Salt and Water Balance in the East African Fresh-Water Crab, Potamon Niloticus (M. Edw.)

Among the Crustacea there are a number of fresh-water forms, of which the crayfishes are the best known examples, which are typical fresh-water animals in the generally accepted sense. The features which characterize an animal as such are: (a) the fact that it cannot survive in sea-water solutions much stronger than its normal blood concentration, (ó) a body surface relatively impermeable to salts, (c) a relatively low blood concentration, and (b) the production of a dilute urine. However, there are other fresh-water crustaceans which do not possess all of these characteristics. Examples of these are the crabs, Eriocheir sinensis and Telphusa fluviatile (= Potamon edule) (Duval, 1925 ; Schlieper, 1930; Schlieper & Herrmann, 1930; Scholles, 1933) and the prawn, Palaemonetes antennarius (Parry, 1957). The two crabs have a high blood concentration and produce an iso-osmotic urine. Eriocheir is also relatively permeable to salts (Krogh, 1938) and so is the prawn, which also produces an iso-osmotic urine. Both the crabs are able to survive in full-strength sea water.

It is perhaps not surprising that the discovery of these facts concerning the freshwater Crustacea has led to the view that whereas the crayfishes are physiologically well-adapted to life in fresh water, the other forms are not. This is supported by the fact that Eriocheir has to return to the sea to breed and it may well be that the other two are only relatively recent invaders of fresh water.

Beadle (Beadle & Cragg, 1940; Beadle, 1943) has suggested that the invasion of fresh water by the Crustacea has proceeded in two stages. The first stage, represented by Eriocheir and Telphusa, is characterized by the development of a mechanism for the maintenance of a high blood concentration in the dilute external medium. The second stage, represented by the crayfish, is characterized by the reduction in the blood concentration and the production of a dilute urine. As Beadle points out, our knowledge is, as yet, based on very few species and even on these, physiological studies in some cases are far from complete. It is therefore of interest to investigate other fresh-water species in the hope that this may throw further light on those factors which are important in the efficient adaptation of the animals to fresh water.

In the tropical regions of the world fresh-water crabs of the family Potamonidae are generally well established. The family is represented by a very large number of genera and species and the animals are found in a great variety of habitats. In most parts of Africa crabs of the genus Potamon and of the allied genus Potamonautes are extremely abundant and to be found in most fresh waters. There is no evidence of their recent arrival from the sea, nor is there any reason to suppose that they are in the process of establishing themselves in the inland freshwaters. They are clearly a well-adapted and successful group by any standards. In many respects the tropical Potamonidae take the place of the crayfishes of the more temperate regions and must surely be judged as successful, if not more so, in their own geographical locations as the latter group in theirs. Therefore a comparison of the mechanisms of salt and water balance in the crayfishes and the potamonids is of considerable interest. Any differences between the two groups in this respect cannot be regarded as being due to either of them being in the process of adaptation to life in fresh water, but rather due to the adoption of different solutions to the same problem.

Potamon niloticus is a member of the Potamonidae which abounds in the lakes and rivers of Kenya and the eastern part of Uganda, although it is not entirely confined to these parts. There are many other species which would equally well repay study. Indeed, the variety of habitats in which other species are found are of such a kind that their inhabitants may show interesting differences in their salt and water regulating mechanisms.

The animals were collected from the shores of Lake Victoria at Entebbe and then kept in an aquarium with circulating lake water. Blood samples were obtained by pipette from a puncture of the arthrodial membrane at the base of the leg. Excretory fluid was collected in the following way. The animal was held by a laboratory clamp and out of water with its anterior end upwards. While viewed through a binocular microscope, the flap closing the excretory opening was lifted with a hooked needle, a pipette inserted and the excretory fluid withdrawn. Often as much as 0·2 ml. could be collected at one time. Blood and urine samples were analysed by the following methods.

