Two unusual new species of Caleremaeus (Acari: Oribatida) from eastern North America, with redescription of C. retractus and reevaluation of the genus

The oribatid mite genus Caleremaeus (Caleremaeidae) is widely distributed in the northern hemisphere but has been represented by only three extant and one fossil species. We redescribe the North American C. retractus (Banks, 1947) based on adults and nymphs; it is distinguishable from the European type species, C. monilipes (Michael, 1882) by its smaller adult size and minor differences in cuticular structure, and by the elongated, tapered form of seta h1 in nymphs. Two new species are proposed: C. nasutus n. sp. from forest soil in Alabama is unique in having adults with a large anterior rostral lobe (juveniles unknown) bearing lamellar setae; the arboreal C. arboricolus n. sp. from eastern USA and Canada is unique among described extant species in having adults with femoral saccules, a transverse ridge bearing lamellar setae and relatively large notogastral setae, and juveniles with a bothridial seta similar to that of the adult. Based on all available data, Caleremaeus is redescribed and considered the sole genus in Caleremaeidae. The higher classification of the family is reviewed, and past placement in Ameroidea is rejected in favor of the monofamilial Caleremaeoidea.


Introduction
Caleremaeus Berlese, 1910 (Caleremaeidae) is a distinctive genus of middle-derivative, brachypyline oribatid mites with adult traits that make identification rather easy, even at modest magnification: enlarged first tibiae (see epigram) and a notogaster with a T-shaped pattern of two conspicuous dorsal bulges separated by a foveate sulcus. Only three extant and one fossil-based Caleremaeus species have been named, collectively distributed in the western Palaearctic and eastern North America.
The extant type species, Caleremaeus monilipes (Michael, 1882), is widely distributed in Europe but it seems absent east of the Ural Mountains (Krivolutsky et al. 1995). We For comparison and preparation of the generic redescription, we studied European specimens of C. monilipes in the first author's collection, including adults from Germany (near Berlin), Spain (Barcelona) and Sweden (Bohuslän), and a single deutonymph accompanying the adults from Sweden. Several specimens of Hungarobelba visnyai Balogh, 1938 from Poland were examined for comparison.
Sources and depositories for specimens include the following: CNC -the Canadian National Collection of Insects, Arachnids and Nematodes, Agriculture and Agri-Food Canada, Ottawa, Ontario, Canada; MCZ -the Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts, USA; RNC -the personal collection of Roy A. Norton, Syracuse, New York, USA; USNM -the National Museum of Natural History, Smithsonian Institution, Washington, DC, USA (currently housed with the U.S. Department of Agriculture collections in Beltsville, Maryland).
Preparation and documentation -Most observations and data are from specimens temporarily mounted in cavity slides in a medium of lactic acid diluted with water (2:1; Grandjean 1949). Dissected mouthparts, legs and fragments of body regions requiring close study were permanently mounted in Hoyer's medium for observation with oil-immersion lenses. Compound microscopy employed bright-field, polarized, and Nomarski (DIC) illumination using a Nikon Eclipse E800 compound microscope; line drawings were made with the aid of a drawing tube. Light micrographs were obtained, usually as image stacks, with an AmScope MU800 digital camera mounted on the compound microscope. Image stacks were combined using the Helicon Focus Pro (v. 5.0) suite; the stacks varied widely in number of individual images, usually only several for highly magnified (1000 x) images and 15-30 for lower magnifications. As needed, images were adjusted with Adobe Photoshop (CS3) for contrast and color balance.
Specimens for scanning electron microscopy (SEM Quanta 600 FEI Company TM, Brno, Czech Republic) were removed from alcohol and cleaned by soaking in Terg-a-zyme® solution for 6-12 h, followed by brief (1-2 s) submersion in an ultrasonic bath. Specimens were critical-point dried using the EM CPD300 (Leica Microsystems, Vienna, Austria), mounted on Al-stubs with double sided sticky tape, and gold-coated in a Hummer sputter apparatus.
Taxonomic context -The general context is the classification used by Norton and Behan-Pelletier (2009) and Schatz et al. (2011), except as noted. A principal digression relates to our use of the superfamily names Eremaeoidea and Zetorchestoidea: these names are fully or partly synonymous according to different concepts in the literature and we use them both below, according to context. Herein, Eremaeoidea is used in the sense of Balogh and Balogh (1992)-including only Eremaeidae and Megeremaeidae-which is how it also seems to have been perceived by Franklin and Woas (1992). Recent classifications (e.g. Schatz et al. 2011) have grouped these families with Zetorchestidae under the older synonym Zetorchestoidea, but the molecular study of Lienhard et al. (2013) casts doubt on the monophyly of such a taxon. Author and date for species-group taxa are given at the first use of the name; those of supraspecific taxa can be found in Subías (2004).
Terminology and conventions -Morphological terminology is mostly that of F. Grandjean (see Vachon 1975 for references, Norton 1977 for leg setal nomenclature and Travé et al. 1996 or Norton andBehan-Pelletier 2009 for overview). Terms are translated from French (Hammen 1980) but Grandjean's original abbreviations and figure notations are often retained. Paired structures are described in the singular unless noted otherwise. Throughout, there are references to numbered 'Remarks on morphology' that conclude the text; each reference is parenthetic, in the form '(R1, R2, etc.).' However, several specific terminology issues are explained here.
Surface sculpturing. Impressed surface-sculpture is referred to as foveate if circular depressions are relatively large and separated by less than their diameter; foveolate sculpture refers to circular depressions that are relatively small and separated by more than their diameter; scrobiculate refers to closely-spaced, parallel elliptical depressions (Harris 1979). Projecting structures are referred to as knots if relatively small, simple, dome-shaped, or tubercles if relatively large and conspicuous, especially if conical, triangular or tooth-like.
Lamella and tutorium. Like Grandjean (1965b), we interpret the two pairs of longitudinal ridge-like structures that are variously developed on the prodorsum of Caleremaeus species as homologues of the lamella and tutorium. We refrain from referring to the middle pair of ridges in Caleremaeus as 'costulae'. The latter term appears to have originated in the first of several important synopses by Balogh (1961); he introduced a simplistic dichotomy between a costula (rib-like) and a lamella (blade-like) that still troubles the descriptive literature (Norton and Behan-Pelletier 2009). Clearly, both lamella and tutorium originated as longitudinal ridges, then hypertrophied to flattened, projecting blades in derived groups. The plesiomorphic form is retained in Caleremaeus and Megeremaeidae (see R1). When applied to the prodorsum, we believe the term costula is best restricted to describing simple, secondary supporting ridges in taxa that plesiomorphically lack lamellae, as in Oppioidea, for example.
Measurements and counts -Body length was measured in dorsoventral aspect, from the tip of the rostrum to the posterior edge of the hysterosoma; width refers to the maximum hysterosomal width in dorsal aspect. Measurements of specific structures or distances are given either as a single number meant to be representative of an average-sized individual, or an estimated range taken from a small sample of several individuals; setae were measured when oriented in a single observational plane (e.g., perpendicular to the surface). Setal and solenidial formulas represent counts per segment for appendages (from leg I to IV; famulus included for tarsus I); epimeral setation is given as number of pairs per podosomal segment (I-IV).
