Is there any tool or API or trick known to the community to find mismatched GBIF links? Example: tonight I was looking at Formica incerta and noticed no GBIF records in the iNat mapped range. When I clicked through, it was linked to Formica incisa instead, so I updated the GBIF link under Taxon Schemes.
2 Likes
I don’t quite understand where the mismatch was:
Under https://www.inaturalist.org/taxa/81745/history
I see that there are no recent changes for F. incerta, but maybe your change wouldn’t appear here. Where have you been able to correct the issue?
See the Taxon Schemes link in Taxonomy tab. e.g. for F. incerta this is the link:
https://www.inaturalist.org/taxa/81745/schemes
It doesn’t have an edit history, and is only curator-editable:
1 Like
thanks, now I understand: exports to GBIF are using the scientific name, and GPIF does the mapping. That is was what confused me at first. You are not referring to exports.
In the present case the GBIF id mentioned in the taxon scheme was wrong, and a non-curator wouldn’t have access to this information, or at least couldn’t edit, as you say.
I think the same problem was discussed here:
https://forum.inaturalist.org/t/something-wrong-with-gbif-map-overlays-for-numerous-insect-species/26126
And this seems to be the answer - not entirely satisfying:
https://forum.inaturalist.org/t/something-wrong-with-gbif-map-overlays-for-numerous-insect-species/26126/11
2 Likes
Another example https://forum.inaturalist.org/t/gbif-dots-margined-tiger-beetle/6270
I will continue to correct them when I see “bad” maps. Thanks for finding the previous thread.
That would be nice to implement …
1 Like
I don’t know of any tool or trick. Once I started noticing that GBIF links are quite often missing or incorrect, I have tried to include a manual check whenever I work on an iNat taxon (e.g. to add photos or an atlas).
I would think the optimal way to do this would be to reference some type of meta-database of references to know that Generis specificis A.Botanist, corresponds to iNat taxon 12345678, GBIF taxon 87654321, POWO taxon 87612345, EOL taxon 148142, etc. We could then maintain and update the info in a single place and reduce duplicated effort.
In fact, Wikidata already does this (see Formica incerta) so maybe that change could be implemented without massive effort?
3 Likes
It would be great to have access to a list that relates taxonomy to GBIF IDs. If there is any way to obtain it, I’ll keep an eye on this thread.
Aside from the manual review that @rupertclayton mentions, I occasionally check, via API, of the species in Mexico, which ones lack this GBIF id:
On the page that displays, I look for the “taxon_schemes_count” tag:
When this indicates zero taxon schemes (“taxon_schemes_count”: 0) I locate some species for which the corresponding data is missing.
3 Likes
besides the cases where there is no taxon scheme for GBIF set up, how do you know there is a mismatch? will the name in GBIF always match the name in iNat? i think GBIF also sometimes has multiple entries of the same name for different authors. so in such a case, how do you know that you’ve matched the right one?
When the map doesn’t match iNat observations, that’s my hint to check the gbif number and see what it represents.
that’s not exactly the kind of answer i was looking for, but it doesn’t really matter at this point.
i was thinking of adding an optional column or two to https://jumear.github.io/stirfry/iNatAPIv1_taxa that would show the GBIF taxon ID set up in the iNat taxon scheme setup, and then i was going to highlight “mismatches”, but because of the CORS setup for iNat pages, it turns out i can’t do this in a browser-based way. i’d have to do this in a non-browser-based way, and i figure that would make it inacesible to the majority of folks, and so i’m not going to go down that path.
1 Like
The other side of this is: how does GBIF “know” where to store the iNat data when there are
I’m not knowledgeable enough about websites or databases to solve this effectively.
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.