About the project

In thsi project we are working to verify the biological relevance of established welfare indicators that are suitable for automatic registration, in order to build the knowledge base for digital welfare registrations to be used as a tool for managing production biological parameters and outcomes.

• Background

  The project builds directly on and will implement results from the project team's previous and ongoing R&D work in the development of operational welfare indicators (WIs) for farmed salmon, automatic recording of WIs and production parameters using machine vision and artificial intelligence (AI) technology, as well as basic research on the relationship between stress coping ability, welfare and health status in fish.

  Over the past few decades, Norwegian salmon farming has established some of the world's most efficient production systems for fish, and is today characterised by innovative and highly technology-driven developments in production forms, management and efficiency. However, along with space constraints, intensive production can create sub-optimal production environments, where disease, parasites and stress occur. Close follow-up and monitoring of welfare is therefore an important part of operations in today's aquaculture industry. This work is typically based on internal and external scoring of fish welfare using established welfare indicators.

  However, animal-based welfare indicators such as physiological measures, appearance, external injuries, condition and behaviour usually only have an impact after a problem has arisen, and it takes a lot of experience-based knowledge to make production-related decisions. Sensors and systems for continuous and automated monitoring of farmed fish are currently being developed rapidly. In this field, however, there is a crucial lack of knowledge about how relevant such data actually is in relation to the salmon's biological condition, reliability, repeatability and fitness for purpose as indicators to fulfil the welfare needs of the fish in a particular farming system or farming routine. 

  This project therefore aims to verify the biological relevance of established welfare indicators that are suitable for automatic registration, in order to build the knowledge base for digital welfare registrations to be used as a tool to manage production biological parameters and outcomes. Neurobiological and molecular biological indicators of reduced stress coping ability will be related to digital welfare registrations, which in turn will be further verified through environmental registrations and production data.

  The results will be continuously communicated to industry actors and the administration through arenas established in this and parallel projects in the consortium, with a view to direct implementation of common limit values and frameworks that will help to maintain animal welfare at an acceptable level.

• Objectives

  Main objective

  To provide the necessary biological knowledge base for the implementation of automated health and welfare registrations from salmon in aquaculture.

  Sub-objectives (linked to work packages)

  1. To verify that selected indicators are relevant to health and welfare, defined as the fish's immune status and ability to cope with its environment. 2. to obtain automated and manually verified measurements of selected WIs from practical large-scale aquaculture, and analyse their relation to environmental and production data.
  2. To obtain practically implementable threshold values for when different indicators reflect the normal situation, cf. conditions requiring action for salmon in food fish production.
  3. To ensure effective communication and information, dissemination and administration of the project
• Expected utility value

  Fish health is a product of the interaction between the fish's own biology and the environment it lives in. Understanding good fish welfare is therefore not only important to ensure ethical and sustainable production, but also to promote disease prevention. Strong fish health and good fish welfare have been identified as a key competitive advantage for the Norwegian aquaculture industry going forward, both from an economic perspective and with regard to sustainability goals. The issue of animal welfare is also becoming increasingly important to consumers, both in Norway and internationally, and more and more detailed documentation is required on how welfare is understood and safeguarded by the individual company.

  The project background, composition of the consortium, and plan for implementation and dissemination are defined with a view to the ultimate goals of verifying the relationship between automatically recordable parameters and actual health/welfare status, as well as providing practically implementable threshold values for when different indicators reflect a normal sustainable situation, cf. conditions where extra measures are required.

  The approach is initially based on digitised visual monitoring of single indicators, and the project is not dimensioned for an in-depth evaluation of a larger number of potential VI at individual or group level. However, the basic biological principles and machine learning algorithms developed in the project are technology-neutral and can be applied to other indicators and other types of sensor data as they approach a satisfactory degree of technology maturity (TRL level). The direct benefit lies in the use of new technology that provides high data quality and increases the degree of automation in the collection, standardisation and sharing of data, as well as AI technology that tracks batches and individual fish and can provide advance warning of problems.

  The knowledge front will also be significantly advanced for PFF (precision fish farming) and help to improve the practice of the 3R principle (guiding principles) for more ethical use of animals in product testing and scientific research, i.e. reduce, refine and replace) in trialling and testing.

  In the final phase of the project, it is prioritised to effectively compile and disseminate the project's results. New knowledge will be communicated both interactively and creatively through specialised days for industry players, through graphic presentations or animations available in the media, as well as traditional dissemination via FHF and the consortium's own network, associated websites, scientific and academic publishing and reporting.

Participants

External participants

Erik Höglund, forsker, NIVA og Professor UiA

Mette Remen, forsker, SINTEF Ocean SA

Christian Schellewald, forsker, SINTEF Ocean SA

Harald Ian Muri, forsker, SINTEF Ocean SA

Tosca B.L Koers, research techniques and master student, NMBU

Espen Høgstedt, research techniques and master student, NTNU