Coral Reefs (2015) 34:535–547 DOI 10.1007/s00338-015-1277-z

REPORT

Spatially distinct and regionally endemic assemblages in the threatened -building faveolata

Dustin W. Kemp • Daniel J. Thornhill • Randi D. Rotjan • Roberto Iglesias-Prieto • William K. Fitt • Gregory W. Schmidt

Received: 28 October 2014 / Accepted: 19 February 2015 / Published online: 27 February 2015 Ó Springer-Verlag Berlin Heidelberg 2015

Abstract Recently, the Caribbean reef-building coral Or- with of Symbiodinium in clades A (type A3), B (B1 bicella faveolata was listed as ‘‘threatened’’ under the U.S. and B17), C (C3, C7, and C7a), and D (D1a/Symbiodinium Act. Despite attention to this species’ trenchii). Within-colony distributions of Symbiodinium conservation, the extent of geographic variation within O. species correlated with light availability, cardinal direction, faveolata warrants further investigation. O. faveolata is and depth, resulting in distinct zonation patterns of en- unusual in that it can simultaneously harbor multiple ge- dosymbionts within a host. Symbiodinium species from netically distinct and co-dominant species of endosymbiotic clades A and B occurred predominantly in the light-exposed dinoflagellates in the Symbiodinium. Here, we inves- tops, while species of clade C generally occurred in the tigate the geographic and within-colony complexity of shaded sides of colonies or in deeper-water habitats. Fur- Symbiodinium-O. faveolata associations from Florida Keys, thermore, geographic comparisons of host–symbiont asso- USA; Exuma Cays, Bahamas; Puerto Morelos, Mexico; and ciations revealed regional differences in Symbiodinium Carrie Bow , Belize. We collected coral samples along associations. Symbiodinium A3 was detected in Me- intracolony axes, and Symbiodinium within O. faveolata soamerican coral colonies, but not in colonies from the samples was analyzed using the nuclear ITS2 region and Florida Keys or Bahamas. Likewise, Symbiodinium B17 was chloroplast 23S rDNA genotyping. O. faveolata associated unique to Mesoamerican O. faveolata, whereas Symbio- dinium B1 was found at all localities sampled. However, using cp23S genotyping paired with ITS2 analysis revealed Communicated by Biology Editor Dr. Simon Davy geographically endemic haplotypes among Symbiodinium D. W. Kemp (&) Á W. K. Fitt clades A, B, and C. Since Symbiodinium spatial hetero- Odum School of Ecology, University of Georgia, Athens, geneity among this coral species is greater than most , a GA 30602, USA question arises as to whether all western Atlantic populations e-mail: [email protected] of O. faveolata should be considered equally ‘‘threatened’’? D. J. Thornhill Alternatively, geographically and spatially distinct coral– Department of Biological Sciences, Auburn University, Auburn, symbiont associations may benefit from specialized man- AL 36849, USA agement protocols. R. D. Rotjan John H. Prescott Marine Laboratory, New England Aquarium, Keywords Biogeography Á Ecological Á Boston, MA 02110, USA Endangered Species Act Á Á Symbiosis Á Symbiodinium R. Iglesias-Prieto Unidad Acade´mica de Sistemas Arrecifales, Puerto Morelos, Instituto de Ciencias del Mar y Limnologı´a, Universidad Nacional Auto´noma de Me´xico, Puerto Morelos, Mexico Introduction

G. W. Schmidt Department of Plant Biology, University of Georgia, Athens, Coral reefs are among the most biologically rich and pro- GA 30602, USA ductive ecosystems in the world. Reef-building corals are 123 536 Coral Reefs (2015) 34:535–547 fundamentally important, serving as ecosystem engineers Reed 1985; Aronson et al. 2008). However, declines of this that form the structural basis for the ecosystem (Jones et al. species throughout the Caribbean (Miller et al. 2009) 1994). However, coral reefs are among the most globally caused the Center for Biological Diversity to seek protec- threatened ecosystems (Hoegh-Guldberg and Bruno 2010). tion of O. faveolata as a threatened species under the ESA. Rapid coral declines have resulted in substantial habitat Furthermore, O. faveolata routinely hosts multiple ge- loss revealing organismal sensitivity to environmental netically distinct and co-dominant Symbiodinium spp., and perturbations (Gardner et al. 2003). Current conservation these symbionts may likewise differ across the range of the and management efforts are focused on ways to minimize host (LaJeunesse 2002; Garren et al. 2006). Symbiodinium species decline and aid ecosystem recovery that may po- species associated with O. faveolata include members of tentially promote ecosystem resilience (capacity for stress clades A, B, C, and D that are distributed across the resistance and recovery) to local stressors (i.e., over-fish- landscape of the colony (Rowan and Knowlton 1995; ing, nutrient influx, sedimentation) and global climate Toller et al. 2001b; Garren et al. 2006; Kemp et al. 2008). change (Mumby and Steneck 2008; Anthony et al. 2014). Various types of Symbiodinium clades A and B tend to Reef-building corals confront managers with multiple dominate the light-exposed colony tops, with clade C types challenges due to biologic complexity and geographic found in greatest abundance on the colony sides, and specificity of host-symbiotic associations. Reef-building Symbiodinium ITS2 type D1a (S. trenchii) found in stress- corals are highly sensitive to environmental perturbations; exposed colonies (Rowan and Knowlton 1995; Toller et al. however, within-species physiologic differences (e.g., 2001a; Thornhill et al. 2006b; Kemp et al. 2008, 2014; thermal and high irradiance tolerance) can result from in- LaJeunesse et al. 2009, 2014). Flexibility in the symbiosis herent and acclimatization protection processes of coral ecology of O. faveolata has been documented spatially and and endosymbiotic algae (Warner and Berry-Lowe 2006; in response to environmental stress; such dynamism can Reynolds et al. 2008; Howells et al. 2013; Palumbi et al. alter coral physiology and stress responses (Rowan et al. 2014). 1997; Toller et al. 2001b; Thornhill et al. 2006b; Warner The relationship between reef-building corals and en- et al. 2006; LaJeunesse et al. 2009; Kemp et al. 2014). dosymbiotic dinoflagellates (genus Symbiodinium) enables Although the symbiosis of O. faveolata is well studied, no coral reefs to thrive in oligotrophic tropical and sub-tro- large-scale survey has investigated Symbiodinium species pical oceans (Muscatine and Porter 1977). Symbiodinium and type diversity within this species across the Caribbean can provide [90 % of the coral’s metabolic requirements basin. Recent evidence from Symbiodinium clade C via photosynthetically fixed carbon that is translocated to demonstrated that O. faveolata and its congeners associate the coral as carbohydrates, fatty acids, sugars, starches, and with highly specific endosymbionts—Symbiodinium species amino acids (reviewed by: Yellowlees et al. 2008). Sym- that apparently do not associate with other host taxa biodinium is a diverse group of dinoflagellates comprising (Thornhill et al. 2014). These host-specific symbionts can be nine major clades (A-I) (Coffroth and Santos 2005; Pochon further divided into regionally endemic lineages and haplo- and Gates 2010) that can be further divided using rapidly types (Thornhill et al. 2014). Such an evolutionary pattern evolving molecular markers into various phylotypes and likely resulted from long-term processes of lineage diversi- species (Sampayo et al. 2009; LaJeunesse and Thornhill fication due to ecological speciation (by host taxa) as well as 2011; LaJeunesse et al. 2012, 2014). Different Symbio- regional differentiation due to allopatry (Santos et al. 2004; dinium spp. often have biochemical differences leading to LaJeunesse 2005, 2010; Thornhill et al. 2009). However, species- and phlyotype-specific physiological responses to such patterns have only been documented in Symbiodinium environmental conditions (Warner et al. 1999; Tchernov clade C for Orbicella. The prevalence of other Symbiodinium et al. 2004; Reynolds et al. 2008; Takahashi et al. 2009). clades and types associated with O. faveolata has been Notably, differential thermal and irradiance tolerances inadequately determined, although preliminary results for have been identified in genetically distinct Symbiodinium, clades A and B—but not D—suggest similar patterns which can lead to different holobiont responses to stress (Thornhill et al. 2010a; Pinzo´n et al. 2011). Thus, we used (Rowan and Knowlton 1995; Baker 2003; Iglesias-Prieto here a micro-sampling technique combined with nuclear ri- et al. 2004). bosomal and chloroplast genotyping of Symbiodinium from Recently the National Oceanic Atmospheric Adminis- O. faveolata colonies to further explore the genotypic var- tration (NOAA) extended Endangered Species Act (ESA) iation in clades and types with respect to geographic corre- protections to the mountainous star coral, Orbicella lates. Colonies were examined in shallow (higher irradiance) faveolata, and 19 other species (NMFS 2014). O. faveolata and deeper (lower irradiance) reefs at four different regions is found from the intertidal zone to 80 m deep and is often of the western Atlantic. the most abundant reef-building coral in forereef environ- We document predictable and stratified heterogeneity of ments throughout the Caribbean (Goreau and Wells 1967; Symbiodinium spp. associated with O. faveolata that likely 123 Coral Reefs (2015) 34:535–547 537 reflects long-term ecological and evolutionary host–sym- coral tissue was collected using syringes along four fixed biont specializations. Defining variables that underlie the axes laid from top to bottom of the colony at north, east, establishment of Symbiodinium-Orbicella associations are south, and west compass headings following the micro- highly relevant because differential physiological re- sampling procedures of Kemp et al. (2008). The mi- sponses to thermal stress have been directly linked to crosampling approach allowed us to collect multiple and functional differences of the symbiotic algae (Kemp et al. precise polyp samples from corals, minimizing damage to 2014). Furthermore, generalized conservation strategies the coral colony. Samples were collected over a 5-yr period that encompass large geographic regions may not equally (2005–2010) from 2 to 3 polyps every 10–20 cm using 2-cc be applicable due to the dynamic physiological and genetic syringes with 16-gauge needles, resulting in 8–50 tissue nature of reef-building corals. collections per coral colony, with all collections from a single colony occurring on the same day. Upon returning to the laboratory, syringe samples were transferred into 2-ml Methods microcentrifuge tubes and centrifuged at (*5000g) for 2 min. Supernatants were decanted and either 80 % ethanol Collection sites and micro-sampling procedure alcohol or DMSO buffer (20 % dimethylsulfoxide, 0.25 M EDTA in saturated aqueous sodium chloride) was added Coral tissue samples were collected from O. faveolata for preservation (Seutin et al. 1991) and archived for later colonies from four geographic regions: Florida Keys, USA; analysis. Exuma Cays, Bahamas; Puerto Morelos, Mexico; and Orbicella faveolata was sampled from three shallow Carrie Bow Cay, Belize (Fig. 1). O. faveolata colonies reefs (B4 m) in the upper Florida Keys from 2006–2008: were collected from nearby shallow (1–4 m) and deeper two colonies at Turtle Reef (TR; 2–4 m depth; 25.294°N, (12–15 m) reefs in each location to examine variation 80.219°W), four colonies at Admiral Patch Reef (ADM; among habitats and regions. For each colony sampled, 2 m depth; 25.045°N, 80.395°W), and four colonies at

