Environmental Biology of Fishes Vol. 20, No. 1, pp. 49-65, 1987 0 Dr W. Junk Publishers, Dordrecht.

Locomotor activity patterns of nineteen fish and five crustacean species from the Baltic Sea

Lars Westin & Gunnar Aneer Ask8 Laboratory,

Keywords:

Institute of Marine Ecology, University of Stockholm,

Brackish water, Biological rythm, Eye structure, Light-dark

S-106 91 Stockholm,

Sweden

cycle, Light intensity

Synopsis

Activity patterns of some common Baltic fish species and macrocrustaceans have been investigated in natural light/dark conditions - in most cases for periods of at least one full year. The fauna of the northern Baltic proper consists of species of both marine and freshwater origin, with both types represented in the study. The different patterns found are discussed in relation to light period, light intensity and to the structure and function of the eyes. The following species were considered nocturnal: ldothea baltica, Mesidothea entomon, Gammarus oceanicus, Palaemon adspersus, Anguilla anguilla, Spinachia spinachia, Pholis gunnellus, Zoarces viviparus, Scopthalmus maximus, and Platichtys flesus. These species were considered diurnal: Clupea harengus, Scardinius erythrophthalmus, Gasterosteus aculeatus, Perca fluviatilis, Gobius niger, Taurulus bubalis, and Cyclopterus lumpus. The next three species showed a crepuscular pattern of activity: Praunus flexuosus, Esox lucius and Gymnocephalus cernus. Four species were considered to show inversions of the activity patterns: Pungitius pungitius, Pomatoschistus minutus, Myoxocephalus scorpius and M. quadricornis.

Introduction

Over the past 20 years a wealth of data has accumulated on biological rhythms of activity in both terrestrial and aquatic animals (for reviews on all aspects of rhythmicity see Harker 1964, Aschoff 1965, Btinning 1967, Menaker 1971, Mills 1973, Palmer 1976, Schwassmann 1980). Many studies have demonstrated the persistence of rhythmicity under constant laboratory conditions, indicating the presence of endogenous control. However, in nature, rhythms are both synchronized and phased by environmental cycles. From an ecological point of view it is essential to understand the impact of the external cycle on the behaviour of the animal in the natural habitat. To

identify how rhythms are integrated is often difficult but is of extreme importance in understanding how the animal lives and also in predicting when and where certain events are most likely to occur as well as when an animal may be predisposed to respond to a particular stimulus. As environmental factors of high latitudes undergo large seasonal variations our intention has been to cover the annual locomotor activity patterns for each species and also to discuss the patterns found in relation to light period, light intensity and eye structure. The present paper describes the locomotor activity patterns of 19 fish and 5 crustacean species occurring in the northern Baltic proper.

50 Materials

and methods

This investigation was undertaken in the Baltic at the Asko Laboratory (58” 50’ N, 17” 38’ E), south of Stockholm, Sweden, between December 1969 and July 1976. The Baltic is one of the largest brackish water areas in the world with low but stable salinity conditions. However, a salinity gradient exists from about 2%,S in the northernmost Bothnian Bay, to about 15-20%0 S in the southwestern part of the Baltic proper (Voipio 1981). As salinity is an important controlling factor, only a limited number of euryhaline marine and freshwater organisms can establish themselves in the Baltic. Tables 1 and 2 give information regarding the investigated species and their main distribution in relation to salinity. Fish were collected in the vicinity of the laboratory and immediately placed under experimental conditions. The method used was that employed by Miiller & Schreiber (1967): a circular, doublewalled aquarium, equipped with a window for a photocell to the light beam. The lamps were shielded with infra-red filters. This method records the fish only at one particular point. The photocell was connected to a printer counter (Elmeg) which recorded the number of passages per hour. The die1 variation in activity was calculated as percentage deviation of 2 h values from the corresponding 24 h mean (Fig. 1). The aquaria were made of fibreglass reinforced plastic. Three different sizes were used, 1.0,0.6 and 0.3 m in diameter with inner diameters of 0.3, 0.2 and O.lm, respectively. The aquaria Table 1. Activity patterns found for investigated experiments. (M) = marine, (F) = freshwater, Species

& origin

Mysidacea Praunus ,flexuosus (M) Isopoda Idothea baltica (M) Mesidothea

Amphipoda Gammams Decapoda Prrlaemon

0

0

4

IGg. I. Explanatory die1 activity pattern monthly 24 h mean. (= 5 lux level).

8

12 16 HOURS

20

diagram for Figure 4. The diagram shows given as percentage deviation from the Vertical lines indicate sunrise and sunset

were supplied with running sea water pumped continuously from an inlet at about 10m depth. The salinity was about 6.5%0 S. In winter the water was slightly heated to prevent freezing in the inlet tubes. Otherwise the temperature varied according to the conditions in the sea (Fig. 2). The continuous flow of sea water through the tanks was arranged in such a way that currents were minimized. Natural habitat bottom material and vegetation as shelter were provided in each aquarium. The experiments were carried out in a building where the windows faced east and west and only natural light (nLD) was provided. During two winter periods, white sheets were used to cover the

crustaceans, numbers (B) = brackish water.

of specimens

in the expeAments,

time of year and duration

Time of experiments (duration, months)

Number of specimens

Activity

October

25

Crepuscular

(1)

entomon

(FB)

All year (20) All year+(l6)

20 5-10

Nocturnal Nocturnal

oceanicus

(M)

All year

20

Nocturnal

5

Nocturnal

adspersus

(M)

June

(1)

24

(14)

of the

pattern

51 windows to simulate natural winter light level conditions in the aquaria (maximum light intensity <20 lux) as described by Muller (1970). A twilight relay connection to a Minigor Z event-recorder was used to record the time for passing the 5 lux level at dawn and dusk on the inside of the building. The 5 lux level was used as an indicator of the

shift between day and night (Andreasson & Muller 1969). The test animals were fed at irregular intervals. The food consisted of immobilized food objects of species normally occurring in the food of each speties. For a restricted period salmon and trout food pellets of varying sizes were also used.

