2.1 Study area

The Mongolian Plateau is located deep in the belly of the Eurasian continent, in the northeast of Asia. It ranks seventh among global plateaus, and is a continental plateau with an average altitude of about 1500 m and a maximum altitude of about 4355 m. The specific location is 87.64°E–126.10°E longitude, and 37.20°N–53.07°N latitude (Wang et al., 2023; Zhang et al., 2023b). The countries that include parts of the Plateau are Mongolia, China, the Republic of Tuva, the Republic of Buryatia and the Trans-Baikal Krai in the south of Russia, with Mongolia and China's Inner Mongolia Autonomous Region being the main ones (Fig. 1). For the convenience of this study, the study area is defined as Mongolia and China's Inner Mongolia Autonomous Region.

2.2 Overview of the natural environment

The Mongolian Plateau has a temperate continental climate, with average annual precipitation of about 200 mm, although some mountainous and hilly areas are at 400 mm, and the spatial and temporal distribution of the precipitation is uneven. The warm and humid ocean currents and the Mongolian high pressure system lead to a transition from arid to semi-arid from west to east, with the precipitation increasing from west to east and from south to north. The precipitation is mostly concentrated from June to September. This region has cold and long winters, short springs and autumns, high temperatures in summer, large temperature differences between day and night, and abundant sunshine, often accompanied by sudden weather changes (Li, 2019).

Fig. 1

Main study areas, topography and provincial administrations of the Mongolian plateau

The terrain is undulating, gradually decreasing from west to east, and there are many different types, including mountains, hills, plateaus, and plains. The northwestern part of the country is relatively high, the climate is relatively humid, the terrain is mainly mountainous, and the vegetation types are mostly mountain coniferous forests, typical grasslands and meadow grasslands (Bao et al., 2018). The climate in the southwest is arid, the terrain is mainly plains and hills, and the vegetation types are mostly desert steppe and desert steppe. The central part is a mainly semi-arid area, dominated by plains, terraces and hills, and the vegetation type is typical grassland. The eastern part is a humid and semi-humid zone, with mountainous and hilly terrain and various vegetation types, including broad-leaved forests, coniferous forests, shrubs, meadows and grasslands. The northeast is the foothills of the Great Khing'an Mountains, with coniferous and broad-leaved forests. The southern part is semi-arid, with mainly plains, hills and some sandy lands, and the vegetation types include desert steppe, shrub and farmland (Yao et al., 2011).

2.3 Data sources

The imagery data used in this study was obtained from the Landsat series of images, which were downloaded from the USGS website ( https://earthexplorer.usgs.gov/) of the United States Geological Survey, including Landsat-8OLI, Landsat-7ETM and Landsat5-TM. The Landsat series of land satellites launched by NASA (National Aeronautics and Space Administration) has a history of more than 40 years.

In this study, we took lakes with an area of more than 1 km² in the Mongolian Plateau from 2000 to 2020 as the research object to study their dynamic evolution and driving forces from 2000 to 2020. The timing of vegetation phenology, that is, the growing season, was taken into account, and the images were screened for those with cloud cover less than 15% from 2000 to June to September 2020. They were sampled for 10-year intervals of 2000, 2010 and 2020. There were 198 images in the study area, of which 101 images with lakes in the Mongolian Plateau were taken as the reference for each year, so 303 corresponding images for 2000, 2010 and 2020 were downloaded. Spatial and temporal analyses of each type of wetland in the lakeshore zone for the three phases were carried out, and the rate of change of each type was analyzed by calculating its kinematic magnitude, and the transformations between types were analyzed by calculating the transfer matrix.

The DEM (Digital Elevation Model) used ASTER GDEM data downloaded from the Geospatial Data Cloud ( https://www.gscloud.cn/), with a resolution of 30 m. This dataset is based on spaceborne emission and reflection radiometer data, and is widely used in land use, geographic planning, geological exploration and other fields (Luo et al.,2020). The vector data were obtained as a vector file for the administrative region of Mongolia and the administrative region of the Inner Mongolia Autonomous Region, which came from the Population Data Open Access and Application Organization ( https://hub.worldpop.org/) of the University of Southampton, UK.

