One of the primary questions in organismal biology is how evolution has acted to shape the species that we see in nature. Beginning to address this incredibly complex question requires a diverse set of approaches that can be difficult to accomplish in the wild, in part because it requires a relatively robust knowledge of the evolutionary history of given populations. Though it cannot tell us how existing species have evolved, experimental evolution is a powerful tool because it allows one to... Show moreOne of the primary questions in organismal biology is how evolution has acted to shape the species that we see in nature. Beginning to address this incredibly complex question requires a diverse set of approaches that can be difficult to accomplish in the wild, in part because it requires a relatively robust knowledge of the evolutionary history of given populations. Though it cannot tell us how existing species have evolved, experimental evolution is a powerful tool because it allows one to track phenotypic and genotypic changes in populations over time in response to a controlled selection pressure. By imposing a particular selection pressure on populations with a known origin, I can test hypotheses about organismal evolution generated by studying patterns in nature. Here I will discuss a series of experiments conducted on populations of Drosophila melanogaster that have been evolved under predatory selection by nymphs of the Chinese mantis (Tenodera aridifolia sinensis). I first investigated the ability to use phenotypic selection analysis to determine long term evolutionary outcomes. To do this, I measured selection acting on wing size and shape in the base population and then again in the evolved populations after 30 generations of selection, and used this to determine other important morphological and behavioral traits that have likely been targets of selection. I show that evolutionary trajectories are largely predictable, but that unmeasured traits can have profound effects on evolutionary outcomes. I also test the predictions of the risk allocation hypothesis as it pertains to courtship, aggression, and anti-predator behavior. Unlike many previous studies that have focused on learned responses to predation risk, I tested whether populations evolved under differences in variation in predation risk would evolve behavioral patterns consistent with the risk allocation hypothesis. I found that while the hypothesis captured several important aspects of the evolutionary response, the specific predictions failed to accurately describe the actual outcome. my results suggested that the riskiness of different behavioral types played a large role in determining whether they conformed to the predictions of risk allocation. In my final chapter, I investigate a unique behavior in which flies evolved in the presence of predators reduce their propensity to initiate flight. Though my results cannot conclusively determine the cause of the evolution of this new escape strategy, they do suggest that associations with allometric scaling relationships are important in determining the fitness of the divergent strategies observed in the predator-evolved populations. Show less