The Pennsylvania State University the Graduate School Eberly College of Science PHYLOGENETICS, POPULATION GENETICS and ECOLOGY T

The Pennsylvania State University

The Graduate School

Eberly College of Science

PHYLOGENETICS, POPULATION GENETICS AND ECOLOGY TO UNDERSTAND THE EVOLUTION OF CORAL-ALGAL MUTUALISMS

A Dissertation in

Biology

Jorge H. Pinzón

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

May 2011 The dissertation of Jorge H. Pinzón was reviewed and approved* by the following:

Andrew Stephenson Professor of Biology and Assistant Department Head for Research Chair of Committee

Todd C. LaJeunesse Assistant Professor of Biology Dissertation Adviser

Iliana Baums Assistant Professor of Biology

David Geiser Professor of Plant Pathology

Douglas Cavener Professor and Head of Biology

*Signatures are on file in the Graduate School.

ABSTRACT

Scleractinian corals form mutualistic associations with dinoflagellates in the genus Symbiodinium. This association has allowed scleractinians, over geological time, to create the framework of one of the most diverse and productive ecosystems on the planet; the coral reefs. While major environmental changes appear to have caused significant changes in partner combinations, there is considerable data to indicate that co- evolutionary processes lead to high host-symbiont specificity and possibly speciation. The study of coral-Symbiodinium associations can be use to understand processes of speciation and the relative importance of biotic (host-specialization) and abiotic factors (isolation by distance, environmental extremes) in the evolution and diversification of both symbiotic partners. The objective of this dissertation is to elucidate the ecological and environmental/geographic factors that are important in the co-evolution of coral-algal mutualisms and how these interactions result in the formation of specific associations and the diversification of corals and algae. The first part of the dissertation is focused on examining the genetic diversity of corals in the genus Pocillopora throughout the Indo- Pacific Ocean and how their symbionts differ with regard to the genetic identity of the animal in combination with geographic location and regional environmental conditions. The second part is designed to assess the influence of host identity and geography on the gene flow among populations of Symbiodinium fitti in the Caribbean. Pocillopora is a common coral genus widely distributed throughout the Indo- Pacific Ocean and the Red Sea. Pocillopora spp. are major reef building corals in the large, isolated Tropical Eastern Pacific (TEP) region. Shallow coral communities in the TEP extend from the Gulf of California to Ecuador. Approximately seven cosmopolitan (P. damicornis, P. verrucosa, P. meandrina, P. ligulata, P. woodjonesi. P. eydouxi and P. capitata) and two endemic (P. effusus and P. inflata) species of Pocillopora recognized based on morphology are found in the TEP. Many species within the genus are phenotypically plastic to the extent that colony and branch morphology between different species may overlap creating uncertainty in their taxonomy. To examine this diversity more carefully (Chapter One), genetic analyses using nuclear (ITS2) and mitochondrial (ORF) DNA sequence data, and microsatellite markers were employed to investigate species boundaries. Colonies were collected at several locations over a broad latitudinal range throughout the TEP (Sea of Cortez, Banderas Bay and Oxaca in Mexico, Clipperton Atoll, Panama and Galapagos). The results identified only three distinct genetic entities of Pocillopora that do not corresponded to morphology. The most common genetic type, designated Type 1, was found throughout the entire region. A second type (Type 3) was identified in equatorial regions (Panama and Galapagos) and a third (Type 2), most likely a rare endemic, occurred only at the isolated Clipperton atoll. Each of these genetic “types” associated with different species of symbiont: Type 1 associated with Symbiodinium C1b-c and S. glynni, Type 3 with S. C1d-c and Type 2 with S. C1ee. Genetic, ecological and biogeographic analyses suggested the diversity of Pocillopora in the TEP is less than previously defined exclusively on macromorphological features (i.e. colony size and branching patterns). Based on these

iii findings, it appears that two of the three species are potentially susceptible to climate change, as they do not associate with the thermally tolerant S. glynni. Contradictions between morphologically defined species and their genetic make up in Pocillopora from the TEP indicated the need for a major revision of the taxonomic status of Pocillopora spp. In an attempt to assess the diversity of the genus Pocillopora (Chapter Two), collection and analyses of samples from various locations in the Pacific (Hawaii, Australia, Thailand, Palau and Taiwan) and Indian Ocean (Tanzania) were performed. The results indicate Pocillopora is composed of at least eight well-defined species with little or no correspondence to morphological features. Inconsistencies between morphology and genetics led to revise the taxonomy for the Pocillopora. Accurate species delineations are important to understand the biology, ecology and evolution of life as well as to improve conservation. An example of this is presented by studying the clonal and reproductive structure of Pocillopora type 1 from two close coral communities in the Gulf of California, off La Paz, Mexico (Chapter Three). Microsatellite loci data analyses revealed that although gene flow is present, P. type 1 has a different reproductive structure on each reef. The exposed reefs, with more high action and a wider platform, is dominated by sexual reproduction while the enclosed, more protected area is highly clonal. Although comparisons with previous studies are not possible (i.e. different methodologies or molecular markers), both reproductive strategies have been found in Pocillopora and none seems related to marginality of the populations. Host selection leads to the diversification of specialized symbiont populations and can be an important step in sympatric speciation. Species radiations in at least two identified Symbiodinium clades (B and C) suggested the diversity of the group alternates between low (mostly generalist) and high (mostly specialist) number of species. Symbiodinium clade A showed a similar pattern of species diversification. Analysis of 10 microsatellite loci showed structured populations of Symbiodinium fitti harbored by different coral hosts (Chapter four). Populations found in Stephanocoenia intersepta were genetically different from those harbored by Acropora species. Among Acropora spp, symbiont populations were not completely isolated from each other with some clonal lines present in both A. palmata and A. cervicornis. Analyses of populations harbored by the same host at different locations showed geographical isolation between S. fitti populations from different locations (Colombia, Panama and Mexico). Symbiodinium speciation seems to be driven by a combination of host selection and environmental/geographical isolation. The conclusions of this research are: (1) information from the dinofalgellate algae can be use to delineate coral species in combination with phylogenetic and population genetic analyses of the host. (2) Coral-algal associations change over geographical ranges suggesting a geographical mosaic of co-evolution with host populations widely distributed and symbiont populations more isolated. (3) Sympatric coral species effectively reduce gene flow between populations of Symbiodinium, and (4) Host identity is a major factor in the diversification of Symbiodinium species.

iv TABLE OF CONTENTS

LIST OF TABLES...... vii

LIST OF FIGURES...... viii

ACKNOWLEDGEMENTS ...... xiii

INTRODUCTION ...... 1

CHAPTER 1: SPECIES DELIMITATION OF COMMON REEF CORALS IN THE GENUS POCILLOPORA USING NUCLEOTIDE SEQUENCE PHYLOGENIES, POPULATION GENETICS, AND SYMBIOSIS ECOLOGY ...... 5 Abstract ...... 5 Introduction...... 6 Methods ...... 10 Results...... 18 Discussion ...... 27 References ...... 40

CHAPTER 2: A GENETIC ASSESSMENT OF SPECIES BOUNDARIES IN THE CORAL GENUS POCILLOPORA ALTERS KNOWLEDGE OF BIODIVERSITY, BIOGEOGRAPHIC PATTERNS, AND THE EVOLUTION OF REPRODUCTIVE STRATEGIES...... 47 Abstract ...... 47 Methods ...... 52 Results...... 59 Discussion ...... 70 References ...... 75

CHAPTER 3: HIGH VARIANCE IN CLONALITY BETWEEN LOCAL POPULATIONS OF THE ECOLOGICALLY DOMINANT POCILLOPORA IN THE GULF OF CALIFORNIA ...... 82 Abstract ...... 82 Introduction...... 83 Materials and Methods...... 85 Results...... 90 Discussion ...... 98 References ...... 105

CHAPTER 4: HOST SPECIALIZATION MORE THAN GEOGRAPHIC ISOLATION DRIVES THE SPECIATION OF CORAL ENDOSYMBIONTS IN THE GENUS SYMBIODINIUM (ZOOXANTHELLAE) ...... 111 Abstract ...... 111 Introduction...... 112 Materials and Methods...... 115

v Results...... 120 Discussion ...... 129 References ...... 137

CONCLUSIONS...... 143

BIBLIOGRAPHY...... 145

APPENDICES...... 150 Appendix A. Allelic frequencies per locus for Pocillopora types I, II, and III and across types (Overall). Samples size, number of unique genotypes and number private alleles (bold) are shown by type and overall...... 150 Appendix B. List of Pocillopora species found in Veron 2000, the * indicates species with restricted  distributions or endemics. Species included in the list by Cairns, 1999, these may correspond to the most commonly known names of the genus...... 152 Appendix C. Pocillopora spp samples used for morphometric analysis and accession numbers from the Museo de Historia Natural de la U. Autónoma de Baja California Sur (MHNUABCS)...... 153 Appendix D. Phylogenetic reconstruction showing the relations between populations of Symbiodinium fitti harbored by four coral hosts (Acropora palmata, A. cervicornis, M. annularis and Stephanocoenia intersepta) in different geographic locations through the Caribbean (Triangles = Mexico, Squares = Panama and Circles = Colombia; Ctg = Cartagena and SM = Santa Marta). The tree was build with a unweighted pair group method with arithmetic mean (UPGMA) based on the Chord-distance (Cavalli- Sforza and Edwards, 1967), using ten microsatellite loci genotypes (n = 94)...... 154

vi LIST OF TABLES

Table 1. Morpho-species diversity in the genus Pocillopora from the Tropical Eastern Pacific, as reported by regional diversity inventories and coral atlases...... 8 Table 2. Characteristics and amplification conditions of the microsatellite loci used to analyze the population structure of eastern Pacific Pocillopora spp. (L = Locus name, N = Number of alleles, Ho = observed heterozygosity, F = frequency, R = Range and Ta = annealing temperature)...... 17 Table 3. Modes of reproduction in four morphologically identified Pocillopora species in different locations from the Indian Ocean, Indo-Pacific, and Tropical Eastern Pacific...... 72 Table 3. Estimates of the indices of clonal structure for each of the circular plots samples in two reefs off La Paz, Gulf of California ...... 91 Table 4. Number of ramets and average distance (m) between ramets for each clonal line found in two reefs of La Paz, Mexico, Punta Galeras (PG) and Isla Gaviota (IG). Clones are organized to match the symbols from left to right in the legend in figure 19...... 94 Table 5. List of loci with significant null alleles on each reef population and overall, 2 inbreeding coefficient (g ) as calculated by RMES and estimated Rst between reefs (IG and PG) with (Rst*) and without (Rst ) corrected frequencies for null alleles. ...96 Table 6 Total number of colonies genotyped (n), number of unique multilocus genotypes (MLG - G) and richness (R) of S. fitti populations (defined by host and location) from four coral hosts and three locations thought the Caribbean. Only two populations (*) showed high levels of clonality (i.e. large numbers of individuals with the same MLG)...... 122 Table 7. Levels of differentiation between populations of S. fitti delimited by location, host and through the Caribbean (G = number of unique multilocus genotypes – MLG, K = optimal number of clusters/populations as suggested by Structure and Φst = Phi Statistics – calculated with GenAlex, * = all Φst calculations were significant at the p = 0.01 level, Ctg = Cartagena) Give variance for phi st values...... 125

vii LIST OF FIGURES

Figure 1. Sampling locations for Pocillopora in the Tropical Eastern Pacific (Gulf of California = GoC; Banderas Bay = BB, Oaxaca = OX, Revillagigedo Islands = REV, Gulf of Panama = PA, Clipperton Atoll = CLIP, Galapagos Islands = GAL).11 Figure 2. Variation in morphology among type 1 colonies in the Gulf of California including morphospecies (a) P. capitata, (b) P. damicornis (above) and P. verrucosa (below), (c) P.verrucosa, (d) P. meandrina, (e) P. capitata, and (f) P. elegans...... 12 Figure 3. Phylogenetic reconstructions of TEP Pocillopora based on (a) the mitochondrial open reading frame (ORF; n = 301) and (b) the ITS2 region (n = 50). Each lineage was designated type 1 (medium grey), type 2 (dark grey) and type 3 (light grey). Topologies are drawn to the same scale and circle size is proportional to the number of samples possessing the same sequence (with the exception of type 1 ORF, n = 257). Numbers on the branches correspond to the support value of each branch after 1000 bootstrap replicates (Maximum likelihood / Neighbor Joining / Maximum Parsimony / Posterior probability MrBayes)...... 19 Figure 4. Estimated population structure using Bayesian clustering (a) and UPGMA phylogenetic reconstruction (b) of 342 MLGs (based on 7 microsatellite loci) of TEP Pocillopora spp. The separate populations identified correspond to the genetic groupings defined by the ORF and ITS2 sequences. Red dots mark 7 putative hybrids of type 1 and type 3 identified by the software NewHybrids (abbreviations of geographic locations correspond to those in figure 1 and the colours blue, green, and yellow coincide with type 1, 2 and 3, respectively). The plot figure shown for a given K is based on the composite probabilities of 5 independent statistical runs at that K...... 21 Figure 5. Estimated population structure of type 1 colonies sampled throughout the far Tropical Eastern Pacific depicting limited population subdivision over a latitudinal gradient. The plot figure shown for a given K is based on the composite probabilities of 5 independent statistical runs at that K. Groupings with the high statistical support (black arrowhead) K=2 appear to distinguishes "high latitude" and "low latitude" populations of type 1...... 22 Figure 6. Genetic structure of Pocillopora populations containing (a) types 1 and 2 from the Clipperton Atoll and (b) types 1 and 3 from the Gulf of Panama. Despite being sympatric, each population exhibited strong genetic differentiation corresponding to the mitochondrial ORF and nuclear ITS2 lineages present. There was no evidence of mixed genotypes to indicate hybridization ...... 23 Figure 7. Principal coordinate analysis (PCoA) (a) and Structurama (b) analyses of the multilocus genotypes (n = 342) obtained from samples collected in the TEP. Colors correspond to three genetic groupings (blue = type 1, green = type 2, yellow = type 3). The branches of three putative “F-2 hybrids” are in black...... 24 Figure 8. (a) Representative DGGE ITS2 fingerprints of Clade C Symbiodinium found in symbiosis with TEP Pocillopora. (b) Phylogenetic relationships between the three

viii Clade C Symbiodinium (C1b-c, C1ee, and C1d) based on the hyper-variable non- coding region of the psbA minicircle. The location origins of each sequence are provided showing distinct phylogeographic patterning. Symbiodinium glynni (D1) was found to associate with only type 1 colonies (not shown). Numbers above each branch indicate the number of sequence changes (including indels) and below are bootstrap support values based on 1000 replicates...... 26 Figure 9. Correspondence between colony morphology and genetic identity. (a) Pie charts representing the discordance between morphology and genetics among types 1 (blue), 2 (green) and 3 (yellow). (b) Morphological variability and similarity among colonies identified as types 1 and 3 in the Gulf of Panama, respectively...... 28 Figure 10. Phylogenetic similarity based on maximum parsimony of ITS2 sequences originating from bacterial cloning of PCR amplifications published by Combosch et al. (2008) in their analysis of TEP Pocillopora with sequences generated by direct sequencing reported by this study (see Fig. 3b). The sequences corresponding to ITS2 clade III (Combosch et al. 2008) are encompassed within the dashed box. ....32 Figure 11. Tentative geographic map of types 1 (blue), 2 (green), and 3 (yellow) in relation to biogeographic sub-provinces defined by Glynn and Ault (2000) (NEP = Northern Province, EEP = Equatorial province and IEP = Island Province). The star indicates the location of Malpelo Island in the IEP...... 38 Figure 12. Locations of the 15 sampling sites covering part of the geographical distribution of Pocillopora spp in the Indian Ocean (IO – Tanzania TAN), the Indo Pacific (IP - Western Australia WA, Taiwan TAI, Palau PAL, Heron Island (HI), and Lizard Island LZI), the Central Pacific (CP - Hawaii HAW) and the Tropical Eastern Pacific (TEP – Mexico; La Paz - GoC, Revigagilledo Islands - REV, Banderas Bay - BB, Oaxaca – OX; Clipperton Atoll CLIP, Panama PA and Galapagos GAL)...... 53 Figure 13. Phylogenetic reconstructions with a most parsimonious analysis (consensus tree) of the mtDNA ORF and rDNA ITS2 regions from Pocillopora species collected in different reefs in the Indian and Pacific Oceans. Dotted lines are encircling the proposed genetic types (numbered in each trees) in the genus after branch support was tested, with black branches highly supported (> 66 %) in all analyses (Maximum likelihood, Neighbor Joining, Maximum Parsimony and Bayesian - MrBayes), while gray branches showed low support. Circles in the nodes and leaves represent sequences obtained here; dark gray circles in the ITS2 tree are groups showing incongruence between phylogenies. Pocillopora 1mea, 1eyu, 3 and 5 showed wide distributions and P. 2, 4, 5, and 6 might be brooders contrary to the other species (See Discussion)...... 60 Figure 14. Clustering of multilocus genotype data from 630 individual colonies of Pocillopora spp determined by Structure (V.3.2) assuming location prior, admixture and correlation between loci. Individual colonies are organized by ORF/ITS2 type (top) and location (bottom) and the y-axis represents the probability of membership. Optimal number of clusters after Evanno et. al. (2005) was K = 2, but further analyses at higher values of K, separate all the 8 types in distinct groups...... 62

ix Figure 15. Distance base phylogenetic reconstruction (a) and Bayesian tree build with Structurama (b). Both trees show relationships between 630 unique multilocus genotypes of Pocillopora from different types and geographic locations...... 63 Figure 16. Sympatric Bayesian analysis performed with Structure. Analyses resulted showed the formation of distinct clusters corresponding to the genetic delineation of the types in the genus Pocillopora. Only locations with different types and enough samples per type are included. Colors correspond to the ORF/ITS2 types as depicted in the legend. Y-axes represent the probability of membership...... 65 Figure 17. Geographic distribution of 8 Pocillopora types as delineated by genetic markers and ecological associations. Triangles demark locations were each type was collected while the dotted lines show the most likely geographical range of each type. Circles show other sampled locations where types were not found in this study...... 68 Figure 18. Morphometrical analysis of Pocillopora spp in Hawaii and the Gulf of Mexico. (a) Picture depicting the distances measured on five randomly selected calices, located approximately 2cm from the tip of the branch. (b) Kruskall-Wallis analyses and significant values (in parentheses) for three of data sets. (c) Diameters (in mm) of eight different groups of colonies from the GoC (light gray) and Hawaii (dark gray) bars correspond to the standard deviation, and letters indicate significantly different groups (p < 0.05)...... 69 Figure 12. Geographical location of the sampling sites showing the circular plots in Punta Galeras (a) and Isla Gaviota (b) off La Paz, Mexico, in the Gulf of California (GoC). The exact location of the colonies and whether or not they represent a single ramet (squares) or several ramets (i.e. clones and hexagons) of a genet (circles) is also presented. Each of the black squares represents a unique multilocus genotype (MLG) while circles with the same code represent clonal lines. Clone symbols are organized from left to right to correspond to the alphabetical order in table 5...... 92 Figure 13. Spatial autocorrelation analyses of the distribution of Pocillopora genotypes in each circular plot off La Paz. Dotted arrows indicate the 97.5% significance interval. Arrows in Punta Galeras 1 (PG1) and 2 (PG2) indicate the only two instances when the distributions deviate from random...... 95 Figure 14. Bayesian assignment of individual genotypes (a) detected in Pocillopora type I from La Paz, Baja California. Phylogenetic tree bases Nei’s unbias distance (b) of the same samples. Repeated multilocus genotypes were removed for both analyses...... 97 Figure 15. Patterns of sexual (or clonal) structure of various population of Pocillopora spp through the Indo-Pacific Ocean. Colors inside the squares represent sexual populations (white) versus clonal populations (gray), cross bars indicate locations where differences between the populations were found (i.e. low gene flow between reefs) and the numbers on top the approximate range (km) of each study. *In two locations in NWA (Pilgramunna and Coral Bay) sexual reproduction was dominant while the other reefs showed clonal populations of Pocillopora. +At One tree Island reef, P. damicornis colonies release large amounts of asexually produced larvae,

x however the population structure is that of a sexual population (Ayre, Miller, 2004). (a. South Africa / Mozambique (Ridgway et al., 2001) b. Northwestern Australia - NWA (Whitaker, 2006)c. South Western Australia - SWA (Stoddart, 1984b; Stoddart, 1984c) d. Southern Taiwan (Yeoh, Dai, 2010) e. Ryukyu Archipelago, Japan (Adjeroud, Tsuchiya, 1999) f. Great Barrier Reef - GBR (Ayre et al., 1997) g. One Tree Island (Benzie et al., 1995) h. Lord Howe Island (Miller, Ayre, 2004) i. Kaneobe Bay (Stoddart, 1984a) j. La Paz Mexico (This study)) ...... 101 Figure 16. Figure 1. Map of the Caribbean sea showing the location of the sampling sites in Mexico (Puerto Morelos), Panama (Bocas del Toro) and Colombia (Cartagena and Santa Marta)...... 116 Figure 17. Chromatograms of two loci (A3Sym_027 and A3Sym_032) amplifications using various ratios (1:0, 1:9 and 1:1) of DNA mixtures obtained from samples of Symbiodinium fitti isolated from two different hosts (Genotype A from Acropora palmata and Genotype B from Stephanocoenia intersepta). The emergence of a second peak in the 1:9 reactions indicates that genotypes representing more than 10% of a mixture should be detectable with these primer sets. The same results were observed with the other primer sets. The 114bp peak in locus A3Sym_27 corresponds to a randomly amplified product whose sequence did not matched that of the originally isolated locus and did not have a repeat region...... 121 Figure 18 Principal coordinate analysis (PCoA) of Symbiodinium fitti associated with the scleractinian coral Stephanocoenia spp in three locations in the Caribbean. The subdivisions of the populations from Mexico and Colombia may correspond to different microenvironments and/or host species respectively (see text for an explanation)...... 123 Figure 19. Genetic clustering analyses of Symbiodinium fitti individual multilocus genotypes showed correspondence with geographic location of the populations (Colombia, Panama, Mexico) associated with the same host species (a. Acropora palmata Φpt = 0.52, p = 0.01, n = 25 and b. Stephanocoenia intersepta Φpt = 0.35, p = 0.01, n = 36) ...... 124 Figure 20. Structure plots showing distinctions between populations of S. fitti harbored by sympatric coral host in three locations in the Caribbean (a. Mexico K = 4, Φpt = 0.57, p = 0.01, n = 27, Panama K = 2, Φpt = 0.50, p = 0.01, n = 39 and Cartagena, Colombia K = 2, Φpt = 0.36, p = 0.01; n = 28)...... 126 Figure 21. Bayesian population structure analysis of the populations of Symbiodinium fitti harbored by different hosts (a. Acropora palmata, b. A. cervicornis, c. Montastraea annularis, and d. Stephanocoenia spp) and geographic locations (Ctg = Cartagena). The influence of the host (at least genera, Acropora/Montastraea vs. Stephanocoenia) in population differentiation appears stronger than that of geographic isolation, as suggested by the optimal number of populations in the data set (K = 2, Φpt = 0.36, p = 0.01, n = 94)...... 127 Figure 22. Networks showing the associations between populations of S. fitti across host and locations (a) and for Stephanocoenia spp. (b) and A. palmata (c) associated populations. Analyses included exclusively unique multilocus genotypes (G)...... 129

xi Figure 23. Theoretical models showing different probable scenarios (a to d) of the evolution of Symbiodinium species, as influenced by the host and/or geographic isolation...... 134

xii ACKNOWLEDGEMENTS

I would like to express my gratitude to my advisor Todd LaJeunesse, for his unconditional support and advice, before and after being part of his laboratory at Florida

International University and Penn State. I also appreciate the advice and orientation given by the members of my graduate committee (A. Stephenson, I. Baums and D. Geiser) who provided suggestions, advice and support during my PhD. Various members of the graduate faculty and friends at FIU and Penn State (T. Collins, W. Goldberg, M.

Donnelly, M. Ostentoski, S. Raju, L. Remily. T. Pettay, E. Sampayo, T. Ridgway) who provided support and advice during the early stages of my Ph.D. This research would not have been possible without the support and cooperation of many collaborators including:

H. Reyes-Bonilla, M.A. Medina, F.J. Vergara and A. López-Pérez (Biología Marina - U.

Autónoma de Baja California Sur) for sample collections and logistics in Mexico. P.

Medina-Rosas (U. of Guadalajara) and Kristie Kaiser provided samples from Clipperton

Atoll and Revillagegado Islands. M. Torchin and C. Schloeder (Smithsonian Tropical

Research Institute-Panama) facilitated collections in Panama. I. Baums and A. Baker provided samples from the Galapagos Islands. J.A. Sanchez, C. Castro and Olga L.

Torres (U. de los Andes) provided field and laboratory support in Colombia. Z. H.

Forsman, T. Hopkins, W. Castro and D. Ruiz provided constructive evaluations on early versions of this dissertation. I also appreciate the help and advice given by the personnel from the DNA facilities at FIU (P. Sharp) and Penn State (D. Grove and A. Price). This research was funded in part by Florida International University, The Pennsylvania State

University, the National Science Foundation (IOB 544854), and an Alfred P. Sloan

Scholarship.

xiii

DEDICATION

It is hard to dedicate this work, especially when I feel surrounded by many people that have contributed one way or the other to my success. My son and his brother, as well as my wife deserve a bigger dedication than this, they have cope with my roller coasters and adjust to my multiple trips – including moving to State College from Miami (FL) in the middle of the winter!!!! To them this is just the first step; there is a long, but hopefully pleasant road ahead. Thanks for your love and patience. I love you!!!

Life changes as we grow, but if we do not pay attention, it is unperceivable. During this journey I became a father, a US citizen, and matured my relationships with my family. I also lost an important person in my life; unfortunately I did not know it until he was gone. Dad whenever you are this is for you too…. Thanks for your help, company and support. I love you!!!

Last but not least to my mother. Mom your support has been very important in my life. I believe this is a dream come true for you too. …. Sorry it took me this long…. I love you!!!

xiv INTRODUCTION

Symbiosis, the association and living together of different species of organism (Douglas,

1994; Trench, 1997), is a universal factor in biology (Harley, Smith, 1983; Margulis,

Fester, 1991; Zook, 1998) and a cornerstone event in the evolution of species and ecosystems (Margulis, 1993). Mutualistic symbioses have largely been seen as a cooperative endeavor between partners (Bronstein, 1994), but they can also be the result of reciprocal exploitation (Axelrod, Hamilton, 1981). In a mutualism, regardless of the type of interaction between partners (i.e. cooperation or exploitation), both organisms/species benefit from the association (Margulis, Fester, 1991). A significant benefit is the development or acquisition of novel characteristics or traits, such as photosynthesis, nitrogen fixation and bioluminescence (Douglas, 1994; Venn et al., 2008;

Wernegreen, 2004). These novel qualities expand the diversity, niche and metabolic capabilities of the participating organisms (Cavanaugh, 1994; Margulis, Fester, 1991) and play important roles in the dynamics of the ecosystems (Smith, 2001). Unfortunately there are many gaps in our understanding of these associations; one example is how the symbiont may influence the evolution and diversity of the host (Smith, 2001) and vice- versa.

Marine symbioses are commonly established between invertebrates (i.e. sponges, tunicates, bryozoans and scleractinian corals) and various types of microorganisms (i.e. bacteria, archaea, cyanobacteria, and eukaryotic algae) (Dunlap, Dunlap, 2007). In some cases, the mass of the symbiont can account for up to 40% of that of the whole holobiont

(Friedrich et al., 1999). The relationship increases the productivity of entire ecosystems

1 by providing carbon fixation in (Muscatine, 1990) and fixating nitrogen in seagrass beds

(Johnson et al., 2002).

Coral reefs represent unique marine ecosystems that provide habitat for a large diversity of organisms and protect coastal environments and economic resources for millions of humans (Spalding et al., 2001). The key factor in the structure of coral reefs is the mutualism between scleractinian corals and the dinoflagellate algae in the genus

Symbiodinium. In this relationship, the coral provides an ideal environment and inorganic nutrients (i.e. phosphate and ammonium) to the algae, which returns 95% of its photosynthetic products (i.e. peptides, amino acids complex carbohydrates and sugars) to the coral (Muscatine, 1967; Muscatine, Porter, 1977; Pearse, Muscatine, 1971; Stat et al.,

2006; Sutton, Hoegh-Guldberg, 1990; Swanson, Hoegh-Guldberg, 1998; Trench, 1979).

The most important outcome of this relationship is the reef’s structural framework, built by the corals using the energy acquired by the algae.

In recent decades coral reefs have been in decline due to a variety of environmental and factors and human enrichment (Borger, 2005; Borger, Steiner, 2005; Bruno et al.,

2007; Bruno et al., 2003; Castro, Pires, 2001; Coles, 2003; Cooney et al., 2002; Hughes,

1994; Hughes et al., 2003; Porter et al., 2001; Porter et al., 1999; Porter, Meier, 1992;

Santavy et al., 1999). Increase in sea surface temperatures have a direct impact on coral- algal associations, having been linked to bleaching, an event in which the coral loses coloration due to the release of the symbiotic algae or its pigments (Douglas, 2003;

Glynn, 1993; Glynn, 2001; Hoegh-Guldberg, Smith, 1989; LaJeunesse et al., 2007). This major disruption of the coral-algal association can result in partial or total mortality of coral colonies and the collapse of entire reef ecosystems (Glynn, 1993; Hughes et al.,

2 2003). It is important to understand the evolutionary biology of coral symbioses if we wish to understand how they will respond to environmental factors such as climate change.

There is uncertainty about the extent to which host-specificity occurs between coral species and their photosynthetic symbionts. With over one thousand described species, scleractinian corals are extremely diverse (Cairns, 1999) and it is only now being recognized that their Symbiodinium associates are also species-rich, grouping into eight evolutionarily divergent clades (Baker, 2003; Coffroth, Santos, 2005; LaJeunesse, 2005;

Sampayo et al., 2009; Stat et al., 2006), meaning there are many possible coral-algal species combinations. The number of different coral-algae combinations can be an indication of the plasticity or specificity in the formation of the association, the environmental adaptability of the coral-algae combinations, and the co-evolution of both partners. For example, it has been proposed that corals may benefit from bleaching in order to acquire stress tolerant symbionts (Buddemeier, Fautin, 1993; Buddemeier,

Smith, 1999; Coles, Brown, 2003), an indication of low specificity and no co- evolutionary patterns. Nevertheless, numerous coral species form and maintain partnerships with a limited number of Symbiodinium species (LaJeunesse, 2005). A multiyear and multispecies survey of the coral-algae symbiosis in the Caribbean showed that most associations experienced almost no seasonal or year to year variation in the dominant symbiont (Thornhill et al., 2006a; Thornhill et al., 2006b). The specificity and stability of this mutualism are clear indications of co-evolutionary processes influencing the formation and stability of coral-algal symbioses.

3 Traditionally the classification and identification of scleractinian corals has been based on colony morphology (Benzoni et al., 2010; Budd et al., 2010; Weil, 1992).

Corals show a high degree of phenotypic plasticity and closely related species tend to have overlapping morphologies (Amaral, 1994; Ayre et al., 1991; Bruno, Edmunds,

1997; Budd, 1988; Budd et al., 1994; Todd, 2008; Todd et al., 2001) (e.g. Acroporids and

Pocilloporids in the Indo Pacific and Poritids in the Caribbean). In the 1990’s, studies of coral taxonomy lead to the delineation of closely related species through multivariate studies that included morphological, biological, ecological and in some cases molecular approaches (Weil, 1992; Weil, Knowlton, 1994). More recently, the status of several groups including families have been debated based on the use of molecular markers at large geographical scales and high phylogenetic (i.e. Family) levels (Fukami et al., 2004;

Fukami et al., 2008; Romano, Palumbi, 1996). Morphology and genetics have agreed in some groups (Budd, Klaus, 2001; Stefani et al., 2007; Stobart, 2000), but in others morphological features were found to be less informative and in disagreement with molecular markers (Forsman et al., 2009; Huang et al., 2009).

