Chrysaora colorata (Russell) Gershwin & Collins, 2002

Gershwin, Lisa-Ann & Collins, Allen G., 2002, A preliminary phylogeny of Pelagiidae (Cnidaria, Scyphozoa), with new observations of Chrysaora colorata comb. nov., Journal of Natural History 36 (2), pp. 127-148 : 134-142

Chrysaora colorata (Russell) comb. nov.

(®gures 3± 7)

Pelagia colorata Russell, 1964: 133 ; Russell, 1970: 87; Haderlie et al., 1980: 51, pl. 3.21; Nybakken, 1982: pl. 1; Brandon and Rokop, 1985: 38; Pearse et al., 1987: 156; Sommer, 1988: 170 A; Larson and Arneson, 1990: 131; Cairns et al., 1991: 12; Martin and Kuck, 1991: 97; Fautin and Lowenstein, 1992: 13; MacInnis, 1992: 74; Gowell, 1993: 4, 13, 34; Nybakken, 1993: pl. 1; Niesen, 1994: 46, ®gures 4±31; Thuesen and Childress, 1994: 84; Lange and Kaiser, 1995: 62; Rigsby, 1996: 4, 7±9; Arai, 1997: 121, 140, 225; Martin et al., 1997; Rigsby, 1997: 189; Wrobel, 1997: 48, 52; Wrobel and Mills, 1998: cover, 21, 23, 54.

Pelagia noctiluca: Fox and Millott, 1954: 392 ; Millott and Fox, 1954: 169; Kramp, 1961: 330; Halstead, 1965: pl. 47, 48; Hinton, 1969: 49, 51, ®gure 23; Hand, 1975: 95; Austin, 1985: 71; McConnaughey and McConnaughey, 1985: 465, ®gure 383; Brusca and Brusca, 1990: 212, ®gure 1E; Gotshall, 1994: 24, ®gure 39.

Pelagia panopyra: Fewkes, 1889a: 122 , pl. 5, ®gure 1; Fewkes, 1889b: 592, ®gure 1; Mayer, 1910: 575 (in part); MacGinitie and MacGinitie, 1949: 122, ®gure 21; MacGinitie and MacGinitie, 1968: 122, ®gure 21, p. 458; Reish, 1972: 25, ®gure 27; Gladfelter, 1973: 256; Galbraith and Boehler, 1974: 53, pl. 2, 3; Hough, 1974: 54; Gabil and Rose, 1975: 8, ®gure 4; Allen, 1976: 22, 75, ®gure 68; North, 1976: 153, pl. 4d; Hauser and Evans, 1978:

19; Gotshall and Laurent, 1980: 40, ®gure 38; Parsons, 1986: 18; Snyderman, 1987: 168, pl. 31; Reish, 1995: 39, ®gure 32.

Pelagia cyanella: Hartman and Emery, 1956: 307 .

Pelagia sp. : Galigher, 1925: 94; Johnson and Snook, 1927: 80; Wells, 1942: 147; Gotshall et al., 1965: 149; Johnson and Snook, 1967: 80; Niesen, 1982: pl. 17; Abbott, 1987: 25; Campbell, 1992: front cover, 2, 4.

Chrysaora hysoscella: Stephens, 1966: 67 .

Jelly®sh: Malnig, 1985: 41; StefoOE, 1997: 2, 7, 17, 20, 21.

Material examined

Live specimens. Fourteen adult medusae (13 females and one male), 30±60 cm subumbrellar diameter; 11 juvenile medusae, 6±12 cm bell diameter; approximately 100 each scyphistomae, strobilae and ephyrae.

Preserved specimens. Adult medusae: CAS IZ Cat. #111015, female, 27 March 1997, San Pedro Channel , CA ; CMA teaching collection, female, diameter 48.0 cm (exumbrellar), 30.0 cm (subumbrellar), 11 April 1997, San Pedro Channel , CA ; CAS IZ Cat. #119475, female with two exumbrellar barnacles, diameter 49.5 cm (exumbrellar), 26.5 cm (subumbrellar), 4 April 1997, San Pedro Channel , CA . Scyphistomae : CAS IZ Cat. #111019, approximatel y ten, 3±4 mm live length, 19 April 1997, MBA culture. Ephyrae : CAS IZ Cat. #111018, approximately ten, 2 mm live 14 diameter, 19 April 1997, MBA culture. Juvenile : CAS IZ Cat. # (awaiting specimen #), 6 February 1998, MBA culture, at least 13-months-old, diameter ca 8.5 cm.

at

Diagnosis

Chrysaora with eight tentacles alternating with eight rhopalia. Bell massive; white or silvery with purple markings of an apical ring, 16 radial stripes, and numerous of

¯ecks and blotches. Rhopalium within thickened region of mesoglea at intersection of subumbrellar pocket and deep exumbrellar and lappettal cones. Oral arms spiralled and entwined, concealing mouth. Quadralinga prominent, with cap three-leaf clover-shaped. Lappets 32, broadly rounded, webbed in pairs between rhopalia or tentacles. Gastric septa S-shaped in distal half, terminating at perradii between by

lappets. Scyphistoma conical to goblet-shaped, pedicelled, typically with amphico-

ronate, spiral tentacles in life. Strobila and ephyra bright purple.

