AL: androconium length; AB: androconium breadth; A: ratio AL/AB; FW: forewing; HW: hindwing; MM: Morten Schneider Mølgaard; N°: number; N/A: not applicable; oc.: ocelli; SC: Sylvain Cuvelier; SD: standard deviation; subm.: submarginal; UNS: underside; UPS: upperside; Var: variable.

In two localities in the Albanian province Korçë (Boboshtiçë and Gjergjeviçë) males and females of P. amymone and P. mniszechii tisiphone were netted by both authors. A search for all potential references on the butterflies of Albania was made during the preparation for the trip. The relevant publications were gathered from different sources in order to build a database including as much historical data as possible (Abadjiev and Beshkov 1996a, b; Alberti 1965; Beshkov 1994; Beshkov 1995; Beshkov and Misja 1995; Gaskin 1990; Misja and Kurrizi 1984; Moucha 1963a, b; Murraj 1972; Płóciennik et al. 2009; Popescu-Gorj 1971; Rebel 1913; Rebel 1918; Rebel and Zerny 1931; Verovnik and Popović 2013a, b). Only the data from all species with rather precise indications of the locality were included in an Excel spreadsheet with coordinates in decimal degrees that were defined with Google Earth and an online coordinate conversion tool (Montana State University 2014). All data from our personal observations were included in this database. During our surveys, coordinates were obtained in the field with a GPS (Garmin Etrex Legend). A map of Albania was adapted for use with DMAP distribution mapping software to produce distribution maps per species and one global coverage map of all Albanian butterfly species. During history the borders of Albania have changed. Some historical data now in fact concern localities that are situated in Montenegro, Kosovo and Macedonia. To be as complete as possible, we have maintained these observations in the coverage map.

Since the discovery of P. amymone by Brown (1976), as far as is known to the present authors, there have been only 10 male and two female voucher specimens collected and included in publications. Pseudochazara species are very variable and difficult to identify. No comparative studies on the external characters of a good number of P. amymone specimens have been published so far and we had no precise idea about the variability of the external characters of both sexes.

The male holotype, figured in black and white, was for a long time the only documented picture of this species (Brown 1976). In his article, Brown described the external characters based on a very small series of four males: “Upperside similar to graeca but wings more rounded and with notably broad clear orange postdiscal bands more or less broken by grey-brown ground colour along v4 of forewing and enclosing blind black oc. in S 2, 5 and minute black oc. in S 3, 4 on forewing. Sex brand inconspicuous. Hindwing sometimes with small black ocellus in S 2 and dark grey submarginal line broken by orange along veins. Marginal grey band thin (1–2 mm wide). Underside ground colour pale yellow-grey but variable. Hindwing irrorate with darker scales and indistinct striae. Forewing length 26–27 mm. Female. Unknown.”

Luckily, the two original photographs from this publication were available and they allow a better comparison. For this purpose, Jos Dils (Belgium) kindly provided the two photographs of the male holotype (Brown 1976) (Fig. 1). On the underside of the photograph it is written: “WATSONI ? Clench & CHS SCH” looking like a link to a Pseudochazara species from Afghanistan (Fig. 1). However, as we never received a reply from Brown, it is not possible to fully understand this detail.

Recently, two prepared male specimens were figured by Eckweiler (2012) that look different from the holotype but with such a small sample size it is difficult to estimate if this is within the normal range of variability. Concerning females, for a long time there was only one figure (Tolman and Lewington 1997) but recently a first photograph of the upper- and underside of a single female from Albania became available (Eckweiler 2012). Even fewer photographs of the butterfly in nature have been published and for all these documents there remained a degree of uncertainty concerning the final determination (Cuvelier 2010, Eckweiler 2012).

In the field, fresh males of P. amymone look quite different when flying than P. mniszechii tisiphone and identification is possible in a fair number of cases. As Pseudochazara species almost never sit with open wings, reliable identification in the field is often based on the underside of the wings and for both sexes it is difficult and depends on the freshness of the butterflies. Therefore we also sampled P. mniszechii tisiphone (Figs 5a–h) at Boboshtiçë, where this butterfly was extremely common, in order to compare the two taxa and to obtain determination keys for males and females of both species.

