SOME POPULATIONS OF THE AUSTRALIAN "", trutta by the Late W. S. air bridge*

ABSTRACT Australian salmon occur in large shoals, usually A preliminary account is glvell of the lnvesti atioll close inshore. They have the habit of coming close of thc Australun Salmon (Arripis truth) p~?cipaf ly as into the beach while in these shoals, especially in an examination of the distribution (m a1 hfephases) certain localities, and it is then that the fishermen of different stocks of the species along the Australian catch them. coast, and the taxononnc status of thesc stocks. ,Mor- bolo iml and othcr evidence 1s presented to support an This is usually regarded as an onshore species. EyptLis that the stocks constitute distinct ncs with Shoals are not commonly sighted beyond the conti- separate breeding grounds. nental shelf; but there are reports, apparently reliable, The Australian salmon is a percomorph, and of salmon shoals being sighted, and specimens caught, not in any way related to the . The well out in the , popular name was presumably given to it by the Figure 1 shows the general world distribution early settlers in this country, on the strength of some of the species. In it extends up the east superficial resemblances to the true and and west coasts to about latitude 30°S, though . occasionally it is taken very much further north. The present paper is only a progress report. The There is a large* alld apparently quite untapped, results are not yet fully tested and definite, but it population of in New It Occurs was thought that delegates might be interested as also in the waters around and Lord much in the lines along which the work is progress- HOwe Island. ing as in the final results. The story which is When the present investigation was started in emerging is, as a matter of fact, somewhat un- 1947 it was believed-for want of contrary evidence expected; and it is of considerable interest to workers --that the dmon of south-east Australia were prob- in Australia. ably all of one stock, probably distinct from the The subject is that of raciation. salmon of south-west Australia. It was believed that the Western Australian fish were in some way In the so-called Australian "Sdmon" we have different from those of the south-east because of a common, 'pecies, of which there are geographical separation, and because the salmon of several very full taxonomic descriptions. Yet when the south-west are very much larger, and the scales we come to investigate it, we find that an apparently indicate a very faster growth-rate. homogeneous and continuous stock is divisible into two distinct subspecies; and one of those subspecies Thc first and most important discovery of the is further divisible into two separate popu- present investigation was that the salmon from south- lations. ~h~ morphologic differences which we have west Australia have fewer gillrakers than the sal- found between the populations are of standard taxo- mon New South The Western nomic characters, which however the early taxom salmon have 25 to 30 gillrakers, and the New South nomists apparently overlooked. Wales salmon have 33 to 3 8. Each group shows a spread of about 5 gillrakers, but there is no overlap An idea of the appearance of the fish can be and, in fact, there is a sizeable gap between the two got from Plate 2 (upper). groups. In general the species is used for canning. It This shows that there is indeed a definite dif- is not greatly fancied as fresh fish. ference between the salmon of the west and of the There are two main fishing centres-namely in east. the south-east and the south-west corners of Austra- Following quickly on this discovery it was La. The fishery in the south-east is capable of pro- found that some of the iuvenile salmon from Tasma- ducing about 4 million pounds a year. The fishery nian and F'ictorian waters also belonged to the group on the south-west is of quite recent development, with the low count. In other words, they were and last Year the production was 5 $5 million lb. distinct from the salmon of . * Late Blologat. C.S.I.11.0. Division of FFllerles, Australia, (Hereafter, these groups will be referred to as the The next step in this work was to obtain sam- western and eastern subspecies-the western sub- ples of salmon from and from Lord species has 25 to 30 gillrakers, and the eastern bas Howe Island. It had been effectively established by jj to 3 8. These juveniles of the western subspecies the work on the gillrakers that there was no appre- which arc found in the south-east of the continent, ciable interchange between the stocks supporting the were fish of from 6 cm. length (mere fingerlings) fisheries of the south-east and of the south-west, but up to 30 or 40 cm. in length (at which size they it was possible that there was interchange between are severaI years old). It was found that in Tasma- New Zealand and New South Wales waters. nian and Victorian waters, and in Bass Strait gener- Two good samples of New Zealand salmon ally, the two subspecies often occurred mixed in the (about $0 fish) were obtained; and a small sample one shoal. Such shoals might be preponderantly of was obtained frim (about 10 fish). one or the other