Glossodoris buko, Matsuda & Gosliner sp. nov.

Figures (1A–C, 2A, 3A–E, 4C–D)

Glossodoris pallida (Rüppell & Leuckart 1830), misidentification, Rudman 1990: figs. 9c, 10e–f; Gosliner et al. 2008: 240, third photo; Turner & Wilson 2008; Gosliner et al. 2015: 237, upper right photo.

Glossodoris xantholeuca Ehrenberg 1831: 92; Rudman 1984.

Glossodoris sp. A Matsuda & Gosliner 2017.

Type Material. Holotype: CASIZ-223284 (ex CASIZ- 191102 B) 4 mm preserved, Papua New Guinea, Madang Province, Bilbil Island, coll: V. Knutson, 10 November 2012, orig. fixative 95% EtOH, GenBank: KT600713 (COI) .

Paratypes: CASIZ- 191102 A, 13 specimens,1 dissected, 3–8 mm, same collection data as holotype. CASIZ- 0 86381, 6 specimens, 1 dissected, 6.5–11 mm, Papua New Guinea, North Coast, North of Madang, approx. 1 km South of Cape Croisilles, South side of The Quarry, coll: T.M. Gosliner, 13 June 1992, orig . fixative Bouin’s solution. CASIZ-181594, one specimen 6 mm preserved, Philippines, Bohol Island, Panglao, Pontog Lagoon I, reef wall with small caves, coll: T.M. Gosliner, Y. Camacho, J. Templado, M. Malaquias, M. Poddubetskaia, 2 Jul 2004, Panglao Expedition 2004, 17–25 meters, orig. fixative 95% EtOH. CASIZ-177264, one specimen, Philippines, Luzon Island, Batangas Province, Tingloy, Caban Island, Layaglayag, coll: T. Gosliner, A. Valdés, M. Pola, L. Witzel, B. Moore, A. Alejandrino, 16 Mar 2008 .

Comparative material of Glossodoris pallida (Figs. 1D–1E): CASIZ-173393, one specimen, dissected, 9 mm preserved, Madagascar, Iles Radama, Nosy Valiha, W of Nosy Valiha, coll: T.M. Gosliner, 20 Oct 2005, 12– 13 m, orig. fixative Bouin’s solution. CASIZ-173395, one specimen, dissected, 9 mm preserved, Madagascar, Iles Radama, Nosy Kalakajoro , West of Nosy Kalakajoro and Nosy Beratia, coll: T.M. Gosliner, 19 Oct 2005, 13– 15 m, orig. fixative Bouin’s solution. CASIZ-176997, one specimen, 14.5 mm preserved, Mozambique, Inhambane Province, Jangamo, Pandane Beach, coll: M. Pola and J. Reis, 6 Feb 2008, 1.5 meters, orig. fixative 95% EtOH. CASIZ-175548, one specimen 4 mm preserved (large portion missing from tissue sample), Madagascar, Iles Radama, Nosy Kalakajoro, coll: S. Fahey and T. M. Gosliner, 13 Oct 2005, CAS-WCS Radama Islands Expedition, 15-20 meters, orig. fixative 95% EtOH. CASIZ-194338, one specimen, dissected, 6 mm preserved, Madagascar , South Madagascar, “Pointe Evatra, crique fond rocheux et gazon d’algues”, coll . South Madagascar Expedition, 30 Apr 2006 – May 2010, 3–8 meters, orig. fixative 95% EtOH. CASIZ-175554, one specimen, dissected, 2 mm preserved, Madagascar, Iles Radama, Nosy Faly, NW of Nosy Faly, coll: S. Fahey and T.M. Gosliner, CAS-WCS Radama Islands Expedition, 13–16 meters, orig. fixative 95% EtOH.

Etymology. The name Glossodoris buko comes from buko (young coconut) owing to the resemblance of this species to the cream-colored coconut meat from the Philippines, where this species is found.

Geographical Distribution. Specimens identified by Matsuda & Gosliner (2017) range from the Philippines to Papua New Guinea and Australia (Turner & Wilson 2008).

External Morphology. Glossodoris buko has a long and slender body that is transparent white in color (Fig. 1A–C). There is an opaque white band that starts anteriorly on the mantle, narrows between the rhinophores and then widens again and narrows between the two major folds in the middle of the mantle, and ends circling the gills (Fig. 2A). In most specimens, the band is continuous, however there was a break in the band posterior to the rhinophores in some. A white opaque band runs the length of the foot on both sides and connects posteriorly. The mantle edge is rippled with the semi-permanent undulations that are characteristic of all Glossodoris . One primary pair of undulations midway on the mantle is identifiable by an indentation of the white dorsal band. A thin, light yellow marginal band runs the length of the outer edge of the mantle and opaque white mantle dermal formations that appear as a thick white band that lies partly under the yellow band. The degree and number of the smaller semi-permanent undulations varies between individuals, but larger specimens have more pronounced undulations. A yellow band borders the foot, however no white band is visible due to the absence of mantle dermal formations. The rhinophores are elongate and conical with 11–12 lamellae. The bases of the rhinophores are white and the tips are yellow. The gill forms a semicircle surrounding the anus opening posteriorly, consisting of approximately 5–8 unipinnate branches. The lamellae are white with yellow tips and are shorter at the ends of the arc. The genital pore is located on the right side of the body below the mantle and behind the rhinophores.

