| Scleractinian corals are the physical and energetic foundations of coral reef communities. Coral reef community dynamics are functions of complex interactions between corals, their mutualistic partners, and their competitors. These dynamics are difficult to parse using traditional approaches, and new computational and theoretical tools could shed light on ecological and evolutionary patterns. In this dissertation, I explored the ecology of coral reef communities using network theory, large-scale simulations, and machine learning. In chapter one, I used network theory to make the first quantitative assessment of coral- Symbiodinium community structure. In doing so, I found that symbiont transmission mode (horizontal or vertical) was correlated with symbiotic specificity and other community patterns. In chapter two, I applied network theory to "shallow" and "deep" coral-Symbiodinium datasets to demonstrate that the detection of community patterns was dependent on sampling coverage. Here, dominant Symbiodinium were not sufficient to detect correlations between transmission mode, symbiotic specificity, and species strengths. Subdominant Symbiodinium drove the correlation between life history traits and community patterns, suggesting that subdominant symbionts are functionally important to their hosts. In chapter three, I used a simulation approach to identify biological and ecological characteristics that are likely to determine how coral-Symbiodinium communities respond to environmental changes. My results suggest four factors that may aid coral persistence, including more robust generalist species. Finally, in chapter four, I combined simulations and machine learning to identify how coral persistence depends on both life-history traits and environmental regime. Severe bleaching events had more detrimental effects on corals than frequent bleaching events, and corals responded differentially to the two depending on their life-history traits. Together, this dissertation shed light on new ecological patterns and demonstrated that quantitative approaches have much to offer coral reef research. |
