REVISION OF THE INDO-PACIFIC SPHENOCARCINUS WITH A SINGLE ROSTRUM AND DESCRIPTION OF TWO NEW SPECIES

Two species of the genus Sphenocarcinus A. Milne Edwards 1876, with a single rostrum exist in the Indo-Pacific and they are principally only known by their original description: S. cuneus (Wood-Mason 1891) and S. aurorae Alcock 1899. Each species is herewith redescribed, based on the examination of the types. Two new species are established. S. difficilis sp. nov., represented by numerous specimens from Madagascar (ORSTOM collection), proved to be an intermediate species between S. cuneus and S. aurorae. S. pinocchio sp. nov., (collected in the Makassar Strait, Indonesia (mission Corindon II), is characterized by its very Jong and strongly curved rostrum. One male specimen, also collected in the Makassar Strait, probably modified by a sacculinid parasite, and with a more curved, snub rostrum, can be regarded as an unusual form of S. pinocchio sp. nov. A key of the four Indo-Pacific Sphenocarcinus with a single rostrum is presented.


INTRODUCTION
While studying the Indonesian material we came across to several specimens from Madagascar that proved to be another new species. We thought therefore it would be necessary to reexamine all the species belonging to single-rostrum group of the Indo-Pacific Sphenocarcinus. Genus Sphenocarcinus A. Milne Edwards, 1875.
The genus Sphenocarcinus A. MILNE EDWARDS (1875: pi. 17, fig. 5;1878;135) is represented by: 1) two American species: S. corrosus A. Milne Edwards 1875, the type-species, and S. agassizi Rathbun 1895, two "analogous species on opposite sides of the continent" (RATHBUN 1925: 187); both are characterized in having a rostrum formed of two long contiguous spines; 2) eleven Indo-Pacific species, into which we iiave just added three new species, which means that a total of 14 species are known from this area (cf. GUINOT & RICHER DE FORGES in press).
Regarding the Indo-Pacific Species, only two, very rare species, have a single rostrum: Sphenocarcinus cuneus (Wood-Masoii, 1891) and S. aurorae Alcock 1899; the rest of the Indo-Pacific species has a bifurcated rostruttn, with spines more or less long and divergent (we Notice that S. auritus Rathbun 1916, has a rostrum with two short flattened lobe's, separated [by a narrow silt (cf. GRIFFIN, 1976: 211, fig .10b). S. auneus is known by the original record's, from the Andaman Sea, and by two other records in the Western Indian Ocean (Providence I., Mombasa), S. aurorae is only known from the Travancore coast, first by the original report (ALCOCK 1899: 84), then by KEMP & SEWELL (1912: 30); both were collected-by the "Investigator" in two different surveys.
The recent discovery of Indo-Pacific Sphenocarcinus with a single rostrum is therefore interesting. The type material of S. cunem and S. aurorae, deposited in the Zoological Survey of India, Calcutta, was made available through the kindness of Dr. Maya Deb, Curator of Crustacea in this Institution, which enables us to redescribe and figure these two nice-looking species. This examination also allowed us to describe as a new species, S. difficilis sp. nov., a Sphenocarcinus from Madagascar represented by numerous samples, which seems to be intermediate between S. cuneus and S. aurorae. On the other hand, the discovery of Sphenocarcinus in the Makassar Strait lead us to establish a new other species, 5. pinocchio sp. nov. In a station, 5. pinocchio (one holotype, two paratypes) has a very long ro,strum, especially in the male; in another station of Makassar Strait, a sacculinized male speci-men has a very curious curved rostrum: probably the sacculinid may have modified this individual, the rostrum has then becaime (? abnormally) feminized? A such supposition requires further confirmation.