2.4 Meteorological data

Meteorological data include annual precipitation, average annual temperature, and an aridity index that was calculated based on raw precipitation and evapotranspiration data, which is known as the Self-calibrating Palmer Drought Severity Index (scPDSI). The annual precipitation and annual average temperature data were obtained from the National Oceanic and Atmospheric Administration (NOAA) ( https://www.ncei.noaa.gov/), and the following data were from the measured meteorological datasets of global meteorological stations over the years, which were selected from the Inner Mongolia Autonomous Region and Mongolia for download. The scPDSI data were obtained from the Centre for Climate Research, University of East Anglia, UK (CRU4.01;  http://ww.cru.uea.ac.uk/data/). Compared with a single precipitation or temperature parameter, the drought index more accurately represents the combined effects of precipitation and evaporation in a region, so it more intuitively indicates the degree of water loss or storage in the region (Li, 2019). It is calculated based on a time series of precipitation and temperature, as well as fixed parameters related to the soil/surface characteristics at each location (Guo et al., 2022), to characterize the water supply and demand relationship in the area, and to describe the drought degree quantitatively and objectively. This calculation used CRU TS 4.05 and the data period was from 1901/01/01 to 01/01/2020. Combined with the time nodes in this study, the relevant data for 2000, 2010 and 2020 were selected for calculation and analysis.

2.5 Extraction of information for the wetlands in the lakeside zone

In 2000, a total of 564 lakes with an area of more than 1 km² were identified in the Mongolian Plateau. With a large number and a vast area, they reached 17105.20 km² in 2010, representing 675 lakes, 287 of which were in Mongolia. In 2020, the number of lakes decreased to 479, 243 of which were distributed in Mongolia.

Mongolian Plateau lakes occupy an important position among the global plateau lakes and play a pivotal role in regulating and stabilizing the global ecological environment (Fig. 2). In 2000, most of the lakes larger than 1 km² in the Mongolian Plateau were concentrated in the northwest and east, and only a small number of lakes were distributed in the central and southern regions.

Fig. 2

Distribution of lakes larger than 1 km² on the Mongolian Plateau in 2000

In this study, we consider the obvious differences in the topography of larger lakes and smaller lakes, that is, the larger lake basins have a wide area and large topographic fluctuations, and the topography of the lake basins changes significantly compared with the surrounding areas. In contrast, the topography of the smaller lakes does not change obviously, so the method for determining the range of the lakeside zone was divided into two approaches according to the size of the lake. In other words, the extraction method of the lakeside zone for larger lakes was determined by taking the altitude range of the bare land around the lake as the threshold, that is, the threshold extraction of DEM data was taken to determine the range of the lakeside zone. For example, Fig. 3 shows that the potential of the Urege Lake site changes significantly, gradually rising from the center to the perimeter. Combined with the 2000 remote sensing image as a benchmark, a suitable elevation threshold that better encompasses the wetland types was selected for the extraction of the lakeshore zone extent, such as for Lake Halwusu, Lake Kusugul, Lake Eyigun, Urakoue Highwall, Lake Hulun, and the Daihai Sea. On the other hand, small lakes were manually delineated and extracted according to the classification results, that is, the lakeshore zone was delineated with the wetland type at the outermost part of the lake. As shown in Fig. 4, according to the remote sensing images in 2000 and the classification results in 2000, the classification results are more accurate, and the wetland land type is obviously different from the surrounding area. Among the two types, the delineation of larger lakes and smaller lakes was based on the lake area in the base year, and the lakes with an area of more than 20 km² were considered larger lakes, while those with an area of less than 20 km² were considered smaller lakes.

Fig. 3

DEM threshold extraction of the Wuleige Lake shoreline

Fig. 4

Manual delineation and extraction of small lake shoreline zones

Various sets of series data were downloaded, including remote sensing data and driver data. The image data were Landsat series image data, and the DEM data were ASTER GDEM data. The vector data were the administrative boundary data for the Inner Mongolia Autonomous Region and Mongolia. The meteorological data were the Adaptive Palmer Drought Index (scPDSI). Combined with the high-definition images of Google Earth and field survey sample points, interpretation markers for the images were established, the sample dataset was constructed, and the optimal classification features were selected according to the sample set and classification effect. The classification features were NDVI, NDWI, a humidity variable (wetness) and a short-wave infrared band (SWIR2) in the spike cap transform. The threshold of each classification feature was determined according to the sample dataset. For the corresponding classification process, rules were established, the decision tree classification model was used for classification, and the accuracy was verified. The classification results were supplemented by manual correction.