Fig. 1 a Western Atlantic reef locations investigated in this study. South Perry (SP). d Reefs sampled in Belize: Carrie Bow Cay Lagoon b Reefs sampled in the Florida Keys: Turtle Reef (TR), Little Grecian Reef (LG) and Carrie Bow Cay Reef Slope (RS). e Reefs sampled in Reef (LG), Admiral Patch Reef (ADM), and Alligator Reef (AG). Mexico: La Bocana Reef (LB) and Ship Channel Reef (SC) c Reefs sampled in the Bahamas: North Norman’s Pond (NNP) and 123 538 Coral Reefs (2015) 34:535–547

Little Grecian Reef (LG; 1–3 m depth; 25.119°N, that were deployed on the reef at approximately 3–4 m. For 80.302°W). Additionally, two colonies were sampled from all sites, temperatures were recorded hourly, and monthly the deeper Alligator Reef (AG; 10–12 m depth; 24.842°N, mean ± SD were calculated for the years 2003–2006 80.624°W) in the middle Florida Keys, for a total of 224 (Fig. 2). samples from the Florida Keys (8–44 samples per colony). Four colonies were sampled from the Bahamas, two each Genetic identification of Symbiodinium from North Norman Pond (NNP; 2–4 m depth; 23.791°N, 76.137°W) and South Perry Reef (SP; 10–12 m depth; Genetic identities of the dominant Symbiodinium spp. in 23.775°N, 76.090°W) in 2006 and 2007 that, with the in- each tissue sample were determined following the proto- tracolony sampling, resulted in 65 total samples (11–27 cols of LaJeunesse et al. (2003). DNA was extracted using samples per colony). In Mexico, O. faveolata tissue sam- a modified Promega Wizard genomic DNA extraction ples were collected from six colonies at La Bocana Reef protocol. Denaturing-gradient gel electrophoresis (DGGE) (LB; 2–4 m depth; 20.875°N, 86.512°W) from 2006 to was used to analyze the ITS2 region of nuclear ribosomal 2009 totaling 206 samples (24–50 samples per colony). O. RNA genes (LaJeunesse 2002). This method detects the faveolata colonies from the deeper Ship Channel Reef (SC; Symbiodinium types that comprise approximately 10 % or 9–10 m depth; 20.837°N, 86.870°W) in Mexico were not sampled as thoroughly as the other locations resulting in 25 total samples. Only one colony was sampled and demarked 32 with a tape measure, while three other colonies were only Florida Keys 31 sampled at the top, middle, and bottom (3–16 samples per Exuma Cays 30 Puerto Morelos colony). In Belize, four colonies were sampled from Carrie Carrie Bow Cay Bow Cay Lagoon Reef (LR; 1–3 m depth; 16.802°N, 29 88.083°W) and two additional colonies were sampled from 28 a deeper reef off Carrie Bow Cay Reef Slope (RS; 10–12 m 27 depth; 16.802°N, 88.080°W) in 2005 and 2006 resulting in 26 141 total samples (17–42 samples per colony). A summary 25 of all samples and the sites from which they were collected Average Temperature (°C) Temperature Average 24 is compiled in Table 1. 23 22 Temperature measurements Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec

Seawater temperatures were recorded from the Florida Keys Fig. 2 Mean monthly seawater temperatures from 2003 to 2006 (LG), Exuma Cays (NNP), Puerto Morelos (LB), and Carrie (±SD). Data loggers were placed in situ on North Normans Pond, Exuma Cays, Bahamas (depth 3 m); Little Grecian Reef, Upper Bow Cay (RS) using HOBO Pro v2 underwater temperature Florida Keys, USA (depth 4 m); La Bocana Reef, Puerto Morelos, loggers (Onset Computer Corporation, Bourne, MA, USA) Mexico (depth 4 m); Carrie Bow Cay, Belize (depth 4 m)

