Optimal Allocation of Slope Ecological Restoration for the Climate Change Mitigation and Natural Function Improvement

The key to dealing with extreme problems at watershed or region scales in the context of climate change is the “de-extremalization” of hydrological processes. The foundation lies in how to optimize the allocation of ecological restoration on slopes to mitigate the extreme impacts of climate change on hydrological processes and improve ecological service functions. Previous studies focused on analyzing the direct effects of slope land use, vegetation distribution, or historical pattern of ecological restoration on runoff processes. This study developed a slope ecological restoration evaluation system to evaluate the naturalness, functional types, and functional coordination of slope ecological restoration at different historical stages and quantitatively identified the mitigation effect on climate change in the future, which was applied to the Huangshui River Basin (above Minhe County) in Qinghai Province, China. Based on the land suitability evaluation methods, a set of layout schemes were constructed. The runoff mutation and ecological function of different schemes under climate change were evaluated, and the highly suitable scheme was selected as the optimal scheme. Compared with the current situation, the coupling coordination degree index of the scheme would increase from 0.32 to 0.59. Meanwhile, the runoff and coefficient of variation would decrease by 30% and 60%, respectively, during the wet season under the high-emission scenario RCP8.5. This study closely links the ecological construction of slopes with the response to extreme climates, which provides technical methods and practical support for the optimization of regional ecological patterns and scientific water governance modes.


INTRODUCTION
In the context of climate change, the frequency and intensity of extreme drought and flood events in most mid-latitude terrestrial and tropical humid regions around the world have shown an increasing trend (IPCC, 2014). The key to dealing with large-scale extreme problems in watersheds or regions lies in the "de-extremalization" of hydrological processes (Yan et al., 2017;Zhang et al., 2020), as well as repairing the terminal governance and process separation problems in traditional water governance modes , and improving the mitigation capacity of slope measures to extreme hydrological processes. Slope ecological restoration is an important measure for the construction of Nature-Enriched-Attributes-Coordinated-Watershed (Yan et al., 2017;. On the one hand, it directly affects vertical hydrological processes such as the atmospheric water vapor flux (Fathizadeh et al., 2017;Ghimire et al., 2017;Liu J. et al., 2018b), soil water movement Huang et al., 2017;Yu et al., 2018;Bo et al., 2021;Xu M. et al., 2021), and evapotranspiration (Li et al., 2016;Wang et al., 2018). On the other hand, the canopy structure and community succession of vegetation indirectly affect the horizontal hydrological processes such as runoff generation and confluence on the slope (Amenu and Kumar, 2008;Yang et al., 2010;Dong et al., 2021) by changing the underlying surface conditions and the surface roughness. Aiming at mitigating the extreme impacts of climate change on hydrological processes and improving ecological service functions, how to optimally allocate ecological restoration on slopes has become one of the key scientific issues in the intersecting fields of climatology, ecology, and hydrology.
At present, the evaluation index of ecological restoration focused on the regulation of ecological function and hydrological process (Nunes et al., 2016;Li et al., 2018;He Z. et al., 2019), such as peak reduction of the runoff, landscape pattern, and ecological service value. Most researchers evaluated ecological restoration, respectively, from the above two aspects. A comprehensive index system, which includes scale, composition, pattern, and eco-hydrological function at the basin scale is needed. Allocation of ecological restoration was based on ecology theory, which was mainly conducted at the rural and urban scales (Li Y. et al., 2017;Zhang and Chen, 2017;Liu J. et al., 2018b), with the common optimization goals of urban pattern, storm water resource management, and ecological service function improvement (Li and Li, 2019b;Liu J. et al., 2018b;Liu, 2019;Jiang et al., 2021), using traditional landscape ecology methods, geographic information technology, and morphological pattern analysis methods (Li and Li, 2019b;Liu, 2019). However, few studies focused on the impacts of the optimal layout of largescale slope ecological restoration on extreme hydrological processes. Regarding the research of the mitigation effect of slope ecological restoration construction on climate change, scholars mainly adopt three kinds of methods. The first one is the theoretical derivation method based on the law of conservation of mass. This method mainly reveals the climate impact from the macro-mechanism level, while it is difficult for this method to accurately quantify the historical distribution of slope vegetation or the extent and magnitude of the influence of ecological restoration (Zhou and Wang, 1999). The second method is the statistical relationship analysis method based on prototype observation, reanalysis data, and remote sensing interpretation (Li X. et al., 2017;Wang et al., 2019;Jin et al., 2020). This method is relatively convenient, but the limitation of data sequence length and spatial accuracy leads to uncertainty in the analysis of mitigation effects. The third one is land surface-hydrological numerical simulation method based on climate models (López-Bravo et al., 2018;Zheng et al., 2020), which uses land use change or forest and grass cover scenarios to simulate and analyze runoff changes under climate change (Yang et al., 2019;Li et al., 2009;Bai et al., 2020;Wang et al., 2020) or feedback of the land cover on local climate (Hua et al., 2015;Li et al., 2017;Zheng et al., 2020). The above methods focused on analyzing the direct effects of slope land use, vegetation distribution, or historical pattern of ecological restoration on runoff processes under the impact of climate change. Few studies investigated the mitigation effect of optimizing the layout of slope ecological restoration on climate change.
To solve the aforementioned key scientific issues, this study took the Huangshui River Basin (above Minhe County) in Qinghai Province, China, as the study area. An evaluation index system for slope ecological restoration was constructed, which includes the naturalness, functional types, and functional coordination of slope ecological restoration in different historical stages. A set of layout schemes was acquired based on the land suitability evaluation method. The runoff mutation and ecological function of different layout schemes in the context of climate change were evaluated based on the distributed hydrological model (WEP) (He S. et al., 2019) and geographic information technology, and the optimal layout schemes were selected. This study closely links the ecological construction of slopes with the mitigation to extreme climates, which provides technical methods and practical support for the optimization of regional ecological patterns and scientific water governance modes.

