Search

This study assessed the bleaching response (BR) of coral colonies within the central reef complex in Sodwana Bay, South Africa. Bleach surveys were conducted at 16 sites on 8 reefs over the period of 2007 to 2013. A total of 12 858 coral colonies from 30 taxa were randomly sampled and colonies were placed into 7 categories of bleaching response. This allowed for the calculation of taxon-specific BR as a weighted percentage of coral cover bleached. Continuous temperature records from a permanent temperature gauge on Two Mile Reef were used to assess thermal stress over this period. The percentage of coral colonies that bleached in 2007, 2008, 2011, 2012 and 2013 were 37.4%, 17.4%, 23.8%, 33.6% and 38.8% respectively. A binomial GLM model framework was used to separate the effects of year, reef and taxon on the bleaching response. Due to inconsistent sampling of sites over time, only data from the seven sites on Two Mile Reef (TMR) and the two sites on Nine Mile Reef (NMR) were included in the model. A total of 6758 coral colonies from the nine most abundantly sampled taxa were used in the assessment of bleaching response for TMR and NMR over the sample period. Taxon was shown to explain most of the variability in the bleaching response of TMR and NMR over time (40.9%). The standardized reef bleaching response of TMR and NMR indicated the same temporal trends with a range of 5% to 28% of live coral surface bleached. Standardized reef-specific BR showed periods of high (2007, 2012 and 2013) and low (2008 and 2011) bleaching response. Low BR in 2008 and 2011 did not correspond to thermal stress (≥27.5°C) suggesting that local upwelling buffered the effects of thermal stress experienced. Standardized taxon-specific bleaching response for TMR and NMR displayed large variability over time and ranged from 2.5% to 45% of live coral surface bleached, with Montipora being the most susceptible and Galaxea and Playgyra being the least. Mean site-specific BR averaged over all years, including all sites, found that Coscinaraea, Montipora, Astreopora and Anomastrea were the most susceptible taxa. This study and other recent studies draw attention to a trend of either episodic or an increasing frequency and intensity of bleaching in southern African reefs.

Includes bibliographical references (p. 209-274)., The biogeography and ecology of benthic shallow subtidal reef communities in the western Indian Ocean is poorly known, particularly in north-eastern South Africa and southern Mozambique. This thesis uses quantitative information to resolve biogeographic patterns, define reef community types, elucidate potential abiotic determinants of community composition, and evaluate whether subsidies of riverine-derived particulate organic matter (POM) support filter-feeder biomass and drive biogeographic patterns. A large-scale biogeographic analysis was conducted using quantitative biomass data derived from 55 shallow subtidal reefs spanning five countries in the western Indian Ocean. Two statistically distinct marine provinces, Tropical Indo-West Pacific and Subtropical Natal, were recognised by differences in community composition and separated by a biogeographic break in the vicinity of Cape Vidal, South Africa. The biogeographic break took the form of a transitional or overlap area corresponding in location to the Delagoa Bioregion, one of three bioregions also revealed by post-hoc analyses. Significant differences in total average biomass and trophic structure were evident among bioregions, with a number of inter-bioregional trends in trophic groups being apparent. In total, 12 reef community types were recognised, based on similarity profile permutation tests. Most reefs in the Subtropical Natal Bioregion were dominated by a community type characterised by a high biomass of the filter-feeding ascidian Pyura stolonifera and various species of articulated coralline algae. In the Delagoa Overlap Bioregion, a comparatively high diversity of community types was defined, many dominated by algal turf, P. stoloniferaand various Alcyonacea and Scleractinia. Further north, P. stolonifera diminished and the contributions of Scleractinia, especially Porites spp., Pocilloporaspp. and Galaxea spp. increased. Many of these community types are not represented within protected area networks, particularly those in southern Mozambique. When the biomass data were correlated with nine abiotic variables, likely determinants of community composition emerged at both inter- and intra-regional scales. Sea surface temperature, significant wave height, chlorophyll-a and suspended inorganic sediment were the variables highly correlated with community composition and therefore most likely to drive biogeographic differences. Within each bioregion, different sets of abiotic variables were found to be important in driving community differences among sites, including turbidity, chlorophyll-a, reef susceptibility to sand inundation, reef heterogeneity and sea surface temperature. Striking differences in the oceanographic conditions of bioregions were evident, especially between Subtropical Natal and Delagoa Overlap bioregions. In particular, the strong influence of wave height emerged as a novel and unexpected correlate at a biogeographic scale. These differences initiated a trophic study conducted in the Subtropical Natal Bioregion, aimed at determining the importance of riverine-derived POM subsidies in supporting the high filter-feeder biomass in this bioregion. Using carbon, nitrogen and sulphur isotopes and a three-source Bayesian mixing model to calculate proportional contributions, I determined that marine-derived POM formed the bulk of the diets of four species of filter-feeders, but the assimilation of riverine-derived POM was nevertheless notable, ranging from 8 to 33 %. I concluded that riverine POM is likely to play an important but secondary role to factors such as increased levels of turbidity and productivity in explaining the high filter-feeder biomass in the Subtropical Natal Bioregion. These findings provide the first evidence of riverine-inshore-pelagic coupling in filterfeeder communities in this bioregion, and throw light on the factors linked to large-scale biogeographic patterns.

[This corrects the article DOI: 10.3897/zookeys.1066.69697.]., (Zoleka N. Filander, Marcelo V. Kitahara, Stephen D. Cairns, Kerry J. Sink, Amanda T. Lombard.)