3.1 Spatial changes in land types in the lakeshore wetland system

Based on the classification results of the lakeside zone in 2000, the overall situation of the lakeside zone in the Mongolian Plateau was assessed. The total area of the lakeside zone was 55216.47 km^2, including 27579.62 km² in Mongolia, and 27636.85 km² in the Inner Mongolia Autonomous Region. There were 564 lakes larger than 1 km² in the Mongolian Plateau in 2000, including 294 in the Inner Mongolia Autonomous Region and 270 in Mongolia. In the southern and eastern regions, the larger lakes were concentrated in the northern part of the Mongolian Plateau. Overall, the lakeside zone of the relatively flat lakes was larger, and the main types were meadow, herbaceous swamp and woody swamp. The area of the lakeside zone of the lakes with steeper terrain was smaller, and the main types were meadow, bare land, woodland, and others.

On the spatial scale, the topography of the Mongolian Plateau was undulating, diverse, and gradually decreasing from west to east. The climatic difference was significant, showing the characteristics of transition from arid to semi-arid from west to east, and precipitation increased from west to east and from south to north. Therefore, by analyzing the spatial variation characteristics of the changes in land type of the lakeside wetlands in different periods and the rate of lake area change in different study periods, the types of lakeside wetlands in the Mongolian Plateau were analyzed on the spatial scale.

From Fig. 5, the main types in 2000 were lakes, meadows, bare ground, and salt marshes with the largest areas, and woody swamps in the Inner Mongolia Autonomous Region were mainly concentrated in the western, central and northeastern regions, while those in Mongolia were predominantly in the central part of the country. Herbaceous swamps were more dispersed and smaller in area, mainly in the northwestern and eastern regions of the Mongolian Plateau. Meadows and bare ground were vast and widely distributed, both in the Inner Mongolia Autonomous Region and Mongolia. Saline land was mainly distributed in the northwestern, southern and eastern parts of the Mongolian Plateau, with the main distribution area in the Inner Mongolia Autonomous Region, indicating that wetland degradation and salinization were more serious in the Inner Mongolia Autonomous Region than in Mongolia. Cultivated land and land for construction were rare and sporadically distributed, reflecting low human interference and impact.

Fig. 5

Distribution maps of wetland land classes in the lakeside zone of the Mongolian Plateau in 2000, 2010 and 2020

In 2010, the main types of wetlands in the lakeshore zone of the Mongolian Plateau were lakes, woody marshes, meadows, bare land, saline and alkaline land. The overall area of lakes was significantly reduced, especially Orog Nuur in Mongolia and Chagan Nuur in the Inner Mongolia Autonomous Region, and the degradation of the lakes and wetlands in the Ulaqqaigobi was most obvious, as shown in Fig. 6. Among them, Orog Nuur showed a significant reduction in area due to the warm and dry climate and no other recharge water sources. Uragui Gobi had dried up as bare land and saline, and after reviewing the information, the upstream industrial interception of water was the main reason for its drying up. Chagan Nur West Lake had been degraded to saline and alkaline land, as the warm and dry climate led to a reduction in river recharge, coupled with the man-made closure of the East Lake lake water recharge (Liu et al., 2015).

Woody swamps increased significantly, mainly in the west-central, southern and eastern parts of the Mongolian Plateau. Herbaceous swamps had decreased, mainly in the southern, east-central and north-western parts of the plateau, with greater distribution in Inner Mongolia Autonomous Region in terms of administrative regions. The saline and alkaline land increased, mainly in the north-western, central and southeastern parts of the plateau, with the increase in Mongolia as the main source of the overall increase. Meadows and bare land were more widely distributed and did not change very much, along with the distribution of most of the lakes. The areas of cultivated land and construction land increased markedly, with a high concentration of land in the Inner Mongolia Autonomous Region, which shows that the Inner Mongolia Autonomous Region increased the degree of land use, while Mongolia is mountainous and Gobi, with a greater distribution of forested land.

