External providers - time for an update?

Curators
1

Hi all.

As most reading likely know, users can import names to the taxonomy using the “search external providers” option when attempting to suggest an identification - but when the name that want seems lacking from core system.

A couple of months ago, i highlighted what i thought was a glitch when trying to help a user (non-curator) import names they wanted.
See https://forum.inaturalist.org/t/glitch-with-search-external-name-providers/53225/2

An explanation i got from @tiwane was that “iNaturalist uses CoL 2012”. I’m sorry to him that i’ve since left it for a couple of months without further comment beyond that chat, but i’ve since been exasperated on a daily basis by noticing when users are importing names from external providers (notably, or perhaps exclusively CoL) but importantly names that do not match with any recent iterations of CoL.

So, to open a discussion. Is iNat still using the version of CoL from 2012? If so, that’s absolutely absurd. While ability of any user to import names from external providers still seems to me a good feature, being able to do so from a source that’s 12 years out of date seems counterproductive. Within the above link, @ cooperj gave a viewpoint from fungi worth noting. However, my concern stems from those for Animalia, where despite many flaws CoL has updated taxonomy on multiple animal lineages over the last 12 years. So why the heck does iNaturalist let users still import from a version of the CoL database that’s so horribly outdated?

We can make curatorial case for why names on such old databases shouldn’t be allowed to be imported, here’s a starter:

Imported taxon goes to ungrafted.
(this seems related to if wider group said to be ‘locked’ but also if the hierarachy is poorly established)
Imported taxon gets misgrafted (seems common with homonyms)
Imported taxon gets grafted but the combination outdated - i.e. a synonym in later studies, and sometimes seems to add the valid combination as a synonym!
Imported taxon is mispelt or wrong suffix

[Edit further] I’d welcome other curators etc add in viewpoint about when and where they see the volunteer curator workload being increased by the external provider(s) being outdated, i.e. leading to the need for curator to step in an edit/annotate what users (mis)create. With addition of new taxa, i’ve often added in links to taxon framework or external sources, but increasingly with recently added external names, then i’m also regrafting, swapping to merge synonyms etc

5 Likes
2

External providers seem pretty useless for any mollusks I try to add. That’s usually because I’m adding new names or combinations that aren’t in CoL. So I’m for having an updated external provider (WoRMS would be great for many groups), but the current system only adds work in that I have to manually create each name.

3 Likes
3

Personally, as iNat’s taxonomic backbone “matures,” I would like to see the External Providers option go away altogether, and have all requests for new taxa come in the form of flags for curation. At this point I doubt that the External Providers option is reducing Curator workload very much, if at all.

13 Likes
4

I do plants, and go straight to POWO to get the useful link.
Or to Google to work around tiny typos, or common names not yet on iNat.
The few times I have tried ‘external providers’ it has been fruitless.

2 Likes
5

for me personally, as I go to add a new taxon entry? it probably saves me about 10 seconds on average, if the taxon is in the 2012 CoL listings. in terms of across my own and others’ work? it probably adds 30 seconds on average, taking into account outdated taxa that keep getting imported over and over again.
so no, I don’t feel that it’s that useful to keep the option, especially since it is in fact still pointing to the 2012 edition of Catalogue of Life…

4 Likes
6

I didn’t realize our CoL taxonomy dump was from 2012. If so, that’s definitely not ideal for arthropods, as CoL’s arthropod taxonomy was abyssmal in 2012. Back then they were using several unreliable and self-published databases for input, some of which just did their own synonymies willy-nilly without relying on officially published sources.

1 Like
7

Agreed. I think the “Import from external name providers” button should be replaced with a link to open a new flag

3 Likes
8

I think it should be updated if it causes problems. I’ve never used it for Chironomidae.

9

This option works without need for a curator and should be kept.
Sometimes it takes a long time to resolve a flag and the ID might never happen.

3 Likes
10

Thanks for feedback on this everyone. Usually with our own preferences of taxonomic focus, whether or not such external things have an influence or not is important, so the viewpoints from different angles is welcome.

I would not be against simply discontinuing external import by users from Catalog of Life, but as per the last comment ( [traianbertau), my view from my bias towards arthropods is that continuing to allow users to add names has utility. The great point is they can quickly add a name (many described arthropod species are of course still missing from the iNat scheme), then link their observation - the curation to fix grafting, naming, and links beyond to GBIF etc can come later. However, if they’re imported outdated names/combinations, then i’d suggest that curators stepping in and “fixing the mess” looks very shoddy from those users who e.g. ‘simply want to put a taxon name on their observation’. Increasingly, i’d like to see that IF a casual user can add a taxon name, then curators have potential to get notification that a taxon has been added (perhaps limited to particular taxa of their defined interest), which raises option it can be verified or expanded on, i.e. external links, synonyms, common names etc can be added if the curator wishes to do that.

Recently, i’ve only haphazardly discovered “new names” added by users in particular taxa of my interest - and increasingly from non-curators those are outdated names BECAUSE the CoL scheme is so outdated. Many of the names i see added could be linked externally etc, or more importantly, an increasing percentage which are added by non-curators then need swaps etc as redundant with other names, often with inactivated names, or misspellings, etc, etc.

Well then how to proceed? Am I/are we howling at the wind here, or can we now get some feedback from admin/staff on this?