The purpose of this project is to examine the genetic diversity of both the coral and the algal symbionts at a fine scale to determine patterns of co-evolution between both partners. This project will focus on two important and widely distributed coral-algal systems: Pocillopora spp. from the Tropical Eastern Pacific (TEP) and Symbiodinium fitti-coral (Acropora palmata, A. cervicornis and Stephanocoenia intersepta) from the

Caribbean. In the Pocillopora system the coral associates with two distinct types of

Symbiodinium that are transmitted vertically to new generations. In the S. fitti-coral association, the symbiont is acquired from the environment (i.e. horizontal transmission).

4 CHAPTER 1: SPECIES DELIMITATION OF COMMON REEF CORALS IN THE GENUS POCILLOPORA USING NUCLEOTIDE SEQUENCE PHYLOGENIES, POPULATION GENETICS, AND SYMBIOSIS ECOLOGY

*Published in Molecular Ecology (Pinzón JH, LaJeunesse TC (2010) Species delimitation of common reef corals in the genus Pocillopora using nucleotide sequence phylogenies, population genetics, and symbiosis ecology. Molecular Ecology, 20(2):311- 325)

Abstract

Stony corals in the genus Pocillopora are among the most common and widely distributed of Indo-Pacific corals and, as such, are often the subject of physiological and ecological research. In the far Tropical Eastern Pacific (TEP), they are major constituents of shallow coral communities, exhibiting considerable variability in colony shape and branch morphology and marked differences in response to thermal stress. Numerous intermediates occur between morphospecies that may relate to extensive hybridization. The diversity of the Pocillopora genus in the TEP was analyzed genetically using nuclear ribosomal (ITS2) and mitochondrial (ORF) sequences, and population genetic markers (7 microsatellite loci). The resident dinoflagellate endosymbiont (Symbiodinium sp.) in each sample was also characterized using sequences of the internal transcribed spacer 2 (ITS2) rDNA and the non-coding region of the chloroplast psbA minicircle. From these analyses, three symbiotically distinct, reproductively isolated, non-hybridizing, evolutionarily divergent animal lineages were identified. Designated types 1, 2, and 3, these groupings were incongruent with traditional morphospecies’ classification. Type 1 was abundant and widespread throughout the TEP; type 2 was restricted to the Clipperton Atoll; and type 3 was found only in Panama and the Galapagos Islands. Each type harboured a different Symbiodinium “species lineage” in Clade C, and only type 1 associated with the “stress- tolerant” Symbiodinium glynni (D1). The accurate delineation of species and implementation of a proper taxonomy may profoundly improve our assessment of Pocillopora’s reproductive biology, biogeographic distributions, and resilience to climate warming, information that must be considered when planning for the conservation of reef corals.

5 Introduction

Modern approaches to ecology, evolution, and systematics increasingly rely on the power of molecular genetic analyses. Acquisitions of genetic data have improved our understanding of many ecological interactions and evolutionary relationships, and have frequently resulted in substantial changes in the systematics of organism lineages.

Indeed, molecular genetic data have supported and challenged the foundations of numerous taxonomic hierarchies based originally on traditional morphological analyses

(Andreakis et al., 2004; Borrero-Pérez et al., 2009; Knowlton et al., 1992; Radashevsky,

Pankova, 2006; Ruiz-Sanchez, Sosa, 2010; Suatoni et al., 2006; Verbruggen et al., 2007).

Cnidarians in the order Scleractinia (i.e. stony corals) comprise an ecologically important group whose systematics requires major revision. The evolutionary relationships among and within morphologically defined families and genera do not always agree with nucleotide sequence phylogenies (Forsman et al., 2009; Fukami et al., 2004; Fukami et al., 2008; Medina et al., 1999; Romano, Palumbi, 1996). For example, it is now evident that extant reef-building corals comprise two divergent clades, the “robust” and

“complex” corals (Romano, Cairns, 2000; Romano, Palumbi, 1996; Romano, Palumbi,

1997). Masked by evolutionary convergence, family groupings that comprise these ancient clades contain representatives that are para- and polyphyletic (Fukami et al.,

2008). For example, the Atlantic faviids and mussiids are genetically more similar to each other than faviids and mussiids in the Pacific, indicating that Caribbean and Indo-

Pacific coral communities comprise evolutionarily distinct assemblages (Fukami et al.,

2004). Clearly the revision of scleractinian systematics has important implications for deducing the ecology and evolution of these organisms.

6 Species delimitation within various coral genera faces similar challenges to those found at higher taxonomic ranks. Irradiance levels, wave exposure, sedimentation, nutrient availability, and competition may significantly influence colony morphology, calice size, and polyp shape (Chornesky, 1983; Lesser et al., 1994; Todd, 2008). The influence of external environmental factors on inter-individual variability can therefore impede the species resolution using morphometric analyses (Amaral, 1994; Ayre et al.,

1991; Bruno, Edmunds, 1997; Budd et al., 1994; Budd, 1988; Todd, 2008; Todd et al.,

2001). To overcome these challenges, comprehensive studies utilizing a combination of morphological, ecological and genetic data help to resolve, substantiate and/or revise taxonomic affinities among closely related species (Sites & Marshall 2003; deQueiroz

2007). Given the importance of corals to the ecosystems they support, the ability to properly describe and deduce ecological patterns and processes demands that they be accurately and precisely classified (Hey et al. 2003; Bickford et al. 2007).

Corals in the genus Pocillopora are common and widely distributed throughout the

Indo-Pacific (Veron 2000). Approximately 16 species (Veron, 2000; Veron, 2002) are classified based on morphology. Eight of these species are reported in the far Tropical

Eastern Pacific (TEP; Table 1, P. capitata, P. damicornis, P. effusus, P. eyudoxi, P. inflata, P. meandrina, P. verrucosa/elegans and P. woodjonesi) where they are especially dominant and vital to the productivity of shallow marine communities (Cortes, Guzmán,

1998; Glynn, Ault, 2000; von Prahl, Erhardt, 1985). Instead of brooding their larvae,

TEP Pocillopora broadcast their gametes during spawning presumably representing an adaptive shift caused by the isolation and relatively inhospitable environmental conditions characteristic of the region (Glynn et al. 1991). Recently, the analysis of the

7 internal transcribed spacer 2 (ITS2) region of the nuclear ribosomal DNA tandem array suggested that introgressive hybridization was common among TEP Pocillopora

(Combosch et al., 2008), and could explain the morphological intermediates found between various morphospecies. Hybridization is thought to be important in the evolution of corals and that hybrids are potentially better adapted to environmentally variable and extreme habitats (Willis et al. 2006). Still, published studies of coral hybridization are limited to a few exceptional genera suggesting that general conclusions about the importance of hybridization to scleractinian ecology and evolution require further substantiation.

Table 1. Morpho-species diversity in the genus Pocillopora from the Tropical Eastern Pacific, as reported by regional diversity inventories and coral atlases.

a, b f g, h, i m Mexico El Costa Panama Colombia Galapagos Clipperton

Salvador c Rica d, e j, k, l

P. damicornis X X X X X X X P. elegans X X X X X X P. meandrina X X X X X P. eydouxi X X X X X P. inflata ✙ X X X X P. capitata X X X X P. effuses ✙ X X X P. verrucosa X P. woodjonesi X X P. danae X Total 9 5 5 5 5 7 2 a Reyes-Bonilla 2003, b Reyes-Bonilla et al 2005, c Reyes-Bonilla and Barraza 2003, d Cortés and Guzmán 1998, e Cortés and Jimenez 2003,f Maté 2003, g von Prahl and Erhardt 1985, h Diaz et al 2000, i Zapata and Vargas-Ángel 2003, j Glynn and Wellington 1983, k Glynn 2003, l Baums personal communication, m Glynn 1999. ✙ Endemic of the Eastern Pacific

8 Pocillopora colonies in the TEP are known to differ in their response to anomalous warm- and cold-water episodes created by El Niño Southern Oscillations (ENSO) and the reverse processes of these events (Glynn et al. 2001; Reyes-Bonilla et al. 2002;

LaJeunesse et al. 2010b). Colonies usually associate with one of two distantly related species of Symbiodinium, one in Clade C and the other in Clade D whose distributions appear random from colony to colony irrespective of morphospecies and external environmental factors (LaJeunesse et al. 2008). This variation in symbioses among colonies explains why one individual losses its endosymbiotic algae, or “bleaches,” while an adjacent colony appears normal (Glynn et al. 2001; LaJeunesse et al. 2010b). Indeed, associations with S. glynni (D1 sensu LaJeunesse et al. 2008) may be an important determinant for why some colonies survive during periods of severe physiological stress

(LaJeunesse et al. 2010b). Therefore, the symbiosis ecology of TEP Pocillopora is presented as an example of how variation in partner combinations is important for determining how coral populations may respond to climate warming.

This study was initiated to investigate questions about hybridization and symbiosis specificity among TEP Pocillopora using the mitochondrial open reading frame region

(ORF, Flot & Tillier 2007) and ITS2 sequences in combination with population genetic data (7 microsatellite markers). We combined these analyses with the genetic identification of the resident algal symbiont. The evidence gathered was consistent with a single parsimonious, albeit unanticipated, conclusion that may finally explain unusual geographic differences in reproductive biology among Pocillopora while raising additional questions regarding the genetic basis of morphological variability among

9 individuals. This investigation offers a model with which to approach future systematic analyses of closely related Scleractinia.

Methods

Field collections for genetic analyses

Colonies of Pocillopora with different morphologies were collected at locations throughout the TEP, including Mexico (Gulf of California, GoC; Banderas Bay, BB;

Revillagigedo Islands, REV; and Oaxaca, OX), the Clipperton Atoll (CLIP), Panama

(PA) and the Galapagos Islands (GAL) (Fig. 1). Colonies were identified in the field by their general morphological characteristics including branch shape and thickness, size and uniformity of verrucae and overall colony morphology as described in the text and pictured in Veron (2000) (Fig. 2). Local experts verified these identifications.

Intermediate or variant morphologies that could not be assigned to a particular species morphotype were recorded as Pocillopora sp. Small fragments (< 1 cm long) were acquired from the colonies using bone clippers. Fragments were fixed in ethanol or

DMSO-NaCl preservation buffer (Seutin et al., 1991) and stored in the freezer at -20℃.

Figure 2. Variation in morphology among type 1 colonies in the Gulf of California including morphospecies (a) P. capitata, (b) P. damicornis (above) and P. verrucosa (below), (c) P.verrucosa, (d) P. meandrina, (e) P. capitata, and (f) P. elegans

12 DNA extractions

Genomic DNA extractions were performed on small skeletal fragments containing animal tissue using a non-toxic protocol modified from LaJeunesse et al. (2003). A small piece of skeleton and tissue was combined with glass beads (Ceroglass, Columbia USA) and

600 µl of a cell lysis solution (0.2 M Tris, 2 mM EDTA, 0.7% SDS, pH 7.6), and shaken at high speed using a BioSpec beadbeater. The extract was then incubated with proteinase

K (20 mg / ml) at 65˚C for 1 hour. After incubation, proteins were precipitated from solution using Ammonium Acetate (9 M) and the sample stored at -20˚C overnight. The frozen extract was centrifuged (10,000 g), the supernatant removed and placed into a new tube, and the DNA precipitated from solution with 100% isopropanol, and centrifuged

(10,000 g for 5 min.). The DNA pellet was washed with 70% ETOH, air dried, and resuspended in 75 µl of distilled water and stored at -20˚C.

PCR amplification and Sequencing of animal DNA

The mitochondrial open reading frame (ORF) was amplified with the FATP6.1 and the

RORF primers (Flot et al., 2008). For a subset of samples, the entire ITS2 as well as a portion of the 5.8S, and LSU RNA genes of the nuclear ribosomal array were amplified using Scler5.8Sbforward and ITSrev primers (LaJeunesse, Pinzón, 2007). Amplified products were sequenced with the forward primer using Big Dye 3.1 terminator mix

(Applied Biosystems) following the manufacturers protocol on an ABI Hitachi 3730XL genetic analyzer.

DNA sequence chromatograms were reviewed and edited using Geneious Pro 5.0

(Drummond et al., 2009). Sequences were aligned by eye or using the software ClustalW

13 with a gap-opening penalty of 15 and extension penalty of 6 (Thompson et al., 1994) and alignments exported into PAUP* (Swofford, 2000). Separate unrooted phylogenies were constructed for the mitochondrial ORF and nuclear ITS2. Heuristic search and bootstrap

(1,000 replicates) were performed using maximum parsimony, neighbor joining and maximum likelihood methods. Additionally, a Bayesian analysis was run in Mr Bayes

(Huelsenbeck, Ronquist, 2001), using the HKG85 substitution model with a chain length of 1,100,000 and 100,000 burn-in.

Population genetic markers: microsatellite genotyping and data analysis

Allelic diversity and frequency were assessed using six published microsatellite loci;

Pd3-002, Pd3-005, Pd2-006, Pd2-007, Pd3-008 and Pd3-009 (Starger et al., 2007) and a seventh new locus (Poc40). Previously published loci were optimized for the TEP samples by adding BSA (0.5mg/ml). In three primer sets (Pd2-007, Pd3-008 and Pd3-

009) the PCR buffer (NEB = 10 mM, Tris-HCl pH 8.3, 50 mM KCl, 25 mM MgCl2), was replaced with the 10X PCR Buffer #3 in the SIGMA PCR optimization kit (Product

#P2206; Buffer #3 = 100mM Tris-HCl pH 8.3, 250mM KCl, 35 mM MgCl2). Poc40 was found during the development of microsatellite markers for Clade D Symbiodinium

(Pettay, LaJeunesse, 2009). Amplification profiles for all loci followed the same conditions ((94˚C 2 min)1 – (94˚C 15 sec – Ta 15 sec – 72˚C 30 sec)31 - (72˚C – 5 min)1) with the appropriate annealing temperature (Ta) (Table 2). Each microsatellite locus, with a labeled primer, was amplified in a separate PCR reaction then pairs of products from loci with different florescent labels and sizes were co-loaded on an ABI Hitachi 3730XL,

14 using LIZ500 as a size standard. Allele sizes were scored in the software GeneMarker

(Softgenetics, State College PA, USA).

The frequencies of null alleles were estimated with two software packages;

Microchecker (van Oosterhout et al., 2004) and INest (Chybicki, Burczyk, 2008). INest includes an estimate of the selfing rates within populations in the calculation of null allele frequencies (Chybicki, Burczyk, 2008). Additional tests of heterozygote deficiency were performed in Genepop on the Web (Raymond, Rousset, 1995; Rousset, 2008). The probability of identity (power of the markers to distinguish different genotypes) was calculated prior to detection and removal of matching multilocus genotypes (GenAlex

6.3, (Peakall, Smouse, 2006). All further analyses were performed with one representative of each MLG characterized.

A Bayesian analysis using the software Structure V. 2.3.3 (Falush et al., 2003; Hubisz et al., 2009; Pritchard et al., 2000) was performed assuming admixture, correlated allele frequencies and specifying a location prior. Simulations included five iterations for each

K value (K = 1 to 10), with 100,000 burn-in and 1,000,000 chain length. The most probable number of genetically homogeneous groups (K) was determined following the

ΔK statistics procedure (Evanno et al., 2005) as implemented in Structure Harvester

(Earl, 2009). Graphics were generated with Distruct (Rosenberg, 2004) after aligning all multiple runs for each K with Clumpp (Jakobsson, Rosenberg, 2007). An additional bayesian estimate of the number of clusters (K) was performed with the software

Structurama (Huelsenbeck, Andolfatto, 2007). Structurama assigns individuals to populations implementing the Gibbs variant of the Markov Chain Monte Carlo (MCMC)

15 assuming a Dirichlet process. We ran this analysis with 1,000,000 MCMC chains treating the number of populations as a random variable (Scale = 1; Shape = 1).

Phylogenetic relations between multilocus genotypes were constructed using an

Unweighted Pair Group Method with Arithmetic Mean (UPGMA) in the program

Neighbor from the PHYLIP package (Felsenstein, 1995). The distance matrix was created using the Cavalli-Sforza and Edwards distance logarithm (Cavalli-Sforza,

Edwards, 1967), implemented in the software Microsat (Human Population Genetics

Laboratory, Stanford University).

Principal coordinate analyses (PCoA) based on genetic distances implemented in

GenAlex 6.3 (Peakall, Smouse, 2006), were performed to detect patterns of association among genotypic data. Probability tests were performed to detect deviations from Hardy-

Weinberg Equilibrium (HWE) in the PCoA clusters using Genepop (Raymond, Rousset,

1995; Rousset, 2008). The significance of differentiation between the PCoA defined groups was assessed with an AMOVA, Rst, and Fst calculations (GenAlex 6.3, (Peakall,

Smouse, 2006).

The software NewHybrids was used to identify hybrid genotypes (Anderson,

Thompson, 2002). NewHybrids uses an MCMC procedure to distinguish individuals belonging to three categories; pure (Species 1 and Species 2), hybrids (i.e. F1 and F2) and backcrosses (i.e. Species 1 – F1). Analyses were run without prior or any individual- specific assumptions, with 50,000 burin and 500,000 sweeps.

16 Table 2. Characteristics and amplification conditions of the microsatellite loci used to analyze the population structure of eastern Pacific Pocillopora spp. (L = Locus name, N = Number of alleles, Ho = observed heterozygosity, F = frequency, R = Range and Ta = annealing temperature)

L Primer sequences Motif Dye N Ho F R Ta

ATCCGAATACAAGCGAAACG AAC HEX 10 0.459 0.289 180- 55 a 002 - CAAAGCTTCTATCAGAAAATGCAA 207 Pd3

AGAGTGTGGACAGCGAGGAT TGA FAM 16 0.348 0.35 191- 55 a 005 - GTTCCTTCGCCTTCGATTTT 257 Pd3

ATCTCCATGTGATCGGCATT CA FAM 12 0.503 0.308 186- 55 a 006 - GTTCCCCCAGCTGAGAAGTT 208 Pd2

* AAGAAGGTGTGGTATTTCAGAGGG AC FAM 17 0.646 0.255 236- 60 a 007

- GGTGGATAAAGTATTTCTCACTCTTGG 468 Pd2

* AGTTGAGGTTGTTGAAACATG CTG HEX 5 0.278 0.39 154- 55 a 008

- TCCATGCAGAACCCC 166 Pd3

* CCAATGCGTCCGTAGCTCTC (CAA) FAM 8 0.164 0.606 327- 52 a 009

- ATCACCTAAAAATTTCAGTCCCTTACC (GAG) 348 Pd3

GTTATTATATGGGTGTATGC CAA HEX 11 0.345 0.367 289- 55 b

CTCAAAGTGCGATTAAAGCC 316 Poc40 aStarger et al 2007 bThis project. *S3 = 10X PCR Sigma Buffer #3, Product #P2206, 100mM Tris-HCl pH 8.3, 250mM KCl, 35 mM MgCl2 Other wise the New England Biolabs PCR Buffer (10mM, Tris-HCl pH 8.3, 50mM KCl, 25mM MgCl2 ) was used. PCR Protocol for all loci (94℃ 2 min)1 (94℃ 15sec – Ta 15sec – 72℃ 30sec)31 (72℃ 5min)1

Molecular genetic identification of Symbiodinium

For identification of the resident symbiotic dinoflagellate in each sample, DGGE fingerprinting of the ITS2 region was employed to screen and sequence the numerically dominant intragenomic variant (for details and significance of this approach, see

17 LaJeunesse 2002; LaJeunesse & Pinzón 2007; Thornhill et al. 2007). Symbiodinium specific primers (ITSintfor2 and ITS2clamp) were used to amplify the ITS2 region with a touchdown PCR protocol (LaJeunesse & Trench 2000). PCR products were analyzed using Denature Gradient Gel Electrophoresis (DGGE) on an 8% acrylamide gel with 45 -

80% denature gradient (urea and formamide). Diagnostic bands in each distinctive ITS2-

DGGE profile were characterized through excision, re-amplification, and direct sequencing (LaJeunesse, 2002; LaJeunesse et al., 2004). The genetic diversity of the resident symbiont was further analyzed using the non-coding sequence of the psbA minicircle following the protocol and using the primers 7.4-Forw and 7.8-Rev specified by Moore et al. (2003). Sequencing reactions were conducted using the forward primer

(7.4-Forw). Sequences were aligned by eye and a phylogeny constructed based on maximum parsimony using PAUP* (Swofford 2000).

Results

Mitochondrial and nuclear markers defined three distinct lineages of Pocillopora in the

TEP

The most phylogenetically informative characters of the mitochondrial ORF are in the first 600 bases (Flot, Tillier, 2007), therefore, we sequenced using the forward primer and aligned the first 830 bases. Four unique haplotypes (HQ378758 - HQ378761) where identified from 301 samples sequenced. Two of these sequences were similar and differed by 2 base pair substitutions (0.2 %) while these differed from the other

18 haplotypes by 14 and 18 base changes (1.7 - 2.7 %), respectively. These haplotypes partitioned phylogenetically into three distinct groups with statistically supported branches (Fig. 3a).

Figure 3. Phylogenetic reconstructions of TEP Pocillopora based on (a) the mitochondrial open reading frame (ORF; n = 301) and (b) the ITS2 region (n = 50). Each lineage was designated type 1 (medium grey), type 2 (dark grey) and type 3 (light grey). Topologies are drawn to the same scale and circle size is proportional to the number of samples possessing the same sequence (with the exception of type 1 ORF, n = 257). Numbers on the branches correspond to the support value of each branch after 1000 bootstrap replicates (Maximum likelihood / Neighbor Joining / Maximum Parsimony / Posterior probability MrBayes).

A total of six unique ITS2 haplotypes (HQ378752 - HQ378757) ranging in size from

399 to 414 bases were recovered from direct sequencing of this spacer region from a total of 50 samples. Phylogenetic analyses of these sequences produced three well-supported lineages (> 84%; Fig. 3b). These phylogenetic groupings matched those produced by the

19 mitochondrial ORF. Samples that shared a particular ITS2 sequence also possessed the same corresponding mitochondrial sequence. Other than the existence of minor and rare sequence variants, no recombinant genotypes were found. Compared with mitochondrial sequence phylogenies, the Pocillopora diversity in the TEP comprises three genetically separate lineages designated as types 1, 2, and 3 respectively (Figs 3a and b).

Microsatellite multilocus genotypes

The extent of genetic exchange among each phylogenetically defined type was examined with 7 microsatellite loci. A total of 392 samples were analyzed and from these 342 different multilocus genotypes (MLG) were identified (G = 173 in the GoC, 71 in BB, 2 in REV, 15 in OX, 20 in CLIP, 42 in PA, 19 in GAL, abbreviations as in Fig 1). The power of these markers to accurately distinguish between closely related genotypes and those produced by asexual reproduction was relatively high (probability of identity = 4.2 x 10-6; (Waits et al., 2001)). The largest number of repeated MLGs were found in the

GoC (n = 8), BB (n = 5) and PA (n = 8) indicating the contribution of clonal propagation in population demographics of these corals. On two occasions a MLG from the GoC was also found in BB.

Figure 4. Estimated population structure using Bayesian clustering (a) and UPGMA phylogenetic reconstruction (b) of 342 MLGs (based on 7 microsatellite loci) of TEP Pocillopora spp. The separate populations identified correspond to the genetic groupings defined by the ORF and ITS2 sequences. Red dots mark 7 putative hybrids of type 1 and type 3 identified by the software NewHybrids (abbreviations of geographic locations correspond to those in figure 1 and the colours blue, green, and yellow coincide with type 1, 2 and 3, respectively). The plot figure shown for a given K is based on the composite probabilities of 5 independent statistical runs at that K.

Bayesian analysis performed in Structure and Structurama revealed three distinct populations, each corresponding to types 1, 2, and 3, respectively (Fig. 4a and S1b).

Analyses using a greater cluster number (i.e. K = 4 to 10) did not further subdivide the data into separate populations beyond those already established. The largest population of related genotypes (type 1) comprised samples from all locations (n = GoC 173, BB 71,

REV 2, OX 15, CLIP 9, PA 22, GAL 16) and displayed significant deviations from HWE

(p < 0.05) for all loci. The other two clusters comprised genotypes that corresponded to type 3 (PA, n = 20; GAL, n = 3) and genotypes that corresponded to type 2 (CLIP, n =

11). HWE deviations were not significant in any locus from types 2 and 3. A second

21 STRUCTURE analysis performed using the genotypic diversity of only type 1 colonies revealed limited geographic partitioning of populations across a latitudinal gradient in the

TEP (Fig. 5). However pair-wise comparisons of Fst (0.05 p = 0.01) and Rst (0.042 p =

0.02) values, after bonferroni corrections, showed no statistical significance among type 1 populations. Finally, analyses targeting sympatric populations of types 1 and 2 from

CLIP (n = 22) and types 1 and 3 from PA (n = 42) showed the same strong genetic differentiation as indicated by the region wide analysis of figure 4a (Fig. 6).

Figure 5. Estimated population structure of type 1 colonies sampled throughout the far Tropical Eastern Pacific depicting limited population subdivision over a latitudinal gradient. The plot figure shown for a given K is based on the composite probabilities of 5 independent statistical runs at that K. Groupings with the high statistical support (black arrowhead) K=2 appear to distinguishes "high latitude" and "low latitude" populations of type 1.

The Cavalli Sforza genetic distance matrix between MLG used to generate a UPGMA phylogeny with Neighbor (Felsenstein, 1995) revealed that these genotypes grouped similarly to the ORF and ITS2 phylogenies. The resulting tree topology formed three

22 clusters matching the three lineages described by the mitochondrial and nuclear ribosomal DNA (Fig. 4b).

Figure 6. Genetic structure of Pocillopora populations containing (a) types 1 and 2 from the Clipperton Atoll and (b) types 1 and 3 from the Gulf of Panama. Despite being sympatric, each population exhibited strong genetic differentiation corresponding to the mitochondrial ORF and nuclear ITS2 lineages present. There was no evidence of mixed genotypes to indicate hybridization

A PCoA analysis separated MLGs into three significant clusters, or groups (Fig. 7a).

An AMOVA (Rst = 0.368; p = 0.01) revealed differences between the three groups with

Fst pairwise values ranging from 0.360 to 0.454 all significant to the p = 0.01 level. These clusters match with those defined by Structure and by the ORF and ITS2 sequence data.

That is, colonies with a particular ORF and ITS2 sequence type also shared closely related MLG.

Bayesian analysis of genotypes implemented in NewHybrids revealed the presence of

7 individuals (2%; n = 342) with various levels of admixture that may be the product of recent hybridization events, all of them belonging to the F2 category (with Q-values

23 between 0.1198 to 0.9928). Structure showed 6 of these individuals (dots in Fig. 4) as mixtures of type 1 and 2 and Structurama placed them close to type 2 (Fig. 7b) but the

UPGMA distance-based tree considers these genotypes well within the type 1 lineage

(Fig. 4b).

Figure 7. Principal coordinate analysis (PCoA) (a) and Structurama (b) analyses of the multilocus genotypes (n = 342) obtained from samples collected in the TEP. Colors correspond to three genetic groupings (blue = type 1, green = type 2, yellow = type 3). The branches of three putative “F-2 hybrids” are in black.

24 Symbiodinium ‘species’ associated with Pocillopora

ITS2-DGGE analysis of all samples identified four distinct Symbiodinium. Of these, three belonged to Clade C (C1b-c, C1d, and C1ee; Fig. 8 a and b) and another was a member of Clade D (S. glynni), all except Symbiodinium C1ee were previously characterized from

Pocillopora (LaJeunesse et al., 2010a; LaJeunesse et al., 2004). In most of the samples, the DGGE fingerprinting detected a single dominant species with mixtures of C1b-c and

S. glynni detected in approximately 3% of colonies. Symbiodinium C1b-c and S. glynni were found in association with only type 1 while colonies characterized as type 3 harbored only C1d. Finally, type 2 colonies from the Clipperton atoll (n = 11) harbored

C1ee exclusively. Phylogenetic reconstructions using the first half of the non-coding region of psbA minicircle (~ 500 to 600 bases) differentiated these DGGE-ITS2 types into three well-supported lineages (Fig. 8b). Inter-individual variability further separated these symbionts into distinct phylogeographic groupings. For example sequences of

C1b-c from Panama were more similar to each other than to those from the Clipperton atoll (Fig. 8b; Genbank accession numbers for C1b-c: HQ336237 - HQ336255; C1d:

HQ336231 - HQ336236; and C1ee HQ336225 - HQ336230).

Figure 8. (a) Representative DGGE ITS2 fingerprints of Clade C Symbiodinium found in symbiosis with TEP Pocillopora. (b) Phylogenetic relationships between the three Clade C Symbiodinium (C1b-c, C1ee, and C1d) based on the hyper-variable non-coding region of the psbA minicircle. The location origins of each sequence are provided showing distinct phylogeographic patterning. Symbiodinium glynni (D1) was found to associate with only type 1 colonies (not shown). Numbers above each branch indicate the number of sequence changes (including indels) and below are bootstrap support values based on 1000 replicates.

26 Discussion

The identification of species diversity is particularly critical when investigating questions pertaining to physiology, ecology, and evolution. Wrongly identified species, due to the presence of cryptic or overestimated diversity, has often contributed to false perceptions about biological patterns and processes (Bickford et al. 2007). Among scleractinians, for example, the apparent incongruence between genetic and morphological data has led to a questioning of traditionally diagnostic morphological traits in numerous cases, often indicating the need for significant changes in coral taxonomy and systematics (Forsman et al., 2009; Fukami et al., 2004; Fukami et al., 2008; Knowlton et al., 1992; Medina et al., 1999; Radashevsky, Pankova, 2006; Romano, Palumbi, 1996; Stefani et al., 2007;

Suatoni et al., 2006). Presently there is considerable interest in the biology of reef- building corals especially in determining their capacity to adjust to climate warming (e.g.

Hughes et al. 2003). The findings discussed below and contributions from recently published studies demonstrate that the taxonomic classifications of many coral lineages may need revision and that when instituted may improve our understanding of these important organisms. Indeed, the convergence of data based on phylogenetic patterns, population genetic makers, and symbiosis ecology unequivocally partitions TEP

Pocillopora into three natural groups that do not correspond to traditional taxonomic schemes.

The recognition that this genus comprises three distinct lineages (i.e. species) that relate little to morphology (Fig. 9a) indicates that the taxonomy of Pocillopora based on the morphospecies concept is flawed. These findings suggest 1) research into the

27 underlying causes of morphological variation is needed, 2) certain coral species may ultimately require molecular genetic analyses for identification and that 3) a taxonomic revision of the genus Pocillopora based on the congruence of various genetic markers may establish their true diversity, improve understanding of ecological and geographic distributions, and explain differences in reproductive biology (Souter 2010; Glynn et al.

1991).

Figure 9. Correspondence between colony morphology and genetic identity. (a) Pie charts representing the discordance between morphology and genetics among types 1 (blue), 2 (green) and 3 (yellow). (b) Morphological variability and similarity among colonies identified as types 1 and 3 in the Gulf of Panama, respectively.

28 High inter-individual variation in colony morphology

Investigations of Pocillopora ecology in the TEP have relied on morphospecies classifications for decades (Cortez 1997; Glynn & Ault 2000; see references in Table 1).

As stated in the introduction, formal descriptions of scleractinians are traditionally based on colony and skeleton morphology, yet phenotypic plasticity, inter-individual variation, and biogeographic variation sometimes confound accurate species identifications (Veron,

1995). Tropical Eastern Pacific Pocillopora spp. appear to exhibit high levels of morphological variability in colony shape, branch size, as well as verrucae shape and density (Figs 2a, b and 9), and substantial variation in colony morphology exists to the extent that one morphospecies grades to another among members of the same population and from the same environment (Vaughan, 1907; Veron, Pichon, 1976).

