As we move into the Anthropocene, a variety of local and global stressors are reorganising coral reef communities, that can push them away from coral dominated states, potentially reducing their capacity to provide ecosystem services. Quantifying ecosystem functions is becoming an increasingly popular tool in helping us understand ecosystem resilience. This thesis focuses on one of the primary ecosystem functions on coral reefs: the grazing of algal turf communities by roving herbivorous fishes. In Chapter 1 and Chapter 2, I use a combination of long-term monitoring UVC data and stereo-cameras to estimate grazing and bioerosion rates, and look at a habitat and management features that influence these both temporally and spatially. In Chapter 1 I show that two functions associated with parrotfish grazing respond differently following a large-scale disturbance, with scraping rates increasing initially and bioerosion rates lagging behind as fish require time to attain large body size. In Chapter 2 I show that, other than greater species diversity, there are varying spatial drivers to grazing rates by functional groups. These chapters both show that converting community data into associated grazing functions sheds light on just how complex drivers to ecosystem functions can be.
The second half of the thesis focuses on characterising benthic changes that occur following mass coral bleaching events, and how these can act as bottom-up drivers to grazing rates by affecting the foraging landscape for grazing fishes over time. I use a range of approaches to track the successional stages of algae and microbial communities in the first three months post-bleaching, how these changes relate to turf algal features known to influence grazing rates. In Chapter 3 I use an experimental approach to show how micro-habitat variations; such as coral life histories and morphology; can influence successional stages and associated features. In Chapter 4, I apply similar approaches to corals which suffered natural mortality during a bleaching event, showing similarities between natural and experimental approaches. Finally, I use a range of experiments to show how successional turf algal communities can vary through time in their attractiveness to grazing fishes, and even their ability to sustain them energetically.