The family Parasitidae (Acari: Mesostigmata) - history, current problems and challenges

The family Parasitidae comprises two subfamilies, Parasitinae and Pergamasinae. A generic concept of this family is not stable and requires a revision. The number of genera varies depending on authors and their view on the systematics of the family, especially on the rank of taxa. We recognize 23 genera within Parasitinae and 22 genera within Pergamasinae, 4 of them are subdivided into subgenera. A dichotomous key for separation of genera and subgenera is provided. A new genus, Coprocarpais n. g. with type species Parasitus copridis Costa, 1963 is proposed and a diagnosis is given. The genus Erithosoma is not assigned to any subfamily because a description of females is missing. The most pressing taxonomic problems at generic and subgeneric levels are discussed. Fourteen new combinations are proposed.


Introduction
Mites of the family Parasitidae Oudemans, 1901 are free living predators which can be found in soil, leaf litter, caves, dung, compost, nests of birds or small mammals, on carrions and other similar habitats. Some species are associated with beetles or bumblebees. If adults live in temporary habitats, deutonymphs are often phoretic on insects (Hyatt 1980). This different lifestyle of adults and deutonymphs leads to a situation, that from many species only deutonymphs are known. However, this is only a small problem compared to other taxonomic issues within the family.
Over lats two decades, several authors (Karg 1998, Alberti et al. 1999) analysed higher classification and a position of Parasitidae within Parasitiformes but this very interesting and important topic is not crucial for practical work with parasitid mites. The identification of species or classification in the genus is often confusing or difficult because the identification keys are outdated (Tikhomirov 1977, Karg 1993 or not existing, information on the taxonomy are scattered over hundreds of publications, the generic concept and the boundaries of the genera have been changed many times. Latreille (1795) designed the genus Parasitus but he changed the name first to Carpais (1796) and later to Gamasus (1802). That is the reason why Oudemans (1901) mentioned the genus Parasitus Latreille, 1795 while Berlese (1882) recognized the genus Gamasus Latreille, 1802. These names were synonymized by Trägårdh (1910). The subfamily Parasitinae was proposed by Oudemans (1901) but several different genera (e.g. Macrocheles Latreille, 1829, Pachylaelaps, Berlese, 1888, Cyrtolaelaps Berlese, 1887 were included in it. In following papers Oudemans (1902Oudemans ( , 1904Oudemans ( , 1923 mentioned the family Parasitidae, each time with a modified concept. The first review of the Parasitidae was made by Berlese (1905), where he recognized 8 subgenera within the genus Gamasus, the content of which is identical to present family Parasitidae. From that time, several revisions of the family were made and different generic concepts were proposed (Holzmann 1969, Micherdziński 1969, Tikhomirov 1977, Karg 1993. Holzmann (1969) recognized 7 genera and combined Eugamasusand Parasitusspecies into the genus Eugamasus Berlese, 1892. She included Parasitellusspecies in the genus Parasitus. Most of Pergamasinae-species were included in the genus Pergamasus Berlese, 1903. Micherdziński (1969 also combined genera Eugamasus and Parasitus into one genus, but this time it was Parasitus and he recognized several species-groups within this genus. Micherdziński (1969) recognized two morphological types -Parasitus, where he mentioned 4 genera and Pergamasus with 2 genera. Juvara-Balş (1972) also distinguished 2 morphological types within Parasitidae -Pergamasus and Parasitus -for which she designed 2 subfamilies, Parasitinae and Pergamasinae. These subfamilies were previously recognized by Athias-Henriot (1971) as cohorts. Woolley (1988) even mentioned the families Parasidae and Pergamasidae and Athias-Henriot (1980a) mentioned the family Pergamasidae.
