandstorm prevention (SP), and habitat quality (HQ). The study aimed to (1) quantify the changes in these ESs and landscape patterns in the EQHP from 2000-2018; (2) analyze the spatial mapping links between landscape patterns and ESs and identify the response of ESs to different natural and human factors; and (3) propose recommendations for the development of ecological management policies.

2 Materials and methods

The conceptual framework of our study is depicted in Figure 1 and aims to investigate the response of ESs to landscape change in the EQHP and the way this information can be used for ecological management. Specifically, we assessed five ESs, namely, NPP, WY, SC, SP, and HQ, and identified three ESBs. We then quantified landscape pattern characteristics using the landscape index method and analyzed changes in ecosystem structure and function from a spatial mapping perspective through bundle-landscape combinations. Finally, we examined the interactions between ESs and natural and human factors to provide policy recommendations for the ecological restoration and management of ESs.
Figure 1 Study flowchart

2.1 Study area

The EQHP area is located in the northeastern part of the Qinghai-Tibet Plateau and is a transition zone between the Yellow River Basin and the plateau (Figure 2). It has a total area of 116,700 km2 and is divided into 26 county-level administrative districts in 5 cities, including Haidong city, Xining city, Haibei Tibetan autonomous prefecture, Hainan Tibetan autonomous prefecture, and Huangnan Tibetan autonomous prefecture. In 2018, the population of the EQHP was approximately 4.8887 million, with more than 80% of the population concentrated in Qinghai province and with a total GDP of approximately 206.795 billion yuan, making it the political, economic, and cultural centre region of Qinghai province. The elevation of the study area ranges from approximately 1681 m to 5280 m, gradually decreasing from the southwest to the northeast, and the main land use types are forestland, grassland, and unused land. The annual average temperature is -9°C to 9°C, and the annual average precipitation is between 260 mm and 800 mm. The eastern part of the area has a lower topography than the western part, and the hydrothermal conditions are better than those in other regions (Jiang et al., 2020a). The region faces ecological problems such as grassland degradation, soil erosion, and desertification due to human activities and climate change (Qi et al., 2020; Li et al., 2022). The government has implemented ecological engineering measures such as the Returning Grazing to Grassland Project and the Natural Forest Protection Project to address these issues (Yan et al., 2022; Zhao et al., 2022a).
Figure 2 Geographical location (a), vegetation coverage (b), and elevation (c) of the eastern Qinghai province

2.2 Data sources

We used a multisource dataset (Table 1), which mainly includes meteorological data, soil data, vegetation data, topographic data, socioeconomic data, and other auxiliary data. Meteorological data such as precipitation, temperature, wind speed, and solar radiation were interpolated as 1 km grid data using ANUSPLIN software. The soil data used in this research mainly included soil organic matter, sand, silt and clay content, soil depth, root depth, etc. The vegetation coverage was calculated using the pixel dichotomy method for NDVI data. DEM data were used to extract slope and topographic relief in ArcMap 10.2. The land use and land cover (LULC) data were obtained from the China land use/cover dataset (CLUD) (Ning et al., 2018). The CLUD was generated based on multisource remote sensing data and field survey data. The comprehensive accuracy was over 90%, which met the needs of this study.
Table 1 Multiple source datasets
Data type Specific data Time resolution Spatial resolution Data sources
Meteorological data



Soil data		 Precipitation, temperature, solar radiation, wind speed day — China National Meteorological Science Data Center (http://data.cma.cn)
Actual evapotranspiration year 1 km (https://doi.org/10.6084/m9.figshare.12278684.v5) from reference (Yin et al., 2021)
Harmonized World Soil Database (HWSD) — 1 km https://www.fao.org/
Vegetation
data		 Normalized vegetation dataset (NDVI) month/year 1 km Resources and Environment Science and Data center, Chinese Academy of Sciences (http://www.resdc.cn/)
Topographic data Digital elevation model (DEM) — 90 m Geospatial Data Cloud (http://www.gscloud.cn/)
Socioeconomic data Resident point
Road network		 — — Resources and Environment Science and Data center, Chinese Academy of Sciences (http:// www.resdc.cn/); OpenStreetMap (https:// download.geofabrik.de)
Other data Long-term series of daily snow depth dataset in China day 25 km National Qinghai-Tibet Plateau Scientific Data Center (http://data.tpdc.ac.cn/zh-hans/)
Land use and land cover data (LULC) year 1 km Resources and Environment Science and Data Center, Chinese Academy of Sciences (http:// www.resdc.cn/)

2.3 Quantifying multiple ecosystem services

2.3.1 Carbon sequestration

Carbon sequestration (CS) plays an important role in maintaining the carbon cycle of terrestrial ecosystems and regulating global climate change. The study area has a strong carbon sequestration capacity due to its large forest and grassland area. This study calculated the net primary productivity (NPP) of vegetation based on the CASA model (Potter et al., 1993) to characterize the carbon sequestration capacity of vegetation in the EQHP. The main calculation formula is as follows:
N P P ( x , t ) = A P A R ( x , t ) × ε ( x , t )
A P A R ( x , t ) = S O L ( x , t ) × 0.5 × F P A R ( x , t )
ε ( x , t ) = T ε ( x , t ) × W ε ( x , t ) × ε max
where NPP(x,t) is the net primary production of pixel x at time t (gC·m‒2); APAR(x,t) is the photosynthetically active radiation absorbed by vegetation (MJ·m‒2); ε(x,t) is the actual light use efficiency (gC·MJ‒1); SOL(x,t) is the total solar radiation (MJ·m‒2); FPAR(x,t) is the fraction of photosynthetically active radiation; Tε(x,t) is the temperature stress coefficient; Wε(x,t) is the water stress coefficient; and max is the maximum light use efficiency of the specific biome under ideal conditions.

2.3.2 Water yield

The Qinghai-Tibet Plateau, known as the ‘Asian Water Tower’, is the source of nine major rivers in Asia and provides freshwater resources to over 1.5 billion people (Liu et al., 2018), making water yield (WY) services an important provisioning service in the area. This study used the water yield module of the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) system model to quantify WY based on water balance principles (Zhang et al., 2001; Zhang et al., 2004; Wang et al., 2021a). The main calculation formula is as follows:
W Y ( x ) = 1 − A E T ( x ) P ( x ) × P ( x )
P E T ( x ) = K C ( x ) × E T 0 ( x )
A E T ( x ) P ( x ) = 1 + P E T ( x ) P ( x ) − 1 + P E T ( x ) P ( x ) w 1 / w
w ( x ) = A W C ( x ) × Z P ( x ) + 1.25
where WY(x) is the annual WY (mm) of a certain land use type in pixel x (mm); AET(x) is the annual evaporation of pixel x; P(x) is the annual precipitation (mm) of pixel x; PET(x) is the potential evapotranspiration of pixel x; ET0(x) is the reference (vegetation) evapotranspiration; KC(x) is the plant evapotranspiration coefficient; AWC(x) is the available water content of plants; w(x) is the empirical parameter; and Z is the Zhang coefficient (Zhang et al., 2001; 2004).

