Suprachiasmatic nuclear lesions eliminate circadian rhythms of drinking and activity, but not of body temperature, in male rats

Male Long-Evans rats were maintained in light proof cabinets while drinking, activity, and telemetered body temperature (Tb) data were collected. After suprachiasmatic nuclear (SCN) lesions, the rats were exposed to a 12:12 light-dark cycle, a 6-hr delay in the lighting cycle, and constant dark. Lesions that abolished the drinking and activity rhythms did not eliminate the Tb rhythm. However, the amplitude, phase, and free-running period of the Tb rhythm were altered. Lesions that only partially damaged the SCN had similar, though lesser effects. In some cases, Tb rhythms remained normal, activity rhythms were only temporarily disrupted, and drinking rhythms were eliminated in the same animals. These results support the conclusion that Tb can remain rhythmic after lesions that permanently or temporarily disrupt other circadian rhythms. Of the three rhythms, it appears that drinking rhythms are most easily and Tb rhythms least easily disrupted by SCN lesions.     

Body size and the daily rhythm of body temperature in dogs

Practically every physiological variable exhibits daily rhythmicity. The daily rhythm of body temperature, like that of many other variables, is often weak in newborns and gains strength as the animals grow. Because of the natural association between physiological maturation and gain in body size, these two processes are naturally confounded. To differentiate between the effects of maturation and the effects of body growth, we took advantage of the large variation in body size that exists among different breeds of the domestic dog. We compared the body temperature rhythms of developing puppies of different dog breeds. Puppies of none of the breeds exhibited statistically significant daily rhythmicity for several days after birth. Regardless of breed or sex, rhythmicity matured over several weeks and attained a stable level by 6 weeks after birth. Body size did not seem to be an important element in the development of rhythmicity because the development was similar in three breeds that differed greatly in body size (Basset Hound, Boxer, and Neapolitan Mastiff). On the other hand, the difference in body size associated with the different breeds had a strong impact on the absolute level of body temperature regardless of age: we found a strong inverse correlation between temperature and body size among the puppies and dams of the three breeds and among 115 adult dogs from 19 different breeds ranging from 2-kg Yorkshire Terriers to 80-kg Great Danes.     

Plasticity of Circadian Activity and Body Temperature Rhythms in Golden Spiny Mice

Most animals can be categorized as nocturnal, diurnal, or crepuscular. However, rhythms can be quite plastic in some species and vary from one individual to another within a species. In the golden spiny mouse (Acomys russatus), a variety of rhythm patterns have been seen, and these patterns can change considerably as animals are transferred from the field into the laboratory. We previously suggested that these animals may have a circadian time‐keeping system that is fundamentally nocturnal and that diurnal patterns seen in their natural habitat reflect mechanisms operating outside of the basic circadian time‐keeping system (i.e., masking). In the current study, we further characterized plasticity evident in the daily rhythms of golden spiny mice by measuring effects of lighting conditions and access to a running wheel on rhythms in general activity (GA) and body temperature (Tb). Before the wheel was introduced, most animals were active mainly during the night, though there was considerable inter‐individual variability and patterns were quite plastic. The introduction of the wheel caused an increase in the level of nighttime activity and Tb in most individuals. The periods of the rhythms in constant darkness (DD) were very similar, and even slightly longer in this study (24.1±0.2 h) than in an earlier one in which animals had not been provided with running wheels. We found no correlation between the distance animals ran in their wheels and the period of their rhythms in DD. Re‐entrainment after phase delays of the LD cycle occurred more rapidly in the presence than absence of the running wheel. The characteristics of the rhythms of golden spiny mice seen in this study may be the product of natural selection favoring plasticity of the circadian system, perhaps reflecting what can happen during an evolutionary transition as animals move from a nocturnal to a diurnal niche.     

Diel rhythms of adipokinetic hormone, fat body response, and haemolymph lipid and sugar levels in the house cricket

Abstract. The adipokinetic hormone (Grb-AKH) content in the corpora car-diaca of the house cricket, Acheta domesticus , varied during a day with two peaks in the scotophase and one peak in the photophase. There were two distinct peaks of total lipid concentration in the blood, one early in the photophase and the other early in the scotophase. Fat body sensitivity to adipokinetic hormone also varied in close synchrony with the lipid rhythm. It was not possible to attribute the rhythm of blood lipid titre unequivocally to either the rhythm of adipokinetic hormone content in the corpora cardiaca or to the rhythm of sensitivity of the fat body to the hormone. In adult crickets the blood carbohydrate titre had two peaks in adult females, one towards the end of scotophase and another in the late photophase (c. 12h apart), but a single peak at the end of scotophase was apparent in last instar larvae. The blood carbohydrate rhythm persists in DD and is therefore endogenous. Carbohydrates were not mobilized by the Grb-AKH, therefore could not be involved in the blood carbohydrate rhythm. Exposure to various day lengths caused shifts in the patterns of the carbohydrate rhythms, but imposition of a cyclic temperature regime had no effect on the rhythm.     

