Review of the genera Berlesia Canestrini, 1884, and Katydiseius Fain & Lukoschus 1983, the subfamily Katydiseiinae Fain & Lukoschus, 1983, and their family group relationships (Acari: Mesostigmata: Gamasina), with description of three new species parasitic on gryllacridid crickets (Orthoptera) Author Lindquist, Evert E. Author Oconnor, Barry M. 0000-0003-4588-0788 Department of Ecology and Evolutionary Biology, Museum of Zoology, University of Michigan, Ann Arbor, Michigan, 48108, USA. bmoc @ umich. edu, https: // orcid. org / 0000 - 0003 - 4588 - 0788 Author Shaw, Matthew D. 0000-0002-5037-3740 Terrestrial Invertebrates, South Australian Museum, North Tce, Adelaide 5000, South Australia, Australia matthew. shaw @ samuseum. sa. gov. au, https: // orcid. org / 0000 - 0002 - 5037 - 3740 Author Sidorchuk, Ekaterina A. Arthropoda Laboratory, Paleontological Institute, Russian Academy of Sciences, Profsoyuznaya ulitsa 123, 117997 Moscow, Russia. International Complex Research Laboratory for the Study of Climate Change, Land Usage, and Biodiversity, Tyumen State University, 10 Semakova Str., 625003, Tyumen, Russia. text Zootaxa 2020 2020-09-28 4857 1 5 70 journal article 9059 10.11646/zootaxa.4857.1.4 c4a1490d-15c3-446f-8676-e6c80096bcb6 1175-5326 4396474 F0AF75AD-BAE2-4B7D-9CCB-3D9477F350BD Review of Iphiopsididae as a definable family-group The genus Berlesia was first placed in the controversial family Iphiopsidinae (now Iphiopsididae ) by Vitzthum (1943) where it has remained. This placement is unsatisfactory for several reasons. Firstly the Iphiopsididae itself needs examination. Kramer (1886) mentioned the new family as Iphiopsidae (sic), just once, after stating that the available gamasid families were not sufficient to encompass known species, and thereby granting family status to Iphiopsis mirabilis ( Berlese, 1882 ) . Kramer (1886) did not specifically include any other genera in his Iphiopsidae , and his proposal of a family for Iphiopsis was overlooked until noted by Vitzthum (1943) . Berlese (1892 , p.16) provided a diagram of relationships between genera (Genera Mesostigmatum Secundum Ordinem Naturalem Disposita) which clearly showed that Berlesia (along with the ant-associated Neoberlesia Berlese, 1892 ) was considered to be distinctly separated from Iphiopsis, Iphis and genera considered similar at that time. The associated key (p. 37) did not diagnose Berlesia , as the key was explictly restricted to European taxa. However the diagnostic information in that key, separating Iphiopsis based on loss of peritreme, re-appears later where it is then used in a much broader geographic context ( Vitzthum 1943 ), and thereafter for many more genera ( Evans 1955 ). Oudemans (1904) was struck by the similarity of female Berlesia chelicerae to those of Varroa Oudemans, 1904 , and their modifications for parasitism (although which cheliceral digit regresses was noted as dissimilar); however, he rejected a close relationship between these genera because of differences in ventral shields. Vitzthum (1943) modified Kramer’s (1886) monobasic Iphiopsididae , adding Jacobsonia Berlese, 1910b and Berlesia to form a subfamily Iphiopsidinae within Laelaptidae (sic). The only significant character given in sup- port of this subfamilial grouping was the almost absent peritreme (along with a weak attribute of the idiosoma being colourless or brownish). However, this criterion was sometimes clearly subordinated to others, since the genus Dinogamasus Kramer, 1898 , which lack peritremes, was placed in the newly-created Hypoaspidinae, while Neolaelaps Hirst, 1926 , with similarly vestigial peritremes, was placed in the Laelaptinae. Keegan (1950) added Dynatochela , a beetle-associate, to the Iphiopsidinae . Its reduced peritreme was the sole basis for subfamily placement, despite other notable discordant features of Dynatochela (including well-developed claws on all tarsi). Clearly, the concept for any supraspecific grouping must include the type species. Evans (1955) invoked this point at the genus level, noting that the most important characters of Iphiopsis were the reduction of the peritreme and the absence of claws on leg I (the species original description by Berlese, 1882 , indeed states that the first pair of legs are destitute of claws, and his accompanying figure 1, plate 64, clearly shows that leg I lacks claws albeit provided with a well-developed ambulacrum). Potential implications at the subfamily level were avoided by Evans (1955) noting that claws of one or more pairs of legs are usually reduced in Iphiopsidinae (our italics). This qualification allowed Evans to retain Berlesia and Dynatochela , and include Dinogamasus in Iphiopsidinae. However , by considering primarily myriapod hosts, Evans’ discussion did not account for the growing heterogeneity in this taxon. Also, iphiopsidines were stated to be “paraphages of arthropods”, thus overlooking statements by previous authors ( Oudemans 1904 , Vitzthum 1943 ) that Berlesia is parasitic. Evans (1955) makes a strong assertion that Iphiopsidinae “with little doubt arose from” the Hypoaspis -group—if so, Hypoaspidinae is rendered paraphyletic, and this creates another question as to the validity of Iphiopsidinae . The Iphiopsidinae was again defined according to a reduced peritreme for the placement of Dyscinetonyssus hystricosus Moss & Funk, 1965 ; however, this was qualified with acknowledgement that some authorities considered the characters used to define Iphiopsidinae as convergent ( Moss & Funk 1965 ). In the first published cladistic morphological analysis of the Laelapidae, Casanueva (1993) recovered a phylogenetic estimate where all sampled myriapod associates clade