Hebella furax Millard, 1957

De Andrade, A. E. Migotto L. P., 2000, The life cycle of Hebella furax (Cnidaria: Hydrozoa): a link between a lafoeid hydroid and a laodiceid medusa, Journal of Natural History 34, pp. 1871-1888 : 1873-1883

Hebella furax Millard, 1957

(®gures 1±5; tables 1 ±5) Hebella furax Millard, 1957: 200 ±202, ®gure 8; 1964: 10±11, ®gures 2B±D; 1975: 181, ®gures 60A±C; Millard and Bouillon, 1973: 59; Boero et al., 1997: 19 ±22, ®gures 10±11.

Examined material

HOLOTYPE of Hebella furax : colonies epizoic on Lytocarpus W lamentosus (Lamarck, 1816) (5 Macrorhynchia W lamentosa), oOE Zwartklip, False Bay, South Africa, 7 December 1903, without gonangia, SAM H34, preserved in alcohol; also two microslides (same deposit number).

Non-type. AM1188 : Saco Grande , 23 ss 49.72 ¾ S; 045 ss 25.52 ¾ W, SaÄo SebastiaÄo Channel, SaÄo SebastiaÄo, SP, Brazil, on Sertularia marginata growing on rocks, 1.0±2.0 m, 5 January 1995, with gonangia (also BMNH 1998.2669 ) ; AM1196 : idem, 9 January 1995, without gonangia ; AM1195 : idem, 16 January 1995, without gonangia ; AM1189 : idem, 13 February 1995, with gonangia ; AM1184 : idem, 29 April 1995, without gonangia ; AM1183 : idem, 15 May 1995, without gonangia ; AM1194 : idem, 17 January 1996, without gonangia, produced gonangia in the laboratory ; AM1197 : idem, 27 February 1996, with gonangia ; SAM H4842 : 12 March 1996, with gonangia; AM1191 : idem, 1 April 1996, without gonangia ; AM1182 : idem, 3 April 1996, with gonangia ; AM1192 : idem, 10 May 1996, without gonangia ; AM1234 : idem, 17 August 1996, without gonangia ; AM1190 : idem, 22 January 1997, without gonangia ; AM1185 : idem, on Obelia bidentata Clarke, 1875 , 8 February 1995, without gonangia ; AM1193 : Farol dos Moleques, 23 ss 49.72 ¾ S; 045 ss 24.78 ¾ W, SaÄo SebastiaÄo Channel, SaÄo SebastiaÄo, SP, Brazil, on Sertularia marginata growing on rocks, 1.0±2.0 m, 17 January 1996, without gonangia ; AM1228 : Parcel da Praia Grande , 23 ss 51.18 ¾ S; 045 ss 25.11 ¾ W, SaÄo SebastiaÄo Channel, Ilhabela, SP, Brazil, on Sertularia marginata growing on rocks, 1.0±2.0 m, 28 October 1996, without gonangia ; AM1236 : 19 February 1996, two 25-day-old medusae, liberated on 25 January 1996 from material AM1194 ; AM1241 : 19 February 1996, three 18-day-old medusae, liberated from material collected from Saco Grande on 1 February 1996; AM1229 : 2 April 1996, one 28-day-old medusa, liberated on 5 March 1996 from material AM1197 ; AM1233 : 2 April 1996, one 26-day-old medusa, liberated on 7 March 1996 from material AM1197 ; AM1226 : 8 April 1996, one 34-day-old medusa, liberated 5 April from colony AM1197 ; AM1227 : 8 April 1996, one 32-day-old medusa, liberated 7 March 1996 from colony AM1197 ; SAM H4843 , 8 April 1996, one 34-day-old medusa, liberated 5 March 1996 from colony AM1197 ; AM1239 : 8 April 1996, one 33-day-old medusa, liberated 6 March 1996 from colony AM1197 ; AM1240 : 8 April 1996, one 34-day-old medusa, liberated 5 March 1996 from colony AM1197 ; AM1243 : 8 April 1996, one 34-day-old medusa, liberated 5 March 1996 from colony AM1197 ; AM1242 : 9 April 1996, one 36-dayold medusa, liberated 4 March 1996 from colony AM1197 ; AM1235 : 9 April 1996, two 34-day-old medusae liberated 6 March from colony AM1197 ; AM1244 : 9 April 1996, one 34-day-old medusa liberated 6 March from colony AM1197 (also BMNH 1998.267 0) .

