Ecohydrological Service Characteristics of Qilian Mountain Ecosystem in the Next 30 Years Based on Scenario Simulation

Abstract

Mountain ecosystems have special ecohydrological services, and the study of water conservation and soil conservation services in the Qilian Mountain Ecosystem (QLME) in China has important theoretical value for scientific understanding of the ecological processes and mechanisms of mountain ecosystems. In this study, we quantitatively estimated the spatial-temporal changes of water conservation and soil conservation services in the QLME based on the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model and estimated the future ecosystem services (ESS) of the QLME under RCP4.5 (Representative Concentration Pathways) and RCP8.5 scenarios using the coupled Geosos-FLUS model. Firstly, the QLME ecohydrological service increased from 1985 to 2018, and its spatial heterogeneity was high in the east and low in the west. Among them, water conservation first decreased and then showed a trend of fluctuating increase, and soil conservation services decreased sharply from 2010 to 2015. Secondly, there are differences in the ecohydrological services of the QLME under different land-use types. The water conservation capacity in descending order is glacier snow, grassland, forest land, wetland, and cultivated land. The soil conservation intensity from strong to weak is woodland, grassland, arable land, glacier snow, and bare land. Thirdly, under different scenarios, QLME water conservation and soil conservation functions will increase to varying degrees over the next 30 years. The water conservation in the RCP4.5 scenario is higher than that in the RCP8.5 scenario, and the higher discharge scenario will lead to the decline of the water conservation service function. The increased rate of soil conservation was greater under the RCP8.5 scenario. With the development of Nationally Determined Contributions (NDCs) and scenarios below 2 °C, the future of QLME ecohydrological services will be further understood.

1. Introduction

Different ecosystems are a unified whole formed by interacting with specific natural environments and living things. Meanwhile, different ecosystems have different ecological structures, with different ESS under specific backgrounds. ESS is a research field that has gradually developed in recent decades. The Study of Critical Environmental Problems (SCEP) first mentioned the concept of environmental service in 1970 [1]. In 1981, Ehrlich formally proposed the concept of EES, and it was widely recognized and used. Since then, numerous studies have been carried out from various aspects [2,3,4]. As a matter of fact, the quantitative assessment of EES needs to solve a series of basic theoretical problems and also needs the support of relevant methodologies. The interaction between biological elements and environmental elements and their ecological, economic, and social effects are the important bases for measuring ESS. At the same time, the ecological processes and mechanisms of material circulation, energy flow, and information transfer in different ecosystems are the important theoretical basis for defining the concept of EES. At present, ESS assessment methods can be divided into two categories, one is the common accounting methods, such as value quantity assessment method, material quality assessment method, emergy analysis method; the other is the ecological model method, these methods have their own advantages and disadvantages in the research and application. So far, modeling is becoming more and more important in the quantitative assessment of ESS. Common ecosystem models include InVEST, Aries, SoLVES, and other models, among which InVEST model is widely used [5]. Global climate change and human exploitation have adversely affected ESS. Numerous studies have also pointed out that climate change has negative effects on most ESS and will continue to increase in the future [6].

The Qilian Mountains are rich in various natural resources. The Qilian Mountain Ecosystem (QLME) undertakes many important EES in China, such as water conservation, biodiversity, water and soil conservation, and other important functions [7]. It plays an extremely important strategic role in maintaining local ecological security. In recent years, affected by climate change and frequent human activities, problems such as vegetation degradation of natural forests and grass, water and soil erosion, shrinking snow and ice-covered areas, and reduction of biodiversity have occurred in the Qilian Mountains [8]. Therefore, research on climate change under the background of the QLME and ESS dynamic change can provide scientific support and effective guidance for the health of the QLME protection and social-economic sustainable development.

Ecohydrological service function is an important function of ESS, whether water conservation or soil conservation, which is important in QLME. Although there are many types of ESS, this study focuses on these functions. As an important ecological barrier in the arid region of western China, it is of great practical significance for the ecological construction, environmental protection, and social and economic development of this region and related regions to maintain the regional ecosystem function under the background of climate change and intense human activities. Although temperature controls soil organic carbon storage in middle and high latitudes, the hydrological climate in tropical regions may be the main driving factor of soil carbon storage [9]. The carbon pool of the global forest ecosystem is extremely complex [10], and forest water use efficiency is closely related to atmospheric carbon dioxide [11]. Model simulation methods play an important role in estimating ecosystem parameters such as Net Primary Productivity (NPP) [12,13]. At present, many methods have been explored around the evaluation of ESS value and the estimation of ecosystem carbon storage [14]. The value-quantity evaluation method and material-quality evaluation method are important evaluation methods, while emergy, shrinking ice and snow, and reduced biodiversity have seriously restricted the no-analysis method. In recent years, the ecological model law has become an active hotspot. With the development of ecological information science and geographic information on science, assessment methods of ESS and estimation methods of carbon storage, carbon budget, and carbon source and sink are gradually associated with ecological models and are further deepened [15]. At present, the modules of InVEST model are sorted into different types. The different EES can directly contribute to human society [16,17].

