A PHYLOGENETIC ANALYSIS OF THE WOODLOUSE-FLIES (DIPTERA, RHINOPHORIDAE) by THOMAS PAPE Zoological Museum Copenhagen, Denmark Abstract The Rhinophoridae are redefined on the basis of the apomorphic structure of the aedea- gus. Evidence is provided for the exclusion of four genera, viz., Angioneura Brauer & Ber- genstamm, 1893, Melanomya Rondani, 1856, Morinia Robineau-Desvoidy, 1830, and Ter- mitoloemus Baranov, 1936; all four are transferred to the Calliphoridae. The genera of Rhi- nophoridae are analyzed phylogenetically with the aid of the results of the present mvestigation and the sparse information available on the morphology of the larval stages. The structure of the aedeagus provides several set-defining characters and the aedeagus of many species is depicted for the first time. Cirillia Rondani, 1856, is proposed as a synonym oi Phyto Robineau-Desvoidy, 1830. Introduction Within the calyptrate fhes the species with a row of bristles on the meron (hypopleuron) constitute a well-corroborated monophyletic group, the Tachinoidea (Tachinidae sensu Girschner (1893), Calliphoroidea sensu Hennig (1958), Tachinidae (sensu lato) sensu Griffiths (1972), or Oestroidea sensu McAlpine et al. (1981)). Most recent authors, including the pre- sent, accept five major groups in the Tachinoi- dea, viz., Oestridae (sensu lato), Calliphoridae, Sarcophagidae, Tachinidae, and Rhinophoridae. Although Crosskey (1965) restricts the name Tachinoidea to the Calliphoridae, Sarcophagi- dae, Tachinidae, and Rhinophoridae, synapo- morphies not shared by the Oestridae (which would be their sister group), to my knowledge have not been provided for these four families by any author. The family Stackelbergomyiidae Rohdendorf, 1948, was obviously established because no evi- dence for incorporatmg the smgle aberrant spe- cies into any of the existing families could be found. An investigation by Herting (1981) sug- gests that it should be included in the Tachini- dae. More interesting are the Neotropical Me- sembrinellinae (Calliphoridae). Crosskey (1965) is of the opinion that an improved classification of the Tachinoidea (in his definition as given above) would result if "peculiar groups such as Mesembrinellinae were treated as families" (p. 43). Guimaraes (1977) follows this recommen- dation and raises the group to family status: Mesembrinellidae, founding his decision on five "consistent differences" between Mesembrinel- linae and the remaining Calliphoridae. These differences corroborate the monophyly of the MesembrinelHnae, but the Calliphoridae sensu Guimaraes are characterized solely on symple- siomorphies and fail to support a family status of the Mesembrinellinae. An argument for split- tingup the Calliphoridae would be that the sim- ple, non-opercular lappet of the mesembrinel- line metathoracic spiracle is plesiomorphic, as this would separate the Mesembrinellinae (still monophyletic) not from the Calliphoridae but from all other Tachinoidea, the monophyly of which would be corroborated by their opercu- lar metathoracic spiracle. This may be the rea- son for Crosskey 's (1965: 43) note that the Me- sembrinellinae "may not be Tachinoidea at all". I hesitate to place the Mesembrinellinae as sister group to all other Tachinoidea and prefer to treat them as Calliphoridae. The structure of the mesembnnelline aedeagus with strong, for- wardly curved dorsolateral processes (paraphal- li) seems a reasonable synapomorphy with the Calliphoridae (and perhaps with the Rhinopho- ridae?). A small digression may be made here, brought about by the recent (and past) dis- agreement of family status criteria. Some au- thors, e.g., Steyskal (1974) and Hackman &: Väisänen (1982), have mentioned the inconsis- tency of Griffiths' (1972) splitting of the Musci- dae sensu Hennig (1958, 1965) into Muscidae and Fanniidac when he unites all tachinoid flies in a single family: Tachinidae (sensu lato). 15 16 Tijdschrift voor Entomologie, deel 129, afl. 2, 1986 However, although Hennig (1965) states that: "Eine der sichersten Feststellungen, vielleicht die gesichertste, die man über das phylogene- tische System der Muscidae treffen kann, ist die, dass zwischen den Fanniinae auf der einen seite und der Gesamtheit aller übrigen Muscidae... ein Schwestergruppenverhältnis besteht" (p. 9), he does not bring conclusive evidence of the monophyly of the "Muscidae sensu lato". Therefore, a separation is to be preferred. If the tachinoids are considered a monophyletic group, they are best treated in common when used for outgroup comparison, and the formal rank — whether family or superfamily — is of minor importance in a phylogenetic sense. Only the ranking of the group relative to the other Calyptratae is important as this constitutes a phylogenetic hypothesis. The monophyly of the Tachinoidea seems fairly corroborated. Griffiths (1972) mentions the following synapomorphies with respect to the groundplan of the Calyptratae: (1) hypopleuron with strong bristles below metathoracic spiracle, (2) eighth sternum ( 9 ) entire, (3) vein mj.,., sharply bent towards r4+5 apical- (4) anal vein not reaching wing margin, (5) sixth tergum (a) shortened, less than half as long as 5th tergum, (6) eighth tergum vestige (cî) lost. The loss of the "eighth tergum vestige" in males is based on a questionable interpretation of a median ventral sclerotization in the postab- domen of some Anthomyiidae and Scatophagi- dae (Griffiths, 1972: fig. 61); this sclerotization more likely is a secondary acquisition. Another character which may be autapomor- phic to the Tachinoidea is: (7) lappet of metathoracic spiracle divided, pos- terior lappet shaped as an operculum. This opercular metathoracic spiracle, absent in all other calyptrates, is present in the majority of the Tachinoidea; the non-opercular metatho- racic spiracle present in the Mesembrinellinae and a few other Calliphoridae, some groups of Tachinidae, Macronychiinae of the Sarcophagi- dae, almost all Rhinophoridae, and many Oes- tridae (sensu lato) may be secondarily derived. The sister group relations of the Tachinoidea within the Calyptratae are still largely unsolved, and the characters mentioned by Griffiths are not necessarily autapomorphies for the Tachi- noidea, viz., Items 4 and 5 mentioned above, which also occur among other calyptrate groups. A shortened anal vein (A,) is character- istic of both the Muscidae and Fanniidae. A few genera in the Tachinoidea (e.g., some Oestridae (sensu lato) and Tachinidae, Bengalia Robi- neau-Desvoidy in the Calliphoridae) possess an extended anal vein, a character which both Hennig (1958) and Griffiths (1972) consider to be secondary. It is interesting, however, that Andersen (1982) reports aerial swarming of male Siphona Meigen as the first example within the Tachinidae and suggests (Andersen, 1982, 1983) that an extended anal vein may be assig- nable to the groundplan of the Tachinidae (and then possibly to all the Tachinoidea). The shortened abdominal tergum 6 in males is of general occurrence in the Muscidae and An- thomyiidae as well. The Rhinophoridae are typical members of the Tachinoidea as defined above (fig. 1), but the affinities to other tachinoid families are still unclear. Many earlier authors placed the rhino- phorids with the blow-flies and flesh-flies in a Calliphoridae (sensu lato), but in a phylogenetic sense this constitutes an entirely unacceptable non-group arising by the splitting off of the flies possessing a swollen subscutellum — the Tachi- nidae. Mesnil (1939) derived most of the subfa- milies of Tachinidae from different rhinophorid stocks, thereby rendering the Rhinophoridae paraphyletic (and the Tachinidae polyphyletic), but at present most authors give the Rhinopho- ridae family rank, acknowledging their uniqueness and the present lack of evidence for a closer relation to any of the other tachinoid families. Kugler (1978), and especially Crosskey Fig. 1. Stevenia deceptoria (Loew); a typical wood- louse-fly. Pape: Rhinophoridae 17 (1977), give a more detailed review of previous differences of opinion regarding the family affi- nities. Larval biology and morphology Although the family is small, an unambiguous demarcation of the Rhinophoridae has not been possible. This is due in part to the existence of deviating tropical forms, e.g., Bequaertiana Curran, and in part