Taxonomic mismatches when using `lsid` for some atlases

[mjwestgate]

Affected atlases:

• Estonia (https://elurikkus.ee)
• France (https://inpn.mnhn.fr)
• Sweden (https://biodiversitydata.se)

If I run:

galah_config(atlas = "Estonia")
galah_call() |>
      galah_identify("Mammalia") |>
      atlas_counts()

I get a count of 44 records returned, whereas:

galah_call() |>
      galah_filter(class == "Mammalia") |>
      atlas_counts()

Returns a count of 117,448 records, which seems more likely. This appears to be an issue with elurikkus.ee not recognising (lsid:11618) as a query via the fq parameter in the https://elurikkus.ee/biocache-service/occurrences/search API.

[vjrj]

cc @shahmanash @sylmorin-gbif

[vjrj]

Maybe, better to use openobs https://living-atlases.gbif.org/participants/openobs/ , right @sylmorin-gbif ?

[sylvain-morin]

Hi, I agree with @vjrj , please use OpenObs (https://living-atlases.gbif.org/participants/openobs/) for France instead.

[mjwestgate]

OK we'll swap this over. Thanks for the tip!

[mjwestgate]

It appears that updates the taxonomy sometimes break links between the taxonomic and biocache services, regardless of which living atlas is involved. Closing this, but have added tests to pick it up when it happens