Allonothrus tuxtlasensis Palacios-Vargas & Iglesias, 1997

Redescription of Allonothrus tuxtlasensis Palacios-Vargas & Iglesias, 1997 View in CoL

Taxonomy

Family Trhypochthoniidae (sensu lato; see below)

Genus Allonothrus van der Hammen, 1953

Type species: Allonothrus schuilingi Hammen, 1953 ; generic diagnosis: van der Hammen (1953), Wallwork (1960)

Background

Allonothrus tuxtlasensis is a Caribbean species, one of seven known from the Neotropics ( Szywilewska-Szczykutowicz & Olszanowski 2006). It was described based on collections from the state of Veracruz, Mexico ( Palacios-Vargas & Iglesias 1997) and its known distribution now includes the Mexican states of Quintana Roo, Chiapas, and Tlaxcala ( Ermilov & Yurtaev 2023, and unpublished records from J.G. Palacios-Vargas). It probably occupies all the Antilles, having been reported from several islands, extending from Cuba in the north to Trinidad in the south ( Prieto Trueba & Schatz 2004; Ermilov et al. 2016; Ermilov & Smit 2017). All collections have been from decomposing leaf litter, but the habitats range widely, from forests (natural jungle, secondary woodlands) to organic residues on sandy soil of an abandoned coconut grove.

The original description of A. tuxtlasensis suffices for identification of the adult, but it lacks important details concerning the structure of gnathosoma and the chaetome of body and legs. These details are provided below, supported by line drawings, light photographs, and SEM images. Juveniles of A. tuxtlasensis are not yet known, so we present previously unpublished developmental data for an Indian species— A. giganticus Haq, 1978 —that was cultured by Palmer & Norton (1990), to allow inferences about leg setation.

Diagnosis

Based on the original description ( Palacios-Vargas & Iglesias 1997) and supplemental data in our redescription we propose the following diagnosis for adult A. tuxtlasensis . Genus-level characters are mostly omitted.

Allonothrus species with body length: 495–585. Rostral seta medium-sized, thickened, densely ciliate; lamellar seta longest on prodorsum, heavily barbed, thick basally, becoming sub-spatulate (phylliform) distally; interlamellar seta short, usually phylliform, densely barbed, at base of small tubercle; bothridial seta long, distal half weakly clavate, densely barbed. Dorsocentral part of notogaster foveolate except near setae; without distinct posterior concavity but pygidial region with slight semi-elliptical depression. Notogastral setae heavily barbed, widest distally, dimorphic; c 1, c 2, c 3, cp, d 1, d 2,e 1, e 2 short, phylliform (mostly sub-spatulate); f 2, h 1, h 2, h 3, p 1, p 2, p 3 long, with thick isodiametric basal half, gradually broadened distal half. Epimeral, genital and adanal setae slightly thickened, densely barbed; 10 or 11 pairs of homogeneous genital setae; anal setae attenuate, roughened by inconspicuous minute barbs. Pretarsi homotridactylous; genua I and II with minute setiform organ σ m in place of genual pore; genu III with genual pore.

Redescription of adult ( Figs 1–7 View FIGURE 1 View FIGURE 2 View FIGURE 3 View FIGURE 4 View FIGURE 5 View FIGURE 6 View FIGURE 7 )

Measurements. Body length 495–585; width 255–315.

Integument. Color of preserved specimens medium reddish to yellowish brown. Body and legs covered by encrusting cerotegument with weakly and irregularly microtuberculate surface ( Figs 2a, b View FIGURE 2 ; 7a, f View FIGURE 7 ), including localized regions of more distinct, larger surface granules (diameter up to 4). With additional adherent debris, especially in posteromedial part of notogaster between longer setae. Exocuticle densely porose, most conspicuously on prodorsum, coxisternum, and subcapitulum ( Figs 3a, f View FIGURE 3 , 6a View FIGURE 6 ).

Prodorsum ( Figs 1 View FIGURE 1 , 2 View FIGURE 2 , 3a View FIGURE 3 ). Triangular in dorsal view; rostrum evenly rounded except for small, paired lateral tooth (lt). Middle third with two topographic features: pair of curved medial carinae (mc), nearly meeting to form inverted-V shape; and pair of marked lateral swellings (sw) underlain by muscle sigilla and delineated by sharp groove and strong contours (lc) in transmitted light. Posterior third of prodorsum slightly raised behind medial carinae, with small, sharply defined medial depression (md) overlying several muscle sigilla, and pair of bulging, almost capsular bothridial lobes. Bothridial opening surrounded by raised rim ( Fig. 2b View FIGURE 2 ); internally with cluster of short brachytracheae opening into region just external to seta bs insertion ( Fig. 3c View FIGURE 3 ). Bothridial seta (86–94; Figs 2b, c View FIGURE 2 , 3c View FIGURE 3 ) slightly and gradually broadened by isotropic cuticle in distal third, forming weakly clavate-spatulate head; barbs minute near base, distally becoming larger and overlapping. Rostral (ro), lamellar (le), and interlamellar (in) setae all heavily barbed but differing greatly in size and form: ro (41–49) nearly isodiametric, curving ventromediad, pair separated by about half their length; le (90–105) inserted on strong tubercle, proximally thick, isodiametric, distal third sub-spatulate, gradually flattening and doubling in width, pair close together, directed almost straight anteriad; in (13–15) inconspicuous, phylliform (fan-shaped), approximately as broad as long, cupped, with dense barbs on outer face ( Figs 2d View FIGURE 2 , 3c View FIGURE 3 ), inserted at base of small, triangular tubercle. Exobothridial setae both represented by vestiges; ex 1 on rim of bothridial lobe lateral to opening, represented by simple, inconspicuous microseta (1) hardly emerging from alveolus ( Figs 2b View FIGURE 2 , 3c View FIGURE 3 ); ex 1 ventral to bothridial lobe, represented only by alveolus, larger than that of ex 1.

