Coeliades forestan forestan (Stoll)

Cock, Matthew J. W., 2010, Observations on the biology of Afro-tropical Hesperiidae (Lepidoptera) principally from Kenya. Part 1. Introduction and Coeliadinae, Zootaxa 2547, pp. 1-63 : 27-38

Coeliades forestan forestan (Stoll) ( Figures 18–30 )

The nominate subspecies occurs throughout tropical Africa, except Madagascar where it is replaced by subspecies arbogastes Guenée. This is the most common and widespread of the genus in Kenya, its range extending into some of the driest country in the North. The adult is most conspicuous, at least in the rainy season, in the drylands between Nairobi and the Coast. For example, van Someren (1939) reports it very numerous at low altitudes in the Chyulu Hills at c. 900m (3,000 ft.), and extending up to c. 1800m (6,000 ft.), and Sevastopulo (1974) found it common in and around Makadara Forest in the Shimba Hills throughout the year. However, caterpillars can be found all over the country on various food plants, for example I have found caterpillars several times in Kakamega Forest on Combretum sp. , a sprawling vine, but over many visits have never seen adults in the forest. This is also true of several other localities which I have only visited once or twice.

Adult behaviour

This species will fly and feed throughout the day—from 07.00h (Bahari Beach, north of Dar es Salaam) until dusk (Mazoe Dam, north of Harare)—a habit also noted by Trimen (1889) and Dickson & Kroon (1978). Migdoll (1988) states that it is more frequently seen in the early morning and late afternoon. Flight is rapid, and C. forestan is not readily separable from Pyrrhiades a. anchises and C. pisistratus in flight, except sometimes when close up or when at rest ( Figure 18 ). I have observed oviposition at 15.30h at Diani Beach and in the morning at Kibwezi Forest. On the latter occasion (23 Dec 1990, 90/127), one shoot of Combretum sp. (093) had three ova: two unhatched and one hatched, two on the abaxial surface of young leaves, and one against a bud on the shoot. Oviposition at this sort of density will lead to caterpillar density and damage such as that shown in Figure 22 (left) of the same food plant at the same location.

Coeliades f. forestan is an avid visitor to flowers, and I have seen it at Lantana camara , wild sunflower, Vernonia sp. and several other species; of these L. camara seems particularly preferred. Dickson & Kroon (1978) associate this species with hill tops, but that has not been my experience in Kenya. I have seen C. f. forestan at hill tops occasionally, but invariably it has been just passing through. Henning et al. (1997) state that they hilltop, but not for long. Coeliades forestan is noteworthy as the first species of Lepidoptera reported to use scent-marking. Williams & Woodhall (2006) describe males patrolling a linear territory, stopping briefly on leaves 1–2m above the ground and marking leaves by means of scale covered papilla-like structure extruded from the dorsal side of the terminal abdominal segment. The function of this behaviour has not been established, but is likely to involve repelling other males, or attracting females or both. It is unlikely that C. forestan is the only skipper showing this behaviour, but additional observations on the behaviour or the organ used have not been forthcoming as yet.

Food plants

This species is particularly striking for the number of different food plants it uses from different families ( Table 2 ), although some of these seem unlikely and would benefit from confirmation, e.g. Peacock’s (1913) record from Brassicaceae . In Kenya, I have recorded it from the species listed in Table 3 ( Figures 19–20 ). In addition, on a visit to the Ukiriguru Agricultural Research Institute, near Mwanza, Tanzania, I noted a specimen of C. f. forestan in the insect collection reared at Ukiriguru, 1.vi.1958 on “ C. procera ” by I.A.D. Robertson (ref. 3933). The food plant is only likely to be Calotropis procera (Apocynaceae) .

1 Also found on this food plant near Mangochi, Lake Malawi, Malawi, Jul 2001 .

Caterpillars will not necessarily transfer easily from one plant species to another. For example caterpillars collected on Indigofera tinctoria ( Figure 19 , right) at Diani Beach initially refused to feed on I. arrecta ( Figure 19 , left) from Nairobi (on which I subsequently found a caterpillar in my Nairobi garden), or on another Indigofera sp. (060) from Ngong Forest. Conversely however, caterpillars collected on I. arrecta (Magadi Road, 89/79) and Combretum sp. A (Kibwezi Forest, 90/10) readily accepted I. tinctoria grown at Nairobi.

