1. Introduction
It was because of a series of records prepared in connection with a study of the salt-water minnows that so successfully cope with the salt-marsh mosquitoes of the north (Chidester,. 1916) that the writer was invited in 1919 by the U.S. Commissioner of Fisheries, Dr H. M. Smith, to engage in an attempt at the physics and chemistry of the migrations of fish.
The subject is one that several experienced zoologists have refused to follow for any length of time, as it is difficult of field observation, and certainly not one to be solved by laboratory experiments.
After several summers of spare time spent in experimental work, amplified by field observations, it has been deemed advisable to survey the literature, and this paper will, it is hoped, aid in crystallising our knowledge of the migrations of anadromous fishes.
2. Parent Streams
To a zoologist who is looking for a problem that will eventually have a solution, it is most disconcerting to discover that there is much evidence that fish like the salmon are able to return in an apparently inexplicable manner to the streams where they were bred.
We are deeply indebted to Dr C. H. Gilbert of Stanford University for most careful work on fish scales that has shown that salmon return to the river of their nativity for spawning, and that they are even able to locate in that river the spot where they were once fingerlings.
Jordan mentions certain observations of Dr Gilbert on the Chinook salmon and Red salmon of the streams near Walla Walla, Washington. Dr Gilbert found that at a point where the salmon have a choice of two streams that come together under a bridge, the Chinooks (king salmon) go up either one of the streams indiscriminately, but the Red salmon (bluebacks) turn always to the stream with a lake.
In a recent paper Dr E. E. Prince, Dominion Commissioner of Fisheries, Ottawa, Canada, states (1916) that his previous report—published in 1896 and repeated in 1912—to the effect that “each river has its own race of salmon,” is borne out by more recent observations. A dark-fleshed race of Sockeye salmon inhabit a small creek near the Skeena River, and the salmon canners rarely net them as the meat is of a “dark, repulsive colour,” although the distance is not great from the waters that furnish the attractive pink salmon of the regular catch.
Dr David Starr Jordan is inclined to question whether migration up the parent streams may not be due to the fact that the young fish do not travel far from the mouths of the rivers that gave them birth, and consequently they find the parent waters pouring into the ocean and quite naturally return to them. Jordan also believes that the salmon are not unfailingly true to their native rivers.
Jordan has pointed out the fact that salmon do not apparently care about the quality of the water, and that while the Chilcoot River comes from a glacier and the water is milk white with glacial debris, the fish migrate into it. Jordan (1920) also cites a case of apparent attractiveness of lake water.
The first lake on the Yukon River, Lake Labarge, is 1800 miles from the mouth of the river. At Boca de Quadra, also a noted salmon stream, is a little stream about ten feet wide and less than a mile long, the outlet of a lake. At the head of the lake it is fed by clear springs. The fish go up the small stream just as they go up the Yukon, although they have only about five miles to go. Their start is accordingly a late one. Jordan does not explain this condition.
Further interest in lake-fed streams is aroused when we cite the case (mentioned by Jordan) of observations by J. P. Babcock of the British Columbia Fish Commission (Jordan, 1919). Babcock observed that in the case of water piped from a lake-fed stream across to another stream, the Red salmon gathered around the mouth of the pipe which contained the lake water.
The writer cannot refrain from remarking at this juncture that a fish responds to currents of water, and that it is quite likely that the Red salmon would have gathered around almost any non-poisonous fluid of the optimum temperature, providing it came with a little force through a tube.
Our knowledge of salmon behaviour has been clearly stated by Dr C. H. Gilbert in a personal letter, in which he answers several questions asked by the writer, and with his permission the answers will be quoted.
“Dear Sir,—I have no knowledge concerning possible factors governing migrations of fishes, and am thus unable to prepare any discussion of the subject that you may care to use. As regards migration of the salmon, however, there are certain unquestioned facts that must be taken adequately into account by any theory which claims to explain their movements.
“1. Salmon which are schooling together in the sea on feeding grounds far removed from their final spawning districts, will at the proper time separate and go their own way rather directly to the mouth of the distant stream from which they originated. At the time they separate, it would seem they must assuredly be exposed to identical environmental conditions.
“2. Salmon ascending a river together will react differently on approaching a given tributary, though again it would seem they must be exposed to the same conditions. There is a fair body of evidence to show that in the main they will re-enter the tributary in which they were hatched.
“3. In both the above instances, we seem wholly at a loss to suggest any purely external factors which condition and guide the migration, and yet lead to such diverse results when applied to fish of different early history.
“4. The ‘parent-stream theory,’ in so far as it has validity, is not a theory in any sense of the term, but a bald statement of fact. The salmon either do, or they do not, return in the main to their parent stream at maturity. We hold that they do, but we are far from claiming that this phrase, which is merely descriptive of their conduct, affords any explanation of it. When the quest is for the causative factors of migration, the ‘parent-stream theory’ is a confession of ignorance pure and simple. But it seems far more whole some and more hopeful of future results to confess ignorance where ignorance undoubtedly exists, than to set up spurious claims to achievement as is all too frequently done in this field. —Very truly yours,
C. H. GILBERT.”
Such clean cut statements as these backed by the observations of Dr Gilbert and his associates, Mr Henry O’Malley and Mr W. H. Rich (1919), who tagged adult Sockeye salmon and studied their migration, would seem to indicate that we have a long way to go in explaining the factors that influence migration.
It may be that we shall have to come to the idea of emanations from the mother, transmitted to the offspring, which serve to guide the fish of a particular race back to their parent stream.
It may not be out of place at this juncture to mention the statement of J. O. Snyder, who reported on the return of king salmon (1922).
In 1919, 2500 king salmon were marked by removing the adipose and the right ventral fins. They were expected to return to the Klamath River in 1921, on the basis of previous scale data. At that time twenty-three were recaptured. The age calculated from the scales was identical with the known age. The interpretation given is that the fish remained in the river during the first year, and most of them must have been in contact with the same environmental conditions of the sea during the second year. Snyder concludes that associations of individuals formed in youth may continue throughout life.
The remaining discussion will instance experiments and observations that show some advance in our knowledge of the determining factors that may in some way influence the migrations of fish.
3. External Influences
a. Physical
Character of the Bottom
The anadromous fishes migrate past all sorts of obstacles and over bottoms of various types in order to reach favourable spawning grounds.
It is the belief of Gurley (1902) that the selection of spawning grounds is by no means a question of chance, but has been determined by the egg type “via natural selection.” Gurley mentions the fact that spawning grounds of fishes are mainly of three kinds : mud, weeds and sand, gravel and rock, and that the habits of the fishes in deposition and the physical condition of the eggs have determined the survival of the fishes. One is immediately led to inquire whether the migratory fish do not occasionally become dispersed to new favourable localities for spawning and thereafter establish a race for that particular stream. We must have further observations and records to fully substantiate this statement so frequently given as a fact.