This was measured by the method devised by Ramsay (1949) and the apparatus was of the same general design as that described by Ramsay & Brown (1955). Owing to the unavailability of solid carbon dioxide, the apparatus was modified to include an outer cooling jacket filled with alcohol cooled by a refrigerating unit.

These were measured by means of an EEL flame photometer.

This was estimated by precipitation with excess silver nitrate and back titration with sodium thiocyanate.

Measurements were made on diluted samples by means of a Milliard conductivity bridge and conductivity cell.

All the methods (with the exception of those for potassium and calcium) were checked and calibrated against a standard solution of sodium chloride.

These were estimated colorimetrically by methods given by H. W. Smith (1956, Appendix V).

This was measured by a flow respirometer constructed by Prof. L. C. Beadle. The apparatus had facilities for varying the oxygen tension of the inflowing water and regulating its temperature.

The composition of the blood of animals taken from lake water is shown in Table 1. The characteristic features are a relatively high total concentration, accounted for largely by sodium and chloride ions and a high potassium concentration. A comparison of the composition of the blood with that of some other fresh-water Crustacea is shown in Table 2. There is an obvious general similarity in the composition of the blood of the four animals, although the blood of Potamon niloticus shows two features of special interest. The total blood concentration occupies an intermediate position between the very high values of the other crabs and the lower concentration of the crayfish blood. This is reflected also in the concentrations of the sodium and chloride ions. The potassium concentration of the blood, like that of Telphusa, is relatively high as opposed to the rather low concentrations generally found in fresh-water animals.

The ability of the animals to withstand increases in the concentration of the external medium was investigated. The animals were each placed in about 500 ml. of the new medium and this was changed frequently. Owing to the non-availability of natural sea water an artificial sea water was used. The experimental solutions were made up from a standard artificial sea water (as detailed in Shaw, 1959 a) and dilutions of 25, 50, 75 and 100% were employed.

The time of survival varied with the concentration of the solution. In 25 and 50 % sea water the animals appeared to be able to survive for long periods (longer than 3 weeks). In 75% sea water survival was more variable: certain individuals lived in the solution for a long time whereas others were not so successful. The survival time varied from about 7 days to more than 3 weeks. In 100% sea water the animals showed a markedly different behaviour. Survival was very poor, some animals dying within a day or so and the remainder rarely surviving for longer than about 4 days.

It is interesting to compare this behaviour with that found in the other freshwater Crustacea. It is rather similar to that of Astacus fluviatilis. Herrmann (1931) showed that these animals could survive immersion in about 66% sea water for long periods but died in higher concentrations in a few days. The behaviour of P. niloticus, however, differs rather strikingly in this respect from its near relative, Telphusa (P. edule). Duval (1925) found that this species could withstand great changes in the concentration of the external medium. It could, for example, be transferred directly from a fresh water to sea water and survive in the full-strength solution for at least a month.

In Potamon niloticus an attempt was made to increase the survival time in fullstrength sea water. In one series of experiments the transfer to the 100% solution was made gradually. The animals were adapted first to 50 % sea water, then transferred to the 75% solution for some days and finally to the 100% solution. This treatment had no noticeable beneficial effect, however, for as before the animals rapidly died off within a few days. It still remains possible that a slower adaptation over the critical concentrations range from 75 to 100% might bring about a longer survival. There was another possibility that it was the ionic composition of the strong sea water rather than the total concentration which was responsible for the lethal effect on the animals. The artificial sea water differed from normal sea water in being sulphate-free, and of the ions which were present in high concentration magnesium seemed most likely to be affected. Survival was, therefore, also tested in magnesium-free sea water solutions but no improvement was found. It seems thus most likely that the death of the animal results from an increase in the total concentration of the blood above a certain critical level. The effect of an increased blood concentration on the muscle cells of these crabs has been discussed in another paper (Shaw, 1959 a).

