Revision of the genus Pseudobryobia McGregor , 1950 ( Acari , Tetranychidae )

The species of the tetranychid mite genus Pseudobryobia are revised mainly based on data published in the literature. Following a survey of the classification history of the genus, emphasizing the changes that occurred in the morphological characters used in its definition, we briefly discuss their relative importance for the genus delimitation. As a result, we provide amended diagnoses for the genera Pseudobryobia and Bryobia and transfer 7 species from Pseudobryobia to Bryobia. Species belonging to the genus Pseudobryobia bear two setae on coxisternal plate II, one pair of tenent hairs on all the empodia and are distributed in the Nearctic biogeographic zone. Conversely, some species previously assigned to the genus Pseudobryobia bear one seta on coxisternal plate II, two rows of tenent hairs on empodia II-IV, and are recorded in the Palearctic and Afrotropical biogeographic zones. These species are transferred to the genus Bryobia. The following nomenclature changes resulting from the revision are proposed: B. (Bryobiopsis) abbatielloi new combination, B. (B.) anacantha reinstated combination, B. (Allobia) bucharica reinstated combination, B. (B.) eurotiae reinstated combination, B. (A.) japonica reinstated combination, B. (B.) neoephedrae new combination, B. (A.) nikitensis reinstated combination. We recognise six subgenera for Bryobia, as previously considered by other authors, and also treat Nuciforaella Vacante (1983) as a junior synonym of Bryobia (Allobia) Livshits and Mitrofanov, 1971. A key to the species of Pseudobryobia is also provided.


Introduction
Among the Bryobiini, the genus Pseudobryobia is the second largest genus in terms of species number, but is far behind the genus Bryobia, with 18 and 136 species, respectively Dorkeld, 2006-2017). This genus was established by McGregor (1950) mainly to accommodate some morphological characters of a new species of bryobiine mite collected in Mexico, Pseudobryobia bakeri. Since its formation, this genus was alternatively synonymized with the genus Bryobia, considered as a subgenus of Bryobia, and reinstated as a full genus; its diagnosis was also amended several times (e.g. Wainstein (1960); Tuttle and Baker (1968); Livshits and Mitrofanov (1972); Baker and Tuttle (1972); Meyer (1987)). Wainstein (1960) and Livshits and Mitrofanov (1971), who were actively involved in the taxonomical status of Pseudobryobia, both mentioned that the separation between the genera Pseudobryobia and Bryobia needed further research. Confusion regarding the definition of these genera was confirmed recently when B. longisetis Reck, 1947 was moved from the genus Pseudobryobia to the genus Bryobia (Auger and Migeon, 2014) when, according to Vacante (1983) and to Barbar (2018), the assignment of P. nikitensis to Pseudobryobia may be inappropriate because of an unusual location of the fourth pair of dorsocentral setae. More recently, Zeity and Srinivasa (2019) went further by proposing the reinstatement of the genus Nuciforaella, and moving P. nikitensis (Livshits & Mitrofanov, 1969) and P. japonica (Ehara & Yamada, 1968) from the genus Pseudobryobia to Nuciforaella. Based on the morphological similarities of current designations, we think that the genus assignation of additional species of Pseudobryobia should be re-examined and probably could be assigned to the genus Bryobia. However, instead of changing a few species, we believe that a full revision of the genus Pseudobryobia is desperately needed to clarify the taxonomy of this group.
To this end, we conduct a thorough historical taxonomic review of the genus Pseudobryobia, paying particular attention to the characters used for its delineation. Our findings result in a significant revision of the genus Pseudobryobia.