together. Thus, the earlier concept of the Iphiopsidinae was recovered in part, but became substantially modified by including Julolaelaps Berlese, 1916 and Scissuralaelaps Womersley, 1945 . The former iphiopsidine Dyscinetonyssus was recovered in a separate clade that included all beetle and cockroach associates, while all hymenopteran associates including Neoberlesia and Dinogamasus were recovered in a greatly expanded Melittiphidinae. Berlesia was not analysed. These are interesting results, which might be intuitively appealing if one believes dermanyssines show a largely inflexible pattern of cophylogeny with their hosts at most levels ranging from species to family. However, the combination of character and taxon sampling does not permit full confidence in the results. For instance, while the loss of hypostomatic seta hp3 has relevance to some myriapod associates, it was used to define a clade with Julolaelaps Berlese, 1916 subtending the myriapod-associated Iphiopsidinae . The sampling entailed a strong assumption that the exclusion of Julolaelaps which retain hp3 (e.g., J. moseri Hunter & Rosario, 1986 ) would not substantially alter the results, including grouping with the similar Hypoaspis as previously hypothesized ( Evans 1955 ). This result and similar ones appear to be reinforced by sampling taxa in a way that creates superficially distinct characters. For instance, the defining synapomorphy reconstructed for most of the Hypoaspidini is loss of podonotal z3 which is a poor, mostly unrepresentative attribute to characterize many taxa intended for this clade including Hypoaspis sensu stricto . Podonotal z3 is an ontogenetically weak deutonymphal seta whose suppression is subject to much homoplasy ( Lindquist & Evans 1965 ), so it would be surprising if its loss could define any sizeable clade. Also the five putative synapomorphies that define a clade equivalent to Iphiopsidinae (i.e. myriapod associates other than Julolaelaps and Scissuralaelaps ) plus an additional four other putative synapomorphies that largely define a yet more inclusive clade, are all regressive leg characters that are also found in hypotrichous melittiphines, notably Myrmozercon brevipes Berlese, 1902 (with this genus actually not included in the analysis). Note that the leg chaetotaxy of Berlesia species is relatively holotrichous compared to most of the myriapod-associated Iphiopsididae sampled by Casanueva (1993) and bears only two of the nine reductive characters reconstructed by Casaneuva (1993; Fig 8 , Stem 3) as putative synapomorphies for the relevant group containing the type species (Group VII Iphiopsini). However, an alleged sister group to the main iphiopsidid clade proposed by Casanueva (1993) is relevant to consider with respect to Berlesia as it contains potential relatives but also reveals possible pitfalls in uncovering these with morphological analysis. In this clade the mygalomorph-associated genus Ljunghia Oudemans, 1932 and the crab-associated genus Cyclothorax von Frauenfeld, 1868 are recovered together. Casaneuva (1993) noted that the synapomorphies linking Ljunghia to iphiopsidids were weak with low confidence in this expansion proposed for the Iphiopsididae . Ljunghia has similarities with Berlesia which may or may not indicate relationship. Deserving attention amongst these are: the usually pos- teriorly-positioned paranal setae, the shared absence of gv3 on the anal shield, and the stout movable digit. For the latter, however, it is important to note that the movable digit of Berlesia appears rotated laterally, with its teeth flared away from the fixed digit, whose base is swollen (somewhat as in Ixodida ). This contrasts with Ljunghia where the teeth on the movable digit still oppose what remains of the fixed digit, and the fixed digit base is not swollen. Thus Ljunghia movable digits lack some special features found in Berlesia that would make it more difficult to argue for similarity due to relationship under a structural criterion ( Riedl 1978 , p 34). The above highlights the rampant homoplasy due to convergence and parallelism that plagues all potential candidates for Iphiopsididae . When both Moraza & Kazemi (2012) and Seeman & Alberti (2015) tried to apply characters proposed by Casaneuva (1993) to define/diagnose Iphiopsididae , they found that these characters had widespread occurrences elsewhere in the Laelapidae , such that both sets of authors rejected the family-ranking of Iphiopsididae . Nemati et al. (2015) also found the inclusion of Julolaelaps to be problematic. Characters such as hypotrichy, reduced peritremes and reduced claws found in the “ Iphiopsididae ”, are known to be correlated with a wide variety of intimate symbioses in various Dermanyssoidea and are regressive in nature (as realized 65 years ago by Evans 1955 ). Iphiopsididae , at whatever rank, will remain an unstable concept, unless future molecular analyses, backed by unexpected morphological attributes indicate otherwise. Currently, it can only be considered satisfactory for taxa that are highly similar morphologically to the type genus Iphiopsis . Better knowledge of the type species would be useful here ( Halliday & Juvara-Bals 2016 ), insofar as clarifying the core taxa related to Iphiopsis . Other piecemeal renovations to Iphiopsididae will not help, unless they include new diagnostic characters that reliably link included iphiopsidid taxa. However, new suprageneric groupings of taxa that are clearly not related to Iphiopsis , when these are supported by convincing synapomorphies, will advance overall systematic perspectives partly because they will reduce the number of taxa to be considered for placement in Iphiopsididae .