2.3.3 Soil conservation

Soil conservation (SC) is an important regulating service that reduces soil erosion and maintains land fertility. In this study, the Revised Universal Soil Loss Equation (RUSLE) model (Wischmeier and Smith, 1965; Renard et al., 1991; Shao et al., 2023) was used to calculate soil water erosion, and the difference between potential soil erosion and actual soil erosion was used to quantitatively assess SC. The main calculation formula is as follows:
S C = A P − A r
A P = R × K × L S
A r = R × K × L S × C × P
where SC represents the average annual soil conservation (t·ha‒1·a‒1); AP represents the potential soil erosion (t·ha‒1·a‒1); Ar represents the actual soil erosion (t·ha‒1·a‒1); R is the rainfall erosivity factor (MJ·mm·ha‒1·h‒1·a‒1); K is the soil erodibility factor (t·ha‒1·h·ha‒1·MJ‒1·mm‒1); L and S are the slope length and slope factor, respectively; and P is the conservation support practice factor.

2.3.4 Sandstorm prevention

Sandstorm prevention (SP) is the suppression and fixation of sandstorms by vegetated ecosystems and is the most important service provided by desert ecosystems in arid and semiarid regions. Based on the Revised Wind Erosion Equation (RWEQ) (Chi et al., 2019; Teng et al., 2021) model, we calculated soil wind erosion in the EQHP and quantitatively evaluated SP based on the difference between the potential soil wind erosion and the actual soil wind erosion. The main calculation formula is as follows:
S P = S L P − S L = 2 Z S P 2 Q max P × e − ( Z / S P ) 2 − 2 Z S 2 Q max × e − ( Z / S ) 2
Q max P = 109.8 × ( W F × E F × SCF × K ′ )
S P = 150.71 × ( W F × E F × SCF × K ′ ) − 0.3711
Q max = 109.8 × ( W F × E F × SCF × K ′ × C )
S = 150.71 × ( W F × E F × SCF × K ′ × C ) − 0.3711
where SP represents the average annual sandstorm prevention (kg·m‒2); Z represents the downwind distance (m); SLP and SL are the potential and actual wind erosion (kg·m‒2), respectively; QmaxP and Qmax are the potential and actual maximum sand transport capacity, respectively; S is the length of the key plot (m); SP is the length of the potential key plot (m); WF is the climate factor (kg·m‒1); EF is the soil erodibility factor; SCF is the soil crust factor; C is the vegetation coverage factor; and K' is the surface roughness factor.

2.3.5 Habitat quality

Habitat quality (HQ) refers to the ability of an ecosystem to provide suitable and sustainable living conditions for individuals and populations (Hall et al., 1997; Chen et al., 2020). HQ can characterize the status of regional biodiversity and is an important support service. Important parameters such as weights, suitability, and sensitivity were based on (Gou et al., 2021; Teng et al., 2022; Zhao et al., 2022b). The main calculation formula is as follows:
Q x j = H j 1 − D x j z D x j z + k z
D x j = ∑ r = 1 R ∑ y = 1 Y r W r ∑ r = 1 R W r r y i r x y β x S j r
where Qxj is the habitat quality in grid cell x with land use type j; Hj is the habitat suitability of land use type j; k is a semisaturated constant, and z is a model default value; Dxj is the total threat level in grid cell x with land use type j; R is the number of threat sources; Yr is the total raster number of threat source r; Wr is the weighting of threat r; ry is the threat value in grid cell y; irxy is the impact of the habitat of threat r in grid cell x in grid cell y; y indices all grid cells on r’s raster map; βx is the accessibility in grid cell x; and Sjr is the sensitivity of land use type j to threat r.

2.3.6 Total ecosystem service

We normalized the five ESs to a range between 0 and 1, aggregated multiple ESs in this study into one indicator, and then added them together to obtain the total ecosystem service (TES) (Maes et al., 2015; Ma et al., 2022). The TES reflects the overall supply of these key ESs in the EQHP.
T E S = ∑ m = 1 5 E S m , n − E S m , min E S m , max − E S m , min
where ESm,n is the value of ES m in pixel n, and ESm,min and ESm,max are the minimum and maximum values of ES m, respectively.

2.4 Identifying the ecosystem services bundles

The self-organizing map (SOM), proposed by the Finnish scholar Kohonen (1982), is an adaptive, self-organizing, and self-learning unsupervised artificial neural network (Kohonen., 1990). The SOM integrates the ability to downscale and cluster (Cord et al., 2017; Shen et al., 2020). It preserves the topology of the input data and adds spatial information to the analysis through the nearest-neighbor relationship function, applying a competitive learning strategy with only one output unit active at each moment. The SOM relies on cortical neurons competing with each other to optimally adjust the network structure appropriately to achieve classification processing and improve the objectivity and accuracy of the classification. Therefore, the SOM method was used to identify ESBs. We normalized the data of the five ESs and then trained them based on the selforgmap function in MATLAB R2018b, and finally used the Calinski‒Harabasz index to determine the optimal number of ESBs.

2.5 Landscape pattern indices

A landscape index highly condenses landscape pattern information and is a simple quantitative index reflecting landscape structural composition and spatial configuration characteristics (Wu., 2007). Different types of landscape indices reflect the changing characteristics of landscape structure. To more comprehensively express the changing characteristics of landscape patterns in the EQHP, nine indices (Table 2) were selected for quantitative analysis in terms of landscape composition, landscape fragmentation, landscape complexity, landscape aggregation, landscape connectivity, and landscape diversity (Bi et al., 2021; Li et al., 2021; Ma et al., 2022). In this study, the PLAND index was used as a driving factor to analyze changes in ESs, while the remaining indices were used to analyze overall changes at the landscape level. All indices were calculated by Fragstats 4.2 software based on the moving window method.
Table 2 Landscape pattern indices used in this study
Landscape indices Indices description Units Range
Percentage of Landscape (PLAND) Percentage of a certain landscape
type in the total landscape area		 Percentage 0≤ PLAND ≤ 100
Patch Density (PD) Number of patches per unit area Number per 100 ha PD>0
Mean Patch Area (AREA_MN) Average area of each patch in the landscape or class ha AREA_MN>0
Edge Density (ED) Proportion of the sum lengths of all edge segments involving each patch type to the total landscape area m per ha ED≥0
Area-Weighted Mean
Shape Index (SHAPE_AM)		 Sum of the average shape factor of each patch type multiplied by the weight of the patch area in the
landscape area		 Dimensionless SHAPE_AM≥1
Aggregation Index (AI) Nonrandomness or aggregation
degree of different patch types in
the landscape		 Percentage 0≤ AI ≤ 100
Contagion Index (CONTAG) Agglomeration degree or extension trend of different patch types in the landscape Percentage 0< CONTAG ≤ 100
Shannon’s Diversity Index (SHDI) Probability and diversity of patch types in the landscape Dimensionless SHDI≥0
Shannon’s Evenness Index (SHEI) Distribution uniformity of each
component in the landscape		 Dimensionless 0≤ SHEI ≤ 1

