Revision of Afrotropical Chonocephalus Wandolleck (Diptera: Phoridae)

Chonocephalus justini sp. nov., C. modestus sp. nov. and C. pudicus sp. nov. are described from the Seychelles; C. cummingae sp. nov. is described from Madagascar, South Africa and Zimbabwe; and C. cautus sp. nov. and C. steineri sp. nov. from Madagascar. C. dimakae Paulian is synonymized with C. depressus Meijere. A key to males and a partial key to the females of the 14 Afrotropical species are provided.


Introduction
The tiny scuttle flies of the genus Chonocephalus Wandolleck are noted for their sexual dimorphism, with the females being flightless but the males being normally winged. Many species breed in ripe fruits, while some exploit other sources of fermenting detritus or plant material, and some are fungivores. The males transport females to suitable larval pabula during nuptial flights (Disney 1994).
Many casual collectors seem unaware of the sexual dimorphism in this genus and consequently have tended to collect only one sex. The naming of undescribed species on the basis of one sex only created the current taxonomic chaos. The growing list of undescribed males and females had become such that I am now naming undescribed males but assigning code letters only to undescribed females as yet unassociated with their males. Such a policy runs the risk that we may describe as new some named species already known in the female sex only. As these are associated with their males, we will thereby create some synonyms. This is probably a smaller price to pay than a confusing proliferation of species (described from males alone) known by code numbers only alongside a smaller list of females known only by code letters. Museum collections abound in misidentified specimens and many species awaiting description. In order to provide the basis for moving forward, all named species in literature were recently reviewed and keys to the Nearctic and Palaearctic species were provided. Furthermore, this revision of the 43 previously recognized species on the world list reduced this to 32 valid species, with the possibility that some only known in the female sex might be the missing females of some species only known in the male sex (Disney 2002). In this paper I describe six new species of Afrotropical species, propose one synonym and provide keys to both sexes for the species of this region. Some females only known in this sex are given code letters only.

Methods
The identification of members of this genus requires slide-mounted specimens. Indeed the females can only be reliably identified by means of internal features of the abdomen (see below). The preferred mounting medium is the so-called Berlese Fluid gum chloral medium, used with coverslips of 10 and 6 mm diameter (Disney 1983(Disney , 1994. Upton (1993) has advised against use of this mountant, but some of his criticisms are based on a number of dubious assumptions and misapprehensions (Disney 2001).
For males, the recognition of the species depends primarily on the details of the hypopygium. These are best detached from the rest of the fly and mounted under a separate 6 mm coverslip with the underside uppermost. When several specimens are available then it is useful if some are mounted to display the left face of the hypopygium. Females are mounted whole with the dorsal face uppermost. When several specimens are available it is useful to detach the legs from at least one specimen and mount these under separate coverslips.
The genus is readily identified in the most recent key to world genera (Disney 1994), with the modification that the genus Epichonocephalus Schmitz was synonymized with Chonocephalus in my recent revision, which provides a detailed diagnosis of the genus in both sexes (Disney 2002). The wings of the males are distinctive ( Figure 40) as is the general appearance of the females (Figure 32).

Males
The recognition of species depends primarily on the details of the hypopygium, which is at first sight very complex. However, if one ignores the, often elaborate, penis complex (X), the details of the epandrium (E) and hypandrium (H) that are referred to in the key below are indicated in Figure 1. This represents a generalized, stylized, hypopygium. While the anal tube (A) is not utilized in the recognition of species it is a useful point of reference. Any one of the labelled structures may be modified and may sometimes be very elaborate in form. In the inferred groundplan state the rear margin of the hypandrium bears four processes. These are the gonopods (g) and a pair of median lobes (l). In the plesiomorphic state the gonopods articulate with the dorsal face of the hypandrium, but more frequently they arise from it. One (usually the left) or both may be reduced or absent. The median lobes (l) may form a single bilobed structure or be reduced to a single lobe (which may be greatly elongated) or be lost altogether. Ventrally each side of the epandrium (E) typically bears an anterior process (a) and a posterior process (p). In some species the anterior processes (a) may appear to be appendages (attaching to the inner face of the epandrium) rather than simple projections of the epandrium. A presumed true left surstylus or clasper (c) is frequently present (e.g. Figure 18), variably fused to the epandrium, and is sometimes elaborate. Occasionally it is fused to the bridge (b), which is sometimes present and probably represents sternite 10. The most anterior (lower) bristle of the left side of the epandrium is sometimes modified into a spine (s), which may be further modified.