subspecies. Unfortunately it has so far been impossible to get In Figure 1, the distribution of the eastern sub- any fish from Norfolk Island. The samples obtained species is distinguished from that of the western were compared with New South Wales specimens subspecies by hatching, and the area where the mixed by means of 18 proportional measurements, ahd by shoals occurs is shown similarly. There is no inform- counting scales, fin-rays, vertebrae and so on. ation yet from the west coast of . Though Though there were so few in the sample from it is known that there are salmon there, it is not Lord Howe Island, it was apparent that one had to known to which subspecies they belong. The zone do here with a very different looking fish. The head of mixing of the two subspecies in Victorian waters- was shorter than in the others, and the body was between Wilson's Promontory and the south-east much stouter; in addition, the flesh was extremely oily. corner of the continent-is of some interest. It hap- Oil is very low in Australian salmon at all times of pens 'to be the approximate division between two the year, and was low in both of the New Zealand previously well-established zoogeographical areas. samples. All these characteristics of the Lord Howe Island fish point to good feeding and fast The South Australian specimens came from the and that conclusion is confirmed by the scales. Of "Gulf" area. It was found that they agreed very more importance, it was found that the gillrakers closely with the samples from Western Australia in were substantially different in number in comparison everything but the girth scale count-and that ap- with the finding for the eastern subspecies of Austra- parent difference may prove not to be significant lia: they varied between 36 and 41, and this feature when larger samples are obtained. seems to distinguish the Lord Howe Island salmon Even more significant perhaps, was the growth- from all others. rate of the South Australian salmon, as shown by The New Zealand salmon were much less distinct their scales. The faster growth and larger size of the from the New South Wales stock than were the West Australian salmon, compared with the New Lord Howe Island salmon. However it is almost South Wales salmon, has been previously mentioned. certain that they are distinct from the New South Though the South Australian fish were young- Wales salmon, because there was a difference of about mostly 2 and 3 year 014s-their scales showed a 1 scale in the girth and about 0.34 in the dorsal ray powth-rate that agrecd very well with the West counts. The material is not yet full enough to Australian salmon. eliminate any chance that sex and year-group may We may now examine the tables of gillraker be affecting these meristic counts; but the data do counts, girth scales and dorsal rays. These figures show that if either variable has an effect, it is not do not embrace all the material available, but have of such magnitude as to invalidate the differences in been drawn up simply for the purpose of illustration. the girth scale count. It must be emphasized that the work is to be regarded The proportional measurements have not yet as very incomplete as yet. been tested statistically, but there appears to be The samples in this. table have been divided up absolutely no difference between the New Zealand into tht possible or suspected discrete populations. and the New South Wales salmon. The two are undoubtedly very closely allied. There are consider- It will be seen that, as far as gillrakers are able differences, as stated above, between the New concerned, there is no doubt about the break be- South Wales and the Lord Howe Island salmon, tween the eastern and western subspecies. There is mainly involving the short head and the great girth probably no significant difference between New Zea- of the Lord Nowe Island fish. Such differences are land and New South Wales, nor between Western not necessarily genetic, but might be merely the Australia, and south-east Australia. effect of particularly good conditions in the waters There is a single instance of a salmon with a around Lord Howe Island. However there are other high gillraker count taken in Western Australia. differences, such as the length of .the and fin and Presumably there is some admixture between the two the diamekr of the eye, which do not seem to be subspecies. They have very similar habits, and no merely the effect of good feeding, but probably mark doubt cross-breeding might occur between isolated off the Lord Howe Island salmon as a separate breed- specimens and the main body of fish. ing population from the other groups studied here. These tables also give the distribution of the two Further samples were obtained from Western other meristic features which show variability. Australia and, after some difficulty, several dozen Taking the dorsal rays first, the material ob- specimens from South Australia. viously falls into two groups, corresponding with the It should be pointed out that all along this study two subspecies distinguishable from the gillrakers. has been hampered by the fact that this species has The girth scales, however, are separable into