Eagris sabadius astoria, Holland, 1896

Cock, Matthew J. W. & Congdon, T. Colin E., 2011, Observations on the biology of Afro-tropical Hesperiidae (Lepidoptera) principally from Kenya. Part 2. Pyrginae: Tagiadini 2893, Zootaxa 2893 (1), pp. 1-66 : 9-18

publication ID

https://doi.org/ 10.11646/zootaxa.2893.1.1

persistent identifier


treatment provided by


scientific name

Eagris sabadius astoria


Eagris sabadius astoria Holland 1896 ( Figures 1–9 View FIGURE 1. 1–3 View FIGURE 2 View FIGURE 3 View FIGURE 4 View FIGURE 5 View FIGURE 6 View FIGURE 7 View FIGURE 8 View FIGURE 9 )

Eagris sabadius astoria is found in wooded and disturbed situations (e.g. gardens), and in Kenya is commonest in the Central Highlands, especially around Nairobi, and extending to 2000m or more on Mount Kenya. It is widespread in western Kenya, but not as common. It is common around Nairobi, but although the caterpillars can almost always be found (see Table 3 of collections), the adults are not that regularly seen. van Someren (1939) reports it from the Chyulu Hills: a few specimens at mid altitudes of 1,525m (5,000 ft.) and plentiful in the mixed forest of the lava ridges, and in the Shimba Hills, Sevastopulo (1974) found it uncommon around Makadara Forest, but these records may refer to ssp. ochreana ( Larsen 1991) . In Tanzania ssp. astoria occurs from northern Tanzania east and south to the Njombe highlands. TCEC reared ssp. astoria from Mufindi, Iringa Region (approximately 8°30'S, 35°40'E) in Jul 2007.

Adult behaviour

The males when seen are usually perched or flying quite high up (3–4m). They fly rapidly and erratically, but usually can be recognised by the yellow colour of the underside ( Figure 1.2 View FIGURE 1. 1–3 ) which fades in the field in old specimens. In Ngong Forest they can be spotted flying around trees adjacent to the open tracks, whereas in Kakamega Forest they seem to prefer to perch lower down on bushes beside tracks. At rest, adults hold their wings flat or inclined slightly downwards. Only occasionally do they come to feed at flowers.

Food plants

This species is polyphagous within the families Sapindaceae and Malvaceae (including the former families Bombacaceae , Sterculiaceae and Tiliaceae ) ( Figure 2 View FIGURE 2 ), and I have collected early stages from at least 14 different food plants (one more, det. as? Prunus sp. (001), was not adequately identified, but note there is an unpublished record from this genus at the end of this paragraph). I once found a pupa in Ololua Forest in a typical shelter on Achyranthes aspera ( Amaranthaceae ); however, this is not a food plant and doubtless the caterpillar strayed from the adjacent Hibiscus calyphyllus on which leaf shelters were present. Caterpillars seemed able to transfer between food plants quite readily, and most of those which I collected were reared through on Pavonia urens ( Malvaceae ) which I grew in my garden. The food plants, collection localities and months of collection are summarised in Table 3. In Tanzania, TCEC found caterpillars on Dombeya burgessiae ( Malvaceae ) and Allophylus spp. ( Sapindaceae ), usually on the latter. Published records of food plants are summarised in Table 4. Two unpublished food plants that would benefit from confirmation are Prunus africana ( Rosaceae ) and Celtis sp. ( Cannabaceae ) (I. Bampton, pers. comm.).

1? = Collection not reared to adult. Ovum

Ova are pale with a translucent shell and turn dark before eclosion. They are 3/4-spherical, 0.9 mm diameter at the base (n=8, range 0.81–0.96mm) with 13 ribs (n=5, range 11–14) which join near the top to form a narrow ring ( Figure 3 View FIGURE 3 ). The ova are laid on the leaf upper surface, close to the mid rib, often in pairs. As she lays the eggs, the female covers them with long, thin, pale scales from her abdomen tip ( Figure 3.1 View FIGURE 3 ). The scales turn darker with age, so that parasitized eggs, for example, appear brown ( Figure 3.2 View FIGURE 3 ). Eclosion is by eating a hole through the top of the egg, leaving most of the shell and its covering intact.

