Artitropa erinnys vansomereni Riley, 1925

Cock, Matthew J. W., Congdon, T. Colin E. & Collins, Steve C., 2015, Observations on the Biology of Afrotropical Hesperiidae (Lepidoptera). Part 8. Hesperiinae incertae sedis: Dracaena Feeders, Zootaxa 3985 (3), pp. 301-348 : 327-332

Artitropa erinnys vansomereni Riley, 1925

Riley (1925) described this subspecies based on 26 specimens reared in Nairobi by V.G.L. Van Someren. It was subsequently misspelt vansommereni ( Evans 1937, Larsen 1991). It appears to be endemic to the Central Highlands of Kenya. Larsen (1991) refers to it as occurring in the Central Highlands, including Nairobi, Meru and Ngong, and MJWC has found it at Thika and Tigoni, as well as Nairobi.

Food plants. Riley (1925) gives the food plant as Dracaena, and quotes from a letter by V.G.L. Van Someren that A. e. vansomereni feeds on D. reflexa . Van Someren (1974) lists Dracaena reflexa , D. fragrans , D. afromontana , D. mannii (as D. nitens ) as the food plants for A. erinnys ‘and subspecies’. Here we try to interpret these records in terms of the different subspecies of A. erinnys . Van Someren reared subspecies vansomereni from Nairobi, Ngong and Meru and in the 1950s he reared radiata from Teita and Rabai Hills (BMNH accessions). Dracaena reflexa is not a Kenyan species ( Beentje 1994, Govaerts 2013), and this record may refer to an ornamental planting or a misidentification; since Riley (1925) quotes Van Someren that D. reflexa is a food plant of A. e. erinnys , we assume that Van Someren’s (1974) record can be attributed to this subspecies. In Kenya, D. fragrans is restricted to the south-west around Lake Victoria; as Van Someren did not rear any Artitropa spp. from this area, the record probably also represents a misidentification or a cultivated plant. Van Someren (1974) does not list D. steudneri , the common food plant of A. e. vansomereni around Nairobi (MJWC observations, Larsen 1991), and since D. reflexa , D. fragrans and D. steudneri are all superficially similar, we suggest one or both records of D. reflexa and D. fragrans may have been a misidentification for D. steudneri , as the food plant of subspecies vansomereni. Dracaena afromontana is a species of moist forest or bamboo mostly found in central and western Kenya, but there is a record from the Teita Hills ( Beentje 1994), so this food plant could apply to either or both A. erinnys vansomereni or A. e. radiata , but is more likely to be the former. In contrast, D. mannii is only found in the coastal area of Kenya, and so this record must apply to A. e. ehlersi , unless it was being grown outside its natural distribution, e.g. at Nairobi. SCC has observed that D. mannii growing in his Nairobi garden is readily accepted by A. erinnys vansomereni .

Caterpillars of A. e. vansomereni are common around Nairobi as its food plant, D. steudneri , is frequently planted in public places (e.g. labelled specimens at the National Museums of Kenya), gardens and as a pot plant. At times the caterpillars are a pest of these ornamental plants, particularly small plants growing in shaded positions. MJWC had a potted plant under cover by his Nairobi house which was almost completely defoliated three times, once following the release of newly eclosed caterpillars onto it. After three years it was still less than a metre tall, spindly with only a few, small leaves. An identical plant was grown in direct sunshine in the garden and after three years, it was more than 2m tall with many large leaves (Figure 2.1).

Life history. V.G.L. Van Someren (in Riley 1925) provides a description of the biology: “The eggs are laid singly for the most part, though two may occasionally be deposited close together. They are laid either on the leaves, upper or under surface, frequently near the base, or on the trunk of Dracaena reflexa . They are generally salmon-pink when laid, but I have seen a female depositing pure white ones. The eggs are hemispherical and attached by the flattened side. There is a slight circular depression, on the top, and a series of ridges radiating from the central spot. The eggs turn greyish as development proceeds, and the larvae emerge within 10 days usually, though some carry over for longer periods. The newly hatched larva is 5 mm. long and yellow with a black head. With the second moult the colour changes to a pale greenish-white and the head piece becomes yellow with black spots arranged as shown on Plate XXXV, fig. 6 [similar to Figure 26 ], thus differing from the Uganda race [i.e. A. reducta ], which has only two spots.

