Heads up: Some or all of the identifications affected by this split may have been replaced with identifications of Pyrgini. This happens when we can't automatically assign an identification to one of the output taxa. Review identifications of Pyrgus 49760

Taxonomic Split 72689 (Committed on 2020-02-27)

Based on accepted changes to the Pelham catalogue, which were based on this paper: https://www.pnas.org/content/116/13/6232. See the Appendix for a breakdown.

Added by nlblock on February 25, 2020 08:25 PM | Committed by loarie on February 27, 2020
split into

Comments

@maractwin, @loarie, does this look right to you? It will reclassify all those individuals identified as the "old" Pyrgus to Pyrgini. This isn't ideal considering many of those individuals are in areas with only Burnsius, but I don't see a way to use atlases to automatically reassign those IDs to the correct new genus (because genera can't have atlases?). That only seems to work at the species level or below, right?

Posted by nlblock about 4 years ago

Carving off Burnsius and Chirgus from Pyrgus is analogous to the Oreothlypis described here:
https://inaturalist.github.io/trees/tree_figure.html?tfr_id=336093
as referenced here:
https://www.inaturalist.org/pages/how+taxon+changes+work
to do that you'll need

Create new inactive taxa for the narrowed Pyrgus, Burnsius and Chirgus (done)
Create a draft taxon split from the active Pyrgus to these new taxa (done)
Create atlases for the split outputs (not done)
Create draft taxon swaps for all downstream taxa where the names change (analogous to. O. virginiae -> L. virginiae in the Oreothlypis example) (not done)
Move all other children of the active Pyrgus to the inactive Pyrgus. (not done)
Commit the swaps and then the split (not done)

I like to use 'change groups' to keep track of all the swaps associated with this split (e.g. https://www.inaturalist.org/taxon_changes?change_group=Oreothlypis+split)

Does this all make sense?

Since Chirgus and Burnsius are already active this has components of a 'retroactive split' (see https://www.inaturalist.org/pages/how+taxon+changes+work) but are all the Pyrgus s.l. species swapped over to Chirgus or Burnsius? Or are some lingering in Pyrgus s.l. (and thus the need for step 4)?

Posted by loarie about 4 years ago

Yes, thank you! I've been trying to learn that taxon change summary site during this process.

All the relevant species have been swapped over to Chirgus and Burnsius, so I think step 4 is complete.

I was waiting on your feedback before moving all the remaining children of the active Pyrgus to the inactive Pyrgus, so I will definitely do step 5 before committing anything.

As far as Step 3, this is the one that's confusing me for a genus-level split like this. I'm not seeing a way to apply atlases to this case because genera don't have atlases. I did create atlases for all species in Chirgus and Burnsius, but will those "roll up" and affect existing genus-level IDs in a split like this one?

Posted by nlblock about 4 years ago

they won't roll up - the atlases are for the 12,330 IDs on Pyrgus (taxon_id: 49760) spread across 6,956 observations (e.g. https://www.inaturalist.org/observations/18634522).
Without atlases all of those IDs will be replaced with Pyrgini.

If we have atlases, any observation located within a unique atlas among the 3 outputs will have IDs of Pyrgus (taxon_id: 49760) replaced with that output (e.g. Burnsius taxon_id: 1038402). So its only really useful to the extent that the output atlases don't overlap

Thanks for pointing out that the Atlas new button doesn't show up for non species. I didn't realize that. A hack is to enter the URL manually, e.g. for Burnsius taxon_id: 1038402 use
https://www.inaturalist.org/atlases/new?taxon_id=1040783

by the way this all looks great - nice work

Posted by loarie about 4 years ago

Okay, great! That makes sense and is kind of how I figured it would work. There are definitely some areas with no overlap, so I will get to work constructing those atlases using the manual URL method. Thanks so much, Scott. I appreciate your patience and assistance.

Posted by nlblock about 4 years ago

What would be the fate of Old World Pyrgus (which are, according to the current knowledge, "real" Pyrgus, i.e., Pyrgus s. str.)? There are quite a lot of records identified to genus only, and it would not be nice if they all will be automatically switched to Pyrgini. As far as I understand from the above, creation of atlases will prevent this -- am I right?

