Collohmannia, Sellnick, 1922

Norton, R. A. & Sidorchuk, E. A., 2014, Collohmannia Johnstoni N. Sp. (Acari, Oribatida) From West Virginia (U. S. A.), Including Description Of Ontogeny, Setal Variation, Notes On Biology And Systematics Of Collohmanniidae, Acarologia 54 (3), pp. 271-334 : 317-321

Relationships of Collohmannia

Ideas about the phylogenetic relatives of Collohmannia – most of which are implied by classifications – have varied, but have focused on dichoid families of middle-derivative oribatid mites that were eventually grouped by Grandjean (1969) as the unranked taxon Mixonomata . The latter was ranked as an infraorder by Schatz et al. (2011), though it is clearly paraphyletic ( Norton 1998; Schaefer et al. 2010; Pachl et al. 2012). Sellnick (1922) assigned Collohmannia to no family when it was first proposed, but he noted resemblance to Lohmanniidae ; the genus name also implies a connection (col - Latin prefix meaning together), though he did not explain the etymology. Vitzthum (1931) seems to have made the first formal family assignment, by including Collohmannia (and Perlohmannia ) in a broad concept of Lohmanniidae , and this was followed in Radford’s (1950; misspelled as ’Callohmannia’) checklist of mite genera. While Grandjean’s original sense of Mixonomata included Lohmanniidae , this family has since been transferred to Enarthronota, specifically to Hypochthonioidea , based on both morphological ( Norton 2010) and molecular ( Pachl et al. 2012) characters. Baker and Wharton (1952) tentatively included Collohmannia instead in the mixonomatan family Eulohmanniidae , but this seems to have been subsequently ignored. It appears that neither these latter authors nor Vitzthum studied specimens, nor were their actions supported by discussion. When Grandjean (1958a) proposed Collohmanniidae , he compared it mainly with Perlohmanniidae but not as a close relative; soon after (but without referring to his paper), Bulanova-Zachvatkina (1960) included Collohmannia in Perlohmanniidae . Again, it seems that neither author had studied specimens at that time. The idea that Collohmanniidae and Perlohmanniidae are closely related survives in some recent catalogues ( Subías 2004, Shtanchaeva and Subías 2010, Subías et al. 2012), in which the two families comprise the Perlohmannioidea.

A different idea focused on the similarity of Collohmanniidae to ptychoid mixonomatans. This was first suggested by Sellnick (1918), who thought the fossil species Embolacarus pergratus Sellnick was ptychoid, but – surprisingly – in 1922 he did not suggest any relationship with Collohmannia (see Norton 2006). ŠtorkAEn (1925) specifically compared Collohmannia gigantea to ’phthiracarid’ mites, even though he saw it alive and knew it was not ptychoid. He discussed C. gigantea at the end of the section on ’Phthiracarinae’ (sensu lato, essentially = modern concept of Ptyctima), which could be interpreted as his intention to include Collohmannia in the group. However, he made no clear statement in this regard and also discussed similarities with Lesseria (now = Epilohmannia ; Epilohmanniidae ). ŠtorkAEn’s (1925) uncertainty probably is captured best by the supposed junior synonym that he claimed to have published in 1923 – Phthiracaroides incertus . He gave no reference and we can find no paper proposing this name; it is not the ŠtorkAEn (1923) paper about mites in mole’s nests, as is sometimes suggested. Therefore, we consider it unavailable: a nomen nudum. In describing C. nova (later synonymized with C. gigantea ), Sellnick (1932) also noted similarities with ’Phthiracarus’ (used in a very broad sense) but considered Collohmannia a genus of Lohmanniidae . Looking beyond the similarity in facies between Collohmannia and Ptyctima, Grandjean (1966, 1969) examined specific characters, particularly five that in cladistic terminology could be considered synapomorphies supportive of a close relationship between Collohmannia and Ptyctima. These are summarized below, followed by five others.

1. Coxisternum — This is unusually structured in Collohmannia ; it is independent of surrounding sclerites, divided by a cross of articulating furrows (scissures) into four plates (each formed from plates of two fused demi-epimeres), and the epimeres narrow from front to back, such that legs IV are unusually close together. It is the same in all Ptyctima and is necessary for folding of the coxisternum during retraction of the podosoma ( Sanders and Norton 2004).

