Optimization of Forest and Grass Vegetation Distribution in the Aksu River Basin by Water Resources Carrying Capacity

Abstract

In arid areas, vegetation is the basis for maintaining the virtuous cycle of ecosystems, while water resources are the key factor restricting the survival of vegetation communities. The balance between water resources and vegetation is related to the sustainable development of ecological benefits and economic benefits in arid areas. Although research on the carrying potential of vegetation in arid areas has been emphasized, there is still a lack of spatial analysis of different vegetation types over large areas. Therefore, this study takes precipitation as the basic source of water resources, calculates the amount of available water resources in the basin from the available effective precipitation and available irrigation water, and it analyzes the spatial distribution of forest and grass vegetation types under the water resources constraint, combined with the ecological water demand of different vegetation types and based on the principle of “Water to determine the Vegetation”. The results showed that the ecological water demand of each vegetation type was as follows: Forest > Shrub Vegetation > Grassland Vegetation > Desert Vegetation. The range of comprehensive available water resources of vegetation was from 0 to 221.71 mm, which decreased with the altitude gradient. Then, the spatial distribution pattern of vegetation types constrained by the water resources in the Aksu River Basin showed a striped distribution from north to south, with shrub vegetation in the high-altitude mountainous area, grassland vegetation in the low-altitude area, and desert vegetation in the desert plain area, respectively, accounting for 29.05%, 10.74%, and 53.45% of the total basin. This research approach provides a scientific basis for the planning of forest and grass vegetation construction in arid regions.

1. Introduction

The term “Carrying Capacity” originally refers to the maximum load of an object without loss and is a physical concept that was later cited and practiced in other fields, including the ecological field. At the same time, some concepts related to the field of ecosystem have emerged, such as grassland carrying capacity, land carrying capacity, environmental carrying capacity, water resources carrying capacity, vegetation carrying capacity, etc. [1,2,3,4]. It is mainly to reflect the ability of natural resources to carry certain objects, which has a limit, beyond which it will have an irreversible impact on the sustainable development of the ecological environment [5,6,7,8]. For inland river basins in arid areas with fragile ecologies, the phenomenon of land desertification is widespread, and the vegetation cover is small and sparse. In order to slow down the degree of desertification and maintain the stability and sustainable development of the ecological environment in arid areas, reasonable vegetation construction and the development of vegetation communities are necessary means [9,10].

Vegetation carrying capacity is commonly used to analyze the growth potential of vegetation, which specifically refers to the maximum load of plants per unit area under specific environmental conditions, based on the principle of not damaging the ecosystem [11,12,13,14]. According to the different research purposes, vegetation carrying capacity mainly considers soil properties, water resources, ecosystem conditions, etc [15,16,17]. For inland river basins in arid areas, water resources are the core elements of local economic and social developments and the key to constrain the growth of vegetation. Water and vegetation affect and restrict each other. Vegetation is the key factor of water conservation, but it is also the main consumer of water resources. Reasonable vegetation layout can play a role in water storage, while excessive or inappropriate vegetation construction will increase ecological water consumption, promote the reduction in available water resources, and eventually break the balance between water supply and demand, resulting in imbalance or even degradation of the ecological environment [18]. Therefore, the carrying balance between water and vegetation is the focus of the ecological environment in arid areas [19,20,21], and the study of vegetation carrying capacity in arid areas cannot be separated from the analysis of water supply and consumption.

In recent years, for research on the carrying capacity of vegetation in arid areas, some studies have analyzed the water consumption of several dominant shrub and grass vegetation types in small areas, which is not suitable for the whole river basin [22,23,24]. There are also some studies from the perspective of soil moisture, through the soil moisture model to simulate the soil moisture content of different depth levels, and then fitted with vegetation coverage to analyze vegetation density carrying different soil moisture [25,26,27]. However, when implementing vegetation construction for a large area, it is not enough to only know the vegetation density. In arid areas, the consumption of water resources by forest trees and shrub grasses is different, and the planning benefits are also different [28,29,30]. So, for inland river basins in arid zones, analyzing the types of forest and grass vegetation that can be carried by water resources based on actual spatial differences in water supply can help to optimize the spatial distribution of vegetation, which can provide a scientific basis for vegetation planning and sustainable development of ecosystems in arid zones [31,32]. With the development of science and technology, a large number of available remote sensing satellite data and meteorological data products are available, providing technical support for the implementation of this idea [33,34]. Therefore, this study intends to analyze the ecological water production and consumption of the whole Aksu River Basin and explore the vegetation carrying potential combined with the ecological water demand threshold of different vegetation types of forests, shrubs, and grasslands.