Posted by kharkovbut about 4 years ago

Yes, I believe so. I've created a preliminary atlas for the "real" Pyrgus (https://www.inaturalist.org/atlases/26381) that obviously has no overlap with Burnsius or Chirgus in the Old World. As near as I could tell, the atlas I created encompasses all Old World sightings in iNat, so they should not be reassigned to Pyrgini (I won't pretend to know if I've covered the full range of the genus in the Old World). Let me know if you think the atlas should be updated!

Posted by nlblock about 4 years ago

that looks great nlblock -
kharkovbut, as nlblock said, ID's of Pyrgus species will be fine since those taxa will be moved from Pyrgus s.l. to Pyrgus s.s. just before this split is comitted
ID's on the genus Pyrgus through will get rolled back to Pyrgini unless they are in location that overlaps with just one of the 3 atlases nlblock is making

Posted by loarie about 4 years ago

@nlblock Thanks! One can add Pakistan, India, Israel, and Jordan. Most probably that's it. But, since there are no records for the above mentioned countries, I think it is not a big deal.

Posted by kharkovbut about 4 years ago

Sounds good! Thanks!

Posted by nlblock about 4 years ago

@loarie Thanks. If everything will work as you described, it should be fine for the Old World. :)

Posted by kharkovbut about 4 years ago

Yes definitely a problem that there aren't links to those genus level atlases available - I'll make the green 'atlased' links clickable and have the red 'No Atlas' links got to the new atlas link for that species

In the meantime, nlblock can you include links to those other two atlases here so we can check them

Posted by loarie about 4 years ago

Here are the three atlases:
Chirgus - https://www.inaturalist.org/atlases/26383
Burnsius - https://www.inaturalist.org/atlases/26382
Pyrgus - https://www.inaturalist.org/atlases/26381

There are a few gaps in the New World Pyrgus map (when compared to the real ranges of the species) to "force" IDs after the split to convert to Burnsius in those locations. I checked all genus-level individuals in those areas to make sure none were actual Pyrgus (they're quite rare/local in those states). As soon as I finish moving all the "true" Pyrgus species to the inactive genus, I think I can commit the split?

Posted by nlblock about 4 years ago

Haha, I was preparing that post before I even saw your responses! :-)

Posted by nlblock about 4 years ago

yes - if you don't mind can I make the split? Just let me know when you're ready. There's two bugs that have been happening with large splits that I've been trying to catch and it would be useful to monitor this in hopes that I can find one of them (assuming the bug happens)

Posted by loarie about 4 years ago

Yes, please do! :-)

Posted by nlblock about 4 years ago

ok great - is it ready?

Posted by loarie about 4 years ago

Yes, I think so! All the children have been moved now.

Posted by nlblock about 4 years ago

great - I'll commit this tomorrow - thanks for all the work onthis

Posted by loarie about 4 years ago

Oh dear, this is interesting...and has reverted a lot of observations to tribe level ID.

Posted by silversea_starsong about 4 years ago

Weird. My examples of Pyrgus are probably all the same species, certainly the same genus. Anything observed on or prior to Sep 30 2017 ended up either in Burnsus communis or in Pyrgini; any observation more recent than that date ended up in genus Pyrgus (inactive taxon).
I'd love to hear an explanation for that.
I don't understand what someting like this is supposed to be: https://www.inaturalist.org/observations/8003374
Is it a species or a tribe, or what? What's the level of agreement?

Posted by ellen5 about 4 years ago

I wonder if the date thing might be related to one of the bugs?

If I understand it correctly, all Pyrgus IDs in states where both "true" Pyrgus and Burnsius occur will have reverted one taxon up to Pyrgini (this is what happened in your example, @ellen5). I know almost all of those will have been Burnsius, but I don't think there's much that could have been done beyond making all the atlases go down to the county level for every Pyrgus species (which still would leave plenty of locations where both genera occur!). In states where Pyrgus does not occur, I think everything should have been converted to Burnsius.

