ABSTRACT
For the egg of any given species of animal there exists a range of temperature within which the embryo is capable of developing into a normal healthy organism. If the temperature of incubation be raised, the velocity of development of a cold-blooded embryo is increased ; if the temperature be lowered the velocity is decreased. In both cases the result is qualitatively similar; the mutual proportions of the various tissues and organs are, as far as is known, much the same in the final products. It is improbable that the velocity of every process which occurs during development is equally affected by changes in temperature; some processes will tend to respond to a greater extent than others. If, therefore, development were simply the summation of each separate process we would expect to find that the relative size of each organ would vary with the temperature at which the egg had been incubated. Consider, for example, two cells each capable of reproducing itself in 12 hours at 10° C. but differing from each other in their temperature coefficients of growth. If a rise of 10° C. increased the rate of development of one cell by 100 per cent., and that of the other by 200 per cent., then at the end of 12 hours’ development at 20° C. there would be five cells of one type for every seven of the other, whereas at 10° C. the numbers would be equal. The very fact that marked variations in the proportionate size of organs and tissues do not apparently characterise variations in the temperature of incubation indicates that, to some extent, the velocity of development of any one type of tissue or the velocity of any one embryological process is dependent on the velocity of other parallel processes, or that they are all controlled by the velocity of one fundamental reaction. By varying the temperature of incubation of the eggs of the trout, a convenient means exists of determining how far this picture of the embryological development of a fish is true, or how far some processes can be regarded as independent reactions.
The present paper deals solely with the effect of temperature changes on the egg of Salmo fario within the range of “normal” development; in other words, on the behaviour of the eggs at temperatures capable of producing healthy functional larvae. The range considered is from 2·8° C. to about 17° C., although above 15° C. the mortality of the larvae is high during the later stages of development. Owing to the fragility of the very young embryos it is difficult to rear larvae in the laboratory from the time of fertilisation until the completion of the larval phase of the life-history, so that most of the observations have been made on eggs which were incubated at about 10° C. for four or five weeks, and which were then subjected to varying conditions of temperature. From the records kindly supplied to me by the owners of various hatcheries it has been possible, however, to construct Fig. 1 and Table I, which show the effect of constant temperatures on the time which elapses between the moment of fertilisation and the time of hatching from the shell.
The application of Arrhenius’ formula to these figures gives a value for μ equal to 24,500 for the range 4° C. to 10° C.; the data given by Maitland (1887) for the range 5° C. to 9° C. gives 20,000. These figures agree tolerably well with those calculated by Crozier (1926) for other types of teleostean fish, and we may conclude that the period of incubation is controlled by temperature in a way essentially similar to that of other species.
It is open to doubt, however, how far an observation of the period of time occupied by the pre-hatching phase of larval life throws any real light on the processes of development. As soon as the eggs hatch, it becomes evident that the morphological form of the embryo which hatches from the egg depends on the temperature of incubation, and is not a fixed or predetermined character, which is the same under all conditions. This fact is illustrated by the behaviour of eggs which were all incubated at 10° C. for 35 or 38 days, and which were then segregated into groups each of which completed its development at a higher or lower temperature. As soon as the larvae hatched they were removed from the hatchery and weighed. From Table II it can be seen that when the temperature of incubation is low, the act of hatching is delayed until the embryo has reached a size which is considerably greater than that at higher temperatures. At high temperatures the eggs hatch precociously early. Thus in Experiment 1, Table II, the embryos on hatching from eggs incubated at 5° C. weighed 2·70 gm. per hundred, from similar eggs incubated at 9° C. one hundred newly hatched embryos weighed only 2·00 gm., and at 17° C. only 1-43 gm. per hundred. In each case the weights were determined within a few hours of hatching, and in every case subsequent development was normal except that prolonged incubation at 17° C. is marked by high mortality. A consideration of column III for the first experiment in Table II shows that the temperature coefficient of the actual hatching mechanism is very much higher than that of the other processes of development which take place between fertilisation and hatching and which are included in Table I.
The first sign of approaching emergence from the egg is a marked increase in the transparency of the shell, and this is almost certainly due to the operation of the enzyme described by Wintrebert (1912). The surface of the larva secretes the enzyme which dissolves the tough ovarian membrane, rendering it thin and transparent. Eventually the larva is able to burst through. If a number of eggs (reared at 10° C.) whose membranes are seen to be transparent are suddenly exposed to a temperature of 15–17° C. they invariably hatch within a few hours, whereas similar eggs exposed to 5° C. may not hatch for three days. It may be concluded that the precocious hatching at high temperatures is due to the very high temperature coefficient characteristic of the production or of the action of the enzyme which softens the egg-shell. It may well be that the hatching enzyme is not liberated until the embryo has reached some critical stage in its development, but this point has not yet been investigated. It is clear, however, that, by varying the temperature of incubation, it is possible to differentiate between two distinct processes : (a) the development of the embryo, and (b) the process of hatching out of the egg-shell. The latter has a much higher temperature coefficient than the former, and the two processes are to some extent independent of each other, but are not completely so. When speaking of the velocity of development, it is imperative to define the precise nature of the process observed since the temperature coefficient of development will vary according to the criterion used as a measure of development, and on the degree to which the velocity of this process is affected by the velocities of others.
The effect of temperature on the rate of growth of the embryo is illustrated in Fig. 2. As might be expected, the higher the temperature the more rapid is the rate of growth, but, as pointed out in a previous paper, Gray (1928), the final size of the embryo at the end of larval life is smaller at higher temperatures than at lower. The reason for this is discussed in the previous paper ; it is due to the fact that at higher temperatures a larger proportion of the yolk is required for maintenance of the embryonic tissue, and consequently less is available for conversion into new tissue. Thus when the temperature of a developing organism is changed, the velocity of each one of many processes is altered and a new state of dynamic equilibrium is established between them. For this reason it is impossible to attribute to development a temperature coefficient in the sense that a coefficient can be determined for a simple non-reversible chemical reaction. A biological temperature coefficient is analogous to that expressing the effect of temperature on a mixture in which a series of catenary, side, and opposing reactions are going on simultaneously. The facts suggest that in no specific instance do the processes of development respond to changes in temperature as a series of parallel but independent chemical or physical reactions. The precise relationship between the effect of temperature on each single process when this is isolated from the others and the effect of temperature on the same process when this is forming part of a series of interdependent processes is by no means clear.
As already pointed out, the possibility of producing normal organisms throughout a considerable range of incubation temperature indicates that the velocity of each embryological process is dependent on the velocities of all the other processes. A deeper insight of the mechanism whereby this is effected must, probably, be looked for outside the range of temperature here considered.
SUMMARY
Normal trout larvae can be raised from eggs incubated at any temperature between 2.8° C. and 13° C. without high mortality. Above 15° C. the mortality is high.
The morphological development of the embryo at the moment of hatching depends on the temperature of incubation. When eggs are incubated at low temperatures the embryos, at the moment of hatching, are significantly larger than those hatching from eggs which have been incubated at higher temperatures. Precocious hatching at high temperatures is probably associated with a high temperature coefficient of the hatching enzyme which softens the egg-shell.
Increasing the temperature of incubation increases the growth-rate of the embryo, but, at the end of larval life, the full-time embryo is smaller than after slower development at a lower temperature. At higher temperatures a larger proportion of the available yolk is required for the maintenance of the embryo, leaving a smaller proportion available for the formation of new tissue.