The Pocillopora sampled in the Gulf of California and mainland Mexico appear to embody a single lineage comprising damincornis-like, meandrina-like, verrucosa-like, and capitata-like colony morphologies (Figs. 2 and 9a). This variation in morphology and existence of numerous intermediates between morphospecies, especially among colonies in the GoC, may relate to relaxed competition for resources (Davidson, 1978;

Naugler, Ratcliffe, 1994). This “character release” has been proposed to explain the variability of the Montastraea annularis species complex based on fossil and living specimens from the Caribbean (Pandolfi et al., 2002) and/or hybridization (Fukami et al.

2004a).

In Panama, where types 1 and 2 coexist, there was little correspondence between colony morphology and genetics (Figs. 9b and 6b). Therefore we were unable to assign traditional morphospecies binomials to these genetic groupings. Pocillopora type 2 may

29 have greater morphological uniformity and actually correspond to Pocillopora effusus proposed originally to be endemic to the Clipperton Atoll (Veron, 2002); although there are reports of this morphospecies in Mexico (Reyes-Bonilla, 2003; Veron, 2000; Veron,

2002). Ultimate determination of the genetic basis of branch geometry, size, and distribution of verrucae may eventually show that, in some Pocillopora populations, these traits are more akin to inter-individial variants (allelic variation) than markers for species identity. The application of genomic approaches may provide information necessary to pursue these questions (Ball et al. 2002).

Congruence between phylogenetic and population genetic data

Members of lineages designated types 1, 2, and 3 possessed a characteristic ORF sequence and a corresponding ITS2 sequence (Fig. 3). The congruence, or reciprocal monophyly, of nuclear and mitochondrial sequences indicates that these lineages are evolutionarily divergent with little or no effective genetic recombination and/or hybrid introgression occurring between them (Avise & Wollenberg 1997). The analysis of population genetic data substantiates that types 1, 2, and 3 constitute reproductively isolated populations, or species (Figs 4a, b and 5). Numerous alleles at various loci were unique to a particular type and/or allelic frequencies differed significantly between types

(Appendix 1). Colonies of type 1 throughout the TEP were far more similar to each other in allelic composition than to colonies of types 2 or 3 at locations where they exist sympatrically (Galapagos and Panama for type 1 and 3, and Clipperton Atoll for types 1 and 2; Figures 6a and b). Indeed there was no clear subdivision among type 1 populations throughout the TEP indicating that gene flow and/or dispersal occurs at a

30 high enough regularity across the region to homogenize allelic frequencies (Fig. 4a and

5).

In search of hybridization

It is presumed that hybridization is common among closely related coral lineages (Veron

1995; Willis et al. 2006) especially among synchronous broadcast spawning Acropora

(Vollmer, Palumbi, 2002). These conclusions, however, are based on a small number of reports involving few genera. Combosch et al. (2008) recently added to these studies and concluded that hybridization was common among morphospecies P. damicornis, P. inflata, P effusus, P. elegans and P. eydouxi from Panama and Clipperton based on ITS2 sequence variation and polyphyletic tree topologies. It should be noted that much of the sequence variability characterized by Combosch et al (2008) was generated from cloning and sequencing PCR products of rDNA, a method shown to commonly recover low copy number intragenomic variants and pseudogenes (LaJeunesse & Pinzón 2007; Figure 10 provides a comparison, based on maximum parsimony, of the ITS2 data reported in

Figure 3b with that reported by Combosch et al. 2008. Note that types 1, 2, and 3 discussed in this paper do not correspond to their ITS2 clades I, II, and III). While one or two sequences are usually most common among rDNA, significant intragenomic variability in rDNA often exists in eukaryote genomes. Either direct sequencing or employing techniques that target the numerically dominant sequence variant are recommended when analyzing and comparing rDNA from closely related taxa

(LaJeunesse, Pinzón, 2007; Thornhill et al., 2007). Furthermore, it is probably inappropriate to use ITS sequences as population genetic markers because ribosomal

31 genes exist in multiple copies and evolve differently and more slowly than single copy nuclear loci (Dover 1982).

Figure 10. Phylogenetic similarity based on maximum parsimony of ITS2 sequences originating from bacterial cloning of PCR amplifications published by Combosch et al. (2008) in their analysis of TEP Pocillopora with sequences generated by direct sequencing reported by this study (see Fig. 3b). The sequences corresponding to ITS2 clade III (Combosch et al. 2008) are encompassed within the dashed box.

Population genetic data presented here indicate that Pocillopora types 1, 2, and 3 rarely if at all hybridize (~ 2%; n = 342, see also Figs 4b and 7b) and is in direct contrast to the conclusions of Combosch et al. (2008). If the different Pocillopora morphospecies examined by Combosch et al. (2008) were instead members of the same species, as is likely the case, then it is not surprising that their genetic data indicated significant

32 evidence for recombination. Future assumptions of species identity may benefit by analyzing population genetic data using a program like STRUCTURE (Pritchard et al.

2000) that distinguishes gene pools irrespective of morphological, geographic, or taxonomic labels. Based on these results, an a posteriori assessment of species diversity and/or estimate of hybridization would follow.

Indeed, most genetic based investigations of coral reproduction, ecology, and evolution begin with the position that morphology is the de facto metric by which to identify species (e.g. Miller & Benzie 1997; van Oppen et. al. 2001; Diekmann et al.

2003; Souter et al. 2009). This a priori assumption may wrongly influence interpretations of hybridization without considering the alternative possibility that more than one morphospecies constitutes a species, or that a single morphospecies comprises multiple cryptic taxa (Souter 2010). For example, allozyme studies involving several

Platygrya morphospecies collected from the Great Barrier Reef (GBR) by Miller and

Benzie (1997) found high genetic exchange and no relationship between genotype and morphospecies. In addition to these findings it was observed that Platygryra on the GBR exhibit identical ecological distributions, all are sympatric, and fertilization success is similar within and between morphospecies (Miller & Babcock 1997). While these data point to one logical conclusion, the possibility that Platygryra morphospecies on the

GBR may actually comprise a single species was never directly proposed and underscores how the sanctity of traditional taxonomic schemes restrains the interpretation of modern genetic data.

Heterogeneity in the reproductive strategies of pocilloporid corals, including differences in gamete buoyancy, temporal separation in spawning times, and differences

33 in their mode of spawning (broadcast spawning vs. brooding) substantially reduces the probability of sexual recombination among species of Pocillopora (Kinzie III, 1996).

This contrasts with community assemblages of Acropora where numerous species synchronously mass spawn (Harrison & Wallace 1990) thus providing frequent opportunities for hybridization. Indeed, experimental crosses show that many Acropora morphospecies can hybridize (Willis et al. 1997) and genetic data indicate that such events occur with some frequency (van Oppen et al. 2000, 2001; Vollmer & Palumbi

2002). While hybridization may have contributed to the significant radiation of Acropora diversity since the Miocene-Pliocene boundary (Fukami et al. 2000; van Oppen et al.

2001), a lack of hybridization may explain why diversity in the genus Pocillopora is significantly lower (Kinzie et al. 1996).

These findings may resolve speculation about why ‘P. damicornis,’ a species that typically broods its larvae, undergoes broadcast spawning in the TEP (Glynn et al. 1991).

Based on ORF sequence comparisons, ‘P. damicornis’ in Hawaii is genetically different than the ‘P. damicornis’ identified in the TEP (Flot et al. 2008; data not shown), a study of the reproductive biology of types 1, 2, and 3 may identify different strategies that correspond instead to these lineages rather than local adaptation to biotic and abiotic factors as previously proposed (Stimson 1978; Richmond 1985). Indeed, Pocillopora meandrina in Hawaii appears to be a broadcast spawner (Stimson 1978) and is a member of the type 1 group (Flot et al. 2008, unpubl. data).

34 Symbiont specificity provides insight into coral species identity

One of the most important ecological interactions in the lives of a reef corals is their association with dinoflagellates in the genus Symbiodinium. Based on the combined analyses of ITS2-DGGE fingerprinting and sequencing of the non-coding region of the psbA minicircle (Fig. 8b), four “species lineages” of symbiont, three in Clade C and one in Clade D, were identified from the samples examined during this study. Most importantly, each of these symbionts associated with only one type of Pocillopora (Fig.

8). Host cell environments (i.e. host-habitat) and interspecific competition likely exert strong selective pressures that contribute to ecological specialization and subsequent speciation among symbionts (LaJeunesse, 2005; LaJeunesse et al., 2010a). It might be that genetic differences between types 1, 2, and 3 influence the molecular/cellular interactions, contributing to the observed differences in host-symbiont specificity

(Goulet, 2006; Goulet, 2007; Harii et al., 2009; LaJeunesse et al., 2004; Rodriguez-

Lanetty et al., 2006; Weis et al., 2001). Among the Clade C Symbiodinium identified in the TEP, Symbiodinium C1d occurs also in ‘Pocillopora damicornis’ from Hawaii

(LaJeunesse et al., 2004). Symbiodinium C1b-c is probably a geographic variant of C1c associated with western and central Pacific Pocillopora spp. (central and southern GBR,

LaJeunesse et al. 2003, 2004a; Sampayo et al. 2007; and found in P. meandrina/eydouxi from Hawaii, LaJeunesse et al. 2004b). Symbiodinium C1ee harboured by type 2 colonies is unique and, together with its host, may be endemic to the TEP (see below).

The high fidelity exhibited by these symbioses may limit how certain partnerships respond to climate warming. Only type 1 associated with Symbiodinium glynni (found in approximately 230 out of 290 colonies). Colonies with this symbiont are more resistant to

35 disassociation, or “coral bleaching,” when exposed to environmental stressors (Glynn et al., 2001; LaJeunesse et al., 2007). This physiological resistance to thermal stress and apparent tolerance to episodes of turbidity may in part explain the ecological dominance of type 1 Pocillopora across the TEP (LaJeunesse et al., 2010b). If the frequency and intensity of warm or cold-water events in the TEP increases, colonies of types 2 and 3 would probably be among the first to die out.

The biogeography of Pocillopora types in the TEP and evidence of broader Indo-Pacific distributions

Distinctive geographic provinces in the TEP are defined by similarity in community assemblages (Heads, 2005). The TEP partitions into three sub-provinces, the northern EP

(NEP; Gulf of California, central and southern Mexico and the Revillagigedo Islands), the equatorial EP (EEP; from Ecuador and Galapagos to Costa Rica) and the island EP

(IEP; Clipperton Atoll and Malpelo Island) (Glynn, Ault, 2000). Based on the scope of sampling in this study, the geographic distributions of types 1, 2, and 3 appear to correspond with these sub-provinces. In the EEP, Pocillopora types 1 and 3 were common inhabitants. At the Clipperton atoll (part of the IEP) types 1 and 2 were both abundant, while type 1 occurred alone in the NEP (Fig. 11). Additional analysis of samples from other coastal-shelf and island regions encompassed by Costa Rica,

Colombia, and mainland Ecuador are needed for the substantiation of these distributions.

By determining the genetic identity of the morpho-species P. capitata and P. eydouxi that dominate communities at Malpelo Island (Garzón-Ferreira, Pinzón, 1999), comparisons

36 with the diversity found at the Clipperton Atoll would test whether IEP is a distinct biogeographic province.

Comparisons with published data identify types 1 and 3 in other Indo-Pacific regions.

Using the mitochondrial ORF sequence as a tentative proxy for “species” identification, type 1 exists in Hawaii and corresponds to the morphospecies of Pocillopora meandrina/eydouxi while type 3 also occurs in Hawaii, but corresponds to the morphospecies of P. molokiensis (Flot et al., 2008). The ORF of type 3 also matches with sequences of the “NF-type” of ‘P. damicornis’ described by Souter (2010) in the

Western Indian Ocean. Based on these preliminary data, types 1 and 3 are widely distributed suggesting that further sampling throughout the Indo-Pacific is needed to determine the extent of their geographic range and degree of genetic connectivity.

Figure 11. Tentative geographic map of types 1 (blue), 2 (green), and 3 (yellow) in relation to biogeographic sub-provinces defined by Glynn and Ault (2000) (NEP = Northern Province, EEP = Equatorial province and IEP = Island Province). The star indicates the location of Malpelo Island in the IEP.

Multifaceted approach in delimiting species of Scleractinia

The application of genetic analyses in the systematics of corals is steadily increasing and many have identified inconsistencies between morphology, gene sequence similarity, and ecology (Flot et al., 2008; Knowlton et al., 1992; Weil, Knowlton, 1994). The recent analysis of ‘Pocillopora damicornis’ from the western Indian Ocean concluded that this morphospecies actually comprises two distinct species (Souter, 2010). In contrast,

Hawaiian morphospecies of Pocillopora are accurately differentiated by distinct mitochondrial ORF sequences with the minor exception that P. meandrina and P. eydouxi

38 appear to comprise the same haplotype lineage (Flot et al., 2008). Furthermore, each lineage harbours a distinct symbiont (LaJeunesse et al. 2004b). While further verification is required, the Pocillopora populations in Hawaii appear to breed true to their morphology. Collectively, these data further indicate that other Pocillopora types requiring additional genetic classification exist. Furthermore, each type may exhibit distinctive morphological appearances that relate to different geographic regions, thus demanding the need for a comprehensive Indo-Pacific wide study of Pocillopora examining the congruence between morphology, phylogenetic similarity, population genetics, and symbiosis ecology.

As demonstrated here, the challenges of understanding species boundaries and/or the importance of hybridization among closely related corals may be overcome by using population genetic markers to examine genetic exchange and/or symbiont specificity when phylogenetic patterns and morphology are in conflict. Based on our current data, the actual diversity of Pocillopora in the TEP appears to be substantially lower than indicted by morphology. The possibility that the Clipperton Atoll contains a rare coral species should significantly influence decisions about regional conservation. The propensity of type 1 to associate with Symbiodinium glynni (D1) (LaJeunesse et al., 2008;

LaJeunesse et al., 2010b) may explain its dominance in the TEP, whereas populations of type 2 and 3 colonies, apparently unable to associate with S. glynni, are probably more susceptible to climate warming (Glynn, 1993; Reyes-Bonilla, 2001; Reyes-Bonilla et al.,

2002).

39 References

Amaral FD (1994) Morphological variation in the reef coral Montastraea cavernosa in Brazil. Coral Reefs 13, 113-117.

Anderson E, Thompson E (2002) A model-based method for identifying species hybrids using multilocus genetic data. Genetics 160, 1217-1229.

Andreakis N, Procaccini G, Kooistra W (2004) Asparagopsis taxiformis and Asparagopsis armata (Bonnemaisoniales, Rhodophyta): genetic and morphological identification of Mediterranean populations. Eur J Phycol 39, 273- 283.

Ayre D, Veron J, Dufty S (1991) The corals Acropora palifera and Acropora cuneata are genetically and ecologically distinct. Coral Reefs 10, 13-18.

Borrero-Pérez GH, Pérez-Ruzafa A, Marcos C, González-Wangüemert M (2009) The taxonomic status of some Atlanto-Mediterranean species in the subgenusHolothuria(Echinodermata: Holothuroidea: Holothuriidae) based on molecular evidence. Zoological Journal of the Linnean Society 157, 51-69.

Bruno J, Edmunds P (1997) Clonal variation for phenotypic plasticity in the coral Madracis mirabilis. Ecology 78, 2177-2190.

Budd A, Johnson K, Potts DC (1994) Recognizing morphospecies in colonial reef corals: I. Landmark-based methods. Paleobiology 20, 484-505.

Budd AF (1988) Large-scale evolutionary patterns in the reef-coral Montastraea: The role of phenotypic plasticity. Int. Coral Reef Symp 3, 393-398.

Cavalli-Sforza LL, Edwards AWF (1967) Phylogenetic analysis: models and estimation procedures. Evolution 21, 550-570.

Combosch D, Guzmán H, Schuhmacher H, Vollmer SV (2008) Interspecific hybridization and restricted trans-Pacific gene flow in the Tropical Eastern Pacific Pocillopora. Mol Ecol 17, 1304-1312.

Cortes J, Guzmán H (1998) Organismos de los arrecifes coralinos de Costa Rica: descripción, distribución e historia natural de los corales zooxantelados (Anthozoa: Scleractinia) del Pacífico. Revista de Biologia Tropical 46, 1-43.

Chornesky E (1983) Induced development of sweeper tentacles on the reef coral Agaricia agaricites: a response to direct competition. Biological Bulletin 165, 569-581.

Chybicki IJ, Burczyk J (2008) Simultaneous Estimation of Null Alleles and Inbreeding Coefficients. Journal of Heredity 100, 106-113.

40 Davidson D (1978) Size variability in the worker caste of a social insect (Veromessor pergandei Mayr) as a function of the competitic environment. The American Naturalist 112, 523-532.

Drummond A, Ashton B, Cheung M, et al. (2009) Geneious v4.8, Available from http://www.geneious.com/.

Earl DA (2009) Structure Harvester v0.3, from website: http://users.soe.ucsc.edu/~dearl/software/struct_harvest/.

Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology 14, 2611-2620.

Falush D, Stephens M, Pritchard JK (2003) Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics 164, 1567-1587.

Felsenstein J (1995) PHYLIP (Phylogeny Inference Package) Version 3.57c. http://www.nih.go.jp/~jun/research/phylip/main.html.

Flot J, Magalon H, Cruaud C, Couloux A, Tillier S (2008) Patterns of genetic structure among Hawaiian corals of the genus Pocillopora yield clusters of individuals that are compatible with morphology. Comptes Rendus Biologies 331, 239-247.

Flot J-F, Tillier S (2007) The mitochondrial genome of Pocillopora (Cnidaria: Scleractinia) contains two variable regions: The putative D-loop and a novel ORF of unknown function. Gene 401, 80-87.

Forsman Z, Barshis D, Hunter C, Toonen R (2009) Shape-shifting corals: molecular markers show morphology is evolutionarily plastic in Porites. BMC Evolutionary Biology 9.

Fukami H, Budd A, Paulay G, et al. (2004) Conventional taxonomy obscures deep divergence between Pacific and Atlantic corals. Nature 427, 832-835.

Fukami H, Chen CA, Budd AF, et al. (2008) Mitochondrial and Nuclear Genes Suggest that Stony Corals Are Monophyletic but Most Families of Stony Corals Are Not (Order Scleractinia, Class Anthozoa, Phylum Cnidaria). PLoS ONE 3, e3222.

Garzón-Ferreira J, Pinzón JH (1999) Evaluacion rapida de la estructura y salud de las formaciones coralinas de la Isla de Malpelo (Pacifico Colombiano). Boletin de Investigaciones Marinas y Costeras 28, 137-154.

Glynn P, Mate J, Baker A, Calderón MO (2001) Coral bleaching and mortality in Panama and Ecuador during the 1997-1998 El Niño-Southern oscillation event: spatial/temporal patterns and comparisons with the 1982-1983 event. Bull Mar Sci 69, 79-109.

41 Glynn PW (1993) Coral reef bleaching: ecological perspectives. Coral Reefs 12, 1-17.

Glynn PW, Ault JS (2000) A biogeographic analysis and review of the far eastern Pacific coral reef region. Coral Reefs 19, 1-23.

Goulet T (2006) Most corals may not change their symbionts. MEPS 321, 1-7.

Goulet T (2007) Most scleractinian corals and octocorals host a single symbiotic zooxanthella clade MEPS 335, 243-248.

Harii S, Yasuda N, Rodriguez-Lanetty M, Irie T, Hidaka M (2009) Onset of symbiosis and distribution patterns of symbiotic dinoflagellates in the larvae of scleractinian corals. Marine Biology 156, 1203-1212.

Heads M (2005) Towards a panbiogeography of the seas. Biol J Linn Soc 84, 675-723.

Hubisz MJ, Falush D, Stephens M, Pritchard JK (2009) Inferring weak population structure with the assistance of sample group information. Molecular Ecology Resources 9, 1322-1332.

Huelsenbeck J, Andolfatto P (2007) Inference of population structure under a Dirichlet process prior. Genetics 175, 1787-1802.

Huelsenbeck J, Ronquist F (2001) MrBayes: a program for the Bayesian inference of phylogeny. Bioinformatics 17, 754-755.

Kinzie III R (1996) Modes of speciation and reproduction in Archaeocoeniid corals. Galaxea 13, 47-64.

Knowlton N, Weil E, Weigt L, Guzman H (1992) Sibling species in Montastraea annularis, coral bleaching, and the coral climate record. Science 255, 330-333.

LaJeunesse TC (2002) Diversity and community structure of symbiotic dinoflagellates from Caribbean coral reefs. Marine Biology 141, 387-400.

LaJeunesse TC (2005) "Species" Radiations of Symbiotic Dinoflagellates in the Atlantic and Indo-Pacific Since the Miocene-Pliocene Transition. Mol Biol Evol 22, 570- 581.

LaJeunesse TC, Pettay D, Sampayo EM (2010a) Long-standing environmental conditions, geographic isolation and host–symbiont specificity influence the relative ecological dominance and genetic diversification of coral endosymbiondt in the genus Symbiodinium. Journal of Biogeography 37, 785-800.

42 LaJeunesse TC, Pinzón JH (2007) Screening intragenomic rDNA for dominant variants can provide a consistent retrieval of evolutionarily persistent ITS (rDNA) sequences. Molecular Phylogenetics and Evolution 45, 417-422.

LaJeunesse TC, Reyes-Bonilla H, Warner ME (2007) Spring “bleaching” among Pocillopora in the Sea of Cortez, Eastern Pacific. Coral Reefs 26, 265-270.

LaJeunesse TC, Reyes-Bonilla H, Warner ME, et al. (2008) Specificity and stability in high latitude eastern Pacific coral–algal symbioses. Limnol. Oceanog. 53, 719- 727.

LaJeunesse TC, Smith R, Pinzón JH, et al. (2010b) Host-symbiont recombination vs. natural selection in the response of coral-dinoflagellate symbioses to environmental disturbance. Proceedings of the Royal Society of London, Series B. Biological Sciences, in press.

LaJeunesse TC, Thornhill DJ, Cox E, et al. (2004) High diversity and host speciﬁcity observed among symbiotic dinoﬂagellates in reef coral communities from Hawaii Coral Reefs 23, 596-603.

Lesser M, Weis V, Patterson M, Jokiel P (1994) Effects of morphology and water motion on carbon delivery and productivity in the reef coral, Pocillopora damicornis (Linnaeus): Diffusion barriers, inorganic carbon limitation, and biochemical plasticity. Journal of Experimental Marine Biology and Ecology 178, 153-179.

Medina M, Weil E, Szmant A (1999) Examination of the Montastraea annularis species complex (Cnidaria: Scleractinia) using ITS ad CO2 Sequences. Marine Biotechnology 1, 89-97.

Naugler CT, Ratcliffe L (1994) Character release in bird song: a test of the acoustic competition hypothesis using american tree sparrows Spizella arborea. Journal of Avian Biology 25, 142-148.

Pandolfi J, Lovelock C, Budd AF (2002) Character release following extinction in a Caribbean reef coral species complex. Evolution 56, 479-501.

Peakall R, Smouse PE (2006) Genalex 6: genetic analysis in Excel. Population genetic software for teaching and research. Mol Ecol Notes 6, 288-295.

Pettay D, LaJeunesse T (2009) Microsatellite loci for assessing genetic diversity, dispersal and clonality of coral symbionts in'stress-tolerant' clade D Symbiodinium. Molecular Ecology Resources 9, 1022-1025.

Pritchard JK, Stephens M, Donnelly P (2000) Inferece of population structure using multilocus genotype data. Genetics 155, 945-959.

43 Radashevsky V, Pankova V (2006) The morphology of two sibling sympatric Polydora species (Polychaeta : Spionidae) from the Sea of Japan. J Mar Biol Assoc Uk 86, 245-252.

Raymond M, Rousset F (1995) GENEPOP (version1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity 86, 248-249.

Reyes-Bonilla H (2001) Effects of the 1997-1998 El Niño-Southern Oscillation on coral communities of the Gulf of California, Mexico. Bulletin of Marine Science 69, 251-266.

Reyes-Bonilla H (2003) Coral reefs of the Pacific coast of Mexico. In: Latin American Coral Reefs (ed. Cortes J), pp. 331-350. Elsevier, Amsterdam.

Reyes-Bonilla H, Carriquiry JD, Leyte-Morales GE, Cupul-Magaña AL (2002) Effects of the El Niño-Southern Oscillation and the anti-El Niño event (1997–1999) on coral reefs of the western coast of México Coral Reefs 21, 368-372.

Rodriguez-Lanetty M, Wood-Charlson E, Hollingsworth L, Krupp D, Weis V (2006) Temporal and spatial infection dynamics indicate recognition events in the early hours of a dinoflagellate/coral symbiosis. Marine Biology 149, 713-719.

Romano S, Cairns S (2000) Molecular phylogenetic hypotheses for the evolution of scleractinian corals. Bulletin of Marine Science 67, 1043-1068.

Romano S, Palumbi SR (1996) Evolution of scleractinian corals inferred from molecular systematics. Science 271, 640-642.

Romano S, Palumbi SR (1997) Molecular evolution of a portion of the mitochondrial 16S ribosomal gene region in Scleractinian Corals. J. Mol. Evol. 45, 397-411.

Rosenberg N (2004) DISTRUCT: a program for the graphical display of population structure. Molecular Ecology Notes 4, 137-138.

Rousset F (2008) Genepop’007: a complete re-implementation of the genepop software for Windows and Linux. Molecular Ecology Resources 8, 103-106.

Ruiz-Sanchez E, Sosa V (2010) Delimiting species boundaries within the Neotropical bamboo Otatea (Poaceae: Bambusoideae) using molecular, morphological and ecological data. Mol Phylogenet Evol 54, 344-356.

Seutin G, White BN, Boag PT (1991) Preservation of avian blood and tissue samples for DNA analyses. Canadian Journal of Zoology 69, 82-90.

Souter P (2010) Hidden genetic diversity in a key model species of coral. Mar Biol.

44 Starger CJ, Yeoh SSR, Dai C-F, Baker AC, DeSalle R (2007) Ten polymorphic STR loci in the cosmopolitan reef coral, Pocillopora damicornis. Molecular Ecology Resources 8, 619-621.

Stefani F, Benzoni F, Pichon M, Cancelliere C, Galli P (2007) A multidisciplinary approach to the definition of species boundaries in branching species of the coral genus Psammocora (Cnidaria, Scleractinia). Zool Scripta 37, 71-91.

Suatoni E, Vicario S, Rice S, Snell T, Caccone A (2006) An analysis of species boundaries and biogeographic patterns in a cryptic species complex: The rotifer - Brachionus plicatilis. Mol Phylogenet Evol 41, 86-98.

Swofford DL (2000) PAUP*, Phylogenetic Analysis Using Parsimony (*and other methods), Version 4.0b.

Thompson J, Higgins D, Gibson T (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position- specific gap penalties and weight matrix choice. Nucleic Acids Research 22, 4673-4680.

Thornhill DJ, LaJeunesse TC, Santos SR (2007) Measuring rDNA diversity in eukaryotic microbial systems: how intragenomic variation, pseudogenes, and PCR artifacts confound biodiversity estimates. Mol Ecol 16, 5326-5340.

Todd PA (2008) Morphological plasticity in scleractinian corals. Biol. Rev. 83, 315-337.

Todd PA, Sanderson PG, Chou LM (2001) Morphological variation in the polyps of the scleractinian coral Favia speciosa (Dana) around Singapore. Hydrobiologia 444, 227-235. van Oosterhout C, Hutchinson WF, Wills DPM, Shipley P (2004) Micro-checker: software for identifying and correcting genotyping errors in microsatellite data. Mol Ecol Notes 4, 535-538.

Vaughan T (1907) Recent madreporaria of the Hawaiian Islands and Laysan.

Verbruggen H, Leliaert F, Maggs CA, et al. (2007) Species boundaries and phylogenetic relationships within the green algal genus Codium (Bryopsidales) based on plastid DNA sequences. Mol Phylogenet Evol 44, 240-254.

Veron JEN (1995) Corals in Space and Time. The biogeography and evolution of the scleractinian Southwood Press, Australia.

Veron JEN (2000) Corals of the world Australian Institute of Marine Science, Twonsville.

Veron JEN (2002) New species described in Corals of the world Australian Institute of Marine Science, Twonsville.

45 Veron JEN, Pichon M (1976) Scleractinia of Eastern Australia Part I. Australian Institute of Marine Science Canberra, 86.

Vollmer SV, Palumbi SR (2002) Hybridization and the evolution of reef coral diversity Science 296, 2023-2025. von Prahl H, Erhardt H (1985) Colombia: corales y arrecifes coralinos Universidad del Valle, Bogota.

Waits L, Luikart G, Taberlet P (2001) Estimating the probability of identity among genotypes in natural populations: cautions and guidelines. Molecular Ecology 10, 249-256.

Weil E, Knowlton N (1994) A multi character analysis of the Caribbean coral Montastraea annularis (Ellis and Solander, 1786) and its two sibling species, M. faveolata (Ellis and Solander, 1786) and M. franski (Gregory, 1895). Bulletin of Marine Science 55, 151-175.

Weis V, Reynolds W, deboer M, Krupp D (2001) Host-symbiont specificity during onset of symbiosis between the dinoflagellates Symbiodinium spp. and planula larvae of the scleractinian coral Fungia scutaria. Coral Reefs 20, 301-308.

46 CHAPTER 2: A GENETIC ASSESSMENT OF SPECIES BOUNDARIES IN THE CORAL GENUS POCILLOPORA ALTERS KNOWLEDGE OF BIODIVERSITY, BIOGEOGRAPHIC PATTERNS, AND THE EVOLUTION OF REPRODUCTIVE STRATEGIES

Abstract

Scleractinian systematics has traditionally relied on morphological traits to delineate species. In recent years, inferences based on genetic data have prompted a revision of coral taxonomy at every level. The genus Pocillopora occurs throughout the Indian and Pacific Oceans, and is an important constituent of many coral reef ecosystems. Extensive morphological variation among colonies in particular regions and habitats makes the identification of species difficult. While regional studies have compared genetic and morphological differences, there has been no comprehensive phylogenetic analysis of this group across the Indo-Pacific. In this study, a combination of molecular phylogenetics (using nuclear – internal transcribe spacer 2 ITS2 - and mitochondrial open reading frame – ORF- sequence data), population genetics (seven microsatellite loci) and symbiosis ecology (symbiont specificity) was used to reassess the diversity, geographic range, and morphological variation of Pocillopora spp. The data indicate that the genus is composed of five, genetically distinct, widely distributed lineages as well as at least three regionally endemic taxa. Accordingly, the genus may be less diverse than previously defined using morphology. Modes of reproduction in the genus seem to transition from broadcast spawning among putative ancestral lineages to brooding in more derived species. Host- symbiont specificity is dynamic and partner combinations change over broad geographic ranges influenced by long-standing regional and local differences in environment. Although gross morphology distinguishes reproductively isolated species in certain places such as Hawaii, genetic groupings often did not relate to species boundaries defined based on morphological characters. Collectively these findings indicate that future research should adopt a molecular genetic approach to identification before any physiological ecological and/or evolutionary processes are inferred.

47 Introduction

Classifications and identifications of scleractinian corals are primarily based on skeletal features, including colony- (macroskeletal) and/or calice- (microskeletal) level characteristics (Benzoni et al., 2010; Veron, Pichon, 1976). Recently, taxonomic studies of hard corals have turned to multivariate analyses including other features, such as reproductive biology and ecological distributions (Budd, Stolarski, 2009; Knowlton,

1992; Locke et al., 2007; Pinzón, Lajeunesse, 2011; Stobart, 2000; Weil, Knowlton,

1994; Weil, Vargas, 2009). This trend has continued by including various molecular techniques and markers, in integrative studies delineating species in several groups of corals (Benzoni, 2006; Benzoni et al., 2010; Benzoni et al., 2007; Combosch et al., 2008;

Forsman et al., 2009; Forsman et al., 2010; Fukami, 2008; Fukami et al., 2004; Fukami et al., 2008; Lopez et al., 1999; Lopez, Knowlton, 1996; Medina et al., 1999; Pinzón,

Lajeunesse, 2011; Stefani et al., 2008a; Stefani et al., 2008b; Weil, 1992; Wolstenholme et al., 2003).

