Halistemma transliratum Pugh & Youngbluth, 1988

Pugh, P. R. & Baxter, E. J., 2014, A review of the physonect siphonophore genera Halistemma (Family Agalmatidae) and Stephanomia (Family Stephanomiidae), Zootaxa 3897 (1), pp. 1-111 : 49-57

Halistemma transliratum Pugh & Youngbluth, 1988

Halistemma transliratum Pugh & Youngbluth, 1988, pp. 1–14 , figs. 1–6, 6A, 7.

Diagnosis: Nectophores without mouth plate. One pair of complete vertical lateral ridges. Lateral ridges extending from apico-laterals almost to ostium, ending on lateral ostial processes. Apico-laterals divide close to the ostium, the inner pair continuing on to the ostium, while the outer pair runs out laterally. Several patches of ectodermal cells on upper and lateral surfaces. Four types of adult bract, three of which have a characteristic transverse ridge. Terminal filament of tentillum ends in small glandiform process.

Holotype: Specimen collected at a depth of 472 m during JSLI Dive 628 (3 October 1981; 26°13.5'N, 77°41.8'W). Preserved in 5% buffered formalin and presented to the United States National Museum (Smithsonian Institution), USNM 78496 .

Paratype: Specimen collected during JSLII Dive 972 at a depth of 563 m (21 October 1984; 26°24.3' N, 77°49.8'W). Presented to the British Museum (Natural History), Regd. No. 1987.4.1.1 .

Material examined. The paratype specimen has been re-examined, together with pieces of other specimens found in 25 recent Discovery hauls (see below).

Description.

Pneumatophore: Borne on a long stalk, with no characteristic features.

Nectophores. ( Figures 50–53 ) Fifteen nectophores, which measured up to 13 mm in length and 13.4 mm in width, were present with the holotype, and seven with the paratype. The number of nectophores found in individual Discovery samples ranged from 1 to 93 (see Table 1 below), but it was impossible to say exactly how many specimens were present in each of the 25 samples collected. The basic Halistemma ridge pattern was present. The pairs of upper and lower laterals joined at the lateral corners of the axial wings and continued for a short distance as a single ridge. The upper laterals converged as they passed over the upper surface of the nectophore and in many of those from the holotype specimen, shortly above the ostium, they overlapped one another before diverging again ( Pugh & Youngbluth, 1988). A short distance above the ostial level they divided, giving off a long lateral branch and a much shorter descending one. In the original description Pugh & Youngbluth (1988) suggested that the latter branch continued down to the ostium, but the recent examination of the paratype suggests that this an artefact, and that the ridge is very short ( Figures 50 & 51 ). The lateral branch passed outwards, parallel to the ostium and petered out just before reaching the tips of the lateral ostial processes. During their course the pair of upper lateral ridges gave rise to the vertical lateral and lateral ridges. Each vertical lateral ridge ran, slightly diagonally, down the lateral margin of the nectophore to join with the corresponding lower lateral ridge. Similarly, each lateral ridge ran obliquely toward the ostium and ended at the tip of the corresponding enlarged lateral ostial process. The lower lateral ridges, from their junction with the upper laterals, ran along the lower lateral margin of the nectophore to reach the ostium.

In young nectophores the thrust block was small and deeply emarginated centrally ( Figures 52 & 53 ). However, as the nectophores matured the thrust block increased considerably in both width and height and could extend further than the axial wings, which did not increase much in size as the nectophore enlarged. A small mouth plate was present on the lower side of the ostium in the younger nectophores, but this virtually disappeared as the nectophores matured. The lateral ostial processes were quite extensive, but it is not known if these bore nematocysts as very young buds of nectophores were not found. The patches of ectodermal cells were more numerous on the young nectophores of the paratype specimen than those previously described from the holotype. A row of up to five patches was arranged parallel to the vertical lateral ridges. In addition, a pair of patches, often quite elongated, appeared above the infra-lateral ridge on the lateral sides close to the ostium, and several patches could occur just lateral to the apico-lateral ridges in the distal half of the nectophore. All these patches became very indistinct on the mature nectophores of the paratype.

The ascending branch of the mantle canal was approximately twice the length of the descending one. Pugh & Youngbluth (1988) noted what they thought to be a thin canal arising from the base of the descending mantle canal, which ran along the mid-line of the lower side of the nectophore apparently to reach the ostial ring canal. This, of course, is not a canal but a giant nerve fibre that, for instance, Mackie (1973) described in the nectophore of Nanomia bijuga .