Remarks
Sphenocarcinus cuneus (Wood-Mason) (in: WOOD-MASON & ALCOCK 1891: 261) has been established for a sample containing two males and one ovigerous female, collected in the Andaman Sea (sta. 56, 220-240 fath.); a female individual (after tihe legend) is figured in Illustr. "Investigator" (ALCOCK & ANDERSON 1896, pi. 21, figs. 1, la). ALCOCK (1895: 193;1893: 50) described the species with more details, from the Andaman Sea (is it the same material ?) and indicated "161 to 250 fath." We could examine a specimen indicated "type" from the Andaman Sea, a male measuring 15 X 11 mm (rositrium only: 8 mm). This male type ( Fig. 1 A-D, pi. I, Figs. A-B) resembles very much the female figured in the Illustr. " Investig," (loc. cit.) in having the same disposition of the plates on the dorsal surface, as follows: an elongale-oval gastric plate; a triangular cardiac plate (Fig, ID), the anterior angle of which is touching the posterior border of the gastric plate; the preocular plate with the internal border salient (PL 1 fig. B) and with rounded angles; the branchial plate somewhat semilunar and very produced laterally. All the plates are strongly raised and separated by deep but not very wide channels. The surface of the plates is punctate, eadh puncture containing a very short spherical seta. The rostrum, of the male type seems to be a little longer than in the female of the Illustr. "Investig."; it is enlarged at its base and bears short setae; it is distinctly bifid at the tip. Some hook-shaped setae are present on the branchial lateral borders, on the gastric region (not on the gastric plate) and in the frontal basal region. A small tuft of hairs stands on each side of the cardiac plate. Alcock (ibid.) did not indicate the existence of variations in. the disposition of the plates of the three known individuals of S. cimeus; he noticed only that the rostrum is longer fn the males than in the female. The abundant material of the Sphenocarcinus collected at Madagascar shows to us the constancy of the characters (except the sexual dimorphism) in the species herewith described under the nafme S. difficilis sp. nov. (cf. infra). We, therefore, consider that S. cuneus is a species with a moderately long single rostrum, with a tria ngular cardiac plate nearly contiguous to the gastric plate; all the plates are separated by relatively not very wide channels.
Perhaps additional material will permit to verify all these characters and find other differences between S. cuneus, S. aurorae and S. difficilis sp. nov. (cf.infra), with the consideration of the presence oft the sexual dimorphism.

Distribution
Sphenocarcinus cuneus is known from the Andaman Sea where three (to be verified) specimens have been collected by the "Investigator". The reports from the Providence Island (Rathbun, loc. cit.) and from the East-African Coast (Mombasa: Griffin, loc. cit.) are questionable.

Remarks
In the same work where fee provided an extensive description of Sphenocarcinus cuneus (Wood-Mason 1891), ALCOCK (1899: 50) gave an addendum where he established, without a sketch, an additional species (ibid.: 84) originating "from off the Travancore coast, 224-284 fath.", S. aurorae, which closely resembles S. cuneus". Alcock had at his disposal 27 specimens of S. aurorae to compare with three specimens of S. cuneus and recognized several particulars separating the two species. Shortly after, KEMP & SEWEEL (1912: 30, pi. 1, fig. 10) mentioned 12 males and 15 females collected by the "Investigator" during the survey season 1910-1911,, alt ,sta. 391, 9°14'10"N -75°45'E, 260 fath. (about the same area than Alcock and the same number of specimens) and gave a drawing of a male S. aurorae. S. uarorae has not been hitherto reported. We could examine one specimen of the original Indian 'material indicated "type", an ovigerous female measuring 15.5 x 13 mm (rostrum only: 5 mm). This female apparently is identical by all characters with the male illustrated by KEMP & SEWELL. We believe that S. aurorae is a valid species. We are able to compare it with S. cuneus, also-one of the types, a male (cf. supra, Fig. 1 A-D, PL I figs. A-B). In S. aurorae (Fig. 2 A-D, PI. 1 figs. C-D) the length of the single rostrum, distincly bifid at the tip is more related in the male than in the female (relatively shorter in the female for both species); in the female type the rostrum is swollen in its basal part. The carapace is more enlarged and inflated in S. aurorae than in S. cuneus: The plates on the dorsal surface are smaller with much wider channels in S. aurorae (PL I, fig. C), whereas the plates are close-set in S. cuneus. The cardiac plate is heart shaped and well separated from the gastric plate in S. aurorae (Fig. 2D), triangular and touching the gastric plate in S. cuneus (Fig. ID). In S. aurorae, the branchial plate is narrower in the middle and in the lateral expansion than in S. cuneus. The preocular plate, not very salient, ends sub-acutely and therefore is much more spine-like in S. aurorae than in S. cuneus. The postocular plate offers a wide posterior margin, extended internally in S. aurorae, narrower in S. cuneus. In S. aurorae (PL I fig. D) , the pterygostomial plate has an irregular border and is widely separated from the postocular plate, whereas these two plates are closer in S. cuneus (PL I fig. B).
ALCOCK (1899: 84) wrote about S. aurorae that "as regards the male sex only: the pair of carinae on the dorsal surface of the carpopodites of the chelipeds and legs are blunt". In the female types, we observe that the carinae of legs are strong, more or leas as in S. cuneus.
In spite of the only two specimens exattnined by us for each species, we conclude that S. aurorae is a valid species, characterized especially: by the very short rostrum (even in the male) ; by the wide channels between the raised plates; by the cardiac plate heart-shaped and remote from the gastric plate; by the preocular plate with angular external borders.