The relation between the concentration of the external medium and that of the blood is shown in Fig. 1. Here the freezing-point depression of the blood, expressed as a concentration of NaCl which would depress the freezing point by the same amount, is related to the concentration of the sea water in which the animal was living. The animals from 100% sea water were removed after only 3 days in the solution and therefore had probably not reached a steady state. This probably accounts for the fact that in some cases the blood appears to be hypo-osmotic to the medium at this concentration.

The behaviour of the blood concentration is quite typical of the majority of truly fresh-water animals : it shows a slight increase as the external concentration approaches the original blood concentration and thereafter the rise is roughly proportional to the increase in the external concentration, the blood remaining slightly hyperosmotic.

In many respects the changes in blood concentration following increases in the concentration of the external solution shown by P. niloticus resemble those found in Telphusa fluviatile (Duval, 1925). They differ in the extent to which the blood concentration can rise without causing the death of the animal.

The effect of an increase in the concentration of the external medium on some of the individual ions of the blood is shown in Table 3. There is no evidence that the animal is able to prevent the increase in concentration of any of the blood ions—in fact their concentrations are always above those of the external medium, and thus indicate a persistence of the ion uptake mechanisms which are operating to maintain the normal ionic composition of the blood when the animal is in its normal environment.

The composition of the normal excretory fluid is shown in Table 4. The fluid resembles that of Eriocheir and Telphusa in that it is practically iso-osmotic with the blood. The difference between the means for the blood and the urine freezingpoint depression is barely significant (t = 2-2 ; P = 0·05). In three cases where blood and urine measurements were made on the same animals the ratios of the total concentration of the blood to that of the urine were 297:247, 283:260 and 265:215, respectively; the urine concentration was always slightly lower. It is possible, therefore, that the excretory fluid is very slightly hypo-osmotic to the blood. Apart from this very small difference the most obvious feature of the urine composition is the low potassium concentration compared with that of the blood. The urine potassium concentration is less than half that of the blood and this difference is statistically highly significant (t = 7·8; P<0·01). The potassium ion may well be reabsorbed in the interests of the maintenance of a relatively high blood potassium concentration.

The composition of the urine of animals from sea-water solutions shows the same general picture. This is illustrated in Table 5 where the blood and urine concentrations are compared. As the blood concentration increases the sodium and chloride concentrations of the urine are also increased. These generally remain slightly below the corresponding concentrations in the blood. In the case of sodium there was one animal where the concentration was as low as 70% of that of the blood, but in all other cases the concentrations lay between 90 and 104 % of the blood levels. The behaviour of the chloride ion was almost exactly the same as that of sodium. The concentration of potassium was still maintained at a fairly low level despite the considerable increase in the blood potassium concentration.

Measurements of the rate of urine production were made by the usual methods. These may be listed under three heads: (a) weight changes after blocking the excretory openings or after transferring the animals to blood iso-osmotic sea water, (b) the collection of measured amounts of excretory fluid at intervals between which the excretory openings are sealed, and (c) the injection of non-metabolized substances into the blood and the measurement of the rate of appearance of these substances in the external medium. None of these methods is ideal by itself, but when they are all used on the same animals then results obtained by their use can be interpreted with confidence. The application of these methods to a number of animals is illustrated in Table 6. The results were surprising and unexpected.

All the methods pointed to a very low rate of fluid production. The weight measurement methods were clearly not sensitive enough to allow an estimate of the urine production rate to be made. The urine collections revealed that at the most only very small amounts of fluid were being formed and the rate of production did not appear to be greater than about 0-05 % of the body weight per day. In the injection experiments, inulin and indigocarmine could not be detected in the water at all during several days, although, in the case of the dye, it could easily be demonstrated in the excretory fluid itself. The methods for the estimation of creatinine are more sensitive than those for inulin and indigocarmine and a real attempt was made by injecting creatinine solutions to obtain a quantitative measure of urine production. Unfortunately, it was found that owing to the very slow rate at which creatinine appeared in the water it was destroyed by bacterial action before a sufficiently high concentration for an accurate measurement had been built up. In certain experiments the bacterial action was prevented by keeping the animals in a diluted buffer at pH 4, although this entailed sacrificing normal physiological conditions. Under such conditions it was possible to demonstrate the appearance of creatinine in the water after an injection into the blood. Following an injection of 5 mg. creatinine it appeared in the water at a rate of about 100 μ g./day. If it was assumed that all the creatinine came out through the excretory system, then a urine production rate of about 0·6 % body weight/day was indicated. This value is almost certainly too high, since it includes loss from all sources including leakage through the injection wound. It is possible also that the low pH of the water had an adverse effect on the permeability properties of the body surface.