to an external morphology intermediate between that of typical calHpho- rids and typical tachinids. More important, however, is the lack of information concerning the morphology and biology of the larval stages. All known first-stage larvae possess a distinctive cephalopharyngeal skeleton with the anterior part of the pharyngeal sclerite greatly elongated and with two or more teeth on the dorsal arc of the mandibles — evidently synapomorphic characters. The larval habit of parasitizing woodlice (Isopoda) is likewise unique to the Rhinophoridae, and interesting insofar as very few biological relationships between Diptera and Crustacea are known (see Roubaud, 1903; Mercier, 1921; Oldroyd, 1964, and Burger et al., 1980). Only seven genera of rhinophorids actually have been recorded as woodlouse parasites, viz., Stevenia Robineau-Desvoidy, Tricogena Ron- dani, Rhinophora Robineau-Desvoidy, Melano- phora Meigen, Paykullia Robineau-Desvoidy, Phyto Robineau-Desvoidy, and Cirillia Ronda- ni (note that Cirillia is a synonym of Phyto, see discussion below). Specific host records for the Palaearctic species are given by Herting (1961) with supplements in Kugler (1978). Parker (1953) mentions breeding of the introduced Melanophora roralis (Linnaeus) in Brazil. No host records exist for any of the Nearctic, Afro- tropical, or Oriental species. Table 1. List of non-isopod hosts of the Rhinophoridae. There has been some doubt as to whether the Rhinophoridae could be parasites in inverte- brates other than isopods, and the tendency has been to disregard any such record. Obviously, the report of Melanophora helicivora Goureaux being bred from the gastropod Helicella con- spurcata (Draparnaud) is based on a mis-identi- fication. As judged from the description and drawings (Goureaux, 1843: figs. 1, 2), the spe- cies does not belong to Melanophora at all, but may be a calliphorid. Lundbeck (1927) mentions a specimen of Melanophora roralis bred from egg-cocoons of the spider Araneus cornutus Clerck. I have seen this specimen, a female deposited in the Zoolog- ical Museum, Copenhagen, and it is correctly identified by Lundbeck. In addition to this there are several reports of rhinophorids parasitizing insects (table 1), and it is probable that rhinophorids occasionally (acci- dentally?) may parasitize arthropods other than isopods.' Very little has been written on the morpholo- gy of the larvae of the woodlouse-flies. Thomp- son (1934) treated in detail the larval stages of eight species, viz., Paykullia maculata (Fallen), Phyto angustifrons (Rondani), Phyto discrepans (Pandellé), Phyto melanocephala (Meigen), Melanophora roralis, Stevenia atramentaria (Meigen) (as species B), Tricogena ruhricosa (Meigen), and Rhinophora lepida (as species A). However, Thompson obtained all his material from dissections of woodlice as most of his at- tempts to obtain eggs from female flies caught in the wild and hatch these to first-instar larvae failed. Furthermore, he often assumed that rhi- nophorid larvae from a single colony of wood- hce were conspecific. This has resulted in some erroneous identifications in his earlier works (Thompson, 1917, 1920; corrected in 1934: parasite/predator specimens host zie. rererence Melanophora roralis Stevenia umbratica Rhinophora lepida Rhinomorinia sarcophagina 1 9 eggs of Araneus cornutus Clerck (Araneae) ? iPyralis farinalis (Linnaeus) (Lepidoptera, Pyralidae) ? Callidium violaceum Linnaeus (Coleoptera, Cerambycidae) 1 6 Paranthrene tabaniformis (Rottemburg) (Lepidoptera, Aegeriidae) 1 6 Saperda carcharias (Linnaeus) (Coleoptera, Cerambycidae) 1(5, 1 9 Malacosoma neustria (Linnaeus) (Lepidoptera, Lasiocampidae) Lundbeck (1927) Bezzi & Stein (1907) Bezzi & Stein (1907) Kolubajiv (1962) Kolubajiv (1962) Kolubajiv (1962) 18 Tijdschrift VOOR Entomologie, deel 129, afl. 2, 1986 380). Of the first-instar larva of Phyto angustif- rons Thompson had only a single defective specimen (the cephalopharyngeal skeleton and a skin fragment). The depicted cephalopharyngeal skeleton (Thompson, 1934: pi. 19, fig. 47) is of the heavy, sclerotized type found in Stevenia, Tricogena, and Rhinophora and very unlike the cephalopharyngeal skeleton of Paykullia, Mela- nophora, and other species oi Phyto. The first-stage larva assigned to Phyto angus- tifrons probably belongs to another species (very likely a Stevenia). The cephalopharyngeal skeleton of the second- and third-stage larva of P. angustifrons (Thompson, 1934: pi. 20, figs. 48, 56) is more in accordance with that of Phyto species. Bedding (1973), in an extract of his Ph. D. thesis, described eggs and larval stages of all English species — actually the same species as those described by Thompson (1934) except for P. angustifrons. The larvae, especially first instars, possess several features which are very useful in a phylogenetic context, but at present the larval stages are known for only a small fraction of the species described. In addition, the uniqueness of many of the features makes any outgroup comparison almost inapplicable in the distinction between apomorphic versus ple- siomorphic larval characters within the family. The first-stage larvae known at present com- prise two distinct groups (see figs. 8 — 44 in Bedding, 1973): A. Phyto, Paykullia, Melanophora (1) mandibles with normal degree of sclerotiza- tion, with three or more small teeth on the dorsal arc, (2) elongated anterior part of pharyngeal scle- rite with an incision, (3) setal bases unmodified, (4) posterior end of larva highly modified for supporting the larva in erect posture; with a dorsal tongue, terminal sac-like lobes, and ventral ridges. B. Stevenia, Tricogena, Rhinophora (1) mandibles heavily sclerotized, with two strongly developed teeth, (2) elongated anterior part of pharyngeal scle- rite without an incision, (3) setal bases protruded into proleg-like struc- tures, (4) posterior end of larva simple, with inflated ventral vesicles. Bedding notes that the two morphologically distinct groups of first-stage larvae possess dif- ferences in their biology (referring to a paper (in prep.) which unfortunately has not yet been published). The toothed mandibles of the first-stage lar- vae are probably an adaptation for penetrating the body wall of the host, analogous to the ser- rate median tooth of tachinid larvae which enter the host through a strongly sclerotized cuticle (Clausen, 1940: fig. 210 A). This character is clearly an autapomorphy for the Rhinophoridae as toothed mandibles occur very sporadically in other Tachinoidea, e.g., the warblefly of the lechwe antelope (Howard, 1980). The two types of cephalopharyngeal skeleton can not be separated into an apomorphic and a plesiomorphic state at present; indeed, it is pos- sible that both types are apomorphic with re- spect to the groundplan of the Rhinophoridae, but this may be the least parsimonious solution to the problem. The proleg-like setal bases must be consid- ered an apomorphic character as these are ab- sent in the majority of the Tachinoidea and nothing indicates their suppression in other rhi- nophorids. This character is found in Stevenia, Tricogena, and Rhinophora and may be a syna- pomorphy of the Stevenia group (see discussion below), thereby corroborating the monophyly of this group. The two types of modified posterior end of the first-stage larva present a problem some- what analogous to that of the cephalopharyn- geal skeleton. However, until more information on the sister group relations of the Rhinophori- dae within the Tachinoidea becomes available, it is reasonable to assume that the sister group possesses first-stage larvae with unmodified posterior ends. The terminal lobes, the dorsal tongue, and the free, posteriorly oriented ven- tral ridges will then be apomorphic characters, and the terminal lobes will be the apomorphic homologues of the inflated vesicles. This will corroborate the hypothesis that Paykullia, Phy- to, and Melanophora are part of a monophyletic group (the Phyto group) not containing Steve- nia, Tricogena, or Rhinophora. Recognition of the Rhinophoridae Crosskey (1977: 7) gives an excellent dis- cussion of the status and recognition of the fam- ily, but he admits that his recognition couplet does not ensure a certain identification. A fur- ther complication is the recently described ge- nus Baniassa Kugler. This