Study location and transects

Permanent transects were established in August of 2013 with the objective of tracking symbiotic associations over time. Three twenty-five meter transects were positioned along a western north-facing barrier reef containing diverse and healthy coral assemblages typical of the equatorial West Pacific Ocean (Rebotel reef located at 7.2497°N, 134.2288°E). The transects were spaced over an area spanning approximately 100 meters of reef at a depth ranging approximately between 8 and 12 meters (at high-tide). To mark the endpoints of each transect, steel rebar stakes were driven into the carbonate substrate using a sledgehammer.

Colony tagging and temporal sampling of transect colonies

Transect tape was strung between rebar stakes to guide colony selection and sampling efforts. To ensure tagged corals were easily identifiable at later time points, selected colonies were within three meters from the central transect line. Numbered thermoplastic polyurethane tags were attached to corals using a hammer and steel nails (mounding and encrusting coral species) or by UV-stabilized nylon cable ties (branching species). Individual colonies were photographed to verify coral species. A fragment of skeleton and tissue (approximately ~2 cm²) was removed for genetic analysis using a small chisel and hammer and placed in numbered plastic bags with seawater. Coral fragments were transported by boat to the Palau International Coral Reef Center (PICRC) where they were preserved in 20% Dimethyl sulfoxide salt (DMSO-EDTA-NaCl) buffer and stored at -20°C.

A variety of taxonomically diverse corals were selected for long-term monitoring. Coral host species were identified visually from photographs. Transect collections include 162 individual coral colonies representing 10 families and 21 genera; Acropora and Isopora (Acroporidae), Stylophora, Seriatopora, and Pocillopora (Pocilloporidae), Symphyllia (Lobophylliidae), Diploastrea (Diploastreidae), Podabacia (Fungiidae), Goniastrea, Favites, Platygyra, and Leptoria (Merulinidae) (Budd et al. 2012, Huang et al. 2014a, Huang et al. 2014b) as well as three additional genera, Turbinaria (Dendrophylliidae), Psammocora (Psammocoridae), Leptastrea (Leotastreidae), and Pachyseris (Pachyseridae) (WoRMS 2024 https://www.marinespecies.org).

Return visits to resample colonies referenced waterproof site maps showing the relative position of each colony, along with colony photos, and tag number that verified their identity. The position of the sample along the surface of the colony was randomized with each sampling. Sampling times were dependent on travel and were varied, initially occurring at approximately six month to one-year intervals. Sampling was conducted at nine independent time points, spanning a total observation period of nine years for several colonies.

Spatial sampling of transect colonies

Intra-colony diversity was investigated by spatial collections. In July 2016, two (paired) samples were collected from each tagged colony. The first sample was obtained from the left side of the colony when facing the colony with the reef slope behind it. The location of the second sample was randomized along the top axis of the colony using a percent distance method. When viewed from above, the colony was visually segmented into ten equal-sized partitions, each representing 10% of the colony’s transverse axis. Next, a random number was generated to assign the numbered partitions between paired samples (^#1–^#10). Using this method, the distance between paired samples ranged from samples taken adjacent to each other to samples collected at the opposite end of the colony. Skeletal fragments were individually placed into separate pre-labeled collection bags and preserved following the methods described above.

DNA extraction and analysis of diversity using conventional genetic markers

DNA extractions were performed on 5x5 mm skeletal fragments containing animal tissue with associated symbionts using a modified Promega Wizard genomic DNA extraction protocol described by LaJeunesse et al. (2003). Samples were analyzed by denaturing gradient gel electrophoresis (DGGE) profiling of the ribosomal internal transcribed spacer 2 (ITS2). Prominent and discrete bands in DGGE of diagnostic profile fingerprints were excised in a sub-set of samples and directly sequenced as described by LaJeunesse (2002). Thus, by targeting only the prominent bands through direct sequencing, we isolated the numerically dominant ITS2 rDNA sequence in the nuclear genomes of resident symbionts. DGGE-fingerprinting was also used to identify samples containing a mixed population of two or more symbionts when present at greater than 10% of the total population (Thornhill et al. 2006). For additional verification of symbiont species identity, the nuclear large-subunit ribosomal DNA (LSU) was amplified following conditions specified by Zardoya et al. (1995). All sequencing was conducted on an Applied Biosciences sequencer (Applied Biosciences, Foster City, CA) at the Pennsylvania State University Genomics Core facility.

