Effects of afforestation on water resource variations in the Inner Mongolian Plateau

Study area

Inner Mongolia is located in the arid and semi-arid climate region. The average annual temperature is between −5∼10 °C, and the average annual precipitation is between 35∼530 mm, Under the comprehensive influence of temperature and water conditions, vegetation zones mainly show the spatial differentiation characteristics in the near meridional direction. From east to west, they are successively forest, forest grassland belt, typical grassland belt, desert grassland belt and desert belt. The dominant land type in the study region is grassland (Fig. 1), which includes steppe and sparsely distributed grass, comprising of approximately 73.2% of the plateau, mostly distributed throughout the central and eastern sections. With the increase of population and the development of economy, the disturbance degree of human activities on the grassland gradually enhance, resulting in a certain degree of ecological degradation in this region. In recent years, the Government of China has implemented several ecological protection and construction programs to alleviate grassland degradation in Inner Mongolia (Fig. 2), however, these programs could also significantly affect the eco-hydrological process of grassland ecosystems. The ecological protection and construction planning programs mainly include the Three-North Shelter Forest Program, the conversion from farmland to forest and grassland, sandy land control, and ecological relocation projects.

Data sources

In order to determine afforested area within the Inner Mongolian Plateau, we used afforestation data across Inner Mongolia from 1980 to 2015. We obtained afforestation data from annually published China forestry statistical yearbooks from 1980 to 2015 (State Forestry Agency of the People’s Republic of China (1980–2015)). All cartographic data were converted to the same coordinate system (Albers equal-area conic projection) and the same spatial resolution (1 km).

We obtained Chinese ecosystem data between 1980 and 2015 (the study period) from the Data Sharing Infrastructure of Earth System Science at the Chinese Academy of Science (http://www.geodata.cn). These data were produced at a resolution of 1 km, using visual interpretations of Landsat (MSS/TM/OLI) images with an average overall accuracy greater than 94%.

We obtained meteorological between 1980 and 2015 from the National Metrological Information Center of the China Meteorological Administration (http://data.cma.cn). Digital elevation model (DEM) data originated from the Shuttle Radar Topography Mission (SRTM), with a resolution of 90 m. Administrative divisions were provided by the Satellite Environment Center, the China Ministry of Environmental Protection (MEP). Lastly, we obtained the continental Global Inventory Modeling and Mapping Studies (GIMMS) normalized difference vegetation index (NDVI, NDVI = (NIR-R)/(NIR + R), NIR is near-infrared band, R is red band) dataset, with a spatial resolution of 8 km and covering a period from 1982 to 2015, from the Ecological Forecasting Lab at the NASA Ames Research Center (https://ecocast.arc.nasa.gov).

Methods

(1) Water resource quantification:

Considering the special climatic and geographical environment in Inner Mongolia, there is a close interaction between climate factors, hydrocycle, and evaportranspiration. Budyko (1974) proposed a theoretical framework to describe the effect of precipitation and evapotranspiration on water storage based, which is an effective method to study the hydrological response to climate change. To quantitatively analyze water resources, we applied a water balance equation based on the Budyko model to the hydro series (from 1980 to 2015).

For this analysis, we do not consider the interaction between surface and groundwater on horizontal direction, due to the interaction usually happens in the vertical direction (Xiao & Xiao, 2019).

The water balance of a catchment can be described as follows (Xiao & Xiao, 2019): (1)${\displaystyle Q=P-E{T}_a-\Delta S}$ where Q is the water storage (mm) as a proxy of the water resource; P and ET_a represent precipitation (mm) and actual evapotranspiration (mm); and ΔS represents a change in catchment water storage (mm), which is usually assumed to be zero over a long period of time (Xiao & Ouyang, 2019).

(2) Evapotranspiration factor (ET):

Following an assumption similar to that made by Budyko (1974), actual evapotranspiration can be estimated as follows (Shen et al., 2015): (2)${\displaystyle ET=\frac{P\times E{T}_p}{{\left(P^n+E{T}_p^n\right)}^{1∕n}}}$ where ET is the actual evapotranspiration (mm); n is the model controlling parameter that determines the shape of the Budyko curve, which mainly represents the integrated effects of catchment land-cover characteristics on the water balance. ET_p is the potential evapotranspiration, and it can be estimated by the Priestley–Taylor (PT) equation (Priestley & Taylor, 1972) as follows:

(3)${\displaystyle E{T}_p=\alpha \times \frac{\Delta }{\Delta +\gamma }\times \left(R_n-G\right)}$ (4)${\displaystyle \Delta =\frac{\text{4,098}\times \left(0.6108exp\left(\frac{17.27T}{T+237.3}\right)\right)}{{\left(T+237.3\right)}^2}}$ (5)${\displaystyle \gamma =0.665\times 10^{-3}\times 101.3\times {\left(\frac{293-0.0065H}{293}\right)}^{5.26}}$

where α is the PT coefficient of 1.26 for open water and saturated land (Priestley & Taylor, 1972); Δ is the slope of the saturated vapor pressure curve (kPa °C⁻¹); γ is the psychrometric constant (kPa °C⁻¹); R_n is the net radiation absorbed at the surface in MJ m⁻², which can be estimated using calibrated radiation in the FAO 56-PM model; G is the downward-directed soil heat flux in MJ m⁻², where G = 0.26 R_n ; T is the mean air temperature (°C); H is elevation above sea level (m).