Internal Anatomy. Radula and buccal armature (Fig. 3A–E). The radular ribbon (Fig. 3E) is short and wide with a radula formula for a preserved specimen of 3 mm of approximately 28 x 14.1.14 (CASIZ- 191102 A) and 13.1.13 for a 6.5 mm preserved specimen (CASIZ-086381). The rachidian tooth (Fig. 3A) is very reduced and quasi-triangular. The first lateral tooth is almost bilaterally symmetrical. It is wide and has a relatively short triangular pointed central cusp. There are approximately five well-defined denticles, each about half the length of the central cusp, that point down and outward on the inner and outer sides of the central cusp. The mid-lateral teeth (Fig. 3B) are longer and have a shorter central cusp with approximately 6–8 loosely packed and well defined denticles solely on the outer edge. Unlike the first lateral, the central cusp on the mid-laterals is almost indistinguishable from the denticles next to it. The denticles in the mid-laterals are almost indistinguishable in size and shape from the central cusp, and maintain this shape and their size integrity to the edge of the ribbon (Fig. 3C). The jaw rodlets are short and well-spaced with distinct gaps between rodlets. They are predominantly bifid (Fig. 3D) with a few trifid rodlets. The ventral side of the buccal mass has a glandular sheath covering the oral tube that contains numerous densely packed white opaque glands (Fig. 4D).

Reproductive system (Fig. 4C). The penial bulb is long and folded and leads to a coiled vas deferens followed by an approximately equal in length prostate gland. The receptaculum seminis duct and the vagina are relatively short. The receptaculum seminis is slightly smaller than the bursa copulatrix, and they are found adjacent to each other rather than being aligned linearly.

Remarks. At first glance, G. buko and G. pallida could be easily confused, as there are few external morphological differences (Fig. 1; Rudman 1984: fig 1b; Gosliner et al. 2015: 237, upper right fig.; Matsuda & Gosliner 2017: fig. 1). The holotype of G. pallida was collected from the Red Sea and subsequently examined by Rudman (1984) together with a specimen from Tanzania. Both specimens share the same color pattern, radular structure and reproductive system morphology as the five specimens of G. pallida we comparatively examined here from Madagascar and Mozambique (Figs. 1D–F, 3F–J). In Rudman’s (1990) G. pallida description, he noted that specimens from East Australia have yellow gills and rhinophores, which are consistent with G. buko, whereas his Tanzania and Sudan specimens, the rhinophores and gills are white. However, he further remarks that yellow tips were reported from specimens in the Red Sea and Reunion Island, indicating that yellow tips may not be a consistent identifier for G. buko . The rhinophores and gills of our G. pallida specimens from Madagascar and Mozambique all have frosted yellow tips, although the yellow is not as bright as in the G. buko specimens.

The most striking differences between the two species are found in the radula, jaws and buccal mass. The radular ribbon of G. pallida (Fig. 3J) is elongate (~ 85 x 23.1.23 CASIZ-173395), whereas it is short and squat in G. buko (Fig. 3E) (~ 28 x 14.1.14) for specimens of comparable size. This is confirmed in Rudman’s G. pallida specimen from the Red Sea (type locality), which has a radular formula of 108(+4) x 39.1.39 (15 mm specimen alive) and 23.1.23 for his Australian specimen (9 mm preserved) (Rudman 1984), and while no length was reported, the number of lateral teeth and presence of a rachidian tooth suggest that this specimen is likely G. buko . Glossodoris pallida has a more prominent and pointed rachidian tooth (Fig. 3F), whereas the rachidians are significantly reduced and amorphous in G. buko . There are also significant dissimilarities in the lateral teeth. The East African G. pallida specimens examined here share the same lateral tooth structure as Rudman’s Red Sea specimen (1984), with a long narrow hook-shaped central cusp with well-defined short denticles resting flat against the outer edge (Fig. 3G, H). Glossodoris buko has a lateral tooth that is broad and concave with a central cusp that is almost indistinguishable from the denticles in size and shape that shares no similarities with G. pallida . These differences are also visible in the jaw. Glossodoris pallida’ s rodlets are long, curved, and tightly packed with bifid tips where one of the points protrudes from slightly below the adjacent rodlet (Fig. 3I). Glossodoris buko has short and loosely packed rodlets that are less consistent in shape (Fig. 3D). Finally, the large glandular sheath on the ventral side of the buccal mass in G. buko (Fig. 4D) is not present in G. pallida (Fig. 4B).

The reproductive system appears similar to the description by Rudman (1983) for G. pallida from Tanzania and the Red Sea and examined here from Madagascar (Fig. 4A). The only noteworthy distinction in the reproductive systems of the two species is that the ejaculatory portion of the vas deferens of G. pallida (Fig. 4A) contains many more convolutions than does that of G. buko (Fig. 4C).

Glossodoris buko is distinct from G. pallida in both internal morphology as shown here, and based on molecular analyses (Matsuda & Gosliner 2017). This distinction is also maintained geographically. Glossodoris pallida has only been recorded off the coast of eastern Africa and the Red Sea, and G. buko is solely from the western Pacific. Matsuda & Gosliner’s (2017) phylogeny of Glossodoris provides support for the splitting of the previously hypothesized Glossodoris pallida into two distinct species. This is further supported through p-distance values (Matsuda & Gosliner 2017). Within G. buko, a grade and one clade are supported (a grade from the Philippines and a clade from Australia and Papua New Guinea). However, there were no observed morphological differences and only a 5–6% p-distance between them and these were not recovered as distinct lineages in the ABGD analysis conducted by Matsuda and Gosliner (Fig. 5). This strongly supports that the western Pacific specimens represent a single species distinct from the Indian Ocean specimens (Matsuda & Gosliner 2017).