Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 67, 1938
This text is also available in PDF
(2 MB) Opens in new window
– 341 –

Remarks on the Brown Trout Stock in Lake Ellesmere

[Read before the Canterbury Branch, May 5, 1937; received by the Editor, July 27, 1937; issued separately, December, 1937.]

In the introduction to a paper on nematode parasites of brown trout in Lake Ellesmere by Stokell (5, pp. 81–2), there are certain passages in the earlier pages which necessitate a further inquiry into their value as an interpretation of the conditions dealt with. Reference is made to the poor condition of trout taken from the lower Selwyn River by anglers at the opening of the 1932 season (commencing October 1, 1932). “The portion of the lake adjacent to the Selwyn, and the deep section of the river, together with all deep pools as far upstream as permanent water extended, contained large emaciated trout, and the presence of dead fish in all post mortem stages from specimens that had died within 48 hours to others that had been reduced to mere skeletons showed that these trout had been dying off for several months.

“A group of 20 specimens captured for examination had an average length of 18.2 inches, and an average condition factor of 31. All were mature Lake Ellesmere trout, and all except two were females. These invariably contained loose eggs in the body-cavity, in some instances lodged between the liver and the diaphragm (sic), and all showed some degree of injury to the ovarian membranes. There was frequent rupture of the peritoneum and occasional rupture of the lower intestinal mesentery, while in one specimen the posterior portion of the spleen, amounting to about half of the total bulk, was almost completely severed from the remainder. Severe inflammation usually existed in the ventral region, and where this was associated with peritoneal injury, there was considerable disintegration of tissue. Such fish could not have recovered, but a few specimens, notably the two males, showed only slight traces of recent inflammation, and appeared to be on the point of recovery.

“The disinclination of these trout to return to the lake in the customary manner is fully accounted for by their pathological condition, and the abnormal absence of floods during the winter of 1932 can be regarded only as instrumental in revealing, to general observation, conditions that were by no means peculiar to this season. The injuries described in no way resemble the effects of parasitism, but have obviously been inflicted in the process of extruding the eggs when the fish were being artificially spawned for cultural purposes. The incidents of a lake trout's existence include none that is capable of causing disruption of internal structure without disclosing some evidence of it in the form of scars or missing scales, but pressure applied through some non-rigid agency, such as human hands, would cause all of the injuries described without leaving an outward mark. It is impossible to explain the presence of eggs between the liver and the diaphragm except as the result of severe

– 342 –

abdominal pressure, or of the fish being held head downward when out of the water, under which circumstances the mesovarium could be expected to burst anteriorly, and the eggs gravitate to the position mentioned.

“As this material was useless for the present investigation, further collecting was deferred, but similar conditions existed at the opening of the 1933 season, and in 1934, although no injured fish were to be found in the river in consequence of their having been washed away by winter floods, many of those taken in the lake revealed evidence of recent injury to the mesovarium, or had fully developed eggs imbedded in the liver. It is noteworthy that such fish, which apparently represent the survivors of the winter's stripping, form a much greater part of anglers' catches than they did prior to 1930, and although the number of eggs collected each year has increased greatly since that date, the increase has not been proportionate to the increase in the percentage of injured fish. It would thus appear that trout are becoming fewer, and a greater percentage of the lake population is now being dealt with.

“It is necessary to add that the Lake Ellesmere material was heavily culled, and all specimens showing evidence of past injury were rejected, as were those which, by scale examination, proved to be not of the typical lake stock, but more or less interlopers.”

In examining the value of the above quotation, it is relevant to consider the spawning habits of trout and the stripping for egg-collection. Trout and salmon normally swim upstream while in the later stages of sexual maturation. They arrive at the spawning grounds and proceed to shed their sexual products. On becoming spent they normally proceed downstream, perhaps to their original habitat of the previous summer, perhaps to some other place. The marked change of response to current during this period is very striking. Trout commonly move into the Selwyn River from Lake Ellesmere, during maturation, from February onwards. The spawning period in that river lasts during some five months, beginning in late April. Usually, during the earlier part of the spawning run, there is permanent water extending some 12 miles from the lake, but sooner or later a continuous stream flows across the Canterbury Plain, which becomes broken in its intermediate portion after a dry season sets in. This break may last for weeks or months. Thus it is possible for fish to travel, on occasions, for many miles upstream from the lake, and also for them to become isolated in pools in the dry portion, and ultimately to die there.

