Review of sexual dimorphism in brachypyline oribatid mites

Expressions of strong sexual dimorphism have been found in 77 species of Brachypylina, representing 36 genera, in the superfamilies Gustavioidea, Ameroidea, Oppioidea, Limnozetoidea, Ameronothroidea, Licneremaeoidea, Oripodoidea, Oribatelloidea, Ceratozetoidea and Galumnoidea. There are many examples of convergences, e.g., modifications of tarsus I setae in Cosmogneta (Autognetidae), Hydrozetes (Hydrozetidae) and Erogalumna (Galumnidae), and of possible behavioural constraints, e.g., the paraxial position of modified setae in sexually dimorphic species in these genera. Similarly, there is strong convergence in position and modification of presumed secretory porose organs in species of Autogneta (Autognetidae), Mochloribatula (Mochlozetidae), Symbioribates (Symbioribatidae), Oribatella (Oribatellidae), Zachvatkinibates, Nuhivabates (Punctoribatidae), Xiphobates (Chamobatidae) and Psammogalumna (Galumnidae). The number of superfamilies with sexually dimorphic species and the range of expression of sexual dimorphism suggest multiple independent origins in Brachypylina, as congeneric species in 20 of these 36 genera do not show such modifications. Despite 1% of brachypyline species being strongly sexually dimorphic, the evidence for courtship behaviour is limited to the Galumnidae and an undescribed species of Mochloribatula (Mochlozetidae). Evolution of strongly sexually dimorphic species in Oribatida seems to be in response to intermittent dryness, or aquatic habitats, or spatially discrete microhabitats. The littoral habitat is represented by 11 species showing strong sexual dimorphism, coastal vegetation by 6, the semiaquatic by 5, dry soil by 4 species and crustose lichens by 3 species. Arguably, these 28 species and some of the 19 species reported from arboreal habitats (including lichens and moss) live in microhabitats that can be intermittently dry, with wet-dry periods of varying lengths and intensity. Seven sexually dimorphic species of Hydrozetes are found in aquatic habitats; males of these all show modifications of one or more paraxial seta on tarsus I which may be used to orient the female. The 5 sexually dimorphic species of Autogneta, and Unguizetes mauritius (Jacot) are associated with decaying wood, bark and fungal sporophores, suggesting evolution of sexual dimorphism in this spatially discrete habitat. Undoubtedly, there are many other undiscovered cases of sexual dimorphism in Brachypylina, as microhabitats where they predominantly occur are rarely studied.


INTRODUCTION
Oribatid mites are unusual among Acari in that the gender of adults is not readily determined in most species without examining genitalia, which are always hidden in the genital vestibule when not in use. The majority of oribatid species are bisexual (Norton et al. 1993), about 10% reproduce parthenogenetically, and only 1% are distinctly sexually dimorphic. The rarity of distinct sexual dimorphism makes its nature and distribu-tion interesting, as selective forces leading to appearance and convergence of traits may be more readily detected. The subject has been partially reviewed several times. Newell (1957) provided a brief first overview of sexual dimorphism in Oribatida in the introduction to his description of Tuberemaeus papillifer (Newell, 1957). Travé (1959) provided the first summary of sexual dimorphism in Brachypylina in his study on Pirnodus detectidens Grandjean, 1956. Behan-Pelletier & Eamer (2010 described the first sexually dimorphic species of Oribatella and again reviewed sexual dimorphism in Brachypylina. Since 2010 a number of sexually dimorphic species have been described, and these highlight the diversity of expression even within a single genus. My goal in this paper is to review sexual dimorphism in Brachypylina and the little that is known of courtship behaviour in the sexually dimorphic species, highlight the ecological conditions where sexual dimorphism is most commonly expressed and discuss why sexually dimorphic species are apparently restricted to occasionally dry habitats, aquatic habitats and spatially discrete habitats. Sexual dimorphism per se is evident in all bisexual Oribatida; the spermatopositor (male) and ovipositor (female) are morphologically and functionally different (Alberti & Coons 1999, Ermilov 2010, Norton & Behan-Pelletier 2009). In addition, females are usually larger on average and have relatively larger genital plates than conspecific males. In this review I focus on species showing additional morphologically distinct expressions of sexual dimorphism. A large number of the species showing distinct sexual dimorphism have the porose organs of the octotaxic system modified in the male, or have presumed porose organs on the ventral or anal plates. As Norton and Alberti (1997) outlined in detail, when these organs are modified it is always in the male and involves hypertrophied porose organs "or new organs may exist in males that have no known homologues either in females of the species or in males of related taxa". They suggested two possible reasons for this sexual dimorphism, either a higher requirement for cuticular maintenance in males perhaps in response to more environmental exposure, or that the hypertrophied or novel porose organs of the males play a role in intraspecific communication producing "sex attractant or aggregation semiochemicals that are involved directly or indirectly in reproduction".