2.6 Identifying driving factors of ecosystem services change

ESBs are generated by the spatial and temporal cooccurrence of ESs, which are influenced by a combination of human activities and natural environmental factors (Shen et al., 2021; Teng et al., 2022). We selected 19 factors to explore the temporal and spatial evolution of ESs. The human factors included landscape composition factors and socioeconomic factors. The landscape composition factors included the percentage of cropland, forestland, grassland, water body, built-up land, and unused land landscape. The socioeconomic factors included the distance from residential points and the distance from the road. The natural factors included meteorology (precipitation, temperature, solar radiation, actual evapotranspiration, and wind speed), vegetation coverage, topography (slope and surface roughness), and soil texture (soil organic matter, sand, silt, and clay). Redundancy analysis (RDA) is an analytical method applied to multiple linear regressions with multiple explanatory and response variables (Mouchet et al., 2014; Teng et al., 2022; Zhang et al., 2022c); therefore, RDA was used to explain the relationships between ESs and landscape patterns as well as socioecological factors. It is important to note that the largest gradient length values of detrended correspondence analysis were all less than 3; thus, RDA was used in this study, and the multicollinearity among factors was eliminated by a forward selection step, which was performed by Canoco 4.5 software.

3 Results

3.1 Spatial and temporal changes in ecosystem services

The spatial distribution of multiple ESs in the EQHP area from 2000 to 2018 showed significant differences (Figure 3), with NPP, WY, SC, and HQ showing increasing trends; however, SP did not show this trend. NPP increased from 421.05 gC·m‒2 to 558.59 gC·m‒2 from 2000 to 2018, an increase of 32.67%, with high NPP areas mainly present in Haibei prefecture, Xining city, and Hainan prefecture, where there was high forestland and grassland cover and NPP showed a spatial distribution pattern of being high from north to south and low in central areas. WY increased at a rate of 12.19 mm·a‒1, from 256.84 mm in 2000 to 488.38 mm in 2018. The high-value areas were located in areas with higher altitudes and were mainly distributed in Huangnan prefecture, southern Hainan prefecture, and northern Haibei prefecture. SC increased from 75.48 t·ha‒1 to 201.63 t·ha‒1 over 19 years, showing a distribution pattern of high in the surrounding area and low in the middle, with high values distributed south of Hainan prefecture, Xining city, Haidong city, and eastern Haibei prefecture. SP decreased from 7.36 t·ha‒1 in 2000 to 5.42 t·ha‒1 in 2018, and the high-value areas were mainly located surrounding the Qinghai Lake Basin and in the central and northern parts of Hainan prefecture. HQ increased slightly over 19 years, from 0.69 in 2000 to 0.70 in 2018. Benefiting from the rise in NPP, WY, SC, and HQ, TES showed an upwards trend overall, increasing from 1.59 to 1.69, with high values mainly distributed in Haibei prefecture, Huangnan prefecture and southern Hainan prefecture, and the low-value areas were significantly reduced.
Figure 3 Spatiotemporal distribution of ESs in the eastern Qinghai province in 2000 and 2018

3.2 Spatiotemporal changes in and characteristics of ecosystem service bundles

The Calinski‒Harabasz index value, which corresponds to the number of bundles, indicated the optimal number of bundles in this study. Based on the SOM method, we identified three ESBs, and the spatial distributions of ESBs in 2000 and 2018 are presented in Figure 4. There was similarity in the spatial and temporal distributions of ESBs and TES, with ESB3 having the highest overall level of service provision and ESB1 and ESB2 having a lower overall level of service provision. The proportion of ESB1 decreased from 19.44% to 17.10% during the study period, with a low level of supply for all ESs except WY. ESB1 was scattered across the northern-central part of the EQHP. The proportion of ESB2 decreased from 25.03% to 18.39%, and ESB2 was mainly concentrated in the central part of the EQHP, including Hainan prefecture, southern Xining city, south-central Haidong city, and Qinghai Lake. These areas had a high level of HQ supply, but NPP, WY, and SC supply levels were the lowest of the three bundles. ESB3 had the highest percentage of area, increasing from 55.53% in 2000 to 64.51% in 2018, with NPP, WY, SC, and HQ service provision levels being the highest in the three bundles. ESB3 was mainly in the northern and southern parts of the study area, including north-central Haibei prefecture, Huangnan prefecture and parts of the other three cities. The spatial distribution pattern of ESBs changed significantly over 19 years, with the largest change occurring in the ESB3 area, which increased by 10,444 km2, and 6848 km2 and 8893 km2 of ESB1 and ESB2 were converted into ESB3, respectively.
Figure 4 Spatiotemporal distribution and transfer of ecosystem service bundles in the eastern Qinghai province in 2000 and 2018

3.3 Spatiotemporal change characteristics of the landscape in the eastern Qinghai

Figure 5 displays the spatial distributions of LULC in the EQHP in 2000 and 2018 and shows significant changes in land use types over the study period. The predominant land use types in the study area were forestland, grassland, and unused land, which together accounted for approximately 87% of the EQHP. Grassland, with the highest proportion of approximately 60%, was the most widely distributed land use type across the study area, while unused land was the second largest land use type and was primarily found in Haibei prefecture, Hainan prefecture, and Huangnan prefecture. Forestland, accounting for approximately 13% of the study area, was primarily distributed in northern Haibei prefecture, southern Hainan prefecture, and northern Huangnan prefecture and around the Huangshui River basin. Cropland and built-up land accounted for a relatively small proportion of the study area, at approximately 7% of the EQHP, mainly in the Huangshui River basin, around Qinghai Lake, and in parts of Hainan Prefecture. In general, grassland increased by 2754 km2 due to the conversion of forestland and unused land. Water bodies increased by 203 km2. Cropland and built-up land expanded, increasing by 251 km2 and 173 km2, respectively. Forestland experienced a slight decrease of 0.14%, while unused land had the most significant reduction (2.76%), with 3219 km2 mostly converted to grassland.
Figure 5 Spatiotemporal distribution and transfer of land use/land cover in the eastern Qinghai province in 2000 and 2018