Material and methods
Our study is based on a literature review dealing with the genus Pseudobryobia. The literature concerning species that were assigned to this genus at least once and works allowing a better delineation of the genus were also accessed. In some cases, particularly when important data from the literature were unclear or missing, we had the opportunity to access some type specimens or specimens belonging to the species required. This occurred several times and concerned P. abbatielloi Smiley & Baker, 1995, P. canescens Baker & Tuttle, 1972, P. curiosa (Summers, 1953, P. drummondi (Ewing, 1926), P. ephedrae (Tuttle & Baker, 1968), P. filifoliae (Tuttle & Baker, 1968), P. japonica (Ehara & Yamada, 1968), P. knowltoni Tuttle & Baker, 1976 and P. namae (Tuttle & Baker, 1964). Type specimens of P. japonica (National Museum of Nature and Science, Tsukuba : NSMT Ac-1303637, paratype) and specimens of P. abbatielloi (3 and 11 slides of mites collected in Yemen and Saudi Arabia, respectively) were examined by Dr Tetsuo Gotoh (Ibaraki University, Japan) and by Dr Fahad Jaber Alatawi (King Saud University, Saudi Arabia), respectively. Other type specimens were examined by the senior author in Dr Baker's collection (USDA, Systematic Entomology Laboratory. National Museum of Natural History, Smithsonian Institution, Maryland Beltsville, Washington DC.) Results and discussion Historical taxonomic review of Pseudobryobia In 1950, McGregor established the genus Pseudobryobia with P. bakeri as type species and he also moved Petrobia drummondi to this genus. Despite a detailed diagnosis of Pseudobryobia, McGregor did not analyse relatedness to other closely related genera. In McGregor's description, the genus is defined as follows: forelegs barely exceeding body, other shorter; peritreme enlarging distally somewhat septate or lobed, not protruding externally; empodia with a pair of tenent hairs; dorsum with striated integument bearing short setae. Pritchard and Baker (1955) criticized the McGregor's genus description as insufficient for generic distinction, focusing only on the peritreme as its defining character state. They regarded this feature as intraspecifically variable, possibly due to slide mounting variation. For this reason, Pritchard and Baker disagreed with McGregor and argued that Pseudobryobia is a synonym of Bryobia. Wainstein (1960) considered that there were too many differences between typical species belonging to the genus Bryobia and both P. bakeri and P. drummondi. Nevertheless, as among the genus Bryobia some species are morphologically "intermediate" between the two genera, he demoted the genus Pseudobryobia to subgeneric rank among the genus Bryobia. Species belonging to this subgenus had the following characters: peritremal enlargement not obvious, sometimes weakly developed; the empodium of leg I padlike with one pair of tenent hairs, those on legs II-IV with one pair or two rows of tenent hairs and the prodorsum without anterior lobes or with these lobes greatly reduced. Based on these criteria, in addition to P. bakeri and P. drummondi, Wainstein also placed in this subgenus Bryobia curiosa Summers, 1953 andB. longisetis Reck, 1947. We note that Wainstein's (1960) "anterior lobes" are equivalent to the prodorsal or propodosomal lobes that bear setae v 1 and v 2 in some bryobiine mites, and are not homologous with the "narrow semihyaline plate […] devoid of setae" noted by McGregor (1950) for P. bakeri. Tuttle and Baker (1968) continued to treat Pseudobryobia as a synonym of Bryobia. They proposed an amended diagnosis of the genus Bryobia and noted variability of some key characters that were previously used to separate this genus and Pseudobryobia: i.e. the shape of peritremal ends varying from simple to anastomosed and the presence or absence of propodosomal projections. In line with this, they described and placed among the genus Bryobia two new species of Bryobia that bear no propodosomal lobes: B. ephedrae and B. filifoliae. Livshits and Mitrofanov (1971) were the first to use a new morphological character, the coxal setal count of leg II, to distinguish between the genera Pseudobryobia (two setae on coxa II) and Bryobia (one seta on coxa II). This was made possible thanks to a correspondence with E. W. Baker who informed them that all the American species previously placed or described among the genus (or subgenus) Pseudobryobia (P. drummondi, P. bakeri, P. curiosa, P. agropyra (Morgan, 1960), P. ephedrae, P. namae, and P. filifoliae) have two setae on coxa II. Thus, to accommodate species without or with poorly developed prodorsal lobes but having one seta on coxa II, Livshits and Mitrofanov (1971) erected the new subgenus Allobia among the genus Bryobia into which they transferred three species: P. japonica, P. longisetis and P. nikitensis. The diagnosis of Allobia is quite close to that of Pseudobryobia but differs by the number of setae on coxa II and the arrangement of dorsal setae f 1 and f 2 : i.e., coxal formula 2-1-1-1; prodosomal lobes absent or poorly developed; distance between setae f 1 greater than distance between setae f 2 with setae f 1 and f 2 well separated, positioned laterally.
