Something strange going on with taxon after a taxon swap

I’ve noticed some strange results after a taxon swap
Melampsorella caryophyllacearum [334444] was replaced by Melampsorella elatina [1400103]. View Taxon Change

However, Melampsorella caryophyllacearum still has 220 observations despite being inactive, and even stranger, many of those observations are “Research Grade” despite being identified as an inactive taxon:


As a particular example, this one: https://www.inaturalist.org/observations/75559099

Not only is it research grade as an inactive taxon, but it’s still identified as the inactive taxon despite both identifications applied to it having automatically updated to the new one. Even weirder, the automatically changed IDs are now listed as maverick, even though there’s nothing disagreeing with them! And the community ID is shown as 0 out of 2.

This observation (https://www.inaturalist.org/observations/74915964) was the same way:

However, when I agreed with one of the maverick IDs for M elatina, it then swapped over to that taxon and became RG (and the other IDs ceased to be mavericks):

I experimented with a second observation (https://www.inaturalist.org/observations/15740856) and instead of adding an ID, I just checked and unchecked “cannot be improved” in the DQA.

This caused it to immediately switch over to being RG as M elatina, however, it leaves all the IDs for M elatina at “Maverick” status and the Community ID box still reads Fir Broom Rust Melampsorella caryophyllacearum Inactive TaxonCumulative IDs: 0 of 2

1 Like

I noticed that it was strange when I comitted the swap. I thought the observations where just not updated and it would take more time for the backend to process all the observations. However, I don’t know what could be the cause of that strange occurence.

This indexing problem happens every so often. I’m not sure if there’s a better way to fix it than manually reindexing each observation (with an ID, DQA vote, etc.).

I don’t think this one is just the indexing though - if it were, shouldn’t doing the DQA vote have fixed all the issues with the final observation I linked to? Instead it’s now RG as the correct taxon, but all the active taxon IDs are still “maverick” and the community ID still shows as the other species.

That was already reported, so likely should be merged into previous bug report, that is happening for months with swaps, nothing helps, but new id.

1 Like

Ah, I didn’t see previous reports about this, feel free to merge.

1 Like

I found it and it’s sadly not a report and was closed (!), while old observations like that didn’t change, so this one should stay. https://forum.inaturalist.org/t/slow-updates-to-a-recent-taxon-change/31986

I suppose it’s a bigger issue where an ID is required (still indexing though?). Reminds me of this bug:
https://forum.inaturalist.org/t/community-taxon-doesnt-exist-even-though-there-is-more-than-1-id-rumina-decollata/19220

We’ll take a look.

1 Like

Resurrecting…
Another instance (obs retaining the old inactive taxon despite the successful swap):
https://www.inaturalist.org/observations?place_id=any&subview=table&taxon_id=514775&verifiable=any

Two obs still listed as ‘Lens culinaris ssp. orientalis’ (inactive) even though it has been correctly changed to ‘Vicia orientalis’. Even confirming that new ID does not help.

I added another confirming ID on one of those, and it did not change the taxon.
Then I marked and unmarked “cannot be improved” in the DQA, which I find sometimes forces it to update - it changed to the correct taxon at that point, but now remains at “needs ID” despite having 3 agreeing species-level IDs and zero dissenting ones. Marking “Cannot be improved” has no effect on this, and it still says

The Community ID requires at least two identifications.

Definitely odd!

1 Like

Just withdrew my first ‘Vicia orientalis’ ID (formerly given by clicking the ‘Agree’ button), then re-added it after yours.
And it’s now correctly displayed as RG (green mark)… although not in the ‘Community taxon’ (still claiming that it needs 2 identifications).
Weird²

edit: for the record, the affected obs https://www.inaturalist.org/observations/78482420

1 Like