Redescription and reassignment of Ondina semicingulata to the Pyramidellidae, with review of the occurrence of genus Evalea in the Western Atlantic (Gastropoda)

Acteon semicingulatus Dall, 1927, previously known only by its original description is reassigned to the Pyramidellidae, in Ondina, based on the collecting of several new specimens along the coast of Brazil, in the same bathymetry as the type locality. Its shell shape variation is discussed and Odostomia (Evalea) ryclea Dall, 1927 is considered a synonymy. Other Western Atlantic species, previously allocated to other genera are transferred to Ondina: Aclis striata Verrill, 1880, Odostomia (Iolaea) hendersoni Bartsch, 1909, Evalea stocki De Jong & Coomans, 1988 and Odostomia (Evalea) emeryi Bartsch, 1955 based on conchlogical comparison to the revison by Høisæter (2014), from Northeastern Atlantic. The genus Evalea is considered to be absent in the Atlantic Ocean.


Introduction
The Pyramidellidae Gray, 1840 is a notoriously rich and taxonomically complex gastropod family. It was included in the "big-five" group of the richest mollusks families by Albano et al. (2002) after extensive surveys in New Caledonia. It includes about 3,000 accepted names at the species level which are classified in around 140 accepted genera (MolluscaBase 2018).
Although recent advances and changes in the phy¬ logenetic position of the family within Heterobranchia have been proposed (e.g. Dinapoli and Klussmann-Kolb 2010), the alpha taxonomy of several genera remains to be revised and knowledge of diversity and distribution of species is still far from satisfactory. Taxonomic stud¬ ies in regions where the Pyramidellidae was poorly stud¬ ied, generally reveal several new species (e.g. Penas and Rolan 2010, in the South Pacific region). Knowledge of Pyramidellidae diversity in Brazil grew considerably af¬ ter several taxonomic works (e.g. Pimenta and Absalao 2001a, b, 2002, 2004a, 2004b, Pimenta et al. 2000, 2008, 2009, 2011, Pimenta 2012, but several genera remain to be revised. Additionally, old published names should be revised in both nomenclature and taxonomy. This is the case of Ondina de Folin, 1870, a genus with 20 valid species (MolluscaBase 2018), mainly from Eu¬ ropean and west African waters but so far not recorded in the Tropical western Atlantic. The nomenclature of On¬ dina was revised by van Aartsen (1984) and the Eastern Atlantic species (from Europe and Africa) were re-de-scribed in a series of papers (e.g. van Aartsen 1987, van Aartsen et al. 1996, Penas et al. 1996, Penas and Rolan 1999, Waren 1991, Hoisaster 2014.
During a visit to the USNM collection, the type se¬ ries of Acteon semicingulatus was examined, which led to its reassessment in the Pyramidellidae, genus Ondi¬ na :, as well as in the review of the previous records of the Pyramidellidae genera Ondina and Evalea in the Western Atlantic.

