Environmental stressors such as increased sea surface temperatures are well-known for contributing to coral bleaching; however, the effect of increased temperatures and subsequent bleaching on coral-associated microbial communities is poorly understood. Colonies of the hard coral Acropora millepora were tagged on a reef flat off Magnetic Island (Great Barrier Reef) and surveyed over 2.5 years, which included a severe bleaching event in January/February 2002. Daily average water temperatures exceeded the previous 10-year average by more than 1 1C for extended periods with field-based visual surveys recording all tagged colonies displaying signs of bleaching. During the bleaching period, direct counts of coral zooxanthellae densities decreased by B64%, before recovery to pre-bleaching levels after the thermal stress event. A subset of three tagged coral colonies were sampled through the bleaching event and changes in the microbial community elucidated. Denaturing gradient gel electrophoresis (DGGE) analysis demonstrated conserved bacterial banding profiles between the three coral colonies, confirming previous studies highlighting specific microbial associations. As coral colonies bleached, the microbial community shifted and redundancy analysis (RDA) of DGGE banding patterns revealed a correlation of increasing temperature with the appearance of Vibrio-affiliated sequences. Interestingly, this shift to a Vibrio-dominated community commenced prior to visual signs of bleaching. Clone libraries hybridized with Vibrio-specific oligonucleotide probes confirmed an increase in the fraction of Vibrio-affiliated clones during the bleaching period. Post bleaching, the coral microbial associations again shifted, returning to a profile similar to the fingerprints prior to bleaching. This provided further evidence for corals selecting and shaping their microbial partners. For non-bleached samples, a close association with Spongiobacter-related sequences were revealed by both clone libraries and DGGE profiling. Despite Vibrio species being previously implicated in bleaching of specific coral species, it is unsure if the relative increase in retrieved Vibrio sequences is due to bacterial infection or an opportunistic response to compromised health and changing environmental parameters of the coral host. This study provides the first molecular-based study demonstrating changes in coral-associated bacterial assemblages during a bleaching event on a natural reef system.

Coral bleaching is a major threat to coral reefs worldwide and is predicted to intensify with increasing global temperature. This study represents the first investigation of gene expression in an Indo-Pacific coral species undergoing natural bleaching which involved the loss of algal symbionts. Quantitative real-time polymerase chain reaction experiments were conducted to select and evaluate coral internal control genes (ICGs), and to investigate selected coral genes of interest (GOIs) for changes in gene expression in nine colonies of the scleractinian coral Acropora millepora undergoing bleaching at Magnetic Island, Great Barrier Reef, Australia. Among the six ICGs tested, glyceraldehyde 3-phosphate dehydrogenase and the ribosomal protein genes S7 and L9 exhibited the most constant expression levels between samples from healthy-looking colonies and samples from the same colonies when severely bleached a year later. These ICGs were therefore utilised for normalisation of expression data for seven selected GOIs. Of the seven GOIs, homologues of catalase, C-type lectin and chromoprotein genes were significantly up-regulated as a result of bleaching by factors of 1.81, 1.46 and 1.61 (linear mixed models analysis of variance, P < 0.05), respectively. We present these genes as potential coral bleaching response genes. In contrast, three genes, including one putative ICG, showed highly variable levels of expression between coral colonies. Potential variation in microhabitat, gene function unrelated to the stress response and individualised stress responses may influence such differences between colonies and need to be better understood when designing and interpreting future studies of gene expression in natural coral populations.