The life history has been described by Aurivillius (1925), described and figured by Trimen (1889), Murray (1932) and Henning et al. (1997), and illustrated in detail by G.C. Clark (in Dickson & Kroon 1978) from South Africa.

Ovum

Ova are laid singly on the underside of leaves (sometimes upper side in the case of Indigofera spp. ) or the young stems of the food plants. Normally oviposition on the leaf lamina under surface is near the margin, but on a large leaf of Terminalia catappa , an ovum was once found on the midrib. The ovum ( Figure 21 ) is creamy white, 0.9 x 0.6mm wide x high, with 15–16 ribs; microsculpture similar to P. a anchises . It takes about 9 days to hatch according to G.C. Clark (in Dickson & Kroon 1978, Plate 1) in South Africa.

Leaf shelters

There is some variation in the detail of the leaf shelters, particularly the third and fourth shelters, linked to the texture and dimensions of the leaves of the varied host plants.

On the large, thin, soft leaves of Quisqualis indica the first shelter is made by cutting two ¾ oval flaps from each side of a major vein, and folding them upwards to make a rounded shelter on top of the leaf 4–5mm long x 2mm high; in the example examined, one flap cut from the edge of the lamina and the other from within the lamina. Where the shelter is made well within the lamina, just one three-quarter oval major cut would be needed for each flap making this a two–cut shelter, whereas if the shelter were made on the mid-rib near the leaf tip (as for P. a. anchises above) it would probably be a four–cut shelter. The example described above would be a three–cut shelter in the terminology of Greeney & Jones 2003). These are all functionally the same shelter, but formed in different ways depending upon where on the leaf the shelter is made, and how big the leaf is. The shelter is subsequently skeletonised by the feeding of the young caterpillar, perhaps because the caterpillar can thereby avoid leaving the shelter and being exposed to predators and other risks. The stage 2 leaf shelter is similar but larger, 11mm long, 5 mm high and 5 mm wide; again the shelter is skeletonised by the feeding caterpillar. The stage 3 shelter is made using the mid-rib adjacent to the bridge, either folding one flap onto the top of leaf (a type 4 one–cut shelter or a type 5 two–cut shelter), or folding two flaps together, similarly to the stage 1 and 2 shelters. Similar stage 1 shelters are made on the thicker, tougher leaves of Combretum sp. A ( Figure 22 ) and probably other Combretum spp.

On the large thick leaves of Terminalia catappa the first three shelters are similar ( Figure 23 , left), but stage 4 leaf shelters are also made by folding a broad leaf margin over or under in a type 1 no–cut shelter ( Figure 23 , right), e.g. Diani Beach (91/49), Lake Malawi, Malawi (01/102).

On the tiny leaflets of Indigofera spp. ( Figure 24 ) the first shelter is a type 6 four–cut shelter similar to those on thicker leaves, because the leaflets are so small. For the stage 2 shelter, two or three leaflets of one leaf are spun together, and for the stage 3 shelter, leaflets of two or more leaves are spun together—both type 2 multi–leaf shelters.

The first occasion on which I found caterpillars of C. forestan was on Lonchocarpus bussei in Arabuko Sokoke Forest on 23 Sep 1989 (89/85). A 9mm caterpillar was found in stage 1 leaf shelter made by cutting two ¾ ovals from each side of the mid-rib, and folding them upwards; one oval reached the leaf lamina, but the other didn’t, thus making this effectively a three–cut shelter; the edge of the flaps had been eaten so as to make them irregular, and at least one small perforation was eaten in one flap ( Figure 25 left). A final instar caterpillar had made a simple shelter between two leaves, the lower slightly convex to make a gap between them—a type 2 multi–leaf shelter. A similar stage 3 shelter containing a fifth instar caterpillar was observed on Combretum sp. B at Kakamega Forest (91/34), except that one of the leaves was of an adjacent Rubus sp. (Rosaceae) .

Caterpillar

There are five caterpillar instars in this species, unlike some other Coeliadinae treated here which may have six or seven instars. Caterpillars are variable, particularly with regard to the size of the black spots on the head. Caterpillars on Combretum sp. A in Kibwezi Forest, always had the head marked heavily with black spots often conjoined ( Figure 28 , lower row), especially in the first four instars ( Figure 26 ), whereas those on Indigofera tinctoria at Diani Beach were lightly marked ( Figure 28 , upper row); samples from other localities were too small to generalise in this way, but tended to be intermediate in markings. Henning et al. (1997) illustrate a final instar caterpillar similar to those treated here, with a lightly marked head comparable to Figure 28 (upper middle).