Table 1 Summary of Symbiodinium frequencies micro-sampled from Orbicella faveolata colonies Site Depth # # Clade Clades A Clades A Clade Clades B Clade Clade (m) Colonies Samples A &B &C B &C C D

Admiral Reef, FL, USA (ADM) 1–2 4 92 0 0 0 91 1 0 0 Turtle Reef, FL, USA (TR) 2–4 2 16 0 0 0 7 3 6 0 Little Grecian Reef, FL, USA (LG) 1–3 4 73 0 0 0 73 0 0 0 Alligator Reef, FL, USA (AG) 10–12 2 43 0 0 0 29 11 3 0 North Norman Pond Reef, Exuma Cay, 2–4 2 37 0 0 0 31 1 2 3 Bahamas (NNP) South Perry Reef, Exuma Cay, Bahamas (SP) 10–12 2 28 0 0 0 9 8 11 0 La Bocana, Puerto Morelos, Mexico (LB) 2–4 6 206 6 37 0 75 45 40 3 Ship Channel Reef, Puerto 9–10 4 25 1 0 2 12 1 9 0 Morelos, Mexico (SC) Carrie Bow Cay Lagoon Reef, Belize (LR) 1–3 4 99 7 15 8 42 17 10 0 Carrie Bow Cay Reef Slope, Belize (RS) 10–12 2 42 3 1 2 8 21 7 0

123 Coral Reefs (2015) 34:535–547 539 more of the endosymbiont community (Thornhill et al. were discarded as burn-in, and a 50 % majority-rule con- 2006b; LaJeunesse and Thornhill 2011). A touchdown sensus tree was calculated from the 7501 remaining trees. thermal cycle described in LaJeunesse et al. (2003) was Symbiodinium type A3 was used as an outgroup (following used for PCR amplification of the ITS2 region using the Nylander 2004). Posterior probabilities (PP) were recorded forward primer ‘‘ITSintfor2’’ (50-GAATTGCAGA ACTCC to assess reliability of recovered nodes. GTG-30) and the reverse primer ‘‘ITS2CLAMP’’ (50-CGC To further explore evolutionary relationships, cp23S CCGCCGC GCCCCGCGCC CGTCCCGCCG CCCCC (including the HVR) and ITS2 haplotype networks were GCCC GGGATCCATA TGCTTAAGTT CAGCGGGT- constructed with TCS v1.21 (Clement et al. 2000) using 30), with a 39-bp GC clamp (italicized; LaJeunesse and default settings and assumptions. Plausible branch con- Trench 2000). Products were electrophoresed on 45–80 % nections between haplotypes were tested using a 95 % urea-formamide denaturing gradient gels (100 % consists connection limit and gaps were treated as missing data. of 7 M urea and 40 % deionized formamide) for 10 h at 150 V at a constant temperature of 60 °C using a C.B.S. ScientificTM system (LaJeunesse 2002; Thornhill et al. Results 2010b). For verifying that DGGE banding patterns corre- sponded exactly with phylotype designations, prominent Intracolonial and geographic diversity of Symbiodinium bands were excised from gels, PCR amplification using spp forward and reverse primers (ITSintfor2 and ITS2rev) was performed as described by LaJeunesse (2002), and prod- ITS2-DGGE and exhaustive sequencing of DNA from ucts were sent to GENEWIZ, Inc. (South Plainfield, NJ) for prominent bands yielded seven distinct ITS2 haplotypes in Sanger sequencing. Sequences were deposited in GenBank total, from four different Symbiodinium clades associated under accession numbers (KP730718-KP730733) and al- with O. faveolata (Table 1). The nucleotide diversity phanumeric nomenclature (sensu LaJeunesse 2002) of the among haplotypes and clades was high (p = 0.400), re- representative DGGE banding patterns was assigned. sulting in difficulties aligning the 273 bp of ITS2 se- Symbiodinium phylotypes identified using the ITS2 re- quences, particularly across clades. gion were further analyzed using the Domain V (including Microsampling and molecular identification of Symbio- the hypervariable region) of the chloroplast (cp23S) rDNA. dinium inhabitants revealed stratified, within-colony Following the protocol of Santos et al. (2002), PCR am- heterogeneity in O. faveolata (Fig. 3). Additionally, there plification using the primer pair 23S1 (50-GGCTGTAAC- were geographic differences in the Symbiodinium ITS2 and TATAACGGTCC-30) and 23S2 (50-CCATCGTATTGAA cp23S types across all localities sampled (Figs. 3, 4). CCCAGC-30) was performed (Santos et al. 2002). Ampli- However, due to the limited number of corals sampled, fied products were sent to GENEWIZ, Inc. (South Plain- geographic variation should be interpreted cautiously. Most field, NJ) for Sanger sequencing using the forward and colonies revealed within-colony, spatial heterogeneity, of reverse primers (23S1 and 23S2) and sequences were de- Symbiodinium influenced by irradiance intensity (i.e., high- posited in GenBank under accession numbers (KP730734- irradiance top of colony versus low-irradiance bottom of KP730749). colony; Figs. 3, 4). A notable exception to this trend was found in shallow-water O. faveolata colonies from the Sequence analyses and phylogenic reconstructions upper Florida Keys, USA. Regardless of within-colony sampling position, only 5 % of samples from upper Florida ITS2 and cp23S chromatograms were visually inspected in Keys (TR, LG, ADM) in\4 m depth were detected with a Sequence Navigator. The edited sequences were aligned symbiont type other than ITS2-type B1 Symbiodinium using ClustalW2 (http://www.ebi.ac.uk/Tools/msa/clus (n = 11 colonies; 181 samples; Table 1; Figs. 3a, 4). talw2/), with additional corrections to the alignment made However, coral colonies from the deeper reef (AG; by visual inspection of the output file. 10–12 m) had more diverse Symbiodinium communities Phylogenetic reconstructions of cp23S, excluding the than shallow-water conspecifics, including 67 % of the hypervariable region (HVR) (Zhang et al. 2000), were samples having ITS2 type B1, 5 % C3, and 26 % B1 and conducted using Bayesian Inference using MrBayes v3.2.1 C3 mixtures (Figs. 3b, 4). implementing the Hasegawa–Kishino–Yano (HKY) ? C Similar to the Florida Keys, shallow-water O. faveolata model of substitution, as indicated by the Akaike Infor- colonies from the Exuma Cays, Bahamas, harbored limited mation Criterion by MrModeltest v2.3 (Huelsenbeck and Symbiodinium diversity within the detection limits of Ronquist 2001). The analysis was run for 106 generations DGGE (four symbiont types across 65 samples; Table 1). and sampled every 100 generations. Due to convergence of Eighty-four percent of the intracolony samples from North chains within 2.5 9 105 generations, the first 2500 trees Norman Pond (NNP) associated solely with type B1 123 540 Coral Reefs (2015) 34:535–547

