Diversity, Phylogeography, and Taxonomy of Hard-Corals in the Genus Porites from the Arabian Peninsula

Dissertation by

Tullia Isotta Terraneo

In partial Fulfillment of the Requirements

For the Degree of

Doctor of Philosophy

King Abdullah University of Science and Technology, Thuwal,

Kingdom of Saudi Arabia

Tullia Isotta Terraneo

EXAMINATION COMMITTEE PAGE

The dissertation of Tullia Isotta Terraneo is approved by the examination committee.

Committee Chairperson: Prof. Michael L. Berumen Committee Co-Chair: Prof. Andrew H. Baird Committee Members: Prof. Christian R. Voolstra, Prof. Mark Tester External Examiner: Prof. James D. Reimer

ABSTRACT

Diversity, Phylogeography and Evolution of Hard-Corals in the genus Porites from the Arabian Peninsula

Tullia Isotta Terraneo

The genus Porites is one of the most important scleractinian genera in terms of species diversity and panmictic tropical distribution. However, Porites is notorious for challenging taxonomic identification based on colony gross morphology, micromorphology, and single gene analyses, suggesting that the current classification poorly represents real evolutionary relationships.

This research integrates skeletal morphology data and single locus genetic evidence with genome-wide analyses and alternative line of evidence to taxonomy (i.e. symbiotic association data), with the aim of clarifying biodiversity, biogeography, and taxonomy of Porites from the

Arabian Peninsula.

In this dissertation, I evaluated the diversity of Porites in the Red Sea and the Gulf of Aden, providing a basic morpho-molecular background to the taxonomy of Porites in the region, and highlighting that a) the current taxonomic and phylogenetic position of 15 Porites morphological species needs to be reassessed, and that b) coral biodiversity in the Arabian region needs to be re-evaluated. To address this, I reconstructed the complete mitochondrial genomes of two endemic Red Sea species, contributing a

4 solid framework for clarifying the phylogeny and taxonomy of Porites in future molecular studies. I implemented the morpho-molecular results with high-throughput sequencing data, generating a comprehensive hypothesis of species boundaries and biogeography of Porites in the seas surrounding the Arabian Peninsula. These results suggested that 15 morphological species from this region, were clustered into eight molecular lineages, two of which previously unknown. Finally, using the nuclear

Internal Transcribed Spacer II (ITS2) marker in a high-throughput sequencing framework, I presented evidence derived from symbiotic association of

Porites with dinoflagellates in the family Symbiodiniaceae. Symbiont diversity showed patterns of geographic-specific association at multiple levels, including at the level of Symbiodiniaceae genera, majority ITS2 sequences, and ITS2 type profile levels. Specific associations with host genotypes were also recovered, providing a further line of evidence that the current taxonomy of Porites is in need for revision.

This dissertation highlights the utility of an integrated approach to taxonomy in elucidating species boundaries and phylogenetic relationships in Scleractinia and represents a framework that could be applied to other taxa awaiting revision.

ACKNOWLEDGMENTS

I would like to express my sincere gratitude to my supervisors Prof. Michael

Berumen and Prof. Andrew Baird for funding this research, and giving me the chance to grow on a scientific and personal level over the time of my

PhD. The endless opportunities that I got during the past 3.5 years would have not been possible without your support and guidance. Thank you.

Thanks to my committee members Prof. Christian Voolstra, Prof. Mark

Tester, and my external examiner Prof. James Reimer for their contributions and great suggestions during the preparation of this dissertation.

I am grateful to Dr. Roberto Arrigoni for the time and patience he invested in teaching me all I know about molecular biology and making me love evolution.

Thanks to Prof. Francesca Benzoni for molding the scientist I am from the beginning. Grazie for bringing me in a world I barely knew existed. All I learned and achieved in science is because of the trust, guidance, and infinite help you gave me.

Last but not least, thanks to my partner Luis, and my amazing family, mamma, papá, Elia, Ludy & Kaila. Thanks for believing in me and giving me the freedom to be the person I wanted to be.

