Echinolittorina miliaris ( Quoy & Gaimard, 1833 )

Reid, David G., 2011, 2974, Zootaxa 2974, pp. 1-65 : 43-47

Echinolittorina miliaris ( Quoy & Gaimard, 1833)

( Figures 18C, D, 19–21)

Littorina miliaris Quoy & Gaimard, 1833: 484–485 , pl. 33, figs 16–19 (Île de l’Ascension; lectotype (here designated) MNHN 5764, seen, Rosewater 1970: pl. 381, figs 1, 2, Rosewater 1981: pl. 2J, Fig. 20H herein; 3 dry paralectotypes MNHN 5796, seen; 3 paralectotypes in alcohol MNHN 5796, seen). Deshayes, 1843: 208–209. Nevill, 1885: 154 (in part, includes E. natalensis (Krauss in Philippi, 1847), E. radiata (Souleyet in Eydoux & Souleyet, 1852)). Locard, 1886: 287. Kobelt, 1888: 169–170. Smith, 1890a: 283 (in part, includes E. helenae ). Smith, 1890b: 319.

Litorina miliaris —Philippi, 1847: 160, Litorina pl. 3, fig. 5. Weinkauff, 1882: 101, pl. 14, fig. 12.

Tectarius miliaris —H. Adams & A. Adams, 1854: 315. Tryon, 1887: 259, pl. 48, fig. 78 (in part, includes E. granosa , E. reticulata , E. feejeensis ( Reeve, 1857)) . Stearns, 1893: 334.

Litorina (Tectaria) miliaris — Weinkauff, 1883: 226.

Nodilittorina (? Granulilittorina ) miliaris — Rosewater, 1970: 424, 494.

Nodilittorina (Granulilittorina) miliaris — Rosewater, 1975: 10–11, fig. 3 (in part, includes E. granosa , E. helenae , E. vermeiji , Tectarius striatus ).

Nodilittorina (Granulilittorina) miliaris miliaris — Rosewater, 1981: 33–34, pl. 1J, K, 2I, J, pl. 3I (operculum), 4D (radula), pl. 6B, D, F, I, J (shell sculpture) (in part, includes E. vermeiji ).

Nodilittorina (Nodilittorina) miliaris — Bandel & Kadolsky, 1982: 36–37, fig 49 (shell, radula) (in part, fig. 50 is E. galapagiensis , here renamed E. lemniscata , see below).

Nodilittorina (Echinolittorina) miliaris — Reid, 1989: 99 (in part, with doubtful inclusion of E. helenae and E. vermeiji ).

Nodilittorina miliaris — Rosewater, 1978: 32–33 (in part, includes E. vermeiji ). Reid, 2002a: 259–281.

Echinolittorina miliaris — Williams et al., 2003: 83. Williams & Reid, 2004: 2227–2251, fig. 6F (map). Williams & Duda, 2008: fig. 1 (phylogeny).

Echinolittorina (Echinolittorina) miliaris — Reid, 2009: figs 1, 37 (phylogeny).

Litorina echinata Anton, 1838: 53 (no locality; types “not clearly identified” in MTD ( Schniebs 1995); synonymy fide Philippi 1847, Bandel & Kadolsky 1982).

Littorina granularis Gray, 1839: 140 (no locality; holotype BMNH 1887.4.26.9, seen, Rosewater 1970: pl. 381, figs 3, 4).? Reeve, 1857: Littorina sp. 68, pl. 13, fig. 68a, b.

? Litorina granularis —Philippi, 1848: 63, Litorina pl. 7, fig. 7.

? Tectarius granularis — Tryon, 1887: 260, pl. 48, fig. 84 (in part, includes E. millegrana (Philippi, 1848) , E. radiata , E. cinerea ( Pease, 1869) , E. vidua ( Gould, 1859)) .

Littorina nodosa — Reeve, 1857: Littorina sp. 13, pl. 2, fig. 13 (in part, includes E. australis ( Gray, 1826) ; not Gray, 1839 = E. australis ).

Litorina (Tectaria) nodosa — Weinkauff, 1883: 226 (in part, includes E. australis ; not Gray, 1839).

Tectarius nodosus — Tryon, 1887: 259, pl. 48, fig. 83 (in part, includes E. australis , E. natalensis , E. subnodosa (Philippi, 1847) , E. lemniscata ; not Gray, 1839).

