Abstract:
Scarine parrotfishes have been identified as key herbivores in coral reef health and recovery, but there is confusion over what governs parrotfish-coral dynamics. The prevailing view of scarines as ecosystem engineers is based on the removal of turf, macroalgae and sediment. However, recent work synthesising behavioural observations, trophic anatomy, digestive physiology, stable isotope analysis and fatty acid analysis suggests that most scarine parrotfishes are 'microphages' that target microscopic photo-autotrophs, especially cyanobacteria. An approach that resolves microscopic prey items is therefore required to investigate the 'microphage hypothesis', clarifying the functional role parrotfish play in coral reef ecosystems. This thesis aims to test the dietary targets of parrotfishes by posing two hypotheses. The first is a direct test of the microphage hypothesis: do parrotfishes target microscopic photo-autotrophs, and what is the composition of the microscopic prey community? Second, does diet composition vary inter-specifically? To identify and quantify the dietary targets of parrotfishes five species of parrotfishes (four scraping Scarus species and the excavator Chlorurus microrhinos) were followed at Lizard Island on the northern Great Barrier Reef until focused feeding was observed. Feeding sites were photographed and a bitecore (22mm diameter) was extracted and fixed for laboratory analysis. Bitecores were photographed to 1000x magnification to identify and quantify the epilithic biota, including rhodophytes, phaeophyceae, chlorophytes, cyanobacteria and diatoms. To identify and quantify euendoliths, bitecores were vacuum-embedded with epoxy-resin then decalcified to expose the euendolith microborings. Filamentous cyanobacteria and pennate diatoms were present on all 100 bitecores, supporting the microphage feeding hypothesis. Significant quantitative and qualitative differences in micro-phototroph assemblages demonstrated trophic resource partitioning among the five parrotfish species. S. spinus targeted cyanobacterial biofilm on live crustose coralline algae, whilst S. dimidiatus targeted cyanobacterial tufts on very recently bleached dead coral. S. frenatus and S. rivulatus had a trophic overlap, targeting a significantly different and more species-rich micro-phototroph assemblage to S. spinus and S. dimidiatus. The euendolith results revealed that C. microrhinos targeted highly bioeroded substrate where the siphonous microchlorophyte Ostreobium was the dominant microborer. The average maximum turf height for the 100 BiteCores was 2.6mm confirming that parrotfish target and maintain turfs in early successional states.