Study area
Huangshui River is a first-degree tributary of the upper reaches of the Yellow River, which rises in the Haibei Tibetan Autonomous Prefecture of Qinghai Province. Huangshui River Basin lies in the transition zone between the Qinghai-Tibet Plateau and the Loess Plateau with a typical feature of ecological fragility. The study site covers most of the Huangshui River basin from the upper reaches of the Yellow River to Minhe County. The main stream length of the region is 278 km with a drainage area of 15,558 km 2 (Figure 1). The terrain of the region is lower in the southeast and higher in the northwestern, which has an elevation change of 3,107 m from the river source to the mouth . The region experiences an arid and semi-arid continental climate with the mean average precipitation of 381.1 mm (1960-2017 records), and the mean average air temperature from 3.1°C to 7.9°C (1960-2017 records) (He S. et al., 2019). The mean annual runoff at the Minhe hydrological station (the outlet of the basin) was approximately 2,053 million m 3 (1956-2000 records). According to the land use classification results of 2017, forest was the main type of land use, followed by grasslands and cropland, occupying 92.91% in total. The main vegetation types were forests, temperate shrubs, grasslands, and meadows (Feng et al., 2013).

MATERIALS
The basic data required for this study include topography data, meteorology data, soil data, administrative division data, land use data, hydrological data, and future climate scenario data ( Table 1). This study uses three main future climate change scenarios in the IPCC-AR5 report: high-emission scenario RCP8.5, medium-emission scenario RCP4.5, and low-emission scenario RCP2.6 (IPCC, 2012; Xu and Xu, 2012). Five general circulation models provided by ISI-MIP (The Inter-Sectoral Impact Model Intercomparison Project) were selected, including GFDL-ESM2M, HADGEM2-ES, IPSL-CM5A-LR,  MIROC-ESM-CHEM, and NORESM1-M (Piani et al., 2010;Warszawski et al., 2014). The involving meteorological parameters are precipitation, average temperature, average relative humidity, near surface average wind speed, and total solar radiation. The time range is from January 1, 2021 to December 31, 2050.

Methods
An evaluation index system for slope ecological restoration was constructed to evaluate the naturalness, functional type, and functional coordination. Evaluation criteria and comprehensive evaluation methods were proposed for the evaluation of slope ecological restoration in different historical stages in the Huangshui River Basin. A set of slope ecological restoration layout schemes was developed as the underlying surface scenarios based on the improved land suitability evaluation method. Different climate model data and underlying surface scenarios were adopted to drive the distributed hydrological model (WEP) to obtain runoff processes under different composite backgrounds. The impacts of climate change on the volume and variability of runoff under different slope ecological restoration layout schemes were quantitatively analyzed. The overall mitigation effects of slope ecological restoration on climate change were identified. Additionally, considering the evaluation results of the slope ecological restoration under different layout schemes, the optimal layout schemes were selected ( Figure 2).