123 Coral Reefs (2015) 34:535–547 541 b Fig. 3 The distribution of Symbiodinium phylotypes (ITS2) along associated only ITS2 type C7a from NNP. Additionally, fixed axes located at cardinal headings of north, east, south, and west 8 % of NNP samples contained Symbiodinium ITS2 type from Orbicella faveolata. Intracolony generalizations of Symbiodini- um zonation patterns of shallow and deeper-water colonies from the D1a (=S. trenchii or D1-4; Figs. 3c, 4). Coral samples from (a, b) Florida Keys, USA; (c, d) Exuma Cays, Bahamas; (e, f) Puerto the deeper Bahamas reef (SP; 10–12 m) were found to Morelos, Mexico; (g, h) Carrie Bow Cay, Belize. Identical sampling associate with type B1 in the tops (higher irradiance) of the protocols were followed at all locations and depths colonies (32 % of intracolony samples). The sides and lower regions of colonies from SP were found to have greater abundances of Symbiodinium type C7a (68 % of intracolony samples, Figs. 3d, 4) and mixtures of type B1 and C7a (29 % of intracolony samples) than the shallow- water conspecifics (Fig. 4). In striking contrast to the Florida Keys and Exuma Cays, O. faveolata from Mesoamerican reefs of Puerto Morelos, Mexico (n = 10 colonies; 231 samples; Table 1) and Carrie Bow Cay, Belize (n = 6 colonies; 141 samples; Table 1) harbored Symbiodinium assemblages with greater diversity. Coral colonies from Mesoamerican reefs simul- taneously associated with up to five co-dominant Symbio- dinium ITS2 types, including various combinations of types A3, B1, B17, C7, and occasionally D1a (Mexico only). Intracolonial spatial heterogeneity correlated with prevailing irradiance patterns across the colony. Generally, Symbiodinium ITS2 type A3 occurred in high irradiance areas on the coral (i.e., top and south facing sides of the colony; Fig. 3e–h), while type C7 was found in lower ir- radiance areas on the coral (i.e., the lower colony edge; Fig. 3e–h). Symbiodinium type B17 was the most fre- quently detected Symbiodinium spp. found in shallow-wa- ter corals from Mexico (72 % of all samples; Fig. 4) and Belize (76 % of all samples; Fig. 4). In shallow-water colonies Symbiodinium type B17 was generally found in higher irradiance regions of the coral and was frequently found mixed with types A3 and C7. Despite its abundance in shallow-water habitats, Symbiodinium type B17 was not found in any of the deeper-water corals (Figs. 3, 4). In- stead, Symbiodinium ITS2 type B1 occurred in deeper- water colonies of O. faveolata from Mexico and Belize (Figs. 3f, h, 4). In deeper-water colonies, the prevalence of Symbiodinium type C7 (Mexico 48 % of all samples; Fig. 4 Generalized Symbiodinium ITS2 phylotype assemblages of Fig. 4) increased in abundance (Belize 71 % of all sam- Orbicella faveolata. Pie graphs represent proportion of identified ples; Fig. 4). Similar to shallow-water counterparts, type Symbiodinium from all sampled corals at each region. Florida Keys, C7 was found most prominently in low-irradiance sides and USA: Turtle Reef, Admiral Reef, and Little Grecian Reefs (2–4 m bottoms of the colonies (Fig. 3f, h). depth; n = 10 colonies; 164 samples) and from Alligator Reef (10–12 m depth; n = 2 colonies; 43 samples). Exuma Cays, Bahamas: North Normans Pond (2 m depth; n = 2 colonies; 37 samples) and O. faveolata-Symbiodinium associations using South Perry Reef (12–14 m depth; n = 2 colonies; 28 samples). Puerto the cp23S marker Morelos, Mexico: La Bocana Reef (2–4 m depth; n = 5 colonies; 197 samples) and Ship Channel (n = 4 colonies; 25 samples). Carrie Bow Cay, Belize: Lagoon Reef (2–4 m depth; n = 4 colonies; 96 samples) Data from the cp23S marker largely mirrored the findings and Reef Slope (10–12 m depth; n = 2; 42 samples) based on ITS2-DGGE. The cp23S identified O. faveolata symbionts belonging to four different clades (i.e., clades A Symbiodinium (Figs. 3c, 4). Mixtures of Symbiodinium through D; posterior probability 1.00 for all nodes) and type C7a (aka C12, see Thornhill et al. 2014) and B1 were nine different sub-cladal types, each with a distinct geo- found in 2 % of samples, whereas 5 % of the samples graphic distribution (Fig. 5). The cp23S alignment 123 542 Coral Reefs (2015) 34:535–547

a cp23S without the HVR b cp23S with the HVR c ITS2

A3 A3 A3 Clade A

B1 B1 B17 1.00 Clade B B1 B17

1.00 B1

C3 C7 & C7a C3 C7 C7a 1.00 Clade C 1.00

D1a D1a Clade D

0.1

Mexico the Bahamas the Florida Keys Belize Unsampled Haplotype

Fig. 5 Bayesian Inference topology and parsimony network analyses probabilities. Clade A Symbiodinium were used as the outgroup of evolutionary relationships among Symbiodinium types and species following Pochon et al. (2006). b, c Parsimony network analyses of associated with Orbicella faveolata. Geographic location is color the cp23S with inclusion of the hypervariable region (a) and ITS2 coded in all figure panels. a Bayesian inference reconstruction of (c) at the 95 % connectivity level and with indels treated as missing evolutionary relationships based on the cp23S without inclusion of the data. Gray dots represent unsampled haplotypes in the analysis hypervariable region. Nodal support values are indicated as posterior comprised 680 nucleotide positions, with a nucleotide di- This shared clade C haplotype likely reflects the lower versity (p) of 0.136. Although the overall patterns between resolution power of the cp23S compared to the ITS2 for nuclear and chloroplast markers were similar, the cp23S clade C Symbiodinium genotypes (Fig. 5). Symbiodinium provided greater within-lineage resolution for clades A and type D1a only occurred in the colonies sampled from the B, but a lower lineage resolution for clade C compared to Bahamas and Mexico according to both the cp23S and the ITS2 (Fig. 5). Whereas the ITS2 marker resolved 1 ITS2 datasets (Fig. 5; but note that previous studies have type of clade A, 3 types of clade B, 3 types of clade C, and detected this symbiotic species in Orbicella colonies in the 1 type of clade D, the cp23S resolved 2 clade A haplotypes, Florida Keys and Belize, e.g., Rotjan et al. 2006; Thornhill 5 clade B haplotypes, 2 clade C haplotypes, and 1 haplo- et al. 2006b; Finney et al. 2010). type of clade D. The geographic distribution of Symbiodinium clades, types, and species was similar for both the ITS2 and cp23S. Discussion Clade A Symbiodinium was detected only in O. faveolata from Mesoamerican reefs and not in O. faveolata corals Over the last two decades, there have been precipitous de- from the Florida Keys or Exuma Cays. Although ITS2 clines of Orbicella spp. throughout the western Atlantic resolved only type A3 in O. faveolata, the cp23S resolved (Edmunds and Elahi 2007; Alvarado-Chaco´n and Acosta different regionally specific haplotypes of type A3 in 2009; Miller et al. 2009). Despite these losses, Orbicella spp. Mexico compared to Belize (Fig. 5b). Clade B Symbio- (including O. faveolata, O. annularis, and O. franski) remain dinium spp. were found in all locations sampled and could among the most ecologically important reef-building corals be divided into various regional haplotypes. These included in the Caribbean. Members of these taxa have slow growth type B1 in the Florida Keys, type B17 in Mexico and rates (Hudson 1981), late reproductive maturity (Szmant Belize, and three different regionally endemic cp23S hap- 1986), low recruitment rates (Hughes and Tanner 2000), and lotypes of type B1—a Florida Keys haplotype, an Exuma high susceptibility to thermal stress (Fitt et al. 1993). It was Cays haplotype, and a Mesoamerican reef haplotype for these ecological, biological, and physiological reasons (Fig. 5b). Clade C Symbiodinium occurred at all reefs that NOAA designated O. faveolata and its congeners as sampled. Type C3 associated with O. faveolata from the ‘‘threatened’’ under the ESA (NMFS 2014). Florida Keys and another cp23S haplotype, comprising ITS Our results suggest that O. faveolata exhibits distinct types C7 and C7a, was found at other locations (Fig. 5b). symbiotic associations in different regions of the Caribbean