Prince has stated (1920) that the eggs and young of but few fishes are found on the sea bottom, and that the majority of marine fishes deposit floating eggs or else spawn inshore. He cites the work of Professor M’lntosh of St Andrews, Scotland, who proved that the eggs of the salmon will not develop in salt water, but become rubbery in consistency.
The migration of fishes seems to be unaffected by the character of the bottom, but their spawning places are quite certainly regulated by such factors. Observations of polluted streams prove, however, that fish will spawn in places that were formerly clean but have become so polluted that the eggs are not able to develop.
Stream Pressure
There is no question that the factor of stream pressure is extremely important in connection with the migrations of fish.
Chamberlain (1907) writes of the tendency in dry seasons of salmon to tarry inshore in bays and the mouths of small streams, and only to proceed up to their spawning grounds when the floods have come rushing down.
Prince (1920) mentions the importance of a down-floating stream, but emphasises the necessity for shallow, clean, gravelly rapids in the case of the salmon, and gives weight to the added fact that the cold waters of the salmon rivers are usually well aerated, since coldness increases the power of the water to absorb gases. Further mention will be made of both of these points.
The significant experiments of E. P. Lyon (1904, 1909) performed with the killifish, Fundulus heteroclitus, Linn., the scup, Stenotomus chrysops, Linn., the stickleback, Gasterosteus bispinosus, Walb., and the butterfish, Poronotus tricanthus, Peck., indicate clearly that fish respond to stream pressure through the optic and tactile senses.
By an ingenious series of devices, Lyon proved that the fish Fundulus responds to relative motion rather than to a current; that the moving “optical field” is the stimulus, and that in the complete absence of pressure the fish will respond to movement outside glass containers. He concluded that the current only stimulates by moving the fish away from their environment.
Blinded fish on the bottom of a long box with open ends (screened with coarse netting) oriented themselves in the sand. Other fish that were blinded and placed in the tideway leading from the eel-pond at Woods Hole, Mass., were quite irregular in their motion until they touched the bottom and then they headed upstream. Even contact with a bit of eel-grass was enough to give them orientation.
The general cutaneous nerves were the ones involved.
Dr Caswell Grave has emphasised to the writer the supposition that variation in turbidity of the water may therefore be a most important difference in the orientation of the fish to streams, and that debris in the water at the time of floods will serve to accentuate tactile stimuli.
Lyon showed that in the case of concentrated streams such as those from a small bore tube or flume, there is sufficient difference in the velocity of the water at the centre and the outside of the stream to furnish the requisite stimuli. In his experiments with fish blinded in one eye, Lyon proved (1909) that they react to currents like the normal fish.
The writer found in 1919 and 1920 (Chidester, 1922) that by increasing the stream pressure of fresh water or of toxic substances in solution and presenting them in a trough that was almost parallel to its twin with sea water flowing into a long receiving trough, it was possible to lure killifish (Fundulus heteroclitus, Linn.) into the unfavourable stream. It is quite possible that the idea of intoxication by chemicals, suggested by Shelford, is not always tenable, and that occasionally the fish in his experimental tank moved towards the toxic substances because they were attracted by the force of the entering stream. His control and experimental streams were situated at opposite ends of the tank, with the outlet at the middle.
It must of course be recognised that the factors of temperature and oxygen content of the water are significant, but we must also appreciate the fact that just as birds respond to air currents, fish not primarily confined to the bottom respond most definitely to the force of moving water.
As pointed out so aptly by Lyon (1904) the fish in rushing torrents orient because of the difference in velocity of the adjacent parts of the stream. In the wider and deeper streams, sight or contact with the bottom or with floating particles in the water furnish the requisite nerve-stimuli.
Rutter (1902) has shown that the Chinook salmon run into the rivers against ebb tide, and then, when the tide turns, run out against flood tide. The flood tide is not continuous for so long a time as the ebb tide, and so the fish keep ascending until the tidal movement is replaced by the river current.
Nothing could be more striking or more worthy of application to lessons on human behaviour than the fact that weakened and spent salmon go down to the sea tail first, feebly fighting the current.
Lights and Shadows
Fish vary considerably in their movements according to light or to darkness. Salmon move both by night and by day. Atkins (1874) mentions the fact that the Atlantic coast salmon spawned at night or on a dark cloudy day. Alewives rest in small pools at night and travel by daylight. Herrings apparently require clear water and a clear sky for their migratory movements.
Shad are said to be timid and easily frightened by shadows. M’Donald cites the case (1884) of the shad, and indicates that floods and muddy water arrest their movements.
Allen, in discussing the migration of mackerel (Mackerel and Sunshine, Allen, 1909), has pointed out that sunshine produces more of the plant food of Copepoda, such as diatoms and Peridinidæ, and that an abundance of Copepoda offers rich food for the mackerel. He attempted, therefore, to “correlate the average quantity of mackerel per fishing boat taken in May with the total number of hours of sunshine recorded during the first quarter of the year.” The effort was unsuccessful, however, and he states that “whilst the 1905 temperature maximum agrees with the maximum total catch of mackerel the temperature of 1903 is accompanied by low catches of mackerel.” Fishermen of the Atlantic coast find that they are more successful in catching mackerel on dark days, and according to Mr Ernest Romling of the U.S. Bureau of Fisheries, they believe that this is due to the fact that their scattered bait (chum) is present in the absence of the living Crustacea that appear near the surface on bright days.
Reeves has concluded (1919) that the longer wave lengths of light have a physiological or psychological effect upon sunfish (Eupomotis gibbosus, Linn.) and horned-dace (Semotilus atromaculatus, Mitchill) that differs considerably from light of shorter wave length and from white light.
The fish in her experiments were trained to respond to food stimulus with the accompaniment of blue light. Then it was tried with red light of brightness approximating that of the blue for the human dark adapted eye. The fish responded about as often to red as to blue at first and then became discriminating and red had to be reduced in intensity.
Parker and Larchner studied the responses of Fundulus to white, black, and darkness (1922). They prepared three boxes, one lined with dull white paper and exposed to the light from a 100 watt lamp: the second lined with black paper and similarly exposed to the 100 watt light: and a third, absolutely light-proof. Funduli placed in the light one remained light in colour ; those placed in the dark one remained dark in colour ; but the ones in the light-proof box remained light until removed and then became dark for five minutes, later becoming light again. Temporarily blinded fish also remained light coloured. These experiments, while of no special interest in connection with the problem of fish migration, serve to emphasise the optical factor in behaviour, and to corroborate the work of Sumner and of Mast along the same line.
