The species of Moloha Barnard, 1946, from the western Indian Ocean, with the description of a new species from India (Crustacea: Brachyura: Homolidae)

. The taxonomy of the deep-water homolid crabs Moloha grandperrini Guinot & Richer de Forges, 1995 and M. alisae Guinot & Richer de Forges, 1995 is re-examined, and the types redescribed and ﬁ gured. Moloha alisae is reported from South Africa for the ﬁ rst time. A new species with an in ﬂ ated carapace, M. tumida sp. nov., is also described from southern India and compared with its closest congeners.


Material and methods
The measurements for the carapace are as follow: tcl = total carapace length, including spines tcw = total carapace width, including spines cl = carapace length, measured at bases of spines cw = carapace width, measured at bases of spines The terminology used follows that by Guinot & Richer de Forges (1995).The abbreviations P1-P5 are used for the fi rst to fi fth ambulatory legs, including the chelipeds; G1 and G2 for the male fi rst and second gonopods, respectively.The counts for the spines on the margins and surfaces of P1-P5 include tubercles and distinct sharp granules, which are sometimes not easy to distinguish because they vary in sizes and degree.As such, we use the term spines for all these structures.For the dorsal row, it includes the large distal spine.The spines along the ventral (fl exor) margin of the merus actually form approximately two rows proximally but merge towards the distal end.As such, the number of spines on the ventral margin is a total count.Specimens examined are deposited in the Department of Aquatic Biology & Fisheries, University of Kerala (DABFUK), India; The Natural History Museum (NHM), London; Muséum national d'Histoire naturelle (MNHN), Paris; and the Zoological Reference Collection (ZRC) of the Lee Kong Chian Natural History Museum (formerly the Raffl es Museum of Biodiversity Research), National University of Singapore.

Diagnosis
Carapace with pseudorostral and supraocular spines long, subequal; supraocular spine with distinct submedian accessory spine; gastric region with small but distinct sharp granules in addition to 3 major spines; branchial regions convex; subhepatic region swollen, with 2 large dorsal and 2 small ventral spines; protogastric region with 2 major spines; basal antennal spine triangular, relatively broad; P2-P4 long, slender, subcylindrical, merus with 6 or 7 spines on dorsal margin, outer surface with 3-10 small spines, ventral margin with 15-21 spines; P5 with 1 or 2 spines on dorsal margin, 4 small spines on outer surface, 4 or 5 spines on ventral margin, subchelate structure stout, propodus with 3 large basal spines, rest of margin with distinct, closely arranged spines of similar size.G1 stout, short, groove on ventral surface median, dorso-median surface concave, distal part rounded, opening relatively smaller, fl ap-like, not auriculiform, directed towards median part of sternum.

Distribution
The species was described from the Maldives and has not been reported elsewhere.

Colour
In life, the carapace, chelipeds and ambulatory legs are orange with patches of white (Fig. 17A).

Distribution
The species was described from the Seychelles; the present record from South Africa is new.