genus has a well-de- veloped metathoracic opercular spiracle, but the absence of a distinct operculum has hitherto Pape: Rhinophoridae 19 provided one of the most important single char- acters for rhinophorid recognition. Besides the structure of the metathoracic spiracle, the char- acters most helpful in recognizing the family have been the tongue-shaped or oval lower ca- lypteres which are widely removed from the scutellum, the bend of vein M which never is greatly concave, and the combination of bare prosternum, proepisterna, greater ampullae, postalar walls, laterotergites, and supra-squamal ridges. In the majority of the Tachinoidea the struc- ture of the aedeagus (and other structures of the terminalia) provides important characters in the diagnostic segregation of species and is often used in the construction of evolutionary trees and in the definition of taxonomie categories above the species level. Some illustrative exam- ples are the works of Mueller (1926) on the Ta- chinoidea, Roback (1954) on the Sarcophaginae, Verbeke (1962) on the Tachinidae, Kurahashi (1966) on the Luciliinae, Lehrer (1970) on the Calliphoridae, and Lehrer (1973) on Sarcophaga (sensu stricto). The distiphallus of male rhino- phorids, however, is seldom depicted, not even in the revisions of the Palaearctic (Herting, 1961) and Afrotropical (Crosskey, 1977) spe- cies, and the information stored in this structure is largely unknown. Mueller (1926) made an early attempt to construct a "Stammbaum . . . auf Grund der Penisform" of the Tachinoidea, but only a few rhinophorids were included and the drawings are more or less incorrect. Séguy (1941) made a preliminary division of the Rhi- nophoridae (as a subfamily of the Calliphoridae, sensu lato) into four groups on the basis of the male genitalia, but he dissected only a few rep- resentatives and his definition of the (sub)family included several tachmid, sarcophagid, and cal- liphorid genera. The structure of the aedeagus may provide additional characters to be used in the recogni- tion of the family; and in order to use this struc- ture in a redefinition of the family and in the re- construction of the phylogeny at the generic level, the following hypothetical groundplan of the tachinoid aedeagus is accepted (terminology as in Hennig, 1976 and McAlpine et al., 1981) (fig-2). Like most other calyptrate flies a well-devel- oped basiphallus, distiphallus and epiphallus are present. The distiphallus is more or less tubular, somewhat swollen basally, and possesses spin- ules on the ventral surface. The distiphallus is connected to the sclerotized basiphallus by Fig. 2. Stevenia atramentaria (Meigen); aedeagus, lateral view. Abbreviations: aph = acrophallus, bph = basiphallus, d.pl = dorsal plate, dl.pr = dorsolateral processes, eph = epiphallus, spd.scl — spermduct sclerotization, v. pi = ventral plate. means of the dorsal plate, which divides distally into a pair of dorsolateral processes. The dorsal plate is extended ventrally on each side, forming two ventral plates. The acrophallus, carrying the phallotreme, is a simple, membraneous ex- tension of the distiphallus, probably encircling the three openings of the female spermathecal ducts during copulation. The aedeagus of many rhinophorids, e.g., Phyto spp. (figs. 15, 16), has not diverged markedly from this ancestral state, and the view is in agreement with that of Rikhter (1980), who mentions an epiphallus, basiphallus, a distiphal- lus immovably connected to the basiphallus, and "relatively" simple structure of distiphallic parts as the groundplan of the Tachinidae. Two features of the rhinophorid aedeagus de- serve mention. A possible autapomorphy for the Rhinophoridae is the well-developed ventral plates clearly set off from the dorsal plate and fused along the ventral margins, thus forming a sclerotized ring. Only the genus Paykullta pos- sesses unfused, but closely apposed, ventral plates, and this may be considered a reversal, as discussed below. It may seem somewhat odd to 20 Tijdschrift voor Entomologie, deel 129, afl. 2, 1986 attach any importance to this character consid- ering the enormous variability of the distiphal- lus within the Tachinoidea, and certainly it is possible to enumerate several cases of non-rhi- nophorids (especially among the Calliphoridae) with fused ventral plates. However, most or all of these instances will be easily rejected as con- vergencies and I think the distinctive ventral plates will be of great value in the proper recog- nition of any rhinophorid. The other character to be mentioned is the sclerotization of the ventral part of the sperm- duct extending from the ventral plates to the phallotreme. In Phyto and Parazamimus this sclerotization is interrupted basally and does not reach the ventral plates (figs. 14 — 16). All other rhinophorid genera possess a sclerotiza- tion fused to the ventral plates and continuing to the phallotreme. The use of outgroup comparison for assessing the level at which this character is apomorphic is difficult to apply as the sister group of the Rhinophoridae is unknown. A similar sperm- duct sclerotization is of general occurrence in the Calliphoridae (the mesohypophallic sclero- tization of Salzer (1968)) but absent in most Sar- cophagidae and Tachinidae. If the interrupted spermduct sclerotization of Phyto is considered to be plesiomorphic within the Rhinophoridae then Phyto must be the sister group to all other genera. This hypothesis seems falsified by the several synapomorphies in the imaginai mor- phology of Phyto and Baniassa, and by the apo- morphic larval morphology of Phyto, which is also found in Paykullia and Melanophora. Probably the possession of a spermduct sclero- tization fused to the ventral plates is a ground- plan character in the rhinophoridae, and the spermduct sclerotization may be an important argument for a close affinity to the Calliphori- dae. To sum up, the characters which I regard as the most useful in the recognition of the family are the following: Larval characters: (a) cephalopharyngeal skeleton of first-stage larvae with toothed mandibles and elongated pharyngeal sclerite, (b) parasites of woodlice. Imaginai characters: (c) aedeagus with well-developed ventral plates that are fused (or closely apposed) along the ventral margins, (d) lower calypteres tongue-shaped, diverging from the scutellum. (e) metathoracic spiracle without a distinct o- perculum (except in Baniassa), (f) prosternum, proepisterna, greater ampullae, postalar walls, laterotergites, and suprasqua- mal ridges bare, (g) bend of vein M never greatly concave. It is important to note that the characters given not necessarily are rhinophorid autapo- morphies as some of them are found in other ta- chinoids as well. Character (f) is obviously ple- siomorphic within the Tachinoidea and is pro- vided to facilitate the exclusion of rhinophorid- like Calliphoridae. Genera misplaced in the Rhinophoridae The previous lack of an unambiguous defi- nition of the family has resulted in some moving about of a few genera. Crosskey (1977) in his review of the Rhinophoridae gives evidence for the exclusion of genera like Shannoniella Townsend (Tachinidae), Bezzimyia Townsend (Tachinidae), and Opsodexia Townsend (Calli- phoridae), all of which earlier have been consid- ered to belong to the Rhinophoridae (or to the Rhinophorinae as a subfamily of the Tachini- dae). This exclusion is accepted in the present paper and only the genera listed by Crosskey (1977), with the additions of Kugler (1978), will be treated in detail. Some of these clearly de- viate from the definition given above and ought to be excluded from the Rhinophoridae. Angioneura Brauer & Bergenstamm. Angioneura has long been treated as belong- ing to the Rhinophoridae, but North American authors, especially Downes (1955, 1965), have transferred it to the Calliphoridae, this view be- ing accepted by Wood (1979). Crosskey (1977) discusses this genus in the paragraph "included genera possibly not Rhinophoridae" but accepts its rhinophorid status. It is noteworthy that the genus Angioneura contains some species with enlarged lower calypteres, viz., A. obscura (Townsend), the only Nearctic species seen, and A. acerba (Meigen). The lower calypteres of the other species investigated, although distinctly diverging from the scutellum, are