ABSTRACT
Minnows kept in a tank are active during the day and quiet at night. Their behaviour is reversed if they are given hollow bricks in which they take cover and so avoid bright light. When cover is available they are very active at sunrise and sunset.
Minnows have no inherent daily rhythm of locomotory activity.
Blind minnows respond to daily variations in light intensity, and are more active at night than during the day.
Minnows will not cross a light-dark boundary when the intensity on the light side is greater than 0.17-0.08 m.c
A minnow shoal disperses between intensities of 0.024 and 0.0034 m.c.
Minnows catch Dapknia better in bright light than in the dark. The change from visual to ‘dark’ feeding takes place between 0.0007 and 0.00007 m.c.
Minnows appear to avoid bright light by a comparison of intensities if the light-dark boundary is sharp, but they may also respond to light photokinetically.
It should be noted that these results, obtained in the laboratory, may not be true for minnows under natural conditions.
INTRODUCTION
Many freshwater and marine animals appear to have an optimum range of light intensity. Under natural conditions pelagic species keep within their range by making a vertical migration from deep to shallow water during the evening and returning to deep water the next day. Reversed migrations are also known. While the pattern of migration can be correlated with changes of light intensity, the reactions of individuals cannot be studied easily in the field, and there is the question as to how the animals find their optimum and remain within it as it moves up and down the water column. Cushing (1951) has reviewed the problem for the planktonic Crustacea, but so far little work has been done with fish. Observations were therefore made on the locomotory activity, shoaling and feeding behaviour of minnows in relation to light intensity to add to an understanding of the responses of fish to light.
MATERIAL
The minnows, Phoxinus phoxinus (Linn.), used in the experiments were caught in the River Cam and kept in a stock aquarium exposed to normal variations in light intensity. They were fed on chopped liver, living Tubifex and Daphnia. As judged from their lengths (Frost, 1943), they were in their second or third years and were sexually mature.
LOCOMOTORY ACTIVITY
Locomotory activity was measured by an automatic recorder (Jones, 1955). The swimming of the fish disturbed a celluloid vane suspended in the water by a work-hardened resistance wire. The movements of the wire completed or broke a circuit, and the number of times that this occurred was automatically counted and recorded on a rotating kymograph drum. The activity of the fish was determined from the number of makes and breaks each hour or half hour. The fish were kept in a glass tank placed in an easterly window well exposed to daylight, and at night to the much lower intensities from street lighting and occasional road traffic. The tank was surrounded by black paper so that the fish would not be disturbed by seeing anyone enter the room. To prevent sudden temperature rises in the summer months water was circulated through the tank, and on very hot days it was shaded from direct sunlight. The rate of flow was so adjusted as to be without effect on the activity recorder. In some experiments the water temperature was recorded continuously by a thermograph, and measurements of light intensity were made with an R.C.A. 931 A photo-multiplier (Jones, 1955).
Experiments were made with individuals or groups of two to eight fish and lasted from 2 to 10 days, and were carried out over a period of 18 months so as to cover the complete reproductive cycle. During each experiment food was usually provided by a stock of living Tubifex. Six series of experiments were made, the conditions of each being as follows :
Cover available in the tank, provided by two small hollow bricks in which the fish could hide.
No cover available.
No cover at first, bricks added later.
Cover available, tank later blacked out by light-proof box.
No cover, tank later blacked out by light-proof box.
Fish blinded by removal of eyes, no cover available.
