The island continent of Madagascar spans nearly 13.5o of latitude in the SW Indian Ocean. Its coastline includes a number of well developed coral reefs, ranging from tropical Nosy Bé (NW Madagascar, 12oS) and Vohemar (Volhmarina, NE Madagascar, 13oS) to subtropical Tuléar (Toliara , SW Madagascar, 23.5oS), as well as temperate coral communities at Fort Dauphin (Tolagnaro, SE Madagascar, 25oS). Given the range of environmental conditions experienced by reef corals at these different sites, Madagascar represents an ideal location to study the distribution of algal symbionts (Symbiodinium spp.) in these coral hosts. To investigate the effect of latitudinal gradients in temperature on Symbiodinium distributions, 220 samples from 27 coral genera in 12 families were collected from these 4 sites in September 2001. To test the stability of these distributions over time, a further 337 samples were collected from the Nosy Bé and Tuléar regions in March 2007 and November 2009. Symbiodinium communities were screened using Denaturing Gradient Gel Electrophoresis (DGGE) to analyze the internal transcribed spacer-2 (ITS-2) region of Symbiodinium ribosomal DNA, with individual symbiont taxa identified by sequencing individual DGGE bands. Significant differences were found in the Symbiodinium cladal composition of reef corals at different sites, with corals at northern sites containing a higher relative frequency of Symbiodinium in clade D (occurring as mixed clade C+D communities) than southern sampling sites. Nominal logistic analysis of the distribution of symbionts found a significant effect of coral taxa and site, but not of sea surface temperature metrics (environmental data obtained from NOAA’s Coral Reef Watch satellite-derived data) in determining the distribution of different symbionts. Rarefaction analysis indicated there were no differences in Symbiodinium richness (at either the clade or the subtype level) between different sites, or between different sampling intervals. Differences existed in the subcladal composition of dominant ITS-2 types found in congeners at different latitudes, with corals in the genus Acropora being dominated by Symbiodinium C3 (specifically subtype C3z) in northern sites, and C1 in southern sites. Symbiont communities changed between 2001 and 2007/2009, with increases in mixed Symbiodinium C+D assemblages occurring at southern sites that had experienced temperature stress during the intervening period. Decreases in mixed Symbiodinium communities occurred at northern sites, which were not as severely affected by thermal stress. It is suggested that the latitudinal gradients in Symbiodinium found in Madagascar, and the environmental controls on community structure described here, provide important insight into how coral species in this understudied area can adapt or acclimatize to changing environmental conditions through shifts in the composition of their symbiont communities. This will help improve our understanding of how projected climate change in the SW Indian Ocean will affect survival trajectories for coral reefs in the region.

Coral reef environments have been degrading steadily over the last few decades. It is generally accepted now that coral reefs are one of the first marine ecosystems to show measurable perturbations driven by global warming. Some of theses perturbations are consequence of extreme stochastic disturbances like hurricanes or extreme thermal anomalies and therefore, can be easily identified using broad indicators like coral cover. These indicators are easy to measure and provide a general description of the system in question. The detection and interpretation of more subtle perturbation in coral communities is more complicated, both logistically and conceptually. However, detecting chronic perturbations at an early stage can increase significantly the success of early mitigating management strategies. This thesis focuses on the development and application of highly sensitive indicators that can detect subtle perturbations in coral communities. It also studies the ecological interpretation of mild perturbations and their effect on the future structure of coral reef environments. The mild thermal disturbance experienced by the Heron Island reef in the south of the Great Barrier Reef during the 2005-2006 summer, and an environmental gradient along the central coast of Venezuela (south Caribbean Sea), were used as models to test the sensitivity of coral reefs to mild disturbances at different organisation levels (organism, population, and community). At the organismal level my results showed that the intensity of bleaching that a colony shows during a mild thermal disturbance is affected by the morphology of the colony as well as the substrates surrounding the colony. Colonies surrounded by sand bleached more than colonies surrounded by dead coral or rubble. I propose that reef specific information on the relative cover of sand, rubble, and dead coral can improve the fine scale accuracy of bleaching predicting models. Studying the same mild thermal disturbance at Heron Island, I showed how demographic parameters of some populations are more sensitive to mild thermal disturbances than coral cover (the most widely used response variable in coral monitoring programs around the world). Furthermore, my results suggest that the response of coral populations to mild thermal disturbance is different in nature and intensity to the respond observed during extreme thermal disturbances. Some taxa like Stylophora pistillata, are highly sensitive to extreme thermal disturbances, and tend to be the first to die under these conditions. In contrast, this taxon was extremely tolerant to the mild thermal disturbance of early 2006 in comparison to other taxa identified previously as more tolerant than S. pistillata. This difference in the response of coral environments to milder more common disturbances can have great influence in the trajectory that coral reefs follow in a rapidly changing world. Finally at a community level, I demonstrated how the power to detect the effect of environmental conditions on coral environments is drastically increased when the coral host-endosymbiotic algae relationship is interpreted holistically. The percentage of the variability of the community structure that can be explained by environmental variables in the central coast of Venezuela is doubled when the unit of selection considered is the holobiont instead of the coral host or the endosymbiont independently. This approach can be crucial for the detection of subtle changes in coral communities as the frequency and intensity of disturbances increase rapidly. The increase in detection power provided by the different approaches developed in this project can both improve our understanding of the structuring role of mild disturbances in coral communities as well as help identify the effect of disturbances in an early stage before the perturbation reaches catastrophic proportions. This early identification of perturbations can be used for the development of adaptive management strategies that can increase the resistance and resilience of coral reefs in the future.

No description available.