Revision of Diplocirrus Haase, 1915, including Bradiella Rullier, 1965, and Diversibranchius Buzhinskaja, 1993 (Polychaeta, Flabelligeridae)

Abstract Diplocirrus Haase, 1915, includes flabelligerids having cylindrical to club-shaped bodies, with cirriform papillae, multiarticulate chaetae in both parapodial rami, 8 branchial filaments of two types (thick and rarely lamellate, or cirriform), gonopodial lobes in chaetigers 5 or 6, or multiple gonopores along some anterior chaetigers. Bradiella Rullier, 1965, has included only the type species: Bradiella branchiata Rullier, 1965, described from Eastern Australia. The original description has been overlooked and it lacked enough details on branchial and chaetal features. Diversibranchius Buzhinskaja, 1993, with Diplocirrus nicolaji Buzhinskaja, 1994, as the type species, was introduced for a similar species from the Japan Sea. These two monotypic genera share the same morphologic features with Diplocirrus, and are herein regarded as its junior synonyms. As herein redefined, Diplocirrus includes, besides its type species, Diplocirrus glaucus (Malmgren, 1867)from Scandinavia : Diplocirrus branchiatus (Rullier, 1965), comb. n. from Queensland, Australia, Diplocirrus capensis Day, 1961 from South Africa, Diplocirrus erythroporus Gallardo, 1968 from Vietnam, Diplocirrus hirsutus (Hansen, 1882) from Arctic and subarctic regions, Diplocirrus incognitus Darbyshire & Mackie, 2009 from South Africa, Diplocirrus kudenovi sp. n. from off Western Mexico, Diplocirrus longisetosus (von Marenzeller, 1890) restricted to the Bering Sea, Diplocirrus micans Fauchald, 1972 from deep water off Oregon and Western Mexico, Diplocirrus nicolaji (Buzhinskaja, 1994), comb. n. from the Japan Sea, Diplocirrus normani (McIntosh, 1908), comb. n. from Scandinavia, Diplocirrus octobranchus (Hartman, 1965), comb. n. from off New England, and Diplocirrus stopbowitzi Darbyshire & Mackie, 2009 from the Irish Sea.


Multiarticulated capillaries, body papillae, soft-bottoms introduction
The delineation of flabelligerid genera has been problematic since Grube (1877a); especially because the eversible anterior end, carrying the branchiae and palps, is rarely exposed. Branchial and chaetal features were employed to propose most genera, but their delineation was not clear-cut in most instances, especially because the branchiae are rarely everted. Thus, Diplocirrus was proposed by Haase (1915:194) following some earlier indications by von Marenzeller (1889:130), and by de Saint-Joseph (1898:366). These two authors have commented on the need to separate the species placed in Stylarioides delle Chiaje, 1831 by using the branchial arrangement. The species transferred to Diplocirrus have four pairs of cirriform, heteromorphic branchiae: the four distal filaments are shorter and thicker, whereas the proximal two pairs include thinner, longer filaments. The branchial filaments in the distal or posterior row differ from those found on the proximal or anterior row; they are basally prismatic due to the fact that when specimens are alive, they are closely packed making a branchial wall. Handling specimens often causes the branchial filaments to separate from the others, such that their lateral connections are not noticed. Further, although the posterior row filaments are thicker than the proximal row filaments, they are dehiscent. The proximal filaments are cirriform separated as two lateral pairs, but are completely free from each other, such that in living or preserved specimens, they look loose and and are deshiscent as well.
It is noteworthy that D. capensis Day, 1961 was described as having all branchiae of the same size and width, neurochaetae distally falcate, and without a cephalic cage. This combination of characters made Day expand the generic diagnosis with some hesitation (Day 1961:510). The whole body was later illustrated (Day 1967:665, Fig. 32.4e), and the emended diagnosis was confirmed. On the other hand, Hartman (1965:178) described Ilyphagus octobranchus and made some comments on its affinity with D. capensis; later, Day (1973:106-107) repeated her observations and because of their proximity, regarded Ilyphagus Chamberlin, 1919 as a junior synonym to Diplocirrus. As stated elsewhere (Salazar-Vallejo et al. 2008:204), this synonymy cannot be supported because of, among other things, the striking differences in body shape, cephalic cage development, and type of neurochaetae. Further, Darbyshire & Mackie (2009:96), after studying the type material, have found that it has the Diplocirrus, typical branchial pattern, and they compiled a table with the morphological characters for most species in the genus. Chamberlin (1919b:397) introduced Saphobranchia for Stylarioides longisetosus von Marenzeller, 1890; however, he overlooked the revision by Haase (1915) who had established Diplocirrus, including this species into his generic definition. Thus, Saphobranchia is a junior synonym of Diplocirrus.