3.4 Analysis of landscape pattern change characteristics in the eastern Qinghai

Window scale selection when using the moving window method to calculate the landscape indices has an impact on the results: a smaller window will not reflect the variability within the landscape, while a larger window will result in the loss of spatial information about the landscape indices. To determine the appropriate window size, we calculated the PD, ED, AI and SHDI at different scale windows of 3000 m, 5000 m, 7000 m, 9000 m, 11,000 m, 13,000 m and 15,000 m to determine the appropriate window scale for the study area by comparing the changes in landscape indices at the different scales. Boxplots show the distribution, error, median and mean of the data, allowing comparison of data changes in the landscape indices at different window scales. Figure 6 shows that the distribution of PD and AI data is unstable at the 3000-5000 m scale, with the mean value decreasing with increasing window scale and stabilizing in the 7000-11,000 m scale range, while ED and SHDI increase with increasing window scale; in addition, the decreasing rate of increase tends to slow down at the 7000-11,000 m window. Therefore, taking into account the stability of the landscape index, the spatial resolution of the LULC and the extent of the study area, we chose 9000 m as the window scale for calculating the landscape indices to analyze the changing landscape pattern characteristics of the EQHP.
Figure 6 Boxplots of the landscape indices at the landscape scale
Eight indices were selected to analyze the changing characteristics of the EQHP landscape pattern from 2000 to 2018. All eight indices decreased, except AREA_MN, AI and CONTAG, which increased from 1474.63, 69.80 and 34.62 to 1656.73, 72.33 and 37.16, respectively. The distribution of the landscape indices exhibited significant spatial heterogeneity, as depicted in Figure 7. PD, ED, SHAPE_AM, and SHDI showed similar spatial patterns, with high-value areas predominantly concentrated in the eastern and northern parts of the EQHP. These areas were characterized by low elevations, high human activity levels, and diverse LULC types, including cropland, built-up land, forestland, and grassland. The expansion of cropland and built-up land in the eastern and northern regions contributed to landscape fragmentation, while the conversion of unused land to grassland in the southwest resulted in the aggregation of landscape patches and the increasing of patch size, leading to significant increases in AREA_MN and AI in that region. Moreover, the conversion of unused land to grassland also had temporal and spatial impacts on ED and SHAPE_AM, resulting in a reduction in landscape complexity in the EQHP. SHDI and SHEI served as useful metrics for describing the diversity of the landscape and showed an increasing trend from the southwest to the northeast. In comparison to the other regions, the northeastern region exhibited a greater diversity of LULC types, including cropland, built-up land, forestland, and grassland, whereas the southwestern region was characterized by a lower SHDI and SHEI, indicating a lower diversity of LULC types. Changes in LULC types significantly affected the landscape diversity level, and the conversion of unused land to grassland in the southwest resulted in a reduction in landscape diversity as grassland patches combined. In summary, the temporal and spatial trends in the landscape indices in the EQHP area revealed that changes in LULC types resulted in an increase in landscape aggregation and connectivity and a decrease in landscape fragmentation, complexity, and diversity.
Figure 7 Spatiotemporal distribution of the landscape indices in the eastern Qinghai province in 2000 and 2018

3.5 Relationship between ecosystem service bundles and land use/land cover

We utilized a method of multiplying the ESB code by 10 and overlaying it with the LUCC data to obtain the corresponding ‘bundle-landscape’ combination code. The number of ‘bundle-landscape’ spatial combinations remained constant in both 2000 and 2018, and we identified 11 combinations of ‘bundle-landscape’ in the EQHP (Figure 8). Among these ‘bundle-landscape’ combinations, those with increasing areas were 11, 15, 24, 26, 32, and 33, whereas those with decreasing areas included 16, 22, 23, 34, and 36. The combination codes that displayed significant area changes were predominantly 16 (-3138), 23 (-8074), and 33 (+10,821), which indicated that changes in ESB1 were highly correlated with unused land, changes in ESB2 were mainly related to water bodies and grassland, and changes in ESB3 were mainly caused by alterations in grasslands. Specifically, ESB1 corresponded to cropland, built-up land, and unused land, while ESB2 and ESB3 corresponded to forestland, grassland, water bodies, and unused land. This result implies that the same landscape can provide more than one ESB, but at the same time, the ESBs offered by a particular landscape can also be distinct. For instance, cropland and built-up land were exclusively linked with ESB1 in this study, while unused land was associated with all three ESBs. The spatial distribution of the LULC and TES supply showed that forestland and grassland provided higher ESs, also indicating that the combined supply level of ESs in ESB3 was the highest of the three bundles. This ‘bundle-landscape’ spatial combination study found that changes in forestland, grassland, and unused land were important for the supply of ESs in the EQHP, and this was reflected in changes in ESBs.
Figure 8 Spatial combinations of ‘bundle-landscape’ in the eastern Qinghai province in 2000 and 2018. Code refers to the code of the ‘bundle-landscape’ combination; Area is the area of each ‘bundle-landscape’ combination (km2); CL represents cropland; FL represents forestland; GL represents grassland; WB represents water body; BL represents built-up land; UL represents unused land.

3.6 Interaction of ecosystem services with natural and human factors

The results from the RDA indicated that natural factors and human activities accounted for 60.4% and 61.3% of the changes in ESs in 2000 and 2018, respectively. The explanatory power of natural factors was found to be stronger than that of human activities in terms of the contribution rate. RDA1 and RDA2 together explained 52.5% of the variance in 2000 and 2018, as shown in Fig. 9. The impacts of the driving factors on changes in ESs were similar between the two years, with vegetation having the highest contribution (>0.3) among the driving factors, followed by meteorology, soil texture, and landscape composition while the contribution of landscape pattern was found to be increasing. Vegetation, forestland landscape, and grassland landscape were identified as the main factors contributing to changes in NPP, and these factors were positively correlated with NPP. Precipitation was found to be positively correlated with WY, while temperature and actual evapotranspiration were negatively correlated with WY. SC was found to be influenced by a combination of factors such as soil texture, topography, and meteorology. SP was negatively correlated with vegetation and actual evapotranspiration and positively correlated with the percentage of unused land landscape. The change in HQ was mainly influenced by landscape pattern, with higher proportions of cropland and unused land landscapes reducing HQ, while higher composition of forestland and grassland increased HQ. Finally, changes in TES in response to various driving factors were found to be highly similar to those in NPP.
Figure 9 Redundancy analysis biplots and contribution rate of driving factors for the ESs in the eastern Qinghai province (Driving factors: V1, vegetation coverage; M1, precipitation; M2, temperature; M3, solar radiation; M4, actual evapotranspiration; S1, soil organic matter; S2, silt; S3, sand; T1, slope; T2, surface roughness; D1, distance from residential points; D2, distance from road; L1, percentage of cropland landscape; L2, percentage of forestland landscape; L3, percentage of grassland landscape; L6, percentage of unused land landscape)
The contribution of the vegetation factor to the change in ESs in each ESB was generally high, similar to that in EQHP. However, the remaining factors showed significant differences in their impact on the change in ESs in each ESB (Table 3). In ESB1, socioeconomic factors and landscape composition had a high contribution rates, with socioeconomic factors showing a greater influence on the change in ESs than other factors (>0.15). ESB2 played a strong role in landscape composition and soil texture. In ESB3, in comparison to that of the other factors, the influence of meteorological factors on ESs was higher, and the impact of landscape composition was significantly lower compared that in the other bundles.
Table 3 Contribution rate of driving factors to changes in ESs in each ESB
ESBs 2000ESB1 2000ESB2 2000ESB3 2018ESB1 2018ESB2 2018ESB3
Contribution rate
of variables		 V1 0.33 0.39 0.17 0.58 0.24 0.15
M1 0.30 0.15 0.26 0.21 0.05 0.26
M2 0.36 0.10 0.17 0.33 0.01 0.08
M3 0.09 0.11 0.05 0.14 0.05 0.02
M4 0.24 — 0.11 0.23 0.37 0.12
M5 — 0.09 0.02 0.03 0.01
S1 0.10 0.10 0.04 0.06 0.05 0.03
S2 — 0.25 0.01 0.12 0.24 0.04
S3 0.13 0.24 0.03 0.13 0.33 0.04
S4 — 0.19 0.03 0.16 0.22 0.03
T1 0.09 0.11 0.07 0.07 0.12 0.04
T2 0.13 — 0.08 0.05 0.19 0.06
D1 — 0.08 — 0.28 0.10 0.02
D2 0.16 0.05 0.03 0.23 0.03 0.03
L1 0.33 — — 0.40 — —
L2 — 0.04 0.09 — 0.02 0.08
L3 — 0.22 0.03 — 0.33 0.01
L4 — 0.35 — — 0.42 —
L6 0.31 — 0.08 0.48 0.09 0.05

