The Magelonidae ( Annelida : Polychaeta ) from the Seychelles . 2 . Description of four additional species , three new to science

In an earlier paper, Mortimer and Mackie (2003) initiated a study of the Magelonidae collected during the Seychelles marine biological expedition carried out by the National Museum of Wales in 2000, under the auspices of the Royal Geographical Society (with IBG) / Royal Society’s Shoals of Capricorn programme, Western Indian Ocean 1998-2001. To date, the polychaetes of the Seychelles, in the western Indian Ocean, have received relatively little taxonomic attention. However, the biodiversity of the infaunal polychaetes and molluscs around Mahé, the main island, has been shown to be very high (Mackie et al., 2005). Presently, there are approximately 64 recognised magelonid species. Almost all are included in the genus Magelona Müller 1858, however, two species have been described for Meredithia HernándezAlcántara and Solís-Weiss, 2000 and one for Octomagelona Aguirrezabalaga, Ceberio and Fiege, SCIENTIA MARINA 70S3 December 2006, 125-137, Barcelona (Spain) ISSN: 0214-8358 SCIENTIFIC ADVANCES IN POLYCHAETE RESEARCH R. Sardá, G. San Martín, E. López, D. Martin and D. George (eds.)

One Seychelles species, Magelona gemmata, was recently recorded off the island of Natuna Besar in the South China Sea (Al-Hakin and Glasby, 2004).
This second study of the Seychelles magelonids (Shoals of Capricorn publication P055) includes detailed descriptions of three new species, as well as M. pygmaea Nateewathana and Hylleberg, 1991, originally described from Thailand.

MATERIALS AND METHODS
All material was collected during the National Museum of Wales expedition to the island of Mahé in March 2000.Samples were collected for both taxonomic research and biodiversity assessments.
All samples in this paper were obtained from a series of sublittoral stations around Mahé (Fig. 1) using the modified 0.1 m 2 Van Veen grabs (Station replicates designated a, b, c…) first successfully deployed in the Irish Sea in 1997 (Wilson et al., 2001) or Tjärnö dredge.Samples were sieved (0.5 mm mesh) and fixed in formalin, mostly stained with Rose Bengal.Specimens were subsequently preserved in 80% alcohol with 2% propylene glycol (Mackie and Oliver 1996).
All drawings were made using a camera lucida attachment on a Leica MZ9.5 zoom microscope, or Nikon Labophot-2 compound microscope.
In order to keep consistency in the descriptions within this family, specimen descriptions follow the same format as those described in Mortimer and Mackie (2003).For example, specimens are herein recorded as complete (c), anterior fragments (af), posterior fragments (pf) or fragments (f).Lateral pouches on abdominal chaetigers were recorded here according to the morphologies described by Fiege et al. (2000): Σ configurationanteriorly open pouches (often convoluted) bounded dorsally and ventrally by large cuticular flaps; C configuration -posteriorly open pouches, often inconspicuous, ventral part sometimes folded over flattening pouch against body.Parapodial structures are primarily described relative to basic polychaete terminology (e.g.postchaetal, prechaetal etc.), however, certain of the terms introduced by Jones (1971Jones ( , 1978) ) are also given (e.g.dorsal medial lobe -DML, ventral medial lobe -VML) to facilitate cross-comparisons with other publications.
Description.A large, stout species (Fig. 2A); thorax width similar to abdomen.Holotype an anterior fragment and three median fragments; pygidium lacking.Thorax (including prostomium) 4.7 mm long, 1.5 mm maximum width, total length 12.5 mm for 26 chaetigers.Abdominal fragments each of nine or ten chaetigers.
Abdominal chaetigers with sharply-pointed triangular lateral lamellae, of about equal size in both rami (Fig. 2G-H).In anterior abdomen, lamellae longer than those of thorax.Lateral lamellae do not extend postchaetally, prechaetal ridges low and indistinct.No dorsal (DML) or ventral (VML) processes on abdominal chaetigers.
Colour.Observations made on a preserved holotype originally stained with Rose Bengal; now cream-white in alcohol.Methyl Green staining of glandular areas rather diffuse, no intensely stained speckles.Darkest staining in mid-dorsal and midventral areas of thorax, with cream-coloured patches noticeable dorsolaterally, laterally and on ventral pads.Cream pigmentation also particularly noticeable in abdomen as intense interparapodial patches and on mid-ventral line.
Etymology.The specific name is from 'symmetros' (Gk), meaning symmetrical, and refers to the uniformity of the noto-and neuropodial lamellae in both thorax and abdomen.
Habitat.Only found in coral sand at one shallow station east of Mahé.
Remarks.Magelona symmetrica n. sp.approaches a suite of five species [M.japonica Okuda, 1937, M. koreana Okuda, 1937 (originally M. japonica var. koreana;see Jones 1971), M. cornuta Wesenberg-Lund, 1949, M. alleni Wilson, 1958 andM. equilamellae Harmelin, 1964], based on the shape of both thoracic and abdominal lamellae.Magelona symmetrica is most similar to M. koreana in possessing a small ventral process on the ninth neuropodium, but differs in not having prostomial horns and sub-equal abdominal lamellae.It also differs from M. japonica and M. cornuta in lacking prostomial horns.The remaining two species, M. alleni and M. equilamellae, have thoracic pigment bands that are not seen in the new species.Further, M. equilamellae has rudimentary horns and M. alleni has unequal abdominal lamellae.