Notogaster ( Figs 1–3 View FIGURE 1 View FIGURE 2 View FIGURE 3 ). About 1.1–1.3 x longer than wide in dorsal view, broadest in posterior quarter, at level of opisthonotal gland ( Figs 1a, d View FIGURE 1 , 2a, e View FIGURE 2 ). Anterolaterally concave above legs III, IV ( Fig. 2e View FIGURE 2 ) and below suprapleural carina ( Fig. 1d View FIGURE 1 , spc) running approximately between levels of lyrifissures ia and im; spc variably expressed according to body distension. Central region mostly with strong, well-spaced foveolae, round but of various diameters (up to 10); foveolae absent in general vicinity of setae ( Fig. 3d, e View FIGURE 3 ) and laterally, ventral to spc. Without distinct larger concavities but pygidial region with shallow, semi-elliptical depression between setal pairs h 1 and ps 1 observable in dorsal view of cleaned specimens. Notogastral setae heteromorphic, all heavily barbed; c 2 directed anteriad, others generally posteriad. Setae c 1, c 2, d 1, d 2 (37–49), c 3 (22–26), cp, e 2 (30–41), e 1 (45–56) strongly phylliform-spatulate ( Figs 2a View FIGURE 2 , 7a View FIGURE 7 ); narrow at base, gradually broadened, flattened and cupped distally; inserted on small tubercles. More posterior setae— f 2 (86–94), h 1, h 2, h 3, p 1 (123–139), p 2 (75–101), p 3 (71–94)—thick and isodiametric in basal half or two-thirds, gradually broadened, flattened and cupped distally to various degrees but often less conspicuously so than shorter anterior setae ( Figs 2a View FIGURE 2 , 7b View FIGURE 7 ); inserted on large tubercles. Seta f 1 represented only by small alveolar vestige in smooth cuticle, anterior to seta h 1 ( Figs 3e View FIGURE 3 , 4a View FIGURE 4 ; see R4). Opisthonotal gland opening on low protuberance, anterodorsal to seta f 2 ( Figs 1a, b View FIGURE 1 , 4a View FIGURE 4 ). Lyrifissures ia, im, ip, ih in normal positions and except for ih of normal size and slit-like shape; ih unusually short and inconspicuous ( Fig. 1a, c, d View FIGURE 1 ), sometimes visible only as alveolus; ips inconspicuous (especially in contracted specimens), sub-cupular in form and located on softer cuticle of broad plicature region, seen directly or by transparency in distended or contracted specimens, respectively (cf. Figs 1c View FIGURE 1 , 3h View FIGURE 3 ; see R5).

Epimeral region. Coxisternum clearly in two units, with fused epimeres I/II separated from III/IV by narrow, but distinct ventrosejugal articulation (vsj, Figs 1c View FIGURE 1 , 3f, g View FIGURE 3 , 7d View FIGURE 7 see R6). Apodeme ap.1 strongly oblique, ap.2 and ap.3 nearly transverse; all three meeting sternal apodeme (ap.st) at midline; ap.st inconsistently developed ( Figs 1c View FIGURE 1 , 3f View FIGURE 3 ), weakening posterior to ap.2, in mid-region of epimere II), and both anterior and posterior to ap.3. With narrow but distinct surface groove external to ap.2 and ap.3 ( Fig. 7d View FIGURE 7 ). Epimeral setal formula 3-1-3-3 (I–IV); setae of different lengths (1a, 2a, 3a, 4b: 13–15; others: 30–34) but all slightly thickened, densely barbed throughout.

Anogenital region ( Figs 1c, d View FIGURE 1 , 2a View FIGURE 2 ). Genital plates collectively trapezoid, 1–1.3 x wider than long; anteriorly with weak ridge along medial edge of each plate, effacing in posterior half. With row of 10–11 genital setae (30–37) aligned along medial edge but not isolated from rest of plate by sharply defined line or carina. Setae (30–37) slightly thickened, densely barbed throughout. Ovipositor ( Fig. 6c View FIGURE 6 ) with general structure normal for family ( Ermilov 2011a); length of distal part (beyond fold) 56, of which distal lobes comprise 19; setae ψ 1, and another (unidentified) τ-seta elongate, acute, narrowly spine-like (19); ψ 2 and two unidentified τ-setae short (7), thick, thorn-like, strongly bent in middle at nearly right-angle ( Fig. 6e View FIGURE 6 ); six coronal setae (k) also short (9), thick, thorn-like, but nearly straight ( Fig. 6d View FIGURE 6 ). Three pairs of adanal setae (22–26) setae appear slightly thickened by dense barbs; two pairs of anal setae (19– 26) attenuate, without conspicuous barbs. Lyrifissures iad and ian distinct, near anterior end of respective plates.

