Use int taxon identifiers in Darwin Core Archives

[markus]

The very useful " Darwin Core Archive " for iNats taxonomy uses URLs as taxon identifiers instead of plain integers as the iNat API and forum discussions do. The data gets used in GBIF and ChecklistBank where I think it would be much better to have the true iNat integers as identifiers instead of the URLs. Especially when users want to map their own names to iNaturalist taxon identifiers.

@loarie would this be an option for future versions? Links to the iNat taxon pages can still be supplied in a dedicated link field dc:references.

id,taxonID,identifier,parentNameUsageID,kingdom,phylum,class,order,family,genus,specificEpithet,infraspecificEpithet,modified,scientificName,taxonRank,references

1,https://www.inaturalist.org/taxa/1,https://www.inaturalist.org/taxa/1,https://www.inaturalist.org/taxa/48460,Animalia,,,,,,,,2021-11-02T06:05:44Z,Animalia,kingdom,http://www.catalogueoflife.org/annual-checklist/2013/browse/tree/id/13021388

2,https://www.inaturalist.org/taxa/2,https://www.inaturalist.org/taxa/2,https://www.inaturalist.org/taxa/1,Animalia,Chordata,,,,,,,2021-11-23T00:40:18Z,Chordata,phylum,http://www.catalogueoflife.org/annual-checklist/2013/browse/tree/id/13021451

[markus]

Note also that dc:references currently points to COL and I suppose to other sources for other taxa. Consumers of the data would rather expect a link to the actual iNat taxon page in such an export. dc:source, which is not explicitly mentioned in the official DwC terms, could be used instead to capture the link to COL or other sources that are the primary references for this taxon in iNat.

[pisum]

unless i’m misunderstanding, the integer taxon identifier is stored in id.

[markus]

Yes, that is correct. But it is not used for dwc:taxonID and the related dwc:parentNameUsageID. So officially the identifier is the URL, not the integer. The id column is informal only and opaque to the dwc standard and tools that build upon that.