Instar 1. The newly hatched caterpillar (28/3) has the head light brown, very diffuse rows of conjoined spots; darker laterally and on posterior margin; 0.8 x 0.7mm wide x high (n=7); T1 green with dark transverse dorsal plate; body green.

Subsequent instars ( Figure 26 ) increasingly resemble instar 5 (below). The head capsules measure: instar 2, 1.3 x 1.3mm (n=6); instar 3, 1.6 x 1.6mm (n=11); instar 4, 2.5 x 2.5 mm (n=22) wide x high; most caterpillars in the early instars have the dorsal row of spots overlapping ( Figure 28 ), but in the latter and final instars these become increasingly separated in some individuals, but not in others ( Figures 27–28 ).

Instar 5 ( Figures 27–28 ). Individual 88/85A was collected on Lonchocarpus bussei in Arabuko Sokoke Forest on 23 Sep 1989, and went on to pupate on 3 Oct. The following description was prepared on the day of collection, when the caterpillar measured 31mm. Head 4.1 x 4.8mm wide x high (compared to an average of 4.2 x 4.2mm, n=17), light tan brown; two rows of black spots across face, the upper row of six with the four central spots oval, extended dorsally, the outer spots smaller, round and slightly lower; the lower row of five with three intermediate spots in centre, and smaller spots laterally and slightly ventrally; posterior margin narrowly black. T1 anterior 2/3 black, posterior 1/3 white. T2–T3 anterior ½ black, posterior ½ white; T3 with a narrow black transverse line in white band. A1–A7 anterior 1/3 black, posterior 2/3 white, with 3 or more transverse narrow black lines; an elongate transverse white streak dorsolaterally in black section, just posterior to middle of black area; sublaterally a white longitudinal line through spiracles; below this line greyish white with obscure darker markings. A8 anterior ½ black, posterior half white, orange-tan dorsolaterally; a single narrow black line in white band. A9 anterior margin black, wider dorsally; posterior to this a white band, angled towards anterior margin laterally, where it becomes orange-tan; posterior to this a black band to posterior margin, except at dorsum which is white. Shortly before pupation the white areas acquired a green tint and the ventrolateral areas a yellow tint.

Pupa

The final shelter may be formed from the last leaf shelter, or by folding under (or sometimes over) the margin of a large leaf such as those of Terminalia catappa ( Figure 23 , right), or between two leaves, one above the other as I have seen on Combretum sp. A .

The pupa of individual 89/8A was formed 3 Oct 1989 and a male emerged 17 Oct; it was described on 12 Oct ( Figure 29 , above). 24mm; dull green covered with white waxy bloom; short slightly upturned frontal spike, 1mm; spiracle T1 an erect blunt spike, black anteriorly and posteriorly; other spiracles black, but small and not conspicuous. Between the eyes a row of four black spots, with the widest gap in the middle. Wings pale under wax bloom; spot at end cell F. Underside of cremaster and the outline of a heart shape anterior to it, black.

There is minor variation in the black markings of the pupa. Thus, the pupa of individual 90/10F collected on Combretum sp. A at Kibwezi Forest had additional black spots: at the base of space 4 F, at base of proboscis, middle of leg of T1; as well as the abdomen having small patches bare of wax covering. Similarly, individual 90/10B collected on the same occasion had a black tip to the frontal spike, and black streaks on F costa near base, and vein 4 near base ( Figure 29 , below). Henning et al. (1997) illustrate a pupa, but in the dorsolateral view the degree of spotting cannot be seen.

I would expect pupation to be in these final shelters, although I have only found empty pupal cases in the field on T. catappa (twice). In captivity, the pupa is not normally formed in a leaf shelter, but often on the lid or between a leaf and the side of the container. This coupled with the fact that I have seldom found pupae in leaf shelters in the field leads me to suspect that pupae may also be formed elsewhere, probably amongst leaves or debris on the ground.

From material reared through at Nairobi, I have recorded 13 pupal durations which averaged 13.5 days with a range of 9–18 days.

