An updated checklist of the European Butterflies (Lepidoptera, Papilionoidea)

Abstract This paper presents an updated checklist of the butterflies of Europe, together with their original name combinations, and their occurrence status in each European country. According to this checklist, 496 species of the superfamily Papilionoidea occur in Europe. Changes in comparison with the last version (2.6.2) of Fauna Europaea are discussed. Compared to that version, 16 species are new additions, either due to cryptic species most of which have been discovered by molecular methods (13 cases) or due to discoveries of Asian species on the eastern border of the European territory in the Ural mountains (three cases). On the other hand, nine species had to be removed from the list, because they either do not occur in Europe or lost their species status due to new evidence. In addition, three species names had to be changed and 30 species changed their combination due to new evidence on phylogenetic relationships. Furthermore, minor corrections were applied to some authors’ names and years of publication. Finally, the name Polyommatusottomanus Lefèbvre, 1831, which is threatened by its senior synonym Lycaenalegeri Freyer, 1830, is declared a nomen protectum, thereby conserving its name in the current combination Lycaenaottomana.


Introduction
Butterflies constitute one of the best-known groups of insects and have become important models to study speciation, community ecology, biogeography, climate change, and insect-plant interactions. With close to 19,000 described species [18,768 presumably valid species recorded by 2011; that figure is higher today, i.e., ca. 19,000 species], they represent about 12% of currently known species of Lepidoptera (Van Nieukerken et al. 2011). According to current molecular systematics (Mutanen et al. 2010;Heikkilä et al. 2012;Espeland et al. 2018), the single butterfly superfamily Papilionoidea comprises 7 families (Table 1, Fig. 1) and includes the Hesperiidae (skippers) and Hedylidae (moth butterflies). The skippers have previously been thought to represent the sister group to the butterflies and were often placed in a separate superfamily Hesperioidea, but the molecular results indicate that the family Papilionidae is the sister to the remaining butterflies, which also include the small Neotropical family Hedylidae with only 36 species. Apart from the latter family, all butterfly families are represented on all continents except Antarctica, although most species of Riodinidae are confined to the Neotropical Region. Butterfly diversity is particularly high in the tropics, especially the Neotropics, and only 496 species are found in Europe according to the present checklist.
The taxonomy of butterflies started in 1758 with the Swedish naturalist Carl von Linné (Latinised to Carolus Linnaeus), who introduced binominal nomenclature and described the highest number of European butterfly species, all of them in a single genus Papilio. Seventy-one of them currently still hold the names given by Linné, albeit mostly in different genera. Other authors who described many new species during the 18 th century were the German entomologists Eugen Johann Christoph Esper and Jacob Hübner, the Danish entomologist Johann Christian Fabricius, as well as the Austrian lepidopterist Johann Ignaz Schiffermüller (the latter in an anonymous publication usually referred to as [Denis & Schiffermüller], but see Kudrna and Belicek (2005), Sattler and Tremewan (2009) and  for a controversial debate on this topic). By 1820, half of the European butterfly fauna had been validly described, and species were placed in a growing number of genera (starting with Hesperia Fabricius, 1793 as the second-named genus for the skippers). During the 19 th century, more than 60 European lepidopterists continued the inventory of Europe's butterfly fauna, and the first overview of Palearctic butterflies (and other Lepidoptera) was published by Seitz (1907Seitz ( -1909. At that time, already 90% of Europe's butterfly species had  been described and the rate of newly discovered species slowed down (Fig. 2). Another milestone for butterfly research in Europe was the field guide of , which included distribution maps of Western Palearctic butterflies, and led to a growing interest in butterflies across Europe. This field guide was also translated into other languages (e.g., German, French, and Spanish) and updated several times (most recently by . However, despite their somehow misleading titles, these guides excluded large parts of eastern Europe (i.e., Belarus, Ukraine, Moldova and most of Russia (apart from Kaliningrad enclave) and therefore all the species from the Ural mountains). The proliferation of butterfly field guides by various authors across Europe also led to an increasing confusion of butterfly nomenclature due to different taxonomic concepts. The first step to standardize European butterfly taxonomy and the precursor of our list was the book (and accompanying CD) by Karsholt and Razowski (1996). It constituted a country-level checklist of all European Lepidoptera, but excluding the Mid-Atlantic islands (i.e., Canary Islands, Madeira, and Azores) and contained 440 butterfly species. This book was also the basis for the list of Lepidoptera in the online database Fauna Europaea, a project under the auspices of the European Commission, which started in 2000 (De Jong et al. 2014) and aimed to provide checklists for all European animal species. This database, which went online on 16 December 2004, also included Cyprus and the Mid-Atlantic islands, which are hotspots of narrow endemics. At about the same time, the first distribution atlas of all European butterflies was published by Kudrna (2002), and finally a butterfly field guide appeared which covered most of the West Palearctic region including all of Europe . The last comprehensive update of the butterfly checklist in Fauna Europaea happened 7 years ago (Karsholt and Nieukerken 2011), and the checklist presented here was first developed as an update to the online database. Unfortunately, funding for Fauna Europaea was discontinued after the initial 4-years funding period and the outdated Fauna Europaea website was only saved due to the commitment of the Natural History Museum in Berlin that set up a new one. However, its functionality is still very limited and the update process severely hampered due to shortage of funding. For this reason, we decided to publish this updated distributional checklist in order to address the need of the lepidopterological community and the public at large. It intends to cover the significant progress in butterfly systematics and faunistics, which was brought about in particular by the advancement of molecular methods.