Driving factors: V1, vegetation coverage; M1, precipitation; M2, temperature; M3, solar radiation; M4, actual evapotranspiration; M5, wind speed; S1, soil organic matter; S2, silt; S3, sand; S4, clay; T1, slope; T2, surface roughness; D1, distance from residential points; D2, distance from road; L1, percentage of cropland landscape; L2, percentage of forestland landscape; L3, percentage of grassland landscape; L4, percentage of water body landscape; L6, percentage of unused land landscape

4 Discussion

4.1 Impacts of natural and human activity factors on ecosystem services

There are both commonalities and disparities in the impacts of the driving factors on both the overall EQHP region and alterations in ESs within each ESB. The heterogeneity in the contribution rates of these drivers to ESs accentuates the potent spatial variability in their outcomes. Among these factors, vegetation and meteorological factors generally played a stronger role in the EQHP and ESBs, as vegetation provides a large number of ESs (Riis et al., 2020; Pérez-Silos et al., 2021), and vegetation development is closely intertwined with climatic conditions, particularly changes in precipitation and temperature (Chen et al., 2020). While the explanatory power of socioeconomic factors in EQHP was low, it was markedly higher in ESB1 than in the other two bundles and the entire area, and this effect was further accentuated over time. This result may be attributed to ESB1 encompassing the region with the highest level of human activity, where the impact of socioeconomic factors gradually intensified as human activity escalated, emphasizing the pivotal role of socioeconomic factors in shaping ESs.
Numerous studies have confirmed the combined effect of natural factors and human activities on changes in ESs of ESBs (Liu et al., 2019; Haberman and Bennett, 2019; Shen et al., 2021). Soil, as the basis for supporting vegetation growth, has strong explanatory power in terms of the variation in ESs of ESB2 based on its texture. The wide variation in soil texture across the ESB2 area may limit the supply of water and nutrients required for vegetation growth in different areas (Jiang et al., 2020b; He et al., 2021). Landscape composition factors also represent certain human activities. From 2000 to 2018, the LULC types changed to varying degrees with an obvious impact on ESs within the area of ESBs (Quintas-Soriano et al., 2016). Landscape composition factors in ESB1 and ESB2 had a high contribution rate. The proximity of ESB1 and ESB2 indicated that human activities such as the social economy and landscape composition also have a certain impact on ESB2. Clarification of the interactions between socioecological factors and ESs is essential for sustainable ecosystem management (Yang et al., 2021; Wang et al., 2022), and the relationship between ESs and socioecological factors may not be consistent across ESBs with different characteristics (Li et al., 2022; Teng et al., 2022). The identification of factors that influence ESs and how the influence of various factors on ESs changes over time can help us understand the mechanisms of ESB formation and assist managers in formulating reasonable ecological policies for different regions. This approach can promote a balance between socioeconomic development and ecological conservation.

4.2 Implications for ecological management of the eastern Qinghai

The Qinghai-Tibet Plateau, also known as the ‘Third Pole’ of the Earth, provides a multitude of ESs for human well-being (Li et al., 2018; Hou et al., 2020). However, the plateau’s delicate ecosystems are highly sensitive to human activities due to its extreme altitude and cold environment (Li et al., 2020; Wang et al., 2021b). With rapid economic development and increased human activities, the conflict between resource exploitation and ecological conservation has become increasingly apparent, necessitating protective measures for the plateau’s fragile ecosystems. Regarding landscape changes in the study area, the conversion of unused land to grassland has contributed to increased landscape aggregation and connectivity. However, the continuous expansion of built-up land and cropland, albeit small in a proportion, has weakened ecological links between different regions and created a fragmented and complex landscape, hindering the improvement of ESs, as observed in ESB1. To address these challenges, we propose controlling the expansion of built-up areas and cropland, protecting ecological land such as forests and grasslands, properly developing unused land, strengthening ecological corridor construction, and enhancing landscape aggregation and connectivity to improve the supply of ESs.
Our findings indicate that various ecological management policies should be tailored to different regions. ESB1 encompasses cropland and urban areas and lies at a lower altitude with favorable hydrothermal conditions. It also has the highest level of human activity among the EQHP areas. Socioeconomic factors play a crucial role here, but the supply of ecosystem services is low. Therefore, in the ESB1 region, we must focus on the reasonable utilization and development of land resources. We must also enhance water resource infrastructure construction to ensure safe irrigation and drinking water, as well as actively promote novel agricultural technologies. In light of the above information, a rigorous protection system for general farmland should be instituted to ensure food security and maximize the ecological benefits of cropland. At the same time, certain ecological protection measures must be taken, such as urban greening and ecologically sound industrial waste disposal. ESB2 encompasses Qinghai Lake, a national nature reserve and globally significant wetland, as well as a national 5A tourist attraction. In such regions, ecological and economic benefits should be simultaneously improved. A reasonable ecological compensation mechanism must be established based on the comprehensive consideration of ecological protection costs, development opportunity costs, and ecological service values. While pursuing economic benefits from tourism, biodiversity must be safeguarded, and the harmonious development of human society and nature should be promoted. ESB3 is endowed with vast forest and grassland resources and provides significant ecological functions. In recent years, the government has implemented a range of ecological engineering measures, such as the Returning Grazing to Grassland Project (Zhao et al., 2022a) and the Natural Forest Protection Project (Yan et al., 2022), which have led to significant ecological improvements. Ecological conservation and restoration measures should be persistently implemented in areas in ESB3. In severely degraded areas, grazing bans or short-term fencing must be enforced (Yin et al., 2019; Sun et al., 2020), while herders should receive subsidies to decrease grassland degradation through the implementation of grazing ban subsidies and grass-livestock balance incentives (Yin et al., 2019; Zhao et al., 2022a).

