Dr Tobias Frøslev1, Thomas Stjernegaard1, Dr Cecilie Svenningsen1
1Global Biodiversity Information Facility, Copenhagen, Denmark
Biography:
Extensive work with molecular taxonomic identification in relation to phylogenetics, systematics and biodiversity. Using DNA metabarcoding to assess and distinguish complex biological samples has been a main focus, as well as designing and developing reliable lab routines and bioinformatic workflows to distinguish biological signal from noise. As Scientific Officer in the GBIF Secretariat, a major focus is on improving and expanding the integration of DNA-derived biodiversity data, supported by data quality improvements and improved coverage and representation of DNA-detected organisms in GBIF.
Abstract:
Environmental DNA (eDNA) metabarcoding is becoming a mainstream tool for biodiversity monitoring, but a bottleneck limits its long-term value: taxonomic assignments in published eDNA datasets are frozen at the time of analysis and grow outdated as reference libraries improve. Compounding this, the reference libraries that enable eDNA identification remain fragmented, unstandardized, and lack any mechanism to credit their creators. The Global Biodiversity Information Facility (GBIF) is addressing both problems. We are building a continuous taxonomic reinterpretation pipeline that uses curated reference libraries to systematically refresh taxonomic assignments across the millions of eDNA occurrence records already mediated through GBIF. In parallel, we are prototyping a global reference library catalogue: a findable, accessible, interoperable and reusable registry where library creators can publish standardized reference databases, gain discoverability, and receive citation credit for their contribution to downstream usage. The reinterpretation engine has been benchmarked against existing GBIF eDNA records, demonstrating systematic taxonomic updating as reference data evolve. The catalogue’s metadata specification is defined and aligned with existing data standards, and a new download format already supports bioinformatic reuse of reference databases and occurrence data. Together, these components turn eDNA data mediated through GBIF from a static archive into living, continuously improving biodiversity evidence, while giving reference library builders lasting credit for their work. We invite maintainers and data providers to help shape this infrastructure.