I focus on the Brachypylina as non-genitalic sexual dimorphism is most diverse and varied in this hyporder. It is also well recognized, though less common, in non-brachypyline taxa. For example, Grandjean (1954) noted that in the palaeacaroid Aphelacarus acarinus (Berlese, 1910a) the palptarsal eupathidium is forked in the female but not in the male. Wallwork (1962) studied three species of the epilohmanniid Epilohmannia in which the border of epimere IV is less strongly angled in the male than in the female. Schuster (1962) described the morphology and behaviour of Collohmannia gigantea Sellnick, 1922 which has a hypertrophied seta on genu IV of the male for holding nuptial fluid, as does the recently described C. johnstoni Norton & Sidorchuk, 2014.

RESULTS
Strong sexual dimorphism is known in 77 species of Brachypylina, representing 16 families and 36 genera in the superfamilies Gustavioidea, Ameroidea, Oppioidea, Limnozetoidea, Ameronothroidea, Licneremaeoidea, Oripodoidea, Oribatelloidea, Ceratozetoidea and Galumnoidea (Table 1). It is not yet reported for the other 15 brachypyline superfamilies (Schatz et al. 2011). In sexually dimorphic species modifications are often striking, are only found in the male, and encompass either modifications of the octotaxic system of dermal glands (
Ameroidea: Amerobelbidae: Among this morphologically diverse superfamily sexually dimorphic species are known only in the amerobelbid genera Hellenamerus (monotypic) and Mongaillardia. Males of H. ionicus have a ventrolateral porose area on the ventral plate. In males of 2 (of 5 described) species of Mongaillardia, the lamellar seta is crochet shaped, with spines, and seta pv' of tarsus II is highly modified in shape. This is the only known modification of a seta on tarsus II; all other sexually dimorphic tarsal setae are on tarsus I (Table 3). Grandjean (1961) speculated that males of these species walked on the posterior 2 pairs of legs during courtship, in a "promenade à deux", and that the highly modified seta pv' could be used for positioning of spermatophores, but there were no observations of living mites.

Oppioidea: Autognetidae:
In this large superfamily, strongly sexually dimorphic species are known only in the autognetid genera Autogneta (5 of 13 described species) and Cosmogneta (2 of 4 described species). Travé (1959) made preliminary observations on an unidentified sexually dimorphic species of Autogneta from Madeira in which males have thickened integument in the humeral region, but the species was neither described nor illustrated. Male Autogneta longilamellata have a long, oval porose area on the notogaster posteriorly. Four other Autogneta species show porose areas posteriorly, either flattened or in a concavity and associated with modified setae h1 and p1 (Behan-Pelletier 2015) (Fig. 1A, B). In contrast, males of the 2 strongly sexually dimorphic species of Cosmogneta have a highly modified seta a' on tarsus I. Limnozetoidea: Hydrozetidae: Seven of the 32 described species of Hydrozetes are sexually dimorphic, usually with one or more setae of tarsus I being modified in males ( Table 3). The exception is H. ringueleti in which the male has a modified shape of femur II, modified claw with a large tooth on leg III, and a large ventral spine on femur IV (Fernandez 1984). Fernandez (1986) noted small variation in femur shape between male and female Hydrozetes escobari Fernandez, 1986, but found no consistent differences that would support the species being considered strongly sexually dimorphic. Surprisingly, different tarsal setae can be modified in males of dimorphic Hydrozetes species, with one or more of setae it' pl' pv, v' and ft' modified ( Table 3). Males of two species of Hydrozetes, H. dimorphus and H. ringueleti are also larger than females (Fernandez 1984), which is the opposite of the usual pattern.
