Live monitoring of mussel farms in the Baltic Sea – to predict toxin production and minimize negative environmental impact

Mussel farming is in many ways a sustainable food production where nutrients are taken up from the water and converted into food which is then taken out of the water at harvest. This is especially beneficial in the eutrophic Baltic Sea and therefore it would be a success if this industry that uses filter-feeding organisms gives mussels a chance to grow and establish themselves. However, mussel farmers today spend a lot of time travelling back and forth to their farming sites to ensure that the equipment is working properly and to manually collect samples to both ensure good ecological status around the farms and low levels of phytoplankton toxins in the water. Our project aims to develop and implement an integrated monitoring system that combines satellite data, advanced sensors that measure pigments in phytoplankton as an indicator of toxicity and in relation to environmental variables. These systems are not available on the market for the marine environment today, similar systems are used to a lesser extent in lakes for drinking water in southern Sweden (Sydvatten) and here we can draw inspiration and exchange knowledge.

We will use both integrated and data-driven approaches combining satellite and sensor data, we will also take samples for microscopy monitoring of phytoplankton species and toxin levels in mussels to validate and match against real-time monitoring for optimized sampling. We will be able to follow developments in the water and at the bottom during the cultivations in real time and with increased precision using AI tools on the web-based platform that we will develop for the project and future use by the growers. We will also investigate the environmental impact to clarify how negative impacts from the cultivations can be minimized. This will be done through investigations of sediment and oxygen levels at the bottom, as well as phytoplankton communities around blooms and in comparison with reference areas.

With the remote monitoring systems available through satellites and linking the data provided by the probes using machine learning algorithms, it is possible to predict toxin production, the most effective time for harvesting and so on. Setting up this type of monitoring would save money for growers and for the government agencies that handle monitoring and analysis. These systems would also be helpful on the West Coast where many growers are located and where they have to close for periods each year. The data-driven optimization through the collection of samples and materials for monitoring allows us to specify and refine strategies for cultivation sites, improving efficiency and sustainability at a level that traditional methods cannot. The integrated data we collect combined with AI tools is innovative and will generate completely new angles on the issue of flowering around cultivations and provide new ways to predict environmental impact.

Based on the results of the study, growers will be able to optimize cultivation depth and location to minimize negative environmental impact, as well as streamline production and monitoring with minimized emissions from boats needed to regularly check the crops. This creates increased sustainability in the system. By comparing these alternatives with our solution, we see that our system not only combines the advantages of real-time monitoring and data analysis, but also uses satellite data to obtain a broader and more detailed picture of environmental conditions. This generates a unique and comprehensive solution available on the market today.