A review of the genus Philodicus Loew, 1848 in Southern Africa (Diptera: Asilidae)

ABSTRACT The history of taxonomic work undertaken on Afrotropical Philodicus Loew, 1848 is summarised. Details relating to the southern African fauna are provided. Five species occur in the sub-region: P. cinerascens (Ricardo, 1900), P. dubius Ricardo, 1921, P. fraterculus (Walker, 1855), P. swynnertoni Hobby, 1933 and P. tenuipes Loew, 1858. A lectotype and paralectotype are designated for P. dubius. A key to the species is provided along with new illustrations of male terminalia in order to facilitate identification. All known records are provided for these species along with distribution maps and notes on their biology. All species are found near places where free water may be found. All known prey records are provided and discussed, and a strong preference for grasshoppers (Acrididae) is noted.


INTRODUCTION
Philodicus Loew, 1848 is an attractive genus ( Fig. 1) of fairly large robber flies with some 20 recognised Afrotropical species (Oldroyd 1980). The genus is also well represented in the Oriental Region with 19 recognised species being listed by Oldroyd (1975). Two species are catalogued for the Palaearctic Region (Lehr 1988), while a single species is listed as occurring in the Australasian and Oceanian Regions (Daniels 1989).
The taxonomic history of Afrotropical Philodicus can be briefly summarised as follows: Wiedemann (1819)   The taxonomy of the genus was competently handled by Blasdale (1957). Unfortunately he had very little southern African material and this persuaded Londt (1978) to study the genus in a southern African context using material largely deposited in southern African collections. Southern Africa is here defined as all territory south of the Kunene and Zambezi rivers. In the more than 35 years since Londt's 1978 paper, many more specimens have accumulated and so a reappraisal is considered useful. Although no new taxa have emerged, a far better understanding of the distributions of those species inhabiting the region is now possible and more information is available relating to habitat preferences and prey choices. With the work of Blasdale (1957) and Londt (1978) being available it is not considered necessary to redescribe the southern African Philodicus species. However, in order to make identifications easier some new illustrations of terminalia are presented along with an updated key.

MATERIAL AND METHODS
Much of the material previously recorded by Londt (1978) has been re-examined and included in the lists of material supplied for each species and marked with an asterisk (*). While it was not considered necessary to recall previously cited specimens from collections outside of South Africa, information concerning these is briefly provided and the locality data plotted on maps. Genital illustrations are designed to complement those of Londt (1978) and Blasdale (1957) and to feature characteristics helpful in identification of the species. The key provided is a modification of that published by Londt (1978). With the focus being primarily on the distributions and biology of species, available material is listed and analysed as fully as possible. Square brackets are used for comments, notes or updated information. While more recently collected specimens usually provide detailed information relating to the localities and habitats from which material was obtained, it has been necessary to attempt to establish precise geographic coordinates for many localities. Although Google Earth and other online gazetteer sources have been used extensively to accomplish this, there remain a few specimens whose origin is doubtful or unknown. Material studied is listed under the countries of origin and ordered according to latitude. Abbreviations for institutions housing material cited are as follows (older citations being supplied in brackets), the names of people who have assisted me in recent times are acknowledged in brackets. In some instances material cited earlier (Londt 1978) was not reworked and so no acknowledgement is necessary. Abbreviations used by Londt (1978)  Diagnosis (compiled using key characters as published by Londt (2005)): Head: Antennal stylus composed of only 2 elements (an attenuated segment-like element tipped with a seta-like sensory element); face with a slight gibbosity ventrally; postocular setae dorsally short, straight or only slightly proclinate. Thorax: Dorsocentral macrosetae confined to mesonotal region posterior of transverse suture; scutellum with fewer than 8 apical macrosetae, disc with setae only (no macrosetae); wing with complete supernumerary crossvein present between R 2+3 and R 4 ; supernumerary crossvein shortish and at most running parallel to R 4+5 for a short distance; cell r 4 long and diverging gradually towards wing margin; hind margin of wing with a double row of microtrichia diverging from the plane of wing membrane.