Daily rhythms of emergence of small invertebrates inhabiting shallow subtidal zones: A comparative investigation at four locations in Japan

Many small invertebrates inhabit the shallow subtidal zone and some of them emerge at times into the water column. The daily timing of their emergence is affected by the day/night and tidal cycles, and shows various patterns of synchrony with these cyclical factors, depending on the species. To detect possible regional differences in their emergence patterns, sampling was carried out at four locations in Japan: a boreal sea (Akkeshi), a temperate sea (Sugashima), an inland sea (Ushimado) and a subtropical sea (Iriomote-jima). The emergence patterns of major taxa were examined by visual inspection and by two statistical methods (periodogram and autocorrelogram). The composition of the taxa collected by the pump system, mostly crustaceans, was similar in each location. The number of taxa that emerged revealed a day/night rhythm in every location. This characteristic was clearest at Iriomote-jima and least clear at Sugashima. The daily fluctuation in the number of individuals in each taxon varied widely, from very clearly nocturnal to weakly diurnal patterns. In Iriomote-jima, the major taxa all showed well-demarcated nocturnal patterns, so these patterns were classified as either level N₂ or N₃ with regard to the degree of synchrony with the day/night cycle. With regard to the synchrony with the tide, the majority of patterns in all locations showed a double-tidal interval. Many patterns were slightly modified by the tidal cycle. These patterns were classified as level T₁ or T₂ with regard to the degree of synchrony with the tidal cycle. The synchrony with the tide was comparatively strong at Ushimado. The synchrony with day/night and tidal cycles varied even within the same species or closely related species. In benthic invertebrates, hiding or resting in the bottom substrates and swimming in the water column would occur alternatively. In planktonic animals, aggregation near the bottom and dispersal in the water column would occur alternatively. The daily timing of such activities may be synchronized with the day/night and tidal cycles to various degrees among species or populations, resulting in a wide variety of emergence patterns in subtidal small invertebrates. This type of behavior is not daily (diel) vertical migration; it should rather be called daily emergence/dispersal. Strong winds, rough waves and unknown seasonal factors would also affect emergence patterns. Furthermore, the transparency of the seawater may also strongly affect these patterns. Nocturnal patterns may be an adaptation to avoid vulnerability to sighted predators. Variation of synchrony with the tide indicates that by definition, the tidal rhythm can only be distinguished from the day/night rhythm. Hence, the daily patterns that are weakly modified by the tides (levels T₁ and T₂) should be called the tidal rhythms. As the period of such rhythms cannot be determined exactly by using statistical methods, lengthy field investigations and visual inspection of each pattern is essential to assess the influence of tides.     

Maternal entrainment of tau mutant hamsters.

Maternal entrainment of the circadian wheel-running activity rhythm was examined in Syrian hamsters heterozygous for a single gene mutation (tau) that affects the free-running period of circadian rhythms. Heterozygous tau pups were born to and raised by wild-type mothers under constant dim light. The pups' wheel-running activity was recorded after weaning on postnatal day 18 or 24. Pups weaned on day 18 had an average free-running period of 21.70 hr, demonstrating that the tau phenotype was fully expressed at this age. Using the activity onset of the postnatal free-running rhythms as a phase reference, we estimated the phase relationships between the pups and their mothers on days 18 and 24. In contrast to results with wild-type pups, the activity rhythms of tau pups were not in phase with the rhythms of their wild-type mothers; that is, activity onsets of mothers and pups did not coincide. The pups did, however, show synchrony among themselves, indicating that they had been exposed to a synchronizing signal sometime during development. It is likely that this synchronizing signal was provided by the mothers, since pups from different litters showed phase relationships similar to those of their mothers. Thus the mothers provided a signal that was sufficient to cause entrainment, despite the 2-hr difference in free-running period between the mothers and pups. Although the pups' activity rhythms appeared to have been entrained by the mothers, they were clearly free-running by postnatal day 18. The mechanism for entrainment is lost during the course of development, despite continued interaction between the mothers and pups.     