Some revisions (Bhattacharyya 1963, Evans and Till 1979, Hyatt 1980 were made only for either the family or one of the subfamilies of British Islands and again different generic concepts can be found in it. A series of three papers written by Athias-Henriot (1967a, b, c) can be considered as a revision of the subfamily Pergamasinae. The most recent identification key to Parasitidae genera and some species can be found in Karg (1993) but this key includes only German genera and provides another different generic concept. The identification keys to the majority of described genera of the subfamily Pergamasinae were published by Juvara-Balş (2002, 2003. The majority of genera belonging to the Parasitinae are not included in any identification key. The characters used to separate and characterized genera were changed almost with every revision. Berlese (1905) separated genera Eugamasus and Parasitus on the basis of male corniculi, Hyatt (1980) on the basis of the shape of paraxial setae on palpgenu and Athias-Henriot (e.g. 1978, 1980b often used the adenotaxy and poroidotaxy do define genera. The last mentioned characters are missing in older diagnoses of genera and that is one of problems for making a complete revision of the family. Over the last decades, many genera were designed (Athias-Henriot 1969, 1979a, b, c, 1980b, c, 1981, 1982Juvara-Balş 1972, 2002Juvara-Balş and Athias-Henriot 1972;Hennessey and Farrier 1989;Tseng 1995) but no summary work comprehending these genera was done. The most recent mention of the generic concept of the Parasitidae can be found in Witaliński and Podkowa (2016). They recognize 16 genera within Parasitinae and 23 genera within Pergamasinae.
The phylogenetic relationships within Parasitidae were never tested with molecular phylogenetic methods. Such future studies could reveal real phylogenetic relationships between species and genera and lead to a completely different generic concept. The aim of our paper is to provide an identification key to all valid genera of Parasitidae, to summarize the biggest taxonomic problems within this family, to provide information to facilitate the work with these mites and thus provide the basis for future revision of the family Parasitidae.

Materials and methods
In the present paper, a dichotomous identification key to the genera and subgenera of the family Parasitidae is presented. The information presented here is derived from a study of the literature, not from the examination of type specimens. Idiosomal setal notation follows that of Lindquist and Evans (1965) with modifications as given by Lindquist (1994). The system of setal notation for the palpi follows Evans and Till (1979). The notation of adenotaxy follows Athias-Henriot (1969). The part of the identification key concerning Pergamasinae is partly adopted and modified from Juvara-Balş (2002, 2003 and Athias-Henriot (1967a).
The subgenera are included in the key since relatively fast changing taxonomic rank of some taxa within the Parasitidae indicates practical needs of having this information in one place.
All valid genera and subgenera are included in the presented key except the genus Erithosoma Athias-Henriot, 1979. This genus is not assigned to any subfamily because a description of female is missing. However, all characters, by which was the genus defined, are mentioned.
Type species: Parasitus copridis Costa, 1963 Diagnosis -Dorsal shield of female divided into podonotal and opisthonotal shield, dorsal shield of male with transverse suture, podonotal shield of adults with twenty-twentythree pairs of heterogeneous setae, opisthonotal shield of adults with more than forty pairs of setae, dorsal shields with distinct foveate sculpture; paraxial seta on palpfemur spatulate or paddle-like, with smooth or denticulate posterior margin, paraxial setae on palpgenu spatulate; opisthonotum extending far on ventrum; endogynium of females often with curled structures; corniculi of males small, deeply split; male palptrochanter with big ventral protuberance bearing two setae modified into flat blades; arthrodial membrane of male chelicera fringed; gnathotectum with three prongs, the middle prong is clearly stronger; tritosternum of male rudimental; only second pair of legs of male with spurs.
The genus Coprocarpais n. g. is characteristic also by its ecology. Deutonymphs are phoretic on beetles from the subfamily Scarabeinae and adults live in dung. All known species are distributed in Asia except P. copridis with a Palaearctic distribution.

Discussion
Division of Parasitidae into Parasitinae and Pergamasinae as proposed by Juvara-Balş (1972) is widely accepted. Juvara-Balş (1975) proposed also a subdivision of Pergamasinae into 3 tribes: Leptogamasini, Pergamasini and Paragamasini. Genera included in Leptogamasini are Leptogamasus, Ernogamasus, Tomeogamasus, Mixogamasus, Phityogamasus and Zelogamasus, genera within Pergamasini are Pergamasus, Holoparasitus, Heteroparasitus and Ologamasiphis and genera within Paragamasini are Amblygamasus and all genera mentioned in our identification key within Paragamasus s. l. Juvara-Balş (1972) mentioned for the first time a division into tribes but she gave a diagnosis only of the tribe Leptogamasini, the other tribes are neither named nor characterized in the paper. Athias-Henriot (1973) mentioned the tribe Pergamasini for the first time but without any diagnosis. The first diagnoses of tribes Pergamasini and Paragamasini were given by Juvara-Balş (1975). We decided to avoid this division until phylogenetic relationships within the subfamily will be solved.