The writer has shown (Chidester, 1923) that Fundulus heteroclitus, Linn., when given a choice of temperature and light combinations, is influenced somewhat by the attractiveness of intense light, and may be attracted towards water that is several degrees warmer than that which is its optimum at that time, without the added factor of light.
We must conclude that fish vary in their responses to light so much that it can be stated that some are primarily influenced in their reaction to currents by their optical sense, while others are primarily responsive to turbidity and to powerful stream pressure increased by floating debris.
Temperature
By far the most evidence has been gathered to support the theory that temperature is the primary influence in fish migration.
Gurley (1902) has prepared a highly instructive paper discussing the relations between temperature and the other factors in fish migration. He holds that there is probably a temperature-responsive nerve-mechanism. This mechanism would explain why spawning in cooling water is associated with migration from warmer to cooler water.
Gurley points out that the time of spawning has been determined by natural selection, and that natural selection has therefore determined the time of migration.
To his statement that spring spawning does not necessarily mean spawning in warmer water, the writer most heartily subscribes. In connection with observations on the killifishes of New Jersey (Chidester, 1920), it was observed that the early spring migrants to the marshes came in April when the waters inshore were cooler than they were at the time in the Raritan Bay.
Jordan says (Science Sketches, 1888, pp. 51-52) that Blue-back salmon and Humpback salmon ascend only snow-fed streams with sufficient volume to send their waters well out to sea. Gurley adds to this the statement that spring freshets mean heavy spring runs and lighter fall runs. Gurley believes most salmonids to be spawners in cooler water. He states that migration to cooler water favours the development of the gonads. Gurley has also pointed out that the fish probably starts to migrate as the result of a temperature stimulus, then continues in response to the pressure stimulus from currents, and that it runs upstream until the gonads have ripened their products and spawning takes place.
Natural selection is the factor that determines the survival of eggs spawned in water of a certain character.
Chamberlain claims (1907) that fish leavę cooler water for warmer. He also points out that streams with lakes have a higher temperature in the summer than streams of similar volume without lakes. Certainly in the case of some fish such as the menhaden it would seem that they prefer warmer waters offshore in the summer. They are attracted by water above 10° C., and may even migrate inland until the water reaches a temperature of 23° C. (Goode, 1879).
Stevenson (U.S. Fisheries Report, 1898) quotes M’Donald as stating that “warm rains produce vast schools of shad.” Floods and muddy water arrest their movements, however.
Mr R. A. Goffin and Mr W. L. Howes of the Woods Hole Laboratory of the Bureau of Fisheries record the fact that alewives come into Oyster Pond soon after the ice leaves in the spring when the temperature is not more than 5° C. The swift currents flowing out of the pond cause the fish to rush in great shoals and almost fill the stream. They are travelling from warmer to cooler water.
Shelford and Powers showed (1915) that herring are sensitive to temperature differences as small as 0. 2° C.
Galtsoff (1923) in a paper on the migration of mackerel in the Black Sea shows that as the temperature decreases the fish press inshore. They are caught in horizontal nets close to the shore. Apparently the fish are not affected by absolute temperature but by the relative temperatures. The same phenomenon occurs when the temperature drops from 24° C. to 21° C. as happens in the drop from 17° C. to 10° C., as observed near Sebastopol. Galtsoff holds that the difference in salinity is not significant, as the same phenomenon occurs in regions of the Black Sea with quite different content of salts.
Johnstone (1908) concludes that it is difficult to separate the influences of salinity and temperature, but there is no doubt that temperature is a factor in itself.
He believes that there is sufficient evidence to conclude that seasonal migrations of fish are related to seasonal changes in temperature, and cites the work of D’Arcy Thompson, who has correlated the volume of catches made by Aberdeen trawlers with the temperature of the sea on the fishing grounds.
Ward (1920) discusses the migration of the Sockeye salmon and finds that the volume and the swiftness of the streams are not apparently determining factors in the selection of one tributary to a stream in preference to another. He further shows that great difference in turbidity is ineffective, and that there is no difference in the apparent chemical character of the waters. The food supply is apparently uniformly poor.
Temperature is the factor that he believes of paramount importance, and he subscribes to the theory that the salmon prefers steadily flowing cool water for spawning.
Whatever influences may be important for migration besides temperature, the writer holds that we must acknowledge the fact that it bears a most important relation to chemical processes, and thus involves not only the environment of the fish but the metabolism of the body of the animal itself. Even if we decide that the fish has some mysterious sense that man cannot discover, it is quite evident that the temperature of the water will have a pronounced effect on the utilisation of that sense.
b. Chemical
Food
Fish vary considerably in their habits as respects feeding when they migrate inshore.
Young immature fish continually travel from the deep waters of the ocean to the shallower waters of the coast and up into the marshes in search of food. Although several species may be in the same shoal, in general it is found that those travelling together far into the inland streams during the period prior to maturity are of the same species. In the ocean the young of several species will be found together.
At the same time that immature minnows come to feed on the teeming insect life of the salt marshes (Chidester, 1916, 1920), one notes that older fish of the same species are migrating inshore primarily for the purpose of spawning.
During the period of travel from the deeper waters, there is usually little feeding done by the mature animals that are to spawn in brackish or fresh water. In no fishes is this characteristic more striking than in the salmon.
Gurley (1902) believes, and it would seem rightly so, that abstinence from food in fishes is in direct ratio to the length of time required as a minimum and the amount of time available as a maximum, for the species to reach their spawning grounds.
Prince holds (1920) that the salmon has never changed its habits but “repairs to the ancestral breeding localities, regardless of the geological and topographical changes wrought in the course of long centuries.” He states that the surroundings of the spawning beds have been changed and the former salinity has been changed to freshwater conditions ; but the connecting channels still remain, and the salmon persists in its habit of going to the old spawning locality.
Whether we accept this as the explanation of the habits of the salmon or not, we must acknowledge that the fish travels for 2000 miles up the rivers of the Pacific Coast to spawn, and that it fasts during the period after it leaves the ocean for fresh water.
Greene (1914) is responsible for a most careful study of the storage of fat in muscular tissue of the salmon. He found that the salmon do not feed in fresh water, no matter how far they travel into it to spawn. The stored fat is apparently the only source of the fats to build up the ovaries, and it also must furnish the energy for muscular activity during the long pilgrimage to suitable spawning grounds.
It is stated by other competent observers, however, that certain salmon will seize bait while migrating up the Columbia River.
The explanation of the fact that a salmon will seize bait when migrating is probably that while the animal is too strongly impelled towards its spawning grounds to stop for feeding en route, it is quite responsive to the optical stimulus of something that is apparently edible.
In the case of fish that spawn inshore, such as the cod, it is stated by Mr R. Hamblin of the U.S. Fisheries Laboratory at Woods Hole, Mass., that although the manuals of fish-culture indicate that they do not eat during the spawning season, both the females and the males will eagerly seize food placed in their detention pools.