It was not possible therefore to arrive at a definite measure of the rate of urine production. The creatinine experiments may be regarded as giving a maximum value and the true rate probably lies between 0·6 and 0·05% body weight/day. Whatever the actual value it is clear that the rate of urine production is exceptionally low and of a different order from that recorded previously for fresh-water animals.

It is possible that there is some other route for the removal of water taken up by osmosis, although such mechanisms have not been demonstrated in other Crustacea or, for that matter, in any other coelomate animals. The fact that substances injected into the blood do not readily appear in the water indicates that if such an alternative route did exist it could not involve the formation and discharge of a blood filtrate, but must presumably take the form of a mechanism for the active secretion of water. Although such a hypothetical mechanism is possible, in the absence of any evidence it seems more plausible to accept the alternative explanation that the water permeability of the body surface is very low.

Although the urine is iso-osmotic with the blood, the very low rate of production insures that the salt loss through this channel is very small. Nevertheless, an appreciable loss of salts does occur and this takes place presumably through the body surface. Normally this loss is balanced by an equal uptake of salts brought about by active uptake mechanisms. The rate of loss was measured in two ways. The first method involved washing the animals in flowing distilled water at such a rate that the concentration of salts in the water itself did not build up sufficiently rapidly for an appreciable active uptake to occur. The washing water was collected and the rate of flow measured together with the sodium and potassium concentrations. From this the rate of loss of these two ions from the animal was calculated. In the second method the uptake mechanism was inhibited by bubbling through the external solution a mixture of air and carbon dioxide in the proportion of 10:1. The rate of loss of sodium was calculated by measuring the increase in sodium concentration of a known volume of the external solution. The results obtained by the use of the two methods are shown in Table 7.

The mean value for the rate of loss of sodium was 8·0 μ M./10 g./hr. Sealing off the excretory openings had no effect on the rate of loss over a 3 hr. period. The mean rate of loss of potassium amounted to 0·5 μ M./10 g./hr. Potassium was lost at about one-sixteenth of the rate of sodium loss. However, since the blood sodium concentration is about thirty times greater than that of potassium the permeability of the body surface to potassium must be nearly twice as great as to sodium.

Since the excretory organ is not involved to any appreciable extent in the loss of salts, the inorganic ions must be diffusing out through a permeable body surface. This may well be the main route for salt loss in many fresh-water Crustacea. Thus in EriocheirKrogh (1938) estimated that 86% of the salt loss occurred across the body surface and in AstacusWikgren (1953) and Shaw (19596) found that at least 90% of the sodium and chloride loss appeared to be extrarenal. It is interesting to compare the rate of sodium loss in these three animals. In Potamon niloticus the sodium loss is about five times greater than in Astacus pallipes (Shaw, 19596) but is rather smaller than in Eriocheir. No accurate measurement of salt loss has yet been made in the latter animal, but from figures given by Krogh (1938) it appears that the loss rate is in the order of 30 μ M./10 g./hr.