“The larva, even in the early stages, adopts the habit of protecting itself by cutting the edge of a leaf and bending this over, securely fastening it down with silk threads in two or three places; from this tunnel it emerges half out and devours the leaf substance immediately surrounding. On the slightest alarm it withdraws itself into· its protective covering. These tactics are adopted throughout the whole larval period—the whole larva is seldom exposed except when changing and reconstructing a tunnel—nevertheless a fair proportion of larvae are parasitized. I have not yet completed the various parasites, but so far two species of ichneumons have been bred out.

“The larva pupates in a tunnel, very similar in construction to its ' hide up' but in addition it spins a fairly close web over the whole of the inside and blocks up the exit with a flimsy web which can be easily torn by the emerging butterfly. The pupa is attached at its anal end and by a body band. A white powdery substance is secreted by the larva during the 24 hours before it finally pupates. The larval period varies according to the freshness of the leaf––between 14 to 20 days, and the pupal stage lasts 20 days or more.

“Judging by the number of eggs, etc., found on one plant one would be led to think that the insect was common, but although I have kept a sharp lookout, very few imagines have been taken or even seen. They are crepuscular, extremely shy, and rapid in flight.”

Van Someren’s collection in the BMNH accessions includes a female and a reared male from Meru, 11 adults from Ngong and 30 from Nairobi, nearly all reared. There is a reference label to the Dry Early Stages Collection (numbers 4691–4695) adjacent to the series from Nairobi. Specimen 4695 is an emerged pupa with associated head capsule which resembles that documented here ( Figure 26 ). Specimens 4691–4694 are emerged pupae with associated head capsules, but the head capsules resemble those of A. milleri ( Figure 12 ). The Van Someren collection in the BMNH accessions also includes material of A. e. radiata , and there is a reference label to the Dry Early Stages Collection (numbers 4696–4703) associated with these; specimen 4696 is an emerged pupa with no associated head capsule, and specimens 4597–4703 do have head capsules associated and these are all similar to those of A. e. vansomereni ( Figure 26 ). None of these emerged pupae was labelled by Van Someren, and would have been pinned near the reared specimens to show their affiliation. It seems likely that the unlabelled pupae got mixed up between A. erinnys and A. milleri at some stage. Certainly if these associated early stages remains had been labelled appropriately (cf. Cock 2010), this confusion should not have arisen.

The information presented below is based on material from Nairobi and Thika on D. steudneri unless otherwise indicated. The information on the life history summarised by V.G.L. Van Someren in Riley (1925) above generally agrees with that presented here, except that our observations differ on the colour of ova and how they change.

Adult behaviour. The females ( Figure 24 ) fly in the late afternoon, or earlier on overcast days, to lay ova. MJWC watched one female around mid-afternoon busily laying ova in Karen against a backdrop of a black sky and approaching rain (the basis of the observations referred to in Larsen (1991)). V.G.L. Van Someren (in Riley (1921) above) refers to them as crepuscular, but while they are active until dusk, we know of no early morning observations.

Ovum. Ova (Figure 25.1–2) are smooth, and hemispherical in shape. They are laid singly on the leaves, and MJWC and SCC observed that when freshly laid they are white, but progressively change to dark pink as they mature over the next 10–12 days. Van Someren reports that they are usually salmon pink when laid and turn greyish (above), but we do not accept this. Unfortunately MJWC only preserved one ovum that was unequivocally A. e. vansomereni , and that had a diameter of 1.61mm. Eggs collected at Thika, where both A. e. vansomereni and A. m. coryndon occur together, may have been of either species; they averaged 1.51 (range 1.29–1.64) x 0.92 (range 0.78–1.06) mm diameter x height (n=44), and showed no indication of separating into two size classes. The ova have numerous fine ribs, becoming more irregular dorsally, which are difficult or sometimes impossible to count in dorsal view. In lateral view, they can be seen more clearly and the width of five gaps between ribs was measured and used to extrapolate the number of ribs based on the circumference calculated from the diameter of each ovum. Checking the estimates where the number of ribs could be counted in dorsal view suggested that the individual estimates were within 10% of the correct number. For the collection of eggs from Thika, the estimated mean number of ribs is 49 (range 37–67). The newly hatched caterpillar eats the shell except for the base.

