ABSTRACT
The factors responsible for seasonal variations in the growth of brown trout in any one environment are examined.
The probable basic importance of the water temperature is established. A critical temperature of 6° C. below which no growth takes place is deduced.
It is shown that the seasonal variations in growth observed do not result in corresponding variations in the width of the scale rings: such variations appear to be the direct result of fluctuations in available food supply.
Under natural conditions both water temperature and available food supply apparently act as limiting factors during the winter: during the summer, provided that the water temperature does not rise to a lethal level, only the available food supply is limiting.
The possible importance of light intensity as a limiting factor is indicated.
Differences in the growth rate of trout from different environments are also considered.
The role of dissolved calcium is investigated. It appears unlikely that the amount of dissolved calcium is responsible for the differences in trout growth in “hard” and “soft” waters.
It is suggested that such differences may be effected not by differences in the concentration of any one specific ion, but by departures from the optimum ionic balance brought about by variations in the relative concentration of any of the ions present.
INTRODUCTION
The factors influencing the growth of salmonid fishes may be divided into two groups—physico-chemical and biotic. In the first group light, water temperature and the chemical constitution of the medium are probably the most important: in the second group a single factor—the food available to the fish—would appear to be of greatest significance.
Seasonal variations in the growth rate of salmonid fishes in any one environment have long been recognized, but it is only recently that the specific factors responsible for these variations have been investigated. Allen (1940) studied such variations in salmon parr (Salmo salar L.) in the field and concluded that water temperature is probably the basic factor in determining whether or not growth takes place. It is clear, however, that no definite conclusions can be drawn as to which of the various factors mentioned above is the basic one until these seasonal variations have been investigated under controlled conditions. Unfortunately the greatest difficulty has been experienced in effecting these variations in the laboratory. Pentelow (1939), for example, working with brown trout (S. trutta L.), could detect no significant seasonal variations in growth rate. Appropriate aquarium facilities were, however, available in the Natural History Department, Aberdeen University, and it was decided to attempt a determination of the basic growth factor under experimental conditions. The results of this work are described in the first part of this paper.
Besides seasonal variations in a specific environment absolute differences in the growth rate of salmonid fishes from different environments are known to exist. In the British Isles these differences can best be demonstrated by comparing the size attained by brown trout in “peaty” hill streams with that reached by the same species in lowland chalk streams. In the former environment even four-to five-year-old fish may reach only 4 oz. (0.11 kg.) in weight: in the latter a weight of 3 lb. (1’36 kg.) at three to four years of age is by no means exceptional. The specific factors responsible for these differences are at present rather more obscure than those concerned in seasonal variations, and it is only possible to assess them in the broadest of terms “acid” and “alkaline” waters, “hard” and “soft” waters.
Southern (1932, 1935) studied the growth of brown trout in certain “acid” and “alkaline” waters in Eire and showed that in “alkaline” lakes and rivers on limestone rocks fish are large, grow rapidly, become sexually mature late and are noted for their longevity: in “acid” water bodies derived from Old Red Sandstone covered with peat on the other hand, fish are small, grow slowly, become sexually mature early and are short lived. Subsequent studies on other Irish waters confirm Southern’s results (Frost, 1939). In Britain, however, some exceptions have been found to this general rule. In Sutherland and Caithness “acid” lochs have been found to yield trout of 3 lb. or more (Salmon and Freshwater Fisheries Reports, 1936, 1937). In the Lake district of England large differences in the growth of brown trout from uniformly “acid” waters have been shown to exist (Swynnerton & Worthington, 1939). Raymond (1938) on the other hand substantially confirmed Southern’s findings by correlating the growth of trout in various British waters with the neighbouring geological strata. It should be pointed out, however, that in this latter work no chemical analyses of the various waters dealt with are given: it is thus somewhat speculative to speak of any correlation between the constitution of the waters (as opposed to the surrounding rocks) and the degree of growth achieved. In spite of the above exceptions British data are on the whole not at variance with Southern’s conclusions that trout grow large in “hard” (alkaline) waters and remain small in “soft” (acid) waters.
