polaris Zeller , 1880 ( Lepidoptera , Plutellidae ) rediscovered at Svalbard , Norway , with comments on its taxonomic position

After 142 years, Plutella polaris Zeller, 1880 was rediscovered at Svalbard, Norway, in 2015. The locality and its vegetation are described. The specimen is the first female ever recorded, and its genitalia are illustrated. The taxonomic position of P. polaris relative to other Plutella Schrank, 1802 species in the Northern hemisphere is discussed.


Introduction
In 1873, seven specimens of an unknown microlepidopteran species were collected by the entomologist Reverend A.E. Eaton at Wide Bay, Svalbard, Norway, between 21 st July and 24 th July.The specimens were subsequently handed over to Henry Tibbats Stainton, who at the time was Britain's leading authority on Microlepidoptera.Stainton sent two of the males to his German colleague Philipp Christoph Zeller for his opinion.In a letter to Stainton dated May 29 th 1874, Zeller described the species and named it Plutella polaris.Stainton (1880) later quoted Zeller's original description and made it clear that the description was Zeller's despite the fact that Stainton himself was the author of the paper (Stainton 1880) and so Zeller's authorship retains the later date.
Since the species was first discovered in 1873, no further records had been reported until 9 July 2015, when a female was collected by Geir Søli at Ringhorndalen, Wijdefjorden, Svalbard.The Ringhorndalen specimen (Fig. 1) agrees well externally with illustrations of a type specimen of P. polaris that was given by Bengtsson and Johansson (2011), and certainly not with P. xylostella (Linnaeus, 1758), a vagrant species observed on Svalbard every three or four years (Peter Coulson, pers. comm.).Geir Søli's new specimen is preserved in the collection of the Natural History Museum, University of Oslo (NHMO).
In the present paper, we provide supplements to the description of the species and its habitat at Svalbard, along with an updated account of its taxonomic status.

The locality
Ringhorndalen is part of the Indre Wijdefjorden National Park and one of several valleys on the eastern side of the Wijdefjorden.Wijdefjorden is the longest fjord on Svalbard, being 108 km long and cutting deep into Spitsbergen from the north (Fig. 2).The vegetation around the inner part  ©Societas Europaea Lepidopterologica; download unter http://www.soceurlep.eu/und www.zobodat.at of the fjord has been characterized as high arctic steppe with several very rare plant species (e.g.Elvebakk and Nilsen 2016 and references therein).Very low precipitation and saline soils are the main reasons for the formation of this extremely rare and unique vegetation type found nowhere else in Europe.The inner, less exposed parts of the Ringhorndalen valley are less arid and locally the summer temperatures can be quite high.
As Wijdefjorden is not easily accessible during summer months, its flora has just recently been more thoroughly investigated (e.g.Elvebakk and Nilsen 2002).These studies have documented a unique vegetation with several rare and thermophilic plant species, some of them new to Svalbard (e.g.Elvebakk andNilsen 2011, 2016;Eidesen et al. 2013).Based on these findings, the rich flora of Ringhorndalen may hold species of relict character from a warm postglacial period.
The single specimen of P. polaris was found on a south facing slope in the inner part of the valley (79.3358°N 16.1289°E) (Fig. 2).These slopes have rich vegetation with several species of Brassicaceae that are possible food plants for the larvae, e.g.Draba alpina L., Braya glabella Richardson or Coclearia groenlandica L. (as suggested by Bengtsson and Johansson (2011)).
The Lepidoptera fauna of Svalbard is very poor, with three resident species only: P. polaris, Apamea exulis (Lefèbvre, 1836) and Pyla fusca (Haworth, 1811) (Coulson et al. 2014).Interestingly, all three species were collected at Ringhorndalen in July 2015, of which P. fusca was even very common at the site.

