Paraonidae (Polychaeta) from the Capbreton Canyon (Bay of Biscay, NE Atlantic) with the description of eight new species

nineteen species of paraonid polychaetes, belonging to five genera, were collected from bathyal depths at the capbreton canyon, bay of biscay, ne Atlantic. Levinsenia kantauriensis n. sp., Aricidea sardai n. sp., A. bifurcata n. sp., A. mirunekoa n. sp., A. maialenae n. sp., A. nekanae n. sp., Paradoneis bathyilvana n. sp. and P. mikeli n. sp., are new to science; Aricidea (Allia) antennata, Levinsenia flava, Paradoneis drachi and Paraonides myriamae are new records for the Atlantic, and Paradoneis eliasoni to the iberian coasts.

inTroDucTion Deep-sea polychaetes are still largely unknown.Some environments peculiar to the deep sea, such as hydrothermal vents, mud volcanoes, methane clathrate deposits, and whale bones, are receiving a great deal of scientific attention, which is revealing a richness of species sometimes beyond expectations (e.g.: Desbruyères and Toulmond, 1998;Dahlgren et al., 2004;rouse et al., 2004;glover et al., 2005;Desbruyères et al., 2006;ravara et al., 2007;Hilário and cunha, 2008;Pleijel et al., 2008;rouse et al., 2008).However, although polychaete fauna represent the largest part of the deep-sea floor community there is little information on them in other deep sea regions, such as abyssal plains, continental rises, and canyons.on many occasions, the material is obtained as a by-product of other studies, and is not treated for taxonomic work.This leads to the specimens being in poor condition or fragmentary, which is usually useful for general faunistic surveys, but not for fine taxonomic research.
From 1987 to 1990, four oceanographic cruises were conducted in the capbreton submarine canyon onboard the rV-"côte d´Aquitaine".in these cruises, the bathyal macrofauna was carefully sampled and sorted, which yielded a large amount of polychaete specimens in excellent condition for fine taxonomic studies, which is evidenced by the increasing number of published papers based on this material (San Martín et al., 1996;Aguirrezabalaga et al., 1999Aguirrezabalaga et al., , 2001Aguirrezabalaga et al., , 2002;;núñez, et al., 2000;Aguirrezabalaga and ceberio, 2003Aguirrezabalaga and ceberio, , 2005aAguirrezabalaga and ceberio, ,b, 2006;;Aguirrezabalaga and carrera-Parra, 2006), and, in the present paper, by the good SeM pictures.

