Hippasteria, NOZAWAI GOTO, 1914

Mah, Christopher, Neill, Kate, Eléaume, Marc & Foltz, David, 2014, New species and global revision of Hippasteria (Hippasterinae: Goniasteridae; Asteroidea; Echinodermata), Zoological Journal of the Linnean Society 171 (2), pp. 422-456 : 441-448

HIPPASTERIA NOZAWAI GOTO, 1914

Goto, 1914: 344; Mah et al., 2010: 285, 286 (as H. nozawai ).

Type species

Hippasteria nozawai Goto, 1914 , type specimen status unknown.

Comments

Based on the original description in Goto (1914), this species appears to share most similarity with H. leiopelta . Most of the characters, pedicellariae, marginal plate series, marginal plate surface granulation pattern, etc., are very similar with only the overall R: r body shape differing. The R: r ratio in H. nozawai is 2.0 vs. H. leiopelta , which has R: r = 1.5. It seems likely that the two species are closely related and/or possible synonyms. It is tentatively placed with H. leiopelta within the H. phrygiana cluster. No material of this species was available for examination.

Occurrence

Hokkaido, northern Japan (no depth information available).

Material examined

None.

HIPPASTERIA PHRYGIANA ( PARELIUS, 1768)

( FIGS 7A–J , 8A–J )

Linck, 1733: 21 (as Pentaceros planus ).

Parelius, 1768: 425 [1770: 349] (as Asterias phrygiana ).

Lamarck, 1816: 555 [non A. equestris Retzius, 1805 ] (as Asterias equestris ).

Gray in Johnston, 1836: 146 (as Asterias johnstoni ).

Gray, 1840: 279, 1866: 9 (as Hippasteria europaea , Hippasteria johnstoni , Hippasteria cornuta ).

Gray, 1840: 279; 1866: 9; Perrier, 1875: 270 [1876: 65], Danielssen & Koren, 1882: 268; Sladen, 1883: 159; Perrier, 1888: 764; Sladen, 1889: 341; Koehler, 1909: 88, 1924: 179 (as Hippasteria plana ).

Forbes, 1841: 125 (as Goniaster equestris ).

Müller & Troschel, 1842: 52; Düben & Koren, 1846: 246 (as Astrogonium phrygianum ).

Forbes, 1843: 280 (as Goniaster abbensis ).

Barrett, 1857: 47 (as Astrogonium aculeatum ).

Norman, 1865: 128 (as Goniaster phrygiana ).

Dons, 1937: 17 (as Hippasteria [Euhippasteria] phrygiana and Hippasteria [Nehippasteria] insignis ).

Verrill, 1874: 413; 1885: 542, Ganong, 1893: 56; Grieg, 1895: 6; Verrill, 1895: 137, 1899: 148; Döderlein, 1900: 218; Hartlaub, 1900: 191; Ludwig, 1900: 457; Whiteaves, 1901: 50; Grieg, 1902: 21; Pearcey, 1902: 308; Simpson, 1903: 40; H.L. Clark, 1905: 1; Nordgaard, 1905: 160, 235; Grieg, 1907: 28, 32; Süssbach & Breckner, 1911: 215; Grieg, 1912: 6; 1913: 115; 1917: 8; 1921: 6; H.L. Clark, 1923: 270; 1926: 13; Mortensen, 1927: 88, 1932: 8; 1933: 245; Haubold, 1933: 200; Einarsson, 1948: 10; A.H. Clark, 1949: 373; Djakonov, 1950: 53; Blacker, 1957: 18, 45; Buchanan, 1966: 25; Wolff, 1968: 82, Walker, 1978: 361; Franz et al., 1981: 406, 415; O’Connor & Tyndall, 1986: 96; Clark & Downey, 1992: 247; Picton, 1993: 22, 23; Gulliksen et al., 1999: 73, 105; Anisimova & Cochrane, 2003: 117, 121; Moore et al., 2004: 246; Dilman, 2006: 181; Mah et al., 2010: 287; Mah, 2011: 27; Clark & Jewett, 2011: 48; Foltz et al., 2013: 1285 (as Hippasteria phrygiana ).

Bell, 1892: 76 (as Hippasterias phrygiana ).

Perrier, 1891: K128; A.M. Clark, 1962: 22 (key); Codoceo & Andrade, 1978: 156, 168; Branch et al., 1993: 44, 60; Larraín et al., 1999: 437; Stampanato & Jangoux, 2004: 7 (as Hippastera hyadesi ).

Verrill, 1909: 63; Fisher, 1911: 224; Goto, 1914: 349; Verrill, 1914: 301; Djakonov, 1950: 53 (1968: 45); Baranova, 1957: 162; Alton, 1966: 1702; Carey, 1972: 38, 39; Lambert, 1978a: 4, 14; Maluf, 1988: 34, 118; A.M. Clark, 1993: 259; Lambert, 2000: 64; Clark & Jewett, 2011: 53 (in key) (as Hippasteria spinosa ).