Historical evaluation of slope ecological restoration (1) Evaluation index system
Based on the ArcGIS platform, the slope ecological restoration scale at different construction stages (1980-2000, 2000-2017, 1980-2017) were acquired and used for subsequent evaluation (He S. et al., 2019). Focusing on the concept of Nature-Enriched-Attributes-Coordinated-Watershed , based on the principles of respecting and conforming to nature, ecological restoration evaluation index system involving naturalness, functional types, and functional coordination was developed. The naturalness is commonly defined as the divergence of the ecosystem's biodiversity or current state from its natural steady state (Eichner and Tschirhart, 2007;Winter, 2012). In this study, we evaluated the naturalness from the scale, composition and the quality of the slope ecological restoration. Meanwhile, the evaluation index system aims to evaluate the hydrological adjustment function, ecological support function, climate regulation function, and ecological service function (serving the economy and society) of the whole watershed. The hydrological adjustment function mainly focused on the plant transpiration and runoff variation. The ecological support function referred to the water conservation, wind prevention, and sand fixation function. The climate FIGURE 2 | Workflow for the optimal allocation of slope ecological restoration for the climate change mitigation and natural function improvement.
Frontiers in Earth Science | www.frontiersin.org March 2022 | Volume 10 | Article 837311 A k is the area of the kth land use type, CR fk is the value of ecosystem service function of the kth land use and fth function for climate regulation Ecological service function serving the economy and society The value of ecosystem services that serve the economic and social aspects A k is the area of the kth land use type, VC fk is the value of ecosystem service function of the kth land use and the fth function for economic and social aspects Functional Coordination

Coupling and coordination
The coordination level among different functions / D C · T √ D is the coupling and coordination, C is the coordination, T is the hydrological regulation-ecological support comprehensive evaluation index, F(x) is the comprehensive evaluation index of hydrological regulation function, G(y) is the comprehensive evaluation index of ecological support function, k = 2, α β 0.5 Frontiers in Earth Science | www.frontiersin.org March 2022 | Volume 10 | Article 837311 regulation function and ecological service function were calculated based on different types of land use. The functional coordination was evaluated from the degree of coordination between the functional types ( Table 2). The classification of coupling and coordination levels is shown in Table 3 (He S., 2020). According to the entropy method, the normalization and weight determination were carried out, and the overall naturalness, functional types, and comprehensive evaluation index of the slope ecological restoration at different stages were obtained.
(2) Evaluation standard According to the meanings of various indicators and the goal of slope ecological restoration construction, the evaluation criteria of various indicators of naturalness, function type, and function coordination were determined ( Table 4).

Configuration of slope ecological restoration based on watershed land suitability evaluation
According to the land suitability evaluation system promulgated by the Food and Agriculture Organization of the United Nations in 1976 (Food and Agriculture Organization, 1976), an evaluation index system was established by considering factors such as DEM, precipitation, air temperature, soil, irrigation conditions, slope, and roughness. The weight and score of each index was determined and calculated based on the analytic hierarchy process. Combined with the geo-statistical classification method, the land suitability evaluation grade of each plot unit on the slope was obtained. Specifically, the evaluation methods of forest land and residential construction sites were as follows.
According to the distribution of natural forest land, artificial forests and grasses (the sum of which are slope ecological restoration), the evaluation criteria, and grade of forest land were modified. Considering the urbanization process, the distribution range of counties and cities, the traditional evaluation criteria, and grade of residential construction sites were adjusted. Based on the evaluation results of the current status of land suitability, following the principles of "livability, cultivated land protection, forest protection, and grass regulation," the layout schemes of slope ecological restoration belonging to three suitability grades of general suitable (S Low ),  Naturalness Green area ratio The larger the green area ratio, the larger the scale of slope ecological restoration, and the more conducive to the improvement of naturalness Biological abundance index The larger the biological abundance index, the higher the biodiversity and the more perfect the ecological restoration composition of the slope Ecological integrity The landscape pattern indices generally develop toward decreasing trend of landscape fragmentation indicating that the ecological integrity is enhanced, and the ecological restoration quality is improved Functional types Annual plant transpiration per unit area Calculate the annual average plant transpiration per unit area from 1965 to 2017 and compare it with the frequency of incoming water to discover whether the ecological restoration construction on the slope can effectively improve the transpiration during the wet and dry years and enhance the effectiveness of hydrological regulation Annual runoff variation coefficient The decrease in the variation coefficient of annual runoff from 1965 to 2017, reflecting the enhanced controllability of hydrological regulation functions Water conservation function The larger the changes in water conservation amount, the stronger the water conservation function than before Wind prevention and sand fixation function The greater the percentage of forest land conversion area, the stronger the wind-proof and sand-fixing function than before The total value of ecological services for climate regulation The greater the value of the climate regulation function, the stronger the climate regulation function Estimate the value of ecosystem services that serve the economic and social aspects The greater the value of the ecological service function, the stronger the function moderately suitable (S Med ), and highly suitable (S High ) were constructed with the goal of eliminating unsuitable plots and minimizing the area of critically suitable plots.