Hydrozetes also includes 5 species that are suspected parthenogens based on the absence or rarity of males in populations, including H. lemnae and H. lacustris (Norton et al. 1993). Whereas Fernandez (1984) noted an equal number of males and females in the population of H. ringueleti he studied, Grandjean (1941) reported 408 females to 3 males for H. lemnae (as H. terrestris) and 83 females to 1 male for H. lacustris, and considered males of these species non-functional and evolutionarily atavistic. He pointed out that rearing is essential to resolve whether a species of Hydrozetes could be both sexual and parthenogenetic. Fernandez & Athias-Binche (1986) considered Hydrozetes lemnae parthenogenetic in their detailed study of the demography of this species, and Ermilov (2006) confirmed parthenogenesis for the population of this species he reared. However, the juxtaposition of an atavistic male with evident dimorphism does raise the question of whether sexual reproduction occasionally occurs, or whether the modification of the tarsal seta in the male in this species is a genetic relic, and the closest relative of H. lemnae a sexual species. Based on adult and immature morphology, Seniczak & Seniczak (2009) and Seniczak et al. (2009) placed Hydrozetes lemnae in the "confervae" species group with H. confervae and H. thienemanni, whereas H. lacustris was placed in the "lacustris" group with other European species, H. parisiensis Grandjean, 1948, H. octosetosus Willmann, 1932, and H. longisetosus Seniczak and Seniczak, 2009. Clearly, sexually dimorphic species are found in both groups. The ecology of some of these species has been studied, e.g., in relation to water quality (Seniczak 2011), but there are no data on reproductive behaviour such as courtship.    Hydrozetes is the only brachypyline genus that includes sexual species, sexually dimorphic species and parthenogenetic species (Norton et al. 1993). Sexually dimorphic species are unknown in the sister family Limnozetidae, where all species are suspected parthenogens based on absence of males in samples and collections (Norton et al. 1993).

Ameronothroidea:
Distinct sexually dimorphic species are found in 2 of 4 ameronothroid families.
Fortuyniidae: Two species of Fortuynia (of 11 described) are strongly sexually dimorphic, but the expression differs significantly between species: males of F. atlantica have 4 pairs of porose areas on the notogaster, a pair of lateral notogastral protuberances, and lanceolate notogastral setae la, lm, whereas in males of F. yunkeri tibia IV is shortened and setae l' and (v) on this segment are modified (Hammen 1963, Krisper & Schuster 2008. Ameronothroidea: Ameronothridae: One species in each of the genera Halozetes (of 16 described), Podacarus (monotypic) and Alaskozetes (of 3 described), and 2 (of 13 described) species of Ameronothrus are sexually dimorphic. Species in the former 3 genera show aggenital neotrichy in the male (Podacarus also with epimeral neotrichy); whereas in Ameronothrus, legs are longer than body width in the male of both A. lineatus and A. nigrofemoratus (Schubart 1975).

Licneremaeoidea:
The octotaxic system is present to varying extent in most species of the 9 families comprising this superfamily, but only Glanderemaeus hammerae in the unplaced monotypic genus Glanderemaeus, is sexually dimorphic. The expression of dimorphism is unusual, with porose organ A3 being a saccule in males but a porose area in females (Norton & Alberti 1997). Oripodoidea: Of the 17 families comprising this superfamily, 6 include distinctly sexually dimorphic species with males having some modification of the octotaxic system (Table 2), combined or not with other modifications (Table 3).
