Remote sensing analysis to forest changes of the biodiversity hotspots in Southwest China

The Mountains of Southwest China is one of the biology diversity hotspots in the world with more than 12,000 species of higher plants, and the habitat diversity is the foundation of the biodiversity. This study uses maximum likelihood classification (MLC) to compare the Landsat-5 images in 1990 and Landsat-7 images in 2000 in order to obtain the forest and degradation areas by ERDAS software. Taking the giant panda as the flagship species, we analyze the habitats' change including their size, shape and component, and describe the relationship among the three of habitat scale, fragment and time in order to define the habitat effect to animal survival ability. The classification accuracy is up to 84.1%. Most of change areas centralize in Sichuan Province and South Tibet and a few areas in Qinghai, Yunnan and Guizhou Province. Most forest degradations are near the riverside, the roads, the uptowns and other human activity areas. In conclusion, it is an essential strategy by protecting the natural habitat and decreasing the human breakages to restore the nature. Remote sensing technology, in biodiversity study, is an important approach to monitor the forest sight and analyze the changing factors.