Abdominal chaetigers with slender foliaceous lateral lamellae, initially longer than those of preceding thoracic chaetigers.Lamellae basally constricted, of about equal size in both rami (Fig. 3H), becoming narrower and more lanceolate (Fig. 3I) towards posterior.Lateral lamellae do not extend pre-or postchaetally.No processes (DML and VML) on abdominal chaetigers.
Colour.No living animals observed, all but 2 grab samples (Stn 4c and 48a) originally stained with Rose Bengal; cream-white in alcohol.Methyl Green staining evident dorsally as light speckles on posterolateral parts of prostomium, the peristomium and chaetigers 1-3.Chaetigers 4-7 densely speckled, becoming lighter on chaetiger 8 and absent on chaetiger 9. Non-staining cream-coloured longitudinal strips present dorsolaterally on chaetigers 1-4 and 8, small cream patches on chaetiger 7. On ven-ter, stained speckles sparse on chaetiger 1, light on chaetigers 2 and 8, increasing in density over chaetigers 3-5 and reaching a maximum on chaetigers 6 and 7.Staining of glandular areas particularly noticeable on abdominal chaetigers as intense interparapodial patches.
Etymology.The specific name refers to the type locality off the island of Mahé.
Remarks.Magelona mahensis n. sp.resembles three species; M. capax Hartman, 1965, M. variolamellata Bolívar andLana, 1986, and an undescribed species Magelona sp.I of Uebelacker and Jones, 1984 in exhibiting a reduction in the size of the thoracic neuropodial lamellae.Magelona capax differs in having a cape-like prostomium with horns, and dorsal processes on thoracic chaetigers.The thoracic lamellae of M. variolamellata reduce from chaetiger 1 to 9, whereas those of M. mahensis only reduce from chaetiger 4 to 8, and M. variolamellata has tridentate hooks.Magelona sp.I differs markedly in having polydentate hooded hooks.
Pouches such as those seen on M. mahensismore or less folded, with thicker cuticle on edges of the fold and thinner cuticle inside -were previously also mentioned in Mortimer and Mackie (2003: 164).Fiege et al. (2000: 217) note that, for C configuration pouches, "Their size and degree of protrusion is variable …" however, the folded 'C' form could be indicative of more variation than acknowledged by either the Σ or C configuration, and there could possibly be intermediates between the two.More detailed examination of intra-and inter-specific variation in pouch structure is required to resolve this.The exact function of these pouches has been much discussed (Jones, 1963(Jones, , 1968;;McIntosh, 1878McIntosh, , 1911;;Fiege et al., 2000), however, nothing is proven.In situ observations on living animals may ultimately prove fruitful.
Prostomium as wide as long (L:W ratio 0.97-1.00),rounded, somewhat onion-shaped, anterior margin smooth and straight or slightly rounded with rudimentary prostomial horns, eyes absent; four prominent longitudinal dorsal muscular ridges, inner pair diverging at both ends and outers abutting inners (Fig. 4A).Proboscis an oval-shaped sac (paratype), with conspicuous longitudinal ridges inferiorly and less conspicuously on superior surface.Palps arising ventrolaterally from base of prostomium, short, reaching chaetigers 9-10; papillated almost to base (non-papillated region reaching to chaetiger 1).Palps, heavily papillated with 8 rows of long papillae (length decreasing only at proximal end) proximally, decreasing to 2 towards distal region, either side of inconspicuous ventral groove (on holotype right hand palp retained, distal tip missing?).
Peristomium of roughly equal size to chaetiger 1. Notopodia of chaetigers 1-8 similar, with subrectangular postchaetal lamellae gradually increasing in size from chaetiger 4 onward (Fig. 4B-G fused inferiorly to short slender digitiform process.All thoracic chaetae simple capillaries.Abdominal chaetigers (Fig. 4I-J) with round spatulate lateral lamellae, basally constricted, of equal size in both rami, becoming more pointed towards median region.Lateral lamellae extend postchaetally in anterior abdomen but by median chaetigers this structure is reduced; postchaetal part less conspicuous in smaller specimen.Short triangular processes (DML and VML) evident on abdominal chaetigers.
Colour.No living animals observed, both specimens originally stained with Rose Bengal; creamwhite in alcohol.Methyl Green stained speckles sparse on lateral prostomial areas, dorsally and ventrally.Thorax generally covered with stained speckles, heaviest staining on chaetigers 3-5.Dorsally, speckles less at anterior of each segment and in mid-dorsal line.Ventrally, staining absent on the two longitudinal lateral lines that bound the paired ventral pads; chaetiger 9 with single small trapezoidal pad.White glandular inclusions in centre of each pad on chaetigers 3-8.No stained speckles in abdomen; white glandular patches in interparapodial areas.
