184.108.40.206.4.4. ORDER DIPTERA Linné, 1758. THE TRUE FLIES
(= Muscida Laicharting, 1781)
by V.A Blagoderov, E.D. Lukashevich, M.B. Mostovski
(a) Introductory remarks. The true flies (Fig. 304) are the fourth most diverse, and one of the most important and popular insect orders comprising more than 100,000 described living species. The concept of the order and its history is taken here generally from V. Kovalev (1984, 1987), Kalugina & Kovalev (1985) and Shcherbakov et al. (1995). Vein nomenclature is given after Wootton & Ennos (1989) with correction according to Shcherbakov et al. (1995). Most references of the first records of nematoceran families was given by Shcherbakov et al. 1995. Further sources are cited when relevant.
(b) Definition. Minute to medium size, rarely large insects, usually with good flight abilities. Antennae variable, formed by two basal segments and multi-segmented (usually up to 20 flagellomeres) to one-segmented flagellum bearing sensory style or arista in the latter case. Mouthparts licking or biting suctorial. Pro- and metathoraces small, immovably fused with large mesothorax. Mesothoracic wing (Fig. 305), unless rarely lost or reduced, with venation not rich, at most with C, simple SC, simple R, both RS and M dichotomous with 4 branches each, simple CuA, closely followed with concave secondary pseudovein iCu (not homologous to CuP), CuP (A1 auct.), 1A (so-called A2 auct.) (alternative is homology by A. Rasnitsyn, pers. comm., of CuA to M5, and of iCu to CuA, similarly to those in hind wings of some other Papilionidea; see Figs. 269, 279). Metathoracic wing reduced to halter. Male lacking 8th abdominal spiracle and with cerci short, one or two-segmented. Larva legless (sometimes with secondary, not articulated appendages), sometimes headless as well, with mouthparts modified and retracted into thoracic segments, holopneustic to apneustic. Pupa with appendages more or less adherent, sometimes encased into puparium (seed-like, hardened, unhatched larval integument). Adults free-living, rarely parasites, larvae terrestrial or aquatic, sometimes parasitic, concealed in various organic substrate, diet of both highly variable.
(c) Synapomorphies. Labium modified into proboscis. Pro- and metathoraces small, immovably fused with mesothorax. Fore wing with venation simplified, with convex vein CuA followed by concave iCu. Hind wing modified into halter. Male lacking 8th abdominal spiracle. Larva legless.
(d) Range. Middle Triassic till now, worldwide.
(e) System and phylogeny, as shown at Fig. 306, is followed generally Hennig (1968, 1973) and V. Kovalev (1987), with correction according Shcherbakov et al. (1995). System of Cyclorrhapha (= Muscomorpha) is given generally after McAlpine (1989) with some modifications. Additional apomorphies for subgroups of Brachycera Orthorrhapha are accepted after Sinclair (1992) and Ovtshinnikova (1998).
(f) History. Diptera originated from some extinct group of mecopterans. Among scorpionfly families, Permotipulidae and especially Robinjohniidae (Permian; both previously assigned to Paratrichoptera, the suborder discarded as polyphyletic by Novokshonov & Sukatsheva 2001) approach generalised Nematocera most closely in the wing structure. It was suggested that transformation of hind wing into a halter took place in a bittacid-like ancestor related to Robinjohniidae, with narrow-based homonomous wings and long cranefly legs, and that several peculiarities of dipteran wing are mecopteran hind wing characters transferred on to the fore wing (Shcherbakov et al. 1995). Alternatively, Diptera are considered descendants from the more generalised scorpionfly family, Permochoristidae (Novokshonov & Sukatsheva 2001). This hypothesis would imply gradual shortening of the hind wing with the compensatory widening of the fore wing, like in some scorpionflies (Permotanyderidae, Liassophilidae etc., Chapter 220.127.116.11.4.1).
There is no reliable record of Diptera until the early Middle Triassic (Anisian), but already in the earliest known assemblage from Vosges, France, various Nematocera co-exist with a single undescribed Brachycera (Krzemiński et al. 1994; Krzemiński 1998; Marchal-Papier 1998). Diptera were quite abundant in this locality (uncommon situation for the Triassic, see below) and were estimated as about 280 specimens or 5 % of total number of collected insects due to numerous aquatic immatures which are not described yet; possibly some of these larvae do not belong to dipterans at all. Up to now a single monobasic family Grauvogeliidae is described from this locality. An undescribed specimen of Diptera was recorded from the Anisian of Mallorca, Spain (Colom 1988).
