Spatial Pattern Changes of Landscape in Ordos Plateau——Taking the Yinjinhuoluo County as an example

Yinjinhuoluo county were located in the Ordos Plateau, the Inner Mongolia Autonomous Region of China,in which the soil erosion of wind and water is very serious. Because the county's economy is a typical mixed economy of agriculture and animal husbandry, the ecologyical environment have the frangibility. In order to explore its characteristics of environment changes, the methods of landscape pattern analysis were selected. Under the land use classification, the land use map of 1984, 1991 and 1998 at the scale 1∶100000 were used in the research process. Based on the land use map, the patch types divided into 9 sorts, which were natural grassland, artificial grassland, wood land, shrub, farmland, river, lake, residential site and shift sand land. To quantify the landscape pattern, the following indexes in entire region and different patch type scale were selected. They are the number of patches (NUMP), class area (CA) , mean patch size (MPS) , patch size standard deviation (PSSD) , edge density (ED) , mean patch edge (MPE) , mean patch fractal dimension (MPFD) , area weighted mean patch fractal dimension (AWMPFD) , mean shape index (MSI) , Shannon's diversity index (SDI) , Shannon's evenness index (SEI) and dominance index (DI) . The results show that the landscape matrix did not changed in three years, which were natural grassland. Landscape pattern changes in different year were all occurred on this same matrix. In the entire landscape scale, the PSSD, ED, MPE, MSI, MPFD, AWMPFD and DI all were increased from 1984 to 1998, but the SDI and SEI were declined meanwhile. The changes reveal that the fragmentation degree is increasing, complexity and heterogeneous is declined. The landscape dynamics were controlled by few patch type, which were shrub and farmland in Yijinhuoluo county. In the patch type scale, The different patch type have the same changing trend. The complexity is increasing, but the extend is limited in a little ranges. Comparing the two scale change trend in different year, it shows that the changing trend is not in the same step.