In 2020, the main types of the Mongolian Plateau were dominated by lakes, meadows, bare ground, and woody swamps, in which the overall area of lakes had increased, such as the east side of Lake Hulun in the Inner Mongolia Autonomous Region, the new Dalai Lake, the Uliangsu Sea, and Uvsu Lake in Mongolia, and others, as shown in Fig. 6. After reviewing the information, those changes occurred due to the drought indicated by the climatic data and the decline in the water level of Lake Hulun, so that the new Dalai Lake and the Hulun Lake cut off the flow until Lake Hulun dried out. In 2009, the implementation of an ecological water replenishment project to restore the natural connection channel with the Hailaer River, and in 2014 and 2017, the implementation of the second phase of the river diversion project and downstream channel bank protection project, respectively, allowed the water level of Lake Hulun to rise year by year, the new Dalai Lake could restore its connectivity, and the new Dalai Lake experienced a re-injection of water. Since the 1990s, the natural recharge of the Wuliangsu Sea water has been constantly declining, along with urban wastewater and industrial wastewater. Since the 1990s, the natural water recharge of Wuliangsuhai has been decreasing, urban sewage and industrial wastewater discharge has increased, the area of the lake has decreased dramatically, and the eutrophication of the lake water body has been serious. Since 2012, the implementation of projects for ecological water replenishment, the implementation of the construction of a grid watercourse, and the lake substrate disposal and other ancillary projects have significantly improved the ecological environment through a comprehensive approach, and algae outbreaks have been reduced in large areas.

Fig. 6

Evolution process of wetlands in the lakeside zone of some individual lakes

The areas of saline and alkaline land decreased significantly, mainly in the north-western, southern and eastern parts of the Mongolian Plateau, and they were mostly converted into bare land and meadows. Herbaceous marshes increased, mainly in the north-western and eastern parts of the Mongolian Plateau, with the Inner Mongolia Autonomous Region dominating in terms of administrative areas. Woody marshes decreased, mainly in the north-western and southern parts of the Plateau, with Mongolia dominating in terms of administrative areas. Meadows increased and bare land decreased in the northwestern and eastern regions, with the administrative areas dominated by the Inner Mongolia Autonomous Region. Cultivated land and built-up areas continued to increase, although the area of forested land changed relatively little.

3.2 Spatial analysis of lake area changes

The distribution of lake area change rates was plotted based on the rates of change in area corresponding to each center-of-mass circle, as shown in Fig. 7. This map reflects the spatial distribution of the rate of change in the areas of different lakes in the Mongolian Plateau, where the legend is a green to red transition color band. The greener the color, the higher the rate of change in the area of lakes in the region, i.e., a higher proportion of lakes showing a reduction in area compared to the initial year; and the redder the color, means the opposite. As shown in the figure, the reduction in lake area between 2000 and 2010 was mainly concentrated in the western, east-central and southern parts of the Mongolian Plateau. In terms of administrative regions, lakes with a larger percentage of area reduction are more densely distributed in the Inner Mongolia Autonomous Region, and concentrated in the central and eastern parts of Mongolia, the western and central parts of Inner Mongolia Autonomous Region, and a small portion of the northeastern part of the country.

The lakes that increased in size are mainly located in the north-western, north-eastern and southern parts of the Mongolian plateau, with the main part of Mongolia concentrated in the north-western and eastern parts of the country. Lakes with a large percentage increase in area are located mainly in the western and central parts of Mongolia and the western and eastern parts of the Inner Mongolia Autonomous Region.

Fig. 7

Distribution of lake area changes from 2000 to 2020

The maps in Fig. 7 show that between 2010 and 2020, the reduction in lakes is mainly concentrated in the western, central and eastern regions of the Mongolian Plateau, and the reductions in the lakes with a large proportion of area are mainly located in the center and east. For the administrative regions, they are mainly distributed in the central, eastern and a small part of the western regions of Mongolia and in the central and eastern parts of the Inner Mongolia Autonomous Region (IMAR), while the distribution of lakes that decreased in size is dominated by the Inner Mongolia Autonomous Region (IMAR). The lakes that increased in area are mainly concentrated in the northwestern, eastern and southern regions of the Mongolian Plateau, with a small proportion in the central part of the country. In terms of administrative regions, the lakes that increased in area are mainly distributed in the Inner Mongolia Autonomous Region, except for the northeastern end of the region, and they are also partly distributed in the central, western and southern regions of Mongolia. Lakes with a large percentage increase in area are predominantly in the western, eastern and southern parts of the Mongolian Plateau.

The changes in lake area from 2000 to 2020 show that the area of lakes located in the central part of the Mongolian Plateau has been shrinking. In terms of administrative regions, the area of lakes in the central and eastern parts of the Inner Mongolia Autonomous Region has been shrinking, while the very few lakes in Mongolia are concentrated in the central part of the plateau. The lakes that have been increasing in area are mainly located in the western, southern and eastern parts of the Mongolian Plateau. In terms of administrative regions, they are mainly located in the central and western parts of Mongolia and the western and eastern parts of the Inner Mongolia Autonomous Region.