Results typical for each series of experiments are given in Figs. 1-5. When cover was available (Fig. 1, Ser. 1) the fish were more active at night than during the day. At sunset there was a burst of activity which declined during the night to be followed by a second burst at sunrise. When there was no cover (Fig. 1, Ser. 2) the pattern of activity was reversed, activity being greatest during the day, but dawn and dusk peaks are not obvious, although there were suggestions of them in some records. The high level of activity recorded during the day was not due to the fish schooling with their reflexions on the side of the tank, as it was still recorded when the inner surface was painted matt black or when an earthenware crock with an unglazed inner surface was used in its place. The activity of the minnows in these two series of experiments is related to changes in light intensity (Fig. 2), while the third series (Fig. 3) shows that the pattern can be reversed by altering the experimental conditions. Individuals or groups gave the same results, and the presence or absence of food appeared to have no effect; experiments were not made with starved fish. From April to June, during the breeding season, the minnows failed to respond to the changes in light intensity. They never took to cover during the day when it was available, and their level of activity was generally very high. The fourth and fifth series of experiments (Fig. 4) failed to reveal any trace of an inherent or endogenous rhythm of locomotory activity. Similar negative results were obtained when fish were taken from the stock aquarium and kept in the dark in a sound-proof room for several days.
After sunset, the blind fish (Fig. 5) showed a great increase in activity, which declined during the night. There is a suggestion of an increase at dawn, after which the activity remained low throughout the day. The pattern of activity does not appear to bear any relation to temperature. In the record shown in Fig. 5 the rhythm weakened towards the end of the experiment. In two of the blind minnows the melanophores were expanded, as is usual, but in the third they were contracted and the fish was pale. When experiments were made with the fish individually, the pale fish gave the most clear and consistent records of low daytime activity followed by an increase at sunset, and possibly at sunrise, with a return to the day level.
When cover was available the minnows hid in the hollow bricks during the day. Sometimes they came out for a few minutes, but they only appeared to swim freely in the tank at night. Daytime observations showed that the fish usually remained just inside the bricks, occasionally moving forward until the head and eyes cleared the shade, pausing and then backing into cover again. The fish appeared to be making a comparison between the intensities in the brightly lit tank and the shade of the hollow bricks. Their emergence from cover at night and when the tank was blacked out (Fig. 4) suggests that they avoid bright light, and the peaks of activity at sunrise and sunset may be in response to the rate of change of light intensity which is of the order of 1 log unit in 15 min. However, in one instance, shown in Fig. 2, the peak occurred before any rapid fall in intensity. In one experiment, measurements were made of the light intensity at which the minnows emerged from and retired to cover.
On three successive nights, intensities of 1·8, 5 and 55 m.c. were recorded for emergence and 68, 78 and 4·6 m.c. for retiring.
The low night and high day activity recorded when no cover was available, and the sudden drop in activity when the tank was blacked out (Fig. 4), could be the result of a photokinesis which might, under natural conditions, be part of a mechanism to ensure that they reach cover of lower intensities. Once cover is reached, however, it seems likely that they would remain there by a comparison of intensities, provided the light-dark boundary or gradient was sharp enough.
While no inherent rhythm of locomotory activity was found in minnows kept in continuous darkness, Spencer (1939) found that goldfish had a diurnal activity rhythm which persisted, although weakening, for several days in continuous light. On the other hand, Harder & Hempel (1954) found that the normal activity rhythm of plaice was broken by exposure to continuous light. While slight variations in the melanophore index of minnows persist in the dark (Pauli, 1926), and the movements of the rods and cones of Ameiurus nebulosus can be detected for at least 2 days under similar conditions (Welsh & Osborn, 1937), melanophores and rods and cones remain in the light-adapted phase under continuous light. Clausen’s (1936) results on the oxygen consumption of the black bass, Huro salmoides, are sometimes quoted as demonstrating the persistence of metabolic activity under constant environmental conditions. But as his fish were kept in chambers covered by a cloth ‘loose enough around the upper end of the chamber to allow an inspection for air bubbles’ it is possible that they could have been stimulated by variations in light intensity.
