Abstract

Monitoring animal behaviour provides critical insights into species ecology and offers essential information for guiding management and conservation efforts. Among the various approaches used to study behaviour, bio‑logging —the use of animal-borne data recorders—has emerged as a valuable tool for observing animals in their natural habitats while minimising human disturbance. A major advancement in bio‑logging has been the integration of accelerometers, which enable high‑resolution analysis of movement and activity in free‑ranging animals. By identifying movement and postural patterns captured by accelerometers, it is possible to associate specific signal sequences with distinct behaviours. To avoid the labour-intensive task of manually labelling accelerometer data, these sequences can be automatically classified into clusters or predefined behavioural categories using machine-learning algorithms. The first ecological study applying machine learning to identify animal behaviours from accelerometer data was published in 2009. Since then, numerous studies have expanded this approach across a wide range of species, employing diverse methodological frameworks that can make it difficult to identify best practices. This literature review aims to provide a clear roadmap for practitioners and researchers, whether new or experienced, engaging in accelerometer-based behavioural identification using machine learning. We summarise the range of species and applications investigated, highlight key methodological frameworks, address common questions faced by new users, and provide practical guidance for implementation. Additionally, we identify knowledge gaps and outline future directions that the research community should aim to address to fully harness the potential of accelerometer-based behavioural identification in ecology. Based on 125 studies, we show that current practices largely rely on a general framework combined with species-specific adaptations, which limit methodological generalisation and explain the prevalence of species-focused methodological papers. However, 2024 marks a turning point, with a growing number of studies applying deep‑learning approaches that hold promise for improving model generalisation. Although the adoption of deep learning in ecology still lags behind its use in human and livestock behaviour recognition, leveraging advances from these domains and fostering interdisciplinary collaboration will be essential to accelerate progress. In particular, developments in real‑time monitoring offer strong potential to enhance conservation efforts, an important next step in bio-logging where machine learning can provide substantial benefits.