One of the central questions in the field of circadian biology concerns the regulation of multiple circadian outputs such as feeding and locomotor activity. While multiple outputs conceivably may be regulated by a single oscillatory system, there is some evidence that separate oscillators control different outputs in humans (Aschoff and Wever, 1976), rodents (Pittendrigh and Daan, 1976b) and birds (Ganshirt et aI., 1984). In birds it is unclear whether the circadian system that regulates locomotor rhythms also regulates other behavioral rhythms. While much is known about the regulation of locomotor rhythms, other rhythmic outputs, such as feeding, have not been well characterized.
The daily rhythm of feeding has only been measured in house sparrows, starlings and pigeons. Studies in which either surgical or environmental manipulations were performed suggest that the three species differ in their responses to similar treatments. For example, following pinealectomy (P-X) the feeding rhythm is abolished in house sparrows (Gwinner, 1989; Chabot and Menaker, in preparation), but not in starlings (Gwinner et aI., 1987). There is evidence for starlings, but not for house sparrows, that the regulation of feeding rhythms may be fundamentally different than the regulation of locomotor rhythms (Ganshirt et al., 1984). In pigeons there is an indication that feeding rhythms do not persist in constant conditions (Yamada et al., 1988).
We have begun to address the question of circadian regulation of feeding in the homing pigeon by monitoring both feeding and locomotor activity simultaneously and exposing pigeons to conditions known to affect locomotor rhythms. The purpose of this series of experiments was to determine if feeding activity in the pigeon is under circadian control, and if so, to determine if procedures known to affect locomotor rhythms in this species affect feeding rhythms in a similar manner.
Experimental Birds and Housing Conditions
The homing pigeons (Columba livia; males and females, 400 - 600 grams) used in this experimenf were maintained either outdoors under natural lighting conditions or indoors under an artificial light-dark (LD) cycle of 12 hours light: 12 hours dark (12:12 LD) prior to experimentation. The birds were obtained from suppliers in Waco, TX (Louis Nicosia) and Eugene, OR (University of Oregon). During experimentation, pigeons were individually housed in tilt bottom cages (35 cm X 38 cm X 26 cm) suspended over a pan filled with bedding (Bed-O-Cob, The Andersons, Maumee, OH) within light tight, wooden boxes in a temperature controlled (23 degrees Celsius) dark room. White noise (93 dBA) was continuously present (Menaker and Eskin, 1966). Food and water were available ad libitum at opposite ends of the cage and were replaced at approximately weekly intervals. An infrared viewer (FJW Optical Systems, Elgin, IL) was used when pigeons were held in constant darkness (DD). Bedding was replaced at approximately biweekly intervals. Pigeons were always allowed at least 10 days in a 12:12 LD cycle when first placed in the cage. Locomotor activity was monitored with two microswitches connected to the tilt bottom and feeding activity was monitored with an infrared emitter/detector pair mounted across a food access hole. Behavioral data were collected on an event recorder (Esterline-Angus,
Indianapolis, IN) or on a computer data acquisition system (Data Quest, Data-Sciences, Inc., Roseville, MN; EZ Paste software, M. Vogelbaum, Charlottesville, VA).
Experiment 1: Circadian control of feeding and the role of the pineal gland.
Pigeons were divided into two groups: Pineal-intact (n=l1) and P-X birds (n=8). Intact birds were used to determine if the feeding activity of normal pigeons was rhythmic DD and to determine if the feeding and locomotor rhythms in DD would exhibit similar characteristics [ie., period, length of daily activity bout (alpha)]. Intact pigeons were first entrained to an LD cycle and then placed in DD until the experiment was terminated, 45 150 days later. Pigeons were P-X to determine the role of the pineal in the maintenance of feeding rhythms and to test for differential effects of P-X on locomotor and feeding rhythms. Pinealectomy was performed using one of two methods: 1) Four pigeons were P-X as described by Ebihara et al., (1984); 2) In four pigeons the procedure was changed in the following ways: The rostral portion of the caudal sinus was ligated, cut, and retracted caudally. The pineal stalk was grasped with mouse-toothed forceps, pulled up and out of the "interhemispheral fissure" with choroid plexus attached and retracted caudally. The sinus was then completely removed along with the pineal body and stalk. In both cases the removed tissue was examined under a dissecting scope to verify that the whole gland was removed. Because there were no apparent differences in the responses of the pigeons to the two P-X procedures, the data were pooled for analysis. After entrainment to an LD cycle for at least 10 days, the P-X pigeons were then placed in DD for the remainder of the experiment (50 - 150 days).
