RISSOIDAE GRAY, 1847

FAMILY RISSOIDAE GRAY, 1847 View in CoL

The taxa included in this analysis are a broad representation of this diverse, entirely marine family. Six main clades can be delimited in the phylogeny (A – F in Figs 1, 2 View Figure 2 ), and some genera have been found not to be monophyletic, as briefly discussed below.

Clade A contains four genera. One species, Simulamerelina cf. wanawana (Kay, 1979) , is attributed to Simulamerelina following Hasegawa (2000) as it is rather similar to the type species, S. corruga ( Laseron, 1956) . This taxon was previously treated as a subgenus of Alvinia Monterosato, 1884 [type species A. weinkauffi (Weinkauff, 1868) ] by Ponder (1985a). However, both taxa are found not to be intimately related. Consequently, we tentatively treat Simulamerelina as a distinct genus, although acknowledging that examination of the type species is required to confirm this. Also included in this clade are the type species of Subonoba Iredale, 1915 , Austronoba Powell, 1927 and Subestea Cotton, 1944 . These taxa were included in Onoba H & A. Adams, 1852 (type species O. semicostata Montagu, 1803 ) ‘somewhat tentatively’ by Ponder (1985a), the first two as synonyms of Onoba s.s., and the third as a subgenus. The general shell morphology and anatomy of the taxa Ponder (1985a) attributed to Onoba are rather similar with their conical to elongate-conical shells bearing spiral threads and, sometimes, weak axial ridges. Their head – foot, radulae, opercula and anatomy are also all rather similar and these features may be plesiomorphic. For example, none has a posterior pedal gland and if a metapodial tentacle is present it is short and triangular. The type species of Onoba was included in our analysis but is widely separated and is the basal member of clade F (containing Alvania and related taxa). On the basis of these results we treat Subonoba and Subestea as distinct genera. Whether Austronoba should be maintained as a genus, a subgenus of Subonoba or a synonym will have to await further study. The only substantive difference is that Austronoba has a more slender shell with axial ridges.

Clade B is basal to a monophyletic group that includes clades C and D + E. It consists of two unidentified Japanese species ( Hasegawa, 2005, fig. 5I, L) attributed to the deep-water genus Benthonella (type species B. tenella Jeffreys, 1869 ). The anatomy of the type species of Benthonella was described by Ponder (1985a) and is particularly unusual in that it appears to practise renal copulation. This unusual means of transferring sperm is not unique as it is also known, as noted above, in a few members of three other truncatelloidean families ( Ponder, 1988).

Clade C is sister to the D + E clade and contains three unusual genera, all of the type species of which were included in the analysis: Lucidestea (type species L. vitrea Laseron, 1956 ), Parashiela (type species P. ambulata Laseron, 1956 ) and Voorwindia (type species V. umbilicata Ponder, 1985 ). These taxa are all characterized by small, conical shells that range from being sculptured with axial and spiral ridges to fine spiral threads or being smooth. Lucidestea species, uniquely for Rissoidae , have a small peg on the inner side of the operculum. The central tooth of the radula of Lucidestea and Voorwindia has two pairs of cusps while that of Parashiela has only a single pair, as in most other rissoids. The radula of the European genus Obtusella Cossmann, 1921 is similar to that of Lucidestea but that taxon was not included in our analysis. These three included genera share a single short, narrow metapodial tentacle that emerges from the foot behind the opercular lobe. Lucidestea , Voorwindia and Obtusella also have a large anterior sperm sac, another unique feature within Rissoidae . The female anatomy of Parashiela has not been studied.

Clade D includes Setia turriculata Monterosato, 1884 , Haurakia hamiltoni Suter, 1898 and Vitricithna marmorata (Hedley, 1907) ; the latter two species represent the type species of Haurakia and Vitricithna , respectively. Ponder (1985a) treated Vitricithna as a synonym of Haurakia , which itself was given subgeneric status within Pusillina [type species P. dolium (Nyst, 1843) ]. This treatment is not supported in the molecular analysis of this study ( Figs 1, 2 View Figure 2 ) and Haurakia and Vitricithna are here treated as distinct genera. Pending more taxon sampling, we tentatively treat S. turriculata as being more typical of Setia (type species S. pulcherrima Jeffreys, 1848 ) than the other species in our analysis, ‘ S. ’ ambigua (Brugnone, 1873) , which is also usually included in that genus.

Clade E contains several species attributed to Rissoa (type species R. ventricosa Desmarest, 1814 ) and Pusillina , as well as ‘ Setia ’ ambigua . There are two groups within this subclade in the BI tree; one includes Rissoa lia Monterosato, 1884 (the type species of Liavenustia Nordsieck, 1972 ), synonymized with Rissoa by Ponder (1985a) and R. variabilis Megerle von M uhlfeld €, 1824 and four species attributed to Pusillina , none of which is the type species. Pusillina inconspicua is the type species of Mutiturboella Nordsieck, 1972 , and P. radiata is the type species of Radiata Nordsieck, 1972 , both of which were treated as synonyms of Pusillina by Ponder (1985a). In the ML tree, R. variabilis is included in the second group with the other species of Rissoa .

