Heads up: Some or all of the identifications affected by this split may have been replaced with identifications of Junonia. This happens when we can't automatically assign an identification to one of the output taxa. Review identifications of Junonia coenia 48505

Taxonomic Split 72690 (Committed on 15-06-2021)

Based on changes to the Pelham catalogue, which were based on this paper: https://onlinelibrary.wiley.com/doi/abs/10.1111/syen.12335

Added by nlblock on 25 February, 2020 21:10 | Committed by nlblock on 15 June, 2021
split into

Comments

@nlblock, not sure if you're still working on this, but FYI, as far as I can tell, J. grisea is not atlased: it has an atlas, but no presence places, so this split would be pretty disruptive in its current state. Unfortunately the paper you cited is closed-access, so the only people who can update the atlas would be those with access to the paper.

I'm now seeing a length discussion at https://www.inaturalist.org/flags/472648. Maybe it would be wise to put a notice in this change's description asking people not to commit it until the situation with BoA has been resolved.

Posted by kueda over 3 years ago

@loarie, here is the draft split. The children still need to be moved, though.

Posted by nlblock about 3 years ago

This seems to show a tremendous number of Common Buckeye (in the post-split sense) in CA. As I'm sure you've read, the lead author of the study doesn't believe any remain there (there's evidence of a historical population, which might have been a short-lived introduction).

Posted by snapdragyn about 3 years ago

All of the CA Commons would be re-identified as Gray Buckeyes with this split, based on the atlases not overlapping there.

Posted by nlblock about 3 years ago

Just pinging again, @loarie. From my understanding, it seems like almost all the buckeyes in AZ are Gray, so do you think it makes more sense to remove AZ and Sonora from the Common atlas? Any rare corrections to the resulting Gray IDs could be then made by identifiers later, if needed. That would leave just CO, NM, west TX, and Chihuahua where buckeye IDs would revert to the genus level.

Tagging some others who may have more input on the atlases: @hypanartia, @upupamartin, @kwillmott, @djringer, @brian_banker

Posted by nlblock about 3 years ago

What are you planning to do with Junonia coenia ssp. bergi and Junonia coenia ssp. coenia? Do they stay with Junonia coenia sensu stricto?

Also - instead of making a new Junonia coenia 1040786, lets use the input Junonia coenia 48505 as described here
https://www.inaturalist.org/blog/40417-using-a-taxon-split-input-as-an-output

This will leave all Junonia coenia 48505 ID's alone where the atlases overlap so it will be less disruptive. Does that make sense? Does that makes sense?

I already switched the taxa on the change, but Junonia coenia 48505 atlas needs an atlas for the sensu stricto sense. Can you make one? (e.g. recreate what you did for Junonia coenia 1040786)

Posted by loarie about 3 years ago

That makes sense! I think I set this up before I knew about the taxon-input-as-output method that is now preferred. I just made the adjustments to the coenia (48505) atlas.

And yes, ssp. bergi and coenia will stay with coenia sensu stricto.

Posted by nlblock about 3 years ago

nlblock this looks great. Thanks for your patience and hard work on this split. Once committed this should:
1) replace all IDs of Junonia coenia 48505 uniquely within the Junonia grisea 1021509 atlas with Junonia grisea 1021509
2) skip all IDs of Junonia coenia 48505 uniquely within the Junonia coenia 48505 atlas
3) replace all IDs in >1 or 0 atlases with Genus Junonia 48506

Please go ahead and commit. It will take a while, but once its done crunching (maybe 24 hours later) let me know if you see anything weird

Thanks!

Posted by loarie about 3 years ago

Great! Thanks so much, @loarie!

Posted by nlblock about 3 years ago

I love the way you guys work together for the good of all! Nuff said.

Posted by mokennon about 3 years ago

Thanks for your comment, Vernon. Can you explain what specific problem(s) you might have about this particular split (which has no basis in anecdotal opinions, BTW)? The papers published on Junonia in the past couple years are pretty good science, in my professional opinion as someone with a lot of professional background in systematics. However, this split was not committed because of my opinion. iNaturalist taxonomy is largely based on accepted taxonomic authorities and changes are not made willy-nilly. In this case, North American butterfly taxonomy on iNat generally follows the Pelham catalogue, which has long been recognized as a rigorous authority.