Females
The morphology of the females has been discussed elsewhere (Disney 1986(Disney , 2002. It was concluded that the first, long, abdominal tergite is a composite of T1+T2 fused together, and that the last, also relatively long, fully developed tergite is T7 ( Figure 32). The microtrichia, lying between the hairs on these tergites, are frequently reduced in size and consequently tend to be more densely packed (cf. Figures 38, 39). Otherwise the taxonomically useful features are mainly the form and size of the modified abdominal tergite (T8) and sternite 8 (S8), the internalized sternite 9 (furca) and the spermatheca. Typically the remnants of T9 are variably fused to the anterolateral apodemes of T10. These structures are most fully developed in C. blackithorum (Figures 2, 3). The form and dimensions of sternite 8 have proved to be especially valuable taxonomically. This sternite has seemingly evolved from an isosceles triangular sternite whose anterior end continues internally as an anteriorly directed apodeme (e.g. Figure 3). In many species the external sternite is greatly reduced or lost while the internal apodeme has increased in length. In some species a small sternite is retained as an externally visible projection (e.g. Figure 9). Tergite 8 is frequently reduced or seemingly absent. In the latter case it is present but only very lightly pigmented. Its presence may be discerned in many such cases by locating a minute hair at each posterolateral corner (e.g. Figure 8). However, these hairs are sometimes absent. There can be some confusion between the furca and spermatheca. Both are subcircular structures lying close to each other between T8 (when present) and S8. The spermatheca is typically a lightly pigmented bowl-shaped structure, as in C. blackithorum (Sp in Figure 3) while the furca is hoop-shaped and paler (S9 in Figure 2). However, in many species one or other or both of these structures may be very pale and it may require critical adjustment of the microscope's lighting in order to discern them both. In observing a female mounted on a slide with its dorsal side uppermost, as one focuses downwards the sequence of encounter is T8-furca (S9)-spermatheca-S8 (e.g. Figure 26). Where an egg is in place ready for laying, then it lies below T8 and the furca but above the spermatheca and S8. Indeed, the furca is often displayed to view most clearly against the background of such an egg. When no egg is in this position the furca may be tilted sideways on (as viewed from above or below). Key to females of Afrotropical species It should be noted that abdominal tergite 3 (T3) is the second apparent tergite as T1 is vestigial or fused to T2.
Caution: these keys to females not only include species given code letters only (until they can be linked to their males) but omit several species, as the number of these coded females is smaller than the number of species still only known in the male sex.
1 Microtrichia, between the hairs, on frons as strongly developed as those on abdominal tergites and not so densely packed (e.g. Figure 39). Without obvious submarginal dark bands at rear of abdominal tergites. Tip of sternite 8 (S8) and spermatheca as

Review of species
The descriptions of the new species are brief as the significant features of the males are largely restricted to the details of the hypopygia. Likewise in the females there are relatively few features that distinguish the species apart from the details of the abdominal terminalia. The formal diagnoses, therefore, are the features used to distinguish the species in the keys above.

Etymology
The name means careful or wary and refers to the fact that unless one is careful one may confuse this species with C. steineri.

Male
A generally brown species but mostly not dark. Antennae and palps light brown, the latter with numerous hairs but apical one only a little differentiated. All femora and tibiae light brown, but hind femur with a darker ventral band in basal half. Wing 1.

Etymology
The species is named after Meg Cumming, who collected the holotype and the only known females.

Male
Palps ovoid, 0.14 mm long and 0.07 mm wide, with about 24 hairs on outer face, the longest (apical) one being 0.07 mm long. Left side of epandrium with a strong spine and about six hairs, and a surstylus present (Figure 18). The right side has as many hairs but the most ventral one is developed as a differentiated longer bristle. The right hypandrial process dark brown and ending in a long curved point, the adjacent gonopd ending in a bristle ( Figure 19). The left process paler and not curved or so strongly tapered, and the adjacent gonopod more parallel-sided and with two pale hairs at tip (Figure 20). Legs brown, the femora being darkest and the tarsi palest, except the distal two-thirds of mid tibiae are white. Wings 1.0-1.3 mm long. Costal index 0.5-0.6. Membrane strongly tinged grey.

Female
Length 1.1-1.4 mm. Frons, thoracic notum and abdominal tergites brown, but not dark apart from dark brown rear margins (embracing a pale spot at base of each hair of posterior row) of T1+2-T6. These dark bands do not extend to lateral margins of the tergites. Eyes small, with 11-13 ommatidia. Antennae pale brown. Palps straw yellow, broad, with several hairs of which the apical and a pre-apical of the lower margin are strongly differentiated. Legs straw yellow variably lightly tinged brown, especially the femora. All tarsi with a ventral hair palisade on basitarsus only. Up to 20 hairs at rear margin of T3. T8 and S8 are a simple rod-shape, the furca is pale and smaller than the large spermatheca (Figures 21, 22). Microtrichia of abdominal tergites and thoracic notum dense and minute, but still larger than those on frons.
Chonocephalus depressus Meijere, 1912 (Figures 23-27, 38) Chonocephalus depressus Meijere 1912, p 151 (female only). Disney 1991, p 208 (male Beyer 1964, in Hardy and Beyer 1964, p 298. Disney 2002, p 15. Chonocephalus simiolus Beyer 1964, in Hardy and Beyer 1964, p 300. Disney 2002 I previously proposed four synonyms of this tramp species (Disney 2002). I now propose a further synonym. The type material of C. dimakae is lost. Furthermore, Paulian's (1958) descriptions are inadequate in the light of current knowledge. However, his figures of the male hypopygia are the most useful parts of his descriptions. Despite being shown in a tilted side view, his figure of the hypopygium of C. dimakae suggested it might be C. depressus. I therefore mounted the male hypopygia of this species in various ways until some achieved the same orientation as Paulian's figure for C. dimakae. Allowing for some simplifications in Paulian's figure (as with his figure of C. vadoni, which was correctly synonymized with C. heymonsi by Prado) these hypopygia are evidently the same. I therefore propose the synonymy of C. dimakae with C. depressus.  This is another tramp species whose original distribution is unknown.