High-resolution genotyping of the symbiont population

Microsatellite loci were used to resolve multilocus genotypes of clone-dominated symbiont populations (i.e. synonymous with resolving individual strains, or clones). A total of 10 microsatellite loci developed for Cladocopium spp. were used including 9 loci (Sgr 34, SgrSpl 13, SgrSpl 22, SgrSpl 24, SgrSpl 25, SgrSpl 26, SgrSpl 78, Spl 1, and Spl 16) described by Wham et al. (2014) and 1 locus (C1.05) described by Bay et al. (2009). Genotyping was applied to those colonies harboring Cladocopium madreporum, Cladocopium patulum, Cladocopium C21 and Cladocopium C^Stylophora. These loci are unlinked and using 10 provides sufficient variability to unambiguously identify distinct clonal cell lines. Each locus was amplified in reaction volumes of 10 μl each comprised of 1 μl of 10 mM dNTPs (deoxynucleotide triphosphates), 1 μL 25 mM MgCl2, 0.2 U Taq DNA Polymerase, and 1 μl of standard Taq Buffer (New England Biolabs, Ipswich, MA), 1 μL of forward and reverse primers at 10 μM and 1 μL of 100-50 ng DNA template). PCR amplifications were performed according to the specifications given by Wham et al. (2014) and Bay et al. (2009). Microsatellites were analyzed on an ABI 3730 Genetic Analyzer (Applied Biosystems) using a 500-bp standard (LIZ-labeled) at the Penn State University Nucleic Acid Facility. Fragments sizes were scored visually using Geneious (Geneious version 6.1.8 created by Biomatters, Newark, NJ, USA).

The number of alleles amplified at each locus identified samples homogeneous for only one genotype as opposed to a mixed population. Two alleles at each locus, expected for Cladocopium, represent a single genotype (Wham et al. 2014, Wham and LaJeunesse 2016), while three or more alleles may be indicative of a combination of multiple co-occurring genotypes or non-target (i.e. non-specific) PCR amplifications. Distinctive symbiont genotypes characterized for each symbiont species (i.e., unique multi-locus genotypes or MLGs) were assigned a number designation (Table S2 from Lewis et al., 2024). Matching symbiont clones were identified when alleles were identical across all microsatellite loci.

While genotypes of mixed populations in a colony could not be characterized for some colonies, there were instances when both mixed and pure single genotypes were observed at different time points for a particular colony. Thus, over the course of temporal and spatial sampling, in instances when a sample was homogeneous for one, the MLG for each was deduced.

Use of high resolution psbA^ncr sequences to verify clone identity

The non-coding region of the psbA minicircle (psbA^ncr) was amplified and sequenced according to the conditions described by Moore et al. (2003). Data from these hyper-variable nucleotide sequences was used to verify symbiont species identity, further assess intra-species genotype identity (i.e. clone), and display this clone diversity on a phylogeny (LaJeunesse and Thornhill 2011). Sequences were aligned using the internet version of ClustalOmega followed by manual editing.

Sea surface temperatures

We obtained average monthly surface seawater temperatures from HOBO temperature loggers deployed at Ulong Rock near the offshore reef site (7.29042° N; 134.24105° E) provided from the Coral Reef Research Foundation. Additional temperatures were obtained from the Daily 5-km Satellite Coral Bleaching Thermal Stress Monitoring Product Suite data from the National Oceanic and Atmospheric Administration’s Coral Reef Watch website. Daily surface seawater temperatures (SST) for Palau were determined by calculating the average of the daily low (SST_MIN) and high (SST_MAX) SST between July 2013 and July 2022 (NOAA Coral Reef Watch). Both surface seawater measurements were plotted for the period of observation to evaluate seasonality and identify potential anomalous temperature events (see Figure 4b in Lewis et al., 2024). The temperature data is presented in the Supplemental file below titled 'palau_hobo_temps_offshore_2013_2022.csv'.