Taxonomy is in constant flux in any group of organisms; it's never "true". It's just an interpretation of a body of evidence that is constantly being added to. As you seem to disagree with how iNat taxonomy changes happen, I'd love to hear any suggestions you have about how to improve the process. iNat's success, after all, is based largely on having a community of users providing input on such matters.

And for what it's worth, if your comment "Persons making such suggestions as meaningful haven't a clue about insect taxonomy, and such suggestions are more akin to school-age minds" is somehow directed at me, you're incorrect.

Posted by nlblock about 3 years ago

@vabrou could you clarify? Are you concerns about:
1) the taxonomy that iNaturalist follows (Junonia coenia s.s. vs. Junonia coenia s.l.)?
or
2) (assuming our intent to split Junonia coenia) whether it's wise or not to use distribution to replace certain IDs of Junonia coenia with IDs of Junonia grisea (vs replacing them Junonia is prefereable)?

Posted by loarie about 3 years ago

Thanks for the work on this.

After this split the J. grisea is not showing up as an option in the computer vision suggestions within its range. Currently (in California) it shows Common Buckeye but indicates it's out of range. Does that have to wait until the latest model run to show up?

Posted by iorek almost 3 years ago

bad news is yes, good news is we're focusing on getting our model retrain interval to under 6 months and we're starting a new train this month

Posted by loarie almost 3 years ago

Thanks @loarie. Good to know. Do you happen to know if there's an enhancement for this (taxon split shows local split species in vision)? Would that be worthwhile if the training intervals are coming down?

Posted by iorek almost 3 years ago

were focusing on getting the training intervals down now before focusing on an enhancement

Posted by loarie almost 3 years ago

@vabrou, there's no such thing as a species being "in fact true and valid" because taxonomy is inherently subjective. Besides, none of that is relevant to how iNaturalist handles its taxonomy in butterflies. If you have an issue with the official taxonomy, take it up with Jonathan Pelham and his collaborators because his catalogue is the basis for iNaturalist's butterfly taxonomy for the U.S. and Canada.

Regardless, have you read the recent papers surrounding Junonia taxonomy? Because it sounds like you haven't. If you'd like copies, I'd be happy to send them to you. To imply that this work is "unsubstantiated meaningless drivel" is incredibly insulting because it couldn't be further from the truth.

You also seem misinformed about the scientific publishing process. Yes, pay walls are very frustrating, and I have lots of gripes with them, too. But that doesn't make the data "private", by any means, as anyone can access these papers using procedures like Interlibrary Loan. Or just try Sci-Hub: https://sci-hub.st/.

Posted by nlblock almost 3 years ago

Thanks everyone for keeping the conversation respectful and civil. Looking at iNaturalist as a whole, butterflies stand out as a clade with tons of attention/observations/species but no global external taxonomic reference.

In contrast, most other observose clades have a a reference (Clements for Birds, Reptile Database for Reptiles, POWO for vascular plants, World Odonata List for dragonflies, etc.)
And other clades without a reference don't have nearly as many observations or attention (eg. Bryophytes).

I think this is the source of these taxonomic issues on iNat. Since we're forced to stitch together regional lists (e.g. Pelham) and the primary literature, as nlblock said its harder to get community consensus on a single taxonomy that we all share when making obs and IDs which is what iNat depends on.

In light of the state of butterfly taxonomy, try to be patient and understanding and come with a spirit of compromise and consensus building with butterflies. Try not to take taxonomic changes personally and realize curators are doing the best we can given the messy state of butterfly taxonomy

From my perspective, it seems if we can even get consensus around tribes/genera globally by comparing the various major regional sources and bringing the players together that would be huge. Or if we could help get Pelham's changes reflected in https://www.butterfliesofamerica.com/L/Neotropical.htm that would be a huge help. I'm sure there's other ideas too...

Posted by loarie over 2 years ago

Add a Comment

Sign In or Sign Up to add comments