Identifying the mitigation effects of different layout schemes on climate change
Previous studies (Peng et al., 2015;Zhou et al., 2018;Li et al., 2019;Liu H. et al., 2020;Yang et al., 2020;Wang et al., 2021) verified the applicability of WEP model in different regions of China including the Huangshui River Basin (He S. et al., 2019). The meteorological data from the five general circulation models were integrated to a set of comprehensive climate model data based on an aforementioned research (He S. et al., 2019). The comprehensive climate model data and underlying surface scenarios were adopted to drive the distributed hydrological model (WEP) to obtain monthly runoff processes from 2021 to 2050 under different RCP scenarios. The change rates (e.g., (S Low -S 8017 )/S 8017 ) of runoff and coefficient of variation (CV) of runoff under different layout schemes and current conditions (S 8017 ) were calculated in the whole hydrological period or in the wet, normal, and dry seasons. The mitigation effect of layout schemes of the slope ecological restoration on the extreme runoff process under future climate change was quantitatively identified.

Historical evaluation of slope ecological restoration
From the perspective of subfunctions, as the scale of slope ecological restoration increased, naturalness showed a trend of improvement, while the functional types decreased gradually before 2000, leading to a lower comprehensive evaluation index in 1980-2017 than that in 2000-2017. It may be due to the slight decrease in water conservation function index. That is because annual evapotranspiration in 2000-2017 was higher than that in 1980-2017 (Table 5). This indicated that the current slope ecological restoration needed to be further configured and improved both for the comprehensive evaluation index and the individual functions to ensure the improvement of overall function of the basin. The results of the comprehensive evaluation index showed that as the scale of slope ecological restoration increases, the comprehensive functions were gradually improved, especially after 2000 ( Table 6).

Evaluation of current land suitability in the watershed
In general, the proportions of highly suitable, moderately suitable, generally suitable, critically suitable, and unsuitable areas in the Huangshui Basin were 14.3%, 21.1%, 40.9%, 21.2%, and 2.5%, respectively. In terms of spatial distribution, the valley areas in the middle and lower reaches of the main stream of the Huangshui River Basin belonged to highly suitable, moderately suitable, and generally suitable grades. The mountainous areas on both sides and the southwestern mountainous areas were mainly classified as critically suitable and unsuitable grades (Figure 3). From the  perspective of land use, the high suitability grade area of forest land accounted for 17.27% of the total area; followed by residential and construction land, and grassland for 16.37% and 15.42%, respectively. The proportion of high suitability grade area of cultivated land was only 7.2%, and the sum of general suitability and critical suitability accounts for 97.92%.

Layout scheme of slope ecological restoration
Compared with the evaluation results of 2017, the forestland area increased by 3,450.73 km 2 , and the grassland area decreased by 3,450.73 km 2 in the highly suitable scheme (S High ). The forestland increased by 1,735.06 km 2 , and the grassland decreased by 1,735.06 km 2 in the moderately suitable scheme (S Med ). The forestland decreased by 554.213 km 2 and the grassland increased by 554.213 km 2 in the generally suitable scheme (S Low ) (Figure 4).

Evaluation of layout schemes of slope ecological restoration
Compared with S8017, most indices of the naturalness and functionality of the three slope ecological restoration layout schemes were enhanced. The coupling coordination degree indices showed growing trends, especially the S High scheme with an increase by 84.3%. According to the evaluation criteria of the coupling coordination degree, the coordination between functional types of the three schemes were close to the constrainedly coordinated ( Table 3). The complete imbalance state of current situation was changed, and the overall functional coordination was enhanced ( Table 7).

