(Table 1): 413 specimens from 33 stations collected at 8–1350 m, bottom temperature -1.53–7.37 °C. Ten specimens from Alaid station 6 deposited at MNCN: 16.01/17777.

Thelepus antarcticus ZIN IV.1.2 (5 specimens)

Largest specimen 140 mm in length and 5 mm in width, although some fragments distinctly larger (up to 7 mm width); maximum size estimated at over 200 mm; larger specimens had been collected at shallow depths, less than 100 m. Number of segments increased with body size; number in investigated specimens: 113.

Buccal tentacles numerous, equal to body length, grooved. Eyespots rounded subepithelial spots, black or dark brown, numerous, usually in several transverse rows on back of upper lip (Fig. 2A, B). Even smallest specimens (<0.5 mm width R/V Sevastopol st. 1769) with numerous eyespots. Specimens from deepest sample (R/V Sevastopol, st. 1580, 1350 m) also with numerous eyespots.

Branchial filaments numerous, long and tangled (Fig. 2A, C–G). Due to tangling it was impossible to count number of branchial filaments in large worms (>5–6 mm width) without removing them one by one. Maximum number of BS1 filaments ca. 20–30, extending laterally to a point level with midpoint or lower edge of row of U1 uncini; outermost filaments usually 2–3 times shorter than those most developed. BS2 with a maximum of 15–20 filaments. One specimen (from Alaid 30.13) had four filaments on BS3 on right hand side of body; length of these was equal to notopodia of same segment. Filaments attached to a transverse elevated stump in 1–2 irregular rows depending on number of filaments. Number of filaments increases with body size; small worms (1–2 mm width) have fewer than 10 filaments on BS1. Smallest specimen (Sevastopol 1769, width <0.5 mm) with no filaments. Extension of filaments laterally depends upon worm size, with filaments extending only to level of upper margin of uncinal row in small worms. Wide medial gap separating left and right groups of filaments. Lateral lobes absent. Dorsum with warts or subepithelial honeycomb, forming more or less regular rows (Fig. 2C, F); number of rows increases with size of segments and worm. Segmentation distinct. Ventrum glandular, more so with increased “wrinkling” (Fig. 2H). Poorly visible, small nephridial papillae on S4–S7 above neuropodia; those on S5–S7 largest and usually only ones visible (Fig. 2D, F, arrowed).

Notopodia commence from BS2, with anterior notopodia large and transverse. Notopodia raised on body surface or flattened, depending on whether fixation occurs whilst within or outside of tube. Notopodia of BS2 equal to or only slightly smaller than those most developed. Notopodia numerous and present on almost all segments except 10–20 posteriormost developing segments; in investigated material present on up to 106 segments. Last notopodia poorly developed, several times shorter than those most developed and almost without rami, with only a few notochaetae; last neuropodia also reduced (Fig. 2I). Part of worm without notopodia not exceeding 10% of whole body length. Notochaetae in few (ca.10) anterior segments in two transverse rows: posterior row with long chaetae, distal half (winged part) becomes stained with methylene blue, anterior row with short chaetae; other notopodia with a single row of notochaetae. Notochaetae with narrow brims (Fig. 11B).

Neuropodia from C3, tori increasing in size to U10, then becoming progressively shorter. Uncini in a single row with well-developed prow and crest and one tooth in profile (Fig. 3); within a neuropodium main fang develops first, crest develops later (Fig. 3D).

Pygidium with crenulated margin, without cirri or papillae.

Morphologically, T. cincinnatus is closest to T. antarcticus Kinberg, 1866. The original description of T. antarcticus is very brief. The most complete re-description is by Benham (1921). It looks very similar to T. cincinnatus; however, I do not believe that it is the same species, since direct comparison of material from the northern and southern hemispheres is necessary to find differences. For the present time it can be stated that, although both species are of equal size (up to 200 mm length and 7 mm in diameter), T. cincinnatus has at least twice as many branchial filaments as T. antarcticus. The five specimens investigated (length up to 5 cm) have no more than 15 branchial filaments on BS1, distinctly fewer eyespots and slightly different uncini, with a hump (Fig. 3E).

