Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Tuatara: Volume 12, Issue 3, November 1964

A Consideration of the Question “What are Whitebait?”

page 134

A Consideration of the Question “What are Whitebait?”


Each Spring, countless thousands of tiny fish migrate from the sea into the rivers of New Zealand. These fish are slender and transparent, about 50mm long, and on entry into fresh-water their stomachs are empty. They are caught near the river mouths during their upstream migration and are known to the fishermen as whitebait.

The term whitebait is used widely throughout the world, referring to small, usually marine, fishes. Graham (1956) reported that ‘English whitebait is made up mostly of young sprats, mixed with the young of shad, herrings, sticklebacks, gobies and shrimps. In Japan the young sea perch are called whitebait, and in Germany the young of various sea fishes go under that name. In Italy it is the same, but in each case, the fish are of a bony nature and inferior to the New Zealand whitebait’. In Tasmanian rivers the young of the aplochitonid Lovettia seali are caught and are known as Tasmanian whitebait.

In New Zealand the name whitebait usually refers to the migratory juvenile of Galaxias attenuatus. In some parts of the country and at different times of the year, usually towards the close of the whitebait season, the whitebait run includes the young and adults of Retropinna, juvenile Gobiomorphus, and, during the last century, reputedly Prototroctes. A second-class whitebait is sold on the Auckland fish market, and this is obtained from the sea. It includes the young of pilchards and other marine fishes and also juvenile Retropinna. On the whole the name whitebait in New Zealand refers to the young of Galaxias attenuatus. G. attenuatus is also found in south-eastern Australia, Tasmania and in South America. Although present in small numbers in some Tasmanian whitebait catches, G. attenuatus does not occur in quantities of commercial significance in any region other than New Zealand.


One of the subjects of discussion in the early study of New Zealand's fresh-water fish was the origin, nature and life-history of the New Zealand whitebait. The small transparent fishes migrating into the rivers of New Zealand were known to both page 135
Fig. I: Galaxias whitebait. (a) G. attenuatus (54mm. Length Caudal Fork); (b) G. brevipinnis (46mm. L.F.C.); (c) G. fasciatus(41mm. L.C.F.); (d) G. postvectis (40 mm. L.C.F.).

Fig. I: Galaxias whitebait. (a) G. attenuatus (54mm. Length Caudal Fork); (b) G. brevipinnis (46mm. L.F.C.); (c) G. fasciatus(41mm. L.C.F.); (d) G. postvectis (40 mm. L.C.F.).

page 136 Maori and Pakeha as a useful and easily obtained food, and because of this, its habits were the subject of discussion, and probably argument. That there was much confusion in identifying this fish in the early days can be inferred from the remarks of Powell (1869) in a paper entitled ‘On four fishes found in the Avon River, with a consideration of the question What is whitebait?’. Powell described the whitebait as the ‘young of the Galaxias commonly known as the smelt’ and presented a figure of what he knew as the adult whitebait, under the name G. fasciatus. The figure is quite a good representation of the adult G. attenuatus. In trying to determine the identity of whitebait, Powell had thus confused smelt (Retropinna sp.) and at least two species of Galaxias.

McKenzie (1902) expressed the opinion that the whitebait ‘are a distinct species and not the fry (young) of another fish’, whilst the rather remarkable opinion, that whitebait have been kept in captivity ‘until they developed into mullet’, comes from Gibson (1902). Hector (1902), as if to attempt a settlement of the controversy, commented that ‘the question of the true nature of the so-called “New Zealand whitebait” has been so fully worked out and published that it is hardly necessary to say more about it. It belongs to the genus Galaxias, of this minnow there are probably several species. G. attenuatus is the adult form of the true whitebait of New Zealand’. Despite the apparent finality in Hector's note, the controversy continued. Hope (1928) reported on some experiments he had carried out to ‘satisfy’ himself on the identity of whitebait. He kept live whitebait in captivity and found that they rapidly grew into inanga and he figured a series of growth stages of G. attenuatus from whitebait to adult. Even today there is still much controversy in the public mind about the nature of whitebait, and questions such as ‘What are whitebait?’, ‘What do they grow into?’, ‘Where do they come from?’, ‘Where do they breed?’, frequently arise.

