How to address the change to the GBIF identifier scheme?

Curators
1

GBIF recently refreshed the design of their site, and the result generally seems to be an improvement. It seems to be easier to conduct faceted searches, the UI seem cleaner, and I’m sure there are a bunch of other improvements that I’ll discover over time. But the changes have made it hard for me to link iNat taxa to their equivalents in GBIF and I’m not sure whether fixing this needs a workaround or a change to iNat code.

Background: iNat has the ability to include GBIF observations in the distribution that is shown on the “Map” tab for a taxon, along with the mini distribution maps that show up on each Observation page and in the Identify and Compare modals. This is the source of the small pink squares in the screenshot below.

Screenshot 2026-06-22 at 9.40.56 AM
Screenshot 2026-06-22 at 9.40.56 AM902×590 153 KB

For the GBIF Occurrences checkbox to be shown requires a curator to specify the GBIF Species ID for that iNat taxon. The process for this was as follows:

  1. Search for the taxon on GBIF and copy the identifier portion of the URL (e.g. 3750498 for Hedyosmum mexicanum, which could be viewed at https://www.gbif.org/species/3750498).
  2. Open the corresponding iNat taxon page and click the Taxonomy tab.
  3. Click Taxon Schemes on the right.
  4. If there is no association with GBIF, click Add to a Scheme and select GBIF; if the association already exists, click Edit to the right of GBIF.
  5. Paste the GBIF species ID into the Source Identifier box; ignore the Taxon name box as no-one understands how this is used.
  6. Click Save. Then navigate back to the Map tab for the iNat taxon. Wait a couple minutes, then refresh the page and you should see the option to show GBIF Occurrences.

Since their redesign, GBIF seems to have moved to using alphanumeric taxon identifiers (e.g. https://www.gbif.org/taxon/3JZW3) in place of the previous numeric species identifiers. It seems that these taxon identifiers are based on those from the Catalogue of Life. On the GBIF site, the old species URLs redirect to the new taxon-specific URLs. And it seems the distribution lookups from iNat continue to work.

The problem I’m facing is that iNat can’t use the new alphanumeric taxon identifiers to retrieve distribution maps. And I can’t find a way on the GBIF site to determine the correct old-style species identifiers. For example, I want to add GBIF distribution info for Epilobium conjungens, but all I have is the new taxon identifier (6G588) and that isn’t working on the iNat side.

Anyone have a smart workaround? Or do we need to ask iNat to work with GBIF to fix this lookup functionality?

2

Temporary solution is Google search, which leads to https://www.mindat.org/taxon-3666683.html

3

Not sure how long they will keep it online, but the old.gbif.org page still has the correct match (‘taxonID’ 6G588 > ‘View in GBIF backbone’ button pointing to 3666683). I suppose there’s some API at work, able to provide the correspondence?

4

Thanks @GeolTel and @egordon88! Those workaround options are a big help and work fine for now.

@loarie and @rademacherdan: It would be great if you could review how iNat links to GBIF. I’d hate for this to suddenly break for all taxa. And if GBIF plans to deprecate species identifiers, I’d expect that coding an alternative approach and transitioning to it might take a while.

5

@rupertclayton

Thanks for great overview! I was also impressed by the GBIF redesign, but dismayed by the sharp transition. I’ve heard various things now for several years in response to GBIF taxonomy issues about the “CoL merger” ongoing, but as an end user and someone who has been trying to get taxonomy schemes better integrated, them just having their backbone as the CoL scheme seems like three steps forward and two back …maybe i didn’t yet discover on GBIF how to get other hidden aspects of previous valuable data to display again!

At this end, i also tried to add a couple of “new taxon identifier” today to some newly added taxa requested in flags, and yes, also found they did not link the mapped data on GBIF.

Anyway - great now to see there’s still a functional “old.gbif.org page” thanks @GeolTel That gives a quick workaround for many - adding the old numbers as before seem to function on those as redirects. My concern with that is how long the “old page” overall will continue to function, and as stated above, and how long any of those numeric only identifiers will work for. [Post edit: The main GBIF blurb on the initial post above says “with occurrences organized to the old backbone will remain temporarily available …” so great - another in a long line of taxonomy based initiative soon to cause us a wealth of dead links across sites].

Beyond this, i’ll add that elsewhere, those old “numeric only identifiers” are often already in corresponding wikidata pages for many taxa. So, curators can search for those “old numbers” there in many cases, at least until for example someone makes a bot to replace them etc!

More generally though, rupert said"

Well, i’d point out that to me it looks like it’s already broken for quite a few thousands already if not hundred of thousands. It’s worth considering that the extensive CoL listings are poor for some lineages (sometimes critically so) or just outdated for others (sometimes horribly so!).

To issue - there’s a bunch of taxa that used to have functional identifiers on GBIF (with museum records etc) that i cannot now find how to access though their new interface (seemingly relying on Catalog of Life [CoL]). Here, for example the CoL scheme was pretty dire for the beetle family Chrysomelidae, where genus Luperus Geoffroy, 1762 has fourteen species. [Only if using their XR version, with names from diverse sources] I mention it as GBIF now has only same fourteen on main page, but used to have about 45 species! For anyone checking, there are 63 valid species in authoritative Bezděk & Sekerka, 2024, " Catalogue of Palaearctic Coleoptera, Volume 6/2/1" (and that’s just Palaearctic - i didn’t check if any described from outside the Palaearctic zone)

Here’s some not on the current short CoL species list
Luperus aetolicus Kiesenwetter, 1861
https://www.wikidata.org/wiki/Q2195914
https://old.gbif.org/species/5751987
https://www.gbif.org/species/5751987

Luperus flaviceps Apfelbeck, 1912
https://www.wikidata.org/wiki/Q2469534
https://old.gbif.org/species/5752005
https://www.gbif.org/species/5752005

Luperus graecus Weise, 1886
https://www.wikidata.org/wiki/Q2454976
https://old.gbif.org/species/5751982
https://www.gbif.org/species/5751982

Basically, the last modern link for each goes now to a “Dead” page. I’d hope that those are a placeholder for those which may come back once taxa are verified and updated in the CoL core scheme. These all above i think are currently ‘valid’ by beetle taxonomists. They all used to have museum data etc linked on GBIF.

Finally just to note, i looked at Luperus as previously that genus used to be misduplicated in GBIF due to it having two different version of the claim authority/year. Else the species “graecus” has an online misspelling that was causing a mis-duplication. I guess by making many 'dead" links in the new GBIF format they’ve incidentally got rid of some of their past problems by delinking those too ;)

EXTRA NOTE - After posting, I just thought of a different case for a recently described taxon. That’s in Catalog of Life and just updated in GBIF but was not in their system before, as others it has a “new alphanumeric taxon identifier” But for this i don’t see any old style numeric only number (it was never in the ‘old’ database!). Is every new taxon on GBIF only going to have an “alphanumeric taxon identifier” ?!?!

6

It’ll be good to figure out integration with the new GBIF codes. I frequently handle flags related to outdated/incorrect codes, including one just now. I used the workaround suggested above, for the time being.