Thelepus cincinnatus differs from other new species described herein as indicated: from T. davehalli sp. n. by the presence of eyespots and the absence of numerous completely developed posterior segments without notopodia; from T. marthae sp. n. by the absence of eyespots and by the higher number of branchial filaments and segments with notopodia; and T. parapari sp. n. has a crest of uncini on TU1 with two rows in profile, while T. cincinnatus has only one. Other species of Thelepus with two pairs of branchiae and eyespots have at least three times fewer branchial filaments and all but T. parapari sp. n. have half the number of segments with notopodia (Table 2).

The investigated material included almost 2000 specimens (from more than 100 stations) from the high Arctic to the Mediterranean, from depths between 2 m and almost 2 km. The type locality of T. cincinnatus is outside the ranges of all investigated species, but T. cincinnatus s. str. investigated specimens perfectly agree with the description of topotypes (Pettibone 1954). It is supposed that Pettibone’s description is that of the true T. cincinnatus.

In some samples, specimens lacked eyespots; however, this is likely to be due to fading because specimens in same samples (with several specimens present) have eyespots, but they are paler, smaller and less numerous than is typical. This fading seems to depend on preservation method: all material with faded eyespots had been stored in formalin for over ten years. The age of samples does not influence fading significantly; although all specimens without eyespots were collected over 50 years ago, other specimens collected a century ago and kept in alcohol had retained eyespots. So absence of eyespots should not be considered to be a characteristic of this species.

Three subspecies (varieties according to original descriptions) of T. cincinnatus have been described (Bellan 2008) and, based on the discussion below, none are considered valid.

Thelepus cincinnatus var. andreanae McIntosh, 1922. McIntosh wrote “dorsal cephalic collar with eye-specks”; as all other Thelepus with two pairs of branchiae from the area near the type locality also lack eyespots, this name should be accepted as a junior synonym of T. cincinnatus s. str. as believed by Bellan (2008).

Thelepus cincinnatus var. canadensis McIntosh, 1885; has eyespots according to the original description. Type locality: 43°04'N, 64°05'W, 51 fms. Specimens collected near the type locality of this subspecies (R/V “Persey-3” see Table 1) did not show differences from other specimens, confirming Hartman’s (1959) acceptance of T. cincinnatus var. canadensis as a junior synonym of the stem subspecies.

Thelepus cincinnatus var. profundus Roule, 1896. The description is too short to be informative: ‘Un seul individu, différant du type par sa taille e plus petite, par son tube plus mince et couvert extérieurement d’un enduit peu épais formé de vase grise, et par la forme de ses plaques onciales; ces dernières sont plus étroites, et leurs trois dents plus espacées’. No figures are given so it is impossible to determine which species he was describing and as no type material was deposited in Paris (Solís-Weiss et al. 2004) this subspecies should be treated as a nomen dubium.

Other literature reports of Thelepus cincinnatus include:

Fauvel (1927) reported for T. cincinnatus; “nombreux points oculiformes”; however, most or all the area covered by the “Faune de France” seems to lie outside the range of T. cincinnatus, but includes the range of T. parapari sp. n. with eyespots, so he probably observed T. parapari sp. n.

Zatsepin (1948) and Holthe (1986); despite their descriptions agreeing well with T. cincinnatus s. str., they probably observed the other species described here, because these species’ ranges fall within those covered by their papers. The same is true for our papers (Jirkov 2001; Jirkov and Leontovich 2013), where we overlooked T. marthae sp. n., T. davehalli sp. n., and T. parapari sp. n. but, in this case, it is supported by re-investigation of the material.

Hartmann-Schröder (1996) reported eyespots for T. cincinnatus (Abb. 258), but her figures showed too few branchial filaments and no visible eyespots (they cannot be confirmed or observed in the figure shown). Either the specimen in the figure is too young (there is no scale) or she was studying a different species.

(Table 1): 444 specimens from 27 stations, collected at depths from 94–495 m, 1.73 °C–11.3 °C. Holotype st. Sevastopol 2587. Material is deposited at the KGB, three paratypes from Sevastopol st.1453 are deposited at the MNCN 16.01/17772. Material from Aveiro (DBUA0000389.01) and Naples (MNCN 16.01/488) is not included in the type series as it was collected too far away from the type locality, despite seeming to be morphologically identical.