The unstable history of coral taxonomy reflects a great deal of uncertainty in the delineation of species (Budd et al., 2010; Lopez, Knowlton, 1996; Weil, 1992). The lack of confidence in species boundaries has been attributed to hybridization, recent divergence between species and morphological variability within several scleractinian genera (Benzoni et al., 2010). Phenotypic plasticity has been observed in many taxa and is believed to allow adaptation to different habitats in response to environmental heterogeneity (Miller, 1994; Todd, 2008). Unfortunately, changes in skeletal features create morphological intergroup-clines (Huxley, 1938), complicating our understanding of species differentiations among scleractinia (Veron, Pichon, 1976). Multivariate

48 assessments of species boundaries utilizing morphological and other characters have resulted in taxonomic reorganizations and a more accurate understanding of the evolutionary history of scleractinia (Budd et al., 2010). Still, coral species differentiation is challenging and is performed in absence of full knowledge of the effects of environment or genetics on morphological characteristics (Benzoni et al., 2010).

Our understanding of any biological process requires a clear definition of the operational taxonomic units being studied (Claridge et al., 1997; Wiens, Servedio, 2000), and conservation programs demand a complete knowledge of the intricate processes involving interacting and correctly delimited species (Bickford et al., 2007). The aim of this study was to use a multivariate approach to assess the genetic diversity of the ecologically important scleractinian genus Pocillopora through its geographic distribution, and use this information to elucidate species boundaries. Our analysis incorporates nuclear (internal transcribe spacer 2 - ITS2 and seven microsatellite loci) and mitochondrial (open reading frame – ORF) molecular markers, ecological

(dinoflagellate alga identity), biogeographical information and microskeletal measurements, to present a comprehensive delineation of species boundaries and initiate a revision of the classification of Pocillopora spp.

Genus Pocillopora (Lamarck, 1816)

The genus Pocillopora comprises corals with branching colonies, although some species rarely show crustose and massive forms. The calices, with poorly developed septa and columella, are usually associated with short protuberances denominated “verrucae”

(Veron, 2000; Veron, 2002; Veron, Pichon, 1976). The species of the genus are

49 ecologically important and common through the Indian and Pacific Oceans (Veron,

2000), including areas where conditions are considered suboptimal for corals (i.e. high latitude and high sedimentation environments). In some locations (i.e. Tropical Eastern

Pacific - TEP) Pocillopora species are common and are the key reef building corals

(Cortes, Guzmán, 1998; Glynn, Ault, 2000; von Prahl, Erhardt, 1985).

Morphological studies of Pocillopora have resulted in inconsistent inferences regarding the actual number of valid species. The most recent compilation of Pocillopora is in the book by Veron (2000), in which 17 Pocillopora are presented as valid species

(Appendix 1), five of these corresponding to new species for this genus (P. effusus, P. fungiformis, P. indiana, P. kelleheri and P. zelli) (Veron, 2002). However, perhaps the most commonly accepted species were included in a previous list by Cairns (1999 - P. capitata, P. damicornis, P. elegans, P. eydouxi, P. meandrina, P. verrucosa, and P. woodjonesi - Appendix 1). The difference in “valid” species may have result from the complexity on the actual delineation of Pocillopora spp.; a situation recognized by early taxonomist working on this genus (see Veron and Pichon, 1976 for a complete compilation of quotations), some considering the number of valid species much lower to that of actual descriptions (Crossland, 1952). More recently, the possibility of P. verrucosa and P. elegans being synonymous (Glynn, Ault, 2000), constitutes another example of the lack of appropriate species delineations. Additionally some of the species are considered rare and possibly endemic (e.g., P. molokensis in Hawaii; P. inflata and P. capitata in the TEP, P. fungiformis in Madagascar, P. effusus in Clipperton and Mexico,

P. ligulata CP and TEP, P. danae, P. kelleheri and P. indiania in the west Indian Ocean,

50 P. zelly in the coral triangle – Indo Pacific), even though they may represent rare forms of the more morphologically variable and widely distributed species.

Problems with morphological based delineations in the genus are evident when

“ecomorphs” from different species are morphologically similar to each other than to the

“normal” morphologies of each species (Veron, Pichon, 1976). Attempts to resolve morphological boundaries between Pocillopora species have been restricted to local studies. Numerical taxonomy was employed in Colombia, but the results revealed clusters that disagreed with traditional taxonomic divisions of the genus in this location

(Cantera et al., 1989). Similar findings were subsequently obtained in a skeletal analysis of three Pocillopora species in Panama (Budd, 1990). Other studies have included

Pocillopora specimens but were not specific to this genus and included species from other genera (Powers, 1970; Powers, Rohlf, 1972).

Although phenotypic plasticity has obscured the taxonomy of the group, no comprehensive genetic analyses have been done and few studies have attempted to solve this problem. The use of molecular markers has resulted in conflicts between genetic and morphological studies about the status of some Pocillopora species. Results include; hybridization in Panama and Clipperton (Combosch et al., 2008), cryptic species in South

Africa (Souter, 2010), and agreements between morphologically-based classification and genetics of some species in Hawaii (Flot et al., 2008). These conclusions have added more complexity to the already complicated and entangled systematic of this scleractinian genus.

In a previous study, we found that Pocillopora through the TEP was composed of only three genetically differentiated species, denominated Types 1, 2 and 3, with low

51 levels of hybridization and no concordance with traditionally morphologically defined species (Pinzón, Lajeunesse, 2011). In addition to differences in nuclear (ITS2), mitochondrial (ORF) sequence data and multilocus genotypes (seven microsatellite markers), each Pocillopora type forms unique associations with particular Symbiodinium species and showed different particular geographical distributions. Reduction in species diversity in the TEP and lack of correspondence between morphological features and genetics prompted our interest in the status and differentiation of Pocillopora species in other locations in the Indo-Pacific. Here we present a comprehensive multivariate analysis that integrates genetic, ecologic, morphologic and biogeographic information to study the diversity of Pocillopora through its geographical distribution and proposed a reorganization of the species in the genus.

Methods

Locations, sample collections and DNA extraction

Selection of field locations was aimed to cover as much as possible the geographic distribution of Pocillopora therefore 15 sites were sampled; Tanzania (TAN) in the IO,

Thailand (THA), Western Australia (WA), Taiwan (TAI), Palau (PAL) Heron Island

(HI), and Lizard Island (LZI) in the IP, Hawaii (HAW) in CP and Mexico (La Paz - GoC,

Revigagilledo Islands - REV, Banderas Bay - BB, Oaxaca – OX), Clipperton Atoll

(CLIP) Panama (PA) and Galapagos (GAL) in the TEP (Fig. 12). Although in the TEP we already demonstrated that gross morphology does not correlate with other characters

(i.e. molecular, populational and ecological) (Pinzón, Lajeunesse, 2011), genetic markers

52 have found some correlation between morphology and genetics in Hawaii (Flot et al.,

2008), and therefore initial identification of the colonies followed the descriptions given by Veron (2000, 2002). Morphologies that did not correspond to a recognizable species morphotype were recorded as Pocillopora sp. Samples consisted of small fragments (< 1 cm long) acquired from the colonies using bone clippers. Fragments were fixed in ethanol or DMSO and stored at -20℃ until further analyses.

Figure 12. Locations of the 15 sampling sites covering part of the geographical distribution of Pocillopora spp in the Indian Ocean (IO – Tanzania TAN), the Indo Pacific (IP - Western Australia WA, Taiwan TAI, Palau PAL, Heron Island (HI), and Lizard Island LZI), the Central Pacific (CP - Hawaii HAW) and the Tropical Eastern Pacific (TEP – Mexico; La Paz - GoC, Revigagilledo Islands - REV, Banderas Bay - BB, Oaxaca – OX; Clipperton Atoll CLIP, Panama PA and Galapagos GAL).

Genomic DNA extraction

DNA extractions were performed on small skeletal fragments containing animal tissue using the Wizard Genomic Purification Kit (Promega, Madison USA) according to the manufacturer’s instructions. A small piece of skeleton and tissue was pulverized with glass beads (Ceroglass, Columbia USA) in the nuclei lysis solution (provided with the kit). The coral tissue was then incubated with proteinase K (20 mg / ml) at 65˚C for 1 hour. After incubation, the solution was centrifuged and the DNA precipitated from the

53 supernatant with isopropanol/ethanol. The sample was air dried and the resulting DNA diluted in 75 µl of distilled water, with final concentrations between 5 to 80 ng / µl, and stored at -20˚C (LaJeunesse et al., 2003).

Pocillopora genetic identification

Genetic characterization of the Pocillopora species was performed using two genetic markers, the mitochondrial open reading frame (ORF) and the Internal Transcribed

Spacer 2 (ITS2) region of the nuclear ribosomal gene repeat. PCR amplifications were obtained as described in Pinzón and LaJeunesse (2011). PCR products were sequenced using Big Dye 3.1 terminator mix (Applied Biosystems) following the manufacturer’s protocol on an ABI Hitachi 3730XL genetic analyzer at the Penn State Huck Institutes

Genomics Core Facility. DNA sequence chromatograms were reviewed and edited using

Geneious Pro 5.0 (Drummond et al., 2009), duplicate sequences were removed from the data set. ClustalW was then used to align sequences using a gap-opening penalty of 15 and extension penalty of 6 (Thompson et al., 1994). Gaps were coded using Gapcoder

(Young, Healy, 2003) and alignments exported into the PAUP* phylogenetic analysis software package 4.0 (Swofford, 2000). Separate phylogenies were constructed for the mitochondrial ORF and nuclear ITS2. Phylogenetic analyses were performed using maximum parsimony, neighbor-joining and maximum likelihood methods, each accompanied by bootstrap analyses employing 1000 replicates. Additionally, a Bayesian analysis was run in Mr Bayes (Huelsenbeck, Ronquist, 2001), using the HKG85 substitution model with a chain length of 1,100,000 and a burnin value of 100,000.

Attempts to use closely related species as outgroups failed due to high differentiation

54 (more than 50% sequence difference) between Pocillopora sequences and those from

Seriatopora and Stylophora specimens.

In addition to the molecular markers, seven microsatellite loci, already tested and known to differentiate species and populations of Pocillopora (Pinzón, Lajeunesse, 2011;

Starger et al., 2007) were use to determine the allelic diversity in the genus.

Microsatellites were amplified as described in Pinzón and LaJeunesse (2011). Each microsatellite locus, with a labeled primer, was amplified in a separate PCR reaction, then pairs of products from loci with different florescent labels were mixed and loaded on an ABI Hitachi 3730XL, using LIZ500 as a size standard. Allele sizes were scored in the software GeneMarker (Softgenetics, State College PA USA).

Microsatellite data was used to establish population/species differentiation with the

Bayesian model implemented in Structure V2.3.3. (Falush et al., 2003; Hubisz et al.,

2009; Pritchard et al., 2000). The analyses in Structure assumed admixture, correlated allele frequencies and a location prior. Simulations included five iterations for each K value (K = 1 to 20), with a burnin value of 100,000 and 1,000,000 chain length. The most probable number of genetically homogeneous groups (K) was determined following the

ΔK statistics procedure (Evanno et al., 2005) as implemented in Structure Harvester

(Earl, 2009). Graphics were generated with Distruct (Rosenberg, 2004) after aligning all multiple runs for each K with Clumpp (Jakobsson, Rosenberg, 2007).

Probability tests were performed to detect deviations from Hardy-Weinberg

Equilibrium within the clusters delineated with Structure, using Genepop on the web

(Raymond, Rousset, 1995; Rousset, 2008). The significance of differentiation between the defined groups was assessed with an AMOVA, Rst, and Fst calculations performed

55 using GenAlex 6.3 (Peakall, Smouse, 2006). Relationships between multilocus genotypes were evaluated using an Unweighted Pair Group Method with Arithmetic Mean

(UPGMA) in the program Neighbor from the Phylip package (Felsenstein, 1995). The distance matrix was created using the Cavalli-Sforza and Edwards distance logarithm

(Cavalli-Sforza, Edwards, 1967), implemented in the software Microsat 1.5 (Human

Population Genetics Laboratory, Stanford University).

An additional Bayesian estimate of the number of clusters (K) was established with

Structurama (Huelsenbeck, Andolfatto, 2007). This software assigns individuals to populations implementing the Gibbs variant of the Markov Chain Monte Carlo (MCMC) assuming a Dirichlet process. We ran this analysis with 1,000,000 MCMC chains treating the number of populations as a random variable (Scale = 1; Shape = 1) and the resulting distance matrix (calculated as the negative of the natural log of the probability that the two individuals are clustered into the same population) was exported to PAUP*, where a

UPGMA tree was constructed.

NewHybrids (Anderson, Thompson, 2002) was used to explore the presence of hybrid colonies between each pair of genetically defined Pocillopora clusters, globally

(across locations) and in sympatry (clusters within same location). NewHybrids uses an

MCMC procedure to distinguish individuals belonging to three categories; pure (Species

1 and Species 2), hybrids (i.e. F1 and F2) and backcrosses (i.e. Species 1 – F1). Analyses were run without prior or any individual-specific assumptions, with a 50,000 burnin value and 500,000 sweeps.

56 Symbiosis ecology – symbiont identification

The identification of the resident symbiotic dinoflagellate in each colony was performed using denaturing gradient gel electrophoresis (DGGE) fingerprinting of the ITS2 region.

This technique screens and sequences the numerically dominant intragenomic variant

(LaJeunesse, 2002; LaJeunesse et al., 2003; Thornhill et al., 2007). Symbiodinium specific primers (ITSintfor2 and ITS2clamp) were used to amplify the ITS2 region with a touchdown PCR protocol (LaJeunesse, Trench, 2000). PCR products were analyzed using

Denature Gradient Gel Electrophoresis (DGGE) on an 8 % acrylamide gel with a 45 % to

80 % gradient of urea and formamide. Based on previous studies, diagnostic bands were identified for each sample (LaJeunesse, 2002; LaJeunesse et al., 2004). In order to confirm the identity of each Symbiodinium ITS2-DGGE profile, prominent bands were excised from the gel and re-amplified using a similar set of primers, but without the clamp sequence on the reverse primer, in a standard PCR protocol with an annealing temperature of 52℃. Sequencing was preformed as previously described (LaJeunesse,

2002).

Morphological analysis

Microskeletal traits were measured and compared among species of Pocillopora from

Hawaii and the Gulf of California (GoC) to explore the concordance between genetic and morphological traits. In Hawaii, four morphologically identified species were sampled:

Pocillopora damicornis (n = 10), P. eyudoxi (n = 7), P. ligulata (n = 19) and P. meandrina (n = 10). In the GoC, collections included 39 colonies with different morphologies from reefs in Punta Galeras and Isla Gaviota north of La Paz, Mexico. In

57 all locations, three branches (3 cm each) were sampled randomly from each colony, bleached in 10% sodium chloride and the skeleton air-dried. A smaller fragment was stored in DMSO for genetic analyses. In addition to this samples, 20 specimens of three

Pocillopora species (P. damicornis, P. capitata, and P. meandrina) from the Museo de

Historia Natural de la Universidad Autónoma de Baja California Sur (MHNUBCS) were also analyzed. The museum specimens originated from various locations throughout the

GoC (Appendix 2).

Photographs of the skeletons were taken, approximately 2 cm from the growing tip of each branch, using a Leica 4EZD stereoscope mounted with a digital camera at a magnification of 4X. Each photograph was analyzed using the software ImageJ

(Abramoff et al., 2004). From each colony, five calices were selected randomly and both the diameter (d) and its distance to the closest calice (di), between calice edges, were measured. Length calibrations were made using a reticle micrometer. Prior to any comparisons five measurements were averaged per sample. Normality and homogeneity of variance tests failed for the measured characters, as well for a transformed data set, therefore statistical significance in calice size and distance separating individual calices were estimated using a non-parametric Kruskal-Wallis One Way Analysis of Variance

(KW-ANOVA) in the SPSS software (IBM, Somers New York). Microskeletal differences among species from the museum were initially evaluated followed by a second KW-ANOVA that included data from recent field collected colonies.

58 Results

Genetic diversity in the genus Pocillopora

Sequences from the most informative region (830bp from the FATP6.1 primer) within the mitochondrial ORF (Flot et al., 2008; Pinzón, Lajeunesse, 2011) were obtained from

666 colonies from diverse locations (TAN 130, THA 12, TAI 9, WA 31, HI 44, LZI 31,

PAL 31, HAW 78, CLI 21, GAL 23, MEX 202, PA 54; see locality abbreviations in

Figure 12). Search for duplicates within this dataset resulted in 15 unique sequences. All reconstructions of relationships between these mitochondrial haplotypes revealed seven well-supported (bootstrap values > 66 %) and distinct groups (Fig. 13), among them those from Pocillopora types 1, 2 and 3 found in the TEP. The remaining groups were designated Pocillopora type 4, 5, 6 and 7 continuing with the previously established nomenclature (Pinzón, Lajeunesse, 2011). Four of the types (2, 5, 6 and 7) were represented by a single sequence, while the other three (Type 1 with 2 sequences, Type 3 with 6 and Type 4 with 2) presented some levels of variability (i.e. more than one sequence).

Figure 13. Phylogenetic reconstructions with a most parsimonious analysis (consensus tree) of the mtDNA ORF and rDNA ITS2 regions from Pocillopora species collected in different reefs in the Indian and Pacific Oceans. Dotted lines are encircling the proposed genetic types (numbered in each trees) in the genus after branch support was tested, with black branches highly supported (> 66 %) in all analyses (Maximum likelihood, Neighbor Joining, Maximum Parsimony and Bayesian - MrBayes), while gray branches showed low support. Circles in the nodes and leaves represent sequences obtained here; dark gray circles in the ITS2 tree are groups showing incongruence between phylogenies. Pocillopora 1mea, 1eyu, 3 and 5 showed wide distributions and P. 2, 4, 5, and 6 might be brooders contrary to the other species (See Discussion).

Data from the 5.8S-ITS2 rDNA (395 to 414 bp long) from a subset (n = 124) of samples (TAN 32, THA 3, TAI 1, HI 1, LZI 1, PAL 15, HAW 21, CLI 15, MEX 29, PA

6) showed more diversity, with 23 unique ITS2 sequences. Phylogenetic analyses resulted in a topography with six well supported groups. Four clusters showed the same phylogenetic relations observed for the ORF (Pocillopora type 2, 3, 5 and 6) while

Pocillopora type 1 appears divided in two corresponding to Hawaii samples identified as

P. eyudoxi (1eyu) and P. meandrina (1mea) (Fig. 13). Colonies identified with the ORF as Pocillopora type 3 showed the largest ITS2 diversity with 9 sequences and containing

ITS sequences from other ORF-delineated groups (Pocillopora 4 and 7) and splitting P.

1mea (Fig. 13). A single colony of Pocillopora type 7 was found therefore in further analyses this type is not used.

Multilocus genotypes (MLG) from seven microsatellite loci were determined from

630 colonies (TAN 109, HI+LZI (GBR) 72, PAL 34, HAW 82, CLIP 20, GAL 19,

Mexico 261 and PA 42). Bayesian analyses of population structure determined by

Structure and Structurama as well as distance-based phylogenetic reconstruction, revealed similar genetic distinctions between all Pocillopora types (Fig. 14 and 15) to those seen with the sequence data.

Figure 14. Clustering of multilocus genotype data from 630 individual colonies of Pocillopora spp determined by Structure (V.3.2) assuming location prior, admixture and correlation between loci. Individual colonies are organized by ORF/ITS2 type (top) and location (bottom) and the y-axis represents the probability of membership. Optimal number of clusters after Evanno et. al. (2005) was K = 2, but further analyses at higher values of K, separate all the 8 types in distinct groups.

Figure 15. Distance base phylogenetic reconstruction (a) and Bayesian tree build with Structurama (b). Both trees show relationships between 630 unique multilocus genotypes of Pocillopora from different types and geographic locations

63 Sympatric analyses with MLG data from 12 locations (TAN, HI, LZI, PAL, HAW,

CLI, GAL, GoC, BB, OX, REV, PA) were performed using Structure to explore the possibility of hidden population structure within each site and/or type. Data from Lizard and Heron Island as well as from all TEP locations was clumped in two groups; Great

Barrier Reef (GBR) and TEP respectively. When possible, all types found on each location were included, but the number of MLGs obtained in some types and/or sites was too small (< 5) for this analyses. Structure results supported the division in discrete groups previously observed, with the number of genetically homogeneous groups (K) matching the number of ORF/ITS2 types found on each location. For example, in

Tanzania, where type 5 was not found, two groups were defined (K = 2; Rst = 0.179; p =

0.01) each containing colonies from either Pocillopora type 1 or P. type 3. In the GBR, four types were found and MLG analyses separated colonies in four clusters (K=4; Rst =

0.174; p = 0.01). In other locations, individuals were separated in K = 3 (PAL, Rst =

0.215; p = 0.01), K = 4 (HAW, Rst = 0.29; p = 0.01) and K = 3 (TEP Rst = 0.311; p =

0.01) (Fig. 16).

Low evidence of hybridization between Pocillopora types

The levels of exchange of genetic material between the different Pocillopora types were low or not existent as indicated by pairwise analyses on NewHybrids. Within locations only in the TEP a few individuals (7 out of 342; 2 %) showed levels of admixture, while in all other locations there was no evidence of hybridization. Comparisons combining

MLG of colonies from the same type across locations revealed an even lower number of

64 hybrid individuals (5 out of 630; ≈1 %). All of these interspecies individuals showed higher levels of admixture (i.e. > 93 %) and were classified as F2 hybrids.

Figure 16. Sympatric Bayesian analysis performed with Structure. Analyses resulted showed the formation of distinct clusters corresponding to the genetic delineation of the types in the genus Pocillopora. Only locations with different types and enough samples per type are included. Colors correspond to the ORF/ITS2 types as depicted in the legend. Y-axes represent the probability of membership.

Highly supported phylogenetic reconstructions, clustering analyses and lack of hybridization among Pocillopora types suggest this scleractinian group is composed of at least eight distinct and reproductively isolated species.

Biogeography of the genus Pocillopora

The Pocillopora types delineated here showed distinctive distributions (Fig. 17). Four

Types, the most commonly found (1eyu, 1mea, 3 and 5), showed wide distributions

65 through the Indo-Pacific (Fig. 17). Among these, Pocillopora type 3 showed the widest distribution, being found in almost all sampled sites from Southern Tropical Eastern

Pacific (PA and GAL) to Eastern Africa (TAN), but was absent in Mexico, the Clipperton

Atoll and the GoC. Pocillopora type 5 was found in all but the TEP locations and Type

1eyu in all but TAN and Western Australia. Type 1mea was only present in reefs from

HAW and TAN, but it is likely to be present in intermediate locations (i.e. Indo-Pacific and Central Pacific). The other four types are more restricted in their distributions, with types 2 and 6 appearing endemic of Clipperton Atoll and Hawaii respectively. Type 4, exclusively found in Australia, also showed a limited distribution range. Only one colony of P. type 7 was found in Thailand, restricting our analysis for this particular species.

Pocillopora-Symbiodinium associations change across geographic locations

Based on the sequences of the ITS2 region and the non-coding region of the chloroplast psbA minicircle, fourteen species (9 from clade C and 5 from clade D) of Symbiodinium spp were found within Pocillopora. All Pocillopora types associate with at least one species of Symbiodinium clade C (i.e. Symbiodinium C1c, S. C1h, C42a, C1ee, C34, C1d,

C1c-f, C125/126, C1g) while P. types 1eyu, 3, 4 and 5 also associate with Symbiodinium

D species (i.e. Symbiodinium glinny, S. D1-4-6, D1a, D5 and D5a) (Fig. 17). All association appear specific at each location, but across regions the partnership is changed, for example Pocillopora type 1eyu, associates with two clade D species, S. glynni (TEP) and S. D5a (THA), and two clade C species, C1b-c (TEP) and C1c (HAW and GBR).

Calice-level morphometrics agree with genetics in HAW but not in the TEP

66 The morphometric analysis on individuals from four morphologically described species in Hawaii (P. damicornis, P. meandrina, P. ligulata and P. eyudoxi) indicated significant

(KW-ANOVA 25.04 p > 0.0001) differences in diameter of the calice (d). P. damicornis had the largest (1.190 ± 0.14 mm) calice diameters while P. eyudoxi showed the smallest

(0.750 ± 0.06 mm) (Fig. 18). The calice diameter was the only microskeletal character that resolved colony-level defined species; distance to the nearest calice (KW-ANOVA

22.74 p > 0.0064) distinguished P. ligulata from P. meandrina (LSD p = 0.0004) but none of the other species.

In the GoC, the average calice diameter (P. capitata 0.889 ± 0.160 mm, P. meandrina

0.889 ± 0.23 mm, P. damicornis 0.902 ± 0.16 mm) and distance to nearest calice (P. meandrina 0.273 ± 0.11 mm, P. damicornis 0.293 ± 0.10, P. capitata 0.294 ± 0.13 mm) showed little variation and no significant differences (KW-ANOVA d 0.15 p = 0.927 and di 1.77 p = 0.412) between morpho-species from the museum collections (Fig. 18).

Colonies collected from the field (and identified as Pocillopora type 1eyu) showed a slightly smaller calice diameter (0.888 ± 0.17 mm) and shorter distances between calices

(0.258 ± 0.08 mm), to those colonies in the museum, but none of these measurements showed significant differences between species (KW-ANOVA d 0.3 p = 0.960 and di

5.59 p = 0.133) (Fig. 18).

Figure 17. Geographic distribution of 8 Pocillopora types as delineated by genetic markers and ecological associations. Triangles demark locations were each type was collected while the dotted lines show the most likely geographical range of each type. Circles show other sampled locations where types were not found in this study.

In the combination of both datasets (i.e. Hawaii and the GoC), both diameter (KW-

ANOVA 36.21 p = 0.0004) and distance (KW-ANOVA 20.55 p = 0.0045) showed significant differences between species. The length of the diameter showed more differences between species than the distance to nearest calice. Specimens from the

MHNUBCS, La Paz, and those of P. ligulata from Hawaii did not show significant differences in diameter or distance to the nearest calice (Fig. 18). Agreements between morphology with the genetic separation (based on sequence and microsatellite loci data) of the Pocillopora spp. in Hawaii, and the genetic homogeneity observed in the GoC

(where only Pocillopora type 1eyu is found) are clear, but a morphological disparity

68 across locations is also evident with P. type 1eyu in the TEP showing similar diameters to

P. ligulata from Hawaii instead of the more genetically similar P. eyudoxi.

Figure 18. Morphometrical analysis of Pocillopora spp in Hawaii and the Gulf of Mexico. (a) Picture depicting the distances measured on five randomly selected calices, located approximately 2cm from the tip of the branch. (b) Kruskall-Wallis analyses and significant values (in parentheses) for three of data sets. (c) Diameters (in mm) of eight different groups of colonies from the GoC (light gray) and Hawaii (dark gray) bars correspond to the standard deviation, and letters indicate significantly different groups (p < 0.05).

69 Discussion

In a previous assessment of the diversity of Pocillopora (in the TEP), it became clear that there was a great deal of incongruence between the traditional morphologically based taxonomy of this genus and the distribution of genetic diversity (Pinzón, Lajeunesse,

2011). Three independent analyses (genetic, ecological and morphological) on

Pocillopora spp presented here reduced the species diversity in the genus to at least eight species.

Pocillopora diversity

Phylogenetic and population analyses showed strong congruence in the distinction of genetic groups within Pocillopora (Fig 13 and 14). We implemented different phylogenetic analyses (parsimony, likelihood, neighbor joining and bayesian for sequence data and bayesian and distance based for MLG) on two kinds of data (sequence and microsatellite MLG) obtaining similar trees, suggesting these distinct groups correspond to different species.

Only in few instances, phylogenetic and population genetic analyses showed incongruences in their results. Hybridization, incomplete lineage sorting and ancestral polymorphisms have been suggested to explain conflicts between phylogenetic analyzes in corals. Hybridization is believed to be one of the major forces of diversification of corals, creating an intricate reticulate pattern of evolution (Veron, 1995). In the

Caribbean, Acropora prolifera has been found to be a hybrid between A. palmata and A. cervicornis (Vollmer, Palumbi, 2002) an evolutionary event also observed among other

70 morphologically defined species (Knowlton et al., 1997; Levitan et al., 2004; Szmant et al., 1997; Wallace, Willis, 1994). In Pocilloporids hybridization was proposed between morphologically defined species in the TEP (Combosch et al., 2008), however analyses based on genetic, ecological and biogeographic observations (not skeletal characters) revealed little evidence of hybridization between genetically defined species in the region

(Pinzón, Lajeunesse, 2011). The current analysis supports our findings in the TEP and confirms recombination between Pocillopora spp. is not likely to happen (Kinzie III,

1996), contrary to other coral genera (i.e. Acropora) (Wallace, Willis, 1994), therefore the species delineation shown here is accurate and Pocillopora is composed of at least eight types or species.

Reproduction in Pocillopora spp making sense of inconsistencies

In a recent review on scleractinian reproduction, the species of Pocillopora appear along with species from three other genera (Oulastrea, Heliofungia and Goniastrea) as the only coral genera (4 out of 110) where colonies of the same species show mixed modes of reproduction (i.e. the same species has colonies broadcasting gametes and/or larvae)

(Baird et al., 2009). Among these genera, Pocillopora (Veron, Pichon, 1976), Goniastrea

(Ow, Todd, 2010) and Oulastrea (Yamashiro, 2000) exhibit a high degree of phenotypic plasticity in their skeletal characteristics in response to environmental conditions.

Unfortunately these morphological changes can obscure species delineations in these groups.

71 Table 3. Modes of reproduction in four morphologically identified Pocillopora species in different locations from the Indian Ocean, Indo-Pacific, and Tropical Eastern Pacific. Species* Location Mode of Observations Source Reproduction South Africa Broadcast Eggs with Kruger and (IO) symbiont Schelyer, 1998 Red Sea Broadcast NO symbiont on Sier and Olive P. verrucosa eggs 1994 Maldives Broadcast Fadllalah 1985 (IO) and Shlesinger, and Loya, 1985 P. meandrina HAW Broadcast Stimson, 1971 P elegans TEP Broadcast Glynn et al 1991 WA Brooder and Active zone – all Ward, 1992 Broadcaster corals broods Calm zone – broadcast, 2/5 brood WA Brooder Stodart 1983 GBR Brooder Tanner, 1996 LI Brooder Intercoloy Harriot, 1983 P. damicornis variability in planula release HAW Brooder Harrigan, 1972 Richmond and Jockiel, 1984 TEP Broadcaster Eggs with Glynn et al 1991 symbiont GoC Broadcaster Chavez-Romo and Reyes, 2007 * All species reported to be hermaphrodites. IO = Indian Ocean, HAW = Hawaii, WA = Western Australia, GBR = Great Barrier Reef, LI = Lizard Island, TEP = Tropical Eastern Pacific, GoC = Gulf of California.

To our knowledge, at least thirteen studies (Table 3) deal with reproductive patterns in four Pocillopora morphologically defined species (P. damicornis, P. elegans, P. meandrina and P. verrucosa). All Pocillopora spp. are hermaphrodites with polyps harboring both male and female gonads in different septa (Kruger, Schleyer, 1998; Sier,

Olive, 1994). In contrast to the majority of the species in the genus, P. damicornis

72 exhibits a “shift” in mode of reproduction, from brooding in the GBR and Hawaii to broadcasting in the TEP. Differences in mode, timing and seasonality of reproduction between species have been attributed to environmentally induced variation and errors in data interpretation of species identifications (Sier, Olive, 1994).

A pattern in mode of reproduction in Pocillopora is evident when geographic distributions and phylogenetic relationships of Pocillopora are considered. First, there is a correlation between brooding and restricted/endemic distributions, and broadcasting and wide distributions. Broadcasting is the reproductive mode in P. type 1mea (P. meandrina Hawaii) and P. type 1eyu and/or P. type 3 in the TEP, while brooding is for P. type 5 (P. damicornis in Hawaii) and likely for types 2, 4, 6 and 7. Larvae from broadcast spawning coral species are long lived and better adapted for long dispersal (Harrison,

Wallace, 1990) (Graham et al., 2008) than larvae from brooders. Brooding species are known to disperse close to maternal colonies (Underwood et al., 2007).