The specific factor responsible for this size difference was for a long time thought to be differences in food supply in the two environments. Southern (1932, 1935), however, has shown that in the “acid” waters of Lough Atorick in Eire where the trout never exceed 6 oz. (0·17 kg.) in weight the food supply is apparently in excess of the requirements of the fish. In view of this Southern suggested that the factor responsible might well be the difference in chemical constitution of the two types of water.
“Hard” and “soft” water differ in their chemical constitution in a number of important respects: the main differences are shown in Table 1. It is clear that for all the ions mentioned, “hard” water contains a greater concentration than “soft”. The importance of the calcium ion in various biological processes has, however, long been recognized and it has therefore been suggested that the difference in concentration of this ion, either alone or in combination with other ions, is the basic factor responsible for the size difference of fish living in the two types of water. This suggestion has been borne out to some extent by the work of Tunison & McCay (1931, 1935) in the U.S.A., who showed that a large proportion of the calcium requirements of brook trout (Salvelinus fontinalis Mitchell) can be obtained direct from the water. Apart from some short term experiments described in the Salmon and Freshwater Report, 1936, however, no attempt has been made to study the effect of different calcium concentrations on the growth of salmonid fishes. It appeared desirable, therefore, to undertake some experiments in this connexion: the results obtained are described in the second part of this paper.
MATERIALS AND METHODS
Yearling brown trout (Salmo trutta L.) were used in the work. They were obtained from the Howietoun and Northern Fisheries, Stirling, in February 1939 (14 months after hatching) and kept in the laboratory throughout the period of the experiments.
Two identical aquarium tanks were used to accommodate the fish. These were rectangular in shape (14 ft. 3 in. × 1 ft. 9 in. × 1 ft. 8 in. [4·34 m. × 0·53 m. × 0·50 m.]) and constructed of in. (3·81 cm.) teak fronted with in. (0·96 cm.) plate glass along one of the longer sides. The tanks were arranged back to back underneath a large skylight to ensure adequate and equal illumination. Each tank was supplied with four water jets arranged more or less equidistantly along its length. All water jets were fed from settling tanks located in the roof, in which the water level was kept constant by ball valves: in this way the quantity of water passing to each of the tanks was kept constant and equal. The water level in the aquarium tanks was maintained at 8 in. (20-32 cm.) from the bottom by means of a standing waste. An emergency standing waste was also employed to avoid flooding in the event of the normal waste becoming blocked. In order to prevent fish escaping, each tank was covered with in. wire mesh arranged in four sections separated by the water inlets. The detailed structure and arrangement of the tanks and water supply are shown graphically in Fig. 1.
On arrival in the laboratory the fish were divided into two batches of twenty-four each, transferred to the tanks described above and left for a period of 6 weeks to settle down. During this period no experimental readings were taken.
The effect of different calcium concentrations on the growth of the fish was studied by adding, at a constant rate, drops of a concentrated solution of calcium chloride (0-55 kg. per litre of solution) to the water in one of the tanks and using the other tank containing untreated tap water as a control: the chloride was used because of its cheapness and great solubility. Calcium determinations by a method based on that of Kramer & Tisdall (1921) as modified by Clark & Collip (1925) were made on the water of both tanks at periodic intervals. That of the control tank remained very constant at 0-4 mg. calcium per 100 c.c. water (the calcium content of a very “soft” water). In the calcium tank, although small variations in the rate of dropping of the calcium chloride solution caused corresponding fluctuations in the total calcium content of the water, this was maintained more or less constant at 5 mg. calcium per 100 c.c. water (the calcium content of a typical “hard” water). Other than this difference in the dissolved calcium content there was no distinction between the experimental treatment of the two tanks.
The water temperature of both tanks was recorded daily. No significant difference between the two tanks could be detected at any one time. The hydrogen-ion concentration of the water was estimated periodically during the course of the experiments: it was found to remain fairly constant at pH 7·0.
The fish were found to survive captivity reasonably well. Throughout the course of the experiments the mortality was six fish from the control tank and four fish from the calcium tank. It was found necessary, however, to treat the tanks with a solution of potassium permanganate once every two or three weeks to prevent the growth of fungus: this procedure was especially important during the summer months.