semicircular and not of the typical tongue-like shape charac- teristic of the Rhinophoridae. The larvae, still unknown from the first stage, seem to be parasites of snails rather than wood- lice. Two of the five Nearctic species of Angio- neura are recorded as having been bred from snails (Reinhard, 1929; Downes, 1965) and A. cyrtoneurina (Zetterstedt) from the Palaearc- Pape: Rhinophoridae 21 tic Region has been bred from the snail Succinea elegans Risso (Cepelâk & Rozkosny, 1968). Bedding (1973) collected thousands of woodlice from about 50 localities in southern England in order to breed all native species of Rhinophori- dae. He did not, however, obtain any specimens of A. acerba or A. cy rione urina, the only Eng- lish representatives (Kloet & Hincks, 1976). On this evidence I find it highly unlikely that any species oi Angioneura parasitizes woodlice. The presence of species with enlarged or semicircular lower calypteres, the life habit of the larvae as parasites in snails, and the ventral plates of the distiphallus which, although rather well-developed, are completely free of and widely removed from each other (fig. 3), clearly corroborate the exclusion of Angioneura from the Rhinophoridae, and I follow Downes (1965) in regarding Angioneura as a calliphorid. It is interesting that the exclusion of Angio- neura leaves the American continent without indigenous species of rhinophorids. Two spe- I cies, however, have been established on this pi continent, both probably introduced from Eu- I rope: Phyto discrepans, which occurs in south- ' ern Canada, and Melanophora roralis, which is recorded from southern Canada, the eastern United States, the West Indies Qamaica, St. Thomas), and Brazil. Examined species: Angioneura acerba (Mei- gen, 1838), A. cyrtoneurina (Zetterstedt, 1859), A. fimbriata (Meigen, 1826), A. obscura (Townsend, 1919). Melanomya Rondani. This genus is apparently closely related to Angioneura, and Downes (1965) treats Angio- neura as a subgenus of Melanomya. No host re- cords are known for the single European spe- cies, Melanomya nana (Meigen), but as with Angioneura, the absence of any specimens of Melanomya nana in the material studied by Bedding (1973) reduces the probability of a woodlouse parasitizing habit. In addition, the ventral plates of the distiphallus are rather widely separated (fig. 4). The similarity to Angioneura will then indi- cate a position in the Calliphoridae. The metathoracic spiracle of M. nana differs from the typical, somewhat triangular, rhino- phorid type of spiracle (Crosskey, 1977: figs. 41 — 44) in being broad with a well-developed anterior fringe. This may provide further evi- dence for a calliphorid status as the majority of the Calliphoridae possess a rather large meta- thoracic spiracle, most often with a distinctly enlarged anterior lappet. Examined species: Melanomya nana (Mei- gen, 1826). Figs. 3 — 5. Aedeagus of Calliphoridae, lateral view: 3, Angioneura fimbriata (Meigen). 4, Melanomya nana (Meigen). 5, Morinia melanoptera (Fallen). 22 Tijdschrift voor Entomologie, deel 129, afl. 2, 1^ Morinia Robineau-Desvoidy. This genus is accepted as belonging to the Rhinophoridae by Crosskey (1977) in spite of the presence of distinct hairs on the postalar walls, a character used by Crosskey to exclude rhinophorid-Uke Calliphoridae. Haired postalar walls occur in many Calliphoridae and in the subfamily Sarcophaginae of the Sarcophagidae (very seldom in subfamily Miltogramminae), but I have not found this trait in any tachinid or rhinophorid. The presence of haired postalar walls and the lack of well-developed ventral plates (fig. 5) make an inclusion under the Rhinophoridae somewhat improbable. Two other characters that may corroborate an exclusion are the well- developed metathoracic spiracular operculum (although an operculum is present in a single rhinophorid genus) and the presence of a weak- ly developed facial carina, these characters being most conspicious in the Japanese species M. ni- gerrima (Herting). A facial carina is not found in any rhinophorid but occurs freguently in the Calliphoridae and Tachinidae. On this sparse evidence I find a position in the Calliphoridae most corroborated. Examined species: Morinia melanoptera (Fallen, 1810), M. nigerrima (Herting, 1961). Termitoloemus Baranov. The only known species, T. marshalli Bara- nov, was originally described as belonging to the tribe Bengaliinae in the Calliphoridae. This was based on a similarity in life habits between Bengalia and Termitoloemus, predators of ants and termites, and similarities in the structure of the proboscis and palpi. Sabrosky & Crosskey (1970) transferred Termitoloemus to the Rhino- phoridae because of the possession of a simple metathoracic spiracle and tongue-like lower ca- lypteres. However, the lower calypteres of Ter- mitoloemus differ strikingly from all rhinopho- rids in having a distinct notch at the posterior base (fig. 6). The lappet of the metathoracic spi- racle is provided with stiff bristle-like hairs among the usual hairs. This condition is not found in the Rhinophoridae, but several groups of Calliphoridae possess stronger hairs on the anterior lappet. I have investigated the slide-mounted genita- lia of the male holotype of T. marshalli. The ae- deagus is highly apomorphic and very unHke that of any rhinophorid (or any other tachinoid) and its ventral plates are not fused (Baranov, 1936: fig. 1). This evidence, indeed, does not Fig. 6. Termitoloemus marshalli Baranov. Semidia- grammatical drawing of right lower calyptere of holo- type S . give much hint of the family affinity of Termito- loemus. The lower calypteres are not of the typ- ical tongue-like rhinophorid type but more sim- ilar to the plesiomorphic, enlarged type, and the metathoracic spiracle can be taken as evidence for either a caUiphorid or a rhinophorid status. I do not find a rhinophorid assignment the most corroborative and I have chosen to consider Termitoloemus to belong to the Calliphoridae. Examined species: Termitoloemus marshalli Baranov, 1936. An inventory of the genera accepted as Rhi- nophoridae in the present paper is given in table 2. Note that Cirillia is treated as a junior syno- nym of P^jîo. ■ The phylogeny of the rhinophorid genera Very few attempts to create a suprageneric classification of the Rhinophoridae have been made, and these are often of little utiHty owing to the inclusion of several non-rhinophorid gen- era. Townsend (1935, 1938) divided his Mela- nophoridae (of which more than half of the gen- era were non-rhinophorids) into the five tribes Villeneuviellini, Melanophorini, Acampomin- thoini, Eggisopsini, and Moriniini. Séguy (1941), still with a rather broad (sub)family con- cept, arranged the few genera of which he had investigated the male genitalia into four groups based on perceived similarity. In the first group, Morinia (as Calobataemyia) is placed with Nyc- tia Panzer (Sarcophagidae), as Séguy apparently has dissected a specimen of Nyctia erroneously taken for a Morinia specimen (see his fig. 445, p. 343). Two other groups, both monogeneric, contain Stevenia and Melanomya (as Morinia), and the last group consists of Phyto, Rhinomo- Pape: Rhinophoridae l'i Table 2. Inventory of genera accepted as Rhinophoridae in the present paper. Following each generic name is the number of species described at present (in brackets) and an indented list of species investigated in the present study. Acompomintbo Villeneuve, 1927 (1 sp.) A. lobata Villeneuve, 1927 Azaisia Villeneuve, 1939 (2 spp.) A. obscura (Villeneuve, 1939) A. setitarsis Villeneuve, 1939 Baniassa Kugler, 1978 (2 spp.) B. fascipennis Kugler, 1978 B.paucipila Pape, 1985 Bequaertiana Curran, 1929 (2 spp.) B. argyriventris Curran, 1929 B. basilewskyi Peris, 1957 Callidesia Kugler, 1978 (1 sp.) C. pictipennis Kugler, 1978 Comoromyia Crosskey, 1977 (1 sp.) (C. griseithorax Crosskey, 1977; not seen) Macrotarsina Schiner, 1857 (1 sp.) M. longimana (Eggers, 1856) Melanomyoides Crosskey, 1977 (1 sp.) M. capensis (Zumpt, 1959) Melanophora Meigen, 1803 (2 spp.) Melanophora roralis (Linnaeus, 1758) Metoplisa Kugler, 1978 (1 sp.) M. carbonaua Kugler, 1978 Oplisa Rondani, 1862 (5 spp.) O. aterrima (Strobl, 1899) O.pollinosa Kugler, 1978 O. tergestina (Schiner, 1862) Parazamimus Verbeke, 1962 (1 sp.) P. congolensis Verbeke, 