Oripodoidea: Mochlozetidae: There are 10 genera in this family, yet strongly sexually dimorphic species are found only in Mochloribatula and Unguizetes. All described species of Mochloribatula are sexually dimorphic with modification of the octotaxic system in the male. In contrast, only 1 of 18 species of Unguizetes, U. mauritius, is sexually dimorphic, with A2 and A3 of the male longer than in the female (Table 2). Oripodoidea: Sellnickiidae: Sellnickia caudata is strongly sexually dimorphic with a modified rostrum in the male (Grandjean 1958). It is possible that the second described species in this genus, S. heveae Oudemans, 1927 is also sexually dimorphic, but only the male was described (Oudemans 1927). This species was considered a junior synonym of S. caudata by Subías (2004) but Grandjean (1958) noted that Oudemans (1927) had described the shape of the palptibial seta to be thick, lanceolate and unilaterally ciliate unlike that of S. caudata which are setose, thus I reject this synonymy.
Oripodoidea: Oripodidae: Of the 27 oripodid genera, distinctly sexually dimorphic species are found in 3, Cryptoribatula (1 of 2 described species), Oripoda (2 of 37 described species), and Pirnodus (2 of 3 described species) ( Table 1). The discovery of sexual dimorphism in Pirnodus detectidens by Travé (1959) and the subsequent description of Pirnodus soyeri by Travé (1969) highlighted the extent of morphological difference between the sexes in this family and in the expression of dimorphism between congeners, although this is evident in other congeners, e.g., sexually dimorphic species of Oribatella (Tables 2, 3). Both Pirnodus species show a modified hysterosomal shape in the male, as well as shifting of notogastral setae of l and h series, lyrifissures im and ips, notogastral saccules and the opisthonotal gland opening medially, relative to the female. In addition, in male Pirnodus soyeri notogastral setae c and h1 differ in shape from those of the female, and in contrast to the female the coxisternal region is elongated to the extent that the genital plates abut the anal plates, the preanal plate is weakly sclerotized and thin and the genital and anal openings maybe separated only by a fine membrane (Travé 1969). It is possible that the third described species of Pirnodus, P. andinus Baranek, 1985 is also sexually dimorphic, but only females were collected and described. Differences in body shape between sexes are possibly at their most extreme in Cryptoribatula euaensis which may be the senior synonym of Euaella gitteae Hammer, 1973 as suggested by Balogh and Balogh (1984) and accepted by Subías (2004). Cryptoribatula euaensis was described on the basis of a female specimen just prior to the description of E. gitteae in the same paper, based on a male specimen, both from the same habitat (Hammer 1973). Whereas the female has a typical oripodid habitus, the male has a pearshaped body, different positions of notogastral setae, a coxisternum that is strongly directed posteriorly and genital plates that strongly abut the anal plates.
Oripodoidea: Parakalummidae: Sexual dimorphism has been described for only one species in this family, Neoribates macrosacculatus (1 of 38 described), in which the notogastral saccules of the male are much larger than those of the female (Aoki 1966a).
Oripodoidea: Scheloribatidae: This diverse family, with 49 genera, has 3 genera with sexually dimorphic species, and the expression is different in each genus. One of 4 species of Parapirnodus, P. coniferinus, is sexually dimorphic with notogastral porose areas more elongated in the male and Aa positioned differently on the notogaster (Behan- Pelletier et al. 2002). In Tuberemaeus papillifer the posterior single pair of saccules are on well-developed tubercles in the male (1 of 30 described species) (Newell 1957). Tuberemaeus is considered a subgenus of Hemileius by Subías (2014), but is retained as a distinct genus by most authors, e.g., Ermilov and Anichkin (2013), and herein. The sexually dimorphic species of Nasozetes are particularly striking, with modification of the rostrum in the male forming a well-developed thumb-like projection (Table 3).