Material
One male, Oman, Ruwi Waltayeh, 31 March to 2 April 1988, M. D. Gallagher (NMW) (5-145). Thirteen males, Yemen, Al Kadan, Malaise trap, light trap, 7 February to 31 March 1998, October to November 2001, April 2002; 19 males, Al Kowd, July to August 1999, February, May to July, October to December 2000, April to May 2001 19,20,23,24,29,31,74) This tramp species has been carried around the world by man. It has accordingly been repeatedly misidentified. It has been recorded from every biogeographical region, but is seemingly most abundant in Africa south of the Sahara. For example, it is known to be common in Kenya (Disney and Darlington 1998) and abundant in Nigeria (see below). It is clearly the dominant species in mainland Africa. Through the agency of man it is now reported throughout the warmer parts of the world and in glasshouses in temperate regions (Disney 1983(Disney , 2002.

Natural history
The larvae and puparium are described by Borgmeier (1935). A female caught on the edible paddy straw mushroom Volvariella (Plutaceae) was probably ovipositing on an overripe sporophore, as was the case with the females recorded on rotting Termitomyces (Amanitaceae) on a termite mound. It has also been reared from bread fruit (Artocarpus altilis (Z.) Frost, Moraceae) (Disney 1994).
Chonocephalus hibisci Paulian, 1958 ( Figure 46) Chonocephalus hibisci Paulian, 1958, p 13. No fresh material of this species has been seen. The female is not recognizable from the perfunctory description by Paulian.

Natural history
Reared from stem of Hibiscus sp. (Paulian 1958). The type series has been lost. One of a freshly collected series of males from the Comoros Islands has been designated the neotype (deposited in ZSM). The female is not recognizable from the perfunctory description by Paulian.

Natural history
Reared from fungi (Paulian 1958

Etymology
The name refers to the lack of any striking features.

Male
Frons dark brown, abdominal tergites paler, thorax somewhat paler and postpedicels and palps even paler brown. The latter with many hairs and a long, robust apical bristle. Legs largely brown with paler tarsi. Wing length 0.95-1.1 mm. Costal index 0.5-0.6. Membrane tinged grey. Hypopygium yellowish brown. Processes of epandrium, hypandrium and gonopods as Figure 53-

Female
Length of eye subequal to diameter of pale brownish yellow postpedicel. Palps straw yellow with up to 20 hairs of which the two subapicals are clearly longer. Frons, thoracic notum and abdominal tergites light brown; T1+2-T6 with dark bands at rear margins, but these not extending to the edges of the tergites; thus T3 with 20-24 hairs at rear margin, but only half lie within the dark band. Microtrichia of frons, thoracic notum and abdominal tergites all minute and dense. Legs straw yellow but femora, especially those of the mid and hind legs, very lightly tinged brown. Five males of C. modestus were caught in the same trap. So species H is possibly the female of this species.
with dark bands at rear margins. T3 with about two dozen hairs at hind margin. Microtrichia of frons minute and dense, those of thoracic notum longer and those of abdominal tergites even longer. Legs largely yellowish brown. Hind tarsus with ventral hair palisade on first two segments. Details of abdominal terminalia as Figures 65, 66.

Natural history
Most of these females were procured from rotting jack fruit (Artocarpus heterophyllus, Moraceae) by Collin and in fruits of Phoenicophorium borsigianum (Palmae) by Justin Gerlach. These fruits are probably the larval pabulum.

Discussion
Of the 14 species now known from the Afrotropical Region, five are tramp species transported around the world by man. Of these C. heymonsi is the dominant species in mainland Africa and this was probably its original distribution. Likewise, the prevalence of C. fletcheri in Arabia suggests that this may be its original locality. Of the nine species apparently restricted to the Afrotropical Region only C. cummingae is recorded from mainland Africa so far. It also occurs in Madagascar, along with four of the other Afrotropical species. The other four are so far only known from the Seychelles. More intensive collecting may alter this picture.
The new species are mainly only known in the male sex. But the two species only known as females (G and H), which cannot be formally named until associated with their males, probably belong to two of these. I would therefore welcome any reared series or pairs caught in copula. Such specimens are best preserved in 70-80% ethanol.