TABLE OF CONTENTS

EXAMINATION COMMITTEE PAGE ...... 2 ABSTRACT ...... 3 ACKNOWLEDGMENTS ...... 5 TABLE OF CONTENTS...... 6 LIST OF ABBREVIATIONS ...... 10 LIST OF FIGURES ...... 15 LIST OF TABLES...... 19 1. CHAPTER 1: INTRODUCTION AND LITERATURE REVIEW ...... 21 1.1 Scleractinian corals ...... 21 1.2 Traditional taxonomy and systematics in the Scleractinia ...... 24 1.3 The molecular revolution ...... 35 1.4 The use of alternative traits to clarify scleratinian evolution ...... 45 1.5 The genus Porites ...... 47 1.6 Project summary and objectives ...... 55 1.7 References ...... 57 2. CHAPTER 2: MORPHOLOGY AND MOLECULES REVEAL TWO UNKNOWN SPECIES OF PORITES (SCLERACTINIA, PORITIDAE) FROM THE RED SEA AND THE GULF OF ADEN ...... 72 2.1 Abstract ...... 73 2.2 Introduction ...... 74 2.3 Materials and methods ...... 78 2.3.1 Sampling ...... 78 2.3.2 Abbreviations ...... 80 2.3.3 DNA extractions, amplification and sequence analyses ...... 81 2.3.4 Phylogenetic relationships reconstruction ...... 83 2.3.5 Species delimitation analyses ...... 84 2.4 Results ...... 85

2.4.1 Red Sea and Gulf of Aden Porites identification ...... 85 2.4.2 Taxonomy ...... 86 2.4.3 Molecular results ...... 95 2.5 Discussion...... 100 2.5.1 Comparison of morphological structures ...... 100 2.5.2 Molecular considerations ...... 102 2.5.3 Species complex in Porites ...... 103 2.5.4 Species biogeography and distribution ...... 106 2.6 Conclusions ...... 109 2.7 References ...... 117 3. CHAPTER 3: USING ezRAD TO RECONSTRUCT THE COMPLETE MITOCHONDRIAL GENOMES OF PORITES FONTANESII AND PORITES COLUMNARIS (CNIDARIA: SCLERACTINIA) ...... 123 3.1 Abstract ...... 125 3.2 Introduction ...... 126 3.3 Materials and methods ...... 126 3.4 Results and discussion ...... 127 3.5 Conclusion...... 129 3.6 References ...... 134 4. CHAPTER 4: PHYLOGENOMICS AND PHYLOGEOGRAPHY OF PORITES FROM THE ARABIAN PENINSULA ...... 136 4.1 Abstract ...... 137 4.2 Introduction ...... 138 4.3 Materials and Methods ...... 142 4.3.1 Collection and identification ...... 142 4.3.2 DNA extraction and quantification ...... 144 4.3.3 PCR amplification and sequencing...... 144 4.3.4 Restriction enzyme digestion and ezRAD libraries preparation .. 145 4.3.5 ezRAD data processing ...... 146 4.3.6 Reference assemblies and phylogenetic analyses of mitochondrial genomes, histone, and rDNA regions ...... 149 4.3.7 Genetic clustering analyses ...... 151 4.3.8 Species delimitation and species tree inference ...... 152