It has been customary for many years for a trap to be inserted in the lower part of the Selwyn River some 4 ½ miles from the mouth. This has been done in the latter half of May and has served as an effective barrier to ascending fish, except during floods. Fish entering the trap, maturing for the first or subsequent time, have been stripped of eggs and milt, the fertilised eggs then being taken for incubation in a hatchery. The stripper separates ripe fish from unripe by pressing slightly on the abdomen, when a few eggs or a little milt may be forced out at the vent. Such as do not emit genital products are returned to the stream to ripen. The ripe fish are stripped by the pressure of the hand along the abdomen

– 343 –

towards the vent, the tail being gripped with the other hand and the fish being held, belly downwards, across the front of the stripper's body. An efficient stripper may extract about 90% of the eggs. It is understood by efficient strippers that only a ripe female will easily emit eggs on suitable pressure, therefore all other females are replaced. Stripped fish are returned to the water below the trap as they are now negatively rheotropic. It is extremely rarely that a stripped female returns upstream to the trap. Males may return several times as more milt is produced.

The process of ripening requires further study. However, from January onwards eggs increase in size, but the ovary appears earlier as a solid body. In April the eggs have reached close on their full size and can be easily moved separately.

Kendall (3) has described the membranes, ovaries, and oviducts of salmonoids, dealing particularly with Onchorhynchus spp. In these latter he describes the ovary as “in a general way boat-shaped with a decided list to port or starboard according to whether it is the left or right ovary” (loc. cit., p. 149). The list brings the exposed egg-surface in apposition to the mesovarium, while the suspensory membrane is continued round the surface of the ovary except on the free egg surface. In the summary (loc. cit., p. 200), Kendall states that in the three families mentioned (Salmonidae, Coregonidae, and Osmeridae), the ovaries “are structurally similar, consisting of a membraneous covering continuous with the mesovarium and almost completely enveloping the ovigerous stroma.” He further states, “a practically complete envelopment is formed by the position of the ovary and the mesovarium. The ovary is so inclined that the otherwise uncovered portion is protected by the mesovarium. The prolongation backward of the mesovariums and ovarian investments form the oviducts, which in the Salmonidae and Coregonidae are trough-like, open above, the inner wall consisting of the mesovarium and the free outer wall supported by the abdominal wall. Near the outlet the two troughs unite into one above the intestine at the point of termination of the dorsal mesentery. At a short distance from the genital orifice each outer wall of the common channel is deflected and is attached to the respective wall of the abdomen.” In a further paper (4, p. 44) he says, in reference to a quotation from Boulenger (1, pp. 567–8), “this statement, like every published statement pertaining to the subject, in anatomical, zoological, ichthyological and fish-cultural works, is simply the acceptance of the erroneous conclusions of earlier writers.” This quotation refers to the statement that the eggs of the Salmonidae fall into the body cavity. Bridge (2, p. 405) also says that the eggs are set free into the coelom. The relevant works of earlier writers to whom Kendall refers are reviewed by him in (3), where a list of references is given. Kendall states (4, p. 46), “when the ova ripen and burst from their inclosing follicles, instead of falling into the abdominal cavity they are deposited in the groove or angle between the upper edge of the laminae, or free egg surface, and the mesovarium covering; the inner end of each lamina is lower than the outer.” He further states that the eggs pass out along the “oviducal trough,” and later, “that the ripe eggs do not naturally fall into the body cavity,

– 344 –

and if by any means they are displaced into the cavity, they cannot be extruded or expressed” (4, p. 48).

It will be understood that Kendall believed that the descriptions given by him of the ovaries and membranes of Salmonidae were adequate to apply throughout the sexual cycle. That he was unjustified in this belief, and that his criticism of earlier writers was unjust, will be shown below by reference to certain dissections.

