Studies on Adaptation to Salinity in Gammarus Spp: I. Regulation of Blood and Tissues and the Problem of Adaptation to Fresh Water

Invasion of brackish and fresh waters by marine animals has not always been accomplished by means of the same type of osmoregulatory mechanism. This is clearly shown by the great differences in the ability of fresh-water animals to produce a urine hypotonic to the blood (Table II) and to concentrate ions from the medium (Krogh, 1939, p. 203). As a result the blood concentration is maintained at very different levels in different species (Table II). It should therefore be of interest to investigate several species of the same genus which have become adapted to habitats of different salinity range and which, we can be reasonably certain, form an evolutionary series. For this reason we have chosen four species of the genus Gammarus. In the present work a survey is made of the relation between the external medium and blood, when the salinity of the former is rapidly changed. In addition, estimations of tissue chloride show that the chloride in the tissues can be regulated against changes in that of the blood and that this mechanism is as important as that for regulation of blood concentration in facilitating adaptation to changes of salinity.

In a later paper results of experiments will be recorded in which certain species are acclimatized by very gradual change to salinities not found in their normal habitats and outside the range tolerated when transfer is rapid. Another interesting problem is raised by the brackish water species G. duebeni which is apparently in the process of becoming adapted to fresh water and is actually the common inland water Gammarus of Ireland (Reid, 1939). Comparative experimental work on samples of G. duebeni from different habitats will be published in the second paper of this series. The G. duebeni used in the present experiments were taken from brackish water only. The value of Gammarus as material for this kind of work is obviously limited by their small size, but ultramicro-methods are now available by which some useful information can be obtained.

G. locusta was supplied from Millport by Mr R. Elmhirst and was obtained by dredging. G. obtusatus¹ was collected under stones close to low-tide level at Cullercoats. Both these species were kept for at least 48 hr. in.aerated sea water (540 mM.) before the experiments.

G. duebeni was collected from Meggies Burn, Blyth, Northumberland. The salinity of this stream is influenced by the daily tidal inflow but never becomes very low owing to saline water continuously pumped into it at its head from a coal mine nearby. Measurements made at irregular intervals during the past four years indicate that it fluctuates between about 10 and 50% sea water. They were kept at least 48 hr. in 50% sea water (270 mM.).

G. pulex collected from the Cor Burn, Northumberland, and kept in tap water. The above stock cultures were fed with elm leaves, but feeding was stopped during the experiments. Dilutions of sea water were made with tap water and aerated by thorough shaking. No further aeration was provided during the 24 hr. taken for each experiment. To obtain the points in Fig. 1, about 25 animals were transferred direct from the culture to about 150 c.c. of dilute sea water of the required salinity, and the measurements were made after 24 hr. Except in the case of G. pulex (see below) only salinities were chosen in which the animals showed no signs of distress after this time.

Extraction of blood, after drying between filter papers, was effected by holding the animal between the thumb and forefinger of the left hand and pricking the dorsal vessel with a needle held in the right hand. A drop of blood immediately appears which can be sucked up into a waxed micropipette for transfer to the thermocouple loop for vapour-pressure determination, or directly into the graduated micropipette for chloride estimation.

Blood osmotic pressure was measured, to the nearest 2 mM./l., by Baldes’s modification of the Hill vapour-pressure method (Baldes, 1934).

Blood chloride was estimated to the nearest 3 mM./l. by Wigglesworth’s micro-method (Wigglesworth, 1938a). By the method adopted by Wigglesworth we found it impossible to obtain a stable coating of wax on the micropipettes and burettes. For the benefit of others who may find the same difficulty we suggest the following procedure by which we were invariably successful. With the object of removing possible moisture and adsorbed air from the inner surface, a column of absolute alcohol followed by one of xylol was drawn up the capillary. Immediately after discharging the latter, a column of molten wax was sucked up and discharged before the final coating operation. This was done as described by Wigglesworth except that the molten wax was as cold as possible.