1962 Paykullia Robineau-Desvoidy, 1830 (8 spp.) P. brevicornis (Zetterstedt, 1844) P. kuglen (Herting, 1961) P. maculata (Fallen, 1820) Phyto Robineau-Desvoidy, 1830 (22 spp.) {Cirillia Rondani, 1856, syn. n.) P. angustifrons (Rondani, 1856) comb. n. P. cingulata (Zetterstedt, 1844) P. discrepans Pandellé, 1896 P. melanocephala (Meigen, 1824) P. pauciseta Herting, 1961 Queximyia Crosskey, 1977 (1 sp.) Q. flavipes Crosskey, 1977 Rhinomorima Brauer & Bergenstamm, 1889 (12 spp.) R. capensis (Brauer & Bergenstamm, 1893) 7?. sarcophagina (Schiner, 1862) R. xanthocephala (Bezzi, 1908) Rhinophora Robineau-Desvoidy, 1830 (1 sp.) R. lepida (Meigen, 1824) Stevenia Robineau-Desvoidy, 1830 (18 spp.) 5. angustifrons Villeneuve, 1913 S. atramentaria (Meigen, 1824) S. deceptoria (Loew, 1847) S. fernandezi Baez, 1978 S. hirtigena Herting, 1961 S. umbratica (Fallen, 1820) Tricogena Rondani, 1856 (1 sp.) T. rubricosa (Meigen, 1824) Tromodesia Rondani, 1856 (2 spp.) T. angustifrons Kugler, 1978 Ventrops Crosskey, 1977 (> 1 sp.) V. milichioides Crosskey, 1977 V. spp. undescribed. Pape (in prep.) rinia (as Metopisena), Angioneura, Rhinophora, and Melanophora. Herting (1961), in his revision of the Pal- aearctic species, divided the (sub)family into two tribes: Azaisiini (containing Azaisia and Acompomintho), with long antennae and elon- gate second aristal segment, and the clearly par- aphyletic Rhinophoriini, whithout these charac- ters. In the following is presented a phylogenetic analysis of the rhinophorid genera based on principles of phylogenetic systematics. Apo- morphies (numbers refer to the cladogram, fig. 30) are only given for genera with more than one species, as autapomorphies of single species (if present) are not necessary for cladogram construction. The species investigated are listed in table 2. As rhinophorids are sparse in museum collec- tions, most of the species were seen in only few (1 — 5) specimens. The monophyly of the Rhinophoridae, as de- fined above, seems well corroborated by at least three synapomorphies: (1) cephalopharyngeal skeleton of first-stage larvae with toothed mandibles and an elon- gated pharyngeal sclerite, (2) parasites of woodlice, (3) distiphallus with well-developed ventral plates, which are fused along the ventral margins (secondarily free in Paykullia). Two monophyletic subgroups, the Phyto group and the Stevenia group, can be erected on larval morphology, as previously discussed. The monophyly of the Phyto group is corroborated by the apomorphy: (4) eighth abdominal segment of first-stage lar- vae with terminal lobes, a dorsal tongue, and paired ventral ridges. No shared apomorphic characters of the adult morphology have been found for the group, and as the first-stage larva is known for representa- tives of only three of the eight genera, the Phyto group is admittedly somewhat weakly founded. 24 Tijdschrift voor Entomologie, deel 129, afl. 2, 1986 The first split in the Phyto group separates Pay- kullia + {Melanophora + Bequaertiana) from the remaining genera, this group possessing the synapomorphies : (5) female terminalia of the reduced non-teles- copic type, (6) wing cell r4+5 long petiolate. Herting (1961) states that in the Palaearctic fauna only Paykullia and Melanophora possess shortened female terminalia (character 5), and Crosskey (1977), in his revision of the Afro- tropical fauna, notes that the female terminalia of the Afrotropical species apparently is of the normal telescopic type, although he did not dis- sect any specimen. Females of Bequaertiana are still unknown, but the assumed presence of non-telescopic ter- minalia seems well founded in the close affinity between Melanophora and Bequaertiana, as dis- cussed below. Character 6 is rather weak as the petiolate condition has arisen independently several times in the Rhinophoridae, and in Bequaertiana and Melanophora asetosa Kugler the bend of M is missing and an ancestral petiolate condition has to be assumed. Paykullia is a well-defined genus with the fol- lowing apomorphies: (7) distiphallus stout, possessing a strongly spinose pad on the ventral margin of each of the ventral plates and with the dorsal wall more or less prolonged (fig. 7), (8) male abdominal sternite 5 simple. As most male calyptrates possess a more or less excavated abdominal sternite 5, the simple, almost rectangular shape in Paykullia must be an apomorphic character. The monophyly of Bequaertiana + Melano- phora is corroborated by the synapomorphies: ( 9) parafrontalia with several (about A — 7) proclinate orbital setae, (10) male antennae with characteristic bottle- brush-like hairing (Crosskey, 1977: figs. 17 and 27). (11) hind coxae elongated. The hind coxae of Bequaertiana males (fe- males still unknown) are distinctly elongated; in both sexes of Melanophora roralis they are only slightly so. In addition, Bequaertiana and Mela- nophora possess very similar distiphalli (figs. 8, 9). The family affinities of Bequaertiana have been much discussed, Zumpt (1956) even sug- gesting an acalyptrate assignment. Crosskey (1977) doubts whether Bequaertiana is a rhino- phorid and although he notes the resemblance of the head to that of Melanophora roralis he is more inclined to accept a relation to Parazami- mus, another aberrant genus from the rainfo- rests of Zaire. The striking agreement in the apomorphic structure of the male antennae, the head, and the hind coxae of both Melanophora and Bequaertiana, however, leaves no doubt of their close affinity. Actually a case can be made for treating them as congeners. Melanophora asetosa, of which only the female is known, seems to be a typical Melanophora (as judged from the description in Kugler (1978)) except for the absence of the bent part of vein M, which is an apomorphic character of Bequaer- tiana\ In the collection of the Zoological Mu- seum, University of Copenhagen, there is a sin- gle female Melanophora from Kenya, Naro Mo- ru, likewise with the bend of vein M missing. The terminalia appear to be of the short non- telescopic type found in Melanophora and Pay- kullia (as seen in situ, the specimen is not dis- sected). On this evidence it seems most proba- ble (with a parsimonious concept) that the re- duced terminalia are a synapomorphy for the group Paykullia + {Melanophora + Bequaer- tiana). The discovery of a female Bequaertiana and a male Melanophora asetosa may be most inter- esting, and if, as I think is most probable on the present evidence, the genus Melanophora is par- aphyletic with respect to Bequaertiana, it will be necessary either to fit Bequaertiana into the generic limits of Melanophora or to place M. a- setosa in the genus Bequaertiana. Melanophora asetosa and Bequaertiana share the apomorphy: (12) bent part of vein M absent (Kugler, 1978: fig. 15; Crosskey, 1977: fig. 28). A similar wing venation occurs in Oplisa aterrima but is obviously a convergence. The genus Bequaertiana possesses some re- markable autapomorphies: (13) tibiae in males without clearly differ- entiated bristles, (14) male abdomen covered with thick silvery pollinosity, (15) wing vein Ri strongly haired along its length. Melanophora (in the restricted sense with M. roralis as the only representative) is charac- terized by the distinctive white wing tips in fe- males. The sister group to Paykullia + {Melanopho- ra + Bequaertiana) is somewhat ill-defined and Pape: Rhinophoridae 25 Figs. 7 — 12. Aedeagus of Rhinophoridae, lateral view: 7, Paykullia maculata (Fallen). 8, Melanophora roralis (Linnaeus). 9, Bequaertiana argyriventris Curran. 10, Callidesia picttpennis Kugler. 11, Tromodesia angustifrons Kugler. 