Mammides Christos, Ieronymidou Christina, Papadopoulos Harris
Nature Conservation Unit, Frederick University, Nicosia, 1036, Cyprus.
Southeast Asia Biodiversity Research Institute, Chinese Academy of Sciences & Center for Integrative Conservation, Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences, Mengla, Yunnan, 666303, China.
Sci Data. 2025 Mar 19;12(1):461. doi: 10.1038/s41597-025-04807-1.
There is growing interest in using novel technologies for large-scale biodiversity monitoring. Passive acoustic monitoring (PAM) represents a promising approach for surveying vocalizing animals. However, further development of PAM methods is needed to improve their accuracy. The availability of extensive ecoacoustic datasets from biodiverse areas can facilitate this development. In this study, we present a large ecoacoustic dataset (1.58 TB) collected at sixty-one study sites on the island of Cyprus between March and May 2023. The dataset comprises >313,000 audio files, representing over 5,200 hours of recordings. It can be used for a range of applications, such as developing and refining species identification algorithms, acoustic indices, and protocols for processing acoustic data to exclude non-focal sounds, e.g., those produced by human activities. It can also be used to explore fundamental ecological questions. To facilitate its use, the complete dataset has been made available on the Hugging Face repository and the ARBIMON platform, operated by Rainforest Connection, which offers a range of free tools for ecoacoustic analyses.
人们对使用新技术进行大规模生物多样性监测的兴趣与日俱增。被动声学监测（PAM）是一种用于调查发声动物的有前途的方法。然而，需要进一步发展PAM方法以提高其准确性。来自生物多样性地区的大量生态声学数据集的可用性可以促进这一发展。在本研究中，我们展示了一个大型生态声学数据集（1.58 TB），该数据集于2023年3月至5月在塞浦路斯岛的61个研究地点收集。该数据集包含超过31.3万个音频文件，代表了超过5200小时的录音。它可用于一系列应用，例如开发和完善物种识别算法、声学指数以及处理声学数据以排除非目标声音（例如由人类活动产生的声音）的协议。它还可用于探索基本的生态问题。为便于使用，完整的数据集已在Hugging Face存储库和由雨林连接运营的ARBIMON平台上提供，该平台提供了一系列用于生态声学分析的免费工具。