Mitigation effect of layout schemes on climate change
The simulated runoff at Xinachuan, Qiaotou, Xining, and Minhe Stations matched well with the observation. The Nash coefficients exceeded 0.55, the correlation coefficients exceeded 0.75, and the relative errors were about 15% during the calibration period. The Nash coefficients were above 0.5, the correlation coefficients were about 0.7, and the relative errors were about 15% during the verification period (He S. et al., 2019). The simulated runoff by the WEP model was used for further analysis.

Relative change rate of runoff
From 2021 to 2050, the change rates of the runoff under different layout schemes relative to S8017 were calculated ( Figure 5). As shown in Figure 5A, in the same RCP scenario, the annual average runoff under the three layout schemes all increased by about 10% compared with the current runoff volume. From the perspective of different periods, the runoff during the wet season can effectively be reduced by about 20% under the three layout schemes. The reduction effects of S High scheme were particularly obvious, and the effects were enhanced with the increase in During the normal and dry seasons, the runoff increased significantly, which effectively supplemented the insufficient flow during the dry season under future climate change, and improved the regional water conservation capacity.

Runoff variability
In the same RCP scenario, the CV values of the annual runoff decreased under the three layout schemes compared with that under current slope ecological restoration at seven typical hydrological stations except for Chaoyang Station, especially for the S High scheme with a reduction of about 60% (Figure 6). Under different RCP scenarios, the CV value of annual runoff under current slope ecological restoration gradually decreased as the emission concentration increased at seven typical hydrological stations except for Minhe Station.
In different hydrological periods, the influences of three layout schemes on CV values of runoff varied. During the wet seasons, the influences showed similar patterns with that in the annual scale. In the normal seasons, the CV values of the runoff decreased under the three layout schemes (compared with the current slope ecological restoration) at all seven typical hydrological stations. During the dry seasons, most stations showed the decreasing trend of CV of runoff under the three layout schemes (compared with the current slope ecological restoration) except for Baliqiao Station.
As for the influences under different climate scenarios, different hydrological seasons showed different patterns (Figure 7). During the wet seasons, the CV values of runoff under RCP 4.5 were the largest compared with those under RCP 2.6 and RCP 8.5 at most stations. In the normal seasons, the CV values of runoff under RCP 2.6 were the largest compared with those under RCP 4.5 and RCP 8.5 at most stations. During the dry seasons, the changes in CV values of runoff under layout schemes were not the same with that under the current slope ecological restoration. Under the current slope ecological restoration, most stations showed the largest CV under RCP4.5 compared with those under RCP 2.6 and RCP 8.5, while under the three layout schemes, the largest CV appeared under RCP2.6 compared with those of other climate scenarios at most stations.

Selection of the optimal layout scheme
The S High scheme has the highest score in the comprehensive evaluation index of all layout schemes (Table 8), which indicated that this scheme meets the needs of a new mode for basin water governance typically for naturalness and functional coordination. Combined with analysis of different layout schemes for future climate change mitigation, this scheme has the most obvious effect on avoiding extreme hydrological events. Therefore, the

FIGURE 7 |
The CV values of runoff in different hydrological periods under different layout schemes and different climate scenarios (A-C) represents the wet seasons, the normal seasons, and the dry seasons, respectively. CV8017_26/45/85 represents the variation coefficient of runoff in each hydrological period with the current slope ecological restoration under the emission scenarios of low-emission scenario (RCP2.6), medium-emission scenario (RCP4.5), and high-emission scenario (RCP8.5), respectively; CVSLow_26/45/85 represents the variation coefficient of runoff in each hydrological period with the layout schemes of slope ecological restoration at generally suitable grade under the emission scenarios of RCP2.6, RCP4.5, and RCP8.5; CVSMe_26/45/85 represents the variation coefficient of runoff in each hydrological period with the layout schemes of slope ecological restoration at moderately suitable grade under the emission scenarios of RCP2.6, RCP4.5, and RCP8.5; CVSHigh_26/45/85 represents the variation coefficient of runoff in each hydrological period with the layout schemes of slope ecological restoration at highly suitable grade under the emission scenarios of RCP2.6, RCP4.5, and RCP8.5, respectively.
Frontiers in Earth Science | www.frontiersin.org March 2022 | Volume 10 | Article 837311 S High scheme is determined as the optimal layout scheme for the slope ecological restoration construction in the Huangshui River Basin. The plan is to adjust the unsuitable grades of residential construction sites and arable land to general suitable grades, and the areas suitable for forestry and grass will be developed as forest land.

