Circadian Rhythm of Locomotion and Its Temperature Dependence in Blattella Germanica

A number of investigators have reported on the diurnal rhythm of locomotion in cockroaches, which shows a very similar picture in the different species. It is well documented that the releasing factor is a change from light to darkness (Gunn, 1940; Harker, 1956; de Roberts, 1960). The rhythm persists under constant conditions, and this so-called free-running rhythm is remarkably temperature-independent, although part of the mechanism might be of a hormonal nature (Harker, 1964).

Based on several years experiments on stridulation in Orthoptera Ensifera (Nielsen & Dreisig, 1970), and later confirmed by experiments on cockroaches and on light-emission of glowworms (H. Dreisig, in preparation), the hypothesis was advanced that activity in nocturnal animals is released by a reduction in illumination if the animal is in a state of specific sensitization. This state was thought to be caused by an unknown agent, a sensitizer (perhaps a hormone) produced continuously, the rate of production being dependent on the temperature. During the daytime quiescent period the sensitization grows until a level is reached at which the activity starts, the threshold of release. This threshold is solely dependent on actual environmental factors. This paper is intended to elucidate the effect of light and temperature on the threshold of release.

The experiments with nocturnal insects have shown that the threshold is high in light and low in darkness and high at high temperatures and low at low temperatures. This means that there is a delay in the start of activity in constant light or if the temperature is raised just before the expected start, and that there is an advance in constant darkness or if the temperature is lowered just before the expected start of activity.

This hypothesis is helpful in explaining the various phenomena connected with a diurnal rhythmic behaviour, and the experiments reported below were made to review the. circadian rhythm in a common cockroach, Blattella germanica, in the light of the concept outlined above. Especially we wanted to test the validity of the assumption that the temperature-independence of the free-running rhythm in constant darkness could be explained by a temperature-dependent threshold of release.

The experiments were carried out at Zoologisk Laboratorium, Copenhagen, and Molslaboratoriet, Jutland. All experiments took place in a constant-temperature room. If temperature was not a necessary part of the procedure, it was kept constant at 22± 1°C, but otherwise it was possible to vary the temperature as required. The light was from a common 60 W light bulb 150 cm above the animal, and ‘daylight’ in the cage was thus about 10 lux. The cockroaches used in the experiments were obtained from the Government Pest Infestation Laboratory and only males were used. The animals were placed in a small transparent cage, 1 cm high, with wire-netting on top. The cage was divided into two parts connected by a narrow passageway. Above this a small lamp was placed with an infra-red filter in front of the lens, so that light only escaped through the filter. Below the cage and illuminated by the beam of infrared light a photocell and a relay were placed, the latter connected to a writing device recording on a revolving drum. When the animal passed through the passageway it interrupted the beam and this caused an increased resistance in the photocell, thus activating the relay. The activity was expressed as the number of passages per hour through the passageway.

Throughout the paper the abbreviation LD x:y signifies a condition of alternating x hours of light and y hours of darkness. DD means constant darkness and LL constant light. All the experiments reported below were repeated several times.

Fig. 1 shows the average activity with standard conditions LD 12:12, with the light on from 09 to 21 h and darkness from 21 to 09 h. Nearly all the activity is during the dark period. The activity may commence immediately after the light is switched off or after a period of latency (Nielsen & Dreisig, 1970) which may last up to 30 min, but is usually 15−20 min. The maximal value of activity is reached after a period of increment, most often in the second hour after the introduction of darkness (Fig. 2). Soon after the maximal value is attained, activity begins to diminish and comes to a halt before or at dawn (in most cases several hours before).

As is normal for nocturnal animals, the period of the free-running rhythm in Blattella germanica was found to be less than 24 h in constant darkness. It varied somewhat in different individuals (Table 1). As reference point the onset of activity was used, as being the one most easily defined. The average of the periods in Table 1 is 23 h 15 min.

Other observers have reported the activity to be delayed if the cockroach is kept in constant light. This could not be shown in B. germanica if the usual light was used in the light period. The animals remained inactive and there was no rhythmic activity. Only with a very faint illumination was there a delay in the recurrence of the activity. The amount of light was so small that it could not be measured. It was further discovered that a red filter instead of an infra-red one had the same effect as a weak light. The activity in continuous weak light was only kept up for a few days.

In a series of experiments the onset of darkness was advanced 4, 8 and 12 h (Fig. 3). It will be seen that the activity did not start immediately after the light was switched off as usual but was only slightly advanced in relation to the normal time of onset of darkness (21 h). At 4 h advance of darkness the activity was advanced about 1 h 10 min, at 8 h advance about 2 h 15 min. At 12 h advance of darkness the preceding period of darkness could of course not have been of 12 h duration as this would have meant continuous darkness. This period was therefore only 6 h, then followed by 6 h of light. Under these circumstances there was a little activity after the light was switched off, but as seen there is a definite period of activity which is only advanced 3−4 h.

If the onset of darkness was delayed 4, 8 and 12 h the activity started at once after the change, but the amount of activity was smaller the more the dark period was delayed (Fig. 4).

In Fig. 5 the light was kept on for 24 h on the first day. The light was switched off the following day at 09 h and the amount of activity was small. The new light-régime was kept on and it is seen that the activity only adapted after several days (only 2 days of the new cycle are shown).

