Flag for Taxon: Old World Mesquites (Genus Prosopis)
| Flagger | Content Author | Content | Reason | Flag Created | Resolved by | Resolution |
|---|---|---|---|---|---|---|
| kevinfaccenda | Old World Mesquites (Genus Prosopis) |
Most species moved to other genera in POWO |
Apr. 10, 2023 21:56:47 +0000 | kevinfaccenda |
all done |
Comments
Hey Kevin, so that probably is a good split, but whew, I'd be quite hesitant in making this -- that would impact tens of thousands of observations here on iNat. I'm North American-biased, so I like to look at Flora of North America for these keys. Not saying that's a final authority or anything (iNat usually uses POWO), but at least it gives some further reasoning behind how to key these beans.
My two cents aren't really worth that much though, I just know that these taxonomic splits require a lot of extra curation and explanation to those curious.
Also, thanks for that paper link -- I need to digest that a bit more! :)
Personally, I hate the idea that this name change exists, but the evidence seems good as far as I can tell. Phylogenomic evidence using 997 genes is a lot of evidence to ignore. At least they provide a good key for separating the different genera. I hope it actually works well in all cases.
As Sam alluded to, if this change is to be made, it should be done with care considering how many observations and species this affects. I'd try to track down as many species as possible and change them all at once (at least the New World species). That will make the backlash easier to manage and reduce future difficulties of uncurated species.
In addition to citing the article, you might want to put a direct link to the article and an explanation in the comments. If I were on the other end of this swap (as I will be if this moves forward), I'd want to see enough to know why the decision was made (i.e., Prosopis was polyphyletic and the old defining character, indehiscent fruits, evolved via convergence) and understand that there is a morphological basis (spine and prickle architecture) to the new genera provided in the key in the article. For species switching to Neltuma and Strombocarpa, I'd mention that all New World Prosopis species are one of those two genera. I can help with what information to include if you'd like. Then you can just copy and paste into all the species that need to swap.
Alternatively, this paper is very recent. I suppose you could wait a year or two to see how widespread the acceptance of this is.
POWO's acceptance argues for acceptance in iNat. One issue is the many North American observations labeled "Prosopis" w/o species names are going to have to be dealt with individually. Locally (Arizona), Prosopis velutina and P. glandulosa (and the cultivated P. alba and P. chilensis) would transfer to Neltuma, P. pubescens to Strombocarpa (genera separated at couplet 5 in the article's key). P. glandulosa var. torreyana transfers to Neltuma odorata. We saw something of the same thing with Acacia not long ago. No objection from me, but I'm no taxonomist. I suggest waiting until they chime in.
I want to follow up on the phylogenomic evidence as I didn't look at it as closely as I should have. The evidences isn't as strong as I thought it was. I think there's pretty good evidence that Prosopis is not monophyletic (figure 1c and 1d show that well enough for me), but that doesn't mean the relationships are well-resolved. Three basal nodes have less than a quarter of the gene trees supporting them and there are actually more gene trees in some of those that suggest an alternative species tree. Most of the important basal nodes have less than 3/4 of the gene trees supporting. This might be enough support to suggest this alternative taxonomy, but I'm a little hesitant about it. I might check with a phylogenomicist who thinks about what constitutes good and bad support more than I do.
As for var. torreyana being recognized at species level as P. odorata, I'd have to see evidence for that! I've seen more than enough intermediates to think that this requires a lot more explanation than the author's provide. If there's a different species delimitation paper that suggested this, I'm willing to hear it out, but the lack of explanation for a group that complicated kind of rubs me the wrong way.
And one just did! Thanks, Nathan!
This change is definitely going to be a challenge to implement. I think it will involve doing a taxon split on Prosopis after all the species are moved to their new genera so that all the genera level Prosopis IDs don't becoming conflicting IDs.
I totally agree that we should make some nice template messages to attach to all the swaps to let everybody know why this is happening.
Raising P. glandulosa var. torreyana to species level does seem somewhat bizarre, I'd like to see whatever evidence supports that decision.
There are already some observations of P. odorata, and from what I have seen it differs slightly from P. glandulosa var. torreyana. The observations of P. odorata show more pairs of longer/slimmer leaflets compared to those of P. glandulosa var. torreyana. It also occupies a different range. P. odorata is found in northeast Mexico, while P. glandulosa var. torreyana is found in the southwestern US and northwest Mexico. So I think that P. glandulosa var. torreyana should still be treated as a variety, for example, "Neltuma odorata var. torreyana" or "Neltuma glandulosa var. torreyana".