In the Middle and Late Triassic the dipterans became widespread and diverse, but not numerous, being found in Central Asia (four localities in Kyrgyzstan and Kazakhstan; Shcherbakov et al. 1995), Australia (Queensland; V. Kovalev 1983; Blagoderov 1999, Lukashevich & Shcherbakov 1999), North America (Virginia, USA; Krzemiński 1992) and Europe (Great Britain; Krzemiński & Jarzembowski 1999). It is impossible to draw any conclusions about dominant groups because of scanty finds: in Triassic assemblages other than Vosges Diptera usually form less than 1 % of insect fossils. About 80 specimens are described from the Triassic beds, half a hundred belong to Grauvogelia Krzemiński, Krzemińska et Papier, but usually descriptions of new taxa are based on holotypes only.
The true flies never had complicated wing venation pattern with numerous additional (intercalary) veins, or a lot of cross-veins. Descriptions of all such phenomena produced by old authors were based on misinterpretations due to poorly preserved material. In the light of this, the suborder Archidiptera erected by Rohdendorf (1961) for several taxa should be abolished, because the families included belong in fact to Tipulomorpha and Bibionomorpha. Comparative wing morphology and palaeontology indicate that modern Tipulomorpha belong to the earliest branch of Diptera, and, moreover, that the ancestral dipterans could be assigned to the same infraorder (considered paraphyletic). The most generalised of Triassic Tipulomorpha, Vladipteridae (see Fig. 305), known from Asia (Kenderlyk locality), give us a possibility to reconstruct the groundplan of dipteran wing as already stalked, without alular incision, with sc-r near RS origin and with short oblique R2 entering wing margin close to R1, and with anal veins free, not yet looped. Such Triassic dipterans as Vladipteridae Psychotipinae (Dzhayloucho locality), combining groundplan characters (free R2), tipulomorphan autplesiomorphies (long convex 1A), and non-tipulomorphan apomorphies (anal loop and alular incision) are difficult to classify, but it is practicable to assign them to Tipulomorpha.
A single large wing in the Triassic (10 mm instead of usual 2-6 mm) was found in Australia (Mt. Crosby). This family Tilliardipteridae, despite of the numerous 'tipuloid' features, should be included in Psychodomorpha sensu Hennig on account of loss of the convex distal 1A reaching wing margin and formation of the anal loop. Similar hypothetical forms with short free R2 (absent in Tillyardiptera Lukashevich et Shcherbakov) could represent a group ancestral to all other Psychodomorpha and to Bibionomorpha as well. Discovery of Tillyardiptera annectent between the most primitive Tipulomorpha and Psychodomorpha strengthens the hypothesis that separation of the latter from the former was the first divergence in the history of Diptera.
One of the two main phyletic lineages of Psychodomorpha, Psychodoidea s. l. + Blephariceroidea s. l., is still to be recorded from the Triassic. The second, ptychopteroid lineage, is represented by Nadipteridae, Eoptychopteridae (ancestors of the living Ptychopteridae; Triassic find is doubtful) and Hennigmatidae; possibly, Grauvogeliidae belong to it too. The most archaic member of Ptychopteroidea, Nadipteridae, at present seem to be closest to the ancestors of both Culico- and Bibionomorpha.
The main novelty suggested by V. Kovalev (1987) in his dendrogram, was the independent and heterochronous derivation of both Culicomorpha and Bibionomorpha. Diversity of these infraorders in the Triassic agrees well with the assumption of the later separation of Culicomorpha. Only one chironomid is known from the Late Triassic of Europe (though Culicoidea is still not recorded, but should be represented already too). At the same time bibionomorphan Diptera were rather diverse in the Triassic represented by Procramptonomyiidae (most primitive Bibionomorpha ), Alinkidae (described as Brachycera but plausible Bibionomorpha close to Procramptonomyiidae), Paraxymyiidae (initial group for Sciaroidea), Protorhyphidae (ancestral to Anisopodidae and Brachycera) and Crosaphididae (in some respects more advanced than some extant members of Bibionomorpha). The only extant families known from the Triassic age are Limoniidae (Tipulomorpha) and Chironomidae (Culicomorpha).
