Gemmula alwyni Kilburn 2005, sp. n.

Gemmula alwyni Kilburn, sp. n.

Figs 3, 4

Gemmula kieneri (non Doumet, 1840): Cernohorsky, 1987: 126, figs 9, 10.

Gemmula unedo (non Kiener, 1840): Medinskaya, 2002:136, 143, figs 1F, 9 (foregut anatomy), 22E (radula).

Etymology: Named in honour of Mr Alwyn Marais, writer of many popular articles on South African shells.

Description: Shell fusiform, b/l 0.32–0.36, a/l 0.42–0.53, approx. 10 teleoconch whorls, last whorl trigonal, siphonal canal moderately long, narrow and tapering, whorls with an angular, more or less median shoulder, suture narrow, slightly notched. Outer lip strongly convex in side view, with a slight stromboid notch, anal sinus openly linguiform, directed slightly adapically.

Subsutural region not riding up previous whorl, without a differentiated cord but a single fine, angular thread with 1–3 finer threads at suture itself; sulcus not defined but area between suture and peripheral (sinus) cord broadly concave and sloping, with 3–7 widely-spaced main spiral threads and an occasional finer intermediary. Peripheral cord nearly basal on early whorls, slightly above periphery on penultimate whorl; crest of sinus cord flatly declivous, with steep sides, bearing gemmules crossing spiral ridges; ridges rounded, 2–3; gemmules prominent and tubercular on early whorls, but become low and more lunulate with age, in axial profile undulating, sometimes weak or flattened on terminal half-whorl; 19–20 gemmules on early whorls, 23–32 on penultimate whorl. Base of each whorl initially with 1 spiral ridge, increasing to 3–4 main spiral ridges, their intervals wide-set, with fine spiral threads. Ridges on base of last whorl somewhat angular, sometimes irregular in strength, wide-set with 1–3 finer intermediaries; approx. 25–30 main ridges on base of last whorl, weak to obsolete towards end of rostrum. Basal ridges rendered rugose but not nodose by coarse collabral threads.

Protoconch (Madagascar paratype) conical, apical whorl missing, remainder with coarse, orthocline axial riblets, 13 on last whorl, breadth 0.93 mm.

Pale brown to brownish-white, with axial streaks of brown, forming dots or blotches on main spiral ridges, gemmules white with orange-brown interstices.

Operculum obovate with emarginal nucleus, brownish-yellow.

Dimensions: 71.0 x 23.4 mm, aperture length 32.2 mm (holotype); largest paratype 81.8 x 29.7 mm, aperture 38.6 mm, smallest 49.1 x 15.8 mm, aperture 25.8 mm.

Comparison and remarks: Of described species, G. alwyni sp. n. resembles most closely G. vagata (E.A. Smith, 1895) from the northern Indian Ocean and the reportedly widely distributed G. unedo (Valenciennes in Kiener, 1840). A third member of the species complex may be mentioned, namely G. invicta (Melvill, 1910), described from the Arabian Gulf. Although invariably cited as a synonym of G. unedo, having a peripheral cord that is gemmulate only in juveniles, the holotype of Turris invicta (BMNH 1911.6.21.10) differs from G. unedo s.s. in its strongly angular subsutural cord.

Gemmula vagata: A syntype (Fig. 5) of Pleurotoma vagata has been examined (ZSI M 181/1), also a more recent series collected by me from west of Cape Comorin, SW India, in 300– 400 m. G. vagata resembles G. alwyni in form, but bears even stronger gemmules, particularly in immature shells, the peripheral cord is narrower and much more projecting, and the subsutural ridge complex is low but more distinct.

Gemmula unedo: Dr B. Oliveira of the University of Utah regards G. unedo auctt. as comprising a complex of species, mainly undescribed, in the central western Pacific (pers. comm.). Based on my examination of the syntype of Pleurotoma unedo, preserved in the MNHN collection, I find that G. unedo s.s. differs from G. alwyni in the peripheral nodules becoming obsolete after a few teleoconch whorls; in the peripheral cord being outwardly (not adapically directed); in its nodose basal cords, and in possessing a distinct subsutural cord that rises high up the preceding whorl. This syntype (Fig. 6) is somewhat worn, with a damaged siphonal canal, but recently collected examples from Indonesia agree well.

Cernohorsky (1987), evidently misled by the row of inter-gemmular brown spots, identified Madagascar material of the present species as Gemmula kieneri (Doumet, 1840), and Medinskaya (2002) misidentified it as G. unedo. A specimen from the same ORSTOM samples, presented to NMSA by R.T. Abbott, agrees well with SE African specimens of G. alwyni, although smaller (largest of 95 shells seen by Cernohorsky = 60 mm; attains 65.1 mm according to Medinskaya). It is here designated as a paratype (NMSA L6785/T1614). The type material of Pleurotoma kieneri appears to be lost, but specimens agreeing with the type figure of this widely distributed species differ from G. alwyni in their narrower form, much less angular whorls, well-differentiated subsutural cord, canaliculated suture, more sharply defined gemmules and much larger protoconch.

Holotype: NMSA V9764 /T1874 (donated by J. and A. Marais). SOUTH AFRICA:off Amanzimtoti (approx. 30°03'S: 30°53'E), Natal South Coast, KwaZulu-Natal, 330–475 m, trawled.

Paratypes: between Amanzimtoti and Umhlanga, 330–475 m, trawled, NMSA W2799 /T2025, one; off Umhlanga, Natal North Coast, KwaZulu-Natal, 330–475 m, trawled, in J. Marais collection, one; off Xai Xai – Zavora region, Mozambique, 400–480 m, trawled, J. Rosado, NMSA L6785 /T1614, one; off Xai Xai, Mozambique, approx. 300–400 m, trawled, NMSA L4649 /T2023; Mozambique Channel, off NW Madagascar (12°36'0"S: 48°17'3"E), 300 m, NMSA J1575 /T2024, ex ORSTOM, one.

Distribution: Continental slope of KwaZulu-Natal and southern Mozambique, N to NW Madagascar, in approx. 280– 700 m.