Corallivory and the microbial debacle in two branching scleractinians

Abstract

The grazing activity by specific marine organisms represents a growing threat to the survival of many scleractinian species. For example, the recent proliferation of the corallivorous gastropod Drupella now constitutes a critical case in all South-East Asian waters. If the damaging effects caused by this marine snail on coral polyps are relatively well known, the indirect incidence of predation on coral microbial associates is still obscure and might also potentially impair coral health. In this study, we compared the main ecological traits of coral-associated bacterial and viral communities living in the mucus layer of Acropora formosa and Acropora millepora, of healthy and predated individuals (i.e., colonized by Drupella rugosa), in the Bay of Van Phong (Vietnam). Our results show a substantial impact of the gastropod on a variety of microbiological markers. Colonized corals harbored much more abundant and active epibiotic bacteria whose community composition shifted toward more pathogenic taxa (belonging to the Vibrionales, Clostridiales, Campylobacterales, and Alteromonadales orders), together with their specific phages. Viral epibionts were also greatly influenced by Drupella corallivory with spectacular modifications in their concentrations, life strategies, genotype richness, and diversity. Novel and abundant circular Rep-encoding ssDNA viruses (CRESS-DNA viruses) were detected and characterized in grazed corals and we propose that their occurrence may serve as indicator of the coral health status. Finally, our results reveal that corallivory can cause severe dysbiosis by altering virus-bacteria interactions in the mucus layer, and ultimately favoring the development of local opportunistic infections.

Fig. 1

Location of the sampling site in the Van Phong Bay (Vietnam)

Fig. 2

The scleractinian corals Acropora formosa a, b and Acropora millepora c, d colonized by the marine gastropod Drupella rugosa in the Van Phong Bay, Vietnam

Fig. 3

Comparison of different microbiological traits between grazed, healthy (ungrazed) corals, and surrounding water in the Bay of Van Phong (Vietnam), * significant differences, Kruskal–Wallis, p < 0.05. Viral abundance a, bacterial abundance b, proportion of active CTC+ cells c, viral lytic production d and proportion of lysogenic bacteria e

Fig. 4

Heatmap of bacterial 16S rRNA gene OTU 97% obtained from the mucus of healthy (_H) (Acropora millepora, MIL; Acropora formosa, FOR; and Fungia sp., FUN) and grazed (_G) corals (Acropora millepora and Acropora formosa), from Drupella’s gut (GUT) and from the surrounding water column (WAT). Independent biological replicates of each type of samples are labeled sequentially. Dendrogram distances are based upon OTU relative abundances within the data matrix. Only the 100 most abundant OTUs are shown in the figure. The left and top dendrogram are based on Bray–Curtis distances and Ward hierarchical clustering method. Colors correspond to normalized relative abundances of a bacterial OTU (column) in a sample (row), from blue (low abundances) to red (high). Pie charts represent the community composition at the order level averaged for all the healthy and grazed corals, GUT and WAT. Differentially abundant OTUs detected in healthy or grazed Acropora corals are indicated in green and orange, respectively

Fig. 5

Bacterial OTUs showing significant differences in abundance between healthy (in green) and grazed (in orange) corals. The bar plots on the left side display the mean proportion of OTU belonging to each order. The dot plots display the difference in mean proportions between the healthy and grazed corals with the associated p-value of the t-test statistic (Bonferroni corrected). Error bars on both the bar and dot plots represent the 95% confidence intervals. Taxonomy was assigned to OTU representatives using Mothur with the Silva 123 database

Fig. 6

a Relative coverage of major viral families in annotated reads by sample (≥1% in one sample). b Relative abundance of major viral genera—or the deepest taxonomic level available—by sample

Fig. 7

Heatmap of viral genotypes relative coverage by sample, with the hierarchical clustering resulting in the grouping of all healthy corals (FUN, FOR_H, MIL_H), with the water sample (WAT), and both grazed corals (FOR_G and MIL_G) with GUT (clade indicated by the snail icon)

Fig. 8

Genomic maps of the four CRESS-DNA viruses detected in the grazed corals