The ecological system is the foundation of human survival and development. Some research shows the impact of climate change on streamflow and hydrology of a mountainous region in Canada [18]. Meanwhile, the research estimates the soil moisture of a watershed in the Qilian mountain ranges via vegetation and land surface temperature [19]. Global change research has an important status and role, while the mountain ecosystem, with its unique geographical perpendicular band landscape characteristics contains a unique topography, soil, vegetation, hydrology, climate, and other natural geographical factors, have become an important area global change research. The QLME is an important ecological barrier in China because of the artificial vegetation habitat damage produced by intense activity, natural grassland degradation, soil erosion, shrinking ice and snow and reduce biodiversity, has seriously restricted the normal play of its function; however, it also seriously restricted regional low carbon and green development. In recent years, the state has made great efforts to improve the ecological environment of the Qilian Mountains. In particular, relying on the original nature reserve of the Qilian Mountains, the state has continuously promoted the pilot work of the national park, which has greatly accelerated the process of ecological restoration. At present, with the attention of ecological and environmental issues, the QLME has also been concerned about the impact of climate change and frequent human activities. In this context, the research background of climate change, the QLME response to climate change, but also the health of the Qilian mountains, ecological environment protection, and social, sustainable economic development will provide a scientific basis for policymaking, pilot project construction, and management of the QLME in China, which has important theoretical significance and practical application value. In the first section of this paper, the main progress of the research on ecohydrological services function caused by climate change is described. In the second section, the data sources and methods are introduced. In the third section, the simulation of ecohydrological services under future scenarios is mainly described. Section 4 and Section 5, respectively, discuss and present the study’s conclusions.

As for climate change scenario models, many models worldwide have certain applicability to simulate the change of relevant climate factors at different scales. In particular, recent developments in NDCs and Shared Socioeconomic Paths (SSPs) scenarios below 2 °C have provided important insights for estimating QLME ecohydrological service functions [20]. Based on the RCP4.5 and RCP8.5 scenarios of the BCC model, this study estimated the future trends and spatial distribution of water conservation and soil conservation functions in the QLME.

2. Data and Methods

2.1. Geographical Position of the QLME

The QLME is located at the intersection of the Qinghai-Tibet Plateau, Loess Plateau, and Mongolia-Xinjiang Plateau. The geographic coordinates are 95°06′~103°01′ E, 36°40′~39°42′ N, spans Qinghai and Gansu provinces, and its total area is 50,200 km², as shown in Figure 1. The whole study area belongs to mountainous ravines, and the average elevation is 3799 m. The landscape includes woodland, grassland, wetland, desert, and snow glaciers and is quite rich. They are important areas for biodiversity conservation in China and also help with water runoff to the inland river system. The study area is an important ecological barrier in China (even in Central Asia) and one of the newly established pilot nature reserves with national parks as the main body, which has extremely important ecological value. At the same time, the application of scenario simulation in mountain ecosystem research is relatively weak, and this study is prospective.

The northern part of the study area has a typical temperate continental climate along the Hexi Corridor, while the southern part has an alpine and semi-arid climate, with annual precipitation ranging from 96.4 mm/year to 729.6 mm/year. As a watershed area of the Heihe River, Shule River, Shiyang River, and other water systems, the glacier, snow cover, woodland, and other ecosystems provide important water conservation services. The QLME is rich in soil resources, covering more than 30 soil types, such as black felt soil, cold calcium soil, and cold frozen soil. Its vegetation presents a vertical distribution rule in different elevation gradient regions. From bottom to top, it can be roughly divided into mountain forest zone, mountain grassland zone, sub-alpine shrub zone, sub-alpine sub-snow, and ice sparse vegetation zone. The vegetation population mainly consists of trees such as Qinghai spruce and Qilian cedar, as well as shrubs such as flagellaria and willow [21].

2.2. Data Source and Processing

This study needs the support of multi-source data. In addition to the basic information obtained from the actual survey, a series of data are adopted in the estimation and analysis of carbon storage. The raster data of land use/cover, plant available water content, and soil maximum root depth were obtained from the National Tibetan Plateau Scientific Data Center and were reclassified and resampled by ArcGIS software. The raster data of annual precipitation and annual potential evapotranspiration were obtained from China Meteorological Science Data Sharing Service Network and were interpolated by Anusplin software and calculated by the Penman–Monteith method. The secondary watershed boundary vector data and Digital Elevation Model (DEM) raster data are derived from the resource and environmental science data cloud platform and processed by the ArcGIS software hydrological analysis module. Slope and slope direction raster data were obtained based on DEM data by ArcGIS software 3D Analyst tool. Vector data of roads, railways, and urban settlements are derived from the national basic geographic information database, and the distance analysis module of the ArcGIS software is used to obtain the Euclidian distance of the elements. According to the basic data requirements of the InVEST and Geosos-Flus model, all kinds of data were preprocessed in different ways and then related ecological elements were further retrieved through the model.

The research used meteorological data, future climate scenario data, land use/cover data, topographic data, soil data, and literature statistical data, and obtained input data of the InVEST and Geosos-FLUS models after various preprocessing. The reliability of input data is closely related to the reliability of output data.

2.3. Feature of the InVEST Model

InVEST is a suite of models used to map and value the goods and services from nature that sustain and fulfill human life. It helps explore how changes in ecosystems can lead to changes in the flows of many different benefits to people. If properly managed, ecosystems yield a flow of services that are vital to humanity, including the production of goods, life-support processes, and life-fulfilling conditions, and the conservation of options. The modules of the InVEST model can be divided into two categories. As mentioned above, one is the module that supports ESS, such as habitat quality and habitat risk assessment. The other type is the final ESS module, including carbon storage, water conservation, water purification, soil conservation, and other service functions, which can directly benefit human society [22]. Despite its importance, this natural capital is poorly understood, scarcely monitored, and in many cases, undergoing rapid degradation and depletion.

