Other formats

    Adobe Portable Document Format file (facsimile images)   TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Zoology Publications from Victoria University of Wellington—Nos. 54 to 57

Development of the Lumpfish, Trachelochismus Melobesia (Pisces: Gobiesocidae)

page 1

Development of the Lumpfish, Trachelochismus Melobesia (Pisces: Gobiesocidae)

Publication of this paper is assisted by a grant from the Victoria University of Wellington Publications Fund.


In the Wellington area eggs of T. melobesia Phillipps, 1927 are laid during spring in clusters of three eggs to several hundred on the under-surface of permanently tide-covered stones. The eggs are oval with a flattened, adhesive base, many oil globules and average 1.65 mm × 1.35 mm. Egg clusters are attended by one parent until hatching. Under laboratory conditions this occurs at 12 days in temperatures of about 15°C. The emergent prolarva is 4.8 mm-5.5 mm standard length and has a prominent yolk sac with a single large oil globule. There are scattered melanophores above the gut and a conspicuous yellow tinge surrounds the brain, extending in the lateral muscle mass to the 8th post-anal somite. Larvae of 6.5 mm-7.85 mm standard length, and more advanced in development, occur in the Island Bay plankton from early September to late December.


Trachelochismus melobesia Phillipps, 1927, is an endemic lumpfish (Family Gobiesocidae) not uncommon on the shore in semi-exposed rock and rubble areas in the Cook Strait region of New Zealand. It is found throughout the inter-tidal zone but appears to be most abundant near the upper low-tide level. Adults rarely attain lengths of more than 30 mm s.l. Trachelochismus pinnulatus (Forster) is also found in areas inhabited by T. melobesia. However, the adult of T. pinnulatus (maximum size 71.2 mm s.l.) is larger than that of T. melobesia, and lacks the reddish-purple patch on the dorsal surface, a character which readily identifies the latter species. Flattened papillae occur on the central region of the sucker disc of T. melobesia (Fig. 4), but are absent in T. pinnulatus. The two species are further distinguished by fin ray counts (Briggs, 1955: 19–20) as follows: —

T. melobesia D 10 (9–11), A 8 (7–8), PI 23 (22–24), C 12.

T. pinnulatus D 8 (7–9), A 6 (5–7), PI 25 (24–26), C 12 (11–12).

The present study describes the egg and larval development of T. melobesia. The life history of T. pinnulatus has been studied by Coakley (1964), and Graham (1939, 1953) has briefly described the egg and prolarval stages of Diplocrepis puniceus (Richardson, 1846), a further gobiesocid known from the New Zealand seashore. The family is otherwise poorly known in respect of early life history. However, comprehensive accounts of the life history of the South American clingfish Gobiesox strumosus Cope, 1870 are given in Runyan (1961: 113–141) and Dovel (1963: 161–166).

page 2

Materials and Methods

On 20.10.69 four egg masses on stones were collected from the western shore of Lyall Bay and kept in rectangular plastic containers (30 × 28 × 13 cm) at the Island Bay Marine Laboratory. The water was changed daily and kept constantly aerated using an eddy-current air pump. The stones with eggs were wedged at a 45° angle against the sides of the containers, with the egg masses on the ventral surface. Porous "air stones" were placed beneath the eggs so that the streams of small bubbles issuing from the "air stones" flowed over the egg masses. This method prevented the accumulation of detritus on the eggs and ensured an adequate supply of oxygenated water. Despite these precautions many eggs died and these were removed each day to prevent the build up of fungi and bacteria. The eggs adhered very closely to the stones and those required for observation were very difficult to remove without damage to the egg membrane. Limited success was obtained by sliding a sharp scalpel between the egg and the stone and then pipetting the dislodged egg into a petri dish. The identity of the above eggs as those of T. melobesia was confirmed by comparing these with thirty eggs laid on 30.10.69 by one of three females (22, 23, 26 mm s.l.), which together with two males (27, 29 mm s.l.) were kept alive in an aquarium. Development of these eggs and those obtained earlier was studied and sketches of the eggs and larvae were made using a binocular microscope equipped with a grid eye-piece. A micrometer ocular was used for measuring eggs and larvae. Measurements of the larvae were based on those recommended in Hubbs and Lagler (1958: 24–26).