4.3.9 Divergence time analysis ...... 155 4.4 Results ...... 157 4.4.1 Morphological identification of Porites ...... 157 4.4.2 Mitochondrial CR, mitochondrial genomes, histone, and rDNA phylogenetic reconstructions...... 157 4.4.3 Holobiont and coral phylogenomic analyses ...... 161 4.4.4 Clustering analyses ...... 165 4.4.5 Species delimitation analyses ...... 169 4.4.6 Divergence time analysis ...... 173 4.5 Discussion...... 176 4.5.1 Diversity of Porites from the Arabian Peninsula ...... 176 4.5.2 Biogeography of Porites in the Arabian Peninsula ...... 180 4.6 Conclusions ...... 183 4.7 References ...... 185 5. CHAPTER 5: ENVIRONMENTAL LATITUDINAL GRADIENTS AND HOST SPECIFICITY SHAPE SYMBIODINIACEAE DISTRIBUTION IN RED SEA PORITES CORALS ...... 194 5.1 Abstract ...... 196 5.2 Introduction ...... 197 5.3 Material and Methods ...... 201 5.3.1 Sampling and identification ...... 201 5.3.2 Coral host DNA extraction and PCR amplification ...... 203 5.3.3 Coral host phylogeny reconstructions ...... 203 5.3.4 Symbiodiniaceae MiSeq sequencing library preparation...... 204 5.3.5 Symbiodiniaceae MiSeq data processing ...... 205 5.3.6 Environmental data ...... 206 5.3.7 Statistical analyses ...... 206 5.4 Results ...... 208 5.4.1 Traditional and molecular identification of Porites ...... 208 5.4.2 Symbiodiniaceae community structure ...... 211 5.4.3 Symbiodiniaceae diversity in relation to environmental variables ...... 218 5.5 Discussion...... 218 5.5.1 Porites morpho-molecular diversity ...... 218

5.5.2 Symbiodiniaceae biogeography across Red Sea gradients ...... 219 5.5.3 Host-symbiont specificity ...... 222 5.6 Conclusions ...... 223 5.7 References ...... 225 6. CHAPTER 6: CONCLUSIONS ...... 234 6.1 General conclusions...... 234 6.2 Future perspectives ...... 239 6.3 References ...... 241 LIST OF PUBLICATIONS ...... 246 LIST OF APPENDICES ...... 249

LIST OF ABBREVIATIONS

ABGD Automatic Barcode Gap Discovery

AIC Akaike Information Criterion

AG Arabian Gulf

ARC Australian Research Council

ATP6 ATP Synthase subunit 6

ATP8 ATP Synthase subunit 8

ATPs β ATP Synthase subunit β

Au–Pd Gold-Palladium bp base pair

BF Bayes Factor

BFD* Bayes Factor Delimitation

BI Bayesian Inference

BIC Bayesian Information Criterion

CalM Calmoduline

Chl-a Chlorophyll-a

COI – cox1 Cytochrome c Oxidase subunit 1 cox2 Cytochrome c Oxidase subunit 2 cox3 Cytochrome c Oxidase subunit 3 cv Maen depth

Cytb Cytochrome b

DistLM Distance-based Linear Modelling

Df Degrees of freedom

DNA Deoxyribonucleic Acid ddRAD double digest Restriction Site-Associated DNA

Sequencing dsDNA double stranded DNA

EMBL The European Molecular Biology Laboratory

ENA European Nucleotide Archive ezRAD easy Restriction Site-Associated DNA Sequencing

FEG SEM Field Emission Gun Scanning Electron Microscope

GMYC Generalized Mixed Yule Coalescent

GoA Gulf of Aden

GoO Gulf of Oman

GTR General Time-Reversible model

ITS Internal Transcribed Spacer

ITS1 Internal Transcribed Spacer 1

ITS2 Internal Transcribed Spacer 2

KAUST King Abdullah University of Science and

Technology

K2P Kimura 2-Parameter model

JC69 Jukes-Cantor model

Ma Million years ago

MAFFT Multiple Alignment using Fast Fourier Transform

MCMC Markov Chain Monte Carlo md Minimum distance

MNHN Muséum National d’histoire Naturelle, Paris, France

ML Maximum Likelihood

MLE Maximum Likelihood Estimate ms Missing data mtCR mitochondrial Control Region mtDNA mitochondrial DNA