The Aksu River is the main water source area of the Tarim River, and the Aksu River Basin is also an important agricultural irrigation area in Xinjiang. In recent years, the phenomenon of river disconnection has occurred frequently, the groundwater level has decreased, the lake and natural vegetation have shrunk to varying degrees, and the ecological function of the basin has been extremely weakened [35,36,37]. The reason for this situation is that, on the one hand, because the needs of human survival and economic development have led to a high demand for arable land and an increase in water required for agricultural irrigation, resulting in the occupation of some ecological water and the reduction in water for vegetation, and on the other hand, because part of the existing vegetation has experienced water overdraft, especially some plantation forests with high water consumption [38]. Zhang Yuxing [39] has found a similar situation in the ecological study of the Loess Plateau region, which is also an arid and semi-dry region, where vegetation plays a positive role in the ecological recovery of the Loess Plateau, but large-scale afforestation exacerbates the shortage of water resources, which affects the sustainable development of the ecological environment. Zhou Guoyi et al. [40] also pointed out that vegetation restoration would have a certain impact on the sustainable development of water resources at the same time. Part of the water source of the Aksu River Basin is surface water, and its basic supply is precipitation, which falls to the high cold area to form glaciers and falls to the plain and basin to be directly used [41], and the other part comes from groundwater, which is also provided by precipitation and other forms of infiltration. In recent years, the massive exploitation of groundwater in the Xinjiang region has led to a decline in the groundwater level, and it is an inevitable trend to reduce or even shut down some groundwater mining points [42,43]. Therefore, in the analysis of the ecological water yield, this study only considers the part of the precipitation that can be directly used by vegetation, regardless of the groundwater.

In general, this study takes Aksu River Basin as the research area, analyzes the amount of water resources available to vegetation from the two parts of directly usable precipitation and indirectly generated available irrigation water, combines the ecological water demand of different vegetation types such as forests, shrubs and grasslands, and researches the types of vegetation that can be carried by the region and their spatial distribution based on the guidance of the principle of “Water to determine the Vegetation”. This study can provide a scientific basis for the construction of forest and grass vegetation in arid inland river basin, and provide support for the sustainable development of the basin.

2. Study Area

The Aksu River Basin (40°11′~41°45′ N, 76°16′~81°17′ E) (Figure 1) is located in southwest Xinjiang, China, and at the northwestern edge of the Taklamakan Desert [44], with large differences in topography. The northwest of the study area is the Tianshan South Vein, with the highest elevation of 7019.0 m, and the southeast is close to the Tarim Basin, mostly plain areas, with the lowest elevation of 943.0 m. The overall pattern is higher in the northwest and lower in the southeast [37]. The climate of the Aksu River Basin is a temperate continental arid climate, with mostly sunny days, less precipitation, and high evaporation. The average annual sunshine hours in the region are above 2500.0 h, and the average annual evaporation is 202.8 mm. The average annual precipitation is only 89.3 mm, and the precipitation season is mainly in July and August, with more precipitation in the northeast mountainous area and less precipitation in the southwest plain area. The Aksu River is formed by the convergence of two major tributaries, the Kumalake River and the Taushgan River. It is the main source of the Tarim River, accounting for about 70% [45]. The high mountain area of Aksu River Basin is the gathering areas of precipitation and glaciers, which are funnelled along the valley to the runoff of the basin for regulation, which flows into the runoff of the basin along the valleys and are controlled by humans, while the plain and desert areas are the main loss areas of water resources, which mainly rely on precipitation [46,47]. There are irrigation areas distributed in the plains in the eastern part of the basin, accounting for 13.5% of the basin, mainly growing crops such as cotton, wheat, melons, and fruits, which are the main reliance of the local economic development. In recent decades, with the expansion of the agricultural irrigation area and the development of the local economy, water demand in the basin has continued to increase, and the amount of water coming from the rivers during the dry season has decreased, resulting in the intensified conflict between supply and demand of water resources [48], which makes the rational use of water resources in the basin particularly important.

3. Data and Methodology

3.1. Data

The main categories of data required for this study include the following:

(1)

Precipitation Data: The data were collected from the Data Center for Resource and Environmental Science, Chinese Academy of Sciences (http://www.resdc.cn, accessed on 11 October 2022), mainly including daily observations of meteorological element stations in Xinjiang from 1990 to 2020, and the annual values of each meteorological element station were obtained by summing the station data for each year. The annual total precipitation raster data were then generated at a spatial resolution of 1 km using Anusplin interpolation.

