Benthic diatom eDNA metabarcoding for ecological assessment of an urban river: A comparison with morphological method

Benthic diatoms have been widely utilized for assessing the ecological status of rivers, due to their ability to reflect the environmental changes from natural and anthropogenic disturbance. However, the recognized limitations of traditional morphological method for diatom identification hinder its widespread application in local department monitoring and assessment efforts. Recently, eDNA metabarcoding methods have emerged as a promising alternative for efficient diatom identification. To reveal the feasibility of eDNA metabarcoding methods in diatom-based evaluation, this study comprehensively compared them with the traditional morphological method. The results revealed that eDNA metabarcoding effectively reflected diatom composition at genus but not species level. It recognized 80.4 % of genera identified morphologically through rbcL identification and 66.7 % through 18S rDNA identification. Despite species-level discrepancies in taxonomic composition, eDNA metabarcoding accurately indicated alpha diversity of diatoms, while also provided better illustration of beta diversity. Additionally, comparisons were performed in taxonomy-based and trait-based indices. It was suggested that no significance was observed between identification methods in any index. Linear regression analyses further supported rbcL identification as a preferred option over 18S rDNA identification due to its stronger congruency with morphological method. In all, this study presented two key suggestions: 1) rbcL identification is preferable to 18S rDNA identification in diatom taxonomy, and 2) eDNA metabarcoding is proposed to be a promising technique to assess river ecological status.