Fisher, 1911: 230 (as Hippasteria leiopelta armata ).

Fell, 1958: 11, pl. 1, figs A, G; 1959: 136, fig. 21; 1960: 61, pls. 2,3; 1962: 33; McKnight, 1967: 300; H.E.S. Clark, 1970: 3; A.M. Clark, 1993: 259; Rowe & Gates, 1995: 65; Koslow & Gowlett-Holmes, 1998: 44 (as Hippasteria trojana ).

Fisher, 1911: 226 (as Hippasteria spinosa kurilensis ).

Goto, 1914: 338; Hayashi, 1952: 147; 1973: 6, 15; Imaoka et al., 1990: 50; Mah, 1998a: 67; Fujikura et al., 2008: 272 (as Hippasteria imperialis ).

Clark, 1926: 13; A.M. Clark, 1952: 196; A.M. Clark & Courtman-Stock, 1976: 63 (as Hippasteria phrygiana strongylactis ).

Mortensen, 1933; A.M. Clark, 1952: 196; A.M. Clark & Courtman-Stock, 1976: 63 (as Hippasteria phrygiana capensis ).

Djakonov, 1950 (1968: 47), Gish, 2007; R.N. Clark & Jewett, 2011: 53 (in key) (as Hippasteria kurilensis ).

Djakonov, 1950: 22 (1968: 46) (as Hippasteria derjungini ).

Djakonov, 1950: 54 (1968: 45), 1952: 409 (as Hippasteria mammifera ).

Djakonov, 1950: 54 (1968: 45); 1952: 410 (as Hippasteria pedicellaris ).

Djakonov, 1950: 55 (1968: 47); 1952: 411; Baranova, 1957: 162 (as Hippasteria colossa ).

Bernasconi, 1961: 1 (as Hippasteria argentinensis ).

Djakonov, 1950: 22 (1968: 46) (as Hippasteria spinosa f. armata ).

Clark & Downey, 1992: 248 (as Hippasteria phrygiana argentinensis ).

R.N. Clark & Jewett, 2011:52 (as Hippasteria armata ).

R.N. Clark & Jewett, 2011:48 (as Hippasteria aleutica ).

Type specimens

Asterias phrygiana Parelius, 1768 . Clark & Downey (1992) speculated that the type may be in the MNHN collections, but not located. A search by one of the authors (C. M.) further concluded that the type could not be located in the MNHN collection. Type specimens for recently synonymized names are below. The status of Hippasteria species described by Goto (1914) remains unknown pending further search of Japanese type repositories. Although the British Columbia syntype of H. spinosa is listed below, the Puget Sound syntype has not been located. Its status remains unknown. Djakonov’s Hippasteria specimens are deposited at the Zoological Institute in St. Petersburg Russia (ZI − abbreviation by C.M.).

List of relevant type specimens from recent synonymy: H. aleutica : Holotype: LACM 2008-030.001; Paratypes USNM 1125124, E47392 ; CASIZ 180539; University of Alaska, Fairbanks INV 8140; NOAA Auke Bay Laboratory AB11-002; LACM 1997-221.001, 1997- 154.013, 1999-183.001, R.N. Clark personal collection (ten miscellaneous specimens). Hippasteria argentinensis : holotype, Argentine Museum of Natural Sciences 9069; H. colossa ZI holotype no. 1/27102; Hippasteria derjungi ZI holotype no. 1/27101; H. hyadesi holotype, MNHN EcAh 2840; H. imperialis type status unknown; H. mammifera ZI holotype no. 1/7103; H. pedicellaris ZI holotype no. 1/7104; H. phrygiana capensis : syntypes, Copenhagen Museum, H. phrygiana strongylactis , holotype, South African Museum A22531 , paratypes MCZ 2831, 2864; H. spinosa , syntype, National Museum of Natural Sciences, National Museum of Canada NMCICI900-2933; H. spinosa armata (= H. leiopelta armata ) USNM 27887; H. spinosa kurilensis (= H. kurilensis ), holotype USNM 27885, paratypes USNM 32103, 32106; H. trojana Holotype, Canterbury Museum, Christchurch.

Synonymy

Results from Foltz et al. (2013) listed several previously accepted species as synonyms of H. phrygiana ( Parelius, 1768) , the type name for species within Hippasteria ( Clark, 1993; Mah & Hansson, 2012). Sampled taxa that were shown to be synonyms of H. phrygiana by Foltz et al. (2013) include H. spinosa Verrill, 1909 , H. armata ( Fisher, 1911) , H. kurilensis Fisher, 1911 , H. armata Fisher, 1911 , and H. aleutica Clark & Jewett, 2011 . Foltz et al. (2013) further removed any doubt regarding synonymy of the Patagonian H. hyadesi Perrier, 1891 or the New Zealand H. trojana Fell, 1958 into H. phrygiana .