The fish were fed daily on a diet made up of minced liver, chopped earthworms and Daphnia suspensions: both tanks received the same quantities of food. Prior to each feeding all food left over from the previous day, together with excreta, were removed from the tanks.
The growth of the fish was recorded as follows. The fish from each tank were weighed collectively and their length measured individually once a month from 3 April 1939 to 7 March 1940. They were removed from the tanks with a hand net, the majority of the water allowed to drain off, and transferred to a previously weighed can containing sufficient water to accommodate the fish without overcrowding. This was then weighed to the nearest 0-5 g. the original weight of can and water subtracted and the total weight of fish recorded. The fish were then removed from the can one by one, placed on a towel saturated with water and their length measured to the nearest 0-5 cm. As soon as this latter measurement was completed the fish were returned to the main tank. Readings were expressed as the average weight or length of a single fish.
In an attempt to corroborate these growth readings scale samples were taken from six fish in each tank at the end of the experiments. The scales were macerated for 14-21 days in tap water, after which they were scrubbed clean with a stiff camel hair brush, washed thoroughly in distilled water and mounted in “Euparal” for microscopic examination.
The effect of different concentrations of calcium in the water on the amount of this element stored in the body of the fish was studied by making calcium analyses of the vertebrae of the fish used for scale samples. The analyses were done as follows. The fish were killed by guillotining and the backbone roughly separated from the carcass. Complete separation of the individual vertebrae from the surrounding tissue was effected by the method of Subrahmanyan et al. (1939).
After disarticulation was complete the vertebrae were washed several times with hot distilled water. A number (weighing c. 0·1 g.) were then roughly dried on filter paper, transferred to a previously weighed crucible and dried overnight at 40 ° C. After recording exactly the dry weight of each sample the vertebrae were ashed to “whiteness” in the usual way. When cold the material was digested with a small quantity of 2NHC1, transferred to a 100 c.c. flask and made up to the mark. Calcium analyses of aliquot portions of this solution were then made by the method referred to above: results were expressed in milligrams calcium per gram of dried bone.
EXPERIMENTAL
One batch of fish was used to investigate the role of water temperature and food supply in effecting seasonal variations in growth. The other batch was employed to determine the effect of higher concentrations of dissolved calcium on growth, the first batch being used as a control.
(1) Water temperature and food supply
Although either of these two factors might be the basic one in controlling seasonal variations in the growth of trout, in view of preliminary observations both in the field and in the laboratory indicating the importance of the water temperature, it was decided to investigate this factor first. Accordingly the water temperature was not controlled and varied with the season of the year. The food supply, on the other hand, was maintained always in excess of the requirements of the fish, so that this never constituted a limiting factor.
A general idea of the temperature variations during the experiments can be obtained from the twice monthly readings given in Table 2 and summarized in Fig. 2. The extremes recorded were 2·5 ° C. on 21 January 1940 and 17·1 ° C. on 28 August 1939.
The amount of food given per week was expressed in arbitrary units—grams of minced liver, grams of chopped earthworms and measures (300 c.c.) of Daphnia suspensions. During the very cold weather from 26 January to 15 February 1940 Daphnia suspensions could not be obtained. They were replaced by either Tubificids or Enchytraeids: the amounts given were expressed in terms of Daphnia measures, 20 g. of the worms being taken as equivalent to 1 measure of Daphnia suspensions. In order to maintain the food supply in excess the amounts given were progressively increased from 1 June 1939 to 15 August 1939, after which they were maintained constant at the level reached on the latter date (Table 3, Fig. 2).
The growth of the fish both in weight and length is given in Table 4 and summarized in Fig. 2. It will be seen that the shape of the two curves (weight and length) is more or less identical and thus they confirm one another. During the spring (April and May) the growth achieved is small: during the summer it rapidly increases, reaching a maximum in early autumn (September) after which it gradually falls off. In the winter months little growth takes place: during January there is actually a loss in weight.
The rapid increase in growth during the summer is correlated with a large increase in the food intake which reaches a maximum about August. Similarly the diminution in growth during the winter months is associated with a corresponding decrease in food consumption as shown by the progressively larger amounts of surplus food found in the tanks after feeding. Allen (1940) and Frost & Went (1940) found from an examination of the scales of salmon parr that a falling off in growth occurred during the period July-August: no such diminution was observed in the experiments described above.