In the QLME, due to the strong influence of natural and human activities, the ecosystem structure and function have undergone a series of changes in recent years, and the ESS has been seriously disturbed. In this context, applying InVEST principles and methods to estimate the current, especially future changes in ecosystem functions, plays an important role in ecological restoration and sustainable development. In fact, InVEST includes a number of services, according to the connotation and classification of InVEST, the service value of InVEST is estimated mainly from the aspects of soil and water conservation, and the possible trend of the service value in the next 30 years is predicted under the future scenario.

2.3.1. Water Conservation Function Assessment in QLME

In the InVEST model, the water conservation service function comprehensively considers influencing factors such as precipitation, potential evapotranspiration, vegetation cover, topographic factors, and soil water conductivity in the QLME. Among them, many studies believe that precipitation, potential evapotranspiration, and vegetation cover have a greater impact on the change of water conservation function. Under the background of climate change, the change of precipitation directly affects the total amount of water resources in the QLME, while the change of temperature indirectly affects its water production and water conservation function by affecting evapotranspiration. Vegetation cover affects water conservation function through transpiration and interception. The water conservation service function is evaluated based on the water yield module in the InVEST model, which is based on the water balance estimation method, i.e., the difference between precipitation and evaporation on the grid cell to obtain the water supply, by assuming that the water yield of each grid is collected from the collection point through different channels. Then the total amount and average value of water yield in the secondary basin are calculated.

2.3.2. Soil Conservation Function Assessment in the QLME

The soil conservation module of InVEST model generates the spatial distribution of sediment yield in the catchment area based on the input DEM data [23]. The model first calculates the potential soil erosion amount, i.e., the soil erosion amount under the protection of nature and vegetation (RKLS), and then calculates the actual soil erosion amount, i.e., the soil erosion amount under the artificial management and conservation measures (USLE). The soil retention amount (SD) is the difference between the potential soil erosion amount and the actual soil erosion amount. The equation is as follows:

(1)$USLE = R \times K \times LS \times C \times P$

(2)$RKLS = R \times K \times LS$

(3)$SD = RKLS - ULSE$

where R is the rainfall erosivity factor, K is the soil erodibility factor. LS is the slope length factor; C is the coverage and management factor; P is the factor of soil conservation measures.

2.4. Features of the Geosos-Flus Model

The Geosos-Flus model is a multi-type land-use change scenario simulation model developed based on the principle of the Flus model [24]. It is the development and inheritance of its predecessor, geographic simulation, and optimization system Geosos. The Geosos-Flus model provides users with the function of simulating spatial land-use change. When simulating future land-use change, users need to apply other methods (System Dynamics model, or Markov chain) or use preset scenarios to determine the amount of future land-use change as the input of Geosos-Flus.

The Geosos-Flus model is mainly divided into two modules. One is the suitability probability calculation module based on Artificial Neuron Network (ANN). The other part is the cellular automata module based on the adaptive inertia mechanism. Firstly, the classified data of land use and the driving factors of land-use change (such as terrain, location, road distribution) were input into the suitability module, and the suitability probability of each region was calculated by ANN. The model principle is expressed as follows:

(4)$p(n,l,t) = {\displaystyle \sum _i{w}_{i,t}} \times sigmoid[ne{t}_i(n,t)]$

(5)${\displaystyle \sum _{l}p(n,l,t)} = 1$

In Equations (4) and (5), p(n,l,t) represents the suitability probability of the crop class l at time t on the grid pixel n; W_i,t and sigmoid are the weights and excitation functions of the hidden layer and the output layer; Neti (n,t) is the response value of pixel n of the class l hidden layer in time.

(6)$I_l^t = \{\begin{array}{cc}{I}_l^t,& \left|M_l^{t-2}\right| \le \left|M_l^{t-1}\right|\\ I_l^t\times \frac{M_l^{t-2}}{M_l^{t-1}},& M_l^{t-2} < M_l^{t-1} < 0\\ I_l^t\times \frac{M_l^{t-1}}{M_l^{t-2}},& M_l^{t-1} > M_l^{t-2} > 0\end{array}$

where I is the adaptive inertia coefficient of the difference between the number of land terminations and the number of land decisions; M_l^t−1 and M_l^t−2 represent the difference between the actual grid number and the number of termination targets at the time of t − 2 and t − 1, respectively.

(7)$T{P}_{n,l}^t = p(n,l,t) \times O_{n,l}^t \times I_l^t \times (1-s{c}_{c-t})$

In Equation (7), TP^t_n,l represents the total probability that grid n at time t is transformed into land type l, and SC_c−t represents the cost that land use type c is transformed into land type l. O^t_n,l stands for the domain role.

(8)$O_{n,l}^t = \frac{{\displaystyle \sum _{N\times N}con(c_p^{t-1}=l)}}{N\times N-1}{w}_l$

where ${\displaystyle \sum _{N\times N}con(c_p^{t-1}=l)}$ represents the total number of grids in class l after the end of the last iteration of the Mole Domain window of N × N, and w_l is the weight of the influence degree of the field.

