Section 9 — Population Genetic* 



from the "Southacre" stock under the same 

 selection regime. Divergence was more rapid in 

 the second experiment than in the first and the 

 two halves of the population became completely 

 distinct by the seventh generation. 



Mating preference tests in this new line using 

 chaeta number as a marker have indicated an 

 isolation index of +0.54 (where would 

 represent random mating and +1 complete 

 positive assortative mating). There is therefore 

 evidence of strong mating preferences in the line. 

 It does not however seem sufficiently strong to 

 provide a complete explanation of the lack of 

 hybrids in the line itself. Further evidence sug- 

 gests that the hybrids that are formed may 

 compete poorly with the non-hybrids. 



1. Nature 193, 1962. 



9.50. Selection for Xanthine Dehydrogenase Activity 

 Levels in Drosophila melanogastei . Edward C. 

 Keller Jr. (Chapel Hill, U.S.A.). 



In an attempt to obtain information about the 

 quantitative genetic control of individual pro- 

 teins, a series of selection and breeding experi- 

 ments were undertaken using xanthine dehy- 

 drogenase levels as the selected phenotype. 

 Various methods of selection have been effective 

 in isolating strains which have very high or very 

 low levels of this enzyme. One high activity line 

 (three or four times the "standard" Oregon-R 

 strain) is of a dominant or overdominant nature 

 and it has not as yet been possible to localize the 

 genetic cause or causes of this high activity. A 

 random sample of about 100 wildtype strains of 

 diverse geographic origin were also assayed for 

 their mean xanthine dehydrogenase activities and 

 many differences in activity levels were found. 

 Analyses of the data revealed that these differen- 

 ces were mainly due to genetic causes. Indeed, 

 several strains contained an autosomal recessive 

 gene which when homozygous limited the xan- 

 thine dehydrogenase activity levels to about 25 

 per cent of the „standard" strain. Subsequent 

 analyses have shown that this gene is located 

 near the center of the left arm of the third 

 chromosome. Therefore, this new mutant is not 

 an allele of the rosy or maroon-like eye color 

 mutants, which have no detectable amounts of 

 xanthine dehydrogenase. 



9.51. Selection for Interpopulation Heterosis. Ken- 



ichi Kojima (Raleigh, U.S.A.). 



The base populations of this study were two 

 divergent long-term cage populations of D. pseu- 

 doobscura. The trait used was egg production. 

 Within each cage there was no significant 

 additive genetic variance, although there existed 

 substantial non-additive hereditary variances. 

 Their hybrid progenies did not show heterosis, 

 but crossbred half-sib covariance was significant. 

 Reciprocal recurrent selection, a scheme in 

 which two populations are selected for their 

 populational combining ability, was employed 

 in order to increase interpopulational heterosis. 



Approximately 40 per cent gain in crossbred 

 progeny performance over the original hybrid 

 was obtained by the 11th cycle of selection, with 

 only slight increase in purebred performance of 

 the two populations. The increase in crossbred 

 performance was accompanied by a reduction 

 of additive genetic variance in crossbred proge- 

 nies. During the next 9 cycles the crossbred 

 performance more or less plateaued, while 

 significant improvement was observed in pure- 

 bred performance. Thus, in reference to the 

 average performance of the base populations, 

 the level of heterosis was increased from zero to 

 40 per cent. With reference to the performance of 

 purebred progenies at various stages of selection, 

 however, the degree of heterosis changed from 

 nil through the highest to a low value of 10 per 

 cent. 



This research was supported in part by a 

 research grant (GM 08202-03) from the National 

 Institutes of Health. 



9.52. A Chromosomal Analysis of Genetic Variation 

 in Drosophila melanogaster. J. F. Kidwell, 

 J. W. Gowen and Janice Stadler (Ames, 

 U.S.A.). 



The marked inversion outcross technique 

 was used to form two lines of Drosophila that 

 were isogenic for all four chromosomes. The 

 81 possible combinations among females were 

 made. Egg production during the sixth, seventh 

 and eighth day and the number of chaeta on 

 the fourth and fifth abdominal segments were 

 measured. Theanalysis was that of a 3 4 factorial 

 experiment, which provided a complete partition 

 into additive, dominance and epistatic genotypic 

 effects. Additive and dominance effects, and some 

 two and three factor interactions, were important. 

 Single, double and triple heterozygotes showed 

 increasing heterosis. The quadruple heterozygote 

 exhibited less heterosis than the triple heterozygote 

 for chromosomes 1, 2 and 3. 



160 



