| Beneficial mutations are the fuel for adaptation but remain poorly studied. Extreme value theory has been used extensively as a model for understanding the distribution of fitness effects (DFE) among beneficial mutations prior to selection, however there is little empirical data available to test the most compelling predictions of the theory. This thesis aims to test the prediction that the DFE among beneficial mutations is a negative exponential, with many mutations of small effect and few of large effect. To do this, I collected a wide range of Pseudomonas aeruginosa mutants resistant to the antibiotic ciprofloxacin and measured their fitness in the absence of the drug. I reject the exponential as an adequate descriptor of the data and show that the DFE among beneficial mutations is better described as being located in the Weibull domain of attraction, which is characterised by having a right-hand bound on the magnitude of the largest fitness effects. I also sequenced a number of genes known to be targets of fluoroquinolone resistance to shed light on the range and number of genetic targets involved in resistance. I found a number of mutations, and often multiple mutations, in known genetic targets. Evidence suggests that mutations in regions of the genome not sequenced are also important in determining fitness under permissive conditions. Taken together, these results provide valuable insight into one of the most fundamental problems in evolution, the nature and fitness effects of beneficial mutations. |
