Hexapodidae, Miers, 1886

GUINOT, DANIÈLE, TAVARES, MARCOS & CASTRO, PETER, 2013, Significance of the sexual openings and supplementary structures on the phylogeny of brachyuran crabs (Crustacea, Decapoda, Brachyura), with new nomina for higher-ranked podotreme taxa, Zootaxa 3665 (1), pp. 1-414 : 120-124

Family Hexapodidae

Traditionally and only until recently, the suppression of the P5 has been thought to be a fundamental character of all the Hexapodidae , with the sole exception of the enigmatic Paeduma cylindraceum ( Bell, 1859) , supposedly having “aborted” P5. Diagnoses of Hexapodidae , including the original one ( Miers 1886) and most recent ones ( Gordon 1971: 106; Guinot 1979a: 114; Manning & Holthuis 1981: 166; von Sternberg & Cumberlidge 2001a: 332; Schweitzer et al. 2000: 55; Schweitzer & Feldmann 2001a: 331, table 1; Huang et al. 2002; Schweitzer 2005: 289), mentioned a P5 that is altogether “aborted” (as in most hexapodids) or rudimentary (as in Paeduma Rathbun, 1897 , substitute nomen for Amorphopus Bell, 1859 ; see Guinot 2006). Sternite 8 also was presumed to be absent in Hexapodidae ( Alcock 1900b: 287, 330; Balss 1957: 1658). Assuming the suppression of P5 and its sternite (sternite 8), most carcinologists admitted a sternal location for the male gonopores in Hexapodidae . According to Alcock (1900b: 330), “The efferent ducts of the male sex open on the 4th sternal segment [thoracic sternite 7] inside the fossa into which the abdomen fits”. The Hexapodidae was considered having sternal male gonopores ( Barnard 1950: 283, key; Balss 1957: 1658; Sankarankutty 1975: 3) and later assigned to Thoracotremata ( Guinot 1978a, with reservation; Schram 1986; Schweitzer et al. 2000; von Sternberg & Cumberlidge 2001a). Conversely, the family was considered to be heterotreme ( Saint Laurent 1989; Guinot & Richer de Forges 1997; Guinot & Bouchard 1998; Martin & Davis 2001; Guinot 2006). Von Sternberg & Cumberlidge (2001a: 332) complicated the issue by speculating that hexapodids were “morphologically thoracotremes although they are derived from heterotrematous precursors not shared by the other thoracotremes”. Even more perplexing was Števčić’ s (2005: 115) diagnosis ascribing a coxo-sternal condition.

The nature of hexapodids has remained uncertain for a long time, and their status a source of discussion. They have been associated to or included in Goneplacidae or in Goneplacoidea (e.g., Manning & Holthuis 1981; Pereyra Lago 1988; Dai & Yang 1991; d’Udekem d’Acoz 1999; Collins in Collins et al. 2003; Schweitzer et al. 2010: 135), or assigned to Xanthoidea ( Martin & Davis 2001: 75; Schweitzer & Feldmann 2001a: 331; Feldmann & Schweitzer 2004: 19, fig. 3B–E; Schweitzer 2005: 289). Hexapodids, however, deserve their own superfamily ( Guinot 1978a: 214; 284; Morris & Collins 1991: 28; Schweitzer et al. 2000: 55; Števčić 2005: 115; Ng, Guinot & Davie 2008: 86; De Grave et al. 2009: 33). A derivation of Hexapodidae from Dorippoidea was suggested by Glaessner & Secretan (1987: 10). Guinot (2006), De Angeli et al. (2010), and Guinot et al. (2010) have reviewed the taxonomic history of Hexapodidae .

The putative suppression of the P5 and the reduction or loss of its sternite (sternite 8) posed a crucial problem since the location of the male gonopore must be either on the P5 coxa or on sternite 8. A short sternite 8 with the external continuation of the vas deferens ending at its inner end was, nevertheless, recognised by Barnard (1950: 300, 301, fig. 56g, k, under the name “sternite 5”). According to Gordon (1971: 106, figs. 1, 2), sternite 8 was divided into two portions by a slight line, with the penis emerging at the inner end of the sternite and the papilla “rather twisted near its base”. Guinot (1979a: 114, 115, 215, figs. 32, 33E, F) proposed that a portion of the reduced sternite 8 may have an appendicular origin. The presence of a vestigial P5, however, escaped the attention of all these authors.