A group of closely relative genera named by Athias-Henriot (1979a, 1980b as neogamasidian series can be recognized within Parasitinae. The content of the neogamasidian series fits with the subgenus Neogamasus Tikhomirov, 1969 in the sense of Tikhomirov (1977). These genera are Neogamasus, Anadenosternum, Colpothylax, Dicrogamasus, Dyneogamasus and Cycetogamasus. Dyneogamasus was originally proposed as a monotypic subgenus of Neogamasus (Athias-Henriot 1979a). That is the only existing mention of this subgenus. However, at least three more species, Neogamasus bisiculus Tseng, 1995, Neogamasus pinatus Tseng, 1995and Neogamasus scirpiculatus Tseng, 1995 presumably belong to Dyneogamasus. For practical reasons, mainly to facilitate a work with parasitid mites, we decided to mention it as a genus but we are aware of the fact that this can be changed with a future phylogenetic analysis. These neogamasidian genera can be easily separated according to shape of paraxial setae on palpgenu, although Hennessey and Farrier (1989) described several species with a combination of characters of more genera within neogamasidian series.
Most taxonomic problems within Parasitinae are associated with the genus Parasitus. The content of the family Parasitidae and that of the genus Parasitus as was originally defined (Latreille 1795), are overlapping. It was only a logical outcome of increasing knowledge, that new genera were excluded from the genus Parasitus. Most of these genera are widely accepted today. It is not surprising, that mainly species without obvious synapomorphies remained in the genus Parasitus, so it becames presumably an artificial group and requires a revision. Tikhomirov (1969) separated species with homogenous dorsal setae into genus Vulgarogamasus but this genus also requires a revision. Among other things, on the basis of studying original descriptions of species described in the last few decades as Vulgarogamasus, mainly in China, we are of opinion, that some of that species belong to the neogamasidian series. Athias-Henriot (1979b designed two new genera, Phorytocarpais and Rhabdocarpais, previously recognized as species groups within Parasitus. The number of species belonging to these genera was described as species of the genus Parasitus, so their revisions are required. Karg (1998) mentioned them as subgenera of the genus Parasitus and designed a new subgenus of South American species. These taxa are considered as genera in our identification key within Parasitus s.l. Several authors (Costa 1975, Tikhomirov 1977, Makarova 1996 recognized a group of closely related species within genus Parasitus. These species possess several morphological synapomorphies, have similar ecology and they are fairly different from type of the genus Parasitus, P. coleoptratorum (Linneaus, 1758). Due to above mentioned reasons we propose a new genus, Coprocarpais n. g., which designation and list of its species are given in the present paper.
Division of the family Parasitidae into two subfamilies is based mainly on the dorsal shield of females, so the classification of the genus based only on males can be mistaken. We assume that the genus Erithosoma belongs to Pergamasinae. This genus was designed for a single known male specimen of Erithosoma pilosum Athias-Henriot, 1979 and its inclusion in a subfamily is uncertain. That is why we do not include this genus in the identification key. Nevertheless, as the aim of this work is to complete knowledge of taxonomy of Parasitidae, we consider necessary not to skip the characteristic of the genus as given by Athias-Henriot (1979c) which enables to distinguish this genus: dorsal shield strongly reticulated, entire, without transverse suture, neotrichous; dorsal setae long and fine, spiny apically; gland pores gd2, gd5 and gv3 absent, gv1 present, gv2 double; anterolateral angle of sternal shield very long, fused to exopodal plate; paraxial setae on palpgenu entire and truncate, paraxial seta on palpfemur divided into five branches, the proximal one is lobed; gnathotectum significantly sculptured, with three prongs; hypostome with eight simple grooves, proximally with field of small denticles; leg setae long and fine, spiny; movable digit of chelicera pluridentate, fixed digit edentate, spermatotreme simple and long, arthrodial oncophysis fringe and very short; femur II with short weak spur, axillary process on femur and seta av1 on genu II and tibia II reduced to conical vestige. Athias-Henriot (1978) designed the genus Paracarpais on the basis of characters which can be found in many species of Parasitidae and sometimes vary within a single genus (See, e.g., Athias-Henriot 1977): paraxial seta on palpfemur fringe, peritrematal shield of female fused to opisthogastric shield, gland pores gv2 double, gd8 and gv3 absent, gv1 small and synarthrodial oncophyses of male chelicerae not transformed. She included four species into genus Paracarpais different enough to design a separate subgenus for each of them. Traditionally these species are located in three different genera (Parasitus, Vulgarogamasus and Porrhostaspis). Witaliński and Podkowa (2016) found as the lowest as well as the highest number of ribbons in sperm of all 27 studied species within the genus Paracarpais. That can indicate a great intrageneric variability of this character or greater phylogenetic distance between these species. This is just one of numerous taxonomic challenges within Parasitidae awaiting a modern phylogenetic study. For the above-mentioned reasons we tended to the traditional attitude to these species and we do not mention the genus Paracarpais in our identification key.
