Characteristics of the circadian activity rhythm in common marmosets (Callithrix j. jacchus)код для вставкиСкачать
American Journal of Primatology 17:271-286 (1989) Characteristics of the Circadian Activity Rhythm in Common Marmosets (Callithrix j . jacchus) HANS G. ERKERT Institut fur Biologie III, Universitat Tiibingen, Tiibingen, West Germany The circadian activity rhythm of the common marmoset, Callithrix j . jacchus was investigated by long-term recording of the locomotor activity of 15 individuals (5 males, 10 females) from 1.5 to 8 years old, both under constant illumination and under LD 12:12. The mean period of the spontaneous circadian rhythm was 23.2 & 0.3 h. Neither sex-specific differences nor a systematic influence of light intensity on the spontaneous period were observed, but the period was dependent on the duration of the trial and on the age of the individual. Due to the short spontaneous period, in LD 12:12 there was a distinct advance of the activity phase with respect to the light time and a masking of the true onset of activity by the inhibitory direct effect of low light intensity during the dark time. After an 8 h delay shift of the LD 12:12, re-entrainment of the circadian activity rhythm required an average of 6.8 0.7 days; the average re-entrainment time after an 8 h phase advance of the LD cycle was 8.6 +- 1.3 day. This directional effect is ascribed to characteristics of the phase-response curve. No ultradian components were observed, either in the LD-entrained or the free-running circadian activity rhythm. * Key words: activity, circadian rhythm, ultradian rhythm, activity pattern, re-entrainment, CaZZithrix INTRODUCTION Callithrix jacchus, the common marmoset, is a small neotropical monkey which is often thought to have a limited need for living space, and which reproduces at the high rate of 3-5 offspring per year. For these reasons it has been considered for years to be a particularly suitable laboratory primate for biomedical research [Deinhardt et al., 1967; Gengozian, 1969; Poole et al., 19781. As a result, a wealth of information about the biology of the common marmoset has accumulated. Data are available concerning its behavioral biology and ethology [Epple, 1970; Rothe, 1975; Stevenson & Poole, 1976; Box, 19771, its reproductive biology and endocrinology [Hearn, 1980,1983; Hearn et al., 1978; Dixson, 1983; Kendrick & Dixson, 19851, its developmental biology and teratology [Posvillo et al., 1972; Abbott, 1978; Siddall, 1978; Heger & Neubert, 1983; Moore et al., 19851, Received for publication August 10, 1988; revision accepted January 11, 1989. Address reprint requests to Hans G. Erkert, Institut fur Biologie 111, Universitat Tiibingen, Morgenstelle 28, D-7400 Tiibingen, West Germany. 0 1989 Alan R. Liss, Inc. 272 I Erkert and its physiology [Hawkey et al., 1982; McNees et al., 1982; McIntosh et al., 1985; Rothwell & Stock, 19851. By contrast, little is known about the chronobiology of Callithrix jacchus. Hetherington [19781 found a marked daily variation in rectal temperature in LD 12:12, with an average range from 35/36 to 39°C. Saito et al. [19831showed that in LD 14:lO marmosets have a daily drinking rhythm, with maximal fluid intake toward the end of the light phase. Kholkute  demonstrated a marked bimodal pattern of the plasma androgen level in Callithrix males, with higher values during the dark phase of an LD 14:lO. The first recordings of the locomotor activity of marmosets living under constant environmental conditions in our laboratory revealed a free-running circadian periodicity of less than 24 h. Furthermore, these recordings showed that acoustic social contact not only triggers additional activities masking the freerunning endogenous activity rhythm, but also produces variations of the circadian period length called relative coordination [Erkert et al., 19861. In the experiments described below additional characteristics of the circadian activity rhythm and of the endogenous circadian timing system by which it is controlled are presented. The variability of the spontaneous period is analyzed, and the degree to which the free-running circadian activity rhythm can be entrained by LD cycles is examined, together with the re-entrainment of this rhythm after 8 h phase shifts of the entraining signal. MATERIALS AND METHODS The activity recordings were carried out in a total of 15 common marmosets (5 males and 10 females) aged 1.5-8 years. All the animals were born in captivity. Some were placed at our disposal from the German Primate Center in Gottingen; others were obtained from Dr. H. Rothe in Gottingen, from Dr. Ch. Welker in Kassel, and the rest from our own colony. During the experiments the monkeys lived in wire mesh cages measuring 85 x 85 x 155 cm or 75 x 105 x 95 cm; 7 animals were kept singly, 4 as pairs, and 4 others both singly and as pairs (Table I). Data from animals kept as pairs were treated as a unit. The recording cages were equipped with a sleeping box and sitting board mounted on the rear as well as with several climbing sticks attached to the wire mesh of the two sides. In the first series of experiments the marmosets were isolated from one another visually, but not acoustically or olfactorily, in a 21 m2 room under constant conditions. Later, to improve the acoustic isolation, the cages of the single animals were placed within thick-walled (10 cm) wooden boxes measuring 130 x 130 x 160 cm, which greatly attenuated external sound. The top of each box was covered with a 30 cm-high light box comprising 7 fluorescent tubes (Osram 40 W/25) and two light-scattering matt plates. The ambient temperature in all experiments was 25°C 2 1"C, and the relative humidity was 60% -t- 5%. All the animals were fed ad libitum with a protein-rich rice-flake porridge and mixed fruit as described