A devastating novel coral disease outbreak, referred to as Stony Coral Tissue Loss Disease (SCTLD), was first described in 2014. It is thought to have originated offshore of Miami-Dade County, FL, but has persisted and spread, affecting new reefs along the Florida Reef Tract and reefs of at least 8 other Caribbean jurisdictions. We investigated the microbial communities of clinically normal and diseased specimens of five species of affected corals using targeted 16S ribosomal DNA sequencing (Illumina MiSeq). Fifty-nine bacterial sequences were identified using contrast analysis that had enriched abundance in diseased coral host microbiomes relative to the microbiomes of clinically normal hosts. Several sequences from known bacterial pathogens were identified in this group. Additionally, we identified fifty-three bacterial species that had differentially elevated numbers in clinically normal coral host samples relative to samples from diseased host corals. The bacterial consortia composing the clinically normal and diseased coral microbiomes were clearly distinct taxonomically. Predicted functional profiles based on taxonomy, however, were found to be quite similar. This indicates a high level of functional redundancy among diseased and clinically normal microbiome members. Further examination of the direct sequencing data revealed that while some bacteria were differentially distributed according to disease status, others were not. Fifty-one bacterial species were found in both diseased and clinically normal coral host samples and not differentially abundant in either disease state. These still may be important in explaining the presentation of disease. IMPORTANCE: Determining causation is a management top priority to guide control and intervention strategies for the SCTLD outbreak. Towards this goal we examined bacterial taxa that were differentially elevated in numbers in diseased corals as compared to clinically normal corals at Looe Key, FL in August 2018. Many of the bacterial species we detected are known to be pathogenic to humans, animals, and (or) plants, and some of these have been found associated with diseased corals in other studies. Microbes that were present (or conspicuous by their absence) in both diseased as well as clinically normal corals were also examined because “healthy” corals from a diseased location such as Looe Key may have been exposed but may not have been showing overt disease at the time of sampling. Although untangling of causation is not possible currently, certain bacterial cliques and excess nutrients appear to be potential risk factors in SCTLD pathology.

The field of metabolomics generally lacks standardized methods for the preparation of samples prior to analysis. This is especially true for metabolomics of reef-building corals, where the handful of studies that were published employ a range of sample preparation protocols. The utilization of metabolomics may prove essential in understanding coral biology in the face of increasing environmental threats, and an optimized method for preparing coral samples for metabolomics analysis would aid this cause. The current study evaluates three important steps during sample processing of stony corals: (i) metabolite extraction, (ii) metabolism preservation, and (iii) subsampling. Results indicate that a modified Bligh and Dyer extraction is more reproducible across multiple coral species compared to methyl tert-butyl ether and methanol extractions, while a methanol extraction is superior for feature detection. Additionally, few differences were detected between spectra from frozen or lyophilized coral samples. Finally, extraction of entire coral nubbins increased feature detection, but decreased throughput and was more susceptible to subsampling error compared to a novel tissue powder subsampling method. Overall, we recommend the use of a modified Bligh and Dyer extraction, lyophilized samples, and the analysis of brushed tissue powder for the preparation of reef-building coral samples for ¹H NMR metabolomics.