3. Simulation of ESS in the QLME under Future Scenarios

3.1. Analysis of Future Land Use Simulation Results and Potential Changes

InVEST models are spatially-explicit, using maps as information sources and producing maps as outputs. InVEST returns results in either biophysical terms (e.g., tons of carbon sequestered) or economic terms (e.g., the net present value of that sequestered carbon). Based on the characteristics of InVEST and after related quantitative processing, the following results are obtained in this study.

Sensitivity Analysis (SA) investigates how the variation in the output of a numerical model can be attributed to variations of its input factors [25]. SA is increasingly being used in environmental modeling for a variety of purposes, including uncertainty assessment, model calibration, diagnostic evaluation, dominant control analysis, and robust decision-making. The Morris method can be used to conduct sensitivity analysis on parameters of hydrological, ecological function in the InVest model-related modules [26]. The annual precipitation, annual potential evapotranspiration, and evapotranspiration coefficient of vegetation are the influences. Precipitation erosivity and soil erodibility are the main parameters affecting soil and water conservation functions. This study can clarify the uncertainty of the InVest model and provide a reference for the application and improvement of the InVest model.

3.1.1. Accurate Test of the Geosos-Flus Model

The land use in 2000 was used to simulate the land use and cover in 2010 to calibrate the Geosos-Flus model and obtain its suitability probability. Then, the land use in 2015 was simulated, and the actual land use data were combined for comparative verification (Figure 2). The Kappa coefficient of the verification results is 0.72, and the overall accuracy of the simulation is 0.81. The results show that the model can well simulate the change of land use types in the QLME, so it can be used to predict the future land use situation in QLME.

3.1.2. Analysis of Future Land-Use Change

The land-use data in 2000 and 2015 in QLME were taken as the initial and termination year data, and the transfer probability of each land use type was calculated by the Markov Chain model, and then the number of land use structures in the QLME in 2050 was predicted. On this basis, the land-use data and driving factor data in 2018 are taken as the input data of the Geosos-Flus model. Combined with the actual situation of the QLME, the transformation matrix selects the natural protection scenario to simulate the land use type in the QLME in 2050 (Figure 3).

From 2018 to 2050, the landscape type transfer matrix of the QLME is shown in Table 1. From Table 1, it can be seen that the landscape type of the whole region remained basically unchanged from 2018 to 2050, with only a few areas having land type transfer, and the transfer trend was mainly to woodland, grassland, and bare land type transfer. Specifically, 1.96% of cultivated land was converted to grassland and 9.8% to bare land. 8.79% of grassland was converted to woodland, and 0.18% to glacial snow cover. 1.17% of the bare land was converted to woodland, 1.09% to grassland, and 2.03% to glacial snow cover. In addition, 0.24% of glacial snow cover was converted to grassland and 21.38% to bare land.

Combined with Figure 4 of the spatial distribution of land cover transfer from 2018 to 2050, it can be seen that due to the large area base of grassland and bare land types, the conversion of grassland types and bare land types in the QLME is quite obvious. Among them, in the central and eastern regions of the QLME, a large amount of grassland was converted to woodland. In addition, the bare land in the QLME has been largely restored to woodland and grassland, and the area of vegetation land types has increased significantly. However, in the western high altitude area of the QLME, the exchange of bare land and glacial snow cover land was mainly manifested, as well as the restoration of glacial snow cover land and bare land to grassland. Most of them were scattered alpine meadows and low coverage grassland. Overall, the spatial change of the land-use in the QLME will develop well from 2018 to 2050, and the area of forest land will increase.

3.2. Changes of ESS under Future Scenarios in the QLME

The Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report (AR5) concluded that the warming of the climate system is unequivocal and human influence on the climate system is clear. This will rely on information from theory, observations, and Earth system model (ESM) simulations that are coordinated as part of the World Climate Research Programme (WCRP) Coupled Model Intercomparison Project (CMIP). CMIP provides the basis for multimodel evaluation and has, over the years, revealed a variety of systematic differences between models and observations, with many persisting from one model generation to the others [27]. At present, IPCC AR6 WG I further emphasizes the seriousness of climate change. In CMIP5, based on RCP4.5 and RCP8.5 scenarios, the QLME has a series of characteristics of the ecohydrological function.

3.2.1. Water Conservation Service Function in the QLME

Based on the land-use type data in 2050 simulated by FLUS and combined with the potential evapotranspiration data under RCP4.5 and RCP8.5 scenarios in CMIP5, the distribution of water conservation service functions in the QLME in 2050 under the two scenarios was estimated by InVEST model (Figure 5). It can be found that, under the scenario of RCP4.5, the average water conservation is 110.06 mm/km² in the QLME in 2050, and the total modulus of water conservation is 5.46 × 10⁶ mm, which is 8636.56 mm higher than that in 2018. Under the scenario of RCP8.5, the average water conservation per unit area in the QLME in 2050 is 108.45 mm, and the total water conservation modulus is 5.38 × 10⁶ mm, which is increased compared with that under the scenario of RCP4.5. The water conservation amount and water conservation capacity under RCP4.5 were higher than those under RCP8.5.

