[Origin and evolution of parasitism in mites of the infraorder Eleutherengona (Acari: Prostigmata). Report I. Lower Raphignathae].

The evolution of animal parasitism in mites of the infraorder Eleutherengona (Prostigmata) is discussed. Parasitism has arisen independently in numerous phyletic lineages or superfamilies of this infraorder. Mites of the family Pterygosomatidae are parasites of terrestrial arthropods and lizards, and species of Myobiidae are exclusively associated with placental and marsupial mammals. Most families of the superfamily Cheyletoidea comprise permanent parasites of vertebrates, and mites of the sister families Cloacaridae and Epimyodicidae, whose phylogenetic relationships with other eleutherengones are unknown, are endoparasites of turtles and small mammals, respectively. Moreover, some families of the diverse cohort Heterostigmata include insect parasitoids or true parasites of insects. Some phylogenetic lineages, such as Heterostigmata and Cheyletoidea, present a series of life forms transitioning between free-living predators, nidicolous predators, fungivorous mites, and, finally, highly-specialized ecto- and endoparasites. The representatives of some phylogenetically distant eleutherengone lineages developed similar adaptations to predation and parasitism. However, in spite of some similarities in these adaptations, the evolutionary trends and pathways for switching to a parasitic mode of life are quite different in particular eleutherengone lineages. With a few exceptions, each eleutherengone lineage is associated with a particular host group. Temporary parasitism by the larval stage only, a life-history pattern characteristic of the Parasitengona, does not occur in these mites, and all active stages are parasitic and live on host, except for the cheyletid tribe Chelonotini where only adult females are parasites. Species in most eleutherengone lineages that parasitize hosts of particular groups are mono- or oligoxenous parasites, and, therefore, good potential models for co-phylogenetic studies. Mites of the family Pterygosomatidae are permanent parasites of lizards and various arthropods that typically live in concealed habitats. This family represents the only example of mites belonging to one exclusively parasitic eleutherengone group occurring on both invertebrate and vertebrate hosts. The most probable explanation of the occurrence of this family on such varied host groups is host switching. It is hypothesized that Pimeliaphilus-like mites parasitized different arthropods living in concealed spaces and switched to parasitism on lizards that sought refuge in similar spaces. This host switch would have been ancient to account for the diversity of lizard parasites seen today. Mites of the family Myobiidae are ectoparasites of mall marsupial and placental mammals. The development of effective structures for attaching to mammalian hair, primarily involving the first pair of legs, has increased in complexity during myobiid evolution. The subfamily Xenomyobiinae includes a single species Xenomyobia hirsuta parasitizing the Peruan marsupial Lestoros inca. Legs I of this species are only slightly modified. Mites of the subfamily Archemyobiinae, tribe Archemyobiini parasitize other South American marsupials. In these mites, the genua of the female legs I bear paired ventral clasping organs. Mites of the tribe Australomyobiinae parasitize Australian marsupials. The last tribe and all other myobiids belonging to the subfamilies Protomyobiinae and Myobiinae have an unpaired attachment organ on genu I. The subfamily Protomyobiinae is separated onto 3 tribes. In more derivative members of the tribes Acanthophthiriini (parasites of bats), Protomyobiini (parasites of "insectivores"), and in all Elephantulobiini (parasites of elephant shrews), the tibia and tarsus of legs I are fused apically. Mites of the subfamily Myobiinae parasitize rodents and are the most morphologically specialized, with 3 apical segments of legs I fused. Basing on the wide distribution of myobiid mites on mammalian taxa, it could be concluded that the origin of parasitism occurred in these mites not later than in the Lower Cretaceous, the assumed time of divergence of marsupial and placental mammals (Carroll, 1993). The host associations of the superfamily Cloacaroidea is still unresolved problem. Mites of this superfamily show a high level of host specificity; however, cloacaroid families and subfamilies are associated with phylogenetically distant lineages of vertebrates, turtles (Cloacarinae), birds (Pneumophaginae), and mammals (Epimyodicidae). The common ancestor of cloacarines may have become associated with turtles before the late Triassic. We believe that parasitism of Pneumophagus bubonis, the single species of Pneumophaginae, in the lungs of Bubo virginianus (Aves: Strigiformes) is a result of an ancient host shift following predation by some birds on live or. dead turtles. All 4 currently known species of the Epimyodicidae are endoparasites in the subcutaneous tissues of voles (Rodentida: Cricetidae), moles, and shrews (Soricomorpha: Talpidae and Soricidae). The sister relationship between Cloacaridae and Epimyodicidae could reflect parasitism by their common ancestor on archaic amniotes before the divergence of the Synapsida (ancestors of mammals) and Diapsida (ancestors of turtles and birds). On the other hand, the ancestors of epimyodicids, retaining some features of their free-living ancestors, could have switched from diapsids onto synapsids considerably later.