Hydroid

Colonies stolonal. Hydrorhiza creeping over hydrorhiza, hydrocaulus and hydrocladium of the sertulariid hydroid Sertularia marginata Kirchenpauer, 1864 (®gure 1D) and occasionally on Obelia bidentata Clarke, 1875 . Hydrothecae pedicellate, arising from hydrorhiza at more or less regular intervals, almost always at the back and basal half of host hydrocaulus, infrequently extending onto hydrocladia. Up to 10 hydrothecae (usually two to six) per cormoid of host. Pedicel indistinctly or distinctly spirally annulated. Pedicel from host hydrorhiza long, given oOE at a right angle from hydrorhiza (®gures 2B, 2F); those from host hydrocaulus and hydrocladia short and given oOE obliquely from hydrorhiza (®gures 2A±D). Hydrotheca with smooth perisarc, deeply campanulate to almost cylindrical, with straight or curved walls. Hydrothecae arising from hydrorhiza of host usually symmetrical; those from erect part of host often asymmetrical, with the wall closer to host more convex. Margin entire, usually perpendicular to axis of hydrotheca, with up to four renovations (®gures 1C, 2A, B). Rim ¯aring; hydrothecal ori®ce circular, without operculum (®gures 2A±F). Base of hydrothecae with annular perisarcal thickening and thin diaphragm (®gures 2A, D). Desmocytes present near annular perisarcal thickening, above diaphragm, arranged in one to ®ve irregular rings (®gure 1E). Hydranth long, cylindrical (®gure 1A), milky white to light pink in colour; hypostome low, conical; tentacles ®liform, amphicoronate, without basal web. Hydranths attached to the skeleton only in the region of desmocytes.

Gonotheca on short pedicel, conical, smooth or with waved perisarc, and straight walls, truncated distally and with a four-valve operculum (®gures 2G, H, 3C); fully developed gonangia containing three medusae in a row, all in diOEerent stages of development. Colonies collected without gonangia and kept in the laboratory developed gonangia after 1±3 days; gonangia appeared as small rounded projections from hydrorhiza and gradually became conical and elongated to achieve their ®nal shape; in the laboratory, at 25 ss C, the ®rst medusa released about 48 h after beginning of gonangia formation.

Morphological character variation of many colonies is given in table 1.

Newly released medusa

Medusae released individually from the gonotheca, oral side ®rst, at about 12-h intervals. Colonies collected having gonangia, and colonies in which the gonangia developed in the laboratory, released medusae with three large perradial tentacular bulbs with tentacles and one perradial bulb, slightly smaller than the other bulbs, without any tentacles (®gure 3A, B). Tentacles and tentacle bulbs hollow; tentacles ®liform, long, armed with microbasic mastigophore nematocysts. Each tentacular bulb with a small, black adaxial ocellus. Umbrella high; jelly thin. Exumbrella with microbasic mastigophore nematocysts in an equatorial band. Umbrella margin with four small interradial bulbs without tentacles, less developed than the perradial ones and without ocelli. Four broad radial canals and circular canal present. Stomach large, ¯at and square. Manubrium with four lips, cruciform and short, reaching about half the height of subumbrellar cavity.

Some newly released medusae diOEered from the description above in having four perradial tentacles, one clearly shorter than the others, and small adradial bulbs between interradial and perradial bulbs. A few medusae released with two tentacles and with smaller dimensions were considered abortive and their measurements were not included in table 2.

