A major scientific effort coordinated through the EU-funded AQUAEXCEL3.0 project has proposed Welfare Indicator (WI) toolboxes for use in European aquaculture research involving fish. Over 50 researchers have collaboratively produced a comprehensive review, soon to be published in Reviews in Aquaculture, outlining updated, fit-for-purpose WIs for five key species: Atlantic salmon, rainbow trout, European seabass, gilthead seabream, and common carp.

To get some key insights into the background of this innovative work, a series of key questions were posed to some of the authors.:

What is new about the Welfare Indicator (WI) toolbox?

Chris Noble, (Nofima): The toolbox builds upon some good foundational work that has been done by previous authors, and expands both the range of WIs in the toolbox and the depth of information on each WI for each of the relevant species. We also do our best to consider life-stage specific factors where these are available. Factors to consider when collecting WI data in research settings are also highlighted, in addition to outlining ways that digitalisation can aid the collection and analysis of this data.

Why do you have to customize the welfare toolboxes for each species and life stage?

Marie-Laure Bégout,(Ifremer): Over many years and several research projects, including research performed within AE3.0, it has become apparent that WIs often need to be life- stage and species specific since welfare needs and the species ecology and biology are very diverse. Nonetheless, general traits can be identified and then adapted for each species e.g. in relation to how the fish behaviourally respond to different husbandry and rearing practices (such as their feeding motivation, swimming, space utilization).

How do these new, refined indicators specifically support the ‘3Rs’ principle of Refinement?

Marie-Laure Bégout, (Ifremer): Welfare indicators are used to measure, monitor and document the fulfilment of various welfare needs. A robust indicator montoring programme, that includes indicators that address the majority of an animals welfare needs, can help an end user identify how and when a scientific procedure impacts upon their welfare. A stakeholder can then use this information to apply e.g. an intervention or mitigation action to limit or alleviate these impacts. Hence, new tools and approaches for measuring animal welfare are timely and the development and implementation digital tools including e.g. IoT are surely the future of early warning systems towards better welfare integration.

Could you briefly explain how the article addresses digitalisation and its impact on collecting and analysing welfare data? What’s the main takeaway for researchers in this area?

Petr Císař, (University of South Bohemia) and Santhosh K. Kumaran, (Nofima): The article highlights how digital technologies, such as sensors, imaging systems, and automated monitoring tools, are transforming data collection and the type and amount of information available for welfare monitoring. These technologies allow continuous, objective, and large-scale monitoring of welfare indicators that were previously difficult or impossible to assess manually. This process opens the way for better data-based decisions and the real-time/online monitoring of welfare indicators, which has utility in terms of welfare documentation, as well as helping shape interventions to mitigate against welfare risks. The key message is that digitalisation is both an enabler and a challenge. It allows more precise, scalable, and dynamic welfare monitoring. However, researchers require interdisciplinary collaboration (linking animal science, data science, and engineering technologies) to accurately interpret complex data and ensure meaningful insights into animal welfare.

Further, most technologies require a huge amount of data annotated by experts to provide trustworthy results, which is a very challenging task. However, by building generalised models, we would need less effort on annotation related tasks. Automated systems can also introduce bias if training data is not representative, making continuous validation and calibration essential. Digitalisation also brings challenges like data ownership and privacy (especially when video involves people). Implementing these tools can also be costly and complex, so solutions need to be economically viable. Digital tools could therefore complement, not replace, expert knowledge, and successful welfare research in the digital era depends on integrating technological innovation with biological understanding. After resolving these issues, the potential impact on the aquaculture industry can be very significant.

As part of the AQUAEXCEL3.0 consortium, what was the most valuable aspect of this collaborative process, and how will this joint effort create a lasting impact beyond the project’s lifetime?

Joint answer from Chris, Marie-Laure, Santhosh and Petr: For us, a really valuable aspect to this work was bringing together a wide range of different backgrounds and specialities into the indicator toolbox and then applying these perspectives to each species/life stage and also in relation to precision monitoring. The work contributes to improving our welfare knowledge on each species and this knowledge and the proposed toolboxes can then be used to help design rearing systems, steer husbandry operations and document how each experiment impacts upon the health and the welfare of the fish.

For the European research community and the wider aquaculture industry, we hope this work can be a valuable resource regarding welfare documentation. It represents a collective commitment—backed by the expertise of over 50 scientists—to provide updated information on how to document the welfare of research animals.

The full review will soon be available online via its Digital Object Identifier (DOI): DOI:10.1111/raq.70109.