Tissue chloride. The method used was based on Patterson’s cold nitric acid digestion method for blood (Patterson, 1928). The main advantage of this method is its extreme rapidity. Five to ten Gammarus were dried on filter paper, cut in two, and the blood drained off by pressing between two filter papers. After weighing (100-200 mg.) they were transferred to a hard glass test-tube (⁠⁠.) containing a drop of distilled water, and were ground to a paste by means of a bulb-ended glass rod. 3 c.c. of conc.HNO₃ were added and stirred thoroughly with the rod. After a few minutes a clear yellow solution containing skeletal debris was obtained. The rod was washed into the tube with the least possible distilled water from a wash bottle, 1 or 2 c.c. (according to the amount of Cl present) of 0 25 % AgNO₃ were added and the mixture shaken, followed by 3 c.c. of pure acetone. After thoroughly cooling under the tap the excess AgNO₃ was titrated with 0·05 NNaSCN from a Bang 2 c.c. microburette, using 6 drops of saturated iron alum as indicator. The NaSCN was previously standardized by substituting for the tissue 1 c.c. of c. 8·4 mM. NaCl, this being the only originally accurate standard required. The method was tested on a piece of fresh fish muscle (haddock), and the following seven results were obtained: 33, 31, 34, 34, 33, 33, 31 mM./kg.Cl. These figures give a standard deviation of ± 1-6 %.

The main results are shown in Fig. 1. These curves require no detailed description since they are largely self-explanatory. We will therefore proceed to discuss their more significant features.

Both of these can withstand considerable dilution of the medium, and over their entire tolerance range maintain their blood concentration at a higher level than G. duebeni, which can actually survive much greater dilution. In fact, the degree of dilution which each of these three species can tolerate bears an inverse relation to the osmotic pressure of the blood. In view of the wider tolerance range of G. locusta than that of G. obtusatus it is interesting to note that locusta has often been recorded from brackish water, whereas obtusatus has never been found except on shores inundated by pure sea water (private communication from Mr G. M. Spooner).¹ It is well known that G. marinus commonly occurs in brackish water at the mouths of streams and estuaries. Though its tolerance range has not, so far as we know, been determined, Widmann (1935) was able to acclimatize it to 186 mM. sea water, in which its blood concentration was 470 mM. Its osmoregulatory mechanism is thus even more effective than that of G. locusta (Fig. 1). But since some marine animals such as Arenicola marina can survive in 25 % sea water with no mechanism for maintaining hypertonic body fluids (Krogh, 1939, p. 47,) the possession of such by the marine Gammarus species does not appear to be functionally significant.

This species can be transferred direct from 100 to 2 % sea water and vice versa without harm. It will be shown in a subsequent paper that adaptation to extreme dilutions and to fresh water depends upon the composition of the water. From the form of the blood curve (Fig. 1) it is evident that there is a critical external concentration (259-350 mM.) below which the osmoregulatory mechanism comes into play, and above which the hypertonicity of the blood is relatively slight.

The blood concentration is here much lower than in the other species. Widmann (1935) measured the freezing-point of the blood of G. pulex in fresh water, by accumulating blood samples from a large number of animals. The figures do not entirely agree with ours. She found seasonal differences in blood concentration ranging from 185 mM. in August to 265 mM. in February. Our figures, obtained in April, were lower than these (142-165 mM.). The reasons for this seasonal fluctuation, which we have not yet confirmed, are quite obscure, but we are not at present concerned with them since all our experiments were done at the same time of year on animals with a similar level of blood concentration.

G. pulex can withstand sudden transfer to 20% sea water (107 mM.), but in 215 mM. a few died after 24 hr., though the majority remained normal. This species showed more individual variation than the others. In 50% sea water (270 mM.) all had practically stopped moving after 24 hr. The blood concentration is raised by increasing the salinity of the medium, and a critical point seems to be reached when the concentration of the medium approaches that of the blood (Fig. 1). Widmann (1935) acclimatized G. pulex by gradual addition of sea water during 17 days to 155 mM. sea water. At this season (May) the blood concentration in fresh water was c. 220 mM. and at the end of the experiment it had risen to 283 mM. These figures would fall on a curve parallel with ours but at a higher level.

The curves for blood chloride (Fig. 1) are mainly parallel with those for osmotic pressure. This suggests that the change in total concentration is due to salt exchange and not to movement of water under osmotic forces. The absence of any significant change in total water content was also shown by weight measurements on G. pulex and duebeni following a sudden change in external salinity. These are given in Table 1. The difference between total osmotic pressure and chloride concentration (non-Cl fraction) is relatively constant for ‘each species, but there is an obvious difference between the bloods of the two marine species and those of G. duebeni and pulex. The approximate average concentrations of the non-Cl fractions are as follows : G. locusta 120, obtusatus 95, duebeni 30, and pulex 35 mM. It remains to be seen whether a low non-Cl fraction in G. duebeni and pulex is in any way connected with their ability to live in low salinities.