It is here shown that the supposed bipartite thoracic sternite 8 was a misinterpretation. The inner portion is truly sternal, but the lateral portion actually represents the pleurite of thoracic somite 7 (pleurite 7), which is concealed under the carapace. Pleurite 7 is modified, and becomes aligned at the same horizontal plane as sternite 7 ( Figs. 27 , 28 ). Conversely, the pleurite of thoracic somite 8 (pleurite 8) is completely lost. Sternite 8 does exist, although extremely reduced and confined to the body medial axis: it is concealed partially under the carapace and partially under the abdomen, except for a small triangular portion visible dorsally. The relative sizes of pleurite 7 and sternite 8 vary among hexapodid genera. The vestigial sternite 8 is dramatically different from the welldeveloped, subrectangular and largely exposed sternites 5–7, and it is obliquely oriented in relation to these sternites (Guinot 2006: figs. 2E, 4D). The dorsally visible sternal portion varies in size, shape, and orientation among the species. The Hexapodidae thus shows a strong modification of somite 8 and of the posterior part of the axial thoracic skeleton.

Hexapodids are easily recognised by their seven exposed sternites (instead of the eight normally present in Brachyura ), with subparallel and similarly developed sternites 5–7 in contrast to a reduced sternite 8. Sternite 4 is laterally extended, forming a marked process on each side in extant ( Guinot 1979a: fig. 33) as well as in fossil hexapodids ( Guinot et al. 2010: fig. 2C, D), but there is no junction with the pterygostome. The Milne Edwards openings show as long, curved slits, each being occupied by a developed, lamelliform coxa. The interlocking mechanism between the carapace and the cephalothorax involves sternite 7, with episternite 7 expanding and overlapping a depression on the posterolateral margin of the carapace. The first two abdominal somites are subdorsal. Additionally, the first abdominal somite is so dorsal that it may be completely covered by the carapace and concealed.

Based on dissections of Thaumastoplax anomalipes and Parahexapus africanus (Balss, 1922) , Saint Laurent (1989: 154, footnote) hypothesised that the penis could have included a vestigial coxal part. Our dissections of male Hexapus sexpes , Hexaplax megalops , and Thaumastoplax anomalipes confirmed the existence of a mobile vestigial P5 coxa, concealed by the abdomen and articulated to the vestigial sternite 8 ( Figs. 27A, C , 28A, B ).

Dissected Hexaplax megalops males show a penis that emerges from the vestigial P5 coxa, on the sloping side of the sterno-abdominal cavity, at the sternite 7 level. A thick duct prolongs from the gonad into the penis, which protrudes as a thick, soft, long papilla. The P5 coxo-sternal condyle is visible, supposedly with a penial condylar protection, and the minute coxa, which bears numerous long setae, is mobile. Although all hexapodid genera were not examined, it may be assumed that a vestigial P5 is shared by all Hexapodidae . Our dissections of females have shown the presence of a small sternite 8 but a complete loss of the P5 coxa. It is not surprising that the P5 is retained in male hexapodids for reproduction (P5 coxal gonopores), whereas in the females (sternal gonopores or vulvae) the P5 is completely lost.

The description of Paeduma cylindraceum with an aborted P5 “in the form of a minute tubercle inserted in a little notch at the base of the first joint of the fourth pair, and scarcely discernible by the naked eye” ( Bell 1859: 28), created a dilemma. The recent discovery and examination of Bell's supposedly lost holotype has shown that the “tubercle” in question actually corresponds to the apodeme of the P4 coxa ( Fig. 27C ) (Guinot 2006: figs. 1A, 2D; De Angeli et al. 2010; Guinot et al. 2010). The supposed rudiment seen by Bell (1859) definitely is not a P5 vestige. The apodemal platelet corresponds to the exposed portion of the apodeme, which shows externally as a sclerotised strip embedded in a notch at the proximal margin of the coxa; it prolongs itself through the arthrodial cavity of the leg into an internal plate providing attachment points for muscles. In? Hexapus sexpes the apodemal platelet is particularly visible when the P4 are extended. The platelet corresponds to the “tubercle” observed by Bell (1859: 29) on the small figure of De Haan (1835: trench pl. 11, fig. 6) at the base of the P4. Similar apodemal platelets are present in notches on the proximal margins of the coxae of other pereopods (especially P2 and P3); they also exist ventrally on the P2–P4 basis-ischia. Apodemal platelets are present in all hexapodids that were examined. Such visible apodemal platelets, also located in notches of pereopod articles, variously exposed, are in fact present in other eubrachyuran families, but they have unfortunately not been studied.

