Catocala lineella/amica

A barcode phylogeny composed of jair, the new species, lineella, and amica, shows that lineella and amica are not distinguishable. Some samples in BOLD labeled lineella and amica have identical sequences, in fact. The two samples of jair that are in the same bin as lineella and amica are likely misidentified lineella/amica (not a hybrid, as mitochondrial DNA does not recombine).

Basically, the data are consistent with lineella/amica being a single monophyletic species, sister to the monophyletic lineage labeled "Catocala n. sp. Schweitzer". There's nothing within the DNA barcode data to suggest lineella and amica can be told apart, matching what Jason has described above. It's always possible that these two named species are very recently evolved and thus can't be told apart using a single mitochondrial gene, but you would need quite a bit of nuclear DNA data to confirm that.

If only it were that simple. But that's not really how BOLD works. Specimens are determined based on the DNA analysis, not prior to (other than guesses sent in by contributors, which are not included in the final determination, nor for purposes of this discussion). It almost wouldn't matter anyway given that, as I stated before, the genetic difference between the two is many times not even great enough to meet criteria for subspecific status, much less species status - it's sometimes statistically identical. (This is not to argue for single species, but to simply illustrate the difficulty separating them even on the molecular level.)

But it's not about whether they're separate species or not - they are treated that way currently and that's fine. But they cannot be determined with any significant margin of accuracy based on photos. 30 years ago it was easy to assume patterns of variation in wing maculations alone were enough to separate them, but it is not that neat and tidy anymore, now that we've had the opportunity to sequence numerous specimens of each species and pour over the data.

As such, like I said in a previous comment, there are a few who think it's "easy" to separate them by appearance (which of course I'd expect Larry to defend his descriptions), but many others who think (and know) otherwise. Personally, I would find it irresponsible to pretend we can separate them here on BugGuide with > 90% accuracy when the data plainly speaks otherwise, which is why I made the change a couple of weeks ago and combined amica/lineella into one species complex page. I'm not trying to lump them and argue for a single species, but simply created a page that more accurately reflects the unlikelihood of distinguishing (less than 50% probability of accurate distinction) one from the other by photo, sans DNA analysis (even genitalia structure is often too ambiguous between the two species to separate them).

I hope that explains why the change was made.

Jason, You are simply incorrect about how the mtDNA determinations work. If a specimen is labeled as Catocala lineella at the USNM, then it goes into the BOLD database as that species even if in fact it is Catocala amica. This kind of mistake happens all the time. There is no apriori way of determining what the mtDNA of Catocala lineella is going to be without having specimens identified by other means such as FW pattern or genitalia or larval characters. It turns out that if you actually get the ID correct based on FW pattern, then mtDNA is pretty good at separating the two species. I don't really care what you do at bugguide, but lumping all the photos into a single page does not reflect the current understanding of the taxa, which are easy to determine most of the time, and misleads people who are trying to figure out their local fauna. It would be better to have separate pages which individuals that certainly belong to each species, and then another page with indeterminate individuals.

That's not quite a proper characterization of how BOLD works. In the initial stage, when a genetic bank for a species is being developed, then yes, a specimen determined by other means in a curated collection will be granted that determination and the subsequent DNA sequence banked as that particular species. However, this only occurs in the initial stages when there hasn't yet been a sequence established for a given species. Once the sample size grows, and in this case a clear delineation between species sequence data is demonstrated (if it exists), all subsequent submissions are able to be determined by their match to the existing bank of sequences, regardless of what determination they carried prior. I've literally seen hundreds, including university specimens, of species determinations updated based on the sequence data, not simply lumped in with the pre-determined species despite the data, just because that's how they were identified when submitted. I mean if we're dealing with a species like Actias luna, then sure - ID them all by appearance and include whatever the sequence data shows in that species sequence bank; but when dealing with two species that are so extremely close genetically, it's a little more involved than that.

This entire objection reminds me of discussions that were had regarding Xanthotype species. There were a few people who steadfastly swore that distinguishing species (namely sospeta and urticaria) by photo was "easy." However, now it's nearly an identical situation to the above, where the genetic data is so close it's nearly indistinguishable itself, and DNA identified specimens show no reliable means of visually distinguishing them. The "it's easy" crowd eventually drifted into silence.

I've been hesitant for years to bring this up, primarily because of previous "it's easy" objections of the past, such as in Xanthotype and Apantesis discussions. However, this is the result of a group discussion going on for the better of three years in email exchanges between myself and a dozen other people in the field. I've given over a decade to BugGuide and I don't make changes hastily and without a lot of input, both internal and external to BugGuide. And as the discussion has gone on, and the BOLD database for these two species has grown, the case has only gotten stronger for a combined page here, lest we continue leaving specimens perpetually stuck at genus level. I know some folks who can't stand BOLD - don't trust it, and don't like how it's disrupted classic means of species determination. But when someone can lay out the criteria for visually determining amica, and then I can present 6 side-by-side, genetically identified amica/lineella pairs which each match the criteria, then the criteria isn't reliable. That's essentially all this is about, and why a combined page makes more sense than unreliable criteria to keep BugGuide looking neat and tidy while beneath the surface the DNA tells a different story.

However, having said that, I'm not going to get a headache over this. I'd just as soon change it back before I have this dragged out into another "classic determination" (wing pattern) verses "modern determination" (DNA) fight. I trust the people and authors that I've conversed with about this over the years, which is why I suggested and then made the change. But if there's a chorus of folks here that share consensus that it's "easy" to separate these two species by photo alone, then step forward and I'll revert everything to how it was previously.

Thanks for the tips. I'll tackle this soon.

It's something I've felt needed to be done for a while, but have been hesitant to initiate. Over the years, I've spoken with a number of other Catocala specialists/experts and have received a variety of opinions - one said he could sort them by photo with ease (but was somewhat dismissive of BOLD specimens that "didn't fit"), others say maybe but without certainty, and several others agree with me that there's no diagnostic criteria for visual appearance that withstands scrutiny. In the end, the data simply didn't give us any reliable means to distinguish them by photo (not to mention the nearly identical DNA issue), so I decided to pull the trigger and get some input on combining them here.

I'll probably wait a little longer before making the change, in case anyone disagrees and provides some good data that shows otherwise. But when I do combine them - this is going to sound bad - how do I do that? I don't know how to merge the pages. Is it just a matter of moving images/data, changing the amica to amica/lineella, and deleting the lineella page? Or is there a way to tag and merge them? Thanks for the help on that. :-)

There are experts on Catocala that still insist they are 2 different species that are relatively easy to ID but even if this isn't true, the way you set up the combined page is problematic. It looks like you just changed the name on the C. lineella page to Catocala amica-lineella. This explains why I knew nothing about this until yesterday when I photographed something that looks like C. amica. A new 'no taxon' page should have been created so when the images were moved there, I would have been notified of the change. To add to the confusion, there's still a page for Catocala amica.