Table2. Activity patterns found for investigated fish species, numbers of specimens in the experiments, time of year and duration of the experiments. (M) = marine, (F) = freshwater, (B) = brackish water.

Species and origin

Time of experiment (duration, months)

Number of specimens

Activity pattern

Clupea

Jul (0.75)

l&4

Diurnal

All year but Ott & Dee (10) Jul (0.3)

1

Diurnal-crepuscular

l&5

Diurnal

Jul-Nov (5)

2&2

Nocturnal

All year (41)

l&10

Diurnal (one winter nocturnal)

All year (18)

4-5

Inversion

All year but Jul-Sep (13) All year (21)

1

Nocturnal

1

Diurnal

All year (25)

1

Crepuscular

Aug (0.25)

1

Nocturnal

All year (29)

1

Nocturnal

Nov-Mar

(5)

1

Diurnal (inversion?)

minutus

All year (21)

5

Inversion

scorpius

All year (43)

l&2

Inversion

All year (17)

l&2

Inversion

All year (31)

1

Diurnal

All year but Aug-Sep (14) All year (15)

1

Diurnal

1

Nocturnal

All year (27)

1

Nocturnal

harengus

Baltic herring t M) Esox

lucius

Pike (F) Scardiniw

erythrophthalmus

Rudd (F) Anguilla

anguiila

European eel (M) Gasterosteus

aculeatus

Three-spined stickleback (M) Pungitius

pungitius

Nine-spined stickleback (F) Spinachia

spinachia

Fifteen-spitted stickleback (M) Perca

fluviatilti

Perch (F) Gymnocephalus

cernuus

Ruffe (F) Pholis

gunnellus

Gunnel (M) Zoarces

vivipaws

Eel-pout (M) Gobius

niger

Black goby (M) Pomatoschistu::

Sand goby (MJ Myoxocephalu!

Father-lasher (M) Myoxocephalus

quadricornis

Fourhorn sculpin (FB) Taurulus

bubalis

Sea-scorpion (longspined) Cyclopterus

lumpus

Lumpsucker (M) Scopthalmus

maximus

(M)

Turbot (M) Platichtys

fleslrs

Flounder (M)

52

MONTHS

2. Mean sea surface temperature at the Asko Laboratory (1969-1973). Vertical bars indicate standard deviation. Fig.

It is well known (e.g. Eriksson 1978, Helfman 1978, Manteifel et al. 1978) that activity patterns of individual fish undergo drastic changes during ontogeny and therefore the test animals were adults of sizes suitable for the respective aquarium. In most fish species specimens were kept singly. Five sand gobies were kept in one of the 0.3 m aquaria. Four to five three-spined sticklebacks were kept in an identical tank. Most species were studied for periods of more than one year in order to check whether alternations in displayed activity patterns occurred at different light intensity levels in wintertime. Thus, some species were studied for periods of up to 5 years. During these studies test animals were replaced only when in bad condition or in cases of mortality. Where results have been obtained for the same time period for two or more years, the period with most observations, or, in cases with only a few observations per period, a weighted mean has been presented. In species where more than one type of activity pattern was observed for a certain time of year the differing patterns are discussed. Level of activity, expressed as number of passages across the photo-cell gate, was also recorded. For all fish species except the Baltic herring, pike, rudd, gunnel and black goby, daylength (average daily duration of light levels >5 lux per month) has been plotted against monthly mean of duration of activity (the time an animal is more active than the die1 mean level of activity, Figala & Mtiller 1972). Linear regression analysis has been

DAYLENGTH

IN H

Fig. 3. Explanatory diagram for Figures 5-8. The abscissa shows the daylength given as hours of light intensity >5 lux. The ordinate shows duration of activity given as hours of activity exceeding the 24 h mean. The points on the curve represent monthly means. Arabic numerals indicate months. Broken lines between points (not in this diagram) indicate months without data. The straight line fitted to the points shows best fit according to linear regression analysis.

used to fit a straight line to the observed points (Y = aX+b) and correlation coefficients are given. In cases with more than one annual activity pattern, diagrams are given for the different types. Figure 3 explains the daylength/duration of activity diagrams. The probable deviation of the regression lines from the 1: 1 relationship between daylength for diurnal species and night length for nocturnal species, respectively, and duration of activity, was tested by means of F-test analysis.

Results Activity patterns Invertebrates Praunus frexuosus O.F. Miiller. - A shoal of this species was studied for almost one month (26 days in October 1972). The activity pattern displayed two major peaks, the first at 0800 and the other at 1800h. Due to the limited study period no daylength/duration of activity diagram can be presented. Praunus flexuosus displayed a crepuscular pattern of activity during the limited study period. Zdothea baltica Pallas. - In this part of the Baltic

53

Fig. 4. The four major different types of activity patterns: A = Diurnal (perch), B = Nocturnal (flounder). C = Crepuscular (ruffe), and D = Inversion (fourhorn sculpin).

54

three species of Idothea occur. The biggest and most common is Idothea baltica. This species was examined for 20 months. At least 20 adult specimens were kept in the aquarium. The activity pattern was clearly nocturnal during the twenty months with the exception of May and June when the activity was well spread out during day and night. The slope of the regression line in the daylength/ duration of activity diagram (Fig. 5) was negative (-0.15) which points to a nocturnal mode of life. The correlation (r = -0.26) was not good due to the aberrant May and June values (upper right hand corner of the diagram). The slope of the regression line was significantly (at the 99% level)

different from the 1: 1 relationship between night length and duration of activity. Table 3 summarizes the results of the linear regressions on the daylight/ duration of activity relationships for species where linear regression analysis was applicable. Idothea baltica is therefore considered to be a predominantly nocturnal animal throughout the year. Mesidothea entoomon L. - Five to ten specimens were studied for 16 months. The activity pattern of Mesidothea entomon was nocturnal throughout the seasons. In the daylengttiduration of activity diagram (Fig. 5) the slope of the fitted line was negative (-0.42) which indicates a nocturnal activity