Melatonin implantation during spring and summer does not affect the seasonal rhythm of feeding in anadromous Arctic charr (Salvelinus alpinus)

Plasma melatonin levels in the high-latitude teleost Arctic charr (Salvelinus alpinus) are constantly low during summer when feeding activity is high, and high during the dark winter when they eat little and loose weight. The question arises if melatonin is involved in the phase-setting of annual rhythms of feeding and growth and if low summer melatonin production is permissive for high summer growth in this species. The present study was therefore set out to compare the seasonal appetite and growth rhythms in Arctic charr with constantly high plasma melatonin levels from February throughout the Arctic summer (melatonin implanted, average mid-day plasma melatonin levels 1,106 ± 147 pg/ml) with those of fish with natural plasma melatonin levels (vehicle implanted and untreated fish with average mid-day plasma melatonin levels of 94 ± 13 and 58 ± 6 pg/ml, respectively). Feed intake, body mass or body length, as well as the timing of the seasonal growth rhythm, were not affected by the high summer plasma melatonin level. Further, Arctic charr fasted for 3 months had a 24 h plasma profile of melatonin which was consistently higher throughout the scotophase compared to fed charr. Although the daily melatonin production seems to be affected by the energy status of the fish, melatonin does not seem to be directly involved in regulation of the seasonal feeding and growth rhythm in the high-latitude, anadromous Arctic charr.     

The daily pattern of heart rate, body temperature, and locomotor activity in guinea pigs.

We studied the characteristics of the rhythmicity of heart rate (HR), body temperature (BT), and locomotor activity (LA) in conscious and unrestrained guinea pigs using a telemetry system. HR and/or LA in some guinea pigs clearly showed circadian rhythms, but in others there were no significant daily patterns; BT did not show significant daily rhythms. These results suggest that guinea pigs might have different individual characteristics of rhythmicity, and we should, therefore, be careful when using guinea pigs in chrono-biomedical research. We believe that the results of this study may be useful for future biomedical studies using guinea pigs.     

大绒鼠及高山姬鼠体温调节和蒸发失水的日节律

为比较横断山区同域分布物种大绒鼠(Eothenomys miletus)和高山姬鼠(Apodemus chevrieri)的日节律特征,对两种鼠在24 h中4个时间段(04:00～06:00时、10:00～12:00时、16:00～18:00时和22:00～24:00时)的体温和蒸发失水进行了测定.结果显示,大绒鼠、高...     

Thermoregulatory circadian rhythms in the pouched mouse (Saccostomus campestris)

1. Circadian rhythms of body temperature (Tb), oxygen consumption (VO2), and minimal thermal conductance (C) were studied in the pouched mouse, Saccostomus campestris under natural photoperiod during February at a constant ambient temperature of 28 degrees C. 2. Circadian rhythms of body temperature were also studied under natural photoperiod and laboratory temperatures (Max: 28.1 degrees C; Min: 23.2 degrees C) during February. 3. The results of the present study suggest that changes in ambient temperature are not the main "zeitgeber" for body temperature rhythm, and it seems that photoperiod plays a major role in this species. 4. The relationship between the rhythms of Tb, VO2, and C are further discussed.     

Development of Adrenocortical Cyclic Nucleotide (Cyclic AMP and Cyclic GMP) and Corticosterone Orcadian Rhythms in Male and Female Rats

The daily rhythm of the adrenocortical cyclic nucleotides (cyclic AMP and cyclic GIMP) was studied in infant male and female Wistar rats before and after the establishment of an adult-like daily rhythm of plasma corticosterone. As in this strain the rhythm of corticosterone is known to be present on postnatal day 18, pups of 2 and 3 weeks of age were studied. The dams and the pups as well as the young adult animals were kept on a controlled 12L-12D photoperiod. Groups of 8-10 pups were killed at 4-hr intervals throughout the day. Plasma corticosterone levels and adrenal cyclic AMP and cyclic GMP concentrations were simultaneously measured and the daily patterns established. Pups of 2 weeks of age showed neither plasma corticosterone nor adrenal cyclic AMP rhythms whereas pups of 3 weeks of age exhibited a typical adult-like circadian rhythm for both variables. The patterns for adrenal cyclic GMP differed according to sex: In female pups no cyclic GMP circadian rhythm could be detected at either 2 or 3 wk. In male pups of 3 wk a typical mature rhythm for adrenal cyclic GMP was evident whereas in younger male pups (2 wk) a circadian rhythm was detected. This circadian rhythm, however, differed from mature circadian rhythm in that its peak was located at 1300 hr instead of 0700 hr. These results demonstrate that, unlike that of cyclic AMP the adrenal cyclic GMP circadian rhythm does not appear at the same time as the plasma corticosterone circadian rhythm. Moreover, a circadian rhythmicity for adrenal cyclic GMP can be found in the absence of any corticosterone circadian rhythm. These facts argue against the view of cyclic GMP being a mediator of ACTH-stimulated steroidogenesis.     

