Coral reefs are among the most diverse and productive ecosystems on earth, and are critical to the survival of tropical marine ecosystems and sustenance of local human populations. However, coral reefs are quite vulnerable to disturbances, both natural and anthropogenic. This thesis looks at how coral reef communities have responded to climactic disturbances, particularly the 1997-98 El Niño Southern Oscillation (ENSO) and subsequent coral bleaching and mortality that affected much of the Indian Ocean, including the coastal waters of Tanzania, where the study was conducted. In particular, it investigates the effects of coral bleaching, habitat degradation and reef spatial arrangement on reef fish assemblages. Habitat structural complexity and spatial arrangement of reefs had an effect on reef fish communities. Fish communities showed patterns in distribution among habitats and between patch and continuous reefs. Fishes preferred live to bleached/dead or eroded coral, but trophic groups reacted differently to patch and continuous reefs. There were slight changes in fish abundance and significant changes in fish diversity on experimental, bleached branching Acropora coral plots over a period of one year. While fish abundance on one site increased shortly after a bleaching event, 6 years later fish abundance had decreased significantly. Conversely, coral reef communities in northern Tanzania had changed little over an 8-year period, with minor changes associated with the 1997-98 ENSO and the presence or absence of fisheries management. The coral reefs in the region were found to show high variability in community structure and responses of associated fish and invertebrate communities. The findings of this thesis indicate the importance of habitat structure and spatial arrangement of reefs, the detrimental effects of coral bleaching, and the possibility that some reefs and some (generalist) reef fish taxa may exhibit resilience to climate change. / At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 3: Manuscript. Paper 4: Manuscript.

Image analysis was used to measure basal areas of selected colonies of Montastraea annularis and Porites astreoides, following the colonies over a three-year period from 2002 to 2004. Existing digital images of permanently-marked quadrats in the Caye Caulker Marine Reserve, Belize, were selected based on image quality and availability of images of selected quadrats for all three years. Annual growth rates were calculated from the basal-area measurements. Mean growth rates (radial skeletal extension) for M. annularis and P. astreoides were 0.02 cm yr-1 and -0.20 cm yr-1, respectively. Basal area measurements demonstrated a large degree of variability. Increases were approximately balanced by declines giving the impression of stasis. By removing negative values and correcting by 25% to allow for comparison with vertical growth rates, mean values increased to ~0.5 cm yr-1 for M. annularis and ~0.8 cm yr-1 for P. astreoides. Basal area as a growth measure was compared to methods used in earlier studies. A new growth index based on basal area and perimeter was proposed and modeled. This growth index can be useful for reporting growth measured from basal areas and comparable other methods. The index also measures negative growth, or mortality, which conventional methods cannot do.