Search for: All records

ABSTRACT Coral reefs have undergone extensive coral loss and shifts in community composition worldwide. Despite this, some coral species appear naturally more resistant, such asMadracis mirabilis(hereinMadracis).Madracishas emerged as the dominant hard coral in Curaçao, comprising 26% of coral cover in reefs that declined by 78% between 1973 and 2015. Although life history traits and competitive mechanisms contribute toMadracis’s success, these factors alone may not fully explain it, as other species with similar traits have not shown comparable success. Here, we investigated the potential role of microbial communities in the success ofMadracison Curaçao reefs by leveraging a low-bias bacterial and viral enrichment method for metagenomic sequencing of coral samples, resulting in 77 unique bacterial metagenome-assembled genomes and 2,820 viral genomic sequences. Our analyses showed thatMadracis-associated bacterial and viral communities are 12% and 20% richer than the communities of five sympatric coral species combined. TheMadracis-associated bacterial community was dominated byRuegeriaandSphingomonas, genera that have previously been associated with coral health, defense against pathogens, and bioremediation. These communities also displayed higher functional redundancy, which is often associated with ecological resilience. The viral community exhibited a 50% enrichment of proviruses relative to other corals. These proviruses had the genomic capacity to laterally transfer genes involved in antibiotic resistance, central metabolism, and oxidative stress responses, potentially enhancing the adaptive capacity of theMadracismicrobiome and contributing toMadracis’s success on Curaçao’s reefs. IMPORTANCEUnderstanding why some coral species persist and thrive while most are in fast decline is critical.Madracis mirabilisis increasingly dominant on degraded reefs in Curaçao, yet the role of microbial communities in its success remains underexplored. This study highlights the potential role ofMadracis-associated bacterial and viral communities in supporting coral resilience and competitive success. By identifying key microbial partners and viral genes that may enhance host stress tolerance and defense against pathogens, we broaden the understanding of how the coral holobiont contributes to species persistence under environmental stress. These insights are valuable for predicting key microbial community players in reef interactions and may inform microbiome-based strategies to support coral conservation and restoration. 

Mapping data used in fractional abundance calculations for bacterial and viral communities. Mapping was performed with SMALT v0.7.6. 

Pooled viral genomes and genome fragments from coral and seawater metagenomes collected in Curaçao. Viral contigs were identified by Genomad and Vibrant and binned by VRhyme. Viral bins are N-linked with 1000 Ns between contigs belonging to the same bin. All viral genomes were dereplicated at MIUViG standards (95% ANI over 80% AF) with CheckV's rapid genome clustering based on pairwise ANI. All viruses which had no viral genes and were reported to have a CheckV quality of “Not-determined” were removed from this database. 

Map Figure (R Script) - V1 (Fig1A)Microscopy Box Plots (R Script) - V1 (Fig1C)Interaction Outcome Plots (R Script) - V1 (Fig2C; Fig2D)Bacterial Community Figures (R Script) - V1 (Fig3A; Fig3B, Fig3C)Viral Community Figures (R Script) - V1 (Fig4A; Fig4B, Fig4C; Fig5A; Fig5B)Provirus Genes (R Script) - V1 (Fig5C)Viral Genome Plots (R Script) - V1 (Fig5D)All associated data sheets 

Bacterial Metagenome Assembled Genomes (bMAGs) from stony coral metagenomes. Bins from MaxBin2 v2.2.7 (Wu et al., 2015), MetaBat2 v2.15 (Kang et al., 2019), and CONCOCT v1.1.0 (Alneberg et al., 2014) were consolidated and improved with the metaWRAP v1.2.1 (Uritskiy et al., 2018). Bins were then quality filtered to ≥50% completion and ≤10% contamination with CheckM2 v1.0.2 (Chklovski et al., 2023). This dataset has been dereplicated using fastANI (--similarity-threshold 0.95). 

QC reports of sequence data generated by FastQC/MultiQC (Andrews, 2010).