Tricky one to ID

[David M]

Is this Pale or Berger's Clouded Yellow?

Seen near Lake Como, Italy:

[bugboy]

I've zero experience of either but the roundedness of the forewings would point towards a Bergers, assuming they're the only 2 candidates, this is based using Pete's latest book.

[selbypaul]

I've previously read that they are virtually impossible to tell apart without having DNA or seeing the caterpillar that led to the butterfly.

However, having read Pete's book over the weekend, supposedly the roundedness of the forewings is one of a few factors that can lead to a 95% chance of successful ID. On that basis, it is more likely to be Berger's Clouded Yellow. But others will I'm sure have views.

[John Vergo]

For me it looks most likely like a Berger´s, do to the roundeness of the forwing apex, and the hindwing ground colour is more warm yellowwish than the C. hyale

[petesmith]

Berger's for me too. For the same reasons as others have mentioned.

[Padfield]

To all appearances, Berger's. I've seen Berger's with the wingshape of pale clouded, but I've never seen pale clouded with the wingshape of Berger's - and this definitely has the wingshape of Berger's.

Habitat is a better clue - and even more, seeing what females are laying on. Berger's is a sedentary butterfly, flying where its foodplants (most commonly, horseshoe vetch) are found. It flies in the company of other species with the same foodplant and is far and away the commoner of the two in the Alps and further south. Pale clouded is much more of a migrant. I find it most reliably on transient fields of clover. Both males and females are strongly attracted to clover. But beware, of course, that clover often flowers in Berger's habitat and this species too will nectar from it.

Guy

[Benjamin]

Just to add this image that was included with the other two when the images were posted in our WhatsApp group.

It seems to cancel out the roundness suggested from the other angles, but still points towards Berger’s in some regards. I spent a chunk of time trying to test the various suggested clues against reliably identified photos but didn’t arrive at anything in any way solid. I think Roger’s summary on his hyale page is excellent.

Interestingly I’ve been experimenting with ObsIdentify recently with some pretty mixed results but with hyale/alfacariensis it seems to be very accurate!
I’ve tested it against my own photos of both species (IDs by location) and also against a good number of Guy’s and it has been correct every time, from undersides as well as uppers.

ObsIdentify considers all 3 of these images to be hyale with a certainty of >95%. Make of that what you will!

[David M]

Thank you all for your input. As the thread title says, this is a tricky one to have a degree of certainty about and given I have zero experience with hyale I thought it best to seek the opinions of others.

Looks like humans and Obsidentify disagree with each other!

[Benjamin]

Actually - scrap that - I’ve just checked some more of both Roger’s and Guy’s images and the app is pretty hit and miss. This fits in with my previous experience of being presented with wrongly identified images by people relying on apps.

[Padfield]

The problem with AI apps for ID is that they’re only as good as the data set (image set) they learn from. Some friends of mine have been preparing an ID app for Swiss butterflies. They have procured literally hundreds of thousands of images from keen amateurs like me but are nevertheless going through every single one to confirm ID (or reject the picture) before feeding the machine. Every so often I get an e-mail saying ‘did you see the uns of this butterfly?’ or ‘Is it possible this is such and such?’ I wonder what set of images Obsidentify was fed with, and who guaranteed the ID of each input …

For what it’s worth, I still incline to alfacariensis from the upperside image. The orange spot looks bright and there is little dark shading at the base of the wings. Yes, the butterfly is on lucerne (I think that’s lucerne), a foodplant of hyale, but everything nectars at lucerne! I saw alfacariensis nectaring at it only yesterday.

Guy

[Roger Gibbons]

I would not rely on an ID app that was based on images submitted by unconfirmed sources. You only have to put a species name into Google to see that quite a selection the images are not correct, and this is true even on sites that have images for licencing.

The CEN-PACA Atlas took a more rigorous approach by having a validation committee (I was on it, so could see first-hand the rigour applied) to filter out the dubious.

I had very little confident experience of hyale as it doesn’t fly in the PACA region of south-eastern France, but last year I chanced on a Lucerne field in central France which had around 50 “hyale” flying there. Many clearly were, as the females were egg-laying and there were many examples of courtship, and nearly all fitted the “classic” ID clues for hyale. However, it is not impossible (as Guy says) that some were alfacariensis, although I didn’t see any that definitely weren’t hyale. I went back there this year and it was still a thriving colony.

I looked into all the published ID clues which I outline on my hyale page, link as below (simpler than reproducing the text here). I would be interested in anything that could be added (or subtracted).

https://www.butterfliesoffrance.com/htm ... 0hyale.htm

FWIW, for me the degree of curvature of the forewing margin points strongly toward alfacariensis.

Roger

[David M]

I have been asked to add a further image and to state that the butterfly was flying in a field full of lucerne:

Of course, this doesn't mean that it must be hyale by default.

[Benjamin]

Yes - I think Guy’s already covered this.

Rearing lots of both and having a really good look has been added to the To-Do list!

Tim Cowles wrote up an interesting study of the immature stages but unfortunately I can’t seem to get his site to load currently.

[Padfield]

Re flying in a field of lucerne: this would be a strong indicator if (a) there were lots of potential hyale there and (b) if the field wasn't embedded in alfacariensis country. Otherwise, there's nothing to stop alfacariensis nectaring there. In my experience, if there is a good potential breeding spot for hyale - a clover field or lucerne field - they will find it en masse, and do so year after year, just as long-tailed blues always find the same spots year after year.

From all the pictures posted, I wouldn't actually have considered hyale - they all look really like alfacariensis.

I photographed this alfacariensis on lucerne yesterday (in Leysin). The clues? There's lots of horseshoe vetch along the track; alfacariensis is the commonest local Colias; Adonis and chalkhill blues fly there; it looks like alfacariensis.





Guy

[Roger Gibbons]

I sent Tim's life-cycle paper to Benjamin via a pm a few days ago, but when I just checked it hadn't gone because my outbox was full, so re-sent and attached here as it may be of wider interest. I think Tim's site may be unlikely to be restored with Orange, but may resurface with another host.

Colias hyale_life cycle.pdf

(853.49 KiB) Downloaded 24 times

Roger