INBRED LINES FOR HETEROSIS TESTS? 349 



^ = .5 at each segregating locus. Progress per cycle when q^ = .5 should 

 approach 



qiii - q\) '^ — . 



lap 



Since qi would be increased above initial equilibrium of (1 + k)/2k, maxi- 

 mum progress per cycle should be 



{\-]-k){k-\) sd-' 



W lap' 



and rate of progress would decline as ^i became larger than (1 -(- k)/2k. The 

 maximum rate of progress, then, for selecting population L to complement 

 the cross of two highly complementary strains M and iV, is expected to be 

 little more than one-half that for selection to complement a homozygous 

 tester. 



Other Considerations 



Under heterozygote advantage and selection toward complimentary 

 strains by either the reciprocal or the homozogous tester method, the strains 

 themselves may be expected to decline in performance for characters that are 

 depressed by inbreeding. The less favorable allele would tend to become 

 fixed at about half of the loci segregating in the foundation stocks. The effec- 

 tiveness of this sort of selection in moving gene frequencies toward opposite 

 homozygous extremes in the complementary strains would be greater for 

 those traits in which heterozygote advantage (^ > 1), and hence inbreeding 

 depression, is larger. That portion of the inbreeding depression arising from 

 loci at which there is no heterozygote advantage {k < 1) would not be pro- 

 duced by selection for cross performance without inbreeding, because selec- 

 tion would favor the dominant allele in both strains. Therefore, any serious 

 decline in performance of the strains themselves, while under selection for 

 cross performance, is indicative of heterozygote advantage and should be 

 accompanied by compensatory improvement in performance of the cross. 



In order to develop complementary strains whose own performance would 

 make them usable in commercial production of crosses, some compromise 

 may be necessary between selection based on test-cross and on individual 

 performance. There is much opportunity for selection in choosing young 

 breeders, especially males, to be tested in the strain-cross. Individual selec- 

 tion for characters little affected by inbreeding would be least apt to impair 

 the effectiveness of the complementary selection. Some selection for indi- 

 vidual performance characters important for both the strains and their 

 cross may become necessary to prevent fixation of rare genes with major 

 detrimental effects in the homozygote, but advantageous in the heterozygote. 

 Selection for fertility and maternal influences (e.g, hatchability, prolificacy, 

 or suckling ability) in test-cross matings should help maintain usable strains. 