Further, eight genera in the polychaete family Flabelligeridae de Saint-Joseph, 1894 have been regarded as monotypic: Bradabyssa Hartman, 1967, Bradiella Rullier, 1965, Coppingeria Haswell, 1892, Diversibranchius Buzhinskaja, 1993, Flabelliderma Hartman, 1969, Pantoithrix Chamberlin, 1919, Poeobius Heath, 1930, and Therochaetella Hartman, 1967. The proposal of some of these genera may be explained by the lack of a revisionary work that clarifies the generic delimitations in the family. For example, Flabelliderma has been redefined recently and it is no longer a monotypic genus (Salazar-Vallejo 2007). Coppingeria has been merged into Stylarioides delle Chiaje, 1831, as indicated elsewhere (Salazar-Vallejo 2011a), and Therochaetella has been regarded as a junior synonym of Trophoniella Caullery, 1944(Salazar-Vallejo 2011b. On the other hand, Bradiella has been only known by its type species, B. branchiata Rullier, 1965, which was described from Moreton Bay, Queensland, Australia. Spies (1975) studied some specimens from the type locality, but they were identified as Diplocirrus cf. capensis Day, 1961. This was an unfortunate decision because B. branchiata was overlooked by posterior scientists working in the same area. Specimens from a similar species were found in the Sea of Japan by Buzhinskaja (1994); she documented several interesting morphological features, and concluded they were different enough from D. cf. capensis. Thus, she proposed a new genus: Diversibranchius, because of the strikingly different branchial filaments.
With this contribution, we revise Diplocirrus and regard Bradiella and Diversibranchius as junior synonyms based on review of type, topotype, and additional materials. Diplocirrus is amended and now contains 13 species that live on soft bottoms in sublittoral depths throughout the world.

Materials and methods
The relative size of notochaetae and their articulation pattern are based on median chaetigers, about chaetiger 10. As in other contributions in this series, specimens were photographed using available equipments; specimens were often temporarily stained with an over-saturated alcoholic methyl green solution. When available, head are depicted in frontal views once branchiae and palps are removed. Plates were prepared by combining several photographs by hand or by using HeliconFocus. Type and non-type materials belong to the following institutions.

Museum acronyms AM
Australian Museum, Sydney.

BMNH
The Natural History Museum, London.

CAS
California Academy of Sciences, San Francisco.
Branchiae. Branchial filaments are made in two different types and can be separated in two series in relation to the prostomium. The proximal series is the anterior row and the distal one is the posterior row. The posterior row includes prismatic or cuneiform filaments, whereas the anterior row is made of cirriform filaments. The posterior row includes four filaments laterally fused to each other, forming a branchial wall that was illustrated by Haase (1915:29, 197, Fig. 5). This wall is formed because each branchia has two lateral sockets keeping them together, and making their separation difficult. Once separated, each filament is more or less triangular in cross section, but there are two basic modifications; filaments have ciliary bands in most species, whereas in a few of them, filaments are convoluted, with transverse ridges along their surface. Further, because in the latter species the dorsal side is often projected with a long wing, whereas the ventral side might be widened by the presence of multiple blades or lamellae, each filament has a prismatic or cuneiform appearance. These lamellae might be restricted to the proximal or cirriform branchial filaments, even in those species lacking the complex features seen in some posterior branchial filaments. Because they are variable within species, the shape of the filaments and the presence of ciliary bands are not useful for distinguishing genera. These blades are made by either a single series of convoluted filaments, or by a series of transverse filaments arranged as twin blades, but marginally independent of the following blade. Further, these blades can extend over different regions along the back of each branchial filament. The anterior row includes four filaments too, but they are separated in two lateral pairs. Each filament is cirriform, thinner, usually provided with a series of transverse ciliated ridges and, if provided with filamentous blades, hence lamellate, they are more or less restricted to the basal region. As in other flabelligerids carrying two series of branchial filaments as in Pherusa, there are some basal branchial knobs between the posterior branchial filaments. They resemble some short, rounded reinforcements present in sabellids or serpulids, and their relative development, whenever evident, might be useful to separate similar species.
Chaetae. All chaetae in Diplocirrus are multiarticulated with notochaetae thinner than neurochaetae. The multiarticulated notochaetae provide useful diagnostic features by their relative size, in relation to body width, or by the relative size of articles along the chaetae. Thus, articles are regarded as short if they are wider than long, mediumsized if they are as long as wide, and long if they are longer than wide. This variation in the articulation pattern is also present in neurochaetae and because it is a conservative feature, is often used to separate similar species.
