Trhypochthoniidae (Norton, 1998)

Overview of leg setation in Trhypochthoniidae View in CoL

With ontogeny known for at least one species in each genus, we propose the following as traits that could be added to a description of the family. (1) Primilateral setae absent from tarsus I (and all other tarsi). (2) Primiventral setae (pv) absent from tarsi II and III, present on tarsus IV, and with various states on tarsus I (usually present). (3) Antelateral setae (a) usually present on tarsi I–III (rarely the adaxial seta is absent), with various states on tarsus IV. (4) Seta ftʹ absent from both tarsi III and IV. (5) Iteral setae absent from all tarsi. (6) Proximal accessory setae present (usually) or absent; lateral setae usually present on I, absent from II-IV; ventral setae usually absent from I, present on II–IV. (7) Larval tibia I with four setae, including d, lʹ, vʹ and cʺ; larval tibia II with three, lacking cʺ.

Loss of primilateral setae. Grandjean (1959) summarized the distribution of the eustasic primilateral setae in oribatid mites and noted that the typical pattern in Nothrina (his Nothroidea) and Brachypylina (Circumdehiscentiae) is to have pair (pl) present on tarsus I but absent on the other tarsi. Various groups of Brachypylina have lost this pair, but among Nothrina only Trhypochthoniidae and Malaconothridae lack them entirely.

Loss of primiventral setae from tarsi II and III. The most interesting, and perhaps controversial, aspect of leg setation in this family relates to the diminished fundamental setation on tarsi II and III. Most uncertainty relates to what can be referred to as the ‘ventral quintet’: seta s and pairs (a) and (pv). When one or more of these setae are absent it is difficult to be certain which setae remain. Wauthy & Fain (1991, with a contribution by J. Travé) explained the difficulty with regard to Malaconothridae , where three of the quintet are lost from tarsi I–III. Their solution (like that of Knülle 1957) was that setae s and (a) were lost, leaving pair (pv). By contrast, Ermilov & Rybalov (2023) believed s and (pv) were lost, leaving (a). But alternative possibilities exist, since these authors illustrated the ʺ seta (pvʺ or aʺ according to author) as lying on the ventral midline, in the usual position of s. Wauthy & Fain (1991) explained the midline position of the purported pvʺ as the result of rotation (basculation) of the setal verticil. However, s is a very stable seta among oribatid mites and it seems no less likely that the remaining setae are s, pvʹ (or s, aʹ).

Trhypochthoniidae have lost two or three of the setal quintet. In an early, comparative application of his chaetotaxic model for leg setae, Grandjean (1941a) noted that larval tarsus II of Trhypochthonius tectorum lacks one pair of setae that occur in related families (it is also true of III, but this tarsus was not discussed); he was uncertain but believed that the missing pair was (pv). We agree with his assessment and believe all members of the family share this loss on tarsi II and III.

Overall, the loss of primiventral setae seems to be quite rare. To our knowledge, aside from Trhypochthoniidae — and Malaconothridae in the above-noted opinion of Ermilov & Rybalov (2023) —the only losses of (pv) reported in the literature relate to: their partial or complete regression in a small group of protoplophoroid Enarthronota ( Grandjean 1946c; Norton et al. 1983); their absence from tarsus II (only) in Nehypochthoniidae ( Norton & Metz 1980) ; their reported absence from all legs of a species of Parapirnodus ( Scheloribatidae ; Behan-Pelletier et al. 2002); and their absence from tarsus IV of Epilohmannia cylindrica as interpreted by Grandjean (1946b).

In Trhypochthoniidae , the key to identifying the missing setae as pair (pv) lies in comparing the spatial relationships of the quintet— s, (a), (pv)—on tarsus I of the larva, where they typically are all present, to those on tarsi II and III, where one pair is absent. It is perhaps most clear in Archegozetes longisetosus . On larval tarsus I setae (pv) have their typical position—on the ventral face and distinctly proximal to s —while setae (a) also have their typical position, low on the lateral face and slightly distal to s. The same features of s and pair (a) are true on larval tarsi II and III, but no ventral setae are proximal to s, so clearly it is pair (pv) that are absent in this species. It is also clear in Trhypochthoniellus longisetus , where on larval tarsi I–III pair (a) are at the level of s or slightly distal to it; since there are no setae proximal to s, pair (pv) are undoubtedly absent. The other trhypochthoniid genera differ in that s is slightly distal to pair (a) on larval tarsus I, with (pv) placed normally, proximal to these three setae; on tarsi II-III, the relationship of s and (a) is similar to that on I, but no setae are proximal to (a). On tarsus I pair (pv) have a slightly but distinctly more ventral position than do (a), and in the adult (pv) align well with the ventral accessory setae, whereas pair (a) do not.