The long pedicular canal reached the nectosac on its lower side just below its median apex, but did not give rise directly to all four radial canals, only the upper and lower ones. The lateral radial canals arose a short distance along the upper canal either together, as in the younger nectophores of the paratype, or asymmetrically, as for the more mature ones. For the paratype nectophores ( Figure 53 ) the course of the lateral radial canals was often quite irregular and, occasionally, small diverticula could be seen branching off at various places along their length. At first the lateral canals were directed out toward the lateral margins of the nectosac but, before they reached it, they made a short loop down toward the ostium, before looping out and up on to the lateral surfaces of the nectosac. There they made a large loop up and down, followed by a small loop on the lower surface before returning to the mid-height of the nectosac and then running directly to join the ostial ring canal.

Siphosome:

Bracts. At least four kinds of adult bract were found with the holotype specimen.

Type A bracts ( Figure 54 ) were elongate and lanceolate, measuring up to 41 mm in length. They resembled the larval bracts of other species, but were easily distinguished by the inverted U-shaped ridge running transversely across the upper surface. There was also a median ridge on the upper surface running from just distal to the transverse ridge down to the distal tip of the bract. In addition, toward the proximal end of the bract there was a small flap on one side on its lower side. Numerous patches of ectodermal cells were found on the upper surface, often with an almost biserial arrangement. The distal portion of the bract, which was triangular in cross section, was a long tapering process that occupied about the distal third of the total length. Its proximal end was demarcated by a pair of lateral teeth and a third tooth, on the inner side, was present.

The bracteal canal ran from close to the proximal end of the bract down to the tip of the long distal process. It penetrated into the mesogloea about half way down the distal process and ended below a small, terminal cupshaped indentation. No nematocysts were noted amongst the cells present in this indentation, but their presence could not be ruled out.

Although these bracts resembled the larval bracts of other species, they were by far the commonest of all the types of bract, which suggests that they may be present on all cormidia. At least three specimens were collected at Discovery St. 11794#27, and of the 100 bracts found 64 were Type A, 20 Type B, 2 Type C, and 14 Type D.

Type B bracts ( Figure 55 ) measured up to 12.6 mm in length and occurred in enantiomorphic pairs. They were distinctly asymmetrical, with a transverse ridge on the upper side, delimiting a thickened distal facet. On one side, the ridge continued out laterally, but less distinctly, to reach the side of the bract, while on the other side there was a lateral cusp. On either side, a branch of the transverse ridge ran distally, gradually becoming less distinct. A symmetrical pair of rounded lateral cusps was present distally, with a short distal process between them. A median ridge ran, on the upper surface, from just distal to the transverse ridge to the distal end of the bract. Patches of ectodermal cells were present on the upper side of the distal facet. On the lower surface there were two flaps. One transverse one at the same level as the transverse ridge, and a median longitudinal flap in the proximal half of the bract.

The bracteal canal commenced within the groove formed by the median flap on the lower side of the bract. It penetrated into the mesogloea, roughly on a level with the two lateral, distal cusps and ran obliquely up to end below a small cup-shape indentation on the upper side of the bract close to its distal extremity. Again, this indentation contained cells, some of which may have been nematocysts.

Type C bracts ( Figure 55 ) were the least common, and the two found with the type specimen probably represented an enantiomorphic pair, although they were by no means mirror images of each other. They measured up to 13.5 mm in length. An asymmetrically placed longitudinal ridge ran most of the length of the bract, but was indistinct proximally, where an additional, ridge ran out to one of a pair of proximal lateral cusps. Another pair of cusps, asymmetrically arranged, was present on the distal half of the bract. No patches of ectodermal cells were noted.

The bracteal canal arose in a central indentation at the proximal end of the bract and ran down to the tip of the distal process, having penetrated into the mesogloea slightly proximal to this process.

Type D bracts ( Figure 55 ) measured up to 13.1 mm in length, and formed enantiomorphic pairs. In outline they appeared almost symmetrical, and distally there were two pairs of rounded lateral cusps, and a short median ridge on the upper side. However, the bracteal canal did not run down the centre of the bract, but arose on the inner lateral margin of the bract at about one-third to one-half its length. It then ran down to the end of the pointed distal process. A single patch of ectodermal cells was found on the upper surface of one of the three bracts of this type found with the type specimen, but Pugh & Youngbluth (1988) did not state its position, and none was present on the Discovery material.

Tentillum. Each tentillum ( Figure 56) consisted of a long pedicel, a coiled cnidoband, usually with 6 coils, and a very long terminal filament that ended in a characteristically shaped terminal process. A cup-shaped process at the base of the cnidoband represented the vestigial involucrum. The cnidoband, in life deep orange-red in colour, contained a few stenoteles, usually confined to the first few coils, and numerous anisorhizas. The terminal filament and the proximal cupulate part of the terminal process contained acrophores and desmonemes, while the distal part of the latter was devoid of any nematocysts, but included numerous platelets.