elsewhere [Erkert et al., 19861. Normally the porridge was provided shortly after the onset of activity of the marmosets, and the mixed-fruit meal 4-6 h later. During the constant-condition (LL) experiments the mush and fruit meals often coincided and were given at randomly varying phases of the circadian cycle with the exception of the last 2 h of rest time and activity time, during which the animals' circadian activity rhythm has been shown to be very susceptible to disturbances. The locomotor activity of the marmosets was recorded by an electroacoustic method, whereby the vibrations (sounds in solids) of the wire mesh of the cage generated by the various activities of the animals (e.g., moving around, intense Circadian Activity Rhythm in Marmosets I 273 TABLE I. Summary of the Conditions of the Constant-LightExperiments in This Paper (LLrLL8) Showing the Light Levels, Durations, Numbers of Animals (n), and the Individuals Used in Each Experiment* Illumination intensity (lux) Experiment ir * SD 88 % 11 64 ? 35 2.6 1.9 340 ? 160 0.16 lr 0.07 280 ? 80 280 ? 80 420 ? 120 350 % 80 * Minimum-maximum 78- 100 12- 100 0.4- 5.1 160- 480 0.08-0.23 160- 350 160- 350 240- 500 200- 410 Test animals Duration (days) nl/nz Individuals 20 20 35 65 20 51 170 150 210150 313 414 414 414 514 915 915 414 616 Fi;Mo;Li Fi;Mo;Li;To Fi;Mo;Li;To Fi;Mo;Li;To Fi;Mo;To;T/A;(Chi) AnlBe;DilCl;T/A;GllGr;Mo An/Be;Di/Cl;T/A;Gl/Gr;Mo T;A;Di;Cl G1;Gr;Em;Na;Mo;Ki *The intensity variation within a given experiment is due to differences in illuminance among the various recording cages. Under “Test Animals,” the first figure (nl) gives the total No. of individuals, and the second (n,) gives the No. of recording units. Where the two differ, some of the animals were kept as pairs. Under the “Individuals,” each of the individual test animals which joined in the respective trial is identified by the first (two) letter(s) of its name. Names divided by a diagonal stroke mark the individuals which were kept and recorded together. Further characterizations of the individual test animals are as follows: F i = ym; Mo= of; Li=of;To=om;Di=om; Cl=of; An=of; Be=om;T=ym;A=yf; Gl=yf; Gr=yf;Em=yEKi=yf; Na=yf, where ym = young male; yf = young female (53 years); om = old male; of = old female ( 2 4 years). scratching, scent marking) were picked up by a transducer-like special microphone (Merula TVJ-XL, Holthausen) that is highly sensitive t o sounds in solids and very insensitive to sounds propagated by air. Therefore, it was very rarely activated by the marmosets’ vocalizations. The electrical signals of the microphone were then amplified to trigger a series of constant-frequency square wave pulses that were added by electronic counters (Demel V224). The count was sampled at 1-min intervals by an Apple II+ PC, totaled every 5 min, with the sum then stored on disk; hard copy was produced by an Epson MX 100 matrix printer. In addition, the half-hour values were printed out every 24 h. For further analysis the data were transferred, via Kermit program, to an IBM-compatibleOlivetti PC M-24 with two floppy disks, a 10 megabyte hard disk, and attached plotter (HP 7475 A). The period length, T , of the free-running circadian activity rhythm was calculated by periodogram analyses [Dorrscheidt & Beck, 19751of the consecutive 30 min values (cf. Fig. 3). Data from four animals were also tested for the presence of an ultradian periodicity by applying both periodogram analysis and the maximum entropy spectrum analysis of Marple  t o the consecutive 5 min values. The latter, according to Leonard [19841, is a particularly suitable method for the analysis of biological time series. In order to both evaluate the recording method and ascertain the time course of several behavioral activities such as feeding and drinking, monitoring, locomotion, scent marking, play, and grooming, as a pilot study we also carried out certain video observations and tapings of single test animals lasting 12-14 h each. By using an infrared-sensitive low-light camera (GRUNDIG FAE 70) it was also possible to check the marmosets’ behavior before the beginning of light time of the light-dark cycle. To determine the spontaneous period in various constant light levels and for different durations of exposure to constant conditions, the marmosets were kept in constant light (LL)with illumination intensities between 0.1 and 500 lux, for times ranging from 20 t o 210 days. Table I summarizes the conditions in the successive constant-light trials such as LL light intensities, duration of experiment, number 274 I Erkert of test animals, and individuals used. The light intensity in each case was measured with a Tektronix J 16 digital photometer with a cosine-corrected photocell (56511). In all measurements the sensor was positioned in the middle of the recording cage, pointing vertically toward the illuminated matt plates of the light box. Between consecutive experiments the animals were kept for several months in LD 12:12 (102:10p1lux), to allow full entrainment and recuperation. In order to keep the influence of the after effects of entrainment [Pittendrigh & Daan, 1976; Aschoff, 19791 as nearly constant as possible in the evaluation of the influence of light intensity on the spontaneous period, only the data from the first 20 days of the LL conditions subsequent to an LD 12:12 were used. RESULTS Activity Pattern Common marmosets are strictly diurnal animals. In artificial LD cycles with >lo0 lux in the light time and light intensities in the dark time ranging from physiological darkness (<lop6 lux) t o full-moon brightness (0.3 lux), their daily activity was confined entirely to the light time. As determined by video observation, the monkeys awakened some time before the beginning of the light time and remained in their sleeping box monitoring