Loss of ftʹ from tarsus III. Grandjean (1941a, p. 39) noted the deficiency of one fastigial seta on tarsus III in the three genera of Trhypochthoniidae he had studied ( Archegozetes , Trhypochthonius and Trhypochthoniellus ), and it appears to be a feature of the entire family. While he did not identify the specific missing seta (perhaps because the remaining seta is close to the dorsal midline), clearly it is ftʹ that is absent. 8 This mirrors the common and nearly ubiquitous situation on tarsus IV of oribatid mites ( Grandjean 1946b). We know of no other Nothrina that have lost ftʹ from tarsus III, but the loss occurs in some Enarthronota and Ptyctima.

Loss of iteral setae. Grandjean(1961, 1964a) reviewed the distribution and ontogeny of the accessory, amphistasic iteral pair (it) among oribatid mites. The vast majority possess iteral setae on one or more legs, but their complete absence is not rare. Since they always are post-larval, their disappearance is easily explained as paedomorphic—i.e., retention of the larval state. Among Nothrina, Grandjean indicated their absence from those Malaconothridae and Trhypochthoniidae known to him, and this pattern remains true after the subsequent addition of much more data. He also considered iteral setae absent from Nothridae , but he had studied only species of Nothrus ; iteral setae have been reported in the more plesiomorphic genus Novonothrus ( Casanueva & Norton 1997, 1998). 9

Setal ontogeny of trochanter III

Among Oribatida , trochanter III most often bears two setae, vʹ and lʹ. The instars in which these setae first appear are usually constant within a species but are quite varied among species. Each may appear in any instar except the larva, their sequence can be reversed, and in rare cases one or both never appear at all (see references on leg setal analyses in Norton & Ermilov 2014, 2024).

What seems to be the ancestral oribatid mite pattern is for v ʹ to form in the PN and lʹ in the DN. This is the principal pattern found in Palaeosomata, Enarthronota, Parhyposomata and some Mixonomata, as well as the outgroup Endeostigmata (e.g., Petralycus ; Grandjean 1943). The pattern is known from all families of Nothrina except Crotoniidae but in each family there are other variants.Among Brachypylina it is the typical (but not exclusive) pattern in Neoliodoidea, Damaeoidea and Ameroidea and is found in some Plateremaeoidea and Cepheusoidea. But most brachypylines, including all the poronotic superfamilies, have different (presumably derived) patterns.

Among Trhypochthoniidae , the ancestral pattern is found in Archegozetes and Trhypochthonius , but three variant ontogenies are known. In Mainothrus and Allonothrus seta vʹ is delayed one instar, so that both setae form in the DN. In Afronothrus and Trhypochthoniellus vʹ is delayed further, to the TN, thereby reversing the ancestral sequence. 10 Mucronothrus presents a third variant, in which lʹ is PNal and vʹ DNal, also reversing the ancestral order; if our concept of the ancestral pattern is correct, this would represent a derived ontogenetic acceleration of lʹ. It is rare for lʹ to form in the PN but both setae are PNal in Perlohmannia dissimilus ( Grandjean 1958) and Eulohmanniidae species ( Norton & Ermilov 2022). Seta lʹ is not known to form earlier than the DN in more basal groups, such as Enarthronota and Parhyposomata.

Setal priorities on genua III and IV

In a special study of the priority—i.e., resistance to regressive loss ( Grandjean 1941b)—of setae on genua of Acariformes, Grandjean (1942a) determined that ancestrally the fundamental chaetome of genua III and IV are identical. In particular, setae d and lʹ are both present when the legs first form. This was evident from studying Endeostigmata, where it seems to be the rule ( Grandjean 1942b, p.105). From studying the ontogeny of a wide range of oribatid mite taxa, Grandjean (1942a, p. 50) generalized that the usual priority of setae on genua III and IV in this group was (in decreasing order) d, lʹ, vʹ, vʺ, lʺ. Over evolutionary time, setal regressions (ontogenetic delays or losses) acting on these genua would first affect the end of the list, and then move toward the beginning.