Gnathosoma. Subcapitulum ( Figs 4b View FIGURE 4 , 6a View FIGURE 6 . 8a, b View FIGURE 8 ) longer than broad: 90–101 × 71–86. With elements of both stenarthry and diarthry: oblique paired labiogenal articulation (scissure lgo) incomplete anteriorly, stopping just before reaching inferior commissure of mouth (Ji). Additional transverse articulation (scissure lgt) running laterally from Ji toward palp; genal seta m posterior to lgt (see R7). Rutellum atelobasic, with weak ventromedial expansion continuing distally as well-developed lamellate cutting edge (ce), fully covering terminal rutellar teeth from below; lateral lips and three pairs of homogeneous, relatively large (11–13), attenuate, mediodistally barbed adoral setae, all well exposed between Ji and rutellar expansion; rutellar brush (comb) comprising two ciliary rows on dorsomedial face; manubrial zone defined proximally by distinct fissure αf, distally by collum (c). Subcapitular setae a (15–19), m (7–9), h (15–22) attenuate, roughened by inconspicuous minute barbs, m thinner than a and h. Palp ( Fig. 4c View FIGURE 4 ) length 49–52, tarsus about 1.5 x length of tibia; setation 0-1-1-2-9(+ω); ω short, baculiform; setae acm, sul, and (ul) with typical shape of eupathidia (smooth, blunted) but hollow interior not confirmed. Postpalpal seta (6) narrowly cylindrical ( Fig. 6b View FIGURE 6 ), roughened by inconspicuous minute barbs. Chelicera ( Fig. 4d View FIGURE 4 ) length 82–94; setae finely attenuate, cha (41–45) distinctly barbed, chb (17–22) nearly smooth (minutely roughened).

Legs. Proportions shown in Fig. 5a–d View FIGURE 5 ; leg I about half body length. Pretarsi homotridactylous; claws strong, similar in size, minutely barbed on dorsal side. Formulas of leg setation and solenidia: I (1-6-5-6-16) [2-2-3], II (1-7- 5-5-14) [2-1-2], III (2-4-3-4-12) [1-1-0], IV (1-3-3-4-12) [0-1-0]; homology of setae and solenidia indicated in Table 1 View TABLE 1 (assumptions about tarsal setations and that of tibia I based on ontogeny of Allonothrus giganticus ; Appendix 1). All or most setae on femora, genua and tibiae thick, heavily barbed. Famulus relatively long (13), acuminate. Tarsus I solenidion ω 1 short, curving, blunt (baculiform); ω 2 and ω 3 thinner but longer, tapered (piliform), close together proximal to seta p ʺ; φ 1 subflagellate but shorter than closely coupled flagellate seta d; other solenidia comparatively short, baculiform to slightly tapered but rounded apically, with those of genua and tibiae much smaller than coupled seta d. Setae (p) eupathidial on tarsus I; s normal, with minute barbs. Genual pore present on genu III ( Fig. 6j View FIGURE 6 ); replaced on genua I and II by minute (2–3), spiniform σ m (presumed solenidion; Figs 6f–i View FIGURE 6 , 7e, f View FIGURE 7 ; see R9)

Discrepancies with original description

Our specimens of A. tuxtlasensis from Mexico, Cuba, Grenada, and Trinidad are morphologically similar among themselves and are similar to the original description of Mexican type material ( Palacios-Vargas & Iglesias 1997) except as relates to two traits. (1) The interlamellar seta is strongly expanded (phylliform) and barbed in all our material, but setiform in the original description. (2) Some of the ventral setae illustrated in the original description have somewhat different relative length than in our material. At our request, Dr. Palacios-Vargas (pers. com. 2024) reinvestigated type specimens, and reported that the morphology of seta in varies in the type material (setiform to phylliform), and that apparent differences in the length of the ventral setae are mainly due to different drawing styles and methods of fixing specimens when drawing (temporary cavity slides versus permanent slides).

Remarks on morphology

1. Exobothridial setae

Ancestrally in oribatid mites there are two pairs of exobothridial setae; each inserts in various positions posterolateral, lateral, anterolateral or (rarely) anterior to the bothridium. The different relative positions of the two have led to the application of several notations. When one is clearly more dorsal, they have been labeled exs and exi (or xs, xi; superior, inferior); when one is clearly more anterior the notations exa, exp (or xa, xp; anterior, posterior) have been used. Alternatively, they are simply numbered ex 1, ex 2. Homologies among the notations are sometimes uncertain but we believe that in most instances notations exs, exa, and ex 1 represent the same seta, with the other seta bearing the notations exi, exp, ex 2. 1 Since relative positions vary, we believe it is best to standardize notations using the numerical form. Of the two, ex 2 is most prone to loss and in such cases the remaining ex 1 usually is simply denoted ex, or x. Some generalities can be made in this regard: (1) among Palaeosomata and Parhyposomata both setae always are well-formed; (2) in Brachypylina ex 2 never forms but can be represented by a distinct vestigial alveolus (exv) in some early- to middle-derivative groups ( Norton & Ermilov 2021, their Remark #4). Within the other major groups various states exist, with the plesiomorphic state being both setae present, but either one (ex 2) or both setae can be absent or reduced to a vestige.

In Nothrina adults ex 2 is reduced to an alveolar vestige. In Trhypochthoniidae usually this vestige (mark ‘ m ’ of Grandjean 1939a; his Fig. 3C, D View FIGURE 3 ) is distinct in the sclerotized adult, located near the margin of the aspis, while in juveniles it is less distinct and usually not on the aspis (which is less developed lateroventrally than in the adult). The vestige is visible in all instars in some genera ( Trhypochthonius , Mainothrus and Archegozetes ), but in Trhypochthoniellus it may be absent until the DN or even the adult.

By contrast, ex 1 has more diverse development in this family. In the larva of all genera it is a distinct small- or medium-sized seta, and it retains this plesiomorphic form through ontogeny in Mucronothrus and Trhypochthoniellus , as it does in most other families of Nothrina . In other trhypochthoniid genera it becomes highly reduced ( Mainothrus ; Ermilov 2021) or vestigial ( Trhypochthonius , Allonothrus , Archegozetes ) or may disappear entirely ( Afronothrus ). The nature of the vestige varies as well. In Trhypochthonius it ranges from a simple pore-like canal ( T. tectorum ; Grandjean 1939a) to a minute but emergent seta ( T. triangulum Nakamura et al. 2013 ). In Allonothrus tuxtlasensis a minute setal vestige barely emerges from the reduced alveolus on the projecting lateral wall of the bothridium ( Fig. 2b View FIGURE 2 ).