11

Thank-you all above for commenting

Until something changes, to circumvent closure of comment due to inactivity, then i’m now aiming occasionally add some notification of various taxa where cutation action flags (edit: “seem like”) arise from the outdated CoL import. Today it’s going to be for the imported brown algae “Kuckuckia spinosa” which for whatever reason that i don’t choose to investigate had mis-grafted under the fish genus Barbus

https://www.inaturalist.org/taxa/1576855/flags
Kuckuckia spinosa
per https://www.marinespecies.org/aphia.php?p=taxdetails&id=144103

As an aside,
there’s still misgrafted taxa under Barbus, namely “Dipterosiphonia dendritica” and “Liebmannia leveillei”
https://www.inaturalist.org/taxa/49620-Barbus
and under others under Coregonus, namely “Dipterosiphonia rigens” and “Halydictyon mirabile”
https://www.inaturalist.org/taxa/87626-Coregonus

It’s arguably at best a few minutes of volunteer curation effort to regraft these to accord with modern databases, but it looks core updates by paid staff might be implemented to reduce expections of volunteers

2 Likes
12

Interesting, that bug of grafting automatic imports somewhere into fish taxonomy has been around for many years. I’m surprised it has never been tracked down.

1 Like
13

Random boring case from Booklice, just to prevent closure and expand.

Genus Pericaecilius Mockford, 2000 and species (3x in genus) requested by user in flag. (Sept 2024). Was of course present in the modern CoL 2024 database from authoritative source Psocodea Species File.

1 Like
14

As i usually filter to at least down to the mega-diverse ‘arthropoda’, i’d not seen before, gAh.

15

Does anyone have positive suggestion how to motivate the discussion here beyond us few? I’m strongly in favor of the above idea that “External Providers” option gets removed entirely - e.g. to be replaced with ‘flag for curation’. I suspect many users fail to understand how to ask for curator help anyway, and i wonder how much confusion/frustration is felt when the taxonomy doesn’t match the names they have from their other sources.

1 Like
16

um, if you’re worried about that, the ‘external providers’ link is the least of your concerns. We should get rid of any link to POWO for a start. No one in the non-inat, non-taxonomist world uses the same names as POWO. The link to external sources option does seem depreciated, but clearly isn’t linked to that issue…

17

Maybe submit a Feature Request for doing so? Don’t know how much it would expand the participants in the discussion, but at least it might push the idea along among iNat staff.

1 Like
18

Thanks for the views above. “Which external databases” to link any import onto of course should be considered if admin begin to implement any changes.

To echo on the “… that “External Providers option gets removed entirely”. I’d also favour the front end “import” feature just being removed, e.g. changed to a “create flag” (for curation). This will also move debate about who/what to import from (or mirror) to be issues for the backend. Turning this into “feature request” is a good idea, thanks for that.

Anyway, new case.
User created a taxon flag for the spider Meioneta habra, commenting that “it seems to be in the wrong genus.” Indeed, that combination being (mis)grafted under genus Agyneta - below a long list of correctly formatted Agyneta spp. The entry Meioneta habra (i.e. 1521489-Meioneta-habra) was created late 2023 by another non-curator user, via CoL import. For the taxonomy, the species “habra” had been recombined from Meioneta to Agyneta about 10 years ago in a study by Dupérré, 2013 to be Agyneta habra (Locket, 1968), and various databases have long reflected that revised combination - including already on iNaturalist (i.e. correctly as Agyneta habra). The import of 2023 therefore led to a needless (mis)duplication, with the redundant “outdated combination” misgrafted under Agyneta. The issue itself was quickly rectified with curatorial action to merge the duplicates under the single valid taxon name. To highlight though, it’s now taken two site users and a curator to fix this, so those of you who may feel that user-import either speeds up or simplifies such things then please bear cases like this it mind. The outdated name was wrongly on display for a year and taken into checklists - it needed another user to highlighted the mistake with a flag.

1 Like
19

Please do not remove the import from external provider thing.

It comes very handy to be able to get a species quickly available for Id without waiting for a flag to be resolved (a few flags for add species requests do take weeks or months to be done).
Anything that is possible without involvment of curators should be kept - there are not enough curators to deal with everything quickly.

3 Likes
20

Just on this issue about whether the import feature is useful (or not), i’ve stated my position that i don’t find it useful - and go further to clarify that i even regularly find the current setup to be counter productive in that it allows import from outdated source(s).
That said, I see the potential delays of “adding missing taxa” as a very important concern. Part of this (i fear) comes from what i see as a general lack of value given to the taxonomic framework (which i see by user comments elsewhere on this forum etc), but perhaps part stems from lack of drives towards the community for engagement, either in feedback, but especially in failure to recruit active curators. The number of unresolved curatoral flags is impressive. I find it shocking how many relate to really quite simple and seemingly uncontroversial requests. Many who commented above (and elsewhere on the forum) have often made great progress to resolve such flags, but many unresolved flags remain. I’d go further to say there are many unresolved for more than several months, which I find rather terrible. There’s been some discussion of ways to resolve such backlog and move things forward, such as curator mentorship, but none of these ideas seems to have gone anywhere. Can we perhaps have a cull of the “Curators” who have never done any curation? How about switch some of those who are active users (as identifiers etc) but not interesting in direct curation to be “taxon advisors” and such, alongside encourage and support those who state some intent and interest for direct curation? If there’s more active -and hopefully informed - curators, then there could be much quicker resolution of flags, especially the simplistic ones to “add missing taxa” and such. As i see it, the discussion and formalisation of the preferred taxonomic combination (i.e. that said to be “valid” or current) can come after the end user has got an available taxonomic name to link their new observation onto.
[note, slight edits above later, mostly to fix some poor english, sorry!]

2 Likes