Under the scenarios of RCP4.5 and RCP8.5, the average potential evapotranspiration in 2050 is higher than that in 2018, which leads to the higher actual evapotranspiration. The actual evapotranspiration per unit area under the two scenarios is different, respectively, both higher than that of 266.34 mm in 2018. Therefore, the average water conservation per unit area in QLME in the 2050 scenario is slightly lower than that in 2018, but due to the large increase in forest area in the natural conservation scenario, its water conservation capacity is strong, so the total amount of water conservation in the two scenarios will be improved. However, the water conservation amount under the RCP8.5 scenario is lower than that under the RCP4.5 scenario, indicating that the service function of water conservation will be reduced under the higher emission scenario.

3.2.2. Service Function of Soil Conservation in the QLME

The coupling of FLUS and InVEST models was used to estimate the soil conservation services in the QLME in 2050 under the two scenarios (Figure 6). It can be found that, under the scenario of RCP4.5, the average soil conservation amount is 5243.11 t in 2050, and the total soil conservation amount is 257.876 million tons, which is increased compared with that in 2018, with a change rate of 2.18%. Under the RCP8.5 scenario, the soil conservation amount per unit area of the QLME is 5520.97 t, and the total soil conservation amount is 271.55444 million tons, with a growth rate of 7.60% compared with that in 2018.

The temporal variation of soil conservation is shown in Figure 7. It can be seen that the soil conservation service function of the QLME presents an overall growth trend. From 2010 to 2015, due to the ecological destruction of the QLME, the forest and grass area has been greatly reduced, leading to a significant decline in soil conservation. After the national policy and the restoration and management of its ecological environment, its soil conservation function has been significantly restored and improved in 2018. Under the scenarios of RCP4.5 and RCP8.5, the soil conservation amount in 2050 will increase to varying degrees. Although the precipitation in 2050 is higher than that in 2018, the rainfall erosion caused by it will also increase, but at the same time, the woodland area of the QLME under the nature conservation scenario has increased greatly, so the contribution capacity of soil conservation amount will also be strengthened.

4. Discussion

Under the setting of future scenarios, the water conservation function in the QLME was lower than that in RCP4.5 under the higher emission scenario; however, the water conservation in the QLME was still increased compared with that in 2018 due to the increase in vegetation cover under the nature conservation scenario. These results indicate that if certain environmental protection policies are adopted in the future to increase vegetation, the impact of climate change on water conservation function can be alleviated to a certain extent. InVEST is the model of soil conservation service function that mainly calculated the soil erosion caused by rainfall, topography, soil properties, vegetation cover, and land-use. Among them, soil and topography greatly influence water erosion, but they are relatively stable. Therefore, some scholars attributed the EES of soil conservation mainly to precipitation and land-use/cover. The greater the vegetation density, the greater the interception of raindrops, and the smaller the influence of rainfall on soil surface particles. Therefore, land-use and vegetation cover exceed the influence of rainfall intensity and topographic factors to some extent. It is considered to be the key factor leading to the change of soil erosion and soil retention.

Space-time environmental heterogeneities for predicting the multiscale ecological and environmental dynamics underlying the patterns are considered in ecohydrology. Model-checking is a research and testing process to determine the correctness, validity, and credibility of the model. Its essence is to verify that the built model can reflect the behavior characteristics of the real system. In the QLME study, both the InVEST model and RCP model are uncertain due to natural and anthropogenic impacts. With further study of the model mechanism, the reliability of prediction will be further improved.

At present, the EES research concept and method are gradually improved and deepened. A number of experts have systematically studied and combed the state of ESS research on an international scale, revealing eight themes and 28 approaches to ESS [28]. Biodiversity conservation, economic value, ecosystem services research, human well-being, landscape planning, land-use are the main topics, and research and assessment methods are more diverse such as measurement, computation, decision support, economics, matrix, meta-analysis/bibliometric analysis, non-monetary valuation, software models, spatially explicit, statistical, and conceptual, mixed/integrated, and indicators. Around method exploration, it also involves conservation management, ecosystem-based governance, operation, planning, and stakeholder participation. More methods involve the interaction between people, the environment, landscape ecology, etc. No matter which aspects are focused on, the purpose of the methods is to solve the problems related to EES, deepen people’s understanding of EES, better improve ESS, and continuously guarantee human well-being. For a long time, human activities such as illegal mining and construction in the QLME have resulted in reducing natural vegetation and serious damage to the ecological environment. Since 2015, the state has formulated policies to carry out a series of ecological restoration projects in the Qilian Mountains, including vegetation restoration and protection, environmental governance in mining areas, and transfer of residents in the protected areas. The restoration measures such as artificial afforestation and mountain closure for forest cultivation have greatly increased the forest area, significantly changed the land cover in the study area, and the ecological environment has also begun to recover and improve, and the water conservation and soil conservation functions have been significantly improved. As in recent years, establish a national park as the main body of the suggestion of the natural reserve system and the implementation of the related ecological protection engineering (i.e., natural protection scenario), the ecological environment of the QLME may be further improved, which makes the ecological environment problems such as soil erosion can be further solved. The development of ESS-related theories and methods has important implications for understanding the function of the QLME ecohydrological service and maintaining and improving water conservation and soil conservation.

