I.  Preliminary Experiment

        In this experiment, Buri attempted to determine whether genotypes at the bw locus differed in fitness.  He performed a series of crosses to determine whether the frequencies of genotypes in offspring differed from Mendelian expectations, which would be an indication of selection on this locus in the laboratory environment. (see handout Measuring Fitness of Drosophila Genotypes in the Laboratory).  Buri performed three series of crosses, corresponding the Roman numerals in the table.  Series I involved females of a single genotype and males of a single genotype.  These crosses tested for differences among genotypes in viability.  Series II involved males of a single genotype and females of two genotypes and tested for differences among genotypes in a combination of survival and female fecundity.  Series III involved females of a single genotype and males of two genotypes and tested for differences among genotypes in a combination of survival and male fertility.

        As can be seen in the table, in only three trials (II-C-a, III-C-a, and III-C-b) did the observed frequencies of genotypes in the adult offspring differ significantly (more than by chance) from the frequencies expected if there were no selection acting. (Significance of deviation from expected frequencies is assessed by the t statistic, given in the last column of the table.  Significant deviations are marked by asterisks.) When these three trials were repeated, the results showed no differences from expectations under no selection.  Consequently, this series of experiments indicates the absence of detectable natural selection at this locus under laboratory conditions.

Table from P. Buri.  1956.  Gene frequency in small populations of mutant Drosophila.  Evolution 10: 367-402.

    II.  Main Experiment

        In the main experiment, Buri initiated 107 replicate population cages in which the initial frequency of each allele was 0.5.  Each generation, eight males and eight females were chosen randomly from the emerging adult offspring of the previous generation and transferred to a new cage to found the next generation.  The genotypes of these 16 individuals were also scored to determine the gene frequency.  The experiment was run for 20 generations.

        The figure shows the frequency histogram of gene frequency of the bw⁷⁵ allele for the 107 populations for each of the 20 generations.  Note that the variance of the histogram gradually increases and that the number of populations fixed for one allele or the other also increases.  Finally, note that the number of populations fixed for the  bw⁷⁵ allele is approximately the same as for the bw allele, as expected from the initial frequencies of the two alleles.
 

Figure from D. L. Hartl and A. G. Clark.  1989.  Principles of Population Genetics.  Sinauer, Sunderland, MA.
 

            [ Back to Lecture 7 | Lecture Outlines | Bio 120 Home Page | Department of Biology | Duke University ]