T~rble.3. Values of a, b and r (correlation coefficient) for each species when linear regression (y = ax + b) analysis was performed on the daylength (h)/duration of activity (h) relationships. The right hand column shows if the calculated regression line diverted from the 1: 1 relationship between the two parameters. Species

a

b

Mesidothea Gammarus Pdaemorl

haltica etItomot1 oceanicus adspersfts”

Baltic herring”’ Pike’ Rudd’ European eel yellow” European eel silver’” Three-spined stickleback (diur.) Three-spined stickleback (invers.) Nine-spined stickleback Fifteen-spincd stickleback* Perch Ruffe Gunnel” Eel-pout Black goby:” Sand goby (noct.) Sand goby (invers.) Father-lasher (noct.) Father-lasher (invers.) Fourhorn sculpin Sea-scorpion Lumpsucker” Turbot Flounder

Significance

-

Prafmtrs ,fle.rmws”’ Iiiothea

r

-0.15 - 0.42 - 0.37

13.25 15.27 14.36

-

-

0.30 - 0.30 - 0.27 0.25 0.20 0.11 - 0.52 0.46 -0.19 -0.10 - 0.20 - 0.04 - 0.52 - 0.19 0.12 0.25 0.30 - 0.24 - 0.23

” Denotes species where values were insufficient to cover a whole year. -Not possible to calculate values.

10.25 13.16 12.03 8.79 9.47 9.52 15.31 7.43 12.01 11.56 13.39 10.95 17.43 12.05 10.64 8.04 7.65 1I .70 13.46

- 0.26

xx

- 0.86 - 0.86

X

0.64 - 0.94 -0.50 0.55 0.42 0.36 - 0.95 0.88 - 0.58

xxx xx xx xxx xxx

- 0.30

xxx

- 0.47 -0.13 - 0.83 - 0.41 0.31 0.67 0.68 - 0.58 - 0.71

xxx xxx xx xx xxx xxx xxx xxx

55

fl 12 !

0 ‘*

.-

l -

,

1

Mesidothea

ldothea baltica

eel

i

6

I

I

I

I

Turbot

Flounder

5. DaylengthIduration

I

Fifteen-Jpinedsticklebac

I

I

Sand goby

I

Sammarus oceanus

1

ilver eel

Eel-pout

Fig.

7

entomon

I

1

ielI&

-0

I

I

1

Father-lasher

of activity diagrams for nocturnal species

56

I

1

k&pined

I

I

I

I

Sea-SCO$Otl

I

I

I

Black goby

stickleback Perch

Lumpsucker

Fig. 6. Daylengthiduration

of activity diagrams for diurnal species.

pattern. The correlation was very good (r = -0.86). The regression line differed significantly from the 1: 1 relationship (Table 3). This species was also clearly nocturnal. Gammarus oceanicus Segerstrble. - In this part of the Baltic proper at least five species of Gammarus are to be found. G. oceanicus is the most common, At least twenty specimens were kept in the smallest type of aquarium and studied for 14 months. A clear nocturnal pattern of activity was found. Figure 5 shows the clear negative slope of the fitted line (-0.37) and the correlation (r = -0.86) was very good. The regression line differed significantly from the 1: 1 relationship (Table 3). Thus, Gummurus oceanicus is also a nocturnal animal. Puluemon udspersza (Rathke). - Five adult specimens were studied in June 1971 for a period of only 18 days. As a result of the short period of study, no daylengthduration of activity diagram could be made. This species exhibited a clear nocturnal activity pattern during the observation period.

Fishes Baltic herring, Clupea harengus L. - Due to its sensitivity, it was unfortunately not possible to study the Baltic herring for more than one month, i.e. July. The results have already been presented (Aneer 1979). One single specimen and a group of four herring all of about 18 cm in length were studied. The limited study period indicated a bimodal diurnal activity pattern for the single specimen with the major part of the activity between noon and dusk. A minor peak in activity was found around 6 o’clock in the morning. The group of four showed a unimodal pattern with more pronounced activity between 0900 and 2200 h (Aneer 1979). No daylength/duration of activity diagram could be made from this limited study. Pike, Esox Iucius L. -Single pike were kept for ten months (excluding October and December). They were between 25 and 30 cm in length. Due to technical failures and a lack of cooperation from the

57 to the silver eel (Table 3), the F-test showed the regression time for yellow-eel to differ significantly from the 1: 1 relationship.

I

Ruffe Fig. 7. Daylength/duration species.

I

Pike of activity diagrams for crepuscular

pike, there are only a few days of observations in each of the summer months (June-August). Predominantly crepuscular activity patterns with a diurnal tendency were found (Fig. 7). Rudd, Scurdinius erythrophthalmus (L.) - One single and a group of five adult rudd, between 15 and 20cm in length, were studied for nine days altogether, the single specimen for two days at the end of May and the group for seven days at the beginning of July 1973. From this limited number of observations we conclude that rudd was diurnal in May and July. European eel, Anguilla anguillu (L.) - Two female yellow eel (0..7 m and 0.65 kg) and two female silver eel (migrating state) same dimensions were studied for five mont:hs (July-November) in outdoor tanks (two 6001 aquaria with a bottom size of 0.9m X 0.9 m). At the end of the study period the temperature was just above the freezing point and the recorded activity was negligible. The activity patterns of yellow and silver eel have already been published (Westin & Nyman 1979). The period of investigation was five months but the inactive period of eel is known to last from the end of October until the beginning of May (Tesch 1977). The study period is thus representative for the active period of eel. Daylengthlduration of activity diagrams for yellow and silver eel are shown in Figure 5. The negative slopes, together with the activity patterns thus show a nocturnal mode of activity. In contrast