2. Aggenital and adanal plates — On either side these plates are fused laterally, leaving only a medial incision to separate the plates. Sellnick (1932) illustrated these plates as independent, but this may have been an artifact of dissection. The partial fusion exists in all specimens of Collohmannia we examined, and Grandjean (1969) considered this the most common situation in C. gigantea , while allowing that plates may be separate in some individuals. Among Ptyctima fusion of these plates characterizes most Euphthiracaroidea, some of which retain a plesiomorphic incision marking the plate boundary. No such fusion exists in the highly derived euphthiracaroid subfamily Temburongiinae ( Synichotritiidae ; Norton and Lions 1992), or in Phthiracaroidea.

3. Sagittal apodeme — This unpaired apodeme (’nervure sagittal’) in the posterodorsal part of the aspis is unknown outside Collohmannia, Phthiracaroidea and most Euphthiracaroidea. Of the few euphthiracaroids that lack it as adults, their juveniles (if known) have a distinct sagittal apodeme ( Grandjean 1969 and cited papers).

4. Bothridium — In Collohmannia , a large section of the bothridium has numerous sausage-shaped, contiguous, smooth-walled locules. Grandjean noted that a bothridium of similar form was known only in Ptyctima, but some large Epilohmanniidae , e.g. Epilohmannia praetritia Berlese and some undescribed species, also have such structures (R.A.N., unpublished).

5. Famulus — The spine-like shape seen in Collohmannia is common, but a rugose surface is known only in Collohmannia and Ptyctima. Outside the infraorders Enarthronota and Palaeosomata, it is rare for any oribatid mite to have a famulus with discernable surface structure, but Nehypochthoniidae , a family proposed as a possible near-outgroup of Ptyctima, has a distinctly annulate famulus ( Norton and Metz 1980).

6. Preanal apodeme — This narrow, blade-like apodeme extends anteriorly from the anal valves of Collohmannia to lie in the ventral midline and provide the origin for some of the genital adductor muscles and part of the extensive set of lateral compressor muscles that insert at the lateral edge of each adanal plate (see also Norton 2006). Grandjean (1969) noted this similarity, but did not include it as support of his idea. The blade-like preanal apodeme is unique to Collohmannia and Euphthiracaroidea. In the latter group this apodeme seems to be a ’kingpost’ for a laterally acting compressor system that controls hemolymph pressure ( Sanders and Norton 2004; Schmelzle et al. 2009). As with character 2 (above), an apodeme of this form is not present in Phthiracaroidea, and seems absent from temburongiine Synichotritiidae as well.

7. Lateral rib — Juvenile Collohmannia and Euphthiracaroidea possess a paired lateral rib (’nervure laterale’), which borders the aspis and ends distally in a condyle that articulates with the subcapitulum. It is present in all instars of Collohmannia , but is most conspicuous in juveniles, where its pigmentation and thickness contrast with the paler surrounding cuticle of the aspis. TravØ (1975) viewed this similarity as support for a relationship between the groups, but a similar rib exists also in endeostigmatid mites and Palaeosomata ( Grandjean 1954). We know of no examples from Enarthronota or Parhyposomata, nor do we know examples in Mixonomata , outside of Collohmannia and Euphthiracaroidea (see Remark 10). Considering its taxonomic distribution (absent from Parhyposomata and other Mixonomata ), we think the rib re-evolved in a common ancestor of Collohmannia and Euphthiracaroidea, but such homoplasies are weak support for a phylogenetic hypothesis.

8. Transverse hysterosomal lines in juveniles — In both Collohmannia and Ptyctima there are transverse lines (linear grooves) in the cuticle of the gastronotic region of juveniles (TravØ 1975). These are in addition to the indistinct line as of Grandjean (1966), which simply marks the posterior boundary of flexible, striated cuticle extending from the sejugal furrow (we presume the notation as relates to the aesthenic zone of Grandjean 1954). Two lines are present in Collohmannia juveniles, designated r2 and r3 by Grandjean (1966); he did not discuss the notations, but they seem to relate to his hypothetical demarcation of primitive segments (ar2, ar3 of Grandjean 1947b). Line r2 would presumably mark the boundary between segments D and E, and r3 that between E and F. It is a reasonable idea, considering the apparent segmentation of the endeostigmatid genus Terpnacarus , thought to represent a close outgroup of Oribatida ( Grandjean 1939; Norton et al. 1993). Mixonomata are quite phylogenetically distant from endeostigmatid mites, and many intervening taxa lack such transverse lines. Disregarding the complicated cases of transverse scissures between sclerites in Enarthronota ( Norton 2001), simple transverse lines can be found in some soft-bodied members of the most primitive infraorders, e.g. Pediculochelidae (Enarthronota) , Aphelacaridae (Palaeosomata) and Parhypochthoniidae (Parhyposomata) . Transverse lines that occur in juveniles of a few species of the derived infraorder Brachypylina (e.g. Pirnodus detectidens Grandjean, 1956 ) cannot reasonably be attributed to primitive segmentation boundaries.