There are also another taxonomic challenges within small, only rarely mentioned genera. We must point out, that genera Psilogamasus, known only from Tanzania, and Taiwanoparasitus, found in Taiwan and China, are very similar and they share several characters which are unique or rare within Parasitinae: hypotrichous dorsal shields, very short setae in anterior part of podonotum, movable digit of female chelicera with four teeth, peritrematal shields free posteriorly. Moreover a big endogynial sac is present in one of Taiwanoparasitus species and in Psilogamasus. However, a presence of a trichocystic seta on telotarsus IV is not mentioned in any Taiwanoparasitus species and a description of adenotaxy and poroidotaxy in Taiwanoparasitus is not available. The mentioned genera also differ in some characters as the number of setae on opisthonotum -five in Taiwanoparasitus vs. six setae in Psilogamasus but this can be an intrageneric variability. A study of types can possibly leads to their synonymization. Furthermore, we are of opinion, that the species Vulgarogamasus brachysternalis Ma and Lin, 2005 and Vulgarogamasus longiscidiformis Ma & Lin, 2005 belong to the genus Taiwanoparasitus as their descriptions fit in most of characters with a diagnosis of the genus Taiwanoparasitus. The species Vulgarogamasus tenuipilosus Karg, 1998 may also belong to one of these genera and should be examined.
Great example how a taxonomic rank can change with growing knowledge is genus Pergamasus. Athias-Henriot (1967a, b, c) published a series of three articles focused on Pergamasus-species. She separated the genus into three subgenera -Pergamasus s. str., Amblygamasus and Paragamasus. Athias-Henriot (1967b) divided subgenus Pergamasus into three sections, which she raised (1971) to a subgeneric level. That is a concept accepted in several papers (Karg 1993, Stănescu andJuvara-Balş 2005) as in the present identification key. Although, within one subgenus can be recognized a subdivision into four species groups proposed by Juvara-Balş (1976). The second subgenus, Amblygamasus, was split by Athias-Henriot (1967c) into seven types of organization, which were never raised to a subgeneric level.
Athias-Henriot (1967a) divided a subgenus Paragamasus into nineteen types of organization, which she raised to generic or subgeneric level (Athias-Henriot 1971, 1980b. However, some species groups raised by Athias-Henriot (1971) to subgeneric level were raised to generic level by Juvara-Balş (1981, 2002. Athias-Henriot (1967a) subdivided some types of organization into sections, which were raised to subgeneric level by Juvara-Balş (1981). Athias-Henriot (1971) raised subgenus Paragamasus to a genus and divided it into eight subgenera but she included only some types of organization mentioned by Athias-Henriot (1967a). Athias-Henriot (1973) described another one subgenus within genus Paragamasus. This division was accepted by some authors (Juvara-Balş 1977, 2002, Schmölzer 1995, Dabert et al. 2011) and Witaliński and Podkowa (2016) mention these taxa as genera included in Paragamasus s. l. except Tanygamasus which they mention as a separate genus. We agree with Juvara-Balş (2002) that these genera are artificial and the Paragamasus group needs a revision. For practical reasons, we consider a taxonomic group Paragamasus s. l. with nine genera in presented identification key. It must be also mentioned, that Karg (1971) proposed a subgenus Lysigamasus, what is a junior subjective synonym of Anidogamasus (Juvara-Balş 2002).
A recent comprehensive revision of the family is still missing, however partial revisions of some genera have been already made. Athias-Henriot (1971) proposed three subgenera within genus Leptogamasus on the basis of her types of organization within genus Paragamasus (Athias-Henriot 1967a). Juvara-Balş (1981) redefined genus Leptogamasus, upgraded existing subgenera to the generic rank (Tomeogamasus and Ernogamasus) and designed new subgenera as mentioned in the present identification key. These subgenera were previously recognized by Athias-Henriot (1967a) as three sections of her organization type parvulus, Juvara-Balş (1972) mentions them as lineages of the genus and Witaliński (1978) also pointed out one of them as a species-group within genus Leptogamasus. Karg (1993) divided genus Leptogamasus into two subgenera -Leptogamasus s. str. and Valigamasus Karg, 1993, which is a junior synonym of Ernogamasus based on the same type species. This ambiguity about subgenera within Leptogamasus solved Juvara-Balş (2003) by publishing an identification key to genera Tomeogamasus, Ernogamasu and Leptogamasus including subgenera.