Three-spined stickleback, Gasterosteus uculeutus L. - Single specimens and groups (10 specimens) of three-spined sticklebacks about 3.5 cm in length were kept for 41 months altogether. During the first two years single specimens were studied but at the end of the period a group was studied. The number of fish within the group underwent minor changes as new specimens were sometimes difficult to obtain. The species exhibited a predominantly diurnal pattern throughout the year but during one winter period a diverging nocturnal pattern was observed. The relationships between daylength and duration of activity are shown in Figures 6 and 8. The slope of the line for the diurnal pattern was 0.25 and the correlation coefficient was 0.55. For the shifting pattern the slope was 0.20 and the correlation coefficient was 0.42 (Fig. 8). The regression lines for both patterns differed significantly from the 1: 1 relationship (Table 3). The three-spined stickleback is therefore judged as displaying a predominantly diurnal pattern. Nine-spined stickleback, Pungitiuspungitiw (L.) Four to five specimens were kept for 18 months. In winter the nine-spined stickleback was diurnal but an inversion took place in May. An earlier attempt to invert the pattern was noted in March. In October the pattern changed and the winter pattern was diurnal once more. The daylength/duration of activity diagram is shown in Figure 8. The slope of the fitted line was positive, 0.11 and the correlation was not so good (r = 0.36). The regression line differed significantly from the 1: 1 relationship (Table 3). The activity pattern of the nine-spined stickleback is therefore judged to be of the inversion type. Fifteen-spined stickleback, Spinuchiu spinachia (L.). - Single specimens, about llcm in length, were studied for 13 months. The period July-September, however, could not be covered. The monthly patterns were clearly nocturnal during the period of investigation.

58

Three-spined

stickleback

Nine-spined

stickleback

Sand goby

7

Father-lasher Fig. 8.

Daylength/duration

Fourhorn

sculpin

of activity diagrams for species with phase-inversion.

The daylength/duration of activity diagram is shown in Figure 5. The linear regression analysis has been performed for the nine months covered. The slope of the fitted line was clearly negative (-0.52), indicating a nocturnal mode of activity. The correlation was very good (r = -0.95). Due to missing data for the summer-autumn period, no test of significance was carried out. The fifteenspined stickleback displayed a clear nocturnal mode of activity. Perch, Perca fluviatilk L. - This species was examined for 21 months. Specimens of about 20cm in length were used singly. The activity pattern was centered around noon in winter with a peak at 1000 h (Fig. 4). From February-March, a second peak was formed around 1600 h. This peak gradually widened and from April-May the major part of the activity was displayed in the afternoon. There was, however, still a minor peak in the morning close to the 5 lux level. In August the pattern changed again and gradually returned to the winter type found in October.

When the linear regression analysis was made on the daylength/duration of activity relationship (Fig. 6) the correlation was found to be very good (r = 0.88) and the slope of the fitted line was clearly positive (0.46). The regression line did not differ significantly from the 1: 1 relationship (Table 3). The perch is thus a clearly diurnal species. Ruffe, Gymnocephalus cernuus (L.) - This species was studied for 25 months. Specimens of about 17cm in length were used singly. The annual pattern of activity was markedly concentrated to the twilight hours (Fig. 4). In Figure 7 the daylength/duration of activity relationship is shown. The slope of the fitted line was negative (-0.19) but the correlation coefficient (r) was only -0.58. The regression line differed significantly from the 1: 1 relationship (Table 3). Ruffe is therefore a species with a crepuscular activity pattern. Gunnel,

Pholis gunnellus (L.) - A single adult

59 specimen was kept for only one week in August 1972. The nocturnal pattern of activity peaked during darkness at 2 a.m. No activity at all was recorded between 0800 and 1400 h. This short study indicates a nocturnal activity pattern at least in August. Eel-pout, Zoarces viviparus (L.) - Single specimens of about 20 cm length were used in the experiments. This species was studied for a total of 29 months. Activity was concentrated to nighttime with the peak centered slightly before midnight for eight of the twelve months of the year. In Figure 5 the daylength/duration of activity is shown. The distribution of points is markedly bent and the correlation of the fitted line was not good (r = -0.30). The slope of the line was negative (-0.10). The line differed significantly from the 1: 1 relationship (Table 2). The eel-pout is thus nocturnal throughout the year. Black goby, Gobius niger (L.) - One adult specimen, about 6cm in length, was kept for five months. The observed pattern was diurnal. An incomplete daylength/duration of activity diagram is shown in Figure 6. Since the study was limited to the dark winter months it would be meaningless to apply any linear regression to the observed points. The black goby thus displayed a diurnal activity pattern during the winter months. The possibility of an inverted activity pattern during summer cannot be exclud.ed. Sand goby, P,omatoschistus minutus Pallas. - Sand gobies, 3-4cm in length and five at a time, were kept for 21 months. The results can be interpreted in two ways. During the first year of study a night active (nocturnal) pattern was displayed with but one exception for November when a diurnal mode of activity was apparent. This divergent pattern in November forced us to prolong the investigation for a further winter period. By reducing the illumination the species was forced into a diurnal pattern during the winter with the exception of the December results. The dayle ngth/duration of activity diagrams (Fig. 5, 8) show somewhat different relationships.

The correlation coefficients were not good, -0.47 and -0.13 for the nocturnal and the phase-inverted patterns, respectively. The slopes of the fitted lines were both negative, -0.20 and -0.04 for the respective activity patterns. Both lines differed significantly from the 1: 1 relationship (Table 2). The sand goby thus displayed an inversion of the activity pattern, diurnal through the winter half and nocturnal during the summer half of the year. Father-lasher, Myoxocephalus scorpius (L.) Single females, about 25 cm in length, were used. They were studied for 43 months and during the spawning periods in 1972 and 1973 a male was introduced. The father-lasher was predominantly nocturnal but during October-December 1971 and December 1972-April 1973 an inverted diurnal pattern was found. No spawning took place during the presence of males but in 1973 both specimens were diurnal in contradiction to 1972. The daylength/duration of activity diagrams (Fig. 5,6) show that the slope of the fitted line for the nocturnal pattern (-0.52) was steeper than for the inverted pattern (-0.19) and the corresponding correlation coefficients were -0.83 and -0.41, respectively. Both lines differed significantly from the 1: 1 relationship (Table 3). Thus, the father-lasher also displayed an inversion of the activity pattern, being diurnal the winter half and nocturnal during the summer half of the year. Fourhorn sculpin, Myoxocephalus quadricornis (L.) - Single specimens, both maie and female about 18-20cm in length, were studied for 17 months. The resulting annual patterns (Fig. 4) have already been described by Westin (1971). In both sexes a clear inversion of the activity took place. The daylength/duration of activity diagram (Fig. 8) shows that the slope of the fitted line was slightly positive (0.31). A significant difference from the 1: 1 relationship was observed (Table 2). The fourhorn sculpin also displayed an inversion of activity.