In locality data, counties within USA states and Canadian provinces are indicated by Co. Berlese (1910) provided no diagnosis when Caleremaeus was proposed, but simply referred to the type species, C. monilipes. In early synopses (e.g. Sellnick 1928, Willmann 1931, genus-level characters were not distinguished. Nor did Grandjean (1965b) distinguish generic traits when he gave an extended diagnosis for the newly proposed Caleremaeidae, based only on C. monilipes. Generic characters seem not to have been isolated until Bulanova-Zachvatkina (1975) listed a single generic trait: the unique notogastral topography. Norton (1978) proposed including Veloppia as a second genus of Caleremaeidae and considered Grandjean's (1965b) diagnosis of this family as a diagnosis of Caleremaeus. Subsequently Subías and Arillo (2001) and Weigmann (2006) distinguished Caleremaeidae and Caleremaeus with brief diagnoses. Below, we use the latter three works and our own observations to propose a new diagnosis and expanded description of Caleremaeus. Parts relating to adults are based on all species known to us, both described and undescribed; those relating to nymphs are based on only C. monilipes, C. retractus (sensu stricto; additionally the 'retractus' group from New York) and C. arboricolus; those relating to the larva are from C. arboricolus and the 'retractus' group from New York, as well as the larval exuvial scalp of C. retractus, and literature data on C. monilipes. For C. monilipes, published information on adults is from references given above (Introduction), while that on juveniles is mainly from Michael (1882), Grandjean (1965b) and Seniczak and Seniczak (2019; but see R12).

Redescription of Caleremaeus
Caleremaeus Berlese, 1910 Type species: Damaeus monilipes Michael, 1882 (p. 16). The original combination often has been given in the literature as Notaspis monilipes, but this was a later recombination by Michael (1888). Etymology - Berlese (1910) did not indicate the etymology of Caleremaeus, but the stem eremaios is Greek (meaning solitary) and is the basis for the older genus name Eremaeus Koch, 1835. The prefix 'cal', if also based on Greek, would be from kalos, meaning beautiful. While less likely, Berlese might have mixed languages: if 'cal' were Latin-based it could relate to the habitat of the type species C. monilipes, which Michael (1882) collected from rotten wood (Latin: cala, piece of wood).
Diagnosis -Brachypylina with small to medium-sized adults (length 306-475 µm), overall shape elongate-pyriform in dorsoventral view. Integument with enveloping cerotegument, usually with dense, dome-to mushroom-shaped excrescences; sclerotized procuticle partly foveate to foveolate. Prodorsum with or without paired, ridge-like lamella and tutorium; with or without prodorsal enantiophysis. Rostrum with strongly developed submarginal crest. Bothridial seta with basal stalk and flattened, expanded head. Dorso-and pleurophragmata absent. Notogaster without porose organs; anterior margin nearly straight, with small dentate tubercles or knots; with distinct humeral process opposing tubercle(s) on posterior wall of bothridium to form humeral enantiophysis; with strong topography consisting of relatively flat lateral region and two strong bulges (transverse anterior bulge and longitudinal posterior bulge) separated by foveate transverse sulcus; with 10 pairs of setae, marginal to submarginal. Pedotectum I present, II absent; propodolateral apophysis absent; discidial ridge usually present, distinct discidium present or absent; circumpedal carina absent; lateral, parastigmatic and aggenital enantiophyses present; coxisternum with distinct medial fossa between setae 4a. Subcapitular rutellum atelobasic . Legs relatively short, tibiae I, II unusually large, with narrow  basal stalk and swollen distal bulb, I with dorsodistal process; pretarsi monodactylous; seta d  absent from genua I-III and all tibiae; iteral setal pair present on tarsi I-II, only it on III, none  on IV. Nymphs plicate, eupheredermous, gastronotum with papilliform attachment cornicle; setal pair h 1 adjacent on extension of pygidial sclerite; paraprocts atrichous in larva, proto-and deutonymph.

Figures 6-7, 12
Facies, cuticle -Preserved specimens colorless to light tan. Unsclerotized regions of gastronotic cuticle plicate, except smooth underneath exuvial scalps of nymphs; plicae generally vertical laterally, vaguely circumferential around opisthonotal gland opening and paraprocts. Hysterosomal line of dehiscence not evident. Cerotegument distinct, enveloping, with short excrescences of various sizes generally similar to those of adults or merged into irregular masses.
Prodorsum -With several ridges or folds, longitudinal laterally and transverse medially; one of latter bearing setal pair le. Rostrum usually truncate anteriorly (Fig. 7F). With normal setation, all but bothridial seta (bs) short to medium length. Bothridium and bs fully developed in all instars; bothridium projecting, cylindrical to slightly funnel-shaped, thin-walled, without tracheal organ; bs similar to that of adult, or narrower and proportionally longer.
Anogenital region -Genital setation varies with species: protonymph with one pair, deutonymph with two or three, tritonymph with four or five. Aggenital seta first formed in deutonymph. Paraprocts atrichous in larva, proto-and deutonymph: setal rows p, ad, an first formed in proto-, deuto-and tritonymph, respectively. Ontogeny of cupules normal, cupule added with respective setal row; ian formed in tritonymph but half size of others (about diameter of setal alveolus), lost in adult.
Gnathosoma -Similar to that of adult, except for weaker sclerotization of subcapitulum making labiogenal articulation indistinct. Palp femoral seta inf first formed in protonymph.
Legs -With size proportions as in adult but most segments somewhat more tubular; tibia I with strong, cylindrical dorsodistal apophysis. Ontogeny of leg chaetome given in Table 1, with salient features as follows. Seta d present on all genua and tibiae, coupled in same alveolus with respective solenidion on all but genu IV; except on tibia I, seta and solenidion very short, equal in length or d very slightly longer (Figs 7N, 12J, 18D), solenidion difficult to see in some orientations. Seta d and φ 1 of tibia I both long, inserting on dorsodistal apophysis; φ 1 flagellate, d 2/3 to 3/4 as long, with barbs or cilia having clear, velum-like coating of various distinctness (Figs 7L, 12I). Protonymphal leg IV setation 0-0-0-0-7, with typical tarsal complement of (p), (u), (pv), ft″. On tarsus I, setae (p) appear to be eupathidial from larva (uncertain in early instars), but s formed as normal seta, becoming eupathidial only in adult; s located proximal to pair (a) in nymphs, but displaced distal to them when eupathidial, as usual. Ontogeny of iteral setae complex: pair forms in tritonymph on tarsus I and in adult on II, with solitary it″ tritonymphal on III (R7, R12).
Type locality -The type specimens (below) derived from a study of soil animals of the Duke Forest (Durham Co., North Carolina) by Pearse (1946). The Duke Forest is somewhat fragmented and, since four different locations and habitats were studied, the exact origin of the types is unknown.