It was not surprising to find that blind minnows react to diurnal variations in light intensity as their pineal organs and adjacent regions of the mid-brain are light sensitive (von Frisch, 1911 ; Scharrer, 1928). Their reactions, however, are the reverse of the reactions of normal fish without cover, and it is interesting to note that while the response of naturally blind cave-dwelling characins is dependent on the degree to which the pineal area is exposed to light, removal of the apparently functionless optic cysts makes them indifferent to light (Breder & Rasquin, 1947). Removal of the eyes in minnows might reduce their overall sensitivity to light, which would account for their low activity during the day, while the increase in activity at sunset, and perhaps at sunrise, suggests that the pineal complex is particularly sensitive to changes in light intensity. If this were so, a fish whose melanophores were contracted, so increasing the exposure of the pineal area, might be expected to respond better than a fish whose melanophores were dispersed, as was found to be the case. But clearly more experimental work is needed here.
REACTIONS AT A LIGHT-DARK BOUNDARY AND SHOALING BEHAVIOUR
The results of the experiments on locomotory activity suggested that minnows avoid high light intensities, and observations were therefore made on their reactions at a light-dark boundary to determine the value of the upper limit. A large aquarium tank, 120 cm. in length, 60 cm. wide and 45 cm. deep, filled with water to a depth of 15 cm., was used in these experiments, which were carried out in a sound-proof and light-proof room. Half the tank was roofed and curtained to the water surface. A 100 W. lamp, was suspended 1 m. above the middle of the tank and its intensity was varied by an autotransformer. Measurements of light intensity were made with a neon discharge tube photometer, similar to that described by Poole & Poole (1930). The photometer was calibrated against a uniplanar tungsten filament substandard lamp. Several experiments were made with five to six minnows in the tank which were provided with Tubifex as food.
In bright light the minnows remained in the shaded half of the tank, up against the far wall, and never approached the light-dark boundary. When the light intensity was very slowly reduced, the fish swam, in a group, more freely in the shaded half of the tank, but were very hesitant on approaching the boundary and never crossed over while the intensity of the light falling on the water surface in the unshaded half of the tank was greater than 0·17-0·08 m.c. Even then there was some hesitation on approaching the boundary, and it was not until the light was reduced to 0·024 m-c-that they swam freely from one half to the other. At intensities of this order the minnows no longer shoaled together as they did in bright light and further observations were therefore made to determine the intensity at which the shoal broke up. For these experiments the roof and curtain were removed. Shoaling behaviour was clear at 0·08 m.c., but at lower intensities the fish became more and more independent of one another, and the shoal appeared to break up at intensities between 0·024 and 0·0034 m.c. At lower intensities the minnows were very active and excited, breaking the surface as if feeding. They could be heard breaking the surface when the light was too low for observation, but after several minutes this stopped. When the light was turned up, the fish were seen to be resting, independently of one another, on the bottom.
While the method for reducing the light intensity was not entirely satisfactory as there was a shift to the red in the spectral composition of the light as the voltage was reduced, the results suggest that minnows avoid intensities greater than 0·17-0·08 m.c., and that their shoaling behaviour breaks down at intensities between 0·024 and 0·0034 m.c. While it is well known that shoaling behaviour is mainly dependent on vision (Morrow, 1948), and that shoals break up at night, no measurements are available to compare with the present results. Feeding is another activity which, in the minnow and other fish which actively catch their food, could depend on vision and might be related to light intensity. Herring, for instance, are not filter feeders but catch their prey (Hardy, 1924) and are said to be able to feed at intensities of the order of moonlight, but not by starlight or in the dark (Mŭzinić, 1931 ; Battle, Huntsman, Jeffers, Johnson & McNair, 1936; Johnson, 1939). Verheigen (1953) found that herring shoals break up gradually with decreasing light intensity, but that feeding continues at lower intensities; when feeding stops, the herrings are still able to avoid obstacles. To complete the picture for the minnow, experiments were therefore made to determine what effect light had on their feeding ability.
