Comparing DNA metabarcoding with light microscopy to identify eukaryotic phytoplankton in the Baltic Sea, Kattegat and Skagerrak

Abstract Marine phytoplankton monitoring has long relied on microscopy, but DNA metabarcoding has recently emerged as a complementary approach. This study assessed the applicability of DNA metabarcoding of the 18S ribosomal RNA gene in marine monitoring and compared its results with conventional microscopy. We analyzed data from 232 surface water samples from 17 monitoring stations in the Baltic Sea, Kattegat, and Skagerrak. Metabarcoding detected more orders, genera, and species than microscopy, with a 43% overlap in the most common genera identified by both methods. Despite attempts to normalize sequence reads to spike-in DNA or DNA concentrations, the correlations between abundances derived from the two methods were weak, though varied considerably between taxonomic groups and geographical areas. Correlations were consistently stronger when using carbon and biovolume concentrations than cell abundances. Our results highlight the potential of metabarcoding to expand biodiversity assessments and advance our understanding of microbial biodiversity in marine ecosystems. As a complement to microscopy, it can enhance existing monitoring efforts. Future improvements in reference database completeness, adoption of long-read sequencing technologies, and better characterization of gene copy number variability per cell are needed to further extend the applicability of metabarcoding for quantitative analyses.