Experiment 2: The effects of melatonin implants on feeding and locomotor rhythms.
Crystalline melatonin (Sigma) filled silastic (1.0 cm length of Dow Corning No. 602-235, 1.47 mm i.d. and 1.96 mm o.d.) implants were prepared as described by Turek et aI., (1976). The melatonin release rate from these capsules was demonstrated to be directly proportional to capsule length in house sparrows (Turek et aI., 1976; 4 uglday/5 mm length of capsule). Intact pigeons (n=4) were entrained to an LD cycle and then placed in DD for 25 35 days. The birds, free-running under these DD conditions, were then removed from their cages into the light and melatonin filled silastic capsules were placed in the peritoneal cavity. Two pigeons received 11 capsules, one received 8, and one received 6 as described by Ebihara et aI., (1984). The daily dose, based on the rate determined for house sparrows (Turek et aI., 1976), was 88, 64, and 48 ug/day respectively. Mer recovery the pigeons were returned to their cages in DD. Several weeks later, each pigeon was again removed from its cage and the melatonin capsules were removed. Mer recovery, the pigeon was replaced in its cage in DD until the experiment was terminated several weeks later.
Experiment 3: The effects of constant light (LL) on feeding and locomotor rhythms.
Intact pigeons (n=4) were entrained to LD for at least 10 days and then were placed in one of several intensities of LL (range 0.17 to 1.0 lux; measured at head height). Two to nine weeks later, after a stable response to the initial intensity of LL was achieved, the intensity was changed; it was decreased if the initial response was arrhythmic feeding and locomotor behavior or increased if the response was rhythmic behavior. Three to five weeks later, after the behaviors had once again stabilized, the intensity was either increased or decreased depending on the observed activity patterns as in the previous set of treatments. In total, three birds were subjected to three different intensities of LL (0.5, 0.7, 0.17 lux) and one bird was subjected to four different intensities of LL (0.17, 0.35, 0.7, 1.0 lux).
Data Analysis
The period, alpha, and phase difference between locomotor and feedingrhythm offsets and/or onsets were measured for all of the free-running locomotor and feeding activities which exhibited clear offsets or onsets. Behavioral offsets were used in nearly all cases for period and phase angle determination because they were usually found to be clearer than onsets. The period was determined from the slope of a best eye-fit line (Eskin, personal communication) drawn through the offsets of the activities. Phase angle differences between locomotor and feeding activity offsets were calculated by drawing a best eye-fit line through the offsets of both behaviors and subtracting the feeding offset from the locomotor offset. A positive phase angle indicates that locomotor activity ended later than feeding activity. Alphas of feeding and locomotor activity were calculated by drawing best eye fit lines through the onsets and the offsets of activity and determining the length of time between the two lines. Means and standard error of the means (SEM) were calculated separately for period, phase angles, and alphas of all four (intact, P-x, melatonin implants, and LL) experimental groups. Differences between periods and alphas of feeding and locomotor rhythms were analyzed using Student's t test. Values were considered significantly different if P < 0.05. The locomotor activity was suppressed in two out of the eight P-X birds. Therefore, neither the feeding nor locomotor records of these two P-X pigeons were included in the quantitative analyses.
Experiment 1: Intact and Pinealectomized Pigeons in LD and DD
The feeding and locomotor activities of both intact and P-X pigeons entrained to an LD 12:12 cycle with nearly all of both activities taking place in the light portion of the cycle. Both rhythms free-ran in DD with similar periods (Figures 1, 2 and Table 1). In two out of the eight P-X pigeons the locomotor rhythm was disrupted in DD, while their feeding activity was clearly rhythmic (data not shown). While the behavioral data from the P-X bird (Figure 2) are less clear than the data from the intact (Figure 1) these results were not consistently observed. Several of the behavioral records from P-X birds were clearer than the behavioral records from two intact birds. There were no significant differences between the feeding and locomotor free running periods in either intact or P-X birds. There were no significant differences between the free-running periods of intact versus P-X pigeons. The locomotor activity alpha was larger than the feeding activity alpha in DD for both intact and P-X pigeons and the phase angle difference between the two behavioral rhythms was positive in both (Table 1).
