Abstract:The bamboo is the exclusive food of the wild giant pandas. Detection of the bamboo forest in giant panda habitat will help scientists further understand the spatial distribution pattern of giant pandas and their habitats. Moreover, it provides crucial scientific evidence for estimating habitat suitability, the level of habitat fragmentation as well as its ecological carrying capacity for pandas. However, it is a big challenge to map bamboo forests using direct remote sensing approach as most of them grow underneath the forest canopy. In this study, two dominant bamboo species, Bashania fargesii and Fargesia qinlingensis in Foping panda reserve were investigated. While remote sensing and GIS techniques were used to map spatial continuous environmental variables, for the first time, the overstory and understory light climate was introduced as a potential factor for explaining bamboo distribution. The model for predicting undestory bamboo density was developed based on statistical analyses between different bamboo species and its relevant environmental factors. Finally, the bamboo density map was produced with the support of GIS spatial analysis. The results suggest that satisfactory mapping accuracy of understory bamboo detection can be achieved by using an indirect remote sensing approach. The overall accuracy of both bamboo species was 78%

Detection of Understory Bamboo in Giant Panda Habitats Using an Indirect Remote Sensing Approach