Paroeneis palaearcticus Staudinger

Paroeneis palaearcticus Staudinger ( Figs 3E View Figure 3 , 4E View Figure 4 )

Valve bilobate, with entirely different shape as compared with other genera. There are no sclerotized teeth on the valve. Tegumen massive. Uncus not massive. Subuncii broad and not pointed; they have approximately the same thickness in the entire length. Juxta small, not strongly sclerotized. Aedeagus long. Vesica simple, membranous, without sclerotized teeth.

Karanasa kirgisorum Avinov & Sweadner ( Figs 3F View Figure 3 , 4F View Figure 4 )

Valve elongated in lateral view, not trapezoidal, with rounded tip; there are sclerotized teeth on the dorsal part of valve near the tip. Tegumen massive. Subuncii have approximately the same thickness in the entire length. Juxta massive and strongly sclerotized. Aedeagus relatively short, curved and armed with spikes. Vesica with two cornuti, each possessing a single sclerotized tooth. Generally, the male genitalia in Karanasa are similar to those in Oeneis (Oeneis) , except for the massive and strongly sclerotized juxta and the aedeagus, which is curved and armed with spikes.

Thus, the following characters of the male genitalia are unique for one or two of the studied genera and subgenera and represent potential (syn)apomorphies. (1) Trapezoidal, nearly triangular shape of valve in lateral view (character 1) and vesica with two large, oval sclerotized cornuti, each cornutus with several claw-like teeth forming a strongly sclerotized ridge (character 2) (unique characters for Davidina and Oeneis (Protoeneis) . (2) Valve distinctly narrowed in distal portion, not trapezoidal, with rounded tip (character 3) [unique character for Oeneis (Oeneis) and Karanasa ]. (3) Specific shape of valve with a hump-like convexity on costa (character 4). This character is found in Neominois ridingsii ( de Lesse, 1951; Warren et al., 2008; this study) and seems to be unique for Neominois . However, it is not genus-specific, since it is not found in the second species of Neominois , N. carmen Warren, Austin et al., 2008 ( Warren et al., 2008). (3) Bilobate valve (character 5), found by us for Paroeneis and previously reported for Aulocera ( de Lesse, 1951; Chou, 1998; Sharma & Rose, 2014). (4) Massive and strongly sclerotized juxta (character 6) and aedeagus armed with spikes (character 7) (unique character for Karanasa ).

PHYLOGENETIC ANALYSIS OF THE MORPHOLOGICAL CHARACTERS

8. Valve with the distal process ( JakŠić, 1998: fig. 61, 4–9, fig. 62, 1–5; Nekrutenko, 1985: fig, 59, 64) (1); valve without the distal process (0).

9. The distal part of the valve pointed and curved up ( JakŠić, 1998: fig. 61, 1; figs 68–70; Nekrutenko, 1985: fig, 58, 65, 66) (1); the distal part of the valve not pointed and not curved up (0).

10. The distal process of the valve long ( JakŠić, 1998: fig. 61, 4–9; Nekrutenko, 1985: fig, 59) (1); the distal process of the valve short and massive ( JakŠić, 1998: fig. 62, 1–5; Nekrutenko, 1985: fig. 64) (0).

11. Valve broad, with distal extension ( JakŠić, 1998: figs 63–66) (1); valve without distal extension (0).

12. Jullien organ present ( JakŠić, 1998: figs 63–66) (1); Jullien organ absent (0).

Maximum parsimony phylogenetic analysis (MP) of the morphological matrix reveals six clades that appear in all nine MP trees. The 100% consensus of the nine MP trees is shown in Figure 5 View Figure 5 . The same topology is revealed by BI analysis of the morphological matrix. Three clades, ( Davidina + Protoeneis ), ( Minois + Arethusana ) and [ Brintesia + ( Minois + Arethusana )] were highly supported in the BI analysis and had a Based on the original data on characters 1–7 and published data on characters 4, 8–12 ( de Lesse, 1951; Nekrutenko, 1985, JakŠić, 1998; Warren et al., 2008), we created the following matrix of the binary characters ( Table 1):

1. Valve trapezoidal, nearly triangular shape in lateral view (1); valve elongated, not trapezoidal or triangular in lateral view (0).

2. Vesica with two large oval sclerotized cornuti, each cornutus with several claw-like teeth forming a strongly sclerotized ridge (1); vesica simple, membranous, without sclerotized teeth or with two single teeth (0).

3. Valve with rounded tip covered by numerous sclerotized teeth (1); tip of the valve not rounded (0).

4. Valve with a large hump-like convexity on costa (1); valve without a large hump-like convexity on costa (0).

5. Valve bilobate (1); valve not bilobate (0).

6. Juxta large, massive and strongly sclerotized (1); juxta small, not strongly sclerotized (0).

7. Aedeagus armed with spikes (1); aedeagus not armed with spikes (0).

medium support in our MP analysis. Three clades, ( Aulocera + Paroeneis ), ( Oeneis + Karanasa ) and ( Satyrus + Pseudochazara + Chazara ), had low support.

WING VENATION

According to Kuznezov (1930), there is a clear distinction between Davidina (based on D. armandi , the type species of the genus) and Oeneis (based on O. norna , the type species of the genus) in the forewing venation. First, in Davidina all radial branches, except R1, form a common and rather long stalk R2 + 3 + 4 + 5; in Oeneis r2 is free, i.e. R2 and R3 + 4 + 5 start independently from the same point of the discal cell. Second, in Davidina M1 rises at a distance from this stalk, the vein r-m1 being, thus, well developed; in Oeneis M1 rises from one point with R3 + 4 + 5, and r-m1 is, thus, absent.

Kuznezov’s observation was based on two specimens only. We analysed 18 specimens of D. armandi and discovered that four types of venation are present in Davidina .

1. All radial branches, except R1, form a common and rather long stalk R2 + 3 + 4 + 5; M1 rises at a distance from this stalk, the vein r-m1 being, thus, well developed [ Davidina type according to Kuznezov (1930), found in 11 samples, Fig. 6A View Figure 6 ].

2. Veins R2, R3 + 4 + 5 and M1 rise from the same point on the discal cell; the vein r-m1 is absent [ Oeneis type according to Kuznezov (1930), found in four samples, Fig. 6B View Figure 6 ].

3. Veins R2 and R3 + 4 + 5 rise from the same point on discal cell; M1 rises at a distance from this point, the vein r-m1 being, thus, well developed (found in two samples, Fig. 6C View Figure 6 );

4. All radial branches, except R1, form a common and rather long stalk R2 + 3 + 4 + 5; this stalk and M1 rise from the same point on the discal cell; the vein r-m1 is absent (found in a single sample, not shown).

Thus, the wing venation is not stable in D. armandi . It is not genus-specific, but variable, even on population level. The supposed differences between Oeneis and Davidina in wing venation ( Kuznezov, 1930) is not confirmed.