Development of Adrenocortical Cyclic Nucleotide (Cyclic AMP and Cyclic GMP) and Corticosterone Orcadian Rhythms in Male and Female Rats

The daily rhythm of the adrenocortical cyclic nucleotides (cyclic AMP and cyclic GIMP) was studied in infant male and female Wistar rats before and after the establishment of an adult-like daily rhythm of plasma corticosterone. As in this strain the rhythm of corticosterone is known to be present on postnatal day 18, pups of 2 and 3 weeks of age were studied. The dams and the pups as well as the young adult animals were kept on a controlled 12L-12D photoperiod. Groups of 8–10 pups were killed at 4-hr intervals throughout the day. Plasma corticosterone levels and adrenal cyclic AMP and cyclic GMP concentrations were simultaneously measured and the daily patterns established. Pups of 2 weeks of age showed neither plasma corticosterone nor adrenal cyclic AMP rhythms whereas pups of 3 weeks of age exhibited a typical adult-like circadian rhythm for both variables. The patterns for adrenal cyclic GMP differed according to sex: In female pups no cyclic GMP circadian rhythm could be detected at either 2 or 3 wk. In male pups of 3 wk a typical mature rhythm for adrenal cyclic GMP was evident whereas in younger male pups (2 wk) a circadian rhythm was detected. This circadian rhythm, however, differed from mature circadian rhythm in that its peak was located at 1300 hr instead of 0700 hr. These results demonstrate that, unlike that of cyclic AMP the adrenal cyclic GMP circadian rhythm does not appear at the same time as the plasma corticosterone circadian rhythm. Moreover, a circadian rhythmicity for adrenal cyclic GMP can be found in the absence of any corticosterone circadian rhythm. These facts argue against the view of cyclic GMP being a mediator of ACTH-stimulated steroidogenesis.     

Birth of viable puppies derived from breeding cloned female dogs with a cloned male

Since the establishment of production of viable cloned dogs by somatic cell nucleus transfer, great concern has been given to the reproductive abilities of these animals (Canis familiaris). Therefore, we investigated reproductive activity of cloned dogs by (1) performing sperm analysis using computer-assisted sperm analysis and early embryonic development, (2) assessing reproductive cycling by measuring serum progesterone (P4) levels and performing vaginal cytology, and (3) breeding cloned dogs using artificial insemination. Results showed that most parameters of sperm motility in a cloned male dog were within the reference range, and in vivo–matured oocytes from a noncloned female were successfully fertilized by spermatozoa from a cloned male dog and develop normally to the 8-cell stage. Three cloned female dogs displayed normal patterns of P4 levels and morphologic changes of the vaginal epithelium. Two cloned female dogs became pregnant using semen from a cloned male dog and successfully delivered 10 puppies by natural labor. In conclusion, these data demonstrated that both cloned male and female dogs are fertile, and their puppies are currently alive and healthy with normal growth patterns.     

Time-lapse analysis of the circadian rhythms of conidiation and growth rate in neurospora

The filamentous fungus Neurospora crassa has frequently served as a model organism for the study of circadian rhythms through its ability to form conidial spores on a daily basis. This phenomenon leaves a spatial pattern of conidiation bands along a solid surface of agar after several days of growth. Using time-lapse video, the authors have quantified the rate of conidiation. They have found that conidia do not form at a specified lag time after the growth front is laid down, but rather the band region tends to simultaneously develop over a short time frame. This produces a sharp peak when the conidiation rate is plotted against time. In addition, the authors used time-lapse video to assay growth rate with greater accuracy than previously reported. It is usually assumed that Neurospora's rate of growth is constant, and this assumption of linear growth has been used extensively to determine period and phase of the conidiation circadian rhythm. The authors have confirmed an earlier report of nonlinear growth rate and have shown that the growth rate varies by a factor of about 2 with each circadian cycle. They have demonstrated that the errors in calculating times of conidiation peaks are maximally 1 to 2 h if linearity is assumed. The conidiation rate and growth rate rhythms are not apparent under conditions (using mutants or high or low temperatures) where the spatial banding rhythm is not observed. In light/dark entraining conditions, the conidiation rate and growth rate rhythms maintain the same phase relationship in different T-cycles. These data are consistent with the hypothesis that the growth rate rhythm is a consequence of the conidiation rate rhythm.     