In Pocillopora, types 1 and 3 are widely dispersed with little or no isolation between geographically distant populations, while Type 5 is highly structure corresponding to geographic locations. The phylogenetic reconstructions in the genus placed types 2, 4, 6

(Figs. 13, 14 and 15) as well as 5, known brooder, in a single cluster, separated from the broadcasters, Types 1 and 3. Suggesting that either reproductive mechanism evolved in the genus after its divergence other sister genera. These observations (i.e. matches), with no a priori conception of the morphological species, seemed more natural and may result in a more comprehensive understanding of reproduction in Pocillopora. However, it is still necessary to re-assess the reproduction patterns of the genetically defined species to confirm these hypotheses.

Do Pocillopora symbioses form a geographic mosaic of co-evolution?

During co-evolution, different species interact with each other causing reciprocal evolutionary changes (Thompson, 2005). This interaction is initially regulated by local and species-specific conditions. As populations disperse to new habitats, a mosaic of interactions forms as the residents and immigrants adapt to one another (Thompson,

2005). The geographic mosaic of co-evolution (GMC) is perceived as a change in specific associations/relations through a distribution range of a species. Coral-algal associations in Pocillopora show patterns of distinctions between locations. For example,

P. type 1mea in the TEP is associated with Symbiodinium C1b-c and S. glynni, in HAW and Australia with S. C1c and in Thailand with S. C1c and S. D5a. Similar situations are observed in other widely distributed Pocillopora spp. (types 1eyu, 3 and 5) suggesting

GMC might be an important evolutionary force for the Pocillopora-Symbiodinium associations.

Recently many interacting biological associations have been proposed to be regulated by the GMC model (Gomulkiewicz et al., 2007), with most failing to prove the role of the three processes required by the model: co-evolutionary hot and cold spots, selection mosaics and trait remixing (Thompson, 2005). In our case the geographic coverage of our samples may be too limited to distinguish hot/cold spots in Pocillopora-

Symbiodinium co-evolution, but the associations do change from location to location, in some cases even to a different Symbiodinium clade (C / D). Changes in the symbiont species can also be perceived as selective pressures for a best fit partnership depending on environmental conditions, for example P. type 5 associates with S. glynni, a thermally

74 tolerant species, in locations were light penetration is low in Palau, while in other areas it associates with species of Symbiodinium clade C The model (i.e. GMC) seemed the best fit to our observations, but a more extensive collection for each type is necessary.

References

Abramoff MD, Magelhaes PJ, Ram SJ (2004) Image Processing with ImageJ. Biophotonics International 11, 36-42.

Anderson E, Thompson E (2002) A model-based method for identifying species hybrids using multilocus genetic data. Genetics 160, 1217-1229.

Baird A, Guest J, Willis B (2009) Systematic and biogeographical patterns in the reproductive biology of scleractinian corals. Annual Review of Ecology, evolution and Systematics 40, 552-571.

Benzoni F (2006) Psammocora albopicta sp. nov., a new species of Scleractinian coral from the Indo-West Pacific (Scleractinia; Siderastreidae). Zootaxa 1358, 49-57.

Benzoni F, Stefani F, Pichon M, Galli P (2010) The name game: morpho-molecular species boundaries in the genus Psammocora (Cnidaria, Scleractinia). Zoological Journal of the Linnean Society.

Benzoni F, Stefani F, Stolarski J, et al. (2007) Debating phylogenetic relationships of the scleractinian Psammocora: molecular and morphological evidences. Contirbutions to Zoology 76, 35-54.

Bickford D, Lohman D, Sodhi N, et al. (2007) Cryptic species as a window on diversity and conservation. Trends in Ecology & Evolution 22, 148-155.

Budd A, Romano S, Smith N, Barbeitos M (2010) Rethinking the Phylogeny of Scleractinian Corals: A Review of Morphological and Molecular Data. Integrative and Comparative Biology, icq062v061.

Budd A, Stolarski J (2009) Searching for new morphological characters in the systematics of scleractinian reef corals: comparison of septal teeth and granules between Atlantic and Pacific Mussidae. Acta Zoologica 90, 142-165.

Budd AF (1990) Longterm patterns of morphological variation within and among species of reef-corals and their relationship to sexual reproduction. Systematic Biology 15, 150-165.

75 Cantera JR, von Prahl H, Escobar JC, Peña E (1989) Sistematica de los corales del género Pocillopora del pacífico colombiano utiizando taxonomía numérica. Revista de Biologia Tropical 37, 23-28.

Cavalli-Sforza LL, Edwards AWF (1967) Phylogenetic analysis: models and estimation procedures. Evolution 21, 550-570.

Claridge MF, Dawah HA, Wilson MR (1997) Species: The Units of Biodiversity, First edn. Springer, London.

Combosch D, Guzmán H, Schuhmacher H, Vollmer SV (2008) Interspecific hybridization and restricted trans-Pacific gene flow in the Tropical Eastern Pacific Pocillopora. Mol Ecol 17, 1304-1312.

Crossland C (1952) Madreporaria, Hydrocorallinae, Heliopora and Tubipora, Great Barrier Reef Expedition 1928-29. Scientific reports 6, 85-257.

Drummond A, Ashton B, Cheung M, et al. (2009) Geneious v4.8, Available from http://www.geneious.com/.

Earl DA (2009) Structure Harvester v0.3, from website: http://users.soe.ucsc.edu/~dearl/software/struct_harvest/.

Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology 14, 2611-2620.

Falush D, Stephens M, Pritchard JK (2003) Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics 164, 1567-1587.

Felsenstein J (1995) PHYLIP (Phylogeny Inference Package) Version 3.57c. http://www.nih.go.jp/~jun/research/phylip/main.html.

Forsman Z, Barshis D, Hunter C, Toonen R (2009) Shape-shifting corals: molecular markers show morphology is evolutionarily plastic in Porites. BMC Evolutionary Biology 9.

76 Forsman ZH, Martinez JA, Maragos JE, Toonen RJ (2010) Zootaxa, Resurrection of Porites hawaiiensis Vaughan, 1907; a Hawaiian coral obscured by small size, cryptic habitat, and confused taxonomy. Zootaxa, 67-68.

Fukami H (2008) Short Review: molecular phylogenetics analyses of reef corals. Galaxea 10, 47-55.

Fukami H, Budd AF, Paulay G, et al. (2004) Conventional taxonomy obscures deep divergence between Pacific and Atlantic corals. Nature 427, 832-835.

Fukami H, Chen C, Budd A, et al. (2008) Mitochondrial and nuclear genes suggest that stony corals are monophyletic but most families of stony corals are not (Order Scleractinia, Class Anthozoa, Phylum Cnidaria). PLoS ONE 3.

Glynn PW, Ault J (2000) A biogeographic analysis and review of the far eastern Pacific coral reef region. Coral Reefs 19, 1-23.

Gomulkiewicz R, Drown D, Dybdahl M, Godsoe W (2007) Dos and don'ts of testing the geographic mosaic theory of coevolution. Heredity 98, 249-258.

Graham E, Baird A, Connolly S (2008) Survival dynamics of scleractinian coral larvae and implications for dispersal. Coral Reefs 27, 529-539.

Harrison P, Wallace CC (1990) Reproduction, dispersal and recruitment of scleractinian corals. In: Coral Reefs, Ecosystems of the World 25, pp. 133-207. Elsevier, New York.

Hubisz MJ, Falush D, Stephens M, Pritchard JK (2009) Inferring weak population structure with the assistance of sample group information. Molecular Ecology Resources 9, 1322-1332.

Huelsenbeck J, Andolfatto P (2007) Inference of population structure under a Dirichlet process prior. Genetics 175, 1787-1802.

Huelsenbeck J, Ronquist F (2001) MrBayes: a program for the Bayesian inference of phylogeny. Bioinformatics 17, 754-755.

Huxley J (1938) Clines: an auxiliary taxonomic principle. Nature 142, 219-220.

Kinzie III R (1996) Modes of speciation and reproduction in Archaeocoeniid corals. Galaxea 13, 47-64.

Knowlton N (1992) Thresholds and multiple stable states in coral reef community dynamics. American Zoologist 32, 674-682.

77 Knowlton N, Mate JL, Guzmán HM, Rowan R, Jara J (1997) Direct evidence for reproductive isolation among the three species of the Montastraea annularis complex in Central America (Panamá and Honduras). Marine Biology 127, 705- 711.

Kruger A, Schleyer M (1998) Sexual reproduction in the coral Pocillopora verrucosa (Cnidaria: Scleractinia) in KwaZulu-Natal, South Africa. Mar Biol 132, 703-712.

LaJeunesse T (2002) Diversity and community structure of symbiotic dinoflagellates from Caribbean coral reefs. Marine Biology 141, 387-400.

LaJeunesse TC, Loh W, van Woesik R, et al. (2003) Low symbiont diversity in southern Great Barrier Reef Corals, relative to those of the Caribbean. Limnol. Oceanog. 48, 2046-2054.

LaJeunesse TC, Thornhill DJ, Cox E, et al. (2004) High diversity and host speciﬁcity observed among symbiotic dinoﬂagellates in reef coral communities from Hawaii Coral Reefs 23, 596-603.

LaJeunesse TC, Trench RK (2000) Biogeography of two species of Symbiodinium (Freudenthal) inhabiting the intertidal sea anemone Anthopleura elegantissima (Brandt). Biological Bulletin (Woods Hole) 199, 126-134.

Levitan D, Fukami H, Jara J, et al. (2004) Mechanisms of reproductive isolation among sympatric broadcast-spawning corals of the Montastraea annularis species complex. Evolution 58, 308-323.

Locke JM, Weil E, Coates KA (2007) A newly documented species of Madracis (Scleractinia: Pocilloporidae) from the Caribbean. Proceedings of the Biological Society of Washington 120, 214-226.

Lopez JV, Kersanach R, Rehner SA, Knowlton N (1999) Molecular determination of species boundaries in corals: genetic analysis of the Montastraea annularis complex using amplified fragment length polymorphisms and a microsatellite marker. Biological Bulletin 196, 80-93.

Lopez JV, Knowlton N (1996) Discrimination of species in the Montastraea annularis complex using multiple genetic loci 2, 1613-1618.

Medina M, Weil E, Szmant A (1999) Examination of the Montastraea annularis species complex (Cnidaria: Scleractinia) using ITS ad CO2 Sequences. Marine Biotechnology 1, 89-97.

Miller KJ (1994) Morphological variation in the coral genus Platygyra: environmental influences and taxonomic implications. MEPS 110, 19-28.

Ow Y, Todd P (2010) Light-induced morphological plasticity in the scleractinian coral Goniastrea pectinata and its functional significance. Coral Reefs 29, 797-808.

78 Peakall R, Smouse PE (2006) Genalex 6: genetic analysis in Excel. Population genetic software for teaching and research. Mol Ecol Notes 6, 288-295.

Pinzón JH, Lajeunesse TC (2011) Species delimitation of reef building corals using nucleotide sequence phylogenies, population genetics, and symbiosis ecology. Molecular Ecology 20, 311-325.

Powers D (1970) A numerical taxonomic study of Hawaiian Reefs Corals. Pacific Science 24, 180-186.

Powers D, Rohlf F (1972) A numerical taxonomic study of Caribbean and Hawaiian reef coral. Systematic Zoology 21, 53-64.

Pritchard JK, Stephens M, Donnelly P (2000) Inferece of population structure using multilocus genotype data. Genetics 155, 945-959.

Raymond M, Rousset F (1995) GENEPOP (version1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity 86, 248-249.

Rosenberg N (2004) DISTRUCT: a program for the graphical display of population structure. Molecular Ecology Notes 4, 137-138.

Rousset F (2008) Genepop’007: a complete re-implementation of the genepop software for Windows and Linux. Molecular Ecology Resources 8, 103-106.

Sier C, Olive P (1994) Reproduction and reproductive variability in the coral Pocillopora verrucosa from the Republic of Maldives. Mar Biol 118, 713-722.

Souter P (2010) Hidden genetic diversity in a key model species of coral. Mar Biol.

Starger CJ, Yeoh SSR, Dai C-F, Baker AC, DeSalle R (2007) Ten polymorphic STR loci in the cosmopolitan reef coral, Pocillopora damicornis. Molecular Ecology Resources 8, 619-621.

Stefani F, Benzoni F, Pichon M, Cancelliere C, Galli P (2008a) A multidisciplinary approach to the definition of species boundaries in branching species of the coral genus Psammocora (Cnidaria, Scleractinia). Zool Scr 37, 71-91.

Stefani F, Benzoni F, Pichon M, Mitta G, Galli P (2008b) Genetic and morphometric evidence for unresolved species boundaries in the coral genus Psammocora (Cnidaria; Scleractinia). Hydrobiologia 596, 153-172.

Stobart B (2000) A taxonomic reappraisal of Montipora digitata based on genetic and morphometric evidence. Zoological Studies 39, 179-190.

Swofford D (2000) PAUP*, Phylogenetic Analysis Using Parsimony (*and other methods), Version 4.0b.

79 Szmant AM, Weil E, Miller MW, Colon DE (1997) Hybridization within the species complex of the scleractinian coral Montastraea annularis. Marine Biology 129, 561-572.

Thompson JN (2005) The geographic mosaic of coevolution The University of Chicago Press, Chicago.

Todd PA (2008) Morphological plasticity in scleractinian corals. Biol. Rev. 83, 315-337.

Underwood JN, Smith LD, van Oppen MJH, Gilmour JP (2007) Multiple scales of genetic connectivity in a brooding coral on isolated reefs following catastrophic bleaching Molecular Ecology 16, 771-784.

Veron JEN (1995) Corals in Space and Time. The biogeography and evolution of the scleractinian Southwood Press, Australia.

Veron JEN (2000) Corals of the world Australian Institute of Marine Science, Twonsville.

Veron JEN (2002) New species described in Corals of the world Australian Institute of Marine Science, Twonsville.

Veron JEN, Pichon M (1976) Scleractinia of Eastern Australia Part I. Australian Institute of Marine Science Canberra, 86.

Wallace CC, Willis BL (1994) Systematics of the coral genus Acropora: Implications of new biological findings for species concepts. Annual Review of Ecology and Systematics 25, 237-262.

Weil E (1992) Genetic and morphological variation in the Caribbean and Eastern Pacific Porites (Anthozoa, Scleractinia). Preliminary results 2, 643-655.

80 Weil E, Knowlton N (1994) A multi character analysis of the Caribbean coral Montastraea annularis (Ellis and Solander, 1786) and its two sibling species, M. faveolata (Ellis and Solander, 1786) and M. franski (Gregory, 1895). Bulletin of Marine Science 55, 151-175.

Weil E, Vargas WL (2009) Comparative aspects of sexual reproduction in the Caribbean coral genus Diploria (Scleractinia: Faviidae). Mar Biol, 1-14.

Wiens JJ, Servedio MR (2000) Species delimitation in systematics: inferring diagnostic differences between species. Proceedings of the Royal Society B: Biological Sciences 267, 631-636.

Wolstenholme J, Wallace C, Chen C (2003) Species boundaries within the Acropora humilis species group (Cnidaria; Scleractinia): a morphological and molecular interpretation of evolution. Coral Reefs 22, 155-166.

Yamashiro H (2000) Variation and Plasticity of Skeletal Color in the Zebra Coral Oulastrea crispata. Zoological Science 17, 827-831.

Young ND, Healy J (2003) GapCoder automates the use of indel characters in phylogenetic analysis. BMC Bioinformatics 4, 1-6.

CHAPTER 3: HIGH VARIANCE IN CLONALITY BETWEEN LOCAL POPULATIONS OF THE ECOLOGICALLY DOMINANT POCILLOPORA IN THE GULF OF CALIFORNIA

Abstract

In nature several species reproduce sexually and asexually. The levels of both modes of reproduction significantly influences population demographics, genetic make-up and evolution in these species. Scleractinian corals have several sexual and asexual reproductive mechanisms, a feature that has helped them to colonize and dominate fairly unfavorable environments. The Pacific Coast of America, from the Gulf of California (GoC) to Ecuador, including oceanic islands (Clipperton Atoll, Revigagilledo, Cocos, Malpelo and Galapagos) encompass a region, known as Tropical Eastern Pacific (TEP), where conditions are considered suboptimal for coral development. Reefs in the TEP are largely dominated by the genus Pocillopora with only one species in the GoC. Pocillopora has been reported in the TEP to reproduce with both sexual and asexual mechanisms, but further research into the relative input of each mechanism in the population structure of TEP reef communities is not known. Here we presented data that supports the existence of sexual reproduction, but also shows how close, genetically connected, populations of Pocillopora can be dominated by different mechanisms of reproduction. These observations suggest that although conditions in the TEP may be disadvantageous for corals, Pocillopora type 1 thrives by using different reproductive strategies.

82 Introduction

Few animals are able to propagate asexually. Several factors, including environmental conditions, local catastrophes, competition/exploitation (Browne et al., 1988; Foster et al., 2007; Jokela et al., 2003; Scheu, Drossel, 2007; Vrijenhoek, 1979; Vrijenhoek,

1998), determine whether individuals of a population engage more frequently in sexual versus asexual reproduction. The possession of both modes of reproduction significantly influences population demographics, genetic make-up and evolution in a population

(Eckert et al., 2008; Halkett et al., 2005; Jackson, 1986). Furthermore, species populations in one geographic region may exhibit one condition of reproduction while in another region the other mode is dominant (Vandel, 1928; Vorburger, 2006). Successful gene combinations best suited to a particular environment are perpetuated by clonal propagation, whereas sex increases diversity by producing new recombinant genomes.

Populations that are primarily clonal differ from those that are primarily asexual in their mode of dispersal, response to environmental change and rate of propagation. While there are many reasons why clonality may inhibit long-term evolutionary success, over the short term and under conditions of environmental stability but where few individuals exist clonal propagation may allow an organism to colonization empty and/or recently vacated habitats (Crow, 1992)

Scleractinian corals are notable examples of animals that exhibit sexual and asexual life history strategies (Harrison, Wallace, 1990). Species with branching colonies can reproduce and spread locally and rapidly through fragmentation. The relative significance of sexual vs. clonal modes of propagation has been widely investigated in

83 branching species of the genus Pocillopora (Adjeroud, Tsuchiya, 1999; Ayre et al., 1997;

Miller, Ayre, 2004; Ridgway et al., 2001; Stoddart, 1984a; Stoddart, 1984b; Stoddart,

1984c; Whitaker, 2006; Yeoh, Dai, 2010). Apparently, variation in clonality exists among and between populations in regions throughout the Pacific. The eastern Pacific is the one region in the world where Pocillopora dominate shallow coral communities down to depths of 6 to 7 meters (Glynn 1976; Cortes 1997). These corals typically broadcast spawn in late summer (Chávez-Romo, Reyes-Bonilla, 2007; Glynn et al., 1991) but the relative importance of reproductive mode in structuring adult populations has not been assessed in the region. Because of the remote location and inhospitable conditions of the

TEP, it is possible that populations of Pocillopora may exhibit high levels of clonality throughout the region.

Recent genetic and ecological analysis of morphospecies in the Gulf of California

(GoC) found that the Pocillopora there comprise a single genetically cohesive population, or species, despite major morphological differences among colonies.

Populations of Pocillopora type 1 (Pinzón, Lajeunesse, 2010) reach their highest latitudinal distribution in the GoC where the annual temperature range is large.

Therefore, it is expected that clonality may be high in populations from this region. To examine this possibility colonies were sampled in several circular plots from two sites subjected to different water current patterns, one in the protected leeward side of a small

Island and the other at an exposed site 10 kilometers away. Population genetic markers involving 7 microsatellite loci were used to determine clone mate abundance and relative their proximity among random colonies sampled in each plot. These data are then

84 discussed in the context of current theory regarding the importance of clonal vs sexual propagation in the ecology and evolution of Scleractinia.

Materials and Methods

Collection sites and sampling

Two reefs (Punta Galeras - PG and Isla Gaviota - IG) off La Paz, Baja California Sur,

Mexico were surveyed for this research (Fig 19). These reefs differ from each other in extension and environmental conditions. Punta Galeras has a larger platform with variable environments resulting from rocky formations mixed with sandy bottoms, the area is expose to wave action and oceanic waters that cross through the San Lorenzo canal (Del Monte-Luna et al., 2005). Isla Gaviota is inside La Paz bay, protected from oceanic waters and wave action. Its platform is flat, shallow (< 3m) and narrow, allowing the reef to extend to a maximum of 20 m from the coastline. In a previous study, it was established that the only Pocillopora species in both reefs was type 1 (Pinzón,

Lajeunesse, 2010), and multilocus microsatellite genotypes suggested the presence of both sexual and asexual reproduction. Sample collection at that time was not designed to assess the clonal structure of these populations. In this project, we randomly deployed 3 circular plots on each reef (10 m radious) separated by at least 30 m following the method described in Baums et al. 2006. Latitude and longitude coordinates of the centers of each plot were recorded with a GPS (Garmin, Etrex Vanture HC).

In order to detect both common and rare genotypes, coordinates (bearing and distance from the center of the plot) for all six plots were randomly generated in Excel (Microsoft,

85 Redmond WA USA). To avoid sampling the same colony more than once, as much as possible, the precision was set to 1 degree and 50 cm. In the field, a compass was used to measure bearings and a measuring tape for the distances. Colonies underneath each coordinate set were sampled, if no colony was found in that point, the coordinates were eliminated and the next pair of random numbers was used. Samples constituted a small fragment (1 - 3 cm) removed from the colonies with bone scissors. A total of 20 samples per plot were collected, fixed in DMSO 20% (Seutin et al., 1991) and store in a freezer at

-20˚C.

DNA extractions and microsatellite genotyping

Genomic DNA extractions were performed using the Wizard Genomic Purification Kit

(Promega). With this kit, a small piece of skeleton and tissue was pulverized with glass beads (Ceroglass, Columbia USA) in nuclei lysis solution. Coral tissue was then incubated with proteinase K at 65˚C for 1 hour. After incubation, the solution was centrifuged and the DNA precipitated with isopropanol/ethanol. The excess of alcohol was air dried and the resulting DNA diluted in 75 µl of distilled water to an average concentration of 5 - 50 ng/ul and stored at -20˚C (LaJeunesse et al., 2003).

Multilocus genotypes were obtained using the same set of seven microsatellite loci

(Pd3-002, Pd3-005, Pd2-006, Pd2-007, Pd3-008, Pd3-009 and Poc40) used in Pinzon and

LaJeunesse (in review). These loci are polymorphic and when combined showed a low probability of identity (4.2 x 10-6) indicating they are very powerful in distinguish different genets and detecting ramets of the same genet. Loci amplifications as well as allele scores were determined following Pinzon and LaJeunesse (in review). Basically all

86 loci with a labeled primer were amplified in different PCR reactions with the same protocol ((94˚C 2 min)1 – (94˚C 15 sec – Ta 15 sec – 72˚C 30 sec)31 - (72˚C – 5 min)1) but using the appropriate annealing temperature (Ta) for each locus. Loci with different dye labels were mixed in pairs before being analyzed on a ABI Hitachi 3730XL genetic analyzer, using LIZ - 500 (Applied Biosystems) as a size standard. Alleles were identified and scored with GeneMarker (Softgenetics, State College PA USA) and exported to Excel (Microsoft, Redmond WA USA) for further analysis.

Clonal structure analysis

Clonal structure was assessed with the software GenClone v. 2.0 (Arnaud-Haond,

Belkhir, 2007), unless otherwise specified. This package includes all statistical procedures suggested for the study of clonality in natural populations (Arnaud-Haond et al., 2007). The first step was to determine the genotypic and clonal identity of each sample, followed by estimating the clonal diversity and the spatial arrangements of the genets.

In order to assign samples to a particular group (i.e. clones or clonal lines), the probability of identical multilocus genotypes (MLG) was calculated using Psex

(probability of identical MLG coming from different reproductive events) with a Monte

Carlo procedure to determine the power of the combined loci to detect distinct MLG.

Furthermore, an analysis of pairwise distances between MLG (under the microsatellite matrix analysis in GenClone) was used to detect scoring errors and/or somatic mutations.

As a result of the identification of non-repeated MLG, single allele differences between

87 MLGs and corrections in scoring errors, samples were assigned to different unique multilocus genotypes, and repeated MLG were removed from further analyses.

The clonal structure of Pocillopora type 1 populations was focused on three major indices: clonal richness (R), Simpson’s evenness (V) and the distribution of clonal membership (Power law - Pareto), since use of similar indices may produce redundant results (Arnaud-Haond et al., 2007).

Genotypic diversity was assessed with the clonal richness index (R = Ng - 1 / N – 1), where Ng is the number of unique MLG, and N is the total number of samples. This index varies from zero, when all samples belong to a single clone, to one, each sample representing a unique MLG (Arnaud-Haond et al., 2005; Arnaud-Haond et al., 2007).

Clonal evenness describes equitability in the abundances of clonal lineages in the populations (Arnaud-Haond et al., 2007) and it was estimated with the Simpson’s evenness complement index (V = (D – Dmin) / (Dmax – Dmin)), where Dmin and Dmax being the approximate minimum and maximum values of the Simpson’s complement, V values range from clonal (zero) to equal abundances (one).

Comparing MLG frequencies distributions to a Pareto distribution can elucidate the relative distribution of clonality on each sample unit. The parameter β, describes the slope of the Pareto distribution and indicates the proportion in which the clones are distributed into different MLG’s. A larger β indicates more even distributions, a smaller β suggests more repeated frequencies in different groups, a signal of a less clonal population (Arnaud-Haond et al., 2007).

In addition to the clonal structure, spatial arrangements of the MLG were assessed using two indices (aggregation - Ac and edge effect - Ee), a spatial autocorrelation

88 analysis and with the average distance between ramets of the same genet. The aggregation index (Ac) tests whether or not individuals from the same MLG are spatially agglomerated. It is based on the probabilities of clonal identities between neighboring individuals. It ranges from zero (random distribution) to one which indicates clonal aggregation (Arnaud-Haond et al., 2007). The edge effect (Ee - a ratio of the difference between average clonal and total distances to the total average distance) tests the possibility of rare MLG to be found towards the edges of the sampled area (i.e. circular plot), therefore overestimating clonal diversity. The null hypothesis of random distribution of all individuals, is tested with a permutation procedure (Arnaud-Haond et al., 2007) set to 1000 permutations.

The spatial structure (i.e. aggregation) of individuals can influence the reproduction strategy and evolution of a population (Smouse, Peakall, 1999). In marginal populations where the lack of gene flow may be high, it is expected to have aggregated distribution of individuals. Here we applied the test of autocorrelation for multilocus genotypes proposed by Smouse and Peakall (1999) as implemented in GenAlex v. 6.3 (Peakall,

Smouse, 2006). Finally, calculations of average distances between colonies belonging to the same genotype were done in Excel (Microsoft, Redmond WA USA).

Population structure

Analyses of population differentiation were performed on a total of 66 unique genotypes after removing repeated MLG. Genetic differentiation between plots and reefs was tested using an AMOVA executed in GenAlex v. 6.3 (Peakall, Smouse, 2006), Rst values were calculated to determine the degree of similarity/differentiation between sampled

89 populations. This procedure was repeated with corrected frequencies after null alleles were detected in three loci (Pd3-002, Pd2-006 and Pd3-009) using Microchecker (van

Oosterhout et al., 2004). Exact tests for Hardy-Weinberg Equilibrium (HWE) and heterozygote deficiencies were performed in Genepop on the web (Raymond, Rousset,

1995; Rousset, 2008). Inbreeding rates per plot, reef and in total were assessed with the

RMES software (David et al., 2007). Additionally a Bayesian protocol executed by

Structure V3.3 (Pritchard et al., 2000) was performed to explore the assignment of individuals to different clusters (i.e. plots/reefs). Conditions were set to 1,000,000 chain length after a burn-in of 100,000 assuming admixture and using the location prior option.

Simulations were run between K = 1 to K =10 with 5 replicates for each K value. The optimal K was detected with Structure Harvester (Earl, 2009) and graphs were drawn in distruct (Rosenberg, 2004) after all replicates for each K were analyzed in Clumpp

(Jakobsson, Rosenberg, 2007). Finally, phylogenetic relations between circular plots were constructed in TPFGA (Miller, 1997) using a UPGMA method and Nei’s unbiased distance (Nei, 1978), branch support values were obtained after a 1000 bootstrap.

Results

Clonal identity and structure

Multilocus genotypes were obtained for 119 samples (one sample from IG2 was lost).

The minimum number of loci suggested for the assessment of clonality in these populations was six; therefore the resolution of the set of microsatellites loci (seven) used in this study is enough to distinguish unique genotypes of Pocillopora type 1 from these

90 reefs. Detection of full genotype matches and possible somatic mutations (1 allele difference between MLG) allowed the identification of 66 unique genotypes in total. The number of genets varied between locations (Table 4; Fig. 19). In PG plots genets were more than double (15 to 17) of those found in IG plots (6 to 7). Ten clonal lines were observed; nine of these were found in different PG plots while only one dominated all plots in IG. Ramet distributions were variable. A single clone dominated IG with large abundances, while in PG genetically identical ramets were restricted to the same plot in lower (2 to 4 colonies) abundances (Fig. 19).

Table 4. Estimates of the indices of clonal structure for each of the circular plots samples in two reefs off La Paz, Gulf of California Isla Gaviota Punta Galeras

IG1 IG2 IG3 PG1 PG2 PG3

Sample size (N) 20 19 20 20 20 20

Richness Unique Genotypes (Ng) 6 6 7 15 17 15

R = (Ng-1 / N-1) 0.263 0.278 0.316 0.737 0.842 0.737

Evenness Simpson (V) 0.000 0.000 0.000 0.857 0.472 0.514

Pareto ------1.756 -1.104

* Edge Effect (Ee) 0.185 -0.052 0.050 0.106 0.055 -0.034 Spatial * Aggregation (Ac) -0.092 0.122 0.059 0.062 0.132 0.055

Genetic Observed (Ho) 0.410 0.261 0.362 0.388 0.500 0.429 diversity Expected (He) 0.550 0.419 0.522 0.434 0.494 0.404

---- = Not estimated; * = significant after Bonferroni corrections

Figure 19. Geographical location of the sampling sites showing the circular plots in Punta Galeras (a) and Isla Gaviota (b) off La Paz, Mexico, in the Gulf of California (GoC). The exact location of the colonies and whether or not they represent a single ramet (squares) or several ramets (i.e. clones and hexagons) of a genet (circles) is also presented. Each of the black squares represents a unique multilocus genotype (MLG) while circles with the same code represent clonal lines. Clone symbols are organized from left to right to correspond to the alphabetical order in table 5.

92 Richness (R) varied from 0.263 (IG 1) to 0.842 (PG 2) with an overall average of

0.529 ± 0.269 (Table 4). Higher richnesses were found in PG (0.772 ± 0.060) while IG showed significant lower (0.286 ± 0.027) values (t-test p = 0.001). Higher richness in PG indicated that sexual reproduction was the predominant mechanism in this population while in IG asexual reproduction was the principal. Dominance of a single clone in IG reduced the average evenness (V) to 0.562 ± 0.026 while PG showed higher values

(0.943 - 0.978) (Table 4). Lower richness and evenness and dominance of a single genotype in IG, as well as the presence of only two frequency groups in PG 1 impeded the estimation of the β parameter for these plots. Thus the slope of the Pareto distribution was only estimated in PG 2 (-1.756) and PG 3 (-1.104), in both cases these values reflected the observed low clonality in Punta Galeras (Table 4).

Significant edge effects after Bonferroni corrections were observed only in PG 1 (Ee

= 0.10; p = 0.002), but this plot did not show significant levels of clonal aggregation (Ac

= 0.062; p = 0.334) (Table 4). Non-random distributions of clonemates were found only in PG 2 (Ac = 0.132; p = 0.006), all other plots did not show significant levels (p < 0.05) of aggregation. Spatial autocorrelation analyses on each plot did not reveal deviation from random distributions (Fig. 20), in any of the plots.