the surroundings until the light had been turned on. Then they immediately began locomotor activity. Activity was usually terminated 1-2 h before the onset of darkness. As a result, in LD 12:12 the average duration of activity time was 11.1 2 0.7 h (n = 12 individuals; n’ = 20 days in each case), and there was a positive overall phase angle difference of the entrained circadian activity rhythm with respect to the center of the light time of the LD cycle. The time course of activity during the light phase varied from one animal to another. Often there was distinct intra-individual variation, although some individuals maintained an extremely constant activity pattern for a long time. Typical examples are shown in Figure 1. The individual patterns found by averaging the corresponding 30 min counts over a long time (>20 days) in LD 12:12 can be unimodal, bimodal, or trimodal. Accordingly, when several such individual patterns are averaged, the resulting “species actogram” can be purely unimodal or tend to have two or three peaks, depending on the composition and size of the sample population, and on the current andlor preceding environmental conditions. This complexity is demonstrated by the patterns in Figure 2, which were obtained from the data for several individuals in three different experiments with LD 12:12 (200-400:0.2 lux). The top diagram represents the mean individual patterns (each of which was an average over 20 days) of 5 old marmosets (2 males, 3 females; >5 yr) kept singly and isolated from one another visually but not acoustically. The middle diagram is based on 4 pairs of old animals ( 2 5 yr), and the one at the bottom is the average pattern of 5 females, 4 young (<3 yr) and one old (>4 yr), which 3 months before had completed a long-term free-running session under constant conditions (LL8) in complete isolation. Since that time they had been in full acoustical contact with occasional direct mutual visual contact. The standard deviations given by the vertical lines in Figure 2 differ greatly from one another. The largest values (i.e., the largest interindividual or betweenpairs differences in activity pattern) are found for the animals kept singly (top and bottom diagrams). Obviously such animals have less influence on the activity of others than do pairs of marmosets on other pairs. The light intensity during the dark time of an artificial LD cycle seems to have relatively little influence on either the activity pattern or the phase position of the Circadian Activity Rhythm in Marmosets I 275 ~~ 2 8 % 4- 0 I 8 K 4 - \ 0 I 8C “j4LWhL 0 I 3 I 15 3 2 8 I I 14 20 I h 2 I - 15 3 Original 1 5 h 3 .Duplicate I d“E R- 31 23.2 0 0 I 12 0 12 24 36 Days I 1 Of 0 Days I 9 I I 21 I I 9 I I I 21 I 9 h CET I 0 I l 180 I 1 I I I I 0 180 0 Degrees(360:‘f) Fig. 1. Intra-individual and interindividual variation of the activity pattern of Cullithrix j. jucchus. Mean actograms are shown for 4 young females ( 5 3 yr; G L K I ) and one old female ( 2 4 yr; NA) in 3 consecutive 20 day segments (left to right) of the entrained circadian activity rhythm in LD 12:12 (420: 0.2 lux). Fig. 2. Mean actograms of common marmosets kept in LD 1212 (200-500: 0.2 lux) under various social conditions (x & SD). Top: Old animals (24yr) kept singly with acoustic contact (n = 5; n’ = 20 days). Middle: Old animals (24yr) kept a s pairs in acoustic contact with other groups (n = 4 pairs; n‘ = 20 days). Bottom: Four young and 1 old animals (data from the 2nd column of Fig. 1) kept singly with acoustic social contact. Fig. 3. Free-running circadian activity rhythm of a female Callzthrix jacchus living under constant ambient conditions (LL 340 lux). Left: Double plot of the original data. Right: Evaluation mode, shown for the data on the left. Top: Results of the periodogram analysis to determine spontaneous period. The curve peak marks the most probable period contained in the time series, and the thin straight line delimits the region above which the probability of chance is P 5 0.01.Middle: Total amount of activity in consecutive circadian cycles. Bottom: Mean actogram of the free-running circadian activity rhythm, obtained by averaging over the calculated spontaneous period 7 = 23.2 h. 276 I Erkert 25 - 20 - 15 - 10 - 5 - 0 22.5 23.0 23.5 24.0 Z ( h ) Fig. 4. Frequency distribution of periods of the free-running circadian activity rhythm of 15 common marmosets for a total of 167 decades (LL 0.2-500 lux). entrained CAR with respect to the LD entraining signal. Of the eight individuals tested in this regard, none noticeably altered its activity rhythm when the dark illuminance was increased from physiological darkness to 0.1-0.5 lux (full-moon brightness). Circadian Period Under constant environmental conditions in the laboratory the marmosets generated a free-running circadian activity rhythm (Fig. 3) that remained stable for a long time (>200 days in tests so far). The period length of this endogenous rhythm was almost always distinctly shorter than 24 h. Among the 15 individuals studied to date, the spontaneous period in LL from 0.1 t o 480 lux varied from 22.7 to 24.0 h. The histogram in Figure 4 shows the distribution of the period lengths as calculated from 167 10 day segments (decades) of the free-running rhythm of these monkeys while living under constant conditions for several months. The overall average of the spontaneous period calculated from these data is q , L = 23.3 2 0.3 h. However, the true mean, i.e., the average calculated from the data of the decades following the 40th day under constant conditions, is slightly shorter (23.2 2 0.3 h; n = 90 decades of experiments LLB and LL,.,). This is due to the fact that the data of Figure 4 include all the decades of each experiment, and the period during the first 2-4 decades under constant conditions is