Regardless of their origin, the lines of Collohmannia differ from hypothetical boundaries in one important respect: seta d 2 lies between r2 and r3, whereas only row e should be present if the lines marked segments. This is exactly the situation in those Ptyctima with juveniles that possess line r2: the phthiracaroid genera Phthiracarus and Steganacarus ( Grandjean 1950; Webb 1977) and the euphthiracaroid genus Paratritia (TravØ 1975). Besides Collohmannia and Ptyctima, we know of no taxon with such an arrangement of lines and setae. In this case, d 2 could have migrated posteriorly across the primitive boundary or (perhaps less likely) the lines (at least r2) may have a different origin. In either instance the similarity is synapomorphic. Its significance is somewhat compromised by our poor general knowledge of juvenile Ptyctima, some of which have apparently lost r2 and retain only r3 or else have neither line (e.g., Walker 1965; Schubart 1967; TravØ 1975; Ermilov 2011b).

9. Plicature zone — There is an extensive infolded plicature zone (pz.1) between notogaster and adanal plates in Collohmannia , a trait also of all Euphthiracaroidea other than Temburongiinae ( Norton and Lions 1992). In Collohmannia the cuticle of this zone is of the transitional type (imbricate and porose, but unpigmented) and anteriorly it bears lyrifissure ips. In Euphthiracaroidea the plicature zone is moderately hardened, but ips is on the notogaster, indicating that the zones are not precisely homologous and making the synapomorphy somewhat equivocal. Members of Perlohmanniidae have a placement of ips similar to that of Collohmannia ( Grandjean 1958a) but Grandjean (1969) argued effectively against the close relationship of these two groups.

10. Posterior notogastral sinus — Posterior to the valves, the plicature zone of Collohmannia diffuses to become a large expanse of pale transitional cuticle, best seen in posterior view ( Fig. 10B ). The sinus may allow the notogaster to flex during contraction of the lateral compressor muscles, performing the same function as the terminal sinus or terminal fissure in Euphthiracaroidea ( Märkel 1964; Sanders and Norton 2004). Its form is consistent with Märkel’s observation that a sinus, rather than a fissure, is typical of those euphthiracaroid taxa having a broader notogaster.

Collectively the ten morphological traits discussed above seem convincing for a close relationship with Euphthiracaroidea, even that of sistergroup. Half (#2, 6, 7, 9, 10) are not shared by Phthiracaridae or (except for #7, since juveniles are unknown) by temburongiine Synichotritiidae , but these differences can be explained as being masked by further derivations. Both groups appear to have been derived within Euphthiracaroidea as the latter is usually conceived. Morphological data support the case for Temburongiinae ( Norton and Lions 1992) and molecular data support that for Phthiracaridae ( Pachl et al. 2012) . At least for Phthiracaridae , all but #7 appear to relate to the evolution in that family of a hemolymph-pressure control system of plates and muscles that act dorsoventrally, rather than laterally as in most euphthiracaroids ( Schmelzle et al. 2009).

Recently an interesting similarity between Collohmannia and Phthiracaridae has emerged that is not shared by any euphthiracaroid mite. The New Zealand species Austrophthiracarus notoporosus Liu and Zhang, 2014 appears to possess numerous notogastral porose areas, although they have no obvious association with setae. It would be unreasonable to list this as a synapomorphy, but it does emphasize that the distribution of notogastral dermal glands is rather mosaic ( Norton and Alberti 1997).