Echininus nodulosus — Packer, 1968: 58 (fide Rosewater 1975; not seen).

Nodilittorina tuberculata — García-Talavera, 1983: 44–45, map (in part, includes E. tuberculata , E. vermeiji , E. granosa , E. helenae , T. striatus ).

Taxonomic history: Rosewater (1970, 1981) incorrectly referred to the ‘holotype’ of L. miliaris , but this is not considered a valid lectotype selection (ICZN, 1999: Art. 74.5), because in the original description Quoy & Gaimard (1833) mention that the species was very common and their figures 16–19 could well be of more than one specimen.

The identity of L. granularis Gray, 1839 is not in doubt, because the holotype is clearly a worn specimen of the present species, as pointed out by Smith (1890a) and Rosewater (1970). However, the name has been variously used by subsequent authors, particularly to refer to the Indo-West Pacific species E. millegrana , E. radiata , E. cinerea and E. vidua (e.g. Tryon 1887; see synonymies in Reid 2007). In the absence of the figured specimens, it is not clear if the figures of ‘ granularis ’ published by Philippi (1848) and Reeve (1857; copied by Tryon, 1887) are indeed of the present species, or of similar shells such as E. radiata or Tectarius striatus . Weinkauff (1882) interpreted Reeve’s (1857: sp. 68) figure as E. radiata (as Litorina exigua ).

Littorina lemniscata Philippi, 1846 was described with the locality indicated as ‘ Cuba?’. It does not resemble E. tuberculata , the only nodulose member of the genus in Cuba and Bandel & Kadolsky (1982) selected a lectotype (BMNH 1968216, seen) from three syntypes and identified it as a synonym of E. miliaris . The two white lines in the aperture of the lectotype exclude it from the Indo-Pacific subgenus Granulilittorina . The three rows of large nodules on the last whorl (one basal, two at and above periphery) and smaller sutural row differ from the four or more equal rows of nodules in E. miliaris , and the narrow shape is rare in that species. In one paralectotype the whorls are rounded and the nodules weak, with microsculpture of strong spiral threads and microstriae. This combination of characters and the range of intraspecific variation correspond only with the species known as E. galapagiensis ( Stearns, 1892) (see Reid 2002b: fig. 7L–Q), and that taxon should now be renamed E. lemniscata ( Philippi, 1846) . The species is found in the Galapagos Islands and there is a single record from the mainland of Ecuador (Reid 2002); it has recently been discovered at Isla del Coco (BMNH; D.G. Reid, unpublished). Interestingly, the name Hamus lemniscatus was used for this species in the first account of the molluscs of the Galapagos Islands, although without explanation or description ( Wimmer 1880).

There has been a long history of confusion and misidentification of E. miliaris . As it was the first of the nodulose Echinolittorina species from the eastern Atlantic to be described, the name L. miliaris (or its combinations) has often been used to include E. granosa and/or E. helenae by those authors with a broad species concept (e.g. Smith 1890a; Tryon 1887; Rosewater 1975, 1981). The western Atlantic species E. vermeiji has also been included under this name ( Rosewater 1975, 1978, 1981; Reid 1989) and there has been occasional confusion with T. striatus ( Rosewater 1975) . Tryon (1887) even included the Indo-West Pacific E. reticulata and E. feejeensis under T. miliaris . The name L. nodosa (a nodulose form of the Australian E. australis ) enters the synonymy following the misidentification by Reeve (1857; as pointed out by Smith 1890a), which was also adopted by Weinkauff (1883) and Tryon (1887).

Diagnosis: Shell sculptured with 4–7 narrow, widely-spaced cords with small, pointed white nodules. Oviduct with expanded muscular chamber between albumen gland and posterior end of bursa. Ascension I. COI: GenBank AJ623015 , AJ623016 .

Material examined: 22 lots (including 11 penes, 6 sperm samples, 10 pallial oviducts, 4 radulae).

