Observations on Food Consumption and Preference in Four Alaskan Mammals

D URING the summer of 1950 a number of living specimens were collected during studies on temperature regulation in Alaskan mamma1s.l These animals were brought back to Wisconsin and maintained in captivity, a number being still alive more than two years after capture. Since an adequate diet is often essential for successful maintenance, some systematic observations on food preference in terms of kind and amount were carried out. Such studies may be indicative of the requirements and food preferences of the animals in nature. The animals studied were the Alaskan ground squirrels, Citellus parryii ablusus (8) and C. osgoodi (2), the Dawson red-backed vole, Clethrionomys mtilus dawsoni (8), the Alaskan collared lemming, Dicrostonyx rubricatus rubricatus (2), and the pika, Ochotona collaris (1). Some details as to habitat, distribution, abundance, and mode of capture of these animals are described elsewhere (Strecker and Morrison, 1952; Strecker et al., 1952). The following observations were made in Madison during October and November, about two months after capture, during which time the weights of the animals were substantially constant, that is there was no steady gain or loss in body weight. The lemmings and the pika were kept in individual cages in a window well in the laboratory while the ground squirrels and the voles were caged in groups in an animal room. The temperature of the latter was always near 22°C while that of the window well averaged about 17°C but showed considerable diurnal variation with low values of 0°C being recorded on two cold nights. However, all animals were liberally supplied with bedding material so that the effective temperature variation was much less. The animals were fed a t 24-hour intervals (48-hour over Sunday) at which time the residual food was removed, weighed, and discarded. The different foodstuffs were kept in separate glass or porcelain dishes of appropriate size. Food that fell through the mesh floor was screened from the sawdust and weighed with the other residual food. Occasionally it was necessary to shake out cached food from the cotton waste bedding. Contamination with urine or feces presented no great problem since the animals tended to deposit excreta a t a definite location in the cage, often in the water dish. Consumption values were calculated by the difference in weight of the residual food from that supplied at the beginning of the period. For wet