Distribution
Known only from the coast of Travancore.  I fig. F). Dorsal surface symmetrically ornamented with raised punctate islets, separated by moderately wide chan ae; some hook-shaped nels; these channels covered with short set setae or gastric regions and on the postero-lateral borders, on the anteri on the base of the ows: one elongate-rostrum. The raised plates as foll oval on the gastri ular on the cardiac c region; one transversely isubcirc region, very well separated from the gastric plate and sometimes with a tubercle on each side; this cardiac plate never with lattero-posterior expansions (Fig. 3D); one (intestinal) in the shape of a Cupid's bow along th er one on the branchial region, wide and extend-e posterior bord ing laterally; one postocular plate (fused with the hepatic plate) a little wider at its base; a preocular plate not very raised but flattened and with rounded angles (except in so'me juveniles where it ends subacutely: Pl. I fig. H); one elongate plate on the subbranchial region; a pterygostomial plate moderately separated from the postocular plate (Pl. I Ambulatory legs cylindrical, shorter and more carinated in the male than in the female; the merus more widened and more crested in the female.

Remarks
The preliminary identification of the material and the first glance suggested that this material <of Madagascar would be Sphenocarcinus cuneus (Wood-Mason). The comparison of tthe numerous Madagascar specimens with the figure in the Illustr. "Investig." (1896, pi. 21, figs. 1, l a), and then with a male type of S. cuneus and with a female type of S. aurorae Alcock showed several differences: we found necessary to esta blish a new species, S. difficilis sp. nov. The encountered difficulties were compensated by the fine and various material collected by A. Crosnier in the north of Madagascar.
Through the examination of these numerous samples from Madagascar we are able to present the following remarks about the variations within the species.
In some male juveniles the rostrum; is extremely short and thick, pubescent, namely of the female type, and even it can show a base very enlarged, with a deeply bifid (PL I fig. H: ♂ 8.2 x 6.1 mm, chalutage 10). In the same male juveniles with a very short rostrum, the plates are spaced by much wider channels and the preocular plate ends su-bacutely with a little spine-like anterior process, as ALCOCK (1899; 84) pointed it for S. aurorae of larger .size. In other specimens, even in the males (especially in a male of the "chalutage" 28 measuring 18 x 12 mm) the plates on the dorsal surface have the punctures furnished with short, close setae and the rostrum is distinctly pubescent; generally the plates seem to be glabrous, smooth, but perhaps it is the effect of a brushing.
Summarily, the male juveniles have a rostrum relatively shorter, thick, sometimes widened at the base, and the ambulatory legs more carinated than in the adults. The females (PL I fig. G) have the rostrum distinctly shorter and wider than the males, the plates punctated, often pubescent; their ambulatory legs are shorter and carinated.
In our material, the uniformity in. the arrangement and the shape of the plates on the carapace is remarkable (except sometimes the channels more widened). The cardiac plate (Fig. 3D) are always transversely oval -shaped and does not touch the gastric plate; the subbranchial plate is prominent, with rounded borders (PI. T fig. F).

Distribution
Only known from Madagascar.

Description (holotype and paratypes)
Carapace triangular: the single rostrum very long and slender in the males, (Fig. 4A, PL II figs. A-B) much shorter in the female (Pl. THE S INDO-PACIFIC SPHENOCARCINU II figs. C-D), and with an extremely deeply bifid in the males, only emarginate in the female (to be verified); the rostrum much more incurved in the males (Fig. 4C, PL II fig. B) than in the female (PL II fig. D).
Dorsal surface (PL II figs. A, C) with several raised islets or plates arranged as follows: one elongate-oval and not very large on the gastric region; a single and elongate branchial plate, the external part extending laterally behind the line of the carapace; one pentagonal plate on the cardiae region (Fig. 4D), with an anterior border truncate and with two characteristic expansions on each side of the hinder part in the males as in the female; one plate in the shape of a Cupid's bow along the posterior border; one postocular-hepatic plate, narrower in the part close to the eye than in the basal part: a preocular plate not very prominent (Fig, 4A) ; a narrow plate on the postero-lateral border; below, a subbranchial plate which is small and not very prominent (PI. II figs. B,, D); a plate on the pterygostomial region small, with an irregular shape and an eroded aspect.
Surface of the plates punctate, without; setae (but the specimens perhaps have been brushed); the punctures much well-marked in the female. The surface between the plates almost glabrous (? brushed) except in the place of the hooked setae which are located on the external side nterior gastric region and along the of the branchial plate, on the a rostrum in the basal part.
Rostrum wit mare numerous h small, short and triangular setae, in the female tal part in the (PL II figs, C-D), diminishing on the dis males (PL II, figs. A-B) ; on the bifi er d extremity, setae a little long (Fig. 4A).
Male abdomen with seven somites. Chelipeds stout in the males (Pl. II fig. A), smooth and unarmed; only some punctures; margins of the triangular merus blunt; carpus with two blunt crests above. Fingers elongate, meeting only at the tips. In the female, chelipeds slender, weaker and pubescent; merus carinated especially on tihe upper border.
Ambulatory legs with a strong sexual dimorphism: in the males (Pl. II fig. A) elongate and cylindrical, except the carpus which is bluntly carinated and flattened. P4 and P5 shorter, with the merus reduced and faintly carinated. In the female (PI. II fig. C), ambulatory legs much shorter, with all the articles (except the dactylus) wide and thick; merus strongly carinated on p2-p5, the paired carinae on the carpus more pronounced than in, the 'males.