123 Coral Reefs (2015) 34:535–547 543

Sea and western Atlantic Ocean (see also Thornhill et al. hypothesis that many phylotypes in this group are high- 2014). It is probable that these different symbiotic com- irradiance and high-temperature specialists (Rowan et al. binations will respond differently to environmental change, 1997; Toller et al. 2001b; Goulet et al. 2005; Reynolds including temperature-induced bleaching events (Rowan et al. 2008; Kemp et al. 2014). Studies have shown that et al. 1997; Kemp et al. 2014). Such information could be specialization to these physiologically stressful conditions useful in designating management recovery units and is facilitated by enhanced photoacclimation and photo- management recovery areas as part of the recovery plan- protective pathways (Robison and Warner 2006; Reynolds ning process for O. faveolata. For example, coral colonies et al. 2008; Takahashi et al. 2009; Kemp et al. 2014). Such with greater proportions of thermally sensitive Symbio- adaptations to high-irradiance environments likely explain dinium types (e.g., C7) could be designated by managers as stratified, intracolonial spatial heterogeneity in O. faveolata ‘‘high risk’’ when bleaching conditions occur (or are from Mesoamerican reefs, but do not explain the absence forecasted) compared to colonies with known high pro- of Symbiodinium type A3 in O. faveolata colonies from the portions of Symbiodinium with greater thermal tolerance Exuma Cays or Florida Keys. (e.g., A3 and D1a; Kemp et al. 2014). Few studies have examined the effects of cold tem- The ability of O. faveolata to simultaneously host up to perature on coral-Symbiodinium physiology and ecological four genetically distinct, co-dominant, Symbiodinium ITS2 responses to cold water perturbation (but see Saxby et al. types in ecologically predictable patterns has not been re- 2003; Thornhill et al. 2008; Kemp et al. 2011). Ecological ported in any other coral genera. Our data support previ- surveys of the Dry Tortugas and Florida Keys documented ously described intracolonial Symbiodinium spatial severe mortality of shallow-water colonies of Porites as- heterogeneity, with the exception of colonies from the treoides (likely associated with type A4 or A4a; LaJe- northern-most reefs sampled (Exuma Cays and Florida unesse 2002; Thornhill et al. 2006a; Kemp et al. 2011), Keys; Fig. 3). Species of the O. annularis ‘‘complex’’ have palmata, and A. cervicornis (both likely associ- commonly been reported to host clade A Symbiodinium ated with type A3; LaJeunesse 2002; Thornhill et al. from reefs in the Caribbean (i.e., Panama, Belize, and 2006b; Baums et al. 2014) immediately following cold Mexico); however, these symbionts are rarely found in fronts that caused seawater temperatures to drop below Orbicella spp. from the Florida Keys and it has never been 16 °C (Davis 1982; Porter et al. 1982). Furthermore, early documented in Orbicella spp. from the Bahamas (Fig. 4; cold tolerance work done by Mayer (1914) found that Rowan and Knowlton 1995; Garren et al. 2006; Thornhill shallow-water colonies of A. cervicornis, A. palmata, and et al. 2006b; Kemp et al. 2008, 2014; Baums et al. 2010). P. astreoides were most sensitive to cold treatments. This Orbicella faveolata individuals were sampled by Baums was corroborated by experimental cold stress experiments et al. (2010) from offshore, forereef, mid-channel, and in- on P. astreoides harboring Symbiodinium type A4a; P. shore reefs throughout the Florida Keys and detected clade astreoides with type A4a experienced greater amounts of A Symbiodinium in only 7 % of 123 sampled colonies. photodamage and were more negatively affected by de- Although clade A Symbiodinium has been reported in Or- creases in temperatures than corals hosting Symbiodinium bicella spp. in the Florida Keys, it is clearly rare. A hy- B1, B5a, or C3 (Kemp et al. 2011). Although the afore- pothesis to potentially explain the rarely reported mentioned symbionts are likely genetically distinct from occurrences of type A3 with Orbicella spp. from the the A3 types associated with Orbicella (see Pinzo´n 2011), Florida Keys and never reported occurrence in Orbicella this body of work indicates that certain Symbiodinium from the Exuma Cays, may be that type A3 Symbiodinium clade A species may be susceptible to the winter low is out-competed by Symbiodinium belonging to clades B temperatures of the Florida Keys and Exuma Cays, leading and C. The higher-latitude reefs of the Upper Florida Keys to displacement by other more thermally robust Symbio- and Exuma Cays, have lower average temperatures, greater dinium spp. within Orbicella spp. seasonal temperature variation, and lower solar incidence The study herein also documents intracolonial and compared to the lower-latitude reefs of Mesoamerica geographically specific patterns of coral-Symbiodinium (Fig. 2; Leichter et al. 2006). Therefore, seasonal cold phylotypes belonging to clades B and C. Symbiodinium water occurrences may cause competitive and phys- belonging to clade C were most prevalent in low-irradiance iological disadvantages to Symbiodinium type A3 com- areas within shallow-water colonies as well as in colonies pared to Orbicella-specific types of clade B or clade C. inhabiting deeper waters (Figs. 3, 4). These findings cor- Symbiodinium belonging to clade A are commonly as- roborate previous studies documenting the dominance of sociated with shallow-water cnidarians in the Caribbean clade C within O. faveolata in low-irradiance areas (Rowan (LaJeunesse 2002). Field and laboratory studies have and Knowlton 1995; Toller et al. 2001a; LaJeunesse 2002; demonstrated many clade A phylotypes to have tolerance Garren et al. 2006). We identified three different ITS2 to high irradiance and high temperatures, leading to the types of clade C with geographic specificity: ITS2 type C3 123 544 Coral Reefs (2015) 34:535–547 from the Florida Keys, C7a from the Bahamas (aka type spp. symbioses actually consist of several regionally en- C12), and C7 from Mexico and Belize. Our data support demic, taxa-specific, symbiotic partnerships that are likely the hypothesis that clade C has diversified throughout the the result of long-term ecological and evolutionarily pro- Caribbean based largely on host identity and geographic cesses. A conspicuous exception to this trend is type D1a location (LaJeunesse 2005; Thornhill et al. 2014). Fine- (S. trenchii), a symbiont species that lacks regional dif- scale chloroplast and microsatellite genetic markers re- ferentiation and associates with many different host taxa cently demonstrated that clade C is highly diverse (see Pettay 2011). The genetic diversity of O. faveolata throughout the Caribbean, including many different host- was not investigated for this study; therefore, we cannot specific lineages that are subdivided by locality (Thornhill ascribe the influence of host diversity to symbiont speci- et al. 2014). This pattern occurred even in clade C sym- ficity. However, O. faveolata population connectivity has bionts harbored by putatively generalist corals in the O. been shown to be genetically well-mixed throughout annularis species complex (Thornhill et al. 2014). Thus, O. Mexico, the Florida Keys, and Puerto Rico (Severance and faveolata is among the most symbiotically flexible coral Karl 2006; Baums et al. 2010), and therefore, the possi- species ever reported, emphasized by its association with bility that host population structure is driving Symbiodini- highly specific lineages of Symbiodinium (i.e., type C7, um diversity patterns that we document is unlikely C7a/C12, and a specific sublineage of type C3) according (Thornhill et al. 2009). to high-resolution chloroplast and microsatellite genetic Whether the ability to host diverse co-dominant markers. Our finding of host-specific Symbiodinium clade assemblages of Symbiodinium is an advantage to host C types in different locations across the Caribbean is en- species requires further investigation. O. faveolata occurs tirely consistent with this pattern. throughout a wide depth range (0–80 m), an unusual trait Symbiodinium belonging to clade B also showed phy- among reef-building corals. The co-occurrence of low and/ logeographic partitioning and were the most abundant or high-irradiance-specific lineages of Symbiodinium spp. phylotypes detected in O. faveolata at all locations. We is likely the result of host–symbiont acclimatization to detected type B1 at all geographic regions and in both high- depth-related niches as has been documented in multiple and low-irradiance habitats within the coral. In shallow- cnidarian taxa (LaJeunesse 2002; Iglesias-Prieto et al. water colonies of O. faveolata from Mexico and Belize, 2004; Kamezaki et al. 2013). Field studies have demon- type B1 was not found in any of the samples; instead type strated differential thermal tolerances of various phylo- B17 was found and appeared to occupy the microhabitat types of Symbiodinium within Orbicella spp. (Rowan et al. otherwise held by B1. Analysis of cp23S revealed further 1997; Kemp et al. 2014) and recovery patterns following diversity of geographically endemic haplotypes within have been shown to cause community ITS2 type B1 at each region (Exuma Cays, Florida Keys, shifts of Symbiodinium associating with Orbicella spp. Puerto Morelos, and Carrie Bow Cay) and allowed genetic (Toller et al. 2001b; Thornhill et al. 2006b; LaJeunesse distinction from S. minutum (LaJeunesse et al. 2012; et al. 2009; Kemp et al. 2014). However, further investi- Thornhill et al. 2013). This host–symbiont distribution gation is needed to understand population and community corroborates previous works using population-level genetic dynamics of coral-Symbiodinium symbioses, functional markers (micro-satellites) that found phylotype B1 to be differences, and potential physiological trade-offs that diverse, having distinct regional and host–symbiont sym- might be associated with hosting genetically distinct, co- bioses (Santos et al. 2004; Thornhill et al. 2009; Finney dominant populations of Symbiodinium. et al. 2010). As ecosystems continue to disappear at This study adds significantly to our understanding of alarming rates, urgency for more conservation is apparent diverse host-Symbiodinium symbiotic associations and needed (Mumby and Steneck 2008; Alvarado-Chaco´n (Thornhill et al. 2009, 2014; LaJeunesse et al. 2010; Pettay and Acosta 2009). O. faveolata is one of seven coral spe- et al. 2011). O. faveolata was one of the first coral-Sym- cies in the Caribbean currently designated ‘‘threatened’’ by biodinium associations to be thoroughly investigated and NOAA (NMFS 2014). This classification should spur de- revealed diverse assemblages of Symbiodinium in eco- velopment of management plans and conservation strate- logically predictable regions of the coral host (Rowan and gies to help avoid eventual loss of species. This study Knowlton 1995; Rowan et al. 1997; Toller et al. 2001a). brings to light the potential management challenges for Worldwide surveys of coral-Symbiodinium symbioses have reef-building coral species that have spatially distinct and revealed hosting diverse co-dominant communities of geographic endemic host-symbiotic combinations. These Symbiodinium in predictable, stratified, assemblages are comparisons are highly relevant, as several studies have unique among symbiotic cnidarians (LaJeunesse et al. shown that corals’ physiological response to environmental 2003, 2010; Stat et al. 2009; Tonk et al. 2013). Further- perturbations can be greatly influenced by different more, this study demonstrates that Symbiodinium-Orbicella assemblages of Symbiodinium (Rowan et al. 1997; 123 Coral Reefs (2015) 34:535–547 545

