Ampittia kilombero Larsen & Congdon

Larsen, Torben B., 2012, The genus Ampittia in Africa with the description of a new species (Hesperiinae; Aeromachini) and three new species in the genera Andronymus and Chondrolepis (Hesperiinae, incertae sedis) (Lepidoptera; Hesperiidae), Zootaxa 3322, pp. 49-62 : 51-53

Ampittia kilombero Larsen & Congdon , sp. nov.

Description Forewing 9–10 mm—one of the smallest of all African skippers. The male forewing upperside is a dark, warm brown with bright ochreous-orange markings consisting of: a long streak just below the costa to the end of the cell, only marginally intruding into the cell; a large orange subapical patch composed of spots in spaces 5 to 8 that are fully fused and extend slightly into neighbouring spaces; a small, narrow streak at the base of space 2; and a fully-fused orange discal band that extends from space 1b to space 3 and of almost equal width throughout. In well-marked individuals (as the type) the discal band almost connects with the large subapical patch. The hindwing has the same ground-colour as the forewing with some slight orange scaling in the cell and a fully fused discal band that stretches from spaces 1c to 6.

The male forewing underside is almost like the upperside with the orange markings generally slightly better developed. In addition to the upperside markings there is a fine ochreous marginal line and a few small ochreous submarginal spots, with those in spaces 4 and 5 best developed. The hindwing underside is dark brown with diffuse golden scaling, especially in the discal area. The orange postdiscal band is longer than on the upperside since it continues inwards to merge with the abdominal fold, which is of a lighter ochreous. The base of space 7 has an orange wedge that spills slightly over into the cell before the origin of vein 6. There is an orange transverse streak in the middle of space 7. As on the forewing there is a fine orange marginal line. There is a small diffusely fused double submarginal spot in spaces 4 and 5. The abdominal fold is light ochreous from base to tornus, narrowly bordered by brown along its edge. The underside contrast between brown and orange is almost as strong as on the upperside. The female is effectively identical with the male. The type series is very consistent.

Male genitalia: The male genitalia of the two species are very similar. The uncus, tegumen, and gnathos branches have the same structure and texture as Ampittia capenas (figure 2) but the lobes are triangular and more deeply incised. The cucullus is long and narrow, somewhat twisted, ending in a slightly spatulate tip, which is toothed. The two preparations in figure 5 both show how difficult they are to mount. The valve of A. kilombero seems to be rather wider than in A. parva as does the spatulate tip of the cucullus. The uncus, tegumen, and gnathos are similarly mounted in near-ventral view in both, showing the uncus to be broader and each of the lobes wider; as in A. capenas the entire structure cannot be mounted on a slide in lateral view (figure 5).

Diagnosis: The only similar species in Africa is A. parva which differs most noticeably from the new species in having the entire hindwing underside ochreous instead of brown, so that the discal band is only faintly defined. On the forewing upperside the subapical patch is smaller than in A. kilombero , while the hindwing discal band is shorter, though wider. This is accentuated on the underside where the band does not extend to space 6 and to the abdominal fold, which is the case in A. kilombero . The sexes hardly differ from each other in either species. Colour and patterns of both species seem to be very stable. We have inspected more than 150 specimens of A. parva from throughout its range, which encircles that of A. kilombero .

Holotype: 3 Mngeta River, Kilombero Valley, Udzungwa, 350 m (S 8o 19’ 5.35”; E 36o 7’ 49.24”), xi.2007 ( ABRI ex CC).

Paratypes: 8 3, 5 ƤƤ data as holotype (v.2008; v.2010) ( ABRI ex CC); 3 Chita, Kilombero Valley, (S 8o 30’ 24.46” E 35o 55’ 51.23”), v.2010 (35 km from Mngeta River) (CC, IB, MH).

Range, habitat and habits: Of the three African Ampittia only A. capenas has a relatively wide range being found from southeastern Kenya through most of Tanzania to the Shaba province of the Democratic Republic of Congo ( DRC), northern and eastern Zambia, Malawi, the eastern mountains of Zimbabwe and neighbouring parts of Mozambique. It is not usually a common butterfly.

Ampittia parva is limited to southeastern Kenya (only from a few recent records) and eastern Tanzania, extending southwards through the Eastern Arc mountains as far south as Songea and Ifakara. Rather surprisingly it is even found on Zanzibar. The distribution is patchy and the species generally rare (it was overlooked in Kenya until recently and not included in the book on Kenyan butterflies (Larsen 1991)). The typical habitats are open, lightly shaded areas in submontane forests as well grassy areas bordering lower level forests, where the grass is not too high. In dry habitats they are often found along forest streams. The flight is slow and low down among the grasses on which both sexes often rest.

Ampittia kilombero is known only from the two colonies in the Kilombero Valley at the southeastern end of the Udzungwa Mountains. There are colonies of the typical A. parva from elsewhere in the Udzungwa Mountains as well as further south and especially north thereof.

On the limited information available, A. kilombero is more strongly associated with swampy ground; both localities where it has been found are swampy. At Mngeta it was flying along the edge of open water. The flight is faster than that of A. parva . Colonies of A. parva can be found in forested areas higher up on mountain slopes under drier conditions. The nearest locality from which A. parva is known with certainty is Masagati Forest 147 km from Chita. From his field observations Congdon believes the two effectively to be ecological vicariants. Further work is needed to establish whether A. kilombero is essentially an inhabitant of the vast Kilombero Valley wetland, which borders the two localities where this species has been found. If this proves to be the case, it would species of more open wetland rather than forest.

Etymology: The species is given the name of the Kilombero Valley, in the Udzungwa Mountains, where it was found by Congdon and his colleagues.