Apantesis vittata and the self-replicatory nature of misidentifications
I've been meaning to post on this subject for a while, and after once again working to correct misidentifications of Apantesis vittata last night, I figure now is as good a time as any.
First off, some background on this species. Apantesis vittata is a moth found in the southeastern USA. It is replaced to the north by the visually extremely similar Apantesis nais. The main differences between these species are the shape of the male genitalia, visible only by dissection under a microscope. The wing patterns are variable in both species, and there is a lot of overlap, especially in the females. In addition, the very abundant Apantesis phalerata occurs throughout the eastern USA, flying with both vittata and nais.
A typical example of a male A. vittata can be seen here: https://bugguide.net/node/view/1492279/bgimage
A typical example of a male A. nais can be seen here: https://bugguide.net/node/view/1755934/bgimage
The trend is for vittata to have reduced forewing markings and spots on the collar of the thorax, which nais lacks. However, this is only a trend, and male nais frequently have these markings, as is the case here: https://bugguide.net/node/view/1755935/bgimage
Here is a typical male of A. phalerata: https://bugguide.net/node/view/10566
Note that the hindwings have a broken black border, are pink at the base, and fade to yellow toward the outer edge. Females have pink hindwings without the yellow portion.
Females of all three of these species have reduced forewing markings and spots on the collar, such as this individual: https://bugguide.net/node/view/633456/bgimage
Based on specimens examined and dissected, A. vittata appears to be restricted to the Gulf Coast and the Atlantic Coast north to the Carolinas. Here is a summary of the species group by @neoarctia, an expert on the Apantesis: https://tinyurl.com/y6ksj78o
Now, here is the major problem with this group. Check the records on iNaturalist, BAMONA, and Moth Photographers Group, and you'll see dozens upon dozens of records of Apantesis vittata from all over the eastern USA, as far north as New England. Could all those records be wrong? How could something like that happen? Well, the answer is, yes, they're all wrong, and how it got to this point is the main reason I'm making this post.
I think it all goes back to BugGuide, one of the earliest sites with user-submitted records of moths. Prior to 2015, there were many photos of Apantesis vittata posted from all over the eastern United States. The difference listed on the page there for A. vittata vs. A. nais was A. vittata's "reduced forewing markings". Of course, females of both species have "reduced forewing markings" compared to the males, and what was actually happening on that page was male nais were being posted as "nais" while female nais were being posted as "vittata". I moved all these female nais off the vittata page in 2015, but unfortunately, the records had already made their way to Moth Photographers Group, where the dots remain to this day, suggesting that vittata occurs in places like Massachusetts, Iowa, Illinois, New York, etc. (which is wrong). Wikipedia just copies its range information from Moth Photographers Group, so it now indicates a very broad range for the species as well.
Meanwhile, with BugGuide, MPG, and Wikipedia indicating that vittata occurs much farther north than reality, Butterflies and Moths of North America (BAMONA) started accepting records of vittata from all over the place, despite some of them being dead-ringers for other species. For example, the main thumbnail on the BAMONA page here: https://www.butterfliesandmoths.org/species/Apantesis-vittata?page=2 has already been corrected to A. nais by an expert and moved to the A. nais page on BugGuide, here: https://bugguide.net/node/view/782932
In fact, nearly all the BAMONA "vittata" photos are unidentifiable at best, and clearly wrong at worst.
Check out the hindwing on this one, clearly phalerata or carlotta, and not nais or vittata: https://www.butterfliesandmoths.org/sighting_details/498894
Here's one that's clearly nais: https://www.butterfliesandmoths.org/sighting_details/902351
Another nais: https://www.butterfliesandmoths.org/sighting_details/527619
Classic carlotta: https://www.butterfliesandmoths.org/sighting_details/507248
And, as with the way they were once sorted on BugGuide, you'll notice from the thinner antennae that most of the ones on BAMONA are just females with the "reduced forewing markings" seen on females of all these species.
Of course, records "verified by BAMONA" make their way onto other websites. For example, vittata likely doesn't occur in Maryland, but check out all the classic phalerata photos listed as vittata here: https://www.marylandbiodiversity.com/viewSpecies.php?species=2496
iNaturalist now gets so many photos submitted under the name vittata that the AI suggests it left and right, even claiming "seen nearby" in places where the species doesn't exist.
So now here we are, in 2020, with all the standard online resources for moths showing a vastly over-representative range for Apantesis vittata, and the posts assigned this species name on iNaturalist popping up faster than anyone can correct them. Part of the problem is that most of the photos on iNat are just not identifiable, and people often get annoyed when you throw things back to the genus level after they think they know what species it is. But a big part of the problem though is that telling someone "vittata isn't found in your area" is a hard sell when they can look at any moth website and see apparent evidence to the contrary. A. vittata certainly might occur in places where it's not yet been recorded, but in the absence of a dissected specimen, it's hard to justify a claim of a range extension based on a photo that might be one of several species.
I don't know what the solution is to this issue, and I'm sure this isn't the only species group where this is an issue. But as someone interested in accuracy of species records, this situation is just overwhelming. Every time I decide to work through identifying these, there are just so many "vittata" to go through, and nearly all of them are either definitely wrong or probably wrong.
