Triskelionia compacta ( Evans, 1951 )

Larsen, Torben B. & Congdon, Colin E., 2011, Triskelionia, a new African genus of the Celaenorrhinini (Lepidoptera: Hesperiidae) and the promotion of T. compacta to species-status., Zootaxa 2931, pp. 53-58 : 55-58

Triskelionia compacta ( Evans, 1951) stat. rev., comb. nov.

Sarangesa tricerata compacta Evans, 1951 . Annals and Magazine of Natural History (12) 4: 1269 (1268–1272).

Type locality: Tanzania: “Mikindani, Tanganyika”. Type depository: 3 Natural History Museum ( BMNH), London (inspected).

Triskelionia tricerata ( Mabille, 1891) is a rather rare skipper with a western distribution in Africa, being known from the Gambia and Guinea to the Central African Republic and western to northern parts of the Democratic Republic of Congo. Only five males in the British Museum (Natural History) were listed by Evans (1937) – a few have been added since. The species has some similarity with the distinctly larger black-and-white members of the genus Celaenorrhinus , as well as certain Sarangesa , but is immediately distinguished by having four subapical spots that are very irregularly placed and the triskelion-shaped, fused cell-spots. Confusion with any other Afrotropical Hesperiid, even in the field, is impossible. It is illustrated in figure 2. We have occasionally collected this species in West Africa and have seen about 70 specimens in museum and private collections from the following countries: Gambia, Guinea, Sierra Leone, Côte d’Ivoire, Ghana, Togo, Nigeria, Cameroun, Central African Republic, Angola, and the Democratic Republic of Congo ( DRC) (the localities are indicated in the map in figure 3). There is only little variation in the material and no indication at all of geographical variation.

Triskelionia tricerata compacta ( Evans, 1951) was described in the genus Sarangesa and adequately described as follows, although the description does not fully show how different it actually is from T. tricerata :

“ compacta , [ssp] nov.: 3 Mikindani, Tanganyika, Jan. – May, 1897: Reimer: type B.M. Paler than tricerata Mabille : upf spots in cell and space 2 very greatly enlarged (2 mm wide) and nearly conjoined: no spot in space 9. Unh brown with dark spots, as in tricerata . Ƥ as in 3. 1 pair from the type locality.”

Both species are shown in figure 2. T. compacta was first figured by Kielland (1990) by the male holotype from London. It will be noted that the male illustrated here has only a single subapical spot, but there may be two or three. The wings are less irregular in outline than those of S. tricerata and there is a sexual dimorphism that is wholly lacking in S. tricerata . The two sexes are united by the overlay of ochreous-brown on much of the hindwing underside – darker in the male. Had a longer series been available to Evans, he would certainly have treated them as distinct species. All material available is from Mikindani and the Rondo Plateau area in Tanzania, not far from the coast and the Mozambique border, and close to each other. With the labelling practices of the time it is even possible that the type was from Rondo.

In addition to the morphological differences, the genitalia differ sufficiently to validate the species rank of T. compacta , as illustrated in figure 1, but they also show the close affinity of the two species.

The genitalia of T. tricerata (SCC 579 Batalimo, Central African Republic) have a less domed tegumen/uncus and a somewhat larger, denticulate gnathos than T. compacta (SCC 587 Rondo, Tanzania), as well as a longer free uncus. The valves have a different shape: the ventral edge of T. tricerata is evenly curved instead of almost straight. The pointed lower lobe of the valve is more massive than in T. compacta ; the dorsal edge of valve is almost straight, until the distal thorn suddenly curves sharply downwards almost to touch the lower lobe; in T. compacta the dorsal edge is gently curved, with a shorter, straighter, and more slender distal tooth. Both species have a small harpe on the valve. The saccus of both species consists of a narrow merging of the two vinculum branches, not solidifying into a large, strongly chitinized saccus as in other similar genera of the Celaenorrhinini . The genitalia differ about as much as one might have expected from the morphological differences between the two species. The valves, gnathos, and tegumen/uncus are compatible with some groups of Celaenorrhinus .

Habitat, habits, and biogeography. The species was first collected from Mikindani in eastern Tanzania in 1897 by a Herr Reimer, apparently a German colonial civil servant who in addition to other duties collected insects and other organisms, as they were encouraged to do by the authorities in Berlin. Mikindani was at the time an administrative centre on the coast of Tanzania, just north of the Mozambique border. Somehow, the holotype ended up in the NHM, London where Evans described it. This area would at the time all have been covered with coastal forest ( Zanzibar – Inhambane Coastal Forest Mosaic).

The African Butterfly Research Institute (ABRI), Nairobi decided that an attempt should be made to find this species again after it had not been recorded for more than a hundred years. Ivan Bampton and Colin Congdon went to the general area, where the forest habitat is now degraded to the point where forest in good condition hardly exists. However, it turned out that the host-plant, Dalbergia armata (Fabaceae) , survives in timber plantations and elsewhere with full canopy cover and three specimens were collected on the wing. A female was found ovipositing on this plant, a forest liana, and further 20 were bred. The early stages are fully described by Cock & Congdon (in press). The plant extends from eastern Tanzania to KwaZulu Natal. The early stages of T. tricerata remain unknown.

The range of T. tricerata covers the main forest zone of western and central Africa, though not reaching the Albertine Rift in the Kivu area or Uganda as do many other species. There are few or no records from the western DRC, the Congo Republic, or Gabon, areas that have been poorly investigated; we have no doubt that the species will be found there as well. This leaves T. compacta in the very special East African coastal forest isolated from its only congener by a gap of nearly 2,000 km (figure 3).

The closest parallel to such a distribution in East Africa seems to be the presence of Catuna sikorana Rogenhofer in the coastal forests eastern Tanzania ( Nymphalidae , Adoliadini ). There are four more or less widely distributed Catuna throughout the main forest zone, with C. crithea Drury reaching as far as Kakamega in Kenya.