In Argentina the changes have been accepted and the delimitations seem to be clear. Still, the changes involve hundreds of observations with highly significant species.
We already use Neltuma and Strombocarpa here, so we are ok to change.
@nathantaylor Are you still going to meet with a phylogenomicist about this? Also, don't forget that there is morphological evidence to compliment the molecular.
@nathantaylor Do you have any compelling evidence that we should reject this new classification? Or should we start drafting the swaps?
I don't have evidence to reject the genus names, but I strongly recommend not accepting torreyana at the species level unless some very compelling evidence with sampling in the Trans-Pecos is done. See Flowering Plants of the Trans-Pecos and Adjacent Areas (the observations there match my own; many apparent intermediates in the region).
Perhaps we can proceed with the non-conflicting taxa and leave the rest for further analysis.
Sorry for all the delays. I realize I've kind of let this one hang unresolved. It's been a busy summer for me. If you'd like to switch the genus names, I don't want to stand in the way of that.
As for my conversation with a phylogeneticist, he seemed to think the evidence was clear enough for him, but wanted to wait to see if other groups working in Fabaceae adopted the names before changing the names in the herbarium he curates.
I don't personally know what to do about the N. glandulasa torreyana / odorata thing, so in the interim I have grafted it to N. glandulosa and kept the Prosopis name.
https://www.inaturalist.org/taxa/58161-Prosopis-glandulosa-torreyana
@nathantaylor Do you have a proposal for what to do here?
All the species are now split! The swaps will probably take a day to process, at least for N. glandulosa.
The question now becomes what to do with this genus. I guess we should go through and so some manual curation to try to ID the american species to Neltuma / Strombocarpa, and then do a split on Prosopis and move all of the new world observations to Mimosoidea?
It would be nice if there was a tribe or something that this genus could be moved to rather than having to push all the new world Prosopis IDs to subfamily.
All the common names will also have to be moved, it seems like most of them will belong to Neltuma
Thanks Kevin
As for the P. glandulosa var. torreyana and P. odorata situation, I don't know enough about the populations in Mexico. Unless someone provides a perspective from the other side of the range, I think both should be kept available for people to use. If P. odorata and var. torreyana really are the same, it becomes a question of whether to treat the entity as a variety or a species. If not, perhaps there's a gradation from odorata to glandulosa. Hard for me to say without knowing the populations in Mexico nor being familiar with the name P. odorata.
Is there any reason they can't just lump Xerocladia and Indopiptadenia into Prosopis rather than breaking apart Prosopis? I hate to see visually-strong generic concepts broken apart. It's good work, but just make the lineages subgenera within a broader Prosopis concept. I'm of the opinion that broad, recognizable, relatable generic concepts are essential to botany. At worst, kick out Prosopis africana from the genus and keep everything else together.
Is it possible to create a Subtribe for all these genera?
I can find no mention of subtribes. I don't think one exists.
http://theworldwidevegetables.weebly.com/subfamily-mimosoideae.html
@gentilcore That would be a preferable solution, but it's not possible. If the genus Prosopis were to be enlarged to include all the genera you list it would also have to include all of Acacia.
Whatever happens needs to take into account that there exist tens of thousands of observations. The taxon change should be automated to the degree possible to avoid disruptions in making the transition. Recent example of what happened with
Gulf Frittiallry (Dione vanillae) comes to mind. That one took two two weeks nearly to resolve. Not being either a taxonomist or a botanist, I leave this to others to resolve. But keep in mind that the majority of persons subscribed to the platform aren't either. Etc.
One of the major takeaways from the recent case of Gulf Fritillary (Dione vanillae) was that if there are in future major taxon changes to be made based on new evidence, etc., then the person/curator wanting to introduce the change should start/flag the discussion first, followed later once all due consideration has been given, by the actual taxon change(s) (inclusive of the various subspecies). In other words, have the discussion first and then make the change(s) and not the other way around. Scott Loarie floated the idea of making possible changes in future in cases like this wherein a group of curators together with administration would be the ones to finally authorize the respective desired change(s), etc. No final decision was made but because of the disruption in question caused there seemed to be several on the forum agreeing with the proposition. As for me, I read the discussion, but that appeared to be a good proposal.