Remarks
There are now three known Indo-Pacific species of the genus Sphenocarcinus with a single rostrum: S. cuneus (Wood-Mason 1891), S. aurorae Alcock, 1899, and a new .species &. difficilis, which has close recemblances with the two preceeding species. From S, difficilis, to which are assigned numerous samples from Madagascar (cf. supra, Fig. 3 A-D , Pl. I figs. E-H), S. pinocchio differs in several .seemingly constant characters whioh in our opinion necessitate the recognition of a new species. S. pinocchio differs from S. difficilis mainly in the following characters : 1) the rostrum is relatively much longer and also more curved in S. pinocchio ( Fig 2) the cardiac plate has two latero-external expansions in C. pinocchio (Fig. 4D); it is subcircular and without lateral expansion in S. difficilis (Fig. 3D), where there is sometimes a tubercle on each part of this plate but always well separated.
3) the latero-posterior plate, situated ventrally below the branchial plate, is oval, distinctly prominent, with regularly, rounded borders in S. difficilis (Pl. I fig. F), narrow and with an eroded aspect in S. pinoc chio (PL II figs. B, D); 4) the postocular-hepatic plate (a single plate) is more prominent laterally, at least in the male, in S. difficilis (Fig. 3A, Pl. I figs E, H) than in S. pinocchio (Fig. 4A, PL II figs. A, C); 5) the basal antennal article, with a longitudinal groove in S. pinoc chio (Fig. 4B), without a groove in S. difficilis (Fig. 3B).
The fir&t male pleofpod is more or less similar in 'both species (S. difficilis; Fig. 6  We wish that these differences could be further verified on a more complete material especially that of S. pinocchio. Anyway of the significant characters of S. aurorae Alcock are better known, we can with certainty compare S. aurorae to S. pinocchio sp. nov.. S. aurorae (cf. supra, Fig. 2 A- In several rtespects, this individual resembles greatly S. pinocchio sp. nov., especially in the arrangement of the plates, but differs in: 1) the rostrum ( Only one individual being at our dispesal, moreover infected with a Sacculina sp., we question the value of these peculiarities. The ambulatory legs (PL II fig. E) are of the female type (Pl. II fig. C), namely shorter and carinated: we can suppose that they became fominized. The chelipeds, completely developed, have the normal male type; the pleopods are present but both are broken at the place where the Sacculina's sac forms a wide knob under the abdomen and turns out the pleopods from the sterno-abdominal cavity. It seems that is it neither feminization of the chelipeds nor of the pleopods.
The question is whether the rostrum, as the ambulatory legs, could be feminized: that would explain the shortened rostrum, with a tip less bifid? The pronounced curvature (PL II fig. F) could be the result of an "abnormal" transformation because of the infection by a parasite, inasmuch as the rostrum of the male holotype of S. pinocchio (Fig. 4 A-C, PL II figs. A, B) is longer and more curved than in the female paratype (PL II figs. C, D).
We are very interested by the shape of the cardiac plate of the male individual parasitized with Sacculina sp. having a similar form to that of the non parasitized S. pinocchio (Fig. 4D, PL II figs. A, C), namely in the expansion on each side at the base of the plate.
For lack of material and furthermore the fact that the single specimen bears a parasite which produces probably a feminization, we leave it in S. pinocchio sp. nov. It was collected more or less in the same locality with other specimens of S. pinocchio, and in similar depths.
These three femaje individuals are near of S. pinocchio but principally differ by the shape of the cardiac plate more transversajly elongate and with smaller postero-lateral expansions. The rostrum of the small female is similar to that of S. pinocchio ovigerous female of analogous size 19,5 x 11.5 mm (cf. PI. II fig. C-D); the rostrum of the larger female is more widened at its baae and is distally broken. The sacculinized female has a very short and wide rostrum.
This male presents a very short rostrum, a gastric plate very narrow and elongate, a cardiac plate with horizontal expansions. These characters are different from those of all another species.