60 Sea-scorpion, Tuurulus bubalis (Euphrasen) . Single specimens, 8-10cm in length, were studied for 31 months. The sea-scorpion displayed a diurnal pattern with some irregularities in summer. The day length/duration of activity diagram is shown in Figure 6. The slope of the fitted line was clearly positive, 0.25 and the correlation was good (r = 0.67). The regression line differed significantly from the 1: 1 relationship (Table 3). The sea-scorpion thus exhibited a day active pattern. Lumpsucker, Cyclopterus lumplls (L.) - Single specimens about 15cm in length were kept for 14 months altogether. No observations were obtained for the period August-September. The activity pattern was diurnal. The daylengthduration of activity diagram is shown in Figure 6. The slope of the fitted line was positive (0.30) and the correlation was good (r = 0.68). The regression line differed significantly from the 1: 1 relationship (Table 3). The lumpsucker exhibited a diurnal pattern with a slight tendency towards crepuscularity. Turbot, Scophthalmus maximus (L.) - Specimens about 20cm in length were studied singly for 15 months. The activity pattern displayed by the turbot was clearly nocturnal. The daylength/duration of activity diagram is shown in Figure 5. The slope of the fitted line was clearly negative (-0.24) and the correlation coefficient was -0.58. The F-test showed a significant difference from the 1: 1 relationship (Table 3). The turbot was a markedly nocturnal species. Flounder, Pfutichtys fle.sus (L.) - Single specimens between 20-25cm in length were kept for 27 months. This was a clearly nocturnal species (Fig. 4). The daylength/duration of activity diagram is shown in Figure 5. The fitted line showed a clear negative slope (-0.23) and the correlation was good (r = -0.71). The regression line differed significantly from the 1: 1 relationship (Table 3). Thus, the flounder is a markedly nocturnal species. Dependence of duratiori of activity upon daylength

An analysis of the diagrams in Figures 5-8 show

that the slope of the regression lines generally reflect the activity pattern of an investigated species. For diurnal species the slopes were positive (xdiurn= 0.31) and for nocturnal species, negative (R,,, = 0.30). For species displaying phase-inversions, the average slope was 0.04 although they varied between 0.20 and -0.19 (Table 3). The only crepuscular species with a sufficiently long study period, the ruffe, had a negative slope of -0.19. In most species with study periods of almost one full year or more, the slope of the regression line differed significantly from a 1: 1 relationship between duration of activity and daylength, alternatively, nightlength for diurnal, nocturnal, phase-inversion and crepuscular species respectively.

Discussion In organisms where no phenomenon such as tide has a marked influence on biological rhythms, different types of activity patterns have been described (e.g. Aschoff 1965, Miiller 1978a, b, Eriksson 1978): (1) A diurnal pattern where organisms respond mainly to die1 changes between light and darkness by increasing locomotor activity during daylight, (2) a nocturnal activity pattern where much activity occurs at night, and (3) a crepuscular pattern where organisms show increased locomotor activities at dawn and dusk. A fourth phenomenon is called phase-inversion (e.g. Eriksson 1975, 1978), which involves a nocturnal pattern during summer and a diurnal pattern in winter. The changeover takes place during a few weeks in autumn and spring respectively. Species with a nocturnal pattern

Crustaceans are often an important food source for a number of fish species. In consequence they run a great risk of being eaten if diurnal; nocturnal behaviour presumably increases chances of survival when under strong predation pressure. Four of the five crustacean species investigated in our study were nocturnal. The exception was Praunus flexuosus, a school-forming mysid which displayed a crepuscular pattern (see below).

61 Idothea

baltica has earlier been confirmed

as (1968) in field

nocturnal by Jansson & Kallander studies. The nocturnal pattern of Zdothea baltica, 1. cheiipes and 1. granulosa has also been shown by Horlyck (1973) who used the photocell method. The littoral G.ummarus oceanicus has earlier been reported to be nocturnal by Jansson & Kallander (1968). Hagerman CSZ Ostrup (1980) confirmed the night active pattern in the littoral brackish water shrimp Palaemon adspersus. Hagerman (1970) also showed that Crangon vzdgaris is nocturnal. Two other important benthic macroinvertebrates in the Baltic which contribute to fish food are Pontoporeia affink and P. femorata. Cederwall (1979) and Lindstrom & Lindstrom (1980) found these two species to be nocturnal, as is another important prey for fish, the polychaete Harmothoe sarsi (Sarvala 1971). A literature review gives few records of activity patterns for the fish species which we found to be nocturnal. Verheijen & de Groot (1967) and de Groot (1971) have presented results for flounder, turbot and other flatfish which are in accordance with our results. These authors discussed the obvious mismatch between food intake period and activity period. It is evident that the activity recorded in this type of study is not necessarily connected with a search for food. The fish species with a nocturnal activity pattern seem to have ,a rather low activity level during the winter period, according to our results, and they migrate seasonally to shallower areas, in this context in agreement with the day-active species. Diurnal patterns of activity

In the Baltic herring, a pelagic schooling fish, most activity occurred by day. Stickney (1972) found two peaks connected with sunrise and sunset in his study of juvenile herring. Our single specimen also displayed two peaks which occurred somewhat after sunrise and before sunset. The group of four herring was primarily active in the afternoon. A light increase in morning activity was found but it was not as pronounced as in single herring (Aneer 1979). Feeding takes place during daytime (De