If the animals are placed in a limited volume of distilled water salts diffuse out from them and the concentration of the external solution rises until it eventually reaches a level at which salt balance is achieved. This may be called the equilibrium concentration. The equilibrium concentration depends on the amount of salt which has been lost from the animal. After excessive salt loss the animal comes into balance at a minimum equilibrium concentration which is the lowest external concentration at which balance can be achieved. This phenomenon has been described and analysed in Astacus pallipes (Shaw, 19596), where the minimum equilibrium concentration was found to be 0-04 mM./l. for sodium. It was also found that the external solution had to exceed a certain volume for balance to be maintained and this was ascribed to the effect of the accumulation of excretory products and the removal of oxygen in the smaller volumes. Potamon niloticus behaved in a similar manner. An animal weighing between 10 and 20 g. required about 1 1. of distilled water before balance could be achieved with certainty. In this volume the animal generally reached a steady state at its minimum equilibrium concentration. The minimum sodium external concentration for a number of animals is shown in Table 8. The mean value is 0·05 mM./l. Na and is similar to that found for Astacus pallipes. In Astacus it was found that a loss of from 5 to 10 % of the internal sodium was required in order to balance at this concentration—in Potamon it is probably even less. To increase the concentration of a litre of distilled water to 0·05 mM./l. requires the addition of 50 µ.M. Na: in a crab of, say 15 g. this would represent a loss of only about 4% of the blood sodium.

The mean value for the potassium equilibrium concentration is 0·07 mM./l. It is interesting that this should be greater than the corresponding value for sodium. The loss of potassium necessary to produce this external concentration in a Etre of distilled water (70 µ,M.) is greater than the total potassium content of the blood of an average weight crab and much potassium must be released from the tissues.

Table 8 also shows the concentrations of potassium and sodium found in the lake water from which the crabs were collected. As in Astacus pallipes, the minimum equilibrium concentration for sodium is well below that found in their natural water and hence a large potential sodium uptake is normally held in reserve. This is not so for potassium: the minimum equilibrium concentration is only just below the normal lake water concentration. Apparently the rate of potassium uptake is not stimulated greatly by potassium loss and, in certain circumstances, the maintenance of the normal blood potassium concentration may become a real problem. The low concentration of potassium in the excretory fluid may well be correlated with the need for potassium economy.

Under conditions of salt balance the rate of salt uptake must just equal the rate of salt loss. Thus, the normal rate of sodium absorption must be 8 µ.M./io g./hr. Sodium uptake was demonstrated by transferring animals which were in sodium balance at their minimum equilibrium concentration, into a limited volume of a more concentrated sodium chloride solution. Under these conditions a net uptake of sodium occurred. The results of these experiments are shown in Table 9. The extent of the net uptake depended on the previous sodium loss. If the loss was great (as, for example, occurred when the animal was washed in running distilled water for several days) then sodium net uptake continued until the external concentration was reduced to near the original minimum equilibrium concentration (see animals 2 and 4 in Table 9). The original equilibrium concentration was not reduced by further sodium loss.

The similarity in the behaviour of Potamon niloticus and Astacus pallipes (Shaw, 19596) with respect to sodium balance and sodium uptake is striking—they probably differ only in the magnitude of the fluxes. If, as it appears, the sodium uptake mechanism in Potamon is influenced in the same way as in Astacus by external and internal sodium concentrations, then their effect may be represented quantitatively as in Fig. 2. The value for the maximum rate at 0·5 mM./l. (16μM./10 g./hr.) was calculated from the sum of the maximum observed net uptake rate (8 /xM./io g./hr. by animal 4, fourth solution, Table 9) and the normal loss rate (8μM./10 g./hr.). These relations account for the observed sodium movements. Thus sodium balance normally occurs at an external concentration of 0·5 mM./l. (curve 1, Fig. 2). A loss of 50 p.M. Na shifts the uptake curve to its maximum position (curve 2) and the animal will be in sodium balance at an external sodium concentration of 0·05 mM./l. If the animal is now placed in 0·5 mM./l. NaCl, the sodium uptake mechanism will be working near its maximum rate and a net uptake of sodium will take place. However, as soon as about 50μM. Na has been absorbed the uptake rate will adapt back to its original level and balance will be restored. If the animal had lost a large amount of sodium and was then placed in 0·5 mM./l. NaCl then again a net uptake of sodium occurs, but the uptake will continue at this rate until the large sodium loss has been made good. Adaptation back to the original rate will then take place as before.

