(1937:200-202), Brumwell (1951:193-200; 213), Dice (1922:46) and Johnson (1926). Lantz (1907) discussed the economic relationships of M. ochrogaster; the section of his report concerning the effects of voles on vegetation was especially useful to me.

Fisher (1945) studied the voles of central Missouri and obtained information concerning food habits and nesting behavior. Jameson (1947) studied M. ochrogaster on and near the campus of the University of Kansas. His report is especially valuable in its treatment of the ectoparasites of voles. In my investigation I have concentrated on those aspects of the ecology of voles not treated at all by Fisher and Jameson, or mentioned but not adequately explored by them. Also I have attempted to obtain larger samples.

The University of Kansas Natural History Reservation, where almost all of the field work was done, is an area of 590 acres, comprising the northeastern-most part of Douglas County, Kansas. Situated in the broad ecotone between the deciduous forest and grassland, the reservation provides a variety of habitat types (Fitch, 1952). Before 1948, much of the area had been severely overgrazed and the original grassland vegetation had been largely replaced by weeds. Since 1948 there has been no grazing or cultivation. The grasses have partially recovered and, in the summer of 1952, some grasses of the prairie climax were present even on the parts of the Reservation which had been most heavily overgrazed. Illustrative of the changes on the Reservation were those observed in House Field by Henry S. Fitch (1953: in litt.). He recalled that in July, 1948, the field supported a closely grazed, grassy vegetation providing insufficient cover for Microtus, with such coarse weeds as Vernonia, Verbena and Solanum constituting a large part of the plant cover. By 1950, the same area supported a lush stand of grass, principally Bromus inermis, and supported many woody plants. Similar changes occurred in the other study areas on the Reservation. Although insufficient time has elapsed to permit analyses of successional changes, it seems that trees and shrubs are gradually encroaching on the grassland throughout the Reservation.

The vole population has changed radically since the Reservation was established. In September and October of 1948, when Fitch began his field work, he maintained lines of traps totaling more than 1000 trap nights near the future vole study plots without capturing a single vole. In November and December, 1948, he caught several voles near a small pond on the Reservation and found abundant sign in the same area. Late in 1949 he began to capture voles over the rest of the Reservation, but not until 1950 were voles present in sufficient numbers for convenient study.

I first visited the Reservation and searched there for sign of voles in the summer of 1949. I found hardly any sign. In the area around the pond mentioned above, however, several systems of runways were discovered. This area had been protected from grazing for several years prior to the reservation of the larger area. In House Field, where my main study plot was to be established, there was no sign of voles. Slightly more than a year later, in October, 1950, I began trapping and found Microtus to be abundant on House Field and present in smaller numbers throughout grassland areas of the Reservation.

GENERAL METHODS

The present study was based chiefly on live-trapping as a means of sampling a population of voles and tracing individual histories without eliminating the animals. Live-trapping disturbs the biota less than snap-trapping and gives a more reliable picture of the mammalian community (Blair, 1948:396; Cockrum, 1947; Stickel, 1946:158; 1948:161). The live-traps used were modeled after the trap described by Fitch (1950). Other types of traps were tested from time to time but this model proved superior in being easy to set, in not springing without a catch, in protecting the captured animal and in permitting easy removal of the animal from the trap. A wooden box was placed inside the metal shelter attached to each trap and, in winter, cotton batting or woolen scraps were placed inside the boxes for nesting material. With this insulation against the cold, voles could survive the night unharmed and could even deliver their litters successfully. In summer the nesting material was removed but the wooden box was retained as insulation against heat.

Bait used in live-traps was a mixture of cracked corn, milo and wheat, purchased at a local feed store. The importance of proper baiting, especially in winter, has been emphasized by Howard (1951) and Llewellyn (1950) who found an adequate supply of energy-laden food, such as corn, necessary in winter to enable small rodents to maintain body temperature during the hours of captivity. The rare instances of death of voles in traps in winter were associated with wet nesting material, as these animals can survive much lower temperatures when they are dry. Their susceptibility to wet and cold was especially evident in rainy weather in February and March.

Preventing mortality in traps was more difficult in summer than in winter. The traps were set in any available shade of tall grass or weeds; or when such shade was inadequate, vegetation was pulled and piled over the nest boxes. The traps usually were faced north so that the attached number-ten cans, which served as shelters, cast shadows over the hardware cloth runways during midday. Even these measures were inadequate when the temperature reached 90°F. or above. Such high temperatures rarely occurred early in the day, however, so that removal of the animals from traps between eight and ten a. m. almost eliminated mortality. Those individuals captured in the night were not yet harmed, but it was already hot enough to reduce the activity of the voles and prevent further captures until late afternoon. When it was necessary to run trap lines earlier, the traps were closed in the morning and reset in late afternoon.

Reactions of small mammals to live-traps and the effects of prebaiting were described by Chitty and Kempson (1949). In general, the results of my trapping program fit their conclusions. Each of my trapping periods, consisting of seven to ten consecutive days, showed a gradual increase in the number of captures per day for the first three days, with a tendency for the number of captures to level off during the remainder of the period. Leaving the traps baited and locked open for a day or two before a trapping period tended to increase the catch during the first few days of the period without any corresponding increase during the latter part of the period. Initial reluctance of the voles to enter the traps decreased as the traps became familiar parts of their environment.

At the beginning of the study the traps were set in a grid with intervals of 20 feet. The interval was increased to 30 feet after three months because a larger area could thus be covered and no loss in trapping efficiency was apparent. The traps were set within a three foot radius of the numbered stations, and were locked and left in position between trapping periods.

Each individual that was captured was weighed and sexed. The resulting data were recorded in a field notebook together with the location of the capture and other pertinent information. Newly captured voles were marked by toe-clipping as described by Fitch (1952:32). Information was transferred from the field notebook to a file which contained a separate card for each individual trapped.

In the course of the program of live-trapping, many marked voles were recaptured one or more times. Most frequently captured among the females were number 8 (33 captures in seven months) and number 73 (30 captures in eight