4.3 Limitations and future perspectives

The research utilized various models to estimate ESs. However, due to the lack of data, some parameters were not localized, leading to uncertainties and errors. Despite this limitation, related studies (Yu, 2020; Ma et al., 2021; Wang et al., 2021; Zhao et al., 2022a) have shown improvements in ecological restoration and ESs in the region. Future studies should improve the accuracy of ES calculations by using accurate data and explore the changes in ecosystem structure and function more accurately. While this study quantified five key provisioning, regulating, and supporting services, cultural services were not included due to difficulties in obtaining and quantifying data. The main reason for this lack of data is that it is more difficult to obtain and quantify data on the temporal dynamics of cultural services in the EQHP. Recent studies on cultural services have first considered how to quantify cultural services in the Qinghai-Tibet Plateau (Hou et al., 2022; Zhang et al., 2022a), and there is still a relative lack of research that incorporates the spatial and temporal dynamics of cultural services in the alpine region into ESBs. In future studies, cultural services should be quantified from spatial and temporal perspectives, and their relationship with ESs should be analyzed more comprehensively. Grazing intensity, ecological engineering, and socioeconomic policies were not included as driving factors due to data limitations, and further studies are needed to investigate the mechanisms of ES changes and ESB formation. Additionally, the study only explored interactions between ESs and landscape patterns at a 1 km scale, and future studies should analyze variations in the effects of ESs interacting with landscape patterns at different scales. Despite these limitations, the study can still provide a theoretical reference for sustainable ecological management on the Qinghai-Tibet Plateau and assist decision-makers in formulating management policies in different regions more rationally.

5 Conclusions

This study explored the response of ESs to changes in landscape patterns using multisource datasets of meteorology, soil, vegetation, topography, socioeconomics and land use. The objective of the study was to provide theoretical references for ecological restoration and management in the Qinghai-Tibet Plateau. First, the spatial and temporal variation characteristics of ESs and landscape patterns in the EQHP region from 2000-2018 were quantified. The links between changes in regional ecosystem structure and function were then explored through the evolution of the spatial assemblages of ESBs and LUCC. Finally, the interactions of ESs with natural and human factors were explored. The main conclusions are as follows:
(1) Between 2000 and 2018, the overall supply of TES in the EQHP increased from 1.59 to 1.69, with the exception of SP, which showed a decreasing trend. NPP, WY, SC, and HQ increased at rates of 7.24 gC·m‒2·a‒1, 12.19 mm·a‒1, 6.64 t·ha‒1·a‒1, and 0.01 a‒1, respectively. The study identified three bundles with different landscape characteristics. ESB1 covered areas with frequent human activities, such as cropland and built-up land. ESB2 included important ecological areas such as Qinghai Lake. ESB3 had rich forest and grassland resources.
(2) The spatial distribution of landscape types in the study area varied significantly, with a significant change in the landscape pattern in the southwest. Forestland, grassland, and unused land dominated the land use types, accounting for approximately 87% of the study area. Eight landscape indices, except for AREA_MN, AI, and CONTAG, showed a decreasing trend, indicating an increase in landscape aggregation and connectivity and a decrease in landscape fragmentation, complexity, and diversity in the EQHP.
(3) The variation in the spatial combination of bundle-landscape suggests that forestland, grassland, and unused land played important roles in the supply of ESs. Natural factors played a stronger role than human activities in the study area, but the strength of the factors varied in different bundles. Socioeconomic factors and landscape composition played important roles in ESB1 and increased over time. Soil texture and landscape composition contributed more to ES changes in ESB2 than in the other bundles. Meteorological factors, such as precipitation and temperature, had the highest explanatory power for ESB3 changes.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Bastian O, Grunewald K, Syrbe R et al., 2014. Landscape services: The concept and its practical relevance. Landscape Ecology, 29(9): 1463-1479.

DOI

[2]
Bennett E M, Peterson G D, Gordon L J, 2009. Understanding relationships among multiple ecosystem services. Ecology Letters, 12(12): 1394-1404.

DOI PMID

[3]
Bi J, Hao R F, Li J et al., 2021. Identifying ecosystem states with patterns of ecosystem service bundles. Ecological Indicators, 131: 108195.

DOI

[4]
Bullock J M, Aronson J, Newton A C et al., 2011. Restoration of ecosystem services and biodiversity: conflicts and opportunities. Trends in Ecology & Evolution, 26(10): 541-549.

DOI

[5]
Cao Y, Cao Y, Li G Y et al., 2020. Linking ecosystem services trade-offs, bundles and hotspot identification with cropland management in the coastal Hangzhou Bay area of China. Land Use Policy, 97: 104689.

DOI

[6]
Chen J H, Yan F, Lu Q, 2020. Spatiotemporal variation of vegetation on the Qinghai-Tibet Plateau and the influence of climatic factors and human activities on vegetation trend (2000-2019). Remote Sensing, 12(19): 3150.

DOI

[7]
Chen T Q, Feng Z, Zhao H F et al., 2020. Identification of ecosystem service bundles and driving factors in Beijing and its surrounding areas. Science of the Total Environment, 711: 134687.

DOI

[8]
Chi W F, Zhao Y Y, Kuang W H et al., 2019. Impacts of anthropogenic land use/cover changes on soil wind erosion in China. Science of the Total Environment, 668: 204-215.

DOI

[9]
Cord A F, Bartkowski B, Beckmann M et al., 2017. Towards systematic analyses of ecosystem service trade-offs and synergies: Main concepts, methods and the road ahead. Ecosystem Services, 28: 264-272.

DOI

[10]
Dick J, Andrews C, Beaumont D A et al., 2011. A comparison of ecosystem services delivered by 11 long-term monitoring sites in the UK environmental change network. Environmetrics, 22(5): 639-648.

DOI

[11]
Dou H S, Li X B, Li S K et al., 2020. Mapping ecosystem services bundles for analyzing spatial trade-offs in inner Mongolia, China. Journal of Cleaner Production, 256: 120444.

DOI

[12]
Duan X Y, Chen Y, Wang L Q et al., 2023. The impact of land use and land cover changes on the landscape pattern and ecosystem service value in Sanjiangyuan region of the Qinghai-Tibet Plateau. Journal of Environmental Management, 325: 116539.

DOI

[13]
Duarte G T, Santos P M, Cornelissen T G et al., 2018. The effects of landscape patterns on ecosystem services: Meta-analyses of landscape services. Landscape Ecology, 33(8): 1247-1257.

DOI

[14]
Frank S, Fürst C, Koschke L et al., 2012. A contribution towards a transfer of the ecosystem service concept to landscape planning using landscape metrics. Ecological Indicators, 21: 30-38.

DOI

[15]
Gonzalez-Ollauri A, Mickovski S B, 2017. Providing ecosystem services in a challenging environment by dealing with bundles, trade-offs, and synergies. Ecosystem Services, 28: 261-263.

DOI

[16]
Gou M M, Li L, Ouyang S et al., 2021. Identifying and analyzing ecosystem service bundles and their socioecological drivers in the Three Gorges Reservoir area. Journal of Cleaner Production, 307: 127208.