Shift Work or Food Intake during the Rest Phase Promotes Metabolic Disruption and Desynchrony of Liver Genes in Male Rats

In the liver, clock genes are proposed to drive metabolic rhythms. These gene rhythms are driven by the suprachiasmatic nucleus (SCN) mainly by food intake and via autonomic and hormonal pathways. Forced activity during the normal rest phase, induces also food intake, thus neglecting the signals of the SCN, leading to conflicting time signals to target tissues of the SCN. The present study explored in a rodent model of night-work the influence of food during the normal sleep period on the synchrony of gene expression between clock genes and metabolic genes in the liver. Male Wistar rats were exposed to forced activity for 8 h either during the rest phase (day) or during the active phase (night) by using a slow rotating wheel. In this shift work model food intake shifts spontaneously to the forced activity period, therefore the influence of food alone without induced activity was tested in other groups of animals that were fed ad libitum, or fed during their rest or active phase. Rats forced to be active and/or eating during their rest phase, inverted their daily peak of Per1, Bmal1 and Clock and lost the rhythm of Per2 in the liver, moreover NAMPT and metabolic genes such as Pparα lost their rhythm and thus their synchrony with clock genes. We conclude that shift work or food intake in the rest phase leads to desynchronization within the liver, characterized by misaligned temporal patterns of clock genes and metabolic genes. This may be the cause of the development of the metabolic syndrome and obesity in individuals engaged in shift work.     

Geological cycles

Abstract

Deep body temperature (DBT) and heart rate (HR) circadian rhythms were determined by radiotelemetry in 4 mares kept under controlled light and temperature conditions. Ovulations were determined by rectal palpation of their ovaries. Mean DBT values ranged from 35.85 ± .04 to 37.22 ± .02°C The circadian range of oscillation was extremely low, approximately 0.5° C, with time of maximum temperature occurring midway through the dark period. Mean HR values ranged from 36.4 ± 1.7 to 53.0 ±3.6 beats per min. The circadian range of oscillation was also low, less than 15 beats per min with time of maximum HR occurring approximately at the time of lights off. The HR circacadian rhythm peaked before the DBT circadian rhythm by 3 to 8 hrs. Ovulation did not appear to consistently affect DBT and HR circadian rhythms or their phase relationships.     

From daily behavior to hormonal and neurotransmitters rhythms: Comparison between diurnal and nocturnal rat species

Mammalian species can be defined as diurnal or nocturnal, depending on the temporal niche during which they are active. Even if general activity occurs during nighttime in nocturnal rodents, there is a patchwork of general activity patterns in diurnal rodents, including frequent bimodality (so-called crepuscular pattern, i.e., dawn and dusk peaks of activity) and a switch to a nocturnal pattern under certain circumstances. This raises the question of whether crepuscular species have a bimodal or diurnal - as opposed to nocturnal - physiology. To this end, we investigated several daily behavioral, hormonal and neurochemical rhythms in the diurnal Sudanian grass rat (Arvicanthis ansorgei) and the nocturnal Long-Evans rat (Rattus norvegicus). Daily rhythms of general activity, wheel-running activity and body temperature, with or without blocked wheel, were diurnal and bimodal for A. ansorgei, and nocturnal and unimodal for Long-Evans rats. Moreover, A. ansorgei and Long-Evans rats exposed to light-dark cycles were respectively more and less active, compared to conditions of constant darkness. In contrast to other diurnal rodents, wheel availability in A. ansorgei did not switch their general activity pattern. Daily, unimodal rhythm of plasma leptin was in phase-opposition between the two rodent species. In the hippocampus, a daily, unimodal rhythm of serotonin in A. ansorgei occurred 7 h earlier than that in Long-Evans rats, whereas a daily, unimodal rhythm of dopamine was unexpectedly concomitant in both species. Multiparameter analysis demonstrates that in spite of bimodal rhythms linked with locomotor activity, A. ansorgei have a diurnally oriented physiology.     