Present Opinion

Up to the present time, little has been known about the life-histories of the New Zealand Galaxiidae, but there has been general agreement amongst ichthyologists that whitebait are simply the juveniles of Galaxias attenuatus. A constantly lurking question, however, has been — do the young of some of the other lowland-dwelling Galaxiidae go to sea? The young of the two lake-dwelling Galaxiidae in New Zealand — G. lynx and G. koaro — both have whitebait-like shoaling juveniles, and this suggests that the same may be true for other species. Clarke (1899), when discussing a fish which he named G. robinsonii (= G. brevipinnis Gunther, see Stokell, 1954) remarked that ‘at occasional intervals I obtained a few isolated specimens of this fish amongst the fry page 137
Fig. II: Galaxias juveniles. (a) G. attenuatus (60mm. L.C.F.); (b) G. brevipinnis (65mm. L.C.F.); (c) G. fasciatus (76mm. L.C.F.); (d) G. postvectis (78mm. L.C.F.)

Fig. II: Galaxias juveniles. (a) G. attenuatus (60mm. L.C.F.); (b) G. brevipinnis (65mm. L.C.F.); (c) G. fasciatus (76mm. L.C.F.); (d) G. postvectis (78mm. L.C.F.)

page 138 of G. attenuatus, Retropinna richardsoni, Eleotris gobioides and Prototroctes oxyrhynchus, the heterogeneous collection making up the so-called whitebait at the end of the inanga season in the Westland rivers, although strange to say, in my long experience I have never found amongst the collection, the fry of any of the larger indigenous Westland Galaxiidae’. It thus seems probable that Clarke considered that the young of G. brevipinnis migrate upstream from the sea in much the same manner as those of G. attenuatus.

Woods (1963) answered the question ‘What are whitebait?’ with the remark that they are the ‘transparent free-swimming and shoaling juveniles of at least five species of Galaxias’, and he listed G. attenuatus, G. lynx, G. koaro, G. campbelli and an unidentified Galaxias from a North Auckland lake. Woods' definition of whitebait as above would seem reasonable, and the difference between shoaling in the sea and shoaling in lakes is not of much significance in such a definition. Woods suggested that the young of the lowland Galaxias, G. brevipinnis and the young of the kokopu, G. fasciatus, G. postvectis and G. argenteus may be whitebait-like in form and habits.

Present Investigations

During the 1963 whitebait season samples of whitebait migrating from the sea were taken from the commercial catch of the Awarua River, South Westland. These samples were collected from the 1st of September until the 21st of November at the author's request by Mr. E. R. Midgely, a professional whitebaiter in the river. Examination showed that in those samples taken between the 1st of October and the 21st of November a small percentage of the fish were not G. attenuatus whitebait. A total of 17 of these ‘non-G. attenuatus’ whitebait were present in 9 of the 17 samples. There were usually only one or two in any sample (making up roughly 3% of these samples) and they occurred fairly evenly through the period when they were present in the samples. From September 1963, until the time of writing (May, 1964), regular samples of G. attenuatus were obtained from a series of stations in the Waikanae River, about 30 miles north of Wellington, and several samples were taken from the Makara Stream, to the west of Wellington. In both these streams whitebait-like juvenile Galaxiidae, which were not G. attenuatus, were taken in small numbers.

These juvenile transparent fishes were easily distinguished from the G. attenuatus whitebait by a number of superficial characters. In the period during which the Awarua samples were being collected the average caudal fork length of the G. attenuatus whitebait varied between 51 and 54mm. and the overall range during this period was 47 to 59mm. By comparison, the fish page 139
Fig. III: Galaxias adults. (a) G. attenuatus (95mm. L.C.F.); (b) G. brevipinnis (137mm. L.C.F.).

Fig. III: Galaxias adults. (a) G. attenuatus (95mm. L.C.F.); (b) G. brevipinnis (137mm. L.C.F.).

page 140 which were not G. attenuatus ranged between 41 and 50 mm., i.e. they were generally smaller than the G. attenuatus whitebait. Perhaps the most obvious feature differentiating the two groups was the form and position of the pectoral fins. These fins in G. attenuatus are small and positioned laterally, with the lamina of the fin vertical. This is not always obvious, especially when the specimens have not been carefully preserved, but when G. attenuatus whitebait are compared with the ‘non-G. attenuatus’ specimens, a difference can be seen immediately. The pectoral fins of the ‘non-G. attenuatus’ whitebait are larger and are held horizontally, i.e. the lamina of the fin is more or less parallel to the ventral surface of the trunk and the fin is positioned low on the body. Further differences were also found in pigment patterns. The migratory juveniles of G. attenuatus have a series of very large melanophores along the dorsal surface of the trunk, anterior to the dorsal fin and sometimes reaching as far forward as the head. Similar melanophores were not found anterior to the dorsal fin in any of the ‘non-G. attenuatus’ whitebait. The anal fin of G. attenuatus is inserted directly below the dorsal fin insertion, but in some of the ‘other’ whitebait, the anal fin insertion is further back, even as far as the middle of the dorsal fin. In some of these fish the lower jaw shows marked recession. Many of the specimens taken in the Waikanae and Makara Streams showed development of pigmentation beyond the ‘whitebait’ stage, and on pigment pattern alone it was quite simple to determine which fish were G. attenuatus and which were not.