Some 14 female Salmo trutta L. in various stages of sexual maturity have been examined for the purpose of checking Kendall's conclusions. All, except one, were taken from streams flowing into Lake Ellesmere, and there was always the possibility that they had been used for the purpose of the artificial extraction of eggs for fish culture. The odd one was obtained from the upper part of the River Avon, Christchurch. Trapping of fish for egg collection in the Avon River has never taken place to the writer's knowledge, so that this fish can be regarded as never having been handled. The Ellesmere fish consisted of 6 which were ripening, 4 which were ripe and 3 which were naturally spent. They were taken from the Selwyn River and the Doyleston Drain. The Avon fish was ripe. The least mature of these fish had ovaries and membranes in very much the same condition as described by Kendall, that is, the free egg surface was apposed to the mesovarium, against which the free edges of

Picture icon

Text-figure 1—Diagrams representing the transverse face of an ovigerous lamina of Salmo trutta in various conditions through the year from one early summer to the next: (1) in quiescent state; (2) in late Autumn; (3) shortly before shedding eggs into the body cavity; (4) after eggs are shed, but before spawning: (5) in quiescent condition in late Winter or early Spring and approaching condition in Fig. 1. Imm. ov.: immature eggs in reserve for next spawning season; I.mes.ov.: lower edge or outer edge of mesovarium; mes.ov.: mesovarium; ov.: maturing eggs. The dots represent the general egg reserve. The stroma is not represented.

– 345 –

the ovigerous laminae abutted. With advancing maturity the free egg surface turned away from the mesovarium until it abutted against the peritoneum of the body wall and its outer edge turned downwards through approximately 180° producing a flattish vertical sheet. This change of relations appeared to take place by the growth in the length of the laminae and a relative decrease in their depth, thus tending to bring the eggs to the surface. The most advanced fish, not yet ripe, had almost the whole of the outer surface of the ovary consisting of free egg surface, except for a narrow lower margin of upturned mesovarial membrane.

In the ripening fish examined, the left ovary extended to the posterior end of the body cavity and the gut was squeezed between the two ovaries which were the most striking components of the abdominal viscera. The so-called oviducts of Kendall were not very obvious unless the specimen was put into water so as to allow of flotation of the parts. The ripe fish, in all cases, had shed their eggs into the body cavity. All the fish had been opened along the side, below the lateral line, by a careful cut from near the vent to near the pectoral girdle; this minimised the risk of internal disturbance. When the flap of body wall was reflected the eggs came into view, lying freely in the body cavity from one end to the other. The so-called oviducts of Kendall and the greater part of the intestine, were hidden by eggs and, whatever may be their function in other species of Salmonidae, they clearly play no significant part as oviducts in Salmo trutta L.

In these ripe fish, with eggs shed into the coelom and occupying all available space, the ovaries had already shrunken to varying extents, but were usually about the length of the quiescent condition, perhaps one-fifth of the length of the body cavity. All the ovaries but one still hung down as flat vertical sheets with the very slight lower margin of reflected mesovarium. The odd one had resumed its quiescent position where the free egg surface had become apposed to the mesovarium. Although the indications are that shedding of eggs from the ovary commences at the posterior end, the condition of the ripe fish examined suggests that in the brown trout all the eggs quickly become loose, perhaps by a rapid shrinking of the ovary. A “hard” fish, one from which eggs cannot easily be pressed, has large ovaries with unshed eggs. Such a fish may become “soft”—will easily emit eggs when pressed—in a few hours. The indication is, therefore, that egg-shedding into the coelom is somewhat rapid and is somehow bound up with the shrinking of the ovary to one quarter or one fifth of its immediately previous length, one third of its depth, or even less, and perhaps one quarter of its thickness.

Kendall must not have examined fish when they were ripe, or else the condition of ripe female brown trout in New Zealand is different from that in North America. This statement is justified on the ground that a wild fish, which had not previously been handled, had precisely the condition here described and not the condition which Kendall described as normal. If the observations here recorded are good and the interpretation sound, it would follow that much is unsound of what Kendall concludes as to the practical significance of his findings.