(2)

Potential Evapotranspiration Data: the data were obtained from the MODIS16A2 PET dataset (https://search.earthdata.nasa.gov, accessed on 15 October 2022) for the period from 1990 to 2020 with a spatial resolution of 500 m, which was adjusted to 1 km by resampling and a temporal resolution of 1 year.

(3)

Digital Elevation Model (DEM) data: ASTER GDEM 30 M resolution digital elevation data were collected from the geospatial data cloud (http://www.gscloud.cn/, accessed on 17 October 2022), with a spatial resolution of 30 m.

(4)

Soil Data: the data include soil type, texture, and soil depth data, all from the Data of the second National Land Survey in 2009 and the Harmonized World Soil Database (HWSD) (https://www.fao.org/soils-portal/en/, accessed on 27 October 2022), with a spatial resolution of 1 km.

(5)

Land Use and 1:1 million Chinese Vegetation Zonation Data: the spatial resolution of the data was 1 km from the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (http://www.resdc.cn, accessed on 3 November 2022), from which the land use data was downloaded from 2020.

(6)

Xinjiang Statistical Yearbook Data: the data for 2020 were derived from the Xinjiang Uygur Autonomous Region Bureau of Statistics (https://tjj.xinjiang.gov.cn, accessed on 27 October 2022), including information on agricultural water consumption, population numbers, domestic water consumption, agricultural land area, industrial water consumption, GDP data, and other data by the region.

3.2. Research Methodology

3.2.1. Method Overview

As shown in Figure 2, the main methods and workflow used in the study are detailed as follows: (1) the ARIMA model was used to predict the precipitation station data for the next 30 years based on the annual precipitation station data from 1990 to 2020, and the precipitation values of the stations in the normal year were interpolated to obtain the raster data of the normal year precipitation, and then the available effective precipitation is obtained by deducting runoff based on water balance method; (2) the surface water production was calculated using the InVEST model, and the water consumption was obtained based on the data of the Xinjiang Statistical Yearbook; (3) the net surplus of the surface water in the region was calculated by calculating the difference between the surface water production and the water consumption, and 30% of the surface water was taken as the available irrigation water; (4) the actual evapotranspiration of the basin was determined, and the ecological water demand threshold of each vegetation type was calculated by superimposing the vegetation zoning map and land use map, and the cumulative value of 10% of the area was taken as the ecological water demand; (5) the theoretical carrying potential distribution of the forest and grass vegetation in the Aksu River Basin was obtained according to the available precipitation, available irrigation water, and ecological water demand of various vegetation types.

3.2.2. Calculation of the Available Effective Precipitation

In the arid zone, precipitation is an important source of water for vegetation growth in the study area, but not all of the precipitation is available for vegetation, and part of it will form runoff that is not utilized by vegetation [49], so this part of precipitation should be deducted from the calculation of the effective available precipitation for vegetation [50,51]. In order to adapt the precipitation analysis to future periods, this study uses the Autoregressive Integrated Moving Average Model (ARIMA) to forecast the annual precipitation for the next 30 years based on the data of the site for the past 30 years. At the same time, in order to reflect the average level of precipitation for many years, the concept of “Precipitation Guarantee Rate” is introduced, which is the cumulative frequency of precipitation greater or less than a certain threshold value in a certain period; it is used to indicate the reliability of the precipitation occurring at that value [52,53]. In this study, using the continuous (partially discontinuous) precipitation data from the past 30 years and the next 30 years, a value with a guaranteed rate of 50% was obtained based on the Pearson Type III Curve Model combined with the Visualizing Optimal Line Method, which is the value of the normal year, and we took the minimum value as the final value. Then, the Anusplin interpolation method was used to obtain the level-water annual precipitation data in the Aksu River Basin combined with the DEM data for spatial interpolation. In hydrology, effective precipitation is the portion of precipitation that generates runoff, which has also been defined as the amount of precipitation that seeps into the soil and is stored in the main root uptake layer of the vegetation [17]. The effective precipitation in this study is the portion of total precipitation that is used to meet the needs of crop growth, which provides the amount of precipitation that plants use for evapotranspiration, thus reducing the crop’s need for irrigation water, which excludes the portion of surface runoff and infiltration to below the root zone of plants [54,55]. This study only analyzes water resources in areas where vegetation can be constructed, so it does not consider areas such as construction land, wetlands, water bodies, and cropland where land properties cannot be changed, and the effective precipitation of vegetation can be directly considered as 0. For the calculation of the effective precipitation in other areas, this study uses the Water Balance Method [56,57,58]. The Water Balance Method takes precipitation as the total input and considers all output components such as precipitation, runoff, infiltration, evaporation operations, and deep infiltration [59], and the expression of the ecosystem water balance equation is as follows:

