So in one of my previous posts (correcting mass misidentifications) I talked about Auriculella perpusilla. While now the snails currently identified as Auriculella perpusilla are not currently lumped in with Auriculella diaphana, there are still a lot of potential misidentifications.
The problem arises from the following: Auriculella perpusilla has a series of features that separate it from Auriculella diaphana despite them now only occurring in the same location. Auriculella perversa is another species, now only found in the same location, that is nearly identical to perpusilla. Distinguishing between the two is nearly impossible. While I actually don’t completely understand the differences, I’m pretty sure you need an apertural view of the shell to even have a chance at being sure of identification. It’s just not possible to ID based on the dorsal view alone.
To those who are not familiar with molluscs, left is the apertural view which shows the aperture, or hole, which the snail’s body comes out of. The right is the dorsal or abapertural view. The shell above is Auriculella perversa.
Now, apertural views of shells of live snails are quite the challenge to get, and to date no perpusilla observation on iNat has the apertural view visible. So theoretically none of these can be ID’d as perpusilla. Of course, we can ID them as a new complex that will be added, the perpusilla group, which includes both species. That is, if the perpusilla group was a real grouping.
In fact, the perpusilla group is polyphyletic. All the members of the perpusilla group are more closely related to some species outside the group than to the other species inside the group. So under iNat’s current rules for making complexes, this species “group” is invalid.
Would you believe it if I said it was actually more closely related to this species, Auriculella tenella, than to Auriculella perversa or Auriculella perpusilla?
So I’m stumped on how to handle this. The current situation is that the obs. are IDd as perpusilla, but in reality we’re (exclusive) saying “this is a member of the perpusilla group, but not necessarily perpusilla itself” which is confusing and not how species IDs are supposed to work. Should they be kicked back to genus and have “Community taxon can be improved?” crossed out? Theoretically yes, right?
I feel like we need a way to have groupings that are non-taxonomic for cases like this. I think the genus is too coarse an ID but the species is too fine an ID for obvious reasons.
I think we use the finest monophyletic group. If there is convergence then we just ignore that when we make the ID. Chances are that there are more traits that gagneorum and tenella have in common than either have in common with Auriculella perversa or Auriculella perpusilla. Even if not, there are probably missing synapomorphies on a physical or genetic level. Making a new taxonomic unit that is non-monophyletic leads to a whole bunch of overlapping groupings like you see on BugGuide here: https://bugguide.net/node/view/7176/bgpage?from=13
Bumping back to genus is fine. You can add a comment indicating the possible options, or you can make an observation field for the species group if there are enough observations you think it is worth keeping track of.
I second Thomas’s suggestion. Bump them to genus and if you need to keep track of them as a group, create an observation field for that. Observation fields are great for tracking morphological similarities that might not correspond to taxonomy.
If you’re certain a species level ID isn’t possible you can always knock an observation back down to genus level with an explination to why. Then on the Data Quality Assessment where it says “Based on the evidence, can the Community Taxon still be confirmed or improved?“ you can just tick “No, it’s as good as it can be”.
The observation will then be kept at genus level, yet also obtains research grade status. I usually do this with observations of red Sarcoscypha in the UK as there are two species that look the exact same and they can only be told apart under a microscope. Without microscope photos on the observation I bump it back to genus and mark it as the ID not being able to be improved. The observation then reaches research grade and (most) people won’t return to the observation to give unverifiable species level ID’s.
This is a problem with Yellow-faced Bumblebees in the Pacific Northwest. One species is abundant but two others occur and we can’t tell them apart except in very unusual photos. They aren’t each other’s closest relatives (they don’t form a clade). We can bump them up to Pyrobombus, which is smaller than their genus Bombus but contains some bumblebees that don’t look quite like the Yellow-faced species. This is not as helpful as I would like but it’s what we can do. So I do it and mentally grumble a bit each time. So it goes.
I kind of wish there were some way to suggest two or three different potential identifications, in general. e.g. if I see an oak gall I can’t identify, I can at least be fairly certain it’s either a midge or a wasp; winged insects seems unnecessarily vague. I guess one could make multiple conflicting IDs, then withdraw them all and ID to something generic, but I’m not sure the intention would be well understood.
Gall projects and Evidence of Presence: Gall help out with the problem of not knowing what type of organism. You can write what possible species it could be in the comment of a broader ID.
There’s no perfect solution, but I agree it’s definitely a bummer when one can confidently narrow an ID down to two or three species but the narrowest valid taxon containing them has dozens to thousands of options. You might find some of the past discussion on this relevant; there is even an active feature request dedicated to this.
Update: I’m done bumping everything back to genus, perpusilla now has 0 observations (I’m still keeping the taxon image though since it does illustrate what perpusilla looks like even though we aren’t sure if it is perpusilla)
Leaving a comment won’t help experts find the observation to help make a more definitive ID. If I ID a gall as “Arthropods” (e.g. it’s either a wasp or a gall mite), that very generic ID won’t help identifiers of either family find the observation.
I ID as Life - to annotate as gall (which is also my comment) - add to the relevant project - ask about the host if observer has not provided that info. And the IDs roll in. Random example https://www.inaturalist.org/observations/277087301
I’m just here to express solidarity from the birding community.
Recently the taxonomy was changed and now Sharp-shinned and Cooper’s Hawks (a commonly confused pair, not always separable from photos) aren’t even in the same genus! You now have to go all the way back up to Subfamily Accipitrinae, which nobody had heard of before.
Taxonomy doesn’t even allow us to definitely distinguish geese from ducks! They are all in Anatidae.
“Accipters” are definitely a frustrating case, there isn’t even a good term we can use now when you see a distant Coop/Sharpie that doesn’t include harriers as well (at least that level exists on iNat, unlike the equivalents for Anatidae). Sandwich Tern is another potential future case.
For waterfowl there are subfamilies that could be used, for example Anserinae contains all the geese and the ducks are divided into pretty sensible subfamilies and tribes, at least based on the list on Wikipedia. But iNat prefers to reduce intermediate ranks as much as possible.
You’re right. I thought that the ident_taxon_id= trick would find withdrawn ids, but it does not. I still think it would be better to be able to propose multiple conflicting IDs than to resort to the lowest common denominator, which in some cases is “Life”.