This type of self-regulating sodium balance system may well turn out to be of common occurrence among the fresh-water Crustacea.

Potassium uptake was not studied in detail. Two experiments, shown in the lower part of Table 9, show that a net uptake of potassium does take place when the external concentration exceeds the equilibrium concentration. It is worth noting that at low external concentrations the sodium uptake rate may be sixteen times that of potassium despite the fact that the concentration gradients very much favour potassium absorption.

Fresh-water animals which are dependent on salt uptake mechanisms for the maintenance of salt balance may be limited in their distribution by environmental factors which may depress the activity of these mechanisms. For example, factors such as the salt content of the water, temperature, oxygen content of the water, pH may all be of great importance. Among the fresh waters of East Africa conditions may be found where any one of these factors might be significant in determining the distribution of the crabs. Although no detailed investigation was made of the effect of these factors on salt uptake, a number of observations indicate their probable importance and suggest some interesting lines for eco-physiological studies in the area.

The importance of the salt content of the water has already been mentioned. Potamon niloticus cannot maintain sodium balance in a concentration less than 0-05 mM./l. Na or potassium balance in a concentration less than 0·07 mM./l. K. Very soft waters are, however, found in the area with concentrations as low as this. For example, the River Sezibwa water (below the Sezibwa Falls, August 1957) contained only o-n mM./l. Na and 0·02 mM./l. K. Although the sodium concentration is just high enough to maintain sodium balance in P. niloticus the potassium concentration is too low. It was therefore interesting that no specimens could be collected from here although another species (probably P.johnstoni) was present. In a few laboratory experiments it was found that this species had a considerably lower minimum equilibrium concentration for sodium than P. niloticus (it was between 0·01 and 0·02 mM./l. Na).

The technique of bringing an animal into sodium balance in a limited volume of distilled water was used to demonstrate the effect of two other factors, temperature and oxygen tension, on the sodium uptake in P. niloticus. The animal was first allowed to come into balance and maintain it for a few days. Then either the temperature was reduced or the water was saturated with a nitrogen : air mixture and the effect of these new conditions on the external sodium concentration was followed. These experiments are illustrated in Figs. 3 and 4. In the case of the temperature change, a reduction of 10° C. upsets the balance and the external sodium concentration increases steadily.

It was of great interest to find an example of the possible limitation by temperature of the distribution of P. niloticus in the field. In Eastern Uganda the River Manafwa arises from near the top of Mount Elgon and flows down into Lake Victoria. Along the river there is a temperature gradient in the water from near o° C. at the source to 20-30° C. at its lower reaches. P. mloticus was found in large numbers in the lower parts of the river. The river was followed for some distance up Mount Elgon and over this section the midday temperature of the water gradually fell from 20 to 9-5° C. at the highest point. The crab became progressively scarcer and where the temperature had dropped to about 13° C. no P. niloticus could be collected. Water analysis of samples taken at various places up the river showed that the water had not changed appreciably in composition. Again it was fascinating to find that as P. mloticus disappeared it was replaced by another species, Potamonautes granviki, which was found up to the highest point. This species has also been recorded from the top of Mount Elgon in water whose temperature was about 2° C. (Prof. L. C. Beadle, private communication).

The effect of oxygen tension of the water on sodium balance is shown in Fig. 4. A reduction of the normal oxygen tension by half has very little effect on the sodium balance providing the supply of oxygen is sufficiently rapid. At lower tensions the animals go out of balance and the external sodium concentration rises.

Over much of Uganda there are large areas of swamp which border the edges of the lakes or form part of the river systems. In these regions oxygen tensions are invariably low (Carter, 1955). Again, it is interesting to find that Potamon niloticus is apparently unable to live in these conditions. Prof. L. C. Beadle has made extensive investigations of the swamps bordering the shores of Lake Victoria but has never found crabs in these areas, despite the fact that the animals are found abundantly on the rocky shores of the lake (private communication).

