Changes in Populations | 143 



a series of experiments in which he reared D. pseiidoobsctira in 

 population cages. His experimental populations were started with 

 known frequencies of chromosome types, and then he repeatedly 

 sampled them to determine what changes in frequency, if any, had 

 occurred. He found that in populations maintained at 16.5 °C there 

 was no change in the chromosome frequency. However, in popula- 

 tions maintained at temperatures above 20 °C the frequencies 

 change, usually arriving at an equilibrium point at which all the 

 original types are still present but in frequencies quite diflFerent 

 from the initial ones. 



The data from one population cage experiment are given in Fig. 

 7.9. The cage colony was constituted on March 1, 1946, with indi- 

 viduals selected so that the population had 10.7 percent ST and 

 89.3 percent CH chromosomes. Throughout the remainder of the 

 year the percentage of ST chromosomes increased, at first rapidly 

 and then more slowly, until at the end of the year ( some 15 genera- 

 tions later) it had leveled off at about 70 percent. This pattern of 

 increase and the establishment of an equihbrium strongly suggested 

 a selective advantage of the structural heterozygotes (ST/CH) over 

 both homozygous types. From the standpoint of population genetics, 

 this would be comparable to overdominance for fitness at a single 

 locus. Similar results could be obtained, however, by negative 

 assortative mating (unlikes mating). This latter possibility was ruled 

 out by taking a sample of eggs from the population cage and raising 

 the larvae under optimum conditions so that almost all survived. The 

 different genotypes proved to be present in the expected Hardy- 

 Weinberg frequency, demonstrating that mating was random and 

 that differential fecundity or fertility was not involved. Samples of 

 adult flies taken from the population, however, showed the following 

 deviations from Hardy-Weinberg frequencies: 



ST/ST ST/CH CH/CH 



These results, along with the establishment of equilibrium at about 

 70 percent, clearly indicate that there is differential ehmination of 

 the homozygous types between the egg and adult stages. It has 

 been demonstrated that in eggs laid by wild flies there is no signifi- 

 cant deviation from the expected Hardy-Weinberg frequencies but 