3.3 Dynamic analysis of wetlands in the lakeshore zone

The data in Table 1 show that, from 2000 to 2020 overall, marsh wetlands had the greatest changes in magnitude and rate, indicating that the wetland types are more fragile and sensitive, and susceptible to changes due to disturbances in the external environment. Among the marsh wetlands, woody marshes had greater levels and fluctuations in magnitude and rate, while herbaceous marshes had less; meadows and bare ground had a small magnitude and a low rate of change; and lake wetlands had the smallest magnitude of change and rate of change. The areas of herbaceous marsh, woody marsh and meadow were generally increasing, while the areas of lake, bare land and saline land were decreasing. These changes indicate that compared with 2000, the salinization of wetlands in the lakeshore belt in 2020 was reduced, the area covered by vegetation increased and the ecology was gradually improved. The area of woody marsh increased by nearly 4488.60 km² from 2000 to 2010, with the largest change in magnitude and fluctuation. The fluctuation of woody marsh was 21.74%, while the fluctuation of meadow and bare land was the smallest, at less than –0.1%. The wetland types had large changes, especially the woody marsh, which increased by nearly 217.39%. Except for saline and woody marshes, the areas of all other types had shrunk by different degrees, indicating that there had been drying and shrinking of some lakes, and degradation and salinization of the wetland types. The area of lakes increased from 2010 to 2020, and woody marshes had the smallest change in magnitude and rate, with a rate of only –1.46%, while herbaceous marshes had the largest rate and change in magnitude, at 180.43% and 18.04%, respectively, followed by saline marsh. The change in wetland type was small, the areas of saline and alkaline land decreased, the area of bare land continued to decrease, and the area of meadow increased, indicating that the area of vegetation cover increased, the degree of salinization decreased, and the wetland type had gradually improved.

Table 1

Table of dynamic change-related indicators

3.4 Wetland transfer matrix analysis of the lakeside zone

As shown in Table 2, among the natural landscape types of wetland ecosystems in the study area from 2000 to 2010, the area of lake wetlands decreased by more than 20%, and was mainly transformed into woody swamps, bare ground and saline soils. The area of marsh wetlands more than doubled, with herbaceous marshes decreasing by about one-third, and the types of transformation out of the area mainly consisted of woody marshes and a small portion of meadows and bare ground. The area of woody marshes increased significantly, and was mainly transformed from lakes and meadows. The area of meadows remained largely unchanged, with a slight increase and a reduction in the area of bare ground. There was an increase in saltmarsh, mainly converted from lakes, meadows and bare ground. In the anthropogenic landscape type, apart from woodland, the areas of arable land and built-up land both increased significantly, with the arable land mainly converted from meadows, lakes and wooded bogs, and the built-up land mainly converted from meadows and bare ground.

Table 2

Land use transfer matrix from 2000 to 2010 (km²⁾

Table 3 shows that among the natural landscape types from 2010 to 2020, the wetland area of the lakeshore zone in the study area increased by about 40000 km² in 2020. This increase was dominated by the increases in lakes and herbaceous marshes, with the area of lakes increasing by about one-fourth, and the conversion types were mainly from woody marshes, bare ground, saline soils, and a small portion of meadows. The herbaceous marsh area increased by more than a factor of two, mainly converted from woody marsh and meadow; while woody marsh decreased, mainly converted to lake and herbaceous marsh. The area of meadows increased, mainly converted from bare ground; while the area of bare ground decreased, and its conversion types were mainly meadows, lakes, and woody marshes. Saline soils decreased significantly, and were mainly converted to bare ground and a small portion of woody marshes. The area of anthropogenic landscape types increased, especially the areas of arable land and building land.

Table 3

Land use transfer matrix from 2010 to 2020 (km²⁾

3.5 scPDSI and lake area changes

The strip number of the Image Marker Symbol System WRS-2 was used as the horizontal coordinate, and the strip number was replaced by the serial number in the figure. Then the difference was measured by the Adaptive Palmer Drought Index for the two periods, i.e., the difference was calculated between the scPDSI at the last year of the study period and the initial year. The difference in the drought index and the rate of change in the lake area were then used as the vertical coordinates, respectively. For convenience in the study and visual display, the aridity index corresponding to the position of the center of mass was approximated as the average aridity index of the strip number, and the rate of change in the lake area was taken as the average rate of change of all lake areas within the strip. Finally, the line graphs from 2000 to 2010 and from 2010 to 2020 were plotted as shown in Fig. 8.