The blood curves in Fig. I present a problem which cannot be entirely solved by reference to blood-regulating mechanisms. Why should the two marine species begin to succumb in dilute media before the blood concentration has fallen to a level normally maintained by G. pulex’s And, conversely, why should G. pulex be unable to withstand a blood concentration as high as the lowest level maintained successfully by the brackish and marine species? Regulation of the blood is of course important only in so far as it leads to the maintenance of healthy tissues. Tissues in general, though in osmotic equilibrium with the blood, are known to differ considerably from it in the concentration of individual ions, a condition which is essential for the normal functioning of the cells. It may therefore be suspected that the intracellular mechanism for maintaining these differences depends upon a certain range of blood concentration which differs in each species. For this reason we estimated tissue chloride, which in most tissues is at a much lower concentration than in the blood. In order to detect a breakdown in the intracellular regulatory mechanism it is necessary to measure the tissue Cl of animals in salinities which they are unable to withstand, but at the same time the concentration of the blood must not be allowed to depart from the range which is normal for other species. There are practical difficulties in this, since from unhealthy animals succumbing to the effects of low salinity treatment, it was impossible to obtain a discrete drop of blood which could be picked up in a micropipette. No doubt this is due to a loss of hydrostatic pressure. Though the blood could not be extracted in the manner necessary for vapour pressure or chloride estimations, it was readily drained off as a preliminary to tissue chloride estimation (see section on methods). For this reason we could not determine the blood concentration of the matine species in water of salinity lower than the tolerance range. But with G. pulex, in salinities above the normal tolerance range, the blood concentration could be predicted with certainty as being at least within the range withstood by G. duebeni (Fig. 1).

The tissue Cl concentration found for G. pulex in fresh water is of an order to be expected from previous work on the muscles of fresh-water animals (Table II). The differences are obviously related to the differences in blood chloride. Scholles (1933) found 152 mM./kg. Cl in the muscle of Eriocheir sinensis in sea water, which is almost identical with our average figure for G. duebeni in sea water. Though the blood and tissue Cl concentrations are not exactly comparable since they were estimated per litre and per kg. respectively, yet, with the greatest possible allowance for the lower water content of the tissues, it is evident that a very large blood tissue Cl gradient is maintained within the tolerance range of all species. G. pulex was presumably evolved from a brackish water ancestor, but it is now less able than G. duebeni to maintain a low tissue Cl when the blood Cl is rapidly raised. On transfer to 40% sea water (215 mM.) the blood Cl is increased by about 56% and the tissue Cl by about 140%. This is not sufficient to immobilize the animals (only healthy ones were analysed), but in 50 % sea water (270 mM.) the still more rapid increase of tissue Cl is associated with obviously deleterious effects on all the animals. So rapid is the penetration of chloride into the tissues in 75 % sea water (400 mM.) that in 4 hr. the tissue Cl had already reached the level finally reached by G. duebeni in 540 mM. sea water. After 8 hr. a further rise had occurred, and the animals had almost ceased moving. It has been maintained that in frog muscle the chloride is confined to the extracellular fluid which is in equilibrium with the blood (Eggleton et al. 1937). Whether or not this is true of Ganmmarus tissues is unimportant in the present connexion. The conception of a blood/tissue Cl gradient is equally valid, and a higher percentage increase in the tissues as a whole than in the blood can only be explained as due to intracellular penetration of chloride.

On the assumption that the tissues remain isotonic with the blood, it is possible to calculate from the data in Fig. 1 the percentage of the total tissue osmotic pressure due to Cl. The results are plotted in Fig. 2 against blood Cl. No previous correction was made to the tissue Cl figures (estimated per kg.) for the water content of the tissues, which may vary, but presumably to the same extent in each species over the same range of blood concentration. These curves illustrate clearly the importance for adaptation to changing salinity of the mechanism for regulating tissue composition against changes in the blood. The two species G. duebeni and locusta, which have the widest tolerance range, have the greatest powers of regulation, whilst the tissues of G. pulex can be maintained only with a relatively low blood chloride.¹ Whether the restricted range of G. obtusatus is due to inability to adapt its tissues to a low blood concentration, or to insufficient power to maintain hypertonic blood in dilute sea water cannot be decided from the data available.

Owing to the roughness of the method for obtaining tissues for Cl estimation, further speculation on the meaning of the curves in Fig. 2 would be useless. It would, however, be worth while doing similar work on large animals such as crabs, from which a sufficient quantity of a single tissue could be got for estimations of several ions.

