INBRED LINES FOR HETEROSIS TESTS? 341 



of genes and of heterozygote su{>eri()rity in food utilization, if not in selective 

 value. 



It seems inevitable that manifold effects of genes and equilibrium between 

 frequencies of alternative alleles are common{)lace, with relative selective 

 values shifting with the characters given emphasis in selection at each stage 

 of development from conception through maturity. 



EFFECTIVENESS OF METHODS OF SELECTING 

 FOR MAXIMUM HETEROSIS 



The evidence presented provides several related assumptions concerning 

 the nature of genetic variability in economic characters of swine as the basis 

 for considering how selection for maximum heterosis can be made most effec- 

 tive. These are: (1) Heterozygote advantage (^ > 1) is important for total 

 performance when its components are characters that have had consistently 

 positive selective values, although lesser degrees of dominance may obtain 

 for individual characters. (2) Average gene frequency approaches an inter- 

 mediate equilibrium near q.i = (1 + k)/2k, whose value and stability depend 

 on the intensity, consistency, and duration of selection. (3) Performance 

 levels attainable by selection in outbred populations are far below the maxi- 

 mum heterozygote, because more than one-half of the individuals are homo- 

 zygous at each locus. (4) Inbreeding decline may be considered as due largely 

 to the reduced number of genes useful to the species that can be carried by 

 the more homozygous individuals, rather than to fixation of unfavorable re- 

 cessive genes. 



Under these assumptions, any method of selecting for maximum perform- 

 ance will involve (1) selection for maximum proportion of heterozygous 

 loci in crosses of complementary strains, and (2) selection based on progeny 

 tests of individuals or lines in crosses. These methods are indicated only 

 when individual and family selection become relatively ineffective, because 

 the intensity of selection per unit of time is much lower for selection based on 

 test-cross progeny performance. 



Importance of Recurrent Selection to Achieve Maximum Heterosis 



Hull (1945) has emphasized the great importance of utilizing cumulative 

 gains from recurrent selection for heterosis in crosses, rather than relying on a 

 single selection among Fi crosses of a group of homozygous lines. This prin- 

 ciple may be illustrated by contrasting the observed distribution for number 

 of heterozygous loci in a population of Fi crosses among inbred lines with the 

 potential range (Fig. 21.4). It can be shown that the standard deviation 

 in proportion of heterozygous loci is: 



-4 



\2q{\-q){\-J)[\-2q{\-q ){\-J) ] 



