Forest Structure across the Geographic Range of the Giant Panda: Up-scaling from Plots to the Entire Region
The main goal of this project is to evaluate forest structure and spatial distribution across the entire geographic range of the endangered giant pandas (located in the provinces of Sichuan, Gansu and Shaanxi, China). Forest ecosystems provide the most suitable cover type for giant pandas. However, due to the rapid growth of human population in the region, coupled with the intensification of agriculture and urbanization, forest degradation and biodiversity loss are accelerated. As a response, the Chinese government has been developing a series of measures to enhance forest conservation and restoration of degraded forest ecosystems, but it is also pushing hard for a rapid economic development of the region. Therefore, it is critical to evaluate the current status of forest ecosystems in order to establish baseline conditions for assessing short- to long-term impacts of these contrasting conservation and development policies.
Our specific objectives are to:
- Assess the spatial distribution of forest types in the entire panda geographic region.
- Evaluate the structural characteristics (e.g., tree species composition, biomass) of the different forest types at plot scales.
- Develop techniques for up-scaling forest structure information, from plots, to nature reserves, to the entire geographic range, by means of synoptic view provided by satellite remote sensing at different spatial resolutions.
- Relate forest structure information with human disturbances at multiple scales, in order to evaluate the potential impacts of human activities on forest ecosystems across the region.
To achieve these objectives, we will take an integrated approach. First, we will establish field plots across the entire region to perform detailed analysis of structural characteristics, and to provide ground truth information for the analysis of remotely sensed data. Second, we will classify the forest areas of the region into functional types, by means of numerical classification of satellite remote sensing data. Third, we will analyze the spatial distribution of structural traits and tree species composition, within the context of the entire geographic range, and in smaller areas (e.g. nature reserves). Fourth, we will relate remote sensing data, acquired at the plot level, in order to device techniques for up-scaling from plots to the entire region. These techniques will be based on:
- Remote sensing of particular tree species using both their characteristics spectral signatures and spatial structures as identification traits in high spatial resolution data, available from commercial satellite systems (e.g., IKONOS, Quickbird);
- Remote sensing of tree species assemblages using phenological signatures as identification traits, in high temporal resolution multi-spectral data (e.g., MODIS).
Fifth, all these components will be incorporated into a GIS as data layers, and will be related with spatial data on human activities (e.g., infrastructure development, agricultural intensification), to establish direct and indirect relationships, in order to find underlying cause-effect processes.
This project builds nicely upon our long-term intensive research in Wolong Nature Reserve (200,000 ha, one of the largest reserves in China) as well as our preliminary work across the entire panda geographic range. Expected outcomes of the project include a solid understanding of the structure and distribution of forest ecosystems in the entire giant panda geographic range, a critical step for their conservation. All these components will advance the theory, enhance the methodology, and widen the application of the geospatial information technology to biodiversity research.
This project was supported by NASA.