"P. odorata is found in northeast Mexico, while P. glandulosa var. torreyana is found in the southwestern US and northwest Mexico. So I think that P. glandulosa var. torreyana should still be treated as a variety, for example, "Neltuma odorata var. torreyana" or "Neltuma glandulosa var. torreyana".
Posted by salvadorcabrera21 "
Don't see how that is possible as the Type of odorata is from mojave, California, not Mexico
https://tropicos.org/name/13032111
@aguilita Do you not believe that is what happened here? I started a flag 5 months ago, recruited feedback, waited even longer in case more feedback came, and then eventually made the changes.
Great work everyone! Something strange I noticed today: I searched for Prosopis in the DFW area, and it showed many observations still as Prosopis. When I clicked on them, it showed the automatic taxon swap to Neltuma, but they didn't change in the heading until I manually added the identification as Neltuma gladulosa. Now it shows those observations as as Subfamily Mimosoideae rather than Neltuma (so it's not removing the Prosopis identification). So then I looked up Prosopis in the Americas, and there are still over 5,000 observations under Prosopis. I don't know if the taxon swap is still working in the background to fix these, or the iNat community will be having fun manually changing all these identifications.
But if needed, I can start working on some of these, at least for N. glandulosa.
EDIT: Here's an example of one identification that now shows as Mimosoideae: https://www.inaturalist.org/observations/177173981
Yes, I was waiting for all the Prosopis glandulosa swaps to finish before I address all of the lingering Prosopis IDs. It actually looks like there are still 300 which haven't finished processing yet: https://www.inaturalist.org/taxa/58160-Prosopis-glandulosa
@tiwane Would you be able to advise on this? My initial plan was to do a split and push all of new world Prosopis to Neltuma, and leave the old world Prosopis as is. But I just realized that there's no way to make an atlas for a genus. So it such a split not technically possible?
Will we have to swap this genus into the subfamily and then manually ID everything to the right genus? That's over 6000 things which would need to be manually ID'd.
I don't really deal with taxonomic issues. Tagging @loarie.
you can make an atlas for a genus, set up the draft taxon split and from that page you can click where it says 'not atlased' e.g. https://www.inaturalist.org/taxon_changes/129154
@loarie Thank you, I was trying to make one via the curation menu on the taxon page but didn't see the option there.
I have drafted a swap. This is not truly an ideal swap but it should give ideal results in at least 95% of observations as almost all of the new-world observations currently identifed as Prosopis belong in Neltuma (but some belong in Strombocarpa). And most of the middle-eastern and north african ones stay in Prosopis.
@loarie is that small amount of error OK? The only alternative I could think of is identifying the remaining 6000 Prosopis observations manually (which I do not have the capacity for) then swapping Prosopis into the subfamily.
Looks pretty good, when I click Analyze IDs I see:
Total IDs of input taxon: 12857
Number of IDs Destination Atlas
275 Prosopis Atlased
12287 Neltuma Atlased
295 Mimosoideae Outside of all atlases
0 Mimosoideae Overlapping atlases
which is a pretty good ratio of IDs that will get assigned to an output vs those that get coarsened
Committed the split. I apologize for the few hundred observations which will get messed up but I'm not sure it's avoidable.
Would anybody be able to take over the task of moving almost all of the common names from the genus Prosopis to Neltuma?
thanks for the curation work kevinfaccenda!
no problem! Do you know if the remaining 307 Prosopis glandulosa observations are still going to switch over? Or if they have glitched?
https://www.inaturalist.org/taxa/58160-Prosopis-glandulosa
Here's an example of one that was swapped 5 days ago but still has Prosopis as the ID
https://www.inaturalist.org/observations/97092412
@loarie Did you see my last message? There are still several hundred observations which haven't updated their community taxon after the swap
I forced an update for those - we have a ticket to make that mopup part of the taxon change process but these should be sorted
https://powo.science.kew.org/taxon/urn:lsid:ipni.org:names:23337-1#children
POWO only accepts three species now, all afro-eurasian
Looks like it's based on this paper: https://phytokeys.pensoft.net/article/75379/download/pdf/
and they split it into the following genera: Anonychium, Indopiptadenia, Neltuma, Strombocarpa, Xerocladia
Their reasoning is based on both molecular and morphological evidence and I can't immediately see any reason why we should deviate from their taxonomy. I can help set up these swaps.
@aguilita @salvadorcabrera21 @joshua_tx @nathantaylor @stevejones @txlorax @sambiology Do have any arguments as to why we shouldn't follow this taxonomy?