Several North Pacific Hippasteria species show few if any morphological differences from H. phrygiana and are almost certainly synonyms. In the North Pacific, this includes H. colossa Djakonov, 1950 , H. derjungini Djakonov, 1950 , Hippasteria mammilifera Djakonov, 1950 , and H. pedicellaris Djakonov, 1950 . R.N. Clark & Jewett (2011) raised H. armata to a species without any apparent justification. Inclusion of H. armata in Foltz et al. (2013) showed that it falls within the synonymy of H. phrygiana along with many other previously established North Pacific Hippasteria spp. Based on the size of specimens and comparisons of appropriately sized specimens of H. phrygiana with the H. armata morphotype, these were almost certainly small-sized individuals.

The results of Foltz et al. (2013) are consistent with historical accounts that have shown H. phrygiana to be either very similar to other species or to be nearly identical but separate because of the distance separating individuals. A.M. Clark (1962) indicated that the type locality outlined for H. hyadesi by Perrier (1891) and Koehler (1926) must be mistaken because the specimen was essentially indistinguishable from H. phrygiana (= H. plana ). Based on our identification of Hippasteria from this region there is no longer a reason to discount the type locality as mistaken. Before being synonymized by H.E.S. Clark & McKnight (2001), Fell (1958) described the South Pacific H. trojana and noted its similar appearance and close relationship to the North Atlantic H. phrygiana . The type description of the North Pacific H. spinosa by Verrill (1909: 63) describes it as ‘Very similar in form and size to H. phrygiana of the N. Atlantic...’ with further remarks on close affinity by Fisher (1911). Other synonyms in the Atlantic almost certainly include H. phrygiana argentinensis and H. phrygiana capensis although we did not include specimens from those regions ( Argentina and South Africa) in our analysis.

Accounts that have described subspecies in Hippasteria [e.g. H. phrygiana argentinensis and others outlined in Clark & Downey (1992)] could not be substantiated by consistent character differences (e.g. spine counts with continuous overlap, etc.). Geographical clusters amongst major H. phrygiana haplotypes were arranged for methodological necessity (see Foltz et al., 2013) and did not reflect any significant morphological differentiation amongst species. Sampling from some localities (e.g. Solomon Islands) was anecdotal and suggests more widespread occurrence may be present in poorly sampled areas, which may further elucidate the phylogeny in Figure 1 .

Occurrence and bathymetry

North Pacific: Japan, Aleutian Islands, Alaska, west coast of North America, including Canada, Washington, Oregon, west to President Jackson Seamount, south to Southern California. South Pacific: New Zealand, Solomon Islands, Straits of Magellan-Chile. South Indian: Kerguelen Islands, Crozet, Marion and Bouvet Isles, South Africa. Atlantic: east coast of North America, south to Cape Cod. North-eastern Atlantic−North Sea: Kattegat, northern Scotland and northernmost Ireland, the UK. South Atlantic: off Argentina, west side of South Africa, Lambert’s Bay to Cape Point.

Collectively throughout its range, H. phrygiana occurs widely from subtidal to continental shelf/upper bathyal depths, 10–1405 m. However, note that regional bathymetric occurrence varies. H.E.S. Clark & McKnight (2001) noted that H. phrygiana in New Zealand waters occurs primarily at> 500 m depths, seldom occurring at shallower depths. By contrast, Lambert (2000) reported this species (as H. spinosa ) along the Pacific north-west coast of North America at depths ranging from 10– 512 m. Maluf (1988) listed the depth of H. spinosa as present between 49–1170 m. Clark & Downey (1992) recorded North Atlantic H. phrygiana as being present in as shallow as 20 m (to 457 m in the Baltimore Canyon), but occurring in the South Atlantic and South Indian Ocean at depths to 310 m. Dilman (2006) noted occurrence of H. phrygiana to the Reykjanes Ridge based on an earlier account by Einarsson (1948). Einarsson (1948) was checked but the reference to Hippasteria could not be verified.

Comments on variation

Variation in H. phrygiana is extensive and although we have attempted to synthesize the extent of this variation, the summary presented here may be imprecise where localized populations are concerned. It remains helpful to refer to regional monographs that offer comprehensive, if older accounts, of H. phrygiana . In the North Pacific, Fisher (1911) summarized H. phrygiana populations (as H. spinosa and others) along the Northwest Pacific coast. Clark & McKnight (2001: 54) pre- sented an overview of variation in H. phrygiana between the New Zealand and the Atlantic specimens. Clark & Downey (1992) provided a taxonomic summary of North Atlantic H. phrygiana . Accounts of high latitude members (e.g. as H. hyadesi ) are more scattered and not as comprehensive (e.g. Perrier, 1891; Bernasconi, 1963; Stampanato & Jangoux, 2004).