Remarks
The holotype male of M. alisae is small (MNHN-IU-2008-11077; cl 36.1 mm, cw 29.7 mm) and, although the G1 and G2 are present, it is clearly still a juvenile.The male pleon is not domed (Fig. 7E) and the gonopods are still not strongly chitinised, being relatively soft (Figs 7C, 16A-C).We have referred the two large specimens from South Africa (ZRC 2008(ZRC .1250) ) to this species because it matches the holotype in most key aspects: the carapace shape is distinctly rectangular, the surfaces between the major spines on the gastric region are smooth and unarmed, the cardio-intestinal groove is deep, and P5 is long, reaching to the base of the pseudorostral spines when folded anteriorly.Another shared character is the proportionately longer P5 propodus of the subchelate structure, which has the teeth on the distal half of the fl exor margin more widely spaced (Figs 11F, K, 12F, H-J).In M. grandperrini and M. tumida sp.nov., the P5 propodus is relatively shorter and the fl exor margin has more closely arranged teeth of similar sizes (Figs 10F, 13I).However, there are a number of differences which we believe are size-related and not signifi cant at the species level.The branchial surfaces of the two large South African specimens are covered with relatively more spinules between the major spines (Figs 2B, 3D, 14D) compared to those on the holotype male (Figs 2A, 3C, 14C).In addition, the P5s of the two largest South African males are still relatively shorter than those of the holotype male from the Seychelles, reaching only to the base of the pseudorostral spines (Fig. 14D) and not to the median part of the spines (Fig. 14C).The armature of P2-P5 is substantially stronger in the two South African males (Fig. 12B-D) compared to that on the holotype male (Fig. 11B-D, G-I; Table 1).There is also a slight difference in the form of the distal part of the G1.In the large South African males, the distal part is more bulbous, with the opening relatively large (Fig. 16D-E), while in the smaller Seychelles male, it is less swollen, with the opening smaller and more folded (Fig. 16A-B).The chelipeds of the two South African males are typical of many large mature homolids, being elongated, stout, the surfaces granular and covered with dense setae (Fig. 8C,  E).Those of the holotype male from the Seychelles (Fig. 7E-F) clearly belong to a juvenile.As such, it is best to refer the two large South African specimens to M. alisae for the time being, at least until more material becomes available from the area, especially belonging to intermediate size-classes.

Colour
In life, the carapace and chelipeds are red; the ambulatory legs are red with patches of white and the ventral surfaces are dirty white (Fig. 17B).

Distribution
The species is known only from southwestern India.