12, Baniassa fascipennis Kugler. may be polyphyletic. The possible monophyly of the group is corroborated by the single syna- pomorphy: (16) surstylar base extended medially (fig. 13). This may seem very conclusive, but several exceptions are found. The median extension is absent m Phyto pauciseta and both species of Baniassa, and indistinct m Phyto angustifrons. The first split in this group separates Tromode- sia + Callidesia from the remaining genera, their monophyly being corroborated by the synapomorphies: (17) clypeus distinctly bulging, (18) distiphallus of characteristic shape with the sclerotization of the spermduct bent dor- sally (figs. 10,11). The two genera are depicted as sister groups on the cladogram (fig. 30), but they are very similar and could as well be treated as a single genus. I have not seen any specimen of Tromo- desia vibripennis Rondani, the type species of Tromodesia, and therefore I have not been able ta evaluate the monophyly of the genus, i.e., to investigate whether T. vibripennis is more 26 Tijdschrift voor Entomologie, deel 129, afl. 2, 1986 M.EXT Fig. 13. Phyto melanocephala (Meigen). Cerci and surstyli, ventrolateral view, showing median exten- sions (m. ext) of surstyli. closely related to T. angustifrons than to any other species (or species group). The monophyly of the sister group of Tromo- desia + Callidesia seems well corroborated by the apomorphies: (19) lunula with setae, (20) notopleuron haired in addition to the usual two bristles, (21) katepimeron haired. All these traits occur sporadically in other rhinophorids, viz., many species of Paykullia and Rhinophora lepida possess some lunular se- tae; Tricogena, some Stevenia, and Rhinomori- nia sarcophagina may have a few additional no- topleural hairs; and some Rhinomorinia may have an occasional hair on the katepimeron (barette). However, the combination of these traits seems to have arisen only once in the Rhi- nophoridae. Baniassa is the possible sister group of Phyto and is well characterized by the apomorphies: (22) strongly holoptic eyes in males, (23) wings darkened apically, (24) wing cell r^^^ petiolate, (25) metathoracic spiracle with operculum. As no other rhinophorids possess opercular metathoracic spiracles (character 25) this may be considered a reversal to the plesiomorphic condition. Baniassa paucipila Pape does not possess any of the synapomorphies 19 — 21 cor- roborating the monophyly of Baniassa + Phyto. However, the reduced hairing of Baniassa pau- cipila may be secondarily correlated with the yellow colouration of the thorax. Many yellow forms, e.g., the totally yellow species oi Paraza- mimus and Bequaertiana from the rainforests of central Africa, possess a deviating, often re- Figs, l'i — 16. Aedeagus of Rhinophoridae, lateral view. 14, Parazamimus congolensis Verbeke; a = dorsal scle- rotization, dorsal view. 15, Phyto angustifrons (Rondani). 16, Phyto melanocephala (Meigen). Pape: Rhinophoridae 27 duced, hairing. This may be correlated with an association to a humid habitat. The aedeagus of B. jascipennis is shown in fig. 12. Phyto (including Cirillia) possesses the apo- morphies: (26) sclerotization of the spermduct interrupted (figs. 15, 16), (27) strong pre-alar bristle. Cirillia is characterized by the strongly devel- oped parafacial setae and a long-petiolate wing cell r4+5. These characters are likewise found in many species of Phyto, e.g., P. hertingi Baez, and as Phyto does not possess any derived char- acters not shared with Cirillia, a generic separa- tion between these seems unnatural in a phylo- genetic sense. Parazamimus is a strange monotypic genus from the tropical rainforests of Zaire. The single specimen known is in somewhat bad condition and the micropin, by which the head is mounted on the body, unfortunately penetrates the lunu- la, making it impossible to see whether setae are present. The structure of the distiphallus with the reduced sclerotization of the spermduct (fig. 14) is very reminiscent of Phyto, and Parazami- mus is tentatively placed as a sister group to Phyto although it does not possess any of the synapomorphies given for Baniassa + Phyto. Returning to the other group that could be erected on larval morphology, the Stevenia group, the possible monophyly is corroborated by the apomorphies: (28) setal bases of first-stage larvae produced into proleg-like structures, (29) acrophallus sclerotized and tripartite. Other genera like Parazamimus and Tromo- desia have the acrophallus partly sclerotized, but apparently developed independently and without the tripartition which is so characteris- tic of the Stevenia group. Typically the acro- phallus is divided into two lateral and one ven- tral sclerotization (the latter being the extension of the spermduct sclerotization), but often a dorsal acrophallic sclerite is more or less dis- tinct. In some genera this dorsal sclerite is sim- ple but in others it is provided with two lateral armlike processes. The three acrophallic scle- rites are more or less grooved and probably guide the sperm into the ducts of the female seminal receptacles; a functional analogue to the acrophallus of many Tachinidae and Sarcopha- gidae (for the latter see Lopes, 1966; Lopes & Kano, 1968). The first split in the Stevenia group separates Melanomyoides, Queximyia, Rhinomorinia, Rhinophora, and Ventrops from the remaining members of the group. All five genera have a general Rhinomorinia-Yikc appearance and their monophyly is corroborated by the synapomor- phy: (30) dorsolateral processes of distiphallus fused into a single median sclerotization (fig. 22a). Queximyia is a monotypic genus from South Africa, easily recognized by the very long an- tennae and characteristic head profile (Cross- key, 1977: fig. 14). The possession of a strong pre-alar bristle suggests an affinity with Phyto, but a bare katepimeron, the lack of lunular se- tae, and the fusion of the dorsolateral processes of the distiphallus suggest this to be unlikely. The long antennae could be taken as evidence for a close affinity to either Azaisia or Acompo- mintho, but no other characters support this po- sition and the present assignment based on the aedeagal structure (fig. 18) seems the best cor- roborative. Ventrops is another well-defined Afrotropical genus, at present with only a single described species, but other species are known. The ae- deagus of V. milichioides is shown in fig. 17. Ventrops possesses the following apomorphies: (31) eyes greatly enlarged, occupying most of the side of the head and with a concave hind margin (Crosskey, 1977: fig. 13), (32) cerei very short and almost concealed be- tween the surstylar bases. The remaining three genera, Melanomyoides, Rhinomorinia, and Rhinophora seem to com- prise a monophyletic group corroborated by their apomorphic head structure: (33) epistome strongly warped forwards (Crosskey, 1977: figs. 8—10, 12) Melanomyoides is a monotypic genus, its rep- resentative M. capensis being originally de- scribed as a species of Chaetostevenia Brauer (= Paykullia) by Zumpt (1959). Crosskey (1977) discusses the affinity of Melanomyoides to other (supposed) rhinophorid genera, and mentions an extreme superficial similarity to Melanomya and an even closer resemblance to Angioneura. These similarities, however, are founded in all three genera being composed of small, shining black flies with holoptic eyes in the male, characters which are not especially convincing; Crosskey concludes by stressing the resemblance in head profile and distiphallus bef^een Melanomyoides and Rhinomorinia. Similarly, a case could be made for a sister 28 Tijdschrift VOOR Entomologie, deel 129, afl. 2, 1986 0.2 mm 22 Figs. 17—22. Aedeagus of Rhinophoridae, lateral view: 17, Ventrops milichioides Crosskey. 18, Queximyia fla- vipes Crosskey. 19, Rhinophora lepida (Meigen). 20, Melanomyoides capensis (Zumpt). 21, Rhinomorinia xan- thocephala (Bezzi). 22, Rhinomorinia sarcophagma (Schiner); a = dorsal sclerotization, dorsal view. Pape: Rhinophoridae 29 group relation between Melanomyoides and Rhinophora, both having somewhat similar wings with a petiolate cell r4+5, but a very short petiole occurs in some Afrotropical Rhinomori- nia. Melanomyoides is easily distinguished by the almost leaflike surstyli, the holoptic male eyes, and the petiolate wing cell r4+5. The aedea- gus o{ M. capensis is shown in fig. 20. Rhinophora is likewise monotypic and is easi- ly separated from Melanomyoides by the di- choptic eyes in males and the presence of lunu- lar setae. The aedeagus is shown in fig. 19. The genus Rhinomorinia is difficult to char- acterize on external adult morphology and I have only found a single character which may establish the monophyly of the genus: (34) Distiphallus ventrally with a greatly en- larged spinous surface (figs. 21, 22). The long and slender cerei and surstyli (Crosskey, 1977: figs. 34, 35) may be another character, but a very similar condition is seen in Queximyia. The sister group to the four Rhinomorinia- like genera possesses the following apomor- phies: (35) acrophallus more complex, the sclerites being longer and more distinctly grooved, (36) dorsal wall of distiphallus extended. A dorsal extension is likewise found in Mela- nomyoides (fig. 20) but this is probably a con- vergence. Two other characters which may be synapo- morphies for this group are: (37) dorsal acrophallic sclerite well-developed, with two lateral arms, (38) hypandrium spoon-shaped. Character 37, however, is not found in Meto- plisa, most Oplisa and most Stevenia. Character 38 is especially distinct in Tricogena, Oplisa, Metoplisa, and Azaisia, and the flat hypandrium found in Stevenia must be secondarily derived. The first split in this group separates Acom- pomintho + {Azaisia + Macrotarsina) from the others. The monophyly of these three genera is corroborated by the following synapomorphy: (39) anal vein (A]) shortened. Acompomintho, the only genus endemic to the Oriental Region, is well defined by the long antennae with prolonged second aristal seg- ment, the well-developed parafacial setae (Lopes, 1938: pi. 1, fig. 2) and the long-petiolate wing cell r^_^.y The aedeagus is shown in fig. 25. 