Type material - Banks (1947) reported two specimens in the type series. The holotype (original designation as 'type') is a slide-mounted specimen in the arachnid collection of the MCZ. The label bears the following data in Banks' handwriting: 'Duke Forest N. Car.; #475; Pearse; Carabodoides retracta Bks; 3018 type.' The mite is broken by crushing but has an estimated total length of about 320 µm. The second specimen is a paratype in the mite collection of the USNM. It is a gravid female, also broken, with an estimated original length of 340 µm and with the following label data in Banks' handwriting: 'Duke Forest, N.C. 238, Pearse, Jan. Carabodoides retracta Bks. Paratype.' The measurements contrast significantly with the '.55 to .6 mm' reported by Banks (1947).

Trochanter
Femur Adult data are based on all studied species and populations, including C. monilipes from Sweden, Germany and Spain. Ontogenetic data are from C. arboricolus n. sp. (topotypical population), C. retractus (near-topotypical population), a species in the 'retractus group' (New York population; see text), and C. monilipes (Norwegian population; see Seniczak and Seniczak 2019 and corrections in R12); from each population all instars were studied, except no larva was available from the C. retractus near-topotypical population. 2 Setae (Roman letters) and solenidia (σ, φ, ω ) are shown where they are first added and are assumed present through the rest of ontogeny, unless noted in brackets. Setae in parentheses represent pseudosymmetrical pairs; dash indicates no addition; underline indicates solenidion is coupled to seta d , in same alveolus.
Legs (Figs 4-5) -Femora I, II similar in form: with abrupt transition from proximal stalk to bulb, junction at nearly right-angle; femur I~2.5, II~2.0 times longer than high in lateral view, stalk occupying~0.4 femoral length; stalk of femur II slightly broader than that of I. All femora with porose area, mostly on adaxial face of bulb. Tibia I with bulb markedly swollen, only~1.2 times longer than high. Tarsus I abruptly tapered in distal half, but without distinct projecting mid-dorsal bulge. Tarsus II without noticeable proximal stalk, depth similar to that of tibia in lateral view. Seta d of femora short, flame-shaped, similar in structure to dorsal body setae (pigmented, barbed, with conspicuous cerotegument nodule at base; Fig. 5D). Seta l of genu and tibia I not conspicuously enlarged.

Comparisons
Adults of Caleremaeus retractus are similar to those of C. monilipes (cf. Weigmann 2006, Miko and and C. divisus in having a distinct lamella and tutorium and lamellar seta inserted on a strong cusp, traits that are absent from C. arboricolus and C. nasutus. According to the crude illustration of C. divisus, the cusp stops well short of the rostral margin, while reaching or surpassing it in C. monilipes and C. retractus; also, the anterior notogastral tubercles of C. divisus were drawn as far larger than those of other species, equal in size to the humeral process (Mihelčič 1952).
Adults of Caleremaeus retractus are distinguishable from those of C. monilipes (Cm) by: (1) having a less sculptured prodorsum, including weakly-developed lamella (stronger in Cm); (2) having nearly erect setae in, the pair separated by 4-5 times setal length (in slightly larger, curved mediad and separated by less than three times setal length in Cm); (3) lacking additional distinct knots or small tubercles along epimeral groove 2 and along the anterior edge of the sejugal groove between tubercle pair Sa (present in Cm); (4) having a modest humeral process that rarely reaches anteriorly to overlap bothridial tubercles (larger humeral process, overlapping with bothridial tubercle in Cm); (5) being smaller, with adult total length 306-340 µm (373-475 µm in Cm). Nymphs are distinguishable by: (1) the narrow, elongate form and closely parallel orientation of setae h 1 in nymphs (distinctly clavate and divergent in Cm; Fig.  18C); (2) having generally smaller, less conspicuous leg setae (generally larger in Cm; cf. setae l in Figs 7M, 18D); having smooth, small, inconspicuous seta ro (larger, conspicuous, with several strong barbs, projecting distinctly forward beyond rostral margin in Cm; Fig. 18D, see also Michael 1882).

Possible species group
As noted in the Introduction, morphometric and genetic evidence suggests that European records of Caleremaeus monilipes represent a complex of species (Krisper et al. 2017) that ultimately may be considered the 'monilipes' species-group. The same may be true of the most similar North American species, C. retractus.
At an early stage of this study, we identified specimens of C. retractus from many locations in eastern North America, including the states of Alabama, Arkansas, Florida, Georgia, Louisiana, Illinois, Indiana, Mississippi, New Hampshire, New York, Vermont, Virginia and West Virginia, as well as the Canadian province of Quebec. Adults of these specimens are presently indistinguishable from those at the type location in North Carolina, except perhaps for a propensity of New York adults to have one or two small knots across epimeral groove 2 from seta 2a (Fig. 18E). But juveniles suggest that more than one species is involved. In addition to the near-topotypical material from Durham Co., North Carolina, we have juveniles from Florida and from New York (Onondaga Co.). The Florida nymphs are identical to the near-topotypes, but New York juveniles are easily distinguished by their shorter, straighter pygidial setae ( Fig.  18G; h 1 ), larger seta l on tibia and genu I, and tibia I seta d with more distinct barbs and velum-like coating. As with near-topotypes, setae h 1 of the New York population are adjacent, parallel and consistently broken from exuviae, unlike those of C. monilipes and C. arboricolus in which they are divergent and retained on exuviae.
The differences could represent geographic variation in these setae, but without further knowledge of juveniles from other locations, and especially without genetic data, we have no basis for judgement. At present, we prefer to assign specimens with the adult traits of C. retractus to a 'retractus' species group unless juveniles are known and correspond with the near-topotypes.

Comparisons
Adults of C. arboricolus are distinguishable from those of all described extant Caleremaeus species in: having a prodorsum lacking enantiophysis eA and having the dorsosejugal tubercle
(dt) represented instead by a linear series of usually three knots; having a notogaster with a series of knots outlining the posterior bulge and setae that are conspicuous, squamose (except p 2 , p 3 ); and saccules on leg femora instead of porose areas.
The most similar described species is C. gleso, which is known only from Baltic amber. Based on Sellnick's (1931) description and sketchy dorsal illustration of an adult, C. gleso shares with C. arboricolus the absence of distinct prodorsal cusps extending forward from a transverse ridge and notogastral setae that are clavate, at least in part. C. gleso differs from C. arboricolus in having: a prodorsum with larger, more clavate lamellar setae; a more sculptured central prodorsum (perhaps with more defined lamellae); and a depression on either side of the transverse sulcus. The notogaster was described as having foveate anterior and posterior bulges, but the sclerotized cuticle of these structures is smooth in all known extant species. Considering the difficulties of observing small amber inclusions, Sellnick may have mistaken the round cerotegument excrescences for 'pits', but this can be confirmed only if the species is rediscovered.