It is probably safe to conclude that fish that spawn inshore in salt water or brackish water will feed until just before they are to spawn, while those that travel far inland press onward, responding by optical and tactile senses to current stimuli, until their eggs have matured and they must spawn. The distance which any particular race of fish travels before spawning is in all likelihood determined by natural selection, together with the factor of time required for the development of the gonads.
Salinity, Osmotic Pressure
It has been shown that the anadromous fishes have blood that is more saline in sea water than it is in fresh water. It has also been found that changing the medium for marine fishes to a less saline one will induce change in their blood. The internal equilibrium of the animal has been preserved by the presence of membranes that are practically impermeable to salts, though permeable to water.
Mather (1881) has listed thirty-three species of fishes that can live in both fresh and salt water. Slow acclimatisation permits these animals to range between the two media with impunity.
There is great variation in the adaptability of fishes to salt or fresh water, and this adaptability is probably due to an inherited resistance.
Sumner (1905, 1906) has studied the effects of changing certain brackish and freshwater fishes to pure fresh water and to water of different degrees of salinity. He mentions changing the young of the Chinook salmon (Onchorhynchus tsawylscha) which weighed from 8 to 30 gms., and which had been reared in fresh water, abruptly to water of a density of 1.013 without any harm, and suggests that a higher salinity would not have injured them.
Rutter (1904) found that if the young of Pacific Coast salmon were transferred alternately from sea water of low to high dilution, it was possible gradually to raise the salinity. This reminds us of the gradual change for the fish that must result when it travels with the ebb and flow of the tide before it finally begins the long journey in fresh water. Rutter found that young quinnat salmon were able to live in higher salinities as they grew older, and from 25 per cent, sea water at six days of age they were able at two months to live in almost pure sea water.
The age of Fundulus heteroclus embryos was found by Loeb (1894) to be definitely correlated with their ability to survive the addition of different proportions of NaCl to sea water.
Bert (1871, 1873, 1883) found that he could gradually acclimatise freshwater fishes to live in water of one-half the salinity of the sea, and that they would survive abrupt transfer to diluted sea water if the proportions were two parts distilled water to one part of sea water. He explained the death of freshwater fishes in salt water by osmotic action on the gills, producing contraction of the capillaries and thus asphyxiation. With scaleless fishes, osmotic action took place over the entire surface of the body.
Fredericq (1885) indicated that the gills, while permeable to gases, are almost impermeable to salts of the sea.
Bert (1883) mentioned the fact that if mucus was removed from skins of freshwater eels, they were quite susceptible to salt water.
Garrey (1905) experimented with Fundulus heteroclitus,, denuding about one-half the body surface of scales from a large number of them. When placed in fresh water, normal sea water, and a mixture of distilled water with an equal quantity of sea water, it was found that the fish survived in the sea water of one-half its normal concentration, but died in the other solutions. Garrey concluded that if the internal and the external media are approximately isotonic, no ill effects result. He holds that injury to the integument furnishes an opportunity for fatal osmotic action.
Sumner (1906) was not able to confirm Garrey’s work.
Trout and other freshwater fishes must be handled with great care when stripped of their spawn.
There must be considerable variation in the susceptibility of fishes to injuries, as many salmon arrive at their spawning grounds with bodies “torn by rough stones, gashed by jutting rocks, maimed by jagged and precipitous obstacles, or by falling, time after time, down almost impassable falls” (Prince, 1920).
Ringer (1883), working with minnows and goldfish, tried the salts of sodium, potassium, and calcium to discover which were capable of sustaining life longer and found that CaCO2 would sustain life longer than a corresponding quantity of sodium or potassium salts. Combinations of KCl1 Na2CO3, and CaC12 aided materially in supporting life. When fish were placed in solutions in which many fish had previously died, the new lot lived as they gained from the organic salts thrown off by the dead fish.
Loeb (1900, 1905, see 1915 for references) has introduced the term physiologically balanced solutions, and defines them as solutions in which the toxic effects are annihilated, which each or certain of their constituents would have if they were alone in solution. He has pointed out that the specific injurious action of NaCl, which destroys the impermeability or semi-impermeability of the membrane, may be counteracted by the addition of CaCl2 or a mixture of KCl plus CaCl2
Temperature is a factor that influences the toxicity of salts. Loeb and Wasteneys (1912) made experiments with Fundulus caught in January in L. I. Sound and kept thereafter in a laboratory at 10° C. The fish were changed to Ringer’s solution and to sea-water, and it was found that the highest temperature they could withstand was 33° C. and the highest concentration was m/4. The higher the concentration of the salts the more resistant the fish were to the higher temperature, up to 33° C. Above that the resistance decreased. Once immunised against the higher temperature the fish withstood transfer to high salinity after being kept in an ice chest.
The authors make two suggestions by way of explanation. First, that rise in temperature brings about changes in the surface of the cells of the body whereby the latter loses its protective impermeability. Second, that under the influence of higher temperature a substance is formed that protects against the effects of higher temperature. Formation of this substance is gradual. Thus the writers explain how animals once immunised against high temperature could be kept alive.
The osmotic pressure of the blood of fishes has engaged the attention of numerous workers, including Fredericq, Mossa, Quinton, Rodier, Botazzi, Garrey, Sumner, and Scott.
As discussed by Scott (1913) the common method of determining the osmotic pressure of a solution is by the determination of its freezing-point. The depression of the freezing-point of the solution below that of pure water is proportional to the osmotic pressure of the solution. The Beckman thermometer is commonly used to determine the freezing-point depression or Δ.
Fredericq (1885 et seq) showed that the osmotic pressure of both elasmobranchs and teleosts was about one-half that of sea water, while the blood of marine invertebrates was nearly isotonic with their medium.
Bottazzi (1896, 1897, 1901) used the cryoscopic method of determining the freezing-point of the solution and then the osmotic pressure, and his results were in accord with those of Fredericq.
Rodier (1899, 1900) and Garrey (1905) have confirmed the results of Bottazzi on teleosts and elasmobranchs.
Garrey found (1905) the freezing-point of the blood of teleosts to have a mean value of — 0.872° C., while the mean value of that of the blood of the elasmobranchs of the Woods Hole region was — 1.92° C. The value of A for the sea water was approximately — 1.82° C.
Greene found (1904) that the freezing-point of salmon blood was for sea salmon, — 0.762° C. ; for brackish water salmon, — 0.7 37° C. ; and for spawning ground salmon, — o.628°C.
This decrease is 3.3 per cent, for the brackish water salmon and 17.6 per cent, for .the spawning ground salmon. This work is significant as it indicates the independence in relation between the water and the blood-composition for the salmon. Greene is not disposed to ascribe the fall of concentration of the blood of 17.6 per cent, to direct absorption of water in fresh water, but to place much weight on the fact that the animals are from eight to twelve weeks in fresh water without food.