$$P^{\prime }=\sum E+\sum R\pm ∆W\pm ∆Q\pm ∆K$$

(1)

where $P^{\prime }$ is the annual precipitation of the normal year; $\sum E$ is the total evapotranspiration; $\sum R$ is the total runoff volume; $∆W$ is the change in water storage in soil; $∆Q$ is the change in the water storage in the vegetation; $∆K$ is the variation in the water storage in the litter layer. In a longer period of time, $∆W$, $∆Q$, and $∆K$ in the water balance equation can be regarded as 0; that is, the precipitation can be divided into two parts: one part evaporates and transpires out in the form of water vapor, and the other part flows in the form of liquid water. Thus, the equation can be simplified as follows:

$$P^{\prime }=\sum E+\sum R$$

(2)

$$\sum E=T+EI+ES$$

(3)

$$\sum R=R_s+R_g+R_m$$

(4)

where T is the vegetation transpiration (including the canopy, understory, and herbaceous layer); EI is the intercept evapotranspiration (equal to the vegetation interception volume); ES is the soil evapotranspiration; R_s and R_g are, respectively, the surface runoff and underground runoff formed by precipitation and glacier melt; and R_m is flowing in the soil, which is very small and largely negligible [60]. The surface runoff and subsurface runoff are exported out of the system after precipitation occurs and cannot be used directly, while the vegetation evapotranspiration, soil evaporation, and vegetation interception volume are water resources essential for the survival and maintenance of vegetation and the ecosystem in which it is located. So, the value of precipitation after deducting the amount of water input to surface runoff and underground runoff is the theoretical maximum effective precipitation that can be obtained by vegetation ecosystem. The ratio of the surface runoff and subsurface runoff to the precipitation can be calculated as follows:

$$W_a=P^{\prime }-R_s-R_g$$

(5)

$$W_a=P^{\prime }\ast (1-\alpha -\beta )$$

(6)

where W_a is the effective precipitation available to the vegetation in the ecosystem; α is the ratio of the surface runoff to the precipitation, which was the surface runoff coefficient (%); β is the ratio of the subsurface runoff to the precipitation, which was the subsurface runoff coefficient (%).

The runoff coefficient is positively correlated with different vegetation types. Runoff coefficients can be determined by regional land vegetation types [61,62]. In the Aksu River Basin, vegetation can be constructed mainly in the forest region, shrub and grass regions, the desert region, and the bare region. In the desert and bare areas of the study area, the proportion of precipitation converted to runoff is almost negligible, and the effective precipitation can be equal to the actual precipitation, so the runoff coefficient is 0. In the case of natural precipitation, the dominant vegetation communities in areas covered by vegetation, such as forest areas and shrub areas, are adapted to the regional environment and have relatively stable community characteristics. The dominant vegetation communities distributed in the climatic zone of the region can be obtained by superimposed on the 1:100 million Vegetation Zones of China data, and the results found that the Aksu River Basin has a temperate continental arid climate, and the top vegetation communities are shrubs and semi-shrubs of warm belt. In this research, by searching the existing literature related to the study of runoff coefficients, we refer to the information of runoff parameters from a region near Tianshan, which has the same climatic zone and the same vegetation community as the study area, and the percentage of precipitation inputs to surface runoff and subsurface runoff are 2.75% and 1.05%, respectively [63].

3.2.3. Calculation of the Availability of Irrigation Water

The available irrigation water in this research is specifically the water that can be used for ecological vegetation construction after the annual output of a region reduces the water consumed in production and living. To obtain the amount of water available for irrigation, the annual water yield and annual water consumption need to be calculated first.