Legs -Femur I~2.7 times longer than maximum height in lateral view, with sharp, nearly right-angle transition between stalk and bulb (Fig. 17A). Femur II (Fig. 17B) similar, but slightly shorter (~2.5); stalk occupying about third length of femur I, about quarter that of femur II. All femora with porose area, mostly on adaxial face of bulb. Tibia I~1.3 times longer than high in lateral view. Tarsus I abruptly tapering in distal half, but without distinct projecting mid-dorsal bulge (Fig. 17C). Tarsus II without noticeable proximal stalk, depth similar to that of tibia in lateral view. Seta d of femora short, flame-shaped (Fig. 17B), similar in structure to dorsal body setae (pigmented, barbed, with conspicuous cerotegument nodule at base). Seta l of genu and tibia I not conspicuously enlarged.

Comparisons
Adults of C. nasutus are unique among described Caleremaeus species in having a prodorsum with a large, tongue-shaped anterior lobe, bearing the lamellar setae. Otherwise, they share several features with C. monilipes and C. retractus, including the presence of enantiophysis eA, the presence of an embossed pattern on the ventral face of the rostral lobe and rostrophragma,   and ridges laterally on epimeres III-IV that outline depressions coapted to the respective trochanters. The form of dorsal setae (with basal cerotegument nodule and pigmented, barbed distal portion) is shared with C. retractus, but not C. monilipes.

Notes on biology
Reproduction -Gravid females of Caleremaeus species in this study carried a maximum of two eggs (one in each oviduct) but a single egg was most common. Eggs are slightly flattened unilaterally, about 1.8 times longer than broad (Fig. 5G). Females were more abundant than males, but all species seem bisexual; males accounted for about one-third to one-half of adults in our samples. This is consistent with data for C. monilipes presented by  but Grandjean (1941) reported a slight male bias (1: 0.8) for this species.
Food -Overall, species of Caleremaeus appear to be opportunistic feeders on fungi and decaying plant organic matter, as is typical of oribatid mites (Schneider et al. 2004). Based on very limited and unquantified information, adults and juveniles of both C. retractus (Fig.  19A-E) and C. arboricolus (Fig. 19F-H) ingest diverse fungal material-both hyphae and spores-and it dominated most boli and pellets we observed. Most other material was not identified but appeared to comprise fragments of plant-based organic matter. Boli and pellets in adults of C. nasutus (Fig. 19I-L) were generally like those of other species.
Despite its wide distribution, little has been written on the feeding biology of C. monilipes. Skubała and Maslak 2010 regarded C. monilipes as a xylophage, but neither of the two cited references (Luxton 1972;Schatz 1983) made such a claim. Fischer et al. (2010) found C. monilipes adults from tree trunks to have stable-isotope signatures more in line with those of soil-dwelling species, rather than lichen feeders; they speculated that they eat algae or other resources available on bark.
Habitat associations -Collectively, Caleremaeus species clearly show some affinity for above-ground microhabitats, primarily in forests. Caleremaeus nasutus and C. retractus (sensu stricto) have not been found outside soil microhabitats, but this may be due to a lack of appropriate sampling where these mites occur. Based on material in the CNC and RNC, members of the 'retractus' group have been collected from diverse non-soil substrates, including arboreal and saxicolous lichens and mosses, bark of both coniferous and deciduous trees (including cankered bark of chestnut trees), rotting wood of logs and tree-holes and decaying woody fungal sporocarps, in addition to soil-litter. Caleremaeus arboricolus is consistently associated with arboreal microhabitats, including bark, twigs and lichens; the single non-arboreal collection was from a regurgitated owl-pellet lying on the soil surface. The similar C. gleso, known only as a Baltic amber fossil, was almost certainly arboreal since trees were the source of original resin-flows. While little is known of C. divisus, it has been reported only from arboricolus moss.
There are scattered reports of Caleremaeus monilipes being collected in low to moderate density from forest soil-litter (e.g. Moraza and Peña 2005), but the original collection was from decaying wood (Michael 1882) and woody substrates seem to be the primary microhabitat, particularly logs in early stages of decomposition (Siira-Pietikainen et al. 2008;Skubała and Maslak 2010 and included references; or rot-holes in standing trees (Skubała and Gurgul 2011;Taylor and Ranius 2014). Other works suggest less fidelity to dead woody substrates. Travé (1963) considered it a predominantly saxicolous species that could also be found on trees. Subías (1977) considered it a muscicole, found with similar frequency on rocks and low on tree trunks; Fischer et al. (2010) also noted the lower-trunk affiliation. By contrast, Arroyo et al. (2013) found moderate numbers in oak-canopy mosses, and  reported an abundance on oak-trunk moss above 1.5 m, but absence from lower regions. Moss substrates also were indicated by Schweizer (1957) and Bonnet et al. (1975). Odd reports (at least some clearly related to transported substrates) include discoveries in nests of mice, birds and ants (Lebedeva and Poltavskaya 2013;Krawczyk et al. 2015;Elo et al. 2018). It remains to be determined how much the microhabitat diversity of C. monilipes reflects possible cryptic speciation (see Introduction).

Family-group classification
Composition Caleremaeidae Grandjean, 1965b, was monogeneric when proposed, but as many as six other genera have been added since, according to author; these include Veloppia and five genera usually grouped as Anderemaeidae (Anderemaeus, Cristeremaeus, Epieremulus, Luxtoneremaeus, Yugaseremaeus). We believe that none of these additions (or others that have been informally suggested in unpublished internet documents: i.e. Megeremaeus, Caucaseremaeus; Subías 2019) can be supported in a phylogenetic context, and that the family should remain monogeneric.
The removal of Veloppia will be discussed in detail separately (Norton et al., in preparation). Woas (2002) doubted its inclusion in Caleremaeidae and perceived a closer relationship to Hungarobelbidae. With increased knowledge, we view the various adult similarities proposed by Norton (1978) as more widespread and possibly symplesiomorphic, and juvenile traits clearly show that Veloppia is not among the plicate eupheredermous groups.