0.2 mm Figs. 23 — 25. Aedeagus of Rhinophoridae, lateral view: 23, Macrotarsina longimana (Eggers). 24, Azaisia obscura (Villeneuve). 25, Acompomintho lobata Villeneuve. Abbreviations: d.a.s = dorsal acrophallic sclerite; l.a.s = lateral acrophallic sclerite. 30 Tijdschrift voor Entomologie, deel 129, afl. 2, 1986 Azaisia + Macrotarsina possess the synapo- morphies: (40) anterior katepimeral bristle much weaker than posterior one, (41) dorsal acrophallic sclerite well-developed, other acrophallic sclerites slender and situ- ated close together (figs. 23, 24), (42) gonopods (pregonites) thickened (only slightly in Azaisia). In Acompomintho (and all other rhinopho- rids) the two katepisternal (sternopleural) bris- tles (character 40) are subequal to equal in size. Macrotarsina is well characterized by the greatly prolonged male fore tarsi. Azaisia is more difficult to characterize; the most conspic- uous trait, which may be autapomorphic for Azaisia, is: (43) antennae long, with prolonged second aris- tal segment. This character is likewise found in Acompo- mintho and may actually indicate a sister group relation between Acompomintho and Azaisia as accepted by Herting (1961), who established a separate tribe containing these two genera. However, I consider the sister group relation between Azaisia and Macrotarsina to be more corroborated by the present evidence. The monophyly of the sister group of Acom- pomintho + {Azaisia + Macrotarsina) is cor- roborated by the apomorphy: (44) ventral plates of aedeagus with a pair of processes, each supporting a spinous pad (figs. 26—29). The first split in this group separates Metopli- sa from Oplisa + {Stevenia -¥ Tricogena). Kug- ler (1978), in his description of Metoplisa car- bonaria, mentioned the superficial similarity to Oplisa, but he erected the genus because the three humeral bristles of Metoplisa form an ob- tuse-angled triangle and not an almost right-an- gled triangle as in Oplisa. The latter configura- tion is used as a key character for the genus Oplisa by Herting (1961) and Kugler (1978), but both Stevenia (with S. hirtigena as an excep- tion) and Tricogena possess this character. As no other rhinophorids possess this arrangement of the humeral bristles, and as the arrangement in an obtuse-angled triangle is of widespread occurrence, the almost right-angled configura- tion is assumed to be a synapomorphy for Ste- venia, Tricogena, and Oplisa: (45) three humeral bristles forming an almost right-angled triangle. Oplisa was divided by Herting (1961) into the two subgenera Oplisa (as Hoplisa) sensu stricto, characterized by latero-reclinate ocellar bristles, and the monotypic Anoplisa with procHnate ocellar bristles. Kugler (1978) described two ad- ditional species of Oplisa, which both would fall into the subgenus Anoplisa, but as this is clearly a paraphyletic group (as defined by Herting) it is not accepted in the present paper. Oplisa is somewhat difficult to characterize by distinct autapomorphies. The enormously enlarged ejaculatory sclerite of O. tergestina, O. aterrima, and O. oldenbergi (Herting) (see Crosskey, 1977: fig. 40; Draber-Moriko, 1978: fig. 18) is unique in the Rhinophoridae, but O. pollinosa possesses a normal-sized ejaculatory sclerite. The following apomorphies corroborate the monophyly of Oplisa: (46) distiphallus with the processes of the ven- tral plate, which support the spinous pads, situated on a stalked extension (fig. 27), (47) male cerei short and blunt, not separated apically, (48) surstyli broadened apically. It seems fairly corroborated that a sister group relation exists between Stevenia and Tri- cogena, which share the apomorphy: (49) parafacial plate with a row of strong setae. Both genera are very similar in external mor- phology and in the structure of the aedeagus (figs. 28, 29). Stevenia is a well-defined genus with the following apomorphies: (50) wing cell r4+5 petiolate, (51) hypandrium flat, (52) mid femur in males with a posteroventral comb of short stout bristles apically. Some species do not, however, possess char- acter 52 (Herting, 1961), which may define an infrageneric subgroup. Genus incertae sedis Comoromyia Crosskey. Crosskey (1977) described the genus on a sin- gle female of C. griseithorax. I have not seen this specimen, which seems to be the only one known at present, and I have not been able to incorporate the genus into the cladogram on the basis of the description alone. Crosskey men- tions a possible relationship with Phyto, as Comoromyia possesses a strong pre-alar bristle, but the bare katepimeron weakens this argu- ment. I prefer to exclude Comoromyia from the cladogram (fig. 30) until more information is available, especially with regard to the structure of the aedeagus as this provides several of the set-defining characters of the present analysis. Pape: Rhinophoridae 31 sp.p Figs. 26 — 29. Aedeagus of Rhinophoridae, lateral view: 26, Metoplna carhonaria Kugler. 27, Oplisa aterrima (Strobl). 28, Tncogena rubricosa (Meigen). 29, Stevenia atramentana (Meigen). Abbreviations: ext = stalked extension of ventral plate; sp.p = spinous pad. Acknowledgements I am grateful to the following colleagues for kindly providing me with a valuable material: Drs. M. Baez (Universidad de la Laguna, Tene- rife), E. de Coninck (Musée royal de l'Afrique Centrale, Tervuren), R. W. Crosskey (British Museum (Natural History), London), R. Dan- ielsson (Zoologiska Institutionen, Lund), A. Freidberg (Tel Aviv University, Tel Aviv), P. Grootaert (Institut royal des Sciences Naturelle 32 Tijdschrift voor Entomologie, deel 129, afl. 2, 1986 de Belgique, Brussels), B. Herring (Staatliches Museum für Naturkunde, Stuttgart), H. J. Müller (Deutsches Entomologisches Institut, Eberswalde-Finow), K. A. Schmidt (American Museum of Natural History, New York), P. Tschorsnig (Staathches Museum für Natur- kunde, Stuttgart), and N. E. Woodley (National Museum of Natural History, Washington, D.C.). I am indebted to Mr. R. W. Crosskey for his help during a visit at the Entomological Depart- ment of the British Museum (Natural History). Special thanks are extended to Dr. L. Lyneborg (Zoological Museum, Copenhagen) for valuable help and advice during my study, and to Drs. V. ! Michelsen and S. Andersen (Zoological Mu- seum, Copenhagen) for many stimulating dis- cussions. Note added while this paper was already in press: The paper by H.-P. Tschorsnig, 1985, Die Struktur des männlichen Postabdomens der Rhinophoridae (Diptera), Stuttg. beitr. Naturk, ser. A 375, pp. 1 — 18, appeared after submitting this paper for publication and a detailed dis- cussion of the hypotheses presented will appear in a future paper. SI -^ -:r CT = O a < Fig. 30. Cladogram of the Rhinophoridae at the generic level. Numbers refer to apomorphies discussed in the text. Genera recorded as woodlouse parasites are marked with an asterisk. Pape: Rhinophoridae 33 References Andersen, S., 19S2. Revision of European species of Sipbona Meigen (Diptera: Tachinidae). — Ent. scand. 13: 149—172. Andersen, S., 1983. Phylogeny and classification of old world genera of Siphonini (Diptera: Tachini- dae). — Ent. scand. 