Handling taxonomic uncertainty

Species identification is a complex process with species typically described from a certain set of characters identified in a published species description and linked to a type specimen held within a scientific collection that be used for validation of species identification. Where taxa are very similar or a set of complex traits are required for correct identification, specific taxonomic expertise not possesed by data publishers may lead to a species misidentification. As users, you must have a clear understanding of how taxonomic determinations for your interest group are made:

If you think that there is a risk that species may be incorrectly identified, you can take a conservative approach to the data you use and only use those data linked to specimens in collections where taxonomic validation would be possible and eliminate other data sources. Another approach may be to use associated data, such as collector information, media, DNA sequences, etc., to validate the taxonomic determination.

Synonymy can arise when the same species has been described several times and a new name is given to the species each time it is described, or, when there is a change in the taxonomy of a species, for example, a species is moved from one genus to another. Only one species name can be accepted, and other names are what we call synonyms. These synonyms may still be in use to a lesser or greater extent, and you should be sure when getting data from GBIF to obtain data for the taxonomic name you need. GBIF’s taxonomic backbone differentiates between accepted scientific names and synonyms, and unique identifiers in the form of taxon keys. Species search allows for filtering for accepted names and synonyms, and taxon keys can be used for programmatic searches of GBIF.

Scientific names can be messy. If you are accessing GBIF-mediated data programatically as opposed to via the website, taxon keys provide an effective way for defining searches based on taxonomy. Taxon keys are issued at the species, genus family, order, phylum and kingdom level. Unique identifiers are issued to accepted names with synonyms of those accepted names issued the same identifier. So, it may make sense to sort out the species by their unique taxon keys provided during the indexation of the dataset by GIBF.

There may be instances where the scientific name does not match any name in the GBIF backbone, perhaps because the species is newly described or is not within a checklist used by GBIF to construct its backbone. These names are flagged with the TAXON_MATCH_HIGHERRANK flag, indicating that the scientific name has not been recognised but that the data point has matched at a higher taxonomic level, e.g., genus or family. This flag can be used for identifying and filtering for these data. When names have been misspelled or badly formatted, there is also a TAXON_MATCH_FUZZY flag that can be used for identifying and filtering names that can only match the taxonomic backbone using a fuzzy, non-exact match.