The genus Holoparasitus was also extensively studied in recent years, the taxonomic rank of its subgenera was raised to genus rank (Juvara-Balş 2002), many new species were described, several revisions of type specimens from old collections were made and new speciesgroups within the genus were defined (Juvara-Balş and Witaliński 2000Witaliński , 2006Witaliński and Skorupski 2002, 2003a, b, 2007Witaliński 1994Witaliński , 2004Witaliński , 2006Witaliński , 2017. The genus Holoparasitus comprised of three subgenera, two of which were redefined and upgraded to the generic rank by Juvara-Balş (2002). In the same paper, she divided both new raised genera -Heteroparasitus and Ologamasiphisin two subgenera, key to which is included in our identification key. However, the placement of monotypic subgenus Medioparasitus in the genus Heteroparasitus is uncertain and requires more studies (Juvara-Balş 2002, Witaliński 2008. Recently, there are no subgenera within Holoparasitus but more than half of species is assembled into eight species groups (annulus, caesus, calcaratus, crassisetosus, inornatus, hemispaericus, mallorcae and peraltus) (Witaliński 2006, Witaliński 2017).
Finally, we want to mention one genus with an uncertain status. The monotypic genus Oocarpais Berlese, 1916 was proposed for a single female specimen from India without claws on tarsus I and with hypertrichous dorsal shield. The original description is very brief and no other records of this genus are known since then. Holzmann (1969) examined the type material and she considered that the species belongs to the genus Pergamasus because the peritremes and jugularia are different than in Berlese's description. However, she did not mention the presence or absence of claws on tarsus I. Since she recognized only two genera, Pergamasus and Ologamasus, among mites included in Pergamasine at present, the study of type material is required. It is relevant to point out that the absence of claws on tarsus I is used as an important character to distinguish genera Oocarpais and Pergamasellus from other genera in the family, although, on the other hand, there can be observed an infrageneric variability in this character in the genus Cornigamasus. Female of a species Cornigamasus ocliferius Skorupski and Witaliński, 1997 does not posses ambulacrum with claws and pulvilli (Witaliński 2014). That is not known in any other species of the genus. Another monotypic genus, Nemnichia Oudemans, 1936, was proposed for a species Zercon elegantŭlus Koch, 1839. Oudemans (1936 stated, that the species described by Koch (1839) did not belong to the genus Zercon, based on three long anal setae it is a nymph of the family Parasitidae. This genus was excluded from Parasitidae by Holzmann (1969) but it is still possible to find it placed in Parasitidae, e.g. in the Biology Catalog (Hallan 2005) so we decided to mention it here.
While preparing this paper, we encountered many species described in other genus than they belong to following this generic concept. However, it was not possible to devote all such species and they have to wait to a revision of individual genera. Nevertheless, a diagnosis of some genera is so evident and different from others, that we dare to propose following new combinations only on the basis of original descriptions: Dyneogamasus bisiculus (Tseng, 1995) comb. nov. Dyneogamasus pinatus (Tseng, 1995) comb. nov. Dyneogamasus scirpiculatus (Tseng, 1995)  We propose and give diagnosis of the new genus Coprocarpais n. g. based on the group of closely relative species, which has been already mentioned (Makarova 1996) regarding the need of creating new genus. There is a large number of similar species groups in the Parasitidae, e. g. species groups within the genus Amblygamasus need an examination. As the phylogenetic relationships within Parasitidae remain unresolved, we are aware that here presented concept can be changed. However, this concept follows recent knowledge and we hope, that it will be helpful in work with Parasitidae and in acquisition of knowledge resulting into a large-scale revision of the family Parasitidae.
The Parasitidae comprises a great number of big, conspicuous, cosmopolitan and frequent mite species but work with them is limited by our insufficient knowledge on their taxonomy. A large revision should figure out this not trivial problem. To make a complete revision of the Parasitidae, the following problems have to be solved: • Re-examination of some morphological characters omitted in early descriptions in some species (e.g. adenotaxy, poroidotaxy, idiosomal and leg chaetotaxy, shape of male cheliceral arthrodial membrane, ...) • • When it will arise from the solutions of mentioned problems, than a preparation of the new identification key to genera