Shell ( Fig. 20): Mature shell height 5.7–22.3 mm. Shape turbinate to high turbinate (H/B = 1.21–1.81, SH = 1.40–2.23); spire whorls rounded; suture distinct; spire profile straight, slightly concave near apex; periphery of last whorl rounded. Columella short, hollowed and slightly pinched at base, usually with an anterior lip; eroded parietal area. Sculpture of 4 narrow spiral cords, from just below periphery to suture, bearing small sharp nodules, equally developed on each cord (occasionally becoming obsolete on last whorl, Fig. 20K); in well-preserved shells a single thread (range 0-3) is present between the cords and entire surface is covered with spiral microstriae; largest shells ( Fig. 20I) with up to 7 nodulose cords at and above periphery; sculpture often eroded; base with 4–6 nodulose ribs. Protoconch 0.29 mm diameter, 2.8 whorls. Ground colour grey, with brownish cords bearing white nodules; aperture dark brown with pale band at base and sometimes a second band at shoulder; columella purple brown.

Animal: Head black; tentacle pale around eye and across base, with two longitudinal black stripes, often fused; sides of foot black. Operculum: opercular ratio 0.40–0.55. Penis ( Fig. 21A–E): filament 0.5–0.6 total length of penis, with annular wrinkles for most of its length (so not clearly differentiated from wrinkled base), smooth only at rounded tip, sperm groove ends terminally; mamilliform gland small, about one-fifth size of large penial glandular disc, borne together on stout projection of base; penis pigmented at base. Euspermatozoa 54–64 µm; paraspermatozoa ( Fig. 21G, H) round to oval, containing 1–2 parallel-sided rod-pieces 9–19 µm, with rounded or truncated ends, sometimes small and irregular, sometimes projecting from cell, which is packed with large round granules. Pallial oviduct ( Fig. 21F): copulatory bursa separates at posterior end of straight section and is separated from spiral of albumen gland by a large, folded expansion of basal chamber leading to seminal receptacle; small amount of additional glandular material present in a swelling around egg groove at anterior end of straight section. Spawn not known; pelagic capsules predicted from form of pallial oviduct ( Reid 2002a).

Radula ( Fig. 18C, D): Relative radula length 4.46–5.73. Rachidian: length/width 1.32–1.56; major cusp long, tip rounded to slightly pointed. Lateral and inner marginal: 4 cusps, tip of major cusp rounded. Outer marginal: 8– 11 cusps.

Range ( Fig. 19): Ascension I. only. Records: English Bay ( BMNH 20110147; USNM 735141); Georgetown ( BMNH 20110156; USNM 735068); North East Bay ( BMNH 20110149; USNM 735026); Boatswain Bird I., E coast Ascension I. ( BMNH 20110154).

Locard (1886) recorded this species from western France, but this is presumably incorrect (Kobelt 1907).

Habitat: On rocks above high tide ( Rosewater 1981).

Remarks: This is the sister species of another island endemic, E. helenae , and the K2P genetic distance between them for COI is 4.39%; reciprocal monophyly is found in analyses of COI and of 28S rRNA sequences, but only the former has significant support ( Williams & Reid 2004; Williams & Duda 2008; Reid 2009). The pallial oviduct of E. miliaris has a consistently larger expansion of the basal chamber of the spiral portion and shell shape is also diagnostic, implying likely nuclear genetic differentiation and supporting recognition of the pair as distinct species. The distance between the islands of Ascension and St Helena is 1290 km, within the known range of larval dispersal of Echinolittorina species (1200 to 2100 km; Reid 2002b, 2007). However, the islands lie inside the South Atlantic gyre where currents are weak, which may have permitted the isolation and divergence of their populations.

The majority of shells of E. miliaris cannot be mistaken for any other eastern Atlantic Echinolittorina species , for the 4–7 rows of small, pointed, white nodules that are not axially aligned are distinctive ( Fig. 20A–F, I). A few shells, generally of large size, have been seen with weaker sculpture ( Fig. 20G, H, J, K) and, as mentioned in the Remarks on E. granosa , confusion with that species is then possible. Another difference between these two is the more well-developed basal chamber of the pallial oviduct in E. miliaris . The occurrence of the present species on an oceanic island, while E. granosa is largely restricted to the African mainland coast, suggests an ecological difference also. Shells of the western Atlantic E. vermeiji are similar, with white nodules of even size, but these are arranged in only 2–4 rows and are of rounded shape ( Reid 2009: fig. 35). In the two other members of the subgenus, E. tuberculata from the western Atlantic and E. lemniscata (formerly E. galapagiensis , see above) from the eastern Pacific, the two rows of nodules at periphery and shoulder are larger than the small row at the suture ( Reid 2009: fig. 31; Reid 2002b: fig. 7L–Q).