Redescription of C. colorata

Adult medusa (®gure 3). Umbrella hemispherical and massive. Size typically 30±50 cm diameter, reported to 150 cm (Hinton, 1969). Exumbrella ®nely granulated, lacking raised (gelatinous) nematocyst warts. Mesoglea greatly thickened, especially on rhopaliar axes; subumbrella and manubrium opaque, stiOEened, cartilaginous; exumbrella transparent under surface pigment and very soft near apex. Tentacles eight, adradial, marginal, alternating with rhopalia; issuing from between lappets; hollow, lacking longitudinal furrows; with bases laterally compressed, typically bulging inward; reddish except in proximal 5±7 cm, where colourless on outer surface with faint purple sheen to inner lining. Rhopalia eight, four each perradial and interradial, located at intersection of subumbrellar pocket and two deep cones, one from exumbrellar surface proximal to the margin, funnelling toward subumbrella (exumbrellar cone), the other from margin between ¯anking lappets, funnelling centripetally (marginal cone). A thorough description of the pelagiid sensory apparatus is given in Bigelow (1890, 1910). Lappets 32, broad, round, thick and darkly pigmented with pale reticulations, with thin, web-like connection between lappets 14

at of by not separated by a tentacle or rhopalium. Oral arms four, perradial, arising from thick manubrium, partially obscuring the subumbrellar ostia and mouth; extremely frilly and typically spiralled and entwined appearing as one large mass tapering with distance from the bell; reported to over 6 m long (Hinton, 1969; and see striking photo in Gowell, 1993: 13). Gonads four, interradial, nearly concealed by thick subumbrellar mesogleal rings surrounding the ostia; femaleÐa ru‚y purple ribbon (or brownish and granular to the naked eye when gravid) visible through ostia, with orange eggs, approximatel y 100 m m in diameter (®gure 4); maleÐsimilar except ribbon is paler with whitish sperm follicles, giving an overall lilac hue, with magni®ed surface appearing as repeated interlocking oblong and crescent shapes, mostly orientated in the longitudinal direction. Coelenteron partitioned unevenly by 16 septa, S-shaped in distal one-third, on either side of tentacular adradii, with the eight rhopaliar pouches being larger than the eight tentacular pouches. Manubrium short and stiOE. Mouth cruciform, concealed by intertwined oral arms. Quadralinga threelobed. Colour: bell white to silvery, with 16 purple to brownish purple stripes radiating from a purple apical ring, leading in alternation to a rhopalium or a tentacle, but stopping short of bell margin; stripes typically ¯anked distally by two arrow-shaped purple blotches and numerous small ¯ecks.

Scyphistoma (®gure 5A). Conical to goblet-shaped, 2±5 mm long, 1.5± 2 mm diameter at tentacle crown, whitish in colour. Tentacles typically 16 but highly variable, spiralled, alternating in oral-aboral orientation from the body (termed`amphicoronate’ by P. F. S. Cornelius). Mouth typically amorphous but quadrate in some. Septa four, each marked by paired ostia on oral disk. Asexual propagation through podocyst formation. Podocysts broadly hemispherical, diameter 200± 500 m m, with crater-like depression in centre, greenish gold in colour.

14 FIG. 4. Chrysaora colorata , female gonad. Eggs are in the lower left corner. Egg diameter is approximately 100 m m.

Strobila (®gure 5B, C). Capable of extreme polydisc strobilation; up to 56 at

ephyrae observed developing at one time. Purple, with basal polyp remaining whitish.

Maturation successive, with distal ephyrae being released before proximal ones form. of Ephyra (®gure 6). Typically with eight arms, but variable; lappets pointed.

Nematocyst patches in characteristic Chrysaora pattern. Mouth crucifom. Diameter

2±3 mm. Richly purple.

Juvenile (®gure 7). Based on captive-raised individuals to about 12 cm bell diameter. Bell hemispherical, colourless to pinkish, some with faintly streaked coloraby tion of adult. Manubrium elongated, tubular. Oral arms 10±100 cm, simple, separate,

transparent to milky in appearance. Tentacles 1±2 m, adradial, basally swollen and laterally compressed, colour distinctly red. Rhopalia four perradial, four interradial, at intersection of exumbrellar and lappet cones, inside subumbrellar pockets.