Gonopodial lobes and gonopores. Adult members of some Diplocirrus species carry two projected lobes in chaetigers 4 or 5 which were regarded as neprhidial papillae since Haase (1915). However, nephridial lobes are restricted to the branchial plate, the projected, segmental lobes have a reproductive role and are consequently regarded as gonopodial. Their position in a given chaetiger, as well as their relative color and shape can be used to separate similar species. Three species lack gonopodial lobes, but have several pairs of ventral, rounded, reddish or dark orange structures of unknown function; pending a histological confirmation, they are herein regarded as gonopodial.
These multiple paired structures have been described for D. erytrhoporus Gallardo, 1968, andD. glaucus orientalis Gibbs, 1971, which is regarded as a junior synonym for the former. Because some species of Diplocirrus lack gonopodial lobes, and because they might be present only during reproduction, their presence or absence could not be employed as a generic diagnostic feature, and the multiple gonopores would be in the same condition. Diagnosis. Body clavate or subcylindrical, often anteriorly swollen. Cephalic cage variably developed. Body papillae abundant, short giving a velvety appearance, or very long, giving a hirsute outlook, sometimes adhering sediment particles. All chaetae multiarticulated capillaries; neurochaetae thicker, sometimes falcate. Branchiae sessile, 4 pairs, distal branchiae thicker, often shorter, proximal branchiae thinner, often longer, sometimes basally lamellate. Gonopodial papillae present in chaetiger 4 or 5, or a series of paired ventrolateral gonopores along some anterior chaetigers.
Remarks. Haase (1915) proposed Diplocirrus for those species formerly included in Stylarioides having two different sizes of branchiae, and multiarticulated capillaries only. Some species currently included in the genus had been previously described in either Trophonia or Stylarioides. However, as an independent genus, it differs by having two different sizes of branchiae, and all chaetae are multiarticulated capillaries. Webster and Benedict (1887:730) proposed Zorus, with Zorus sarsi as the type and only species. They indicated that it had a body anteriorly swollen, becoming thinner posteriorly, only with capillary chaetae, and stated that branchiae and palps arise from an eversible stalk but gave no details on the size relationship of branchiae. Hartman (1961:122) regarded Zorus as a junior synonym of Piromis Kinberg, 1867. However, because of the body form and chaetal features, it rather resembles Diplocirrus, because Piromis has few papillae arranged in longitudinal rows and sometimes bifid neurohooks, which were not found in Zorus. Webster (1879:46) had already stated the differences among capillary chaetae and ventral hooks when he described another flabelligerid; so, there is no room for any such confusion. The only illustration provided by Webster and Benedict (1887, Pl. 5, Fig. 67), shows a cross section of a middle segment with very long chaetae, and long papillae. These features resemble D. hirsutus (Hansen, 1879), which is comm in the Bay of Fundi (Appy et al. 1980:32). However, because there is no type material, the generic definition did not include a size relationship of branchial filaments, and the description and illustration lack critical information, Zorus sarsi has been regarded as indeterminable (Salazar-Vallejo et al. 2008).
These differences prompted Day (1961:510), to propose a misfortunate redefinition of Diplocirrus, because the branchial features have been employed to establish it by Haase (1915:26, 194). Especially because the posterior row of branchiae are not just thicker than the anterior row filaments; rather, they tend to be closely packed, with each filament laterally fused forming a branchial wall. Further, the cirriform thinner branchiae are contractile, and if they were observed completely relaxed by Day, he might have had the impression that they were of about the same thickness. Later, Day (1973:106-107) modified the generic definition of Diplocirrus concluding that it would also include Ilyphagus Chamberlin, 1919. However, as stated above, this second emendation is problematic as well, because of marked differences in neurochaetae, because in the species of Ilyphagus neurochaetae are aristate neurospines whose handle is made of fused or anchylosed, short articles, and a hyaline fragile tip, whereas in Diplocirrus there are only multiarticulate, often falcate, neurochaetae.
Bradiella Rullier (1965:190) was compared with Diplocirrus and Brada Stimpson, 1854. The branchial features were incompletely described (see below); it was regarded as different from these two genera because of the branchiae, and because it lacks gonopodial lobes. The potential differences between Bradiella and Diplocirrus would be that in Bradiella there are no gonopodial lobes in chaetigers 4-5, but gonopodial lobes have not been recorded in some Diplocirrus species at all. Further, the surface of branchial filaments is very complex in Bradiella, because it is provided with lamellate complex filaments, in contrast to the cirriform or tapering filaments which might barely have some ciliated bands, but some Diplocirrus species have a complex lamellar structure along the branchial filaments bases. Spies (1975) studied specimens from the type locality, Moreton Bay, Queensland, Australia, but overlooking the paper by Rullier (1965), identified them as Diplocirrus cf. capensis Day, 1961. He noticed that the branchiae include eight filaments, not just two as stated by Rullier, with four cirriform and four lamellate filaments. Spies (1975, Pl. 6, Fig. 18) illustrated a (lateral) dorsal spoon-like branchia provided with a flat lateral lobe, and a series of independent branchial blades. Thus, because there are variations in the presence of gonopodial lobes and in the development of lamellar structures in branchial filaments, the only difference to separate the Bradiella-like species would be the presence of paired ventrolateral pores. However, because there is no other major difference in chaetal types, Diplocirrus and Bradiella are regarded as synonyms.