Material and methods
Taxonomic identification of the new material from Brazil was based on conchological comparison with type spec¬ imens and with the recent revision by Hoisaster (2014). All available material consists of dry shells; in the "Ex¬ amined material" the number of shells is indicated be¬ tween brackets.
For detailed examination, shells were prepared follow¬ ing the standard methods to preparation of micromollusc shells for SEM of Geiger et al. (2007) and observed by scanning electron microscopy at the Centro de Microscopia Eletronica, Departamento de Invertebrados, Museu Nacional/UFRJ, with a JEOL JSM-6390LV microscope.
Measurements were made with the software ImageJ (Rasband 2012 Redescription. Shell small, thin, up to 3.2 mm, width -50% of length; oblong ovate to biconical; color trans¬ lucent white; spire regularly conical, -30° angled, -40% of shell length. Protoconch heterostrophic, helicoidal, with about one smooth whorl, intorted, oriented -180°t o teleoconch axis, immersed into first teleoconch whorl, with no visible nucleus; width about 210 pm; transition with teleoconch weak, hardly discernible. Teleoconch up to four stepped whorls, each whorl slightly convex, last whorl somewhat globose; suture deep, forming a nar¬ row furrow. Axial sculptured absent, except for growth lines; spiral sculpture with very narrow spiral grooves of variable distribution and number; usually restricted to the periphery, near the area of implantation of the outer lip, extending anteriorly up to about 1/3 of last whorl and on the base; young shells with three-four very thin fur¬ rows; some adult shells with up to 20 furrows covering all base surface and extending anteriorly on last whorl more than half of its length, but not reaching anterior suture; adult shells with visible furrows just above suture; spiral furrows covered by microscopic axial threads. Aperture zse.pensoft.net elliptical-oblong, length about half of last whorl length, anteriorly elongated-rounded, posteriorly narrow and somewhat acute. Columellar margin slightly concave, without tooth. Outer lip thick. Umbilicus deep and wide, ranging from circular to wide chink. Geographic distribution. USA: Florida (type locality); Barbados (present study); Brazil: Amapa, Ceara, Rio Grande do Norte, Bahia and Santa Catarina states (present study).
Remarks. Except for the features of soft parts, absent in the type specimens of Acteon semicingulatus , all char¬ acteristics agree with the diagnose provided by Hoi sac ter (2014) for Ondina, including the oblong-ovate shell, ob¬ long aperture and intorted protoconch. In comparison to the Eastern Atlantic species, which usually have more elongate shells, O. semicingulata has a wider shell, with biconical general shape. The original allocation in Acteon is rejected since Acteonids have a solid shell, with colu¬ mellar tooth and rounded protoconch (Valdes 2008).
Hoisaster (2014) discussed the high variability of the surface of the shell, being smooth or with variable incised spirals. According to this author, in a single species, the spirals may cover uniformly the whole shell, they may be confined to the lower half of each teleoconch whorl or the shell may be smooth and shiny.
Ondina semicingulata (  Odostomia {Evalea) ryclea with type locality off Goergia is here considered a synonymy name of Ondina semicingulata, since it has identical shell shape and pro¬ toconch. Despite the eroded shell surface of the holotype (Figure 2a- Remarks. Aclis striata was described (Verrill 1880) based on two shells: from shallow water in the Bay of Fundy; and from deep-water off Newport, Rhode Island, by the USFC. Two years later, Verrill (1882) referred to the same two shells, adding station number information (USFC sta. 873) to the the deep-water shell. In this work, Verrill (1882: pi. 58, fig. 13) presented the drawing of a shell without indicating which one of the two syntypes.
The type material of Aclis striata was studied by John¬ son (1989: 66, pi. 11, fig. 7), who designated and figured the lectotype (YPM 15757, from the Bay of Fundy), and listed YPM 15704, from USFC sta. 873, as a paralec¬ totype (but see comments below about its type status). After searching the YPM Molluscan Collection (E. Lazo-Wasen pers. comm.), it was noticed that the lectotype was missing (the vial was empty) and the label of the sup¬ posed paralectotype YPM 15704 (Figure 2h) mentions 'USFC sta. 863' (which is a shallow water station in the Vineyards Sounds, 33 m depth).
According to Johnson (1989: 15), Verrill sent the samples (types and other specimens) of the species de¬ scribed by himself to the National Museum of Natural History (USNM), keeping nevertheless, some duplicate specimens with him, which he later sold to YPM (Lazo-Wasen pers. comm.). Thus, it seems that the two original syntypes of Aclis striata were split by Verrill between the USNM and the YPM collections. This is corroborated by the original label of USNM 44820 (Figure 2g) that states: " Aclis striata V. Off New¬ port, R. I. U.S.F.C. 1880". This label has a hand-written indication of "figd. type" that corresponds to Verrill's cal¬ ligraphy (E. Lazo-Wasen pers. comm.); such "figd. type" (Figure 2d-e) matches perfectly the figured syntype by Verrill (1882: pi. 58, fig. 13), and thus, it must be consid¬ ered the paralectotype from deep-water (USFC sta. 873), while the missing lectotype (figure in Johnson 1989: pi. 11, fig. 7), by the other hand, is a smaller shell.
Thus, the shell YPM 15704 (Figure 2h), considered by Johnson to be a paralectotype has no type status. As indicat¬ ed by its label, it was dredged by USFC is sta. 863, in which an additional shell (ANSP 102517) was also collected.
The only remaining question is the reference to USFC sta. 863 in the original label of specimen USNM 44820 (Figure 2g). We believe that such information was mis¬ takenly inserted by someone later since it has a different zse.pensoft.net handwriting and Verrill (1880, 1882) did not list material from that station.
Ondina striata was originally described as belonging to Adis due to its spiral striae and was later transferred to the Pyramidedidae genus Odostomia (Menestho) by Bartsch (1911: 435) without any comments on this tax¬ onomic rearrangement. The species has all features that characterize the genus Ondina (Hoisaster 2014) and is very similar to the type species of the genus.
Material examined. Photographs of the types.
Remarks. Ondina stocki (Figure 3c-e) was originally de¬ scribed in Evalea , based on a sample of shells from Aruba/Curagao; the original illustration is a simple drawning that precludes a precise generic allocation, but the pho¬ tographs of the holotype (Figure 3c) and of a young and an adult paratypes (Figure 3d-e) clearly shows that this species belong in the genus Ondina. The shell shape is similar to Ondina hendersoni , but the shell is narrower, and the spiral grooves are less numerous.
Lee (2009) recorded Evalea stocki from Jacksonville Beach, Florida, (USA) based on a young specimen. The author, in the same work, referred to a color illustration of that species, published by Gundersen (1998), but named by Gundersen (1998) as Amoura hendersoni. The figured shells from Sanibel Island in Gundersen (1998) and that from Jacksonville in Lee (2009) clearly can be ascribed to the genus Ondina , but the images provided by the au¬ thors as well as the drawing of Amoura cf. hendersoni by Ode (1994), do not allow a precise separation between O. stocki or O. hendersoni. Also, the simple drawing of Amoura cf. hendersoni from Texas in Ode (1994) does not allow a conclusive identification, in spite of the pat¬ tern of spiral grooves resembling that found in O. stocki.
Therefore, we consider that the only confirmed occur¬ rence of O. stocki is that restricted to the type locality area.