MTQ Museum of Tropical Queensland, Townsville, QLD,

Australia

NAD2 NADH dehydrogenase subunit 2

NAD3 NADH dehydrogenase subunit 3

NAD4L NADH dehydrogenase subunit 4

NAD5 NADH dehydrogenase subunit 5

NAD6 NADH dehydrogenase subunit 6

NASA National Aeronautic and Space Administration

NCBI The National Center for Biotechnology Information

GenBank nDNA nuclear DNA

NGS Next Generation Sequencing

NHM Natural History Museum London, UK

IUCN International Union for Conservation of Nature

PAUP Phylogenetic Analysis Using Parsimony

PCI Pax-C 46/ 47 Intron

PCA Principal Component Analyses

PCR Polymerase Chain Reaction

POC Particulate Organic Carbon

PTP Poisson Tree Processes qPCR quantitative Polymerase Chain Reaction

RADSeq Restriction Site-Associated DNA Sequencing

RAxML Randomized Axelerated Maximum Likelihood rDNA ribosomal DNA rnl large subunit ribosomal DNA rns small subunit ribosomal DNA

RS Red Sea

SEM Scanning Electron Microscope/ Microscopy

SNPs Single Nucleotide Polymorphisms

SS Sum of Squares

SST Sea Surface Temperature tRNA transfer RNA

UNIMIB University of Milano – Bicocca

USNM National Museum of Natural History, Washington

DC, USA

WoRMS World Register of Marine Species

LIST OF FIGURES

Figure 1.1 Main scleractinian coral skeletal features used in traditional Scleractinia classification. Modified from Wells (1956). Treatise on Invertebrate Paleontology, P. F336...... 28 Figure 1.2 The classification system and evolutionary relationships among Scleractinia proposed by Wells (1956). Branches represent families, patterns represent superfamilies, and columns represent suborders. Modified from Wells (1956). Treatise on Invertebrate Paleontology, P.363 ...... 29 Figure 1.3 The classification system and evolutionary relationships among Scleractinia proposed by Veron (1995). Branches correspond to families. The width of the branches corresponds to the number of genera in each family. Modified from Veron (2000), Corals of the World, P.31 ...... 34 Figure 1.4 Phylogenetic relationships among scleractinian corals based on combined COI and CytB genes. Branch support represent Posterior Bayesian probabilities (>70%), and ML bootstrap values (>50%). Numbers in circles show connections among trees (A to D). Letter codes correspond to traditional coral families. Roman numbers indicate clade subdivisions according to the tree. Colors in the picture refer to coral suborders sensu Veron (1995) as outlined in the legend on the left side of the tree...... 44 Figure 1.5 Schematic pattern of Porites corallites main skeletal features. Modified from Forsman (2015 ...... 53 Figure 2.1 Map of the sampling localities in the Red Sea (Saudi Arabia) and north-western Gulf of Aden (Yemen) of the Porites spp. material examined in this study...... 80 Figure 2.2 Porites sp. 1 (a‒d) and Porites sp. 2 (c‒h) in situ: a, colony of KAUST SA0172 (in the dashed circle), Marka, Farasan Banks, Saudi Arabia (same as Fig. 10 b); b, colony of MNHN-IK-2012-14238, Marka, Farasan Banks, Saudi Arabia (same as Fig. 10 a, c); c, a colony at Aminah, Farasan Islands, Saudi Arabia, the inset shows a detail of the retracted polyps through which five obvious pali (white arrows) can be observed; d, close up of the surface of the colony in b showing the typical beige-green coenosarc and brown polyp colouration; e‒f, colonies at Burum, Yemen; g, colony of MNHN-IK- 2012-14238 (same as Figs. 12 a‒d), south Burum, Yemen; h, close-up of the same colony as g showing the typical white coenosarc and red polyp colouration...... 92