It was difficult to conclude if character variation observed within H. phrygiana could be definitively associated with any observed environmental parameter (i.e. depth, location). Variable characters across the species’ range, which have historically been used to distinguish species, included accessory granules, spines, and all aspects of pedicellariae, including size, shape, abundance, and distribution.

Some deeper-water individuals (e.g. USNM 1215324 from 1405 m depth) displayed finer and more acutely tipped spination with relatively small body size, whereas others from more distant regions (e.g. IE IE- 2007- 1311 in the Solomon Islands) from comparable depths (722 m) were large with robust spines. Fewer samples are available from these settings, and regional vs. bathymetric variation are likely to be key factors .

An unusual aspect of the observed variation of H. phrygiana across its geographical range was the differing R: r ratios (i.e. the degree to which the arms are more distinct relative to the arms) at comparable sizes for different populations. For example, Figure 7G, H shows two individuals at approximately identical sizes ( R = ∼4.4/4.5). One from the North Atlantic ( USNM E46619 ) shows a much stronger stellate shape vs. one from the Crozet Islands (MNHN EcAs 10683), which has a larger disk and broader arms. Environmental factors affecting growth, such as the amount of nutrients and degree of available calcium carbonate, would presumably be significant factors in affecting ontogenetic variation amongst differing populations within this species.

Description

Body in adults, strongly stellate, disk broad, thick ( Fig. 7A, D ). Arms triangular, distinct, short (at R = ∼ 0.5 cm) to long, triangular (in adults, R > ∼ 2 cm). Inter-radial arcs strongly curved to straight. R: r = 2.0– 1.75 (adults) to 1.0 (small individuals). Large individuals R = ∼15.0 cm.

Abactinal surface convex, surface often tumescent. Individual plates round to oval to irregular in outline. Smaller, more oval- to irregular-sized plates occurring distally adjacent to contact with superomarginal series. Larger, more circular plates occurring more proximally especially near disk. Primary circlet and carinal series variably distinct to indistinct. More evident in New Zealand, Chilean/sub-Antarctic individuals, less clearly evident in North Atlantic and North Pacific individuals. Abactinal plates all bordered by accessory granules, two to 30, which vary in size and shape across different populations. These range from small, beadlike granules (<0.5 mm) or sharp, granular spinelets to larger, quadrate, blocky granules (1–2 mm) sometimes with no apparent regional distinction. For example, some New Zealand specimens show both the blocky granules ( USNM E13768) to small, bead-like granules ( NIWA 38153). More blocky granules are more closely arranged vs. the smaller accessory granules, which are more spread apart. Abactinal plates weakly convex and plate surfaces are bare except for spine or pedicellariae. Exceptionally plates lack both spines and pedicellariae and are completely smooth and bare. Shallow but well-developed fasciolar grooves present around abactinal plates extending to marginal and actinal surfaces. Madreporite convex, polygonal in shape, prominent (5–7 mm in diameter) with distinct grooves.

A large (often about 1.0 mm in diameter and up to 5 mm in length), single spine, conical to cylindrical and more blunt-tipped in shape, is present on a majority of abactinal plates but some variable individuals do show more pedicellariae relative to spines. As with the plates, largest spines are present proximally on the disk, becoming shorter and more tubercle like distally, adjacent to the superomarginal contact. Some New Zealand and Chilean/Southern Hemisphere individuals seem to display longer, sharper abactinal spines compared to their more northern counterparts, which have shorter, blunter spines.

Abactinal pedicellariae bivalve (exceptionally trivalve) vary in size and shape but are abundant on abactinal surface, ranging in shape from large-sized with more elongate/bivalve valves (Atlantic) to more elongate with rounded valves bearing weakly expressed but jagged teeth. Atlantic individuals observed have large, wide (up to 3.0 mm length at R = 10.6 cm) valves ( Fig. 7B ). These individuals with smooth, elongate bivalve pedicellariae with smooth valves present on those plates lacking spines, roughly occurring one to every five plates with a spine. Extensive variation is observed in North Pacific, especially Aleutian Island, individuals, with shorter, more rounded, curved more elongate valves (up to 2.0 mm length at R = 11.3 cm) with some individuals bearing pedicellariae that outnumber spines on abactinal disk plates ( Fig. 7J ). The Solomon Islands specimen ( IE- 2007-1311) has pedicellariae significantly smaller than those observed on the other specimens. South Pacific ( New Zealand) individuals have smaller pedicellariae overall with several bearing more elongate (defined as length of valve> width of valve) valves each bearing two to four weakly jagged teeth. Larger bivalve pedicellariae with smooth valves have also been observed.

