University of Twente Student Theses


Reassessing Giant Panda Habitat with Satellite-derived Bamboo Information: A Case Study in the Qinling Mountains, China

Sun, Yiwen (2011) Reassessing Giant Panda Habitat with Satellite-derived Bamboo Information: A Case Study in the Qinling Mountains, China.

[img] PDF
Abstract:Estimating and mapping suitable habitat play a critical role in endangered species conservation planning and policy. Serving as the essential food source of giant pandas, bamboo is the most important ecological variable in giant panda habitat assessment. However, lack or inadequate information about understory bamboo distribution in previous studies has led to the variety in both quantity and quality of panda habitat. In this study, the understory bamboo was mapped using Maxent model based on giant panda occurrence data and multi-temporal MODIS EVI data. By incorporating this satellite-derived bamboo information, together with topography and human factors, the suitability of giant panda habitats in the Qinling Mountains were reassessed. Consequently, the conservation status of the current nature reserve network for giant pandas in the Qinling Mountains was also evaluated. The study results indicated that the panda occurrence data may be used as a surrogate for bamboo distribution modeling at a spatial resolution of 250m with an accuracy of kappa 0.74 and AUC 0.92. The study also showed that deficiency of bamboo information and human disturbance factor may bring about a huge overestimation of the total suitable panda habitat as well as a serious underestimation of the degree of habitat fragmentation. The sharp drop in habitat area with bamboo information indicated overestimations of more than 70% and 80% in suitable habitat and marginally suitable habitat respectively. Human disturbances further led to a reduction of 33% in suitable habitat and a decrease of 63% in marginally suitable habitat, as well as more severe habitat fragmentation. The reassessed giant panda habitat in the Qinling Mountains covers a total area of 1808 km2, which is much less than the area of 3475 km2 that estimated from the third national panda survey. About 54% of the habitat area consisting of large patches with good quality is under protection of the current panda nature reserve network, which is lower than the expected number of 72%. The study suggests that it is necessary to incorporate more accurate bamboo distribution information that derived from remotely sensed data into large-scale giant panda habitat research and management and to avoid overestimation of habitat. Moreover, the protective efficiency of panda habitats varies among different nature reserves; while some suitable habitats outside the nature reserves need further investigation for habitat expansion and linkage. All in all, this study facilitates understory bamboo mapping and has important implications for the long-term and sustainable development of giant panda conservation.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page