Divergent symbiont communities determine the physiology and nutrition of a reef coral across a light-availability gradient

Reef corals are mixotrophic organisms relying on symbiont-derived photoautotrophy and water column heterotrophy. Coral endosymbionts (Family: Symbiodiniaceae), while typically considered mutualists, display a range of species-specific and environmentally mediated opportunism in their interactions with coral hosts, potentially requiring corals to rely more on heterotrophy to avoid declines in performance. To test the influence of symbiont communities on coral physiology (tissue biomass, symbiont density, photopigmentation) and nutrition (δ13C, δ15N), we sampled Montipora capitata colonies dominated by a specialist symbiont Cladocopium spp. or a putative opportunist Durusdinium glynnii (hereafter, C- or D-colonies) from Kāne‘ohe Bay, Hawai‘i, across gradients in photosynthetically active radiation (PAR) during summer and winter. We report for the first time that isotope values of reef corals are influenced by Symbiodiniaceae communities, indicative of different autotrophic capacities among symbiont species. D-colonies had on average 56% higher symbiont densities, but lower photopigments per symbiont cell and consistently lower δ13C values in host and symbiont tissues; this pattern in isotope values is consistent with lower symbiont carbon assimilation and translocation to the host. Neither C- nor D-colonies showed signs of greater heterotrophy or nutritional plasticity; instead changes in δ13C values were driven by PAR availability and photoacclimation attributes that differed between symbiont communities. Together, these results reveal Symbiodiniaceae functional diversity produces distinct holobionts with different capacities for autotrophic nutrition, and energy tradeoffs from associating with opportunist symbionts are not met with increased heterotrophy.

collected, the depth of each colony and the time of collection was noted to correct for changes in tidal height. Final depths were corrected to mean seawater height using NOAA tide data at 6min intervals for Moku o Lo'e (Station ID: 1612480) from CO-OPS API in a custom R code (Innis et al. 2018).

Dissolved nutrients and plankton sampling
Seawater dissolved nutrients and the isotope values of plankton (i.e., isotope end-members or heterotrophic food sources) were sampled 10 August and 19 December 2016 and used to account for site and/or seasonal differences in nutrient loading and end-members across the study. Molar concentrations (µmol L -1 ) of ammonium (NH 4 + ), nitrate+nitrite (NO 3 -+ NO 2 or N+N), phosphate (PO 4 3-) and silicate (Si(OH) 4 ) were analyzed using a Seal Analytical AA3 HR nutrient autoanalyzer at the University of Hawai'i at Mānoa SOEST Lab for Analytical Biochemistry.
Plankton and suspended particles were sampled at the four locations where corals were collected, along with two additional locations in central Kāne'ohe Bay (21°27'28.7"N, 157°49'37.5"W, and 21°27'35.2"N, 157°49'23.7"W). At each location, plankton was sampled by pooling a vertical (<10 m) and surface horizontal plankton tows (63 µm mesh); visible debris or plant materials were removed and plankton were size-fractioned with nylon mesh in two size classes: 100 -243 µm and >243 µm. Additionally, seawater samples (10 L) collected at 3 m depth were fractioned with nylon mesh into three size classes: <10 µm, 10 -100 µm, <243 µm. All samples were filtered onto GF/F filters (0.7 µm) using a vacuum pump at low pressure, rinsed with ddH 2 O, and dried at 60 °C overnight. Plankton samples were removed from filters and ground to a powder with mortar and pestle; seawater fractioned materials were left on the GF/F filter, which was subsampled for carbon and nitrogen isotope analysis.

Photopigment analysis
Photopigments (chlorophyll a and c 2 ) were quantified by centrifuging an aliquot of the tissue slurry to isolate symbiont cells (13,000 g × 3 min), re-suspending the pellet in 100% acetone, and extracting pigments at -4 °C for 24 h in darkness. Chlorophyll concentrations were measured on a spectrophotometer using a glass 96-well plate at 630 nm and 663 nm and quantified using equations for dinoflagellates (Jeffrey and Humphrey 1975).

Stable isotope analysis
The coral tissue slurry was filtered to remove carbonates (20 µm nylon mesh) (Maier et al. 2010) and the host and symbionts were separated by centrifugation (2000 g × 3 min) with filtered seawater rinses (0.2 µm) (Muscatine et al. 1989). Tissues were lyophilized, ground to a powder, and packed in tin capsules for analysis. Analyses of tissue carbon and nitrogen isotope compositions were conducted using a Costech elemental combustion system coupled to a Coral skeleton samples were collected by shaving the uppermost layers of the coral skeleton (ca. 2 mm) using a Dremel tool equipped with a diamond-tip, followed by grinding samples to a powder (Rodrigues and Grottoli 2006 in only one technical replicate were considered absent. In each sample, relative symbiont abundance (i.e., C:D ratio) was determined from amplification threshold cycles (C T ) for Cladocopium and Durusdinium (i.e., C T C , C T D ) according to the formula C: Gene locus copy number and fluorescence intensity were used to normalize symbiont-specific C T values (Cunning et al. 2016).