Effect of Artificial Insemination on Frequency of Undesirable Recessive Genes

Reduction in the frequency of' undesirable recessive genes is examined for several combinations of initial gene frequencies and number of offspring obtained by random mating of young sires to the population. Only sires not having affected offspring are then returned to heavy use. Such selection can veD" rapidly reduce the number of affected offspring, if the initial gene frequency is high or if the number of tested offspring is 200-500 for lower gene frequencies. For some combinations, frequency of affected offspring in the population is higher after two or three generations of selection than in the first generation of selection. With high initial gene frequencies, practically all heterozygous males are detected and culled during the test period, resulting in practically no affected offspring in the first generation. The population gene frequency remains relatively high, due to the influence of the unselected females. In the next generation the heterozygous males are not detected with high probability and the frequency of affected offspring increases above that of the first generation. After the new peak of affected offspring is reached in a few generations, frequency drops, as expected. One criticism of the artificial insemination (AI) industry has been that the frequency of deleterious recessive genes could be increased very rapidly by unknowing use of carrier males. Yarious methods of testing sires for heterozygosity have been proposed~mating to known carriers, mating to own daughters, etc. Another proposal has been to mate sires at random in the population. Such a test would parallel a production proof made on a limited number of progeny. I f the frequency of a recessive gene were high in the population, a carrier of the recessive gene would be detected with high probability and few affected offspring would result when tested bulls were used heavily. I f the gene frequency were low, carriers would not be detected with very high probability, but the frequency of affected offspring would not be Received for publication August 9, 1966. high because of the low frequency of the gene in the female and male populations. Gene frequency might be increased temporarily, but in the next generation carriers would be detected with higher probability and no damage would be done. This concept of balance seems reasonable enough, but has never been formally presented. The pro-pose of this paper is to describe in terms of population genetics the effects of an A.I. testing program on the frequency of undesirable recessive genes. None of the concepts described is original, since the basis of the entire discussion is the Hardy-Weinberg law and the properties of random union of gametes from populations of males and of females. Hopefully, however, this development will make some practical application of these genetic principles to a problem of the A.I. industry.