The average overall pairwise distance between clonemates was 8.00 ± 2.874 m. The maximum (12.523 m) and minimum (3.013 m) distances were found in PG. Clonemates in IG plots were in average 6.637 ± 0.666 m apart (Table 5), but this clone dominates through the reef and the distances between its ramets across the reef, varied from 1.03 m to 90.9 m.

Table 5. Number of ramets and average distance (m) between ramets for each clonal line found in two reefs of La Paz, Mexico, Punta Galeras (PG) and Isla Gaviota (IG). Clones are organized to match the symbols from left to right in the legend in figure 19. Clone Plot Ramets Distance (St Dev)

A PG 1 2 7.653

B PG 1 2 5.196

C PG 1 2 3.013

D PG 1 2 12.523

E PG 1 2 11.724

F PG 2 2 8.000

G PG 2 3 6.387 (4.400)

H PG 3 4 11.542 (5.624)

I PG 3 3 9.864 (4.256)

IG 1* 15 6.222 (3.010)

J IG 2* 14 6.497 (3.453)

IG 3* 14 7.472 (3.484)

* Clone J from IG was found across all plots with a maximum distance on 90.9 m between two colonies one from IG1 and the other from IG3

Figure 20. Spatial autocorrelation analyses of the distribution of Pocillopora genotypes in each circular plot off La Paz. Dotted arrows indicate the 97.5% significance interval. Arrows in Punta Galeras 1 (PG1) and 2 (PG2) indicate the only two instances when the distributions deviate from random.

Population structure

Pairwise comparisons between plots (Rst = -0.030; p = 0.960) and reefs (Rst = -0.009; p =

0.530) did not reveal any significant difference in the population structure between IG and PG. These values can be overestimated by the presence of null alleles, which were significant for three loci (Pd3-002, Pd2-006 and Pd3-009) and were not in HWE (p <

0.05). Significant null allele frequencies did not result from inbreeding (p > 0.05) or amplification mistakes (i.e, homozygous samples were amplified under different conditions resulting in the same genotype). Rst calculations with corrected genotypes for the loci with null alleles did not reveal any difference between reefs (Rst = 0.023; p =

0.100) suggesting samples collected in both reef belong to the same population (Table 6).

Table 6. List of loci with significant null alleles on each reef population and overall, 2 inbreeding coefficient (g ) as calculated by RMES and estimated Rst between reefs (IG and PG) with (Rst*) and without (Rst ) corrected frequencies for null alleles. 2 Reef Loci with g Rst Rst*

Null Alleles

PG Pd3-002 -0.058 (s = 0.000)

Pd2-006 -0.009 (p = 0.530) 0.023 (p = 0.100) Pd3-009

IG Pd2-006 0.146 (s = 0.146)

Overall Pd3-002 -0.027 (s = 0.000)

Pd2-006

Pd3-009

96 The Bayesian simulations and the UPGMA phylogeny showed congruent results with those from the AMOVA and the Rst values, with an optimal number of genetically different clusters (i.e. populations) set to one (K = 1). The maximum unbiased distance

(Nei’s unbiased) detected in the construction of the UPGMA phylogeny was low (0.036), placing PG 2, PG 3, and IG 1 in a single cluster in the tree (Fig 21) with relative low distances from the other plots.

Figure 21. Bayesian assignment of individual genotypes (a) detected in Pocillopora type I from La Paz, Baja California. Phylogenetic tree bases Nei’s unbias distance (b) of the same samples. Repeated multilocus genotypes were removed for both analyses.

In summary, our data suggests both locations (IG and PG) contain a single

Pocillopora type 1 population, but each reefs has a distinct clonal structure. IG is highly

97 clonal, dominated by a single clone that expands through the length of the reef, while PG showed high genotypic diversity with few clones, a sign of a sexual population.

Discussion

The study of the genetic variability and connectivity within and between populations is important to understand the biology, ecology and evolution of species (Hubby, Lewontin,

1966; Stoddart, 1984a; Whitaker, 2006). In scleractinian corals the relative importance of sexuality vs. asexuality has been found to be variable across species, reproductive strategies (i.e. spawning vs. brooding) and geographical locations (Ayre, Dufty, 1994;

Ayre, Hughes, 2000; Ayre et al., 1997; Ayre, Willis, 1988; Baums et al., 2006; Benzie et al., 1995; Sherman et al., 2006). This is the first study to explore the relative predominance of the sexual mechanism in the populations of any Pocillopora species using fine scale, codominant markers following a methodology designed to assess clonality in natural populations (Arnaud-Haond et al., 2007). To our knowledge only two studies have address similar questions and used the same methodology presented here.

The first found populations of Acropora palmata have different reproductive characters driven by local habitat conditions within each population (Baums et al., 2006). The second, also in the Caribbean, studied the massive Montastraea annularirs from

Honduras. In M. annularis, sexual reproduction seemed more important, but low levels of asexual input were found in association with major physical disturbances (Foster et al.,

2007). Environmental differences between reefs (i.e. PG vs. IG) exert an important influence on the reproductive structure of Pocillopora type 1 populations in La Paz.

98 Sexual populations may be favored by more active/expose conditions than those from tranquil locations, while clonality may by favored by calmer conditions (Jokela et al.,

2003).

Sexual and asexual reproduction in Pocillopora spp. and its relative importance

Pocillopora spp. are known for their variable reproductive strategies (Chávez-Romo,

Reyes-Bonilla, 2007; Glynn et al., 1991; Richmond, 1987a; Richmond, 1988; Richmond,

Jokiel, 1984; Richmond, 1987b; Stoddart, 1983; Stoddart, Black, 1985). These species are widespread through the Indian and Pacific Oceans and the Red Sea (Veron, 2000;

Veron, 2002) including locations considered marginal to coral development such as the

Eastern Pacific (Glynn, Ault, 2000). In the TEP, pocilloporids represent the more important reef building corals of the region and dominate reefs to a depth of 6 – 7 m

(Cortes, 1997; Glynn, Ault, 2000).

The genetic structure of Pocillopora has been assessed in various locations through its geographic distribution (Adjeroud, Tsuchiya, 1999; Ayre et al., 1997; Ayre, Willis,

1988; Benzie et al., 1995; Chavéz-Romo et al., 2009; Miller, Ayre, 2004; Ridgway et al.,

2001; Ridgway et al., 2008; Sherman et al., 2006; Souter et al., 2009; Stoddart, 1984a;

Stoddart, 1984b; Stoddart, 1984c; Whitaker, 2006; Yeoh, Dai, 2010). Comparisons between these studies and the present are feasible; however it is important to take into account differences in methodologies. For example, none of the previous studies followed a sampling strategy design to assess the clonal structure of these populations; most of them used allozymes (as opposed to microsatellites) and the geographical coverage of the studies is not uniform varying from few meters to 1000’s of km.

99 The relative importance of sexual vs. asexual reproduction in Pocillopora populations has been documented in South Africa (Ridgway et al., 2001), North (NWA) and South

(SWA) Western Australia (Stoddart, 1984b; Stoddart, 1984c; Whitaker, 2006), the Great

Barrier Reef (GBR) (Ayre et al., 1997; Benzie et al., 1995; Sherman et al., 2006), Lord

Howe Island (Miller, Ayre, 2004), Japan (Adjeroud, Tsuchiya, 1999), Taiwan (Yeoh,

Dai, 2010) and Hawaii (Stoddart, 1984a) (Fig. 22). The general consensus is that even though Pocillopora spp. shows both sexual and asexual mechanisms, the relative dominance of one over the other depends on local geographical and/or environmental conditions as well as on selective pressures on the survival of the recruits and the genetic connectivity of the populations. However, the life cycle of Pocillopora, and most corals, is complex and can involve, at least two sources of asexual recruits, colony fragments and asexual larvae (Baird et al., 2009). In both instances populations will look clonal, but the nature of the clonality is different. During fragmentation, the extension of the clone is independent of the coral, in other words, external factor are “regulating” the separation of fragments, while with the production of an asexual larvae, the coral has to designate resources in its production, therefore is “selecting” to be clonal.

In the GoC, a single population of Pocillopora type 1 in La Paz showed different levels of asexuality/sexuality in relatively close reefs (less than 10 km apart) (Figs. 19 and 22). There is a small scale variation in the contribution of asexual reproduction to population structure. Similarly, in NWA two reefs (Coral Bay and False Passage) located relatively close to each other showed populations with different reproductive structure

(Whitaker, 2006) and in SWA Pocillopora reef, a highly clonal system showed low levels or sexual recruitment (Stoddart, 1984b; Stoddart, 1984c). In One Tree Island, the levels

100 of sexuality change from the lagoon to the crest of the reef, but it was concluded that in both zones Pocillopora damicornis was reproducing sexually (Benzie et al., 1995).

Figure 22. Patterns of sexual (or clonal) structure of various population of Pocillopora spp through the Indo-Pacific Ocean. Colors inside the squares represent sexual populations (white) versus clonal populations (gray), cross bars indicate locations where differences between the populations were found (i.e. low gene flow between reefs) and the numbers on top the approximate range (km) of each study. *In two locations in NWA (Pilgramunna and Coral Bay) sexual reproduction was dominant while the other reefs showed clonal populations of Pocillopora. +At One tree Island reef, P. damicornis colonies release large amounts of asexually produced larvae, however the population structure is that of a sexual population (Ayre, Miller, 2004). (a. South Africa / Mozambique (Ridgway et al., 2001) b. Northwestern Australia - NWA (Whitaker, 2006)c. South Western Australia - SWA (Stoddart, 1984b; Stoddart, 1984c) d. Southern Taiwan (Yeoh, Dai, 2010) e. Ryukyu Archipelago, Japan (Adjeroud, Tsuchiya, 1999) f. Great Barrier Reef - GBR (Ayre et al., 1997) g. One Tree Island (Benzie et al., 1995) h. Lord Howe Island (Miller, Ayre, 2004) i. Kaneobe Bay (Stoddart, 1984a) j. La Paz Mexico (This study))

Our results indicated that both sexual and asexual reproduction are important for population maintenance of Pocillopora type 1 in La Paz. Pocillopora reef, in SWA, showed higher genotypic diversity in smaller (presumably recent recruits) than larger

(older adults) colonies presumably as a result of sexual recruitment (Stoddart, 1984b).

Variable sexual recruitment combined with selective pressure may result in the differences observed between the two reefs in La Paz. Asexuality, for example, depends on the genetics composition of the populations/species, the geographical location and

101 local environmental conditions (Eckert et al., 2008). Areas more susceptible for sexual recruitment over time become more diverse reefs (i.e. PG) than areas where conditions

(i.e. less exposure, higher sedimentation, less habitat availability) only allow the settlement and survival of a reduced number of genotypes (i.e. IG).

In locations where the dominant mode of reproduction was asexual (Hawaii, SWA and NWA), genotypic variability suggested sexual reproduction was also present and may play a role in maintaining the connectivity between populations (Stoddart, 1984a;

Stoddart, 1984b; Stoddart, 1984c; Whitaker, 2006). In NWA, two reefs showed sexual reproduction as the primary mode of reproduction (Whitaker, 2006). Mixed modes of reproduction were also evident in La Paz. PG is largely sexual, and even though a single clone dominates in IG, the other colonies sampled represent unique genotypes, product of sexual reproduction. Sexual reproduction or migration (evidenced by the null alleles) may maintain genetic connectivity between IG and PG.

Is the TEP a marginal location for Pocillopora?

Populations of Pocillopora spp. have shown variability in the relative dominance of reproductive strategies with a pattern that did not follow that of sexual populations in central locations and asexual populations in marginal areas of the distribution (Adjeroud,

Tsuchiya, 1999; Ayre, Hughes, 2000; Ayre et al., 1997; Benzie et al., 1995; Sherman et al., 2006; Stoddart, 1984a; Stoddart, 1984b; Stoddart, 1984c; Whitaker, 2006; Yeoh, Dai,

2010). In several cases evidence of the non-dominant mechanism has been reported (i.e. presence of sexually produced recruits and low levels of repeated MLG) suggesting that local environmental conditions may be favoring settlement of particular recruits (i.e.

102 sexual vs. asexual), therefore populations will appear sexual or clonal depending on the selected composition of the recruits (Ayre, Hughes, 2000; Ayre et al., 1997; Benzie et al.,

1995; Stoddart, 1984a; Stoddart, 1984b; Stoddart, 1984c; Whitaker, 2006).

In species with large distributional ranges, generalized adaptations are stabilized and maintained in central populations (Brown, 1984). As geographic isolation increases and gene flow decreases, genetic diversity within-populations is reduced while differentiation among-populations is increased (Eckert et al., 2008) making peripheral populations potentially locally adapted and genetically isolated (Brown, 1984). Under the centre- periphery hypothesis (CPH – also known as central marginal hypothesis), populations in the periphery are expected to have lower genetic variability with higher extinction probability when compared to those in the center of the species distribution range

(Alberto et al., 2008). In marginal populations, inbreeding and drift are responsible for the lost of genetic polymorphism (Kapralov, 2004) with asexual reproduction playing an important role in the persistence and dispersal of the species (Tatarenkov et al., 2005).

This phenomenon has been observed in corals (A. palmata, Baums et al. 2006) and several other organisms (Eckert et al., 2008; Tatarenkov et al., 2005) suggesting CPH is spread in nature.

In the past, marginal locations for scleractinian included those located at high latitudes (i.e. SWA) or relatively isolated regions (i.e. Hawaii) (Richmond, Hunter,

1990). The isolated TEP with low scleractinian diversity has been considered a marginal location for corals (Cortes, 1997; Glynn, 1976). Pocillopora, however seems to be able to to overcome the “adverse” conditions from this “suboptimal” location (Glynn, Ault,

2000). Similarities between our findings and other populations in the Indo-Pacific

103 suggests that Pocillopora populations in the GoC are far from uncommon, presenting both reproductive modes (i.e. sexual and asexual), contrary to peripheral populations where clonality tends to be favored (Vorburger, 2006).

Having both reproductive mechanisms within the same population can offer an advantage, on one hand, production of offspring without sexual reproduction results in genetically identical individuals (Arnaud-Haond et al., 2007). Clonality may increase the persistence and dispersal of the species (Tatarenkov et al., 2005), but loss of genetic diversity may render these populations more susceptible to drastic environmental changes

(Booth, Grime, 2003; Mingfeng et al., 2004; Schmid, 1985; Vorburger, 2006; Zhu et al.,

2000) and pathogens (Hamilton, Axelrod, 1990). On the other hand, genotypically diverse populations on the other hand, had an increased rate of recovery after a perturbation (Reusch, Ehlers, 2005). Assuming the resulting clonal structure of

Pocillopora is driven exclusively by the production of asexual larvae, balancing both reproductive mechanisms can benefit for the species as a whole, increasing the adaptability of Pocillopora type 1 to TEP environmental conditions.

We hypothesize that while the TEP may be marginal for most corals (i.e. Acroporids,

Favids and Poritids) it is not necessarily marginal for Pocillopora type 1. Local environmental conditions seemed to control how Pocillopora populations behave in terms of its reproductive strategy, thus resulting in the patterns observed for these species through the Indo-Pacific Ocean with no relation to marginal or central areas of distribution. In the TEP, Pocillopora thrive just fine occupying niches (i.e. major reef builders, different morphologies inhabitant different microhabitats within a reef) that in other areas (i.e. central/western pacific and Indian ocean) are occupied by other more

104 dominant (and “successful”) coral species. Testing this hypothesis will require the inclusion of more locations in the TEP (i.e. Mexican Pacific, the Southern TEP - Panama and Oceanic Islands – Clipperton) as well as the other genetically defined species found in the region (Type 2 and 3). It is predicted that all species at all locations will show evidence for both sexual and asexual mechanisms with populations from different reefs having different reproductive behaviors, patterns observed in many invertebrate taxa

(Vrijenhoek, 1998).

References

Adjeroud M, Tsuchiya M (1999) Genetic variation and clonal structure in the scleractinian coral Pocillopora damicornis in the Ryukyu Archipielago, southern Japan. Mar Biol 134, 753-760.

Alberto F, Massa S, Manent P, et al. (2008) Genetic differentiation and secondary contact zone in the seagrass Cymodocea nodosa across the Mediteranean-Atlantic transition region. J. Biogeogr. 35, 1279-1294.

Arnaud-Haond S, Alberto F, Teixeira S, et al. (2005) Assessing genetic diversity in clonal organisms: low diversity or low resolution? Combining power and cost efficiency in selecting markers. Journal of Heredity 96, 434.

Arnaud-Haond S, Belkhir K (2007) GENCLONE: a computer program to analyse genotypic data, test for clonality and describe spatial clonal organization. Mol Ecol Notes 7, 15-17.

Arnaud-Haond S, Duarte C, Alberto F, Serrao E (2007) Standardizing methods to address clonality in population studies. Mol Ecol 16, 5115-5139.

Ayre DJ, Dufty S (1994) Evidence for restricted gene flow in the viviparous coral Seriatopora hystrix on Australia's Great Barrier Reef. Evolution 48, 1183-1201.

Ayre DJ, Hughes TP (2000) Genotypic Diversity and Gene Flow in Brooding and Spawning Corals Along the Great Barrier Reef, Australia. Evolution 54, 1590- 1605.

105 Ayre DJ, Hughes TP, Standish RJ (1997) Genetic differentiation, reproductive mode, and gene flow in the brooding coral Pocillopora damicornis along the Great Barrier Reef, Australia. MEPS 159, 175-187.

Ayre DJ, Miller KJ (2004) Where do clonal coral larvae go? Adult genotypic diversity conflicts with reproductive effort in the brooding coral Pocillopora damicornis. MEPS 277, 95-105.

Ayre DJ, Willis BL (1988) Population structure in the coral Pavona cactus: clonal genotypes show little phenotypic plasticity. Marine Biolgy 99, 495-505.

Baird A, Guest J, Willis B (2009) Systematic and biogeographical patterns in the reproductive biology of scleractinian corals. Annual Review of Ecology, evolution and Systematics 40, 552-571.

Baums IB, Miller MW, Hellberg ME (2006) Geographic variation in clonal structure in a reef-building caribbean coral, Acropora palmata. Ecological Monographs 76, 503-519.

Benzie JAH, Haskell A, Lehman H (1995) Variation in the genetic composition of coral (Pocillopora damicornis and Acropora palifera) populations from different reef habitats. Mar Biol 121, 731-739.

Booth R, Grime J (2003) Effects of genetic impoverishment on plant community diversity. Journal of Ecology 91, 721-730.

Brown J (1984) On the relationship between abundance and distribution of species. American Naturalist 124, 255-279.

Browne R, Davis L, Sallee S (1988) Effects of temperature and relative fitness of sexual and asexual brine shrimp Artemia. Journal of Experimental Marine Biology and Ecology 124, 1-20.

Cortes J (1997) Biology and geology of eastern Pacific coral reefs. Coral Reefs 16, S39- S46.

Crow J (1992) An advantage of sexual reproduction in a rapidly changing environment. Journal of Heredity 83, 169-173.

Chavéz-Romo HE, Correa-Sandoval F, Paz Garcia D, et al. (2009) Genetic structure of the scleractinian coral, Pocillopora damicornis, from the Mexican Pacific. Proc. 11th Inter. Coral Reef Symp. Ft. Lauderdale 1, 429-433.

Chávez-Romo HE, Reyes-Bonilla H (2007) Reproducción sexual del coral Pocillopora damicornis al sur del Golfo de California, México. Ciencias Marinas 33, 495-501.

David P, Pujol B, Viard F, Castella V, Goudet J (2007) Reliable selfing rate estimates from imperfect population genetic data. Mol Ecol 16, 2474-2487.

106 Del Monte-Luna P, Arreguín-Sánchez F, Godínez-Orta L, López-Ferreira CA (2005) Batimetría actualizada de la bahía de La Paz, Baja California Sur, México. Oceánides 20, 75-77.

Earl DA (2009) Structure Harvester v0.3, from website: http://users.soe.ucsc.edu/~dearl/software/struct_harvest/.

Eckert C, Samis K, Lougheed S (2008) Genetic variation across species' geographical ranges: the central-marginal hypothesis and beyond. Molecular Ecology 17, 1170- 1188.

Foster N, Baums IB, Mumby P (2007) Sexual versus asexual reproduction in an ecosystem engineer: the massive coral Montastraea annularis. Journal of Animal Ecology 76, 384-391.

Glynn PW (1976) Some physical and biological determinant of coral community structure in the Eastern Pacific. Ecological Monographs 46, 431-456.

Glynn PW, Ault J (2000) A biogeographic analysis and review of the far eastern Pacific coral reef region. Coral Reefs 19, 1-23.

Glynn PW, Gassman N, Eakin C, et al. (1991) Reef coral reproduction in the eastern Pacific: Costa Rica, Panama, and Galapagos Islands (Ecuador). Mar. Biol., 355- 369.

Halkett F, Simon J, Balloux F (2005) Tackling the population genetics of clonal and partially clonal organisms. Trends in Ecology & Evolution 20, 194-201.

Hamilton W, Axelrod R (1990) Sexual reproduction as an adaptation to resist parasites (a review). Proceedings of the National Academy of Sciences of the United States of America 87, 3566-3573.

Harrison PL, Wallace CC (1990) Reproduction, dispersal and recruitment of scleractinian corals. In: Coral Reefs, Ecosystems of the World 25, pp. 133-207.

Hubby JL, Lewontin RC (1966) A molecular approach to the study of genic heterozygosity in natural populations. I. The number of alleles at different loci in Drosophila pseudoobscura. Genetics 54, 577-594.

Jackson JBC (1986) Modes of dispersal of clonal benthic invertebrates: consequences for species' distributions and genetic structure of local populations. Bulletin of Marine Science 39, 588-606.

107 Jokela J, Lively C, Dybdahl M, Fox J (2003) Genetic variation in sexual and clonal lineages of a freshwater snail. Biological Journal of the Linnean Society 79, 165- 181.

Kapralov M (2004) Genotypic variation in populations of the clonal plant Saxifraga cernua in the Central and Peripheral Regions of the species range. Russian Journal of ecology 35, 413-416.

LaJeunesse TC, Loh W, van Woesik R, et al. (2003) Low symbiont diversity in southern Great Barrier Reef Corals, relative to those of the Caribbean. Limnol. Oceanog. 48, 2046-2054.

Miller KJ, Ayre DJ (2004) The role of sexual and asexual reproduction in structuring high latitude populations of the reef coral Pocillopora damicornis. Heredity 92, 557-568.

Miller M (1997) Tools for population genetic analyses (TFPGA) 1.3.

Mingfeng H, Changliank Y, Hongbo R (2004) Evolution of population with sexual and asexual reproduction in changing environment. International Journal of Modern Physics C 15, 289-299.

Nei M (1978) Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89, 583-590.

Peakall R, Smouse PE (2006) Genalex 6: genetic analysis in Excel. Population genetic software for teaching and research. Mol Ecol Notes 6, 288-295.

Pinzón JH, Lajeunesse TC (2010) Species delimitation of reef building corals using nucleotide sequence phylogenies, population genetics, and symbiosis ecology. Molecular Ecology in review, 1-39.

Pritchard JK, Stephens M, Donnelly P (2000) Inferece of population structure using multilocus genotype data. Genetics 155, 945-959.

Raymond M, Rousset F (1995) GENEPOP (version1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity 86, 248-249.

Reusch T, Ehlers A (2005) Ecosystem recovery after climatic extremes enhanced by genotypic diversity. Proceedings of the National Academy of Sciences of the United States of America 102, 2826-2831.

Richmond R (1987a) Energetics, competency, and long-distance dispersal of planula larvae of the coral Pocillopora damicornis. Mar Biol 93, 527-533.

Richmond R (1988) Competency and dispersal potential of planula larvea of a spawning versus a brooding coral 2, 827-831.

108 Richmond R, Hunter C (1990) Reproduction and recruitment of corals: Comparisons among the Caribbean, the Tropical Pacific, and the red sea. MEPS 60, 185-203.

Richmond R, Jokiel PL (1984) Lunar periodicity in larva release in the reef coral Pocillopora damicornis at Enewetak and Hawaii. Bull Mar Sci 34, 280-287.

Richmond RH (1987b) Energetic relationships and biogeographical differences among fecundity, growth and reproduction in the reef coral Pocillopora damicornis. Bull Mar Sci 41, 594-604.

Ridgway T, Hoegh-Guldberg O, Ayre DJ (2001) Panmixia in Pocillopora verrucosa from South Africa. Mar Biol 139, 175.

Ridgway T, Riginos C, Davis J, Hoegh-Guldberg O (2008) Genetic connectivity patterns of Pocillopora verrucosa in southern African Marine Protected Areas. MEPS 354, 161-168.

Rosenberg N (2004) DISTRUCT: a program for the graphical display of population structure. Molecular Ecology Notes 4, 137-138.

Rousset F (2008) Genepop’007: a complete re-implementation of the genepop software for Windows and Linux. Molecular Ecology Resources 8, 103-106.

Scheu S, Drossel B (2007) Sexual reproduction prevails in a world of structured resources in short supply. Proceedings of the Royal Society B 274, 1225.

Schmid B (1985) Clonal growth in grassland perennials: III. Genetic variation and plasticity between and within populations of Bells Perennis and Prunella vulgaris. Journal of Ecology 73, 819-830.

Seutin G, White BN, Boag PT (1991) Preservation of avian blood and tissue samples for DNA analyses. Canadian Journal of Zoology 69, 82-90.

Sherman CDH, Ayre DJ, Miller KJ (2006) Asexual reproduction does not produce clonal populations of the brooding coral Pocillopora damicornis on the Great Barrier Reef, Australia. Coral Reefs 25, 7-18.

Smouse P, Peakall R (1999) Spatial autocorrelation analysis of individual multiallele and multilocus genetic structure. Heredity 82, 561-573.

Souter P, Henriksson O, Olsson N, Grahn M (2009) Patterns of genetic structuring in the coral Pocillopora damicornis on reefs in East Africa. BMC Ecol 9, 19.

Stoddart JA (1983) Asexual production of planulae in the coral Pocillopora damicornis Mar Biol 76, 279-284.

Stoddart JA (1984a) Biochemical genetics of Pocillopora damicornis in Kaneobe Bay, Oahu, Hawaii. Hawaii Inst Mar Biol, Tech Report 37, 133-150.

109 Stoddart JA (1984b) Genetic differentiation amongst populations of the coral Pocillopora damicornis off Southwestern Australia. Coral Reefs 3, 149-156.

Stoddart JA (1984c) Genetical structure within populations of the coral Pocillopora damicornis. Marine Biology 81, 19-30.

Stoddart JA, Black R (1985) Cycles of gametogenesis and planulation in the coral Pocillopora damicornis. MEPS 23, 153-164.

Tatarenkov A, Bergstrom L, Jonsson R, et al. (2005) Intriguing asexual life in marginal populations of the brown seaweed Fucus vesiculosus. Molecular ecology 14, 647- 651. van Oosterhout C, Hutchinson WF, Wills DPM, Shipley P (2004) Micro-checker: software for identifying and correcting genotyping errors in microsatellite data. Mol Ecol Notes 4, 535-538.

Vandel A (1928) La partheénogense geographique. Contribution a l’etude biologique et cytologique de la parthenogenese naturelle. Bulletin Biologique de la France et de la Belgique 62.

Veron JEN (2000) Corals of the world Australian Institute of Marine Science, Townsville.

Veron JEN (2002) New species described in Corals of the world Australian Institute of Marine Science, Townsville.

Vorburger C (2006) Geographic parthenogenesis: Recurrent patterns down under. Curr Biol 16, R641-R643.

Vrijenhoek R (1979) Factors affecting clonal diversity and coexistence. American Zoologist 19, 787-797.

Vrijenhoek R (1998) Animal clones and diversity. Bioscience.

Whitaker K (2006) Genetic evidence for mixed modes of reproduction in the coral Pocillopora damicornis and its effect on population structure. MEPS 306, 115- 124.

Yeoh S-R, Dai C-F (2010) The production of sexual and asexual larvae within single broods of the scleractinan coral, Pocillopora damicornis. Marine Biology 157, 351-359.

Zhu Y, Chen H, Fan J, et al. (2000) Genetic diversity and disease control in rice. Nature 406, 718-722.

110

CHAPTER 4: HOST SPECIALIZATION MORE THAN GEOGRAPHIC ISOLATION DRIVES THE SPECIATION OF CORAL ENDOSYMBIONTS IN THE GENUS SYMBIODINIUM (ZOOXANTHELLAE)

Abstract

Reef building and reef dwelling cnidarians obligately associate with a limited subset of available symbiotic dinoflagellates in the genus Symbiodinium. This specificity in turn appears to have directly influenced the speciation of these symbionts based on ecological and phylogenetic data. In order to examine the relative significance of niche partitioning relative to geographic isolation, we analyzed gene flow among populations of a putative host generalist symbiont, Symbiodinium fitti (formerly type A3 sensu LaJeunesse 2001), associated with various Caribbean corals, including Acropora cervicornis, A. palmata, Stephanocoenia intersepta, and Montastraea annularis in the Western Caribbean using ten microsatellite markers. Ninety-two percent of samples (144 out of 153 samples) contained a single multilocus genotype. From these, 94 distinct multi-locus genotypes were characterized. Significant population structure existed and there was little or no indication of gene flow between S. fitti populations originating from Stephanocoenia intersepta and those found in colonies of Acropora spp. and Montastraea annularis existing sympatically on the same reef. To a lesser extent, geographic separation appeared to further influence population structuring. This study is the first to use population genetics to examine how ecological specialization to a particular host habitat effectively drives population subdivision ultimately leading to species diversification from a once generalist symbiont.

111

Introduction

During speciation, restrictions of gene flow may arise because of geographic isolation

(allopatric, or parapatric, speciation) or when populations with overlapping geographic distributions begin to occupy different ecological niches (sympatric speciation) (Coyne,

Orr, 1998). Speciation is most often attributed to geographic isolation (Bolnick,

Fitzpatrick, 2007; Mallet et al., 2009; Pimentel et al., 1967), however, under certain circumstances sympatric speciation may play a critical role in generating diversity (Via,

2001). There remains some controversy surrounding this mode of lineage separation and divergence (Fitzpatrick et al., 2009). Speciation under sympatry is usually difficult to demonstrate and details about the processes creating pre- and postzygotic isolation are vague. One viable solution requires a population genetic approach to look for population subdivision and incipient reproductive isolation (Orr, Smith, 1998; Via, 2009).

Early stages of sympatric speciation involve formation of “races,” or

“subpopulations,” that exhibit reproductive isolation (Pimentel et al., 1967). Diversifying selection for opposing character traits can induce population differentiation by directly, or indirectly, creating barriers to genetic recombination (Feder et al., 1994; Hawthorne, Via,

2001; Via, Hawthorne, 2002). Mate choice and physiological attributes correspond with genetic differences among individuals of a species occupying different habitats, indicating that adaptive processes operate during ecological specialization (Hawthorne,

Via, 2001; Via, Hawthorne, 2002).

Coral reefs are extraordinary among the planet’s ecosystems because most of the animals that form the basis of the ecosystem (i.e. corals) are symbiotic with

112

photosynthetic dinoflagellates in the genus Symbiodinium. These mutualistic associations, and the reef systems they build, are in decline as a result of local and global environmental stressors (Gardner et al., 2003; Pandolfi et al., 2003; Selig, Bruno, 2010).

The importance of the symbiont to the physiology and heath of the animal host has facilitated considerable research on their diversity, ecology, and evolution (Rowan,

Powers, 1991; LaJeunesse, 2001; (Baker, 2003; Goulet, Coffroth, 2003; LaJeunesse,

2001; Pawlowski et al., 2001; Pochon et al., 2004; Rowan, Powers, 1991; Santos et al.,

2003; Santos et al., 2004; van Oppen et al., 2001); (LaJeunesse, 2005; LaJeunesse et al.,

2010; Thornhill et al., 2006a; Thornhill et al., 2006b).