14: 1 — 15. Baranov, N., 1936. Eine neue Calliphorine-Art (Dipt.), die Termiten angreift. — Ann. Mag. nat. Hist. (10)17:646— 651. Bedding, R. A., 1973. The immature stages of Rhino- phorinae (Diptera: Calliphoridae) that parasitise British woodlice. — Trans. R. ent. Soc. Lond. 125 (1): 27—44. Bezzi, M. & Stein, P., 1907. Katalog der paläarktis- chen Dipteren. III. Cyclorrhapha Aschiza. Cy- clorrhapha Schizophora: Schizometopa: 827 pp. — Budapest. Burger, J. F., Anderson, J. R. & Knudsen, M. F., 1980. The habits and life history of Oedoparena glauca (Diptera: Dryomyzidae), a predator of bar- nacles. — Proc. ent. Soc. Wash. 82(3) : 360—377. Cepelâk, J. & Rozkosny, R., 1968. Zur Bionomie der Art Angioneura cyrtoneurina Zetterstedt, 1859 (Rhinophorinae, Diptera). — Acta Zootechnica 17:181—191. Clausen, C. P., 1940. Entomophagous insects: i — x, 1—688. — McGraw-Hill. Crosskey, R. W., 1965. A systematic revision of the Ameniinae (Diptera: Calliphoridae). — Bull. Br. Mus. nat. Hist. B. 16: 33—140. Crosskey, R. W., 1977. A review of the Rhinophori- dae (Diptera) and a revision of the Afrotropical species. — Bull. Br. Mus. nat. Hist. B. 36: 1 — 66. Downes, W. L., Jr., 1955. Notes on the morphology and classification of the Sarcophagidae and other calyptrates (Diptera). — Proc. Iowa Acad. Sci. 62: 514—538. Downes, W. L., Jr., 1965. Tribe Melanomyini. Pp. 932 — 933 in: Stone et al.: A catalog of the Diptera of America north of Mexico. — Agricultural Handbook 276, U.S. Department of Agriculture, Washington, D.C. Draber-Mohko, A., 1978. Scatophagidae, Muscinae, Gasterophilidae, Hippoboscidae, Calliphoridae, Sarcophagidae, Rhinophoridae, Oestridae, Hypo- dermatidae und Tachinidae (Diptera) der Pieni- nen. — Fragm. faun. 22(2): 51—229. (In Polish, with Russian and German summaries.) Girschner, E., 1893. Beitrag zur Systematik des Mus- eiden. — Beri. ent. Z. 38: 297—312. Goureau, M., 1843. Sur un diptère dont la larve vit dans V Helix conspurcata. — Annls Soc. ent. Fr. 1: 77—80. Griffiths, G. C. D., 1972. The phylogenetic classifica- tion of Diptera Cyclorrhapha with special refer- ence to the structure of the male postabdomen: 340 pp. — Junk Publishers. The Hague. Guimaraes, J. H., 1977. A systematic revision of the Mesembrinellidae, stat. nov. (Diptera, Cvclorrha- pha). — Arqs Zool. 29(1): 1—109. Hackman, W. & Väisänen, R., 1982. Different classi- fication systems in the Diptera. — Annls Zool. Fenn. 19:209—219. Hennig, W., 1958. Die Familien der Diptera Schizo- phora und ihre phylogenetischen Verwandtschaftsbeziehungen. — Beitr. Ent. 8 (5— 6): 505— 688. Hennig, W., 1965. Vorarbeiten zu einem phylogene- tischen System der Muscidae (Diptera: Cyclorrha- pha). — Stuttg. Beitr. Naturk. 141 : 1 — 100. Hennig, W., 1976. Anthomyiidae. Pp. I— LXXVIII in Lindner: Die Fliegen der palaearktischen Region 7, 1(1), Stuttgart. Herting, B., 1961. Rhinophorinae. Pp. 1 — 36 in Lindner: Die Fliegen der palaearktischen Region, 64e Lfr. 216, Stuttgart. Herting, B., 1981. Typenrevisionen einiger paläark- tischer Raupenfliegen (Dipt. Tachinidae) und Beschreibungen neuer Arten. — Stuttg. Beitr. Na- turk, ser. A (Biologie) 346: 1—21. Howard, G. W., 1980. Immature stages and affinities of the southern lechwe warblefly, Strobiloestrus vanzyli Zumpt (Diptera: Oestridae). — J. Nat. Hist. 14:669— 683. Kloet, G. S. & Hincks, W. D., 1976. A check list of British Insects. Part 5: Diptera and Siphonaptera. — Handbksldent. Br. Ins. 11 (5): ix -I- 139 pp. Kolubajiv, S., 1962. Die Ergebnisse der zuchten von entomophagen (der Parasiten und Räuber) der schädlichen Insekten (vorwiegend der forstschäd- linge) in der Zeitperiode von 1934 bis 1958. — Rozpr. csl. Akad. Ved 72(6): 1—73. Kugler, J., 1978. The Rhinophoridae (Diptera) of Is- rael.— Israel J. Ent. 12:65—106. Kurahashi, H., 1966. Studies of the calypterate mus- coid flies from Japan V. Revision of the tribe Luci- liini (Diptera, Calliphoridae). — Sci. Rep. Kana- zawaUniv. 11(2): 105—124. Lehrer, A., 1970. Considérations phylogénétiques et taxonomiques sur la famille Calliphoridae (Di- ptera). — Annotnes zool. bot., Bratislava 61: 1 — 51. Lehrer, A., 1973. La taxonomie du genre Sarcophaga Meigen (Fam. Sarcophagidae, Diptera). — An- notnes zool. bot., Bratislava 89: 1 — 22. Lopes, H. S., 1938. Sobre um Rhinophoridae e varios Sarcophagidae da colleccâo do "Deutsches Ento- mologisches Institut" em Berlin-Dahlen. — Mems Inst. Oswaldo Cruz 33(4): 555—565, 4 pi. Lopes, H. S., 1966. Sobre Malacophasomyia g. n. (Diptera, Sarcophagidae) cujas larvas vivem em caddveres de Gastropoda (Mollusca). — Revta bras. Biol. 26(3): 315—321. Lopes, H. S. & Kano, R., 1968. Studies on copulation of some sarcophagid flies (Diptera). Taxonomie importance of some features of male and female genitalia. — Revta bras. Biol. 28(3): 295—301. Lundbeck, W., 1927. Diptera Danica, genera and spe- cies of flies hitherto found in Denmark. Part 7. Platypezidae, Tachinidae: 1 — 571. — G. E. C. Gad, Kobenhavn. 34 Tijdschrift voor Entomologie, deel 129, afl. 2, 1986 McAlpine, J. F., Peterson, B. V., Shewell, G. E., Tes- key, H. J., Vockeroth, J. R. & Wood, D. M., 1981. Manual of Nearctic Diptera 1 : 1 — 674. — Bi- osystematics Research Institute, Ottawa. Mercier, L., 1921. La larve de Limnophora aestuum Villen., diptère marin. — C. r. hebd. Séanc. Acad. Sci., Paris 173: 1410—1413. Mesnil, L., 1939. Essai sur les tachinaires (Larvaevori- dae). — Monographies publiées par les stations et laboratoires de recherches Agronomique, Paris: 1—67. Mueller, A., 1926. Der Stammbaum der deutschen Tachiniden auf Gund der Penisform. — Nova Ac- ta Acad. Caesar. Leop. Carol. 108(1): 1 — 30, 1 pi. Oldroyd, H., 1964. The natural history of fhes: i — xiv, 1 — 324. — Weidenfeld and Nicolson, Lon- don. Pape, T., 1985. A neu' species of Baniassa Kugler from Iraq (Diptera: Rhinophoridae). — Ent. scand. 16:209—211. Parker, H. L., 1953. Miscellaneous notes on South American dipterous parasites. — Boll. Lab. Ent. agr. Filippo Silvestri 12: 45 — 73. Reinhard, H. J., 1929. Notes on the muscoid fhes of the genera Opelousia and Opsodexia with the de- scription of three new species. — Proc. U.S. natn. Mus. 76(20): 1—9. Rikhter, V. A., 1980. Some plesiomorphic characters of the male genitalia in the family Tachinidae (Di- ptera). — Ent. Rev., Wash. 59(4): 162—171. (1982). Roback, S. S., 1954. The evolution and taxonomy of the Sarcophaginae (Diptera, Sarcophagidae). — Il- linois biol. Monogr. 23(3 — 4): 1—181. Rohdendorf, B. B., 1948. A new family of parasitic muscoid Diptera from the sands of Transvolga. — Dokl. Akad. Nauk SSR 63: 455—458. (In Rus- sian). Roubaud, M., 1903. Sur des larves marines de Doh- chopodes attribuées au genre Aphrosylus (Wlkr.). — Bull. Mus. Hist, nat., Paris 9: 338—340. Sabrosky, C. W. & Crosskey, R. W., 1970. The type- material of Muscidae, Calliphoridae, and Sarco- phagidae described by N. Baranov (Diptera). — Proc. Ent. Soc. Wash. 72: 425—436. Salzer, R., 1968. Konstruktionsanatomische Untersu- chungen des männlichen Postabdomens von Calli- phora erythr acephala Meigen (Dipt.). — Z. Morph. Ökol. Tiere 63: 155—238. Séguy, E., 1941. Études sur les mouches parasites. 2, Calliphorides. Calliphorines de l'Europe occiden- tale et meridionale. — Encycl. ent. (A) 21 : 1 — 436. Steyskal, G. C, 1974. Recent advances in the primary classification of the Diptera. — Ann. ent. Soc. Am. 67(3): 513— 517. Thompson, W. R., 1917. Sur un diptère parasite des isopodes terrestres {Phyto melanocephala Meig.). — C. r. Séanc. Soc. Biol. 80: 785—788. Thompson, R. W., 1920. Sur les diptères parasites des isopodes terrestres. — C. r. Séanc. Soc. Biol. 83: 450—451. Thompson, W. R., 1934. The tachinid parasites of woodlice. — Parasitology 26 : 378 — 448, 8 pi. Townsend, C. H. T., 1935. Manual of myiology 2: 1 — 296, 9 pis. — Itaquaquecetuba, Sâo Paulo. Townsend, C. H. T., 1938. Manual of myiology 4: 1 — 246. — Itaquaquecetuba, Sâo Paulo. Verbeke, J., 1962. Contribution a l'étude des Tachini- dae africains (Diptera). — Explors hydro-biol. des lacs Kivu, Edouard et Albert 3(4): 81—185. Wood, D. M., 1979. Tachinidae and Rhinophoridae. Pp. 418 — 419 in Danks: Canada and its insect fau- na. — Mems ent. Soc. Can. 108. Zumpt, F., 1956. Calliphoridae (Diptera Cyclorrha- pha). Part I: Calliphorini and Chrysomyiini. — Explor. Parc natn. Albert. Miss. G. F. de Witte 87:1—199. Zumpt, F., 1959. Diptera (Brachycera): Calliphoridae. — S. Afr. anim. life 6: 427—440.