Discussion
Knowledge of species richness, geographic distribution and accurate taxonomic status of marine molluscan fauna from Brazil is still far from satisfactory. Traditionally, the marine molluscan diversity was compiled in catalogues (e.g. Rios 1994Rios , 2009, with the main purpose of working as identification guides, with a brief description of each taxa, and an image. Despite the recent contributions to the Brazilian molluscan diversity (e.g. Pimenta et al. 2004a), several taxa remain poorly known, and the revision of genera usually gives rise to the discover of new records and/or new species (e.g. Fernandes and Pimenta 2011). This is especially true to the marine micromollusks, as discussed by Pimenta and Geiger (2015).
For both families, re-examination of type material is imperative and in the case of Acteon semicingulatus , its reassignment to the Pyramidellidae revealed the first re¬ cord of the genus Ondina in the Western Atlantic.

On the occurrence of Ondina in the Western Atlantic
The most complete and recent account on the taxonomy of Ondina was provided by van Aartsen (1987), Waren (1991 and Hoisaster (2014), based on European species. Hoisaeter (2014) figured the type species Ondina semiornata de Folin, 1870 (= Ondina warrenii) and discussed the intraspecihc variation found in the genus.
According to the works by van Aartsen (1987), van Aartsen et al. (1998), Penas et al. (1996, Penas andRolan (1999), Waren (1991), and Hoisaster (2014), On¬ dina is mainly distributed in the Eastern Atlantic, both in temperate and tropical latitudes, including the Northern European Seas, the Lusitanian, the Mediterranean Sea and West African Transition Provinces, while a single species is known from the temperate northern Pacific, in the coast of Japan (Hori and Fukuda 1999).
The present records of Ondina semicingnlata, O. em¬ ery i, O. stocki and O. hendersoni in the western Atlantic broadens the distribution of the genus including geo¬ graphical areas that go from Georgia (north-western At¬ lantic) and south Brazil (south western Atlantic), which gives the genus a wider latitudinal range in the western Atlantic when compared to its distribution in the eastern part of the Atlantic. Ondina semicingnlata , besides pre¬ senting a wider distribution in the western Atlantic, also presents a larger bathymetric range from 72-500 m depth. Ondina mosti van Aartsen, Gittenberger & Goud, 1998 is the only eastern Atlantic species with similar bathymetry (119^105 m), while the other species are restricted to lit¬ toral and continental shelf depths (Aartsen et al. 1998).

Evaluation of the occurrence of Evalea in the Western Atlantic
Aartsen (1987) evaluated the generic allocation of sever¬ al European odostomids species previously included in Evalea and based on the absence of a well-developed columellar tooth, trasnferred them to Ondina. This criterion was followed by Hoisaster (2014).
According to Aartsen and Menkhorst (1996: 51-52), Evalea was confused by Nordsieck (1972), most likely because of the lack of an illustration of the type species Odostomia (Evalea) elegans A. Adams, 1860. Aartsen and Menkhorst (1996: fig. 11) designated a neotype for the type species of Evalea which shows a shell with coarse spiral striae covering all whorls and a visible columellar tooth.
According to MolluscaBase (2018), Evalea has 18 species and it is not present in European waters (e.g. East Atlantic and Mediterranean), being restricted to the Pacif¬ ic coast of Japan, New Zealand, Indian South Africa, and to the western Atlantic.
In the present work, an attempt was made to eval¬ uate the presence of Evalea in the western Atlantic by checking the previous species recorded in that genus and comparing them to the type species as illustrated by Aartsen (1996). According MolluscaBase (2018), the fol¬ lowing species are recorded from the western Atlantic: Evalea fernandina (Bartsch, 1927), E. ryclea (Bartsch, 1927), E. emeryi (Bartsch, 1955), and E. stocki De Jong & Coomans, 1988;besides that, Ode (1994) recorded an aditional taxon as " Evalea sp. indet. A".
As demonstrated above, Evalea stocki, E. ryclea (= O. semicincgulata) and E. emeryi belong in fact to the genus Ondina. As for the other species we do not have at present enough evidence to critically discuss their ge¬ neric placement.
Odostomia {Evalea) fernandina (Dali, 1927: 85) was originally described in Odostomia and it has an elongate shell with whorls with an almost rectilinear outline. Al¬ though the holotype, USNM 108053, (Figure 4a, b) is eroded and has a partially broken outer lip, it is possible to distinguish from Evalea by the absence of spiral sculp¬ ture and columellar tooth. However, it does not seem to belong to Ondina, because of the different protoconch which is not fully immersed. Thus, until further evidence is available we suggest this species to be kept in the genus Odostomia. The record of Evalea sp. A by Ode (1994: 46, fig. 8) is considered doubtful because in the drawing provided by the author, characters that could relate the specimen to Evalea such as the spiral sculpture through¬ out the shell and the presence of a columellar fold, are difficult to interpret. Therefore, based on previous studies (e.g Hoisaster 2014) and our own results it is here suggested that the genus Evalea is absent in the Atlantic Ocean.