GSUA typically assigns probability distribution functions to all model factors and propagates that into model outputs [29]. That is useful for assessing input factor importance and interaction, regimes, and scaling laws between model input factors and outcomes. The sensitivity of the output to an input variable is therefore measured by the amount of variance in the output caused by that input. Variance-based methods fully explore the input space, accounting for interactions and nonlinear responses. For these reasons, they are widely used when it is feasible to calculate them. Sensitivity Analysis (SA) investigates how the variation in the output of a numerical model can be attributed to variations of its input factors. SA is increasingly being used in environmental modeling for a variety of purposes, including uncertainty assessment, model calibration and diagnostic evaluation, dominant control analysis, and robust decision-making.

In fact, the reasons for the temporal and spatial differences of the QLME ecohydrological services are complex, with different impacts of natural factors and human activities. It is subject to the adjustment of national ecological environment protection policy and the long-term adaptation of regional hydrological, climatic, and soil conditions, and the biomass has a cumulative effect. Under this background, the ecohydrological service function is increasing. However, the QLME disturbed by artificial intensity will be affected by the lag of ecological effect, and it will take a certain period of time to increase ecohydrological service function. At present, the QLME study area is covered by an integrated information network and monitoring sites, greatly facilitating the restoration of ecological functions. With the development of green, low-carbon, and high-quality development models, the QLME’s ecohydrological services are expected to be further enhanced in the future.

5. Conclusions

The QLME has a variety of resources and plays an extremely important strategic role in maintaining ecological security. In recent years, under the influence of climate change, especially overgrazing, illegal construction, mineral mining, tourism and sightseeing, and other intensive human activities, vegetation degradation of natural forests and grass, water and soil erosion, shrinking snow and ice-covered areas, and reduction of biodiversity have occurred in the QLME. Under the background of a series of strategic decisions to implement, and further research dynamic change with climate change scenarios, and estimates and analyses the different scales of ESS spatiotemporal changes in the QLME and the comprehensive environmental effect have important theory value. At the same time, this study to explore the QLME influence factors of carbon storage and reveal the carbon cycle and biogeochemical cycle, improve the carbon storage estimation technology, ESS value estimation technology, ecosystem habitat quality evaluation technology, and a series of ecosystem of the ecological environment evolution and the quality evaluation system of technology also has important theoretical value. In the long run, it is of great practical significance to advocate the construction of natural protected areas and low-carbon development.

Based on the InVEST model, this paper analyzed the spatiotemporal evolution of water conservation and soil conservation services in the QLME and estimated the changes of ESS under the future scenarios by coupling the Geosos-Flus model and reached the following conclusions.

• (1). The ESS of water conservation and soil conservation in the QLME showed an increasing trend in general from 1985 to 2018 but showed spatial heterogeneity. The water conservation in the QLME first decreased and then fluctuated and increased, and the spatial distribution was higher in the northwest and the southeast. The spatial distribution pattern of soil conservation services was high in the eastern and central regions and low in the western regions, with an overall upward trend in time. However, due to illegal mining of mineral resources and construction, soil conservation volume decreased sharply from 2010 to 2015.

• (2). In the future scenario, water conservation and soil conservation functions in the QLME will increase to varying degrees in 2050. In 2050, although the water conservation amount will increase compared with 2018, the average water conservation amount will decrease. The water conservation amount under the RCP4.5 scenario is higher than that under the RCP8.5 scenario, and the EES of water conservation will decline under a higher emission scenario. However, the growth rate of the soil conservation function was greater under the RCP8.5 scenario.

In the context of climate change, there are common points in many studies on the understanding of climate change in the Qilian Mountain region, and there are also some controversies. Due to the differences in research spatial and temporal scales and evaluation methods, different scholars have different discussions on climate characteristics in the Qilian Mountain region. Although the climate effects of different ecosystems have been understood through a large number of monitoring, analysis, and simulation estimates, and the facts and effects of relevant climate change have been confirmed, people still need to understand further the process, severity, and impacts of climate change on different regions. This study systematically studied the temperature and precipitation in the QLME from 1960 to 2018 and summarized the characteristics and effects of the main factors of climate change in the QLME.

• (3). The ESS in the QLME were different in different land types. From the perspective of different land-use/cover, the order of water conservation is grassland, glacier and snow, forest land, bare land, wetland, arable land, and construction land. However, water conservation capacity is as follows: glacier snow cover, grassland, woodland, wetland, arable land, etc. The land type with the highest soil conservation intensity was woodland, followed by grassland, cultivated land, and glacial snow cover. The soil conservation ability of bare land was the weakest.

• (4). In the future scenario, water conservation and soil conservation functions in the QLME will increase to varying degrees in 2050. In 2050, although the water conservation amount will increase compared with 2018, the average water conservation amount will decrease. The water conservation amount under the RCP4.5 scenario is higher than that under the RCP8.5 scenario, and EES of water conservation will decline under a higher emission scenario. However, the growth rate of the soil conservation function was greater under the RCP8.5 scenario.

Author Contributions

R.W., Conceptualization, Writing—original draft; Q.P., Conceptualization, Software, Formal analysis; W.Z. (Weidong Zhang), Methodology, Software, Writing—Review & editing; W.Z. (Wenfei Zhao), Formal analysis, Writing—Review & editing; C.L., Formal analysis, Validation, Writing—Editing; L.Z., Validation and Writing—Editing. All authors have read and agreed to the published version of the manuscript.

Funding

This study was supported by the National Key Research and Development Program (2019YFC0507403).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data presented in this study are available on request from the corresponding author.

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Table

Figure 1. Geographic location of the study area.