Larvae larger than 5.3 mm were obtained during the spring by making regular plankton tows approximately 100 yards offshore at Island Bay from a small boat. The net used was of standard conical design with a two-foot diameter opening and a mesh size of 500 microns.

Spawning and Development

In the Lyall Bay and Island Bay areas near Wellington T. melobesia spawns from early September to late December. Egg masses are found beneath smooth stones which remain covered with water at low-tide. The egg clusters range from 0.5 cm in diameter, containing as few as three eggs, to about 8 cm in diameter and containing several hundred eggs. The eggs are laid close together forming flat, irregular-shaped masses.

The larger egg clusters always contain groups of eggs which differ in colour, ranging from bright crimson to pale pink. This variation in colour represents the progressive depletion of the yolk supply in the eggs of each group as development advances. It is strongly suggested therefore that the eggs are laid at different times on the same stone. The pale yolked eggs, being the most advanced in development, are found mainly in the middle of the egg masses. However, the difference in development of eggs of some adjacent groups is relatively very small, indicating that these egg groups are laid no more than one to two days apart.

The absolute fecundity of T. melobesia is apparently very low, as determined from 10 gravid females each of which was found to contain only 20–30 ripe eggs. Considering the apparent low fecundity and the close development of the adjacent groups within a cluster it is concluded that more than one female is responsible for egg clusters that contain several hundred eggs.

page 3

Approximately 60 egg clusters were examined in rock pools in the Wellington area, of which about 80% had an adult T. melobesia in attendance. In these cases it was not clear whether this was a male or an immature female. When two adults were found close to the eggs the female could be distinguished by her characteristic pink and distended belly (the condition of a gravid female). Parental vigilance occurs commonly in littoral fish. Gibson (1969: 385) observes that predation on the eggs is prevented by the guardian activities of one of the parents, usually the male. Runyan (1961: 118) also states that "...eggs in good condition were usually accompanied by a male Gobiesox strumosus, that kept them constantly aerated by fanning anal and caudal fins...."

Development of the Egg

The mean dimensions of 100 eggs were 1.65 × 1.35 mm. The eggs are oval and dorso-ventrally depressed and are attached to the substrate by a flattened adhesive base. When first laid the eggs are bright crimson and are about 1.2 mm long, but within minutes they expand to a length of 1.65 mm, by uptake of water through the egg membrane. The yolk is central and subspherical and has a mean diameter of 1.3 mm. The yolk of the fertilised egg is a crimson-red colour, and contains between 10 and 100 oil globules, one of which becomes dominant during later development. Development to hatching (Figs. 1–2, nos. 1–10) took 12 days at a water temperature of 15°c.

Two hours (Fig. 1, no. 1). A single large cell appears from beneath the yolk and later expands and moves to one end of the yolk within one of the lateral perivitelline cavities.

Three hours (Fig. 1, no. 2). The first cleavage divides the cell dorso-ventrally relative to the egg base, and at right angles to the yolk surface. The cells formed round off and appear as two swellings protruding from one end of the yolk-sac.

Four hours (Fig. 1, no. 3). The four celled stage is produced by a second cleavage at right angles to the first. Oil globules are free to move and tend to migrate to the uppermost region of the yolk; hence their position depends on the attitude of the egg.

Five hours. In the eight celled stage the blastomeres are still in a single layer. The cells are arranged in two rows of four cells each.

Six hours. The sixteen celled stage is reached with the cells beginning to form a round blastodisc.

Twenty hours (Fig. 1, no. 4). The blastula is well formed and consists of a prominent cap of cells, beneath which lies the blastocoel.

Forty hours (Fig. 1, no. 5). At this stage the blastodisc has spread halfway around the yolk, in so doing obliterating the blastocoel. Epiboly is not obvious, except for the slight thickening of the lateral rim of the blastodisc.