Water Yield

The water yield is the volume of runoff generated per unit basin area in a specified time period. The calculation methods used for the water yield include the Soil Water Storage Capacity Method, Water Balance Method, Annual Runoff Method, and other methods [64]. With the development of remote sensing technology, there are more and more mature simulation models; among them, the InVEST model is the most widely used, which fully considers the interference of different factors and has relatively simple data requirements. It can be used to evaluate the water yield of the system. The specific calculation method is as follows [65,66]:

$$Y_{xj}=\left(1-\frac{{AET}_{xj}}{P_x}\right)\times P_x$$

(7)

where $Y_{xj}$ is the annual water yield of land use type $j$ on raster $x$; $P_x$ is the average annual precipitation of raster $x$; ${AET}_{xj}$ is the average annual actual evapotranspiration of land use type $j$ on raster $x$.

$$\frac{{AET}_{xj}}{P_x}=\frac{1+{\mathit{\omega }}_x\times R_{xj}}{1+{\mathit{\omega }}_x\times R_{xj}+\frac{1}{R_{xj}}}$$

(8)

where ${\mathit{\omega }}_x$ is a dimensionless parameter of nonphysical properties, reflecting the actual conditions of the soil and vegetation in the basin under natural climatic conditions; $R_{xj}$ is the Budyko drying index of land use type $j$ on the raster $x$. The specific calculation equation is as follows:

$$R_{xj}=\frac{k_{xj}\times {ET}_{0x}}{P_x}$$

(9)

$${\mathit{\omega }}_x=Z·\frac{{AWC}_x}{P_x}$$

(10)

$${AWC}_x=Min(D_s,D_r)\times {PAWC}_x$$

(11)

$$\begin{gathered} {PAWC}_x=54.509-SAN\left(0.132+0.030SAN\right)-SIL(0.055-0.006SIL)-CLA(0.738+0.007CLA) \\ -C(2.668-0.501C) \end{gathered}$$

(12)

where ${ET}_{0x}$ is the potential evapotranspiration on raster $x$, mm; $k_{xj}$ is the vegetation evapotranspiration coefficient of land use type $j$ on raster $x$, using standard crop coefficients, obtained according to the Crop Evapotranspiration Factor Guidelines of the Food and Agriculture Organization of the United Nations and the InVEST 3.2.0 User Guide [67]; ${AWC}_x$ is the vegetation water content on raster $x$, mm; Z is the Zhang coefficient, namely, the seasonal parameter, which is an empirical parameter obtained by calibration using the measured runoff from the basin, and Guo Lige et al. [68] determined the optimal value in the calculation of the water yield in the Aksu River Basin in 2020 with several adjustments, and the value of 0.69 was taken as the basis in this study; $D_s$ is the soil depth data; $D_r$ is the root depth data, which were obtained for different feature use types based on studies of the maximum root depth of vegetation on a global scale in the relevant literature [69]; ${PAWC}_x$ is the available moisture of the vegetation on raster $x$, mm; $SAN$ is the sand content of the soil; $SIL$ is the powder content of the soil; $CLA$ is the clay content of the soil; $C$ is the organic matter content of the soil on raster $x$.

Water Consumption

According to the Xinjiang Regional Statistical Yearbook, the annual consumption of water is mainly composed of agricultural, residential, and industrial water consumption, and different types of water consumption raster data are obtained based on the interpolation of statistical data for each region in Xinjiang. The amount of water consumption represented by each raster is as follows [70]:

$$C_x={Agr}_x+{Ind}_x+{Dom}_x$$

(13)

where $C_x$ represents the water consumption of raster $x$, mm; ${Agr}_x$ represents the agricultural water consumption of raster $x$, mm; ${Ind}_x$ represents the industrial water consumption of raster $x$, mm; ${Dom}_x$ represents the domestic water consumption of raster $x$, mm. The specific calculation formulae are as follows:

$${Agr}_x={Area}_{Agr}\times {Com}_{Agr}$$

(14)

$${Ind}_x={Vol}_{GDP}\times {Com}_{GDP}$$

(15)

$${Dom}_x={Num}_{pop}\times {Com}_{pop}$$

(16)

where ${Area}_{Agr}$ represents the area of the agricultural land of raster $x$; ${Com}_{Agr}$ represents the water consumption per unit area of the agricultural water, mm; ${Vol}_{GDP}$ represents the total GDP of raster $x$; ${Com}_{GDP}$ represents the water consumption per 10,000 Yuan GDP, mm; ${Num}_{pop}$ represents the population of raster $x$; ${Com}_{pop}$ represents the water consumption per person, mm.

