Deleterious mutation and evolution: Empirical data and theoretical approaches

The work in this dissertation is an attempt to quantify the impact of deleterious mutations on the evolution of populations by (1) measuring the rate and studying the molecular nature of spontaneous mutation; (2) measuring the impact of induced mutation on quantitative traits; and (3) modeling the long-term dynamics of a partially selfing population under influence of both mutation and selection.;To measure the rate and study the molecular nature of spontaneous mutation, the progeny of flies from a wild population of Drosophila melanogaster were screened for complete loss-of-function mutants at eight eye-color loci. In total, 28 mutants were found in 16 independent events among 887,000 individuals screened. This gives a per locus mutation rate of (2∼15) x 10-6. About 87% of the loss-of-function mutations in the eight eye-color genes in D. melanogaster were due to insertions and deletions. This suggests that D. melanogaster mutations may have a very different molecular nature from mutations in humans, since majority of human mutations are due to substitutions. Based on the molecular nature of mutations found, the per nucleotide mutation rate in D. melanogaster was estimated to be (1.2∼10.0) x 10 -9.;To study the impact of mutation on quantitative traits, groups of outbred flies with different fractions of their genome descended from males treated with EMS mutagenesis were compared in terms of nine quantitative traits. For four life-history related traits-developmental time, viability, fecundity, and longevity, EMS mutagenesis had significant impacts. In contrast, no significant differences were detected in the other five traits, which were all morphological traits. One generation of spontaneous mutation increased the developmental time by 0.09% at 20°C and by 0.04% at 25°C, and reduced viability under harsh conditions, fecundity, and longevity by 1.35%, 0.21%, and 0.08%, respectively.;Finally, to study the effect of mutation on the population dynamics of a population that is under selection, two multilocus models with populations having self-fertilization occurring at either the haploid or diploid stage were investigated. Surprisingly, cyclic dynamics of the population fitness occurred under conditions when genomic deleterious mutation rate, U, is large (U > 2.2 under haploid selfing and U > 10 under diploid selfing), there was an intermediate rate of selfing, mutations were highly recessive, and there was strong selection against homozygous mutations. An investigation of the mechanisms causing these cycles revealed that the cycles can be explained as the result of fluctuations in the efficiency of purging deleterious mutations through selfing over generations due to the interplay between selective interference and identity disequilibrium among homozygous loci.