Figure 2. Land-use simulation and actual situation of the Geo-Flus model in the QLME in 2015 ((a): land use simulation in 2015, (b): actual land use situation in 2015).

Figure 3. Land-use in the QLME in 2018 and 2050 ((a): 2018, (b): 2050).

Figure 4. Spatial distribution of land-use transfer in the QLME from 2018 to 2050.

Figure 5. Distribution of water conservation in QLME in 2050 under different scenarios ((a): RCP4.5 scenario, (b): RCP8.5 scenario).

Figure 6. Spatial distribution of soil conservation under different scenarios in the QLME in 2050 ((a): RCP4.5 scenario, (b): RCP8.5 scenario).

Figure 7. Changes of soil conservation in the QLME from 1985 to 2050 ((a): RCP4.5 scenario, (b): RCP8.5 scenario).

Figure 7. Changes of soil conservation in the QLME from 1985 to 2050 ((a): RCP4.5 scenario, (b): RCP8.5 scenario).

Table 1

Land-use transfer matrix in the QLME from 2018 to 2050 (%).

2018	2050
	Cultivated Land	Woodland	Grassland	Wetland	Waters	Construction Land	Bare Land	Snow/Ice
cultivated land	88.24	0	1.96	0	0	0	9.8	0
woodland	0	100	0	0	0	0	0	0
grassland	0	8.79	90.97	0.01	0.06	0	0	0.18
wetland	0	2.04	0	83.67	2.04	0	12.24	0
waters	0	0	0	0	100	0	0	0
construction land	0	0	0	0	0	100	0	0
bare land	0	1.17	1.09	0	0.04	0	95.67	2.03
snow/ice	0	0	0.24	0	0	0	21.38	78.39

References

1. Gymer, R. Man’s Impact on the Global Environment. Assessment and Recommendations for Action (Report of the Study of Critical Environment Problems); Massachusetts Institute of Technology: Cambridge, MA, USA, 1971; [DOI: https://dx.doi.org/10.1021/ed048pa369.2]

2. Holdren, J.P.; Ehrlich, P.R. Human Population and the Global Environment: Population growth, rising per capita material consumption, and disruptive technologies have made civilization a global ecological force. Am. Sci.; 1974; 62, pp. 282-292. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/4832978]

3. Westman, W.E. How much are nature’s services worth?. Science; 1977; 197, pp. 960-964. [DOI: https://dx.doi.org/10.1126/science.197.4307.960]

4. Ehrlich, E. Extinction: The Causes and Consequences of the Disappearance of Species. Q. Rev. Biol.; 1981; 1, pp. 82-85.

5. Wu, Z.; Dai, E.; Ge, Q.; Xi, W.; Wang, X. Modelling the integrated effects of land use and climate change scenarios on forest ecosystem aboveground biomass, a case study in Taihe County of China. J. Geogr. Sci.; 2017; 27, pp. 205-222. [DOI: https://dx.doi.org/10.1007/s11442-017-1372-x]

6. Runting, R.K.; Bryan, B.A.; Dee, L.E.; Maseyk, F.J.F.; Mandle, L.; Hamel, P.; Wilson, K.A.; Yetka, K.; Possingham, H.P.; Rhodes, J.R. Incorporating climate change into ecosystem service assessments and decisions: A review. Glob. Chang. Biol.; 2017; 23, pp. 28-41. [DOI: https://dx.doi.org/10.1111/gcb.13457] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27507077]

7. Xu, J.; Xie, G.; Xiao, Y.; Li, N.; Jiang, Y.; Chen, W. Dynamic analysis of ecological environmental quality changes in national key ecological function areas in China. Acta Ecol. Sin.; 2019; 39, pp. 3039-3050.

8. Wang, T.; Gao, F.; Wang, B.; Wang, P.; Wang, Q.; Song, H.; Yin, C. Status and suggestions on ecological protection and restoration of Qilian Mountains. J. Glaciol. Geocryol.; 2017; 39, pp. 229-234.

9. Hein, C.J.; Usman, M.; Eglinton, T.I.; Haghipour, N.; Galy, V.V. Millennial-scale hydroclimate control of tropical soil carbon storage. Nature; 2020; 581, pp. 63-66. [DOI: https://dx.doi.org/10.1038/s41586-020-2233-9]

10. Dixon, R.K.; Brown, S.; Houghton, R.A.; Solomon, A.M.; Trexler, M.C.; Wisniewski, J. Carbon pools and flux of global forest ecosystems. Science; 1994; 263, pp. 185-190. [DOI: https://dx.doi.org/10.1126/science.263.5144.185]

11. Keenan, T.F.; Hollinger, D.Y.; Bohrer, G.; Dragoni, D.; Munger, J.W.; Schmid, H.P.; Richardson, A.D. Increase in forest water-use efficiency as atmospheric carbon dioxide concentrations rise. Nature; 2013; 499, pp. 324-327. [DOI: https://dx.doi.org/10.1038/nature12291]

12. Tian, H.; Chen, G.; Liu, M.; Zhang, C.; Sun, G.; Lu, C.; Xu, X.; Ren, W.; Pan, S.; Chappelka, A. Model estimates of net primary productivity, evapotranspiration, and water use efficiency in the terrestrial ecosystems of the southern United States during 1895–2007. For. Ecol. Manag.; 2010; 259, pp. 1311-1327. [DOI: https://dx.doi.org/10.1016/j.foreco.2009.10.009]

13. Son, K.; Lin, L.; Band, L.; Owens, E.M. Modelling the interaction of climate, forest ecosystem, and hydrology to estimate catchment dissolved organic carbon export. Hydrol. Process.; 2019; 33, pp. 1448-1464. [DOI: https://dx.doi.org/10.1002/hyp.13412]

14. Li, H.; Cai, Y.; Chen, R.; Chen, Q.; Yan, X. Effect assessment of the project of grain for green in the karst region in Southwestern China: A case study of Bijie Prefecture. Shengtai Xuebao Acta Ecol. Sin.; 2011; 31, pp. 3255-3264.

