Selection for litter size in mice in the presence and absence of gonadotropin treatment.

HE extent to which reproductive characteristics can be modified by selection Tis of interest in understanding the nature of genetic variation in characteristics closely correlated with fitness. It is also relevant to the question of effecting genetic improvement in an important trait in species producing protein for human food. Different species and races within species have acquired widely different reproductive rates in the course of their evolution, but there is extensive evidence that intrapopulation genetic variability, or at least the component of it providing the basis for response to mass selection, is low. Such evidence has been reported, for example, for Drosophila (ROBERTSON 1957; KOJIMA and KELLEHER 19863) ; mice (FALCONER 1960b; DALTON and BYWATER 1963) ; swine (BOYLAN, REMPEL and COMSTOCK 1960) and sheep (REEVE and ROBERTSON 1953). This low heritability is commonly attributed to the association of reproductive rate with fitness, i.e., to the fact that natural selection has acted directionally with respect to reproductive capacity, and hence most of the additive genetic variation has been fixed (KOJIMA and KELLEHE,R 1963). The experiments to be reported here are part of a series designed to compare the effectiveness of alternative methods of selection for reproductive capacity, using the laboratory mouse as the experimental animal. This paper presents the results of selection for number born in untreated and superovulated lines. Superovulation treatment has been reported ( FALCONER et al. 1961 ) to increase phenotypic variability in litter size; selection following such treatment was carried out to determine if this is an effective means of exposing latent genetic variability.