These preliminary ecological studies show that although P. niloticus is widely distributed there are a number of ecological situations in which it is apparently unable to survive. There may, of course, be many factors which determine its distribution, but it is certainly possible that the provision of suitable conditions for the maintenance of salt balance may be one of the most important.

The normal oxygen consumption of animals in lake water was measured in order to assess the percentage of the total metabolism involved in the maintenance of salt balance. At the same time the effect of the two environmental factors, temperature and oxygen tension of the water, on the general metabolism of the animal was studied. The results are shown in Table 10. The mean value for the normal oxygen consumption at 24° C. is 0·72 mg./io g. body weight/hr. Similar values were found by Schwabe (1933) for Eriocheir (0·67 mg./10 g./hr.) and Carcinus (0·90 mg./10 g./hr.).

Temperature has a marked effect on the oxygen consumption of Potamon: a reduction of 10° C. depresses it to about one-third of the normal value (Table 10). The effect of a reduction of the oxygen tension of the water may also be pronounced. The oxygen consumption is relatively unaffected by a fall in the oxygen tension to about half of the saturation value, but below this it falls off rapidly and is reduced to about one-third of normal at oxygen tensions around 1 mg./l. Since the effect of temperature and oxygen tension on sodium balance is paralleled by their effect on the oxygen consumption, it is probable that the depression of sodium uptake by these factors is not specific but is a result of the general fall in the metabolic level.

When the mechanisms of salt and water balance in Potamon niloticus are considered as a whole it is apparent that the animal only resembles a ‘typical’ fresh-water animal in one respect : namely, its inability to survive in the higher concentrations of sea water. At first sight it may appear that Potamon is ill-adapted to life in fresh water in that: (a) it has a relatively high blood concentration, (b) it is relatively permeable to salts, and (c) it produces an iso-osmotic urine. However, if one compares it with other brackish-water and fresh-water crabs, such as Carcinus and Eriocheir, it becomes apparent that this view is not really tenable ; in fact, Potamon is an example of another type of fresh-water adaptation which differs in certain ways from that shown, for example, by the crayfish. The most important factor in the adaptation of Potamon to fresh water has probably been an all-round reduction in the permeability of the body surface, both to salts and also to water. A comparison between the three crabs, shown in Table 11, brings this out. Here the rate of loss of sodium, the active uptake rate, the minimum osmotic work done in recovering the lost salts and the minimum percentage of the total metabolism involved in osmotic work are compared for these species. The figures for Carcinus are taken from unpublished data. Sodium loss in Eriocheir was calculated in two ways: the higher figure was derived from the loss through the urine (calculated from its concentration and the rate of urine production; Scholles, 1933) and from the fact that the urine loss represents only 14% of the total loss (Krogh, 1938); the lower figure was calculated from the initial rate of loss of sodium into distilled water (Krogh, 1938). The minimum thermodynamic work for the recovery of salts was calculated from the relation W=URT In Na₁/Na₀, where Na₁ and Na₀, are the internal and external sodium concentrations, respectively, W is the work done and U is the uptake rate. It was assumed that chloride values were in all cases the same as the sodium ones and the total work was therefore taken as twice the sodium work. Some interesting points emerge from this comparison. Eriocheir is three or four times less permeable than Carcinus, but nevertheless it uses a very large amount of its metabolic energy for the maintenance of salt balance. This difference in permeability is, however, of great functional importance. Although Carcinus, in 40 % sea water, uses less than 1 % of the total available energy for the purposes of salt balance, if this crab were to move into fresh water without reducing its permeability then the work required to maintain its blood concentration would be over 30 % of its total metabolism—surely an intolerable burden. If a crab like Carcinus were to penetrate into fresh water this would have to be accompanied by a reduction in permeability and, at the same time, a development of the uptake mechanism to operate at low external concentrations. Such a stage would then be similar to the situation found in Eriocheir. Potamon mloticus can be looked upon as illustrating a further development in this direction. With a permeability still lower than in Eriocheir (it is twelve times less permeable than Carcinus) the load on the total metabolism is further reduced to a reasonably low level. The special feature of Potamon is that the low permeability extends both to water and to salts, and with the result that salt loss through the urine is extremely small despite the fact that the fluid is iso-osmotic with the blood.