It is now well established that the genus Symbiodinium comprises numerous evolutionarily divergent lineages, or clades, with some of these containing tens and potentially 100’s of diverging sub-lineages distinguished ecologically by differences in host species association, depth, and geographic range; and when investigated, also exhibit differences in physiology (Berkelmans, van Oppen, 2006; Hannes et al., 2009;

Rodriguez-Lanetty et al., 2004; Rodriguez-Lanetty et al., 2006; Sampayo et al., 2008).

Some Symbiodinium are characterized by their ability to associate with many host taxa

(ecological generalists) while most others exhibit comparatively high specificity for hosts of a particular genus or species (i.e. ecological specialists). The relative abundance of host specific lineages suggests that ecological specialization to a particular host habitat drives much of the speciation in the genus Symbiodinium (Finney et al., 2010;

LaJeunesse et al., 2010).

One of the most ecologically common and geographically widespread Symbiodinium in clade A is type A3 (sensu LaJeunesse 2001; LaJeunesse et al. 2009) where all members

113

of this type have an identical intragenomically dominant internal transcribed spacer (ITS) sequence. Following the development of microsatellite markers, this type was recently designated Symbiodinium fitti (Pinzón et al., 2010). In the Caribbean, S. fitti associates with several coral species including the threatened Acropora palmata and A. cervicornis

(Correa et al., 2009; Finney et al., 2010) and Stephanocoenia sp. (Finney et al., 2010;

LaJeunesse, 2002). The larvae of these corals acquire Symbiodinium fitti cells from the environment (i.e. open or horizontal transmission), suggesting that they associate with genetically similar populations of this putative host-generalist symbiont. Alternatively, if host specialization is important in the diversification of populations in sympatry, gene flow among populations of S. fitti may be restricted to populations associated with a particular host species. To test these possibilities, a population genetics approach was used to examine early stages of speciation not yet apparent in sequence divergence of genes traditionally used to delimit species (Via, 2009). Ten microsatellite markers specific to Symbiodinium fitti (Pinzón et al., 2010) were utilized to genetically characterize populations of this symbiont from various host taxa in sympatry over a latidudinal gradient in the western Caribbean in order to determine the relative influence of host/habitat specialization and/or geographic isolation in structuring the populations of

S. fitti.

114

Materials and Methods

Sites and sample collection

Sampling was performed with SCUBA diving in three locations in the Caribbean; Puerto

Morelos in Mexico, Bocas del Toro in Panama and Cartagena (i.e. I. del Rosario and

Barú) and Santa Marta in Colombia (Fig. 23). Samples consisted of small (less than 1 cm3) colony pieces containing both skeletons and tissue. In the branching corals (A. palmata and A. cerviconis), fragments were excised with bone scissors, while hammer and chisel were used with the massive species (St. intersepta and M. annularis).

Fragments were fixed in DMSO (20%) salt (NaCl) supersaturated (Seutin et al., 1991) and store at -20˚C until DNA extraction. A. palmata (n = 50) and St. intersepta (n = 54) were collected in all locations, while A. cervicornis (n = 43) only in Panama and

Colombia and M. annularis (n = 6) in Panama.

DNA extractions and genetic analyses

DNA was isolated from fixed samples using a Wizard Genomic Purification Kit

(Promega, Madison USA) following the manufacturer’s protocol. A small piece of the sample (tissue and skeleton) was mixed with the nuclei lysis solution, included in the kit, and pulverized with glass beads (Ceroglass, Columbia USA). Then the fragment was incubated with proteinase K (20 mg / ml) at 65˚C for 1 hour followed by a centrifugation to eliminate skeleton pieces and glass beads. The DNA was precipitated from the supernatant with isopropanol/ethanol. The excess of alcohol was air dried and the

115

resulting DNA diluted in 75 µl of distilled water to a concentration between 5 to 80 ng /

µl and stored at -20˚C (LaJeunesse et al., 2003).

Figure 23. Figure 1. Map of the Caribbean sea showing the location of the sampling sites in Mexico (Puerto Morelos), Panama (Bocas del Toro) and Colombia (Cartagena and Santa Marta)

PCR Denature Gradient Gel Electrophoresis (PCR-DGGE) of the full ITS2 region was used to screen the samples for the presence of Symbiodinium fitti. In this method the

ITS2 region is amplified using Symbiodinium specific primers (ITSintfor2 and

ITS2clamp), after successful amplification, the products were run on a 8% acrylamide gel containing a 45 – 80% denature gradient, overnight at 110 volts (LaJeunesse, Trench,

2000). Bands were stained with SYBER-green (Invitrogen, Carlsbad CA, USA) to facilitate comparisons with a known ladder containing ITS2 bands from Symbiodinium fitti (S. A3).

116

Multilocus genotypes (MLG) were obtained using ten previously published microsatellite loci specific to Symbiodinium fitti (Pinzón et al., 2010). Single loci amplifications were performed under the optimized conditions for each of the primer sets.

The PCR profile included an initial denaturing step (94˚C 2 min), followed by 31 cycles of denature, annealing and extension (94˚C 15 sec – Ta 15 sec – 72˚C 30 sec) and a final extension step (72˚C – 5 / 30 min). The appropriate annealing temperature (Ta) was set for each primer according Pinzón et al. (2010). Each reaction contained a dye-labeled primer for each locus. Pairs of PCR products with different dyes were mixed in a single tube in a 1:1 ratio and electrophoresed on an ABI Hitachi 3730XL sequencer (Applied

Biosystems), with LIZ - 500 as size standard. Alleles were scored using the software

Genemarker 1.6 (Softgenetics State College PA, USA).

In order to test for mixtures of clonal lines harbored in the same samples, DNA from two samples (one from A. palmata and one from St. intersepta) with different single identifiable alleles on each locus were diluted to the same concentration (15 ng / µl) and mixed in different proportions (i.e. 1:1 and 1:9). Amplifications and analyses of all loci, from these mixtures and diluted samples, were performed with the same protocol used for all field samples.

Data analysis

Since the presence of two alleles in one locus may indicate mixed symbiont populations in a single coral colony, MLG were loci with more than one peak where remove from further analyses. The power to distinguish identical MLG of the combination of the ten microsatellite loci was assessed on the whole data set (n = 141 – after removing samples

117

with double peaks on a single locus) by estimating the probability of identity (PI) as implemented in GenAlex V 6.3 (Peakall, Smouse, 2006), and the probability of identical

MLG resulting from different sexual events (Psex) using the round-robin method (Parks,

Werth, 1993) implemented in GenClone V 1.0 (Arnaud-Haond, Belkhir, 2007).

GenClone V 1.0 was also used to detect identical MLG, somatic mutations and possible scoring errors. Revisions of suspicious genotypes were done and corrections were made, when necessary. Identification of unique genotypes (G) allowed the estimation of the genotypic richness (R = G - 1 / n – 1) for each population (i.e. Symbiodinium fitti harbored on different host at each location). Finally identical MLG genotypes were removed from the data set from further analyses.

Previous to any assignment analyses, linkage equilibrium was tested with Arlequin

(Excoffier et al., 2005), each population defined as individuals from the same host and same location. Genotypic assignments of individuals were performed using a bayesian approach in Structure V 2.3.3 (Pritchard et al., 2000) with 5 iteraction for each K (K = 1 to 10), 100,000 burn in, 1,000,000 replications under the assumption of admixture between groups and using population identification as location prior (Hubisz et al.,

2009). The optimal number of clusters was determined using the ∆K method (Evanno et al., 2005) executed in Structure Harvester (Earl, 2009). The ∆K method analyzes the change of the log of the probability in sequential values of K and determines the optimal number of populations in the sample based on the rate of the log change between consecutive K’s (Evanno et al., 2005). Finally, plots were constructed with Distruct V

1.1 (Rosenberg, 2004) after compiling the individual and population membership probabilities with Clumpp V 1.1.2 (Jakobsson, Rosenberg, 2007), using the intermediate

118

search algorithm (Greedy) to find the optimal alignment between replicates. To assess the influence of host and location in population differentiation in S. fitti, the data was analyzed in three ways: (1) by host, for those populations collected in all locations (A. palmata and S. intersepta), (2) by location including individuals associated with different hosts and (3) with all the data set.

Genetic differentiation between populations (i.e. clusters defined with Structure) was assessed by AMOVA’s and Φst calculations as implemented in GenAlex V6.3 (Peakall,

Smouse, 2006). Population structure was additionally visualized through principal coordinate analyses (PCoA), Networks and dendrograms using an unweighted pair group method with arithmetic mean (UPGMA). PCoA’s were constructed in GenAlex 6.3 from a binary genetic distance matrix, where a zero is assigned to any two samples (i.e. individuals) sharing the same allele and one to samples not sharing a particular allele.

Additions of these values across loci give the total distance between each pair of samples

(Peakall, Smouse, 2006). Networks were built with the software Network 4.516 (Fluxus) using the median-joining algorithm for all genotypes across hosts and geographic locations and for genotypes found among Stephanocoenia and A. palmata separately

(Bandelt et al., 1999). UPGMA dendrograms were built on the program Neighbor

(Felsenstein, 1995) using a Chord-distance (Cavalli-Sforza, Edwards, 1967) matrix created in MICROSAT (Human Population Genetics Laboratory, Stanford University).

This combination (Chord Distance and UPGMA) has been suggested to give the best tree topologies for microsatellite datasets (Takezaki, Nei, 1996).

119

Results

Coral colonies harbor a single clonal line of S. fitti

A total of 153 samples of S. fitti from different hosts and locations were genotyped with ten microsatellite loci. As haploid organisms (Blank, 1987; Pfiester, Anderson, 1987), amplifications of microsatellite loci in any Symbiodinium spp. are expected to generate chromatograms with single peaks for each locus (i.e. one allele per locus). The presence of two alleles in one locus may indicate mixed symbiont populations in a single coral colony. We tested the minimum percentage of the total concentration of DNA at which single MLG are found by experimentally mixing DNA from different Symbiodinium fittii genotypes. The mixture experiments indicated different S. fitti genotypes should be distinguished from each other, even at low concentrations. For all loci tested, mixtures with 9:1 ratios resulted in the identification and scoring of two peaks (i.e. two genotypes).

Therefore, MLG’s contributing with at least 10% of the total DNA concentration should be identifiable (Fig. 24). The lack of a second peak indicates each colony harbors a single genotype (i.e. clonal line).

120

Figure 24. Chromatograms of two loci (A3Sym_027 and A3Sym_032) amplifications using various ratios (1:0, 1:9 and 1:1) of DNA mixtures obtained from samples of Symbiodinium fitti isolated from two different hosts (Genotype A from Acropora palmata and Genotype B from Stephanocoenia intersepta). The emergence of a second peak in the 1:9 reactions indicates that genotypes representing more than 10% of a mixture should be detectable with these primer sets. The same results were observed with the other primer sets. The 114bp peak in locus A3Sym_27 corresponds to a randomly amplified product whose sequence did not matched that of the originally isolated locus and did not have a repeat region.

In our dataset, the presence of double alleles (populations contributing to more than

10% of the total DNA concentration) was observed in five loci (A3sym_018;

A3Sym_028; A3Sym_031; A3Sym_32; A3sym_041). A total of ten samples (6.54 %; n

= 153; 3 from A. palmata and 4 from St. intersepta in Panama and 3 from St. intersepta in

Mexico) presented double peaks in a single locus, while two colonies of St. intersepta

(1.31 %; n = 53; one from Panama and one from Mexico) had two peaks in two different loci. These observations indicate the majority (92.16 %) of the sampled coral colonies harbor a single clonal line of Symbiodinium fitti.

Genetic diversity among S. fitti populations

Samples with double peaks were removed from the data set; therefore genotypic diversity was assessed from 141 colonies (A. cervicornis n = 43, A palmata n = 47, M. annularis n

121

= 6 and St. intersepta n = 45). All populations (i.e. each host on a particular location) presented repeated multilocus genotypes (MLG), but clones were not found across geographic locations or hosts, except for one colony of A. palmata harboring the same clone of S. fitti found in three samples of A. cervicornis in Colombia. Genotypic diversity across populations was variable with an average richness (R) of 0.68. Two populations (A. palmata - Panama R = 0.30 and A. cervicornis - Colombia R = 0.36) were found to be largely clonal, the other eight were more diverse, with values of R ranging from 0.50 in A. palmata – Colombia (Cartagena) to R = 1.0 in M. annularis – Panama

(Table 7). In total, 66.66 % (n = 94) of all genotyped samples have distinctive unique

MLG.

Table 7 Total number of colonies genotyped (n), number of unique multilocus genotypes (MLG - G) and richness (R) of S. fitti populations (defined by host and location) from four coral hosts and three locations thought the Caribbean. Only two populations (*) showed high levels of clonality (i.e. large numbers of individuals with the same MLG). Host Location n G R

Panama 20 18 0.89 A. cervicornis Colombia 23 9 0.36*

Mexico 9 7 0.75

Panama 24 8 0.30* A. palmata Colombia (Ctg) 5 3 0.50

Colombia (SM) 9 7 0.75

Mexico 23 20 0.86

S. intersepta Panama 10 7 0.67

Colombia 12 9 0.73

122

M. annularis Panama 6 6 1.00

Linkage between loci was found to be significant (after Bonferroni corrections) in 42

(9.9%) of the 450 pairwise tests. Most of the linked loci are found within St. intersepta harbored populations, which showed subdivisions (Fig. 25) that may correlate with environmental and/or host genetic differences (see below). Independently of this subdivisions, the relatively low number of significant pairwise tests indicated the assumption of linkage equilibrium (required by Structure) for these populations was fulfilled (Kirk et al., 2009).

Figure 25 Principal coordinate analysis (PCoA) of Symbiodinium fitti associated with the scleractinian coral Stephanocoenia spp in three locations in the Caribbean. The subdivisions of the populations from Mexico and Colombia may correspond to different microenvironments and/or host species respectively (see text for an explanation).

123

Figure 26. Genetic clustering analyses of Symbiodinium fitti individual multilocus genotypes showed correspondence with geographic location of the populations (Colombia, Panama, Mexico) associated with the same host species (a. Acropora palmata Φpt = 0.52, p = 0.01, n = 25 and b. Stephanocoenia intersepta Φpt = 0.35, p = 0.01, n = 36)

Geographic isolation of S. fitti populations across the Western Caribbean

In order to assess the influence of environmental conditions on S. fitti population differentiation, assignment analyses were performed in populations associated with the same host, but from different locations. Bayesian and distance analyses in both datasets delineated three (K = 3) populations of S. fitti. Each cluster matching the geographic location where samples were collected (Fig. 26), with the exception of the MLG found in association with A. palmata from Cartagena, which seemed a mixture to those found in

124

Panama and Mexico. Φpt values showed significant (p = 0.01) population differentiation between the defined clusters for both hosts’ populations (A. palmata associated populations Φpt = 0.52 and St. intersepta associated populations Φpt = 0.35; Table 8).

Table 8. Levels of differentiation between populations of S. fitti delimited by location, host and through the Caribbean (G = number of unique multilocus genotypes – MLG, K = optimal number of clusters/populations as suggested by Structure and Φst = Phi Statistics – calculated with GenAlex, * = all Φst calculations were significant at the p = 0.01 level, Ctg = Cartagena) Give variance for phi st values

G K Φst*

Location

Mexico 27 4 0.282

Panama 39 2 0.500

Colombia (Ctg) 21 2 0.367

Host

A. palmata 25 3 0.529

St. intersepta 36 3 0.352

All 94 2 0.359

Low gene flow between S. fitti populations harbored by different coral hosts

Genotypic assignments of MLG in each location revealed that populations of S. fitti had significant levels of structure (p = 0.01; Table 8), associated with host identity. Two populations were differentiated in Panama (Φpt = 0.50) and Cartagena (Colombia; Φpt =

0.36), corresponding to populations of S. fitti associated with Acropora spp (and M.

125

annularis in Panama) and those associated with St. intersepta (Fig. 27). In Mexico the optimal number of S. fitti populations was four (Φpt = 0.57), one cluster associated with

A. palmata and the other three with St. intersepta (Fig. 27). The later subdivision among

St. intersepta associated populations seems to have a relation with local environmental conditions, with deeper (> 7m) colonies tending to associate with one of the groups and the shallow (3 to 7 m) ones with the other, unfortunately, samples sizes prevented any statistical analysis. In Colombia two mitochondrial CO1 types of Stephanocoenia have been found (Pinzón and LaJeunnese unpublished data) in association with different S. fitti populations.

Figure 27. Structure plots showing distinctions between populations of S. fitti harbored by sympatric coral host in three locations in the Caribbean (a. Mexico K = 4, Φpt = 0.57, p = 0.01, n = 27, Panama K = 2, Φpt = 0.50, p = 0.01, n = 39 and Cartagena, Colombia K = 2, Φpt = 0.36, p = 0.01; n = 28)

126

Figure 28. Bayesian population structure analysis of the populations of Symbiodinium fitti harbored by different hosts (a. Acropora palmata, b. A. cervicornis, c. Montastraea annularis, and d. Stephanocoenia spp) and geographic locations (Ctg = Cartagena). The influence of the host (at least genera, Acropora/Montastraea vs. Stephanocoenia) in population differentiation appears stronger than that of geographic isolation, as suggested by the optimal number of populations in the data set (K = 2, Φpt = 0.36, p = 0.01, n = 94).

127

Geography/environment and host increase S. fitti population differentiation

Genetic differentiation analyses of the dataset with unique MLG’s (n = 94) suggested the optimal number of clusters (K) to be two (Φpt = 0.36, p = 0.01). The division in two populations corresponded to different host, one cluster grouped all individuals associated with Acropora spp and the other the individuals associated with Stephanocoenia spp.

Individuals harbored by M. annularis appeared as a mixture between both clusters. At higher K’s this pattern in the population structure is maintained, but as K increases, further subdivisions are clearly observed within each cluster (Fig. 28). For example at K

= 3, the cluster of organisms associated with St. intersepta is further divided, grouping individuals from the same locations (Colombian, Panama and Mexico), in a patter that may results from geographical/environmental what?. In K = 5 to 7 individual genotypes of S. fitti associated with Acropora spp. are clustererd in groups corresponding to geographic distributions.

This pattern of host and geography driven population subdivision is also observed with network analyses and the distance-based method phylogenetics (Fig. 29 Appendix

D). The Network with all genotypes (n = 94) clearly separates the populations of S. fitti by host and within host by locations, while in the UPGMA tree two clusters are formed, one encompassing populations associated with Acropora spp and the other populations harbored by both St. intersepta and M. annularis. In both analyses, a geographical component of population differentiation is observed.

128

Figure 29. Networks showing the associations between populations of S. fitti across host and locations (a) and for Stephanocoenia spp. (b) and A. palmata (c) associated populations. Analyses included exclusively unique multilocus genotypes (G).

Discussion

The importance of coral-algal symbioses and recent deterioration, due to climate change and pollution, of coral reefs has prompted the study of the ecology and evolution of

Symbiodinium species (Barneah, Brickner, 2010). In the present assessment we focus on the fine scale differences among closely related groups (i.e. populations) to determine its

129

relation with host/habitat selection and or geographical distribution. Analyses of population markers showed a correlation between host identity and symbiont (i.e. S. fitti) population differentiation that agrees with a pattern of ecological specialization hypothesized as an initial stage of speciation.

Species radiations in Symbiodinium

Species radiations, either adaptative (Sanderson, 1998), ancient (Whitfield, Lockhart,

2007) or recent and rapid (Hughes, Eastwood, 2006; Klak et al., 2004; Richardson et al.,

2001a; Richardson et al., 2001b) are responsible for most of biological diversity (Linder,

2008; Weisrock et al., 2006). Radiations cause the emergence of several related species in a short period of time (Schluter, 2001; Schluter, Conte, 2009). Adaptative radiations,

(e.g. diversification into new habitats) have been proposed as the speciation (and diversification) mechanism in various Symbiodinium clades (Finney et al., 2010;

LaJeunesse, 2005; LaJeunesse et al., 2009), with host specialization and geographic isolation as the main processes in the initial stages (i.e. populations differentiation) of speciation (LaJeunesse, 2005).

In this research we explored the effects of host/habitat selection and/or geographic location on population isolation in the relatively generalist symbiont S. fitti. Population differentiation is a key factor in the process of speciation (Bush, 1994; Pimentel et al.,

1967) and can be linked to adaptative radiation. In other Symbiodinium clades, such as B, sequence data from two microsatellite loci have shown different populations of S.

B1/B184 in the same reefs are associated with different octocoral species

130

(Pseudopterogorgia elisabethae and P. bipinnata) (Santos et al., 2003). Similar findings, also in S. B1, showed patterns of host specificity across various Cnidarian (Hydrozoa,

Scleractinia, and Gorgonacea) hosts (Finney et al., 2010). Additionally to host-associated divergence, Symbiodinium B1 showed population structure associated with geographical patterns of distribution, either in the same general locations (i.e. Florida Keys) (Kirk et al., 2009) or across the Caribbean (Finney et al., 2010). Our data suggest that both host and geography/environment have an impact on population divergence in S. fitti providing additional evidence that ecological specialization to host/habitat of Symbiodinium spp. results in the diversification of the group.

Coral-Symbiodinium co-evolution

The influence of host selection and geographic distribution on the diversification of

Symbiodinium species may be the result of coevolved interactions between these two organisms. In mutualistic relationships, evolutionary changes in the host can exert changes in the symbiont and vice versa, making co-evolution an important force in the specialization and speciation of symbiotic partners (Janzen, 1980; Schluter, Conte, 2009).

Studying the effects of host/habitat selection and geographic isolation on population differentiation of mutualistic organisms is crucial to understand their response to environmental and/or selective pressures. These studies can also elucidate the initial stages of ecological speciation in both allopatry and sympatry (Hoeksema, 2010).

Diversification and evolution of symbiotic organisms in Rhizobium-Legume (Martínez-

Romero, 2009), insect-endobacteria (Clark et al., 2001) and termite-gut microorganisms

(Hongoh et al., 2005; Ohkuma et al., 2001) have been attributed to interactions between

131

hosts and their symbionts. In the S. fitti-coral system from the Caribbean, algal populations are differentiated in correlation to the host species (or at least genera –

Acropora vs. Stephanocoenia) suggesting an interaction between host-symbiont maybe responsible or at least influencing the diversification of Symbiodinium spp.

Historical changes in climate and geological conditions (i.e. glaciations events) have had a profound impact on the distribution of populations and species (Cwynar,

MacDonald, 1987; Muss et al., 2001). As new habitats are created, populations/species tend to expand their distributions. In the case of mutualisms expansions are accompanied by diversification of both partners (Léotard et al., 2009). As the new habitats are colonized and distance and barriers between populations increase, a reduction in gene flow between populations is more likely (McRae et al., 2005). Differentiation between populations of S. fitti associated with the same host, but from different geographical locations (Fig. 26) suggests geography influences the diversification of Symbiodinium species.

Although geographic isolated is observable, the host as a habitat seems to be a stronger barrier to gene flow between S. fitti populations, with particular symbiont populations associated exclusively with distinct host genera/species (Fig. 28). Our data, and previous research on species divergence in the genus (Finney et al., 2010;

LaJeunesse et al., 2010), suggest speciation in Symbiodinium results after host specialization (Fig. 30B) followed by diversification due to geographic isolation (Fig.

30D). The alternative idea, not seen in our data, is speciation happening before the onset of the association (Fig. 30C).

132

Details of coral-alga symbiosis formation are still elusive, but mechanisms of specialization (e.g. elimination of less cooperative and chemical/physical mechanisms) may control the formation of particular coral-Symbiodinium associations (Kawakita et al.,

2010). Dissimilar genetic groups (e.g. clades/types) are more productive (i.e. exploit resources) (Agashe, Bolnick, 2010) and may show lower levels of competition between them, than within genetically similar groups (e.g. genotypes).

133

Figure 30. Theoretical models showing different probable scenarios (a to d) of the evolution of Symbiodinium species, as influenced by the host and/or geographic isolation.

134

Mutualistic relationships have been found to be stable even in cases where gamete dispersion exists (i.e. broadcast spawning) and when the possibility to form associations with multiple genotypes (i.e. horizontal/open transmission systems) of either partner is high (Aanen et al., 2007). In the S. fitti system, horizontal transmission of the symbiont in various sympatric coral hosts is very common, however the relation between individual coral colonies and symbiont clonal lines remains highly specific. This level of specificity in mutualisms is maintained through either partner-fidelity feedback, cooperator associations and/or partner choice (Foster, Kokko, 2006; Foster, Wenseleers, 2006), with partner-fidelity and partner choice having a larger impact. Specifically in S. fitti, and in general in Symbiodinium-cnidarian associations, the presence in low levels of additional genotypes in the same colony, may increase the symbionts genetic variability and could be an indication of cooperator associations (Foster, Kokko, 2006).

Ecological speciation in S. fitti

Ecological speciation occurs in contrasting environments (biotic or abiotic) or across time barriers (i.e. reproductive timing) and finalizes when reproductive isolation is developed as a by-product of the divergence between lineages (Schluter, 2001). Delimiting divergent groups, can elucidate tempo and mode of lineage formation (i.e. new species appearance) (Weisrock et al., 2006). As isolation increase between divergent populations, the populations become more specialized and speciation is likely to take place unless other processes, like extinction and/or reduction in gene flow, alter the course towards

“new” species formation (Hawthorne, Via, 2001; Via, Hawthorne, 2002). S. fitti show

135

levels of differentiation between sympatric (i.e. harbored by different hosts) and allopatric (i.e. different locations) populations that can correspond to early stages of the speciation (Nosil et al., 2002; Via, 2009). Further evidence may be necessary to evaluate whether or not they have diverge enough to constitute different species (i.e. developed reproductive isolation).

Speciation involves two stages; initial build-up of reproductive isolation, usually in allopatry, and reinforcement (Schluter, 2001). If selection is strong and gene flow is low between populations, allopatry may not be necessary (Turelli et al., 2001; Via, 2001). In

S. fitti both stages may be present, since geographic isolation and host selection have induced differentiation between its populations. Initial effects of speciation can be confused with other populations and evolutionary events, such as: genetic drift, speciation by fixation and/or polyploidy (Schluter, 2001). Ecological speciation can be determined by testing for divergent selection, hybrid fitness, ecological traits and or parallel speciation (Funk, 1998; Nosil, 2009; Schluter, 2001). Our data indicates that

Symbiodinium fitti populations may be undergoing either divergent adaptation (i.e. specialization to a specific coral host) (Schemske, Bradshaw, 1999) and/or parallel speciation (i.e. divergence in similar environments or host internal environment – conditions) (Funk, 1998). It is likely that specialization to a particular host and geography/environment is important in differentiation of populations and formation of new lineages in Symbiodinium.

136

References

Aanen D, Ros V, de Fine Licht H, et al. (2007) Patterns of interaction specificity of fungus-growing termites and Termitomyces symbionts in South Africa. Bmc Evol Biol 7, 115. Agashe D, Bolnick D (2010) Intraspecific genetic variation and competition interact to influence niche expansion. Proceedings of the Royal Society B: Biological Sciences 277, 2915-2924. Arnaud-Haond S, Belkhir K (2007) GENCLONE: a computer program to analyse genotypic data, test for clonality and describe spatial clonal organization. Mol Ecol Notes 7, 15-17. Baker AC (2003) Flexibility and specificity in coral-algal symbiosis: Diversity, Ecology, and Biogeography of Symbiodinium. Annual review of ecology 34, 661-689. Bandelt H-J, Forster P, Röhl A (1999) Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution 16, 37-48. Barneah O, Brickner I (2010) MultIbiont symbioses in the coral reef ecosystem. In: Symbioses and Stress (eds. Seckbach J, Grube M), pp. 121-144. Springer. Berkelmans R, van Oppen M (2006) The role of zooxanthellae in the thermal tolerance of corals: a ‘nugget of hope’for coral reefs in an era of climate change. Proceedings of the Royal Society B: Biological Sciences 273, 2305-2312. Blank R (1987) Cell architecture of the dinoflagellate Symbiodinium sp. inhabiting the Hawaiian stony coral Montipora verrucosa. Mar Biol 94, 143-155. Bolnick DI, Fitzpatrick BM (2007) Sympatric Speciation: Models and Empirical Evidence. Annual Review of Ecology, Evolution and Systematics 38, 459-487. Bush G (1994) Sympatric speciation in animals: new wine in old bottles. Trends in Ecology & Evolution 9, 285-288. Cavalli-Sforza LL, Edwards AWF (1967) Phylogenetic analysis: models and estimation procedures. Evolution 21, 550-570. Clark J, Hossain S, Burnside C, Kambhampati S (2001) Coevolution between a cockroach and its bacterial endosymbiont: a biogeographical perspective. Proceedings of the Royal Society of London. Series B: Biological Sciences 268, 393. Correa AMS, Brandt ME, Smith TB, Thornhill DJ, Baker AC (2009) Symbiodinium associations with diseased and healthy scleractinian corals. Coral Reefs 28, 437- 448. Coyne J, Orr H (1998) The evolutionary genetics of speciation. Philos T R Soc B 353, 287-305. Cwynar LC, MacDonald GM (1987) Geographical variation of lodgepole pine in relation to population history. The American Naturalist 129, 463-469. Drummond C, Xue H, Yoder J, Pellmyr O (2009) Host-associated divergence and incipient speciation in the yucca moth Prodoxus coloradensis (Lepidoptera: Prodoxidae) on three species of host plants. Heredity. Earl DA (2009) Structure Harvester v0.3, from website: http://users.soe.ucsc.edu/~dearl/software/struct_harvest/.

137

Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology 14, 2611-2620. Excoffier L, Laval G, Schneider S (2005) Arlequin ver. 3.0: An integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online 1, 47-50. Feder JL, Opp S, Wlazlo B, Reynolds K (1994) Host fidelity is an effective premating barrier between sympatric races of the apple maggot …. PNAS 91, 7990-7994. Felsenstein J (1995) PHYLIP (Phylogeny Inference Package) Version 3.57c. http://www.nih.go.jp/~jun/research/phylip/main.html. Finney JC, Pettay DT, Sampayo EM, et al. (2010) The Relative Significance of Host– Habitat, Depth, and Geography on the Ecology, Endemism, and Speciation of Coral Endosymbionts in the Genus Symbiodinium. Microb Ecol, 1-14. Fitzpatrick BM, Fordyce JA, Gavrilets S (2009) Pattern, process and geographic modes of speciation. Journal of Evolutionary Biology 22, 2342-2347. Foster K, Kokko H (2006) Cheating can stabilize cooperation in mutualisms. Proceedings of the Royal Society B: Biological Sciences 273, 2233-2239. Foster K, Wenseleers T (2006) A general model for the evolution of mutualisms. Journal of Evolutionary Biology 19, 1283-1293. Frank S (2003) Repression of competition and the evolution of cooperation. Evolution 57, 693-705. Funk DJ (1998) Isolating a role for natural selection in speciation: host adaptation and sexual isoation in Neochlamisus babbianae Leaf beetles. Evolution 52, 1744- 1759. Gardner TA, Cote IM, Gill JA, Gant A, Watkinson AR (2003) Long-term region-wide declines in Caribbean corals. Science 301, 958-960. Goulet T, Coffroth MA (2003) Genetic composition of zooxanthellae between and within colonies of the octocoral Plexaura kuna, based on small subunit rDNA and multilocus DNA fingerprinting. Mar Biol 142, 233-239. Hannes AR, Barbeitos M, Coffroth MA (2009) Stability of symbiotic dinoflagellate type in the octocoral Briareum asbestinum. Mar Ecol-Prog Ser 391, 65-72. Hawthorne D, Via S (2001) Genetic linkage of ecological specialization and reproductive isolation in pea aphids. Nature 412, 904-907. Hoeksema JD (2010) Ongoing coevolution in mycorrhizal interactions. New Phytol 187, 286-300. Hongoh Y, Deevong P, Inoue T, et al. (2005) Intra-and interspecific comparisons of bacterial diversity and community structure support coevolution of gut microbiota and termite host. Applied and environmental microbiology 71, 6590-6599. Hubisz MJ, Falush D, Stephens M, Pritchard JK (2009) Inferring weak population structure with the assistance of sample group information. Molecular Ecology Resources 9, 1322-1332. Hughes C, Eastwood R (2006) Island radiation on a continental scale: exceptional rates of plant diversification after uplift of the Andes. Proceedings of the National Academy of Sciences 103, 10334.