Lappets 16, rectangular, with shallow marginal notch; rhopaliar lobe smaller than tentacular lobe. Mesoglea thickened in rhopaliar regions of the bell. Finely granulated nematocyst warts scattered over exumbrellar surface, not uniform.

Remarks on C. colorata

Development. In the laboratory, spawning took place upon changes from darkness to light. Spawned eggs were negatively buoyant. Fertilization is apparently external, as in C. quinquecirrha (see Littleford, 1939). Motile planulae were apparent within several days after spawning. Isolated planulae metamorphosed into scyphistomae within a week, possessing an inverted triangularly shaped body on a thin sheathed pedicel, with a broadened basal attachment disc. By day 8, two tentacles formed, opposite one another, followed by formation of secondary tentacles within several more days, at which time septa became visible through the oral disk. Young polyps readily fed on rotifers and a thin suspension of hard-boiled chicken-egg yolk. Polyps grew quickly when fed a variety of zooplankton and strobilated uninduced in week 10 post-spawn.

14

at of by

Ecological interactions. As with most medusae, crustaceans appear to be the predominant commensal group associated with C. colorata , although there are reports of other taxa as well. Cancer gracilis Dana, 1852 has been reported clinging commensally to the medusa or in the gut cavity (Hauser and Evans, 1978; Haderlie et al., 1980; Wrobel and Mills, 1998); juvenile yellow®n and spot®n croakers have been seen accompanying the medusa (Hinton, 1969); and MacGinitie and MacGinitie (1949, 1968) reported several occurrences of Obelia colonies attached to the exumbrellar surface of the bell. Two Alepas paci W ca were found with rootlike tendrils embedded in the apical surface of a single medusa. Although it does not appear to have many predators, C. colorata is reportedly fed upon by the ocean sun®sh Mola mola ( Linnaeus, 1758) and the blue rock®sh Sebastes mystinus ( Jordan and Gilbert, 1881) (Gotshall et al., 1965; Haderlie et al., 1980).

Bioluminescence. A 20 cm individual recorded as `Pelagia sp.’ from California waters was luminous after being agitated in a bucket of water that had been left in the dark (S. Haddock, personal communication). Due to the size and location of the individual, it is probable that it was C. colorata . If C. colorata is indeed bioluminescent, this is the ®rst report of luminescence in the genus Chrysaora . It is interesting that this small (juvenile?) medusa exhibited luminescence, while large adults we have tested have not. Other reports of bioluminescence appear to be 14

FIG. 6. Chrysaora colorata ephyra, illustrating the characteristic Chrysaora nematocyst pattern of two oblong patches ¯anking each rhopalium and a corresponding ring of round patches on the body. Cruciform mouth is also visible through the body.

at assumptions based on the mistaken belief that C. colorata is conspeci®c with of P. noctiluca , which is brilliantly luminescent, or with one of its congeners.

Variation in symmetry. Typically tetramerous, but highly variable. Clone mates of diOEerent symmetries were reported by Gershwin (1999). A later study of 973 ephyrae revealed a total of 79 non-tetramerous (Gershwin, unpublished notes). Monterey Bay Aquarium has occasionally displayed both hexamerous and trimerous by individuals.

Stinging ability. Sting painful, lasting approximately one hour. Russell (1964) reported that ammonia aggravate d the sting, whereas hydrochloric acid helped to soothe it. In contrast, Hinton (1969) reported that ammonia helped alleviate the pain. Parsons (1986) recommended isopropyl alcohol or household vinegar. However, neither ammonia nor vinegar were helpful to L. G. or ®eld assistants.

Discussion

Although the members of the Pelagiidae have traditionally been grouped on tentacle number, these results illustrate the importance of multi-character analysis in determining relationships. Although Chrysaora colorata has`the de®ning Pelagia character’ according to previous classi®cations, namely eight tentacles, each in alternation with a rhopalium, it shares many more characters with species of Chrysaora . Furthermore, its closest relative appears to be the recently described Chrysaora achylos ( Martin et al., 1997) . We found it surprising that the large and conspicuous quadralinga had not been previously described (®gure 1).

Phylogeny review and taxonomic implications

The historical genus Dactylometra is of particular interest to us. For a long time, this genus was thought to represent the third of four stages in pelagiid development 14

at of by (see Uchida, 1935, for a well-phrased summary). This genus name was later reduced to a junior synonym of Chrysaora because some Dactylometra medusae reach maturity in the `Chrysaora stage’, i.e. with 24 tentacles ( Kramp, 1955; Calder, 1972). This conclusion may have been in error. If we de®ne taxa by a single character, e.g. tentacle number, then any deviant specimen calls into question our classi®cation. However, if we de®ne taxa by numerous characters, then we may be better able to evaluate evolutionary relationships. In our preliminary analysis, the taxa with typically 40 tentacles cluster as a sister group to the taxa with typically 24 tentacles (and C. colorata with only eight). The split is only weakly supported; however, analysis of additional characters may prove instructive on whether Dactylometra is a clade.