The immatures of Triassic Tipulomorpha and Psychodomorpha can be hypothesised (sub)aquatic (developing in submersed organic material, in detritus or in saturated earth, as living Limoniidae, Tanyderidae and Ptychopteridae do), while Bibionomorpha would be terrestrial (developing within decaying plant matter like the recent Anisopodoidea and Mycetobioidea).
In the Early Jurassic Diptera had become one of the most abundant and diverse insect order, and retained the position of dominant or subdominant group later. There are a lot of the Jurassic dipteran assemblages in Western Europe, Central Asia and Siberia (e. g. Handlirsch 1906-1908, A. Bode 1953, Rohdendorf 1962, 1964, Hong 1983a, Ansorge & Krzemiński 1994, 1995, Ansorge 1996a, 1999, Blagoderov 1996, Lukashevich 1996 a, b, Lukashevich et al. 1998, Mostovski 1996, 1997a, b, 1998, 1999a). Limoniidae, Nadipteridae, Eoptychopteridae, Hennigmatidae, Chironomidae, Protorhyphidae (Fig. 307), Procramptonomyiidae (Fig. 308) and Paraxymyiidae are passed from Triassic, but only the first family plays significant role in oryctocoenoces as well as in present. Since that time the living families Trichoceridae (Fig. 309), Tanyderidae (Fig. 310), Psychodidae, Chaoboridae and Dixidae are known.
In the earliest Jurassic (Sogyuty in Kyrgyzstan, probably Sinemurian; German Lias, Toarcian) the records of Culicomorpha are very scanty. But soon culicomorph dipterans began to play an important unless dominant role in many Jurassic freshwater assemblages. In the oligotrophic Jurassic lakes (Chapter 3.3.4) aquatic larvae of these groups became predatory and algophagous (nectic Chaoboridae and benthic Chironomidae, the latter represented with only primitive subfamilies Aenneinae, Ulaiinae, Podonominae and Tanypodinae). Chaoboridae have flourished, and their immatures were often the most numerous fossils among not only the aquatic dipterans but aquatic insects in general. However, the chaoborid generic diversity was not great (usually not more than two genera per locality), and the extinct genera were not too exotic (extraordinary characters are not found, only their unusual combinations). For example, two extremities of the chaoborid larval morphotypes still occur now. The first of them is represented by Eucorethra Underwood which lives near water surface and collects its victims from the surface: larva dark, with head capsule broad (Fig. 311), dorsoventrally flattened, antennae widely separated, air sacs not developed, siphon present. The second type is represented by the highly advanced genus Chaoborus Lichtenstein (nectic predator): larva transparent, with head much elongate, antennae approximated, air sacs developed and siphon absent (Figs. 312, 489). Both types, and some genera with the mosaic of above characters, already existed in the Jurassic.
It should be mentioned that as yet undoubted ancestral forms for any culicomorph family are not described (though several extinct, primitive subfamilies are known in Chironomidae, e. g. Aenneinae and Ulaiinae). Architendipedidae from Sogyuty (alleged ancestor of Chironomidae) appeared to be Eoptychopteridae, Protendipedidae from Karatau (Late Jurassic of Kazakhstan) are too bad preserved to draw any serious conclusions, Rhaetomyidae (Sogyuty and Grimmen) and Dixamimidae (Karatau), supposed ancestors of Chaoboridae and Dixidae, respectively, were shown to be Chaoboridae with standard venation. The single exception among Culicomorpha is the genus Syndixa Lukashevich (Fig. 313), which differs from all dixid genera as well as from all other Culicoidea in having R3 joining R1 like in Chironomoidea, and so could be close to chironomoid ancestors.
The diversity of Bibionomorpha was growing up dramatically if compared to Culicomorpha. Up to date no one extinct culicomorph family is known, but 15 extinct families of Bibionomorpha are described for Mesozoic and one can see a number of steps in the development of various lineages. Procramptonomyiidae (see Fig. 308), a stem-group of the whole the infraorder except for Axymyiiformia, gave several descendants: Elliidae (Late Jurassic - Early Cretaceous, Fig. 314), presumed ancestors of Pachyneuridae and Bibionidae, Cramptonomyiidae (Late Jurassic till now), and Heterorhyphidae (Jurassic). Since Jurassic first rather scarce Axymyiiformia sensu Shcherbakov et al. (1995) are known, i. e. Boholdoyidae (Early Jurassic - Early Cretaceous), Axymyiidae (Late Jurassic till now, Fig. 315), Perissommatidae (Middle Jurassic till now).