Because of the rediscovery of these peculiar branchial features, Buzhinskaja (1994) established Diversibranchius. However, she overlooked Rullier (1965) as well, and compared his specimens with Diplocirrus, stressing its resemblance with D. cf. capensis. She found that branchiae were of two types, cirriform and prismatic, or cuneiform, provided with foliose projections, and illustrated that both have convoluted branchial lamellae giving the impression of a series of independent blades, as was illustrated by Spies (1975). Bradiella and Diversibranchius Buzhinskaja, 1994 resemble each other by having two different types of branchiae, short to long body papillae, and multiarticulated neurohooks. These two genera are herein regarded as junior synonyms to Diplocirrus, such that the type species are redescribed, and transferred and newly combined into Diplocirrus.
Two of these species (D. incognitus and D. stopbowitzi), have been recently described and only their diagnosis and illustrations are included. On the other hand, three other currently undescribed species are informally characterized but not all have been included in the key because the quality of the materials; one is from Morocco, another one from off Sri Lanka, and the other from Antarctica. The species can be separated using several morphological features as stated below. Additional material. Norway. One specimen (MNHN-A183), broken in two, without posterior end, anterior end exposed, appendices lost, Solsvick, no further data. Many specimens , Hardangerfjorden (60°10'00"N, 06°00'00"E) separated as follows: 14 anterior fragments (LACM-AHF 2620), Stat. Z20, 7 Jun. 1957, 25-16 m (up to 36 chaetigers, all with multiarticulated neurohooks; in posterior chaetigers with over 10 long articles). Two posteriorly incomplete specimens (LACM-AHF 2622), Stat.  (Malmgren, 1867). Non-type specimens (LACM-AHF 2683), Norway A anterior end, dorsal view, head exposed B same, close-up of branchiae showing longitudinal striae C another specimen (221), head exposed, anterior end, lateral view D same, cirriform branchiae with basal ridges e same, head, frontal view, branchiae and pals removed (BS: branchial scars, DL: dorsal lip, LL: lateral lip, NL: nephridial lobes, PS: palp scar), another specimen (LACM-AHF 2627) F chaetiger 24, basal, medial and distal notochaetal regions G same, basal, medial and distal neurochaetal regions.
Remarks. Diplocirrus glaucus (Malmgren, 1867) is closely allied to D. incognitus Darbyshire & Mackie, 2009 because both have swollen anterior chaetigers and some sediment particles scattered over the body. They differ in the relative size of lateral papillae and notochaetal articulation; thus, D. glaucus has smaller papillae (up to onefifth notochaetal length), and poorly defined basal articles in notochaetae, whereas D. incognitus has longer papillae (up to one-third notochaetal length) and medium-sized basal articles in notochaetae.
The original description (Malmgren 1867) indicated that the color was variable from bluish-gray to greenish or pale, but the number of chaetae in chaetiger 1 was stated as about 3, which has been used to separate it from similar species. The species was originally described from Bahusiae (Malmgren 1867:192), corresponding with the current Bohuslan (58.88° N, 10.51° E), where the Tjarno Marine Biological Station is, and where some of the specimens used for this description were collected.
Holotype with branchial plate damaged. Posterior branchiae compressed, lateral filaments lost, median filament bent towards the mouth, lamellate; cirriform branchiae lost, two lateral scars per side, placed below a dorsal crest. Slide IRFA-W40' shows a branchial blade made of fused branchial filaments. Another specimen (SAM-GR201), with head slightly exposed (Fig. 2H), branchiae complete of two different types. Posterior row with four prismatic, thicker, lamellate branchiae, tips bare (Fig. 2J); lateral branchiae smaller (one in regeneration), each with dorsal keel reduced, with longitudinal bands, dorsal surface laterally expanded with a thin axis, provided with two rounded lateral lobes; median branchiae larger, dorsal keel large, foliose, markedly corrugated. Distal branchiae with ventral side with a blade made of fused branchial filaments, convoluted, looking like a series of successive blades, but actually made by a single convoluted blade. Anterior branchial row with four thin, cirriform filaments, shorter than palps, arranged in two lateral pairs, each filament with a convoluted lamella along its basal third (Fig. 2I), and successive ciliary bands medial-and distally. Branchial basal lobes between median and lateral branchiae (dorsal), and outside the lateral ones (lateral); dorsal lobes small, rounded, lateral lobes rounded, larger).