usually longer than subsequent periods (see below and Fig. 6 ) . There were no differences attributable to the sex of the animals. The average spontaneous period of the ten females in the study was 23.2 0.3 h, identical with the 23.3 0.3 h found for the five males. Young marmosets had a somewhat shorter circadian period in LL than did older animals. On the basis of decade values (79 for the young and 69 for the old animals) the average T L L of individuals less than 3 years old was 23.2 L 0.3 h, and that of those over 4 years old was 23.4 2 0.3 h (Mann-Whitney U-test: u = -2.98; P <0.01). Despite a slight negative correlation between the period length and the logarithm of LL light intensity (Fig. 5 ) the circadian spontaneous period does not depend on light intensity. The Spearman rank-correlation analysis of the 7 values found for the illuminance range from 0.08 to 480 lux showed no significant correlation between light level and period length (r, = - 0.29; P>0.05). As with all the other diurnal primates studied so far as well as certain other diurnal * Circadian Activity Rhythm in Marmosets I 277 25 f 5 Callithrix jacchus -.. *. . .. . .. . *:- 23 4 22 I 2 I 1 I I 3 1 1 22.6 1 2 3 4 5 6 mammals, Callithrix jacchus seems not to obey Aschoff's circadian rule, which postulates a negative correlation of period length with illumination intensity for diurnal species in general [Aschoff, 19791. The circadian period gradually shortened for some time after the beginning of the LL regimen. Figure 6 shows the average spontaneous periods of 11marmosets in the first six consecutive 10 day segments of a several-month session under constant conditions (LL 200-400 lux) following a prolonged stable entrainment by LD 12:12. The downward trend at the left of the graph indicates that these aftereffects [Pittendrigh & Daan, 19761 of the preceding 24 h rhythm in Callithrix jacchus can persist an average of 20-40 days. This influence of duration of constant conditions on the period of the free-running circadian activity rhythm was confirmed statistically (Friedman test: P<O .01). Wilcoxon tests showed that the spontaneous periods of the first decade were significantly longer than those of decades 3-4 (P <0.01) and 6 (P <0.05), and comparison of the second decade period values with those of the fourth to sixth decades gave the same result (P<0.05). It follows that fairly constant period lengths, i.e., a steady state of the free-running circadian activity rhythm are not reached before the 30th-50th day of constant conditions. Shorter LL experiments can therefore lead to incorrect results. The average duration of the activity time of the free-running circadian rhythm, (YLL, was 11.3 ? 0.4 h, which did not differ from the value found under entrainment by LD 12:12 (Wilcoxon test: P>>0.05). Because the free-running period is always shorter than 24 h, the proportion of the circadian cycle occupied by a was somewhat larger under constant conditions than in LD 12:12 (48.5% 2 1.8% as compared with 46.5% k 4.1%). However, the difference was not statistically significant (Wilcoxon test: P >0.05). Entrainment and Re-Entrainment The main entraining signal, by which the endogenous rhythm of Callithrix jacchus (averaging 23.3 h) is synchronized with the external 24 h day, is the 278 / Erkert 0 10 20 30 40 50 60 70 DAYS I 9 I I 15 I I 21 I I 3 I ] 9 I I 15 I I 21 I I I 3 h[CET1 ] Fig. 7. Entrainment of the circadian activity rhythm of a n old female Callzthrix. Originally free-running in LL a t 280 lux, the circadian activity rhythm is first entrained by LD 12:12 (280:0.2 lux), then re-entrained after a 6 h delay shift of the LD signal, and finally returned to free-running in LL a t 280 lux. The shaded areas mark the dark phase of the LD 12:12. alternation between light and darkness-as it is for other animals. The entraining effect of the LD alternation on the free-running circadian activity rhythm is illustrated in Figure 7 by activity recordings in an old female. In this experiment, after changing the constant LL on day 21 to LD 12:12 (240:0.3 lux), it took only five lengthened transient cycles to bring the circadian rhythm into stable synchrony with the light phase of the LD. On the 40th day of the experiment the LD was phase-delayed 6 h by extending the duration of darkness. After this abrupt shift, the activity rhythm of the animal became re-entrained within 6 days. That this phenomenon is true entrainment in the sense of phase-setting, and not just light-induced masking, is demonstrated by the timing of the onset of activity when constant conditions were again imposed. On the first free-running day (LL 240 lux, beginning on day 62; T = 23.1 h) activity began at nearly the same time as it had Circadian Activity Rhythm in Marmosets I 279 CALLITHRIX JACCHUS [TtA) 0 10 20 30 40 50 60 70 Fig. 8. Re-entrainment of the circadian activity rhythm of a Callithrix pair after an 8 h phase delay (day 19) and advance shift (day 55) of the synchronizing LD 12:lZ (350:O.Zlux). under the preceding LD 12:12 conditions, and then gradually drifted further out of phase. On the other hand, the immediate 2.5 h phase advance of the activity on the first day of LL indicates that masking is also involved. Under the preceding LD conditions the activity time would have begun earlier, if it had not been inhibited by the low dark illuminance. Re-entrainment after a phase shift of the entraining signal was studied in a total of 14 marmosets, 12 kept as pairs and 2 singly. In these experiments the entraining LD 12:12 (280: 0.16 lux) was first phase delayed 8 h by a single lengthening of the L-phase, and 35 days later a corresponding phase advance (by an 8 h shortening of the D-phase) restored the original situation. Figure 8 shows the course of re-entrainment for a pair of marmosets. In this case, after