Remarks
See the Discussion section.
As discussed at length by Guinot & Richer de Forges (1995: 380), Moloha alcocki (Stebbing, 1920) s. str. is a very distinct species, markedly different from all congeners in its laterally fl attened ambulatory meri (Guinot & Richer de Forges 1995: fi gs 29a).In addition, its carapace is distinctly longitudinally rectangular, with the lateral margins straight (Guinot & Richer de Forges 1995: fi gs 29a-b).The ischium of the third maxilliped also bears fi ve tubercles (Guinot & Richer de Forges 1995: fi g. 30B) but on M. grandperrini and M. alisae it only has four (Figs 10A,11A,12A).Moloha alcocki is known only from South Africa so far.Specimens which have been identifi ed as "M.alcocki" by other authors have since been referred to M. majora (Kubo, 1936) and M. grandperrini Guinot & Richer de Forges, 1995(Guinot & Richer de Forges 1995: 380).Guinot & Richer de Forges (1995) described two new species from the western Indian Ocean which were superfi cially similar to M. alcocki but differed in having subcylindrical ambulatory meri -M.alisae and M. grandperrini.
European Journal of Taxonomy 166: 1-25 (2015) In describing M. alisae from a single male (cl 40.6 mm, cw 29.7 mm) from the Seychelles, Guinot & Richer de Forges (1995: 379, 391) noted that it was close to M. grandperrini, which was described also from a solitary male (cl 46.5 mm, cw 39.0 mm) from the Maldives.They argued that M. alisae differed from M. grandperrini in having the dorsal surfaces of the carapace relatively less convex and swollen, the surfaces between the large spines on the gastric region of the carapace are smooth (covered with small sharp granules in M. grandperrini), the spines lining the ventral margin of the ambulatory merus of P2-P4 are fewer and spaced further apart, and the distal edge of the merus of P5 reaches the base of the pseudorostral spine when it is folded anteriorly (reaches only the protogastric region in M. grandperrini).As discussed earlier under M. alisae, the number and strength of spines on P2-P4 may not be a useful character as it is size-associated (see also Table 1).
The two large specimens from South Africa here referred to M. alisae (ZRC 2008(ZRC .1250) ) possess almost all the diagnostic characters stated by Guinot & Richer de Forges (1995) for the species (see discussion for this species).Despite their much larger sizes (cl 67.7 mm, cw 56.2 mm; cl 67.5 mm, cw 56.5 mm), both have carapaces that appear to be somewhat more rectangular in form when viewed dorsally (Figs 2B, 3D, 14D) compared to that of M. grandperrini, with the lateral branchial margins of the latter species slightly more convex (Figs 1A, 3A, 14A).The gastric region of M. grandperrini, other than armed with three major spines (two anteriorly and one posterior), also has several distinct sharp granules on the surface (Figs 1A, 3A, 14A).In M. alisae, the surfaces of the gastric region are completely smooth, other than for the three spines (Figs 2, 3C-D, 14C-D).Regardless of size, the ambulatory legs of M. alisae are also proportionately longer, notably in the lengths of the propodi (Figs 2, 11B-D, G-I, 12B-D) (relatively shorter in M. grandperrini; Figs 1A, 10B-D, G-H); and for P5, the distal edge of the merus (not including the distal spine) reaches the base of the pseudorostrum when it is folded anteriorly, even in the largest specimen (Fig. 14C-D) (reaches only to the anterior edge of the protogastric region in M. grandperrini; Fig. 14A).Another marked difference not mentioned by Guinot & Richer de Forges  (1995) is the depth of the transverse cardio-intestinal groove.In M. alisae, this groove is very deep and distinct (Figs 2, 3C-D, 14C-D), and is evident even on the small holotype male (Figs 2A, 3C, 14C, Guinot & Richer de Forges 1995: fi g. 29f).In M. grandperrini, the transverse cardio-intestinal groove is relatively more shallow (Figs 1A,3A,14A).Another character differentiating the species is the form of the subchelate process of the P5.In M. alisae, the propodus is more elongate and other than the four major spines at the proximal edge, the spines on the rest of the fl exor margin are spaced further apart and relatively smaller (Fig. 11E-F, J-K) (see also Guinot & Richer de Forges 1995: fi g. 51i).In M. grandperrini, the propodus is proportionately shorter and the smaller spines on the inner margin are relatively larger and close to each other (Fig. 10E-F, I).The G1s of the two species are superfi cially similar, but in M. grandperrini the distalmost part is more rounded (Fig. 15A-B) than it is in M. alisae (Fig. 16A-B, D-E).
Moloha tumida sp.nov. is easily distinguished from M. alisae and M. grandperrini in its prominently more convex and swollen branchial regions (Figs 1B, 3B, 5B, 14B).In addition, the basal antennal spine is acute in M. tumida sp.nov.(Fig. 5B), but is relatively broader and stouter in M. alisae and M. grandperrini (Fig. 5A, C-D).The carapace of M. tumida sp.nov. is similar to that of M. grandperrini in possessing secondary small tubercles on the gastric region and in having the cardio-intestinal groove very shallow and barely visible (Figs 1B,3B,14B).The ambulatory legs of M. tumida sp.nov.are proportionately the shortest among the three species, notably in the length of the propodus (Figs 1B, 13B-D, E-G); this also applies to the P5 in which, when folded over the carapace, the distal edge (excluding the spine) only reaches to the proximal part of the subhepatic region (Fig. 14B).There are no obvious differences in the degree of spination on the meri of P2-P5 in adult specimens of these species (Table 1).The holotype male of M. alisae is small and is a young male, and its P2-P5 are relatively less spinate (Table 1), but this is almost certainly because of its small size.The subchelate structure of M. tumida sp.nov. is similar in that of M. grandperrini, differing from M. alisae in the same features (see earlier).Although the G1 of M. tumida sp.nov. is superfi cially similar to that of M. alisae and M. grandperrini, it differs in having the ventral groove closer to the inner margin (Fig. 15D) (ventral groove distinctly median in M. alisae and M. grandperrini; Figs 15A, 16A, D), the surface of the median part of the G1 (when viewed dorsally) almost fl at (Fig. 15F) (surface concave on the median surface in M. alisae and M. grandperrini; Fig. 15B, 16B, E), and the distal opening distinctly more fl ared and auriculiform (Fig. 15D-E

Table 1 .
Spine counts on margins and surfaces of the meri of P1-P5 of Moloha species.The counts are given in the following order: dorsal margin / outer surface / ventral margin.L and R are for left and right sides, respectively.K.L. & KUMAR A.B., Species of Moloha from the western Indian Ocean