Maybe this happens every time a genus is split, but I noticed recently that with the Accipter genus being split into multiple genera (while also retaining the Accipter genus) has resulted in thousands of observations that were formerly Research Grade being knocked back to Needs ID. Most of them are because someone selected Accipiter before the split, and others refined the ID to a species that formerly was an accipiter and now is in a new genus (like Cooper’s Hawk, now in genus Astur).
Those are easy enough to deal with, if annoying, but worse is that observations that were only identified at the genus level are now potentially incorrect, because the definition of genus Accipiter has changed.
Ideally all of the former ID’s of genus Accipiter should have transitioned to Subfamily Accipitrinae, but that doesn’t seem to have happened (maybe it never happens?). Is there any way to deal with this in bulk, or is the only option to go back through all the old genus Accipiter IDs and correct them one at a time?
There is a scheduled split of Accipiter that should solve most of these issues. Of course, it may always be that some of them will need to be done by hand…
These are pretty massive changes in terms of numbers of observations, so they will take a while to finish. It looks like people are on top of things though.
I’m glad to see that at least Accipitrinae exists in the taxonomy to catch these obs, I was concerned at first seeing this thread pop up since I remember subfamilies being rejected for Anatidae.
They definitely need to be swapped, as the vast majority of Accipiter sensu lato species/individuals/observations are not Accipiter sensu stricto. By the way, I didn’t realise that Lophospiza wasn’t in Accipitrinae, even though it’s part of Accipiter sensu lato. That’s crazy. I’m not saying it should be classified in Accipitrinae, as I understand that separating it (as subfamily Lophospizinae) was necessary to make Accipitrinae monophyletic, unless Buteoninae was merged into Accipitrinae, due to it being sister to a clade consisting of Accipitrinae and Buteoninae. [1] The crazy thing is that it was part of the genus Accipiter when it turns out it should have been in a separate subfamily all along.
Anyway, it seems that Accipiter IDs have been swapped to Accipitrinae everywhere, even in Lophospiza’s range, and those ones need to be manually corrected. But I guess that’s better than keeping them at Accipiter.
If it’s causing this much trouble on iNaturalist, think of what it’s doing to field notes. In my field notes, I still tend to write “Accipiter” if I don’t get a good enough look to tell if it is a Cooper’s or a Sharp-shinned.
Part of the confusion is that we also call buteos “hawks” in North America. In Europe maybe you could get away with saying “I saw a couple distant buzzards and a hawk” for a distant sparrowhawk/goshawk?
Here in North America what are we going to call them? Hawks vs. true hawks? Mammal hawks vs. bird hawks? “I saw a couple distant Buteo’s and an Accipitrine”
Maybe we just keep calling them accipiters as a common name for the whole subfamily and harriers get to be an honourary lower-case non-italicized accipiter haha.
Turns out I overlooked Microspizias. It was removed from Accipiter and put in Harpaginae. Guess we also need to swap Accipiter to Accipitridae instead of Accipitrinae in their range too.
Accipiter sensu lato actually spans over three current subfamilies, Accipitrinae, Lophospizinae and Harpaginae. Please don’t tell me there’s ANOTHER one I missed!
It’s very common with the constant taxonomic changes of iNat, though this was a particularly difficult example. iNat has prioritized taxonomic modernity such that they feel it justifies this sort of impact.Z
Creating informal species complexes consisting ONLY of the initial members of the split, and with the name of the previous taxonomic unit, would fix a lot of these problems if adopted as a policy for all splits.
This change is being adopted unanimously among bird taxonomies as far as I know, and similar frustrations are occurring in other places like eBird, although the way eBird’s taxonomy is programmed doesn’t involve a taxonomic tree in the same way so they have a bit more flexibility:
Not having to strictly follow a tree is nice in this case since it allows for taxonomically arbitrary slash options, but it is frustrating in a lot of other situations. For example you can’t search eBird for a higher level like “hawks and eagles” and have the search filter include all photos identified to species level below that level. It doesn’t recognize that “hawks and eagles” is a bucket that includes Cooper’s Hawk and Bald Eagle.
yes, but that was a choice, not a technical limitation. I’ll leave it at that… but this was a choice to follow a certain taxonomic policy, and it has impacts.
