| Grasslands,as the largest land type in the Xilingol League,are important economic and ecological resources with regulating and supporting functions.In recent years,due to excessive interference from human activities and climate change,Xilingol League grassland is seriously degraded,productivity declines,landscape fragmentation intensifies,and increased patchy grasslands have continued to worsen the already sensitive grassland ecosystem issues.Large-scale land use changes and the evolution of small-scale grassland patches are important scientific questions in the study of grassland ecosystems,and are also the basis for the rational allocation of grassland resources and the adjustment of ecological project implementation plans according to local conditions.This study takes the northern steppe of Xilingol League as the research object.Through remote sensing interpretation,field experiment and laboratory analysis.Comparing spatial patterns of land use and landscape ecological changes in the study area in 2010,2015 and 2021 on a large scale,exploration of spatial differentiation patterns and patch edge effects on grassland patches with severe landscape fragmentation at a small scale.The aim is to analyze in depth the land use changes and the evolution direction of patches in different grasslands in northern Xilinguole League,to provide a theoretical basis for the scientific management and sustainable use of grassland ecosystems in the northern part of Xilinguole League.The following research findings were mainly obtained:(1)The extraction of different features relies mainly on spectral features,and when adding topographic features,texture features and NDVI data of different time phases,the overall classification accuracy was 91.6%,90.2%and 93.69%,respectively.Grassland in East Urumqi showed an increasing trend during the study period,increasing by 593.15km~2 from 2010 to 2021,mainly from saline and sandy land conversion.Both Abaga Banner and Sunit Left Banner showed a decreasing trend in grassland area,decreasing by 972.49 km~2 and 112.66 km~2 respectively,mainly converted to sand and saline land.The NP and PD of patches increased in East Urumqi,Abaga Banner and Sunit Left Banner.The AI and CONTAG of grassland patches decreases,the degree of view fragmentation increases,the dominance of grassland gradually decreases,the landscape heterogeneity increases,and it is more influenced by the outside world.(3)Spatially positive correlations existed between landscape ecological risk values in East Ujumqin Banner,Abaga Banner and Sunit Left Banner during the study period,2015 is the period with the strongest spatial variability and the highest degree of variation in the ecological risk in the study area,and after 2015 the ecological risk in the study area has all decreased.The highest risk types in East Urumqi and Sunit Left Banner in the study period area are both medium-risk areas,but after 2015 both high-risk areas have an increasing trend,with 0.38%and 2.59%increase in both at the end of 2021,respectively.(4)The species richness of all three grassland types is inside the patch>edge of the patch,the community complexity is inside the patch>edge of the patch;the TN and SOM contents are inside the patch>outside the patch,and the distribution of other soil and community characteristics indicators are different.The distribution of AN content was consistent with the distribution of species diversity in either grassland,and the distribution of AP content and SOM content were consistent with species richness.The AN,AP,AK and TK of the three types of grasslands showed strong autocorrelation.From the autocorrelation distance,the autocorrelation distance of species diversity was greater than that of each indicator of soil,indicating that the change of soil is a slower process and the influence of vegetation on soil has a certain lag.(5)The biomass of three types of grasslands has been increasing with the increasing distance from the boundary(0~3.5m),and the maximum values of species evenness of all three types of grasslands are near the boundary,between-1~0m.The maximum values of the diversity indices of the three types of grasslands were also the same,between-1~0m;the maximum values of the AP of all three types of grasslands were between-0.5~1m,while the trends of the other abrupt indicators were different.The highest contribution of SOM was 76.8%and 77.6%in typical grassland and desert grassland,respectively,and the highest contribution of AP was 51.1%in meadow grassland,while other indicators showed different.However,in terms of distance,all three were essentially the same,with high AP,AN,and TN as well as AK and TK positively correlated with increased biomass.The accuracy of structural equation model construction for the three types of grasslands was typical grassland>desert grassland>meadow grassland in order,and SEM models for all three types of grasslands showed that biomass was the indicator most significantly affected by boundary effects,while AP and AN were the most significant indicators affecting plant species richness.The boundary peaks calculated with Euclidean distance,vegetation community characteristics and soil indicators did not overlap,and the boundary peaks of community indicators basically occurred between-1~0m,while the peaks of soil indicators all occurred between-0.5~1m. |