Leaf shelters

The first shelter is a one-cut shelter, usually oval (6–7 x 4–5 mm), cut from the middle of the leaf lamina ( Figures 4.1, 4.2 View FIGURE 4 , 5 View FIGURE 5 ), but occasionally cut from the edge. The second shelter may be similar (10–12 x 7–8 mm) ( Figures 4.3 View FIGURE 4 , 5 View FIGURE 5 ), but commonly a two-cut shelter, cut from the edge of the leaf lamina. The final shelter ( Figures 4.4 View FIGURE 4 , 5 View FIGURE 5 ) is normally cut from the leaf edge, but in the form shown in Figure 4.4 View FIGURE 4 with a long stem, and elongate shelter, truncate distally due to the edge of the leaf from which it was cut. Sometimes, however, this final caterpillar shelter is formed with one leaf on top of another (e.g. on Allophylus rubifolius ). In making these shelters, the caterpillar leaves uncut the main vein of the portion of the leaf which will form the lid of the shelter, which thus remains green.


Caterpillars in the first shelter feed nearby by perforating the leaf to eat out small compact holes ( Figure 4.1 View FIGURE 4 ). This feeding is sufficiently distinct to enable separation of the shelter from that of Netrobalane canopus which makes a similar shelter on several of the same Malvaceae food plants, but skeletonises the leaf as it feeds. There are five or, at least in captivity, six instars. Early stages can be found more or less throughout the year ( Table 3), but my collections suggest that they are much commoner from July to November and January around Nairobi; there is no obvious reason for the absence of collections in May and December. Average instar durations (and the range and number of observations) which I have noted include: L2 9 days (1); L3/Ln–2 9 days (7–13, 3); Ln–1 14 days (8–21, 7); Ln 20 (12–29, 8).

Instars 1 and 2 are dark, rich red with a dull black-brown head ( Figure 6.1 View FIGURE 6 ). The head of instar 1 measures 0.60 x 0.57mm wide x high (n=3, range 0.56–0.63 x 0.52–0.59mm), while that of instar 2 is 0.86 x 0.85 (n= 9, range 0.81–0.93 x 0.81–0.89).

Instar 3 is intermediate to the later instars, with a yellow dorsolateral line, and increasingly dark green, but often strongly tinted with red ( Figure 6.2 View FIGURE 6 ).

Penultimate instar (mixture of instars 4 and 5). Light green ( Figure 7 View FIGURE 7 ), due to a sprinkling of white dots on a green ground colour. The head varies from brown to dark brown. The head measures 2.38 x 2.29mm wide x high (n=31, range 2.04–2.81 x 1.89–2.78mm).

Final instar. The body ( Figure 8 View FIGURE 8 ) is similar to that of the penultimate instar. The head measures 3.25 x 3.13 mm wide x high (n=36, range 2.93–3.70 x 2.74–3.52mm). Although the ranges of size overlap, the head of female caterpillars in the final instar is on average slight larger than that of the male: 3.37 x 3.21mm (n=15) compared to 3.13 x 3.03mm (n=14), t 27 =3.87, P<0.001 (based on head capsule width). The colouring of the head is variable from uniform light brown ( Figure 8.2 View FIGURE 8 ), through intermediate forms with increasingly brown ground colour, and increasing amounts of dark brown at the vertex, in a stripe down the middle of the face, dorsally, around the stemmata, and in the darkest form the posterolateral region, leaving just a patch each side of the face that is paler brown ( Figure 8.1 View FIGURE 8 ). Although the body appears smooth to the naked eye, under 40x examination with a binocular microscope it can be seen to be covered with very short, simple, pale setae, with scattered longer, simple, pale setae. Fully mature caterpillars are often strikingly pale, almost white, and this seems to be associated with the uniform pale brown head ( Figure 8.2 View FIGURE 8 ). This variation is not linked to sex, food plant ( Malvaceae vs. Sapindaceae ) or the female colouring. The variation shown by this common species has made it difficult to distinguish early stages of other members of the genus without rearing them


The pupa is formed in the final caterpillar shelter, which is characteristically suspended above the leaf from which the flap is cut ( Figure 4.4 View FIGURE 4 ). As noted above, these shelters are not always formed on the host plant, and I once reared E. sabadius from a pupa in a shelter on Achyranthes aspera in Ololua Forest (see also E. lucetia and E. decastigma , for both of which pupae have been found on plants that are not food plants). The pupa is attached by the cremaster and supported by a Y-shaped silk girdle. There is no loose white waxy powder, but white markings on the pupa are due to a layer of white wax which can be easily scraped off. The pupal stage lasts about 14 days (19 observations, range 9–23).