– 346 –

The question arises as to the meaning of the membranes which form the alleged oviducal groove. They may be simply regarded as extensions of the mesovarium which have established a connexion with the posterior mesentery. When the posterior mesentery has separated from the mid-dorsal line it has carried downwards the associated mesovaria, and the groove so formed is clearly the mechanical result of a growth process and has nothing immediately to do with oviposition. As the eggs lie in the body cavity they are postero-ventrally divided by the ventral mesentery, but are postero-dorsally continuous through the gap behind the dorsal mesentery. There is, in fact, nothing to prevent the movement of eggs from any part of the body-cavity into the genital ducts of the papilla. Kendall's figure 7 (3, p. 196) gives the impression that the mesovarial structures, especially the oviducal groove, are stout and rather rigid. Actually, in the brown trout, they are delicate and flaccid.

A specimen of Salvelinus fontinalis from a small lake in the upper Ashburton River Valley was in the process of ripening. The left ovary was smaller than the right and the free egg surface in both cases was apposed to the mesovarium. The mesovarial membrane of the left ovary, which was adjacent to the body wall, was undergoing a process of vertical shredding so that the eggs were exposed between the strands. The strands appeared to coincide with the ovigerous laminae. If that represents a normal case, it would appear that the eggs would dehisce into the body cavity. However, no other material was at hand to test the case.

Wild fish usually shed all their eggs, but a few, perhaps half a dozen, may be found in the vicinity of the liver attachment near the abdomino-pericardial wall or elsewhere in the body cavity. Hand stripping does not, so far as is known, so effectively remove all the eggs, and a stripped fish may often be recognised in consequence. Stokell does not refer to the number of eggs found by him nor does he speak of the way in which he differentiated between stripped and naturally spawned females. The contents of unshed eggs usually become absorbed and the crumpled shells are commonly found attached to the liver or the stomach. It is possible sometimes to find two, or even three, generations of egg remains in naturally spawned fish, quite apart from the generations of eggs to be found in the ovary.

It is not clear how eggs could be forced forward in the body-cavity when a ripe fish is held upside down preparatory to stripping. The cavity is already full of free eggs and viscera, and the body is firm. What can happen is that eggs may be left there, but not forced there, as described by Stokell. A ripe fish is not a flabby sack, but turgid and active, becoming flabby after eggs are extruded, either naturally or artificially.

Stokell apparently believes that the dead and dying fish which he found in 1932 in Lake Ellesmere and in the Selwyn River as far upstream as permanent water extends were the result of egg-collecting activity in the Selwyn River. In this connection the following facts must be taken into account. The stripping of fish in the lower Selwyn River usually has occupied a period, over all,

Picture icon

Ripening female brown trout from the Selwyn River, showing right ovary in advanced condition. The free egg surface has turned outward, and the edge of the mesovarium is easily visible. The specimen is somewhat tilted, and shows more of the ventral surface than of the dorsal.

Picture icon

Ripe female brown trout from the River Avon, Christchurch, showing eggs free in the body-cavity and the ovary shrunken. The ovigerous laminae can be easily seen with their edges turned onwards. The lower or outer edge of the mesovarium makes an apparently thickened lower edge to the ovary. The spots on the eggs are due to preservation.

Picture icon

Ripe female brown trout from the Doyleston Drain, after the eggs had been removed by flotation. The shrunken ovary shown the edges of the ovigerous laminae, and there is a very narrow ventral border of mesovarium. The mesovarium is traceable back as the intestine is exposed, forming Kendall's “oviducal groove.”

– 347 –

from late May to about mid-July. The spawning activity covers from late April to early September, this from observations on naturally spawned eggs. Spawning takes place over a region beginning roughly four miles up from the lake and extending upwards to the Main South Road, where the river runs through,* a distance of about 20 miles from the mouth, or up to the springs when the river is broken, a distance of about 12 miles from the mouth. There may be from 8 to 16 miles of water above the trap occupied by spawning fish. A stock of fish spawns above the trap-site. The downstream post-spawning movement brings naturally and artificially spawned fish to the vicinity of the lake. Only a portion of the ascending fish which reach the trap-site are handled by the stripper, the rest spawning naturally. This portion is variously estimated as a proportion of the whole stock.