2. Bothridial porose organs

In Mixonomata and Nothrina a variety of porose organs can invaginate from the inner wall of the bothridium. In most instances these are air-filled respiratory organs ( Norton et al. 1997) such as simple pouch-like pockets, saccules, sausage-like brachytracheae, or short tracheae. To our knowledge, the only report of such structures in Trhypochthoniidae was for Afronothrus , in which Wang et al. (1999) found a short, flat saccule invaginating from the bothridium (see our Fig. 9a View FIGURE 9 ). But in fact most Trhypochthoniidae have such respiratory structures, though they may be less developed. Small pocket-like or saccule-like invaginations are present in Archegozetes , Mainothrus and Trhypochthonius ( Fig. 9c–e View FIGURE 9 ), which seem similar to the bothridial saccules of Platynothrus ( Alberti et al. 1997) . Allonothrus species have the greatest development of such structures in the family ( Figs 3c View FIGURE 3 , 9b View FIGURE 9 ). In the seven species studied by us (see Materials and methods) there is a small cluster of short brachytracheae, similar to those described for Nothrus ( Grandjean 1934; Tarman 1961; Călugăr & Vasiliu 1979; Alberti et al. 1997) but fewer and more spreading.

3. Trichobothrial regression

Except for Hermanniidae , in all examined Nothrina the prodorsal trichobothrium is strongly regressed in the larva ( Grandjean 1939a, his Fig. 3D View FIGURE 3 ; 1939b). The bothridial seta (‘sensillus’) is small, minute, or even absent, and the bothridium is undeveloped or shrunken to a narrow tube, such that the structure is similar to an apobasic seta ( Grandjean 1956b; Călugăr & Vasiliu 1979). As in most other Nothrina , in Trhypochthoniidae it usually becomes a normal trichobothrium during subsequent development, with the instar depending on taxon. In Afronothrus , Archegozetes Trhypochthonius and Mainothrus this is the PN, for Allonothrus schuilingi the DN. In Mucronothrus the regression persists until the adult, i.e. it is paedomorphic; the seta is well formed, but its insertion is apobasic, not really bothridial ( Norton et al. 1996). In Trhypochthoniellus the same may be true, according to species or, apparently, to genetic strain ( Weigmann 1999).

4. Notogastral seta f 1

Of the 16 pairs of setae that comprise the fundamental (holotrichous) chaetome of the hysterosomal dorsum in oribatid mites, seta f 1 has been considered the weakest, the most prone to loss ( Grandjean 1954a; Travé 1975; Haumann 1991). This regression has occurred in most Mixonomata, many Nothrina , all Brachypylina except Hermannielloidea , and probably all Astigmata (see summary in Norton & Ermilov 2022). In the latter two groups, as well as the mixonomatan family Eulohmanniidae , f 1 has disappeared without trace, but in most mixonomatans and nothrines, its loss is indicated by a persistent alveolar vestige.

In Nothrina , seta f 1 seems rather labile among higher taxa and even among species. In Crotoniidae (sensu lato) the ancestral condition reflects that of the mixonomatan family Perlohmanniidae ( Grandjean 1958; Suzuki 1977), in which f 1 is present in the larva as a small seta, becoming vestigial in nymphs and adult. This is the state in the crotoniid subfamilies Camisiinae and Heminothrinae but in Crotoniinae f 1 is developed in all instars ( Colloff & Cameron 2009). Since crotoniines are derivative within the family ( Domes et al. 2007b) the transition can be viewed as paedomorphic—retaining the larval presence through ontogeny—and explained by the removal or inactivation of whatever developmental mechanism caused its suppression in ancestors ( Weigmann 2010). Several examples demonstrate this lability at the species level. Species of Platynothrus ( Lee 1985; his seta J4) and Camisia ( Colloff 1993) have been described with f 1 uniformly present in the adult, and Seniczak et al. (1990) reported variable development of f 1 in adults of Platynothrus capillatus ( Berlese, 1914) . Removal of ancestral suppression would be the simplest explanation for the holotrichous setation of Nothridae and Hermanniidae , as well as the brachypyline superfamily Hermannielloidea .

Trhypochthoniidae do not form seta f 1 in any instar, and f 1 vestiges have been reported or illustrated in the literature for adults of all genera except Allonothrus . This may be because Allonothrus species have a notogastral cuticle that is far more ornate than those in other genera and the simple, inconspicuous vestige has been overlooked. We found these vestiges in their usual position, aligned between setae e 1 and h 1, in A. tuxtlasensis ( Fig. 3e View FIGURE 3 ), A. sinicus ( Fig. 9l View FIGURE 9 ) and the other six Allonothrus species we examined (see Materials and methods).

It is clear that the suggestion of Badejo et al. (2002a) —i.e., that Allonothrus species are holotrichous, and that describers have overlooked one pair of developed notogastral setae—is specious. While some descriptions and illustrations are of marginal quality, most are adequate and several are highly precise, with no real possibility that a seta as conspicuous as those of Allonothrus species could be missed. Our finding of a setal f 1 vestige should remove any doubts about the ‘unideficient’ chaetome of Allonothrus . Considering f 1 lability in other groups, the report by Badejo et al. (2002a) of holotrichy in Parallonothrus is not unreasonable, but it has not been conclusively demonstrated (see below).