The occurrence of these fish migrating with G. attenuatus whitebait in the Awarua River indicates that they had been to sea and were migrating upstream again. Although migratory juveniles were not collected from the Waikanae River in an unpigmented state, evidence from these collections also suggested that the young fish had recently entered the river. Samples taken in September and October (1963) contained only G. attenuatus, but the November sample contained 12 whitebait of other species, the December sample 45, and the January (1964) sample 2. Their sudden appearance during October, suggested that the young fish had just entered the river and were making their way upstream. It can thus be inferred that some of the New Zealand Galaxiidae, apart from G. attenuatus, have marine dwelling whitebait.

The fauna of the Awarua River is not well known, but Mr. Midgely (pers. comm.) reported that he has taken G. argenteus and G. fasciatus from the Awarua River. One of the fish in the Awarua whitebait samples could be identified as G. brevipinnis from its general form and pigmentation. In the Waikanae River system all the above species are present, and, in addition, G. postvectis is found. These are thus the Galaxias species which may have marine-dwelling whitebait.

page 141
Fig IV: Galaxias adults. (a) G. fasciatus (176mm. L.C.F.); (b) G. postveetis (205mm. L.C.F.).

Fig IV: Galaxias adults. (a) G. fasciatus (176mm. L.C.F.); (b) G. postveetis (205mm. L.C.F.).

page 142


The identification of these young fish presented considerable difficulty. The New Zealand species of Galaxias were reviewed recently (Stokell, 1949, 1954, 1959) but although the descriptions suffice to separate the species as adults, they are not adequate to identify the juvenile fishes where adult pigmentation and body proportions have not been attained. While there is no question of the distinctness of the three kokopu species — G. argenteus (Gmelin). G. fasciatus Gray, G. postvectis Clarke — the most obvious and useful diagnostic characters are colour patterns and details of dentition and body proportions. These are either undeveloped or difficult to distinguish in the smaller juveniles.

For all the fish captured, details of pigmentation pattern, general body characters like relative jaw lengths, the position of the anal fin relative to the dorsal fin, were recorded and the specimens were cleared in 3% KOH and stained with alizarin dye. Counts were made of the vertebrae and fin rays.

In G. attenuatus whitebait the jaws are equal, the anal fin is inserted directly below the dorsal fin insertion, and the base of the anal fin is much longer than that of the dorsal. Meristic details from the specimens studied were as follows:— vertebrae 59-64; pectoral fin rays 11-15; dorsal fin rays ii-iv, 9-12; anal fin rays ii-iv, 14-18 (Fig Ia).

One of the other species of fish in the Awarua samples was more easily identified than the others. It was very slender, had a flattened head and tapered snout, and the lower jaw was much shorter than the upper. The anal fin was positioned well back from the insertion of the dorsal fin. Details of the body form and pigmentation of fishes of this type which had spent some time in fresh-water indicated that the species was G. brevipinnis, an identification confirmed by examination of meristic characters. These were as follows:— vertebrae 58-62; pectoral rays 11-15; dorsal rays ii-iv, 9-11; anal rays i-iv, 10-13; anal fin inserted below 7-10th dorsal rays (undeveloped rays counted). Meristic characters which separated G. brevipinnis from the other fish were the high vertebral number, the high number of rays in the pectoral fin and the relatively short anal fin. Any doubts about the identification of these fish as G. brevipinnis were dispelled when the characters of the fresh-run Awarua fishes were found to agree with those of the more highly developed and pigmented juveniles from the Waikanae samples. The pigmentation of the Waikanae fishes indicated that they were indeed G. brevipinnis (Fig Ib).

The remaining specimens were divisible into two groups, distinguishable by the relative position of the anal fin. In a large number of the fishes the anal insertion was directly below and in line with the dorsal insertion, but in others the anal fin was page 143 set back from the dorsal insertion. Superficially, these two groups were otherwise similar. The presence in the samples of fishes with developed pigmentation further assisted identification. The fishes with the anal fin directly below the dorsal proved to be G. fasciatus.

