Is camtrap-dp:taxonID a scientificNameID or taxonID?

[peterdesmet]

The definition of Camtrap DP's taxonID is:

Identifier of the scientificName as defined in package.taxonomic.taxonID for that scientific name.

In package.taxonomic.taxonID the identifier has the following definition:

Unique identifier of the taxon according to the taxonomic reference list defined by taxonIDReference.

The term acts as a unique identifier for the following terms in taxonomy:

• taxonIDReference
• scientificName
• taxonRank
• kingdom
• phylum
• class
• order
• family
• genus
• vernacularNames

What is the good Darwin Core equivalent of this term to included in occurrence data?

1. https://dwc.tdwg.org/terms/#taxonID
2. https://dwc.tdwg.org/terms/#dwc:scientificNameID
3. https://dwc.tdwg.org/terms/#dwc:taxonConceptID

See also this discussion.

[peterdesmet]

This also affects the camera trap publication guide and camtraptor.

[peterdesmet]

@mdoering what would you advise here? If I can distill your comments in this discussion, you would advise to use:

• scientificNameID for this type of information in occurrence data
• To populate the term with a link https://www.catalogueoflife.org/data/taxon/QLXL, rather than the naked identifier QLXL

Correct?

[mdoering]

As cpt:taxonID appears to define taxonomic values such as the classification it is per definition a Taxon identifier, not one for the name alone. I would therefore think https://dwc.tdwg.org/terms/#taxonID is the corresponding term in DwC.

The question then is what taxon identifiers exist that you can reuse? There hardly are taxonomic identifiers, but mostly name ids as I've mentioned in the above discussion.

The identifier COL QLXL when not used with a specific release really is a name based identifier. The classification of the name can change between versions of the COL checklist, but not the name. If you want to unambiguously reference the taxonomy of that name in COL you currently would have to use a specific release such as https://www.checklistbank.org/dataset/9923/taxon/QLXL or COL:9923:QLXL as a simpler identifier that does not rely on a URL and resolution. COL is also working towards true stable taxon identifiers, but that will still take a while.

[mdoering]

camtrap-dp NAME information:

• scientificName
• taxonRank

camtrap-dp TAXON information:

• taxonIDReference
• kingdom
• phylum
• class
• order
• family
• genus
• vernacularNames

[peterdesmet]

Thanks for the reply @mdoering.

1. I follow your thinking that we intent in Camtrap DP is a Taxon Identifier (and thus a dwc:taxonID). It acts as a link between the observation and the full taxonomic information in package.taxonomic, so we don't have to repeat the full taxonomic information for every observation record.

2. I think we should update the definition of taxonID slightly from:

   Identifier of the scientificName as defined in package.taxonomic.taxonID for that scientific name.

   To:

   Identifier of the taxon of the scientific name. Foreign key to package.taxonomic.taxonID.

   3. While we use QLXL as identifier, it is implied to be interpreted in combination with the required term taxonIDReference which in the example dataset is https://www.checklistbank.org/dataset/3LR. @mdoering is that fine as a reference? Or should we use your suggested https://www.checklistbank.org/dataset/9923 instead? Note that users may populate this with whatever URL, so it won't always result in a specific version for the taxonomy.

[mdoering]

Agree with all Peter. Just not that the URL to 3LR is like a redirect to the latest current version and thus the content behind it does change over time (monthly here). The one with a fixed dataset key does not. I guess it depends on your intention which one to use.

For long time stable annual releases of COL we also provide these URLs: https://www.checklistbank.org/dataset/COL2022

[peterdesmet]

Thanks Markus! I've made the necessary changes in #352. I will use https://www.checklistbank.org/dataset/COL2023 in the example dataset