Berkelmans and van Oppen 2006; Howells et al. 2012; Coffroth MA, Santos SR (2005) Genetic diversity of symbiotic Kemp et al. 2014). Better understanding of local and dinoflagellates in the genus Symbiodinium. Protist 156:19–34 Davis GE (1982) A century of natural change in coral distribution at geographic acclimatization strategies and coral–symbiont the Dry Tortugas - a comparison of reef maps from 1881 and symbioses could provide useful information that may aid in 1976. Bull Mar Sci 32:608–623 conservation applications of reef-building corals. Recog- Edmunds PJ, Elahi R (2007) The demographics of a 15-year decline nizing coral-Symbiodinium regional and geographic pat- in cover of the Caribbean reef coral Montastraea annularis. Ecol Monogr 77:3–18 terns could be used to forecast organismal physiological Finney JC, Pettay DT, Sampayo EM, Warner ME, Oxenford HA, responses to environmental perturbations (e.g., thermal LaJeunesse TC (2010) The relative significance of host-habitat, anomalies), which may provide utility for prioritizing re- depth, and geography on the ecology, endemism, and speciation covery plans. of coral endosymbionts in the genus Symbiodinium. Microb Ecol 60:250–263 Fitt WK, Spero HJ, Halas J, White MW, Porter JW (1993) Recovery Acknowledgments We thank the Florida Keys National Marine of the coral Montastrea annularis in the Florida Keys after the Sanctuary, Caribbean Marine Research Center, Bahamas, the 1987 Caribbean bleaching event. Coral Reefs 12:57–64 Universidad Nacional Auto´noma de Me´xico, Mexico, and Carrie Bow Gardner TA, Cote IM, Gill JA, Grant A, Watkinson AR (2003) Long- Key field station, Belize, for permits and help obtaining samples. We term region-wide declines in Caribbean corals. Science thank Edgar Mancera for providing temperature data for Puerto 301:958–960 Morelos, Mexico. Jennifer McCabe Reynolds, Xavier Hernandez- Garren M, Walsh SM, Caccone A, Knowlton N (2006) Patterns of Pech, Clint Oakley, and Gordon Hendler provided dive support. We association between Symbiodinium and members of the Mon- thank Scott Santos for cp23S primers and Jake Li from Defenders of tastraea annularis species complex on spatial scales ranging Wildlife for helpful discussion concerning the ESA. National Science from within colonies to between geographic regions. Coral Reefs Foundation grant NSF-0137007 awarded to W.K.F. and G.W.S. and 25:503–512 The World Bank Center of Excellence, Coral Reef Targeted Research Goreau TF, Wells JW (1967) Shallow-water of Jamaica: program, supported this work and grant NSF-1015342 provided salary Revised list of species and their vertical distribution range. Bull to D.W.K. during the completion of this work. This is contribution Mar Sci 17:442–453 number 971 of the Caribbean Coral Reef Ecosystems Program Goulet TL, Cook CB, Goulet D (2005) Effect of short-term exposure (CCRE), Smithsonian Institution, supported in part by the Hunterdon to elevated temperatures and light levels on photosynthesis of Oceanographic Research Fund, contribution number 82 of the Key different host-symbiont combinations in the Aiptasia pallidal Largo Marine Research Laboratory, and contribution number 125 of Symbiodinium symbiosis. Limnol Oceanogr 50:1490–1498 the Department of Biological Sciences at Auburn University. Hoegh-Guldberg O, Bruno JF (2010) The impact of climate change on the world’s marine ecosystems. Science 328:1523–1528 Howells EJ, Berkelmans R, van Oppen MJ, Willis BL, Bay LK (2013) Historical thermal regimes define limits to coral acclimatization. Ecology 94:1078–1088 References Howells E, Beltran V, Larsen N, Bay L, Willis B, Van Oppen M (2012) Coral thermal tolerance shaped by local adaptation of Alvarado-Chaco´n EM, Acosta A (2009) Population size-structure of photosymbionts. Nat Clim Chang 2:116–120 the reef-coral Montastraea annularis in two contrasting reefs of Hudson J (1981) Growth rates in Montastraea annularis: A record of a marine protected area in the southern Caribbean Sea. Bull Mar environmental change in Key Largo coral reef marine sanctuary, Sci 85:61–76 Florida. Bull Mar Sci 3:444–459 Anthony K, Marshall PA, Abdulla A, Beeden R, Bergh C, Black R, Huelsenbeck JP, Ronquist F (2001) MRBAYES: Bayesian inference Eakin CM, Game ET, Gooch M, Graham NA (2014) of phylogenetic trees. Bioinformatics 17:754–755 Operationalizing resilience for adaptive coral reef management Hughes TP, Tanner JE (2000) Recruitment failure, life histories, and under global environmental change. Glob Chang Biol long-term decline of Caribbean corals. Ecology 81:2250–2263 Aronson R, Bruckner A, Moore J, Precht B, Weil E (2008) Iglesias-Prieto R, Beltran VH, LaJeunesse TC, Reyes-Bonilla H, Montastraea faveolata. The IUCN Red List of Threatned Thome PE (2004) Different algal symbionts explain the vertical Species Version 2014.1 distribution of dominant reef corals in the eastern Pacific. Proc R Baker AC (2003) Flexibility and specificity in coral-algal symbiosis: Soc Lond B Biol Sci 271:1757–1763 Diversity, ecology, and biogeography of Symbiodinium. Ann Jones CG, Lawton JH, Shachak M (1994) Organisms as ecosystem Rev Ecol Evol Syst 34:661–689 engineers. Oikos 69:373–386 Baums IB, Devlin-Durante MK, LaJeunesse TC (2014) New insights Kamezaki M, Higa M, Hirose M, Suda S, Reimer JD (2013) Different into the dynamics between reef corals and their associated zooxanthellae types in populations of the zoanthid Zoanthus dinoflagellate endosymbionts from population genetic studies. sansibaricus along depth gradients in Okinawa, Japan. Mar Mol Ecol 23:4203–4215 Biodivers 43:61–70 Baums IB, Johnson ME, Devlin-Durante MK, Miller MW (2010) Kemp DW, Fitt WK, Schmidt GW (2008) A microsampling method Host population genetic structure and zooxanthellae diversity of for genotyping coral symbionts. Coral Reefs 27:289–293 two reef-building coral species along the Tract. Kemp DW, Hernandez-Pech X, Iglesias-Prieto R, Fitt WK, Schmidt Coral Reefs 29:835–842 GW (2014) Community dynamics and physiology of Symbio- Berkelmans R, van Oppen MJH (2006) The role of zooxanthellae in dinium spp. before, during, and after a coral bleaching event. the thermal tolerance of corals: a ‘nugget of hope’ for coral reefs Limnol Oceanogr 59:788–797 in an era of climate change. Proc R Soc Lond B Biol Sci Kemp DW, Oakley CA, Thornhill DJ, Newcomb LA, Schmidt GW, 273:2305–2312 Fitt WK (2011) Catastrophic mortality on inshore coral reefs of Clement M, Posada D, Crandall KA (2000) TCS: a computer program the Florida Keys due to severe low temperature stress. Glob to estimate gene genealogies. Mol Ecol 9:1657–1659 Chang Bio 17:3468–3477