Potential variables of the external characters were selected and included in an Excel workbook containing separate worksheets per species and gender. After this first selection, the colour of the fringes was discarded as a variable due to the difficulty of formulating measurable criteria. Drawings of measurement on UPS and UNS are included in Appendix 1. The following variables were analyzed: UPS FW: Var 1: length of FW from apex to point of attachment to thorax, fringes included (mm); Var 2: visual assessment of the oc. in S2 and S5 (blind= 0, white pupil= 1); Var 3: visual assessment of the number of spots in S3 and S4 (0, 1 or 2); Var 4: width of the submarginal band across the centre of the ocellus in S2 (mm); Var 5: Var 4/Var 1 (%); Var 6: visual assessment of the conspicuous sex brand in cell (absent= 0, present= 1). UPS HW: Var 7: visual assessment of the number of oc. in the submarginal area (0, 1, 2 or 3); Var 8: width of the submarginal band along vein 3 (mm); Var 9: Var 8/Var 1 (%). UNS FW: Var 10: visual assessment of the oc. in S2 and S5 (blind= 0, white pupil = 1); Var 11: visual assessment of the pale area from ocellus in S5 to the cell (uniform= 0, contrasted = 1); Var 12: visual assessment of the number of spots in S3 and S4 (0, 1 or 2); Var 13: visual assessment of the marginal line (diffuse= 0, sharp= 1); Var 14: visual assessment of the fine black line-shape markings in the basal area of the cell (absent= 0, present= 1); Var 15: length of the ocellus in S5 (mm); Var 16: shortest distance from the white centre of the ocellus in S5 to the margin (mm); Var 17: Var 15/Var 1 (%); Var 18: Var 15/Var 16 (%); Var 19: Var 16/Var 1 (%). UNS HW: Var 20: visual assessment of the number of oc. in the submarginal area (0, 1, 2 or 3); Var 21: visual assessment of the median band (absent= 0, present= 1);

We photographed the upper- and underside of all male and female specimens of P. amymone and P. mniszechii, each of us in our personal reference collections. A scale bar was included with each specimen (Appendix 1b). Each digital image was imported into Paint Shop Pro v. 6.02. A straight vector line was drawn on the butterfly in its own layer to measure the exact length of the desired parameter. Afterwards, the vector line was rotated into horizontal position and then moved onto the scale bar under the butterfly, making it possible to measure length in millimetres, at an accuracy level of 0.25 mm. The whole dataset was used for two analyses: a) a variability study between the two P. amymone populations and b) a comparison between all P. amymone and P. mniszechii tisiphone. Statistical analysis was performed with StatSoft STATISTICA 12. The Mann-Whitney U test was used to test for differences at 0.05 significance level in a two-tailed test.

Figure 1. 

Upper and underside of the P. amymone holotype photographs, holotype collected in the mountains just N. of Ioánnina, Epiros, Greece, 650m, 10.vii.1975 (photograph: SC).

Androconial scales were removed from the upperside of the forewings of one P. amymone and one P. mniszechii tisiphone and photographed with a calibrated 5 megapixel Dino-Lite AD-7013MZT digital microscope with adjustable magnifications. The length and breadth of the androconia were measured according to the description by Wakeham-Dawson (2000) but at maximum magnification (×500).

DMAP, distribution mapping software: http://www.dmap.co.uk/. — Dr. Alan Morton, Blackthorn Cottage, Chawridge Lane, Winkfield, Windsor, Berkshire, SL4 4QR, UK.

On 15.vii.2013, late in the afternoon, we started our search for P. amymone in a narrow valley with very steep slopes near Boboshtiçë (province of Korçë). On the dark red-grey slopes our attention was attracted by cushions of pink flowers that we identified as Acantholimon echinus (L.) Boiss. (Plumbaginaceae) and the whole area looked very similar to the published habitat photograph (Eckweiler 2012). The rest of the day we explored this area and found some P. mniszechii tisiphone, but we had not had a glimpse of P. amymone. We were, however, quite certain that we had to be near its habitat.

The next morning, 16.vii.2013, we entered deeper into the river valley and searched at 1100–1200 m altitude on SSW exposed steep rocky slopes with parts of loose gravel. On this ophiolite substrate (Fig. 2a) with characteristic red-grey colour, scattered tall grasses were growing, but otherwise the area was almost devoid of vegetation (Fig. 2b). Here P. amymone had just emerged and males (Fig. 2d) were already flying quite commonly. We observed the males showing a typical territorial behaviour: patrolling and chasing away other males. The females were searching for nectar sources and egg-laying places in the scattered tall grasses that were present in the biotope. The rumours of a rocket speed flight and strange nuptial behaviour appeared not to be true. The species is not shyer than other species in the genus Pseudochazara.

On 18.vii.2013, early morning, we again visited this locality and observed that some males were already getting worn. One male P. amymone (Fig. 2e), seen from afar sitting with open wings on flowers of A. echinus, had been caught by a crab spider (Fig. 2f) that was identified by Rop Bosmans (Belgium) as Thomisus onustus (Walckenaer, 1805), a common species in the Balkans. This butterfly also shows typical injuries on the hind wings caused by lizards that were common in the habitat. Females were already more numerous than two days earlier. Here, P. amymone flies sympatrically with P. mniszechii tisiphone. The males of these two species are easily distinguished in the field, but this is not the case for females. Flying males of P. amymone look smaller and also show a much more pronounced orange and black contrast. Although the females of P. amymone are a little smaller, this cannot be observed in the field. They also do not exhibit the contrasting orange and black colours and this makes it difficult in the field to distinguish females of these two species.