Parapodia poorly developed; chaetae emerge from the body wall. Notopodia and neuropodia with papillae as long as other papillae. Noto-and neuropodia close to each other. Median neuropodia lateral, very close to notopodia.
Chaetal transition from first chaetiger to body chaetae abrupt; notochaetae of chaetigers 2-3 large multiarticulated hooks, distal article hooked, entire. All other notopodia with multiarticulated capillaries. Median notochaetae arranged in a longitudinal line. Notochaetae of chaetigers 1 and beyond the third, multiarticulated capillaries; by chaetiger 11, as long as half body width, 10-11 per bundle (6-7 in smaller specimen), each with long articles throughout the chaeta (Fig. 2F). Neurochaetae multiarticulated hooks from chaetiger 1, arranged in a short J-shaped pattern, 4-5 per bundle, each with long articles of about the same length, tips falcate (Fig. 2G), with a hood-like membrane.
Posterior end invaginated in holotype; other specimens with truncated rounded lobe; notochaetae directed posteriorly; without anal cirri.
Variation. Pigmentation varies from pale orange to dark yellowish, or to dirty pink with gonopodial pores reddish or pale. Further, there are two main variations related to body size: papillae are longer in larger specimens, and gonopores become more pigmented, and are probably present along more segments as body enlarges.
Remarks. Diplocirrus branchiatus (Rullier, 1965) comb. n. is very similar to D. nicolaji (Buzhinskaja, 1994), comb. n. because both species have bodies without sediment particles, ventrolateral gonopores along several anterior chaetigers, short chaetae in the first chaetiger, and their caruncle tapers posteriorly. They differ in the relative development of neurochaetae and of the extent of the lamellate area in their cirriform branchiae; thus, in D. branchiatus median chaetigers have neurochaetae with about 23 articles, tapering to a delicately falcate tip, and the lamellate region might be up to onefifth of the branchial length, whereas in D. nicolaji, neurochaetae are barely tapering, having about 10 articles, their tips are markedly falcate, and the lamellate region might extend up to one-third of branchial length.
Diplocirrus branchiatus (Rullier, 1965) has been known only through the original description. Spies (1975) studied some specimens from the type locality (herein re-examined); they fit the original description but the anterior end was previously removed. Rullier's description is fairly complete, though the presence of multiarticulated hooks in notopodia 2-3 was overlooked, as well as the presence of the gonopores. The anterior end has a symmetrical pattern and the original description does not provide complete details about branchiae; however, the drawings show that there were two larger lamellate branchiae (his figure 7C), and that there were smaller lateral branchiae (his figure 7D), but there are no details on cirriform branchiae; they might have been lost during dissection. As originally shown by Rullier (1965), and confirmed by the observation of one permanent slide, branchial blades include a series of parallel filaments; however, they are not arranged as successive, independent blades but rather as a continuous, convoluted, branchial blade. So far, this special type of branchial pattern is only known for a few species in Diplocirrus. Further, Rullier illustrated that neurohooks are distally tapering (his figure 7G), but he described them as (p. 190) "plus courtes et recourbées à leur extrémité" (shorter and distally curved), which is the correct description. Spies (1975) made some observations and his drawings are slightly inaccurate in several features: the caruncle does not taper posteriorly, and does not reach the posterior margin of the branchial plate, interbranchial lobes were not illustrated, and the lateral palp lobes were not seen.
Cephalic hood not exposed. Prostomium with four small, black eyes. Palps thick, as long as branchiae. Caruncle projected dorsally, not reaching the posterior margin of branchial plate. Lips corrugated, fused. Nephridial lobes in branchial plate not seen. Branchiae very dark, of two types. Posterior row with four wedge-shaped filaments; anterior row branchiae cirriform, separated in two lateral pairs by the caruncle. Interbranchial lobes not seen. Lamellate region difficult to evaluate.
Parapodia poorly developed; chaetae emerge from the body wall. Notopodia and neuropodia with papillae longer than other body ones. Noto-and neuropodia close to each other. Median neuropodia lateral, very close to notopodia.
Posterior end unknown. Day, 1961 is closely related to D. kudenovi sp. n. and D. stopbowitzi Darbyshire & Mackie, 2009 because their bodies do not incorporate sand particles, and by lacking ventrolateral gonopores. However, these two latter species are provided with hemispherical papillae whereas in D. capensis papillae are elongate, often basally swollen, but never hemispherical.