As a result, Livshits and Mitrofanov (1972) and Baker and Tuttle (1972) independently felt that Pseudobryobia should have a full generic status and proposed to reinstate it. Both contributions provided new diagnoses of the genus in which the number of coxal setal count was raised: in specimens belonging to the genus Pseudobryobia the coxal formula is 2-2-1-1 versus 2-1-1-1 in the genus Bryobia. In addition, Baker and Tuttle (1972) re-emphasized the positions of dorsocentral setae f 1 : in Pseudobryobia, the dorsocentral f 1 setae are located in the normal longitudinal dorsal position and are not marginal.
By the same logic, when Mitrofanov (1973) described a new bryobiine mite (Pseudobryobia eurotiae (Mitrofanov, 1973)) meeting the criteria of Pseudobryobia, excepting the coxal formula, he created the new subgenus Bryobiopsis defined as follows: outer lobes minute, in the form as small tubercles and not separated from inner lobes by a deep incision; setae c 3 in line with c 1 and c 2 ; setae f 1 well separated from f 2 and in central position (not lateral), coxal formula 2-1-1- 1. Smith Meyer (1974) decided to follow Baker's opinion and mentioned that Pseudobryobia species can be separated from Bryobia by their coxal formulae. She therefore defined Pseudobryobia as species without propodosomal lobes, a coxal setal formula of 2-2-1-1, and also noted the position of setae f 1 . She defined the latter feature as setae f 1 either closer together than the first pair of setae c 1 , or in line with the latter, unlike the genus Bryobia in which setae f 1 are further apart than setae c 1 and located more laterally.  provided an updated diagnosis of the genus Pseudobryobia that is a composite of that previously proposed by Baker and Tuttle (1972) and by Meyer (1974): in species belonging to this genus, in addition to the absence of prodorsal lobes and to a coxal setal count of 2-2-1-1, setae f 1 are located in normal position (i. e. not laterally, more or less in line with the other dorsocentral setae). In the same year, Tuttle and Baker (1976) described a new bryobiine mite species, P. knowltoni but, surprisingly, they did not provide any information about its coxal setal count. Vacante (1983) noted a discrepancy in the assignment of P. nikitensis, which has no prodorsal lobes but bears only one seta on coxa II and has setae f 1 in a lateral position. He considered that this species should belong to a morphologically "intermediate" genus between the genera Bryobia and Pseudobryobia. Thus, to accommodate the characters of this species, he created the monospecific genus Nuciforaella with the type species N. nikitensis. It must be noted that until recently this new genus was not recognized in later acarological works like Smith Meyer (1987) and Bolland et al. (1998).
In the subsequent contributions dealing with the genus Pseudobryobia by Mitrofanov et al. (1987), Meyer (1987) and Baker and Tuttle (1994), these authors continued to emphasize the absence of a prodorsal projection, the coxal formula and the location of setae f 1 in separating this genus from Bryobia. Despite the interest in these morphological traits, when Baker and Tuttle (1994) described two new species, P. antennaria Tuttle, 1994 andP. konoi Baker &Tuttle, 1994, they did not provide their coxal formulae. In 1987, Mitrofanov et al. discovered an additional character for distinguishing the two genera: duplex setae on tarsus III are associated in Bryobia versus dissociated in Pseudobryobia.
The last contribution dealing with Pseudobryobia by Smiley and Baker (1995) generated confusion. Indeed, in their diagnoses of the genera Pseudobryobia and Bryobia, they reversed the coxal formula between the two genera so that the coxal formula 2-2-1-1 was assigned to the genus Bryobia and 2-1-1-1 to Pseudobryobia. This mistake has not been without consequence because in the wake of the Smiley and Baker's new descriptions, they also synonymized the subgenus Bryobiopsis (see Mitrofanov (1973) above) with the genus Pseudobryobia. This makes sense with the flawed new diagnoses proposed as until this contribution the subgenus Bryobiopsis only differed from Pseudobryobia by the number of setae on the coxa II (1 seta present in species belonging to the subgenus Bryobiopsis). This supports the idea that their mistake in the coxal formulae may be not simply a typing error.