Daily levels and rhythm in circulating corticosterone and insulin are altered with photostimulated seasonal states in night-migratory blackheaded buntings

The circadian rhythms are involved in the photostimulation of seasonal responses in migratory blackheaded buntings. Here, we investigated whether changes in daily levels and rhythm in corticosterone (cort) and insulin secretions were associated with transitions in the photoperiodic seasonal states. Buntings were exposed to short days to maintain the winter (photosensitive) non-migratory state, and to long days for varying durations to induce the premigratory, migratory (shown by migratory restlessness at night, Zugunruhe) and summer non-migratory (photorefractory) states. We monitored activity patterns, and measured plasma cort and insulin levels at six and four times, respectively, over 24 h in each seasonal state. Buntings were fattened and weighed heavier, and exhibited intense nighttime activity in the migratory state. The daytime activity patterns also showed seasonal differences, with a bimodal pattern with morning and evening activity bouts only in the summer non-migratory state. Further, the average baseline hormone levels were significantly higher in premigratory and migratory than in the winter non-migratory state. Both cort and insulin levels showed a significant daily rhythm, but with seasonal differences. Whereas, cort rhythm acrophases (estimated time of peak secretion over 24 h) were at night in the winter non-migratory, premigratory and migratory states, the insulin rhythm acrophases were found early in the day and night in winter and summer non-migratory states, respectively. These results suggest that changes in daily levels and rhythm in cort and insulin mediate changes in the physiology and behavior with photostimulated transition in seasonal states in migratory blackheaded buntings.     

Dopaminergic Regulation of Circadian Food Anticipatory Activity Rhythms in the Rat

Circadian activity rhythms are jointly controlled by a master pacemaker in the hypothalamic suprachiasmatic nuclei (SCN) and by food-entrainable circadian oscillators (FEOs) located elsewhere. The SCN mediates synchrony to daily light-dark cycles, whereas FEOs generate activity rhythms synchronized with regular daily mealtimes. The location of FEOs generating food anticipation rhythms, and the pathways that entrain these FEOs, remain to be clarified. To gain insight into entrainment pathways, we developed a protocol for measuring phase shifts of anticipatory activity rhythms in response to pharmacological probes. We used this protocol to examine a role for dopamine signaling in the timing of circadian food anticipation. To generate a stable food anticipation rhythm, rats were fed 3h/day beginning 6-h after lights-on or in constant light for at least 3 weeks. Rats then received the D2 agonist quinpirole (1 mg/kg IP) alone or after pretreatment with the dopamine synthesis inhibitor α-methylparatyrosine (AMPT). By comparison with vehicle injections, quinpirole administered 1-h before lights-off (19h before mealtime) induced a phase delay of activity onset prior to the next meal. Delay shifts were larger in rats pretreated with AMPT, and smaller following quinpirole administered 4-h after lights-on. A significant shift was not observed in response to the D1 agonist SKF81297. These results provide evidence that signaling at D2 receptors is involved in phase control of FEOs responsible for circadian food anticipatory rhythms in rats.     

Circadian behavioral and melatonin rhythms in the European starling under light-dark cycles with steadily changing periods: evidence for close mutual coupling?

In European starlings exposed to constant conditions, circadian rhythms in locomotion and feeding can occasionally exhibit complete dissociation from each other. Whether such occasional dissociation between two behavioral rhythms reflects on the strength of the mutual coupling of their internal oscillators has not been investigated. To examine this, as well as to elucidate the role of melatonin in this system, we simultaneously measured the rhythms of locomotion, feeding and melatonin secretion in starlings exposed to light-dark (LD) cycles of low intensity with steadily changing periods (T). In birds initially entrained to T 24 LD cycles (12L:12D, 10:0.2 lx), beginning on day 15, T was either lengthened to 26.5 h (experiment 1) or shortened to T 21.5 h (experiment 2) by changing the daily dark period 4 min each day. After 18 and 19 cycles of T 26.5 and T 21.5, respectively, birds were released into constant dim light conditions (LL(dim); 0.2 lx) for about 2 weeks. Locomotor and feeding rhythms were continuously recorded. Plasma melatonin levels were measured at three times: in T 24, when T equaled 26 or 22 h and at the end of T 26.5 or T 21.5 exposure. The results show that, contrary to our expectations, the three rhythms were not dissociated. Rather they remained synchronized and changed their phase angle difference with the light zeitgeber concomitantly and at the same rate. The melatonin rhythm stayed in synchrony with the behavioral rhythms and as a consequence, peaked either during day or at night, depending on the phase relationship between the activity rhythm and the zeitgeber cycle.