Remarks. Diplocirrus capensis
The records of D. capensis by Day (1973:105-107), and Milligan (1984:47.9-11, Figs. 47.5-6) differ from the typical South African form because they have different body color, cephalic cage, larger lateral papillae, and by the relative numbers of chaetae. They might represent a different species but their description as new species must wait for better specimens. There is a similar, apparently undescribed species in the Mediterranean Sea, which has been recorded as D. glaucus by Fauvel (1937:34, non Malmgren, 1867. The materials are damaged (MNHN-406), many chaetae broken, anterior regions smashed or without exposed head, and were collected off Alexandria, Egypt. Better specimens would help clarifiy its affinities with D. capensis.
Posterior end missing in holotype; non-type specimen (NTM-18920) with posterior end tapering to a blunt cone; pygidium with anus terminal, no anal cirri.
The original description (Gallardo 1968) was brief. It indicated that there were six tentacles (branchiae), four larger and two smaller ones, and there were no details on the extent of the cephalic cage. Thus, a redescription was required in order to separate this species from other similar ones in the Indo-Pacific regions. The two additional specimens were one maculated with rounded dark brown spots (11 mm long, 2.5 mm wide, cephalic cage 0.9 mm long, 22 chaetigers, gonopores in chaetigers 4-13), which was dissected to study the anterior end, and another without dark spots (14 mm long, 2.8 mm wide, cephalic cage 1.0 mm long, 23 chaetigers, nephridial pores in chaetigers 4-12; it is a mature female). Diplocirrus glaucus orientalis Gibbs, 1971 was described without illustrations; it has orange globular papillae below each neuropodium in chaetigers 4-14(16). This could include the record of D. glaucus by Fauvel (1932:186-187, Fauvel 1953:353, Fig. 184a-d). It is being regarded as a junior synonym of D. erytrhoporus.
Anterior end observed in a previously dissected specimen and in non-type specimen. Cephalic hood short, smooth, margin smooth. Prostomium low cone, grayish, eyes barely pigmented (Fig. 4E), difficult to be seen in syntype or non-types. Caruncle not observed in syntype, weakly defined in non-types. Palps thick, longer than the only available cirriform branchia; palp bases rounded, projected. Lateral lips projected, thick, well-developed, dorsal and ventral lips reduced. Branchiae mostly lost, scars remain; posterior row with thicker scars, anterior row with a single cirriform branchiae without basal blades (all cirriform, posterior ones slightly thicker, smooth). Nephridial lobes rounded, elevated, separating anterior and posterior branchial rows (taking methyl green stain deeply).
Remarks. Diplocirrus hirsutus (Hansen, 1878) resembles D. longisetosus (von Marenzeller, 1890) andD. normani (McIntosh, 1908), comb. n. because they have bodies provided with long papillae but without sand particles. Their main difference lies in the relative length of notochaetae in median chaetigers, because the latter two species have notochaetae markedly longer than body width, whereas in D. hirsutus they are about as long as body width. Haase (1915:199) noticed the cinnamon-red color for specimens of this species. The available specimens show a concentration of the pigment towards the anterior end, making a thin crust surrounding papillae and chaetae. Thus, it is not the basic color of the organism but rather some adsorbed minerals on these structures and, whenever this pigmentation is present, chaetae are darker, which indicates that the minerals are either ingested and later used for chaetal formation, or adsorbed to chaetae as well as over the tunic. This pigmentation should rely on the minerals available in the sediments, and therefore should not be used as a diagnostic feature.
Etymology. This species is named after Jerry D. Kudenov, who has studied several polychaete families on a world-wide basis, and especially for his series of publications on the polychaetes from the Gulf of California, which have been very useful for many researchers working in the region, including one of us (SISV). The epithet is a noun in the genitive case.
Posterior end tapering to a rounded lobe; pygidium with anus terminal, without anal cirri.

Remarks.
As currently restricted, Diplocirrus longisetosus (von Marenzeller, 1890), closely resembles D. micans Fauchald, 1972 andD. normani (McIntosh, 1908), comb. n. These species have notochaetae longer than the body width, and long papillae with-out sand particles, although D. micans separates from the other two species by having neurochaetae with long articles, and because it lacks gonopodial lobes. Then, D. longisetosus and D. normani differ especially in the relative body color, papillae and gonopodial lobes, and on the relative resolution of neurochaetal basal articles. In D. longisetosus, papillae and gonopodial lobes are pale, and basal neurochaetal articles are well-defined, whereas in D. normani, the body is grayish, and papillae and gonopodial lobes are darker or blackish, whereas neurochaetal basal articles are poorly-defined.