By marking the position of ova on a small tree of D. steudneri in his Nairobi garden, and monitoring their fate, MJWC found that several disappeared for no apparent reason. Because the leaf surface is smooth and seems waxy, the ova are quite easy to remove by exerting a little pressure, and it seems likely they were either carried off by generalist predators such as wasps, or they were brushed off due to the action of wind and/or rain.

Leaf shelters. The newly hatched caterpillar rests on the under surface edge of the leaf and spins a silk pad on which it then rests. The first shelter is then made by cutting a notch basal or distal to this pad and rolling the section of leaf with the pad under the leaf (Figure 25.3). The caterpillar then starts to feed basal to the shelter; if the first notch was distal to the shelter it now appears as if the caterpillar had cut a notch at each end to roll the shelter, and if the initial notch was basal to the shelter this is simply expanded. In Figure 25.3, the shelter would have been initially partially rolled with the silk lines second and third from the left, but is now rolled more tightly using the silk lines far left and far right, which have left the earlier silk lines slack as the leaf roll is pulled tighter to the leaf. Larger caterpillars use shelters made by cutting a notch at one end (usually the distal end), and rolling a flap tapered to the basal end. Fifth, final instar caterpillars may stay in such shelters, or in the case of small, shaded plants with less robust leaves, they may move to the base of a leaf to make a shelter. Pupation is in the final caterpillar shelter; when the shelter is closed, there is a sheet of silk webbing across the anterior end; a stout single strand girdle at A1 of pupa; the shelter is lined with white waxy powder but at most only a light covering on the pupa.

Caterpillar. The head capsules from multiple rearings indicate that there are normally five instars, the head capsules measuring 0.86 x 0.94 (n=1), 1.18 x 1.33 (n=1), 1.67 x 1.86 (n=7), 2.35 x 2.74 (n=6), and 3.66 x 4.41mm (n=5). They are 7–16% larger than those of the corresponding instar heads of the co-occurring A. milleri coryndon .

The first instar is green with a black head, pronotum and anal plate (Figure 25.3); six pale setae on posterior margin of anal plate, the two each side of the centre are longer, 0.8 mm (cf. Figure 32 of A. e. erinnys ). By the end of the instar the body becomes whitish. The second instar resembles the late first instar, but the posterior margin of the anal plate is pale. In most individuals of the third instar, the shape of the head markings is obscurely visible, as a dark four-lobed marking on the face outlined in brown; the anal plate is dark, pale on posterior margin with no obvious setae.

By the fourth instar the black on yellow markings of the next, final instar are apparent on the head. The fifth instar caterpillar ( Figure 26 ) is striking, and the following is based on collections 88/17 and 88/34. The head is shiny, rugose; yellow-straw in colour with black markings (Figure 26.1–2). Pronotum with a narrow black transverse line ending laterally in a large rounded black spot. Body light bluish grey-white; darker diffuse dorsal line, stronger A8–A9; spiracles white; all legs concolorous. Anal plate slightly chordate, about 3.5mm wide x 3 mm long, light translucent brown, central 3/4 of anterior margin black, shading diffusely into posterior 1/2 to 1/3 of plate. The caterpillar grew to 42mm when mature, at which time the wax glands are conspicuous, occupying the ventral area from A1 to A8 apart from the prolegs (Figure 27.1).

The fourth and fifth instars take 5–6 and about 15 (range 12–19) days respectively. The total caterpillar development of 14–20 days given by V.G.L. Van Someren in Riley (1925) seems too short. Of the different subspecies of A. erinnys treated here, the caterpillars of A. e. vansomereni are the most distinctive and easily recognised by the markings of the head, but even here, there are atypical specimens (e.g. Figure 26.3) which could be confused with those of other subspecies. A better understanding of the extent and frequency of individual variation would be helpful to interpret the descriptions and figures that we present below.

Pupa. The pupa (Figure 27.2) and its shelter (above) are similar to those of other members of the genus. Pupa 88/17 collected on D. steudneri , Nairobi, 25 Aug 1988, measured 33mm, but others (e.g. 88/34, measured up to 36 mm); the proboscis sheath protrudes 5mm beyond cremaster tip; thorax and head translucent white; abdomen opaque white; thin dark dorsal line on T2; dark markings at the dorsal posterior part of the thorax; discontinuous dark line around the top of the eyes and frons; T1 spiracle dark (or brown 88/34); other spiracles inconspicuous. Pupation takes on average 19 days (range 17–20) in Nairobi (V.G.L. Van Someren (in Riley 1925) states 20 or more).

