High-resolution remote sensing for biodiversity studies

The Atlantic Forest, now only 7.5% of its original size, has the largest continuous areas of remaining forest in the state of S o Paulo, where palm trees (Arecaceae) represent an important component of its biodiversity. The typical architecture of palm tree canopies makes them more easily identifiable by remote sensing than other trees in the forest canopy. In this context, the purpose of this research was to evaluate the potential of using high spatial resolution remote sensing images to identify palm trees in the region of Ubatuba, on the northern coast of S o Paulo. A QuickBird image (0.6 m spatial resolution) from 2007 was used for the preliminary identification of species by visual interpretation, serving as a basis for both field planning and analysis for digital processing. This approach made it possible to define which palm tree species are identifiable in the high-resolution satellite image, defining the minimum mappable size of canopy palm tree crowns and which morphological attributes best describe each species and aid in identification by remote sensing.