Natural enemies

One of four ova collected 9 Jul 1991 at Diani Beach on Indigofera tinctoria was parasitized by a solitary black egg parasitoid (91/53).

The caterpillars are attacked by a braconid parasitoid which paralyses third and fourth instar caterpillars about to moult. The parasitoid larvae feed externally and then spin cocoons as a group ( Figure 30 ). I have reared this parasitoid from C. f. forestan on Indigofera arrecta in Nairobi (90/4), and Combretum sp. A at Kibwezi Forest (90/10A). The former produced nine males and one female, and the latter six males and nine females. What appears to be the same species was reared from P. a. anchises and Coeliades pisistratus . The biology and form of the cocoons appear to match very closely to those of Bracon spp. (e.g. Bennett 1960).

I also reared a solitary ichneumonid larval–pupal parasitoid from this host once. A 13mm caterpillar collected on Combretum sp. A at Kibwezi Forest (90/10D), 21 Jan 1990, pupated 11 Feb, and an ichneumonid wasp emerged 16 Mar.

Discussion

There seems no reason not to interpret these observations as representing a single highly polyphagous species, with a range of variation in caterpillar head markings from almost none, to very heavy spots running into each other. If the extremes, i.e. very light spots on Indigofera tinctoria on the coast and very heavy spotting on Combretum paniculatum in the South Nyambani Hills, had been examined in isolation, it would have been tempting to see them as separate species, but the range of continuous variation found on a variety of food plants shows otherwise. Clearly, interpretation of small numbers of isolated specimens needs to be done with caution.

Figures

FIGURE 18. Newly emerged female Coeliades forestan forestan, collected as caterpillar on Indigofera arrecta, Magadi Road, 10 Oct 1989 (89/79); adult 1 Dec 1989.

FIGURE 19. Indigofera spp. food plants of Coeliades f. forestan; left, I. arrecta, Nairobi, 12 Nov 1989; right, I. tinctoria, Nairobi, Feb 1990 (transplanted from Diani Beach).

FIGURE 20. Quisqualis indica, an introduced garden plant found in some hotel gardens at the Kenya coast is a food plant of Coeliades f. forestan, Diani Beach, 16 Dec 1989.

FIGURE 21. Ovum of Coeliades forestan forestan on Combretum sp. A, Kibwezi Forest, 23 Dec 1990 (90/127).

FIGURE 22. Leaf shelters of Coeliades f. forestan on Combretum sp. A, Kibwezi Forest, collected 21 Jan 1990 (90/10); left, heavily damaged shoot with various shelters, including one opened to show a large caterpillar; centre, stage 1 four–cut shelter with 4mm caterpillar (90/10E); right, stage 2 two–cut shelter, which contained a 19mm third instar caterpillar ready to moult (90/10D).

FIGURE 23. Leaf shelters of Coeliades f. forestan and characteristic feeding damage on Terminalia catappa, Diani Beach, 7 Jul 1991 (91/49); left, stage 3 type 5 two–cut shelter on mid-rib; stage 4 type 1 no–cut shelter by folding under leaf margin.

FIGURE 24. Leaf shelters of Coeliades f. forestan on Indigofera tinctoria, Diani Beach, 28 Mar 1989 (89/21); left, stage 1 four–cut shelter; centre, stage 2 multi–leaf shelter; right, stage 3 multi–leaf shelter.

FIGURE 25 Leaf shelters of Coeliades f. forestan; left, stage 1 shelter on Lonchocarpus bussei, Arabuko Sokoke Forest, 23 Sep 1989 (89/85), right, stage 2 shelter showing caterpillar visible within on unidentified food plant, Forêt du Yapo, Côte d’Ivoire, 4 Dec 1988.

FIGURE 26. Young caterpillars of Coeliades f. forestan; left newly moulted instar 2, 5mm caterpillar, collected on Combretum sp., Kibwezi Forest, 21 Jan 1990 (90/10E); right, instar 3 (or 2) on Indigofera tinctoria, Diani Beach, 28 Mar 1989 (89/21).

FIGURE 27. Instar 5 caterpillars of Coeliades f. forestan, collected on Combretum sp. A, Kibwezi Forest, 21 Jan 1990, and photographed 22 Jan; left, individual 90/10F; right, individual 90/10C.