Jurassic representatives of the last family were more advanced morphologically then living forms, in particular, Jurassic species of Palaeoperissomma Kovalev (Kubekovo, possibly also in the Early Cretaceous of Turga) lacked discal cell unlike recent Perissomma spp. The first Scatopsoidea appeared in the Early Jurassic represented by Protoscatopsidae (Early to Late Jurassic) and Synneuridae (Early Jurassic till now).
The sciaroid lineage founded by Paraxymyiidae, descendants of Procramptonomyiidae, was continued by Protopleciidae (first representative of Sciaroidea, known since the Early Jurassic, when they were very diverse and numerous, till the Late Jurassic, and completely disappeared in the Early Cretaceous). The descendants of Protopleciidae were Pleciofungivoridae (Fig. 316) and Antefungivoridae that became more diverse then their ancestors in the Middle and Late Jurassic. The primitive sciaroids were small gnats looking like living Sciaridae and probably also developing in decaying plant matter rich in fungal hyphae, the diet being probably the most common among Jurassic terrestrial dipterans. Appeared that time were the further members of Sciaroidea: Mesosciophilidae (Middle Jurassic - Early Cretaceous, a sister group of living Mycetophilidae s. str.), Eoditomyiidae (Early - Late Jurassic, probable ancestor of Ditomyiidae), and Archizelmiridae (Late Jurassic - Late Cretaceous).
Suggested prevalent biology of Jurassic Bibionomorpha was the saprophagy. The larvae of Perissommatidae, Mesosciophilidae and Synneuridae were most probably associated with fungi, the more so that the basidiomycetes entered the fossil record roughly at the same time that the three families in question (Krassilov 1982).
Main groups of Brachycera Orthorrhapha appeared during the Jurassic as well. The principal sources of our knowledge about these flies in the Jurassic are Laurasian deposits. Exceptionally scarce primitive snipe-flies are found in the Early Jurassic beds of Gondwanaland (India, Kotá Formation). Firstly the brachyceran stock has split into Stratiomyomorpha represented by Oligophrynidae, and Asilomorpha. This event occurred not later than in the Sinemurian. In fact, brachyceran flies of at least four lineages did dwell in the Early Jurassic time, i. e. Stratiomyoidea (Stratiomyomorpha), Xylophagoidea + Tabanoidea, Nemestrinoidea and Empidoidea (Asilomorpha). Nemestrinoidea represented by the extinct subfamily Archinemestriinae (Nemestrinidae) (Fig. 317) and extant subfamily Heterostominae (Rhagionemestriidae), and Empidoidea have been found in the Toarcian of Grimmen (German Lias, J. Ansorge, pers. comm.). One can assume archinemestriins to be parasitic like recent tangle-veined flies, and connected with plants possessing flowers or flower-like organs as nectar feeders and possible pollinators. Other parasitic brachyceran flies are recorded since either the Middle (first Archisargidae, Fig. 318, and Mythicomyiidae) or Late Jurassic (Eremochaetidae, Acroceridae, Bombyliidae). Aphagy cannot be excluded for the adults of Eremochaetidae. Generally, the Middle and Late Jurassic deposits yield a vast majority of the brachyceran orthorrhaphous flies, namely Stratiomyidae (Beridinae, Fig. 319), Xylomyidae (= Solvidae), Xylophagidae (Coenomyiinae), Kovalevisargidae, Rhagionempididae, Asilidae, Therevidae (Fig. 320), Vermileonidae, Hilarimorphidae (Fig. 321), Scenopinidae, Apystomyiidae, and Empididae. Representatives of more advanced nemestrinids belonging to Hirmoneurinae appeared as well. All the Jurassic dance-flies may be allocated to the extinct subfamily Protempidinae characterised by full set of plesiomorphies according to Chvála (1983). Remains of Mesozoic wormlion flies are scarce. Vermileonidae from the Shevia locality in the East Transbaikalia (Central Siberia, Ukurey Formation of debatable Late Jurassic or Early Cretaceous age, see Glushkovo Formation in Chapter 4.1. for details) seem to be plesiomorphic if compared with the recent representatives of the family (B. Stuckenberg, pers. comm.), and may represent ancestral forms of vermileonids. Such an early separation from rhagionid-like ancestor is supported by analysis of the male genitalia muscles of the recent flies (Ovtshinnikova 1997).