The North Canterbury Acclimatisation Society has, from time to time between 1911 and 1927 inclusive, tagged brown trout caught in the trap placed in the Selwyn River. The New Zealand Freshwater Research Committee tagged a further lot in the same trap in 1931. A total of 3,689 fish was marked in the 17 seasons concerned, and consisted of approximately 1,144 males and 2,545 females (on one occasion 52 fish were marked without a record of their sex being made). There were reported 203 recaptures, but other tagged fish were taken and not reported. Of the reported recaptures five were certainly taken above the trap-site, among them being one in December, 1915 (tagged June, 1912), another in April, 1933 (tagged June, 1931). Twenty-one were merely reported as being taken in the Selwyn River, but local information leads to the conclusion that they were caught in the lower two miles of the river, below the trap-site. The rest, or 177, were reported as caught in various parts of the river below the trap-site, in Lake Ellesmere or its other tributaries, or in other rivers of the South Island.

There is nothing in the records of the above-named bodies to show that fish stripped and tagged at the Selwyn trap were recovered in the trap during the same egg-collecting season. However, Mr John Digby, who collected eggs from trapped trout in the Selwyn River from 1922 to 1936, informs me that during clear weather with a clear river, apart from re-maturing males, an occasional female returned after being stripped, and that when stripping had been carried on in a rising river, during a big run, ten per cent. would be the highest proportion of stripped fish found returning within a few hours. His experience was that such a return was exceptional, and that a proportion would consist of males which would be ripening again.

Titcomb (4, pp. 52–3) states that a trout which contains “one or two eggs” after stripping will persistently stay about a spawning bed until it has discharged them. The returns of tagged fish from the Selwyn River indicate a marked downstream movement after stripping. It is probable that Titcomb's phrase “one or two eggs,”

[Footnote] * There is also regular spawning in the upper waters whether the river runs through or not.

– 348 –

means perhaps “half stripped or thereabouts.” Partially spent fish commonly continue to run upstream and are often taken in fish traps.

In view of the foregoing facts and of the lack of information provided by Stokell, it is not clear how he was able to satisfy himself that any particular fishes examined by him had been handled in the trap. Nowhere does he state his criteria for the determination of a stripped fish except that eggs were found near the “diaphragm.” The “injury to the ovarian membranes” is not described, nor is a description given of an uninjured ovary. However, personal fortnightly examination of trout through the whole of the fishing season and during the spawning season has shown the present writer that the ovary is flabby for many weeks after the eggs are shed. It begins to harden in late spring and early summer, and by November is often seen as a hard, brightly coloured object, very different from the soft, dull-coloured body of early October and July. It is conceivable that to the inexperienced the appearance of the flabby, somewhat ragged ovary may suggest injury. Since the spawning season is so long drawn out (signs of redd-making were seen late in September, 1933, in the lower Selwyn, above the Ellesmere Railway Bridge), it is quite possible that fish may occur early in the fishing season still showing signs of the early post-spawning condition.

In September of 1933, the present writer was requested by the Council of the North Canterbury Acclimatisation Society to examine numerous fish in poor condition which were reported to be lying in pools some 10 miles up from the mouth of the Selwyn River. There was free egress from the pools, but apparently the fish would not leave. The river was low after a dry spring, but about the end of September a small freshet came and the fish disappeared. Here was in part a repetition of what took place in 1932, which had a drier winter and spring. There were poor fish above the trap-site, as well as below, associated with a low river. A similar phenomenon occurred in the Opihi River in the season 1932–3. The present writer was requested in 1933 by the South Canterbury Acclimatisation Society to examine this river, and discussed the matter with the Council of the Society. The conditions in 1932 were apparently the same as in the Selwyn, dry preceding weather, a low river over a long period, and poorly conditioned fish which continued well into the season. This state of affairs persisted up river as far as Pleasant Point, some 12 miles from the mouth. After a freshet the quality of the fish rapidly improved.