The characters of G. fasciatus whitebait are these:— jaws about equal or lower slightly shorter; anal insertion below dorsal insertion; head rounded with snout blunt. Meristic characters:— vertebrae 56-59, pectoral rays 11-14, dorsal rays i-iv, 9-11; anal rays iii-v, 12-14. Many of the more highly developed of these fishes showed colour patterning definitely identifying them as G. fasciatus (Fig. Ic).

The other of these types was the juveniles of G. postvectis. They showed the shortening of the lower jaw which is typical for G. postvectis but is not found in the other kokopu. The finding of a specimen in this group in the adult habitat of G. postvectis, some miles upstream from where most of the samples came, pointed fairly conclusively to their being G. postvectis. The characters of juveniles of this species are:— marked shortening of the lower jaw, snout otherwise fairly blunt; trunk stout; anal fin set back from insertion of dorsal fin. Meristic characters:— vertebrae 57-60; pectoral rays 11-15, dorsal rays iii-iv. 9-12; anal rays iii-iv, 12-14. The anal fin takes its insertion below the 2nd-7th ray of dorsal fin (Fig. Id).

G. argenteus, the third kokopu species, is obviously closely related to G. fasciatus and G. postvectis, and because the latter two species have a marine whitebait stage it would seem highly probable that G. argenteus would have the same basic life-history pattern. To date, no whitebait-like fish definitely identifiable as G. argenteus have been seen by the author. However, one small Galaxias whitebait with undeveloped pigmentation was taken in the Makara Stream towards the end of November 1963, and differed from other examples in being much shorter (37mm. length caudal fork) and much more slender. The lower jaw of this fish was a little shorter than the upper and the anal insertion was set back a little from the dorsal insertion. Counts of fin rays and other characters agreed fairly well with those of G. argenteus, but until further specimens identifiable with this species are obtained the presence of a marine whitebait for G. argenteus cannot be confirmed. The small specimen known to the author and definitely identifiable as G. argenteus was 48mm. long and showed a heavy covering of melanophores with a few pale vertical bands on the sides of the trunk.

Growth and Pigmentation

As the juveniles grow larger, their pigmentation becomes more intense. In all the species discussed there is initially an overal page 144 covering of melanophores to give the fish a drab grey appearance.

In G. attenuatus this pigmentation rapidly becomes broken up. Initially the trunk becomes mottled dorsally and the melanophores become concentrated along the myotomes, especially ventro-laterally (Fig IIa). Intensification of the mottling leads to the typical adult colour pattern (Fig. IIIa).

In G. brevipinnis there is also concentration of the melanophores along the myotomes and this develops into alternating light and dark V-shaped bands on the lateral trunk. The bands eventually become broken up to form a distinctive blotching pattern, the markings very irregular but clearly definited (Fig. IIb). In the adult G. brevipinnis, the banding becomes even more broken and the pattern is better described as speckling (Fig. IIIb).

After attaining a general covering of melanophores, the whitebait of G. fasciatus develop myotomal banding of rather indistinct nature and then vertical paler bands form along the lateral trunk. The bands are usually broad and distinctive, but as the fish grows larger they become relatively narrower and more numerous and extend over the dorsal surface of the trunk, where a pattern of rings may form as the bands from each side sub-divide and intersect (Fig. IIc). In adult G. fasciatus the bands may become divided by sub-banding and confined to the posterior third of the trunk (Fig. IVa).

After developing myotomal concentration of the melanophores the young G. postvectis do not develop bands as distinct as in G. attenuatus, G. fasciatus and G. brevipinnis. The fish remains drab in appearance and rather obscure dusky-brown bands develop along the sides of the trunk (Fig. IId). The adult likewise has no clear bands, the colour pattern comprising rather indistinct marbling on a slightly paler background (Fig. IVb).

Behaviour and Migration

These ‘non-G. attenuatus’ whitebait were found to shoal freely with G. attenuatus whitebait, although they always made up a minority in the shoals. As an adult. G. brevipinnis is unable to swim freely in mid-water and the fish must constantly struggle to maintain a mid-water position. This difficulty is less marked in the younger juveniles, but rapidly becomes more evident in the fish as they grow larger. Young G. brevipinnis which had considerable difficulty in maintaining a mid-water position were nevertheless found to shoal with the whitebait of other species. The young of G. postvectis and G. fasciatus were better able to swim in mid-water and had no apparent difficulty in shoaling with G. attenuatus. Although G. attenuatus remains a shoaling species throughout life, the other species soon forsake this behaviour and become solitary.