5. Lyrifissures

It seems likely that all Trhypochthoniidae have the usual five pairs of notogastral lyrifissures, but they can have various forms and positions. Weigmann (1997b) considered a relatively large size of ia and ip to be a synapomorphy of Malaconothroidea , but this is not a consistent trait in Trhypochthoniidae , particularly in trhypochthoniid genera that were not part of his study, viz. Afronothrus , Allonothrus and Archegozetes . Lyrifissure ip was not discerned in some Allonothrus studies (e.g. van der Hammen, 1953; Wallwork 1960, 1961), but its presence can be masked by the strong foveolation in this genus; it was found in the seven species we studied (see Materials and methods).

In oribatid mites, lyrifissure ips typically is located on the sclerotized cuticle of the notogaster, near its ventrolateral edge, but probably this is not true in most Trhypochthoniidae . In Allonothrus ips is off the sclerite, in the soft, foldable plicature zone (van der Hammen 1953; Wallwork 1961; Subías & Sarkar 1982; Olszanowski & Bochniak 2015; our Fig. 3h View FIGURE 3 ), where it has a somewhat cupular, rather than slit-like form. Whether one views ips directly or by transparency depends on the degree of distention in the hysterosoma, but most illustrations of contracted specimens do not seem to make this distinction. Illustrations of well-distended specimens are more rare, but they show ips in the plicature zone of at least some species of Afronothrus , Mucronothrus , Mainothrus , and Trhypochthonius ( Ramani & Haq 1992; Norton et al. 1996; Weigmann 1997b; Szywilewska 2004). By contrast, ips seems to have a normal position on the edge of the notogaster in Archegozetes . We are uncertain about Trhypochthoniellus , but ips also lies in the plicature zone in Malaconothridae ( Knülle 1957) , While this position is unusual among Nothrina , ips also lies in the plicature zone of the mixonomatan families Perlohmanniidae and Collohmanniidae ( Grandjean 1958; Norton & Sidorchuk 2014), which leaves open the possibility that it is plesiomorphic if Malaconothroidea is a basal nothrine taxon (see below).

6. Ventrosejugal articulation

Adults of Trhypochthoniidae and Malaconothridae are distinct among Nothrina in having a divided coxisternum; i.e., a narrow transverse articulation (scissure) separates epimeres II and III. As such, they do not conform to the general body-form terminology proposed by Grandjean (1969) for oribatid mites having significant cuticular sclerotization. He proposed the term holoidy for the condition in which the proterosoma and hysterosoma are not independently movable: the usual key trait is that the coxisternum is undivided. Holoidy is found in all Brachypylina and most Nothrina , and was the basis for the collective name of these two groups, Holosomata. 2 He contrasted this with dichoidy, in which a broad, ring-like sejugal (or ‘protero-hysterosomatic’) articulation allowed significant independence of the two secondary body regions, including relative motion—some degree of axial angular displacement in all directions—and some degree of telescoping. As he explained, dichoidy (or its derivative, folding defensive body form, ptychoidy) is found in the other major macropyline groups—Palaeosomata, Enarthronota, Parhyposomata and Mixonomata. Iconic mixonomatan examples include Eulohmanniidae , Epilohmanniidae and Perlohmanniidae .

Grandjean (op. cit.) emphasized his view that dichoidy was not ‘primitive’ in the sense of being ancestral to holoidy, but he was silent about two important issues. First, he did not indicate what type of body form was ancestral to holoidy: if not dichoidy, then what? Certainly, Nothrina and Brachypylina did not evolve their sclerotization patterns directly from an unsclerotized ancestor. Second, he left unexplained the problematic nothrine families Malaconothridae and Trhypochthoniidae , which are neither holoid (since they have a narrow but distinct ventrosejugal articulation) nor dichoid (since this narrow articulation serves only as a hinge, allowing slight dorsoventral flexing but not lateral relative motion or telescoping). Weigmann (1997b) called this narrow articulation between epimeres II and III a ‘weak transversal zone’; for brevity, we coin the term subholoid for this condition (noun = subholoidy).

Without using specialized terminology, Knülle (1957) illustrated his interpretation of the transition from subholoidy to holoidy, referring to a ‘Trhypochthonoidea-stage’ (his Fig. 21) with a narrow ventrosejugal articulation, and a ‘ Nothrus -stage’ with a fused coxisternum (his Fig. 22). In effect, he viewed the ventrosejugal articulation as plesiomorphic in Nothrina (= Nothroidea). Haumann (1991, p. 144) specifically indicated that holoidy evolved from dichoidy, though he did not characterize the condition in Trhypochthoniidae and Malaconothridae as intermediate. For support, he quoted Grandjean (1969), but it was a misinterpretation: the cited text was a hypothetical ‘strawman’ statement that Grandjean immediately rejected in the next sentence (see above).

Overall, the ideas promoted by the three German authors might be distilled as: subholoidy was derived from dichoidy by narrowing of the ventral part of the sejugal articulation, and holoidy derived from subholoidy by elimination of the articulation. In this view, Trhypochthoniidae and Malaconothridae form the sister-group of Holosomata (see Fig. 11a View FIGURE 11 ); we examine this issue further below, in a section that addresses inferences from molecular studies.

An important point is that the subholoid state is not restricted to Trhypochthoniidae and Malaconothridae . Norton & Metz (1980) described it in the mixonomatan family Nehypochthoniidae , though they used no special terminology. Norton & Sidorchuk (2014) followed Grandjean (1969) in applying the term dichoid to another mixonomatan family, Collohmanniidae , but in effect they described the subholoid condition in these mites, which have a very narrow ventral articulation and lack the independent movement of proterosoma and hysterosoma found in truly dichoid families.