The idea that the southern-hemisphere genus Anderemaeus and some other members of Anderemaeidae are confamilial with Caleremaeus recently was discussed and rejected by Norton and Ermilov (2019). In summary, inclusion of Anderemaeus in Caleremaeidae was first formalized by Woas (2002), but it had roots in earlier studies Woas 1992, Subías andArillo 2001). By transferring the type-genus, Woas (2002) effectively subsumed Anderemaeidae within Caleremaeidae sensu lato. Subías (2004) followed this proposal but expanded it by including all genera of Anderemaeidae. Woas' proposal was based on three shared traits, paraphrased here: the presence of a lamella-tutorium complex; the presence of the aggenital enantiophysis (e4, U); and a tubular preanal organ. Norton and Ermilov (2019) considered these to be symplesiomorphies of the relevant taxa, and instead showed that Anderemaeus-and therefore its family, Anderemaeidae-are part of a clade of Brachypylina that is more highly derived than Caleremaeus. Based largely on the morphology of juveniles, they transferred Anderemaeidae to Gustavioidea. Grandjean (1965b) proposed Caleremaeidae as part of a reorganization of what he called the 'groupe E restant', i.e. remaining eupheredermous genera that, due to insufficient knowledge, were not treated in his iconic classification of oribatid mites (Grandjean 1954a). He included six of the families in the newly proposed Eremuloidea (= Ameroidea; R10) but left Caleremaeidae, Eremaeidae and Tenuialidae without higher assignment. Since then, Caleremaeidae most often has been included in Oppioidea (e.g., Balogh 1972;Bulanova-Zachvatkina 1975;Marshall et al. 1987;Fujikawa 1991;Balogh and Balogh 1992), but as concepts of that superfamily became more restricted, Caleremaeidae was omitted (e.g. Behan-Pelletier 1991). Subías (2004) deconstructed the traditional Oppioidea and included his broad sense of Caleremaeidae (see above) in a newly recognized Eremelloidea, but none of the other included families is known to be eupheredermous. This classification has been little used by other authors. Informally, in unpublished but frequently-cited online annual updates, Subías (2016 and following) abandoned Eremelloidea and grouped Caleremaeidae sensu lato with Eremaeoidea. This had been suggested earlier-but not formalized-by Franklin and Woas (1992), though Subías' concept of the latter superfamily seems much broader than theirs. Weigmann (2006) noted that previous classifications of Caleremaeidae were problematic in a phylogenetic context and recognized the monofamilial Caleremaeoidea, though he did not indicate whether his sense of Caleremaeidae was broad or narrow.

Superfamily classification
Previously (Norton and Behan-Pelletier 2009;Schatz et al. 2011), we followed the tentative suggestions of Woas (2002) and included Caleremaeidae in an expanded concept of Ameroidea. It was the most recent of several expansions of the superfamily that began with Grandjean's (1966) addition of Staurobatidae. Curiously, he did not also include Basilobelbidae, despite their having the key traits of aggenital neotrichy and highly reduced proral setae on tarsi II-IV (see Grandjean 1959b, as Hammation), but this family was added by Balogh (1972) along with Heterobelbidae. Subsequently, Miko and Travé (1996) added Hungarobelbidae and Woas (2002) added Rhynchoribatidae, Spinozetidae and Oxyameridae. We relied upon the similarities of Caleremaeus and Hungarobelba noted by Travé (1961) and Miko and Travé (1996) to support adding Caleremaeidae to Ameroidea, but the resulting group of 14 families little resembles Grandjean's (1965b) original concept and even includes some families whose juveniles are not known to be eupheredermous. After our detailed studies of Caleremaeus herein, and reconsideration of traits-plus reference to two limited molecular studies-we no longer feel that this transfer to Ameroidea was reasonable; there are three general reasons.
First, the several purported similarities of Caleremaeus and Hungarobelba are not exclusive enough to be convincing synapomorphies or are not clearly homologous. Paraphrasing from Travé (1961) and Miko and Travé (1996), these are as follows. (1) The two genera share the laterosejugal enantiophysis (eL): however, this is found in several other eupheredermous taxa (R4). (2) They share the prodorsal enantiophysis (eA) and associated groove, though eA can be absent (presumed lost) from species in each genus: but eA is even more widespread than eL (see R4 and summary in Norton and Ermilov 2019). (3) They share an anteriorly truncate notogaster, with some type of humeral projection: again, this is not exclusive, being shared also by numerous other eupheredermous taxa-i.e. Veloppia, Megeremaeidae, Cepheidae-as well as some disparate apheredermous families such as Anderemaeidae, Autognetidae and Quadroppiidae, as well as Spinozetidae (with juveniles unknown). (4) They purportedly share a similar dimorphic type of genital papilla, with the posterior two pairs being more elongated, clavate, and slightly distant from the anterior pair: but these papillae are not perfectly rounded, and their shape can appear slightly different depending on orientation-we found no difference in shape or spacing among the papillae of any Caleremaeus species (see also Behan-Pelletier 1991). (5) They share a ridge near the rostral margin that bears seta ro: these are differently formed structures-that of Caleremaeus is a uniform shelf-like structure that may extend around the front of the rostrum while that of Hungarobelba is a much less conspicuous carina that effaces anteriorly; it is not unusual in Brachypylina for a carina (lateral ridge) to be directed anteriorly from acetabulum I, though typically it lies close to or even merges with the rostral margin (e.g. Grandjean 1960Grandjean , 1962Behan-Pelletier 1990) rather than reaching seta ro. (6) They supposedly share a long famulus on tarsus I: but this is vague, and a matter of degree-that of Hungarobelba is indeed unusually long and tapered but so is that of Basilobelbidae (Grandjean 1959b), and a long, tapered famulus is common in Damaeidae; the famulus of Caleremaeus is more modest in length and baculiform.
Second, the most distinctive trait of Ameroidea-at least in its original context (Grandjean 1965b(Grandjean , 1966)-that might be considered apomorphic among eupheredermous taxa is aggenital neotrichy, and several of the component families also have a more general ventral neotrichy. Caleremaeus exhibits no neotrichy on the body or legs. Other unusual traits of Ameroidea are not universal in the group. Most, but not all families exhibit size regression of proral setae on legs II-IV, where they are small, spiniform and often inconspicuous. In Caleremaeus these setae are plesiomorphic: normally formed and conspicuous. Caleremaeus also has a rather plesiomorphic form of preanal organ, in which the internalized apodematal extension is a simple, hollow, tubular or caecum-like structure. This does not occur in Ameroidea, but no clearly derived state for the group has been characterized. In its expanded context of 14 families, Ameroidea seems unjustifiable by synapomorphies; it has been defined (Norton and Behan-Pelletier 2009) by a collection of some widespread plesiomorphies and others that are present or absent according to family.
Third, the only existing molecular phylogeny study that includes Caleremaeus and members of Ameroidea contradicts the inclusion of Caleremaeidae in Ameroidea. With a focus on relationships among members of Zetorchestoidea, Lienhard et al. (2013) examined tree topology for seven eupheredermous families. Their consensus tree, based on a combined data set of mitochondrial (COI) and nuclear (EF-1α) genes, showed Caleremaeus as the sistergroup of Niphocepheus (Niphocepheidae); in turn, this clade was sister-taxon to the families Ctenobelbidae and Damaeolidae, which are unquestioned members of Ameroidea. Numerous eupheredermous families were not represented, and not all branches had high statistical support, but the strong linkage of Caleremaeus to Niphocepheidae, rather than the ameroid families, seems significant. Like Caleremaeus (but unlike Ameroidea), juveniles of Niphocepheidaeand other members of Zetorchestoidea (sensu Schatz et al. 2011), as well as Neoliodoidea and Plateremaeoidea-have plicate cuticle, a trait that we consider plesiomorphic in Brachypylina.
The only other molecular work that included Caleremaeus was based on the 18S ribosomal RNA gene. In a study encompassing representatives of many Brachypylina families, but none in Ameroidea, Schaefer and Caruso (2019; supplemental online material) presented two detailed trees. A maximum likelihood tree included Caleremaeus within a clade that contained four species in the family Damaeidae (in the monofamilial Damaeoidea), while in a Bayesian inference tree Caleremaeus formed a trichotomy with two different clades of those Damaeidae species. Since no Damaeidae were included in the Lienhard et al. (2013) work, direct comparison is not possible.