The second group includes seven species attributed to Rissoa , including the type species and the types of seven other taxon names included in the synonymy of Rissoa by Ponder (1985a), and ‘ Setia ’ ambigua . This latter species has a smooth, transparent, elongately conical simple shell that resembles a few other species included in Setia but is rather different from the type species of that genus, as indicated above.

Because our analysis did not include the type species either of Pusillina or of Setia , it is not possible to make definitive comments on the validity of these taxon names. We recommend leaving the status quo given that our results are not clear cut, but with the clear realization that the Rissoa -group of taxa needs revision.

Clade F contains several subclades which are consistent in both of our analyses. A basal branch, which is the sister to the rest of the clade, contains several taxa from Japanese waters, some of which are unidentified. These include the shallow-water ‘ Alvania ’ concinna (A. Adams, 1861) and several deep-water species that were recently reviewed by Hasegawa (2014). These latter are ‘ Alvania ’ akibai (Yokoyama, 1926) ( Fig. 5A View Figure 5 ), Frigidoalvania asura (Yokoyama, 1926) , Punctulum flavum (Okutani, 1964) , Punctulum cf. flavum ( Fig. 5C View Figure 5 ) and P. tanshumaruae Hasegawa, 2014 . These taxa are all rather similar in having broad shells, most have axial and/ or spiral ribs and some have a thick periostracum. Given that they are all very similar in our analysis, we suggest that they should all be referred to Punctulum Jeffreys, 1884 which, on the basis of this result, might include Frigidoalvania Waren, 1974 (type species Rissoa janmayeni Friele, 1878 ) as a synonym. There are, however, some differences between the type species of these two genera, for example some details of the anatomy and the metapodial tentacles (see Ponder, 1985a) so we do not formally synonymize them here.

Based on the molecular results, ‘ Alvania ’ concinna is clearly not a member of the genus Alvania Risso, 1826 and there is no generic name that is suitable for it. Alvania rudis (Philippi, 1844) , the type species of Thapsiella Fischer, 1885 , has a somewhat similar, tall shell but it does not agree well in other shell characters so we do not include it there.

At the base of the branch including the remaining taxa in clade F are: a species attributed to Simulamerelina , ‘ S. ’ tokunagai (Yokoyama, 1927), Onoba semicostata (Montagu, 1803) , the type species of Onoba , Manzonia crassa (Kanmacher in J. Adams, 1798) the type species of Manzonia Brusina, 1870 and ‘ Alvania ’ sp. ( Fig. 5B View Figure 5 ), which does not fit any named genus.

Many of the remaining taxa are currently placed within Alvania as recognized by Ponder (1985a) but this genus is rendered polyphyletic by the inclusion of other clearly distinct lineages. The results are detailed below but final taxonomic decisions must await better taxon sampling, and in particular the inclusion of type species of key genus-group names.

The next clade contains four species of Alvania , including the type species, A. cimex (Linnaeus, 1758) . The next branch is Cingula trifasciata (J. Adams, 1800) (the type species of Cingula Fleming, 1818 ) and then two species attributed to Crisilla , but not the type species, C. semistriata (Montagu, 1808) . Crisilla was treated as a subgenus of Alvania by Ponder (1985a). The next clade contains three species, ‘ Alvania ’ circinata A. Adams, 1861 , a spirally ribbed species lacking axial sculpture. It resembles Alvania hedleyi Thiele, 1930 from Western Australia but no generic name is available for these taxa. The taxon Conalvinia is available for the remaining two taxa, the type species Conalvinia novarensis and C. ogasawarana (Pilsbry, 1904) .

‘ Alvania ’ tenera (Philippi, 1844) is then sister to the final clade which includes five species currently attributed to Alvania . ‘ Alvania ’ tenera is a broad species with a conical shell sculptured with spiral ribs and weaker axial threads. The remaining five species include ‘ A. ’ discors (Allan, 1818) , ‘ A. ’ lanciae (Calcara, 1845) , ‘ A. ’ scabra (Philippi, 1844) , ‘ A. ’ aeoliae Palazzi, 1988 and ‘ A. ’ lineata Risso, 1826 . Of those, all have conical shells with strong axial ribs with weaker spirals, with ‘ A. ’ scabra being distinctive in having a more pagodiform outline. It is the type species of Alvaniella Sacco, 1895 , which is the earliest genus-group name that could be used for this subclade. ‘ Alvania ’ lineata is the type species of Alvanolira Nordsieck, 1972 .

