Dr Yassine Kasmi1
1Thuenen Institute of Fisheries Ecology, , Germany
Biography:
Yassine Kasmi is a molecular ecologist and Bioinformatician at the Thünen Institute of Fisheries Ecology in Bremerhaven, Germany. His research develops qPCR- and metabarcoding-based eDNA tools for marine and freshwater fish monitoring, coupling molecular detection with statistical and computational modeling to advance non-invasive, quantitative biomass and biodiversity assessment.
Abstract:
Environmental DNA (eDNA) is a well-established, non-invasive tool for qualitative species detection, offering an alternative to invasive bottom trawl surveys in sensitive habitats such as Marine Protected Areas and Natura 2000 sites. Yet the transition from presence/absence detection to reliable, quantitative biomass estimation remains a central challenge, as eDNA concentrations are shaped by species-specific shedding, degradation, and hydrodynamic dispersal, making direct translation of eDNA copy numbers into stock assessment metrics far from straightforward.
This study draws on two qPCR-based eDNA investigations in the German North Sea and Baltic Sea to examine this qualitative-to-quantitative transition. A Gaussian process regression model was developed for Atlantic cod (Gadus morhua), and a general linear model for common dab (Limanda limanda) and European plaice (Pleuronectes platessa), to relate eDNA copy number to trawl-derived catch per unit effort (CPUE).
Both assays showed 100% qPCR specificity, confirming robust qualitative detection. Quantitatively, non-linear model-based approaches outperformed simple correlation: GPR yielded a significant eDNA–CPUE relationship for cod, while basic linear models on regional North Sea data alone were weak and non-significant. For flatfish, forecast models explained 80% and 60% of variance in plaice and dab biomass, respectively, with good transferability for plaice (RMSE 26% and MAE 34%).
eDNA-based modeling shows real promise for bridging qualitative and quantitative fisheries assessment, but predictive accuracy remains species- and region-dependent. Advancing shedding-rate correction, cross-regional model validation, and standardized detection/quantification limits will be key to establishing eDNA as a stand-alone quantitative tool for stock assessment and marine conservation management.