New Taiwan machaon paper

[adamcotton]

https://doi.org/10.1371/journal.pone.0310318 (open access)

Nazari, V., Yen, S.-H., Hsu, Y.-F., Shapoval, G. Shapoval, N. & V. Todisco 2024. Wiped out by an earthquake? The ‘extinct’ Taiwanese swallowtail butterfly (Lepidoptera, Papilionidae) was morphologically and genetically distinct.

Abstract
For the first time, we obtained for the first time a COI DNA barcode from museum specimens of the Old World swallowtail butterfly endemic to Taiwan, Papilio machaon ssp. sylvina, that has disappeared since the devastating Jiji earthquake in 1999 that shook Central Taiwan. We demonstrate that this population was not only phenotypically distinct, but also had a unique mitochondrial haplotype among all other Holarctic populations of P. machaon. The life history of P. m. sylvina from rearing experiments carried out in the 1990s is illustrated and discussed.

Adam.

[Chuck]

Thanks Adam. I presume that there’s been sufficient field work to believe it is extinct?

[adamcotton]

It certainly seems so. People have looked for it in the highlands of central Taiwan where it was known from but none have been seen since 1999.

It is of course possible that it is still there, but it's strange that no-one has seen it.

Adam.

[wollastoni]

Very interesting article. Thank you Adam for sharing it here.
Always hard to understand how the earthquake could wipe out the whole population.

I hope it will be found again one day... but it's very unlikely a machaon could avoid all the Taiwanese nature photographers since 1999 so I guess it is extinct.

[kevinkk]

It does seem hard to believe until you read that landslides caused the apparent disappearance. 25 years is a long time to be certain, especially on a populated island.

[Chuck]

It does seem hard to believe until you read that landslides caused the apparent disappearance. 25 years is a long time to be certain, especially on a populated island.

Wasn’t E Atala “extinct” for a while?

As far as “populated island” I ran into an exact analogy yesterday at the National WW2 museum, which showed 1942 Guadalcanal in all red, controlled by the Japanese. Totally wrong.

Both Guadalcanal and Formosa are mostly high mountains and difficult terrain, with the urban populations compacted on the coast. Ditto OAhu and Maui and Kauai. The interiors of mountain islands are very lightly tread, and when they are penetrated it’s not by entomologists.

[wollastoni]

Yes, but unlike in Guadalcanal, there are thousands of naturalist photographers in Taiwan (+ tourists) who often go to the mountains to take wildlife pictures. I doubt they would miss a machaon. It is not a small Lycaenid or a moth... it is a big yellow butterfly who loves nectaring along paths.

[adamcotton]

Actually Taiwan's mountainous interior is relatively easily accessible, and Prof. Hsu (one of the co-authors) spends a lot of time exploring the highlands, mainly for Lycaenidae.

Adam.

[Chuck]

Actually Taiwan's mountainous interior is relatively easily accessible, and Prof. Hsu (one of the co-authors) spends a lot of time exploring the highlands, mainly for Lycaenidae.

Adam.

While unarguably SOME parts of Taiwan's mountains are readily accessible- a great joy I have enjoyed- just drive up- that is not true of all of Taiwan.

Look at the iNat records for southern Nantou county- so very few, despite a population of millions of people. https://www.inaturalist.org/observation ... n_id=47157

Anecdotally, the Bog Buckbean Moth, Hemileuca nee-maia, in NY wasn't formally announced until 1978 despite proximity to a campground and state park- because nobody wants to go into a bog. Similarly, John Tennent insisted Polyura jupiter did not occur on San Christobal, until he took my fresh female from me. And, two years ago I was chasing a yellow-and-black swallowtail (bairdii? indra? zelicaon?) in Bourne TX, an exploding suburb of San Antonio; surely someone would have previously seen such a thing, yet there are zero records of any of those taxa for 500 plus miles.

I'm not saying that machaon exists on Taiwan, nor do I know if the conditions and foodplant exist in the relatively unexplored areas, just that Taiwan does indeed have very poorly surveyed areas.

[adamcotton]

Let's hope that it is still there somewhere.

Adam.

[Chuck]

Let's hope that it is still there somewhere.

Adam.

I've learned to spend more time on host plants to find particular leps. In the case of P m sylvina, it has one host plant, Peucedanum formosanum.

Looking at iNaturalist at the same region I formerly pointed out is poorly surveyed, I see that Peucedanum formosanum is recorded from the fringes of that area- specifically, roadsides. Presumably it occurs elsewhere within those mountains.

The paper shows a map of historical records of sylvina, but not which of those records are more recent (e.g., since 1990.) I am not well read on this butterfly, though I have read repeatedly it was wiped out by a landslide, which infers there was but one known remaining population. All the same, the millions of highway iNaturalist picture-takers are highly unlikely to find P m sylvina even if it does still exist, nor would the typical 50+ YO Lep specialist- it would take some very long, uncomfortable ground pounding to say with any certainty that it does not exist elsewhere.

The problem is that science remains where the paper closes: "one can always hope that it still persists in the remote mountain regions in the Taiwan highlands." Yet, Hope is not a strategy.

[Wu Ｍing Hsuan]

Interestingly, based on the COI analysis in the paper, verityi and archias fall within machaon clade rather than being distinct species.

[adamcotton]

Unfortunately the P. machaon group is one of the species groups in which the results of COI analysis are problematic. In a tree of many samples the various populations are rather jumbled up. There are similar issues with other species groups such as Graphium sarpedon. Genomics will hopefully improve clarity on the situation, and what I have seen recently seems to place verityi in a very different position to COI alone.

In this particular paper relatively few samples were included in the final tree, which makes the resulting tree look better. It will be interesting to see what more comprehensive molecular analyses find in future, but unfortunately genomic analysis is relatively expensive.

Adam.

[Chuck]

Interestingly, based on the COI analysis in the paper, verityi and archias fall within machaon clade rather than being distinct species.

What Adam said is correct. With COI the higher number of specimens will sometimes reveal significantly where they land on the tree (i.e. relationship to each other within a taxon, and to other taxa.) Plus it depends on other factors- quality of the MtDNA sample, when and how it was analyzed, and which section of COI was used. On the latter, it was common in the past to use COI 3' section, while today COI 5' section is more common; the two do not overlap so it's somewhat like looking at the top 2' of a person and then the bottom 2' and trying to determine relationship; in any event, it's common that the trees developed from 3' and 5' differ remarkably.

Then SNPs will typically provide yet another "snapshot" of taxonomic relationships, which also is likely to differ from COI trees.

As I am often reminded, COI is "ONE ASPECT" of taxonomic determinations. Plus, as Adam has said, some animals are less inclined to be well defined based on COI.