In the same contribution, they also described a new species, P. abbatielloi, but as they did not provide its setal formula, one has to wonder how many setae are present on its coxa II. Following their reasoning, the synonymy of the sub-genus Bryobiopsis with Pseudobryobia, we may anticipate that P. abbatielloi bears one seta on its coxa II.
Following this contribution several species were moved to the genus Pseudobryobia: i) Ehara (1996) transferred B. japonica to the genus Pseudobryobia because of the absence of prominent prodorsal lobes over the gnathosoma (Gotoh, personal communication); ii) Bolland et al. (1998) moved 4 species from the genus Bryobia to Pseudobryobia: B. anacantha (Strunkova & Mitrofanov, 1983), B. agropyra, B. bucharica (Strunkova & Mitrofanov, 1983) and B. nikitensis. As these four species do not bear the same number of setae on coxa II, the new combinations provided by Bolland et al. (1998) make sense if we accept that these authors did not take into account the coxal setal formula to distinguish between the genera Bryobia and Pseudobryobia. Moreover, in some of these species, the dorsocentral setae are arranged in a longitudinal row whereas in others setae f 1 are near the body margin (e.g. P. agropyra, P. drummondi, P. neoephedrae Bolland et al., 1998 and P. bucharica, P. japonica, P. nikitensis, respectively). Therefore, the dorsal setal pattern is overlooked as a possible or important characteristic to delineate taxa, and a characteristic not used by these authors when creating new taxonomic assignments. The only character shared by all the species Bolland et al. (1998) merged into taxa into the genus Pseudobryobia is the absence of propodosomal lobes over the gnathosoma. Auger and Migeon (2014) transferred P. longisetis (sensu Bolland et al., 1998) back to the genus Bryobia using key morphological characters defined by Livshits and Mitrofanov (1972), by Baker and Tuttle (1972) and by Meyer (1974) to distinguish between the genera Bryobia and Pseudobryobia.
The last contribution dealing with Pseudobryobia reinstated the genus Nuciforaella and transferred P. nikitensis and P. japonica (sensu Bolland et al., 1998) to this genus (Zeity and Srinivasa, 2019). Their proposal was based on Vacante's opinion which considered that the morphological characteristics of the genus Nuciforaella represent an intermediate state between the genera Bryobia and Pseudobryobia (Vacante, 1983). Table 1 presents the potentially informative morphological character states that may separate Bryobia from the species of Pseudobryobia as currently defined by Bolland et al. (1998). As the setal count on coxa II was agreed by most of the acarologists to reliably distinguish the genera Pseudobryobia and Bryobia ( (Livshits and Mitrofanov, 1971); Livshits and Mitrofanov (1972); Baker and Tuttle (1972); Meyer (1974); Mitrofanov et al. (1987); Meyer (1987); Baker and Tuttle (1994)), we organize the taxa according to this morphological trait (displayed in the 3rd column in the Table I): species bearing two setae on coxa II are gathered and provided first. Arranged thus, some character states are shown to be highly variable, especially the shape of the distal end of the peritreme and the shape of the dorsohysterosomal setae. However, the location of setae f 1 , the number of tenent hairs on empodium I-IV and the state of the duplex setae on tarsi III and IV do not vary among the species group bearing 2 setae on coxa II. Some of these traits are shared by species with one seta on coxa II, but two morphological criteria are congruent and distinguish two distinct species groups among this genus: a species group with 2 setae on coxa II, having empodia II-IV with a pair of tenent hairs and exclusively Nearctic in geographical distribution; and a group with one seta on coxa II, bearing empodia II-IV with two rows of tenent hairs and having Palearctic and Afrotropical distributions.