the 8 h delay shift of the LD the animals’ circadian activity rhythm needed seven lengthened transient cycles before the onset and end of activity matched their 280 I Erkert I I I 2 0.5 I I I 5 1.0 r I I , , , , , , , , , P[hl I1 8 e 17 I i ti I 2 I I I I I 5 I I , 8 , , I1 , , , P[hl , , , 17 Fig. 9. Lack of ultradian components in the LD-entrained (above) and free-running (LLs; see Table I) circadian activity rhythm of a female Callithrixjacchus (n = 30 days in each case). Results of a periodogram analysis (left) and of a maximum entropy spectrum analysis (right). All the peaks in the periodogram-analysis plot which exceed the significance limit, P 5 0.01, are fractions of the circadian period and therefore do not indicate ultradian periodicity. original phase position to the entraining signal. After the advance shift, 6 (activity onset) and 9-11 (activity end) transient cycles were required. The average re-entrainment times of all the marmosets tested were 6.6 0.9 (activity onset) and 6.8 2 0.7 (activity end) days for the delay shift, and 6.4 2 0.7 (onset) and 8.6 1.3 (end of activity) days for the advance shift of the LD. If the end of activity is taken as the reference phase (because in these animals end of activity is less subjected to masking effects than the onset of activity), the marmosets are found to be re-entrained more slowly after an 8 h advance of the LD 12:12 than after an 8 h delay shift (Wilcoxon test: P >0.05). Whereas there was no difference between the times needed for re-entrainment of onset and end of activity following a delay shift, the advance shift produced a statistically significant difference in the rate of re-entrainment of these two phases of the circadian activity rhythm (Wilcoxon test; P <0.05). However, the activity records in Figure 8 indicate that this difference is at least partly an artifact of masking the onset of activity by an activity inhibiting effect of darkness. * * Lack of Ultradian Components Additional analyses were carried out to determine whether the circadian activity rhythm of Callithrix is modulated by a superimposed ultradian rhythm. In four adult females the activity data obtained from a 30-day free-running segment (LL 350 lux) and an entrained segment (LD 12:12;350: 0.16 lux) of the same length were analyzed by an expanded periodogram analysis and by maximum entropy analysis. In no case was there evidence of an ultradian component (Fig. 9). Similarly, spectrum analysis of all data from a 210 day free-running experiment revealed no periodicity other than that of the circadian activity rhythm in any of the animals. The secondary peaks in the periodogram-analysis plots coincide exclusively with integral fractions of the calculated circadian period. Hence they represent computational artifacts of the circadian rhythmicity. The circadian activity rhythm of Callithrix exhibits no ultradian components. Circadian Activity Rhythm in Marmosets i 281 But since only the gross locomotor activity was monitored, rather than particular forms of behavior, the results do not rule out the possibility that certain elements or sequences of behavior may be repeated with an ultradian periodicity within the circadian cycle. DISCUSSION The currently available data from chronobiological studies on four prosimian and seven simian species are summarized in Table 11. Nonhuman primates constitute a relatively homogeneous group with respect to the period length of the free-running circadian rhythm under constant environmental conditions. Neither the mean nor the range of variation of the spontaneous period differ characteristically between prosimians and simians, or between the diurnal and the nocturnal species. With a period spectrum ranging from 22.3 to 26.3 h over all the species, the intraspecific range of variation of period lies between 0.9 h in Macaca irus [Hawking & Lobban, 19701 and 2.7 h in Galago senegalensis [Erkert et al., 19841. The range of spontaneous periods found here for Callithrix, 1.3 h, is close to the overall average of 1.6 i. 0.5 h for all the primate species studied thus far. Furthermore, Callithrix is not unique in having the period consistently shorter than 24 h, as the same appears to be true for the nocturnal Galago garnettii [Erkert et al., 19841 and the diurnal Macaca nemestrina [Tokura & Aschoff, 19831. On the basis of the relation between the spontaneous circadian period and the phase position of the entrained rhythm with respect to the LD signal found for other vertebrate species [Aschoff & Pohl, 19781, the positive overall phase-angle difference of the entrained marmoset rhythm in LD 12:12 can be ascribed to its relatively short period (7 << 24 h) combined with the characteristics of the phase response curve. As a result of the short endogenous rhythm the marmosets ordinarily awake before the end of the dark phase of the LD 12:12. This true onset of activity, i.e., the onset preprogrammed by the circadian timing system, then is evidently masked by a direct inhibitory effect of the low D illuminance (0.1 lux) on the motor system. Such masking effects by the direct action of light [Aschoff et al., 1982; Erkert & Grober, 19861 are particularly clear during re-entrainment of the circadian activity rhythm following phase shifts of the LD entraining signal. Whenever part of the daily activity coincides with the dark phase of the delayed or advanced LD 12:12, the level of activity is drastically reduced as long as the darkness continues (Figs. 7, 8). Accordingly, the 2-2Y2 h advance of the activity onset observed immediately after either a sudden advance of the light phase of an LD or a sudden transition from LD 12:12 to LL can be explained, at least in part, by the abrupt cessation of the activity-inhibiting action of the low D illuminance (Fig. 7). The spontaneous period produced