Generally, despite the presence of a number of families persisting until the present, the leading Jurassic terrestrial families (Protopleciidae, Pleciofungivoridae, Antefungivoridae, Protorhyphidae, Rhagionidae, Fig. 322) were more common and taxonomic diverse that time comparing the later one. As a result the fauna was essentially Mesozoic in appearance. Of interest is the exceptional rarity of dipterans in the later Early and Middle Jurassic insect assemblages of Eastern Siberia westward of Baikal Lake as well as in Northwest Mongolia, and the apparent absolute absence of Culicomorpha and Bibionomorpha there which were dominating groups in other regions.
V. Kovalev (1984) distinguished two stages of the Diptera fauna formation during the Jurassic: the Early Liassic (as reflected by the assemblage of Sogyuty in Central Asia) and the rest of the period. He believed the beginning of Liassic as a separate stage because of the dominance of Protopleciidae, that makes it different from the Triassic as well as from the later Early, Middle and Late Jurassic, when Pleciofungivoridae and Antefungivoridae become dominating, and the first brachyceran flies have appeared. Now we know, that brachyceran flies are more ancient than it was believed a few years ago, and difference in the dominance structure between the earlier Early Jurassic and the other Jurassic dipteran assemblages does not seems so crucial. At the same time, taxonomic composition of the Jurassic assemblages differs radically from that in the Triassic and Cretaceous ones (particularly so from the Late Cretaceous).
The Cretaceous dipteran assemblages are numerous and well-recorded in both Laurasian and Gondwanan deposits (e. g. Jarzembowski 1984, Jell & Duncan 1986, Hong & Wang 1988, 1990, S. Waters 1989a, b, Grimaldi 1990, V. Kovalev 1990, Kalugina 1993, Martins-Neto & Kucera-Santos 1994, Blagoderov 1995, 1997, 1998a, b, 2000, Coram et al. 1995, 2000, Mostovski 1995a, b, 1999a, Ren et al. 1995, Ren 1998, Blagoderov & Martínes-Delclòs 2001, Mostovski & Martínes-Delclòs 2000). Another significant source of information about the dipteran assemblages, inclusions in the fossil resins, became available since the Early Cretaceous (e. g. Hennig 1971, 1972, Poinar 1992a, Borkent 1995, 1996, 1997, Szadziewski 1995, 1996, Szadziewski & Arillo 1998, Arillo & Mostovski 1999, Azar et. al. 1999b, Grimaldi & Cumming 1999, Mostovski 1999b, S. Waters & Arillo 1999). Tiny and delicate dipterans are preserved in resins much better than as compression fossils. Moreover, fossil resins yield the insect assemblages originating from palaeoenvironments usually differing strikingly from those producing the compression fossils. Fossil resins, especially of the retinite type, were often originated from riverine forests, with the aquatic insects being enriched there by the rheophilous groups, and the terrestrial ones by the tree-trunk dwellers (Chapter 18.104.22.168.1).
Unlike the Jurassic, Cretaceous fauna consisted mostly of Cainozoic families, with the majority of the Jurassic ones going extinct either near or before the Jurassic-Cretaceous boundary (Nadipteridae, Hennigmatidae, Ansorgiidae, Eoditomyiidae, Oligophrynidae, Archisargidae, Kovalevisargidae) or toward the mid-Cretaceous time (Eoptychopteridae, Protorhyphidae, Procramptonomyiidae, Paraxymyiidae, Elliidae, Boholdoidae, Pleciofungivoridae, Antefungivoridae, Mesosciophilidae, Archizelmiridae, Eremochaetidae). The brachyceran Rhagionempididae are supposed to persist till now being represented by genera currently included in the family Apsilocephalidae (Nagatomi & Yang 1998). As for the endemic Cretaceous families of Nematocera, described only from China, their taxonomic status is too unclear to be discussed here. Re-examination of the type material might result in synonymisation of some of them with recent families (e. g. Gracilitipulidae and Zhangobiidae with Limoniidae, Sinotendipedidae with Blephariceridae).
Despite the extinction of the Jurassic families, diversity of the order was increasing due to appearance of Tipulidae, Ptychopteridae, Blephariceridae, Corethrellidae, Culicidae, Ceratopogonidae, Simuliidae, Thaumaleidae, Cramptonomyiidae, Bibionidae (Pleciinae), Bolitophilidae, Diadocidiidae, Keroplatidae, Sciaridae and Cecidomyiidae (although, the Late Jurassic age cannot be excluded for some of these records).