Morphological characters and genus delineation
The convergence of morphological and geographical criteria demonstrates the importance of the coxal setal formula in the taxonomy of spider mites, and specifically genus delimitation. According to Lindquist (1985), the primitive coxal formula on adult spider mites is 2-2-1-1 but there is a common derivative loss of setae on coxisternal plate II which has only one seta in various genera of Bryobiinae. Among this subfamily, as early as the 1960's, Bagdasarian (1957) and Reck (1959) were the first to reference the coxisternal setal count for genus delimitation and mentioned that 2 setae on coxa I and 1 on coxae II-IV are present in the genus Bryobia. Later, several authors included the coxal formula in the diagnoses of the genera Bryobia or Pseudobryobia (e.g. Baker and Tuttle (1972); Meyer (1974); Vacante (1983); Meyer (1987); Baker and Tuttle (1994)) but also in other genera, suggesting that this character may be fixed among some of them. Unfortunately, this subject is still poorly studied and therefore poorly documented. Although information is lacking on the coxal formula of each known species (data are not always provided in the original descriptions and redescriptions), if we exclude the Pseudobryobia species, the number of setae on coxa II generally does not vary between species belonging to the same genus within the Bryobiinae. However, there are some rare exceptions: i) in the genus Neopetrobia the usual coxal setal formula is 2-2-1-1 whereas it is 2-1- 1-1 in N. (N.) tarkaensis; ii) in the genus Paraplonobia the basic coxal formula is 2-2-1-1 but, in the monotypic subgenus Brachynychus it is 4-3-2-2; iii) Smith Meyer (1987) reports that it could vary between 2-3-1-1 and 2-2-1-1 in some species belonging to the subgenera Anaplonobia. However, these cases are exceptions and therefore not fully comparable with what is observed in the genus Pseudobryobia.
To our knowledge, coxal setal variations, comparable in number to that observed in Pseudobryobia, also occur in the tetranychine genus Eutetranychus. Species of Eutetranychus can be separated in two groups using their setal count on coxa II: one group with 1 setae (16 species) and the other with 2 (11 species) (Kamran et al., 2018). Nevertheless, in the genus Eutetranychus, geographical distribution of the species does not appear related to their coxal setal formula. For example, among the two newly described species collected in Saudi Arabia, one bears one seta and the other two setae on coxa II (Kamran et al., 2018). Finally, no clear morphological differences between the species of the two subgroups have yet been reported.
However, in our new concept of Pseudobryobia the differences in coxal setal formula are related to another discriminant morphological trait: the morphology of the empodia of Table 1 Variations in some morphological key characters (used in the literature) of mites belonging to the genus Pseudobryobia (sensu Bolland et al., 1998) with data on their geographic distribution and their possible genus assignation (as the absence of prodorsal projection over the gnathosoma is shared by all the species, this character is not included in the table  Livshits & Mitrofanov (1971) and Mitrofanov (1973) P. agropyra Coxal formula 2-1-1-1; prodosomal lobes absent or poorly developed; if present, outer and inner lobes not separated by deep incision; distance between f 1 setae members larger than distance between f 2 members; setae f 1 and f 2 well separated, lateral.
legs II-IV. And, these characters are related to their geographical distribution, which suggests allopatric differentiation on a broad biogeographic scale (e.g. Turelli et al. (2001), Barraclough and Vogler (2000)) and reflects two distinct lineages. We therefore propose a revision of the genus Pseudobryobia based on the coxal formula, the single pair of tenent hairs on empodia II-IV, and the geographical distribution of Bryobia and Pseudobryobia.
This revision particularly recalls the works by Livshits and Mitrofanov (1971;Mitrofanov (1973). In addition to their definition of the genus Pseudobryobia we think that their proposal of six subgenera of Bryobia (Bryobia, Lyobia, Periplanobia, Allobia, Bryobiopsis and Eharobia) is robust and, in the present paper, this is particularly the case with the subgenera Allobia and Bryobiopsis. We share the opinion of Zeity and Srinivasa (2019) that the species P. nikitensis and P. japonica belong the same taxonomical unit. However, B. tadjikistanica Livshits & Mitrofanov, 1969 and, according to our data, P. bucharica, should probably also be added to this taxon. In addition, as the morphological characteristics of species placed in the genus Nuciforaella correspond to those of the subgenus B. (Allobia), we do not share the view of Vacante (1983) and Zeity and Srinivasa (2019) and we consider the genus Nuciforaella as a part of B. (Allobia).