Habitat.Found at two sandy stations (33-45 m depth) off the east coast of Mahé.
gemmata Mortimer and Mackie, 2003, and 4 unnamed species Magelona spp.G, J, K and L of Uebelacker and Jones, 1984.In overall morphology, M. cepiceps appears closest to M. berkeleyi and Magelona sp.J.
Magelona pacifica, M. crenuliformis, M. tehuanensis, M. gemmata, and Magelona spp.G, K and L can be readily distinguished from the new species in having well-developed frontal horns, while also lacking superior processes on chaetiger 9. Frontal horns are less conspicuous on M. capensis and M. berkeleyi, but they also lack dorsal processes on chaetiger 9.The remaining two species, M. lenticulata and Magelona sp.J do possess dorsal processes on chaetiger 9, however both have frontal horns, prostomia with crenulated anterior margins and there are differences in the form of certain neuropodial postchaetal lamellae.Nateewathana and Hylleberg, 1991 (Fig. 5) Diagnosis.Small, slender species.Prostomium wider than long, rounded triangular to subtrapezoidal, without prostomial horns.Notopodia of chaetigers 1-9 with smooth-edged triangular postchaetal lamellae.Neuropodial lamellae rounded triangular.Thoracic chaetigers only with capillary chaetae.Abdominal lateral lamellae sublanceolate.Hooded hooks tridentate, in 2 groups, vis-á-vis.No lateral pouches on abdomen.Pygidium with lateral pair of cirriform anal cirri.margin smooth, prostomial horns and eyes absent; two prominent longitudinal dorsal muscular ridges, separated, diverging at both ends.Palps arising ventrolaterally from base of prostomium, long, reaching chaetigers 13-35 (usually to about chaetiger 15).Palps with non-papillated region reaching to chaetiger 2, and 1 row of papillae on either side of inconspicuous ventral groove.
Abdominal chaetigers with sublanceolate lateral lamellae, of about equal size in both rami, basally constricted (Fig. 5G-I).Lateral lamellae do not extend postchaetally or prechaetally, inferior and superior processes (VML and DML) absent.

Colour. Observations made on preserved
Seychelles specimens originally stained with Rose Bengal; cream-white in alcohol.On largest complete Seychelles specimen dorsal staining with Methyl Green evident posterolaterally on prostomium and on chaetigers 2-8, strongest on chaetiger 4. Ventral staining on chaetigers 2-9, strongest on 4-6.Interparapodial regions of anterior abdomen and posteriormost 9 chaetigers with densely stained speckles.Other Seychelles material similar, but thoracic staining variable in intensity and occurrence; prostomium often unstained.The Thai paratype shows no prostomial staining and only light staining on the thorax, particularly toward the posterior half.Lateral abdominal staining is present anteriorly.The Thai non-type material (PMBC 4227 and 4241) generally shows a similar staining pattern and a larger specimen also has some prostomial staining.The remaining material (PMBC 4234) also conforms in having predomi-nantly posterior thorax staining, however, in the abdomen there is additional staining (dense speckles) occurring as narrow dorsal and ventral transverse bands just behind the parapodia, and along a mid-ventral line.
Habitat.Found at six stations (Stn 3,7,9,10,11 and 12), off the southeast and west coasts of Mahé, at depths of between 27-57 m, in muddy sands/clays to coarse sands.The Thai type and non-type material all comes from fine sand (10-30 m) off the west coast of Phuket (Nateewathana and Hylleberg, 1991;Hylleberg and Nateewathana, 1991).
Remarks.Magelona pygmaea is a member of a 'M.papillicornis' group based on the shape and arrangement of the thoracic lamellae.There are five clear members of this group: M. papillicornis F. Müller, 1858, M. californica Hartman, 1944, M. minuta Eliason, 1962, M. pettiboneae Jones, 1963 and M. pygmaea.Magelona pygmaea differs from all of these in possessing tridentate and not bidentate hooded hooks.
The description presented above from Seychelles material agrees well with the original one from Thailand.The small size of this species and a degree of variation in the prostomial shape (some prostomia are more rounded than originally figured) and thoracic postchaetal lamellae (on some specimens the lamellae are less pointed than originally figured) make any separation of material between the two locations difficult.Nevertheless, some of the Thai non-type material appeared to have both bi-and tridentate hooks and one lot (PMBC 4234) showed additional abdominal staining using Methyl Green.The significance of these observations is unclear at present, though a larger investigation of morphological variability within M. pygmaea from the type locality would be useful.Magelona pygmaea was the most widely distributed (Hylleberg and Nateewathana, 1991) of the 8 species described by Nateewathana and Hylleberg (1991).Hylleberg and Nateewathana (1991) indicated that M. pygmaea was more abundant during a period when the silt-clay content of the sediments was reduced from 25% to 5%.In the Seychelles Magelona pygmaea was present in sediments with 0.17-43.26%siltclay.The number of specimens per 0.1 m 2 grab was relatively low (0-18 individuals) and there was a high degree of intra-station variability.No preference for low silt-clay sediments was found.