As a result of some theoretical calculations on hypothetical semipermeable animals, Potts (1954) reached the conclusion that the most important means by which an animal in fresh water can reduce the strain on its osmoregulatory mechanisms is by reducing the blood concentration and by producing a dilute urine. At first it may seem peculiar that P. mloticus, a highly successful fresh-water animal, should display neither of these features. The explanation lies in the fact that the basic assumption behind Potts’s calculations (namely, that animals are semipermeable) seems to be of very doubtful validity. It is certainly not true of P. niloticus’, nor is it true of Eriocheir sinensis, which Potts uses as an example to illustrate his arguments. Krogh (1938, 1939) explicitly states that in this animal only about 14% of the total salt loss occurs through the urine. This was also shown by Koch & Evans (1956) who found substantial sodium loss from animals with their excretory pores sealed.

If an animal is not in fact semipermeable then it is clear that the total osmotic work done in salt uptake will be greater than anticipated on this hypothesis; but a more important fact is that the conclusions which Potts arrived at for semipermeable animals are not valid for permeable ones. Thus, for example, if an animal like Eriocheir, where nearly 90% of the salt loss occurs through the body surface, were to produce a dilute urine, the saving in osmotic work could not exceed 10 % of the total and, in fact, would be less. The same also applies to the importance of the blood concentration : if the animal is salt-permeable then the osmotic work will be approximately proportional to the blood concentration when the animal is in the fresh water (actually it is proportional to (B— using Potts’s symbols). In the semipermeable animal, on the other hand, the work is approximately proportional to the square of the blood concentration (see Potts’s equations 7 and 9) and it is due to this that the blood concentration appeared to be of such special importance.

It is now apparent that the factors whose importance Potts stresses are only of especial advantage to a fresh-water animal which has achieved a differential reduction in the permeability of its body surface, i.e. it has decreased the permeability to salts but not to water, so that it has become essentially semipermeable. This may well have been the situation in an animal like Astacus. Certainly the rate of salt loss in Astacus is much lower than in Potamon and yet the water permeability (measured by the rate of urine production) is much higher. It is, of course, true that the minimum osmotic work done by Astacus is a good deal smaller than that done by Potamon, but this is due to the difference in salt permeability; if Potamon had reduced its salt permeability to the same level as has Astacus it would, in fact, do slightly less osmotic work than the crayfish, despite the fact that its blood concentration is higher and that it produces an iso-osmotic urine.

With these general considerations in mind it would seem worth while to attempt to extend Beadle & Cragg’s hypothesis on the steps involved in the invasion of fresh water by the Crustacea by emphasizing the importance of permeability changes. Beadle’s first stage involved the maintenance of a high blood concentration in dilute solutions. This must be achieved by a reduction in permeability of the body surface, together with the development of the salt uptake mechanisms. The next step involves a further reduction in the permeability. If this takes the form of an all-round reduction in permeability to both water and salts, as in Potamon niloticus, then there is little to be gained by a decrease of either the blood or the urine concentration. If, on the other hand, the reduction in permeability is restricted largely to salts and the animal remains relatively permeable to water, as in Astacus, then the production of a dilute urine or the further reduction in the blood concentration would have a pronounced selective value.

This work was carried out in the Department of Zoology, Makerere College, Uganda during the tenure of a Royal Society and Nuffield Commonwealth Bursary.

It is a pleasure to acknowledge the hospitality shown me by Prof. L. C. Beadle while working in his Department. I should also like to thank Prof. L. C. Beadle for kindly making all of the measurements of oxygen tensions and Prof. C. P. Luck who worked so hard to get the freezing-point apparatus working in time.