Impact of slope ecological restoration scale on historical evaluation
From the perspective of Naturalness, S8000 had no obvious impacts on the indicators compared with S80. Compared with S80 and S8000, the ecological restoration scale and the biological abundance index under S0017 and S8017 increased. This is mainly concerned with the implementation of a series of ecological restoration projects including natural reserves, national forest park, and large-scale forest farms in the Huangshui River Basin since the 1970s (Sun et al., 2015;Zhang Y. et al., 2016;Forestry and Grassland Bureau of Qinghai, 2018;Treacy et al., 2018;Feng et al., 2021). As the scale of ecological restoration has increased significantly, the landscape pattern indices, such as number of patches, edge density, Shannon diversity index, and Shannon evenness index increased, and the largest patch index, average patch area, area-weighted mean patch fractal dimension index, and contagion index decreased. The overall naturalness of the basin increased. While due to the discrete distribution and homogeneity of such ecological restoration measures, the landscape pattern index shows an increasing trend in number of patches, diversity, and fragmentation.
Referring to the indicators of function types, S8000 had no obvious impacts on them compared with S80. Annual plant transpiration per unit area under S0017 and S8017 has slightly increase compared with that under S80. The underlying mechanism maybe complex. On the one hand, the study area experiences an arid and semi-arid continental climate, and continuous drought in spring and summer occurs in 40% of the years. The background value of evapotranspiration per unit area is low (Zhang C. et al., 2016;Li et al., 2020). On the other hand, the growth of vegetation in the study area depends on artificial irrigation, and the local water resources are scarce, which limit the water used for irrigation. The growth of vegetation is mainly restricted by artificial irrigation along the river and from reservoirs, and snowmelt-runoff irrigation (Kou et al., 2021). When the scale of ecological restoration is expanded, annual plant transpiration per unit area will not increase much. Annual runoff variation coefficient under S0017 and S8017 has little changes compared with that under S80 and S8000. In normal seasons, the runoff CV in most hydrological stations all decreased except for the slight increase at the Baliqiao station. In dry seasons, the runoff CV at the Xinachuan station increased slightly, with the rest of the hydrological stations decreasing. The effects of slope ecological restoration on runoff CV in wet seasons are spatially inconsistent. Comprehensive influence of the three seasons explains the slight change in annual runoff variation coefficient under S0017 and S8017 compared with that under S80 (He S. et al., 2019). Moreover, the upstream areas of Qiaotou and Chaoyang Stations are alpine regions and covered with snow and ice. After forest disturbances, greater snow accumulation is estimated (melting earlier and faster) (Winkler et al., 2005), and melt ice and snow during wet seasons replenish runoff, which lead to the insignificant reduction in runoff variation coefficient.
With the expansion of scale of ecological restoration and enrichment of composition, the structure and function of ecosystems changed, and the water conservation function, wind prevention, and sand fixation function, ecological services for climate regulation and those serving the economic and social aspects have been significantly enhanced, and the function type indicators have been improved overall, which were similar to the conclusions of other scholars (Tang et al., 2014;Angela et al., 2015;Xu et al., 2019;Zhou 2019). Functional types increased, functional coordination increased, and comprehensive evaluation index increased significantly. We also propose that increasing tree species diversity which will positively affect soil functioning (e.g., soil biodiversity, nutrient availability), the resilience of the forest ecosystem, and the ecological service functions of watersheds (Vicente-Vicente et al., 2019).
As the scale of slope ecological restoration increased (S8017 compared with S0017), the complexity of ecosystem composition increased, the diversity of landscape patterns were enhanced, and the degree of patch aggregation decreased. The overall functional type index showed a decreasing trend after 2000. Two reasons may be related to phenomenon. Ecological restoration measures in the Huangshui River Basin were scattered, and the types of artificial forests were homogeneous. Although the scale of ecological restoration has expanded, the improvement of ecological service functions was limited. After 2000, due to the reduction in water conservation function and climate regulation function, the overall function type index of S8017 was lower than that of S0017. Although functional type index decreased, functional coordination enhanced with the increase in slope ecological restoration scale, which indicated that the coordination between hydrological adjustment and ecological support functions were improved. The comprehensive evaluation index decreased. This phenomenon indicates that the slope ecological restoration needs to further refine the layout schemes and pay more attention to the simultaneous improvement of the comprehensive evaluation index and the individual evaluation index.