Marginal plates broadly quadrate to polygonal in shape (e.g. Fig. 7C, E, I ) but variable in length vs. width, ranging from more elongate to more equidimensional (L = W). Superomarginals correspond 1:1 with inferomarginals most strongly inter-radially, becoming more offset distally. Number of marginals is variable but ranges (measured arm-tip to arm-tip) from 32 to 52. Marginal plate surface is bare in most save for large spines and pedicellariae with the exception of the individual specimen from the Solomon Islands ( IE- 2007-1311), which has several concentric rows of granules covering the marginal plate surface up to the spine base on each plate. However, even in that specimen, the plate surface around immediately around the spine base is devoid of granulation. Both marginal plate (20–60 in number) series have rows of highly variable accessory granules present on the edges around them, similar to those on the abactinal plates. These granules range in shape from small and spiny, bead-like in outline, sharp and jagged to heavier and more blockish in shape ( Fig. 7C, E, I, J ). As with the others, larger, blockier accessory granules occur closer together than those that are smaller, which are distributed further apart. A shallow but welldefined fasciolar groove occurs between marginals and is continuous from the abactinal surface to the actinal surface.

Marginals often have a large spine present on each plate (e.g. Fig. 7A, C, E, I ). North Atlantic and North Pacific specimens tend to be lacking pedicelllariae on the marginal plate surface. However, other populations, such as from New Zealand, have one or two irregularly occurring bivalve pedicellariae with smoothedged valves, set into shallow pits on the plate surface ( Fig. 7E ). Spines vary from one to three (both supero- and inferomarginal series), usually only one or two, and are similar in overall morphology to those present on the abactinal plates ( Fig. 7A, C, E, I ). Spines vary in shape from blunt and cylindrical to sharp and conical. When more than one spine is present, the two or three spines emerge from a raised boss on the plate surface. Marginal plates with multiple spines tend to be found inter-radially. Sometimes, short tubercle-like spines sit adjacent or offset to spines. Marginal spines tend to be arranged in discrete, serial rows forming a spiny fringe around the marginal edge. Both marginal series but especially in inferomarginals, possess small, round granules present on marginal plate surface.

The actinal intermediate surface is broad and well developed, composed of six to eight actinal chevrons in an irregularly triangular pattern ( Fig. 7B ). Plates round (proximally) to oval or irregular (distally adjacent to inferomarginal contact) in shape. Actinal plates covered by both large, prominent pedicellariae (e.g. Figs 7B , 8B, J ) and blunt to conical tubercles or spines with large accessory granules forming boundary around each plate. Size of accessory granules variable across individuals with larger blockier granules forming a mosaic on the actinal surface to spinelets, which form a prickly array on the oral surface. One to several spines or large, pointed granules present on actinal plates, sometimes occupying all plate surface. Some actinal plates bare with neither pedicellariae nor spine/ tubercle or large granule.

Actinal pedicellariae frequently bivalve, with smooth valves, large, occupying up to 85% of plate surface, sometimes bisecting the plate ( Fig. 7B ). Abundance on actinal surface varies but it is common to see a full series of pedicellariae present on actinal plate series adjacent to the adambulacrals from the postoral to the edge of the disk with pedicellariae becoming more irregularly distributed more inter-radially, variably occupying anywhere from 40 to 90% of the actinal plates present on the actinal intermediate surface ( Fig. 7B ).

Furrow spines blunt, thick to flattened and bladelike, numbering one or two, oval in cross-section but exceptionally three or four ( Fig. 7F ). Three furrow spines sometimes present on postoral adambulacral plate. Subambulacral spine(s), thick, number one or two, sometimes replaced by pedicellariae. Adambulacral plate surface bare or covered with small to large granules, often blunt and round to angular and/or spinelet-like in appearance.

Oral plates with four to five blunt furrow spines, oval to flattened to triangular in cross-section. One large, blunt spine pointing into oral opening. Oral plate surface covered by large, blunt, cylindrical spines similar to those in furrow and subambulacral series in addition to smaller angular to rounded granules, six to nine adjacent to the actinal surface. Pedicellariae sometimes occurring in subambulacral positions behind furrow spines.

Widespread accounts indicate that H. phrygiana is a variably light orange to deep shade of orange or red (e.g. Fisher, 1911; Clark & Downey, 1992; Clark & McKnight, 2001) with white to yellow tube feet.

