Genomic signatures suggesting adaptation to ocean acidification in a coral holobiont from volcanic CO2 seeps

Abstract

Ocean acidification, caused by anthropogenic CO₂ emissions, is predicted to have major consequences for reef-building corals, jeopardizing the scaffolding of the most biodiverse marine habitats. However, whether corals can adapt to ocean acidification and how remains unclear. We addressed these questions by re-examining transcriptome and genome data of Acropora millepora coral holobionts from volcanic CO₂ seeps with end-of-century pH levels. We show that adaptation to ocean acidification is a wholistic process involving the three main compartments of the coral holobiont. We identified 441 coral host candidate adaptive genes involved in calcification, response to acidification, and symbiosis; population genetic differentiation in dinoflagellate photosymbionts; and consistent transcriptional microbiome activity despite microbial community shifts. Coral holobionts from natural analogues to future ocean conditions harbor beneficial genetic variants with far-reaching rapid adaptation potential. In the face of climate change, these populations require immediate conservation strategies as they could become key to coral reef survival.

Fig. 1. Schematic workflow of Kenkel and coauthors and our study with a map of Normandby and Dobu Islands showing site locations, pH value and number of colonies (N) collected.

Inset map shows the location of Normandby and Dobu Islands within the Papua Island region (white square). Sampling sites are colored as follows: teal for Dobu Reef, brown for Upa-Upasina Reef; darker colors for control sites, lighter colors for seep sites. Kenkel and coauthors’ schematic workflow is presented at the top of the figure, shaded in gray. The schematic workflow of this study is below with the analyses conducted for the three holobiont compartment detailed: photosymbionts on the left, coral host in the center and microbiome on the right.

Fig. 2. Genome scans and candidate adaptive SNPs to low pH environment in the coral host.

For the Manhattan plots (a–c) dark gray and light gray circles represent SNPs in odd and even chromosomes, respectively, loci under selection or associated with pH are represented by circles with strokes colored by chromosome. a Redundancy Analysis (RDA). b BayPass run in basic mode (BayPass XtX). c BayPass run in covariate mode (BayPass Env). d BayeScan results showing SNPs under diversifying selection in red circles. e Venn diagram representing overlapping results among the four analyses. The 625 SNPs identified by at least two of the four analyses are represented in bold.

Fig. 3. Population genomic structure of the coral host.

a DAPC results using the dataset of 625 candidate adaptive SNPs. b DAPC results using the dataset of 11,169 independent neutral SNPs. Both DAPC plots are colored following Fig. 1. c STRUCTURE results with K = 2 for the candidate adaptive SNP dataset (see delta K plot in Supplementary Fig. 2b). d STRUCTURE results with K = 2 for the independent neutral SNP dataset (see delta K plot in Supplementary Fig. 2a).

Fig. 4. Top 20 enriched GO terms for each category, grouped by REVIGO representative (see REVIGO treemaps in Supplementary Fig. 3).

a Biological processes, b cellular components, and c molecular functions. Enrichment scores (-log10(p-value)) are plotted for each GO term, with circle size representing the ratio between the number of candidate adaptive genes with a particular GO annotation and the number of genes in the A. millepora proteome with that GO annotation. Dotted lines represent p-value = 0.05, dashed lines represent p-value = 0.01. Numerical source is provided in Supplementary Data 3.

Fig. 5. Population genetic structure among Cladocopium goreaui population pools (i.e., coral colonies).

DAPC results are shown on the first and second (a), and the first and third axes (b). c Heatmap plot of the pairwise F_ST matrix among C. goreaui population pools. Warm colors represent higher pairwise F_ST values, indicating high differentiation between population pools. Cold colors represent lower pairwise F_ST values, indicating low differentiation between population pools. Population pools are colored following Fig. 1 and coded in (c) as follows: Dobu Seep, DS; Dobu Control, DC; Upa-Upasina Seep, US; Upa-Upasina Control, UC.

Fig. 6. Microbiome metatranscriptomic characterization results for the taxonomic and functional classifications.

a NMDS plot on the Bray-Curtis dissimilarities among colonies using the taxonomic classification of metatranscriptomic reads. b Percentages of microbial metatranscriptomic reads belonging to each Phylum at each sampling site. c NMDS plot on the Bray-Curtis dissimilarities among colonies using the functional classification of metatranscriptomic reads. Coral colonies are colored following Fig. 1. Numerical source is provided in Supplementary Data 4.