In [Fig. 1] the catch per day of voles, varying from month to month, reflects chiefly the changing population density. However, other factors also have important effects on the catch. For example, bait acceptance is better in the winter when natural foods, especially greens, are scarce, with the result that a higher catch can be made with the same population density. Interference with the trap line by other animals also affected the catch of voles. In warm weather the traps were checked in both morning and evening, and the catch was correspondingly greater than it was in cool weather when the traps were checked only once daily. The ratios obtained of young to adult voles cannot be accepted at face value as the true ratios in the population, either. For the first several days of each trapping period, the voles caught were mostly adults previously marked and, presumably, conditioned to the grain bait. Later, young voles not previously recorded, came to the traps in increasing numbers. The young, being at first not conditioned to the bait, and also having relatively small home ranges, would generally be less well represented in the catch than would the adults.
In other species of Microtus, so far as known, a 21-day gestation period seems to be the rule (Bailey, 1924:528; Hamilton, 1941:13; Hatfield, 1935:264). M. ochrogaster seems to conform to this pattern, but the data obtained were meager, because breeding activity was usually inhibited in voles kept in confinement.
A female live-trapped on July 23, 1951, appeared to be in breeding condition. When trapped two days later, she had a copulatory plug, and 21 days after this she was found with a newborn litter in a trap. A female thought to have given birth to a litter between successive captures on July 20, and July 21, 1951 (on the basis of appearance of genitalia, and reduction in weight from 53 to 46 grams), appeared to have just completed parturition when she was examined on August 10. A female that gave birth to a litter in confinement on May 18, 1954, bred and was released the same day. She was recorded as pregnant in the first week of June, but on June 7 was no longer pregnant. If this pregnancy terminated normally, a gestation of 20 days or less is indicated.
Greenwald (1956:221) suggested that in M. californicus, oestrus might occur in the period of lactation, because he found recently formed corpora lutea in lactating females. In the course of my field work on M. ochrogaster, I obtained precise or approximate dates of successive litters born at intervals of somewhat more than 21 days apart. In different females, intervals of 23, 23, 24, 26, and approximately 27 (between 26 and 28) days were recorded between successive litters. In four other females intervals between litters were known only approximately because one of two records was based on a capture in late pregnancy judged to be within two or three days of parturition. For these females, intervals of 23, 24, 24, and 26 days were recorded. From the trend of these records, it seems that females often became pregnant within a few days after birth of a litter. Pregnancy from post-partum oestrus would seem to be less frequent than pregnancies beginning a few days after birth of the previous litter, and within the period of lactation.
Jameson (1947:146) found an average of 3.4 young per litter in 58 litters of M. ochrogaster from northeastern Kansas, mostly from Douglas County. Martin (1956:386) recorded a somewhat lower mean of 3.18 ± 0.24 in 65 litters on the Reservation in 1950, 1951, and 1952. For a total of 82 litters recorded from 1950 through 1956, inclusive, I obtained an average of 3.37 ± .075 young per litter. Several litters that were recorded were excluded from this computation as in each instance there was reason to suspect that they were incomplete. These included instances of females found in traps with young several days old, females that may not have completed parturition when they were released with newborn young, and those litters that might have sustained losses through cannibalism by the mother or her trap-mates.
Mean numbers of young per litter were found to vary from year to year and from month to month, as shown by the following lists: 1950, 3.0 (13 litters); 1951, 3.5 (23 litters); 1952, 3.5 (11 litters); 1953, 3.4 (5 litters); 1954, 3.4 (15 litters); 1955, 4.1 (7 litters); 1956, 3.8 (5 litters); January 2.0 (1 litter); February 3.5 (4 litters); March 4.5 (4 litters); April 3.9 (12 litters); May 3.3 (25 litters); June 3.0 (9 litters); July 2.7 (4 litters); August 2.9 (7 litters); September 2.8 (6 litters); October 3.4 (7 litters); November 5.0 (2 litters); December 4.0 (1 litter).
These differences can be logically explained on the basis of changes in the average age of the breeding females in the population. On the average, with greater length, weight and age, females produced progressively larger litters, although individuals did not necessarily conform to this general trend. For 24 females recorded in 1954-1956 and measured within a few days of birth of their litters, average length was correlated with number of young as follows: 6 young, 163.5 mm.; 5 young, 158.0 mm.; 4 young, 157.7 mm.; 3 young, 154.6 mm.; 2 young, 160.5 mm.
For 48 other females, recorded in 1950-1953, that were not measured, but that were mostly assignable to broad age groups on the basis of their individual histories in the trapping records, the following well defined trend was demonstrated.