123 546 Coral Reefs (2015) 34:535–547

LaJeunesse TC (2002) Diversity and community structure of symbi- Pinzo´n JH, Devlin-Durante MK, Weber MX, Baums IB, LaJeunesse otic dinoflagellates from Caribbean coral reefs. Mar Biol TC (2011) Microsatellite loci for Symbiodinium A3 (S. fitti)a 141:387–400 common algal symbiont among Caribbean Acropora (stony LaJeunesse TC (2005) ‘‘Species’’ radiations of symbiotic dinoflagel- corals) and Indo-Pacific giant clams (Tridacna). Conserv Genet lates in the Atlantic and Indo-Pacific since the Miocene-Pliocene Resour 3:45–47 transition. Mol Bio Evol 22:1158–1158 Pochon X, Gates RD (2010) A new Symbiodinium clade (Dino- LaJeunesse TC, Trench RK (2000) Biogeography of two species of phyceae) from soritid foraminifera in Hawai’i. Mol Phylogenet Symbiodinium (Freudenthal) inhabiting the intertidal sea anemone Evol 56:492–497 Anthopleura elegantissima (Brandt). Biol Bull 199:126–134 Pochon X, Montoya-Burgos JI, Stadelmann B, Pawlowski J (2006) LaJeunesse TC, Thornhill DJ (2011) Improved resolution of reef- Molecular phylogeny, evolutionary rates, and divergence timing coral endosymbiont (Symbiodinium) species diversity, ecology, of the symbiotic dinoflagellate genus Symbiodinium. Mol Phylo- and evolution through psbA non-coding region genotyping. genet Evol 38:20–30 PLoS One 6:e29013 Porter JW, Battey JF, Smith GJ (1982) Perturbation and change in LaJeunesse TC, Parkinson JE, Reimer JD (2012) A genetics based coral reef communities. Proc Natl Acad Sci USA 79:1678–1681 description of Symbiodinium minutum sp. nov. and S. psyg- Reed JK (1985) Deepest distribution of Atlantic hermatypic corals mophilum sp. nov. (Dinophyceae), two dinoflagellates symbiotic discovered in the Bahamas. Proc 5th Int Coral Reef Symp 6:249–254 with . J Phycol 48:1380–1391 Reynolds JM, Bruns BU, Fitt WK, Schmidt GW (2008) Enhanced LaJeunesse TC, Smith RT, Finney J, Oxenford H (2009) Outbreak photoprotection pathways in symbiotic dinoflagellates of shal- and persistence of opportunistic symbiotic dinoflagellates during low-water corals and other cnidarians. Proc Natl Acad Sci USA the 2005 Caribbean mass coral ‘bleaching’ event. Proc R Soc 105:13674–13678 Lond B Biol Sci 276:4139–4148 Robison JD, Warner ME (2006) Differential impacts of photoacclima- LaJeunesse TC, Loh WKW, van Woesik R, Hoegh-Guldberg O, tion and thermal stress on the photobiology of four different Schmidt GW, Fitt WK (2003) Low symbiont diversity in phylotypes of Symbiodinium (Pyrrhophyta). J Phycol 42:568–579 southern corals, relative to those of the Rotjan RD, Dimond JL, Thornhill DJ, Leichter JJ, Helmuth B, Kemp Caribbean. Limnol Oceanogr 48:2046–2054 DW, Lewis SM (2006) Chronic parrotfish grazing impedes coral LaJeunesse TC, Wham DC, Pettay DT, Parkinson JE, Keshavmurthy recovery after bleaching. Coral Reefs 25:361–368 S, Chen CA (2014) Ecologically differentiated stress-tolerant Rowan R, Knowlton N (1995) Intraspecific diversity and ecological zonation endosymbionts in the dinoflagellate genus Symbiodinium (Dino- in coral-algal symbiosis. Proc Natl Acad Sci USA 92:2850–2853 phyceae) Clade D are different species. Phycologia 53:305–319 Rowan R, Knowlton N, Baker A, Jara J (1997) Landscape ecology of LaJeunesse TC, Pettay DT, Sampayo EM, Phongsuwan N, Brown B, algal symbionts creates variation in episodes of coral bleaching. Obura DO, Hoegh-Guldberg O, Fitt WK (2010) Long-standing Nature 388:265–269 environmental conditions, geographic isolation and host-sym- Sampayo EM, Dove S, Lajeunesse TC (2009) Cohesive molecular biont specificity influence the relative ecological dominance and genetic data delineate species diversity in the dinoflagellate genetic diversification of coral endosymbionts in the genus genus Symbiodinium. Mol Ecol 18:500–519 Symbiodinium. J Biogeogr 37:785–800 Santos SR, Shearer TL, Hannes AR, Coffroth MA (2004) Fine-scale Leichter JJ, Helmuth B, Fischer AM (2006) Variation beneath the diversity and specificity in the most prevalent lineage of surface: quantifying complex thermal environments on coral reefs symbiotic dinoflagellates (Symbiodinium, Dinophyceae) of the in the Caribbean, Bahamas and Florida. J Mar Res 64:563–588 Caribbean. Mol Ecol 13:459–469 Mayer A (1914) The effects of temperature on tropical marine Santos SR, Taylor DJ, Kinzie RA 3rd, Hidaka M, Sakai K, Coffroth . Carnegie Inst Wash 183:1–24 MA (2002) Molecular phylogeny of symbiotic dinoflagellates Miller J, Muller E, Rogers C, Waara R, Atkinson A, Whelan KRT, inferred from partial chloroplast large subunit (23S)-rDNA Patterson M, Witcher B (2009) Coral disease following massive sequences. Mol Phylogenet Evol 23:97–111 bleaching in 2005 causes 60% decline in coral cover on reefs in Saxby T, Dennison WC, Hoegh-Guldberg O (2003) Photosynthetic the US Virgin Islands. Coral Reefs 28:925–937 responses of the coral Montipora digitata to cold temperature Mumby PJ, Steneck RS (2008) Coral reef management and conser- stress. Mar Ecol Prog Ser 248:85–97 vation in light of rapidly evolving ecological paradigms. Trends Seutin G, White BN, Boag PT (1991) Preservation of avian blood and Ecol Evol 23:555–563 tissue samples for DNA analyses. Can J Zool 69:82–90 Muscatine L, Porter JW (1977) Reef corals - mutualistic symbioses Severance EG, Karl SA (2006) Contrasting population genetic adapted to nutrient-poor environments. Bioscience 27:454–460 structure of sympatric, mass-spawning Caribbean corals. Mar NMFS (2014) Final listing determinations on proposal to list 66 reef- Biol 150:57–68 building coral species and to reclassify elkhorn and staghorn Stat M, Pochon X, Cowie RO, Gates RD (2009) Specificity in corals, final rule 79, Federal Register 53852–54123 communities of Symbiodinium in corals from Johnston . Nylander J (2004) MrModeltest v2. Program distributed by the Mar Ecol Prog Ser 386:83–96 author. Evolutionary Biology Centre, Uppsala University 2 Szmant AM (1986) Reproductive ecology of Caribbean reef corals. Palumbi SR, Barshis DJ, Traylor-Knowles N, Bay RA (2014) Coral Reefs 5:43–53 Mechanisms of reef coral resistance to future climate change. Takahashi S, Whitney SM, Badger MR (2009) Different thermal Science 344:895–898 sensitivity of the repair of photodamaged photosynthetic Pettay DT (2011) Diversity, stability, and connectivity of Symbio- machinery in cultured Symbiodinium species. Proc Natl Acad dinium populations and various spatial scales. Ph.D. dissertation, Sci USA 106:3237–3242 Pennsylvania State University, pp 1–174 Tchernov D, Gorbunov MY, de Vargas C, Yadav SN, Milligan AJ, Pettay DT, Wham DC, Pinzon JH, Lajeunesse TC (2011) Genotypic Haggblom M, Falkowski PG (2004) Membrane lipids of diversity and spatial–temporal distribution of Symbiodinium symbiotic algae are diagnostic of sensitivity to thermal bleaching clones in an abundant reef coral. Mol Ecol 20:5197–5212 in corals. Proc Natl Acad Sci USA 101:13531–13535 Pinzo´n JH (2011) Phylogenetics, population genetics and ecology to Thornhill DJ, Fitt WK, Schmidt GW (2006a) Highly stable symbioses understand the evolution of coral-algal mutualisms. Ph.D. among western Atlantic brooding corals. Coral Reefs disseration, Pennsylvania State University, pp 1–155 25:515–519