I'm curious what others think about this sort of problem. Is this just a feature of crowd-sourcing photo records of things that can't always be identified from photos? Is there a feature that could be added to iNat to make it harder to submit records outside the known range of a species without being really sure of your ID? All thoughts appreciated.
Σχόλια
In my defense the A. vittata on MPG I submitted years ago was from coastal South Carolina. I think I can recognize carlotta from the pure yellow hindwings, but I find it hard to get my collection past Apantesis for the remaining 3 species. Does nais and phalerata occur in the south too?
@paul_dennehy I'd like to see the summary at https://tinyurl.com/y6ksj78o but the link doesn't work. I do not have a Facebook account.
The piece of info I lack that has kept me from tackling curating these species in Maryland is when a hindwing view is needed and when it is not for a species ID. So I haven't felt confident overriding species ID with a genus ID when only a dorsal view of the FWs is shown, without HWs showing and with or without the thorax pattern in view. A simple summary of the 8 species+sex forms and which features are needed for species ID would help understand the identifiability conditions.
There is also a more general need to raise awareness of how the computer vision suggestions should never be blindly accepted as a substitute for an identification.
@rayray If it's from coastal SC, vittata is much more likely than nais. Phalerata is very common in the South, and I've had nights in Florida with many phalerata and a few vittata mixed in. Carlotta usually has all-yellow hindwings with broken black borders, but sometimes (~1/10 here in PA), the hindwings have a peachy color toward the base, reminiscent of phalerata. They're not that hard to separate if you can see the whole moth, other than separating nais/vittata near the boundary between the two species, which usually requires dissection. If you post some of your specimens and tag me in them, I'll give my thoughts on them.
@treichard If you DM me you email address, I'll email you the PDF. After brushing some abdomens and using Chris's pointers in that document, I got all my male specimens to species with a pretty high level of confidence.
And yeah, the computer vision recommendations are good sometimes, but particularly bad in cases like this.
@billhubick @jimbrighton Note the mention of MBP's Apantesis vittata page.
Thanks, @treichard - we'll work with you to get those cleaned up. Perfect example of why this live curation model is so important.
@paul_dennehy - which of the six images at MBP would you sort to A. phalerata and which would you move to genus or other pages? We'll get them moved. Covering all wild Maryland species (now 19,000+) we need to rely on the best expertise available and do very actively curate based on feedback like this, new publications, etc. We are lucky to have Tim, Hugh, and others, and now a worldwide community via iNaturalist. These are very interesting and complicated challenges.
For fun and defending a scrap of honor, MBP's iNaturalist ingest service caught a similarly out-of-range click of yours yesterday ;)
https://www.inaturalist.org/observations/51419104
Thanks for all you do. MBP ingests from iNaturalist based on a list of known authorities and you're on of them. If you ID a Maryland species in iNaturalist, it gets added or updated in our DB. Of course we need to manually curate records from outside iNaturalist.
@hughmcguinness @drkilmer - FYI
Yes, good catch on the Acrolophus! I ought to know that one by now :)
For the A. vittata page, I would call the first moth with two images A. phalerata based on the broken black HW border and light pink hindwing base fading outward to yellow. The second moth is a female that I would not feel safe taking past genus (though Chris Schmidt might be able to ID it to species). The third moth looks like carlotta to me, based on the combination of yellow hindwings and black costa. The last moth has black along the costa, which is a mark against vittata/nais, but I wouldn't feel safe calling between phalerata and carlotta without seeing the hindwings (though my gut reaction is to call it another phalerata).
I of course didn't mean to besmirch MBP; I use that page quite often to compare to my specimens from Pennsylvania. In fact, I have screenshots of the MBP identification images for Petrophila fulicalis and canadensis saved on my phone and computer to keep the two of them straight. (still waiting to get canadensis here in my home county in Pennsylvania along the Susquehanna River, but it'll happen one of these days!)
Thanks, Paul. We'll get those updated per your suggestions. Don't hesitate to contact us anytime with future corrections. Someone on this thread is almost certain to make the edit same day. Please let us know if we can ever be assistance with questions, etc., as well!
Paul, we have one more record for A. vittata at MBP. Would you mind looking at this record: https://www.marylandbiodiversity.com/record/415448
Thanks so much!
That one is a female, so I wouldn’t feel confident taking it to species. It looks like the female nais I get here in Pennsylvania though. I’d show it to Chris Schmidt if you want to be certain. If a male had those markings, it would be suggestive of vittata, but the females of nais and vittata both have the reduced FW markings, bold reddish hindwing, etc.
I can't imagine the Yale Peabody Museum Apantesis are sorted out properly between these four species.
Thank you for this fascinating account.
Why can't iNaturalist prohibit identifications of species where it is known with certainty NOT to exist? With a pop-up message or a requirement that submissions from unlikely to impossible locations must be curated?
As a learner in an area with few moth-ers to learn from, I appreciate when my ignorance is caught early & corrected.
Sandy
Προσθήκη σχόλιου