Ceratozetoidea: Chamobatidae:
A single species in this family is sexually dimorphic, Xiphobates callipygis (1 of 8 described) Only the male of this species is known, but I infer sexual dimorphism because of the unusual, almost transverse position of notogastral porose areas A1 and A2 posteromedially on the notogaster and porose areas A3 being on a posterior protuberance (Pavlichenko (1991, his Fig, 4)), similar to what is known for males of some sexually dimorphic punctoribatid species, e.g., Zachvatkinibates schatzi and species of Nuhivabates (Behan- Pelletier & Eamer 2005, Niemi & Behan-Pelletier 2004. Ceratozetoidea: Punctoribatidae: The 2 described species of Nuhivabates are sexually dimorphic with males differing in size, position and number of notogastral porose areas, and the presence of a pair of posterior tubercles (Niemi & Behan-Pelletier 2004). Similar modifications of the octotaxic system are found in the 5 (of 17 described) species of Zachvatkinibates that are sexually dimorphic (Behan- Pelletier & Eamer 2005, Weigmann 2009) (Table 2) (Fig. 1F-H).

Ceratozetoidea: Zetomimidae:
The single species of Naiazetes and 3 of the 5 described species of Heterozetes are sexually dimorphic. In contrast with other ceratozetoids (the dimorphic Chamobatidae and Punctoribatidae), sexual dimorphism in Zetomimidae does not involve the octotaxic system, which is present in Naiazetes but lost in Heterozetes. Naiazetes reevesi shows modification of the rostrum, rostral setae and genital papillae Va in the male, whereas dimorphic Heterozetes species have either a porose anal plate or a porose ventral plate in the male, or a combination of both. These porose regions are presumed dermal glands, though there is no evidence supporting this as yet.

Morphology of sexual dimorphism
A number of morphological characters are modified in males of sexually dimorphic species, and there is convergence and/or overlap in expression in taxa that do not seem to be closely related (Tables 2, 3). These morphological traits are reviewed below for the different body regions:
Modifications in shape and/or position of anal and genital plates in male: Pirnodus soyeri, Cryprotibatula euaensis (Oripodidae).

Evidence for associative mating
Most oribatid mites have sperm transfer that is not only indirect, but usually done without male-female association, i.e., dissociative. Males produce stalked spermatophores (stalked sperm packages) deposited independently but without male-female association (also known as dissociative sperm transfer) on a substrate, which is usually humid (Alberti 1999, and included references). Despite this, oribatid spermatophores are morphologically complex structures, and often taxon specific (Alberti et al. 1991, Fernandez et al. 1991 and Alberti (1999) also referenced modifications to spermatophores allowing their deposition in both aquatic and xeric habitats.
Although most evidence for spermatophore deposition appears to be dissociative (e.g., Pauly 1952, Woodring 1970, in some species deposition of spermatophores is proximal or associative, that is it is enhanced in the presence of females, for example in Pelokylla malabarica Adolph & Haq, 1982(Haq & Adolph 1981, and Pergalumna sp. (Oppedisano et al. 1995) or not produced when females are absent, e.g., Spatiodamaeus subverticillipes Bulanova-Zachvatkina, 1957 (Shereef 1972). Furthermore, in an undescribed species of Pergalumna, Oppedisano et al. (1995), described a system of physical signals in the form of stalks deposited in a cross formation, guiding the female to the spermatophore in the centre of this formation. How this signal system precisely leads the female to the spermatophore is unclear, but the authors assumed that signaling chemicals are deposited on the signal stalks. However, none of these species is distinctly sexually dimorphic.