While molecular data help clarify the origin of Phthiracaridae , they paint an equivocal phylogenetic picture regarding Collohmannia . Analyses of the 18S ribosomal RNA gene by neighbor-joining ( Lee et al. 2006) and maximum-parsimony ( Dabert et al. 2010) algorithms produced trees consistent with C. johnstoni (as Collohmannia sp. ) and Nehypochthonius porosus Norton and Metz (as predicted by Norton and Metz 1980) being the closest outgroups of Ptyctima, but not with strong statistical support. The maximum-likelihood tree of Lee et al. (2006) had C. johnstoni and Steganacarus magnus (Nicolet) as sister-taxa, but N. porosus was distant. By contrast, maximum-likelihood and Bayesian analyses of Dabert et al. (2010) included C. johnstoni in the midst of Nothrina ( Desmonomata ), as sister-taxon to Nothrus sp. (Nothridae) . Statistical support for this relationship was weak, but the similar result of a Bayesian analysis by Pachl et al. (2012) – a grouping of C. johnstoni (as Collohmannia sp. ) with Nothrus silvestris Nicolet – had strong support.

We can identify no morphological support for this latter relationship – no synapomorphies of Collohmannia and Nothrus – but Collohmannia does have a chelicera with apomorphic traits of the infraorder Desmonomata as a whole (sensu lato, Schatz et al. 2011). One is the encroachment of the cheliceral sheath onto the face of the principal body, resulting in an ’inserted’ chelicera ( Norton 1998). Ptyctima and other mixonomatans have the plesiomorphic state, in which the sheath attaches proximally. Another cheliceral trait is the presence of a large, well-defined Trägårdh’s organ, which also characterizes Desmonomata . There is some evidence that a small, inconspicuous and delicate Trägårdh’s organ is present in Ptyctima but its evolutionary polarity relative to the large structures noted above is equivocal ( Lions and Norton 1998 and cited papers). On the dorsal face of the rutellum, Collohmannia species lack the ciliary comb (= brush) common to nearly all desmonomatans outside Astigmata, but they do possess adjacent hyaline ’spines’, usually two, in this position. Grandjean (1966) considered the spines a special form of the comb, but their form is quite different from that of desmonomatan cilia. We know only one family currently classified in Mixonomata that does have a well-formed rutellar brush – Eulohmanniidae (R.A.N., unpublished) – but the relationships of this strange mite are poorly understood.

Three other characters are interesting, if not conclusive. First, the prelarva of Collohmannia is quite similar to that of Nothrina, but their shared retention of relatively large leg vestiges is an obvious symplesiomorphy. Prelarvae of Ptyctima are more regressive, i.e. more derived, in all known instances ( Sitnikova 1960; Grandjean 1962b; Lions 1973). The second relates to sperm ultrastructure, studies of which have been illuminating in recent decades. According to Alberti and Schuster (2005) mature sperm of C. gigantea are more similar to those of brachypyline oribatid mites than to the unique, small, lens-like sperm of Phthiracaridae (the only Ptyctima that have been studied in this regard). However, developing spermatids of C. gigantea are unique among mites in their length and highly coiled chromatin, and the presence of a cuff (’manchette’) of microtubules accompanying chromatin condensation is shared only with Phthiracaridae among studied species. The third trait relates to general body form. Collohmannia is dichoid, but the ventral part of the sejugal articulation is narrow – little more than a hinge to allow dorsoventral flexing – and the dorsal part merges with the broad articulation surrounding the notogaster. As such, it approaches the holoid body form of Desmonomata .

While the gnathosomal similarities seem significant, we see no other morphological support for grouping Collohmannia with Nothrina or with any particular family in that group. Nothridae do have bothridial outpockets, but they are quite different in form ( Calugar and Vasiliu 1979). For now, Collohmannia must join the sizeable list of taxa exhibiting conflicts between morphological and molecular data.

Figures

FIGURE 10: Collohmannia johnstoni n. sp. adult, ano-genital region: A, B – male, incident light; C, D – female, transmitted brightfield (C) or DIC (D) illumination. A – ventral overview, showing positions of cross-sections; large white arrow points to notch of unsclerotized cuticle between adanal and aggenital plates; B – posterior view of hysterosoma, white arrow points to soft terminus of body; C - cross section at level of postanal apodeme; D – cross-section showing composition and musculature of the mid-anal region. Abbreviations: AG, AN, AD, GE – aggenital, anal, adanal, genital plates, gl – gland, is.1, is.2 – intercalary sclerites, mu – muscle, NG – notogaster, pa.a – postanal apodeme, pr.a – preanal apodeme. Each image layered and color- and sharpness-enhanced (see Material and Methods).