Erms mapping¶

Some rules how ERMS is mapped to CDM¶

Important rules and queries to check are currently to be found in #8792

Accepted taxa, synonyms, parent, child¶

Taxon status¶

The taxon status (accepted, synonym, misapplied, unresolved) is kind of mentioned in the ERMS field tu_status. But information from this field is not exact enough. The better way to retrieve information if a taxon is accepted or synonym is to check tu.id = tu.tu_accfinal (or tu.id = tu_acctaxon). In the following a synonym is defined by tu.id <> tu.tu_accfinal.
This link should be direct, not recursive, but this needs to be checked (#8792)

Note: There are some autonyms (and/or alternate representations?) being the accepted taxon of there own parents. These need to be handled as accepted taxa (but should be cleanup in future). See comments in ErmsTaxonRelationship.isAccepted() and ErmsTaxonImport.getAcceptedTaxaKeys() and in #8792

Note: There are some records with tu.tu_accfinal == null . This should be fixed by VLIZ over time, but for now these taxa are handled as accepted taxa as there is no other possibility to define their status.

Note: The fields tu_acctaxon and tu_accfinal do have a similar semantics but tu_accfinal seems to be more completely (for current data)

In ERMS synonyms are allowed as taxa within the taxonomic tree. Also factual data are attached to synonyms. This is not possible in CDM. Therefore such synonyms are handled as "Taxon" in the CDM and the synonym relation to the real accepted taxon is handled via taxon concept relations.

In ERMS the parent relationship is primarily nomenclatural, not taxonomic. E.g. a species name has as parent the according genus. This is no problem for real accepted taxa, but for synonyms it can't be used. Therefore for synonyms the parent is used only for retrieving the name parts.

Misapplications¶

Misapplications are currently recognized by patterns in tu.tu_unacceptreason. These are not standardized in ERMS. Some MAN are probably not recognized this way. Other ways (e.g. by authorship) are discussed in #8792. But as they are not exact enought further data cleaning or standardization is needed before they can be used.