H. phrygiana from the Solomon Islands /Hawaii

Two specimens from the Central/South Pacific are new records. One from the SALOMONBOA 3 expedition ( IE- 2007-1311, Fig. 8 A−C) is the first representative of Hippasteria collected from the Solomon Islands / Papua New Guinea region (722 m). DNA sampled from this specimen places it in the clade containing the widespread species, H. phrygiana , showing closest affinities with the Kerguelen Island, and to a lesser extent the New Zealand, population ( Foltz et al., 2013). A second specimen ( CASIZ 102006) from Hawaii at 396 m (1300 ft) shares several characters with and is very similar to MNHN IE- 2007-1311. However, CASIZ 102006 was fixed in formalin and unavailable for inclusion in our molecular data set.

Both specimens vary morphologically from North Atlantic/North Pacific/South Pacific H. phrygiana specimens in several respects. Abactinal, marginal, and actinal spines appear to be smaller and are more abun- dant, marginal plates have two rows of flattened granules forming perimeter around plate surface, pedicellariae are present on nearly every actinal plate, and the actinal pedicellariae themselves are composed of much narrower, more angular valves ( Fig. 8C ). Actinal granules have more spinelet-like tips and are present on actinal plate surfaces as on the periphery ( Fig. 8B ). In contrast, several characters are consistent with those of H. phrygiana , including the quadrateshaped marginal plates with only one to three (usually only one or two) spines on each plate, a single subambulacral spine, and only one to two (exceptionally three) furrow blunt, furrow spines. Both specimens also possess a large pedicellariae behind the subambulacrals on each adambulacral plate.

‘H. imperialis’ from Japan

Our data support a specimen ( CASIZ 115700) resembling H. imperialis Goto as part of the H. phrygiana clade. Morphology of this taxon shows some distinct differences from most members of H. phrygiana . The most notable distinctions come from the absence of the numerous conical spines on the abactinal surface ( Fig. 8H ). ‘ Hippasteria imperialis ’ surface is covered by swollen, convex bosses or swellings with relatively few spines. Spines are present on the marginal plate surfaces of nearly all the other H. phrygiana populations, but are apparently absent from the marginal plate surface of H. imperialis ( Fig. 8H, I ). Such a distinct difference in morphology suggests ecophenotypic variation, but further study is warranted.

Material examined

North Pacific. USNM E47392 H. aleutica paratype off north coast Kiska Island , Aleutians. 52°04′17″N, 177°15′20″E, 91.0 m. Coll. R . Clark aboard FV (Fishing Vessel) Dominator, 31.vii.1997 (one dry spec. R = ∼5.2, r = 2.2); CASIZ 150478 ID as H. kurilensis Aleutian Islands , 52°22′46.56″N, 170°39′39.96″W to 52°22′1.1274″N, 170°39′2.52″W 278 m. Coll. M. Airey aboard F V Vesteraalen 31.v.2000, NMFS Aleutian Survey 2000 (one dry spec. R = 6.3, r = 3.5); CASIZ 141112 ID as H. spinosa Aleutian Islands 52°4′32.0514″N, 176°21′17.6394″W to 52°4′55.452″N, 176°20′15.36″W, 149 m. Coll. M. Airey aboard F V Vesteraalen 31.v.2000, NMFS Aleutian Survey 2000 (one dry spec. R = 4.9, 4 = 3.5, arms upturned); CASIZ 117923 ID as H. spinosa . Aleutian Islands 52°54.9624′N, 169°52.056′W, 106 m, coll. B. Frye aboard F V Vesteraalen , 16.v.1999 (one dry spec. R = 8.2, r = 4.5); CASIZ 101829 ID as H. spinosa off Shelter Cove, Humboldt County, CA 39°53′N, 124°07′W, 183 m (100 fms). Coll. R . Bolin , 5.ix.1935 (one dry spec. R = 6.5, r = 4.9); CASIZ 101830 as H. spinosa off Shelter Cove, Humboldt County, CA 39°53′N, 124°07′W, 183 m (100 fms). Coll. R . Bolin , 5.ix.1935 (one dry spec. R = 10.6, r = 6.0); CASIZ 107744 . Id as H. spinosa Hein Bank, Strait of Juan de Fuca , Puget Sound, Washington, 48°19′N, 123°5′W, 45.7 m (25 fms). Coll. E. Swan, 6.viii.1951 (one dry spec. R = 15.6. r = 7.3); USNM 1215324 President Jackson Seamount C, North Pacific. 42°44′50.4″N, 128°5′25.5″W (42.747347, −128.09044), 1405.8 m, coll. MBARI D82 - A21 (one wet spec. R = 3.9, r = 2.1) ; CASIZ 115674 Id as H. spinosa . San Pedro Channel , off Avalon , Santa Catalina Island, Channel Islands 33°23′N, 118°20′W, 146 m (80 fms). Coll. Myers et al. on the Zaca, 17.ix.1938 (one wet spec. R = 4.0, r = 2.5); CASIZ 115695 Id as H. spinosa . Off Santa Catalina Island, Channel Islands 33°22′N 118°20′W, 60 m. Coll. 20.ix.1938 (one wet spec. R = 5.2, r = 3.3); CASIZ 102006 6 miles offshore from Makapuu Point. Oahu , Hawaii. 396 m (1300 ft). Coll. S. Earle, B. Bartko fr. Deep Star II. 25.x.1979 (one dry spec. R = 12.5, r = 7.6) .