Posted by nlblock about 4 years ago

Something changed overnight, now they've all moved to Pyrgini. So I reckon my glitch has been resolved. I can go back and re-identify as Burnsius on the strength of y'all's confidence.

Posted by ellen5 about 4 years ago

Great! I'm guessing the changes were in-progress when you checked yesterday.

Posted by nlblock about 4 years ago

At least two Palaearctic Pyrgus sp. also ended with "Pyrgus (inactive taxon)" ID. @loarie Could this happened because the authors opted out of automatic taxonomic changes application to their observations?

Posted by kharkovbut about 4 years ago

I have seen that happen before with taxon swaps. If the user rejects the community ID, their ID does not get updated with the swap.

Posted by nlblock about 4 years ago

@nlblock Thanks. But this is not because of what you have wrote: that users did not opted out of communbity ID. The cases are https://www.inaturalist.org/observations/23003570 and https://www.inaturalist.org/observations/33813262

Posted by kharkovbut about 4 years ago

Am I crazy? It looks like neither of the US (https://butterfliesofamerica.com/US-Can-Cat-1-30-2011.htm) or neotropical (https://www.butterfliesofamerica.com/L/Neotropical.htm) BoA have adopted this split. I thought everyone was wanting to more rigorously adhere to those lists for American butterflies?

Posted by loarie about 4 years ago

I suppose this is alike the case with POWO, where we follow the global research and treatments over the local ones? At least, where possible? This is what BugGuide tends to do.

Posted by silversea_starsong about 4 years ago

There is another unpleasant issue that became clear with this split. There were a number of observations that were originally (wrongly) identified to species; I placed a genus level identification with a disagreement, so the community ID was shifted to genus. After the split, the generic ID's were automatically reinstated with the new Pyrgus s. str., but the disagreement disappeared! In many cases, the community ID thus was shifted back to (incorrect) species.

In my view, after splits (or other taxonomic changes) the whole identification data should be reinstated as accurate as possible, without loss of valuable information. Probably this is a systemic issue that should be fixed.

Posted by kharkovbut about 4 years ago

example?

Posted by loarie about 4 years ago

https://www.inaturalist.org/observations/28423194
https://www.inaturalist.org/observations/17821222

In the last case, I have already added another ID, fixing the issue.

There are many more cases; hard to find them all.

Posted by kharkovbut about 4 years ago

hm - interesting. thanks for bringing that to my attention

Posted by loarie about 4 years ago

Am I right though that this was not accepted by either Pelham catalogue (https://www.butterfliesofamerica.com/L/Neotropical.htm or https://butterfliesofamerica.com/US-Can-Cat-1-30-2011.htm)? My understanding was that we were going to try to adhere to those for American butterflies. This is now a pretty significant deviation from that, and not sure how to get the word out to curators not to swap it back if they're following Pelham

Posted by loarie about 4 years ago

How long does it take for those sources to update? I don't think it has been "not accepted" rather that it just hasn't been reviewed yet.

Posted by silversea_starsong about 4 years ago

sorry I should have said - 'not currently included in' as opposed to 'not accepted'

Posted by loarie about 4 years ago

Right -- wasn't trying to be difficult or anything, was just actually curious and hoping to clarify. If only these changes happened more quickly. It may be interesting to see how BugGuide handles it.

Posted by silversea_starsong about 4 years ago

@kwillmott may have more insight into what it will take to get butterfliesofamerica.com to update. I would love to see this reflected there so we get back in sync (rather than ahead of) our reference

Posted by loarie about 4 years ago

BOA has not been updated in a couple years, as far as I know, but I might be wrong. Pelham has been updating his taxonomy a lot lately, though: http://butterfliesofamerica.com/US-Can-Cat.htm (I think you had an old link for the catalogue). I'm guessing BOA will keep up when it gets updated?

Posted by nlblock about 4 years ago

Add a Comment

Sign In or Sign Up to add comments