Dakin (1908) found that the freezing-point of the blood of the flounder, Pleuronectes flesus, in the North Sea was 0.83° C., while in the River Elbe in fresh water its blood had a freezing-point of 0.68° C., a decrease of 15 per cent,
Scott (1916), from whom many of the references in this section were obtained, studied the freezing-point of the blood of the white perch, Morone americana, which lives equally well in fresh or in salt water. When taken from Tashmoo Pond, Marthas Vineyard, the blood showed a Δ. of 0.635° C. The water was slightly brackish. After remaining in running tap-water for a day, perch from this same pond were examined and their blood showed a Δ. of 0.571° C., similar to sea water. Still others were placed in sea water for two days and the of their blood was 0.766° C.
Scott concludes (1916) that anadromous fish are ‘‘ able to adapt themselves to a degree to the great changes in the osmotic pressures of the external medium, which they meet in passing from salt to fresh water, or vice versa, by a slight corresponding change in the osmotic pressure of the blood.”
Wells has shown (1915) that starvation may cause certain fishes to seek water of lower concentration of salts and others to seek water of higher concentration of salts. Possibly this may be correlated with the migration of some of the anadromous fishes back to the ocean after spawning. It does not, of course, explain their journey to the fresh water to spawn.
The work of Shelford and of Wells indicates that fishes are probably not so sensitive to salt ions as they are to hydrogen and hydroxyl ions.
Acidity or Alkalinity
Hydrogen-ion Concentration
Wells showed (1915) that freshwater fishes select slight acidity in a gradient, when the other choices are neutrality or alkalinity.
Shelford and Powers demonstrated (1915) that alkalinity and acidity are more important in the behaviour of the herring than is salinity.
Moore, Prideaux, and Herdman showed (1915) that the hydrogen-ion concentration (carbon dioxide tension) of sea water is subject to seasonal variation. So far investigations of such seasonal change in freshwater streams have not been published.
M’Clendon showed (1916) that changing the PH very much killed marine animals more quickly than changed salinity.
Shelford (1918, 1918 a) has emphasised the toxicity of acids from pollution and has also brought out the fact that increased acidity is accompanied by the liberation of CO2-He believes that acidity and alkalinity are extremely powerful factors in fish migration.
Mayor (1919) has shown that it is possible to detect ocean currents by a study of the hydrogen-ion concentration.
Krogh and Leitch (1919) have studied the oxygen-unloading tension of the hæmoglobin of the blood in the plaice, cod, eel, carp, pike, and trout. They found that in the carp, eel, and pike, which are subjected occasionally to low oxygen tension (Powers suggests also CO2 content), the oxygen content for unloading at 1 5° C. was low, 2 to 3 mm., while the oxygen unloading tension of the cod was 18 mm. and that of the plaice and trout was about 10 mm. Low temperature decreases the efficiency of hæmoglobin for carrying oxygen. A small amount of carbon dioxide present serves greatly to diminish the affinity of the hæmoglobin for oxygen, so that in the carp, eel, and pike with a CO2 tension of 1 per cent, the unloading is increased to 7.S mm.
Roule (1914 et seq.) has long contended that salmon directed by an actual need, migrate towards a .richer supply of oxygen.
Powers (1921) cites the experimental work of numerous authors to prove that before fresh water or marine fishes exhibit oxygen want, the oxygen content of fresh water and of sea water as well must be reduced to about 1. 7 to 0.4 c.c. of oxygen per litre. He also shows that the oxygen content of the waters where Roule’s salmon were found was in excess of the amount necessary for the apparent well-being of the fish.
Working with herring and salmon smolt near the Puget Sound Biological Station, Powers found (1920, 1921) that the greatest number of herring apparently preferred water with a PH varying from 7.76 to 7.73, while the salmon smolt were present in water ranging from PH 7.98 to 8.08. His field studies were preceded by careful laboratory experiments with relatively small numbers of fish of several species.
In correlating the work of Krogh and Leitch and others with his own observations on the marine fishes, Powers asks (1921) what part the CO2 tension plays in nature as affecting the unloading tension of the hcemoglobin in the blood of fishes. He states : “The very fact that they found that the fishes which are subjected occasionally to lower oxygen tension and higher carbon dioxide content have a lower unloading oxygen tension than those not subjected to these conditions; and the fact that the experiments and field observations recorded (Powers, 1921) show that the fishes which live on the bottom and among the vegetation did not react to a gradient of hydrogen-ion concentration, while the more freely swimming forms did so and were found also in water having a PH at or near that to which they reacted positively in the experiments, are very suggestive.”
Powers concludes that the hydrogen-ion concentration, or the carbon dioxide tension of the water, has a considerable influence on the movements of pelagic fishes.
Miss Jewell has emphasised (1922) the significant fact that fish are exceedingly sensitive to changes from the hydrogenion concentration to which they were accustomed.
Coker (1923, 1923,a) has made important studies of the relation between acidity and the occurrence of brook trout and other fishes in freshwater streams. He urges the importance of further studies on this important environmental actor. The trout is less adapted than other fishes to live under conditions of low oxygen pressure. Coker has been much struck with the apparent preference of brook trout for acid streams, and with the correlation between the observations of Gardner and Leetham (1914) that trout in saturated water consume less oxygen at the lower temperatures.
It is of interest in this connection to examine the tables of Fox showing the c.c. of oxygen in a litre of water at different temperatures and salinities, when the water is saturated with this gas (Murray and Hjort, 1912, p. 254).
“At 30° C., a litre of water saturated with oxygen contains little more than half as much oxygen as at 0° C. Normally there is more oxygen in the cold water-masses of the Arctic and Antarctic regions than in the warm water-masses of the tropics. Salinity is not such an important factor in the solubility of oxygen as temperature” (Murray and Hjort, 1912).
Significant work has been done by Lillie and Shephard (1923) on the relation between the PH and reactions of the worm Arenicola to light. In this particular animal, pos1t1ve heliotropism requires almost neutral or slightly alkaline reaction.
As we proceed further with studies of alkalinity and acidity we find that they bear important relations to the distribution of plants and animals (M’Gregor, 1921; Arrhenius, 1921; Atkins, 1921, 1923; Coker, 1923; Powers, 1921,et seq.; Shelford, 1923).
Like other factors discussed in this paper, hydrogen-ion concentrations must be reckoned with in field studies of the future on fish migration. The writer inclines to the belief of Shelford, Powers, and Coker in the great significance of this addition to the apparatus of the true field ecologist.