Hexapodids are clearly true decapods, having five pairs of legs. All brachyurans are therefore decapods. Despite the absence of most P5 articles, the recognition of a vestigial coxa from which the penis emerges demonstrates that hexapodids exhibit a coxal condition. The Hexapodidae should be assigned to Heterotremata. All the preceding details do not support the hypothesis that heterotremes derive from thoracotremes. Instead, Heterotremata should be viewed as a polytomy basal to the Thoracotremata (see Monophyletic Heterotremata).

The P5 is absent from the first two zoeae in the larval development of Spiroplax spiralis ( Barnard, 1950) . It is present, however, in the third zoea as a small bud, and in the megalopa as a reduced P5, folded dorsally under the carapace and ending in three long setae. It is absent in the first crab stage (Peyrera Lago 1988: 590, figs. 6H, 7A, J). Information lacks in zoeae studied by Sakhaie et al. (2009).

It was long accepted that the transversally elongated carapace and the absence of marked ornamentation allow hexapodids to move inside restricted tubular spaces, such as polychaete tubes and body cavities of hydrozoans ( Zehntner 1894; Tesch 1918a). Gordon (1971: 106) explained the loss of the P5 as a result of the lack of room when the carapace and sternum are interlocked. In addition to body shape, reduction of sternite 8 and loss of the P5 were interpreted as the result of the hexapodid lifestyle, a convenient argument for a number of species that are symbiotic. Some hexapodids (e.g., Latohexapus granosus Huang, Hsueh & Ng, 2002 ) are so far known to be free living in soft bottoms and exhibit a heavily ornamented body. Spiroplax spiralis , which is recorded from clay bottoms in the Iranian Gulf ( Stephensen 1946: 185), lacks a transversally elongated carapace. Nonetheless, these two species, have the typical hexapodid thoracic sternum, with a reduced sternite 8.

The Hexapodidae , which embraces 13 living genera with 19 species, updated from Ng, Guinot & Davie (2008: 86, figs. 77, 78) (see also Guinot 2006: 553–571, figs. 1–4), is known by a robust fossil record, correctly based on the presence of only seven thoracic somites, and thus exceptional reduction of sternite 8 ( Collins & Morris 1978: 977, pl. 118, fig. 8; Beschin et al. 1994: 191, fig. 8.2; Schweitzer et al. 2000: 55, fig. 16; Schweitzer & Feldmann 2001a: 331; Feldmann & Schweitzer 2004: 19, fig. 3D; Schweitzer 2005: 289; Karasawa & Schweitzer 2006: 58; Guinot 2006: 556; De Grave et al. 2009: 33; De Angeli et al. 2010: 52, 71, 72; Schweitzer et al. 2010: 135; Guinot et al. 2010: 284, 287, 293, 297). The assertion that a transversally elongated, rectangular, and smooth carapace is typical for Hexapodidae applies to a number of fossil hexapodids. Indeed, most of them have generally a carapace that is transversally elongated and smooth, including the oldest known representative, the Palaeocene † Goniocypoda rajasthanica Glaessner & Rao, 1960 , and † G. tessieri Remy & Tessier, 1954 , from the Maastrichtian, but, unfortunately, without a known ventral surface ( Remy & Tessier 1954; Glaessner & Rao 1960; Crane 1981; Crane & Quayle 1986; Schweitzer & Feldmann 2001a). Conversely, others have a subquadrate, coarsely tuberculate carapace (e.g., † Bellhexapus granulatus De Angeli, Guinot & Garassino, 2010 , fom the Middle Eocene). More unusual for a hexapodid is the Middle Eocene † Eurohexapus lobatus De Angeli, Guinot & Garassino, 2010 : it is a true Hexapodidae as shown by the typical thoracic sternum and the completely concealed sternite 8, but with a narrow, as long as wide carapace. These examples put in doubt the hypothesis that the hexapodid configuration is an adaptation for life inside restricted bodies, resulting in the reduction of sternite 8 and the loss (for the most part) of the last appendage. Certain rectangular hexapodids († Holthuisea cesarii ( Beschin, Busulini, De Angeli & Tessier, 1994) , and † Eohexapus albertii De Angeli, Guinot & Garassino, 2010 ), as well as more quadrate († Bellhexapus granulatus and † Eurohexapus lobatus ), have been found together in the same environment (Middle Eocene, northeast Italy) among volcanodetritic marls that are devoid of polychaete tubes but rich in foraminiferans, nummulites, echinoids, molluscs, and brachyurans (in particular † Retropluma eocenica Vía Boada, 1959 , with preserved ventral parts and chelipeds, see Beschin et al. 1996: 92, fig. 4, pl. 1, figs. 2–4; Beschin et al. 2001: 95; Beschin et al. 2009: 75, pl. 3, fig. 6). These marls, which were deposited in warm and shallow waters and rapidly deposited without strong alteration, allowed the preservation of most parts of the fossils, often with complete ventral surfaces (De Angeli et al. 2010). Another rectangular hexapodid, † Palaeopinnixa alontensis De Angeli, Guinot & Garassino, 2010 , was discovered in a calcarenite rich in molluscs and corals, and also with two Palicidae ( De Angeli & Beschin 2000; Beschin & De Angeli 2003).

