Gene Transfer Agents (GTAs) are phage-like pthesiss that are produced by many alpha proteobacteria in late stationary growth phase and are capable of transferring chromosomal genes (termed "constitutive transduction"). Examination of alpha proteobacterial genomic sequences indicated widespread occurrence of GTA-like elements. The goal of this study was to investigate gene transfer potential of GTAs of marine alpha proteobacteria in culture as well as in natural marine environments. Another goal was to determine the potential of bacterial symbionts from zooxanthellae and coral to genetically transfer beneficial properties between symbionts. Ruegeria mobilis (ID 45A6) was isolated from cultures of the coral endosymbiotic dinoflagellate, Symbiodinium spp. A goal of the research was to determine if GTAs from this isolate have the capability of transferring genes to environmental recipients and have an impact on settlement of coral larvae. Little is known about coral settlement cues, yet there may be contributions from the extensive symbiotic relationship of coral reef-associated bacteria. Several gene transfer experiments in different environments were performed using transformed isolates of Ruegeria mobilis containing a transposon marker gene. Experiments were also performed using GTAs from the Ruegeria mobilis isolate to observe any impact GTAs have on coral larval settlement, using larvae from the brooding coral, Porites astreoides, and from the reef building coral, Montastraea faveolata. Gene transfer frequencies from statistically significant gene transfer experiments resulted in an average of 2.92 × 10-1 (transfer recipients to total viable population). Coral settlement experiments resulted in a statistically significant increase in larval settlement with the addition of GTAs for 80% of the executed experiments. The entire study has demonstrated that GTA-mediated gene exchange is much higher than any other mode of horizontal gene transfer and it has been established that these genes can be exchanged between bacterial taxa. GTAs can also have an impact on coral larval settlement mechanisms that are not yet completely understood. GTA-mediated beneficial gene exchange may be an important driver in adaptation to an evolving planet.

Coral recruitment is a key process that contributes to the community structure and resilience of coral reefs. As such, quantification of this process is important to assist with the management of these threatened ecosystems. While coral recruitment has been the focus of numerous studies over the past 30 years, an understanding of this process on the high-latitude reefs of South Africa is limited. In addition, variations in methods used in recruitment studies make the results difficult to compare. A rapid in-situ method for universal application in the detection of early post-settled recruits would thus be useful. In this study, scleractinian coral recruitment was investigated at three study sites on Two-mile Reef, over two six-month sampling periods, covering summer and winter. Two components were investigated by attaching settlement tiles consisting of ceramic and marble tiles, and ceramic tiles conditioned with crustose coralline algae (CCA) onto the reef in a spatially structured experimental design. Firstly, coral recruitment was compared on the three different tile surfaces and fluorescence photography was investigated as a rapid in situ technique to detect early post-settled recruits. Fluorescence photography was then used to compare recruitment on tiles with the surrounding natural substrata. Secondly, the spatial and temporal variation in the abundance, composition and size of recruits was investigated. Additionally, the percentage cover of biota surrounding each recruit within three millimeters of its corallum was visually estimated to quantify the microhabitat surroundings of coral recruits. Overall recruitment on the three tile types differed, yet spatial variation in coral recruitment, regardless of tile surface, accounted for most of the variance in recruitment. While the highest recruitment occurred on CCA tiles, this was not significantly greater than ceramic tiles, indicating that the conditioning of ceramic tiles with Mesophyllum funafutiense CCA did not enhance coral settlement in this study. Although many recruits were not detected with fluorescent photography (73%), it proved useful to reveal recruits as small as 0.75 mm in corallum diameter, and indicated that recruitment on the tiles and natural substratum differ significantly. Spatially, the abundance and composition of coral recruits differed between study sites, within sites, and predominantly occurred on tile edges. Coral recruitment was lowest at shallower sites, and was dominated by pocilloporids regardless of study site. Additionally, the abundance and composition of recruits differed between the two sampling periods, with a 6.6-fold decrease in the mean abundance of recruits from summer to winter, with only pocilloporid settlement occurring in the latter season. The majority of recruits were <3 mm, and their microhabitat was dominated by bare substrata and crustose coralline algae. The results suggest that, while the choice of artificial settlement surface used in such studies can have a profound influence on the results, spatial variation in recruitment can be greater. The recovery of scleractinian coral taxa on Two-mile Reef in the event of a severe disturbance is expected to differ, with greatest recovery in areas of high levels of recruitment. The microhabitat surrounding recruits is described here for the first time, suggesting that further research into coral-crustose coralline algae interactions is warranted. Finally, while fluorescence photography has its limitations, it shows promise as a useful tool for rapid qualitative, but not quantitative, assessment of recruitment. / Thesis (M.Sc.)-University of KwaZulu-Natal, Westville, 2011.