Changes in Populations 151 



with artificial selection. A great deal of work in artificial selection 

 has been concerned with so-called quantitative characters. ( Quanti- 

 tative characters are those influenced by many pairs of alleles at 

 many different loci, such as height and intelligence in man, bristle 

 number in Drosophila, color pattern in water snakes, etc. ) 



The work of Mather and Harrison is a rather typical example of 

 artificial selection. In that case, selection was for changes in the 

 number of abdominal chaetae in Drosophila melanogaster. The 

 diagram of Fig. 7.11 summarizes the results of more than 100 gen- 

 erations of selection for higher bristle number. For 20 generations, 

 selection was extremely effective. At generation 20, reduced fertility 

 and fecundity in the population made it necessary to discontinue 

 selection in order to keep the strain from dying out. In the absence 

 of selection over several generations, the population reached an 

 equilibrium chaeta number higher than the original one but much 

 lower than that achieved at the peak in generation 20. From the line 

 in which selection was suspended, a new selected line was extracted 

 at generation 24, and progress was rapid to a point near that 



Fig. 7.11 I Artificial selection for abdominal bristle number in Dro- 

 sophila melanogaster. Ordinate, mean number of bristles; abscissa, 

 number of generations; lines under selection are solid, lines not under 

 selection are dashed; T indicates deliberate termination of a line, D that 

 it died out through sterility; circled numbers indicate different lines. 

 From line 1, all the selections were for low number of bristles, except 

 that marked H which was for high number. ( From Mather and 

 Harrison, 1949, Heredity 3. ) 





25 30 35 



Generations 



60 