South Pacific. Hippasteria hyadesi holotype. EcAs 2840 south-east of Port Famine (Puerto del Hambre), Chile, 100 m (one dry spec. R = 4.7, r = 2.2, specimen is partly damaged); Id as H. hyadesi USNM E 13586 northwest of Amundsen Sea, South Pacific. 54°49′S, 129°48′W, 549 m. Coll. R V Eltanin (three dry specs R = 3.5, r = 2.0; R = 8.2, r = 3.9; R = ∼9.4, r = 4.7, arms upturned). Id as H. trojana USNM E 13768 east of Banks peninsula, South Island, New Zealand. 43°54′S 174°35′54″E, 585 m. Coll. 27.viii.1971 (one dry spec. R = ∼10.1, r = 6.1, arms upturned); Id as H. hyadesi USNM 1082740 South Pacific (sub-Antarctic) 67°23′S− 67°24′S to 180°00′W – 179°58′W, 595– 516 m. Coll. US Antarctic Research Program (two dry specs R = 8.3, r = 4.3; R = 10.1, r = 6.2); USNM 1121154 Cape Froward, Strait of Magellan. 53°39′24″S, 70°55′W, 82 m (one dry spec. R = 8.4, r = 4.5). MNHN IE- 2007-1311 Solomon Islands 9°12′S, 160°55′E, 722 m, coll. Richer de Forges & Boisselier, SALOMONBOA 3 aboard N/O Alis (one wet spec. R = 12.4, r = 7.6).

South Indian Ocean. EcAs 10692. Id as H. hyadesi . Crozet Islands 45°37′S, 50°34′E, 200 m. Coll. N.O. Marion Dufresne MD 30 , 20.ii.1982 (six dry specs R = 11.4, r = 6.2; R = 8.6, r = 4.6; R = 8.2, r = 4.3; R = 6.7, r = 3.8; R = 6.6, r = 3.3; R = 6.9, r = 3.3); EcAs 10683. Id as H. hyadesi Crozet Islands 46°33′S, 51°47′E, 200 m. Coll. N.O. Marion Dufresne MD 30 , 24.ii.1982 (five dry specs R = 4.6, r = 2.6; R = 3.8, r = 1.8; R = 2.8, r = 1.5; R = 3.0, r = 1.4; R = 2.8, r = 1.6); EcAs 10860. Id as H. hyadesi . Crozet Islands, 210 m. Coll. N.O. Marion Dufresne MD 08 , CP 257 (one dry spec. 13.1, r = 7.7); EcAs 10861. Crozet Islands, 46°26′S, 52°04′E, 50 m. Coll. N.O. Marion Dufresne MD 08 , CP DC 309 (two dry specs R = 7.7, r = 3.6; R = 16.1, r = 7.0); MNHN IE- 2006-1162 Id as H. hyadesi . Kerguelen Islands 48°43′S, 71°06′E, 925– 937 m. Coll. A. Guille, 13.iii.1975 (one dry spec. R = 10.5, r = 6.5); MNHN IE-2006-1154 Id as H. hyadesi . Kerguelen Islands 52°43.26′S, 72°41.73′E, 240 m. Coll. N.O. Marion Dufresne 42, st. 02–12 (one dry spec. R = 13.1, r = 8.2), EcAs 10735. Marion Island. 46°41′S, 38°73′E, 315– 570 m. Coll. N.O. Marion Dufresne MD08 , CP173 29.iii.1976 (one dry spec. R = 8.7, r = 3.9); EcAs 10856 Marion Island. 46°47′S, 38°03′E, 180 m. Coll. N. O. Marion Dufresne MD08 , cp116 26.iii.1976 (one dry spec. R = 8.9, r = 4.7), EcAs 10858 Marion Island. 46°45′S, 37°57′E, 185– 232 m. Coll. N. O. Marion Dufresne MD08 , CP134 27.iii.1976 (one dry spec. R = 11.2, r = 7.1); EcAs 10854 Bouvet Island 53°13′S, 44°39′E, 570– 585 m. Coll. N. O. Marion Dufresne MD24 , CP51 2.ix.1980 (one dry spec. R = 6.4, r = 2.8), EcAs 10852 Bouvet Island 46°1′S 42°32′E, 270– 315 m. Coll. N. O. Marion Dufresne MD24 , CM69 8.ix.1980 (three dry specs R = 6.7, r = 3.4, R = 7.4, r = 3.2; R = 1.5, r = 1.0) .