Pollutions
Much work has been done on the subject of stream pollution, and the situation is becoming increasingly grave as great industrial plants persist in discharging their wastes into creeks and rivers. There is no question that fish are repelled by poisons in the water. Some species are so extremely sensitive that they will not pass through a polluted area, while others will continue on to their native spawning grounds.
Waste substances may be classified as toxic-repellent, toxic -attractive, and non -toxic oxygen -absorbing. The toxic-repellent wastes have a direct effect in driving fish from streams ; the toxic-attractive wastes such as gas wastes may lure the fish to their death (Shelford, 1917, 1918).
Non-toxic oxygen-absorbing wastes, such as the wastes from milk plants and fish canneries or from city sewage, may putrefy and absorb the oxygen in a limited area. These last are believed ultimately to furnish fish food by increasing the organic matter that will support the small organisms upon which many fish feed.
Useful summaries of the literature have been prepared by the U.S. Bureau of Fisheries (Stream pollution, 1919), and by Suter and Moore (Stream pollution studies, N.Y. State Conservation Commission, 1922), while significant investigations by Prince (1899), Knight (1901), and others indicate the belated interest that is being taken in the preservation of the natural resources of our waters.
Suter and Moore (loc. cit.) have prepared a table that shows how much dilution of wastes is necessary before fish life may be preserved in a stream.
Tolerance of fishes to trade wastes (Suter and Moore, 1922) :—
No dilution required—sawdust and treated sewage.
Dilution less than 1: 10—raw sewage, fibre factory wastes.
Dilution from 1 : 10 to 1 : 100—spent dyes, paper mill wastes.
Dilution r : 100 to 1 : 1ooo — gas manufacture wastes, wastes from bleacheries.
Dilution 1 : 1000 to 1: 10,000—caustic lime, bichloride of mercury, etc.
Dilution 1: 10,000 to 1 : 100,000—lime, strong acids, gas tar.
Dilutions greater than 1 : 1,000,000—copper sulphate, bleaching powder.
The disappearance of many of our food fishes from the coastal streams is undoubtedly due directly to the influence of wastes, and in most cases it would be possible at small expense to arrange for the discharge of such waste into tanks or pools where much of value might be reclaimed. The chief difficulty seems to be in arousing the interest of the owners of the factories concerned, as many of them are content to do things in the old way, regardless of the interest of the public or even themselves.
4. Internal Influences
Senses
There is great variation in fishes as regards their use of the senses. Some are predominantly olfactory or gustatory, while others are dependent upon vision for directing their movements. In all, the temperature sense is of paramount importance.
Olfaction, Gustation (Chemical sense)
Parker* has recently summarised the work of many authors on the question of the chemical sense in fishes. In the migratory teleosts with which we are concerned, the olfactory pits are dorsal and not connected with respiration, but are purely sensory in function.
Parker (1910, 1911, 1913), Sheldon, and others have demonstrated that fish have such a well-developed olfactory sense that they can detect odorous substances in the water at quite considerable distances. Parker concludes that fishes have a chemical sense dependent upon free nerve endings ; smell, dependent upon the olfactory nerve which is a rather highly developed distance receptor; and taste which is dependent upon the taste buds.
Shelford and Powers (1915) have successfully shown that the herring and other fishes are sensitive to variations in the salinity of the water. The determination of acidity or of alkalinity (Powers, 1920, 1921; Jewell, 1922; Shelford, 1915, 1923; Coker, 1923) unquestionably is dependent upon the chemical sense and probably in part upon the allied senses of smell and taste.
Shelford states (1915) : “It is not necessary to appeal to instinct to explain the return of certain salmon to certain rivers or the running of herring in certain localities, since their origin in the region and limited tendency to leave it (Johnstone, 1908), coupled with their ability to detect and follow slight differences in water, is a sufficient explanation of all their peculiar migrations.”
Obviously the important work of Parker, Shelford, Powers, and others on the delicate sensibilities of fishes to changes in the salinity and PH of the water, has a most direct bearing on the migrations of fishes and an interesting relation to the problem of pollutions.
Parker has suggested to the writer that it is barely possible that a certain race of fish may give off emanations that differ chemically from those of other races, and that one might attribute the return of individual races to home streams to their power to sense the familiar emanation. While this would seem to be a difficult thing to prove, it is worthy of considertion in view of man’s unwillingness to permit things to go unexplained !
Vision
As has previously been pointed out in this paper, the work of Lyon (1904, 1909) indicates clearly that fish will react sharply to currents even if the currents are not directly in contact with them, providing they are able to see the bank alongside.
We are not concerned with colour vision of fishes, and it is of only passing interest to note that Sumner (1911) and Mast (1916) have discovered that flatfish are unable to simulate their background in pattern or shade unless the eyes are functioning.
Hess believes that fish are colour-blind (1908, 1912). White believes that there is evidence for a certain degree of discrimination (1919). Reeves (1919) holds that in the sunfish and the horned dace at least there is discrimination of light of longer wave-length from light of shorter wave-length and from white light. She associated a feeding response with the stimulus of restricted wave-length. (See Reeves, 1919, for an excellent bibliography, and Washburn, The Animal Mind, Macmillan, 1917, for a discussion of sensory reactions of vertebrates.)
In the section under “lights and shadows” in this paper we have indicated already that while fish vary considerably in their light sensitivity, there is no question that their ability to migrate long distances is in part due to the visual sense. That light may even influence the reactions of a fish to temperature has been shown by the writer (1923) in the teleost Fundulus heteroclus. Other work by Cole (1922) on a salamander, and by Jordan, H. E. (1917), and Crozier (1918), on fishes, indicates the importance of both vision and cutaneous photosensitivity in the behaviour of fishes.
Tactile and Kinæsthetic Senses
In connection with the orientation of fishes to currents, we find that the work of Lyon, Parker, Chidester, and others indicates clearly that the senses involved include touch and the kinæesthetic sense.
Parkr has shown (1902) that in the teleosts, surface waves and current action stimulate fishes in which the lateral line nerves have been cut, and he assumes, therefore, that the general cutaneous nerves for touch are the ones involved.
Lyon, in a paper on compensatory motion in fishes (1900), demonstrated that if the tail of a dogfish is turned to a given side the dorsal and anal fins will bend towards that side, as when the skin is stimulated. Parker comments (1909) on the fact that a 2 per cent, solution of cocaine applied to the skin of the tail will cause this latter reaction to disappear.
Parker (1909) found that in the dogfish tactile stimulation produces movements of the nictitating membrane and the fins. He divided the surface of the body of the dogfish into five tactile regions characterised by the response resulting from their stimulation. Mechanical stimuli induce fin motion and are unquestionably the cause of rapid changes in the direction of swimming.
The writer (Chidester, 1921, 1922) was much impressed by the fact that contact with a swift stream would attract the fish in experimental troughs away from more favourable and otherwise more attractive liquids flowing at that time with less force. Poisonous and normally repellent solutions could be made temporarily more attractive by increasing the force of the current.