Silva 1973) although activity and feeding patterns do not coincide fully. The rudd was studied for a limited period by Siegmund (1969) who found patterns similar to those found by us. Hesthagen (1976) studied black gobies in winter and found a diurnal pattern which became less clear when shelter was offered. It cannot be excluded that the black goby belongs to the phaseinversion species owing to the high activity level in winter. The three-spined stickleback showed two different patterns, one diurnal throughout the year and interestingly, an aberrant nocturnal one in winter. This ‘inverted’ phase-inversion is unknown for other species. A very limited study by Worgan & FitzGerald (1981) also points to a diurnal pattern. In perch, we found a diurnal pattern with a slight tendency towards crepuscularity, in agreement with studies by other authors (Hergenrader & Hasler 1966, Siegmund 1969, Lind 1974, Eriksson 1975, Alabaster & Stott 1978). In contrast to the other sculpins the sea-scorpion was diurnal. Summing up experience of diurnal species investigated in our area,.they are active during the summer half of the year and rather passive and rest in deeper areas during the winter period (Aneer et al. 1978). Ail school-forming species were day active. Crepuscular species

The mysid Praunus flexuosus has a different life pattern from the investigated nocturnal macrocrustaceans. It is a partly carnivorous, schoolforming animal with a fast, jerking, avoidance reaction. Although studied for a short period only, it displayed a pattern which corresponds to the crepuscular type at that time of year. Jansson & Kallander (1968) found a nocturnal pattern when working in the summer. Their findings do not contradict our results. At our latitudes the crepuscular pattern appears ‘nocturnal’ in summer. The ruffe, classed by Siegmund & Wolff (1972) as a night-active fish, was clearly crepuscular and in fact a species displaying a most typical crepuscular pattern.

62 Pike diverged from the diurnal species by showing a tendency towards crepuscularity and by a high level of activity in winter. Lind (1974) maintained that pike showed phase-inversion but his diagrams look more crepuscular than phase-inverting. Diana (1980) found a diurnal pattern while Lawler (1969), Malinin (1969, 1971), Poddubnyi et al. (1970) and Casselman (1978) all found crepuscular patterns. Generally, the crepuscular pattern is rather similar to the inversion type and species show a rather high year around activity level. In the spring and autumn periods the inversion species make a rapid shift between day and night activity. The activity of the crepuscular species is centered around twilight and this results in an illusory shift between day and night activity resembling the inversion patterns at our latitudes. Phase-inversion

species

The sand goby displayed both a nocturnal and an inversion pattern. Hesthagen (1976,198O) reported bimodal patterns for Gobius niger and Pomatoschistus pictus and suggested a crepuscular pattern at least in feeding. The time of year at which Hesthagen carried out his studies, November-April, does not exclude the possibility of inversion in these species. Rumohr (1979), using an automatic camera placed on the bottom in the Kiel area, found gobies to leave the bottom at sunset and to return to it at sunrise, i.e. a nocturnal pattern which does not contradict the possibility of a phaseinversion further north (see next paragraph). The fourhorn sculpin and father-lasher showed phaseinversions. It is clearly shown that even if the most important factor influencing the die1 activity patterns of fish is the alternation between light and darkness, the intensity of light in the case of phase-inversion is of utmost importance. Andreasson (1973) showed that in Cottus poecilopus and C. gobio, phase shifts in activity patterns were dependent on the light intensity and the species were nocturnal throughout the year in southern Sweden while shifting to a diurnal pattern in winter at the Arctic Circle. He also expected such a shift to take place in southern Sweden under

conditions of ice with snow cover, i.e. during periods of low light intensities not always encountered in southern parts of the country. Mtiller (1978b) went a step further when he claimed that nightactive fishes at higher northern latitudes (>60”N) shift their activity to daytime after the autumn equinox and invert from day to night in the early spring. This statement is based on a lot of experimental investigations where the critical maximum light intensity value is determined as around 2.5 lux during the winter period (Mtiller 1970, 1978a, 1978b). In our experiments no inversion at all occurred among the night-active species until winter light intensity was kept below 25 lux but unlike Mtiller (1978b) we did not succeed in forcing all night-active species to shift from night to day activity during the darker winter period. The inversion took place only in species with high year around activity, but not in species with low winter activity like e.g. eel, turbot and flounder. The fish species which display the inversion type of activity in our investigation have in common a high activity level all year round accompanied by a change in habitat in spring and autumn. Correlation ture in fish

between activity pattern and eye struc-

From examinations of eyes in coral reef fish, Munz & McFarland (1973) found that visual structures in the diurnal species differed from those of the nocturnal, as might have been predicted from their contrasting feeding circumstances. In their material there were only two piscivorous predators, but significantly, these constituted a third group which had eyes with visual structures intermediate to those in the diurnal and nocturnal groups and were assumed to be twilight feeders. Ahlbert (1969) found differences between three percid species, perch, ruff and pike-perch. From our investigation we know that the perch is diurnal, the ruffe crepuscular and, according to Kelso (1976) the American pike-perch (Stizostedion vitreum) is crepuscular to nocturnal. According to Ahlbert (1969) a comparison between the ecology of these species (Stizostedion lucioperca instead of S. vitreum) and their cone-mosaic structure in the retinae shows

63 that their different feeding behaviour is well reflected in the structure of their eyes. In another paper, she (A.hlbert 1975) also mentioned that the presence of a tapetum lucidum in pike-perch increases the light sensitivity, but the eye is not so well adapted for acute vision. The ruffe has a tapeturn lucidum in the two upper thirds of the retina in accordance with its habit of living and feeding on bottom organisms in a poor light environment. In the same paper she reported that the perch has a well developed visual acuity but lacks a tapetum lucidum.