15. Wang, R. Study on Coupling Relationship between Soil and Water and Ecological Security of Carbon Dioxide Abatement Forest; China Meteorological Press: Beijing, China, 2021.

16. Clerici, N.; Paracchini, M.L.; Maes, J. Land-cover change dynamics and insights into ecosystem services in European stream riparian zones. Ecohydrol. Hydrobiol.; 2014; 14, pp. 107-120. [DOI: https://dx.doi.org/10.1016/j.ecohyd.2014.01.002]

17. Dawen, Q.; Guangmin, C.; Yangong, D.; Qian, L.; Xiaowei, G. Spatio-temporal dynamics of ecosystem service value in the southern slope of Qilian Mountain from 2000 to 2015. Shengtai Xuebao Acta Ecol. Sin.; 2020; 40, pp. 1392-1404. [DOI: https://dx.doi.org/10.5846/stxb201811292598]

18. Ghaderpour, E.; Vujadinovic, T.; Hassan, Q.K. Application of the Least-Squares Wavelet software in hydrology: Athabasca River Basin. J. Hydrol. Reg. Stud.; 2021; 36, 100847. [DOI: https://dx.doi.org/10.1016/j.ejrh.2021.100847]

19. Bai, X.; Zhang, L.; He, C.; Zhu, Y. Estimating regional soil moisture distribution based on NDVI and land surface temperature time series data in the upstream of the Heihe River Watershed, Northwest China. Remote Sens.; 2020; 12, 1766998. [DOI: https://dx.doi.org/10.3390/rs12152414]

20. Chai, Q.; Fu, S.; Wen, X. Modeling the implementation of NDCs and the scenarios below 2 °C for the Belt and Road countries. Ecosyst. Health Sustain.; 2020; 6, 1766998. [DOI: https://dx.doi.org/10.1080/20964129.2020.1766998]

21. Chen, H.; Ding, W.; Tanjia, B.Z. Evaluation of soil water erosion in Qilian mountain national park based on USLE model. Sci. Soil Water Conserv.; 2020; 18, pp. 38-44. [DOI: https://dx.doi.org/10.16843/j.sswc.2020.04.005]

22. Zhang, F.; Li, X.; Feng, Q.; Wang, H.; Wei, Y.; Bai, H. Spatial and Temporal Variation of Water Conservation in the Upper Reaches of Heihe River Basin Based on InVEST Model. J. Desert Res.; 2018; 38, pp. 1321-1329.

23. Qiu, W.; Zhang, Y.; Yu, J.; Zhang, C.; Zheng, C.; Yu, S. Feasibility verification with field application of a water conservation module using the InVEST Model. J. Zhejiang Agric. For. Univ.; 2018; 35, pp. 810-817.

24. Wang, X.; Pan, P.; Wang, X.; Wang, X. Simulation of landscape pattern for land use in Hebei province based on GeoSOS-FLUS model. Jiangsu J. Agric. Sci.; 2021; 37, pp. 667-675.

25. Pianosi, F.; Beven, K.; Freer, J.; Hall, J.W.; Rougier, J.; Stephenson, D.B.; Wagener, T. Sensitivity analysis of environmental models: A systematic review with practical workflow. Environ. Model. Softw.; 2016; 79, pp. 214-232. [DOI: https://dx.doi.org/10.1016/j.envsoft.2016.02.008]

26. Lei, X.; Han, H.Q.; Gong, X.Y.; Huang, Z.Y.; Du, Q.G.; Wang, H. Parameter sensitivity analysis of InVEST Model for freshwater Module based on Morris. J. Ningxia Norm. Univ.; 2019; 40, 5.

27. Eyring, V.; Cox, P.M.; Flato, G.M.; Gleckler, P.J.; Abramowitz, G.; Caldwell, P.; Collins, W.D.; Gier, B.K.; Hall, A.D.; Hoffman, F.M. et al. Taking climate model evaluation to the next level. Nat. Clim. Chang.; 2019; 9, pp. 102-110. [DOI: https://dx.doi.org/10.1038/s41558-018-0355-y]

28. Torres, A.V.; Tiwari, C.; Atkinson, S.F. Progress in ecosystem services research: A guide for scholars and practitioners. Ecosyst. Serv.; 2021; 49, 101267. [DOI: https://dx.doi.org/10.1016/j.ecoser.2021.101267]

29. Convertino, M.; Muñoz-Carpena, R.; Chu-Agor, M.L.; Kiker, G.A.; Linkov, I. Untangling drivers of species distributions: Global sensitivity and uncertainty analyses of MaxEnt. Environ. Model. Softw.; 2014; 51, pp. 296-309. [DOI: https://dx.doi.org/10.1016/j.envsoft.2013.10.001]