page 145

Many of the G. fasciatus, G. postvectis and G. brevipinnis whitebait were captured below a weir in the Waikanae River. This weir is about six feet high with a continuous torrent of water pouring across it forming a substantial pool at its foot. Shoals of fish comprising G. attenuatus, G. brevipinnis, G. fasciatus and G. postvectis were found to accumulate below the weir and to mill about there. During nine months of sampling in the Waikanae River only one example of G. attenuatus was found to have made its way over the weir. The other species, however, were highly successful in surmounting the weir, and were found to occur in slack water immediately above it, as if resting before moving further upstream into the adult habitat. G. brevipinnis, G. postvectis and G. fasciatus were all found to occur in the upstream tributaries of the Waikanae River system.

The failure of G. attenuatus to surmount the weir to any significant degree may be related to the nature of the pectoral fins. The young of the other three species, with their pectoral fins held horizontally, are adept at climbing up glass surfaces by eel-like wriggling, probably using the surfaces of the pectoral fins against the surface of the glass to increase traction. The pectoral fins of G. attenuatus, being disposed higher on the lateral trunk, are of little or no assistance in climbing, and G. attenuatus has not been observed climbing the walls of the aquaria in the way the other species have.


Under Woods' (1963) definition, that whitebait are the transparent shoaling juveniles of species of Galaxias, the following New Zealand species of this genus can be said to have a whitebait stage in the life history:— G. attenuatus, G. postvectis, G. brevipinnis and G. fasciatus, all having a marine whitebait which occurs in the commercial whitebait catch; and G. koaro and G. lynx, having lake-dwelling whitebait. Scott (1941) recorded that G. truttaceus also has a marine whitebait in Tasmania. Woods asserted that G. campbelli has a marine whitebait stage, and this is likely from its obvious relationship to G. brevipinnis and from its presence on the isolated sub-antarctic islands. However, there is as yet no documented evidence that G. campbelli does have a marine whitebait. It remains probable but uncertain that G. argenteus also has a marine-dwelling whitebait stage in the life-history.

The ‘whitebait’ of the fisherman is thus primarily G. attenuatus, with G. fasciatus, G. postvectis, G. brevipinnis and probably G. argenteus making up a small proportion of the catch.


The author wishes to thank Mr. Eric Midgely for his willing co-operation in providing samples from his whitebait catch in the page 146 Awarua River; also Mr. K. F. Maynard and Mr. G. A. Eldon for assistance in collection of specimens and data and Mr. L. J. Paul for reading and criticising the manuscript.

Literature Cited

Clarke, F. E., 1899. Notes on Galaxiidae, more especially those of the western slopes, with descriptions of new species. Trans. Proc. N.Z. Inst. 31: 78-91, 5 figs.

Gibson, E., 1902. Notes on New Zealand whitebait. Trans. Proc. N.Z. Inst. 35: 311.

Graham, D. H., 1956. Treasury of New Zealand fishes.Wellington, Reed, 424pp, figs, 2nd Ed.

Hector, J., 1902. Notes on New Zealand whitebait. Trans. Proc. N.Z. Inst. 35.5: 312-319.

Hope, D., 1927-28. Whitebait (Galaxias attenuatus); growth and value at trout food. Trans. Proc. N.Z. Inst. 58: 389-391, 2pl.

McKenzie, A. J., 1902. Notes on the whitebait of New Zealand. Trans. Proc. N.Z. Inst., 35: 309-310.

Powell, L., 1869. On four fishes commonly found in the River Avon, with a consideration of the question, what is whitebait. Trans. Proc. N.Z. Inst. 2: 84-86, 2 figs.

Scott, E. O. G., 1941. Observations on fishes of the family Galaxiidae III. Pap. Proc. Roy. Soc. Tasm. 1940: 55-69, 17 figs.

Stokell, G., 1949. The systematic arrangement of the New Zealand Galaxiidae. II. Specific classification. Trans. Roy. Soc. N.Z. 77(4): 472-496, 12 fiigs.

Stokell, G., 1954. Contributions to galaxiid taxonomy. Trans. Roy. Soc. N.Z. 82(2): 411-518, 2 figs.

Stokell, G., 1959. Notes on galaxiids and eleotrids with descriptions of new species. Trans. Roy. Soc. N.Z., 87 (3 and 4): 265-269, 2pl.

Woods, C. S., 1963. Nafive and introduced fresh-water fishes. Wellington, Reed, 64pp, 45 figs.