In sculpins it has been shown (Engstrom 1963) that the diurnal sea-scorpion differs from the father-lasher and fourhorn sculpin which belong to the phase-inversion group in the cone types of the retina. It is peculia:r that the literature gives examples of diverging activity patterns for the same species. For example in flounder, in spite of the observed nocturnal patterns in our study, it is known that the main food intake takes place in the afternoon (Verheijen & de Groot 1967, de Groot 1971). But the eye structure of the flounder points to a feeding pattern at times other than at night which is directly supported by the above reports. Their studies of flatfish locomotor activity patterns also support our findings of nocturnal patterns for both flounder and turbot in aquaria and in nature. The importance of food search activity for the final expressions of a locomotor activity pattern varies significantly between species. For example, in a fast, hunting, pelagic species, food search necessarily occupies a substantial part of the active period. On the other hand, in a predator of the ambush type, such as turbot, food search is often an act of waiting for a suitable prey and the locomotor activity is probably expressed in a search for new hunting grounds and/or in migration. The activity patterns studied in our photo-cell gate-equipped aquaria present only a general picture of the overall locomotory patterns, in other words, the swimming activity. These activity patterns reflect a variety of behaviour such as food search, predator avoidance, migration, territoriality, reproduction and health status. The most important factor affecting activity pat-

terns is the light-dark regime and due to its great variation the organisms need to be flexible in their responses to its changes (Miller 1978a, 1978b). The most interesting part of the light intensity scale, and that which most drastically affects the activity patterns, seems to be below 100 lux where threshold effects determine what pattern will be found (Eriksson 1978). Our double patterns were caused by the different light intensities offered within this threshold region, e.g. the patterns observed in three-spined stickleback, sand goby, father-lasher and fourhorn sculpin. The existence of phase inversions in turbot (Lota lota) and the freshwater sculpins (Cottus poecilopus and C. gobio) are coupled to threshold light intensity values between 20-30 lux (Andreasson 1973, Muller 1978b). Eriksson (1978) showed that in experiments with a constant light/ dark intensity ratio at three different intensity levels, brown bullheads (Zctalurus punctatus) were diurnal at low intensities but remained noctural at high levels. Obviously it is not only the shifts between light and dark which affect the activity patterns but also the intensity of light. If the basic activity pattern were indeed a food search pattern, as reflected by the eye structure, then only small differences would be noted in results between studies. This is however not the case because food search constitutes only a part of the total swimming activity and the final activity pattern must be affected both by the life strategy of the organism and its environmental conditions.

References cited Ahlbert, L-B. 1969. The organization of the cone cells in the retinae of four teleosts with different feeding habits (Perca fluviatilis L., Lucioperca lucioperca L., Acerina cernua L. and Coregonw albula L.). Ark. Zool. Ser. 2,22: 445-481. Ahlbert, I.-B. 1975. Organization of the cone cells in the retinae of some teleosts in relation to their feeding habits. Ph.D. Thesis, Department of Zoology, University of Stockholm. 29 PP. Alabaster, J.D. & B. Stott. 1977. Swimming activity of perch, Perca fluviatilis L. Fish Biol. 12: 587-591. Andreasson, S. 1973. Seasonal changes in die1 activity of Cotrus poecilopus Heckel an C. gobio L. (Pisces) at the Arctic Circle. Oikos 24: 16-23. Andreasson, S. & K. Mtiller. 1969. Persistence of the die1 ac-

tivity phase shit? in Cottus poecilopus Heckel (Pisces) after transfer from 66” to 55” northern latitude. Oikos 22: 171-174. Aneer, G. 1979. On the ecology of the Baltic herring-studies on spawning area, larval stages, locomotory activity pattern, respiration, together with estimates of production and energy budgets. Ph.D. Thesis, Department of Zoology, University of Stockholm. 72 pp. Aneer, G., A. Lindquist & L. Westin. 1978. Winter concentrations of Baltic herring (Clupeu harengus var. membras L.). Contr. Asko Lab., Univ. Stockholm 21: 1-19. Aschoff, J. (ed). 1965. Circadian clocks. North-Holland Publ. Co., Amsterdam. 479 pp. Biinning, E. 1967. The physiological clock. Springer Verlag, Berlin. 167 pp. Casselman, J.M. 1978. Effects of environmental factors on growth, survival, activity and exploitation of northern pike. Amer. Fish. Sot. Spec. Publ. 11: 114-128. Cederwall, H. 1979. Diurnal oxygen consumption and activity of two Pontoporein (Amphipoda, Crustacea) species. pp. 3@+316. In: E. Naylor & R.G.Hartnoll (ed.) Cyclic Phenomena in Marine Plants and Animals. Pergamon Press, Oxford. Diana, J.S. 1980: Die1 activity pattern and swimming speeds of northern pike (Esox luck) in Lac Ste. Anne, Alberta. Can. J. Fish. Aquat. Sci. 37: 1454-1458. Engstrom, K. 1963. Cone types and cone arrangements in teleost retinae. Acta Zool. 44: 179-243. Etiksson, L.-O. 1975. Diel and annual activity rhythms and vertical distribution in the perch, Perca fluviutilis L., at the Arctic Circle. Ph.D. Thesis, Section Ecology, Zool., University of Umei, Umei. 14 pp. Eriksson, L.-O. 1978. Nocturnalism versus diurnalism-dualism within fish individuals. pp. 69-89. In: J.E. Thorpe (ed.) Rhythmic Activity of Fishes, Academic Press, London. Figala, J. & K. Mtiller. 1972. Jahresperiodik und Phasenlage eines tropischen Fisches (Barbus partipentazona) am Polarkreis. Aquilo Ser. Zool. 13: D-20. Groot de, S.J. 1971. On the interrelationships between morphology of the alimentary tract, food and feeding behaviour in flatfishes (Pisces: Pleuronectiformes). Neth. J. Sea Res. 5: 121-196. Hagerman, L. 1970. Locomotory activity patterns of Crangon vulgaris (Fabricius) (Crustacea, Natantia). Ophelia 8: 255 266. Hagerman, L. & J. Ostrup. 1980. Seasonal and diel activity variations in the shrimp Pulaemon &pews from a brackish, non-tidal area. Mar. Ecol. Prog. Ser. 2: 329-335. Harker, J.E. 1964. The physiology of diurnal rhythms. Cambridge University. Press, London. 114 pp. Helfman, G.S. 1978. Patterns of community structure in fishes: summary and overview. Env. Biol. Fish. 3: 129-148. Hergenrader, G.L. & A.D. Hasler. 1966. Diel activity and vertical distribution of yellow perch (Percufluvescens) under the ice. J. Fish. Res. Board Can. 23: 449-509. Hesthagen, I.H. 1976. Locomotory activity of the black goby, Gobius niger L. (Pisces, Gobiidae), under artificial light con-