A caterpillar collected 25 Jun 1987 on? Prunus sp. in Nairobi (87/1) pupated 12 Jul, and the following description was prepared 14 Jul; it turned dark on 26 Jul and a male emerged the following day. 15mm long, including frontal projection; intricately marked in white and brown, some of which has a bronze tone. Frontal spike 0.7mm long, 0.6 wide at apex which is weakly bifurcated. Eyes marked in white and two shades of brown. Dorsally on collar, two dark triangles with apex pointed forwards and rounded; thorax dorsally with obscure brown markings. Abdomen light brown dorsally; weak white dorsal line; narrow transverse dark line across dorsum on A3–A4; a broader, conspicuous bar each side of dorsal line A5; white dorsolateral area A3 and A4; below these a dark interrupted lateral line from A4–A9; below this a white transverse streak on anterior margin A5 and posterior margin of A4–A7; ventrally abdomen whitish yellow. Cremaster dark brown, plain, 3/ 4mm long. Wings with dorsum dark brown, remainder white with variable spots along veins. Other appendages pale whitish brown; apex of proboscis brown. T1 spiracle protuberant; orange-brown.

The frontal projection is variable which seems to be loosely linked to the size of the resultant adults. The length is 0.7mm on averages and varies 0.6–1.0mm, and the width at the apex averages 0.6 and varies 0.4–0.7mm. Pupae that produce small adults have the apex slightly bifurcate into dorsal and ventral halves, whereas pupae that produce normal adults are usually quadrifurcate; intermediate pupae show the dorsal protuberance divided laterally, but not the ventral one. On average, the proboscis sheath projects 2.4mm beyond the wing cases, varying 0.2– 3.2mm.

Natural enemies: I have reared egg parasitoids from this species on three occasions: Ngong Forest on Dombeya burgessiae ( Malvaceae ) 3.ii.1991; Ngong Forest on Grewia similis ( Malvaceae ) 1.x.1989, 16.ix.1990.

I have only once found a caterpillar attacked by a parasitoid. A second instar caterpillar collected from D. burgessiae in Ngong Forest on 3.ii.1991 was being fed upon by a solitary ectoparasitic eulophid larva which pupated naked in the host shelter a few days later, and emerged about six days after that.

The pupae are also attacked rarely, and only once did I rear a tachinid: from a pupa collected on Pavonia urens in Kakamega Forest (16.vi.1991, 91/33). The tachinid larva emerged from the pupa and was successfully reared. The host pupa was certainly that of an Eagris sp. but might have been that of E. lucetia , rather than E. sabadius . Unfortunately, I cannot now locate the tachinid or host pupa remains, although it is possible that they are in the NHM.

Having collected quite large numbers of caterpillars around Nairobi ( Table 3), and only once obtained one that was parasitized, I am confident, albeit somewhat surprised, that this species is not normally parasitized in the caterpillar stage. I am more surprised that the pupa does not seem to be commonly parasitized since in its shelter it seems dangerously exposed to small generalist pupal parasitoids such as Brachymeria spp. ( Chalcididae ) and eulophids, and yet of nine field collected pupae only one was parasitized. Perhaps further collecting will show it not to be as protected as it seems.

The comparison with Celaenorrhinus galenus opalinus (Butler) is interesting, as caterpillars of both are common around Nairobi. Celaenorrhinus galenus opalinus caterpillars are heavily parasitized while those of E. sabadius are barely affected, yet the adults of E. sabadius are equally common, or if anything slightly rarer than those of C. galenus .














Eagris sabadius astoria

Cock, Matthew J. W. & Congdon, T. Colin E. 2011

Eagris sabadius astoria

Holland 1896

ssp. astoria

Holland 1896

ssp. astoria

Holland 1896