Lateral Line
While much has been done by other workers, including Schulze (1870), Bonnier (1896), and Lee (1898), G. H. Parker has contributed the most satisfactory investigation on the function of the lateral line in fishes.
He has shown (Parker, 1902, 1905) that the lateral line organs of several species of teleosts are not stimulated by light, heat, salinity, food, oxygen, carbon dioxide, foulness of water, pressure of water, currents, or sounds. They are stimulated by vibrations of low frequency, about six per second; and experiments as well as histological investigations indicate that the lateral line organs are intermediate in function between the organs of touch and the ear, which is, of course, sensitive to vibrations of high frequency.
Parker would hold that while the lateral line organs are not sensitive to currents, and therefore not particularly important to the migrating fish in its orientation against moving water, they are stimulated by surface wave-movements produced by air on the water or by objects falling into the water, and thus may be important in warning the fish of otherwise disturbing influences.
On the other hand, Hofer (1908) holds that in Parker’s operations on fishes, when he cut the lateral line nerves he also destroyed the nerves supplying the skin of the head. Hofer believes that the skin nerves are the ones affected by slow vibrations in the water, and that the true function of the lateral line organs is response to streaming movements in the water.
Whether we accept the theory that the skin nerves are the ones that determine orientation to currents, or that the lateral line nerves are the ones involved is, of course, debatable ; but the writer holds to the assumption of Parker that the general cutaneous nerves are the ones chiefly involved in rheotropic response, while the lateral line nerves serve in an auxiliary capacity.
Hearlog
Parker (1902, 1908, 1911, 1912) has summarised the work of Kreidl, Lee, and many others on the ear in fishes, and references to their work may be found in his accompanying bibliographies. Parker has concluded that in the Squeteague (Cynoscion regalis) the utriculus has to do with equilibrium and muscular tonus, while the saccular organ is the chief organ of hearing (1908). Parker finds that while most sounds are repellent to fish, some may be lures to particular species (1911). The surface between the water and the air is a screen through which little sound passes, but sounds that actually reach the water are according to Parker quite effective stimuli.
Bernoulli, working with eels and trout, found (1910) that pistol shots as well as the muffled sound of a bell were not stimuli to the fish.
He concluded that the reactions of fishes to stimuli are tactual and visual responses to the mechanical motion of the water, and not true auditory responses.
It is quite probable that the roar and vibration due to waterfalls are attractive to many fishes, rather than repellent, and whether the fish actually hear the sounds as man does, or only sense the vibrations through the pressure of the waves, is of little moment in this problem. The writer, however, believes that Parker has proven his point for the minnow, Ftmdulus heteroclilus, and the Squeteague, Cynoscion regalis.
Temperature
One of the greatest factors in the behaviour of fishes is their response to temperature change.
We find that, by means of heat and cold corpuscles in the skin, fish proceed towards warmer or cooler water, and that it is the relative temperature rather than the absolute temperature that determines their activity. The work of numerous investigators has been summarised elsewhere in this paper, and so we may content ourselves at this time with pointing out the following.
Herring are sensitive to very slight temperature changes, according to Shelford and Powers (1915), reacting to differences as small as o. 2° C.
It has been shown that temperature changes influence the reactions of fishes to light, salinity, PH and to internal factors such as the developing gonads.
Unquestionably this sense is the greatest single factor that we find influencing the metabolism and behaviour of fishes. It must even be considered in connection with the instincts of the species.
Physical Condition of the Body
Size
There is apparently no relation between the length of the fish and its ability to migrate. It. is quite certain that fishes travel upstream and leap waterfalls, even when they are at almost the maximum growth for the species.
As regards the weight, we find that migratory fishes vary from a few ounces in weight to as much as 85 lb., the weight of the Chinook salmon.
Fat
It has been conclusively shown by Greene (1913 a and 1913b) that fat is the immediate source of the energy expended by the salmon during the spawning migration. The following discussion is quoted from Greene’s papers. Salmon fat is stored in the body during the stage of feeding and reaches a maximum at the time the feeding stage ends at the beginning of the migration fast. Storage (according to Greene) is in muscles and the intermuscular connective tissues. Storage tissues of minor importance are cutaneous and other adipose tissues such as the liver, the alimentary tract, and the skeleton.
The superficial dark muscle is loaded with fat at all stages of the life cycle, especially as the spawning migration begins. The deep lateral pink muscle is loaded with intermuscular fat at the time of maturity, but has little traces of fat during the feeding stage.
The pancreas of the salmon is diffuse, scattered over the pyloric cæca, mesenteries of the stomach and intestine, the inner loop of the stomach, and the mesentery of the spleen.
The pancreas is abundantly active during the migration fast, producing an internal secretion that is rich in lipases chiefly discharged into the tissue-spaces, reaching the blood stream.
The pancreatic lipase produced during the fasting period is “chiefly discharged into the tissue-spaces, reaches the blood stream, and is transported by the circulation to the tissues of the body, including the fat-storing tissues and the active fat-using muscles” (Greene, 1913a).
In one of Greene’s investigations he has shown (1913b) that the primary function of the large number of pyloric cæca of the salmon stomach is fat-absorption, with the intestine also an important region of fat-absorption. The cardiac and pyloric types of epithelium of the stomach also absorb fat.
These investigations offer convincing evidence of the power of fat-storage that is so necessary to an animal migrating thousands of miles without feeding en route. That there is a relation between the condition of the body with respect to stored fats, and the time when such an animal begins to migrate into fresh water, goes without saying.
Specific Gravity
In connection with a study of the air-bladder and the specific gravity of fishes, Taylor has pointed out (1922) that a fish which increases in fat-content diminishes in specific gravity. He states that “at 29.34 per cent, fat, the fish without an air bladder would be in equilibrium with sea water, and at 46.55 per cent, fat (if it were possible) would float in fresh water without an air bladder.”
Taylor believes that a fish living at sea and accumulating fat will find itself more at home physically in fresh water without enlarging the air-bladder, and asks the pertinent question, “May not this be the influence which directs salmon and shad from salt water to the mouths of rivers?”
It is possible for fish to adjust their specific gravity by reducing the size of the air-bladder when they migrate from fresh water to salt water, and by secreting oxygen into the bladder, and thus enlarging it, when they move from salt water to fresh water. Taylor concludes that increased fat renders navigation in salt water more difficult and fresh water the better physical medium. This evidence links up well with the observations of Greene, Paton, and Hjort on the body composition of salmon and herring at different stages.
