Introduction
Fish welfare is increasingly recognized as crucial for sustainable aquaculture, and minimizing stress is a key aspect of good welfare. New monitoring technologies are making it possible to observe fish behavior and well-being more closely than ever before. This collaborative study – involving experts from USB (Czech Republic) HCMR (Greece), SINTEF and NOFIMA (Norway), and CSIC (Spain) – explored how emerging tools like underwater video cameras, echosounders (sonar), and electronic biosensor tags can be used to monitor fish welfare. The aim was to identify common welfare indicators by detecting stress responses across multiple key European aquacultural fish species (Atlantic salmon, European seabass, and gilthead seabream) in different aquaculture systems (land-based tanks and net pens). By uniting efforts across institutions and species, the project provides a broad, high-level perspective on using technology to track fish well-being in real time.
Materials and methods
Experiments were carried out in both controlled tank environments and marine open net pens to reflect common aquaculture settings. At each site, the team employed a multi-technology approach to monitoring the fish. Underwater video cameras were used to record fish behavior and schooling patterns, while an acoustic echosounder was deployed in the large net pen setups to track how the fish school distributed themselves in the water column. In the tank trials, some fish were fitted with data-logging tags (including acceleration sensors) that recorded their movement and activity levels. All three species (salmon, seabass, and seabream) were subjected to a mild, short-term stress challenge designed to mimic routine farm events – most often a crowding of the fish into a smaller space, and in the case of the net pens, the additional disturbance of a net cleaning procedure. These stressors were applied in a harmonized way across the different experiments so that results could be compared between species and environments. The combination of video, sonar, and tagging at each experimental site (as available) allowed researchers to capture complementary data streams and observe how each technology detected changes in the fishes’ condition during and after the induced stress.
Results
The monitoring tools successfully detected behavioral changes in the fish when a stress event occurred, although the specific responses varied by species and context. In general, when the fish were crowded or otherwise disturbed, the video recordings revealed noticeable shifts in behaviour – for example, altered swimming speeds, tighter schooling behaviour, or changes in preferred depth or location in the tank/net pen. The echosounder data from the salmon net pen similarly showed changes in how the school was distributed, such as fish moving away from the area of the net pen being cleaned, indicating avoidance behaviour. The sensor tags on individual fish in the tank trials captured spikes in physical activity corresponding to the moment of stress, confirming that those fish were reacting even if it was not immediately obvious on camera. Notably, all three species (salmon, seabass, and seabream) exhibited some form of stress response that these technologies picked up, but the magnitude and form of the response were species- and environment-dependent. For instance, the salmon in a net pen did not show as dramatic a change during crowding as salmon in a confined tank or as other species did, though even in the net pen subtle indicators (like slight increases in swimming activity under certain conditions) were observed. These findings highlight that species may display stress in different ways in different settings, underlining the importance of interpreting welfare indicators in context. Importantly, by using multiple methods together, the study was able to detect these varied responses more reliably – one method often confirmed or supplemented the observations of another, giving a more complete picture of the fishes’ state.
Conclusion
In summary, this study demonstrates the value of a multi-tech approach to monitoring fish welfare in aquaculture. Modern technologies such as video cameras, echosounders, and active tags each contribute unique insights into fish condition: video provides direct observation of fish behaviors and interactions, sonar offers a broad view of group distribution and movement in large enclosures, and sensor tags yield information on individual fish activity or physiology. These methods are complementary, and an integrated monitoring system that combines them can overcome the limitations of any single tool. For example, acoustic monitoring works well in large, sometimes opaque water environments like net pens, while cameras and tags are especially useful in smaller or clearer systems – using them together ensures that signs of stress do not go undetected. The collaborative, cross-species results of this project suggest that building a cohesive monitoring framework (where camera, sonar, and tagging data are used in concert) will enhance our ability to identify stress responses early and accurately. Such an integrated system could enable farmers and researchers to respond more quickly to welfare issues, ultimately improving fish well-being and health. The study’s findings also emphasize that stress indicators are not one-size-fits-all; they depend on the species and context, so a flexible, multi-technology strategy is the most promising route to better fish welfare monitoring across diverse aquaculture conditions.
Acknowledgement: European Union’s Horizon 2020 research and innovation program under grant agreement No. 871108 (AQUAEXCEL 3.0)