TOWARDS A NEW GENERATION OF INDIVIDUAL BEHAVIOUR SENSORS FOR AQUACULTURE RESEARCH

Background

Aquaculture is increasingly driven by the need to better understand fish behaviour as a fundamental feature of health, welfare, and production performance. In cultured systems, behavioural responses emerge from the dynamic interplay between physiological state, environment and social interactions, giving rise to complex collective patterns that remain poorly resolved by conventional monitoring strategies. Current approaches to behavioural monitoring in aquaculture rely largely on population‑level observations or aggregate indicators such as bulk activity, spatial distribution or group movement metrics. These methods are suitable for detecting general trends, but they provide limited insight into individual variability and may fail to resolve subtle or heterogeneous behavioural responses. This limitation is particularly relevant in experimental and pilot‑scale contexts, where understanding individual responses is essential for assessing the effects of genetics, nutrition, environmental conditions or management strategies. Individual‑level behaviour monitoring addresses these limitations by enabling the direct acquisition of behavioural and physiological signals from single fish. By operating at the individual scale, such approaches allow variability, heterogeneity and response dynamics to be captured, rather than averaged out at the population level. This is especially relevant in aquaculture research, where experimental questions often target differential responses to controlled changes in genetics, feeds, environmental conditions or husbandry practices. The availability of miniaturised sensing technologies has facilitated the practical implementation of individual behaviour monitoring in experimental and pilot‑scale aquaculture research.

Sensor‑based individual behaviour monitoring

One way to address these monitoring needs at the individual level is through sensor‑based approaches. FishBIT is a miniaturized individual monitoring device designed for behavioural and welfare assessment in aquaculture research. The device is externally attached to the operculum using a minimally invasive fixation system, allowing stable placement without impairing respiration, feeding, or normal swimming behaviour. Its small size (15 x 6 x 6 mm) and low weight (below 1.8 g in air) allow fish to rapidly recover normal activity, making the device suitable for behavioural studies under experimental and controlled conditions. FishBIT integrates a triaxial accelerometer that captures opercular and body movements with high temporal resolution, enabling simultaneous recording of physical activity and respiratory responses. Opercular movements provide a reliable indicator of respiratory frequency, while body‑related acceleration signals reflect locomotor activity. Both signals are acquired continuously at the individual level, without reliance on visual observation. It has been successfully applied in major European aquaculture species, including gilthead sea bream, European sea bass, rainbow trout and Atlantic salmon. Previous studies have demonstrated the robustness of the device for monitoring behavioural responses to nutritional, environmental, and physiological challenges (Calduch-Giner et al., 2022). The updated version of FishBIT (Figure 1) incorporates advances in sensor design and system integration aimed at improving usability and performance. Battery capacity has been increased (3 V, 30 mAh), enabling extended recording periods of up to 200 hours. The hardware configuration also includes additional inertial sensing components, such as a gyroscope, expanding the range of motion‑related signals available at the individual level. Connectivity has been enhanced through the integration of Near Field Communication (NFC), allowing wireless device programming and data retrieval without physical connectors. In parallel, the use of a new ultra‑low‑power microcontroller and updates to the internal electronics architecture have increased computational capacity and data handling efficiency, supporting more flexible signal acquisition and on‑board data management.

Future Perspectives and Development

FishBIT has been established as a reliable individual monitoring tool for assessing activity, respiratory responses and behavioural changes in aquaculture, supporting applications in welfare assessment and experimental research. Building on these capabilities, the advances presented here enhance device autonomy, connectivity and computational capacity, and extend its applicability in experimental and pilot‑scale aquaculture studies. The incorporation of enhanced on‑board processing capacity facilitates a more effective use of the inertial sensing hardware of the device. In this context, current research is exploring the use of gyroscopic information to support the inference of individual movement trajectories. Such trajectory‑based descriptors could provide additional insight into spatial behaviour and relative positioning of fish within a group, opening new possibilities for the study of collective organisation under controlled conditions. In parallel, ongoing developments in data analysis software aim to facilitate the interpretation of complex behavioural datasets. Machine‑learning‑based approaches are being investigated as tools to assist in pattern recognition, data integration and decision support, facilitating the scalable use of individual behaviour monitoring in aquaculture research and precision management scenarios.

Figure 1. From sensor concept to application.

Acknowledgment

This work was supported by Spanish MCIU project BreamHOLOBIONT (PID2023-146990OB-I00) and Generalitat Valenciana (TOMACUA, CIAICO/2024/281).

References

Calduch‑Giner, J., Holhorea, P.G., Ferrer, M.A., Naya‑Català, F., Rosell‑Moll, E., Vega García, C., Prunet, P., Espmark, Å.M., Leguen, I., Kolarevic, J., Vega, A., Kerneis, T., Goardon, L., Afonso, J.M., Pérez‑Sánchez, J. (2022). Revising the impact and prospects of activity and ventilation rate bio‑loggers for tracking welfare and fish–environment interactions in salmonids and Mediterranean farmed fish. Frontiers in Marine Science, 9, 854888. https://doi.org/10.3389/fmars.2022.854888