(3) Spatiotemporal analyses:

The ‘Spatial Analyst’ tools in ArcGIS 10.3 were employed to reveal the spatial characteristics of water resources in different regions. To detect the variation in trends of water resources during the study period (1980–2015), a least squares linear regression model which is a commonly used method in trend analysis of variation, can be used to obtain the trend of every pixel change by fitting a linear equation of water resource variables as a function of time (year). The trend value was calculated by programming in Interactive Data Language (IDL) 4.8 (Harris Corporation, Melbourne, FL, USA).

(4) Abrupt change analyses:

We mainly conducted analysis of abrupt changes in water resources using the Mann-Kendall (MK) test (Xiao & Ouyang, 2019). We used this method on the premise that the positive sequence curve UF_k crosses the critical ratio reliability line; therefore, if the positive sequence (UF) and the inverse sequence (UB, generated with the reverse data series of UF) have only one obvious crossing point located between the reliability lines, this denotes the catastrophe point and is statistically significant. On the other hand, if the crossing point is located outside the reliability line or there are several obvious crossing points between the lines, a definite catastrophe point cannot be established. In this latter case, we used the MK test on different lengths of sequences separately on the basis of a moving t-test technique. If a catastrophe point was still shown in these different sequences, we could then confirm that this point was the definite catastrophe point (Pettitt, 1979). The Mann–Kendall trend test can be described as follows:

For a time series X = {x1, x2, …, xn}, the test statistic is given by

${\displaystyle S_k=\sum _{i=1}^{i=k}{n}_i.}$

The statistics time series S_k is from i = 1, 2, …, k, for any dual value (x_i, x_j), if j < i, x_i > x_j, an upward trend is stated, n_i is the total number of dual values with growth trend.

The expected variance of S_k is calculated as follow,

E(S_k) = k(k − 1)∕4, Var(S_k) = k(k − 1)(2k + 5)∕72.

and the forward trend sequence UF_k is given as:

${\displaystyle U{F}_k=\frac{S_k-E\left(S_k\right)}{\sqrt{\mathrm{V\; ar}\left(S_k\right)}}.}$

Validation

To verify the reliability of simulated water resources in Inner Mongolia, we validated simulated results against observed data, which are available from the China Water Resources Bulletin from 1998 to 2015 (Ministry of Water Resources of the People’s Republic of China; 1998–2015).

Overall variation trends of water resources and vegetation on the Mongolian Plateau

Before the spatio-temporal analyses, we Validate the reliability of simulated water resources in Inner Mongolia, and the simulations agreed well with corresponding estimates (average r = 0.84; p < 0.001; Fig. 3). We could therefore confirm that simulated water resources could also be considered reasonable.

In recent years, water resources in Inner Mongolia have fluctuated but have generally exhibited a declining trend between 1980 and 2015 (a decrease of 0.2 mm/yr). This declining trend is unfavorable to vegetation growth. NDVI exhibited a similar trend. On a continental scale, 1998 generated the maximum water resource and NDVI values (Fig. 4). At the same time, Figure 4 also provides interannual variations and their influencing factors (such as temperature, evapotranspiration, precipitation, and plantation area) throughout 1980–2015. The decrease in rainfall and the increase in evapotranspiration in the region are not favorable to water resource storage. The characteristic of statistics of the six variables was shown as Table 1.

Declining water resources in the Mongolian Plateau

On a regional scale, almost the entire Mongolian Plateau has experienced water shortages between 1980 and 2015 (Fig. 5A). As a whole, water resources decreased significantly in the eastern section of the plateau, especially in the Horqin District and the Hulunbuir Plateau. By contrast, water resources increased as a whole in the western section of the plateau (Alxa Plateau). To evaluate long-term changes in water resources, we calculated the turning points of trends using the MK trend test and the Pettitt’s test. As shown in Fig. 5B, we identified two distinctly different trends throughout 1980–1998 and 1999–2015.

Causes of water resource variations on the Mongolia Plateau

There were obvious relationships among the various driving factors and water resources (Table 2). Our regression results found significant effects for six parameters associated with water resource changes from 1980 to 2015 (Table 3). For parameter value effects on water resources, the highest contribution was increasing precipitation (accounting for 39.35% of the total water resource changes), followed by increasing evapotranspiration (−24.48%), increasing DEM (22.72%), and increasing village density (−7.10).