(based on holotype and paratypes). Holotype with 97 segments, 32 segments with notopodia, 95 mm length. Paratypes up to 100 mm long and 5 mm wide; number of segments increased with body size, up to 91.

Several tens of grooved buccal tentacles as long as half body length. Eyespots absent. Branchial filaments numerous, long and tangled (Fig. 4A, C). Due to tangling, it was impossible to count number of branchial filaments in large worms (>5–6 mm width) without removing them one by one. Maximum number of BS1 filaments ca. 20 (13 in holotype), extending laterally to a point level with upper edge of row of U1 uncini. BS2 with ca. ten filaments (nine in holotype). Filaments attached to a transverse elevated stump in 1–2 irregular rows but, due to numerous filaments, there appear to be more rows. Number of filaments increases with body size; small worms (1–2 mm width) with ca. 5 filaments on BS1. Lateral extension of filaments depends upon worm size: in small worms, filaments extend only to a point level with notopodia. Lateral lobes absent. Dorsum with warts or subepithelial honeycomb, forming more or less regular rows (Fig. 4A, C); number of rows increases with size of segments and worms. Segmentation distinct. Nephridial papillae on S5–S7 above neuropodia (Fig. 4C, arrowed), usually poorly visible or not visible; papillae on S4 apparently absent. Ventrum glandular, with “wrinkling” (Fig. 4B) increasing with worm size.

Notopodia from BS2. In small worms, more or less similar, almost cylindrical; in large worms, anterior notopodia transversely flattened, those in first few anterior segments several times smaller than those that are most developed (Fig. 4). Largest specimens in each sample with about 30–40 segments with notopodia, the smallest with fewer, but even specimens ten or more times smaller than largest (by size) with over 30; the next 40–60 segments without notopodia, i.e. about 1/3–1/2 of body length without notopodial segments. Notochaetae with narrow brims (Fig. 11A).

Neuropodia from C3; tori increasing in size to U10, then becoming progressively smaller. Uncini in a single row, uncini of U1 with well-developed prow and crest with one tooth in profile (Fig. 4F, G); posterior uncini (U48) very similar (Fig. 4H).

Pygidium with crenulated margin, without cirri or papillae.

Only one previously known species, T. pascua (Fauchald, 1977) from the Atlantic coast of Panama, has two pairs of branchiae and no eyespots. It differs from T. davehalli sp. n. in its lower number of branchial filaments: single filament in BS1 and BS2 in T. pascua; up to 20 filaments in BS1 and up to 10 filaments in T. davehalli. Only one previously known species, T. hamatus Moore, 1905 from Pacific Alaska, has two pairs of branchiae and segments of the posterior half of the body without notopodia. It differs from T. davehalli in the presence of eyespots and a lower number of branchial filaments: five in BS1 and BS2 in T. hamatus; up to 20 filaments in BS1 and up to 10 filaments in T. davehalli. Thelepus davehalli differs from the other species described in this paper and other known species with two pairs of branchiae in the presence of fully developed segments without notopodia in the posterior 1/3–1/2 of the body.

The last biramous parapodia of Thelepus davehalli is well developed (not reduced), following uniramous parapodia with well-developed neuropodia, contrary to other species described in this study (Fig. 4D, E). Anterior segments lack well-developed notopodia, contrary to those in T. cincinnatus and T. marthae.

Thelepus cincinnatus var. andreanae McIntosh, 1922 was described from within the range of T. davehalli. However, McIntosh clearly stated “Dorsal cephalic collar with eye-specks” while this new species has no eyespots.

The species is named after my friend Mr. David Hall, Head of Marine and Freshwater Laboratories, Associate Director APEM Ltd., UK (Fig. 5).

(Table 1): 921 specimens from 38 stations collected from depths between 95–1,510 m, bottom temperature -1.84–2.8 °C. Holotype: R/V Tunetz cruise 105 station 6. Material is deposited at the KGB, fifteen paratypes from Alaid st. 6 are deposited at MNCN 16.01/17773, seven paratypes are deposited at ZIN 1/33266.

