Male-biased reproduction and sex-ratio adjustment in muskrats.

Sex ratios of a population and of litters were sampled in muskrats in Ontario, Canada. Sex ratios of litters sampled from nests were male biased (54% male). Until weaning, no differential costs of producing and rearing male and female young were identified that could account for this greater production of males. Following weaning, however, male-biased dispersal of juveniles from their natal site and more frequent acquisition by females of these sites as breeding sites the following year suggested a greater investment by adult females in female young. Therefore, competition between female siblings for the acquisition of their natal site may be sufficient to result in the greater production of males. In addition, the simultaneous occupation of, and competition between, siblings and parents for the resources of the natal home range may not be necessary for local resource competition to result in a greater production of the dispersing sex. Greater-than-expected binomial variance in sex ratios of litters suggested that adjustment of sex-ratios occurred. However, we were unable to associate the adjustment of litter sex ratios with changes in maternal condition. The greater production of males and the predominance of monogamous associations between adults in this population may have lead to slightly greater variation in male fitness than female fitness. Therefore, a female in better-than-average condition may have benefited by producing more males. Similarly, a lower cost of producing dispersing males may allow nutritionally-stressed females to reduce their total expenditure on offspring by producing more males. Because these experiments were non-manipulative, maternal condition may not have varied sufficiently during this study to detect adjustments of litter sex ratios resulting from either of the above mechanisms acting separately, but the combined effects of small differences in matermal condition and selective pressures operating in the same direction may have resulted in the observed deviation from the binomial.