Materials and methods
This updated checklist is based on the last version of Fauna Europaea (2.6.2). This version is almost identical to the most recent Lepidoptera update in version 2.4 (online on 28 January 2011) but includes some emendations by the staff of the Fauna Europaea office in Berlin that had not been approved by the Lepidoptera group coordinators (Erik van Nieukerken and Ole Karsholt). The geographic area covered remains the same: It includes the European mainland to the eastern slopes of the Ural mountains, plus the Macaronesian islands (excluding the Cape Verde Islands) and Cyprus, with the Caucasus and western Kazakhstan excluded (Fig. 3). Included are the British Isles and all Mediterranean islands under European administration, as well as the Greek offshore islands along the Turkish coastline. Iceland has no native butterfly species. Distributional information is based on political units at country level as in Fauna Europaea, following the ISO-3166 code. However, with the exception of the Macaronesian Islands, the additional regional splits of several countries in Fauna Europaea (mainly for Russia and some island territories) were not adopted.
The following categories are used to explain the distribution: A Absent (never recorded in the respective country or island group or only doubtful records) P Present (native or well-established populations, including alien species such as the South African Cacyreus marshalli) P?
Possibly present (recorded but continued presence doubtful; usually these are species with range limits near the border of the respective country) M Regular migrant (species which has no permanent populations, e.g., because it cannot overwinter, but is observed almost every year; included are extinct species if they are still observed as regular migrants) I Irregular vagrant (irregular vagrants or introductions which do not reproduce or only irregularly, including temporary or recently established populations) Ex Regionally extinct (native species which have become extinct, even though vagrants might be seen occasionally) It should be noted that the "Extinct" category is used in a rather strict sense, in line with the IUCN Guidelines which demand that exhaustive surveys have been undertaken to prove that 'there is no reasonable doubt that the last individual has died'. In some cases, this has led to species being recorded as "Present", even though they are most probably extinct, e.g., Colias myrmidone in Austria (no proof for more than 25 years; H. Höttinger, pers. comm.). In addition, some of the national Red List Assessments are already outdated, even though attempts have been made to update those. An example for an update is the status of the Madeiran endemic Pieris wollastoni, whose last reliable record is from 1986. It was classified as "Critically Endangered (Possibly Extinct)" in its last Red List assessment ( , but is now classified as "Extinct", because extensive surveys in recent years have failed to prove its continued presence. This is the only European butterfly species which is known to have become globally extinct in historical times. According to the concept of Fauna Europaea, changes were only carried out if supported by newly published research. This restriction helps to stabilize nomenclature, but can also lead to inconsistent results, e.g., due to the retention of some weakly differentiated taxa, whose species status is questionable, but for which no new published evidence is available. Potential examples in our list are Lysandra caelestissima (Verity, 1921), Polyommatus nephohiptamenos (Brown & Coutsis, 1978), Hipparchia neapolitana (Stauder, 1921), Hipparchia sbordonii Kudrna, 1984, Satyrus virbius Herrich-Schäffer, 1844, and Pieris balcana Lorković, 1969 The main criterion whether to include or exclude a species taxon based on new (and possibly contradictory) publications was evidence for species status from at least two character sets, e.g., mitochondrial as well as nuclear DNA, or differences in morphology and karyology.
Nomenclatural changes are annotated with reference to the sources and strictly follow the last (fourth) edition of the International Code of Zoological Nomenclature (ICZN 1999). This includes the controversial article 34.2, which mandates that »the ending of a Latin or Latinised adjectival or participial species-group name must agree in gender with the generic name with which it is at any time combined«. Due to its linguistic complexity, this rule has led to many wrong or ambiguous decisions and causes additional instability of nomenclature each time a species name is transferred to another genus. Therefore a majority of lepidopterists, including the group editors of Fauna Europaea, have decided to ignore this rule and use the original spelling instead (de Jong et al. 2014). Difficulties with the gender agreement rule in Lepidoptera are as old as binominal nomenclature, because there is not even an agreement about the gender of the genus Papilio. Therefore Carl von Linné used nouns as species names and avoided the use of adjectives (Welter-Schultes 2013). However, for easy reference to Fauna Europaea and other databases, we also list the original ending and compiled a comprehensive list of original combinations, using various sources such as the LepIndex (Beccaloni et al. 2003), PESI (2018), FUNET (Savela 2018) and . In case of doubts or discrepancies, the original publications were checked as well.
In a few cases, necessary changes due to new nomenclatural findings have not been carried out yet, because they would result in the replacement of a well-established name by an (almost) unknown synonym. Such cases should be referred to the International Commission on Zoological Nomenclature for ruling, and changes implemented only after a decision has been made by the Commission. One such case is the wellestablished name Parnassius phoebus, which has turned out to represent another Asian Parnassius species which is currently known as Parnassius ariadne (Lederer, 1853) (see Hanus and Thèye 2010) and would thus need to be replaced. After the first attempt to preserve this name (Balletto and Bonelli 2014) failed (ICZN 2017), a second proposal has recently been submitted to the Commission (Lukhtanov et al. in press). According to article 82.1 of the code, prevailing usage has to be maintained until the case has been decided by the Commission.
An exceptional case which would cause a large number of changes in the names of Lepidoptera are many of the names published by [Denis & Schiffermüller] (1775) which are lacking a sufficient description, but have already been used for a very long time. In accordance with the opinion of the Fauna Europaea editorial team, we have not replaced these names. The effect on butterfly taxonomy would be rather marginal, however, because only one butterfly species would have to change its name (Nymphalis vaualbum to Nymphalis l-album (Esper, 1781)) and five others only their authorship, see Kudrna and Belicek (2005). We are looking forward to a decision of the ICZN to solve this matter (see . Another case concerns the genus name Muschampia Tutt, 1906 (type species: Papilio proto Ochsenheimer, 1808; currently known as Muschampia proto (Ochsenheimer, 1808)), which appears to be a subjective synonym of the genus name Sloperia Tutt, 1906 (type species: Hesperia poggei Lederer, 1858; currently known as Muschampia poggei (Lederer, 1858)). Both genus names were published in the same paper and Hemming (1967) was the first to note that Sloperia should have precedence over Muschampia, because Warren (1926) as the first reviser chose Sloperia. However, the name Muschampia has remained in prevailing use during the last 90 years and, in addition, there is evidence from molecular data (Wiemers et al. unpublished) that the current classification of the species presently placed in the genera Carcharodus and Muschampia needs to be substantially revised. However, molecular data are still missing for most of the (mainly Asian) species currently placed in Muschampia, and therefore we suggest to postpone a rearrangement until better data become available.
Finally, one of us (GL) discovered that Polyommatus ottomanus Lefèbvre was published in 1831 (and not in 1830) and therefore has to be regarded as a subjective junior synonym of Lycaena legeri Freyer, 1830. This would mean that the well-established name of the species currently known as Lycaena ottomana (Lefèbvre, [1831]) would need to be changed to a name which has not been used for this species during the past century. However, according to article 23.9.1 of the Code, the prevailing usage must be maintained when the senior synonym (i.e., legeri Freyer) has not been used as a valid name after 1899 (article 23.9.1.1), and the junior synonym has been used, as its presumed valid name, in at least 25 works, published by at least ten authors during the last 50 years and encompassing a span of not less than ten years (article 23.9.1.2). In our opinion, the condition of article 23.9.1.1 applies in this case, and evidence that the conditions of article 23.9.1.2 are met, are given in Appendix 1 herein. Therefore, we regard the name Lycaena legeri Freyer as invalid and qualified as a nomen oblitum and declare the name Lycaena ottomana Lefèbvre as valid and qualified as a nomen protectum, which has precedence over the former as long as both names are thought to represent subjective synonyms.