South Atlantic /Sub-Antarctic. Id as H. hyadesi USNM E 43921 66°20′S, 67°47′W, 325 m. Coll. 25.iii.1959 (one dry spec. R = 7.4, r = 3.5) .

North Atlantic. USNM 14361 Georges Bank, Massachusetts , 287 m (157 fms), coll. USFC Albatross , 30.viii.1883 (one dry spec. R = 10.8, r = 6.1); USNM E46615 northern edge of Georges Bank. 42°7′N, 66°23′W, 82 m. coll. R V Albatross IV (one dry spec. R = 6.3, r = 3.3); USNM E46606 , south-east of Cape Elizabeth, Gulf of Maine, Maine. 43°31′N, 69°49′W, 97 m. Coll. R V Albatross IV, 14.vii.1965 (one dry spec. R = 6.2, r = 3.7); USNM E 46610 , Browns Bank , Nova Scotia, 42°34′N, 65°44′W, 93 m, coll. R V Albatross IV, 15.x.1965 (one dry spec. R = 7.3, r = 3.5); USNM E 46611 Browns Bank , Nova Scotia, Canada. 42°45′N, 65°7′W. 97.0 m. Coll. R V Albatross IV, 15.x.1965 (one dry spec. R = 1.7, r = 1.0). USNM E46613 western part of Georges Bank, Massachusetts. 41°24′N, 68°25′W, 68 m. Coll. R V Albatross IV, 28.x.1965 (one dry spec. R = 2.2, r = 1.7); USNM E46619 south of Browns Bank , Nova Scotia 42°11′N, 65°50′W, 247 m. Coll. R V Albatross IV (12 dry specs R = 5.1, r = 2.4; R = 4.4, r = 2.2; R = 5.1, r = 2.8; R = 4.3, r = 2.0; R = 3.3, r = 1.6; R = 3.2, r = 1.6; R = 3.0, r = 1.5; R = 3.0, r = 1.5; R = 3.1, r = 1.2; R = 2.9, r = 1.5; R = 2.7, r = 1.4; R = 2.2, r = 1.2); USNM 46622 south of Browns Bank , Nova Scotia 42°2′N, 65°47′W, 256 m. Coll. R V Albatross IV (two dry specs R = 6.2, r = 2.7; R = 4.2, r = 2.0). USNM E30719 Baltimore Canyon , 38°9′30″N, 73°51′12″W, 457 m, coll. Bureau of Land Management/ Johnson Sea Link. iii.1983 (one dry spec. R = 15.5, r = 8.3); USNM E30720 Baltimore Canyon , 38°9′56″N, 73°51′52″W, 417 m, coll. Bureau of Land Management/ Johnson Sea Link, iii.1983 (one dry spec. R = ∼13.7, r = 7.2) .

Figures

Figure 1. Evolutionary relationships of sampled Hippasterinae (Goniasteridae) based on 654 aligned nucleotide positions of the cytochrome c oxidase subunit I gene. The phylogenetic tree was inferred using the neighbour-joining method (Saitou & Nei, 1987) in MEGA v. 5.2.1 (Tamura et al., 2011). The bootstrap consensus tree inferred from 500 replicates was drawn to scale, with distances computed using the maximum composite likelihood method (Tamura et al., 2004), in units of numbers of base substitutions per site. Ambiguous positions were removed for each pair of sequences prior to calculating the distance matrix. Terminals were labelled with GenBank accession numbers and (where available) with museum voucher numbers. Bootstrap support values are shown above or below branches when greater than 70%.

Figure 7. Hippasteria phrygiana (North Atlantic, South Pacific). USNM 14361 North Atlantic. Scale bar = 1.0 cm. A, abactinal surface. B, actinal surface. C, marginal plate series. USNM E13768 New Zealand. D, abactinal surface. E, marginal plate series. F, actinal surface showing furrow and mouth spines. G and H show USNM E46619 and MNHN (Paris) EcAs 10683, respectively, at R = ∼4.5–5.0. I, CASIZ 113374 British Columbia. J, CASIZ 173038 Aleutians. Lateral view showing abactinal spines, pedicellariae, and marginal plates.

Figure 8. Hippasteria phrygiana variation across its range. Sclae bar = 1.0 cm for all unless specified otherwise. A, MNHN IE-2007-1311. Solomon Islands. Abactinal surface. B, actinal surface. C, marginal plates. D, Hippasteria aleutica paratype USNM E47392. E, abactinal spines and pedicellariae. Each pedicellariae is 2.0 mm in length. F, furrow spines and actinal pedicellariae. Scale bar = 0.5 cm. G, marginal plates. H, CASIZ 115700 Hippasteria imperialis. Abactinal surface. I, marginal plates. J, actinal surface showing pedicellariae. Scale bar = 0.5 cm.