Coral reefs are highly complex and also highly threatened ecosystems. Population growth and the unsustainable use of coral reefs have resulted in 55% of the world’s reefs being considered degraded. Fishing, the primary ‘local’ threat on most reefs, has altered the composition of most reef communities. As a result, very few pristine coral reefs remain. Typically, coral reef research is done via underwater visual censuses, providing abundance estimates but no indication of trophic interactions, therefore we know relatively little about the structure of intact reef food webs. Understanding how human activities affect trophic structure and feeding interactions among resident reef species may be important for coral reef conservation. Here, I apply stable isotope analysis to coral reef piscivorous fishes from Kiritimati (Republic of Kiribati), the world’s largest atoll. I examine dietary niche metrics of five focal species (Cephalopholis argus, Cephalopholis urodeta, Aphareus furca, Lutjanus bohar, and Lutjanus fulvus) and of the piscivore functional group as a whole, across an anthropogenic disturbance gradient that results from the atoll’s heavily skewed geographic population distribution. Using bootstrapped stable carbon (δ13C) and nitrogen (δ15N) isotope values, controlled for body size effects and analysed with Bayesian methods using the SIAR (Stable Isotope Analysis in R) program, I provide evidence of isotopic niche differentiation in C. argus and L. fulvus relative to other sampled species in terms of niche width metrics and mean δ13C and δ15N values. I also analyse the effect of fishing pressure at an individual level (controlling for body size effects on stable isotope signatures for each species), population level (accounting for observed differences in body size distributions across the fishing pressure gradient for each species), and the ‘community’ level (accounting for body size and relative abundance differences of the five piscivores across the fishing pressure gradient). These metrics reveal species-specific changes in niche metrics of three of the focal species at the individual level: C. urodeta, showed regionally distinct niche width metrics but no apparent correlation with fishing pressure, while A. furca and L. bohar, both had broader niche width metrics in heavily fished areas. No significant effect of fishing pressure was found at population or community levels. This study provides the first evidence using stable isotopes that fishing can alter the diets of coral reef fishes. The mechanism by which it can do so, while not entirely clear, would most likely be by expanding a given species’ dietary diversity by either forcing it to switch to non-preferred prey items or changing the diet and/or body size of its prey items, both of which would reflect significant ecological changes within a community. This thesis provides evidence of the utility of stable isotope analyses in answering important ecological questions in coral reef food webs, and reveals that fishing can affect reef communities at the most fundamental level of trophic interactions. / Graduate / 0329 / burrilladrian@gmail.com