As the location moves, the peaks and valleys of the curve for the rate of change in lake area coincide well with the peaks and valleys of the curve for the difference of the aridity index (Fig. 8). The similarity of their fluctuations indicates an obvious correlation between the two: when the aridity index becomes larger, the rate of change in lake area is positive, i.e., the degree of aridity becomes wetter, and the overall lake area increases; and conversely, it is in a state of shrinking. Note that the corresponding lake area change rate is also large in the area with a large difference in aridity index and a large change in aridity degree, while the corresponding lake area change rate is also small in the area with small differences in the aridity index.

By comparing the difference in the aridity index between the two time periods, it is apparent that the rates of change in lake area and aridity index between 2000 and 2010 were small in each region, while the rates between 2010 and 2020 were generally large, reflecting the large degree of climate change between 2010 and 2020, and the fact that the climate trend is more arid than humid in these regions.

Fig. 8

Line plots of scPDSI and lake area rates of change from 2000 to 2010 and from 2010 to 2020

4.1 Results of the lakeshore belt wetland driving force model analysis, Inner Mongolia, 2000–2020

Amos software was used to construct a structural equation model for the driving force analysis of wetlands in the lakeshore zone of the Mongolian Plateau. The model was fitted and analyzed separately for Inner Mongolia Autonomous Region and Mongolia in different study periods. Sequentially, for making assumptions, the structural equation model of the composite system was constructed with wetland-meteorology-human activities as elements, the corresponding measurement indexes were selected and tested through the series of fitting indices, the model was corrected, and the results were analyzed. The SEM models of the Mongolian Plateau from 2000 to 2010 and from 2010 to 2020 are shown in Figs. 9–12.

Figure 9 shows that between 2000 and 2010 in Inner Mongolia, human activities had a significant negative impact on the change in wetland area in the lakeshore zone, with a path coefficient of –0.28, while meteorological factors had a positive and insignificant impact on the wetland area. Among the human activities, agricultural sowing area and livestock number had significant impacts on wetland area, and the highest correlation was –0.8 for agricultural sowing area. Agricultural sowing had a significant negative correlation with marsh wetland, and we assumed that the expansion of agricultural sowing area would encroach on wetland, thereby reducing the wetland area and destroying the wetland equilibrium. The number of livestock had a non-significant negative effect on the marsh wetland area. Meteorological factors had a positive but non-significant effect on the change in the wetland area in the lakeshore zone. Of the meteorological factors, the average annual temperature had a significant negative correlation with the marsh wetland, with a correlation index of –0.33, and precipitation had a positive effect on the lake.

Fig. 9

SEM model of the driving forces of lakeside wetlands in Inner Mongolia Autonomous Region from 2000 to 2010

4.2 Results of the lakeshore belt wetland driving force model analysis in Mongolia, 2000–2020

The model diagram in Fig. 10 shows that between 2010 and 2020 in the Inner Mongolia Autonomous Region, the human activities and wetland area had a complicated relationship. Human activities had a very significant impact on the wetland area of the lakeshore zone, which are negatively correlated, with a correlation index of –0.45. Among the human activities, the industrial output value shows a very significant positive impact, with a loading coefficient of 0.26. The area of crop sowing continues to have a negative influence, and the number of livestock had a positive effect on saline and alkaline land, indicating that an increase in the number of livestock enhances the intensity of grazing and exacerbates the shrinkage of the lake and salinization of the wetland. On the other hand, the relationship between meteorology and wetland area shows that meteorological factors contributed less to the impact on wetlands in the lakeshore zone. Annual precipitation showed a significant positive correlation with the areas of lakes and marsh wetlands, with path coefficients of 0.29 and 0.39, respectively, indicating that an increase in precipitation promotes increases in the areas of lakes and marsh wetlands. The industrial output value had a non-significant positive correlation with the average annual temperature, indicating that industrial development tends to contribute to the increase in temperature.

Fig. 10.

