RGB and Multispectral UAV Mapping of Dugong Foraging Hotspots and Seagrass Beds in Thailand and Mozambique.

Unmanned aerial vehicles (UAVs) are widely used for mapping and monitoring coastal ecosystems due to their high accuracy and efficiency, providing surveys that are less costly and time-consuming compared to vessel-based methods. This study demonstrates the utility of using UAV imagery combined with ground surveys to evaluate the spatial distribution of dugong (Dugong dugon) foraging based on their feeding trails and associated intertidal seagrass beds in (1) Dugong Tower and (2) Juhoi, Libong Island, Thailand, and (3) Saco, Inhaca Island, Mozambique, as well as the temporal distribution of dugong feeding trails on Mook Island, Thailand. Ground survey results showed that small- and medium-sized seagrass species are the most preferred by dugongs. RGB images capture detailed plant characteristics, while NDVI images assess vegetation density, with higher values indicating denser vegetation. In denser areas (e.g., Juhoi), both images detected feeding trails, with RGB identifying distinct trails and NDVI highlighting contrasts. In sparse areas (e.g., Dugong Tower and Saco), NDVI provided clearer detection. However, UAVs may be limited by restricted flight endurance and sea state conditions, as well as by water level, turbidity, and sun glint. This study highlights the potential of drones to survey and monitor dugong populations indirectly, assisting coastal managers in assessing seagrass availability for dugongs and observing dugong behavior in their natural habitat, particularly in hotspot areas. [ABSTRACT FROM AUTHOR]