Belomitra, P. Fischer, 1883

Kantor, Yuri I., Puillandre, Nicolas, Rivasseau, Audrey & Bouchet, Philippe, 2012, 3496, Zootaxa 3496, pp. 1-64 : 60-61

Position of Belomitra within Buccinoidea

Conchologically, Belomitra is rather variable, some species (e.g., B. granulata , with very small shell with reticulated sculpture, poorly pronounced subsutural ramp, and normally developed columellar plaits) being particularly divergent from the “typical” looking species resembling B. quadruplex . Nevertheless, the unique morphology of the lateral radular teeth unambiguously allows attribution of all species to a single clade, the monophyly of which is confirmed (at least for the species tested) by molecular data.

The family position of Belomitra is much less clear. When Belomitra was for the first time excluded from Conoidea, it was rather arbitrarily placed in Buccinidae . The authors ( Bouchet & Warén 1985: 223) wrote: “We have preferred to place Belomitra in Buccinidae simply because there is more variation in the buccinid radula than in the fasciolariid one, as far as one can judge from the literature”. This proposal was accepted in two later publications that dealt with Belomitra species ( Bouchet & Warén 1986; Poppe & Tagaro 2010). The present paper is thus the first attempt to validate the taxonomic position of Belomitra , this time based on molecular and anatomical data.

In our analysis, Buccinoidea were not retrieved as a monophyletic group, with Conus clustering as a part of the ingroup. Within buccinoids, three lineages (besides Belomitra ) were found to be unambiguously monophyletic: they correspond to the families Colubrariidae , Nassariidae and Columbellidae . Melongenidae , usually recognized as a separate family, is represented in our analysis by a single genus, Melongena , and therefore no conclusion can be drawn. Despite this rather limited dataset (the phylogeny of the entire Buccinoidea was outside the scope of the present work), our results are congruent with those presented by Hayashi (2005) based on complete 16S sequences, and Oliverio and Modica (2010) based on four genes (12S, 16S, COI and 28S). In the phylogenies published by these authors, Nassariidae and Colubrariidae were always monophyletic with high support, while “ Buccinidae ” were always polyphyletic. In our opinion, and even if more efforts have to be done in terms of taxon and gene sampling, this result would tend to support the non-monophyly of Buccinidae in its current ( Bouchet & Rocroi 2005) usage.

The families that were monophyletic in the molecular analysis are also well defined anatomically. For example, Nassariidae possess a characteristic multicuspid central radular tooth ( Bandel 1984), a stomach with very long posterior mixing area and very closely spaced ducts of the digestive gland ( Kantor 2003), and also a gastric shield is usually present. In Columbellidae , the central teeth lack any cusp, while the very characteristic and easily recognized lateral teeth are long (their length exceeds their width nearly three times), with a very narrow base attached to the subradular membrane and few large cusps on the inner side of the tooth facing the central teeth ( Kantor & Medinskaya 1991; Bandel 1984). Colubrariidae are characterized by an extremely long proboscis that is folded many times within the proboscis sheath, and a very small radula with lateral teeth having 9–11 cusps and central tooth with 10 cusps ( Oliverio & Modica, 2010). In Melongenidae , the very long proboscis, usually folded within the proboscis sheath, has large and powerful paired retractor muscles that are attached to its most distal part ( Kosyan & Kantor 2004). By contrast, Buccinidae s.l. are very heterogeneous in radula and anatomical characters and lack synapomorphies; and the Fasciolariidae have been so far very poorly studied and are difficult to define anatomically. Thus, the molecular results suggesting polyphyly of Buccinidae and Fasciolariidae do not contradict the anatomical data. Although a more complete molecular phylogeny of the entire Buccinoidea remains highly desirable, our results show that the Belomitra clade shares with other well-defined buccinoidean clades several features (anatomical distinctiveness, support for monophyly, large divergence from other buccinoid lineages) that lead us to rank it as a family, in accordance with the ranking of Nassariidae , Columbellidae , and Colubrariidae . We of course accept that ranking is a subjective issue; back in 1988, Ponder and Warén had proposed a classification of Neogastropoda where Nassariidae , Fasciolariidae and Melongenidae were all ranked as subfamilies within Buccinidae , but this ranking was never adopted by malacologists.

The anatomy of Belomitridae does not preclude its inclusion in Buccinoidea (although this is neither proved, nor disproved by our molecular results). The species of Belomitra examined in this paper are rather uniform anatomically and lack autapomorphies (except for the radula). As in other Buccinoidea , the anterior foregut is characterized by a long proboscis, large paired salivary glands, a large gland of Leiblein and a well defined valve of Leiblein. In Belomitra , the salivary ducts join the walls of the anterior oesophagus shortly after leaving the glands, an arrangement also found in Colubrariidae . The stomach ( Fig. 19F) is relatively very small, U-shaped and lacking posterior mixing area, being externally similar to some Antarctic Buccinoidea , in particular Lusitromina abyssorum (Lus 1993) [attributed to Cominellini Gray, 1857] ( Harasewych & Kantor 2004). The radula of Belomitra is very characteristic and unique in Buccinoidea : the lateral teeth are club-shaped, with very long and narrow base, forming a “handle” that is attached to the subradular membrane almost longitudinally, at a very sharp angle, along its narrow ventral side; close to their tip, these teeth always bear two cusps that differ slightly in shape. Somewhat similar lateral teeth are found in Prosiphonini Powell, 1951, an endemic Antarctic buccinid tribe, in which however the lateral teeth are generally multicuspid. The radula is very similar in all studied species of Belomitra , despite significant conchological differences (eg. between shells of Belomitra granulata n. sp. and others) and can be considered an autapomorphy of the family. Noticeable morphological variations were found only in the shape and number of cusps on the rachidian teeth, which is multicuspid in B. brachytoma , tricuspid in other species, albeit with differences in shape. The species of Belomitra lack accessory salivary gland and rectal gland, similar to other Buccinoidea , but this is not necessarily a synapomorphy as these glands may have been lost multiple times in the evolution of the group.

Although shells of Belomitra show a range of morphotypes and DNA suggests that several lineages could be defined within the Belomitridae , we prefer to classify all species in a single genus mostly for reasons of incomplete coverage, absence of molecular material for many species, and lack of support for most of the relationships between species in the phylogenetic tree. In the future, additional genera will most likely be recognized within the family, but we think that it is not yet desirable, especially since none of the type species of the nominal genera synonimized with Belomitra is available for molecular studies.