Natural enemies. Ova of Artitropa sp(p). collected on D. steudneri at Thika were heavily parasitized by at least two parasitoid species: of nine ova collected 17 Oct 1989 (MJWC 89/58), three ova produced specimens of a large, dark, solitary egg parasitoid, and six produced several specimens (2, 2, 4, 7, 8, 9) of a small, dark, gregarious species; in July 1998 (MJWC 98/207G), one ovum of nine hatched, four produced a large, dark, solitary egg parasitoid and four produced several specimens (2, 4, 4, 7) of a small, dark, gregarious species. Ova which MJWC collected in his Nairobi garden in 1988–89 were unparasitized, but in February 1990 MJWC collected six ova, one of which hatched normally, two were parasitized and three failed to hatch (MJWC 90/23); in June 1990, one ovum out of 12 hatched, five were parasitized, yielding 2, 4, 5, 6, and 8 gregarious egg parasitoids, and six failed to hatch (MJWC 90/62).

Young caterpillars of A. milleri coryndon and A. erinnys vansomereni on D. steudneri at Thika are attacked by Elasmus sp. ( Eulophidae ), a gregarious exoparasitoid, which forms naked black pupae in the leaf shelter of the host caterpillar. A third instar caterpillar of A. e. vansomereni had nine associated pupae (MJWC 88/61A) which produced two male and seven female Elasmus sp., while a younger caterpillar collected on the same occasion (which may have been A. milleri coryndon ) had three associated pupae (MJWC 88/61B) which produced one male and one female Elasmus sp. Two other examples of small caterpillars that could have been A. e. vansomereni or A. m. coryndon were collected at Thika, 7 Oct 1990: a shelter with two pupae but no caterpillar remains (MJWC 88/ 90B), and one with the dried remains of a small caterpillar and one pupa (MJWC 88/90B), both collections yielding one female Elasmus sp.

V.G.L. Van Someren (in Riley 1925) mentions two species of ichneumonid parasitoids. One third instar caterpillar from Thika (MJWC 88/61C) was attacked by a solitary ichneumonid, which span a scruffy, grey cocoon in the leaf shelter, but failed to emerge.

Figures

FIGURE 12. Final instar caterpillar of Artitropa milleri coryndon, collected on Dracaena steudneri, Thika, Kenya. 1 – 2, collected 22 Dec 1988; moulted to final instar 11 Jan 1989; photographed 16 Jan 89; pupated 27 Jan; 88 / 122 E; 1, head, lateral view; 2, head, anterolateral view. 3, anterodorsolateral view; collected 26 Jun 1988; moulted to final instar 12 Jul 1988; photographed 18 Jul, pupated 27 Jul; 88 / 52 B 3.

FIGURE 24. Adult female Artitropa erinnys vansomereni. 1, upper side, reared from Dracaena steudneri, Nairobi, 4 Feb 2004 [SCC, photo TCEC]; 2, underside, collected as pupa on D. steudneri, Thika, Kenya, 17 Sep 1989; emerged and photographed 24 Sep; no voucher; 89 / 58 A.

FIGURE 26. Final instar caterpillar of Artitropa erinnys vansomereni, on Dracaena steudneri. 1, dorsolateral, anterior and posterior views; collected Thika, 2 Jul 1988; photographed 18 Jul; pupated 24 Jul, 88 / 52 A 1; 2, head, anterior view; collected Nairobi; photographed 13 Sep 1988; 88 / 75 A 1; 3, head with atypical light markings, anterolateral view; Nairobi, 17 Jan 2006 [TCEC]; 4, head with atypical dark markings, anterior view; Nairobi, 23 Dec 2002; 5, dorsolateral view, Nairobi; photographed 25 Nov 89; 89 / 94.

FIGURE 32. First instar caterpillar of Artitropa erinnys erinnys, newly hatched from ovum laid on Dracaena aletriformis, Gillitts, Durban, South Africa, photographed 20 Oct 2012 [© S. Woodhall].