FIGURE 28. Heads of instar 5 caterpillars of Coeliades f. forestan; upper row; collected on Indigofera tinctoria at Diani Beach, 2 Apr 1989, individuals 89/21D1, 89/21B1, 89/21C1; lower row collected on Combretum sp. A at Kibwezi Forests, 21 Jan 1990, individuals 90/10F, 90/10B, 90/10C.

FIGURE 29. Pupae of Coeliades f. forestan; left, newly formed pupa reared from a caterpillar collected on Lonchocarpus bussei, Arabuko Sokoke Forest, 23 Sep 1989, pupated 3 Oct, photographed 4 Oct (88/85A); centre and right, lateral and ventral views of pupa reared from a caterpillar collected on Combretum sp. A., Kibwezi Forest, 23 Jan 1990, pupated 5 Feb, photographed 11 Feb (90/10B).

FIGURE 30. Parasitism of Coeliades f. forestan by an unidentified braconid ectoparasitoid; left, paralysed fourth instar caterpillar pre-moult collected on Indigofera arrecta, Kyuna Estate, Nairobi, 14 Jan 1990, photographed 18 Jan (90/4); right, cocoons reared from parasitized fourth instar pre-moult collected on Combretum sp. A., Kibwezi Forest, 21 Jan 1990, cocoons formed 22 Jan, photographed 23 Jan (90/10A).

Tables

TABLE 2. Recorded food plants of Coeliades forestan forestan. Localities preceded by a question mark reflect a food plant record from a national or regional work, which may or may not be based on observations in that country or region. What appear to be repeat records from other regions may also be listed against an earlier record.

Family Food plant	Locality	Reference
Apocynaceae
Marsdenia ?‘ senegalensis ’1	East Africa	van Someren 1974
Marsdenia schimperi (= Dregea schimperi )	East Africa	van Someren 1974
Marsdenia or Dregea sp.	?Kenya	Sevastopulo 1975; Henning et al. 1997
Brassicaceae
Brassica oleracea (Gongylodes group) kohlrabi or German turnip	Nigeria	Peacock 1913; Smith 1965
Chrysobalanaceae
Parinari excelsa (= curatellifolia)	Southern Africa	Pringle et al. 1994; Henning et al. 1997
Combretaceae
Combretum sp.	Malawi	Gifford 1965
	?Southern Africa	Pinhey 1965
	?East Africa	Sevastopulo 1975
Combretum apiculatum	South Africa	Murray 1932, 1959; Dickson & Kroon 1978; Migdoll 1988; Pringle et al. 1994; Henning et al. 1997
Combretum bracteosum	South Africa	Murray 1932, 1959; Dickson & Kroon 1978; Migdoll 1988; Pringle et al. 1994; Henning et al. 1997
Combretum paniculatum	East Africa	van Someren 1974
Combretum racemosum	Côte d’Ivoire	Vuattoux 1999
Quisqualis indica	Uganda (Kampala)	Sevastopulo unpublished2
Quisqualis sp.	?Southern Africa	Henning et al. 1997
Terminalia catappa	Mauritius	Williams 1989; Davis & Barnes 1993
Terminalia	?East Africa	Sevastopulo 1975; Henning et al. 1997
Triaspis odorata	Côte d’Ivoire	Vuattoux 1999
Fabaceae
Canavalia ensiformis (= gladiata)	Malawi	Smee 1944
	Kenya, Uganda	Le Pelley 1959
	Mauritius	Williams 1989; Davis & Barnes 1993

...... continued on the next page

TABLE 3. Food plants of Coeliades forestan forestan recorded by the author in Kenya.

Family Food plant	Locality	Month(s)
Combretaceae
Quisqualis indica	Diani Beach	Dec
Terminalia catappa	Diani Beach & Mtwapa1
Combretum paniculatum	South Nyambani Hills	Apr
Combretum sp. A (081)	Kibwezi Forest	Jan, Mar, Jun, Dec
Combretum sp. B (093)	Kakamega Forest	Jun, Jul
Fabaceae
Indigofera tinctoria	Diani Beach	Mar, Jul
Indigofera arrecta	Magadi Road	Oct
	Nr Kisii (Kadera God Forest)	Aug
	Nairobi	Jan
Indigofera . sp.	Kakamega	May, Nov
Lonchocarpus bussei	Arabuko Sokoke Forest	Sep