Results and discussion
The updated species list of European butterflies includes 496 species, which belong to 110 genera in 21 subfamilies and six families (Tables 2 and 4; Fig. 4). A list of main authors with some additional data is given in Table 5. An electronic version of the checklist that includes a country-based distributional checklist is found in Suppl. material 1, Suppl. material 2.
Compared to the last version 2.6.2 of Fauna Europaea, nine species have been excluded from the list ( Table 6). On the other hand, 15 species were added to the list. Another recently discovered species, Spialia rosae Hernández-Roldán, Dapporto, Dincă, Vicente & Vila, 2016, has already been added to the Fauna Europaea database.
Apart from the changes due to the gender agreement provision (Table 7), only three species names had to be changed due to new nomenclatural evidence: Pyrgus bellieri (Oberthür, 1910) to Pyrgus foulquieri (a name which had already been used in previous field guides), Proterebia afra (Fabricius, 1787) to Proterebia phegea (hopefully solving a longstanding controversy, see e.g., Jutzeler and Lafranchis 2011), and the mandatory change of Pseudochazara hippolyte (Esper, 1783) to Pseudochazara mercurius due to primary homonomy. Thais cerisy Zerynthia cretica (Rebel, 1904) Thais  Taxon Table 3. Annotations to the updated checklist of the butterflies of Europe.
1 Iphiclides feisthamelii is considered a separate species based on differences in adult morphology van Oorschot 2011, Lafranchis et al. 2015) and nuclear genetic markers (Wiemers and Gottsberger 2010;Dincă et al. 2015), despite very local hybridisation along the contact zone in southern France (Lafranchis et al. 2015) and extensive mitochondrial introgression in the Iberian Peninsula (Wiemers and Gottsberger 2010;Dincă et al. 2015). Its distribution includes the SW part of France, the Iberian Peninsula, and northern Africa. A larger number of changes concern the genus names. Most of them are in the family Lycaenidae, where 26 species changed their genus name, mainly based on the molecular study by Talavera et al. (2013), which substantially improved our knowledge of phylogenetic relationships of the subtribe Polyommatina. However, none of the genus names is new and many of them have already been used with the same species. In addition, four species formerly placed in the genus Argynnis were transferred into the genera Fabriciana and Speyeria, based on the study by De Moya et al. (2017). The former genus name had already been used previously for the same species, whereas the latter seems new to European lepidopterists, but is commonly used in North America. Although it could be argued that the change was avoidable by keeping a larger genus Argynnis, a solution originally also favoured by Simonsen et al. (2006), this would have meant to rename a large number of North American butterflies currently placed in the genus Speyeria, and was rejected by North American lepidopterists. Therefore, the recommended changes appear to cause the least changes on a global level and will hopefully contribute to a more consistent taxonomy of Holarctic Argynnini. Finally, quite a number of minor changes have been implemented, which correct mistakes in names of authors, year of publication, or the incorrect use of parentheses for species that have changed generic combinations. An example is the change of year for 6 butterfly names due to a correction of the publication date of Linnaeus' Fauna Svecica. Evenhuis (1997: 480) has shown convincingly that this edition was actually published on [14 November 1760], not "1761" as stated in the title page of the work and Bousquet (2016) also agrees with that year of publication.