MATeriAlS AnD MeTHoDS
The specimens were collected in the capbreton canyon, bay of biscay, ne Atlantic.eighteen sta- tions (Fig. 1, Table 1) were dredged either with a Sanders-Hessler epibenthic dredge (Di) or a Flusha box-corer (KF).Samples were sieved through a 0.5 mm mesh and the sorted specimens were preserved in a 10% formaldehyde-seawater solution.
The types and representative specimens of the species described here are deposited in the Museo nacional de ciencias naturales, Madrid (Mncn) and in the Sociedad cultural de investigación Submarina inSub, Donostia (inSub). in the examined material section, the number of specimens at each station is indicated between brackets.in the descriptions, "l/w" means ratio length / maximum width at base of branchiae, and "l" and "max.l" refer to the length and the maximum length of branchiae respectively.The indicated coloration always refers to preserved specimens.Question marks (?) under the descriptions of Paradoneis mikeli n. sp. and Aricidea (Allia) sardai n. sp.indicate that part of the material identified respectively as Paradoneis lyra and Arici dea mediterranea, in the given references, probably belongs to the newly described species.However, this material was not revised in the present study, and is only presented as possibly belonging to the new species.likewise, question marks in the species distribution designate the possible but unconfirmed presence of the new species in the indicated areas, based on the given references.
Description.All specimens incomplete.body long, slender; wider anteriorly to thinner, filiform from midbody.Anterior segments short, much wider than long (two to three times); longer and biannulate (length = width) at postbranchial region, then cylindrical, longer than wide.
Many specimens with orange-coloured anterior region.
Discussion. our specimens agree well with the Strelzov (1973) description, the only difference being the number and size of branchiae.However, the original description was based on a single specimen, so the natural variability of the taxon was not reflected.According to Strelzov (1973), Levinsenia flava has three pairs of branchiae with a l/w of 2.8, and neuropodial hooks present from chaetiger 15. in our specimens, the number of branchiae varies between 3-5 and l/w between 2-4 (mostly between 2.5-3.5),while the neuropodial hooks appear mainly from chaetigers 14-17.
Anterior segments short, twice as wide as long; longer in postbranchial region (length = width), then cylindrical, longer than wide.
Etymology.The species name refers to the collecting site (cantabrian Sea = Kantauri itsasoa, in basque).
Levinsenia kantauriensis n. sp. is distinguished from L. oli gobranchiata, which has shorter branchiae and dorsal lobes on the posterior branchial segments, and modified neurochaetae with very poorly developed pubescence on the convex side of the shaft.
Distribution.north Atlantic, South Africa, california, Mediterranean, Sea of Japan, red Sea. between 8 and 2,780 m depth.
Prostomium conical, with a terminal sensory organ (Fig. 7A).nuchal organs as a pair of deep nuchal slits in posterior part of prostomium.Posterior buccal lip with longitudinal folds starting from median part of chaetiger 1 (Fig. 7b).
All chaetae capillaries; dorsal-most neuropodial one in median and posterior segments longer.
Discussion.Finding this species in the bathyal region of the capbreton canyon seems to confirm the deep-water record made by Amoureux (1982), who reported the species from the continental slope off brittany, at 1200 meters.Material examined.one specimen from capbreton canyon, bay of biscay, Atlantic ocean (coordinates in Table 1).cb88/Di37.
Prostomium conical, as long as wide.Median antenna absent (Fig. 8A).one pair of nuchal organs as slits in posterior part of prostomium.Posterior buccal lip with longitudinal folds starting from anterior part of chaetiger 1 (Fig. 8b).
Discussion.Paradoneis spinifera (Hobson, 1972) and Paradoneis drachi are characterized by having smooth acicular spines as modified notopodial chaetae, but differ in number of prebranchial chaetigers (4-5 and 6 respectively).Although blake (1996) stated that both species could be the same, the specimen from capbreton agrees well with the description of laubier and ramos (1974) and is here considered as P. drachi.Mackie, 1991 (Fig. 9)
Prostomium conical, rather longer than wide (Fig. 9A).Pair of nuchal organs as deep nuchal slits.Posterior buccal lip with two large tongue-shaped folds starting from buccal segment, and several narrower, longitudinal folds starting from first chaetiger (Fig. 9b).First three notopodial postchaetal lobes short, rounded tubercles; on branchial region, more than two times longer, digitiform, distally rounded (Fig. 9A); on postbranchial region, similar in length but slightly more slender and pointed.
Discussion.The studied specimens agree well with the description by Mackie (1991).
Pygidium with three anal cirri of similar size: two dorso-lateral and one mid-ventral.
Etymology.The name of the species refers to its similarity to Paradoneis ilvana, combined with its different bathymetrical range.
Etymology.This species is fondly dedicated to Mikel Agirrezabalaga Arraras.
Discussion.Paradoneis mikeli n. sp. is compared with other species of the genus in Table 3.The specimens are very similar to Paradoneis lyra, but differ in having branchiae starting from chaetiger 5, with the single exception of the right side of one specimen, where they start at chaetiger 4 (Fig. 13c).Paradoneis lyra always has the branchiae starting at chaetiger 4 according to Mackie (1991), who redescribed the species.However, this is not the case in some deep-water records of Paradoneis lyra, in which branchiae were reported to start at chaetiger 5. Pettibone (1963), in a collection of specimens from 1 to 1060 fathoms depth (1.8 to 1939 m), reported branchiae starting at chaetigers 4-5. in turn, all specimens collected by Amoureux (1982) from 500 to 2,000 m depth had branchiae starting at chaetiger 5, and Katzman and laubier (1975) collected Mediterranean specimens with branchiae starting both at chaetigers 4 (40% of individuals) and 5 (60% of individuals) from 20 to 760 m depth.Although in some of the reported cases it is not possible to determine the exact bathymetric distribution of the two forms, we suggest that the form with 4 prebranchial chaetigers could be of deeper origin, thus corresponding to the new species described here.