Further, D. longisetosus was described from Providence Bay, Russia, in the Bering Sea, with a single anterior fragment. Haase (1915:200) studied the supposed holotype (which is now lost), an additional specimen sent him by von Marenzeller, probably coming from Spitzbergen, Norway, and an additional broken specimen. This combination resulted in a mixture of morphological features and the species has been recorded from several localities in the Arctic Ocean as well as in the Northern Atlantic and Northern Pacific. Consequently, a redescription and proposal of a neotype is needed to clarify if there is more than one species. Støp-Bowitz (1948:32) noticed the nephridial lobes in the branchial plate, but he regarded them as accessory branchiae.
After the International Code of Zoological Nomenclature (1999, Art. 75), a neotype is being designated because there is no name-bearing type specimen, and because of the confusion between the above two species requires a designation to objectively define D. longisetosus. Consequently, in order to satisfy the qualifying conditions (Art. 75.3), it must be stated that this designation will clarify the taxonomic status, a description and illustrations have been presented to ensure the recognition of the species. Further, collection managers in several German museums were contacted in order to find the type material for this species, but none exists. On the other hand, the neotype fits the characteristics originally noticed in the species, it was found in a locality with ecological conditions similar to the ones prevailing in the original type locality, and has been deposited in the California Academy of Sciences.
Posterior end unknown. Fauchald, 1972 resembles other species with abundant papillae and long chaetae such as D. longisetosus (von Marenzeller, 1890), andD. normani (McIntosh, 1908), comb. n. However, D. micans separates from the two other species because its neurochaetae have long articles, and there are no gonopodial lobes, whereas the two other species have distal articles barely longer than wide, and gonopodial lobes.

Remarks. Diplocirrus micans
The record by Fauchald and Hancock (1981:36) was based on a single, damaged specimen collected off Oregon, United States. The specimen (LACM-AHF 2616) resembles D. micans but it is brittle, apparently it has dried out in the past, so the conical lobes in first few chaetigers cannot be confirmed. However, this specimen has many more chaetae per bundle, especially in the anterior end, and articles are much longer than in D. micans, so it may be a different species, but the specimen is in poor shape and more specimens are required to describe it.

Diplocirrus nicolaji
Branchiae of two different types (Fig. 9A, B). Posterior row with four prismatic, thicker, lamellate branchiae, lamella reaching the tips; lateral branchiae of the same size, with dorsal keel rounded, reduced, with longitudinal bands and laterally expanded dorsal surface, with a thin axis, branchial lateral margins with two rounded, sucker-like sockets; median branchiae with dorsal keel as those present in lateral branchiae, not foliose, corrugated. All posterior branchiae with a series of successive transverse blades on their ventral side; in median branchiae, all laterally fused making a single convoluted blade; in lateral branchiae the transverse blades laterally free. Anterior row with four thin, cirriform branchiae, shorter than palps, arranged in two lateral pairs, each filament with a convoluted lamella along its basal third, and successive ciliary bands medial-and distally. Interbranchial lobes small, between median and lateral branchiae (dorsal), and outside the lateral ones (lateral); dorsal lobes small, rounded, lateral lobes rounded, slightly larger).
Parapodia poorly developed; chaetae emerge from the body wall (Fig. 9F). Notopodia and neuropodia with papillae as long as the others. Noto-and neuropodia close to each other. Notochaetae multiarticulated capillaries, all articles long (Fig. 9G). Median notochaetae arranged in a longitudinal line, with 4 per bundle in holotype (11 per bundle in larger; 6-7 in smaller specimens), longest about as long as one-third body width. Median neuropodia lateral, very close to notopodia. Neurochaetae multiarticulated hooks from chaetiger 1 (Fig. 9G), arranged in a short longitudinal line (J-pattern in other specimens), 3-4 per bundle (6-8 in other specimens), each with long articles of about the same length, distal article falcate, finely transversely divided, not articulated, with a hood-like membrane.
Remarks. Diplocirrus nicolaji (Buzhinskaja, 1994), comb. n. is closely allied to D. branchiatus (Rullier, 1965) because the bodies of these species lack sediment particles, have ventrolateral gonopores in some anterior chaetigers, reduced chaetae in the first chaetiger, and their caruncles taper posteriorly. Their main differences rely on the relative neurochaetal development in median chaetigers, and on the area covered by lamellae in the cirriform branchiae; thus, D. nicolaji has barely tapering neurochaetae, with some 10 articles of about the same length, tips markedly falcate, and their cirriform branchiae has a lamellate region extending up to one-third of the branchial length, whereas in D. branchiatus, the neurochaetae are tapering, provided with about 23 articles, decreasing in size distally, tips delicately falcate, and the lamellate region along cirriform branchiae might reach one-fifth of the branchial length.