Character analysis

In analysing the characters of C. colorata , we found the number and shape of the marginal lappets as described by Russell (1964) to be confusing. In adult Pelagia , there are 16 rectangular lappets, whereas in adult Chrysaora there are 32 or more and they are either all rounded or of two types. In species typically with 32 tentacles, each tentacle occupies a position between the lappets, which are rounded and all alike. In species that typically have 40 tentacles, the lappets adjacent to the rhopalia (I and VI in each octant) are divided radially by a narrow membrane; this membrane is the site of generation of the additional tentacles. These rhopalial lappets are of a diOEerent form than those in 24-tentacled species, which possess no such membrane. Instead of broad rounded lappets, the splitting of the membrane leaves the rhopaliar lappets small and pointed, and the other`half ’ of each original lappet larger and asymmetrical (recognized as lappets II and V). In these medusae, the centremost lappets (III and IV) remain unchanged, being broadly rounded. In some individuals, the tentacles begin to grow from the subumbrella prior to splitting of the membrane (L. Gershwin, personal observation). Thus, typically the number of lappets is the same as the number of tentacles1 rhopalia, the tentacles being in alternation with one or the other.

In C. colorata, Russell interpreted the lappets as being 16, rectangular with shallow median notches, and with a thin central area of fusion. However, in the mature specimens we observed, there are 32 thickened and heavily pigmented lobes, all of the same form, such that the thin central area of fusion appears to be a weblike connection between two adjacent lappets rather than dividing two sides of a 14 single lappet. Thus, we contend that mature C. colorata have 32 lappets, with the 16 inter-lappettal webs occupying the spaces in which tentacles are normally observed. Confusingly, however, juvenile C. colorata actually appear to have 16 undivided lappets, as described by Russell; the signi®cance of this discrepancy at

is unclear.

It is also important to note that scyphozoans can be particularly variable in of relative numbers of body parts. Gershwin (1999) found that numerous species of scyphozoans, including C. colorata and C. fuscescens , readily exhibit variations in overall symmetry, adding or subtracting parameres but maintaining the numerical ratio of body parts constant on each paramere. Similarly, the number of rhopalia can vary. Calder (1972) reported variable numbers of tentacles in C. quinquecirrha by corresponding to growth stage. Thus, simply counting the organs of a given specimen may prove an inadequate method of determining its identity.

Outstanding questions concerning C. colorata

Although Chrysaora colorata has been raised in captivity for over a decade, many questions concerning its basic biology remain. For example, we still do not know how long it lives or where it breeds. Whether the medusae migrate into southern California waters, or ascend with upwelling current, or are carrying out their life cycle locally, is unknown. Most sightings are in the late spring and are of mature medusae that have already attained the large size and distinctive coloration pattern. The scyphistoma has yet to be found in the wild, and juvenile medusae have only rarely been observed (Fewkes, 1889a; S. Haddock, personal communication). Captive-raised medusae typically live approximately 2 years and reach only about 12 cm diameter; interestingly, they attain coloration only faintly reminiscent of the wild type (D. Wrobel, personal communication). Calder (1972) noted a similar lack of coloration in his captive-raised C. quinquecirrha . It seems likely that the natural diet of these species is responsible for the coloration.

Another question, for which we no answer, came to us when we sexed the captured individuals. All but one were mature females; the other was a mature male (N 514). Although this was a small sample size, we still wondered if the female bias 14

at of by was meaningful (a male bias was found by Gershwin (2001) in her study of Aurelia labiata ). One possible explanation for the scarcity of males is that the females outlive the males, as is the case in Cyanea capillata ( Linnaeus, 1758) , a planula brooder (Brewer, 1989). However, there is no indication that Chrysaora colorata females brood their planulae, and thus no apparent advantage to the females outliving the males. L. G. has observed that captive C. colorata medusae spawn regularly with changes in light, resulting in viable planulae, indicating that even very large medusae are still in reproductive condition. It is also possible that C. colorata is a hermaphrodite, as is C. hysoscella , with mature individuals being in the female phase. This seems doubtful, however, because the one male found was within the size range of females and possessed testes in the typical scyphozoan form of folded gonads, which contrasts with C. hysoscella , in which the testes are located instead in blisters on the oral arms and gastric ®laments (Claus, 1877; Widersten, 1965).

It is odd that pelagiid systematics are in disarray. The medusae tend to be large, distinctively pigmented, and coastal. In addition, they often sting rather severely. And yet, Chrysaora colorata was only described in 1964, and even then incorrectly classi®ed. Moreover, its close relative, C. achylos , was only described three years ago despite being the largest invertebrate described in the 20th century ( Martin et al., 1997). It seems likely that new pelagiids in other regions have yet to be discovered.