Sixty-six hours (Fig. 1, no. 6). The blastodisc covers the entire yolk-sac. The primitive streak is well defined and lies deeply notched into the yolk, particularly in the cephalic region.

Eighty-five hours (Fig. 2, no. 7). The outline of the embryo is distinct, encircling half the yolk-sac. The myotome rudiments and the otic capsules are just visible. The main oil globule has become larger, apparently at the expense of the smaller oil droplets which have slowly decreased in size. Fore-brain develop ment is quite evident.

page 4

Fourth day. At this time the embryo encircles more than half of the yolk and the tail begins to extend free from the yolk-sac. The oil globule lies directly beneath the tail and remains here for the rest of the development. Optic vesicles contain outlines of the lens, and there is a pronounced enlargement of the hindbrain. Approximately 17 myomeres are present.

Fifth day (Fig. 2, no. 8). The total body length has increased slightly. Nearly all the myomeres are present posterior to the otic capsules and extending well down into the tail. At this time the heart begins to beat very faintly, but blood movement is seen only in the region of the heart. The heart lies well forward beneath the head and is obscured by the yolk.

Seventh day (Fig. 2, no. 9). The yolk is reduced considerably and the ventral aspect of the embryo faces upwards. The tail is well formed and is turned back to lie parallel with the body. The gut is formed and has scattered pigment spots on its upper surface. The heart beats regularly and strongly. Each beat of the heart sends a wave of movement through the yolk. Blood flows in the dorsal and ventral blood vessels, and the vitelline vessels are large and run laterally across the yolk. Pigment has appeared in the chorioid of the eye.

Twelfth day (Fig. 2, no. 10). The mouth and external nares are formed and the gut pigmentation has become darker. There is a further reduction in the amount of yolk. Pigmentation of the chorioid appears complete. The embryo is cramped within the egg and the tail arches forward to overlap the head. Just prior to hatching the embryo becomes agitated and begins to flex its tail. As a result of this activity the chorion is ruptured and the prolarva is released.

Prolarvae (Fig. 3, no. 11–12). Some of the larvae that hatched in the laboratory may have been induced to do so prematurely, as a result of disturbance. This is suggested by the variation in the amount of yolk present in each prolarva immediately after hatching. Prolarval length on hatching ranges from 4.8–5.5 mm. At this stage the gut is long and extends to the base of the 15th myomere. The upper peritoneum of the gut is covered with numerous stellate melanophores, extending from above the yolk to the vent. Posterior to the gut are 16–18 myomeres. Stellate melanophores are present in the myomeres just past the vent, but their number is variable. An obvious yellow tinge surrounds the braincase, and this extends through the myotomes above the gut to the 8th myomere past the vent. All larvae kept in the laboratory died within three days of hatching.

Larvae (Fig. 3, no. 13–14). At 6.5 mm the yolk-sac is almost absorbed and the gut reaches beyond the mid-length of the body. The oil globule is no longer visible. Pigmentation has changed very little except for an increase in the size of the melanophore at the base of the pectoral fin. The jaws are well formed and appear functional, although as yet are without well defined teeth. Two gills and 6–7 branchiostegal rays are visible on either side of the head. Two sucker buds lie ventral to the gills and heart. The longitudinal fin folds have slightly increased in size.

At 7.85 mm (Fig. 3, no. 15–16), the overall shape of the larva changes slightly. Essentially there is a flattening and broadening of the head, and an increase in the depth of the tail. In addition to the existing pigment pattern, numerous grey spots are scattered about the outer edges of the larvae. The dorsal, anal and caudal fins show signs of ray formation. The urostyle is curved upwards and extends a considerable distance into the caudal fin. The ventral sucker is well developed and functional. At this small size the larvae are able to cling to the page 5sides of glass jars. No specimens larger than 7.85 mm were caught in the plankton nets. Because larvae of this size have a fully functional sucker (Fig. 4, no. 18) it is suggested that at this stage they attach beneath stones.