Availability of Irrigation Water The remaining water after subtracting the water consumption from the regional supply cannot be fully used for the ecological construction of vegetation, and a large part of it needs to be used to ensure the river runoff to maintain the ecological balance and purification capacity of the river, which is called the river environmental flow [71]. The available water for irrigation is calculated by the following formula:

$$W_x=\mathit{\delta }\ast \left(\sum Y\left(x\right)-\sum C\left(x\right)\right)$$

(17)

where $W_x$ is the amount of available water for irrigation in the study area for ecological vegetation construction; $Y\left(x\right)$ is the annual water yield on raster $x$ in the study area; $C\left(x\right)$ is the water consumption on raster $x$ in the study area; $\mathit{\delta }$ the proportion of the local net surplus water available for ecological vegetation construction, which is affected by the runoff flow. The runoff flow in arid areas is required to be unaffected by ecological water use and not to change its ecological category and ecological function. The runoff flow in arid areas should first determine its flow management level according to the maintenance standards of its ecological conditions and then further determine the flow required to maintain the health of river ecosystems based on the empirical relationship between flow and ecology, specifically referring to the river flow management level table in the literature of Hughes et al. [72] and Han et al. [73]. In this study, the amount of the remaining water resources available for ecological vegetation construction should be prioritized to meet the ecological conditions of not damaging the original river ecosystem, ensuring the integrity of the water resources development, maintaining from 60% to 80% of the original annual average runoff, with an average value of 70%, so the δ value in Equation (17) is 30%.

Rationalization of Surface Water Resource Estimates

The amount of precipitation that is directly utilized by vegetation and the amount that is indirectly utilized by vegetation in the form of converted irrigation water make up the surface water available to vegetation. However, there are some coefficients that cannot be calculated explicitly in the process of calculation, such as the Z coefficient of the irrigation water yield, runoff coefficient, and vegetation evapotranspiration coefficient, all of these are defined and referenced in the relevant literature and standards guidelines, and the objectivity of the water resources estimation is a problem that needs attention. In this study, the whole Aksu Region is taken as the verification sample, and based on the above research methods, it is calculated that the available effective precipitation in the Aksu Region is 12.42 × 10⁹ m³, the water yield is 0.57 × 10⁹ m³, and the total amount of water resources is 12.99 × 10⁹ m³. According to the statistical yearbook, the value of the surface water resources in the Aksu Region obtained after subtracting the duplicated amount of the surface water and groundwater resources is 12.66 × 10⁹ m³, which is close to the calculated value, and the rationality of the estimation result can be proved on the side.

3.2.4. Definition of the Ecological Water Demand Threshold for Vegetation

In this study, the definition of vegetation ecological water demand refers to the water consumption that maintains the steady growth of natural vegetation and artificial vegetation protection systems, which can be divided into critical ecological water demand, optimal ecological water demand and saturated ecological water demand [74]. The Vegetation Evapotranspiration Method calculated the actual water demand of vegetation by the potential evapotranspiration of vegetation, and the actual water demand of vegetation was used as the ecological water demand of vegetation, which can reflect the actual water demand of vegetation growth. Therefore, this method was selected for the calculation of the ecological water demand in the study. The actual evapotranspiration of vegetation was calculated by using MODIS data and the Penman–Monteith formula, and the formula is as follows:

$$ET=P\times \left\{1+\frac{{ET}_0}{P}-{\left[1+{\left(\frac{{ET}_0}{P}\right)}^m\right]}^{1/m}\right\}$$

(18)

where $ET$ is the actual evaporation of land surface vegetation, mm; ${ET}_0$ is the potential evapotranspiration, mm; m is a parameter representing the underlying surface permeability, vegetation status, and topography, and m = 2 is generally adopted. $P$ is the precipitation, mm. After calculating the actual evapotranspiration of vegetation, the ecological water demand of each vegetation type can be obtained by superimposing the spatial distribution data of regional vegetation types and land use data.

4. Results

4.1. Available Effective Precipitation

It can be seen from Figure 3 that the overall topography of Aksu River Basin is quite different, among which the northwest is a high-altitude mountain area, the southwest desert area has a slight elevation difference, and the southeast area is almost all plain area. Based on the precipitation data of the past 30 years and the next 30 years from 1990 to 2020, the available effective precipitation of the Aksu River Basin was obtained (Figure 4a). It can be seen that the range of the available effective precipitation in the Aksu River Basin was between 0 and 174.50 mm, the average available effective precipitation per unit area was 116.78 mm, and the total available precipitation was 56.49 × 10¹¹ m³. The available effective precipitation of the vegetation also has a similar spatial distribution, with a significant vertical zonal distribution, which increases with elevation. The available effective precipitation of the vegetation in high-altitude mountainous areas is the highest, with an average of 135.34 mm, followed by that in the desert region, with an average of 119.58 mm, and the available effective precipitation of the vegetation in the plains area where the population gathers is the lowest, with an average of 93.23 mm.