Development of medusa

During development the jelly became progressively thicker, especially on the upper part of the umbrella, the umbrella gradually became ¯atter, and the exumbrellar nematocysts became scattered, completely disappearing in adult medusae. Data on growth of many medusae reared in the laboratory are presented in table 3.

Tentacles and cordyli. The fourth perradial tentacle was completely formed 2± 3 days after release. At this age the interradial bulbs attained the same size as the perradial ones, and adaxial black ocelli were apparent. Typically, the interradial tentacles developed after the corresponding canals reached the circular canal. Fiveday-old medusae started to develop rudimentary adradial swellings, which later acquired tentacles and ocelli (sometimes two ocelli in each bulb). Eleven-day-old medusae had the adradial tentacles completely formed; at this age, between every two tentacles one cordylus started to grow. The development of the adradial tentacles was independent of the increase in number of radial canals. The cordyli were clubshaped to amphora-like, with nematocysts concentrated at the tip and with an adaxial black ocellus, and measured approximatel y 250±500 m m in length (®gure 4C).

Radial canals. Three- to 5-day-old medusae started to develop interradial canals, originating from the stomach; usually these canals appeared simultaneously; it took more than about 3 days for the interradial canals to attain the same size as the perradial ones. In most medusae the adradial canals showed no pattern in their appearance; they arose directly from the stomach or as branches from the primary canals; some branches grew near the stomach making it di cult to distinguish their origin. Frequently, as the medusa grew, the branch and the distal part of the primary radial canal beyond the branching point changed their positions in relation to the proximal part of the primary canal; as a result it appeared that the primary canal had stopped its linear growth and bifurcated, producing two branches that grew until they touched the circular canal. In some 34-day-old medusae the beginning of the second branch could be seen coming oOE from the primary canal, on or near the ®rst branching point (®gure 5); there were even two secondary canals branching oOE from the same primary canal. The radial canals were narrow at their terminal end, before reaching the circular canal (®gure 4).

Gonads. The gonads began developing on the lateral walls of the radial canals, near the stomach, appearing as faint granules, in medusae of about 11±14 days of age, and grew along the canal towards the bell margin. In some canals the two gonadal bands fused; usually a middle furrow remained in most gonads of adult medusae. In 20-day-old medusae the gonads were situated along the middle approximately two-thirds of the length of the radial canals, were linear to sinuous, and pale yellow. Nine medusae reached maturity, all of them female (®gure 4). The medusae spawned numerous eggs after 31 days of cultivation.

Adult medusa

Umbrella ¯at, twice as broad as high, with moderately thick jelly. Manubrium short and ¯at, irregular in shape, with lobes; limit between base of radial canals and stomach not well marked. Mouth with four irregular lips. Nine to eleven broad radial canals (primary and secondary), broader near stomach, and narrower near circular canal. Secondary canals originating from stomach or near base of the primary canals; being di cult to distinguish one from the other due to irregular shape of the stomach. Velum narrow, less than one-eighth of bell radius. Gonads linear to sinuous, on walls of the canals. Margin of umbrella with 24±37 tentacles, and 30±57 cordyli; usually one or two cordyli between every two tentacles, sometimes three or none (®gures 4, 5). Marginal bulbs with one to three irregular, black adaxial ocelli. Cordylus club-shaped, with black adaxial ocellus and a few nematocysts at tip. Morphological data of nine medusae that reached maturity are summarized in table 4.

Nematocysts

The hydroid stage has two size classes of microbasic mastigophores; the larger one (seen undischarged) distributed only on the pedicel and stolon; capsule elongated, bean-shaped, with asymmetrical walls. Small microbasic mastigophore s distributed on hydranth body, especially on tentacles; capsule oval; discharged shaft more or less continuous with the longitudinal axis of the capsule, approximately of the same size as the capsule length ( table 5).