Faced with a seemingly mosaic distribution of morphological traits among the main eupheredermous taxa (Miko and Travé 1996;Woas 2002), at present we can offer no strong argument for including Caleremaeidae in any of the larger superfamilies. For example, the characteristic topography of the Caleremaeus notogaster is approximated in certain Plateremaeoidea-Pedrocortesellidae (e.g. Hunt 1996, his Fig. 34A)-but no other similarities seem important and differences are many. The prodorsum of at least some Caleremaeus species seems most similar to that of Megeremaeidae (Eremaeoidea) in having a lamella-tutorium complex, eA, eL, eH and an alveolar vestige of the second exobothridial seta, but most hysterosomal and leg traits differ. The scalp-attachment cornicle of Caleremaeus nymphs is seen also in Damaeoidea, but nymphs of the latter are not plicate, and adults differ in many ways.
Therefore, we adopt Weigmann's (2006) monofamilial Caleremaeoidea, but view Caleremaeidae in its strictest sense, i.e. including only Caleremaeus. Its diagnosis, as well as that of the family, would be identical to that of the genus, given above. Such a redundant classification has no phylogenetic content, but in this instance it seems preferable to retaining Caleremaeidae in an obviously polyphyletic superfamily, such as Ameroidea in the sense used by Woas (2002), Subías (2004), Behan-Pelletier (2009) andSchatz et al. (2011). Caleremaeoidea joins Neoliodoidea, Plateremaeoidea and Zetorchestoidea in a paraphyletic cluster of superfamilies near the base of Brachypylina phylogeny, characterized by eupheredermous nymphs with plicate cuticle.
Remarks on morphology and classification 1. Role and evolution of lamella and tutorium -It seems likely that the lamella and tutorium appeared early in the evolution of Brachypylina, in a ridge-or rib-like form such as in Megeremaeidae (Aoki and Fujikawa 1971;Behan-Pelletier 1990; see also Woas 2002, Norton andErmilov 2019). As known for 150 years (Michael and George 1879;Berlese 1896), the two structures play a role in defense by providing a place for coaptation of leg I, the distal parts of which typically lie between them when legs are adducted following disturbance. In more derived oribatid mites they often are highly developed as blade-like projections that overhang the adducted leg from above and below, respectively (Grandjean 1952;Fernandez et al. 2013).
Regression also has occurred, with transitions starting from both blade -and ridge-like forms. In some families or genera of the derived, poronotic superfamilies Ceratozetoidea and Oripodoidea, large blade-like lamellae and tutoria have regressed to narrow carina or ridges, or have disappeared. Similarly, in Caleremaeus monilipes and C. divisus the ridge-like lamellatutorium complex seems sufficiently developed to provide a coaptive space (see Seniczak and Seniczak 2019, their Fig. 4), while in North American species this complex ranges from partially regressed (C. retractus, C. nasutus) to being almost entirely lost (C. arboricolus). Whether the defensive leg adduction occurs in any Caleremaeus species is unknown, as behavior has not been reported.
2. Rostral bulge -This bulge is a centrally located convexity in the solid limb of the rostral tectum that appears to accommodate distal elements of the gnathosoma-probably the cheliceral tips-when the mite is in a defensive posture with chelicerae retracted and the subcapitulum elevated. It is a common feature in brachypyline oribatid mites and often it seems 'excavated' on the inner face to make the bulge very thin-walled, though this is not true of Caleremaeus species. Also, the pattern of excavation may result in an apparent central tooth projecting into the bulge (Norton and Ermilov 2017). Caleremaeus has no such tooth, but a frontal view of some species may give such an impression, since a unique, raised central rib is seen in optical cross-section (Fig. 2L). The rib is the central element of an embossed 'scorpion-like' pattern on the lower surface of the bulge (Figs 2K; 15F-H; 18A) that is present in all examined species except C. arboricolus. The central rib is crossed by a symmetrical series of about six raised transverse striae directed perpendicularly from it. The transverse striae are concentrated at the base of the rostral bulge, but the 'tail' of the scorpion appears to continue onto the freely-projecting rostrophragma, to which the soft cheliceral frame is attached ( Fig.  15H; rph, ch.fr).
3. Exobothridial seta vestige -Close to the exobothridial seta in the adult of all examined Caleremaeus species is a circular pale spot that in transmitted light looks much like the alveolus of seta ex, except no seta emerges from it. We consider it an alveolar vestige (exv) of the second exobothridial seta, a seta that has been lost from all members of Brachypylina. Among Brachypylina, we know of a similar vestige only in adults of Eremaeidae and Megeremaeidae (Behan-Pelletier 1993; as em). Though interesting, we interpret this similarity as a symplesiomorphy and therefore not evidence of a phylogenetic relationship among these families. A similar vestige is widespread in the outgroup Nothrina, many members of which also have a single exobothridial seta (e.g. Norton et al. 1996), and even in some Enarthronota (Grandjean 1963).
4. Enantiophyses -Adults of Caleremaeus species are rich in enantiophyses, tubercles that oppose each other across a cuticular groove. The grooves and tubercles seem to function in holding and anchoring an air film that has contact with the stigmata of the apodematalacetabular tracheal system (Chen et al. 2004). It is obvious how such localized plastrons are important for intertidal taxa (e.g. Pfingstl and Krisper 2014), but even for fully terrestrial species they could be important whenever the immediate environment is inundated with water for extended periods of time. Grandjean (1954b) proposed the term 'enantiophysis' in reference to Damaeidae and later (1960) refined his ideas and nomenclature, but the structures are widespread, particularly in early-to middle-derivative eupheredermous families. Grandjean (1960Grandjean ( , 1966 doubted that enantiophysis homology among taxa could be established on a general scale, but he did note the widespread taxonomic distribution of two, both of which occur in Caleremaeus (Figs 1,  13), These span the important sejugal groove, which encircles the body and in which one of the three pairs of stigmata open. The parastigmatic enantiophysis (eS or S) spans the groove just below the sejugal stigma, and the humeral enantiophysis (eH or H) spans it behind the bothridium, from which a tubercle on its posterior wall opposes a humeral projection from the notogaster. Since then, it has become apparent that three other enantiophyses are consistently placed and readily identified.
The laterosejugal enantiophysis (eL or L) also spans the sejugal groove, but mid-laterally, above the acetabula (Norton 1978). In the literature it has been reported for relatively few taxa-Caleremaeidae, Hungarobelbidae, Gymnodampia (Ameridae), Veloppia-but it also exists in Megeremaeidae (e.g. Behan-Pelletier 1990, her Fig. 32). Moreover, many descriptions of oribatid mites omit details of the lateral podosoma, so this is certainly an incomplete list. An almost certain independent evolution occurred in some Selenoribatidae, intertidal mites that rely on plastron respiration (Pfingstl 2013(Pfingstl , 2015. In this family the anterior tubercle bears the opening of the coxal gland (z), which is not true of other taxa and reinforces the idea that this is an independent appearance of eL. Grandjean (1968) considered this a parastigmatic enantiophysis but it is distinctly dorsal to the acetabula, whereas in its usual sense eS is below them.