Whether these examples represent separate evolutions of subholoidy from dichoidy is unknown, but three other examples of narrowing in the ventral component of the sejugal articulation are certainly convergent. Grandjean (1969) promoted the idea that dichoidy was ancestral to ptychoidy, the defensive body form in which the aspis can swing down to cover the retracted legs ( Sanders & Norton 2004 and included references). This functionality requires the coxisternum to fold in half precisely, and it has this capability because the ventral component of the sejugal articulation became very narrow, hinge-like. Grandjean (1969) noted three evolutions of ptychoidy in oribatid mites, which implies three different times in which a broad, dichoid-type sejugal articulation evolved into one with a narrow, hinge-like ventral component. Grandjean did not focus on this point, but in essence he accepted such an evolutionary reduction in the ventral component of ptychoid groups, while apparently rejecting such a transition during the evolution of holoidy.

As discussed below, in molecular studies Malaconothridae usually have been recovered as basal in Nothrina , or nearly so. This would be consistent with the idea that subholoidy is a transitional state between dichoidy and holoidy. But a possible alternative origin for subholoidy in Trhypochthoniidae is suggested by some molecular trees in which this family lies in the midst of holoid Nothrina (e.g., Fig. 11e View FIGURE 11 ). Typically, the coxisternal sclerotization of juveniles in holoid groups is fragmented, so it might also have arisen as a paedomorphic (neotenic) trait. Overall, paedomorphic trends could explain many of the setal reductions that characterize Trhypochthoniidae ( Norton 1998) and an effective reversal from holoidy might be another manifestation, if indeed this family evolved in the midst of holoid Nothrina .

7. Subcapitular structure and the evolution of diarthry

A striking feature of Trhypochthoniidae is that—despite the absence of specialized chelicerae—they exhibit the widest range of subcapitular structure of any oribatid mite family ( Table 4 View TABLE 4 ). In his classic synthesis on the morphology and evolution of the oribatid mite subcapitulum (‘infracapitulum’), Grandjean (1957) recognized three general forms that relate to the presence and nature of paired labiogenal articulations (lg) on the ventral face. Such articulations allow small relative movement of the paired malapophyses (the supposed coxal endites of the palpal segment, the ventral surface of which is the gena), each bearing a distal rutellum. The articulation allows flexing, such that the rutellum maintains contact with the retracting chelicera, providing a scissoring action to cut food fragments (van der Hammen 1980; Evans 1992). A stenarthric subcapitulum has an oblique pair of labiogenal articulations leading posterolaterally from the mouth (commissure Ji); these circumscribe a triangular mentum, considered the sternum of the palpal segment. In a diarthric subcapitulum lg runs laterally from the mouth to the base of each palp, circumscribing a rather quadrate ‘mentum’ (see below). An anarthric subcapitulum lacks such ventral articulation, either because it is not significantly sclerotized or—if sclerotized—because other features preclude the need for an articulation. While the forms have a somewhat mosaic distribution, in general anarthry characterizes endeostigmatid outgroups of Oribatida , as well as some Palaeosomata and most Enarthronota, and was interpreted by Grandjean (1957) as the most primitive form. Stenarthry characterizes most middle-derivative groups, and diarthry (or its derivatives) characterizes the Brachypylina.

In this conceptual model, with increased levels of sclerotization stenarthry evolved from anarthry by the evolution of an oblique lg to allow flexing of the malapophyses, and diarthry evolved from stenarthry by a change in the trajectory of lg. Grandjean did not discuss how or why this latter change might have occurred but—since the terminology (mentum, genae, lg) does not vary between the types —presumably he did not question the homology of lg or the sclerites it separates in the two articulated forms. We know no intermediate states that would support the idea that the trajectory of lg changed gradually, so how did diarthry arise?

The subcapitular forms in Hermanniidae suggest one likely pathway to diarthry.While many of its species exhibit typical stenarthry with a long, oblique lg running from the mouth to the posterolateral margin of the subcapitulum (e.g. van der Hammen 1968), others have an incomplete articulation, with lg progressing only a short distance posterolaterad from the mouth (e.g., Figs 19, 23 in Woas 1978, Figs. 1 View FIGURE 1 , 8 View FIGURE 8 in Woas 1981). Such an abbreviated lg may have subsequently extended further laterad toward the palp base, to create a diarthric subcapitulum typical of Brachypylina. 3

Allonothrus shows us another potential pathway between stenarthry and diarthry, and also demonstrates that the two forms of labiogenal articulations may not always be homologous. As Woas (2002) noted, the subcapitulum has a variety of forms in what he referred to as the ‘ Allonothrus group.’ Allonothrus sinicus , for example, exhibits typical stenarthry ( Fig. 8c View FIGURE 8 ). By contrast, the subcapitulum of A. tuxtlasensis (also A. malgorzatae ) has elements of two articulations and might be characterized as ‘duplex’. In A. tuxtlasensis the oblique articulation (lgo) is incomplete anteriorly (opposite to that of some Hermanniidae ), not quite reaching the mouth, while a transverse articulation (lgt) runs laterad from the mouth ( Figs 4d View FIGURE 4 , 6a View FIGURE 6 , 8a, b View FIGURE 8 ). Seta m, generally associated with the gena in oribatid mites, is posterior to lgt in the duplex type, whereas it is anterior to lg in the typical diarthric subcapitulum of Brachypylina and in the shortened, near-transverse lg of some Hermanniidae (see above).