123 Coral Reefs (2015) 34:535–547 547

Thornhill DJ, Xiang Y, Fitt WK, Santos SR (2009) Reef endemism, of four taxa of Symbiodinium on different reefs and across host specificity and temporal stability in populations of symbi- depths. Biol Bull 201:348–359 otic dinoflagellates from two ecologically dominant Caribbean Toller WW, Rowan R, Knowlton N (2001b) Repopulation of corals. PLoS One 4:e6262 zooxanthellae in the Caribbean corals Montastraea annularis Thornhill DJ, Doubleday K, Kemp DW, Santos SR (2010a) Host and M. faveolata following experimental and disease-associated hybridization alters specificity of cnidarian-dinoflagellate asso- bleaching. Biol Bull 201:360–373 ciations. Mar Ecol Prog Ser 420:113–123 Tonk L, Bongaerts P, Sampayo EM, Hoegh-Guldberg O (2013) Thornhill DJ, Kemp DW, Sampayo EM, Schmidt GW (2010b) SymbioGBR: a web-based database of Symbiodinium associated Comparative analyses of amplicon migration behavior in differ- with cnidarian hosts on the Great Barrier Reef. BMC Ecology ing denaturing gradient gel electrophoresis (DGGE) systems. 13:7 Coral Reefs 29:83–91 Warner ME, Berry-Lowe S (2006) Differential xanthophyll cycling Thornhill D, Lewis A, Wham D, LaJeunesse T (2014) Host specialist and photochemical activity in symbiotic dinoflagellates in lineages dominate the adaptive radiation of reef coral endosym- multiple locations of three species of Caribbean coral. J Exp bionts. Evolution 68:352–367 Mar Bio Ecol 339:86–95 Thornhill DJ, LaJeunesse TC, Kemp DW, Fitt WK, Schmidt GW Warner ME, Fitt WK, Schmidt GW (1999) Damage to photosystem II (2006b) Multi-year, seasonal genotypic surveys of coral-algal in symbiotic dinoflagellates: a determinant of coral bleaching. symbioses reveal prevalent stability or post-bleaching reversion. Proc Natl Acad Sci USA 96:8007–8012 Mar Biol 148:711–722 Warner ME, LaJeunesse TC, Robison JD, Thur RM (2006) The Thornhill DJ, Kemp DW, Bruns BU, Fitt WK, Schmidt GW (2008) ecological distribution and comparative photobiology of symbi- Correspondence between cold tolerance and temperate biogeog- otic dinoflagellates from reef corals in Belize: Potential impli- raphy in a western atlantic Symbiodinium (Dinophyta) lineage. cations for coral bleaching. Limnol Oceanogr 51:1887–1897 J Phycol 44:1126–1135 Yellowlees D, Rees TAV, Leggat W (2008) Metabolic interactions Thornhill DJ, Yiang X, Pettay DT, Zhong M, Santos SR (2013) between algal symbionts and invertebrate hosts. Plant Cell Population genetic data of a model symbiotic cnidarian system Environ 31:679–694 reveal remarkable symbiotic specificity and vectored introduc- Zhang Z, Green BR, Cavalier-Smith T (2000) Phylogeny of ultra- tions across ocean basins. Mol Ecol 22:4499–4515 rapidly evolving dinoflagellate chloroplast genes: a possible Toller WW, Rowan R, Knowlton N (2001a) Zooxanthellae of the common origin for sporozoan and dinoflagellate plastids. J Mol Montastraea annularis species complex: patterns of distribution Evol 51:26–40

123