It’s unfortunate that the taxonomic category of subgenus is so rarely used to subdivide genera when new information is found about relationships. It could minimize a lot of the mess created by having to create or resurrect genera to rearrange species. And in reality subgenus serves the same purpose.
Yeah, for example an alternative option here would’ve been to expand the concept of Accipiter to include harriers as a subgenus. But people are so used to thinking of harriers as their own group unrelated to other hawks that it’s just as hard to conceptually accept that lump.
has anyone addressed the question of whether, in groups that are complex and backcross, tree cladograms are even accurate or usable on this level? I know that goes well beyond iNat, but i know taxonomy is just forcing a complex system into less complex labels, and i think with all this genetic data we’ve pushed that beyond what makes any logical sense at all any more.
Well, as the thousands of researchers working on phylogenetics worldwide have determined: yes.
A cladogram (i.e. a phylogenetic tree) is the most accurate visualisation we currently have for relationships between organisms (or at least pluricellular eukaryotes, bacteria are a bit different). Especially in animals and particularly vertebrates, where hybridizations at higher taxonomic levels are pretty much nonexistent.
yes, the researchers in that one field have decided that they know best and should redefine the concept of ‘species’ and force it on everyone else. That’s my entire point. The ‘thousands’ of phylogenetics researchers are a tiny majority of people who identify organisms. I think the results speak for themselves, trying to use this system is making it very difficult to monitor or identify organisms or ecosystems at all, which seems a lot more important to me than cladistics, no? Given that all life depends on those?
But the thing is, Accipiter sensu lato has more than just accipitrines. With lophospizines and harpagines too, the smallest clade containing all of Accipiter sensu lato has Buteoninae. It would be quite ridiculous to expand Accipiter to encompass Buteoninae.
But maybe the splitting was excessive too. They certainly needed to split the genus and remove the non-accipitrines to not have Buteoninae in Accipiter, but they didn’t need to split that much. All the accipitrines in Accipiter could have remained in it, if they would expand Accipiter to include Erythrotriorchis, Megatriorchis and of course, Circus. Perhaps a good approach is to move the least number of species, while ensuring every taxon is monophyletic. The approach I described probably fulfills this.
But is this a reasonable approach? My intuition says that if Accipiter sensu lato had a type species that isn’t in Accipiter sensu stricto, then if a taxonomic approach would be non-trivially different, then it isn’t reasonable. So let’s imagine that Accipiter sensu lato had been called Tachyspiza. It appears that approach would be essentially the same, except that the new Accipiter would be called Tachyspiza. Now what if it were called Lophospiza? In this case, almost every other Lophospiza species would need to be moved, so this approach wouldn’t really tell you much.
If this approach was applied to sauropsid classes, to fix the paraphyly of Reptilia, then the approach would be to limit Reptilia to Lepidosauromorpha, keep birds in Aves, and erect a turtle class and crocodilian class. Though it appears that Reptilia has two type species, Lacerta agilis and Crocodylus niloticus. Perhaps this means it has to have both, so Reptilia should be expanded to include Aves. Or perhaps we can choose one to remain as the type species. If choosing the lizard, the approach is as I described. If choosing the crocodile, it is essentially the same, but Reptilia is Pseudosuchia, so cannot be used for Lepidosauromorpha. Aves has so many species, that putting them in Reptilia is always disfavoured with this “minimum species being moved” approach. In fact, if all sauropsids were to be in one class, it should be called Aves in this approach.
There is no unified species concept in biology, that’s the thing.
The “biological species” concept (based on total reproductive isolation) that was force-fed to students for decades is as flawed and inoperant as any other ;)