Turanana panagaea
Distributed outside Europe in the Asian part of Turkey and replaced by Turanana taygetica in Europe Coutsis 2005

Polyommatus galloi
According to the molecular study of Vila et al. (2010) P. galloi represents an isolated population of Polyommatus ripartii and is not considered as a separate species. (Balletto & Toso, 1979)

Polyommatus menalcas
Distributed outside Europe in Asian part of Turkey

Conclusions
Taking into account the many recent research findings, especially those with molecular methods, we think that the new taxonomy represents a step forward in stabilizing European butterfly taxonomy and nomenclature. Nevertheless, we have to note that some groups, e.g., the genera Euchloe, Callophrys, Pseudophilotes, Melitaea, and Hipparchia, as well as the subgenus Agrodiaetus of the genus Polyommatus are still in need of revision, which will certainly lead to additional changes in the future. Furthermore, we still have large knowledge gaps for species in other regions of the Palearctic region (especially in Central Asia), which might require changes in order to achieve a consistent taxonomy of Palearctic and Holarctic butterflies.  Our thanks also go to Ole Karsholt for his review which helped to improve the manuscript.
VL was supported by grant N 14-14-00541 from the Russian Science Foundation to the Zoological Institute of the Russian Academy of Sciences and ZF by grant 14-36098G from the Czech Science Foundation.
tomanus Lefèbvre. The latter name was thought to represent the valid name and was first used in its original combination (Polyommatus ottomanus) and starting from the 20 th century mostly in the combination of Chrysophanus ottomanus: • Brullé (1832) During the last 50 years we are not aware of any use of legeri Freyer, except as a subjective junior synonym of ottomanus Lefèbvre. The latter name was mostly used in the combination of Heodes ottomanus and later as Lycaena ottomanus or, due to the gender agreement rule of the code, as Lycaena ottomana: