Three, two, one! Revision of the long-bodied sphaerodorids (Sphaerodoridae, Annelida) and synonymization of Ephesiella, Ephesiopsis and Sphaerodorum

DNA extraction, PCR amplification and sequencing

DNA was extracted with QuickExtract DNA Extraction (Epicentre, Madison, WI, USA); a small piece, usually one or two parapodia, were put in 50–100 µl QuickExtract, and treated with 65 °C for 45 min followed by 2 min in 95 °C in a dry block thermostat. We used the primers 16SANNF (GCGGTATCCTGACCGTRCWAAGGTA) (Sjölin, Erseus & Källersjo, 2005) or 16SARL (CGCCTGTTTATCAAAAACAT), together with 16SBRH (CCGGTCTGAACTCAGATCACGT) (Palumbi, 1996) for 16S rDNA; LCO1490 (GGTCAACAAATCATAAAGATATTGG) and HCO2198 (TAAACTTCAGGGTGACCAAAAAATCA) (Folmer et al., 1994) for COI; 28SC1 (ACCCGCTGAATTTAAGCAT) and 28SD2 (TCCGTGTTTCAAGACGG) (Lê, Lecointre & Perasso, 1993) for 28S rDNA (D1–D2 region); and 18SAL (AACCTGGTTGATCCTGCCAGT and CCAACTACGAGCTTTTTAACTG), 18SBO (TGATCCTTCCGCAGGTTCACCT and AAGGGCACCACCAGGAGTGGAG), and 18SCY (CGGTAATTCCAGCTCCAATAG and CAGACAAATCGCTCCACCAAC) (Apakupakul, Siddall & Burreson, 1999), amplifying three overlapping fragments, for 18S rDNA. PCR mixtures contained 0.33 µl of each primer (10 µM), 1 µl of DNA template, and 10 µl of RedTaq 1.1 × MasterMix 2.0 mM MgCl2 (VWR). Temperature profile was as follows: 96 °C/1 min–(95 °C/30 s–52 °C (for COI, 16S rDNA, and 18S rDNA) or 60 °C (for 28S rDNA)/30s–72 °C/60 s) × 29 cycles–72 °C/7 min. PCR products were visualized with UV-light (312 nm) following electrophoresis for c. 15 min on a 1% agarose gel (1 g Agarose DNA Pure Grade (VWR) in a TAE Buffer Ultra Pure Grade (Amresco, Dallas, TX, USA)) containing 1 µl GelRed Nuclear Acid Stain (Bioticum, El Monte, CA, USA) in 50 ml agarose. Each PCR product was purified with 2 µl cleaning solution made from 500 µl mQ-H20, 40 µl FastAP (EF0651), 45µl Buffer FastAP, and 20 µl Exonuclease (EN0581) (Thermo Scientific, Waltham, MA, USA). PCR products with added cleaning solution were run for 37 °C in 60 min, followed by 75 °C in 15 min. Sequencing was performed at Eurofins Genomics, DNA Sequencing Department in Ebersberg, Germany.

Overlapping sequence fragments were merged into consensus sequences using Geneious version 7.0.6 available from Geneious 9 (http://www.geneious.com/). We used MAFFT v7.017 (Katoh et al., 2002) within Geneious 7.0.6 with the following settings: algorithm = E-INS-i, scoring matrix = 200PAM/k = 2, gap open penalty = 1.53 to align the sequences. We used the online GBlocks server v. 0.91b (Castresana, 2002), using the options ‘Allow gap positions within the final blocks’ and ‘Allow less strict flanking positions’, to detect alignment-ambiguous sites (Talavera & Castresana, 2007). Analyses were performed both with and without these alignment-ambiguous sites. Gene partitions were concatenated using Mesquite v. 2.75 (Maddison & Maddison, 2008). Information about the specimens used in DNA sequencing, vouchers and accession numbers can be found in Table 1.

Phylogenetic analyses

The mitochondrial (COI and 16S rDNA and nuclear data sets (18S rDNA and 28S rDNA) were analysed separately and combined using Bayesian inference (BA), and Maximum Likelihood (ML). Bayesian analyses (BAs) of separate and combined data sets were run in MrBayes 3.2 (Ronquist & Huelsenbeck, 2003), and the best-fit models were selected using the Akaike information criterion in JModel (Darriba et al., 2012). The protein coding gene COI was further divided into two partitions, one with the first and second positions, and one with the third positions. The selected best-fit models were a general time reversible model with gamma distributed rate across sites and a proportion of the sites invariable (GTR+I+G) for the COI-partition with first and second positions, 16S rDNA, 18S rDNA, and 28S rDNA, while a Hasegawa, Kishino and Yano model, with gamma distributed rate across sites (HKY+G) was selected for the COI-partition with third positions.

Partitions were unlinked for the parameters statefreq, revmat, shape and pinvar. Rateprior for the partition rate multiplier was set to be variable. Number of generations was set to three million, with four parallel chains (three hot, one cold), sample frequency was set to 1,000, and number of runs set to two. One fourth of the samples were discarded as burn-ins. Maximum likelihood analyses (MLs) were performed in raxmlGUI (Silvestro & Michalak, 2012). In RAxML, the analyses were run with the GTRGAMMAI model, the combined data set was partitioned as in BA, and clade support was assessed using 1000 bootstrap replicates.

Eyes (Figs. 4A–4E, Supplemental Information).

The cerebral eyes in sphaerodorids are attached to the brain and observed beneath the epithelium when this is translucent (Figs. 4A, 4C– 4E, Capa, Bakken & Purschke, 2014). They are often pushed back from the prostomium to as far as the second or third chaetiger (Fig. 4C; Capa, Bakken & Purschke, 2014). They are generally crescent-shaped (semicolon-shaped) spots (Figs. 4C–4E) and have been reported as absent or present in numbers of two, four or more, and colour varying from brown to red (in live specimens) and this variation has been considered as species specific (e.g., Fauchald, 1974; Mòllica, 1994; Hartmann-Schröder & Rosenfeldt, 1988; Imajima, 2003). Eyes have been observed even in individuals collected in hundreds of meters deep (e.g., in Ephesiella abyssorum (Hansen, 1879) from Norway, M Capa, pers. obs., 2016). In fixed, opaque specimens, eyes are often not conspicuous or seem to be absent (e.g., Fig. 4B) and variation to this condition has been observed among individuals of the same population and after longer preservation periods (e.g., when revisiting type material, Supplemental Information). The taxonomic value of presence and absence of eyes, colour, and arrangement should be assessed in live material when the epithelium is translucent. It is obvious that they are not always visible in fixed specimens and that the eye pigment could fade after some time as of all the types with reported eyes revisited had no obvious eyes, except for E. muhlenhardtae Hartmann-Schröder & Rosenfeldt, 1988. Species that were described as lacking eyes include Ephesiella australiensis and Ephesiopsis shivae (Rizzo, 2009), yet in other descriptions the absence or presence of eyes was not mentioned Supplemental Information.

Hooks in anterior chaetigers (Figs. 4F–4K, Supplemental Information).

The function of hooks, anterior stout and curved simple chaetae, is unknown. However, its absence or presence has been one of the main diagnostic features for species discrimination (e.g., Fauchald, 1974; Mòllica, 1994; Hartmann-Schröder, 1982; Imajima, 2003). These structures are often drawn-out of the parapodia and are visible from the lateral or ventral side of individuals (Figs. 4F–4K) but they are sometimes withdrawn, and difficult to observe in fixed and opaque specimens (e.g., Bakken, 2002). Nevertheless, some translucent individuals, or specimens where parapodia have been dissected, seem to lack these chaetae, and this is not related to the gender or size of individuals (M Capa, pers. obs., 2016). The only known specimen of Ephesiella mixta (Hartman & Fauchald, 1971), with nine chaetigers, and the holotype of Sphaerodorum ophiurophoretos Martín & Alvà, 1988, with eight chaetigers bear hooks in the first chaetiger (Fig. 4F). Similarly, adult females and males of what has traditionally been considered E. abyssorum and S. flavum have been observed with and without this type of simple chaetae, so a sexual dimorphism is discarded. Further investigations should verify if the presence and arrangement of hooks in anterior chaetigers are attributes able to discriminate between lineages of this supposed complex of species (E. abyssorum + S. flavum; as shown in Fig. 2).

Hooks can also vary in number. Some specimens bear the typical pair, present in the first chaetiger, while others have another one or two sets of hooks in same parapodia, and the number can vary between the left and right parapodia (e.g., Bakken, 2002). The position of these hooks are generally attributed to the first chaetiger where records of other chaetae have not been reported in their absence or in addition to them. This is the most common arrangement (Figs. 3C, 3E, 3F, 4F–4K). However, a couple of individuals are herein reported bearing hooks in the first two chaetigers (Fig. 4G). The taxonomic importance of this variation is unknown, but may not be species specific.

Some specimens of different putative species bear additional chaetae in the first chaetiger in addition to the hooks, described herein for the first time. These chaetae are not necessarily compound even in the typical “Ephesiella” species but mainly look pseudo-compound as they have only a groove indicating a possible fusion of shaft and blade (Figs. 4L–4O).

Relative position of macro- and microtubercles (Fig. 5, Supplemental Information).

The presence of dorsal macro- and microtubercles is characteristic of all members of the three genera of long-bodied sphaerodorids but variation within its relative position has been considered of taxonomic importance at species level. Ephesiella mammifera is the only species described with microtubercles partially fused to the dorsal edge of the macrotubercles (Fauchald, 1974). This condition was only verified in some segments of the paratypes (Fig. 5B), but not in the holotype which seemed to present macro- and microtubercles separated by a small gap. Moreover, some individuals belonging to other species also presented very close macro and microtubercles, or in contact (Figs. 5B–5G, 5I, 5L). This is the case for the holotype of S. indutum (Fig. 5I), and specimens of Ephesiella cf. oculata from Japan (Fig. 5C), Ephesiella spp. from Papua New Guinea and Brazil (Figs. 5D and 5F, respectively), E. phuketensis (sensu Bakken, 2002) and S. flavum from Norway (Fig. 5L), among others (see more examples in Figs. 5A–5L). Since specimens with this condition (macro- and microtubercles in close proximity) showed also some variation along the body, it is probable that the gap between macro and microtubercles can vary with the contraction of the tegument or the turgor of the macrotubercle (probably due to the amount of glandular content or preservation method).

In most of the specimens examined, the first chaetiger lacks microtubercles and has an undersized macrotubercle (Figs. 5N–5P). This is not the case for the holotype of E. mixta and E. muhlenhardtae and the paratype of E. australiensis (Capa & Bakken, 2015) that have both types of tubercles in the first segment (bearing the hooks). The paratypes of the latter were in poor condition and almost disintegrated when examined and this feature, that could be of taxonomic value, was not corroborated, but should be considered and tested in future taxonomic revisions. Differences in the absence/presence of microtubercles on the first chaetiger have been observed even within the type material of E. australiensis (the holotype lacks microtubercles on first chaetiger), so it could well be a character showing intraspecific phenetic variation.

Shape and size of parapodia (Figs. 6A–6K)

The size and shape of parapodia can fluctuate considerably with contraction of parapodial muscles. Parapodial retractor muscles together with chaetal flexor muscles and acicular retractor muscles fill the whole parapodia (Helm & Capa, 2015) and are responsible for their size and shape. Differences of four times in length have been measured within the same fixed specimen, observation that has been verified in live specimens. Consequently, it does not seem appropriate to use the shape and length of these appendages for species diagnoses and probably not referring to its length when comparing relative length of other structures (i.e., ventral cirri, lobes, papillae, etc.).

One of the differences justifying the description of Sphaerodorum papillifer Moore, 1909 was its large parapodia with parallel sides (Moore, 1909; Fauchald, 1974), a feature with questionable taxonomic utility.

Ventral cirri (Figs. 6A–6K)

The shape, position and the length of the ventral cirri in relation to the acicular lobe have been considered diagnostic features for some long-bodied sphaerodorid species (e.g., Fauchald, 1974). Most descriptions refer to the ventral cirri as digitiform (Fauchald, 1974; Rizzo, 2009), basally inflated (Kudenov, 1987), with a distal protuberance (Moreira & Parapar, 2011) or with a distal articulation (as in Ephesiopsis guayanae (Hartman & Fauchald, 1971)) but direct re-examination of specimens from different species revealed that in most cases the ventral cirri are simple (without a distal articulation), with a wider base and with a distally thinner end resembling a bottle or a bowling pin (Figs. 6A, 6C, 6F–6H). Some exceptions have been found in small specimens or in anterior body segments of longer specimens, where the cirri are simply digitiform (cylindrical with a rounded tip; e.g., Figs. 6D–6J). In the single known specimen of E. gallardoi (Fauchald, 1974), cirri are distinctly conical (not mentioned in the original description), a feature which taxonomic value should be corroborated when intraspecific variability can be checked.

Ventral cirri slightly projecting beyond the parapodia is the general rule (E. antarctica, E. brevicapitis and E. mixta, e.g., Figs. 6D, 6F, 6G) except in anterior segments when it is usually shorter (e.g., Fig. 6C). However, a few species have been described as having shorter ventral cirri (e.g., Moore, 1909; Hartman & Fauchald, 1971; Fauchald, 1972; Fauchald, 1974). This feature may not be relevant for species discrimination since some intraspecific variation has also be noted and could be attributed to the degree of contraction of parapodia. The insertion of the cirri on the parapodia could be a valid character for distinguishing species, but again, its display could be affected by the level of contraction of parapodia. Revision of the types of E. mixta and E. gallardoi confirmed that the ventral cirri are inserted in the middle of the parapodia while in the rest of the species of Ephesiella, Ephesiopsis and Sphaerodorum ventral cirri have a more distal position (Figs. 6A, 6C, 6D, 6F–6H, 6J; Hartman & Fauchald, 1971; Fauchald, 1974; Rizzo, 2009; Moreira & Parapar, 2011).

Parapodial lobes (Figs. 6A–6K)

The sphaerodorid descriptions often refer to parapodial lobes, conceived as parapodial appendages that are longer than papillae, absent or present. The most common terms found in the literature are acicular lobe (not often described but assumed to be present in all specimens as it is housing the tip of the acicula), pre- and postchaetal lobes (implicitly located anterior and posterior to the chaetae, respectively) and dorsal chaetal lobe (implicitly located dorsally) (e.g., Fauchald, 1974). Several issues emerge from this classification of parapodial lobes. Chaetae are, in members of the long-bodied sphaerodorids, not arranged in a well-defined transverse row but in randomly arranged groups and interspersed with the almost indistinguishable lobes and distal papillae (Figs. 6C–6K), so the terms pre- and postchaetal lobes are here considered imprecise. This could explain why the description of specimens from same species are not always congruent in this matter. The tip of the acicula is sometimes difficult to spot and some authors may have referred to it as a terminal papillae or dorsal chaetal lobe (e.g., Fauchald, 1974; Moreira & Parapar, 2011). Some specimens clearly present a longer papilla on their dorsal distal end (e.g., Fig. 6E), but this is not always recognisable in all segments.

According to Fauchald (1974), sphaerodorid parapodial lobes differ histologically from the parapodial papillae, but at least in the long-bodied forms the papillae and lobes are difficult to distinguish externally (e.g., Figs. 6A–6K) and with the exception of the tip of the parapodia (i.e., the acicular lobe), that show strong neuronal innervation, indicating a sensory function, the nature and internal structure of other parapodial structures seem to be similar in all cases. All epithelial papillae (excluding macro- and microtubercles) seem to have a sensory function as they are provided with distinct ciliation, pores and lack glandular content (Helm & Capa, 2015). Since the lobes are difficult to tell apart from papillae in long bodied sphaerodorids, and do not seem to be recognised even after histological sections, its use is discouraged.

Parapodial papillae (Figs. 6A–6K, 7)

The number and arrangement of parapodial papillae are features often used as diagnostic in sphaerodorid species identification. Nevertheless, a few descriptions indicate the intraspecific variability and it is not known if number or shape change with age and size or if papillae are retractile (as mentioned by Moore, 1909). The terminology used to refer to the arrangement of these papillae in the parapodia is ambiguous. On one hand, descriptions refer to the number of papillae per side of parapodium (a conical or cylindrical structure) and not necessarily refer to ‘side’ in the same way, making comparisons between descriptions difficult. Some authors refer to dorsal and ventral sides, others to anterior and posterior, other authors divide the parapodia in four sides and most of the descriptions do not make clear which sides they are considering. On the other hand, it has been shown that parapodia are highly contractile so the display of papillae varies between a retracted parapodium with most papillae appearing distally grouped (e.g., Figs. 6H, 6J, 6K) and a relaxed parapodium with more separated papillae (e.g., Fig. 6I).

The range of variation in the number of parapodial papillae reported within the group varies from none to 20, but most species records indicate 4 −12. The only species considered as lacking papillae is S. recurvatum, but the poor condition of the only known specimen, the holotype, could be the explanation for such interpretation (Fig. 7V). The specimens with larger number of parapodial papillae are generally large. This is the case for types of S. indutum (Fig. 7J), S. olgae (Fig. 7P), S. papillifer (Fig. 7S) and some of the specimens of E. abyssorum and S. flavum, measuring over 10 mm, with around or over 10 parapodial papillae. Contrary, the smaller forms have lower number of papillae and this is the case for the types of E. mixta (Figs. 6A, 6M) and E. australiensis, measuring less than 2 mm and with only 2 −3 parapodial papillae. An exception to this pattern is E. pallida, with specimens measuring over 10 mm reported and with around four parapodial papillae (Fig. 7R). Species with high number of specimens available have shown some interspecific variability (e.g., E. abyssorum and S. flavum, Figs. 6D–6H) but these nominal taxa also show a broad bathymetrical and geographic distribution range and could well be a complex of species.

The position and arrangement of the parapodial papillae, could be one of the morphological features able to discriminate between species. Some variation between specimens from same putative species and collected in same localities has been observed and therefore it would still need to be further investigated. We encourage the use of a unanimous methodology and terminology to refer to the parapodial sides in order to minimise subjectivity (e.g., Fig. 7, as proposed by Capa & Bakken, 2015).

Chaetae (Figs. 8A–8V, Supplemental Information).

Not only does the presence of simple and/or compound chaetae seem not to characterise monophyletic groups (Fig. 2) but also there is no clear boundary between what should be defined as simple or compound chaetae. The shaft and blade of some Ephesiella species are not clearly divided (e.g., E. brevicapitis; Moore, 1909) and in some species of typically simple compound chaetae, a groove has been noted indicating a possible fusion of blade and shaft (e.g., S. ophiurophoretos and S. olgae; Martín & Alvà, 1988; Moreira & Parapar, 2011), and is also herein reported from specimens of S. flavum (Figs. 8U, 8V).

Chaetal morphology has been utilised to distinguish species. The most common traits include the relative length/width of the blades (in compound chaetae) or distal tip (in simple chaetae), the curvature of the distal end, or the presence of a tooth or knob at the distal edge of the shaft (e.g., Capa & Bakken, 2015). Intraspecific variability in most cases has not been assessed and direct examination of material has shown that a broad variation in chaetal morphology can be observed from anterior to posterior segments with a general trend that distal tips shorten and become more curved towards posterior chaetigers (e.g., Figs. 8M–8O).

Ephesiella species described with almost straight chaetae in all chaetigers are typically E. bipapillata, E. macrocirris, E. mixta and E. muhlenhardtae (e.g., Fig. 8D). The types revisited indeed showed no curved chaetae in any segment of any of these species, and blades are also proportionally longer than in other congeners, reaching up to five times the length of the oblique proximal edge, with the exception of E. macrocirris, where chaetae are up to four times the length of the proximal edge. Species with only strongly curved and short blades are E. australiensis and E. mammifera (Fig. 8C; Fauchald, 1974; Hartmann-Schröder, 1982; Capa & Bakken, 2015) with lengths of blades shorter than 2.5× the oblique base of the blades. Detailed study of specimens from the Mediterranean and the Norwegian Sea, potentially E. cantonei and E. abyssorum respectively, showed curved and short chaetae, but some also were longer and straighter in anterior chaetigers (Figs.  8B, 8M–8O).

Serration on one of the distal edges is often fine and difficult to assess under the compound microscope but variation has been observed within specimens under SEM (Figs. 8M–8O). Contrary to expectations, the posterior-most chaetae are often smooth, even though annelids add segments posteriorly and these chaetae should be newer and less eroded (e.g., Figs. 8M–8O).

Ephesiopsis guayanae has both compound and simple chaetae in every segment (Hartman & Fauchald, 1971; Fauchald, 1974) a condition verified in both holo- and paratype (Capa, Osborn & Bakken, 2016; Fig. 8G). The compound chaetae (two per parapodia in mid-body segments) has a curved blade while the simple chaetae (three per parapodia) has a straight and tapering edges tip. Simple chaetae are distally flattened and resemble those present in S. vietnamense, according to the original drawings as Sphaerodorum sp. A. (Gallardo, 1968). Some simple chaetae of E. guayanae seem to be the result of a fusion of the shaft and blade, with the edge between them still marked as a faint ridge (Capa, Osborn & Bakken, 2016).

Variation in the type of simple chaetae among members of Sphaerodorum has also been noted. Some species, such as S. papillifer, S. recurvatum and E. vietnamense were described to have a “spur in the subdistal swelling boss” (Fauchald, 1974) but after looking at this ‘spur’ under the SEM it turned out to be the distal serration of the shaft, and this feature is also present (but not always) in specimens identified as S. flavum (Fig. 8T). Ephesiella recurvatum was described to have strong hooks in first segment while in S. olgae these can be almost straight (Fauchald, 1974; Moreira & Parapar, 2011).

The number of chaetae in long-bodied sphaerodorids is fairly constant and most specimens present 3–6 chaetae per parapodium (Figs. 6C–6K, 7E–7J; Moore, 1909).

Presence of embryos

The reproductive mode of long-bodied sphaerodorids is unknown. References of Ephesiella mixta being hermaphroditic was made after finding an individual (the holotype), with eggs in anterior, sperm in posterior segments and embryos in middle segments (Fauchald, 1974). Revision of this specimen did not provide evidence about this because sperm is no longer distinguishable. The structures that seem to have been considered embryos could be the segmental ganglia (called perikarya according to Filippova et al., 2010), that are well developed in sphaerodorids (Reimers, 1933; Kuper & Purschke, 2001).

Type species: Sphaerodorum flavum Ørsted, 1843.

Emended diagnosis: Body long and slender with blunt anterior end. Two longitudinal rows of macrotubercles over dorsum, one pair per segment, above parapodia. Macrotubercles sessile, with terminal papillae. Two longitudinal rows of microtubercles, with a collar and a terminal papilla, one pair per segment, running parallel between macrotubercles. Additionally, papillae arranged in 4–5 transverse rows on dorsum and ventrum. Head appendages, palps, lateral and median antennae short, spherical or digitiform. Parapodia with simple, compound or pseudocompound chaetae. Simple hooks absent or present on first chaetiger (sometimes also on second or third chaetiger).

Remarks: the emended diagnosis includes the variation of chaetal morphology (including simple, compound or pseudocompound chaetae) previously regarded as distinctive between long-body sphaerodorid genera.

The current nominal species circumscribed in Sphaerodorum, and nomenclatural changes proposed, after this study are the following:

Sphaerodorum abyssorum Hansen, 1879 n. comb.

Type locality: Norway, 63°5′N 3°0′E, 960 m deep.

Sphaerodorum antarctica (McIntosh, 1885) n. comb.

Type locality: Antarctica, 62°26′S, 95°44′E, 3,612 m deep.

Sphaerodorum australiensis (Hartmann-Schröder, 1982) n. comb.

Type locality: Cervantes, Western Australia, intertidal.

Sphaerodorum bipapillatum (Kudenov, 1987) n. comb.

Type locality: Louisiana, Gulf of Mexico, 28° 54′48″N, 89°59′05″W, 33.6 m deep.

Sphaerodorum brevicapitis Moore, 1909 n. comb.

Type locality: California, Eastern Pacific, 3,740 m deep.

Sphaerodorum cantonei (Mòllica, 1994) n. comb.

Type locality: Sicily, Central Mediterranean, 3–6 m deep.

Sphaerodorum flavum Ørsted, 1843

Type locality: Denmark, shallow water, perhaps intertidal.

Sphaerodorum gallardi (Fauchald, 1974) n. comb.

Type locality: South Vietnam, 19 m deep.

Sphaerodorum guayanae (Hartman & Fauchald, 1971) n. comb.

Type locality: Surinam, Western Atlantic, 520–550 m deep.

Sphaerodorum indutum Fauchald, 1974

Type locality: Antarctica, 61°25′S, 56°30′W, 300 m deep.

Sphaerodorum macrocirris (Hartman & Fauchald, 1971) n. comb.

Type locality: New England, Western Atlantic, 1,470–1,330 m deep.

Sphaerodorum mammiferum Fauchald, 1974 n. comb.

Type locality: Baja California, Eastern Pacific, intertidal.

Sphaerodorum mixtum (Hartman & Fauchald, 1971) n. comb.

Type locality: Bermuda, 3,753 m deep.

Sphaerodorum muhlenhartdtae (Hartmann-Schröder & Rosenfeldt, 1988) n. comb.

Type locality: joinville Island, Antarctica, 63°30′S, 54°15′W, 220 m deep.

Sphaerodorum oculatum (Imajima, 2003) n. comb.

Type locality: Johashima, Japan, 100 m deep.

Sphaerodorum olgae Moreira & Parapar, 2011

Type locality: Bellingshausen Sea, Antarctica, 400–1,799 m deep.

Sphaerodorum ophiurophoretos Martín & Alvà, 1988

Type locality: Pas de Calais, English Channel, intertidal.

Sphaerodorum pallidum (Fauchald, 1974) n. comb.

Type locality: South Shetland Islands, Antarctica, 1,437 m deep.

Sphaerodorum papillifer Moore, 1909

Type locality: California, Eastern Pacific, 914 m deep.

Sphaerodorum phuketensis (Bakken, 2002) n. comb.

Type locality: Phuket Island, Andamen Sea, 63 m deep.

Sphaerodorum ramosae (Desbruyères, 1980) n. comb.

Type locality: Plateau de Meriadzek, North East Atlantic, 47°29.2′N 8°30.7′W, 2,156 m deep.

Sphaerodorum recurvatum (Fauchald, 1974)

Type locality: South Africa, Indian Ocean, 29° 45′S, 31°40–39′E, 445 m deep.

Sphaerodorum shivae (Rizzo, 2009) n. comb.

Type locality: Off São Paulo, Western Atlantic, 24°07.637′S 45°51.895′ W, 147 m deep.

Sphaerodorum vietnamense (Fauchald, 1974)

Type locality: South Vietnam, 32 m deep.