Improved fish stock assessment through the integration of environmental DNA

The challenge

Estimating how many fish are in the sea remains one of the most difficult challenges in fisheries assessment. For species such as anchovy, this becomes even more complex because of their high spatial and temporal variability. Although acoustic surveys combined with pelagic fishing are the standard method, this approach is invasive and catches do not always validate everything detected acoustically: nets cannot reach certain depths and some fish may avoid them. This can limit our ability to accurately characterise what is present in the water and has increased interest in complementary methods that can optimise sampling and reduce the impact of research surveys.

The solution

AZTI has been exploring environmental DNA (eDNA) for years as a non-invasive way of observing the ocean. The key challenge is no longer to demonstrate that the technology works, but to integrate it rigorously into fisheries assessment.

Within this framework, AZTI has established how acoustic and eDNA data can be combined to estimate the biomass of European anchovy in the Bay of Biscay, in a study published in the ICES Journal of Marine Science. The research draws on six years of data (2018–2023) from the JUVENA survey, combining acoustic transects with water samples analysed using digital PCR, which quantifies the number of DNA copies per litre.

To bring these two sources of information together, a joint Bayesian model was developed to integrate different signals and their associated uncertainty, using a common underlying true biomass as the latent variable.

The results

The combined estimates were consistent with existing knowledge, showing higher concentrations over the continental shelf. The eDNA analysis also suggested a somewhat broader distribution, providing additional support for species detectability and a quality-control element for acoustic indices. The approach has the potential to reduce uncertainty and identify possible false absences, improving our understanding of the stock’s spatial distribution.

The study is presented as a proof of concept. eDNA sampling is generally less regular and, in this case, data from several years had to be combined, limiting the scope for year-to-year analysis. To move towards more operational integration, since 2024 eDNA sampling during the JUVENA survey has been aligned with fixed hydrographic stations, paving the way for more detailed and efficient annual assessments.