ORIGIN OF DEEP- SEA RISSOIDS

Our molecular phylogeny suggests that the Rissoidae originated in the shallow sea and independently radiated into bathyal waters at least twice ( Figs 1, 2 View Figure 2 ). Bathyal rissoids have been assigned to several genera including Benthonella , Frigidoalvania Waren, 1974 , Onoba , Powellisetia Ponder, 1965 , Punctulum and Pusillina , as well as to the polyphyletic Alvania ( Waren, 1974, 1996b; Ponder, 1983b; Hasegawa, 2005, 2014). The seven bathyal species studied herein (459 – 1919 m; Supporting Information, Table S1) constitute clade B ( Benthonella ) and a subclade of clade F (‘ Alvania ,’ Frigidoalvania and Punctulum ; Figs 1, 2 View Figure 2 ). One of the two studied species of Benthonella had previously been placed in Alvania ( Hasegawa, 2005, fig. 5I) suggesting that the latter genus contains distantly related lineages. Repeated invasion of the bathyal zone has been documented for other gastropod families of shallow water origin (e.g. Williams et al., 2013).

The species of Benthonella have a relatively large, thin shell with a simple outer lip of the aperture (Pon- der, 1985a). This contrasts with the smaller and more solid shells with a more or less thickened outer lip that characterize Punctulum , Frigidoalvania and ‘ Alvania ’ in clade F ( Fig. 5A View Figure 5 ; Hasegawa, 2014, figs 2 – 48). The two deep-sea clades also differ in their biogeographical and bathymetric distributions. Benthonella species have been reported from low- to highlatitude seas ( Ponder, 1985a; Waren, 1996b; Lozouet, 2014) with their depth ranges extending to the lower abyssal plain ( Rex & Etter, 1990). On the other hand, the bathyal species in clade F are components of more nutrient-rich waters off northern Japan under the influence of the south-flowing Kuril (Oyashio) Current. Most bathyal rissoids in the North Atlantic fall into the latter clade ( Waren, 1974, 1996b).

The lack of pigmented eyes is among the bestdocumented morphological features of deep-sea gastropods. Apomorphic loss of the retinal pigmentation may occur in rather short periods of time (i.e. a few million years; Williams et al., 2013). All species of Benthonella , Punctulum and Frigidoalvania so far investigated lack pigmented eyes ( Ponder, 1985a; Hasegawa, 2014) but ‘ Alvania ’ cf. akibai retains eye pigmentation ( Hasegawa, 2014), despite its co-occurrence with F. asura with unpigmented eyes in the same sample (Supporting Information, Table S1). The phylogenetic position of ‘ A. ’ cf. akibai basal to Punctulum and Frigidoalvania , accompanied by small genetic distances among these taxa, implies that the apomorphic loss of the pigmentation occurred only once, and rather recently, in this bathyal subclade of clade F ( Figs 1, 2 View Figure 2 ).

MONOPHYLY OF RISSOOIDEA

Addressing the question of the monophyly of the Rissooidea was not the main focus of this study. However, given that our molecular analyses are based on the most complete sampling available for this superfamily, and as some results are in conflict with those of other studies, this work may shed light on this yet controversial issue. The superfamily Rissooidea was maintained by Criscione & Ponder (2013, fig. 2) with Hebeulima (Vanikoroidea: Eulimidae ) as a sister to the rissoinid – barleeiid clade. A rissoinid – eulimid clade was retrieved in nearly all previous caenogastropod phylogenies ( Colgan et al., 2007; Ponder et al., 2008). However, those studies included no other rissooidean family. Subsets of the rissoidean families were included as outgroups in two molecular studies of other caenogastropod groups ( Wilke et al., 2013; Takano & Kano, 2014). In a phylogeny of the ‘hydrobioids’ (Truncatelloidea), Barleeia and Rissoidae were found to be only distantly related ( Wilke et al., 2013). In a molecular phylogeny of the Eulimidae , rissooideans clustered together in a monophyletic group when five genes were used ( Takano & Kano, 2014, fig. 2). However, a tree based on two genes did not support a close relationship of the rissoids with rissoinids, zebinids and barleeiids ( Takano & Kano, 2014, fig. 1); the latter three families were more closely related to Eulimidae and Vanikoridae , albeit without including Emblandidae or Lironobidae in the analysis. Although with relatively weak support, our phylogenies ( Figs 1, 2 View Figure 2 ) suggest a monophyletic Rissooidea that is only distantly related to Eulimidae . However, the monophyly of this diverse superfamily can only be investigated with a more comprehensive sampling of its component genera, by including several critical outgroup taxa (e.g. Eulimidae and Vanikoridae ), and ideally by combining multi-gene phylogenies with an assessment of key morphological traits. Accordingly, we maintain the tentative recognition of this superfamily as distinct from the Eulimoidea, in accordance with Criscione & Ponder (2013).

CONCLUDING REMARKS

This study represents the first attempt to investigate the relationships within the Rissooidea in a cladistic phylogenetic framework. By producing a phylogeny based on molecular data from the largest number of rissooidean samples to date, we unearthed considerable amounts of previously undetected diversity within the Rissooidea , challenging the current, exclusively phenetic, systematics of the group. Our work is only a glimpse of the evolution of one of the oldest, most widespread megadiverse groups within the Caenogastropoda, which still remains largely neglected. Further studies, based on better taxon sampling and larger amounts of molecular data than the present one, are required to improve the understanding of the rissooidean systematics and evolution.