Furthermore, branchiae are short (body width/ branchial length >1.4 in most of the specimens), always shorter in P. mikeli n. sp.than in P. lyra, and never approximately equal to distance between branchial bases (Mackie, 1991). in addition, the typical subdermal brown eyes of P. lyra are absent in P. mikeli n. sp., and the number of lyriform chaetae is maintained until the last preanal segment (3-4 in the postbranchial region).other differences are the presence of a fold in the mid-dorsum of the postbranchial segments and the bilobed pygidium with ventrally opening anus in P. lyra (Katzmann and laubier, 1975), compared to dorsally opening in P. mikeli n. sp.
in the present samples (as well as in others from the continental slope off galicia), we found several individuals that agree well with the second form of A. mediterranea (i.e., with simple, short and distally inflated, club-shaped antenna).
in turn, there is a unique species of Aricidea, A. (Al lia) monicae, which is characterized by its peculiar interparapodial small digitiform lobes, regularly cylindrical and distally rounded, present in some anterior segments and its very short, cylindrical, distally rounded antenna (laubier, 1967).A. monicae was again collected from the French Mediterranean continental shelf (guille, 1970), and also from Mediterranean bathyal depths (carpine, 1970), during the Polymède mission (laubier and ramos, 1974), in the bay of rosas (Desbruyères et al., 1972) and in the Adriatic (Katzmann and laubier, 1975).However, it has always been described as having a short, simple, cirriform median antenna (gaston and Mclelland, 1996).
in our samples, 12 specimens from capbreton (as well as several more from the continental slope off galicia) agree well with the description of A. monicae.However, one individual, non-distinguishable from A. monicae in most relevant characters, had a bifurcated median antenna very similar to that of Aricidea mediterranea and Aricidea (Aricidea) petacalcoensis león-gonzález, Hernández-guevara and rodríguez-Valencia, 2006.
Therefore, Aricidea monicae, like Aricidea mediterranea, also seemed to have two forms, the predominant one being that with a simple antenna (as our finding is the first reported with bifurcated antenna found).This finding also suggests that the simple antenna may not be the result of a regeneration as stated by laubier and ramos (1974).
Discussion.Aricidea quadrilobata Webster and benedict, 1887 was first collected on the north American Atlantic coast.Since then, several similar species have been described: A. uschakowi zachs, 1925zachs, , A. antennata, and A. longicornuta berkeley and berkeley, 1950zachs, in the Pacific, and A. annae laubier, 1967 in the Mediterranean.According to Strelzov (1973) all of them, except for A. uschakowi, should be considered as A. quadrilobata (including the most recent records of A. uschakowi). A. quadrilobata has biramous notopodial postchaetal lobes in the branchial region (Strelzov, 1973), a typical character of large individuals (width >0.6 mm).conversely, blake (1996) stated that A. antennata (from Pacific coasts) and A. quadrilobata (from Atlantic coasts) were different, the latter being characterized by a wide prostomium and simple notopodial postchaetal lobes, and the former by an anteriorly truncated prostomium, apparently trilobed, and biramous notopodial postchaetal lobes.All the specimens from capbreton canyon agree with A. antennata.Distribution.eastern Pacific, from canada to southern california; Western Pacific, Japan, Sea of okhotsk.capbreton canyon (bay of biscay).492 to 1113 m depth.First record to the Atlantic.Material examined.Two specimens from capbreton canyon, bay of biscay, Atlantic ocean (coordinates in Table 1).cb89/KF50: Holotype (Mncn 16.01/11206) and one paratype (inSub Pol 333).
First two notopodial postchaetal lobes digitiform, becoming larger, cirriform from chaetiger 3 (Fig. 15A), considerably decreasing in size between penultimate branchial and first postbranchial segments, keeping that size in all other segments.neuropodial postchaetal lobes tuberculated from chaetiger 1 to 15. Modified ventral chaetae from chaetiger 61, as capillary chaetae attenuated and prolonged in a fine, long distal filament, with a pubescent area at attenuation level (Fig. 15c).
Etymology.This species is named in honour of rafael Sardá, in recognition of his work on polychaetes of the Mediterranean Sea.
Etymology.The species name refers to the shape of median antenna.
Discussion.Aricidea bifurcata n. sp. is compared with other taxa of the subgenus Aricidea (Allia) in Table 4.The main morphological characters, interparapodial lobes included, resemble those of Arici dea (Allia) monicae, but can be distinguished by its bifurcated median antenna, very similar to the antennae of Aricidea (Allia) mediterranea and Aricidea (Aricidea) petacalcoensis (see the remark section under Aricidea (Allia)).
Description.large specimen 6.68 mm long, 0.63 mm wide for 46 chaetigers; small one in poor condition.
granular brownish pigmented spots in each segment.
Discussion.The number of branchial pairs (from 7 to 14) starting at chaetigers 4 (small specimen) and 3 (large specimen) agree well with Hartley (1981: fig. 4), who reported from 7 to 36 pairs and the same sizerelated variability in the starting branchial chaetiger.

AcKnoWleDgeMenTS
Thanks are due to Dgo (Talence) for the loan of the Flusha box-corer; to biMM-MnHn (Paris) for the loan of the sieving equipment; to the French cirMAT (inSu-cnrS) for logistical support and loan of an epibenthic dredge; to the crew of the rV "côte d'Aquitaine" for their helpful assistance at sea; to A. urzelai, i. esteban and i. zabala (inSub, Donostia) for their helpful contribution to the sorting of capbreton samples; to Dr. J.c. Dauvin (MnHn, Paris) for kindly providing bibliography; to David romero (universidade da coruña) for providing the final versions of the species drawings and to r. Andrade (uPV-eHu) for the SeM photographs.
Distribution.Western Mediterranean -rosas bay, 115 m depth, Adriatic, 525 to 530 m depth.capbreton canyon, bay of biscay, 508 to 576 m depth.First record for the Atlantic.

Table 3 .
list of species in the genus Paradoneis with some comparative morphological characteris Species considered by some authors as being a junior synonym of P. armata.