by the circadian timing system in Callithrix seems neither to be systematically influenced by the light intensity, nor to depend on the sex of the animal, though an effect of age was observed. The finding that young marmosets have a somewhat shorter free-running period than the older animals is inconsistent with the results of Pittendrigh and Daan [19741, who demonstrated a shortening of T with increasing age in three rodent species. On the other hand, Haussler  found that in the neotropical bat M. molossus, in which spontaneous periods shorter than 24 h predominate, individuals over 2 years old had larger 7 values than did the young animals. It might be that these contradictory findings in different mammalian species are due in part to aftereffects of prior entrainment in one case or another, and they may indicate an age-dependence of aftereffects rather than a genuine age-dependence of the circadian period. Thus the present results demonstrate that generalizations in *di = diurnal; no = nocturnal; loc. di di di Macaca nemestrina Macaca irus Pan troglodytes = gross locomotor 4 4 3 2 1 = 1-85 0150 SD = - T ,,, (h) 22.7-24.0 24.5-26.2 24.3-26.3 23.3-25.8 23.8-24.4 23.8-24.9 22.3-23.8 23.7-24.6 23.7-25.1 24.2-26.2 23.5-25.2 22.8-24.6 22.8-25.5 22.4-23.7 7,in T Standard lux). Yellin & Hauty 119713 Martinez [1972l Tokura & Aschoff [1978, 19831 Hawking & Lobban [19701 Farrer & Ternes  Present paper Aschoff & Tokura [ 19861 Sulzman e t al.  Thiemann-Jager [19861 Erkert et al.  Erkert e t al.  Erkert e t al. [19841 Erkert e t al. [19841 Authors 2 care temperature; pD = physiological darkness ( 5 23.32 0.3 25.22 0.4 25.02 0.5 24.92 0.6 24.0k 0.3 24.1-r- 0.3 23.12 0.4 24.2 25.12 0.9 24.52 0.6 23.72 0.5 23.8k 0.6 23.1-r- 0.5 T 2 Period length, feeding; temp. 270 0/0.4-1300 0.003-100 0.1-400 0.1-400 1-600 0.1-360 0.1-240 pD-0.1 pD-0.1 pD-0.1 Range illumination intensityhx activity; feed. loc. loc. feed. temp loc. loc. loc. loc. feed. 14 8 di di di loc. 6 no loc. loc. loc. loc. 3 5 6 5 n no/di no no no diho Macaca mulatta Prosimiae Lemur fulvus Microcebus murinus Galago senegalensis Galago garnettii Simiae Aotus lemurinus griseimembra Callithrix jacchus Saimiri sciureus Parameter TABLE 11. Summary of the Spontaneous Periods Found Thus Far in Nonhuman Primates (Means T Deviation and Overall Range of Variation) * Circadian Activity Rhythm in Marmosets I 283 chronobiology should be based on extensive comparative studies on many species of diverse systematic groups. Unexpectedly, in view of the short spontaneous period and the marked positive phase angle difference between the entrained circadian period and the LD cycle, the circadian activity rhythm of Callithrix becomes re-entrained more slowly after an 8 h advance shift of the LD 12:12 than after an 8 h delay shift (Fig. 8). It appears, therefore, that this “directional effect” [Erkert & Grober, 1986; “asymmetry effect” according to Aschoff, 19781 in re-entrainment after shifting of the entraining signal is unrelated-or at least has no direct causal relationshipto the period length. The alternative possibility suggested by Aschoff  and also considered likely by Schanz and Erkert [19871, is that the magnitude of the directional effect may be determined primarily by characteristics of the phase response of the circadian timing system to changes in light intensity. The variability of the average actograms obtained under identical physical environmental conditions in the laboratory makes it difficult to decide whether the basic activity pattern of Callithrix jacchus is primarily unimodal, bimodal, or multimodal. Since no ultradian periodicity was found, however, a regular multimodal actogram can be excluded. The data in Fig. 2 suggest that in the normal case, i.e., in animals with complete social contact, bimodal activity patterns predominate. This assumption is supported by the findings of Saito et al. 119831 that feeding and drinking activity in Callithrix jacchus do have a bimodal time course as well. Common marmosets are relatively small mammals. Therefore their body temperature rhythm should be expected to run in parallel to locomotor activity. The more unimodal temperature pattern described by Hetherington  seems to be inconsistent with this deduction. However, since Hetherington measured rectal temperature while handling the marmosets, the basic circadian pattern could have been masked by handling-induced rises in core temperature. Actual comparable data can be obtained only by telemetric measurements in undisturbed animals. The proportions of time spent in different activities and the diurnal patterns of these behaviors are generally accepted to be important in primate behavior and ecology [Harrison, 19851. But in spite of this, field studies in which such data were established in a systematic manner and over an adequate time span are still lacking. From field observations lasting only a few days, of a group of Callithrix jacchus inhabiting a small woodland a t a university campus in Brazil, Maier et al.  reported the activity pattern to vary considerably from day to day. By radiotagging single members of two groups of common marmosets inhabiting the same forest area, Hubrecht [19851proved as well that the pattern of activity levels differed markedly between groups. These observations made in the natural habitat coincide quite well with the variability of the activity pattern found in the present study under controlled laboratory conditions. Maier et al.  also reported that the animals started locomotor activity about 15-35 min after sunrise and returned to their sleeping site about 1 h before sunset. From these data, a positive overall phase-angle difference of the marmoset’s circadian activity rhythm to the entraining natural light dark cycle can be deduced. This is consistent with the results