Figure 2.3 In situ images of the Porites colonies examined in this study, circular insets shows polyp detail for each colony: a, P. fontanesii, KAUST SA0438, Qita al Kirsh, Saudi Arabia; b, P. columnaris, UNIMIB P0009, Balhaf, Yemen; c, P. rus, UNIMIB BA0109, Bir Ali, Yemen; d, P. lobata, KAUST SA0054, Farasan Banks, Saudi Arabia; e, P. lutea, UNIMIB BA0098, Sikha Island, Yemen; f, P. echinulata, UNIMIB AD0007, Aden, Yemen; g, P. compressa, UNIMIB BA0025, Bir Ali, Yemen; h, P. solida, UNIMIB P0011, Balhaf, Yemen. 94 Figure 2.4 RAxML phylogenetic reconstructions of Red Sea and Yemen Porites at two molecular loci. a, mitochondrial Control Region, b, nuclear Internal Transcribed Spaces region. Number on the branches represent support values corresponding to Bayesian posterior probabilities (>90%), ML bootstrap values (>70%). Porites porites was selected as outgroup in both the analyses. In c, black curved lines connect the two sequences of heterozygotes individuals. Summary of putative species delimitation drawn by ABGD, PTP, and GMYC, (b, d)...... 99 Figure 2.5 Morphology of the designated reference specimens of Porites sp. 1.: a, MNHN-IK-2012-14238 (same specimen as in Fig. 3), on the left, and KAUST SA0180 on the right; b, KAUST SA0172 (same specimen as in Fig. 2); c, detail of the corallites on the corallum surface of the reference specimen. Scale bars represent: Fig. a‒b, 1 cm; Fig. c, 1 mm...... 111 Figure 2.6 SEM images of a fragment of the reference specimen of Porites sp. 1. MNHN-IK-2012-14238 (same specimen as in Fig 7 b, 9 d, 10 b): a and b, corallites on different parts of the corallum surface separated by well- formed coenosteum; c, side view of two adjacent corallites showing the five exert pali (orange arrows) surrounding the columella sitting deeper in the fossa (light blue arrows); d, a corallite showing the typical septa arrangement and large pali size. Scale bars represent: a‒b, 1 mm; c‒d, 500 µm...... 112 Figure 2.7 SEM images of one of the two fragments of the reference specimen of Porites sp. 2. MNHN-IK-2016-208 (same specimen as in Fig. 8): a and b, corallites on different parts of the corallum surface separated by coenosteum; c, a corallite showing the typical septal arrangement with a short dorsal directive septum and a well-formed radius uniting the two pali found in correspondence of the dorsal lateral pairs (white arrow); d, another corallite in which the radius uniting the two dorsal pali is not formed. Scale bars represent: a‒b, 1 mm; c‒d, 500 µm...... 113 Figure 2.8 Schematic representation of the corallite structures observed in the reference specimens of: a, Porites sp. 1.; b, Porites sp. 2 Dashed lines

17 indicate structures that are well formed in some corallites but absent in others within the same corallum ...... 114 Figure 2.9 SEM images of the corallites of the Porites species examined in this study: a, P. fontanesii, KAUST SA0438 (same as Fig. 8 a); b, P. columnaris, UNIMIB P0009 (same as Fig. 8 b); c, P. rus, UNIMIB BA0109 (same as Fig. 18 c); d, P. lobata, KAUST SA0054 (same as Fig. 8 d); e, P. lutea, UNIMIB BA0098 (same as Fig. 8 e); f, P. echinulata, UNIMIB AD0007 (same as Fig. 8 f); g, P. annae, UNIMIB BA0025 (same as Fig. 8 g); h, P. solida, UNIMIB P0011 (same as Fig. 8 h)...... 116 Figure 3.1 Phylogenetic reconstruction based on complete mitochondrial genomes of Porites fontanesii, P. columnaris and other Scleractinia. Numbers at nodes represent Bayesian posterior probabilities and maximum likelihood bootstrap values. Fungiacyathus stephanus was selected as an outgroup...... 133 Figure 4.1 Comparison of reconstructed species trees. a) coral mitochondrial CR, b) coral histone region, c) coral rDNA, d) coral mitochondrial DNA. Node values represent BI posterior probabilities and ML bootstrap supports. Roman numbers from I to VIII refer to the assigned clade numbers. Colour codes are explained in the legend ...... 160 Figure 4.2 Comparison of RAxML tree based on a) “holobiont-max” dataset, that allowed for 50% missing data and consisted of 54,108 SNPs, b) “coral- max” dataset, that allowed for 50% missing data and consisted of 96,986 SNPs, c) zoom-in Clade V in the “holobiont-max” dataset, d) zoom-in Clade V in the “coral-max” dataset. Values at nodes represent ML bootstrap supports. Roman numbers from I to VIII refer to the assigned molecular clade numbers. Colour codes are explained in the legend ...... 164 Figure 4.3 Principal Component Analysis (PCA) from the “coral-max” dataset that allowed for 50% missing data, and consisted of 96,986 SNPs. a) PCA results based on PC1 (x axis) and PC2 (y axis), b) PC3 (x axis) and PC4 (y axis). Colors in the picture are explained in the legend ...... 167 Figure 4.4 Admixture plots from the “coral-max” dataset that allowed for 50% missing data, and consisted of 96,986 SNPs. From top to bottom: a) K=5, best suggested model, b) K=4, second most probable model, K=8, c) third most probable model. Colour codes are explained in the legend. Clade from I to VIII refer to the assigned molecular clades ...... 169 Figure 4.5 Species tree of Porites. The claudogram represents the posterior distribution of species trees inferred with SNAPP, based on a priori imposition of xx taxa, 1107 unlinked biallelic SNPs and 0% missing data. High colour