Proximal or associative sperm deposition and/or mating is assumed in the seventy-seven brachypyline species that are strongly sexually dimorphic but there is little evidence, although that available is tantilizing. Grandjean (1956Grandjean ( , 1964Grandjean ( , 1966b) observed a courtship dance in the sexually dimorphic Erogalumna zeucta and Centroribates mucronatus, but there was no evidence of sperm deposition. Oliveira et al. 2007 described in detail the intricate courtship behaviour of an undescribed species of Mochloribatula from Brazil, with males having the modified terminal pair of porose areas raised medially. The female palpates the male's terminal porose areas with her palps, the stimulated male deposits a nuptial food on the substrate from his genital opening, walks forward a few body lengths, stands on his front two pair of legs, and the female eats the nuptial food. However, no spermatophores were deposited, and the authors speculated that sperm transfer may be direct. Extensive and prolonged courtship behaviour and deposition of nuptial food on modified seta of genu IV, but not associative mating, has been observed in the nonbrachypyline Collohmannia gigantea and C. johnstoni (Collohmanniidae) (Schuster 1962, Alberti & Schuster 2005, Norton & Sidorchuk 2014. To date, associative mating has only been observed in a non-sexually dimorphic species of Pilogalumna, where the male forces a sperm package into the venter of the female, but does not precede this with courtship (Estrada-Venegas et al. 1996).

Habitat of sexually dimorphic species
Of the 77 strongly sexually dimorphic brachypyline species the microhabitat has been fairly well defined for 62, there is no habitat data for 4 species, and the remaining 12 species have been extracted from "litter and soil" without further specification (Table 1) • Three species are from crustose lichens on rocks: Pirnodus detectidens, P. soyeri, Vaghia carinata.
• Seven species are found in continually aquatic habitats, these are the sexually dimorphic species of Hydrozetes, all of which show modifications of one or more paraxial seta on tarsus I.
• Seven species are from decaying wood, bark and fungus, including all 5 sexually dimorphic species of Autogneta and Unguizetes mauritius. Symbioribates papuensis collected from cryptogamic plants growing on the elytra of the Papuan weevil Gymnopholus is included here.

DISCUSSION
Of the more than 8350 described species of Brachypylina (Subías 2014), only 77 are strongly sexually dimorphic, representing less than 1% of described species. They are found in 36 of 1096 brachypyline genera (i.e., ca. 3%), which represent 16 of 131 families (i.e., ca. 13% of families) found in 10 of 25 superfamilies (40%): Gustavioidea, Ameroidea, Oppioidea, Limnozetoidea, Ameronothroidea, Licneremaeoidea, Oripodoidea, Oribatelloidea, Ceratozetoidea, Galumnoidea (Table 1). We can contrast this with the diversity of parthenogenetic species, about 10%, (Norton & Palmer 1991, Cianciolo & Norton 2006. In Brachypylina most parthenogens are isolated in otherwise sexual genera; there are few wholly parthenogenetic genera such as Tectocepheus and Limnozetes, whereas there are 8 genera where all species are strongly sexually dimorphic. It is possible that parthenogenesis in Brachypylina is not as high as in Oribatida in general; but this has not been reviewed since Norton et al. (1993), though many subsequent studies have looked for males in populations of species, e.g., Cianciolo & Norton (2006), Maraun et al. (2013), Fischer et al. (2014. In genera that also include monomorphic species, such as, Autogneta, Hydrozetes, Ameronothrus, Parapirnodus, Zachvatkinibates, do the dimorphic species form a clade? Analysis is hampered by the dearth of phylogenetic data. To date, there is no morphological or molecular analysis of phylogenetic relationships in Brachypylina per se, and other than Schäffer et al. (2010), who included 2 sexually dimorphic species of Hydrozetes in their morphological and molecular analysis, and Maraun et al. (2004) who included Autogneta longilamellata in their molecular analysis, no molecular studies include strongly sexually dimorphic species. Despite these caveats, the range of taxa showing strong sexual dimorphism and the fact that in many genera only some species are strongly sexually dimorphic indicates multiple origins of these traits. There is no evidence they are confined to any phylogenetic lineage in Brachypylina, other than that sexually dimorphic species are concentrated in the 3 closely related poronotic brachypyline superfamilies, Oripodoidea, Ceratozetoidea and Galumnoidea (Norton & Behan-Pelletier 2009), while being absent from the earlier-derivative poronotic taxa Achipterioidea and Phenopelopoidea.
