iQQTV

Scientists at iQQTV in Oban have trained an AI taxonomist as part of a new method that will revolutionise how seabed health below fish farms is monitored.

The software, known as eDNA-2-IQI, has now been formally adopted by the Scottish Environment Protection Agency (SEPA), the aquaculture industry regulator in Scotland, to assist in its statutory compliance assessment after proving its effectiveness in linking bacterial organisms in seabed samples to corresponding health indicators.

Fish waste and uneaten feed can settle on the seabed and affect the organisms living there. Understanding those changes helps SEPA assess whether operators are meeting the environmental standards required by their permits and whether further investigation or intervention is needed.

iQQTV, a partner of UHI, led the research project that devised the method, which uses artificial intelligence and environmental DNA (eDNA) to quickly check for bacterial diversity in the seabed. The type of organisms present in the sediment is an indicator of the health of that environment.

This means aquaculture sites and the regulator Scottish Environment Protection Agency (SEPA), can sample farm sites and get results within weeks. The current method, which requires larger samples of sediment, the use of toxic chemicals and thorough examination by taxonomists, can take up to three months.

As well as drastically reducing the long turnaround time to get samples analysed, the new method also addresses concerns around the limited number of taxonomists currently available for such work.

iQQTV Associate Prof. Tom Wilding, who led the research effort behind the new sampling method, explained: “This will revolutionise the way that fish farms are sampled and monitored, updating a practice that has been standard for half a century.

“eDNA-2-IQI uses a combination of machine learning and metabarcoding to tell us which bacteria are present in the sediment. A bit like a barcode on a supermarket item, we can scan the DNA sequence to see what’s there.

“Using a random forest, which is a form of artificial intelligence, we have trained the model on 950 samples, so it knows which bacteria are most useful in predicting the health status of sediment.

“This is far quicker than the traditional method of sampling and removes a lot of the time pressure on taxonomists delivering more manual analysis.”

The breakthrough has come from research projects MeioMetBar and BactMetBar, supported by the Sustainable Aquaculture Innovation Centre and the Marine Directorate, and involving partners in industry and other research centres.

The new method is not only quicker, but also more cost-effective and allows farm managers to better adapt farming practice.

Mike Montague, Lead Specialist for Aquaculture at SEPA, said: “Our role as Scotland's environmental regulator is to make sure operators understand and meet the standards required to protect the environment.

“Monitoring gives us important evidence about what is happening on the seabed around marine fish farms. eDNA gives us the potential to understand what that evidence is telling us much sooner.

“That matters because where there is evidence of a potential problem, operators can respond and we can investigate and intervene sooner where required.

“The environmental standards haven't changed and neither has our regulatory role. What has changed is the science available to support that regulation.

“This shows what can be achieved when scientific, regulatory and industry expertise come together to solve a practical problem. We are taking innovative science and putting it to work, helping make environmental regulation more efficient and responsive while maintaining the protections Scotland's environment depends on.”

BactMetBar (Environmental DNA metabarcoding as an alternative to macrobenthic assessments in fish-farm compliance assessment) was a £2m research project to:

1. Refine, standardise and publish standard protocols in relation to eDNA extraction and amplification for Scottish benthic samples taken for eDNA-based analysis over a diverse range of habitats
2. Develop data-processing and RF algorithms that best predict the IQI from bacterial eDNA sequence data by accounting for spatial and temporal variability in the bacterial response to fish-farm associated change
3. Collate the statistical routines into a single, user-friendly ‘R-package’
4. Publicise the standard operating procedures (SOPs) and demonstrate the BactMetBar R-package and embed the innovative alternative approach to macrobenthic-based monitoring in the sector (knowledge exchange).

Partners involved in the project are: Salmon Scotland, Scottish Environment Protection Agency (SEPA), Scottish Association for Marine Science (University of the Highlands and Islands), Institute for Biodiversity and Freshwater Conservation (University of the Highlands and Islands), The Technical University of Kaiserslautern (TUK), Mowi Scotland Limited (Mowi), Scottish Sea Farms Limited (SSF).

The development of the eDNA method was informed by the following research paper -