Palpon: Long, thin, transparent, featureless palpons borne on long peduncles. Long palpacle. No sign of nematocysts on either structure.

Gonophores: Only developing male gonophores, of typical form, were found on the siphosomal stem of the holotype specimen, and no gonophores were present with the paratype.

Nectalia - stage: It is probable that the paratype specimen was at the Nectalia post-larval stage ( Figure 57 ), as the stem bore only a single gastrozooid and a few palpons together with seven nectophores and a Type B bract ( Figure 58 ). The pneumatophore was c. 1.3 mm in length and 0.3 mm in diameter, with the pneumatosaccus and gas gland clearly visible. It was borne on a long, narrow stalk, over 4 mm in length. Two other bracts were found in association with the specimen, which both resembled the Type A bract apart from the arrangement of the U-shaped transverse ridge. On the larger one ( Figure 58 ) the transverse ridge was incomplete, and there was a distinct flap on the lower and inner side of the bract. The bracteal canal entered the mesogloea at about half the length of the distal, triangular process and ended below a small cupulate process that contained small nematocysts. The canal was thickest proximally in the region of attachment, then thinned as it continued along the lower surface of the bract, and again as it penetrated into the mesogloea and ran to the distal end. The smaller of the loose bracts had a complete transverse ridge, with the proximal end of the upper, median, longitudinal ridge joining it.

At first glance, it appeared that the gastrozooid bore a few adult type tentilla, as they had a spiral cnidoband, but on closer examination, they were found to be somewhat different ( Figure 59 ). Although they possessed a long pedicel, a rudimentary involucrum, a coiled cnidoband with 4 turns, and a long, coiled terminal filament. However, the terminal process was not of the adult-type as it consisted of a long, hollow tube devoid, as far as could be seen, of nematocysts. The form of these tentilla is quite different from the larval tentilla found with other Halistemma species but, at present, we cannot be certain that this type of tentillum is a true larval one, or a stage in the development of the adult form.

Distribution: The two JSL specimens came from the Bahamas and were collected at depths of 628 and 563 m. Several specimens have been identified from recent Discovery collections:

(Continued)

Discovery Nects Bracts Date Latitude Longitude Depth range (m) Station N W

8517#01 93 165 14-Jun-1974 44°28.9' 12°51.7' 400–500

8507#42 29 41 02-Apr-1974 44°14.9'. 12°57.2' 900–1000

8507#55 26 21 07-Apr-1974 44°03' 12°46.8' 90–110

11120#02 71 64 27-May-1984 43°00.6 20°56.6' 100–550

10523#13 11 10 12-May-1982 4°54.1'S 0°23.7' 995–1100

Many of these specimens were collected in the region of the Porcupine Seabight and Goban Spur off Ireland (c. 49°N, 13°W), or at 47°N, 19– 20°W ( Pugh, 1990), and mainly in the 100–600 m depth range. One specimen, however, was collected in the Gulf of Guinea (c. 5°S, 0.4°W) between 995 and 1100 m. There are no other known published records, but Dr Dhugal Lindsay (personal communication) has caught a specimen of Halistemma transliratum in Sagami Bay , Japan. Thus the species appears to have quite a wide distribution in tropical, subtropical and temperate waters, and may exist at deeper depths in warmer waters throughout the World's oceans .

Figures

FIGURE 50. Halistemma transliratum Upper views of mature nectophore from holotype. Scale bar 3 mm. From Pugh & Youngbluth (1988).

FIGURE 51. Halistemma transliratum. Upper (A), lower (B) and lateral (C) views of mature nectophore from paratype specimen. For Legend see Figure 14. Scale bar 1 mm.

FIGURE 52. Halistemma transliratum. Young nectophore from holotype. A. upper, B. Lower, and C. Lateral views. Scale bar 3 mm. From Pugh & Youngbluth (1988).

FIGURE 53. Halistemma transliratum. Upper (left) and lower (right) views of a younger nectophore from the paratype specimen. Scale bar 1 mm.

FIGURE 54. Halistemma transliratum. Type A bracts—upper views. Scale 5 mm. Redrawn from Pugh & Youngbluth (1988).

FIGURE 55. Bracts of Halistemma transliratum. Upper views of Type B, Type C, and. Type D. Scale bars 3 mm. Redrawn from Pugh & Youngbluth (1988).

FIGURE 57. Halistemma transliratum Nectalia-stage paratype specimen from JSLII Dive 972. Scale bar 5 mm.

FIGURE 58. Halistemma transliratum. Nectalia-stage bract from JSLII Dive 972. Proximal to the left, distal to the right. Scale bar 5 mm.

FIGURE 59. Left. Young tentillum of Halistemma transliratum from the JSLII Dive 972 specimen. Scale bar 1 mm. Right. Detached terminal filament. Scale bar 0.5 mm.