4.2. Availability of Irrigation Water

The water yield per unit area in the Aksu River Basin ranged from 0 to 168.06 mm, with an average water yield per unit area of 8.31 mm and a total water yield of 40.18 × 10¹⁰ m³. From the spatial distribution pattern of the water yield, the water production in the Aksu River Basin was mainly distributed at the intersection of the high mountains to the low mountains in the east (Figure 4b). The water consumption per unit area in the Aksu River Basin ranged from 0 to 549.51 mm, with an average water consumption per unit area of 75.09 mm and a total water consumption of 36.32 × 10¹¹ m³. From the spatial distribution pattern of water consumption (Figure 4c), the water consumption in the Aksu River Basin was mainly distributed in the rivers and the eastern population gathering area. Irrigation water was obtained based on water production and consumption, only the area greater than zero was analyzed, which meant that there was a surplus of water. The available water for irrigation per unit area in the Aksu River Basin was between 0 mm and 50.29 mm, the average water consumption per unit area was 2.12 mm, and the total water consumption was 10.26 × 10¹⁰ m³. From the perspective of the spatial distribution pattern (Figure 4d), it was mainly distributed in the eastern high-altitude mountainous area, where alpine snowmelt and rainwater were integrated and supplied in many aspects, while the desert region and the plains area where the population gathers were characterized by severe runoff and less water supply.

4.3. Threshold of Ecological Water Demand for Vegetation

From the whole basin, the actual evapotranspiration per unit area in the Aksu River Basin ranged from 30.46 to 839.90 mm, with an average actual evapotranspiration per unit area of 237.64 mm. In terms of the spatial distribution pattern, the actual evapotranspiration in the Aksu River Basin was lowest in the high mountainous area in the west of the north, followed by the actual evapotranspiration in the populated area and the low mountainous area, and it was highest in the surrounding plain desert area (Figure 5). The actual evapotranspiration data of the Aksu River Basin were overlaid with the 1:1 million vegetation type map and the 2020 land use data to obtain the cumulative values of the actual evapotranspiration characteristics of four vegetation types, including forests, shrub vegetation, grassland vegetation, and desert vegetation, where the ecological thresholds ranged from 235.60 to 627.80 mm for forests, 85.52 to 504.80 mm for shrub vegetation, 57.88 to 607.40 mm for grassland vegetation, and 58.31 to 478.40 mm for desert vegetation.

Hao et al. [75] found that the 10% area accumulation rate of the annual ET data could be used as the minimum ecological water demand of vegetation in the study of the arid River Basin, so in this study, the ecological water demand of each vegetation type was set as the value of the ecological water demand of each vegetation type in the Aksu River Basin when the cumulative area of each vegetation type in the whole basin accounted for 10% (Figure 6). Finally, based on the actual evapotranspiration of the Aksu River Basin, the ecological water demand of the forest, shrub, grass, and desert vegetation in the Aksu River Basin were 259.65 mm, 144.65 mm, 136.37 mm, and 80.12 mm.

4.4. Distribution of Vegetation Theoretically Carried by Water Resources

In this study, the comprehensive available water of regional vegetation was obtained by integrating available precipitation and available irrigation water, with the value ranging from 0 to 221.71 mm. Then, based on the water demand threshold of forest and grass vegetation, the spatial distribution pattern of the carrying vegetation was obtained by removing the areas that could not change the land use attributes, such as rivers, construction land, and cultivated land (Figure 7). From the spatial distribution pattern of vegetation types that can be carried by water resources in the Aksu River Basin, it is distributed in a strip pattern from north to south, followed by shrub vegetation, grass vegetation and desert vegetation, among which the water resources conditions in the southern and northeastern regions are relatively poor. According to the statistics, the total area of the shrub, grass, and desert vegetation that could be carried by the water resources theory in the Aksu River Basin was 3.76 × 10⁴ km², and the carrying shrub vegetation area was 1.17 × 10⁴ km², accounting for 29.05%, and the vegetation area of the carrying grassland vegetation was 0.43 × 10⁴ km², accounting for 10.74%, and the carrying desert vegetation area was 2.16 × 10⁴ km², accounting for 53.45%, and the rest was bare land.

