Applicability of DNA metabarcoding approach in the bioassessment of Portuguese rivers using diatoms

Benthic diatom communities are used in the ecological assessment of Portuguese rivers through the calculation of an autecological index, the IPS (Indice de Polluosensibilite Specifique), officially adopted for Portugal. This index requires a high level of taxonomic expertise for morphological identification of individuals at species level. Advances in genomics, such as the DNA metabarcoding combined with high-throughput sequencing techniques offer a promising alternative to this classical approach, decreasing expertise requirement and therefore reducing time and costs. The aim of this study was to test the potential of DNA metabarcoding of diatoms in the bioassessment of Portuguese rivers, by comparing the classifications of the IPS obtained with morphological and molecular approaches. A total of 88 samples were collected in four river types and a quality gradient in central Portugal in the spring of 2017, following the Water Framework Directive (WFD) standards. The morphological approach comprised taxonomic identification of at least 400 valves, under the light microscope. Molecular approach included eDNA extraction followed by PCR amplification and DNA high-throughput sequencing (Illumina MiSeq) using a diatom-specific 312 bp rbcL DNA barcode. Sequences were analysed with Mothur software, producing Operational Taxonomic Units (OTUs) that were taxonomically assigned using the R-Syst::diatom reference library. We also tested the effect of a correction factor (CF) based on biovolume for the quantification of molecular data. When comparing the number of species identified, morphological, molecular without and with CF inventories revealed 306, 125 and 111 species, respectively. Molecular data showed 67% of unassigned OTUs. These results were mainly due to the incompleteness of the reference library, which currently represents the major difficulty in taxonomic assignment of DNA sequences in metabarcoding. Regarding IPS values, the results indicated a good correlation between morphological and molecular methods, particularly when applying the CF. Furthermore, Principal Coordinates analysis plot based on species abundances revealed a quality gradient in all 3 methodologies (morphological, molecular without and with CF). These results support the hypothesis that DNA metabarcoding is a valid approach for ecological quality assessment of Portuguese rivers. Yet, there is still work to be done on this new methodology to enable a smooth transition between the traditional and this new approach without misspend accumulated knowledge from the last decades on water quality assessment.