DOI

[17]
Haberman D, Bennett E M, 2019. Ecosystem service bundles in global hinterlands. Environmental Research Letters, 14(8): 84005.

DOI

[18]
Hall L S, Krausman P R, Morrison M L, 1997. The habitat concept and a plea for standard terminology. Wildlife Society Bulletin (1973-2006), 25(1): 173-182.

[19]
He Y J, Han X R, Wang X P et al., 2021. Long-term ecological effects of two artificial forests on soil properties and quality in the eastern Qinghai-Tibet Plateau. Science of the Total Environment, 796: 148986.

DOI

[20]
Hou Y Z, Zhao W W, Hua T et al., 2022. Mapping and assessment of recreation services in Qinghai-Tibet Plateau. Science of the Total Environment, 838: 156432.

DOI

[21]
Hou Y Z, Zhao W W, Liu Y X et al., 2020. Relationships of multiple landscape services and their influencing factors on the Qinghai-Tibet Plateau. Landscape Ecology, 36(7): 1987-2005.

DOI

[22]
Hu Z M, Dakos V, Rietkerk M, 2022. Using functional indicators to detect state changes in terrestrial ecosystems. Trends in Ecology & Evolution, 37(12): 1036-1045.

DOI

[23]
Jiang L, Xue B, Xing R et al., 2020a. Rural household energy consumption of farmers and herders in the Qinghai-Tibet Plateau. Energy, 192: 116649.

DOI

[24]
Jiang P, Ding W G, Yuan Y et al., 2020b. Diverse response of vegetation growth to multi-time-scale drought under different soil textures in China’s pastoral areas. Journal of Environmental Management, 274: 110992.

DOI

[25]
Karimi J D, Corstanje R, Harris J A, 2021. Bundling ecosystem services at a high resolution in the UK: Trade-offs and synergies in urban landscapes. Landscape Ecology, 36(6): 1817-1835.

DOI

[26]
Kohonen T, 1982. Self-organized formation of topologically correct feature maps. Biological Cybernetics, 43(1): 59-69.

DOI

[27]
Kohonen T, 1990. The self-organizing map. Proceedings of the IEEE, 78(9): 1464-1480.

[28]
Li M Q, Liu S L, Wang F F et al., 2022. Cost-benefit analysis of ecological restoration based on land use scenario simulation and ecosystem service on the Qinghai-Tibet Plateau. Global Ecology and Conservation, 34: e02006.

DOI

[29]
Li P J, Zuo D P, Xu Z X et al., 2021. Dynamic changes of land use/cover and landscape pattern in a typical alpine river basin of the Qinghai-Tibet Plateau, China. Land Degradation & Development, 32(15): 4327-4339.

DOI

[30]
Li S C, Zhang H, Zhou X W et al., 2020. Enhancing protected areas for biodiversity and ecosystem services in the Qinghai-Tibet Plateau. Ecosystem Services, 43: 101090.

DOI

[31]
Li S C, Zhang Y L, Wang Z F et al., 2018. Mapping human influence intensity in the Tibetan Plateau for conservation of ecological service functions. Ecosystem Services, 30: 276-286.

DOI

[32]
Li S C, Zhao Y L, Xiao W et al., 2022. Identifying ecosystem service bundles and the spatiotemporal characteristics of trade-offs and synergies in coal mining areas with a high groundwater table. Science of the Total Environment, 807: 151036.

DOI

[33]
Liu C, Yang M H, Hou Y T et al., 2021. Ecosystem service multifunctionality assessment and coupling coordination analysis with land use and land cover change in China’s coastal zones. Science of the Total Environment, 797: 149033.

DOI

[34]
Liu J, Liu X L, Wang Y et al., 2020. Landscape composition or configuration: Which contributes more to catchment hydrological flows and variations? Landscape Ecology, 35(7): 1531-1551.

DOI

[35]
Liu J, Milne R I, Cadotte M W et al., 2018. Protect Third Pole’s fragile ecosystem. Science, 362(6421): 1368-1368.

DOI

[36]
Liu M X, Gao Y, Wei H J et al., 2022a. Profoundly entwined ecosystem services, land-use change and human well-being into sustainability management in Yushu, Qinghai-Tibet Plateau. Journal of Geographical Sciences, 32(9): 1745-1765.

DOI

[37]
Liu Y X, Liu S L, Wang F F et al., 2022b. Identification of key priority areas under different ecological restoration scenarios on the Qinghai-Tibet Plateau. Journal of Environmental Management, 323: 116174.

DOI

[38]
Liu Y X, Lü Y H, Fu B J et al., 2019. Quantifying the spatio-temporal drivers of planned vegetation restoration on ecosystem services at a regional scale. Science of the Total Environment, 650: 1029-1040.

DOI

[39]
Lyu R F, Clarke K C, Zhang J M et al., 2019. Spatial correlations among ecosystem services and their socio-ecological driving factors: A case study in the city belt along the Yellow River in Ningxia, China. Applied Geography, 108: 64-73.

DOI

[40]
Ma F K, Jiang Q O, Xu L D et al., 2021. Processes, potential, and duration of vegetation restoration under different modes in the eastern margin ecotone of Qinghai-Tibet Plateau. Ecological Indicators, 132: 108267.

DOI

[41]
Ma S, Wang L J, Wang H Y et al., 2022. Spatial heterogeneity of ecosystem services in response to landscape patterns under the Grain for Green Program: A case study in Kaihua County, China. Land Degradation & Development, 33(11): 1901-1916.

DOI

[42]
Maes J, Barbosa A, Baranzelli C et al., 2015. More green infrastructure is required to maintain ecosystem services under current trends in land-use change in Europe. Landscape Ecology, 30(3): 517-534.

PMID

[43]
Millennium Ecosystem Assessment MEA, 2005. Ecosystems and Human Well-being:Synthesis. Washington, DC: Island Press.

[44]
Mouchet M A, Lamarque P, Martín-López B et al., 2014. An interdisciplinary methodological guide for quantifying associations between ecosystem services. Global Environmental Change, 28: 298-308.

DOI

[45]
Ning J, Liu J Y, Kuang W H et al., 2018. Spatiotemporal patterns and characteristics of land-use change in China during 2010-2015. Journal of Geographical Sciences, 28(5): 547-562.

DOI

[46]
Pérez-Silos I, Álvarez-Martínez J M, Barquín J, 2021. Large-scale afforestation for ecosystem service provisioning: Learning from the past to improve the future. Landscape Ecology, 36(11): 3329-3343.

DOI

[47]
Potter C S, Randerson J T, Field C B et al., 1993. Terrestrial ecosystem production: A process model based on global satellite and surface data. Global Biogeochemical Cycles, 7(4): 811-841.

DOI

[48]
Qi Y, Lian X H, Wang H W et al., 2020. Dynamic mechanism between human activities and ecosystem services: A case study of Qinghai lake watershed, China. Ecological Indicators, 117: 106528.

DOI

[49]
Quintas-Soriano C, Castro A J, Castro H et al., 2016. Impacts of land use change on ecosystem services and implications for human well-being in Spanish drylands. Land Use Policy, 54: 534-548.