(based on holotype and paratypes). Holotype with 81 segments, 55 segments with notopodia, 55 mm length. Paratypes up to 80 mm in length, 6–7 mm in width, 100 segments, last segments still in formation and clustered, not fully developed, with poorly-developed neuropodia, so not possible to count total number of segments.

Several tens of buccal tentacles, their length in fixed specimens equal to half of body length. Eyespots absent (Fig. 6A–C). BS1 with up to ten filaments (seven in holotype); BS2 with up to five (four in holotype) (Fig. 6F, H). Number of filaments increases with worm size; smallest worms, width <1 mm, with either no branchiae or with 1–2 filaments on BS1 and none on BS2. However, maximum number of filaments constant in different samples (containing sufficient worms) despite a range of maximum worm sizes across the samples. For example, largest worms from sample SP-22 st. 60 are at least three times larger than those from sample Alaid st. 3, but maximum number of filaments observed is same. Branchial filaments of BS1 extend laterally from level of notopodia of C1, to a maximum level with upper margin of uncinal row of U1. Filaments attached in a single row on an elevated stump. A wide medial gap separates left and right groups of filaments. Lateral lobes absent. Barely visible nephridial papillae on S4–S7 above neuropodia (Fig. 6A, B arrowed), in most specimens, few papillae visible, usually none. Ventrum glandular, with “wrinkling” (Fig. 4B) increasing with worm size (Fig. 6G).

Notopodia from S3, anterior notopodia almost cylindrical. Notopodia on C1, often C2, and sometimes C3 two to three times smaller than most developed notopodia (app. C10), sometimes one notopodium on C1 absent (Sevastopol 1358). Most developed notopodia transversally flattened, then reduced in size and become cylindrical again. In the most posterior segments notopodia very small; notochaetae present but several times shorter than most developed ones with no more than 10 per ramus; neuropodia also reduced to small pinnuli with few uncini. Notochaetae absent in 20–40 developing segments near pygidium (Fig. 6E); exact number difficult to determine as both annulation and neuropodia poorly developed. Some specimens also without notopodia on the 10–20 preceding reasonably well-developed segments. Number of segments with notopodia around 60 (in few complete worms available for this species), with several posterior segments lacking notopodia. However, segments without notopodia form only ca. 10% of the total worm length. Notochaetae of anterior segments two to three times longer than notochaetae of posterior segments. Notochaetae in two transverse rows: anterior row with short chaetae, distal half (winged part) becomes stained with methylene blue, posterior row with long chaetae. Notochaetae with narrow brims (Fig. 11C).

Neuropodia from C3; tori increasing in size to U10, then becoming progressively slightly shorter. Uncini in single row. Uncini of U1 with well-developed prow and crest with one tooth in profile (Fig. 7A, C–E), posterior uncini (U20 from pygidium) very similar (Fig. 7B).

Pygidium with crenulated margin without cirri or papillae (Fig. 6E).

Only one previously known species, T. pascua (Fauchald, 1977) from the Caribbean coast of Panama, has two pairs of branchiae and no eyespots. It differs from T. marthae in the lower number of branchial filaments: single filament in BS1 and BS2 in T. pascua; up to 20 filaments in BS1 and up to 10 filaments in T. marthae. Thelepus marthae differs from T. davehalli (described above) in the typically observed absence of fully developed segments without notopodia; if present, they form no more than 10% of the body length. Thelepus marthae differs from T. crassibranchiatus Treadwell, 1901, T. hamatus Moore, 1905 and T. pascua (Fauchald, 1977) (which have eyespots) in the higher number of branchial filaments and segments with notopodia. Thelepus marthae differs from T. cincinnatus and T. antarcticus in the lower number of branchial filaments and segments with notopodia. Thelepus marthae differs from T. parapari in the shape of its uncini.

One specimen (SP-22 st. 72) has numerous spots (Fig. 6D); together forming a transverse row, as with typical eyespots but, in this case, each individual spot is longitudinal instead of rounded as in T. cincinnatus (Fig. 2B) and other Terebellidae. These spots are in the same place as eyespots, but their very unusual shape makes their interpretation as eyespots doubtful; other interpretations are possible.

Species is named after my friend Dr. Martha K. Leontovich (Fig. 8); she has described several new terebellid species.

