Sample collections:

Triplicate colonies of the corals Orbicella (formerly Montastrea) faveolata (Budd et al., 2012), Montastrea cavernosa, Diploria strigosa, Porites astreoides and Porites porites were sampled via SCUBA diving using hammer and chisel at four sites within the Florida Keys, offshore of Summerland Key at depths ranging from 2.4 – 7.6 m (Reef flat, 24 deg 33.155’, 81 deg 22.88’; Open water patch reef, 24 deg 33.164, 81 deg 26.225; Mid-channel patch reef 24 deg 33.620, 81 deg 30.076; Nearshore reef, 24 deg 36.341, 81 deg 25.756 and an offshore Nursery site, 24 deg 33.745, 81 deg 24.013, where corals had been transplanted from nearby reefs). Fragments from each colony were placed in sterile Whirl-Pak bags (Nasco, Fort Atkinson WI, USA) underwater and transferred to a cooler containing ice after dive completion. Within 1-4 hours, each fragment was rinsed with 0.1 um filtered seawater. Coral mucus was collected by siphoning the mucus from the coral surface with a pipette, and mucus fractions were frozen at -80 deg C. The coral fragment was then divided using a sterilized chisel and hammer into two smaller fragments; one frozen at -80 deg C and the second placed in a 4% paraformaldehyde-0.2 um filter sterilized phosphate buffered saline solution, fixed at 4 deg C for 1 hour, and stored at -20 deg C.

At each site, seawater temperature, dissolved oxygen, and pH were measured from a depth just above the corals (~5 m) and from the surface (0-1 m) using an EXO water quality sonde (YSI Inc., Yellow Springs, OH, USA). Seawater samples (4 l) from the same depths were also collected for microbial biomass, inorganic nutrient analyses and direct cell enumeration, and stored on ice for 1-4 hours. Upon arrival at the laboratory, water for nutrient analyses was frozen at -20 deg C. For direct cell counts, duplicate 1 ml aliquots of seawater were preserved at a final concentration of 4% paraformaldehyde for 20 minutes at 4 deg C and then frozen at -80 deg C. Finally, ~2 l of water was pressure filtered onto duplicate 25 mm, 0.22 um Durapore membrane filters (Millipore, Boston, MA, USA) using a peristaltic pump for collection of seawater microbial biomass and stored at -80 deg C.

Nutrient and pigment analyses and direct cell counts:

Concentrations of dissolved inorganic nutrients (NH4+, NO3– + NO2–, NO2-, PO43–, and silicate) were measured using a continuous segmented flow system with methods previously described (Apprill and Rappé, 2011), with NO3- was derived from subtracting the contribution of NO2-. NH4+ was also measured in the field on unfrozen samples using the ortho-phthaldialdehyde fluorescence method (Holmes et al., 1999;Taylor et al., 2007). Pigment analysis of the phytoplankton community was performed using high-performance liquid chromatography with known standards on extracts from the frozen filters, and quantified as described previously (Van Heukelem and Thomas, 2001). Seawater microbial cell counts were measured using flow cytometry (Apprill and Rappé, 2011).

Preparation of nucleic acids:

In order to examine mucus- and tissue-associated microbes, coral samples were processed using three approaches (Fig. 1A). The first approach harvested the mucus that was collected as previously described (hereafter referred to as ‘mucus’ samples). For the second approach, the frozen coral tissue (which still contained some mucus) was removed from the skeleton using an airbrush with 80 psi air pressure and 0.22 um filtered phosphate buffered saline solution. The tissue homogenate was then vortexed for 2 min and centrifuged at 20,000 x g for 20 min at 4 deg C. After removal of the supernatant, the cells were stored at -80 deg C for 1 week or less (referred to as ‘holobiont’ samples). Lastly, the third biomass substrate analyzed was decalcified coral tissue, referred to as ‘tissue’ samples, which were devoid of mucus and skeleton. To obtain these samples, the coral subsample that was initially preserved in paraformaldehyde solution was placed in a 20% EDTA solution (Acros Organics Thermo Fisher Scientific, NJ, USA) for 2-3 weeks at 4C on a slow rocker. The EDTA solution was exchanged daily until complete skeleton dissolution, (similar to Ainsworth et al., 2015).

Nucleic acids were extracted from the seawater membrane filters (1/4 of the 25 mm filter), the holobiont cells, and the decalcified tissue samples using the PowerPlant Pro DNA Isolation Kit (Mo Bio Laboratories, Inc., Carlsbad, CA, USA), with modifications that included adaptations that included 350 mg of 0.1mm silica beads (MP Biomedicals, Irvine, CA, USA) to the bead solution and 25 ul of proteinase K (Qiagen Inc., Valencia, CA, USA) followed by incubation at 65 deg C for 60 min. Because tissue samples were initially preserved in paraformaldehyde, they were subject to an extended 30-minute proteinase K digestion at 55 deg C followed by an additional high-heat incubation step at 90 deg C for 60 minutes before beadbeating. The PowerPlant Pro extraction method did not result in high-quality DNA from the mucus. Therefore, the PowerBiofilm DNA Isolation kit (Mo Bio Laboratories), containing inhibitor removal steps but otherwise a similar bead mixture and proteinase K digestion at 65 deg C for 60 min, was used for mucus samples. All nucleic acids were quantified using the Qubit dsDNA BR fluorescent assay (Invitrogen Corp., Carlsbad, CA, USA) on a Qubit 2.0 fluorometer.

Sequencing of V4 region of archaeal and bacterial SSU rRNA genes:

Barcoded primers targeting the V4 hypervariable region of the SSU rRNA gene, 515F and 806R, were utilized for sequencing analysis as detailed in Kozich and colleagues (2013). This sequencing and data analysis occurred prior to the modifications of these primers to capture additional SAR11 clade bacteria (Apprill et al., 2015) and Thaumarchaeota (Parada et al., 2015), and therefore SAR11 are likely underrepresented by 15-25% in the seawater samples; Thaumarchaeota are probably not heavily underestimated in this study because the archaeal amoA gene abundances suggest their low abundance in reef seawater. Triplicate 25 ul PCR reactions contained 1.25 U of GoTaq Flexi DNA Polymerase (Promega Cooperation, Madison, WI, U.S.A.), 5X Colorless GoTaq Flexi Buffer, 2.5 mM MgCl2, 200 uM of each dNTP, 200 nM of each barcoded primer, and 0.15 ng - 2.0 ng of genomic template for most samples, with some samples diluted or concentrated to encompass a range of 0.0057 to 4.40 ng. The reaction conditions included an initial denaturation step at 95 ºC for 2 min, followed by 32-39 cycles of 95 ºC for 20 s, 55 ºC for 15 s, and 72 ºC for 5 min, concluding with an extension at 72 ºC for 10 min. The reactions were carried out on a thermocycler (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Reaction products (5 ul) were screened on a 1% agarose-TBE gel. Samples were optimized for PCR at the lowest number of cycles that resulted in an amplified PCR product detected on a gel with the HyperLadder II standard (generally 5 ng ul-1) (Bioline, Taunton, MA, USA), thus minimizing biases from over-amplification. The three replicate reactions were excised from the gel, combined, and purified using the Qiagen QIAquick Gel Extraction Kit (Qiagen Inc., Valencia, California, USA), and quantified using the Qubit dsDNA HS fluorescent assay. Barcoded amplicons were pooled in equimolar ratios (5 ng each) and shipped to the University of Illinois W.M. Keck Center for Comparative and Functional Genomics for construction of libraries and sequenced using 2x250 bp paired-end MiSeq (Illumina, San Diego, CA, USA).