With much more data available now (see leg setal analysis listings in Norton & Ermilov 2014, 2024), the priorities largely still hold for genu III— particularly with regard to lʹ and vʹ —even though there is great variation in the chaetome and the specific ontogeny. We have noted a couple exceptions. In Nanhermannia coronata seta lʹ is lost while v ʹ remains (Seniczak & Seniczak 2023). More striking, in some Oribatulidae —e.g. Zygoribatula exilis and Phauloppia nemoralis —it is vʹ that is fundamental, with lʹ lost ( Ermilov & Kolesnikov 2012; Ermilov et al. 2015).

By contrast, genu IV, which usually forms no setae until the DN, is less stable in this regard. Grandjean (op. cit.) noted several exceptions to the ancestral priority on genu IV, including one genus in each of three major groups, in which v ʹ forms in an earlier instar than lʹ— Palaeosomata ( Aphelacarus ), Parhyposomata ( Parhypochthonius ) and Nothrina ( Nanhermannia ) 11 — but now we see it is more widespread. This derived, inverted priority is found on genu IV of the parhyposomatan genera Elliptochthonius and Gehypochthonius and the mixonomatan Nehypochthonius , in which lʹ is lost but vʹ remains. In Collohmannia lʹ is variable and can form after both vʹ and vʺ. While most Brachypylina retain the ancestral priority of lʹ, vʹ on genu IV, in Eremaeidae and Megeremaeidae vʹ forms while lʹ is lost, and vʹ can even be fundamental.

Among Nothrina , Trhypochthoniidae are unique in showing a consistent pattern of reversed priority of lʹ and vʹ on genu IV. 12 Either vʹ precedes lʹ during ontogeny ( Afronothrus , Archegozetes , Mucronothrus , Trhypochthonius ) or vʹ forms while lʹ is lost ( Mainothrus , Trhypochthoniellus ). In the most striking example, vʹ even takes the place of lʹ as being fundamental on genu IV of Archegozetes . Only in Allonothrus is this reversal equivocal, since both lʹ and vʹ form together in the TN.

Homology and analogy in the chaetotaxy of tibiae I and II

Grandjean (1940a, 1940b, 1954b) demonstrated that tibiae I and II of adult oribatid mites have an ancestral setation forming a whorl of seven setae. Originally proposed as unpaired d, and pairs (ls), (li), and (st), these notations were simplified to d, (l), (c), and (v), when homologies with the more common vertical of five setae— entirely lacking (c)—were established ( Grandjean 1954b). 13 Outside of Palaeosomata, cʹ does not occur and the maximum (except for neotrichous taxa) is an adult chaetome of six: d, (l), (v), cʺ.

For any member of Nothrina with known ontogeny, the presence of six setae on adult tibia I or II indicates that the larva has four setae: d, lʹ, vʹ, and cʺ, with lʺ and v ʺ always being accessory, i.e., added during ontogeny ( Fig. 10 View FIGURE 10 ; square box). This is the case for tibia I of Allonothrus , for example (Appendix 1). However, cʺ can be present in the larva even if the adult chaetome is only four or five setae: it only means that one or both of the possible accessory setae fail to form, as occurs in all other genera of Trhypochthoniidae ( Fig. 10 View FIGURE 10 ). On tibia II of Trhypochthoniidae , cʺ never forms, but the accessory setae (lʺ, vʺ) show the same range of states as they do on tibia I. Under such circumstances, a species may have the same number of setae on tibiae I and II while their exact compositions differ (e.g., Trhypochthonius , Mucronothrus ).

Like many tarsal setae, cʺ on tibiae I and II is both fundamental and eustasic: if not present in the larva, it never forms during ontogeny. Because of its position in the lower third of the posterior face, Grandjean (1940a; 1941b) recognized that cʺ (= liʺ) is easily mistaken for accessory seta lʺ or vʺ, if one or both of these latter setae do not form. Therefore, with regard to homology, a tibia I chaetome of five does not directly translate to a standard verticil of five— d, (l), (v). He gave examples of chaetomes of five in which cʺ appeared to ‘imitate’ lʺ and others where it masqueraded as vʺ. If notations based on homology are the goal, then the literature is replete with errors or inconsistencies. Below, we use examples to examine these issues in a larger taxonomic context, attempting to identify general evolutionary changes to the chaetome of tibiae I and II.