The microbasic mastigophores of the medusa similar in morphology to the small size class present on the hydranth. Young medusae with two, slightly diOEerent size classes, the larger on exumbrella and the smaller on tentacles and tentacle bulbs. Nematocysts of adult medusae larger than those of young ones, distributed only on tentacles, tentacle bulbs and cordyli (no nematocysts were seen on mouth and manubrium) ( table 5).

Remarks

The description of the hydroid stage given above is based solely on material from SaÄo SebastiaÄo. The pedicels from the holotype are longer and have more annuli than specimens from SaÄo SebastiaÄo; this is the most striking diOEerence between the two ( table 1). As can be seen in data from table 1, there is also great diOEerence in the length of pedicel between hydrothecae from the hydrorhiza and hydrocauli of

Length (m m) Width (m m) [mean Ô SD (range)] [mean Ô SD (range)]

Hydranth and hydrorhiza 8.3 Ô 0.9 (6.5±10.0) (n 5 48) 2.9 Ô 0.4 (2.5±4.0) (n 5 48) Hydrorhiza 26.9 Ô 2.3 (19.0±30.0) (n 5 28) 8.1 Ô 0.7 (7.0±10.0) (n 5 28) Newly released medusae

Umbrella 11.5 Ô 0.5 (10.0±12.5) (n 5 90) 4.8 Ô 0.4 (4.0±5.5) (n 5 90)

TentacleÐlarge nematocyst 11.5 Ô 0.7 (10.0±13.0) (n 5 90) 4.5 Ô 0.4 (4.0±5.5) (n 5 90)

TentacleÐsmall nematocyst 8.0 Ô 0.7 (7.0±9.5) (n 5 90) 3.1 Ô 0.4 (2.5±4.0) (n 5 90) 34-day-old medusae

Tip of tentacle 15.1 Ô 0.7 (14.0±16.5) (n 5 80) 5.6 Ô 0.6 (4.5±6.5) (n 5 80)

Middle region of tentacle 15.3 Ô 0.7 (14.5±16.5) (n 5 80) 5.7 Ô 0.5 (5.0±6.5) (n 5 80)

Base of tentacle 14.5 Ô 0.6 (13.0±15.0) (n 5 80) 5.1 Ô 0.4 (4.5±5.5) (n 5 80)

Tip of cordylus 11.0±13.0 (n 5 4) 3.5±4.5 (n 5 4)

the hosts, in specimens from Brazil, but the variation in this length is much greater in the holotype. Boero et al. (1997) also reported that the species may have short or long pedicels and that`...the hydrothecae growing at the base of the stem of the aglaopheniid hosts are sharply cylindrical and have a long pedicel, being so diOEerent from the ordinary hydrothecae that they might be referred to diOEerent species...’. DiOEerent shapes of hydrothecae according to their position on the host are also exhibited by H. scandens (see Migotto, 1996).

Unlike the report of Millard (1975) which described a form that sometimes invades the perisarc of the host, no parasitism was found in specimens from SaÄo SebastiaÄo.

As the stolons of H. furax can follow the hydrorhizal net of the host to reach other hydrocauli, it is capable of colonizing several stems of a colony or clone of the host.

Boero et al. (1997) described the newly released medusa as having`four perradial tentacular bulbs, two opposite ones bearing a long tentacle each, the other two bearing one short tentacle each’. They did not mention any newly released medusa with three tentacles only, as most of ours possessed, but only that one of the short tentacles was shorter than the other. We obtained only a few newly liberated medusae bearing four tentacles, and these came from colonies from one collecting site only (Farol dos Moleques). Colonies from the other localities on the Channel of SaÄo SebastiaÄo produced medusae bearing three tentacles. This indicates that the number of tentacles present during release is a variable character among populations of H. furax . We, therefore, discard the hypothesis that the newly released medusae with three tentacles are abortive, as we obtained hundreds of such medusae from many diOEerent colonies, some collected with fully developed gonangia and some that developed gonangia in the laboratory.