Liptena kiellandi Congdon & Collins, 1998

Libert, Michel & Collins, Steve C., 2021, Taxonomic notes on Liptena eketi Bethune-Baker, 1926 and related species (Papilionoidea: Lycaenidae: Poritiinae), Metamorphosis 32, pp. 30-35 : 33-34

Liptena kiellandi Congdon & Collins, 1998

Type material and distribution

L. kiellandi

1♀ Manki, NE of Foumban, west Cameroon (17.ii.1988) MLC gen. prep. Libert 120–028; 5♂ Nord-Kivu, NE DRC, leg. Ducarme, RDC; 4♂ Kasugho, 2000 m, v.2015, viii.2014 (gen. prep. Libert 117-191, 117-192), xi.2016, viii.2017; 1♂ Maliva, 1800 m, ii.2016, Bold MLIB–2530, MLC; 5♂ Ebogo, S of Yaoundé, Cameroon, ABRI; 1♂ Ebolowa, south Cameroon, ABRI; 1♂ Bertoua, east Cameroon, ABRI; 1♂ Batalimo, nr Bangui, Central African Republic, ABRI; 2♂ Mabira Forest, Uganda, ABRI.

The description of L. kiellandi was based on c. twenty specimens, all collected at the type-locality (Minziro forest, in north-western Tanzania). The authors also described the subspecies kakamegae, from c. thirty specimens, all from the type-locality (Kakamega Forest, western Kenya). All of these type-specimens are in ABRI.

A female captured in western Cameroon, originally identified as L. eketi by Libert (1992: 28) looks closer to L. kiellandi (which was only described in 1998).

More recently, several other populations of L. kiellandi have been discovered:

o Two specimens captured in the Mabira forest , in south-eastern Uganda (ABRI, referred to the nominate subspecies)

o A large population in north-eastern DRC (around twenty-five specimens – ABRI and RDC)

o Western Uganda (three specimens from Budongo and Bugoma in the Nairobi Museum, as L. fulvicans 1)

The range of L. kiellandi therefore extends from western Kenya to western Cameroon ( Fig. 1 ). It is separated from L. eketi by the Cameroon highlands, although Manki and the Rhoko forest are only about 260 km apart.

Facies ( Plate 1 : E–H)

Male holotypes of both subspecies were shown in the original description (upper side, two-thirds the actual size) and in d'Abrera (2009: 653), and their comparison appears to confirm the validity of the Kenyan subspecies. However , there are now about fifty specimens of the Kakamega subspecies in ABRI, several of which are hardly different from the nominate subspecies.

The upper side of L. kiellandi is very variable and not very different, either from L. seyboui , or from L. eketi , but the wide black margin along the costal edge of forewings makes it more similar to L. seyboui .

On the underside of the hindwings, the widening of the second pale transversal band is more strongly marked in L. kiellandi than in the other two species, and forms a discoid spot around the middle of spaces 4 and 5; on the costal edge, the spot in space 7 is also larger.

It was probably the existence of the discoid spot that prompted Vande Weghe (2010: 349) to compare L. kiellandi to L. praestans Grose-Smith, 1901 and, looking at the underside of the hindwings only, the two species could indeed be confused, especially in the Kivu area of the DRC, where L. praestans is not rare. However, both the upper side of L. praestans and male genitalia are very different from those of L. kiellandi .

Given the variability of specimens from different populations, especially for the extension of the orange part on the upper side of hindwings, the DRC population cannot be considered as a distinct subspecies. Furthermore, it is uncertain whether a female of the DRC population has ever been captured, because dissection has shown that three specimens in the Ducarme collection identified as females from a photo were actually males.

Since the abdomens of several specimens from ABRI had been lost during transport, it has not been possible to compare the genitalia of the males of the Cameroon population to those of the more easterly populations. Paradoxically, the only female formally identified as such is from the Cameroon population, i.e. the Manki specimen (Figs 13–14 show its genitalia). Vande Weghe (2010, pl. 106) illustrates two of these specimens, one as male (Figs 41–42), the other as a female (Figs 43–44), but the latter is identical to the male from Minziro which was dissected for this work, and it is not certain whether it is female.

Male genitalia ( Figs 8 –12)

The figures show the genitalia of a male nominate kiellandi from the type-locality. The genitalia of the population from Kasugho (NE DRC) and of subspecies kakamegae (W Kenya) are identical to those of nominate kiellandi . All genitalia dissected were significantly larger than those of L. eketi (about 25%).

The saccus, as massive as in L. eketi , confirms the proximity of the two species; at the distal end, the curled up part has two rounded lateral rounded processes (Figs 11–12), but there is no notch, and we cannot speak here of a guide for the penis (Fig. 12). At the distal end of the valves, the internal process is not very different from that of L. eketi , but the distal one is very reduced, almost vestigial ( Fig. 8 , 10 & 12); on the other hand, the dorsal edge of the valves shows a short, pointed process, which is absent in L. eketi ( Fig. 8 ). The uncus is slightly more indented than in L. eketi (Fig. 9), and the distal part of the subunci is shorter ( Fig. 8 ). The penis, appreciably longer, is hardly curved ( Fig. 8 ); it is not really twisted, but its distal end is turned slightly to the left (Fig. 10).

Female genitalia (Figs 13–14)

As mentioned above, only the female from Manki ( Cameroon) was dissected.

The ostium bursae is located between sternites 6 and 7 (Fig. 13); the strongly sclerotized sternite 7 is constituted by the three plates that form the sinus aginalis, a ventral plate and two lateral ones (Fig. 13).

The three plates seem to be merged, but the lateral plates are slightly separated from the edges of the ventral one, with which they laterally form an angle of about 40°. The sclerotization of sternite 8 is extremely reduced (Fig. 13). The ductus bursae is very short, but the bursae copulatrix is a rather voluminous ball; folded above the lateral plates, it considerably hinders their observation [this is why it is illustrated on a three-quartered view (Fig. 14)].

DNA barcode analysis

Almost all of the material examined for this work is too old for barcode sequencing to have been attempted, and the only sequence obtained was extracted from a male of L. kiellandi from Kivu (Maliva, in the north of the Mitumba Mountains). Although a single sequence is of little interest, Maximum Parsimony and Maximum Likelihood trees constructed with a hundred sequences belonging to numerous Liptena species show that this sequence constitutes the sister-group of [ L. fallax Libert, 2018 (x2) + L. eukrines, Druce, 1905 + L. lualaba diminuta Libert, 2018 (x4)].

With a bootstrap index of only 49, this result is not robust, and it is only given as an indication (the massive saccus of L. kiellandi and L. eketi is found in L. eukrines males but not in those of the two other species).

Figures

Figure 1 – Distribution map for L. eketi, L. seyboui, L. kiellandi

Plate 1 – A, B: L. seyboui, Alépé (Ivory Coast) A: ♂ Recto, B: ♀ Verso; C, D: ♂ L. eketi, Rhoko forest (Nigeria) C: Recto, D: Verso E, F: ♂ L. kiellandi, Maliva (NE DRC) E: Recto, F: Verso; G, H: ♀ L. kiellandi, Manki (Cameroon) G: Recto, H: Verso.

Figs 8 to 14 – Genitalia of Liptena kiellandi Male genitalia: 8 – left lateral view; 9 – dorsal view of uncus; 10 – dorsal view of the valves processes and the penis; 11 – saccus, ventral view; 12 – posterior view of the saccus and the valves processes. Female genitalia: 13 – ventral view of sternites 7 & 8; 14 – three-quartered dorsal view.