It will be seen that in both the Selwyn and the Opihi the slabby character of the fish was altered after freshets. It would appear that fish after spawning need some stimulus, associated with rising water, to cause them to commence feeding. Commonly, rain, at one time or another, during the winter, causes freshets, but the prolonged droughts of 1932 and 1933 left large numbers of spent fish without the stimulus connected with rising turbid water until anglers were on the scene.

The statement that “it would thus appear that trout are becoming fewer” will bear further investigation. The records of the North Canterbury Acclimatisation Society provided the following

– 349 –

information regarding females which were stripped in the trap, weighed, measured and tagged:—

Year. Average Weight. Average Length. Number in Sample.
lb. in.
1916 2.4 19.61 139
1917 2.7 19.55 150
1918 3.6 21.05 156
1919 2.2 17.00 37
1920 2.8 18.78 41
1921 2.6 17.20 117
1922 2.3 18.20 58
1923 3.1 18.60 121
1924 4.3 20.74 85
1926 4.0 20.84 131
1927 2.0 16.36 41

The records of the New Zealand Freshwater Research Committee show:—

1931 3.4 20.3 655
1932 2.9 19.1 666
1937 20.7 90

The Seventy-second (1936) Annual Report of the North Canterbury Acclimatisation Society provides figures of average egg production of stripped female brown trout from the Selwyn River, as follows (many of the fish come in from Lake Ellesmere):—1932—1,119 eggs; 1933—1,419 eggs; 1934—1,413 eggs; 1935—1,457 eggs. In those four years the ratios of females to males handled were:— 1932—1.4:1; 1933—1.6:1; 1934—2.5:1; 1935—1.6:1.

If the inference were sound that an increasing mortality resulted from the effects of stripping in the Selwyn, there should be shown some sign of it in the character of the fish handled in successive years. The year 1930 has been referred to as representing a kind of end of a period after which survivors of the winter's stripping form a greater part of anglers' catches than they did before. The average lengths and weights given above certainly show fluctuations between 1916 and 1927, and in 1931 and in 1932, but there is nothing in them which shows a general downward trend, which would be expected if the death-rate of mature fish increased through stripping. Since mature fish almost entirely are handled in the trap, any casualties would tend to affect the higher age groups, and, unless the growth rate has undergone some remarkable change, the higher length groups, thus reducing the average length and increasing the proportion of earlier spawners. The average egg production from 1932 onwards does not indicate a reduction in productivity. It should be noted that the average productivity published by the North Canterbury Acclimatisation Society is much lower than would be expected from counts made from ripening ovaries by the present writer. Many more or less spent females enter the trap, some of them perhaps half spent. At the same time there is a considerable variation in productivity, as is indicated by the following examples: 16 inches, 2,225 eggs; 17 inches, 1,430 eggs; 17 ¾ inches, 3,320 eggs; 18 ½ inches, 3,050 eggs; 21 inches, 2,510 eggs. The error is probably fairly constant, and the figures of productivity may have some value, but there is need for further investigation.

– 350 –

Further, since the act of stripping females may be regarded as more vigorous than with males, there should be shown some reflection on the sex ratio. Actually, the sex ratios given show no particular depression in the proportion of females. The sex ratios given above probably do not represent the actual facts, as males may ripen more than once and may thus be counted several times. In such circumstances it is probable that there may be some constancy in the error so introduced so that the figures may be not entirely valueless.

The “peritoneal injury” referred to is not described. It is well-known among anglers and others who open trout that after a longer or shorter interval, especially in warm weather, the lining of the body cavity, in the vicinity of the stomach, appears inflamed and the tissues show signs of destruction. The effects are in such cases undoubtedly due to auto-digestion, a phenomenon which is known in the fish-canning industry. Unfortunately, Stokell does not state the time that elapsed between the killing of his specimens and their being opened, and it is quite possible that the “peritoneal injury” was actually an auto-digestive effect. Whether a fish suffering from acute peritoneal damage could live for a period of about 12 weeks is not known with certainty.

