GIS-based Southeast Regional Ethnic Villages Distribution And Land Use-related Research

Based on DEM data and remote sensing maps of Qiandongnan Prefecture in1993,1999, and2009, this study analyzed the spatial distribution of ethnic group villages, the land use cover change and land use change in spatial gradients and the relationship between ethnic group villages and land use cover.The results are as follows,1. The elevation of Qiandongnan ethnic group villages range from244m to1256m, and mainly distributed on600-1000m. As to slope, the range is1.14-40.47°, and primarily on the gradient of7-15°. For the monsoon, ethnic group villages mainly distribute on eastern slope and southern slope. Generally, ethnic group villages are far away from water source, whose distance ranges from2m-1.37km, and most are100-399m. Besides, it is not convenient to reach the roads from ethnic group villages, usually, the distance to highways is furthest, the nearest is to county roads. In aspect of dot pattern spatial analysis, the distribution of ethnic group village on aspect and distance to water source are random, the distribution on elevation and slope are highs cluster, and the distribution on distance to roads are lows cluster.2. We defined the elevation gradients into six grades, and analyze the land use features in proportion, transformation matrixes, land use degree and their changes, and landscape metrics. The result indicated that farmland was the main land use type in lower elevation, while in higher elevation, forest land was the main land use type; with the elevation rising, the spatial variation of land use degree declined, and every spatial variation of land use degree augmented from1993to2009. From the standpoint of landscape, from1993to2009, in the area of0-1200m, the patch density declined drastically and the spatial distribution became more scattered; in the area of1200-2000m, the land use became more complex, its fragmentation increased and spatial distribution became more aggregated. In the range of0-800m and1600-2000m, it experienced a fluctuant change of area shape; while the range of800-1600m and>2000m, the area shape became order.3. We defined the slope gradients into four grades, and analyze the land use features in proportion, transformation matrixes, land use degree and their changes. The results showed that all the land use types can be found at every gradient. With the slope degree increasing, the area of forest gradually augmented, while generally the area of the other land use types, farmland, grassland and building, declined. Moreover, a mass of farmland were shifted from other land use types on slope degree of0°-25°, while a quantity of forest were transformed form the others on slope degree of>25°from1993to2009. Generally, large area of natural land use types were transformed to man-made land use types, for example, forest, grassland and water transformed to farmland and building. In terms of land use degree and changes, it cut down when slope degree increased and indicated that land use condition were all in development in all study periods. Findings from metrics analysis revealed PD values, SHDI values and CONTA values had similar variation trend that is, it decreased when slope degree increased. The values of AWMSI became larger when slope degree increased within0°-25°, while it reduced from the third to the fourth gradient. But there was no obvious variation trend on LPI value.4. Forest, farmland and grassland were the main land use type on buffers whose center were ethnic group villages in Qiandongnan prefeture, and the area of water body and built-up were rare. At the same time, the proportion of forest, farmland and grassland between each county, each elevation grade and each slope grade were so different. With the radius increased, the area of forest and grassland decreased, but the area of farmland and built-up increased, also, the changes on forest and farmland became shrinking. From the standpoint of time scale, on the counties, the area of forest and grassland still declined, and the area of farmland and built-up still increased. But there were no distinct variation trend on elevation grades and slope grades.