Expressions of strong sexual dimorphism are not random; as outlined in Results they are confined to modifications of dermal glands, setal modifications and the shape of legs and body. Furthermore, there is evidence of convergence; for example, tarsal setae are modified in Ameroidea (Mongaillardia), Oppioidea (Cosmogneta), Limnozetoidea (Hydrozetes), Galumnoidea (Erogalumna) and notogastral setae are modified in Oppioidea (Autogneta), Ameronothroidea (Fortuynia), Oripodoidea (Pirnodus). Grandjean (1960Grandjean ( , 1963 noted that in all brachypyline species with tarsal setal modifications in the male, these setae are paraxial. He hypothesised (1963, 1964, 1966b) that during courtship behaviour males of these species walk behind the female on two or three pairs of legs with legs I or II on the flank of the female, and reported this behaviour for Erogalumna zeucta (1956,1964). In contrast, females of a Mochloribatula sp. that touch the sides of the male during courtship do not show tarsal setal modifications (Oliviera et al. 2007).
Similarly, there is strong convergence in position and modification of secretory porose organs in sexually di-morphic species of Autogneta, Mochloribatula, Symbioribates, Zachvatkinibates, Nuhivabates, Xiphobates and Oribatella canadensis, with a porose organ(s) on a posterior notogastral tubercle. In all cases of this type of sexual dimorphism, other than in Glanderemaeus hammerae and Neoribates macrosacculatus the dermal glands modified are surface porose areas, rather than invaginated saccules (Norton et al. 1997, Norton & Alberti 1997 (Table 2).
Sexually dimorphic modifications of prodorsal structures are most developed in Oripodoidea and Galumnoidea, and again there is evidence of convergence. Adults of Sellnickia caudata, a species known from New Zealand and Australia, have a dimorphic rostral lobe with a porose leading edge (Grandjean 1958). The male lobe is strongly curved distally and has the appropriate size and shape to fit over the pygidial tubercle of the female (Norton and Alberti 1997). Males of Nasozetes (Scheloribatidae) have a distinct rostral protuberance (Grandjean 1970), which Bernini (1984) noted is similarly found in the galumnid Kabylogalumna rhinoceros. The dimorphism of rostral setae in species of Symbioribates, with those of the male enlarged and strongly birefringent throughout, is unique among known Oribatida; their morphology has not been examined in detail, but it is possible they function similarly to the modified tarsal setae in dimorphic species of Hydrozetes, steering females to, or in the direction of spermatophores. Proctor (1998) captured the paradox of dissociative mating: "... the improbability that a small female arthropod will stumble across a much smaller droplet of sperm in a vast volume of soil or water". Dissociative sperm transfer and deposition of spermatophores is clearly evolutionarily viable in moist soil and litter habitats with high populations, habitats in which oribatid mites are particularly species rich. But some kind of association would seem to be advantageous in environments that can stress the sperm package with dryness or that are aquatic. In Proctor's (1998) terminology this "pairing with indirect transfer" occurs "when the male courts a particular female before, during, or after spermatophore deposition and often directs her towards his deposited spermatophore". Modifications shown by sexually dimorphic oribatid species would appear to have evolved for this type of association between males and females, and to be linked to particular non-soil microhabitats (Norton & Alberti 1997).

Sexual dimorphism and habitat
Arguably, species that are strongly sexually dimorphic and living in intermittently dry microhabitats, such as those associated with littoral habitats (11 species), coastal vegetation (6 species), semiaquatic habitats (5 species), dry soil (4 species) and crustose lichens (3) suggest that these are conditions where traits for associative mating are evolutionarily advantageous. To quote Norton and Alberti (1997) "sex pheromones and associa-tive mating may be especially important in drier environments. The reason for this pattern is undetermined, but it may relate to instability of spermatophores or to the need for locating mates in low density populations." In species that live in semiaquatic habitats (Heterozetes, Naiazetes), or littoral habitats (e.g., Fortuynia, Zachvatkinibates), spermatophore deposition and reproduction may occur only at times when the habitat is drier, e.g., low tide, or when adults can aggregate in drier parts of the habitat, e.g., on surface vegetation. This has been postulated for Fortuynia atlantica by Pfingstl (2013) who observed adults and immatures of this species retreating into crevices to avoid submergence.