Further asilomorph and cyclorrhaphan families entered the record during the Early Cretaceous, viz. Tabanidae, Athericidae, Bombyliidae, Apioceridae, Dolichopodidae, Opetiidae, although changes of taxonomic composition were more considerable at subfamily level. The eremochaetid subfamily Eremomukhinae with derived wing venation replaced more archaic Eremochaetinae. Archinemestriinae were being gone, and different Nemestrininae appeared within the tangle-veined flies. More or less advanced Atelestinae, Oreogetoninae, Hybotinae and Microphorinae substituted Protempidinae in the family Empididae (Fig. 323). Some Early Cretaceous dance flies demonstrate adaptations to predatory way of life of adults such as raptorial legs and elongated stout proboscis. Representatives of Apioceridae are supposed to be flower-visitors as possessing the long proboscis. The first cyclorrhaphan flies appeared in the Early Cretaceous represented by Platypezidae, Ironomyiidae (Sinolestinae), and by rarer Lonchopteridae, Sciadoceridae and Phoridae (Prioriphorinae).
The Late Cretaceous time was marked with appearance of more high-rank taxa of brachycerans, viz. Parhadrestiinae (Stratiomyidae), Empidinae, Tachydromiinae (see Fig. 304), Trichopezinae (Empididae), Ironomyiinae (Ironomyiidae), Metopininae (Phoridae), Syrphidae, Pipunculidae, Milichiidae, Calliphoridae. The scuttle flies of the subfamily Prioriphorinae became more common in the Late Cretaceous. Findings of Cyclorrhapha were being increased in number toward the end of the Mesozoic. However, the origin of the Cyclorrhapha remains one of the most controversial issues in dipteran systematics (Wiegmann et al. 1993). There is widespread agreement that the affinities of these flies lie within orthorrhaphan Heterodactyla, and most probably within the Empidoidea.
In spite of the taxonomic diversification the adaptive zone of the order only weakly enlarged. There were only some cases of origin of new life forms, e. g. those with necrophagous larvae (Sciadoceridae, Calliphoridae and maybe Phoridae: Prioriphorinae); possibly also the first endophytic Cecidomyiidae. The appearance of undoubted blood-sucking dipterans (Corethrellidae, Culicidae, Ceratopogonidae, Simuliidae in Culicomorpha, Phlebotominae in Psychodomorpha, and Tabanidae, Athericidae, and Palaeoarthroteles (Rhagionidae) in Asilomorpha) was very important and significant event in the Cretaceous coinciding with the appearance of other presumed blood-sucking insects (e. g. Saurophthirus Ponomarenko, Chapter 22.214.171.124.4.5). It is possible that sucking of the body fluids is more ancient diet in lower dipterans because some recent representatives of relict groups which flourished during the Jurassic (Tanyderidae, Chironomidae Podonominae) have long piercing mouthparts as well as some fossil Tanyderidae (Kalugina 1991).
Composition of the freshwater dipteran assemblages changed strongly toward the mid-Cretaceous. In the numerous Asian localities of the Early Cretaceous age insect-bearing beds are literally filled with chaoborids, sometimes with the immatures dominating, sometimes the adults (Fig. 324). Density of their remains may be as high as 300-400 specimens per 100 square cm. Toward the end of the Early Cretaceous Chaoboridae reduced their abundance while Chironomidae were taking their dominant position, which is usually retained till now. Since that time the subfamily Chironominae is known, now the dominant group of chironomids in eutrophic lakes. Supposedly this was an effect of changes in the trophic regime of water bodies resulted from the Mid-Cretaceous biocoenotic crisis (Chapter 3.3.6). This was probably the only example of strong reaction of dipterans on the crisis, and this seems natural. The crisis is hypothesised to be provoked by changes in plants, or at least it concerned most of all the animals closely connected with plants ecologically (Chapter 3.2). Yet the terrestrial Diptera lacked close ecological ties with living plants during the Mesozoic.
Among non-aquatic Diptera the role of dominants were taken by Mycetophilidae, various Empididae, and primitive Aschiza since the Early Cretaceous. The fungus gnats of the family Mycetophilidae appeared diverse and numerous in the beginning of the Early Cretaceous, replacing their ancestors, Mesosciophilidae. Remarkable is the modern taxonomic composition of the Early Cretaceous mycetophilids: at least 6 of their 18 genera known from the Early Cretaceous are extant, an exceptional ratio for all the insects.