In Boboshtiçë, we observed the first ever known copula (Fig. 2g) and a pale form (Figs 3g–h) of a male P. amymone. After our trip, Eckweiler confirmed (oral comm., August 2013) that this was the locality where he originally discovered P. amymone in Albania.

During these explorations, on 18.vii.2013, we were able to extend the known distribution of P. amymone by approximately 25 km to the west, as the crow flies, as we discovered it in another remote mountain range in the westernmost part of the province of Korçë. This mountain range is physically separated from Boboshtiçë by a 10 km broad river valley, at 850 m altitude.

Entering a remote valley, we observed steep rocky slopes orientated to the SSW on ophiolite substrate, as we had seen in Boboshtiçë. On climbing these slopes we immediately observed a large population of P. amymone. The new biotope is situated near Gjergjeviçë (Fig. 2c) and in the upper part of the known altitudinal distribution (Pamperis 2009) of P. amymone, being at an altitude of 1200–1400 m. It has the same characteristics as the biotope at Boboshtiçë, except for one major difference: the presence of scattered bushy vegetation whereas the habitat at Boboshtiçë is completely open. We observed that P. amymone near Gjergjeviçë is also sympatric with and even outnumbering P. mniszechii tisiphone and here a copula of this species was photographed (Fig. 2h). The flight period of P. amymone was apparently the same as in Boboshtiçë, due to the general freshness of most of the specimens and the presence of good numbers of females as observed in the morning of the same day at Boboshtiçë.

Figure 2. 

a–b. Habitat of P. amymone, Boboshtiçë, Albania, 16.vii.2013. c. Habitat of P. amymone, Gjergjeviçë, Albania, 18.vii.2013. d. ♂ P. amymone, Boboshtiçë, Albania, 16.vii.2013 (photographs: MM). e. ♂ P. amymone caught by a crab spider, Boboshtiçë, Albania, 18.vii.2013. f. T. onustus holding a ♂ P. amymone, Boboshtiçë, Albania, 18.vii.2013 (photographs: SC). g. Copula of P. amymone, Boboshtiçë, Albania, 16.vii.2013 (coll. & photograph: SC). h. Copula of P. mniszechii tisiphone, Gjergjeviçë, Albania, 18.vii.2013 (photograph: MM).

To our big surprise, we noticed P. amymone was not rare at all in the two biotopes (Boboshtiçë and Gjergjeviçë) and that the habitat in both localities was very large but difficult, if not impossible, to explore. Nevertheless we sampled enough voucher specimens to get a better idea of the range of variation in external characters and to make a comparative study on the habitus of these two separate populations of P. amymone.

There are two limitations of our dataset which require further attention in the future when new populations are discovered. The sample size of 38 males and 19 females for such variable butterflies remains suboptimal and the sample size of the two localities is not equal. The range and mean of all variables for the two populations is given in Table 1. Males and females are figured (Figs 3a–h, 4a–f). All the variables of the two populations clearly overlap and there was not a single variable showing clear differences between the two populations. For these reasons the measurements from the two populations can be pooled together for the comparison between Albanian P. amymone and P. mniszechii tisiphone.

Two P. amymone specimens exhibited asymmetry between right and left side. This was the case for one male with one spot in S3-S4 on the left FW UPS and no marking on the right FW. One female had one black ocellus on the left side of the HW UNS and two on the right side.

There seem to be a few marked differences between P. amymone from Albania and the original description by Brown (1976). It is clear that the holotype (Fig. 1) is not fresh and probably the paratypes were even more worn. A few butterflies in our study that were less fresh also have a clearer appearance and tend to become more orange in the postdiscal bands of the FW UPS. We have clearly shown that many characteristics are much more variable within a single population than described by Brown (1976). We presume, until more material becomes available, that Greek P. amymone falls within the given range of variability.

Figure 3. 

Variability in P. amymone. a–b. ♂ typical upper- and underside, Boboshtiçë, Albania, 16.vii.2013 (coll. & photograph: SC). c–d. ♂ typical upper- and underside, Gjergjeviçë, Albania, 18.vii.2013 (coll. & photograph: MM). e–f. ♂ aberration, upper- and underside, Boboshtiçë, Albania, 16.vii.2013 (coll. & photograph: SC). g–h. ♂ pale form, upper- and underside, Boboshtiçë, Albania, 16.vii.2013 (coll. & photograph: SC).