SEM model of the driving forces of lakeside wetlands in Inner Mongolia Autonomous Region from 2010 to 2020

Figure 11 shows that from 2000 to 2010, the influence of climatic factors on the area of lakeshore wetlands in Mongolia was very significant and positively correlated, with a path coefficient of 0.29, while human activities had only a small influence on the area of lakeshore wetlands. The drought index of meteorological factors had a negative effect on lakes and a positive effect on marsh wetlands, and presumably a wetter climate promotes an increase in the area of lakes and a subsequent reduction in the area of marsh wetlands, transforming them into lakes. This is also in line with the path of the negative effect of lakes on marsh wetlands in the figure. Human activities and meteorological factors show mutual influences, where the numbers of population and livestock showed a significant positive correlation with the average annual air temperature, with path coefficients of 0.31 and 0.42, respectively. This indicates that increases in population and livestock will make the air temperature tend to increase, leading to the greenhouse effect, which is consistent with the ecological background of global warming. scPDSI had a non-significant negative effect on the area of sown crops, indicating that with more drought, the area sown to crops tends to be less.

Fig. 11.

SEM model of the driving forces of Mongolian lakeside wetlands from 2000 to 2010

In Fig. 12, between the wetland area and human activities in Mongolia between 2010 and 2020, human activities had a significant positive effect on wetland area in the lakeshore zone, with a path index of 0.14. It was dominated by the population size and the value of industrial output, which are significantly and positively correlated with human activities, with loading coefficients of 0.35 and 0.45, respectively. The population size had a weak negative influence on saline area, and the negative influence of meteorological factors on wetland area was not significant.

Fig. 12.

SEM model of the driving forces of Mongolia's lakeside wetlands from 2010 to 2020

4.3 Impacts of natural factors and human activities on wetland changes in the lakeshore zone

Regarding the interactions between human activities and meteorological factors, the population has a significant positive correlation with the annual average temperature, with a path coefficient of 0.51, and the industrial output value has a non-significant positive correlation with the temperature, indicating that increases in population and industrial output value have great relationships with the increase in temperature.

From 2000 to 2010, human activities in the Inner Mongolia Autonomous Region had a significant negative impact on the area of wetlands in the lakeside zone, and meteorological factors had a positive but insignificant impact on the wetland area. Among the types, the sown area of crops had a significant negative impact on the area of swamp wetland, and the annual average temperature had a significant negative correlation with swamp wetland. From 2000 to 2010, the impact of human activities on the area of lakeside wetlands in the Inner Mongolia Autonomous Region was very significant and they were negatively correlated. The industrial output value of human activities showed a very significant positive correlation, and the annual precipitation had a significant positive correlation with the areas of lakes and swamp wetlands. From 2000 to 2010, the impact of climatic factors on the area of lakeside wetlands in Mongolia was very significant and positively correlated, while the impact of human activities on the area of lakeside wetlands was small, and the numbers of population and livestock had significant positive correlations with the annual average temperature. From 2010 to 2020, human activities in Mongolia had a significant positive impact on the wetland area in the lakeshore zone, but the negative impact of meteorological factors on wetland area was not significant.

In order to meet the ever-increasing needs of the population for survival and the demand for raw materials for other industries, the development of agriculture is an important and basic part of the national economy. While guarding the red line of cultivated land, it is also necessary to abide by the red line of urban development and ecological protection, so that all aspects of construction can be developed in an overall and coordinated manner. Attention should also be paid to the protection and role of the ecological environment. As the transition zone between lake ecosystems and terrestrial ecosystems, the lakeshore zone plays an important role in buffering and diluting pollution, conserving water sources, providing wildlife habitats, and maintaining ecological balance. To develop agriculture, it is necessary to increase the output of the existing cultivated land, and achieve efficient production in accordance with local conditions through rational and scientific farming, and the selection and breeding of improved varieties. In terms of water use, it is necessary to use water reasonably and fully, and to irrigate scientifically, increase yields, develop high-quality farmland, achieve green agriculture, and promote sustainable development.

With the increase in the population and the needs of economic development, the expansion of the scale and output value of animal husbandry has become an internal necessity. However, the endless increase in grazing intensity will lead to the gradual depletion of grassland resources, and the grazing will gradually expand to the fullest extent of the land, and then threaten the existence of wetlands. Therefore, the grazing intensity should be kept within the limit of the environmental carrying capacity, the efficiency of grass use should be improved, and seasonal free grazing in the restricted area should be used, so that the grassland has time to recover and rest in rotation. In addition, the storage structure should be adjusted according to the type of grassland to realize the sustainable use of grassland resources (Bao et al., 2010; Qin et al., 2017).