The periodicity of daily activity and its seasonal changes in free-ranging and captive kangaroo rats.

Populations ofDipodomys microps andD. merriami in eastern California (37°11'N. Lat.) are active on the surface throughout the night and during the whole year. These two species, and the males and females within each species, show no significant differences in times of beginning or end of activity. Beginning and end of nightly activity generally fall within the brighter part of the twilights, even though the total time spent on the surface by individuals is at times only an hour or two, or even less, per night. Near the summer solstice in southern Saskatchewan, Canada (50°45'N. Lat.), when sunset and sunrise were 7 h 43 min apart, the time from onset to end of activity in a population ofD. ordii (the northernmost of all kangaroo rats) was only 6 h; these animals were also only on the surface at intervals during the night.During the course of the year, the light intensity at which the first individualD. microps andD. merriami became active on the surface varied between 200 and 2,000 lux, and the light intensity at which the last individual was active in the morning varied between 50 and 20,000 lux. There was no apparent influence of the moonlight cycle upon the onset and end of activity; in fact the light levels at beginning and end of activity are about 2 to 4 orders of magnitude brighter than that at full moon. The onset of activity is fairly synchronous within the population in that most individuals surface within about a fhalf hour of each other; furthermore, two-thirds of the individuals appearing in the first 32 min of activity had already appeared within the first 12 min. There appears to be a similar, synchronous, but less well marked end of activity in the morning.There were systematic seasonal changes in the time relative to sunset and sunrise respectively, and in the corresponding light intensities, at which the animals began and ended activity. The phase relationship between the onset of activity and sunset (ψ onset) showed tow maxima and two minima per year, but the phase relationship between the end of activity and sunrise (ψ end) showed only one maximum and one minimum per year. On the other hand, the annual range of change in the evening phaseψ onset (28 min) was half the annual range of change in the morning phaeψ end (57 min). Such a conspicuous difference in frequency and amplitude ofψ onset andψ end has not previously been reported for any species, nor has it been predicted by models of circadian rhythms and phase relationships. The resemblance of the course of annual change inψ onset to annual change in length of twilight and the resemblance of annual change inψ end to change in length of night are discussed. These differences may underline a circadian system based on two separate, but normally coupled components, which could be separately synchronized by dusk and dawn respectively.The daily running-wheel activity of 10 individually housedD. merriami under natural skylight in Los Angeles (34°05'N. Lat.) began and ended at light intensities two to three orders of magnitude lower than corresponding values for surface activity in the field. The onset and end of activity for the captive population did not show the synchronous peaks typical of the field populations; nor did the nightly activity patterns of individuals show peaks at the beginning and end of the night. The social isolation of the captive kangaroo rats is suggested as a factor contributing to differences between activity parameters in captivity and field.Proximate and ultimate factors shaping the periodicity of activity in kangaroo rats are discussed. It is hypothesized that social interactions involving some form of territoriality may give rise to a population peak in onset of activity in some solitary mammals such as kangaroo rats. A set of criteria is presented for comparing vertebrate species with regard to the periodicity and synchronization of both daily and seasonal functions.