The average measurements of twenty-five prolarvae are as follows:
Standard length 5.4 mm
Total length 5.7 mm
Head length 0.95 mm
Eye length 0.41 mm
Snout to anus 3.0 mm
Larval measurements are as follows:
Standard length 6.5 mm 7.85 mm
Predorsal length 3.2 mm> 5.0 mm
Head length 1.4 mm 2.15 mm
Depth of head 0.9 mm 1.21 mm
Snout length 0.31 mm 0.5 mm
Eye length 0.5 mm 0.57 mm
Interorbital distance 0.35 mm 0.55 mm
Head width 1.1 mm 1.4 mm


I would like to thank Dr. P. H. J. Castle, Department of Zoology, Victoria University of Wellington, for his constructive criticism of this paper.

Literature Cited

Briggs. J. C. 1955: A monograph of the clingfishes (Order Xenopterygii). Stanford Ichthyol. Bull. , 6, 224 pp., 114 fig., 1 tab.

Coakley, A. 1964: Life history and general biology of Trachelochismus pinnulatus (Forster) (Order Xenopterygii). Unpublished M.Sc. thesis, University of Canterbury (not seen).

Dovel, W. L. 1963: Larval development of clingfish, Gobiesox strumosus , 4.0 to 12.0 millimeters total length. Chesapeake Sci. , 4 (4): 161–166, 6 fig., 3 tab.

Gibson, R. N. 1969: The biology and behaviour of littoral fish. Oceanogr.Mar.Biol.Ann. Rev. , 7: 367–410, 5 fig., 3 tab.

Graham. D. H. 1939: Breeding habits of the fishes of Otago Harbour and adjacent seas. Trans. Proc. Roy. Soc. N.Z. , 69: 361–372. 7 pl.

Graham. D. H. 1953: Treasury of New Zealand fishes . (Wellington: A. H. & A. W. Reed.) 404 pp., 153 fig.

Hubbs, C. L. & Lagler, K. F. 1958: Fishes of the great lakes region . (Michigan: Cranbrook Institute of Science.) 213 pp., 251 fig., 6 tab.

Phillipps. W. J. 1927: Notes on New Zealand fishes. Trans. N.Z. Inst. , 58 (1, 2): 125–135, 6 pl.

Runyan, S. 1961: Early development of the clingfish. Gobiesox strumosus Cope. Chesapeake Sci. , 2 (3, 4): 113–141, 33 fig., 6 tab.

page 6
Fig. 1. Trachelochismus melobesia. No. 1: 2 hours; 2: 3 hours; 3: 4 hours; 4: 20 hours; 5: 40 hours; 6: 66 hours.

Fig. 1. Trachelochismus melobesia. No. 1: 2 hours; 2: 3 hours; 3: 4 hours; 4: 20 hours; 5: 40 hours; 6: 66 hours.

page 7
Fig. 2. Trachelochismus melobesia. No. 7: 85 hours; 8: 133 hours; 9: 7th day; 10: 12th day.

Fig. 2. Trachelochismus melobesia. No. 7: 85 hours; 8: 133 hours; 9: 7th day; 10: 12th day.

page 8
Fig. 3. Trachelochismus melobesia. Nos. 11 & 12: lateral and dorsal aspect of a pro-larva; 13 & 14: 6.5 mm larva; 15 & 16: 7.85 mm larva.

Fig. 3. Trachelochismus melobesia. Nos. 11 & 12: lateral and dorsal aspect of a pro-larva; 13 & 14: 6.5 mm larva; 15 & 16: 7.85 mm larva.

page 9
Fig. 4. Trachelochismus melobesia. No. 17: ventral view of the sucker of an adult (27 mm s.l.); 18: ventral view of the sucker of a larva (7.85 mm s.l.).

Fig. 4. Trachelochismus melobesia. No. 17: ventral view of the sucker of an adult (27 mm s.l.); 18: ventral view of the sucker of a larva (7.85 mm s.l.).

J. G. Ruck, B.Sc. (Hons),
Department of Zoology,
Victoria University of Wellington,
P.O. Box 196,
WELLINGTON, New Zealand.