The general taxonomic composition of the Palaeogene dipteran fauna was more similar to the Cretaceous than to Neogene one. Although the acalyptrate and caliptrate flies, the dominants of recent dipteran fauna, reached considerable taxonomic diversity during the Palaeogene (e. g. Hennig 1965; Evenhuis 1994), their share in assemblages of that time was low (Shatalkin 2000 removed Trypaneoides ellipticus described by Hong 1981 from Lauxaniidae and concluded that it resembles most closely dolichopodids). The transition from the Late Cretaceous to Palaeogene faunas was gradual, with continual growth of polydominance in the assemblages, and with retained prevalence of phytosaprophagy and mycetophagy as the characteristic diet of the leading terrestrial groups. Another evolutionary trend observable during the Palaeogene was the increasing taxonomic and ecological diversity of the terrestrial dipteran fauna. The complex of bloodsucking dipterans was expanded by brachyceran Hippoboscidae, Glossinidae and, possibly, their relatives Eophlebomyiidae. Also increasing was the number of the invertebrate parasites (Conopidae, Cryptochaetidae, Sciomyzidae). It is of much interest that the structure of the Palaeogene fauna differs so much from the Neogene and contemporary ones, in spite of the clear dominance of elements of the contemporary fauna. Only two families, Eophlebomyiidae and Proneottiophilidae, have gone during the Palaeogene. The youngest living nematoceran families with aquatic immatures are known since the Palaeogene - Cylindrotomidae (Mo Clay, Fig. 325) and Nymphomyiidae (Baltic and Bitterfeld amber). The absence of fossil records of Deuterophlebiidae, presumably existed in the Palaeogene too, can be explained by their connection with mountain streams (Chapter 126.96.36.199.1).
The current stage of the dipteran history began at the Palaeogene-Neogene (Oligocene-Miocene) boundary. Characteristic of that stage was rapid diversification of the cyclorrhaphous flies (Fig. 326), origination of many new families andmassive speciation of their members (Evenhuis 1994). They have radiated into an enormous range of both larval and adult niches, phytophagy, coprophagy, necrophagy, parasitising different invertebrate and vertebrate hosts, a range much broader than that of all other flies. Particularly important for that process has been hypothesised the development of wide open steppe-like territories in the Oligocene and Miocene, with their extensive herds of large vertebrates leaving huge amount of dung and corpses (Chapter 3.2). The adaptive explosion of Cyclorrhapha has not been reached at expense of the former dipteran dominants (Empididae, Dolichopodidae, Rhagionidae), for the new groups were mastering new niches and resources. Only few Cretaceous and Palaeogene families became extinct or rare in the Neogene. As a result, beginning from the Neogene, the dipteran ecological and taxonomic diversity is certain to expand comparing the earlier stages of the order history. Another important feature of this last period is that it is impossible to single out any particular dominant trophic group.
To summarise, the faunogenesis of the Diptera had several stages in its development. Triassic fauna was nearly completely consisted of extinct families, so it looks very peculiar. Jurassic one is peculiar as well due to dominance of the typically Jurassic taxa (Protoplecidae, Pleciofungivoridae, Protorhyphidae, Archisargoidea), despite appearance of several families persisting till now, even if some of these (e. g. Limoniidae and Chaoboridae) played the significant role in oryctocoenoces. During the Cretaceous and Palaeogene the dominance structure was gradually changing toward the modern one. The typical Jurassic families and subfamilies had become extinct (Eoptychopteridae, Protorhyphidae, Procramptonomiidae, Protoplecidae, Pleciofungivoridae, Eremochaetidae, Prioriphorinae) or relict (Chaoboridae, Perissommatidae, Ironomyiidae, Sciadoceridae), being replaced by the extant ones (e. g. Mycetophilidae, Sciaridae, Bibionidae). By the Neogene that process, along with the rapid radiation of Cyclorrapha, resulted in appearance of the polydominant fauna of modern type. Of special interest is the rapid initial radiation of the order, with both suborders appearing in the Triassic, i. e. at the very beginning of the order evolution. The same is true for some subordinate taxa, particularly for brachyceran flies. For example, the orthorrhaphous Heterodactyla, Aschiza and Schizophora seem to appear suddenly, from nowhere, and in a great diversity at once.