The relation between the water temperature and the degree of growth achieved is significant. With rising water temperatures during spring and summer there occurs a rapid increase in growth: with falling water temperatures during autumn and winter growth decreases. The exact nature of this relation was determined by calculating the rate of growth (grams per gram of starting weight) for each month, and the corresponding average water temperature (average of daily readings). The results are given in Table 5 and shown graphically in Fig. 3. In the graph the points have been separated into two series—one representing the growth rate/temperature relation during a period of rising water temperature, the other the same relation during a period of falling water temperature. The readings for February 1940 have been omitted from the graph as these constitute the beginning of another rising temperature series. It will be seen that the relation is different in the two series: in the rising temperature series the relation is roughly hyperbolic, in the falling temperature series, linear. Thus the growth rate of fish with a low temperature history increases slowly with rising water temperature ; on the other hand the growth rate of fish with a high temperature history decreases rapidly with falling water temperature. In the rising temperature series there is no falling off in the curve even at the highest average temperature (i.e. 16·4° C.): it appears therefore that the lethal temperature is considerably above this level. This is in accord with the experimental work of Gardner & Leetham (1914) who found the upper limit of temperature to be 25° C. and with the field observations of Embody (1921) in the U.S.A, and Phillips (1929) in New Zealand who recorded brown trout at temperatures of 83° F. (28-3° C.) and 77° F. (25·0° C.) respectively.
From the above results it is clear that the seasonal variations observed were brought about fundamentally by variations in water temperature. It should be noted, however, that no attempt was made to control the illumination of the tanks: it is possible, therefore, that the variation in light intensity which was similar to the variation in water temperature (high in summer, low in winter) may have been the responsible factor.
The broken line in Fig. 3 indicates approximately the general relationship between growth rate and water temperature, both series of points being included. It will be noticed that this line cuts the abscissa at 6° C. indicating that this is the critical temperature below which growth does not occur. It is interesting to note that Allen (1940) concluded from field observations on another salmonid (parr of Salmo salar) that 70 C. represented the critical temperature in that species. The similarity between the two temperatures is significant.
Under natural conditions, however, the basic influence of water temperature on growth is modified by the available food supply. When the former is below the critical temperature little or no growth can occur however great the available food supply: when above, growth is limited by the available food supply. Temperature determinations by the author in the Aberdeenshire rivers Dee and Don (unpublished) show that the water temperature remains below 6° C. for about six months of the year (from the end of October to the end of April). Similar determinations by Butcher et al. (1937) in the River Tees (Yorkshire and Durham) and by Allen (1940) in the River Eden (Cumberland and Westmorland) show this period to be a little less (November to March). For approximately 50% of their life, therefore, trout in rivers in Scotland and Northern England are below their critical temperatures and growth is at a standstill: during the rest of their life water temperature is above the critical value and growth is only limited by the available food supply.
Scale rings: it is generally assumed that variations in the growth of salmonid fishes under natural conditions are reflected in the width of the rings which occur on their scales. Examination of scales from the experimental fish, however, showed no variation in ring width even though seasonal variations in growth had undoubtedly taken place. Gray and Setna (1931) found that provided food was abundant no seasonal periodicity in the width of rings occurred in Rainbow trout (Salmo irideus Gibbons). Bhatia (1932) working with the same species showed that variations in temperature had no direct effect on the width of the rings: variations in available food supply on the other hand were found to bring about alternate zones of broad and narrow rings. It is thus clear why no such zones were found on the scales of the experimental fish: no alternating conditions of abundant and deficient nutrition were present—the available food supply was always in excess of the metabolic requirements of the fish. In nature, narrow rings are laid down during the winter months, indicating a deficiency in the available food supply. Field observations, however, tend to show that during the winter the invertebrate fauna which constitutes the greater proportion of the food of trout (Neill, 1938) is present in sufficient, if not ample quantities. It appears, therefore, that throughout this period this potential food supply is not available to the fish in sufficient quantity to meet its metabolic requirements. The specific factores) responsible for this decrease in the availability of the potential food supply are unknown, but it appears probable that the following factors may be involved:
(1) Inactivity of the fish brought about by low water temperatures decreasing area of feeding.
