Temporal Sampling of White Band Disease Infected Corals Reveals Complex and Dynamic Bacterial Communities

White band disease (WBD) is a coral disease that is currently decimating populations of the endangered staghorn coral, Acropora cervicornis and elkhorn coral, A. palmata across the Caribbean. Since it was first reported in 1979, WBD has killed 95% of these critical reef-building Caribbean corals. WBD is infectious; it can be transmitted through the water column or by a corallivorous snail. While previous research shows that WBD is likely caused by bacteria, identification of a specific pathogen or pathogens has remained elusive. Much of the difficulty of understanding the etiology of the disease comes from a lack of information about how existing bacterial communities respond to disease and separating initial from secondary colonizers.

In order to address this lack of information, we performed a fully-crossed tank infection experiment. We exposed healthy corals from two different sites to disease and healthy (control) homogenates from both sites, replicating genotype across tanks. We sampled every coral at three time points: before inoculation with the homogenate, after inoculation, and when the coral showed signs of disease. We then performed 16S rRNA gene sequencing on the Illumina HiSeq 2000.

We saw significant differences between time points and disease state. Interestingly, at the first time point (time one) we observed differences between genotypes: every fragment from some genotypes was dominated by Endozoicomonas, while other genotypes were not dominated by one family. At time two we saw an increase in abundance of Alteromonadaceae and Flavobacteriaceae in all corals, and a larger increase in disease-exposed corals. At time three, we saw another increase in Flavobacteriaceae abundance in diseased corals, as well as an introduction of Francisella to diseased corals. While Flavobacteriaceae and Francisella were proposed as potential pathogens, their increase at time three suggests they may be secondary colonizers or opportunists. In genotypes that were dominated by Endozoicomonas at time one, we saw a decrease of Endozoicomonas in diseased corals, indicating that Endozoicomonas may be beneficial symbionts and/or antagonists of the pathogen(s). We also identified a clade of Pasteurella found in the disease dose and diseased corals at times two and three that is likely important in the WBD etiology.