Methodology from McIlroy & Coffreoth, Coral Reefs (2017):

Coral collection and rearing is described in detail in McIlroy et al. (2016). Briefly, we collected gametes from multiple spawning Orbicella faveolata colonies at two reefs in the Florida Keys, FL at Alligator Reef (24°48.7710 N, 80°40.1670 W) and Looe Key (24°32.6930 N, 81°24.5620 W) in 2011. Eggs and sperm from each site were mixed immediately for fertilization (Miller and Szmant 2006). At the laboratory, gametes and later embryos from all sites were combined and maintained in closed, 18-20-L tanks with 1.0-lm filtered seawater (FSW). Tanks were outdoors under a metal structure for shade with water temperatures ~28-30 degrees C, similar to ambient temperatures at the spawning sites. Ceramic tiles (5 cm x 5 cm, preconditioned for 10 d in 1.0-um FSW) served as a settlement substrate. Full FSW changes were done initially 1-3 times d-1; then every 1-3 d following settlement.

To assess ontogenetic effects on initial symbiont uptake, a set of recruits was maintained aposymbiotic. To measure the effect of initial infection on symbiont uptake, inoculations with 500 cells mL-1 of one of two cultured Symbiodinium strains, A194 or B184 (Table 1), were done repeatedly from the late larval stage through settlement and were terminated when visible infection was observed (~2 weeks). The cultures were Symbiodinium species present in other hosts throughout the Caribbean (LaJeunesse et al. 2012; Mellas et al. 2014) and are readily acquired by O. faveolata in the laboratory (Voolstra et al. 2009; McIlroy et al. 2016), but they are not the Symbiodinium species that are associated with adult O. faveolata colonies in the field (Table 1). Corals may respond differently to symbiont species dominating the symbiosis in the field, but these species have yet to be brought into culture, so this hypothesis could not be tested.

Two weeks post settlement, recruits were transferred to the University of Miami Experimental Hatchery (Virginia Key, Miami, FL) where they were maintained in indoor tanks under 5-20 umol photons m-2 s-1 in FSW (i.e., no exposure to exogenous symbionts), and fed weekly with Zeigler (Gardners, PA) larval diet (\100 lm in size). The light level was similar to that measured in reef crevices at the spawning site (McIlroy et al. 2016), the preferred settlement environment of this species (Miller 2014). Each month we took tiles from each treatment (A194, B184, aposymbiotic), removed 10 recruits per treatment for genotype analysis, and transferred the tiles with the remaining~20 recruits to an indoor seawater system of coarsely filtered (~20 lm) natural seawater with environmental symbionts at the same temperature and under the same low light levels. Other sources of symbionts included 1-2-yr-old, laboratory-reared O. faveolata corals with in hospite symbiont communities dominated by D206 and infected recruits (A194 and B184 treatments) from this study that were no longer physically isolated from each other (Fig. 1). One month after transfer we sampled the recruits (~20 per initial treatment) into 95% ethanol for symbiont genotyping.

To genotype Symbiodinium within recruits, we extracted the total genomic DNA (Coffroth et al. 1992), amplified the hypervariable region of the chloroplast 23S rDNA gene, and ran an electrophoretic fragment length analysis (Santos et al. 2003). All PCRs included positive (cultured Symbiodinium) and negative controls. Where no amplification occurred after three attempts, Symbiodinium was assumed to be absent or below the level of detection of this technique (10-1000 Symbiodinium cells; Santos et al. 2003) and was reported as "none detected". Within the Symbiodinium B184 treatment, we randomly selected four post-exposure recruit DNA extractions, and amplified and sequenced a species-level marker, B7Sym15, which can distinguish our culture species from other potential environmental types including the Symbiodinium B184 that occurs in O. faveolata adults (Finney et al. 2010; LaJeunesse et al. 2012).

For symbiont uptake in the aposymbiotic treatment, we compared the frequency of infected and uninfected polyps after exposure to environmental symbionts with Fisher’s exact test (FET; R package RVAideMemoire v 0.9-52). Where significant, we used post hoc multiple comparison with Bonferroni adjustment. We limited the comparisons a priori to pairs of consecutive months. For the A194 and B184 infected recruits, we used FET to compare the frequency with which the two pre-inoculated recruits harbored additional symbionts.