Temporal pattern of water resource variation

Results from our study clearly show that water resources in Inner Mongolia fluctuated significantly but generally exhibited a declining trend between 1980 and 2015 (a decrease of 0.2 mm/yr). NDVI exhibited a similar trend (Miao et al., 2016; Shi et al., 2019). These trends are not favorable to vegetation growth. Potential causes for these declining trends are that water consumption of artificial vegetation is far higher than that of natural vegetation; thus, the practice of restoring the environment and water resources in the Inner Mongolia Plateau by means of afforestation is likely to further aggravate existing water resource shortages (Liu et al., 2018a; Cao et al., 2019).

Figure 3 also shows interannual variations of influencing factors (such as temperature, evapotranspiration, precipitation, and planting area) from 1980 to 2015. Since 1978, a series of large-scale ecological restoration projects have been implemented in China, including the Three-North Shelter Forest Program, natural forest protection projects, and conversion projects from farmland to forest and grassland to name a few. Greater than 30 yr of practical experience has confirmed that green vegetation cover within the study area has rapidly increased by means of these environmental policies (Lu et al., 2018). However, compared to the natural recovery of abandoned land, afforestation in arid areas has reduced total vegetation cover, increased degraded areas, and aggravated local desertification. The key internal mechanism that causes such water-related problems is as follows, when precipitation is less than evapotranspiration, vegetation (including trees) will be forced to rely on groundwater extraction to survive, leading to the depletion of surface water and a decline in groundwater levels, thereby reducing the availability of water resources for all vegetation (Wu et al., 2017). At the same time, a decrease in rainfall and an increase in evapotranspiration are not favorable to water resource storage.

Our results also showed that this water resource reversal in the Mongolian Plateau is intensifying, namely, that vegetation has begun to further degenerate due to large-scale afforestation initiatives (by an increase of 12.79 × 103 ha/yr) and global climate warming (by an increase of 0.044 °C/yr).

The MK trend test and Pettitt’s test showed that there was an abrupt change in water resources around 1998, namely, water resources declined rapidly after 1998. This was mainly due to the fact that precipitation in Inner Mongolia has decreased yearly since 1998 while afforestation area has increased rapidly since that time, which has caused an annual increase in temperature and increased consumption of local water reserves. Along with this decline in water resources, the status of grassland vegetation has also gradually degenerated.

Spatial pattern of water resources variation

Our research results clearly found a significant reduction in water resources in the eastern region of the Mongolian Plateau, especially in the Horqin District and the Hulunbuir Plateau. In contrast, there has been an increase in the total amount of water resources in the western region (Alxa Plateau) (Yao et al., 2017). The cause of this duel phenomenon is likely the implementation of ecological construction projects, such as natural forest resource protection initiatives (Liu et al., 2018b), the conversion from farmland to forest (Chang et al., 2011), the Three-North Shelter Forest Program (Wang et al., 2007), and compensation for public welfare forests (Yang et al., 2012), which have resulted in increased evapotranspiration and reduced rainfall in the region, while the usual practice of the Government of China is to simply promote a balance between different ecological service policies and practices.

With respect to the current increase in rainfall in the western region of Inner Mongolia, we believe this region should primarily be designated grassland. Namely, the protection and improvement of natural grassland ecosystems should be the key focus of the government, that it should abide by measures that combine protection and construction initiatives, and by sponsoring measures suitable to local conditions to further promote the ecological function of grassland and effectively alleviate grassland degradation and desertification trends, thereby developing a virtuous cycle (Yang et al., 2019b). We should also plan livestock feeding according to the availability of grass, promote a balance between grassland and livestock, convert grazing land to grassland, establish grassland protection areas, build artificial grassland areas in farming and pastoral farming areas, relocate population density beyond its environmental bearing capacity, etc., to strengthen the protection of grassland (Fan et al., 2010).

Policy implications of water resources management

Being a public resource, the Government of China should instigate a comprehensive plan of action for water resources on a regional scale to ensure its sustainable utilization and to minimize negative impacts to the ecological environment. Similarly, prior to the implementation of an ecological policy, the Government of China must also take into account environmental factors, such as global climate change, sand and dust storms, as well as other relevant factors (Chavas, 2017). Concerning water resource variation on the Inner Mongolian Plateau, we found clear evidence that showed that water scarcity issues in the overall region primarily result from precipitation (Wu et al., 2013). Changes in precipitation have yielded a positive feedback on changes in water resources (a 39.35% contribution rate), while changes in water resources were primarily a consequence of increased evapotranspiration (a −24.48% contribution rate). A potential likely reason behind this is the large number of artificial forests that have been planted in the region in recent decades, resulting in increased evapotranspiration. Furthermore, village density in the Mongolian Plateau has had a negative impact on changes in water resources, which suggests that anthropogenic activities (such as the development of stockbreeding, afforestation, etc.) in recent years have had a significant impact on the region.