Whereas, some of the nineteen arboreal species listed in the Results are known from forests in the subtropics or tropics that may experience intermittent dryness, others are from rainforest or cloud forest. For example, Parapirnodus coniferinus is the only known sexually dimorphic arboreal species in temperate rainforests of western North America; a continually moist habitat. Similarly, Nuhivabates hivaoa and N. nukuhiva were collected from cloud forest, and Nasozetes choerognathus was collected from sphagnum moss in high altitude rainforest.
In discussing sexual dimorphism in species of Hydrozetes, Norton and Palmer (1991) suggested that oribatid spermatophores in hypotonic environments maybe osmotically stressed, but also noted that in freshwater Prostigmata dissociative sperm transfer is common, as are evolutionary trends towards more proximal sperm transfer. They suggested that the modified paraxial tarsal setae in Hydrozetes species may allow for direct sexual contact during mating. Or they may be used to guide the female to the spermatophore. It is possible that similar guiding (oribatid choreography) is accomplished by the modified seta on tarsal segments in species of Mongaillardia, Fortuynia yunkeri and species of Galumnidae, and the modified rostral setae in Symbioribates (Table 3).
The five sexually dimorphic species of Autogneta, and Unguizetes mauritius, have similar modifications to those in intermittently dry habitats, i.e., modified dermal glands, with or without associated modified setae, and they possibly also encounter periods of dryness unconducive to dissociative sperm transfer. However, decaying wood, bark and fungal sporophores can provide protected, food rich habitat patches for members of a population within a matrix of unsuitable space. These patches can be spatially discrete from similar microhabitats, populations can be isolated, and dispersal could be the lifehistory trait that is constrained (Holt 2003), thus limiting interaction between members of populations of a species and optimizing conditions for evolution of some kind of associative mating. Possibly the same applies to Symbioribates papuensis living on the fungi, algae and lichens growing on the pronotum and elytra of wingless Gymnopholus weevils in high altitude, moist, moss forests, where pop-ulations of up to 60 mites can be found on a single weevil (Gressit et al. 1965).
Many of the hypotheses on the habitat associations and distribution of sexual and parthenogenetic oribatid species proposed by Norton and Palmer (1991) have been tested in recent years e.g., Cianciolo and Norton (2006), Cianciolo (2009), Fischer et al. (2014, Maraun et al. (2013). Hopefully this review will stimulate ecological studies on sexually dimorphic species. For example, is it moisture per se, or longevity of spermatophores, or the chemistry of pheromones produced by either sex that is the constraining factor for sperm transfer in intermittently dry habitats? What are the differences in ecological conditions of microhabitats of sexual, sexually dimorphic, and parthenogenetic species of Hydrozetes? Is evolution of sexually dimorphic species of Autogneta and Symbioribates related to their spatially discrete microhabitats with dispersal constraints?

CONCLUSIONS
The range of expression of sexual dimorphism in oribatid mites suggests complex evolutionary histories and multiple independent origins in response to selective forces imposed by microhabitat. I suspect that the 1% strongly sexually dimorphic species known in Brachypylina is an underestimation; many species descriptions do not indicate the sex being described, and/or both sexes are not examined for potential differences. Furthermore, microhabitats where sexually dimorphic species are most common (Table 1) are underrepresented in biodiversity studies. Brachypyline families without known dimorphic species, yet present in non-soil habitats, e.g., Microzetidae, Licneremaeidae, Tegoribatidae, warrant close attention, particularly when new species are described. Strong sexual dimorphism clearly has arisen many times in Brachypylina; and the expression of this dimorphism does not appear restricted to any particular lineage, though this is hard to validate as phylogenetic data are generally lacking. Undoubtedly, there are many other undiscovered cases of sexual dimorphism in Brachypylina and knowledge of their morphology is just a precursor to more interesting questions about their biology, evolution and ecology.