Two others are associated with different grooves. The prodorsal (eA, or A) enantiophysis spans a dorsolateral groove on the prodorsum between the levels of acetabula I and II. If a tutorium exists, its posterior end forms the anterior tubercle of eA, with a separate tubercle immediately posterior to the groove. In the absence of a tutorium there can be a separate small tubercle on the anterior side of the groove. This enantiophysis is constant, or nearly so, in some eupheredermous families (Megeremaeidae, Pheroliodidae), but has variable presence in Caleremaeidae, Hungarobelbidae and Veloppia. It may be absent from most members of a family, but present in one or more genera, as in Damaeidae (Tokukobelba, some Kunstidamaeus), Ameridae (Gymnodampia) and the apheredermous family Anderemaeidae.
The aggenital enantiophysis (e4, U, G, co.ag) spans epimeral groove 4, with the anterior tubercle often bearing seta 4b (see R5). It is widespread but scattered among the non-poronotic Brachypylina, as summarized by Norton and Ermilov (2019). Among proven eupheredermous taxa it is present in all species of Caleremaeidae and Veloppia, some Cepheidae (Eupterotegaeus) and some Eutegaeidae (Neoeutegaeus); among 'presumed' eupheredermous families-i.e. with unknown juveniles but classified in Cepheoidea and Polypterozetoidea-it seems to be universal in Microtegeidae, Cerocepheidae and Nodocepheidae. Among apheredermous taxa, it is typical of Anderemaeidae and is found in one genus each of Otocepheidae (Fissicepheus; Aoki 1967) and Nosybeidae (Topalia; Colloff 2019). It is absent from Megeremaeidae and Hungarobelbidae, and is questionable in a few Damaeidae (a tubercle bearing seta 4b may be hypertrophied, but no posterior tubercle is present (e.g. Tokukobelba; Lamos 2016).
5. Epimeral setation -Herein (Figs 3F, 8B, 10D, 13B), we follow the modification of Grandjean's (1934) chaetotaxy for epimeral setae proposed by Norton and Franklin (2018; their Remark #15). The fundamental difference is that Grandjean focused on the order of setal appearance on each epimere, while the modification focuses on positional correspondence and therefore probably reflects metameric homology. This modification especially affects the notations for 4a and 4b, setae on an epimere that is first formed in the protonymph. Like the many leg IV setae that are delayed to the deutonymph, we feel seta 4a (= 4b in the widely used Grandjean chaetotaxy) is also delayed to the deutonymph; when it appears, it is well aligned with larval setae 1a, 2a and 3a on the soft sternal cuticle (Fig. 12H). By contrast, protonymphal seta 4b (= 4a in the Grandjean chaetotaxy) is inserted on the paired epimeral sclerite ( Fig. 12G; out of view in Fig. 12H), as are larval setae 1b and 3b.
6. Preanal organ -This sclerotized structure serves for the origin of paired adductor muscles that act on the genital valves (Grandjean 1969). It appears to have evolved from a simple unpaired plate on the anterior wall of the anal vestibule, as in many Nothrina, to become a largely internalized apodeme or apophysis in most Brachypylina. When the anal valves are tightly closed, the base of the organ either may be entirely hidden in the vestibule or it may be partly exposed; this depends on the size of the base and especially whether or not the anterior tectum of the two anal valves fully meet. In Caleremaeus species the organ is much like that of Damaeidae (Grandjean 1969, his Fig. 5A; LF), though the shape of the internal apophysis differs. The exposed base of the organ is relatively large, subtriangular in ventral view (Figs 8B,13B,14B;pr.o); the muscles originate on the internalized tubular apophysis and insert directly on the genital plate (Figs 3F, H, I, 16G). Weigmann (2006;his Fig. 121b) illustrated the anal plates of C. monilipes as completely covering the preanal organ, but this appears to be an error: in all our specimens of C. monilipes the base of the organ is exposed when anal plates are closed.
7. Leg setation -Variation. The complement of setae and solenidia on legs of both adult and juvenile instars is remarkably consistent in Caleremaeus, such that Table 1 expresses the ontogeny in all examined species to the extent that data exist (see R12). No interspecific difference was noted among adults, and a single example relates to ontogeny: seta l of tibia III is tritonymphal in C. monilipes, but deutonymphal in the other studied species (Table 1). Noted intraspecific variation includes only the following: (1) in the near-topotypical population of C. retractus, genu I seta v was absent from two of nine tritonymphal legs examined; (2) in the topotypical population of C. arboricolus, tibia IV seta l was absent from one of eight tritonymphal legs examined; (3) in the same population, tarsal seta l was absent from one of 10 adult legs I examined and (4) tarsal seta pl was absent from one (different) adult leg. The last example presumably is anomalous, since pl is a fundamental, eustasic tarsal seta that forms in the larva. The previous three relate to delays in the formation of amphistasic setae (setae variable among species or populations regarding the instar of first appearance). In examples 1 and 2 the setae presumably would have appeared in the adult, since these setae were present on all adults examined; in example 3, the delay in an adult-forming seta represented a loss by retardation in the concepts of F. Grandjean (see review in Norton 1977).
Trochanter III. Caleremaeus species are unusual in having both setae of trochanter III-v and l -formed in the deutonymph. The many oribatid mites for which setal development is known exhibit a variety of ontogenies on trochanter III, but usually these setae are formed in successive instars. We know of only four other species with the Caleremaeus pattern. Of these, only Niphocepheus nivalis (Schweizer, 1922) is a member of Brachypylina (Travé 1959); the other three are in the nothrine superfamily Malaconothroidea and include Mainothrus badius (Berlese, 1905), Tyrphonothrus maior (Berlese, 1910; as Trimalaconothrus novus and denoted v 1 , v 2 ) and Allonothrus giganticus Haq, 1978 (data respectively from Seniczak et al. 1998;Knülle 1957 and R.A.N. unpublished).
Lateral setae. On femora and tibiae I and II of Caleremaeus species, the lateral pseudosymmetrical pair have positions that seem far from being 'paired' (Fig. 4A, B). The adaxial member (l ) is high on the inner face, whereas the abaxial member (l ) is low on the outer face. On tibiae I and II l can easily be mistaken for a ventral seta prior to the tritonymph, when v first appears below it. On tibiae III and IV (Fig. 4C), which lack l , it is seta l that is abaxial and very low on the outer face, but in these instances v is already present by the time l forms, so the potential confusion is less. In essence, the setal verticil of these segments seems somewhat rotated on the segment, turned clockwise on legs I and II and counterclockwise on III and IV, but the shifts are most obvious in the lateral setae. Wauthy and Fain (1991) referred to such asymmetries-resulting from one member of a pair rising from its ancestral position and the opposite member falling-as 'basculations'. Those noted here are abaxial (=antiaxial) basculations ( on I/II and on III/IV); they are not uncommon in oribatid mites, but they seem unusually strong in Caleremaeus.
