Fig 1.
Sampling location and example Porites spp. size classes.
(a) Map of Orpheus and Pelorus Islands, Australia. These islands are located on the inner Great Barrier Reef, and their relative location to Australia is marked on the map inset. (b, c) Examples of colonies we considered small Porites spp., less than 30 cm in diameter. (d, e) Individuals we considered large Porites spp., greater than 2 m in diameter. Individual colonies were not measured due to the large, visible disparity between what we considered size classes. Orpheus and Pelorus Island map data edited from © Great Barrier Reef Marine Park Authority 2014 [25].
Table 1.
Site structures Porites spp. host genetics. PERMANOVA results for the model, Genetic Distance ~ Site + Size Class + Site*Size Class, with the marginal importance of each variable given.
Fig 2.
Host genetics are structured by site.
(a) Proportion of individuals from each admixture group at each site. (b) Admixture plot representing all colonies sequenced, with each vertical bar representing one sample. Bars are ordered by maximum group membership. Orpheus and Pelorus Island map data edited from © Great Barrier Reef Marine Park Authority 2014 [25].
Fig 3.
A large degree of genetic structure exists in Porites spp. species complex.
(a) PCoA of genetic distances colored by assigned admixture group. (Inset) Unrooted hierarchical tree of pairwise FST distances between admixture groups. FST values are reported in S1 Fig.
Fig 4.
C15 ASVs dominate Symbiodiniaceae community.
Heatmap of ASV group abundance in each coral colony, ordered by the genetic subcluster each colony was assigned to. Colors along the top of the heatmap indicate which admixture group each colony was assigned to and correspond to the colors in Fig 3. We were not able to recover host genetic data for some colonies which had successful ITS2 amplification. These colonies are marked “NA” and are annotated white. Each row is a different ASV group, labeled by their assigned taxonomy. Most ASV groups belonged to C15 (a Cladocopium spp. subtype), though we recovered one ASV group which was assigned to D4 (a Durusdinium spp. subtype). Groups labeled “NA” did not have a consensus taxonomy call. Fill colors within the heatmap are log standardized values, log10(Abundance + 1), for the abundance of each ASV group.
Fig 5.
Porites microbial community composition across genetic subclusters.
Heatmap of ASV abundance in each coral colony for the top 30 most prevalent ASVs, ordered by the genetic subcluster each colony was assigned to. Colors along the top of the heatmap indicate which admixture group (top) and genetic subcluster (bottom) each colony was assigned to and correspond to the colors in Fig 3. We were not able to recover host genetic data for some colonies which had successful 16S amplification. These colonies are marked “NA” in the annotation colors. Each row is a different ASV, labeled by its assigned genus- and species-level taxonomy. Groups labeled “NA” or “Unk” did not have a consensus taxonomy call but still represent putatively separate species. Fill colors within the heatmap are log standardized values, log10(Abundance + 1), for the abundance of each ASV.
Fig 6.
Host genetics is the primary driver of Symbiodiniaceae and microbial community composition.
(a) PCoA of Symbiodiniaceae community structure, based on Bray-Curtis distances, colored by host colony admixture group. (b) PcoA of microbial community structure, based on Bray-Curtis distances, colored by host colony admixture group. (c) RDA-forest model results giving weighted relative importance for Symbiodiniaceae (gold) and microbial (green) communities. In brief, weighted relative importance represents the proportion of variance explained by a given variable out of the total variance explained by the model. For both microbes and Symbiodiniaceae, genetic subcluster is the most important predictor of community structure. We find qualitatively similar results through PERMANOVA analysis (Tables 2 and 3).
Table 2.
PERMANOVA results for drivers of Symbiodiniaceae community structure support admixture group as most important predictor. Host genetics have the strongest (and only significant) effect on symbiont community, followed by sampling site, then size class.
Table 3.
PERMANOVA results for drivers of microbial community structure show a significant role of admixture group, site, and size class. Host genetics have the strongest effect on the microbial community, followed by sampling site, then size class.