obtained here in an artificial light-dark cycle and can therefore also be traced back to the short circadian period underlying the activity rhythm in Callithrix jacchus. Similar times of onset and end of activity (30-60 min after sunrise and 1-2 h before sunset, respectively) have also been reported for a wild group of the tassel-ear marmoset, Callithrix humeralifer, inhabiting an evergreen dryland forest in the State of Mato Gross0 in Brazil [Rylands, 19861. Proceeding from this observation 284 I Erkert one should expect the activity rhythm of this marmoset species to be based on an endogenous circadian rhythm with a period length of less than 24 h as well. CONCLUSIONS 1. The activity patterns of common marmosets living in artificial light-dark cycles (LD 12:12) reveal marked intra-individual and interindividual variation. In animals with full social contact, bimodal patterns predominate. 2. The daily-periodic activity rhythm of the marmosets is controlled by an endogenous timing system which generates an average circadian spontaneous period of 23.2 5 0.3 h. The period length of the free-running circadian activity rhythm does not seem to be influenced systematically by the illumination intensity or by the sex of the animal, but it does vary slightly with age, and is dependent on the duration of trial. Due to such aftereffects a steady state of the spontaneously free-running rhythm is not reached before at least 30-40 days spent under constant conditions. 3. The light-dark cycle acts as a potent entraining signal which synchronizes the marmoset’s endogenous rhythm with the environmental 24-h periodicity. The positive overall phase-angle difference of the entrained activity rhythm with respect to the LD cycle, as observed under artificial and under natural lighting conditions, can be attributed to the very short circadian spontaneous period. 4. After sudden 8 h delay shifts of an entraining LD 12:12 the circadian activity rhythm of Callithrix jacchus resynchronizes more quickly than after 8-h advance shifts. This directional effect in resynchronization is ascribed t o characteristics of the phase response of the circadian system t o changing light intensities. 5. Neither the free-running nor the entrained circadian activity rhythm of Callithrix jacchus contains ultradian components. ACKNOWLEDGMENTS I am very grateful to Prof. Dr. H.-J. Kuhn and Dr. W. Kaumanns, of the German Primate Center in Gottingen, as well as to Prof. Dr. C. Vogel and Dr. H. Rothe, Gottingen, and to Dr. C. Welker, Kassel, for providing me with some of the marmosets used here. I thank Mrs. Birgit Brodbeck-Vater for carefully tending the animals, and for much-appreciated technical assistance in carrying out and evaluating the experiments. Many thanks are also due to Ms. Ursula Schardt and Ms. Ingrid Wilhelm, who made available to me certain data from their experiments under constant conditions. This work was supported by the Deutsche Forschungsgemeinschaft (Er 59/10). REFERENCES Abbott, D.H. Physical, hormonal and behavioral aspects of sexual development in the marmoset monkey Cullithrix jacchus. JOURNAL OF REPRODUCTION AND FERTILITY 53:155-166, 1978. Aschoff, J. Problems of re-entrainment of circadian rhythms: Asymmetry effect, dissociation and partition. Pp. 185-195 in ENVIRONMENTAL ENDOCRINOLOGY. J. Assenmacher, D.S. Farner, eds. Berlin, Springer Verlag, 1978. Aschoff, J. Circadian rhythms: Influences of internal and external factors on the period measured in constant conditions. ZEITSCHRIFT FUR TIERPSYCHOLOGIE 49:225-249, 1979. Aschoff, J.;Pohl, H. Phase relations between a circadian rhythm and its Zeitgeber within the range of entrainment. NATURWISSENSCHAFTEN 65:SO-84,1978. Aschoff, J.; Daan, S.; Honma, K.-I. Zeitgeber, entrainment, and masking: Some unsettled questions. Pp. 13-24 in VERTEBRATE CIRCADIAN SYSTEMS. STRUCTURE AND PHYSIOLOGY. J. Aschoff, S. Daan, G.A. Groos, eds. Berlin, Springer Verlag, 1982. Aschoff, J., Tokura, H. Circadian activity Circadian Activity Rhythm in Marmosets I 285 rhythms in squirrel monkeys: Entrainment by temperature cycles. JOURNAL OF BIOLOGICAL RHYTHMS 1:91-99, 1986. Box, H.O. Social interactions in family groups of captive marmosets (Callithrix jacchus). Pp. 239-249 in THE BIOLOGY AND CONSERVATION OF THE CALLITHRICHIDAE. D.G. Kleiman, ed. Washington, DC, Smithsonian Institution Press, 1977. Deinhardt, J.B.; Devine, J.; Passovoy, M.; Pohlman, R.; Deinhardt, F. Marmosets as laboratory animals I. Care of marmosets in the laboratory, pathology and outline of statistical evaluation of data. LABORATORY ANIMAL CARE 17:ll-29,1967. Dixson, A.F. The hormonal control of sexual behaviour in primates. Pp. 131-219 in OXFORD REVIEWS OF REPRODUCTIVE BIOLOGY, Vol. 5 . C.A. Finn, ed. Oxford, Oxford University Press, 1983. Dorrscheidt, G.J.; Beck, L. Advanced methods for evaluating characteristic parameters (7,a,0 ) of circadian rhythms. JOURNAL OF MATHEMATICAL BIOLOGY 2: 107-121,1975. Epple, G. Maintenance, breeding, and development of callithrichidae in captivity. FOLIA PRIMATOLOGICA 1256-76, 1970. Erkert, H.G.; Grober, J. Direct modulation of activity and body temperature of owl monkeys (Aotus lemurinus griseimembra) by low light intensities. FOLIA PRIMATOLOGICA 47:171-188,1986. Erkert, H.G.; Nagel, B.; Schanz, F.; Thiemann-Jager, A. Vergleichende Untersuchungen zur Chronobiologie nichtmenschlicher Primaten. VERHANDLUNGEN DER DEUTSCHEN ZOOLOGISCHEN GESELLSCHAFT 77:217,1984. Erkert, H.G.; Nagel, B.; Stephani, I. Light and social effects on the free-running circadian activity rhythm in common marmosets (Callithrix jacchus; Primates): Social masking, pseudo-splitting, and relative coordination. BEHAVIORAL ECOLOGY AND SOCIOBIOLOGY 18:443-452,1986. Farrer, D.N.; Ternes, J.W. Illumination intensity and behavioral circadian rhythms. Pp. 1-7 in CIRCADIAN