ditions, including a false dawn and dusk. Sarsia 62: 9-10. Hesthagen, I.H. 1980. Locomotor activity in the painted goby, Pomatoschistus pictus (Malm) (Pisces), in relation to light intensity. Sarsi 65: 13-18. Horlyck, V. 1973. Seasonal and diel variation in the rhythmicity of Idothea baltica (Pallas) and Idothea granulosa Rathke. Ophelia 12: 117-127. Jansson, B.-O. & C. Klllander. 1968. On the diurnal activity of some littoral peracarid crustaceans in the Baltic Sea. J. exp. mar. BiolEcol. 2: 24-36. Kelso, J.R.M. 1976. Die1 movement of wallye Stizostedion vitreum vitreum in West Blue Lake, Manitoba, as determined by ultrasonic tracking. J. Fish. Board Can. 33: 2070-2072. Lawler, G.H. 1969. Activity periods of some fishes in Heming Lake, Canada. J. Fish. Res. Board Can. 26: 3266-3267. Lind, E.A. 1974. Seasonal variation in weight and activity of the freshwater fishes Percafluviatilis L. and Esox lucius L. Nordic Council Arct. Med. Res. Rep. 10: 65-66. Lindstrom, M. & A. Lindstrom. 1980. Swimming activity of Pontoporeiu uffirzis (Crustacea, Amphipoda) - Seasonal variations and usefulness for environmental studies. Ann. Zool. Fenn. 17: 213-220. Malinin, L.K. 1969. Home range and homing instinct of fish. Zool.Zh. 48: 381-391. (Fish. Res. Board Can. Transl. Ser. No. 2050). Malinin, L.K. 1971. Home range and actual paths of fish in the river pool of the Rybinsk reservoir. Biol. Vnutr. Vod. Inf. Byull. 22: 158-166. (Fish. Res. Board Can. Transl. Ser. No. 2282). Manteifel, B.P., 1.1. Girsa & D.S. Pavlov. 1978. On rhythms of fish behaviour. pp. 215-224. In: J.E. Thorpe (ed.) Rhythmic Activity of Fishes, Academic Press, London. Menaker, M. (ed.). 1971. Biochronometry. Nat. Acad. Sciences, Washington, D.C. 662 pp. Mills, J.N. (ed.). 1973. Biological aspects of circadian rhythms. Plenum Press, London. 319 pp. Munz, F.W. & W.N. McFarland. 1973. The significance of spectral position in.the rhodopsins of tropical marine fishes. Vis. Res. 13: 1829-1874. Mtiller, K. & K. Schreiber. 1967. Eine Methode zur Messung der lokomotorischen Aktivitat von Siisswasserfischen. Oikos 18: 135-136. Miiller, K. 1970. Die Tages- und Jahresperiodik der Buntflossenkoppe Cottus poecilopus Heckel am Polarkreis. Oikos Suppl. 13: 108-121. Miiller, K. 1978a. Locomotor activity of fish and environmental oscillations. pp. 1-19. In: J.E. Thorpe (ed.) Rhythmic Activity of Fishes, Academic Press, London. Mtiller, K. 1978b. The flexibility of the circadian system of fish at different latitudes. pp. 91-104. In: J.E. Thorpe (ed.) Rhythmic Activity of Fishes, Academic Press, London. Palmer, .D. 1976. An introduction to biological rhythms. Academic Press, New York. 375 pp. Poddubnyi, A.G., L.K. Malinin & V.V. Gaiduk. 1970. Experiment in telemetric observations under ice of the behaviour of wintering fish. Biol. Vnutr. Vod Inf. Byull. 6: 65-70. (Fish.

65 Res. Board Can. Transl. Ser. No. 1817). Rumohr, H. 197!). Automatic camera observations on common demersal fish in the Western Baltic. Ber. dt. Wiss. Komm. Meeresforsch. 27: 198-202. Sarvala, J. 1971. !Ecology of Harmothoesarsi (Malmgren) (Polychaeta, Polynoidae) in the northern Baltic area. Ann. Zool. Fenn. 8: 231-309. Schwassmann, H.O. 1980. Biological rhythms: their adaptive significance. pp. 613-630. In: M.A. Ali (ed.) Environmental Physiology of Fishes, Plenum Publ. Co., New York. Siegmund, R. 1969. Lokomotorische Aktivitat und Ruheverhalten bei einheimischen Shsswasserfischen (Pisces; Percidae, Cyprintdae). Biol. Zbl. 88: 295-312. Siegmund, R. &D. Wolff. 1972. Influence of acoustical stimulation on circadian rhythm of locomotor activity in two species of fish. Forma et Functio 5: 33-44. Silva, De S.S. 1.973. Food and feeding habits of the herring Clupea harengus and the sprat C. sprattus in inshore waters of the West Coast of Scotland. Mar. Biol. 20: 282-290. Stickney, A.P. 1972. The locomotor activity of juvenile herring

(Clupea harengus harengus L.) in response to changes in illumination. Ecology 53: 438-445. Tesch, F.W. 1977. The eel. Chapman & Hall, London. 434 pp. Verheijen, F.V. & S.J. de Groot. 1967. Diurnal activity pattern of plaice and flounder (Pleuronectidae) in aquaria. Netherl. J. Sea Res. 3: 383-390. Voipio, A. (ed.). 1981. The Baltic Sea. Elsevier Publ. Co., Amsterdam. 418 pp. Westin, L. 1971. Locomotory activity patterns of fourhorn sculpin, Myoxocephalus quadricornis (L.) (Pisces). Inst. Freshw. Res. Drottningholm 51: 184-196. Westin, L. & L. Nyman. 1979. Activity, orientation and migration of Baltic eel (Anguila anguilla L.). Rapp. P.-v. Reun. Cons. int. Explor. mer 174: 115-123. Worgan, J.P. & G.J. FitzGerald. 1981. Die1 activity and diet of three sympatric sticklebacks in tidal salt marsh pools. Can. J. Zool. 59: 2375-2379.

Received

10.2.1984

Accepted

12.9.1986