Articulating citizen science, semi-automatic identification and free web services for long-term acoustic monitoring: examples from France and UK

Monitoring biodiversity over large spatial and temporal scales is crucial for assessing the impact of global changes and environmental mitigation measures. Bats often have high conservation prioritisation owing to their trophic position, habitat associations and threat level, and many have dedicated management plans. However, poor knowledge of species' ecology, identification issues and surveying challenges mean that large-scale monitoring to produce required distribution and abundance information is less developed than for some other taxa. Exciting possibilities applicable to professional and citizen science are offered by new recording techniques and methods of semi-automated species recognition based on sound detection. Static detectors deployed to record bats throughout whole nights have been recommended for standardised acoustic monitoring but until recently cost and lack of software to support the analyses of such data has prohibited wide uptake. Such monitoring schemes have recently been deployed in both Britain and France allowing the fast and standardized collection of millions of bat records together with very interesting data on non-targeted taxa such as bush-crickets. Such data management led us to develop generic and open tools: (1) the Tadarida software toolbox providing a generic detection and classification of sound events, and (2) an open dedicated web portal (www.vigiechiro.herokuapp.com) to allow participants to manage and upload their data, then being processed trough Tadarida to get a quick feedback on the content of the data. We demonstrate how such data can accurately describe pronounced ecological patterns for numerous species at different scales: spatial variation in activity as a proxy for relative abundance, habitat selection and phenology of seasonal and nocturnal activity. If maintained in the long term, such schemes will also greatly improve estimates of species temporal trends and hence the assessment of conservation priorities. The feedback produced by these two monitoring schemes allows us the opportunity to provide recommendations for the sustainability of long-term acoustic monitoring of bats. These include a database that is adaptively managed to allow all raw data to be re-analysed every time automatic identification makes significant progress, while keeping the link with expert validation to ensure consistency in the semi-automated process. More importantly, there are real benefits of developing long-term acoustic monitoring within a collaborative framework. Specifically, (1) for collaboration among bat scientists for the collection of reference sound data, because diversity and quantity of the reference library remains a limiting factor for automatic identification, and (2) for work on bats to consider the wider acoustic monitoring of other species groups by working with other zoologists to share resources and costs