The influence of weather on activity rhythms may be manifested in many ways, including energetics of maintaining body temperature, initiation of breeding, availability of food, and change in the environment such as formation of ice on lakes and rivers. Gray squirrels at Cedar Creek showed a marked reduction in activity in winter (Tester, 1987; Figala and Tester, 1986). These squirrels were under a strong influence of cold temperatures in January and February. During these months they spent their resting time in tree cavities or in nests made of dry leaves. Protection provided by these nests was probably necessary to maintain their critical physiological temperature. The squirrels were active only during the time of day when ambient temperatures were at maximum.
Badgers (Meles meles) in Britain also exhibited a shorter alpha in winter (Harris, 1982). However, no explanation for this change in activity rhythm was given.
The seasonal effects of temperature on the activity of black bears in southeastern North America revealed marked seasonal changes. In spring, increases in level of activ-' ity were associated with increases in ambient temperature until the' temperature reached approximately 20°C, where activity appeared to stabilize. In summer, activity also increased as temperature increased, until about 23°C when activity declined. In fall, activity increased during warmer periods and decreased when the temperatures exceeded 20°C (Garshelis and Pelton, 1980).
The effect of freezing of lakes can best be illustrated by data on muskrats at Cedar Creek (Tester, 1987). From July 1969 through February 1970, the number of activity periods per 24 hours varied from two to four. In July and August four periods of activity were about evenly distributed throughout the 24 hours. In subsequent months the number of activity periods was reduced to three and then to two at the time of freeze-up of lakes and streams. The day active periods were eliminated and muskrats became essentially nocturnal with the total amount of activity per 24 hours being reduced to approximately half that exhibited in late summer.
The activity rhythm for a breeding male woodcock (Scolopax rusticola) monitored in Britain changed markedly from April to May (Hirons and Owen, 1982). In early spring, woodcock spent the night on pasture where they were active and presumably feeding on worms. They were inactive for most of the day, which was spent in forest. Within one month, as nights became shorter and the grass longer and the ground on pasture fields harder, the birds switched to feeding in the forest during the day and resting at night.
Switches from nocturnal to diurnal feeding during winter in response to a changing food supply were reported for red fox (Vulnes fulva) by Sargeant (in preparation). A switch from a diurnal pattern of feeding to nocturnal feeding was reported for sea otters (Erihvdra lutris) in Alaska by Garshelis (1983) when the otters moved to areas of richer food supply.
Seasonal differences in activity rhythms reported for black bears by Garshelis and Pelton (1980) were reported to be due in part to the abundance and nature of the food supply. Bears lost weight during spring, possibly due to the unavailability of nutritious foods or to inability to digest available foods efficiently after winter dormancy. The low energy and poor digestibility of the high-fiber grasses consumed in the spring resulted in a diet with low nutritional value, which may have restricted the amount of energy bears could expend during mid-day. The crepuscular activity pattern exhibited in the spring could reflect an optimal foraging strategy for bears on a low energy diet (Garshelis and Pelton, 1980). During summer bears fed on berries and fruits. The small size and dispersion of this food may have required foraging to continue throughout the middle of the day. Such activity would be possible with a diet high in caloric value. During fall, increased foraging was necessary to increase body fat in preparation for winter dormancy.
In a comparison of the activity rhythms of badgers living in urban and rural sites, Harris (1982) observed that rural badgers had longer active spans per 24 hours. Emergence from the sett was delayed about an hour compared to badgers in the rural study area, but the time of return to the sett was approximately similar. No explanation is given for this difference, but it might be assumed that the disturbance caused by human activities in the urban area resulted in the later initiation of activity.
A more striking example of the effect of disturbance on activity rhythms is illustrated by wild pigs (Sus scrofa) in Europe and Asia. Wild pigs in zoos or in large fenced areas, exposed to much human activity, appeared to have lost their natural shyness and were day active. In
Siberia, Bromlej (1964) observed that Sus scrofa ussuricus was day active, and attributed this to the lack of human disturbance in this region. In contrast, Gundlach (1968) and Briederman (1971) reported that the main peaks of. activity in Germany, where the pigs were exposed to much human activity, occurred before dawn and after dusk. Both authors suggested the influence of a "command vector", of human activity which serves as a social interspecific synchronizer. Somewhat similar changes in phase of activity as a result of disturbance by humans have been reported for roe deer fCapreolus europeus) by Bubenik and Bubenikova (1967), crocodiles (Crocodilus niLoticus) by Corbet (1970) and pine marten CMartes fbina) by Lohrl (1972).