BMSA
Key to the southern African species of Philodicus (adapted from Londt 1978) 1 Anterior and posterior surfaces of femur 1 lacking macrosetae; antennal stylus at least as long as twice the length of postpedicel; ♀ ovipositor (S8) distally broad and spade-like (alcimoides group  (Walker, 1855) Philodicus cinerascens (Ricardo, 1900) Figs 2-6 Alcimus cinerascens Ricardo, 1900: 176;1922: 42. Philodicus umbripennis Ricardo, 1921: 236. Philodicus cinerascens: Ricardo 1925Blasdale 1957: 139 (pl (Ricardo, 1900  Material not re-examined: Additional southern African material from Zimbabwe was listed in Londt's (1978)  Distribution, phenology and biology: Southern African distribution is limited to a few localities in Zimbabwe (Fig. 6) where the species has been collected in September, November, February, March, April and June (Table 1) -seemingly throughout the year. The species has been collected sympatrically with dubius at Massangena. Little is known of its biology. Malawian specimens were collected in Brachystegia woodland near the shores of Lake Malawi. The female ovipositor is distally broad and spade-like so may be adapted for digging in loose sand. No prey records are available.
Philodicus dubius Ricardo, 1921 Figs 7-11 Philodicus dubius Ricardo, 1921: 179;Blasdale 1957: 139 (pl Ricardo (1921) described the species on 'Type (male) and another, type (female), all from M'fongosi, Zululand (W. E. Jones), March 1911, in Cape Museum Coll' and so did not designate a holotype, but merely listed a male and a female as 'types'. These two specimens, which must be considered syntypes, have been studied, along with others from the same locality, and the male is here designated Lectotype. The female is considered a Paralectotype, while other material has no type status as there is no evidence that Ricardo actually studied them. Material not re-examined: Londt (1978)  Distribution, phenology and biology: Distributed fairly widely throughout the eastern parts of southern Africa (Fig. 11). Adults are active throughout the warmer, and wetter, months of the year, having been recorded from November through to May (Table 1). The species has been collected sympatrically with cinerascens as well as tenuipes on which it has also been caught feeding. Label data suggest that the species inhabits a variety of open woodland habitats, often in the vicinity of or along the banks of rivers and streams. Prey records limited to 5, with only females involved -Orthoptera: Acrididae   Distribution, phenology and biology: Found over much of southern Africa (Fig. 16), records being more concentrated in the north-eastern parts which normally receive summer rainfall. Adults are summer active (Table 1), typically flying from November through to May (a single record for July). May be locally abundant. Typically inhabit grassy places in association with water (rivers, streams, dams, pans etc.). The female ovipositor is distally narrow and spike-like so may be adapted for thrusting eggs into soil. Some 24 prey records are available (7♂ (29 %) 17♀ (71 %)) -Orthoptera: Acrididae (10), Tridactylidae (1). Neuroptera: Myrmeleontidae (2). Diptera: Asilidae (1 -P. fraterculus ♂), Tabanidae (1), Bombyliidae (2), Sarcophagidae (2). Lepidoptera: Lycaenidae (1), unidentified moths (2). Hymenoptera: ? Vespidae (1), ? Halictidae (1). Cannibalism is recorded.

Distribution
Philodicus is represented by five species in southern Africa. Two of these, fraterculus and tenuipes are well represented and widely distributed, found in both the drier western parts of the subcontinent as well as in the more temperate eastern parts, although records are more numerous from the east, probably because more sampling has been done in these more developed parts of the subcontinent. A third species, dubius, is also reasonably well represented, but confined to the eastern parts of the subcontinent in higher rainfall regions. The remaining two species, swynnertoni and cinerascens, are apparently fairly widely distributed species which have their southern limits in the north-eastern parts of the sub-region.

Biology
Philodicus is of particular interest as species appear to require the presence of surface water for their survival. All southern African species are invariably collected near fresh water sources such as rivers, streams, dams and natural pans. In some instances specimens are found in drier and even arid places where surface water is only available after rain. Dry river beds and pans are also good places to collect these flies and so I conclude that developing larvae, probably living in sand or soil, are able to survive in relatively dry places as long as rain falls from time to time. The actual habitat in which adults are found varies greatly from forest margins, woodland, savannah and semi desert. Of interest is the fact that species in Blasdale's (1957) alcimoides species group (e.g. cinerascens, dubius, tenuipes) have relatively broad spade-like ovipositors as opposed to those of his fraternus group (e.g. fraterculus, swynnertoni) with their narrow spike-like ovipositors. It is highly likely that these differently shaped organs reflect the nature of the substrates in which eggs are deposited. It is probable that broad, spade-like ovipositors would be better adapted for digging in loose sand while spike-like ovipositors might be better organs in more compact substrates. A study of actual oviposition sites would no doubt be instructive.
As far as prey is concerned, Londt (2006: 321), who analysed some 2000 Afrotropical asilid prey records (including records previously published by Hobby (1935)), had  this to say about the then known prey of Philodicus, for which he had 183 records -'Orthoptera (62), Hymenoptera (47), Diptera (41), Hemiptera (18), Coleoptera (7), Lepidoptera (5), Neuroptera (2), Mecoptera (1). The first three listed orders constitute 82 % of prey. Significant prey families are Acrididae (62), Apidae (20), Asilidae (18) and Cicadidae (10). Acrididae represent 34 % of the diet of Philodicus.' Although there are a number of new prey records in this paper they merely support earlier findings. Clearly, grasshoppers are a significant food source which suggests that Philodicus is capable of catching flying prey as well as prey that commonly rest on the ground or on vegetation. Of interest is the fact that far more females have been collected with prey than males. Although females tend to be larger than males, and may be more readily collected, I believe the difference is more likely to relate to the need for protein for egg maturation. It is also possible that females live longer and could be more abundant than males at any one time.