It is evident, however, that in consideration of their thoracic sternal organisation all these fossil taxa assigned to Hexapodidae also possessed, as in the living species, only four normal pairs of pereopods (P1–P4) and a reduced P5. It is not known if the P5 of the early hexapodids consisted of several articles and not only the P5 coxa as in living representatives. The legs of the fossil Hexapodidae so far discovered having unfortunately been lost during fossilisation, the origin of the unique organisation of the Hexapodidae remains unknown.

The wide geographical distribution of the relatively small number of extant Hexapodidae suggests a highly specialised group, probably in decline, that has remained conservative and likely consisting of more than one lineage. In addition to the different patterns of the vulva (see above), several significant morphological differences among the hexapodid genera (e.g., mxp3, male abdomen, presence or absence of sternal trenches, G1) are indicative of several lineages that so far have not been translated into the current taxonomy. New insights on the affinities of Hexapodoidea are presented below (see Affinities between Palicoidea , Retroplumoidea , and Hexapodoidea ).

Figures

FIGURE 27. Vestigial P5 coxa, penis, and organisation of posterior portion of cephalothorax in Hexapodidae. A, B, Hexaplax megalops Doflein, 1904, male 13.0 × 19.0 mm, Fiji (MNHN-B29154); C,?Hexapus sexpes (Fabricius, 1798), female 10.6 × 18.0 mm, Persian Gulf (MNHN-B10209): exposed, translucent apodeme prolonging into sclerotised P4 platelet (left). a, apodeme; a.p., apodemal platelet; c, coxo-sternal condyle; cx4, P4 coxa; cx5, vestigial coxa of P5; e7, episternite 7; p, penis; p4, pereopod 4; pl7, pleurite of thoracic somite 7; 7, thoracic sternite 7; 8, reduced thoracic sternite 8; 6/7, thoracic sternal suture 6/7. Scale bar: 5 mm (C).

FIGURE 28. Coxal condition in Hexapodidae. A, Paeduma cylindraceum (Bell, 1859), female, 9.4 × 14.4 mm, west coast of Mexico (USNM 170897): posterior view of cephalothorax, with carapace slightly displaced to expose pleurite of thoracic somite 7 (pleurite 7). B, C, Hexaplax megalops Doflein, 1904, male 12.0 × 17.0 mm, New Caledonia (MNHN-B19244): schematic representation of posterior part of body, abdomen removed. a1–a3, abdominal somites 1–3; c, P5 coxo-sternal condyle; cx4, P4 coxa; cx5, vestigial coxa of P5; e7, episternite 7; p, penis; pl7, pleurite of thoracic somite 7; s8, exposed, visible dorsally portion of thoracic sternite 8; 6–8, thoracic sternites 6–8; 6/7, thoracic sternal suture 6/7. Scale bar: 5 mm (A).