As ocean warming threatens reefs worldwide, identifying corals with adaptations to higher temperatures is critical for conservation. Genetically distinct but morphologically similar (i.e., cryptic) coral populations can be specialized to extreme habitats and thrive under stressful conditions. These corals often associate with locally beneficial microbiota (Symbiodiniaceae photobionts and bacteria), obscuring the main drivers of thermal tolerance. Here, we leverage a holobiont (massive Porites) with high fidelity for C15 photobionts to investigate adaptive variation across classic (“typical” conditions) and extreme reefs characterized by higher temperatures and light attenuation. We uncovered three cryptic lineages that exhibit limited micro-morphological variation; one lineage dominated classic reefs (L1), one had more even distributions (L2), and a third was restricted to extreme reefs (L3). L1 and L2 were more closely related to populations ~4300 km away, suggesting that some lineages are widespread. All corals harbored Cladocopium C15 photobionts; L1 and L2 shared a photobiont pool that differed in composition between reef types, yet L3 mostly harbored unique photobiont strains not found in the other lineages. Assemblages of bacterial partners differed among reef types in lineage-specific ways, suggesting lineages employ distinct microbiome regulation strategies. Analysis of light harvesting capacity and thermal tolerance revealed adaptive variation underpinning survival in distinct habitats: L1 had the highest light absorption efficiency and lowest thermal tolerance, suggesting it is a classic reef specialist. L3 had the lowest light absorption efficiency and the highest thermal tolerance, showing that it is an extreme reef specialist. L2 had intermediate light absorption efficiency and thermal tolerance, suggesting that is a generalist lineage. These findings reveal diverging holobiont strategies to cope with extreme conditions. Resolving coral lineages is key to understanding variation in thermal tolerance among coral populations, can strengthen our understanding of coral evolution and symbiosis, and support global conservation and restoration efforts.

Materials and Methods

Site selection and sample collection

Three extreme and three paired classic sites were selected in Chelbacheb, Palau. Extreme sites were shallow semi-enclosed lagoons with high water retention, increased light attenuation, and distinct assemblages of massive Porites lineages compared to classic sites (Rivera et al., 2022; van Woesik et al., 2012). Between 2010 and 2018, these sites had naturally elevated water temperatures year-round (up to 2°C) compared to nearby classic sites. Seasonal temperature ranges at extreme sites were ~29-32 °C, compared to ~28-31 °C at classic sites (Rivera et al., 2022). Sites were selected to ensure that, aside from the contrast in terms of temperature between classic and extreme sites, variability was minimal: they were all comparable in terms of depth (1 - 6 m), proximity to land (~10 - 20 m from land), and wave exposure (calm, on the leeward side of the archipelago or in a lagoon).

Water temperatures were measured every 30 minutes using triplicate temperature loggers at each site (3-4 m depth) between November 2021 and May 2022, and light levels were measured with duplicate loggers for 16 days in April 2022 (Onset, Wareham, USA). Temperature and light data from each site were then averaged across replicate loggers. Colonies resembling the gross morphology of Porites lobata Dana, 1846 were tagged at all sites in November 2021 in a transect along the shoreline. All colonies were sampled between 1 – 6 m depth, with the majority between 3 - 4 m. All selected colonies were at least 1 - 5 m apart to reduce the risk of sampling clone mates while maximizing the probability that the colonies were exposed to similar conditions within a site. Targeted colonies were also relatively small in size (30 – 50 cm) to facilitate transportation to aquarium facilities for further analyses and experiments. The total area over which corals were collected was 250 - 500 m² per site. Tissue samples were taken from the center of each colony and immediately stored in ethanol at -20℃ (2x2 cm samples; n = 90 total, 15/site). An additional 22 colonies were sampled in April 2022 (n = 2-7 per site). 

Characterization of microbial communities

Characterization of photobiont and bacterial communities associated with each lineage was conducted on samples collected in November 2021 using ITS-2 and 16S sequencing, respectively. Raw ITS-2 reads were processed by Symportal (Hume et al., 2019) to produce defining intragenomic sequence variant (DIV) profiles for each coral colony (final n = 73). All colonies were dominated by C15 DIVs. Quality filtering, denoising, merging, and taxonomy assignments of 16S rRNA gene reads were conducted with DADA2 against the Silva v. 138.1 database (Quast et al., 2012), and contaminant reads were removed in Phyloseq (McMurdie & Holmes, 2013) (final n = 48). 

Analyses of holobiont optical traits

To characterize differences in holobiont structural and optical traits, we quantified polyp densities and light-harvesting characteristics in a subset of coral colonies. A total of 20 tagged colonies of known lineage and reef type were transported to Boston University in May 2023, fragmented, and then acclimated for 63 days in aquariums. Reflectance (R) between 400 and 750 nm was measured (sensu Enríquez et al., 2005; Vásquez-Elizondo et al., 2017). Chlorophyll a densities were quantified and the specific absorption coefficient of Chlrophyll a (Chla; a*_chla), a measure of the light absorption efficiency of the holobiont, was estimated following Enríquez et al., (2005). 

Thermal challenge experiment

A 25-day common garden heat challenge experiment was conducted to test for differences in thermal tolerance between the three Porites lineages (n = 24). L1 colonies were collected from classic reefs and L2 and L3 colonies were collected from extreme reefs. We initially aimed to include L1 and L2 colonies from both extreme and classic reefs, but were unsuccessful. Two cores were extracted from each colony; one core was assigned to the control treatment, and the other to the heat treatment. Mean temperatures in control tanks (n = 3 tanks per treatment) were maintained at 29.5°C ± 0.1°C. Temperatures in the heat treatment tanks were ramped by ~3°C over seven days, followed by a 12-day hold. On day 19, temperatures were increased by an additional ~1°C to simulate an extreme thermal stress event until day 25. All cores were inspected daily for mortality. Maximum PSII photochemical efficiency (Fv/Fm) was measured daily or semi-daily following >90 minutes of dark incubation. All fragments were also photographed with a color standard at six timepoints to measure changes in coloration (paling), quantified as the intensity in the gray channel using ImageJ (sensu McLachlan & Grottoli, 2021). To control for differences in gray intensity at the start of the experiment, we also calculated and analyzed relative changes in gray intensity (D grey intensity; values in heat - values in control for each colony at each timepoint).