10 years on, FAIR practices in ecoacoustics: State of the art review, challenges and opportunities for data stewardship

Academia currently finds itself at a crossroad, with advances in data science and artificial intelligence unlocking great research potential, whilst also creating obstacles for accountability, through for instance lack of transparency and poor-quality datasets. To tackle this, frameworks stipulating data handling are needed, especially when it comes to complex datasets. Here, we present a state of the art on the adoption of the FAIR (Findable, Accessible, Interoperable and Reusable) principles for data management in ecoacoustics, as an outcome of application of a mixed method: a literature review and semi-structured interviews. We synthesise challenges and opportunities concerning FAIR implementation. The first mention of the FAIR principles in ecoacoustic research took place in 2021, five years after they were adopted as best practice in research. An analysis on the implementation of the FAIR principles in the field showed that while there is more focus in the literature on elements of findability and interoperability, they are poorly implemented, partially due to limited resources facilitating this. Key barriers are inconsistent metadata practices, scarce machine-actionable standards, low findability of existing resources, and heterogeneous FAIR interpretations. Going forward, special attention should be placed on the publication of domain-specific standards and semantic types, the development of ecoacoustic ontologies and semantic data models, the implementation of access-control protocols that respect legal and community rights, and we recommend use of frameworks such as GO PLAN and the FAIR data maturity model for project- and resource-level FAIRification.