Mitigation effect of layout schemes on climate change
Under RCP2.6, RCP4.5, RCP8.5 climate scenarios, the impacts of layout schemes on relative change rate of runoff at different time scales varied. The relative change rates of annual average runoff and of runoff in the normal and dry seasons were positive, while the relative change rates of runoff during the wet season were negative. This phenomenon indicates that the layout schemes of slope ecological restoration can effectively reduce the peak flow and increase the dry season runoff under future extreme climate scenarios. The impacts of slope ecological restoration on reduction of runoff in wet seasons and supplement in dry seasons can be demonstrated by relevant studies (Hou et al., 2018;Takata and Hanasaki, 2020), which reduce flood potentials in the wet season and drought severity in the dry season (Guo et al., 2008). As Bosch J.M. and Hewlett J.D. (Bosch and Hewlett, 1982) inferred, coniferous forest, deciduous hardwood, shrub, and grassland have (in that order) a decreasing influence on water yield of the basin in which these covers are manipulated. Therefore, regulation of forest land may be the key to the optimal layout of slope ecological restoration. At Chaoyang Station, there was little difference between the CV values of the annual runoff under three layout schemes and that under current slope ecological restoration, which indicated that the three layout schemes have relatively less impacts on the runoff variability at the Chaoyang Station. This may be due to the fact that Chaoyang Station is located in the middle reaches of the Huangshui River Basin, and the three layout schemes have relatively small differences in the sub-basin area controlled by Chaoyang Station. Hence, annual runoff variability is significantly affected by climate change.
The differences in the CV value of annual runoff under current slope ecological restoration at Minhe Station may be concerned with its location. Minhe Station is located at the exit of the Huangshui River Basin. Under the current slope ecological restoration, its annual runoff CV increases with the increase in the emission concentration of the climate scenario, while the annual runoff CV under the three layout schemes decreases as the emission concentration increases.
Except for Chaoyang and Baliqiao stations, the S High scheme at most stations has significant reduction effect on the CV value of annual runoff. We concluded that S High reduced the occurence of extreme water conditions in wet and dry seasons and effectively increased the stability of the annual runoff .
The decreasing trends of the three layout schemes in the wet and normal seasons are similar with the results of . In general, the three layout schemes of slope ecological restoration have significantly reduced the variability of runoff processes in the entire watershed, at different stations and in different hydrological periods. The layout schemes have also avoided extreme hydrological events, and effectively mitigated the negative effects of climate change.

CONCLUSION
Based on the naturalness, functional types, and functional coordination, the evaluation index system of slope ecological restoration was constructed. The individual index at different historical stages was analyzed, and the entropy method was used to calculate the comprehensive evaluation index. The results showed that as the scale of slope ecological restoration increased, the comprehensive functions were improved gradually, the naturalness and functional types were gradually improved, and the functional coordination was gradually better.
According to the evaluation results of the current status of land suitability, following the principles of "livability, cultivated land protection, forest protection, and grass regulation," the layout schemes of slope ecological restoration belonging to three suitability grades of general suitable (S Low ), moderately suitable (S Med ), and highly suitable (S High ) were constructed with the goal of eliminating unsuitable plots and minimizing the area of critically suitable plots.
On the one hand, the optimized configuration of slope ecological restoration will have the effect of "cutting peaks and replenishing dryness," which significantly reduce the extreme impact of future climate change. On the other hand, it can also exert the naturalness and functional coordination of slope ecological restoration to provide reference for the future watershed governance mode.
In this study, a comprehensive Evaluation Index System of slope ecological restoration was developed, among which, hydrologic regulation function and ecological support function were considered. With the in-depth identification of the impact mechanism of slope ecological restoration on the process of slope runoff generation in the future, the evaluation system needs to be further enriched and improved.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

AUTHOR CONTRIBUTIONS
TQ wrote the original draft of this manuscript. SH and SL designed the manuscript. HN and BD undertook the analysis of data. XL revised the manuscript.