Diplocirrus normani
Cephalic tube long, smooth, margin apparently smooth. Prostomium low, eyes not seen. Caruncle not seen. One palp remaining, thick, longer than remaining branchiae, longitudinal furrow shallow; palp keels reduced. Dorsal and ventral lips reduced, lateral lips thicker. Branchiae cirriform, most lost, sessile on branchial plate, arranged in two concentric rows, distal row continuous with 4 thicker filaments bases, proximal row discontinuous, filaments probably thinner, lower filaments bases smaller. Nephridial lobes very thin, long, placed below the posterior row lateral filaments.
Median notochaetae arranged in a transverse horizontal C-shaped pattern; all notochaetae multiarticulated capillaries, short articles basally and distally, long medially (Fig. 10E). About 10 (-12) chaetae per bundle, at least twice as long as body width. All neurochaetae multiarticulated capillaries, short, poorly-defined articles along basal half or 2/3 chaetal length, better-defined, medium-sized and then long articles along the rest of chaetae (Fig. 10F), tips straight, arranged in a transverse line, 9-10 per bundle.
Posterior end tapering to a rounded lobe; pygidium with anus terminal, blackish, without anal cirri.
Remarks. Diplocirrus normani (McIntosh, 1908), comb. n. was regarded as a junior synonym of D. longisetosus (von Marenzeller, 1890) by Haase (1915:200) because they are very similar. As stated above, they also resemble D. micans Fauchald, 1972, though the latter separates from the other two species because it lacks gonopodial lobes and its neurochaetae have long articles. Thus, once D. longisetosus has been restricted, these species differ regarding coloration of body, papillae and gonopodial lobes, and because of the relative resolution of neurochaetal basal articles. Thus, in D. normani, although the body is grayish, papillae and gonopodial lobes are darker or blackish, and neurochaetal basal articles are poorly-defined, whereas in D. longisetosus, on the contrary, the papillae and gonopodial lobes are pale, and the basal articles of neurochaetae are well-defined.

Diplocirrus octobranchus
Cephalic hood tube long, made of two rings, basal one shorter, both smooth; cephalic hood margin smooth. Prostomium low, eyes not seen (Fig. 11C). Caruncle low, wide. Palps lost (pale, laterally corrugated, 1.5 times longer than branchiae in one paratype); palp keels rounded, elevated. Lateral lips thick, projected outwards, rounded. Ventral lip reduced. Dorsal lip projected as a triangular lobe. Branchiae cirriform of two different widths; posterior row with thicker filaments, rectangular, with a middorsal black band, branchial bases continuous, anterior row with branchiae thinner, cirriform, separated in two lateral pairs. Branchiae of about the same length; size relationships with palps unknown. Nephridial lobes in branchial plate low, whitish.
Posterior end observed in one complete paratype, tapering to a swollen pygidium, with anus dorsoterminal, without anal cirri. One paratype is a damaged female, oocytes 100-150 µm.
Remarks. Diplocirrus octobranchus (Hartman, 1965), comb. n., is closely allied to an undescribed species from Antarctica, and both differ from other species with long papillae because they have sand particles over the body. These two species differ in the extent of sediment particles along the papillae and on the relative length of the neurochaetal anchylosed region. Thus, in D. octobranchus sediment particles are restricted to the base of papillae, and their neurochaetae have an anchylosed region of about onefifth of the chaetal length, whereas in the Antarctic undescribed species, the sediment particles spread along the papillae, and the anchylosed region might be about half or one-third of the chaetal length.
Diplocirrus octobranchus is a typical member of the genus because its branchiae are of two different widths. It does not belong in Ilyphagus because it has multiarticulated neurospines, with long articles in the medial and distal regions, and short articles only basally, whereas in Ilyphagus neurochaetae are aristate spines with very short articles basal-and medially, and distally hyaline. Further, the cephalic cage chaetae in Ilyphagus are clearly dorsal whereas in Diplocirrus they are lateral, or dorsolateral at most. After Hartman amended Ilyphagus (Hartman 1965:177), the correct placement for her new species as a member of Diplocirrus was indirectly stated by comparing it to D. glaucus, the type species for the genus (Hartman 1965:179). This made Day (1973:106) suggest the informal, new combination, which is herein confirmed after the examination of the type material and of the redefinition of Diplocirrus.

Remarks.
As stated above, Diplocirrus stopbowitzi Darbyshire & Mackie, 2009 resembles D. kudenovi sp. n. because in both species the body has hemispherical papillae, but lacks sand particles or ventrolateral gonopores. They especially differ regarding some neurochaetal features in median chaetigers such as their number and the relative length of articles; thus, D. stopbowitzi has 2-3 neurochaetae, each with long articles being about seven times longer than wide, whereas D. kudenovi has 5-6 neurochaetae and each has shorter articles, each being twice as long as wide.