While it is possible that the damaged spleen referred to by Stokell was caused by stripping, it is equally possible that such an injury could be brought about up river by the use of explosives, a method which is not unknown among poachers. Further, a specimen, naturally spawned, which recently came into the hands of the present writer, had a spleen already subdivided into a larger part and a smaller part. The presence of dead fish all along the permanent water is not satisfactory evidence that they were killed as a result of being handled in the trap. Since permanent water occupied by spawners over several months exists for some eight miles above the trap-site, and since stripped and other spent fish usually pass down-stream, it is clear that some other cause for the deaths must be found. It is not uncommon for dead fish to be caught above the trap by the barrier which is first placed there. These latter deaths may be due to exhaustion by spawning or to other causes, but they certainly are not due to stripping.

Stokell speaks of “typical lake stock.” It is known that fish taken from the lake are usually rather silvery in appearance, those taken in the river being usually darker, but there is nothing in the literature available which shows any constant structural difference between lake and river fish. At the same time, silvery fish are taken from the lower part of the river, particularly from late January onwards. How long a lake fish takes to assume the outward appearance of a river fish or vice versa is not reported. It is common statement among the anglers about the lower Selwyn River that fish move into and out of the river, and many of the fish that move into the river during the early part of the spawning run may spend several months there before they return to the lake, but large numbers deposit their genital products in the tributaries. As far as we know, the lake is stocked by trout which pass in from the

– 351 –

streams after spending more or less of their life therein. What the term “typical lake stock” means is not at present clear.

It may be concluded that, on the evidence offered, the view is unsatisfactory that the trout stock has been reduced as a result of stripping operations, and, in fact, other interpretations of the conditions mentioned by Stokell can be more effectively adduced. The recorded samples of the stock through a number of years show no continued reduction in average length, or average egg production, or alteration of sex-ratio, such as would be expected if the death rate of mature fish, particularly females, were increased. The presence of loose eggs is seen to be a common occurrence, and internal changes may be regarded as a consequence of internal causes. The description of the ovary of the brown trout, given above in this paper, does not assist in showing how bursting of the mesovarium could take place during efficient stripping. The poor condition of fish in 1932 was not peculiar to the Selwyn River and was repeated in 1933. It is clearly linked with the weather conditions of the period. The natural behaviour of trout before and after spawning, the long spawning and short trapping season, the proportion of naturally to artificially spawned fish, and the approximately eight miles of available spawning water above the trap-site make necessary much more information about Stokell's method of identifying stripped fish than he has provided.

Indebtedness to the North Canterbury Acclimatisation Society and the New Zealand Freshwater Research Committee is here acknowledged for permission to make use of their records, to Mr. D. F. Hobbs for asistance in the field, and to Mr. E. F. Stead, of Christchurch, for the provision of two fishes. Financial assistance was given by the New Zealand Freshwater Research Committee, to which body the writer is much obliged. Thanks are also due to the Chief Inspector of Fisheries, Mr. A. E. Hefford, through whose assistance part of the cost of the plates was defrayed.

References.

1. Boulenger, G. A., 1904. The Cambridge Natural History, Fishes, Ascidians, etc., London.

2. Bridge, T. W., 1904. The Cambridge Natural History, Fishes, Ascidians, etc., London.

3. Kendall, W. C., 1919–20. Peritoneal Membranes, Ovaries, and Oviducts of Salmonoid Fishes and their Significance in Fish-cultural Practices, Bull. U.S. Bureau of Fisheries, vol. XXXVII, pp. 183–208.

4. Kendall, W. C., 1920. Some Previously Unrecognised Anatomical Facts and Their Relation to Fish-cultural Practice, Trans. American Fisheries Society, vol. L, pp. 43–55.

5. Stokell, G., 1936. The Nematode Parasites of Lake Ellesmere Trout, Trans. Roy. Soc. N.Z., vol. 66, pp. 80–96.

Description of Plates.

The specimens had been preserved in Formol-Alcohol after dissection and before photography in water.

Abbreviations:

mes. ov.: lower edge or outer edge of Mesovarium. ov. 1.: ovarian laminae, o.g.: edge of Kendall's oviducal channel.