a very wide distribution. It is usually sporadic in three or four groups. The West Australian counts its appearance, and its distribution covers some may or may not be significantly different from those completely uninhabited stretches of coastline. of South Australia, but are significantly different As previous:^ stated, there are two centres of from the eastern sub-species. The New Zealand the fishery, south-east and south-west. In the samples also differ significantly from those from south-west, the roads all run inland and not coast- Australia. wise, and the coastline is desolate. East of Esperance Again it is necessary to stress that this material there is ncthing but desert. From Ade!aide to Mel- has not yet been fully tested to determine whether bourne is also sparsdy inhabited, with no salmon age--or year-groups or sex-are causing slight dif- fishery. The west ccast of Tasmania is very desolate, ference between means. But it is known that these with no fishery. There are therefore many gaps factors are n3t causing large differences, such as are in the records of the species which must be filled apparent in these tables. in by supposition, or at times by fewer samples than Th: other obvious meristic features were also one would Iike. counted-vertebrae, and the other fin rays and spines-but all showed great constancy and little spawned somewhere in the vicinity. The set of the likelihood of being useful. Some other scale counts currents in this region is to the easc. By analogy were madeabove the lateral line, and around the with the situation in the east, it is tempting to sup- caudal peduncle; they seemed to show parallel varia- pose that these fry have been carried to the south tion with the girth scales, but because the girth and east from somewhere on the South Australian scales could be counted with greater certainty, the coast. Since there is no fishery for salmon along other scale counts were neglected. this part of the ccast a!l this must be rather supposi- So far, this paper has been concerned mainly tional. But that is certainly what it looks like. with the taxonomic characters by which the various There must be a large population of adults of the subspecies and populations may be revealed in this western subspecies somewhere on this part of the species. But there is a mass of rather scattered coast, to the west, to account for the large numbers information concerning habits, migrations, distribu- of fry in Victorian waters and around Tasmania. tion of young, and so on, which all fails into place It appears therefore that when they reach if interpreted in the light of the foregoing results. maturity, or earlier, the mixed shoals of juveniles First of all, in New Zealand it is known there which occur round Tasmania split up into their are large numbers of salmon fry which are seen at respective subspecies, the members of the eastern times in shallow water. This is pretty conclusive subspecies moving to the north and east, and those evidence that the salmon in New Zealand of the western subspecies moving to the north and waters. The same is probably true of Lord Howe west. Island, where young salmon are taken from time In Western Australia the behaviour of the sal- to time. Since the salmon are also known to spawn mon is strikingly parallel to that of the eastern sub- in New South Wales waters, this is confirmatory species, as we know it in New South Wales and evidence that the New Zealand salmon are distinct Tasmanian waters. and independent of the New South Wales stoclr. As There is a winter fishery at Hopetoun and mentioned earlier, there are reports of salmon shoals Esperance; only a few of the fish are taken west of half way across the Tasman Sea, and it may be that Hopetoun. In summer, with developing roe, the there is some exchange between the populations. fish make a westward movement and spawn in the The adults of the eastern subspecies undertake Cape Leeuwin-Naturaliste area in April and May. a rather scattered migratory movement. They Subsequently there follows a return movement of the spend winter and spring in New South Wales, and in fingerlings and first year fish. A few occur around summer and autumn move south. They spawn off the south-west, but the real concentration is in the southeast corner and apparently move further Esperance area and perhaps eastward of it. These south, but subsequently move north. The fry are movements are to be considered in connection with carried south in the current. The Tasmanian waters the current flow of this area. are the nursery ground for this subspecies; there are Now what is particularly interesting is that no juveniles in New South Wales waters. half-grown salmon are almost unknown in Western But in Tasmanian waters and in Bass Strait Australia. Some shoals of three and four year old there are also large numbers of young of the western fish put in a brief appearance at Esperance in the subspecies-passibly they predominate. As men- summer. But they are never taken in any numbers tioned earlier, the two subspecies often occur in Tas- around the south-west corner. It will be remem- manian waters as mixed shoals-one or other sub- bered that it was stated