(2) Alteration in the habits of the invertebrate fauna reducing their availability as food for the fish.
(3) Changes in certain physico-chemical factors, such as water turbidity and light intensity, rendering the capture of food by the fish more difficult.
The above conclusions concerning the limiting factors to growth under natural conditions must therefore be somewhat modified. Both water temperature and available food supply apparently act as limiting factors during the winter months: the possible dependence of the latter on the former has been indicated above. In the summer, provided that water temperatures do not rise to a lethal level, only the available food supply is limiting. In the absence of data to the contrary, however, the possibility of a third major limiting factor such as light intensity being involved, must not be excluded.
(2) Calcium content of the water
The possible importance of this factor in determining the growth of trout in different habitats has been indicated earlier in this paper (p. 436). The experiments undertaken to test this hypothesis were as follows. The dissolved calcium content of one of the aquarium tanks was raised at the outset of the experiments (p. 438) and the growth achieved by the fish in this tank compared with that in the untreated tank. The growth of the fish both in weight and length in the former (calcium) tank is given in Table 6. The results are summarized and compared with those from the untreated (control) tanks in Fig. 4.
It will be seen that the growth of the control and calcium series is more or less identical during the first four months of the experiments (April to August 1939). During August, however, the growth both in weight and length of the calcium series begins to fall off (or that of the control series to rise).
This diminution in weight increase is continued during September and October: after the latter month the weight increase becomes more or less equal in the two series. The diminution in the increase of length, on the other hand, persists only during September: in October and subsequent months the length increase is approximately the same for both series.
Examination of scales of fish from the calcium tank showed no periodic zones: their appearance was more or less identical with those from the control tank. Evidently higher concentrations of dissolved calcium have no effect on the arrangement of scale rings.
It should be noted that the maximum water temperatures occur round the point where the growth curves of the two series commence to diverge (cf. Fig. 2): it appears possible, therefore, that the initial falling off in growth of the calcium series is due to the combined effect of additional dissolved calcium and high water temperatures.
The diminution in growth in water of increased calcium concentration is contrary to the original hypothesis, for if the higher calcium content of “hard” waters is, as was thought possible, responsible for better growth in those waters, the raising of the calcium content of a “soft” water to the level of that of a “hard” water should result in an increase of growth. It should be noted, however, that the addition of calcium chloride solution, besides raising the dissolved calcium content, also increases the chloride content of the water. It may be, therefore, that the effect observed was that due to the increase in dissolved chloride, the action of increased calcium being masked thereby. It has not yet been possible to test this hypothesis by using another very soluble calcium compound such as calcium nitrate to raise the calcium content and comparing the effect produced with that already found using calcium chloride.
Another explanation of the results is that the calcium level in the control (“soft” water) tank was still too high for it to function as a limiting factor. This supposition is borne out to some extent by an analysis of the calcium content of the bones of fish from the two series. The results are shown in Tables 7 and 8. It will be seen that there is no significant difference between the two series, showing that just as much calcium was stored by the fish in water low in calcium as in water in which the calcium content had been augmented. If this is the explanation it is quite clear that the concentration of dissolved calcium in natural waters is very unlikely to be the factor determining the nature of fish growth in different environments, as there are few natural waters with a calcium content lower than Aberdeen tap water.
It appears probable, therefore, that some factor other than the dissolved calcium content is responsible for the differences in trout growth in “hard” and “soft” waters. The nature of the factor is at present unknown, but in its subsequent elucidation it should be borne in mind that these differences in growth in different environments may be effected, not by differences in the concentration of any one specific ion, but by departures from the optimum ionic balance brought about by variations in the relative concentration of any of the ions present.
ACKNOWLEDGEMENTS
I am indebted to Prof. L. Hogben, F.R.S., for his help and advice. To the Fisheries Division of the Scottish Home Department, especially to W. J. M. Menzies and P. R. C. Macfarlane, I would express my sincere thanks.
The investigation was supported by a series of grants from the Development Commission.
REFERENCES
Owing to difficulties in obtaining publications from abroad, those references marked with anasterisk have been consulted in abstract only.