DOI

[50]
Raudsepp-Hearne C, Peterson G D, Bennett E M, 2010. Ecosystem service bundles for analyzing tradeoffs in diverse landscapes. Proceedings of the National Academy of Sciences, 107(11): 5242-5247.

DOI

[51]
Renard D, Rhemtulla J M, Bennett E M, 2015. Historical dynamics in ecosystem service bundles. Proceedings of the National Academy of Sciences, 112(43): 13411-13416.

DOI

[52]
Renard K G, Foster R G, Weesies G A et al., 1991. RUSLE: Revised universal soil loss equation. Journal of Soil and Water Conservation, 46: 30-33.

[53]
Riis T, Kelly-Quinn M, Aguiar F C et al., 2020. Global overview of ecosystem services provided by riparian vegetation. Bioscience, 70(6): 501-514.

DOI

[54]
Shao Y J, Liu Y S, Li Y H et al., 2023. Regional ecosystem services relationships and their potential driving factors in the Yellow River Basin, China. Journal of Geographical Sciences, 33(4): 863-884.

DOI

[55]
Shen J S, Li S C, Liang Z et al., 2020. Exploring the heterogeneity and nonlinearity of trade-offs and synergies among ecosystem services bundles in the Beijing-Tianjin-Hebei urban agglomeration. Ecosystem Services, 43: 101103.

DOI

[56]
Shen J S, Li S C, Liu L B et al., 2021. Uncovering the relationships between ecosystem services and social-ecological drivers at different spatial scales in the Beijing-Tianjin-Hebei region. Journal of Cleaner Production, 290: 125193.

DOI

[57]
Sun J, Liu M, Fu B J et al., 2020. Reconsidering the efficiency of grazing exclusion using fences on the Tibetan Plateau. Science Bulletin, 65(16): 1405-1414.

DOI PMID

[58]
Teng Y M, Zhan J Y, Liu W et al., 2021. Spatiotemporal dynamics and drivers of wind erosion on the Qinghai-Tibet Plateau, China. Ecological Indicators, 123: 107340.

DOI

[59]
Teng Y M, Zhan J Y, Liu W et al., 2022. Spatial heterogeneity of ecosystem services trade-offs among ecosystem service bundles in an alpine mountainous region: A case study in the Qilian Mountains, Northwest China. Land Degradation & Development, 33(11): 1846-1861.

[60]
Turner M G, 2005. Landscape ecology: What is the state of the science? Annual Review of Ecology, Evolution, and Systematics, 36: 319-344.

[61]
Wang C, Wang X, Wang Y F et al., 2023. Spatio-temporal analysis of human wellbeing and its coupling relationship with ecosystem services in Shandong province, China. Journal of Geographical Sciences, 33(2): 392-412.

DOI

[62]
Wang X C, Dong X B, Liu H M et al., 2017. Linking land use change, ecosystem services and human well-being: A case study of the Manas River Basin of Xinjiang, China. Ecosystem Services, 27: 113-123.

DOI

[63]
Wang X F, Chu B Y, Feng X M et al., 2021a. Spatiotemporal variation and driving factors of water yield services on the Qingzang Plateau. Geography and Sustainability, 2(1): 31-39.

DOI

[64]
Wang X Z, Wu J Z, Liu Y L et al., 2022. Driving factors of ecosystem services and their spatiotemporal change assessment based on land use types in the Loess Plateau. Journal of Environmental Management, 311: 114835-114835.

DOI

[65]
Wang Y, Wang X F, Yin L C et al., 2021b. Determination of conservation priority areas in Qinghai-Tibet Plateau based on ecosystem services. Environmental Science & Policy, 124: 553-566.

[66]
Wischmeier W H, Smith D D, 1965. Predicting rainfall-erosion losses from cropland east of the Rocky Mountains: Guide for selection of practices for soil and water conservation. Agricultural Research Service, US Department of Agriculture.

[67]
Wu J G, 2007. Landscape Ecology:Pattern, Process, Scale and Hierarchy. Beijing: Higher Education Press.

[68]
Xu W H, Xiao Y, Zhang J J et al., 2017. Strengthening protected areas for biodiversity and ecosystem services in China. Proceedings of the National Academy of Sciences, 114(7): 1601-1606.

DOI

[69]
Yan K, Wang W F, Li Y H et al., 2022. Identifying priority conservation areas based on ecosystem services change driven by Natural Forest Protection Project in Qinghai province, China. Journal of Cleaner Production, 362: 132453.

DOI

[70]
Yang M H, Gao X D, Zhao X N et al., 2021. Scale effect and spatially explicit drivers of interactions between ecosystem services: A case study from the Loess Plateau. Science of the Total Environment, 785: 147389.

DOI

[71]
Yin L C, Tao F L, Chen Y et al., 2021. Improving terrestrial evapotranspiration estimation across China during 2000-2018 with machine learning methods. Journal of Hydrology, 600: 126538.

DOI

[72]
Yin Y T, Hou Y L, Langford C et al., 2019. Herder stocking rate and household income under the Grassland Ecological Protection Award Policy in northern China. Land Use Policy, 82: 120-129.

DOI

[73]
Yu R, 2020. An improved estimation of net primary productivity of grassland in the Qinghai-Tibet region using light use efficiency with vegetation photosynthesis model. Ecological Modelling, 431: 109121.

DOI

[74]
Zhang J X, He C Y, Huang Q X et al., 2022a. Evaluating the supply and demand of cultural ecosystem services in the Tibetan Plateau of China. Landscape Ecology, 37(8): 2131-2148.

DOI

[75]
Zhang J Z, Wang S, Pradhan P et al., 2022b. Mapping the complexity of the food-energy-water nexus from the lens of Sustainable Development Goals in China. Resources, Conservation and Recycling, 183: 106357.

[76]
Zhang L, Dawes W R, Walker G R, 2001. Response of mean annual evapotranspiration to vegetation changes at catchment scale. Water Resources Research, 37(3): 701-708.

DOI

[77]
Zhang L, Hickel K, Dawes W R et al., 2004. A rational function approach for estimating mean annual evapotranspiration. Water Resources Research, 40(2): 1-14.

[78]
Zhang X R, Wang Y S, Yuan X F et al., 2022c. Identifying ecosystem service supply-demand imbalance for sustainable land management in China’s Loess Plateau. Land Use Policy, 123: 106423.

DOI

[79]
Zhao L, Liu Z H, Hu Y M et al., 2022a. Evaluation of reasonable stocking rate based on the relative contribution of climate change and grazing activities to the productivity of alpine grasslands in Qinghai province. Remote Sensing, 14(6): 1455.

DOI

[80]
Zhao Y H, Qu Z, Zhang Y et al., 2022b. Effects of human activity intensity on habitat quality based on nighttime light remote sensing: A case study of northern Shaanxi, China. Science of the Total Environment, 851: 158037.

DOI

Options
Outlines

/

〈 〉