Figure 4. 

Variability in P. amymone. a–b. ♀ typical P. amymone upper- and underside, Gjergjeviçë, Albania, 18.vii.2013. c–d. ♀ typical P. amymone upper- and underside, Boboshtiçë, Albania, 16.vii.2013. e–f. ♀ dark form upper- and underside, Gjergjeviçë, Albania, 18.vii.2013. (Coll. & photographs: SC).

Figure 5. 

Variability in P. mniszechii tisiphone. a–b. Typical ♂ upper- and underside, Boboshtiçë, Albania, 16.vii.2013. c–d. ♂ typical upper- and underside, Boboshtiçë, Albania, 16.vii.2013. e–f. ♀ typical upper- and underside, Boboshtiçë, Albania, 16.vii.2013. g–h. ♀ dark form upper- and underside, Boboshtiçë, Albania, 18.vii.2013. (Coll. & photographs: SC).

The maximal range and mean of all variables for the two species are shown in Table 2. An overview of the statistical significance status of all variables is shown in Table 3. Some variables of P. amymone and P. mniszechii tisiphone did not overlap at all. Some differences are present in both sexes, others are only present in males. Below we list the variables that can be used in distinguishing P. amymone and P. mniszechii tisiphone.

Var 13: UNS FW submarginal line in both sexes (Figs 6a–b). This line is always sharp in P. amymone and diffuse in P. mniszechii tisiphone.

Var 14: UNS FW basal area of the cell in both sexes (Figs 6c–d). There are always black linear markings inside this area in P. amymone which are absent in P. mniszechii tisiphone.

Var 6: UPS FW sex brand position in males (Figs 6e–f). P. amymone has a black sex brand over the whole cell and the androconial field is extending to the inner margin of the FW. P. mniszechii tisiphone does not have a sex brand in the cell and the androconial field is covering only half of the cell towards the inner margin of the FW.

Var 7: UPS HW number of oc. in males (Figs 6g–h). There is 1 ocellus in P. amymone and 2 oc. in P. mniszechii tisiphone. Females of the two species have both a range of 1 to 2 oc.

The wingspan of the females was in many cases useful in the field for the identification but one P. mniszechii tisiphone female falls in the upper range of P. amymone.

The dense sex band in the FW cell makes it difficult to isolate the androconial scales (Fig. 7a) of P. amymone. Scales with quite different shapes, sometimes bright silver-grey (Fig. 7b), were found in this area and created confusion. The type of these scales is unclear to the authors.

According to the criteria of Gross, the androconial scales of P. amymone (Fig. 7c) are in general of type 7 (Gross 1978) but few were found where the diameter of the lamina decreased immediately from the basal stalk towards the apex. These are more closely resembling a type 6 androconial scale. AL= 0.42 mm, AB= 0.05 mm and A= 8.4. These data are near the values for P. mamurra in Wakeham-Dawson and Kudrna (2000). Also the shape of the scales from P. amymone falls within the ranges of P. mamurra (Wakeham-Dawson and Kudrna 2000; Wakeham-Dawson and Kudrna 2005). The androconia of P. amymone are much larger and of a different shape than P. graeca (Wakeham-Dawson, 2000). The androconial scales of P. mniszechii tisiphone (Fig. 7d) are different in shape, transitional type 5-6 with dimensions: AL= 0.33 mm, AB= 0.027 mm and A= 12.19.

Figure 6. 

Visualization of the significant difference between P. amymone (a, c, e, g) and P. mniszechii tisiphone (b, d, f, h). Var 13 (a–b), Var 14 (c–d), Var 6 (e–f), Var 7 (g–h).

Figure 7. 

a. UPS FW sex brand of ♂ P. amymone (× 250), Boboshtiçë, Albania, 16.vii.2013. b. UPS FW scales in sex brand of ♂ P. amymone (× 500), Albania, 16.vii.2013. c. Androconial scale of ♂ P. amymone (× 500), Boboshtiçë, Albania, 16.vii.2013. d. Androconial scale of ♂ P. mniszechii tisiphone (× 500), Boboshtiçë, Albania, 16.vii.2013. (Coll. & photographs: SC).

Additional field work in other localities of the province of Korçë and in the provinces Kolonjë, Përmet, Tepelenë and Skrapar fill an important gap in the documented distribution of the butterflies from south-eastern Albania. The coverage map (Fig. 8) shows historical data (red dots), the two areas where P. amymone was found (blue letters B and G) and all new places that were surveyed by the authors (green dots).

Figure 8. 

Coverage of historical data and personal observations: map of Albania indicating Papilionoidea observations from literature (●) and from observations by the authors (●). B: Boboshtiçë; G: Gjergjeviçë: localities of P. amymone.