138

Jakobsson M, Rosenberg N (2007) CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinformatics 23, 1801-1806. Janzen D (1980) When is it coevolution? Evolution 34, 611-612. Kawakita A, Okamoto T, Goto R, Kato M (2010) Mutualism favours higher host specificity than does antagonism in plant-herbivore interaction. Proc R Soc B, 1- 11. Kirk NL, Andras JP, Harvell CD, Santos SR, Coffroth MA (2009) Population structure of Symbiodinium sp. associated with the common sea fan, Gorgonia ventalina, in the Florida Keys across distance, depth, and time. Mar Biol 156, 1609-1623. Klak C, Reeves G, Hedderson T (2004) Unmatched tempo of evolution in Southern African semi-desert ice plants. Nature 427, 63-65. LaJeunesse TC (2001) Investigating the biodiversity, ecology and phylogeny of endosymbiotic dinoflagellates in the Genus Symbiodinium using the ITS region: in search of a "species" level marker. Journal of Phycology 37, 866-880. LaJeunesse TC (2002) Diversity and community structure of symbiotic dinoflagellates from Caribbean coral reefs. Marine Biology 141, 387-400. LaJeunesse TC (2005) "Species" Radiations of Symbiotic Dinoflagellates in the Atlantic and Indo-Pacific Since the Miocene-Pliocene Transition. Mol Biol Evol 22, 570- 581. LaJeunesse TC, Loh W, Trench RK (2009) Do introduced endosymbiotic dinoflagellates ‘take’ to new hosts? Biol Invasions 11, 995-1003. LaJeunesse TC, Loh W, van Woesik R, et al. (2003) Low symbiont diversity in southern Great Barrier Reef Corals, relative to those of the Caribbean. Limnol. Oceanog. 48, 2046-2054. LaJeunesse TC, Pettay D, Sampayo EM (2010) Long-standing environmental conditions, geographic isolation and host–symbiont specificity influence the relative ecological dominance and genetic diversification of coral endosymbiondt in the genus Symbiodinium. J Biogeogr 37, 785-800. LaJeunesse TC, Trench RK (2000) Biogeography of two species of Symbiodinium (Freudenthal) inhabiting the intertidal sea anemone Anthopleura elegantissima (Brandt). Biological Bulletin 199, 126-134. Léotard G, Debout G, Dalecky A, Guillot S, Gaume L (2009) Range expansion drives dispersal evolution in an equatorial three-species symbiosis. PLoS ONE. Linder H (2008) Plant species radiations: where, when, why? Philos T R Soc B 363, 3097-3105. Mallet J, Meyer A, Nosil P, Feder JL (2009) Space, sympatry and speciation. Journal of Evolutionary Biology 22, 2332-2341. Martínez-Romero E (2009) Coevolution in Rhizobium-Legume Symbiosis? DNA and Cell Biology. McRae B, Beier P, LE DEWALD L (2005) Habitat barriers limit gene flow and illuminate historical events in a wide-ranging carnivore, the American puma. Molecular …. Muss A, Ross Robertson D, Stepien C, Wirtz P, Bowen BW (2001) Phylogeography of Ophioblennius: the role of ocean currents and geography in reef fish evolution. Evolution 55, 561-572.

139

Nosil P (2009) Adaptive population divergence in cryptic color-pattern following a reduction in gene flow. Evolution. Nosil P, Crespi B, Sandoval CP (2002) Host-plant adaptation drives the parallel evolution of reproductive isolation. Nature 417, 440-443. Ohkuma M, Noda S, Hongoh Y, Kudo T (2001) Coevolution of symbiotic systems of termites and their gut microorganisms. Riken Review, 73-74. Orr M, Smith TB (1998) Ecology and speciation. Trends in Ecology & Evolution 13, 502-506. Pandolfi JM, Bradbury RH, Sala E, et al. (2003) Global trajectories of the long-term decline of coral reef ecosystems. Science 301, 955-958. Parks J, Werth C (1993) A study of spatial features of clones in a population of bracken fern, Pteridium aquilinum (Dennstaedtiaceae). American Journal of Botany 80, 537-544. Pawlowski J, Holzmann M, Fahrni JF, Pochon X, Lee J (2001) Molecular identification of algal endosymbionts in large Miliolid Foraminifera: 2. Dinoflagellates. The Journal of Eukaryotic Microbiology 48, 368-373. Peakall R, Smouse PE (2006) Genalex 6: genetic analysis in Excel. Population genetic software for teaching and research. Mol Ecol Notes 6, 288-295. Pfiester L, Anderson D (1987) Dinoflagellate reproduction. In: The biology of dinoflagellates (ed. FJR T), pp. 611-648. Blackwell, London. Pimentel D, Smith G, Soans J (1967) A population model of sympatric speciation. The American Naturalist 101, 493-504. Pinzón JH, Devlin-Durante MK, Weber MX, Baums IB, Lajeunesse TC (2010) Microsatellite loci for Symbiodinium A3 (S. fitti) a common algal symbiont among Caribbean Acropora (stony corals) and Indo-Pacific giant clams (Tridacna). Conservation Genet Resour, 1-3. Pochon X, LaJeunesse TC, Pawlowski J (2004) Biogeographic partitioning and host specialization among foraminiferan dinoflagellate symbionts (Symbiodinium; Dinophyta). Mar Biol 146, 17. Pritchard JK, Stephens M, Donnelly P (2000) Inferece of population structure using multilocus genotype data. Genetics 155, 945-959. Richardson J, Pennington R, Pennington T, Hollingsworth P (2001a) Rapid Diversification of a Species-Rich Genus of Neotropical Rain Forest Trees. Science 293, 2242-2245. Richardson JE, Weitz FM, Fay MF, et al. (2001b) Rapid and recent origin of species richness in the Cape flora of South Africa. Nature 412, 181-183. Rodriguez-Lanetty M, Krupp D, Weis V (2004) Distinct ITS types of Symbiodinium in Clade C correlate with cnidarian/dinoflagellate specificity during onset of symbiosis. Marine Ecology Progress Series 275, 97-102. Rodriguez-Lanetty M, Wood-Charlson E, Hollingsworth L, Krupp D, Weis V (2006) Temporal and spatial infection dynamics indicate recognition events in the early hours of a dinoflagellate/coral symbiosis. Marine Biology 149, 713-719. Rosenberg N (2004) DISTRUCT: a program for the graphical display of population structure. Molecular Ecology Notes 4, 137-138. Rowan RA, Powers D (1991) A molecular genetic classification of Zooxanthellae and the evolution of animal-algal symbioses. Science 251, 1348-1351.

140

Sampayo EM, Ridgway T, Bongaerts P, Hoegh-Guldberg O (2008) Bleaching susceptibility and mortality of corals are determined by fine-scale differences in symbiont type. PNAS 105, 10444-40449. Sanderson M (1998) Reappraising adaptive radiation. American Journal of Botany 85, 1650-1655. Santos SR, Gutierez-Rodríguez C, Lasker HR, Coffroth MA (2003) Symbiodinium sp. associations in the gorgonian Pseudopterogorgia elisabethae in the Bahamas: high levels of genetic variability and population structure in symbiotic dinoflagellates. Mar Biol 143, 111-120. Santos SR, Shearer TL, Hannes AR, Coffroth MA (2004) Fine-scale diversity and specificity in the most prevalent lineage of symbiotic dinoflagellates (Symbiodinium, Dinophyceae) of the Caribbean . Mol Ecol 13, 459-469. Schemske D, Bradshaw H (1999) Pollinator preference and the evolution of floral traits in monkeyflowers (Mimulus). Proceedings of the National Academy of Sciences 96, 11910-11915. Schluter D (2001) Ecology and the origin of species. Trends in Ecology & Evolution. Schluter D, Conte G (2009) Genetics and ecological speciation. PNAS 106, 9955-9962. Selig E, Bruno J (2010) A global analysis of the effectiveness of marine protected areas in preventing coral loss. PloS one 5, e9278. Seutin G, White BN, Boag PT (1991) Preservation of avian blood and tissue samples for DNA analyses. Canadian Journal of Zoology 69, 82-90. Takezaki N, Nei M (1996) Genetic distances and reconstruction of phylogenetic trees from microsatellite DNA. Genetics 144, 389-399. Thornhill DJ, Fitt W, Schmidt G (2006a) Highly stable symbioses among western Atlantic brooding corals. Coral Reefs 25, 515-519. Thornhill DJ, LaJeunesse TC, Kemp D, Fitt W, Schmidt G (2006b) Multi-year, seasonal genotypic surveys of coral-algal symbioses reveal prevalent stability or post- bleaching reversion. Mar Biol 148, 711. Toller WW, Rowan RA, Knowlton N (2001) Zooxanthellae of the Montastraea annularis species complex: patterns of distribution of four taxa of Symbiodinium on different reefs across depths. Biological Bulletin 201, 348-359. Turelli M, Barton N, Coyne J (2001) Theory and speciation. Trends in Ecology & Evolution 16, 330-343. van Oppen M, Palstra F, Piquet A, Miller DJ (2001) Patterns of coral–dinoflagellate associations in Acropora: significance of local availability and physiology of Symbiodinium strains and host–symbiont selectivity. Proc. R. Soc. Lond. B 268, 1759-1767. Via S (2001) Sympatric speciation in animals: the ugly duckling grows up. Trends in Ecology & Evolution 16, 381-390. Via S (2009) Natural selection in action during speciation. PNAS. Via S, Hawthorne D (2002) The genetic architecture of ecological specialization: correlated gene effects on host use and habitat choice in pea aphids. The American Naturalist 159, 76-88.

141

Weisrock D, Shaffer H, Storz B, Storz R, Voss R (2006) Multiple nuclear gene sequences identify phylogenetic species boundaries in the rapidly radiating clade of Mexican ambystomatid salamanders. Molecular Ecology 15, 2489-2503. Whitfield J, Lockhart P (2007) Deciphering ancient rapid radiations. Trends in Ecology & Evolution 22, 258-265.

142

CONCLUSIONS

Integrative approaches to the study of coral-algae mutualisms are a powerful tool into the discovery of biological, ecological, and evolutionary traits in both dinoflagellates and scleractinian corals. The analyses presented here show Pocillopora as a genetically diverse scleractinian group composed of at least eight distinct types (i.e. species) each forming symbiotic relationships with Symbiodinium of either clade C or D. The genetic diversity of the group seems to be not correlated with traditional morphological descriptions of the species across geographic locations, rendering morphology as a less useful character in species delineations among this group. It is suggested future studies incorporate genetics to identify the studied species.

In the light of the observed patterns here, previous conflicting conclusions (i.e. shift in mode of reproduction) seem more adequate and natural in explaining the distribution and diversification of Pocilloporid corals in the Indo-Pacific Oceans. The likelihood that brooding/broadcasting evolved within the group after its divergence from the sister genera can allow to “divide” the genus brooding and broadcasting. In this division,.

Species previously tough to be shifters appear now to have just one mode; a more parsimonious explanation to the complexity of the genus.

Pocillopora species have the potential to reproduce sexually or asexually, and patterns of clonality in its populations seemed influenced by local environmental conditions rather that regional patterns. Contrary to what has been predicted by the

143

central-margin theory, populations on marginal conditions (i.e. TEP) populations show unexpected levels of sexuality.

The creation of gene flow barriers between populations is believed to be an important initial stage in the diversification of many taxa (i.e. speciation). In mutualistic organisms, the level of gene flow is usually influenced by intrinsic characteristics of the symbiotic relation. One such aspect in coral-algae relations seems to be the identity of the coral host. Sympatric scleractinian species harbor genetically differentiate Symbiodinium populations, even in host with horizontal transmission of the dinoflagellate. The identity of the host, and/or selection of host, increase isolation between populations of

Symbiodinium and are major driving factors in the diversification of this dinoflagellate group.

144

BIBLIOGRAPHY

Amaral FD (1994) Morphological variation in the reef coral Montastrea cavernosa in Brazil. Coral Reefs 13, 113-117.

Axelrod R, Hamilton WD (1981) The evolution of cooperation. Science 211, 1390-1396.

Ayre DJ, Veron JEN, Dufty SL (1991) The corals Acropora palifera and Acropora cuneata are genetically and ecologically distinct. Coral Reefs 10, 13-18.

Baker AC (2003) Flexibility and specificity in coral-algal symbiosis: Diversity, Ecology, and Biogeography of Symbiodinium. Annual Review of Ecology, Evolution & Systematics 34, 661-689.

Benzoni F, Stefani F, Pichon M, Galli P (2010) The name game: morpho-molecular species boundaries in the genus Psammocora (Cnidaria, Scleractinia). Zoological Journal of the Linnean Society.

Borger JL (2005) Scleractinian coral diseases in south Florida; incidence, species susceptibility, and mortality. Diseases of Aquatic Organisms 67, 249-258.

Borger JL, Steiner SC (2005) The spatial and temporal dynamics of coral diseases in Dominica, West Indies. Bulletin of Marine Science 77, 137-154.

Bronstein JL (1994) Our current understanding of mutualism. The Quaterly Review of Biology 69, 31-51.

Bruno J, Selig E, Freckleton R (2007) Regional Decline of Coral Cover in the Indo- Pacific: Timing, Extent, and Subregional Comparisons. PLoS ONE 2, e711.

Bruno JF, Edmunds PJ (1997) Clonal variation for phenotypic plasticity in the coral Madracis mirabilis. Ecology 78, 2177-2190.

Bruno JF, Petes LE, Harvell CD, Hettinger A (2003) Nutrient enrichment can increase the severity of coral diseases. Ecol Letters 6, 1056-1061.

Budd A, Romano S, Smith N, Barbeitos M (2010) Rethinking the Phylogeny of Scleractinian Corals: A Review of Morphological and Molecular Data. Integrative and Comparative Biology, icq062v061.

Budd AF (1988) Large-scale evolutionary patterns in the reef-coral Montastraea: The role of phenotypic plasticity. Int. Coral Reef Symp.

Budd AF, Johnson KG, Potts DC (1994) Recognizing Morphospecies in Colonial Reef Corlas: I. Landmark-Based Methods. Paleobiology.

145

Budd AF, Klaus JS (2001) The origin and early evolution of the Montastraea "annularis" species complex (Anthozoa: Scleractinia). Journal of Paleontology 75, 527-545.

Buddemeier RW, Fautin DG (1993) Coral Bleaching as an Adaptive Mechanism. BioScience 43, 320.

Buddemeier RW, Smith SV (1999) Coral adaptation and acclimatization: a most ingenious paradox. American Zoologist 39, 1-9.

Cairns S (1999) Species richness or recent scleractinian. Atoll Research Bulletin 459, 45.

Castro C, Pires DO (2001) Brazilian coral reefs: What we already know and what is still missing. Bulletin of Marine Science 69, 357-371.

Cavanaugh C (1994) Microbial Symbiosis: Patterns of Diversity in the Marine Environment. American Zoologist 34, 79-89.

Coffroth M, Santos SR (2005) Genetic diversity of symbiotic dinoflagellates in the genus Symbiodinium. Protist 156, 19-34.

Coles S (2003) Corals of Oman Richard Keech, Ahmed Yusuf Samdani and Steve Coles, North Yorkshire, United Kingdom.

Coles S, Brown BE (2003) Coral Bleaching - Capacity for acclimatization and adaptation. Advances in Marine Biology 46, 183-223.

Cooney RP, Pantos O, Le Tissier MDA, et al. (2002) Characterization of the bacterial cosortium associated with black band disease in coral using molecular microbiological techniques. Environmental Microbiology 4, 401-413.

Douglas AE (1994) Symbiotic interactions Oxford University Press, Oxford.

Douglas AE (2003) Coral bleaching - How and why? Marine Pollution Bulletin 46, 385- 392.

Dunlap, Dunlap (2007) Biomedicinals from the phytosymbionts of marine invertebrates: A molecular approach. Methods 42, 358.

Forsman ZH, Barshis DJ, Hunter C, Toonen RJ (2009) Shape-shifting corals: Molecular markers show morphology is evolutionarily plastic in Porites. BMC Evolutionary Biology 9.

Friedrich AB, Merkert H, Fendert T, et al. (1999) Microbial diversity in the marine sponge Aplysina cavernicola (formerly Verongia cavernicola) analyzed by fluorescence in situ hybridization (FISH). Marine Biology 134, 461.

Fukami H, Budd AF, Paulay G, et al. (2004) Conventional taxonomy obscures deep divergence between Pacific and Atlantic corals. Nature 427, 832-835.

146

Fukami H, Chen CA, Budd A, et al. (2008) Mitochondrial and Nuclear Genes Suggest that Stony Corals Are Monophyletic but Most Families of Stony Corals Are Not (Order Scleractinia, Class Anthozoa, Phylum Cnidaria). PLoS ONE 3, e3222.

Glynn PW (1993) Coral reef bleaching: ecological perspectives. Coral Reefs 12, 1-17.

Glynn PW (2001) Coral reef bleaching: Caribbean events and some problem areas for future research, 33-35.

Harley JL, Smith SE (1983) Mycorrhizal symbiosis Academic Press, London.

Hoegh-Guldberg O, Smith GJ (1989) The effect of sudden changes in temperature, light and salinity of the population density and export of zooxanthellae from the reef corals Stylophora pistillata(Esper) and Seriatopora hystrix. Journal of Experimental Marine Biology and Ecology 129, 279-303.

Huang D, Meier R, Todd PA, Chou LM (2009) More evidence for pervasive paraphyly in scleractinian corals: Systematic study of Southeast Asian Faviidae (Cnidaria; scleractinia) based on molecular and morphological data Molecular Phylogenetics and Evolution 50, 102-116.

Hughes TP (1994) Catastrophes, phase shift, and large-scale degradation of a Caribbean coral reef. Science 265, 1547-1551.

Hughes TP, Baird DR, Bellwood DR, et al. (2003) Climate change, human impacts, and the resilience of coral reefs. Science 301, 929-933.

Johnson M, Fernandez C, Pergent G (2002) The ecological importance of an invertebrate chemoautothrofic symbiosis to phanerogam seagrass beds. Bulletin of Marine Science 71, 1343-1351.

LaJeunesse T, Reyes-Bonilla H, Warner M (2007) Spring “bleaching” among Pocillopora in the Sea of Cortez, Eastern Pacific. Coral Reefs 26, 265.

LaJeunesse TC (2005) "Species" Radiations of Symbiotic Dinoflagellates in the Atlantic and Indo-Pacific Since the Miocene-Pliocene Transition. Molecular Biology and Evolution 22, 570-581.

Margulis L (1993) Symbiosis in cell evolution W.H. Freeman and company, New York.

Margulis L, Fester R (1991) Symbiosis as a source of evolutionary innovation : speciation and morphogenesis MIT Press, Cambridge, Mass.

Muscatine L (1967) Glycerol Excretion by Symbiotic Algae from Corals and Tridacna and Its Control by the Host. Science 156, 516-519.

147

Muscatine L (1990) The role of symbiotic algae in carbon and energy flux in reef corals. In: Ecosystems of the World 25, Elsevier, Amsterdam (ed. Dubinsky Z), pp. 75-87. Elsevier, Amsterdam.

Muscatine L, Porter JW (1977) Reef Corals: Mutualistic Symbioses Adapted to Nutrient- Poor Environments. BioScience 27, 454.

Pearse VB, Muscatine L (1971) Role of simbiotic algae (Zooxanthellae) in coral calcification. Biological Bulletin (Woods Hole) 141, 350-363.

Porter JW, Dustan P, Jaap WC, et al. (2001) Patterns of spread of coral disease in the Florida Keys. Hydrobiologia 460, 1-24.

Porter JW, Lewis SK, Porter KG (1999) The effect of multiple stressors on the Florida Keys coral reef ecosystem: A Landscape hypothesis and a physiological test. Limnol. Oceanog. 44, 941-949.

Porter JW, Meier OW (1992) Quantification of loss and change in Floridian reef coral populations. American Zoologist 32, 625-640.

Romano S, Palumbi SR (1996) Evolution of scleractinian corals inferred from molecular systematics. Science 271, 640-642.

Sampayo EM, Dove S, LaJeunesse T (2009) Cohesive molecular genetic data delineate species diversity in the dinoflagellate genus Symbiodinium. Mol Ecol 18, 500- 519.

Santavy DL, Peters EC, Quirolo C, Porter JW, Bianchi C (1999) Yellow-blotch diseas outbreak on reefs of San Blas Islands, Panama. Coral Reefs 18, 97.

Smith D (2001) Symbiosis research at the end of the millenium. Hydrobiologia 461, 49- 54.

Spalding MD, Raviliuous C, Green EP (2001) World atlas of coral reefs University of California Press, Berkeley.

Stat M, Carter D, Hoegh-Guldberg O (2006) The evolutionary history of Symbiodinium and scleractinian hosts-Symbiosis, diversity, and the effect of climate change. Perspectives in Plant Ecology, Evolution and Systematics 8, 23.

Stobart B (2000) A taxonomic reappraisal of Montipora digitata based on genetic and morphometric evidence. Zoological Studies.

148

Sutton DC, Hoegh-Guldberg O (1990) Host-Zooxanthella Interactions in Four Temperate Marine Invertebrate Symbioses: Assessment of Effect of Host Extracts on Symbionts. Biological Bulletin 178, 175-186.

Swanson R, Hoegh-Guldberg O (1998) Amino acid synthesis in the symbiotic sea anemone Aiptasia pulchella Marine Biology 131, 83-93.

Thornhill D, LaJeunesse T, Kemp D, Fitt W, Schmidt G (2006a) Multi-year, seasonal genotypic surveys of coral-algal symbioses reveal prevalent stability or post- bleaching reversion. Marine Biology 148, 711.

Thornhill DJ, Fitt W, K., Schmidt GW (2006b) Highly stable symbioses among western Atlantic brooding corals. Coral Reefs 25, 515-519.

Todd P (2008) Morphological plasticity in scleractinian corals. Biol. Rev. 83, 315-337.

Todd PA, Sanderson PG, Chou LM (2001) Morphological variation in the polyps of the scleractinian coral Favia speciosa (Dana) around Singapore. Hydrobiologia 444, 227-235.

Trench R (1997) Diversity of symbiotic dinoflagellates and the evolution of microalgal- invertebrate symbiosis 2, 1275-1286.

Trench RK (1979) The Cell Biology of Plant-Animal Symbiosis. Annual Review of Plant Physiology 30, 485-531.

Venn AA, Loram JE, Douglas AE (2008) Photosynthetic symbioses in animals. J. Exp. Bot. 59, 1069-1080.

Weil E (1992) Genetic and morphological variation in the Caribbean and Eastern Pacific Porites (Anthozoa, Scleractinia). Preliminary results 2, 643-655.

Wernegreen JJ (2004) Endosymbiosis: lessons in conflict resolution. PLoS Biology 2, e68.

Zook D (1998) A new symbiosis language. ISS Symbiosis news 1, 1-3.

149

APPENDICES

Appendix A. Allelic frequencies per locus for Pocillopora types I, II, and III and across types (Overall). Samples size, number of unique genotypes and number private alleles (bold) are shown by type and overall. Locus Allele type 1 type 2 type 3 Overall 180 0.002 0.000 0.000 0.001 183 0.015 0.000 0.000 0.013 186 0.036 0.000 0.065 0.037

189 0.063 0.409 0.391 0.096

002 192 0.232 0.045 0.391 0.237 - 195 0.565 0.455 0.130 0.532 Pd3 198 0.008 0.000 0.022 0.009 201 0.003 0.000 0.000 0.003 204 0.075 0.091 0.000 0.070 207 0.002 0.000 0.000 0.001 191 0.005 0.000 0.022 0.006 194 0.065 0.000 0.022 0.060 196 0.036 0.000 0.000 0.032 200 0.023 0.500 0.000 0.037 203 0.740 0.000 0.022 0.668 206 0.058 0.091 0.022 0.057

209 0.015 0.182 0.043 0.022

005 212 0.042 0.000 0.870 0.096 - 215 0.002 0.000 0.000 0.001 Pd3 218 0.003 0.227 0.000 0.010 221 0.002 0.000 0.000 0.001 224 0.003 0.000 0.000 0.003 227 0.002 0.000 0.000 0.001 233 0.002 0.000 0.000 0.001 235 0.002 0.000 0.000 0.001 257 0.002 0.000 0.000 0.001 186 0.000 0.000 0.043 0.003 188 0.002 0.000 0.043 0.004 190 0.261 0.000 0.022 0.237 192 0.360 0.000 0.130 0.333

194 0.302 0.000 0.739 0.322

006 196 0.054 0.136 0.022 0.054 - 198 0.002 0.136 0.000 0.006 Pd2 200 0.005 0.091 0.000 0.007 202 0.000 0.273 0.000 0.009 204 0.010 0.045 0.000 0.010 206 0.003 0.182 0.000 0.009 208 0.002 0.136 0.000 0.006 236 0.045 0.682 0.022 0.064

- 260 0.104 0.318 0.000 0.104 007 Pd2 262 0.002 0.000 0.000 0.001

150

264 0.002 0.000 0.000 0.001 280 0.029 0.000 0.000 0.026 282 0.089 0.000 0.022 0.082 305 0.093 0.000 0.000 0.083 329 0.024 0.000 0.152 0.032 331 0.000 0.000 0.783 0.053 353 0.281 0.000 0.022 0.254

355 0.005 0.000 0.000 0.004 371 0.013 0.000 0.000 0.012 375 0.162 0.000 0.000 0.146 398 0.133 0.000 0.000 0.120 417 0.002 0.000 0.000 0.001 464 0.008 0.000 0.000 0.007 468 0.008 0.000 0.000 0.007 154 0.002 0.000 0.000 0.001 157 0.062 0.000 0.000 0.056 008 - 160 0.781 0.273 0.000 0.712

Pd3 163 0.107 0.682 0.935 0.181 166 0.049 0.045 0.065 0.050 327 0.000 0.091 0.000 0.003 331 0.010 0.045 0.022 0.012

334 0.075 0.000 0.000 0.067

009 337 0.016 0.545 0.891 0.092 - 340 0.825 0.182 0.087 0.754 Pd3 342 0.013 0.136 0.000 0.016 345 0.060 0.000 0.000 0.054 348 0.002 0.000 0.000 0.001 289 0.003 0.000 0.000 0.003 292 0.731 0.000 0.000 0.658 295 0.034 0.091 0.043 0.037 298 0.008 0.000 0.522 0.042

301 0.019 0.727 0.435 0.070 304 0.104 0.000 0.000 0.094 Poc40 307 0.078 0.182 0.000 0.076 310 0.015 0.000 0.000 0.013 313 0.006 0.000 0.000 0.006 316 0.002 0.000 0.000 0.001 Sample size 351 12 29 392 Unique Genotypes 308 11 23 342 Private Alleles 35 2 2

151

Appendix B. List of Pocillopora species found in Veron 2000, the * indicates species with restricted distributions or endemics. Species included in the list by Cairns, 1999, these may correspond to the most commonly known names of the genus.

P. damicornis (Linnaeus, 1758) P. verrucosa (Ellis and Solander, 1786) P. capitata (Verill, 1864)* P. danae (Verill, 1864)* P. elegans (Dana, 1846) P. ligulata (Dana, 1846)* P. meandrina (Dana, 1846) P. eyudoxi (Milne Edwards and Haime 1860) P. molokensis (Vaughan, 1907) P. woodjonesi (Vaughan, 1918)* P. ankeli (Sheer and Pillai, 1974) P. inflata (Glynn, 1999)* P. effusus (Veron, 2000)* P. fungiformis (Veron, 2000)* P. indiania (Veron, 2000)* P. kelleheri (Veron, 2000)* P. zelli (Veron, 2000)*

152

Appendix C. Pocillopora spp samples used for morphometric analysis and accession numbers from the Museo de Historia Natural de la U. Autónoma de Baja California Sur (MHNUABCS).

Pocillopora damicornis Pocillopora meandrina Pocillopora capitata MHNUABCS # Collection MHNUABCS # Collection MHNUABCS # Collection site site site 906 San Gabriel 835 Punta Arenas 837 Punta Arenas 813 San Gabriel 913 San Gabriel 1434 Gaviota 1361 San Jose del 1232 Cabo Pulmo 1432 Gaviota Cabo 1224 Cabo Pulmo 910 San Gabriel 891 Isla Catedral 951 Isla Cerralvo 780 Isla San Jose 1204 Cabo Pulmo 694 Gaviota 1357 San Jose del 1214 Cabo Pulmo Cabo 1926 Isla Cerralvo 785 Isla San Jose 890 Isla Catedral 898 Isla Catedral 622 Punta Perica 1192 Isla Cerralvo 798 Isla San Jose 618 Punta Perica 892 Isla Catedral 686 Gaviota 1229 Cabo Pulmo 894 Isla Catedral 794 Isla San Jose 1236 Cabo Pulmo 889 Isla Catedral 343 Isla Cerralvo 832 Punta Arenas 1435 Gaviota 1210 Cabo Pulmo 1234 Cabo Pulmo 1201 Cabo Pulmo 808 San Gabriel 817 San Gabriel 893 Isla Catedral 896 Isla Catedral 836 Punta Arenas 1348 San Jose del Cabo 683 Gaviota 1353 San Lucas 1213 Cabo Pulmo 862 San Gabriel 736 Isla San Jose 1202 Cabo Pulmo NN1 957 Isla Cerralvo 1205 Cabo Pulmo 684 Gaviota 640 Punta Perica 788 Isla San Jose 687 Gaviota 639 Punta Perica 1930 La Paz

153

Appendix D. Phylogenetic reconstruction showing the relations between populations of Symbiodinium fitti harbored by four coral hosts (Acropora palmata, A. cervicornis, M. annularis and Stephanocoenia intersepta) in different geographic locations through the Caribbean (Triangles = Mexico, Squares = Panama and Circles = Colombia; Ctg = Cartagena and SM = Santa Marta). The tree was build with a unweighted pair group method with arithmetic mean (UPGMA) based on the Chord-distance (Cavalli-Sforza and Edwards, 1967), using ten microsatellite loci genotypes (n = 94).

154 Jorge H. Pinzón C.

VITA Jorge H. Pinzón C. [email protected]

EDUCATION Ph.D. Biology, The Pennsylvania State University, May 2011 M.Sc. Marine Sciences, U. of Puerto Rico – Mayagüez, June 2004 Specialization in Environmental Management of Coastal Zones, U. Jorge Tadeo Lozano – Cartagena, Colombia, June 2001 B.S. Marine Biology, U. Jorge Tadeo Lozano – Bogotá, Colombia, April 1997 PROFESSIONAL AND TEACHING EXPERIENCE Graduate Assistant 2006 – 2011 – The Pennsylvania State University Symbiosis, ecology and evolution laboratory (LaJeunesse). The co-evolution of coral-algae associations

Research Assistant 2002 – 2004 – University or Puerto Rico – Mayagüez Coral Biology and Ecology Laboratory (Weil) Ecological and disease surveys in corals 2000 – 2001 – Environmental specialist Adolfo Sanjuan Environmental Consultant 1995 – 2000 INVEMAR - Marine Research Institute – Colombia Various positions including: Research scientist (1997- 2000), Thesis student (1996-1997), Professional practicing (1995-1996) and Research assistant (1995)

Teaching assistant The Pennsylvania State University 2008 to present – General Biology Laboratory 2008 – Tropical Ecology Course in Costa Rica Florida International University 2004 to 2007 – Introduction to Marine Biology, Ecology, General Biology and Microbiology laboratories U. of Puerto Rico – Mayagüez 2001-2004 – Cellular Physiology, General Chemistry and Organic Chemistry laboratories

GRANTS AND AWARDS 2008 Alfred P. Sloan Foundation, Minority Ph.D. Program in Mathematics Science and Engineering. 2006 Short Term Visit Grant Smithsonian Institution. 2005 Lerner Gray Memorial Grant American Museum of Natural History. 2004 Student Research Award