It is easy to envision the loss (by fusion) of the oblique articulation altogether, and this may have occurred during the evolution of diarthry in Afronothrus ( Fig. 8d View FIGURE 8 ) and Trhypochthoniellus ( Weigmann 1996; his Fig. 1C View FIGURE 1 ). While their illustrations were crude and difficult to interpret, Badejo et al. (2002a) described such conditions in Parallonothrus : P. nigeriensis has a transverse articulation but the additional oblique line was implied to be a nonfunctional ridge; by contrast, P. brasiliensi s purportedly possesses only the transverse articulation. Unfortunately, much literature on Trhypochthoniidae does not distinguish between a functional articulation and a simple ridge or line of fusion,

Weigmann (1996) had a different view: he promoted stenarthry as the ancestral condition in oribatid mites, and envisioned anarthry as a transitional condition during the evolution of diarthry in Trhypochthoniellus ; i.e. its ancestors entirely lost the oblique articulation, then evolved the transverse one anew. We believe such a pathway is unlikely. In essence, it was imposed by a constraint: the relationships perceived in his cladogram (his Fig. 2 View FIGURE 2 ; our Fig. 8b View FIGURE 8 ; see also Knülle 1957; Weigmann 1997b), in which Trhypochthoniellus is sister-group to the anarthric Malaconothridae . The inferred relationship was based primarily on other traits, and if—as we believe— Trhypochthoniellus is instead part of a monophyletic Trhypochthoniidae (sensu lato; see below) the constraint is removed.

This brings up a terminological problem, first touched on by Weigmann (1996; see also Alberti & Coons 1999). Theoretically, if the triangular mentum of a stenarthric subcapitulum represents the sternite of the ancestral palpsegment, and each paired gena represents only the venter of the coxal endite (i.e., the malapophysis; see also van der Hammen 1968, 1980), then the region that is lateral to the triangular mentum and basal to the palp—which must derive from the ancestral palp coxa—has no designation. It remains problematic, but whether this interpretation is correct or not, and despite common usage, the ‘mentum’ of stenarthric and diarthric subcapitula are not fully equivalent.

8. Structure of the rutellum

Based on outgroup comparisons with mixonomatans, the plesiomorphic form of the rutellum in Nothrina is atelobasic—i.e., not reaching the midline at its base, such that lateral lips are at least partly exposed—with a simple dentate distal margin ( Grandjean 1957). Among Trhypochthoniidae this combination is best exhibited by Mucronothrus and Trhypochthonius , but modifications have occurred both proximally and distally. While most Trhypochthoniidae have an atelobasic rutellum, it is somewhat variable in Allonothrus . In A tuxtlasensis , the lateral lips are well exposed ( Figs 4b View FIGURE 4 , 8a View FIGURE 8 ), while in A. malgorzatae the rutellum seems pantelobasic, with no exposure of the lateral lips. Distally, the rutellum can be modified by the development of a variably-developed distal cutting edge (ce), such that the terminal dentition is hidden in ventral view. This edge is strongly developed, lamellate, in Afronothrus ( Fig. 8d View FIGURE 8 ) and Trhypochthoniellus but in Archegozetes it is rudimentary, represented by a simple transverse carina (see Fig. 3 View FIGURE 3 in Alberti et al. 2011). The edge seems similarly variably among species of Allonothrus (cf. Fig 8a, 8c View FIGURE 8 ), and Mainothrus (cf. Weigmann 1997a; Bayartogtokh & Yondon, 2019).

9. Genual pore and setiform organ σ m

As first reported by Grandjean (1940a), in various middle-derivative families of Mixonomata and Nothrina the genu of leg I bears a so-called genual pore. Despite its name, it does not open to the surface: it is a simple, well-defined canal passing through the procuticle (‘ectostracum’) but not the epicuticle (‘epiostracum’). He noted (see also Grandjean 1954c, 1958, 1966) that the genual pore has a somewhat mosaic distribution—present in some Mixonomata (including Ptyctima, Epilohmanniidae , Collohmanniidae and Perlohmanniidae ) and Nothrina ( Nanhermanniidae , some ‘Camisiidae’, some Trhypochthoniidae ) but not others—and that in some groups a metameric homologue is present also on genua II and III, but never IV.

Among Trhypochthoniidae , the pores can be found on genua I–III in representatives of all terrestrial and semiaquatic genera: they were reported previously in Trhypochthonius ( Grandjean 1940a) and Mainothrus (Ermilov 2021) , and with new observations we confirm their presence on these same genua in Afronothrus , Archegozetes , and some Allonothrus species ( Table 2 View TABLE 2 ; Fig. 9 View FIGURE 9 ). Each is located posterior to the coupled d σ, but at various distances (cf. Fig. 9f, g View FIGURE 9 ). Conversely, we confirm that genual pores are absent from Trhypochthoniellus ( Grandjean 1940a) and Mucronothrus ( Norton et al. 1996) , which are the two genera specializing in aquatic habitats ( Behan-Pelletier & Eamer 2007; Schatz & Behan-Pelletier 2008). The habitat correlation suggests that the pores have some function and that the cuticular canal includes a dendritic extension, but this has not been investigated.

Regarding its origin, Grandjean (1940a) speculated that the pore could be a vestige of either a lost solenidion or an ancestral lyrifissure. He later (1954c) dismissed the first idea, but without explanation. He did not dismiss the second but expressed a lack of confidence in the idea. In fact, lyrifissures are not found on the genu of any acariform mite, so to us it seems an unrealistic explanation. Grandjean implied the pore could be a newly evolved structure, rather than a vestige, but in either case it would be apomorphic, since the pore is not present in the more basal oribatid mite groups (Palaeosomata, Enarthronota, Parhyposomata) or outgroups in Endeostigmata. Regardless of its origin, the mosaic pattern in Mixonomata and Nothrina , and absence in the derived group Brachypylina, suggest that the genual pore was lost multiple times.