Experiment 2: Melatonin Implants
Constant release of large doses of melatonin from silastic implantsdisrupted rhythmicity in all four of the pigeons subjected to this treatment (Figure 3). The rhythms remained disrupted while the capsules remained implanted. When the capsules were removed rhythmicity was immediately restored to both behaviors. Thereafter, both feeding and locomotor rhythms in individuals exhibited virtually identical free running periods with stable phase relationships until the experiment was terminated several weeks later. The average locomotor activity alpha was longer than the average feeding activity alpha (Table 1).
Experiment 3: Constant Light
In Figure 4, both feeding and locomotor rhythms are shown to persist in LL of 0.5 lux. This occurred in 3 of the 4 pigeons tested; however in one pigeon clear feeding and locomotor rhythms did not persist at any LL intensity tested (0.7, 0.25, 0.17 lux). In the other three pigeons both rhythms were disrupted when the LL intensity was increased sufficiently (Figure 4). In those free-running patterns which were discernable, the average period of the feeding and of the locomotor rhythms were similar (Table 1). The average locomotor alpha was slightly longer than the average feeding alpha and the phase angle between locomotor and feeding activity was larger for pigeons in LL than in DD (Table 1). Significant effects of LL intensity on free-running feeding and locomotor periods were not observed, perhaps due to the small number of experimental subjects.
Our results indicate that feeding activity in homing pigeons maintains rhythmicity in constant conditions with a period slightly different from 24 hr and is therefore regulated by a circadian oscillatory system. Perturbations known to disrupt locomotor rhythms in the pigeon - exposure to LL and implantation of melatonin capsules - also disrupt the circadian rhythms of feeding activity. These results suggest that melatonin, because it has been demonstrated to be important in the maintenance of locomotor rhythmicity (Foa and Menaker, 1988; Ebihara et aI., 1984), may also be involved in the maintenance of circadian feeding rhythms in the pigeon.
Feeding and locomotor rhythmicities appear nearly identical in intact pigeons, in melatonin implanted pigeons and in pigeons subjected to LL. Thus these behaviors may be controlled by the same circadian oscillatory system. The circadian rhythm of body temperature may also be controlled by the same system; Oshima et aI., (1989b) have shown that by P-X and/or bilateral enucleation affected temperature and locomotor rhythms similarly. Alternatively there may be separate, strongly coupled oscillatory systems which control individual circadian outputs in the pigeon. Although there are no data which require such an interpretation, the information currently available is insufficient to rule it out.
Our results do show some differences between feeding and locomotor rhythms. Under free-running conditions in both LL and DD, locomotor activity began before, and ended after, feeding activity. While P-X did not affect the clarity of feeding rhythms in any of the eight birds subjected to this procedure, the amplitude of locomotor rhythmicity was diminished in two of these eight birds. The feeding rhythm for any given individual was often clearer than it's locomotor rhythm. Although these differences in some of the qualities of the two free-running rhythms were consistent, the periods of the two rhythms were always nearly identical. These results suggest to us that the differences are a reflection of differences in the output pathways to the two behaviors from a single circadian pacemaker.
Our findings in the homing pigeon are similar to those found in studies of multiple rhythms in house sparrows. Perturbational experiments have been shown to affect locomotor, feeding and body temperature rhythms in the same manner in this species. P-X abolished both locomotor and body temperature rhythms (Binkley et aI., 1971) and P-X (Chabot and Menaker, in preparation; Gwinner, 1989), LL, or melatonin implants all affected both locomotor and feeding rhythms in a similar manner (Chabot and Menaker, manuscript in preparation).
In both pigeons and starlings feeding rhythms are more robust and resistant to perturbations than are locomotor rhythms. Feeding rhythms persist in both P-X pigeons and P-X starlings while locomotor rhythms are suppressed in some individuals (Gwinner et aI., 1987). In starlings, P-X (Gwinner et aI., 1987)and exposure to LL (Ganshirt et aI., 1984) have been found to have different effects on locomotor and feeding rhythms. Also in starlings, testosterone implants differentially affect the daily activity bout length (alpha) of feeding and locomotor rhythms (Subbaraj and Gwinner, 1985). While either P-X or LL abolish disrupt locomotor rhythmicity under some circumstances, neither are able to abolish feeding rhythmicity. Gwinner et aI., (1987) have suggested that one circadian system may control both behaviors but that the coupling of the locomotor behavior to the circadian system may be weaker than that of the feeding behavior. This hypothesis is supported by observations that the feeding and locomotor rhythms were never dissociated (ie - they never free-ran with different periods). Our data in homing pigeons is consistent with that hypothesis although the difference in coupling strength appears smaller in pigeons than in starlings.