(Table 1): 177 specimens from 11 stations collected 26.03.1986 between rhizomes of Posidonia, coralligenous formations, calcareous concretions and under stones, 2–15 m, Andalusia, Spain. Holotype MNCN 16.01/17774 (previously part of MNCN 16.01/5706), 5 paratypes previously deposited in MNCN 16.01/5706 now deposited in KGB.

(based on holotype and paratypes). Holotype with 58 segments, 50 of them with notopodia, 50 mm length. Paratypes up to 60 mm in length, 2 mm in width, 60–70 segments, posterior segments clustered and developing with poorly-developed neuropodia, so not possible to count total number of segments.

Several tens of buccal tentacles, their length in fixed specimens equal to half of body. Eyespots absent in most specimens (Fig. 9A), only some with reddish eyespots forming a band without dorsal gap (Fig. 9B). Eyespots probably fade during preservation or variation in this character. Preserved body uniformly beige to yellowish, without distinct patterns of pigmentation; one specimen with eyespots with reddish spots around branchiae. BS1 with up to 12 filaments (11 in holotype); BS2 with slightly fewer filaments (eight in holotype; generally, >70% number on BS1). Filaments thin and very long, reaching more than half of corresponding segment’s width (Fig. 9A, C). Number of filaments increases as worm grows; smallest observed worms (width <0.5 mm) with 1–2 filaments on BS1 and one on BS2. Branchial filaments of BS1 attach in an irregular row on a slightly elevated stump and extend laterally to a point level with notopodia of C1 or sometimes level with upper margin of uncinal row of C3. Filaments of BS2 do not reach notopodia and usually form two rows. A wide medial gap separates left and right groups of filaments. Lateral lobes absent.

Notopodia commence from S3, almost cylindrical anteriorly; those from C1 onwards of equal size. Posterior notopodia poorly developed (almost no rami), with few notochaetae that are several times shorter than most developed notochaetae; neuropodia also reduced. Notochaetae absent only in developing segments near pygidium, approximately ten such segments, exact number difficult to determine as both annulation and neuropodia poorly developed (Fig. 9D). Characteristic number of segments with notopodia less than 60 (based on few available complete worms). Segments without notopodia from only ca. 5% of total worm length. Relatively distinct (in comparison with species described above), small nephridial papillae on S4–S7, above neuropodia (Fig. 9A). Ventrum glandular, without distinct pads (Fig. 9A, E).

Notochaetae sometimes form two distinct transverse rows: anterior row with short chaetae, posterior row with longer chaetae, distal half (winged part) becomes stained with methylene blue, but usually in one row with mixed short and long chaetae; flanges appear to be wider than in species described above (Fig. 11D).

Neuropodia from C3, tori. Uncini in a single row. Uncini of U1 with two teeth in profile above main fang, unlike three species described above (Fig. 9F). However, posteriorly, uncini have only one tooth in profile, in common with species described above (Fig. 9G).

Pygidium with crenulated margin, without cirri or papillae.

Only one previously known species, T. pascua (Fauchald, 1977), from the Caribbean coast of Panama has two pairs of branchiae and no eyespots. It differs from T. parapari in the lower number of branchial filaments: single filament in BS1 and BS2 in T. pascua; up to 11 filaments in BS1 and up to 8 filaments in T. parapari. Thelepus parapari differs from T. davehalli (described above) in the absence of fully-developed segments without notopodia. Thelepus parapari differs from T. crassibranchiatus Treadwell, 1901, T. hamatus Moore, 1905 and T. pascua (Fauchald, 1977) (all of which have eyespots) in the higher number of branchial filaments and segments with notopodia. Thelepus parapari differs from T. cincinnatus and T. antarcticus in the lower number of branchial filaments and segments with notopodia. Thelepus parapari differs from T. cincinnatus and T. marthae (described above) in the shape of the uncini of U1. Thelepus nucleolata (Claparède, 1870), as Heterophenacia nucleolata, was described from nearby (Gulf of Naples), but T. parapari has uncini in a single row, whilst in T. nucleolata they form two rows.

Species is named after my friend Dr. Julio Parapar, Universidade da Coruña, Spain (Fig. 10).