Allonothrus provides us with additional insight, if not a certain solution, regarding the origin of these pores. Ermilov & Bakowski (2021) first noticed a small novel ‘seta’ (their d *) on genua I and II of A. malgorzatae , though they did not discuss it. Allonothrus tuxtlasensis has identical structures that we denote σ m, and these are present on other species of the genus, but not all ( Table 2 View TABLE 2 ). We believe each σ m is homologous with a genual pore because the two structures never co-occur, and σ m occupies the same place on genua I and II—posterior to coupled d σ—as does the genual pore in other species (cf. Fig. 3h, k View FIGURE 3 ).

The presence of σ m suggests that Grandjean’s first idea might have been correct: i.e., the genual pore seems to have some historical connection with a solenidion. For several reasons, we believe σ m may represent the atavistic reappearance of solenidion σ 2. First, σ m is isotropic—not birefringent in polarized light—among the known setiform organs of oribatid mites only solenidia are isotropic ( Grandjean 1935; Alberti & Coons 1999). Like solenidia, σ m also is hollow and appears to attach directly to the soft alveolar cuticle, rather than having an inserted solid ‘root’ that would characterize a seta. In σ m we have not seen the transverse striations that typically appear when solenidia (which are multiporous chemosensilla) are viewed with transmitted light, but these striations often are indiscernible in small solenidia. While solenidia rarely are so small, the palaeosomatan genus Aphelacarus has several that seem similar in form to σ m and Grandjean (1954b) considered these ‘vestigial.’

Positional analogy with the tibia can lend support to this idea. Whether in the form of a pore or σ m, the structure on genu I has the same position relative to the coupled d σ that φ 2 has to the coupled d φ 1 on tibia I. Positional similarity among taxa led A. Seniczak et al. (2023) to independently (without reference to Grandjean’s studies) consider the genual pore of Nanhermannia to be a vestigial second solenidion, σ 2: they offered no reason, but as explained by S. Seniczak (pers. comm. with R.A.N., 2024) it has a position similar to σ 2 in Lohmanniidae .

Uncertainties remain, however, and these may have led Grandjean (1954c) to dismiss his original idea of a solenidial origin for the pore. Perhaps most important is that σ 2 and the genual pore are not mutually exclusive in all groups. In the larva of both Perlohmannia and Collohmannia genu I bears a genual pore in addition to two solenidia ( Grandjean 1958, 1966; Suzuki 1977). If their pore is a solenidial vestige, it must be from a third ancestral solenidion. A third genu I solenidion is rare outside Palaeosomata; it occurs in the mixonomatan family Eulohmanniidae , though it forms in the PN ( Norton & Ermilov 2022; Ermilov & Norton 2023). Also, while a genual pore can occur on genu III in addition to the usual σ, no known oribatid mite has a second solenidion on that segment, and to our knowledge only members of Parhypochthoniidae , perhaps the earliest derivative family of Novoribatida, have a second solenidion on genu II. A distant outgroup, the endeostigmatid family Alycidae (e.g., Pachygnathus , Petralycus ; Grandjean 1937, 1943) includes species with two solenidia on both genua II and III, so σ 2 may have existed on these segments in distant oribatid mite ancestors.

Another complication is that the pore may be doubled on genu I. We know of only one published observation ( Grandjean 1958; Perlohmannia ), but we have also seen an isolated example in a large, undescribed species of Epilohmannia (R.A.N., unpublished). We consider such rare doubling to be developmental anomalies, similar to the occasional doubling of setae.

If σ m is a homologue of solenidion σ 2, it is atavistic, representing a reversal that we presume resulted from the removal or inactivation of a long-standing genetic or epigenetic suppression that effected its initial disappearance ( Weigmann 2010), similar to the reappearance of notogastral setae f 1 (see above). This possibility is relevant also to a different phylogenetic issue, the origin of Astigmata from within oribatid mites. While there are different specific hypotheses for their origin based on morphological ( Norton 1998) or molecular (e.g., Dabert et al. 2010; Klimov et al. 2017) data, each infers that σ 2 reappeared on genu I in the ancestor of Astigmata. Norton (1998) mentioned possible examples of solenidial reappearance on tarsi, but the Allonothrus examples gives more credibility to the specific reappearance of σ 2.

10. Shape of tarsal setae

The shapes of ventrodistal leg setae in Allonothrus species —pairs (p), (u), (a) and s —are typical of terrestrial oribatid mites in being attenuate, with finely tapered tips that probably increase adhesion to substrates in a gaseous medium. The same is true of species in Archegozetes , Trhypochthonius and Mainothrus . The latter two genera have species that inhabit damp environments such as bogs and fens ( Weigmann et al. 2015; Behan-Pelletier & Lindo 2023), but they are not aquatic mites. Afronothrus also have normal leg setae and are usually associated with a variety of terrestrial habitats (see Wang et al. 1999 for overview). An unidentified Afronothrus species was considered ‘aquatic’ by Pepato et al. (2022), since abundant adults and juveniles were active in free water held in the leaf axils of a soil-growing bromeliad in Florida (P.B. Klimov, pers. comm. with R.A.N., 16-i-2023), but we suspect this was an opportunistic association.

By contrast, species of Mucronothrus and Trhypochthoniellus are truly aquatic, associated with both lentic and lotic habitats ( Norton et al. 1988; Weigmann & Deischel 2006; Behan-Pelletier & Eamer 2007; Schatz & Behan-Pelletier 2008). The ventrodistal setae on tarsi II–IV are very short, acute or conical, lacking the fine tips of the terrestrial genera. Such setal shapes also characterize Malaconothridae , many of which are aquatic or live in wet organic soils. This has been considered a synapomorphy with Trhypochthoniellus ( Weigmann 1997b) but, as discussed below, molecular data do not support this relationship.