Gonads
Gurley, in a paper on “Biological-Empirical Psychology” (1909), adds to his earlier notes on the migrations of fishes the interesting statement that in the fishes we have an “intoxication factor,” since paralleling the seasonal development of the reproductive organs, in-shore movement begins, inherited from the parents who responded to similar stimuli and were able to spawn successfully. Obviously the parents selected proper environments for such spawning, else there would have been no offspring to continue the habits.
Johnstone, in a small book Life in tlie Sea (1911), cites the intoxication theory of Gurley, and states that the “elaboration of an internal secretion from the ovary or testis produces an intoxication and impels the fish to seek water of definite physical conditions.” The writer of this review believes that the influence of hormones from the gonads is responsible for increased activity of the fishes in any case, and that their subsequent behaviour is in part dependent upon other factors. These factors, which have been discussed elsewhere in the present paper, have been the basis of much conjecture for years.
Temperature
It is not necessary to do more than mention, by way of reminder, what our literature survey has already shown that every factor under discussion is regulated by that rise in temperature which increases chemical action and also increases, up to a limit of course, the speed of the response of an animal to stimuli. With increased metabolic activity in the living organism a high temperature becomes intolerable after a time, and the fish seeks cooler waters.
Chemical Conditions
Without attempting to discuss the matter in detail again, we may point out a few facts relative to environment, and indicate the relation of the changing chemical condition of the animal to its environment.
Roule has upheld the thesis that salmon migrate towards a richer supply of oxygen, because of an actual need for it. Powers and others have concluded that the carbon dioxide tension of the water determines to a large extent the movements of pelagic fishes.
We must agree that metabolic activity accompanying fat-accumulation and gonad-development will create new needs as regards the environment, and that while the ensuing restless movements of the fishes may not always direct them immediately to the needed new locality, they do result in securing optimum conditions for many who will perpetuate in their descendants the habit of return.
It is likewise evident that fishes inherit a certain degree of permeability, with an optimum salt-content of the blood and of the tissues, and that the ability of the present-day fishes to adjust themselves to changed external conditions is probably dependent upon a capacity that was established in their ancestors during a period following marked continental upheaval, with the receding of saline waters.
We may conclude, therefore, that while the physical and chemical demands of the fully-fed or fattened fish with developing gonads, direct it towards cooler and fresher water, it continues upstream because of the stimulus of running water (stream pressure), favourable in oxygen content.
Unfavourable chemical conditions will kill the eggs of fish that are not vigorous enough to reach the spawning grounds that have been established for that species. Whatever may have been the salinity of the water when our present-day anadromous fishes began spawning inshore, it is certain that now, in the case of many species, their maturing eggs demand both plenty of oxygen and a salinity that is much less than that of sea water.
While this is not the place to discuss the migration of the eel, it must be evident from the work of Schmidt that we have to consider the internal condition of the body in catadromous forms as well.
5. General Conclusions
Gilbert and others have shown that three or more years after they are spawned, mature male and female salmon will separate into races belonging to the parent stream and return to ancestral spawning grounds.
While fish will migrate past obstacles and over bottoms of a distinctly unfavourable type, they will seek out spawning grounds that are most favourable. Survival of the fit determines the habitual characteristics of the grounds for any given race.
Fish respond to relative motion, the moving optical field acting as the stimulus, but they also respond to the force of moving water, reacting to tides and currents.
There is much variation in the reaction of fishes to light, but they share with other animals a sensitivity to such stimulus, and their behaviour in the presence of other stimuli is affected by lights and shadows.
Temperature is the most important factor in fish migration, since it affects not only the chemical processes of the environment but also has a profound influence on the rate of metabolism of fish and their response to external stimuli.
Search for food is an important agent in the dispersal of fishes and leads many species to travel from the deeper waters of the ocean to the shallows offshore, or even up into fresh water, before they are mature. It bears little relation to spawning migration, however.
While variation in salinity is important in determining the range of some species, we have records of thirty-three species (Mather) of fishes that readily acclimatise themselves to fresh and salt water. Anadromous fishes adapt themselves to changes in osmotic pressure by changing slightly the osmotic pressure of the blood.
There is some evidence for the belief that anadromous fishes, as well as those passing their whole life-history in either salt water or fresh water, are susceptible to slight changes in acidity or alkalinity. That hydrogen-ion concentration bears an important relation to the range of dispersal of pelagic fishes seems to be well established. Roule has long contended that the salmon responds to a need for higher oxygen content. Powers and others emphasize the carbon dioxide tension or PH as the important factor, more so than salinity.
Fish are driven from the mouths of rivers emptying into the ocean by the pollutions from factories, canneries, and mines. In some cases they may spawn in regions too polluted to permit the eggs to live, and so a race of fish may be destroyed for that particular stream.
By means of the olfactory, gustatory, and chemical senses, fish trace food and distinguish favourable from unfavourable waters. That they are able to detect emanations from ancestors and thus return to their parent streams is rather unlikely, although it is an interesting conjecture.
Vision is important in connection with the progress of fishes upstream and undoubtedly plays a great part in their return to former haunts.
The tactile and kinæsthetic senses influence the characteristic reaction to fishes to currents which enables them to travel far up rivers and creeks.
The lateral line sense bears little relation to migration in fishes, while the poorly developed sense of hearing is even less important.
Temperature sense, located in the heat and cold corpuscles of the skin, is extremely important in connection with migratory movements, and is most striking in its manifestations, as only slight relative changes in the temperature of the water are sufficient to bring about the movements of vast shoals of fishes.
Size and weight in themselves have no appreciable relation to the migration of fishes, except as they influence the specific gravity and rate of metabolism.
Fat not only furnishes energy for long migrations, but it is, according to Taylor, the cause of movements from salt to fresh water, since it reduces the specific gravity so much as to force some species to seek a less buoyant medium.
Developing gonads exert an influence on the movements of fishes, since hormonic stimulation increases bodily activity and creates a demand for some new, definitely favourable environment. The environment sought is one that ancestral success has selected as the optimum.
The chemical condition of the body, with reference to oxygen content, carbon dioxide tension, lactic acid accumulation, and salt content are all important factors in reactions to stimuli and in creating a need for movement to a new environment.
In all probability, the initial movements of anadromous fishes are from deeper waters to the shallower regions inshore, and until maturity they are primarily directed by the search for food and by temperature differences. At maturity, the course into fresh water is determined by increased gonadial development and stored fat, which create an urgent need for water with less buoyancy and more available oxygen.
Migration up a given stream is determined in part by its accessibility at the time of the greatest internal urge, and steady travel is due to the reaction of a fish to current stimulus under favourable conditions of temperature and with a chemical content of the water that is at that time the optimum. The distance travelled is regulated by the period of development of the eggs after the animal has become fully fed, and this physiological factor is of course determined by an inherited tendency.
References
G. H. Parker, 1922, Smell, Taste, and Allied Senses, J. B. Lippincott Company, Philadelphia, Pennsylvania.