LIGHT INTENSITY AND FEEDING
Using the apparatus shown in Fig. 6, experiments were made to see how many of fifty living Daphnia two minnows could catch in half an hour at different light intensities. The lower light intensities were obtained by using a torch bulb and neutral density filters and the highest by using a 60 W. lamp. Control experiments were made in the dark. The two minnows used in each experiment were drawn at random from a stock of fourteen fish which had been fed only a few Daphnia each day for a fortnight before the series commenced. After the minnows had been left in the light-proof box for half an hour to become adapted to the conditions of the particular experiment, fifty Daphnia were added through the feeding tube. Half an hour later the box was opened, the fish quickly removed and the remaining Daphnia recovered and counted.
The results are summarized in Table 1 and shown graphically in Fig. 7. Over 90 % of the Daphnia were eaten at intensities greater than 0·008 m.c. The minnows did not do so well at lower intensities, but they took 40 % of the Daphnia in the dark.
The results show that minnows depend in part on vision for feeding and that the change from visual to ‘dark’ feeding takes place between 0·0007 and 0·00007 m-c-No attempt was made to find out how the Daphnia were captured in the dark.
DISCUSSION
The results are summarized in Fig. 8, together with values of various light intensities found under natural conditions, and some results of other authors which fit in with the general picture. As judged by their behaviour in the light-dark boundary experiments, minnows avoid intensities greater than 0·17 m.c. ; a shoal would break up at an intensity a little below that of moonlight, and they would just be able to feed by eye at the surface under starlight. The change from visual to ‘dark’ feeding between intensities of 0·0007 and ·00007 m-c-be related to the change from cone to rod vision, which Brunner (1935) concludes takes place at 0·008-0·002 m.c. But it should be noted that her results with the minnow differ from those of Wolf & Zerrahn-Wolf (1936) with the sunfish Lepomis. For this fish it appears that cone vision changes to rod vision at 0·4 m.c., which is about the same intensity at which the change takes place in man.
Minnows appear to keep below their upper limit of light intensity by a comparison of intensities, that is, by a taxis, but the results of the experiments on locomotory activity suggest that they also respond to light photokinetically. Evidence of a photokinesis in teleosts has been given by Shaw, Escobar & Baldwin (1938), who found that the swimming speed of goldfish fully adapted to intensities of 30 and 550 m.c. is about twice that observed when the fish are adapted to intensities less than 0·5 m.c. Similarly, Schlagel & Breder (1947) found that blind cave characins consume more oxygen in the light than the dark, and Woodhead & Woodhead (1955) have found a photokinesis in larval herring. As a comparison of intensities can only be made if the light-dark boundary is sharp enough, it would be interesting to see how minnows behave in a light gradient of gentle slope and whether they keep below their upper limit by a kinesis or a taxis under such conditions.
If minnows kept below their laboratory limit in the field, they would be expected to hide under stones or weeds during the day, and, if swimming over deep water, to keep below 30 m. during a summer day. In Windermere, Frost (1943) found that minnows five in the littoral zone, hiding during the winter months under stones at depths of 1·5-2 m., and in the late spring, summer and autumn swimming pelagically in depths of 0·3-0·6 m. Under natural conditions then, large shoals of minnows appear to behave quite differently when compared with small groups in the laboratory. However, minnows may not spend all their time in such shallow water as during the summer they mainly feed on cladocerans and copepods which migrate to deeper water during the day (Ullyott, 1938), and the minnows themselves are the principal food of perch over 16·5 cm. in length, which at this time of the year are most abundant at depths of 6-12 m. (Allen, 1935). But in shallow streams there is little doubt that minnows live under light intensities which, under laboratory conditions, they would avoid if cover were available. Under such conditions the minnows would be expected to be very active throughout the day. In the laboratory, complete disregard for light intensity was only shown during the spawning season. Here, at least, is a measure of agreement between the experimental and field observations.
ACKNOWLEDEGMENTS
This work was carried out in the Zoological Laboratory, University of Cambridge, and I wish to thank Prof. Sir James Gray, F.R.S., for letting me work in his department. While the work was carried out I was in receipt of a grant from the Development Commission.