Abstract: Bird vocalizations, as an important bioacoustic indicator for interspecific identification, population dynamics, and ecosystem assessment, play a crucial role in biodiversity monitoring. We conducted acoustic monitoring in Hainan Tropical Rainforest National Park from June 2021 to April 2022. We used manual recording and passive acoustic monitoring equipment to collect data, in combination with artificial intelligence recognition algorithms and manual review, to construct a localized vocalizations database containing 46 common bird species from 25 families and 9 orders. The dataset covers a total of 12,187 audio segments automatically identified and filtered by the AI model, as well as standardized vocal samples of 1–20 segments per species manually selected. This database provides basic data support for research on vocalizations of geographically specific populations of tropical rainforest birds, large-scale acoustic monitoring driven by artificial intelligence, and biodiversity assessment of national parks.

Key words: National Park of Hainan Tropical Rainforest, bird diversity, bird sound recognition, passive acoustic monitoring