遇见数据集

Annotated audio snippets of bird and anuran vocalizations from the rupestrian fields of Serra do Cipó National Park, Brazil

收藏
Zenodo2026-09-30 更新2026-10-01 收录
官方服务:

资源简介:

This dataset contains 407 annotated three-second audio snippets of vocalizations from four bird species (Asthenes luizae, Embernagra longicauda, Scytalopus iraiensis, and Zonotrichia capensis) and four anuran species (Boana cipoensis, Ischnocnema juipoca, Leptodactylus camaquara, and Leptodactylus cunicularius) recorded in Serra do Cipó National Park, Minas Gerais, Brazil. The dataset includes a compressed archive containing all audio snippets in WAV format and an accompanying metadata spreadsheet (.xlsx) describing the annotations and associated recording information. Recordings were collected in the Alto Palácio region of Serra do Cipó National Park (19°40′S, 43°30′W), located within the rupestrian fields (campos rupestres) ecosystem of the Espinhaço Range, state of Minas Gerais, Brazil. Passive acoustic monitoring was conducted between June 2024 and May 2025 using 16 AudioMoth autonomous recorders distributed across the study area. Recorders operated at 16-bit resolution following a duty cycle of one minute every ten minutes and one recording day every ten days, following the standardized soundscape monitoring protocol developed for protected areas by Campos et al. (2026). Audio files were processed in YAPAT (Yet Another Passive Acoustic Tool; Gouvêa et al. 2026), where recordings were segmented into three-second snippets and annotated through an iterative workflow combining reference recordings, acoustic similarity search, active learning, and expert validation. The deposited snippets correspond exclusively to the eight focal species listed above. Species identities were independently verified by taxonomic specialists. The metadata file contains the snippet identifiers, file names, taxonomic assignments, annotation labels, and other information required to link annotations to the corresponding audio files. Data collection was entirely non-invasive and relied exclusively on passive acoustic monitoring. No animals were captured, handled, or disturbed during data collection.

提供机构:
Zenodo
创建时间:
2026-09-29
二维码
社区交流群
二维码
科研交流群
商业服务