Plutella in the Northern hemisphere
Presently 26 species of Plutella Schrank, 1802 are recognised worldwide, of which a few must still be regarded as dubious (see Robinson and Sattler (2001) Baraniak (2007), Landry and Hebert (2013)).Eleven species have been recorded from the northern hemisphere north of 30°N (Table 1).Three of these species have a wide distribution, including the nearly cosmopolitan P. xylostella, the Holarctic P. porrectella (Linnaeus, 1758) and the circumpolar P. hyperboreella Strand, 1902(Landry et al. 2013).The two species P. armoraciae Busck, 1912 andP. notabilis Busck, 1904 are restricted to north-western North America, while the remaining six species are Palaearctic, seemingly with a rather restricted distribution confined to montane (P.geniatella Zeller, 1839, P. haasi Staudinger, 1883, P. huemerella (Baraniak, 2007) and P. kyrkella (Baraniak, 2007)) or arctic regions (P.polaris and P. mariae Rebel, 1923) (see also Bengtsson and Johansson 2011).P. polaris is among the most enigmatic of these species, as it has been known by only the original syntypes.
Of the seven syntypes originally designated for Plutella polaris, only two males are still extant in the Natural History Museum, London.What has happened to the remaining five syntypes is not known.In addition to our rediscovery of P. polaris in Wijdefjorden, we were recently informed of another specimen identified as P. polaris.This specimen, a single male, was found during an expedition by Austrian and Russian lepidopterists to the Republic of Altai in Russia in 2016 at a considerably lower latitude of 49.5°N, 88.08°E (Huemer et al. 2017;BOLD), and is commented on below.
Compared to other species of Plutella that occur in northern Europe, the forewing pattern of P. polaris (Fig. 1) resembles that of P. haasi in having the brownish grey ground colour with a paler broad band along the posterior margin that is widened in the tornal area.The posterior margin is marked with a row of dark dots.The costal half of the forewing is paler than in the medial area.The two species P. hyperboreella and P. mariae differ externally from P. polaris by their strongly contrasted forewing pattern.Palaearctic: Norway, Sweden, Scotland Plutella huemerella (Baraniak, 2007) Palaearctic: France: Alpes-Maritimes Plutella hyperboreella Strand, 1902 Holarctic: Norway, Sweden, Finland, Canada, Alaska Plutella kyrkella (Baraniak, 2007) Palaearctic: Russia: SW Altai Plutella mariae Rebel, 1923 Palaearctic: Novaya Zemlya Plutella notabilis Busck, 1904 Nearctic: North-western North America.Plutella polaris Zeller, 1880 Palaearctic: Norway and Russia (Altai) Plutella porrectella (Linnaeus, 1758) Holarctic: North America and Eurasia Plutella xylostella (Linnaeus, 1758) Cosmopolitan The female genitalia of Plutella polaris Baraniak (2007) studied the two syntypes of P. polaris kept in the Natural History Museum, London, figured the wings and the male genitalia, and designated a lectotype.Bengtsson and Johansson (2011) also figured the male genitalia and presented a water colour painting of the moth.
As the female of P. polaris is found for the first time, a short description and illustration of the female genitalia (Fig. 3) are presented: apophyses anteriores longer than apophyses posteriores; tergite 8 forming sub-rectangular plate, about 2,5 times wider than long; sternite 8 formed by two slender sclerites that are narrowed laterally and fused with apophyses anteriores, at touching point with spines on posterior edges; ostium boat-shaped with curved "roof"; sternite 7 with concave posterior edge; ductus bursae narrow in posterior half, curved before middle, anterior portion wide, small sclerite present before widening; corpus bursae oval, without signa; bulla seminalis also oval, smaller than corpus bursae.The genitalia differ from those of P. haasi in having the ductus bursae divided in a narrow posterior part and a broad anterior one.They appear to be more similar to those of P. hyperboreella, but differ by the concave posterior edge of sternite 7.

Genetic analyses
In order to confirm the species identity of P. polaris, and derive more information about its taxonomic position relative to other Plutella species in the Northern hemisphere, we performed an analysis based on DNA barcoding.One leg was sampled from the Wijdefjorden specimen and sent for DNA extraction, amplification and sequencing at the Canadian Centre for DNA Barcoding in Guelph.The DNA barcoding fragment of the mitochondrial cytochrome c oxidase subunit 1 (CO1) was sequenced, using standard primers and bi-directional Sanger sequencing.The DNA sequence and original trace-files of the P. polaris specimen from Svalbard are publicly available in the dataset 'Plutella polaris Svalbard, Norway' (DS-PPSN) (doi: dx.doi.org/10.5883/DS-PPSN) in the Barcode of Life Data System (BOLD) (Ratnasingham and Hebert 2007).Both the Wijdefjorden and Altai specimens of P. polaris belong to the same Barcode Index Number (BIN), BOLD:ADB0013, with 1.32% pairwise divergence between them (609 bp in common).
In addition to the CO1 sequence obtained from P. polaris, 41 CO1 sequences from nine other species of Plutella were obtained from BOLD (Ratnasingham and Hebert 2007)  This solves the previous enigma of the existence of a prior endemic to Spitzbergen not collected for 142 years that existed close to the 80 th parallel.More thorough collecting in northern Eurasia will most likely uncover new localities for the species.Another possibility is that P. polaris has a circumpolar distribution, as recently revealed for P. hyperboreella (Landry et al. 2013).According to Hodges et al. (1983) there are a few additional Plutella species in North America, but there is no modern taxonomic treatment of them.If future taxonomic work results in the synonymization of any of these species with P. polaris, the name P. polaris would still have priority.
The genetic analysis also revealed a conflicting interpretation of the species P. haasi.The specimen forming a sister-group to P. hyperboreella originates from the type locality (Norway: Dovre), while the specimen located further towards the root of the Plutella clade, was collected in North Ural, Russia, and undoubtedly represents a new species.Further studies are needed to elucidate the status of these species.
©Societas Europaea Lepidopterologica; download unter http://www.soceurlep.eu/und www.zobodat.at The exact locality for Eaton's first records of P. polaris in 1873 is not known, as the locality is simply referred to as "Wide Bay", an old, unofficial name for Wijdefjorden, used up to 1934.As previously mentioned, Ringhorndalen is just one of several valleys opening towards the fjord, and Eaton may well have disembarked elsewhere along the fjord.From our own observations in Flatøyrdalen, the first valley south of Ringhorndalen, it is likely that the neighbouring valleys also may offer favourable conditions for the species.Hopefully future expeditions in the region will reveal a more detailed knowledge about the present distribution of P. polaris.

Figure 2 .
Figure 2. Ringhorndalen in Indre Wijdefjorden National Park where Plutella polaris was rediscovered in 2015.The inserted map shows Svalbard with the exact position of the locality (red circle).(Photo: G. Søli).

Figure 4 .
Figure 4.The Maximum Likelihood tree based on DNA barcoding of P. polaris (arrow) and 10 other nominal species in the genus Plutella.Six species representing the two genera Eidophasia and Rhigognostis were used as outgroup.Bootstrap support values (1000 replicates) are listed above the branches.Number of specimens is given in brackets after each species.