An annotated checklist of the tribus Parnassiini sensu Korshunov of the Old World (Lepidoptera, Papilionidae)

It is the first checklist of Parnassiini of the Old World within last 25 years. The list comprises of 54 species and 6 genera. This list takes into account all modern data about the tribus, special attention paid to the molecular investigations. The type localities of 35 species have been corrected or qualified.


Introduction
Butterflies of the tribus Parnassiini sensu Korshunov (genus Parnassius sensu Ackery) are one of the most popular (if not most popular) objects of research and collecting within Lepidoptera. There are lots of papers devoted to their systematics and nomenclature. The number of described subspecies and infrasubspecific forms is countless, it is a reason of huge amount of the synonyms and unjustified emendations as far as constantly juggling the ongoing statuses. In the last decades when the methods of DNA analysis, variative statistics and multivariative morphometric analysis started to be used widely, some of the taxonomic problems inside of this tribus were resolved (Gratton and Sbordoni 2005;Gratton et al. 2006Gratton et al. , 2008Korb 2011Korb , 2012Korb , 2013Todisco et al. 2010Todisco et al. , 2012. However, the system of this tribus in general remain the same ancient state. The problem within the systematics of this tribus is very simple: when other organisms researched by progressive methods in general, the "Parnassiologists" still keep the system of O. Staudinger (all taxa within this tribus are members of the same genus) using mainly wing pattern and refuse to use the benefits of the new techniques. This is very strange in that sense, that all phylogenetic papers clearly showing clusters within the tribus cladograms, which are in the very strong correlation to its morphological features (especially the male genitalia).
It is very important to note that within Parnassiinae no male genitalia variability have been found (Korb 2012). Thus, the genitalia structures within this group are the objects of great importance for the species determination as far as for the system building. The history of building of the system of this tribus is very rich, so I will just touch it here a little bit without the deep analysis which can be found in the last books published on this topic (Sorimachi 1995;Kocman 2009;Weiss 1991Weiss -2005Rose and Weiss 2011;Weiss and Rigout 2016). The first system of Parnassiini prepared by Austaut (1889) and based on the structure of sphragis. The next system produced by Stichel (1906) and based on the wing venation; actually it was a replica of the Austaut's system but with using another morphological structures as the system basis. At the same time the systems, based on the male genitalia and wing pattern, appeared (Moore 1902;Sokolov 1929;Bryk 1935). Ackery (1975), an author of the widely used system of Parnassiini, used the term "group" like "apollo-group", "delphius-group" etc. Finally, the most radical reformation of the system of Parnassiini performed by Korshunov (1988Korshunov ( , 1990: he erected 4 new genus group taxa, raised the status of the genus Parnassius sensu Ackery 1975 to the tribus and established the new subtribes Parnassiina, Koramiina and Sachaenina. Another great parnassiologist, Kreuzberg, used mostly the host plant preferences and trophic chemistry of Parnassiini for building his system; in his opinion, Parnassius sensu Ackery, 1975 is a monophyletic group and consists of 6 subgenera (Kreuzberg and Dyakonov 1990). The other systems proposed in the last century are in fact only reflections of the systems listed above, or their modifications (Ford 1944;Munroe 1961;Eisner 1974;Higgins 1975;Hiura 1980;Hancock 1983;Igarashi 1984;Koçak 1989). The last annotated list of Parnassiini published over 25 years ago by Häuser (1993).
The reviews of Parnassiini published recently (Sorimachi 1995;Sakai et al. 2002;Weiss 1991Weiss , 1992Weiss , 1999Weiss , 2005Rose and Weiss 2011;Weiss and Rigout 2016), but these papers are good examples of the "parnassiologist" point of view on the system of this group: the highest concentration on the subspecies and only few words about the generic (or subgeneric) structure of Parnassini. I must also mark the recently published book about the Parnassiini of Tibet (Kocman 2009), which is almost the same in the topic of subspecies composition.
Within last two decades I managed to find most of the type material of the species of this tribus; to find the type material belonging to all described species group taxa of this tribus (like subspecies or infrasubspecific forms) I see as a mission impossible: there are thousands and thousands of them, partially deposited in the private collections, partially lost or stolen, etc. This is easily illustrated on examples of several species of Parnassiini, which have been recently revised: Parnassius apollo (Linnaeus, 1758) (Glaßl 2017a) with 296 valid (!!!) subspecies and over 600 aberrations or Driopa mnemosyne (Linnaeus, 1758) (Glaßl 2017b) with 175 valid (!!!) ones.
I do not want to discuss the status of generic names within Parnassiini; genera or subgenera -depend from the researcher point of view. In the last part of the book "The Parnassiinae of the World" (Weiss and Rigout 2016) these taxa treated as subgenera with species complexes inside. In my opinion these taxa represent good genera because they form very strong clusters in the phylogenetic tree; they are representing the separate genera because they have very deep and strong differences in the male genitalia; and they are representing the distinct genera because they have very strict host plant associations. Clear advantage to use generic ranks instead of subgeneric is that no need to adopt non-systematic categories like "groups", "complexes" etc. -easy to use subgenera. The similarity in the external morphology is widely distributed and proved biological fact, and it is the only thing which supports the followers of the "monolithic Parnassius", it is very weak evidence and it must be supported by many other features; but facts are opposite.
Actually these genera can be subgenera too, this is not so much important. The main thing that we must recognize in the system is to use the correct taxa ranks but not the surrogates like "species group"; I think, it's time to adopt modern, reasonable and scientific system of this tribus. This paper is about to make it.
I must also pay attention to the highest commercial interest within this group (Rose 1995;. This is the reason that a huge number of species group taxa is described inside this group; this is the reason that the statuses of some of the described taxa are constantly changing; this is the reason that from time to time descriptions of fake species appear. In the last quarter of century the fake species are: choui Huang et Shi, 1994, liudongi Huang, 1999, cheungensis Zhang, 2010. Sometimes fake species descriptions (as far as some subspecies with very weak or invisible diagnostic features) appear in the sources very hard to obtain; for example, the description of cheungensis Zhang, 2010 remained unknown until current time, this taxon known only by several pictures from the online sale resources (like eBay). The male genitalia dissection was performed by the unified method of Stradomsky (2005).
Clarification. Ménétriès (1846) noted that the material collected by Stubbendorf in the Kansk vicinity was collected near the Khorma river; thus this locality must be stated as its type locality according the Recommendation 76A.1.2 of the Code: "Kansk environs near Khorma river".
Remark. In the book of Weiss and Rigout (2016) the taxa hoenei Schweitzer, 1912, esakii Nakahara, 1926 and tateyamai Fujioka, 1997 listed as 'semispecies' of D. stubbendorfii. Molecular data does not support this opinion (Yagi et al. 2001;Wang et al. 2019); these taxa are good island subspecies but not 'semispecies' or any other status making it closer to species than to subspecies. The wing pattern differences are very weak and unclear, the male genitalia were not figured by Weiss and Rigout (2016) and no male genitalia differences were listed. This is quite understandable: there are not any differences.
Distribution. South Siberian mountains (in the borders of Russia, China and Mongolia), Novosibirsk and Kemerovo Provinces in the West Siberian Plain, Far East (northwards to Magadan Prov., southwards to South Korea), Southern Kuril Islands, Sakhalin (Asahi et al. 1999), Japan.
Qualification. Hakodadi is a Japanese harbor located near southern limit of Yesso island.
Distribution. Russia: Vladivostok Prov., Sakhalin island; North Korea; North-East and partially Central China; Japan: southern part of Hokkaido island, northern part of Honshu island. by the designation of a neotype. As the previous Case was rejected, the nomenclature of this species is still under the strict usage of the Code, implemented by Theye (2010, 2013). I think, it is already too late to try to conserve the name Papilio phoebus in its old sense. First of all, it is already widely used by entomologists who use the Code, unlike those who are rooted into the stone age of the systematics. Secondly, in time required for the ICZN decision, its usage will significantly increase. Finally, it is not known, what the Commission will decide, but the system we are using right now.
[an]" (by the lectotype designation) (Korb and Bolshakov 2016: 41) is a subspecies of D. (E.) eversmanni. It has no differences in the male genitalia, and its COI sequences are identical to the one received from Altai, and have only one nucleotide difference from the sequence received from Alaska (the following sequences have been compared: felderi -AM231430, EF473797; eversmanni -KU875779 (Alaska), FJ663893 (Altai); all are openly accessible via GenBank)). The wing pattern and ground color features of this taxon were statistically processed by Glustshenko and Martynenko (1998), the features of the wings of this taxon and 'pure' eversmanni showing up the clinal variability and it is clearly visible on the results of the cited authors. Basing on it as far as on that fact that these taxa (felderi and eversmanni) have no COI differences, the taxon felderi is only a variable subspecies of highly variable eversmanni (Korb and Bolshakov 2016). Distribution

Parnassius apollonius (Eversmann, 1847)
TL. "in Songariae montibus" (original description); "Songarie" (by the lectotype designation) (Korb 2017a: 6) ["foothills of Dzhungarian Alatau near Lepsinsk in South-Eastern Kazakhstan" by the correction]. Correction. Until current time the only information about the type locality of this taxon is that it was collected by Schrenk in Dzhungarian Alatau (Kreuzberg 1985). The exact type locality remained unknown. Schrenk visited this area four times (1840, 1841, 1842 and 1843). In all of these trips he can collect this species, so we must analyze every trip. We know the month of collecting: June ("Volat… Junio" (Eversmann 1847: 72)). In order to determine the exact type locality of this species we must recover Schrenk' s journeys to Dzhungarian mountains in June. 1840: His expedition started to work in 11 th May in the environs of Semipalatinsk, and from June to August he was working in the lakes Balkhash, Alakol and Sasyk and in the Tarbagatai mountains (Schrenk 1845); so in this trip he was unable to collect this species. 1841: This trip was started at 10 th May again from Semipalatinsk, and then to Tarbagatai as in the previous trip; his way back started in July on the southern slope of Dzhungarian mountains (he visited the mountain gorge Tekeli) (Schrenk 1841); thus, this expedition was also not in time to collect P. apollonius. 1842: This year the Schrenk' s expedition started in May in the vicinities of Omsk, then it was moved to Petropavlovsk and in the beginning of June he was visited the Ulutau mountain (48.644173 N, 66.946388 E) in Central Kazakhstan; then on the rivers Sary-su and Chu via Golodnaya Step' to Karkaraly and then back to Omsk (Maslova 1955); thus, in 1842 this species was also not collected because Dzhungarian mountains have not been visited. 1843: It was the last trip of Schrenk in this area and it was started much earlier than three previous ones, in April. Its route was: Omsk -western shore of Balkhash lake -Khan-Tau Mts. -Ili river -Lepsa -Ayaguz -Semipalatinsk; Lepsa visited in June (Bogdanov 1875;Maslova 1955). Lepsa (now Lepsinsk in South-Eastern Kazakhstan,45.534006 N,80.613235 E) located in the foothills of Dzhungarian Alatau Mts. and according the route and timing of Schrenk' s expedition it is definitely only place where Schrenk was able to collect P. apollonius. Thus, I correct the type locality of this species: foothills of Dzhungarian Alatau near Lepsinsk in South-Eastern Kazakhstan.
Distribution. Mountains of Central Asia and South-Eastern Kazakhstan, South-East Altai (within the borders of Kazakhstan), South Kazakhstanian lowlands (local), rocky shores of the rivers Ili and Charyn (including Chinese parts). Distribution. Pamir-Alai, Gissaro-Darvaz, East Badakhshan.

Parnassius bremeri Bremer, 1864
TL. "Amur Ufer" (by the lectotype designation) (Korb 2016: 79) ["Oldoi river shores on its connection to Amur river" after correction]. Correction. In the original description the following information on its type locality is stated: "an der Mündung des Oldoi, an der Dseja und im Bureja-Gebirge; …am Ussuri, von seiner Mündung bis zur Ema" (Bremer 1864: 6). Thus the only Oldoi river coast which is connected to Amur river can be the shores of Amur as it is stated on the lectotype label. According to these data the type locality of P. bremeri must be corrected to «Oldoi river shores on its connection to Amur river» (53.554722 N, 123.329444 E).
Remark 2. The status of taxon sacerdos Vorbrodt, 1912 which by some authors was treated as a separate species (Häuser 1993) according to modern phylogeographic studies is a subspecies of P. corybas (Todisco et al. 2012).
Distribution. European mountains, Urals, mountains of Siberia and Far East. Its record from Transcaucasus (Nekrutenko 1990: 56) based on two old specimens (collected in 1836 and labelled "Adshara") is most likely erroneous.
Remark. The status of this taxon was raised by Michel et al. (2008) basing on a single COI sequence. It was confirmed by Todisco et al. (2012), but also basing on the single COI sequence. The lectotype of this taxon was designated by me (Korb 2017a: 6), it combines features of both P. corybas and P. actius (Eversmann, 1843).
Distribution. Known only from the type locality.
Correction. This species was not recorded from Tarbagatai as far as from Southern Altai and surrounding areas (Toropov and Zhdanko 2014). Thus both its type localities, recovered from its original description (Southern Altai) and from the lectotype designation (Tarbagatai) are incorrect. The specimens used for the description have typical Dzhungarian 'face' and without doubts were collected in Dzhungarian Alatau Mts. There are two evidences for it. Firstly, the type specimen were collected for sure not in 1843: the journal where it was described was ready for publication in 16 th June, 1843 (the Imperial Censorship Committee Resolution printed on the first page of the 3 rd number of the Bulletin de la Societe Imperiale des Naturalistes de Moscou with this date) (the species flies in End of June -July). Secondly, the only expeditions reached the mountains where P. actius occur (Dzhungarian Alatau Mts.) till 1842 are the expeditions of Schrenk (Maslova 1955) (see in details: Parnassius apollonius). Most likely the types of P. actius collected in the second trip of Schrenk when in July (it is actually flight period of P. actius) he started his way back for wintering. His expedition was working 12 days in Tekeli gorge of Dzhungarian Alatau Mts.; there are no other expeditions before 1842 which hits right time to collect this species. Thus, I correct the type locality of P. actius: Dzhungarian Alatau Mts., Tekeli gorge 15-20 km SE of Tekeli settlement.
Parnassius epaphus Oberthür, 1879 TL. "Tibet" (by the lectotype designation) (Ackery 1973: 5) ["Ladak in Himalayas" after correction]. Correction. The name epaphus proposed for the specimens listed by Gray (1853) as P. jacquemontii Boisduval 1836 from "Chinese Tartary at an elevation of 15,000 feet". Under the name Chinese Tartary in the English entomological literature of 18-19 centuries meant Ladak but not Tibet. Elwes (1886: 36) wrote: "I have been able to find out the route which Major Charlton followed in Ladak, or Chinese Tartary as it was called in those days…". Thus the type locality of P. epaphus is Ladak but not Tibet; additional argument to this conclusion is that Major Charlton never did such a long trips, all his trips were limited by the northern and north-eastern limits of the Indian parts of the British Empire between 1825 and 1840 (Paget 1874(Paget , 1907. Guided by these two reasons I must correct the type locality of P. epaphus as "Ladak in Himalayas".

Parnassius dongalaicus Tytler, 1926
TL. "Donga La" (a mountain pass located on the highway Bumthang -Ura in Bhutan,27.394808 N,90.9953358 E). Remarks. It described as a good species by a single female which now deposited in the British Museum. For the long time its status was questionable after the comment of Bryk (1935: 279): "Nach der gekielten Sphragis der weiblichen Type gehört diese Unterart nicht zu Р. epaphus C. Oherth., aber ihre Facies ist so epaphusartig, daß eine Begattung mit dieser Art wahrscheinlich ist, zumal sogar Begattung zwischen sehr weit voneinander entfernten Arten wie Lingamius hardwickei (I.E. Gray) und P. epaphns C. Oberth. festgestellt wurde" [By its sphragis equipped by a keel, the type female of this subspecies cannot be Р. epaphus, however its fascia looks so close to P. epaphus, so it can be a result of crossbreeding between very distant species like Lingamius hardwickei and P. epaphns]. The species rank was returned to it by Weiss (2005) and supported by Kocman (2009).

Parnassius mercurius
Remark. The superspecies P. jacquemonti was recently redrawn by Weiss (2005) and Weiss and Rigout (2016): taxa actinobolides Bang-Haas, 1928, rubicundus Stichel, 1906, tibetanus Rühl, 1893 and mercurius were raised to the species rank. I support this opinion partially: comparison of COI sequences of the closely related taxa actinobolides, tibetanus, rubicundus and mercurius showed up that 3 of them have not enough differences (p-distances less than 1%) to treat them as good species (Korb 2012); the only taxon which COI sequence differs enough from all of the above listed taxa and from P. jacquemonti is mercurius.

Parnassius tianschanicus Oberthür, 1879
TL. "Forêt de Kouldja et du Tianschan" (original description) ["East Tian-Shan between lake Sairam-Nur and city of Kuldja" after correction]. Correction. "Forêt de Kouldja et du Tianschan" -this locality pointed to the Chinese part of Tian-Shan (East Tian-Shan between lake Sairam-Nur and city of Kuldja). Thus the type locality of P. tianschanicus must be corrected to "East Tian-Shan between lake Sairam-Nur and city of Kuldja".
Distribution. Mountains of Central Asia.
Remark. Very variable species with many subspecies and aberrations described. Some of them (richthofeni Bang-Haas, 1927, koiwayai Ohya et Inomata, 1988, gabrieli Bryk, 1934 and others from the nomius Grumm-Grshimailo, 1891 clade) looks like different species but according its DNA and male genitalia they are without doubts the same species.
Correction. The place where Linnaeus lived in Sweden was Uppsala; main part of his collection was collected here (Honey and Scoble 2001); most likely the type specimens of P. apollo too. The materials Linnaeus collected "near his home" were basically without the geography labels, including the lectotype of P. apollo (Honey and Scoble 2001: 298). Due to these reasons I must correct the type locality of this species to "Sweden, near the city of Uppsala".
Remark. Very variable species, hundreds, if not thousands, species group taxa were described. The modern book about this species (Glaßl 2017a) listed 296 subspecies of it.

Tadumia hardwickii (Gray, 1831)
TL. "Nepal" (by the syntype) (Ackery 1973: 15). Note. The species described "…from the collection of General Hardwicke…" (Gray 1831: 32) from Nepal. It is no possibility to recover the exact type locality of T. hardwickii; General Hardwicke was a great English naturalist and he did many trips to the English India and surrounding areas which are not described good enough to determine the right one (Dawson 1946).
Distribution. Himalayas, southern and western limits of Tibetan plateau.
Correction. See P. epaphus. Note. Highly variable species. Many forms and subspecies have been described; some closely related species are in the "acco"-complex. The molecular and morphological data used to delineate the species inside of this complex group (Condamine 2018). I support Kocman (2009) in his opinion that the taxa treated before as the good species przewalskii (Alphéraky, 1887) and baileyi (South, 1913) are subspecies of T. acco.
Remark. In the last part of "The Parnassiinae of the World" (Weiss and Rigout 2016) mentioned 'Parnassius cheungensis Zhang, 2010' which considered to be a synonym of T. acco in the same book.
Distribution. Tibet, Karakorum, North-West Himalayas. Qualification. Chumbi valley (27.383 N,88.883 E) located on the borders of India, China and Bhutan; several authors erroneously placed the type locality of T. hunnyngtoni to the slopes of the Everest Mt., but the Chumbi valley is 200 km away from this mountain in the south-eastern direction.
Distribution. Southern Tibet in the borders of Nepal, Sikkim and Bhutan.

Correction. See Parnassius actius.
Remark. Recently the conclusions of Kreuzberg (1985) regarding the systematics of the delphius-group were critically rearranged (Churkin 2009;Lukhtanov et al. 2016); it was proved that the taxon maximinus is not a good species. The situation with the type material of maximinus is entangled. Tshikolovets et al. (2016: 83) listed the «lectotype ♂» from Museum für Naturkunde, Berlin, Germany, referring to Bergmann (1995). However, Bergmann (loc. cit.) did not publish any lectotype designation; he just wrote that the lectotype was designated by Kreuzberg in 1985. Kreuzberg published no designation of the lectotype of maximinus (Kreuzberg 1985: 45); according its labels, he selected the lectotype in 1989 but this designation has not been published too. Thus, both Bergmann and Tshikolovets et al. treated as a lectotype one of the syntypes; the lectotype designated by me (Korb 2017b: 12).