Concordance of microbial and visual health indicators of white-band disease in nursery reared Caribbean coral Acropora cervicornis

Abstract

Background: Coral diseases are one of the leading causes of declines in coral populations. In the Caribbean, white band disease (WBD) has led to a substantial loss of Acropora corals. Although the etiologies of this disease have not been well described, characterizing the coral microbiome during the transition from a healthy to diseased state is critical for understanding disease progression. Coral nurseries provide unique opportunities to further understand the microbial changes associated with diseased and healthy corals, because corals are monitored over time. We characterized the microbiomes before and during an outbreak of WBD in Acropora cervicornis reared in an ocean nursery in Little Cayman, CI. We asked (1) do healthy corals show the same microbiome over time (before and during a disease outbreak) and (2) are there disease signatures on both lesioned and apparently healthy tissues on diseased coral colonies?

Methods: Microbial mucus-tissue slurries were collected from healthy coral colonies in 2017 (before the disease) and 2019 (during the disease onset). Diseased colonies were sampled at two separate locations on an individual coral colony: at the interface of Disease and ∼10 cm away on Apparently Healthy coral tissue. We sequenced the V4 region of the 16S rRNA gene to characterize bacterial and archaeal community composition in nursery-reared A. cervicornis. We assessed alpha diversity, beta diversity, and compositional differences to determine differences in microbial assemblages across health states (2019) and healthy corals between years (2017 and 2019).

Results: Microbial communities from healthy A. cervicornis from 2017 (before disease) and 2019 (after disease) did not differ significantly. Additionally, microbial communities from Apparently Healthy samples on an otherwise diseased coral colony were more similar to Healthy colonies than to the diseased portion on the same colony for both alpha diversity and community composition. Microbial communities from Diseased tissues had significantly higher alpha diversity than both Healthy and Apparently Healthy tissues but showed no significant difference in beta-diversity dispersion. Our results show that at the population scale, Healthy and Apparently Healthy coral tissues are distinct from microbial communities associated with Diseased tissues. Furthermore, our results suggest stability in Little Cayman nursery coral microbiomes over time. We show healthy Caymanian nursery corals had a stable microbiome over a two-year period, an important benchmark for evaluating coral health via their microbiome.

Keywords: Acropora cervicornis; Coral disease; Coral nursery; Coral reefs; Core microbes; Pathobiome; White band disease.

Figure 1. Representative sampling of Acropora cervicornis colonies in Little Cayman, CI.

(A) Healthy (dark orange circle) coral mucus/tissue slurry samples were taken haphazardly on a coral colony that showed no visual signs of disease or tissue loss. (B) Disease (gray circle) and Apparently Healthy (light orange circle) sampling sites for coral mucus/tissue slurry from the nursery reared A. cervicornis in 2019. Disease shows a distinct white band of exposed coral skeleton adjacent to live coral tissue. Apparently Healthy samples were chosen on the same coral colony 10 cm away from a disease lesion. (C) A graphic representation of sampling of colonies in a group (samples representing all the treatments from one frame). Each frame consisted of colonies from one genotype. In this study, colonies sampled from the same frame were in the same genotype in all 2019 samples.

Figure 2. Relative abundance of bacterial and archaeal taxa for A. cervicornis mucus samples taken in 2017 and 2019 represented at the order level.

Each bar represents a different coral. Only samples with greater than 800 sequences are included in this plot, and ASVs with relative abundances greater than 0.5%. Samples from 2019 included “Healthy”, “Apparently Healthy”, and “Disease” tissue types. Colors indicate Order designated by the SILVA database. Both Healthy 2017 and 2019 samples and 2019 Apparently Healthy samples appear similar in bacterial composition, mainly due to the high relative abundance of the bacterial order Rickettsiales (in bold). Disease samples appear distinct from all other samples in composition and abundance of taxa. The x-axis of the graph shows each sample’s blocking group and genotype designation (indicated by colors red, green, yellow, black, and blue). Samples from 2017 were collected from different colonies (and groups) than in 2019.

Figure 3. Boxplots of the Shannon and Inverse Simpson diversity of microbial communities in A. cervicornis based on tissue type and year.

Each point represents a single tissue sample. Colors represent tissue types and year. The center bar of each box plot represents median diversity. Whiskers extend to ±1.5 of the inter-quartile range. (A) Shannon diversity of microbial communities was not significantly different across tissue types or years in A. cervicornis. (B) The Inverse Simpson index was not significant across Healthy tissue types for each year, but within differed significantly across tissues treatment types in 2019. Disease samples show higher Inverse Simpson diversity than either Healthy or Apparently Healthy tissue types in 2019. Significant differences (p < 0.05) based on Tukey HSD tests are indicated by different letters. Healthy samples from 2017 and 2019 were not significantly different for either alpha diversity measurement.

Figure 4. Beta diversity dispersion in A. cervicornis samples for 2019 tissue types and Healthy tissues across years.

Beta diversity (dispersion) was measured and plotted for (A) Healthy, Apparently Healthy, and Disease samples in 2019 and (B) Healthy corals across years. Each point represents the distance to a centroid value for each sample within a treatment. Middle bars on each plot represent the median distance to centroid for each sample type. Upper and lower ends of each box plot represent the upper and lower range of the data set, respectively. Microbial variance was not significantly different across (A) tissue treatment types for 2019 or (B) Healthy tissues across years.

Figure 5. Principal Coordinate Analysis (PCoA) of A. cervicornis microbial communities for 2017 and 2019 tissue types.

Principle coordinate analysis based on the Bray Curtis dissimilarity on the relative abundances of ASVs in each sample. Each point represents an individual coral mucus sample. The color of each ellipse represents tissue types and year, and the shape of the point represents genotype. Ellipses were created using the geom_mark_ellipse() function in ggplot. Points that are closer together indicate communities that are more similar. Disease tissues typically clustered together (top left ellipse), indicating disease samples were more similar to one another but different from Healthy or Apparently Healthy corals. Healthy and Apparently Healthy samples clustered near each other. PERMANOVA results suggest disease tissues are different from Healthy tissues in 2019. Between years, PERMANOVA results suggest no differences in Healthy tissues (p-value =0.119) and no difference in genotypes (p-value =0.11).

Figure 6. Selected ASVs that were significantly differentially abundant across Disease and Healthy tissue types for 2019 samples.

Each point represents an ASV that was significantly differentially abundant in either Diseased tissue or Healthy tissues (from a healthy colony) according to the DESeq2 analysis. Point colors represent the different tissue types. There was a total of 214 ASVs that were significantly enriched (Table S2).

Figure 7. Comparative disease table.

Bacterial orders from our study were compared to seven other coral disease studies. The table characterizes four diseases, and bacterial orders that are shared with our study are shown using an asterisk next to the order name.