The Joint Danube Survey (JDS) is a regular transnational survey to monitor the quality of the Danube and its main tributaries, in accordance with the EU Water Framework Directive. During the JDS4 in 2019, conventional methods to monitor selected biological quality elements were complemented with DNA metabarcoding.
All together 72 phytobenthos samples were collected along the Danube and its major tributaries within the JDS4, using light microscopy and DNA metabarcoding amplifying a fragment of the rbcL marker gene. The (i) applicability of DNA metabarcoding to identify diatom communities compared to microscopy; (ii) diversity metrics between DNA metabarcoding and microscopy analysis and (iii) the usability of DNA metabarcoding for routine monitoring and assessment of the Danube under future JDS surveys were investigated.
Diatom communities resulting from light microscopy and DNA metabarcoding assessments share 26.5% of all taxa which corresponds to 64.3% when considering relative abundances. Discrepancies originate from biases both from metabarcoding, e.g. missing taxa from the reference library, and from microscopy, e.g. overlooking of hardly visible taxa. Microscopy detected more taxa in total but metabarcoding revealed a higher alpha diversity, detecting also very rare taxa in a given sample. Molecular and microscopy based Specific Pollution Sensitivity Index (IPS) values correlated significantly but differences were detected at several sites due to the differences in community composition and the overestimation of large taxa by metabarcoding. Although both methods showed a decreasing trend of IPS along the Danube, the metabarcoding based IPS covered a higher range of quality classes indicating lower values for downstream sites and the tributaries.
We suggest that metabarcoding provides a standardisable and efficient tool in biomonitoring, being more distinctive among quality classes than microscopy. Due to the high sequencing depth, it is able to detect a higher diversity on genetic level in a time- and cost-efficient manner that should be implemented in future quality assessment tools. We recommend its use in future biomonitoring surveys, for now, as a complementary method to conventional ones.
