As sea surface temperature increases, many coral species that used to harbor symbionts of the genus Cladocopium have become colonized with the thermally tolerant genus, Durusdinium. Here, we asked how symbionts of one genus react to the presence of another symbiont genus within the same coral host, and what effect this has on the host. We used previously published data from Acropora hyacinthus corals hosting Cladocopium and/or Durusdinium symbionts and looked at gene expression in all three symbiotic partners depending on the relative proportions of symbiont genera within the host. We find that both Cladocopium and Durusdinium change their expression most when their proportions are nearly equal (the state that we call “codominance”): both genera elevate expression of photosynthesis and ribosomal genes, suggesting increase in photosynthesis and growth (i.e. higher productivity). At the same time, the coral host also elevates production of ribosomes suggesting faster cellular growth, and, when heated, shows less pronounced stress response. These results can be explained in two ways. One explanation is that increased competition between symbionts heightens their productivity, which benefits the host, making it more resilient to stress. Alternatively, the symbionts’ elevated productivity might be the consequence of the host being particularly healthy. Under this explanation, rapid growth of the healthy host creates new space, lowering the symbionts’ competition and allowing for codominance. The latter explanation is supported by the fact that codominance is associated with lower symbiont densities. Irrespective of the causation, the presence of mixed symbiont communities could potentially be used as an instant indicator of coral well-being, which would be a useful tool for coral conservation and restoration.

Abstract: Interrogation of chromatin modifications, such as DNA methylation, has potential to improve forecasting and conservation of marine ecosystems. The standard method for assaying DNA methylation (Whole Genome Bisulfite Sequencing), however, is too costly to apply at the scales required for ecological research. Here we evaluate different methods for measuring DNA methylation for ecological epigenetics. We compare Whole Genome Bisulfite Sequencing (WGBS) with Methylated CpG Binding Domain Sequencing (MBD-seq), and a modified version of MethylRAD we term methylation-dependent Restriction site-Associated DNA sequencing (mdRAD). We evaluate these three assays in measuring variation in methylation across the genome, between genotypes, and between polyp types in the reef-building coral Acropora millepora. We find that all three assays measure absolute methylation levels similarly, with tight correlations for methylation of gene bodies (gbM), as well as exons and 1Kb windows. Correlations for differential gbM between genotypes were weaker, but still concurrent across assays. We detected little to no reproducible differences in gbM between polyp types. We conclude that MBD-seq and mdRAD are reliable cost-effective alternatives to WGBS. Moreover, the considerably lower sequencing effort required for mdRAD to produce comparable methylation estimates makes it particularly useful for ecological epigenetics.

Interrogation of chromatin modifications, such as DNA methylation, has potential to improve forecasting and conservation of marine ecosystems. The standard method for assaying DNA methylation (Whole Genome Bisulfite Sequencing), however, is too costly to apply at the scales required for ecological research. Here we evaluate different methods for measuring DNA methylation for ecological epigenetics. We compare Whole Genome Bisulfite Sequencing (WGBS) with Methylated CpG Binding Domain Sequencing (MBD-seq), and a modified version of MethylRAD we term methylation-dependent Restriction site-Associated DNA sequencing (mdRAD). We evaluate these three assays in measuring variation in methylation across the genome, between genotypes, and between polyp types in the reef-building coral Acropora millepora. We find that all three assays measure absolute methylation levels similarly, with tight correlations for methylation of gene bodies (gbM), as well as exons and 1Kb windows. Correlations for differential gbM between genotypes were weaker, but still concurrent across assays. We detected little to no reproducible differences in gbM between polyp types. We conclude that MBD-seq and mdRAD are reliable cost-effective alternatives to WGBS. Moreover, the considerably lower sequencing effort required for mdRAD to produce comparable methylation estimates makes it particularly useful for ecological epigenetics.