earlier that there was no species predominating, or else the two occur about fishing in all the area of the Great Australian Bight equally. There is a striking difference in the growth- east of Esperance. It seems very likely that these rate shown by the two subspecies, even at this stage- juvenile salmon are distributed through the Bight, the western subspecies appears to grow almost twice and perhaps even further east. as fast as the eastern subspecies, in spite of the fact that the fish seem to be living under almost identical The similarity of the samples of two and three conditions. The difference in growth-rate is so year old salmon from South Australia therefore be- marked that it may be possib!e later to distinguish comes important. These two age-groups are com- between members of the two subspecies on this monly caught in the Gulfs of South Australia. That character alone, without reference to the gillrakers. is highly suggestive. It seems that the salmon of this area (South Australia) are the juveniles of the In Victorian waters the juveniles appear to be stock of large, fast-grown salmon upon which the a!most all members of the western subspecies. West Australian fishery is dependent; just as the Now the origin of these juveniles of the western stocks of salmon in Tasmanian waters include the species is obscure. They occur in fairly large num- juveniles upon which the fishery of New Sovth bers as mere fingerlings. They must, theref~re,be Wales is dependent. To what population then do the juveniles of the The Tommy Ruff is a considerably smaller western subspecies, which are found in Victorian and species-the two have not been magnified to an equal Tasmanian waters, belong? They have no connec- degree in the plate. It will be seen that this species tion with the New South Wales fish. Where are the is morphologically very different from the salmon. adults that give rise to them? They can hardly have The eye is larger, the snout is shorter, thc caudal been distributed from Western Australia. peduncle is narrower. Also the scale counts and There is, as a matter of fact, no record whence fin-ray counts are very different from any of the these juveniles can have come. populations of salmon. The South Australian-Victorian stretch of All this is in contrast with the salmon-be- coast, as mentioned before, is known to be populated tween the populations of which there is a high degree with salmon, but is virtually unfished. It seems of uniformity in all but one or two characters. that one must postulate at least one further popu- If, therefore, one were to split the salmon- lation which spawns somewhere in this region, and Arripis tmtta, into a number of species, then the whose fry are carried to the south-east, to be found Tommy Ruff-Arrii)is georgian~ts,would have to be later in Victorian and Tasmanian waters. A hypo- removed from the genus and a new genus created thetical boundary between this presumed population, for it, which would then be a monotypic genus. and the supposed eastern limit of the south-west It seems best, then, all things considered, to Australian -population, - has therefore been dotted in. leave the salmon as a single species, but to split it The extent of present ignorance about this popu- up into two subspecies, in each of which there is lation has been stressed, but the fact must not be perhaps more than one population. lost sight of that there arc these juvenile members GIRTH SCALES of the western subspecies, occurring in very large numbers, in Victorian and Tasmanian onshore waters; 26 27 28 29 30 31 32 and one must somehow account for them. Our -- ...... - present ignorance is due chiefly to the lack of Western Australia. 1 5 12- 1 observations along the South Australian-Victorian South Australia. 2 -14 14 4 stretch of coastline. South east Australia This account completes the picture so far as it -westem has been worked out. It can be described partly as subspecies, well-established fact, partly as a working hypothesis, --eastern which will guide and direct future work. The Iimits sol)spccics. . 1 10 31 5 2 of the stock on which the New South Wales fishery Ncw Zealand. )I 916-8 1 is based have been demarcated; and progress has been Lorcl Howe Island. 8i 471 made in demarcating the limits of the stock of the - south-western fishery. DORSAL RAYS There is just one further point. It may have occurred to some readers that the two subspecies . .. would be better described as two, full, distinct species. Wcstern Australia. There is a striking difference in the gillrakers, and South Australia. probably in the rays of the . South east A~istral~n-vcstern No doubt these characters would be sufficient ubspecies. in some genera of fishes to create a new species. --eastern However, in the present case there is one other species subspecies. of the genus Arril~is-that is, Arripis georgianz~.~, New Zealand. the Tommy Ruff (see Plate 2). Lord Howe Island.

- -- .- -- - . . ------.-. - .- .. ? . - . Western Australia. 10 52 28 21 3 1 South Australia. South east Australia ,- -westem subspecies. I 4 29 39 16 1 -eastem - subspecies. 1 11 50 62 77' 25 - 2 New Zealand. 1522E 6 Lord Howe Island. 722351- ( 84 )