A Concept-Aware API for the Taxonomic Checklist Infrastructure in Germany

Biological data refer to taxa (e.g., species) or to populations (e.g., all individuals of a taxon in a region). Scientific names of organisms are commonly used as identifiers for those data, e.g., in databases. However, scientific names are assigned (linked) to taxonomic concepts, which very often change over time or space, e.g., when a species name refers to different infraspecific taxa in different regions (Geoffroy and Berendsohn 2003). For this reason, the names themselves are not sufficient to unambiguously assign factual data to the taxonomic concepts they belong to. The problem has been discussed for many years, and data models such as the Berlin Taxonomic Information Model (Berendsohn et al. 2003) and the Common Data Model of the EDIT Platform for Cybertaxonomy (Güntsch et al. 2018) allow taxonomic concepts to be represented and linked. Taxon concepts are considered here as sets of (potential) specimens. Thus, relationships between taxa can be handled the same as mathematical set relationships (Fig. 1). From a user perspective, however, there are hardly any interfaces that can be used to effectively access taxonomic concepts and their relationships, for example to semantically annotate research data with name references. The German Federal Agency for Nature Conservation (BfN) and the National Research Data Infrastructure NFDI4Biodiversity have cooperatively developed an API (Application Programming Interface) for accessing available German species checklists, where taxonomic concept relationships between checklist versions can flexibly be queried and integrated into other systems. The query for scientific names includes a similarity search and initially returns the taxa with matching names. In a subsequent query, hierarchies, synonyms and the concept relationships to name usages in other checklist versions are output (e.g., congruence, inclusion, pro-parte-inclusion, interference, exclusion, and total exclusion). From this output, it is possible to recognize, in particular, to which taxa an assignment of information is problematic. The API is currently in the testing phase and is used in the context of NFDI4Biodiversity for the annotation and integration of research data.