Giant Panda Habitat Distribution across its Entire Geographic Range: A Preliminary Assessment
Andres
Vina, Michigan State University, vina@msu.edu
(Presenting)
Weihua
Xu, Chinese Academy of Sciences, xuweihua123@gmail.com
Yu
Li, North-West University, Shaanxi Province, China, liyu_chrislee@yahoo.com.cn
Zhiyun
Ouyang, Chinese Academy of Sciences, zyouyang@mail.rcees.ac.cn
Jiaguo
Qi, Michigan State University, qi@msu.edu
Jianguo (Jack)
Liu, Michigan State University, liuji@msu.edu
The world-famous endangered giant pandas (Ailuropoda melanoleuca) depend on forest overstory as shelter and understory bamboo as staple food. Although giant pandas had a wide geographic distribution in the past, they are currently restricted to five major mountainous regions in China. To understand the distribution of giant panda habitat across its entire geographic range, we have acquired relevant field and remotely sensed data. The spatial locations of panda evidences (feces, tracks, and eaten bamboo shoots) were recorded in the field using global positioning systems. They were utilized to develop presence/availability models by means of Ecological Niche Factor Analysis, using time series of different vegetation indices (obtained from MODIS) as predictor variables. We assessed the performance of the models created with each of the predictor data sets using two different validation procedures (Minimal Predicted Area and Prediction Success). In addition, a series of landscape metrics were calculated per mountain region in order to evaluate the degree of fragmentation of the habitat for the pandas. Preliminary analyses reveal that the habitat for the giant panda in its entire geographic range exhibits a high degree of fragmentation, particularly those located in the southern part of the geographic range. In addition, the temporal variability of vegetation indices exhibits a phenological characterization of the land surface that represents a suitable environmental predictor for giant panda habitat mapping. These results suggest that information contained in MODIS data has considerable potential for endangered wildlife species habitat mapping and management.