more abundant. A graveled and heavily used road along one edge of House Field, leading to the Reservation Headquarters, was a barrier which voles rarely crossed. A little-used dirt road crossing the trapping plot in Quarry Field constituted a less effective barrier. The disturbed areas bordering the roads were likewise little used and tended to reinforce the effects of the roads as barriers. There were almost pure stands of Bromus japonicus along both roads. No mammal of any kind was taken in traps set where this grass was dominant.

Because seasonal changes in vole density followed the curve for rate of growth of the complex of grasses on the Reservation, and because years in which there was a sparse growth of plants due to dry weather showed a decrease in the density of voles, the relationships between productivity of plants and vole population levels on the two study areas were investigated. In both fields the composition of the plant cover was similar, and the differences were chiefly quantitative. In June, 1951, ten square-meter quadrats were clipped on each of the areas to be studied. The clippings from each were dried in the sun and weighed. From Quarry Field the mean yield amounted to 1513 ± 302 lbs. per acre; while from House Field the yield was 2351 ± 190 lbs. per acre (Table 1). Using experience gained in making these samples, I periodically estimated the relative productivity of the two areas. House Field was from 1.5 to 3 times as productive as Quarry Field during the growing seasons of 1951 and 1952. Although House Field, being more productive, usually supported a larger population of voles than Quarry Field the reverse was true at the time of the clipping (Fig. 5).

Table 1. Relationship Between Yield and Various Population Data

Although no explanation was discovered which accounted fully for the seeming aberration, two sets of observations were made that may bear on the problem. In June, 1951, the population of voles and cotton rats on Quarry Field was increasing rapidly whereas in House Field that trend was reversed. The trends were reflected by the percentages of immature individuals in the two populations and by the ratios of the June, 1951, densities to the March, 1951, densities (Table 1). Perhaps the density curve was determined in part by factors inherent in the population and, to that extent, was fluctuating independently of the environment (Errington, 1946:153).

The flood in 1951 reduced the population of voles and obscured the normal seasonal trends. Although House Field produced a heavier crop of vegetation, Quarry Field produced a larger crop of rodents, chiefly Microtus and Sigmodon. In House Field, however, the ratio of Sigmodon to Microtus was notably higher. Presumably the cotton rats competed with the voles and exerted a depressing effect on their numbers. The intensity of the effect seemed to depend on the abundance of both species. That this depressing effect involved more than direct competition for plant food was suggested by the fact that in House Field, with a heavy crop of vegetation and a seemingly high carrying capacity for both herbivorous rodents, the biomass of voles, and of all rodents combined, were lower than in Quarry Field which had less vegetation and fewer cotton rats. The relationships between voles and cotton rats are discussed further later in this report.

When the centers of activity (Hayne, 1949b) of individual voles were plotted it was seen that there was a shift in the places of high density of voles on the trapping areas. This shift seemed to be related to the advance of the forest edge with such woody plants as Rhus and Symphoricarpos and young trees invading the area. These shifts were clearly shown when the distribution of activity centers on both areas in June, 1951, was compared with the distribution in June, 1952 (Fig. 1). The shift was gradual and the more or less steady progress could be observed by comparing the monthly trapping records. It was perhaps significant that during the summers the centers of activity were less concentrated than during the winter. The shift of voles away from the woods was more nearly evident in winter when the voles were driven into areas of denser ground cover, which provided better shelter.

Fig. 1. Progressive encroachment of woody vegetation onto study areas, and the accompanying shift of the centers of populations of voles. Activity centers of individuals were calculated as described by Hayne (1949b) and are indicated by dots. The cross-hatched areas show places where the vegetation was influenced by the shade of woody plants.

View larger image

From 1948 to 1950 and again in 1952 and 1953 I trapped in various habitat types in a mixed prairie near Hays, Kansas. Before the great drought of the thirties, Microtus ochrogaster was the most common species of small mammal in that area. Since 1948, at least, it has been taken only rarely and from a few habitats. No voles have been taken from grazed sites. In a relict area, voles were trapped in a lowland association dominated by big bluestem. Since 1948 only one vole has been trapped in the more