A novel framework for phytoplankton biomonitoring: Trait assignment of 23S rRNA sequences

• Direct trait assignment is an innovative way to analyse DNA metabarcoding data. • A higher number of ASVs can be assigned to traits than to genera or species. • Key components for successfull assignment are discussed. • Trait analysis provide good ecological interpretation of community structures. • Trait analysis of barcodes overcome gaps in reference libraries. Phytoplankton is a key biological group used to assess the ecological status of lakes in several legislative water management plans. Two cutting-edge approaches for community characterization are DNA metabarcoding and trait-based analyses. While the former provides a fast, cost-effective and high-throughput methodology for identifying communities, the latter reveals the structure of communities through bio-ecological traits. The main aim of this study was to combine these approaches to directly assign traits to amplicon sequence variants. To achieve this, we used the newly developed Phytool v3 reference database. Using an in silico test, we assessed the efficiency and reliability of our approach. We found: (1) that a greater number of sequences with better reliability can be assigned to traits than to genus or species level and (2) that traits are conserved in the phylogeny with varying extent. Then, we tested the usefulness of direct trait assignment on environmental samples from lakes. The test showed a greater number of successfully assigned sequences and a good ecological interpretation of community structures in the different environments. Furthermore, we identified three factors (completeness of the reference library, sequence similarity and the number of neighbours in the reference database) which, depending on the trait under consideration, interfere with the assignment success of our approach. While DNA metabarcoding data can be exploited in many ways depending on the objectives, our study showed that an innovative framework based on direct trait assignment of sequences could overcome gaps in reference databases and further improve our knowledge of phytoplankton community structure.