5. Discussion

5.1. Water Yield Analysis

The Aksu River Basin is located in the northwestern arid region, and its socio-economic activities are based on irrigated agriculture but also include pastoralism and industry. It is characterized by an alternating geomorphic pattern of mountains and basins, and water is the link between the three areas of mountains, deserts, and construction areas [76]. The vegetation is dominated by halophytes, xeric shrubs, and herbaceous vegetation, which are characterized by low height, sparsity, and low coverage [48]. There are artificial oases distributed around the basin, which are mainly supplied by irrigation water, and the spatial distribution of water resources is extremely unbalanced, with more water in the summer and less in the winter, and the contradiction of water resources in all aspects of the ecosystem is prominent [77]. Therefore, the ecological environment of this basin is typical in arid areas, and the estimation of the vegetation carrying potential based on the water resource constraint is representative of arid inland river basins. In previous research on vegetation carrying capacity in arid areas, it is mostly considered that vegetation cover has a great influence on the ecological water consumption of the natural environment, and analyzed regional carrying capacity from the perspective of vegetation coverage [78,79,80,81]. However, the vegetation planning of arid basins should be based on the theoretical spatial distribution of vegetation types that combines qualitative and quantitative factors, among which the analysis of the ecological water yield is the basis [82,83,84].

Considering that the continuous exploitation of groundwater in recent years has caused certain risks for the stable development of vegetation communities [85,86,87], the ecological water yield in this study mainly analyzed the amount of water resources that can be directly utilized by vegetation from precipitation, which is the main source of surface water and also affects the characteristics of river runoffs. With the diversity of the global climate environment, the annual precipitation is divided into wet years, dry years, and normal years, and the fluctuation of the precipitation will affect the vegetation community [88,89]. In order to solve this problem, this study selects precipitation in normal years to analyze the amount of water resources available to vegetation, which not only ensures the maximum benefit of vegetation construction but also reduces the risk of water resource overdraft caused by excessive vegetation loss in dry years [90]. The water yield module in the InVEST model is based on the water balance method to simulate the spatial distribution. On the one hand, factors such as precipitation, potential evapotranspiration, land use, and vegetation water content are considered, which are generally comprehensive; on the other hand, the input data are relatively small, easy to operate, and the scope of application is wide [91,92,93,94]. Therefore, this model was used in this study for the water yield analysis.

5.2. Ecological Water Demand of Vegetation

The ecological water demand of vegetation refers to the water consumed to maintain stable growth [95,96], that is, the amount of water that the ecological environment must supply for the growth of vegetation, including the biomass method, the area quota method, the potential evapotranspiration, the water balance method, and other commonly used calculation methods [97,98,99,100], which have large calculation errors and uncertainties for arid areas with sparse vegetation [101]. For the analysis of the natural vegetation in a large range of arid areas, it is more appropriate to use evapotranspiration value obtained from remote sensing image data to calculate the ecological water demand of vegetation, which not only considers the relatively complete water consumption of the vegetation in the ecosystem but also reflects its spatial heterogeneity. Hao et al. [75] conducted a consistency test on the ecological water demand of the natural vegetation extracted based on evapotranspiration in the northwestern arid region, and the results showed that the method was more accurate and suitable for different vegetation types, and found that the ET data value when the cumulative area reached 10% could be used as the minimum ecological water demand of vegetation. In arid areas, water resources are affected by local socio-economic development such as agriculture and industry, and water resources should also meet the stability of river ecosystems, the amount of water provided to vegetation is not enough. Therefore, the maximum limit of carrying vegetation is the condition when water resources meet the lower limit of the water demand for vegetation growth, and beyond that, it will lead to the negative succession of the vegetation community [102,103].

6. Conclusions

Considering that the impact of climate change on the arid basin is mainly reflected in precipitation and temperature, the amount of available water resources is calculated based on precipitation, and the ecological water requirement of vegetation is considered from the actual evapotranspiration. The method of water balance was used to analyze the theoretical distribution of forest and grass vegetation under the constraint of water resources by using water to determine the vegetation. Among them, the ecological water demand of the forest vegetation, shrub vegetation, grassland vegetation, and desert vegetation in the Aksu River Basin, respectively, were 259.65 mm, 144.65 mm, 136.37 mm, and 80.12 mm. The water resources of the Aksu River Basin can carry vegetation types including shrub vegetation, grassland vegetation, and desert vegetation; of these, shrub vegetation is mainly suitable for the most water-rich alpine areas, grassland vegetation is suitable for low mountains, and the area that can carry desert vegetation is the largest and is suitable for the desert plain region. At the same time, there are also some areas that can be improved. The calculation of water resources does not take into account the secondary replenishment of alpine glacier runoffs caused by climate warming. The snowmelt submodule of the SWAT model can be considered to enrich the analysis of water resources in the alpine region. The classification of vegetation types is not accurate enough, the spatial resolution is 1000 m, and the water consumption of different tree species is not considered. Therefore, the water demand of the dominant vegetation in the basin can be further optimized and analyzed to obtain more a detailed distribution of the vegetation types.