RHYTHMS IN NONHUMAN PRIMATES. Bibl. Primat., No. 9.F.H. Rohles, ed. Basel, Karger, 1969. Gengozian, N. Marmosets: Their potential in experimental medicine. ANNALS OF THE NEW YORK ACADEMY OF SCIENCES 162~336-362,1969. Harrison, J.S. Time budget of the green monkey, Cercopithecus sabaeus: Some optimal strategies. INTERNATIONAL JOURNAL OF PRIMATOLOGY 6(4):351-376,1985. Haussler, U. Zeitliche Organisation von Ak- tivitat und Verhalten der Bulldoggfledermaus Molossus Molossus Pallas 1766 (Chiroptera, Molossidae). Dissertation, Universitat Tiibingen, 1987. Hawkey, C.M.; Hart, M.G.; Jones, D.M. Clinical Hematology of the common marmoset CaZlithrix jacchus. AMERICAN JOURNAL OF PRIMATOLOGY 3:179199,1982. Hawking, F.; Lobban, M.C. Circadian rhythms in macaca monkeys (physical activity, temperature, urine and microfilarial levels). JOURNAL OF INTERDISCIPLINARY CYCLE RESEARCH 1:267290,1970. Hearn, J.P. Endocrinology and timing of implantation in the marmoset monkey, Callithrix jacchus. PROGRESS IN REPRODUCTIVE BIOLOGY 7:262-269, 1980. Hearn, J.P. The common marmoset (Callithrix jacchus). Pp. 181-216 in REPRODUCTION IN NEW WORLD PRIMATES. J.P. Hearn, ed. Lancaster, MTP Press, 1983. Hearn, J.P.; Abbott, D.H.; Chambers, P.L.; Hodges, J.K.; Lunn, S.F. Use of the common marmoset, Callithrix jacchus, in reproductive research. PRIMATES IN MEDICINE 10:40-49,1978. Heger, H.W.; Neubert, D. Timing of ovulation and implantation in the common marmoset, Callithrix jacchus, by monitoring of estrogens and 6P-hydroxypregnanolone in urine. ARCHIVES OF TOXICOLOGY 54: 41-52, 1983. Hetherington, C.M. Circadian oscillations of body temperature in the marmoset, Callithrh jacchus, LABORATORY ANIMALS 12:107,108,1978. Hubrecht, R.C. Home-range size and use and territorial behavior in the common marmoset, Callithrix jacchus jacchus at the Tapacura Field Station, Recife, Brasil. INTERNATIONAL JOURNAL OF PRIMATOLOGY 6531-548, 1985. Kendrick, K.M.; Dixson, A.F. Luteinizing hormone releasing hormone enhances proceptivity in a primate. NEUROENDOCRINOLOGY 41:449-453,1985. Kholkute, S.D. Plasma progesterone levels throughout the ovarian cycle of the common marmoset (Callithrix jacchus). PRIMATES 25123-126, 1984. Leonard, J.L. A top-down approach to the analysis of behavioral organization. JOURNAL OF THEORETICAL BIOLOGY 107: 457-470, 1984. Maier, W.; Alonso, C.; Langguth, A. Field observations on Callithrix jacchus jacchus L. ZEITSCHRIFT FUR SAUGETIERKUNDE 47:334-346,1982. 286 / Erkert Marple, L. A new autoregressive spectrum analysis algorithm. IEEE TRANSACTIONS ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING 28:441-454,1980. Martinez, J.L. Jr. Effects of selected illumination levels on circadian periodicity in the rhesus monkey (Macaca mulatta). JOURNAL OF INTERDISCIPLINARY CYCLE RESEARCH 3:47-59, 1972. McIntosh, G.H.; Lawson, C.A.; Rodgers, S.E.; Lloyd, J.V. Hematological characteristics of the common marmoset (Callithrix jacchus jacchus). RESEARCH IN VETERINARY SCIENCE 38:109-114, 1985. McNees, D.W.; Ponzio, B.J.; Lewis, R.W. ; Stein, F.J.; Sis, F.R.; Levy, B.M. Hematology of common marmosets (Callithrixjacchus). PRIMATES 23:145-150, 1982. Moore, H.D.M.; Gems, S.; Hearn, J.P. Early implantation stages in the marmoset monke; (Callithrix- jacchus). AMERICAN JOURNAL OF ANATOMY 172:265-278. 1985. Pittendrigh, C.S.; Daan, S. Circadian oscillations in rodents: A systematic increase of their frequency with age. SCIENCE 186: 548-550,1974. Pittendrigh, C.S; Daan, S. A functional analysis of circadian pacemakers in nocturnal rodents. I. The stability and lability of spontaneous frequency. JOURNAL OF COMPARATIVE PHYSIOLOGY 106:223252, 1976. Poole, T.B.; Stevenson, M.F.; Sutcliffe, A.G. Callithrix jacchus jacchus as a subject for behavioural projects in biomedical research. PRIMATES IN MEDICINE 10:232238,1978. Posvillo, D.E.; Hamilton, W.J.; Sopher, D. The marmoset as a n animal model for teratological research. NATURE 239:460462,1972. Rothe, H. Some aspects of sexuality and reproduction in groups of captive marmosets (Callithrix jacchus). ZEITSCHRIFT FUR TIERPSYCHOLOGIE 37:255-273, 1975. Rothwell, N.J.; Stock, M.J. Thermogenic capacity and brown adipose tissue activity in the common marmoset. COMPARATIVE BIOCHEMISTRY AND PHYSIOLOGY. A: COMPARATIVE PHYSIOLOGY 81:683686,1985. Rylands, A.B. Ranging behaviour and habitat preference of a wild marmoset group, Callithrix humeralifer (Callithrichidae, Primates). JOURNAL OF ZOOLOGY (A) 210:489-514,1986. Saito, T.R.; Katsuyama, M.; Murakoshi, H.; Koide, M.; Yamaguchi, T.; Takahashi, K.W. Patterns of drinking behavior in the common marmoset, Callithrix jacchus jacchus. ZOOLOGICAL MAGAZINE 92:216218,1983. Schanz, F.; Erkert, H.G. Resynchronisationsverhalten der circadianen Aktivitatsperiodik von Galagos (Galago senegalensis, Galago crassj-caudatus garnettii). ZEITSCHRIFT FUR SAUGETIERKUNDE 52:218-226,1987. Siddall, R.A. The use of marmosets (Callithrix jacchus) in teratological and toxicological research. PRIMATES IN MEDICINE 10:215-224, 1978. Stevenson, M.F.; Poole, T.B. An ethogram of the common marmoset (Callithrix jacchus jacchus): General behavioural repertoire. ANIMAL BEHAVIOUR 24:428-451, 1976. Sulzman, F.M.; Fuller, C.A.; Moore-Ede, M.C. Tonic effects of light on the circadian system of the squirrel monkey. JOURNAL OF COMPARATIVE PHYSIOLOGY 129: 43-50, 1979. Thiemann-Jager, A. Charakteristika der Circadianen Aktivitatsperiodik von Nachtaffen (Aotus triuirgatus, Humboldt 1811). Dissertation, Universitat Tiibingen, 1986. Tokura, H.; Aschoff, J . Circadian activity rhythms of the pig-tailed macaque, Macaca nemestrina, under constant illumination. PFLUGERS ARCHIV 376:241-243,1978. Tokura, H.; Aschoff, J. Effects of temperature on the circadian rhythm of pig-tailed macaques, Macaca nemestrina. AMERICAN JOURNAL OF PHYSIOLOGY 245: R800-R804,1983.