As has been shown by previous workers and as seen from Table 1 the free-running rhythm is virtually independent of the temperature over the normal range. But as shown in Fig. 6 and 7 it is not independent of a change in temperature before the expected start of activity. Two procedures were followed, the first being to decrease the temperature, the second to increase the temperature.

In the first case cockroaches were kept at a normal change of light and darkness at a high temperature (30°C) for several cycles to bring the rhythm into a steady state. After this, the animals were given constant darkness, and next day, before the expected onset of activity, the temperature was lowered to 15°C, a procedure which took 2−3 h. In the example shown in Fig. 6 the activity was advanced, not by about an hour (as normally in continuous darkness) but by 9 h, i.e. as soon as the temperature had dropped to 15°C. The activity dropped during the following hours, probably because the temperature became too low for locomotion (11·5°C), but increased again when the temperature rose to 15°C. This experiment was repeated four times with different individuals with the same result each time (Table 2).

In the second case the animals were kept at a normal régime of light and darkness, but at a low temperature (15°C). As before, constant darkness was imposed but now the temperature was increased on the first day in darkness (Fig. 7). It is seen that the activity is not advanced as normally in darkness, but is delayed 2−4 h. The figure shows one of five experiments of this kind which were carried out (Table 3).

As outlined in the introduction the activity is thought to be caused by a temperature-dependent sensitization reaching a certain threshold of release, the level of which is dependent on the illumination and the temperature. What causes the sensitization is unknown but hormone production is a possible cause.

Fig. 8 is a schematic interpretation of this relationship between the sensitization and the light-dependent threshold of release when the temperature is kept constant. The heavy line indicates the sensitization and the rising slope of this indicates the rate of sensitization. The threshold of release in light (L) is higher than in darkness (D) and it is seen that as soon as the threshold in dark has been passed the activity is released by a change from light to darkness (Fig. 8 a). This only happens a few hours before the normal time of activity. If the onset of darkness is advanced more than a few hours as shown in Fig. 3, the activity will not be released by the change because the sensitization is too low; the onset of darkness is only slightly advanced in relation to the normal time of activity as if the animal had been in continuous darkness.

The rate of increase of sensitization is unknown, except for it being faster at higher temperatures than at low (shown by experiments on stridulation and light-emission in glowworms), but recent investigations suggest that in fireflies the rate of sensitization is low or zero after the termination of activity and that the rate increases during the daytime. It has been drawn as a straight line in Figs. 8 and 9 for the sake of simplicity.

During the course of the active period the sensitization is reduced to a low level by using-up of the sensitizer. Of this process, which determines the apical angle in Fig. 8 and the length of the active period, we know nothing. In the figure it has been drawn as a straight line for the sake of simplicity. After the termination of activity the cycle is repeated as the sensitization starts increasing again.

Under constant light-conditions the activity is released when the animal is fully sensitized, which happens about the normal time of onset of activity. The threshold of release is lower in darkness than in light (D, L in Fig. 8) and this explains the well-known fact of an advance in the recurrence of activity in constant darkness in nocturnal animals and a delay in constant light. In constant darkness the threshold of release is reached at an earlier hour than normal (24 − h, Fig. 8 c) and in constant light at a later time than normal (24+h, Fig. 8 b). In Blattella the threshold in constant darkness is reached on the average 45 min before the preceding period of activity. The threshold in constant light in most cases is too high to be reached, which means that the activity is absent in LL. It has been shown that if the light is faint the threshold might be reached and the endogenous rhythm becomes overt. This suggests that the change in threshold in light is gradual over a certain range of illumination.

If the threshold in light is not reached and the activity is absent, the sensitization does not continue to rise, but we must imagine a mechanism of overflow which again brings the sensitization to a low level. This is shown by the experiments in which the onset of darkness is delayed (Fig. 4). In this case the sensitization of the animal is gradually diminished after the normal time of start of activity.

The experiments in which the animals were kept in DD and the temperature was changed before the expected onset of activity (Figs. 6, 7) could be explained by assuming that the threshold of release is higher at a high temperature than at a low. This could explain the temperature-independence of the free-running rhythm as reported by several workers and as shown in Table 1 where the recurrence of activity under constant conditions is independent of the temperature. At a constant high temperature (Fig. 9 a) the increase in sensitization (FS) occurs more rapidly than at a constant low temperature (Fig. 9b, SS) but this is compensated by the higher threshold of release at high temperatures, so the activity starts at approximately the same time in DD (24 −h) at both high and low temperatures.

The experiments in Figs. 6 and 7 are shown schematically in Fig. 9c, d. In Fig. 9c the sensitization is fast at a high temperature rising towards the high threshold of release (T). On the second day in darkness the temperature is lowered sometime before the normal time of activity, thereby lowering the threshold of release and advancing the activity, not about 1 h as normally in DD, but as much as shown in Table 2.

In Fig. 9d the sensitization is slower at a low temperature rising towards the low threshold of release (t). On the second day in DD the temperature is raised thereby raising the threshold of release to (T). The activity is consequently not released at the expected time in DD (24 – h) but is delayed until the sensitization has reached the new threshold as shown in Table 3.

We are grateful to Dr Ellinor Bro Larsen for much help, and to Professor H. M. Thamdrup for placing the accommodations of the Mols laboratory at our disposal. The Government Pest Infestation Laboratory provided the animals, for which we are indebted.