Iteral setae. The iteral pair (it) of tarsal setae are amphistasic, always post-larval, and always appear in the same position between the proral and tectal pairs, which are in contrast eustasic larval setae. But the specifics of this development among oribatid mites are surprisingly diverse, with more than a dozen ontogenetic patterns known (Grandjean (1961a(Grandjean ( , 1964a; see also Norton and Franklin 2018). The presence and instar of appearance of iteral setae vary among taxa, among the four tarsi, and rarely even between members of the pseudosymmetrical pair. The iteral ontogeny of Caleremaeus is interesting for two reasons. First, the pattern is complex, differing on each leg. Grandjean (1964a) noted this for C. monilipes by a formula that indicated first appearance of iteral setae on tarsi I-IV: (n3-Ad-[0, Ad]-0). In other words, the iteral pair appears in the tritonymph on tarsus I, in the adult on tarsus II, and fails to form on tarsus IV; tarsus III is asymmetrical in this regard, with it failing to form but it appearing in the adult. The second interesting point is that this complex formula, which is unique among oribatid mites, appears to be a generic trait (though only the adult setation is known in C. nasutus and we have no setation data for C. divisus; also, see R12).
8. Nomenclature of posterior gastronotic setae h 1 vs p 1 -In numerous families of oribatid mites, particularly eupheredermous taxa, the hysterosoma of nymphs terminates in a posterior process, typically with an indistinct pygidial (or caudal) sclerite, on which two pairs of setae insert. One pair usually is closer together than the other and slightly more dorsal, though in some taxa they are at equal height; most often the medial, higher pair is significantly larger. When applying his opisthonotal chaetotaxy to Porobelba spinosa (Damaeidae) nymphs, Grandjean (1954c) was uncertain which of these pairs represented h 1 and which was p 1 This is understandable, given that the pygidial region represents the dorsal part of terminal hysterosomal segments and therefore undergoes the least amount of spreading as segments are added to the caudal bend during anamorphosis (Grandjean 1939). Slightly later, for the similar pygidium of Polypterozetes (Polypterozetidae), Grandjean (1959a) clearly labeled the larger, more dorso-medial pair as h 1 .
By contrast, regarding Caleremaeus monilipes, Grandjean (1965b; p. 719) specifically and clearly considered the larger, more medial setae on the pygidial sclerite (therein called the 'croupion') to be pair p 1 ; we infer therefore that he considered the smaller, more separated, slightly lower pair to be h 1 . We found no explanation in his writings, but this seeming reversal of his hypothesis on homology is consistent with his labeling of setae in studies of at least three other eupheredermous taxa -Mongaillardia (Ameroidea; Grandjean 1961b), Pheroliodes (Plateremaeidae; Grandjean 1964b) and Fosseremus (Damaeolidae; Grandjean 1965a)-though the medial pair are not enlarged in these groups.
Without noting this apparent contradiction, most subsequent authors appear to have followed the earlier Polypterozetes-model in considering the adjacent and usually larger middle pair to belong to segment H (with the notation h 1 ). Some examples relate to Neoliodoidea (Seniczak et al. 2018b, Platyliodes), Plateremaeoidea (Ermilov et al. 2010, Pedrocortesella, present); (4) preanal organ without caecum (the internalized process is hollow, essentially like a caecum); and (5) proral setae of legs short and spiniform (they are normally formed setae). The diagnostic couplet (p. 483, couplet 88) distinguishing Caleremaeidae from Anderemaeidae also is wrong or misleading (cf. Norton and Ermilov 2019). The first character should be deleted: as noted above, Caleremaeidae species have ring-like ridges within the bothridium. The character 'circular depressions between setae c and la' refers to the foveae of the transverse sulcus; more important is the absence in Anderemaeidae of the unique notogastral topography (bulges and sulcus) found in Caleremaeidae. Enantiophysis A (= eA) is said to be absent from Anderemaeidae and present in Caleremaeidae, but in fact it is typical of Anderemaeidae and present or absent in Caleremaeidae. The cheliceral character should be ignored: it does not apply to most taxa currently included in Anderemaeidae.
12. Discrepancies with Seniczak and Seniczak (2019) -This important paper, based on material from Norway, redescribed the adult of the type species of Caleremaeus, C. monilipes, and presented the first complete assessment of ontogeny. However, during our study of C. monilipes specimens from Germany, Spain and Sweden, we noted several discrepancies with their results. After reexamination of these issues, Anna Seniczak kindly provided explanations and in some cases corrections (personal communication with R.A.N., 2019).
Enantiophyses. Their Fig. 3 shows the absence of enantiophysis eL, which is present on all our specimens (and was illustrated but not labeled by Miko and Travé 1996;their Fig. 8C). The specimen illustrated for their Fig. 3 was a light brown (teneral) specimen on which eL was not conspicuous, but eL clearly is present in the darker, more mature adults among their material. Also, they used some unconventional notations for enantiophyses (see R4); most significant was the use of Va-Vp (usually reserved for the ventrosejugal enantiophysis) for the aggenital enantiophysis spanning epimeral groove 4, and E4a-E4p for small tubercles on either side of epimeral groove 3.
Discidium. They illustrated and labeled a discidium (their Fig. 2; dis), whereas Grandjean (1965b) had indicated the absence of a discidium or discidial ridge in this species. According to A. Seniczak, their darker specimens match the figures of Miko and Travé (1996;their Figs 8C, 13B), which show a low, irregular and vaguely defined elevation in the region between acetabula III and IV. Our specimens also match the latter figures: when viewed from below in transmitted light, the irregular elevation appears darker by projection, but the impression disappears in lateral view. As described above, the development of a discidium or discidial ridge is variable among species of Caleremaeus, and that of C. monilipes seems weakest of all.
Pedotectum I. All species of Caleremaeus have a typical, well-defined pedotectum I in the form of a 'cupped' scale that posteroventrally envelops the base of leg I. This is well-shown in their Figs 2 and 4. However, their Fig. 1 is misleading in showing it as an isolated structure, distant from the leg, and their text incorrectly designates it as a propodolateral apophysis (P) instead of a pedotectum.
Leg setation. There are three discrepancies between their description of leg setation (their Fig. 13 and Table 2) and the results of our studies on C. monilipes. Upon reexamination of their material they found the following: only one iteral seta, it , develops on tarsus III and only one antelateral seta, a , develops on tarsus IV; both setae v and l form on trochanter III in the deutonymph. Therefore, there is no conflict with our generic description.
Palp solenidion. In their Fig. 3C, the palp is illustrated as having solenidion ω entirely fused to seta acm, but in correspondence they expressed some uncertainty, not having directly distinguished the two components. In all our studied material of Caleremaeus, including C. monilipes, ω is independent from acm and lies prone on the tarsal cuticle, where it is inconspicuous. Grandjean (1965b) also noted this independent, prone form in C. monilipes (presumably a French population).