months). Among the males, number 37 (21 captures in six months) and number 62 (21 captures in eight months) were most frequently taken. The mean number of captures per individual was 3.6. For females, the mean number of captures per individual was 3.8 and for males it was 3.4. Females seemingly acquired the habit of entering traps more readily than did males. No correlation between any seasonally variable factor and the number of captures per individual was apparent. To a large degree, the formation of trap habits by voles was an individual peculiarity.

In order to study the extent of utilization of various habitats by Microtus, a number of areas were sampled with Museum Special snap-traps. These traps were set in linear series approximately 25 feet apart. The number of traps used varied with the size of the area sampled and ranged from 20 to 75. The lines were maintained for three nights. The catch was assumed to indicate the relative abundance of Microtus and certain other small mammals but no attempt to estimate actual population densities from snap-trapping data was made. In August, 1952, when the live-trapping program was concluded, the study areas were trapped out. The efficiency of the live-trapping procedure was emphasized by the absence of unmarked individuals among the 45 voles caught at that time.

Further details of the methods and procedures used are described in the appropriate sections which follow.

HABITAT

Although other species of the genus Microtus, especially M. pennsylvanicus, have been studied intensively in regard to habitat preference (Blair, 1940:149; 1948:404-405; Bole, 1939:69; Eadie, 1953; Gunderson, 1950:32-37; Hamilton, 1940:425-426; Hatt, 1930:521-526; Townsend, 1935:96-101) little has been reported concerning the habitat preferences of M. ochrogaster. Black (1937:200) reported that, in Kansas, Microtus (mostly M. ochrogaster) preferred damp situations. M. ochrogaster was studied in western Kansas by Brown (1946:453) and Wooster (1935:352; 1936:396) and found to be almost restricted to the little-bluestem association of the mixed prairie (Albertson, 1937:522). Brumwell (1951:213), in a survey of the Fort Leavenworth Military Reservation, found that M. ochrogaster preferred sedge and bluegrass meadows but occurred also in a sedge-willow association. Dice (1922:46) concluded that the presence of green herbage, roots or tubers for use as a water source throughout the year was a necessity for M. ochrogaster. Goodpastor and Hoffmeister (1952:370) found M. ochrogaster to be abundant in a damp meadow of a lake margin in Tennessee. In a study made on and near the campus of the University of Kansas, within a few miles of the area concerned in the present report, Jameson (1947:132) found that voles used grassy areas in spring and summer, but that in the autumn, when the grass began to dry, they moved to clumps of Japanese honeysuckle (Lonicera japonica) and stayed among the shrubbery throughout the winter. Johnson (1926:267, 270) found M. ochrogaster only in uncultivated areas where long grass furnished adequate cover. He stated that the entire biotic association, rather than any single factor, was the key to the distribution of the voles. None of these reports described an intensive study of the habitat of voles, but the data presented indicate that voles are characteristic of grassland and that M. ochrogaster can occupy drier areas than those used by M. pennsylvanicus. Otherwise, the preferred habitats of the two species seem to be much the same.

In the investigation described here I attempted to evaluate various types of habitats on the basis of their carrying capacity at different stages of the annual cycle and in different years. The habitats were studied and described in terms of yield, cover and species composition. The areas upon which live-trapping was done were studied most intensively.

These two areas, herein designated as House Field and Quarry Field, were both occupied by voles throughout the period of study. Population density varied considerably, however (Fig. 5). Both of these areas were dominated by Bromus inermis, and, in clipped samples taken in June, 1951, this grass constituted 67 per cent of the vegetation on House Field and 54 per cent of the vegetation on Quarry Field. Estimates made at other times in 1950, 1951 and 1952 always confirmed the dominance of smooth brome and approximated the above percentages. Parts of House Field had nearly pure stands of this grass. Those traps set in spots where there was little vegetation other than the dominant grass caught fewer voles than traps set in spots with a more varied cover. Poa pratensis formed an understory over most of the area studied, especially on House Field, and attained local dominance in shaded spots on both fields. The higher basal cover provided by the Poa understory seemed to support a vole population larger than those that occurred in areas lacking the bluegrass. Disturbed situations, such as roadsides, were characterized by the dominance of Bromus japonicus. This grass occurred also in low densities over much of the study area among B. inermis. Other grasses present included Triodia flava, common in House Field, but with only spotty distribution in Quarry Field; Elymus canadensis, distributed over both areas in spotty fashion and almost always showing evidence of use by voles and other small mammals; Aristida oligantha and Bouteloua curtipendula, both more common on the higher and drier Quarry Field; Panicum virgatum, Setaria spp., especially on disturbed areas; and three bluestems, Andropogon gerardi, A. virginicus and A. scoparius. The bluestems increased noticeably during the study period (even though grasses in general were being replaced by woody plants) and they furnished a preferred habitat for voles because of their high yield of edible foliage and relatively heavy debris which provided shelter.

On House Field the most common forbs were Vernonia baldwini, Verbena stricta and Solanum carolinense. On Quarry Field, Solidago spp. and Asclepias spp. were also abundant. All of them seemed to be used by the voles for food during the early stages of growth, when they were tender and succulent. The fruits of the horse nettle (Solanum carolinense) were also eaten. The forbs themselves did not provide cover dense enough to constitute good vole habitat. Mixed in a grass dominated association they nevertheless raised the carrying capacity above that of a pure stand of grass. Other forbs noted often enough to be considered common on both House Field and Quarry Field included Carex gravida, observed frequently in House Field and less often in Quarry Field; Amorpha canescens, more common in Quarry Field; Tradescantia bracteata, Capsella bursapastoris, Oxalis violacea, Euphorbia marginata, Convolvulus arvensis, Lithospermum arvense, Teucrium canadense, Physalis longifolia, Phytolacca americana, Plantago major, Ambrosia trifida, A. artemisiifolia, Helianthus annuus, Cirsium altissimum and Taraxacum erythrospermum. Both areas were being invaded from one side by forest-edge vegetation; the woody plants noted included Prunus americana, Rubus argutus, Rosa setigera, Cornus drummondi, Symphoricarpus orbiculatus, Populus deltoides and Gleditsia triacanthos.

In House Field the herbaceous vegetation was much more lush than in Quarry Field and woody plants and weeds were