LK MUTATION, SELECTION, AND POPULATION FITNESS 35 



Table 1. — Genotypic Proportions in a Population. 



columns of Table 1, ireq(A) = p and freq(a) = q, although the geno- 

 typic proportions for the three mating systems are quite different. 

 If the three genotypic proportions are denoted by D, H, R, whatever 

 the mating system, the gene frequencies can always be calculated 

 according to the relation 



p = D + \/ 2 H, q = i/ 2 H + R 



where D + H + R= 1. Should D + H + R- W for some reason, the 

 obvious modification is p = (D + i/ 2 H)/W, and q = (l/ 2 H + R)/W, 

 so that p + q = 1. 



C. Genotypic Selection 



Let w 1} w 2 , w 3 denote the selective fitness values of the three geno- 

 types whose frequencies in the population are D, H, R, as shown in 

 Table 2. The direct meaning of the w's is the probability with which 

 the genotypes survive to reproduce; that is, we assume for the sake 

 of simplicity that selection operates prior to reproduction. Then it is 

 the selected group (that survives to reproduce) that determines the 

 gene frequency of the next generation. The general procedure of 

 calculating the new gene frequency after selection is given in Table 

 2, and the selection effect on gene frequency is Aq = q' - q per gen- 

 eration. In this simple model the w's are assumed to be constants and, 

 as far as the selection effect on q is concerned, only their relative 

 magnitude is relevant. Any other set of three fitness values propor- 

 tional to w 1 :w 2 :w 3 will yield the same value of Aq. Hence, for prac- 

 tical calculation, one of the three w's can always be taken as unity. 



To illustrate the balance between selection and mutation, I will 

 mention only two examples: (a) When selection is against a dominant 

 mutant gene (Table 2A) and (b) when selection is against the recessive 

 genotype (Table 2B). In the former case, the frequency of the domi- 

 nant deleterious gene is so low that we may assume homozygous 



