332 



GORDON E. DICKERSON 



recent studies of crossbreeding using inbred strains (Hazel el al., 1948; Sierk, 

 1948) verify the earlier conclusions. 



Some degree of dominance is the most obvious genetic mechanism by 

 which change in heterozygosity from inbreeding or crossbreeding would affect 

 the level of performance. Inbreeding decline due to dominance would be a 

 function of 2q{l — q)k f, where q is frequency of the dominant allele, / is 

 Wright's inbreeding coefficient, and k is the degree of dominance (Hull, 

 1945) defined in terms of phenotypic scale as (2 Aa-AA-aa)/{AA-aa). 



TABLE 21.1 



RESULTS OF CROSSBREEDING EXPERIMENTS SUM- 

 MARIZED BY CARROLL AND ROBERTS (1942) 



* From the original publications of these experiments. 



If genetic intermediates in one or more primary functions produce maxi- 

 mum performance, the increased total genetic standard deviation (V 1 + /) 

 associated with inbreeding would tend to increase the average deviation from 

 opti mum genotype and hence depress performance roughly in proportion to 

 (vl + / — !)• Inbreeding alone would not alter mean level of performance 

 without dominance, if only epistatic factors of the complementary or dupli- 

 cate sort were involved. 



Inbreeding depression and crossbreeding advantage indicate some degree 

 of dominance or of genetic intermediate optimum, but, alone, they fail to dis- 

 tinguish between the two or to indicate the probable degree of dominance. 



EfFectiveness of Selection within Inbred Lines 



Selection within mildly inbred lines has been only slightly effective. De- 

 cline in performance with mild inbreeding (2 to 4 per cent per generation) 

 has been much the same as would have been expected from inbreeding with- 

 out selection. These statements are based largely on a study' of time trends 



1. To be published in more detail, separately. 