It is remarkable that fresh-water animals differ so greatly in the concentration of their bloods. We have collected as much data as possible to illustrate this (Table II). These variations may be due in part to mere quantitative differences between the osmoregulatory mechanisms. But that these mechanisms differ also in kind is clearly shown by the available data for urine concentrations (Table II). The urine is hypotonic to the blood in all except the crabs, Eriocheir sinensis and Telopusa fluviatile, in which it is iso- or slightly hypertonic to the blood. It is surprising that these two species, in which the excretory organs do not therefore function for osmoregulation, happen to be those with the highest blood concentration (over 300 mM.). It is known that Carcinus moenas also produces an isotonic urine (Schlieper&Herrmann, 1930). In dilute sea water it maintains a strongly hypertonic blood (of the same order as that of G. locusta), and, though apparently incapable of adaptation to pure fresh water, interpolation of the blood concentration curve would give a concentration in fresh water of about 300 mM. (Duval, 1925). If a high blood concentration in dilute sea water is associated with iso- and a low blood concentration with hypotonic urine production, we should expect from our results (Fig. 1) that the marine Gammarus species are of the former and G. pulex of the latter class. There is, of course, no direct evidence on this point, but Schwabe (1933), from a comparison of the structure of the excretory organs of G. locusta and pulex, deduced that only G. pulex can produce an hypotonic urine. A similar investigation of G. duebeni would be of interest.

We suggest the following as a working hypothesis to explain the facts : adaptation to fresh water has proceeded by two main stages. The first was accomplished by a mechanism which is very effective in maintaining a high blood concentration. The excretory organs do not play a part in this mechanism, which probably entails active absorption of ions from the medium. By this means the tissues are not subjected to a very great change in blood concentration. The second stage was marked by a lowering of the blood concentration with a consequent lessening of both blood/external medium (O.P.) and blood/tissue Cl gradients to levels more easily maintained. This was accompanied by the development of the renal salt reabsorption mechanism, which further facilitated adaptation to fresh water. The attainment of this final condition entails the inability to withstand transfer back to more than low concentrations of sea water, for the reason that the tissues have lost the power of maintaining a high blood/tissue Cl gradient, which is essential for resistance to a raised blood concentration (e.g. G. pulex, Fig. 1). Duval (1925) noted that not only Eriocheir sinensis, which is also found in brackish water, but also Telphusa fluviatile, a purely fresh-water form, can be transferred from fresh water direct to sea water without harm, the blood of both becoming ultimately isotonic with the medium. Both of these maintain high blood concentrations and, as shown by Eriocheir (Table II), high blood/tissue Cl gradients. On the other hand, most fresh-water animals have a low blood concentration and presumably, as in the cases of G. pulex and the frog (Table II), a low blood/tissue Cl gradient. This, according to our argument, would account for their restricted salinity tolerance range. Potamobius fluviatilis has an intermediate blood concentration level (Table II). It is therefore significant that Bogucki (1934) found it to withstand a salinity as high, but not higher than c. 370 mM., in which its blood concentration was approaching that of the medium.

Duval (1925), in comparing Telphusa fluviatile, the carp, and Anodonta cygnea, did in fact conclude that the degree of salinity which a fresh-water animal can withstand depends upon the level of its blood concentration. Whilst confirming this we would suggest further that high blood concentration is associated with an intracellular mechanism for maintaining a high blood/tissue Cl gradient, and that it is upon this that the tolerance range primarily depends. The inequality in the distribution of ions is certainly not confined to Cl, but tissue Cl was chosen for the ease with which it is estimated. The way is clear, by carefully controlled experiments, for further testing of this hypothesis. If the seasonal differences in blood concentration of G. pulex found by Widmann (1935) can be confirmed, it would be of great interest to determine whether these are associated with differences in the salinity tolerance range. There are several recorded experiments in which, by very gradual change over several months, purely marine animals have been acclimatized to fresh water and vice versa, and Sexton (1928) succeeded by these means in adapting G. pulex to sea water and G. locusta to fresh water. The physiological mechanism by which this is accomplished is obviously of great interest, and we are starting investigations on this point.

There are of course several animals of fresh-water origin which can live in extremely saline water. But, as investigations on two of such animals show (Artemia salina, Medwedewa, 1927; Aedes detritus larva, Beadle, 1939), a special mechanism has been evolved whereby the blood concentration is maintained at a level characteristic of purely fresh-water animals. By these means they can with-stand a salinity several times that of sea water, conditions which are intolerable to most marine animals. A similar mechanism also exists in a few brackish water animals, by which the blood concentration is kept at a high level (e.g. 330-350 mM. in Palaemonetes varians, Pannikar, 1939), though hypotonic to the medium when this is pure sea water. Whether such animals can withstand an abnormally high salinity has not apparently been determined.