18 densities are representative of high topology agreement in the species tree. Values on branches represent posterior probabilities. Colour codes are explained in the legend ...... 172 Figure 4.6 Species tree calibrated chronogram. Purple bars represent 95% highest posterior densities (HPD). Node symbols represent posterior probabilities as explained in the legend. Values at nodes represent estimate time of node divergence. The scale bar represents millions of years and is based on the ICS International Chronostratigraphic Chart ...... 175 Figure 5.1 Sampling location and specimen collection overview (a) Red Sea map with sampling localities. (b) to (j) Porites morphologies encountered among our 80 collected samples. (b) Porites fontanesii, (c) Porites columnaris, (d) Porites sp. 1, (e) Porites rus, (f) Porites monticulosa, (g) Porites lutea, (h) Porites lobata, (i) Porites annae, (j) Porites echinulata, (k) Porites solida...... 202 Figure 5.2 RAxML phylogeny reconstructions of Red Sea Porites at two molecular loci. (a) mitochondrial Control Region (b) nuclear Internal Transcribed Spaces region. Number on the branches represent support values corresponding to Bayesian posterior probabilities (>90%), ML bootstrap values (>7%). Goniopora was selected as outgroup in both the analyses. In (b) black curved lines connect the two sequences of heterozygotes individual ...... 210 Figure 5.3 Stacked bar charts (100%) comparing Symbiodiniaceae diversity at the (a) (b) (c) genus level, (d) (e) (f) majority ITS2 sequence level (the 20 most abundant shown), (g) (h) (i) ITS2 type profile level (the 20 most abundant shown). For each of these three datasets, bar charts were plotted for three different factors: (a) (d) (g) Porites morphological species, (b) (e) (h) Porites molecular clade, (c) (f) (j) Red Sea localities ...... 215

LIST OF TABLES

Table 1.1 Summary of the main macro- and micro-skeletal features used to distinct scleractinian suborders in the main classifications proposed during the 19th and 20th centuries (except Veron (1995)). (x) refer to fossil taxa. .. 30 Table 1.2 Porites species recognized by WoRMS Editorial Board (2019) and by Veron (2000)...... 53 Table 2.1 Genetic distance values (p-distance) within and between Porites species from the Red Sea based on the mitochondrial CR. Interspecific pairwise comparisons of genetic distance are in bold; intraspecific genetic distances are indicated along the diagonal; standard deviations are reported in brackets or on the upper right-hand portions for each set of comparisons; n.c. not calculable ...... 99 Table 2.2 Genetic distance values (p-distance) within and between Porites species from the Red Sea based on the nuclear ITS region. Interspecific pairwise comparisons of genetic distance are in bold; intraspecific genetic distances are indicated along the diagonal; standard deviations are reported in brackets or on the upper right-hand portions for each set of comparisons; n.c. not calculable ...... 100 Table 3.1 Porites fontanesii complete mitochondrial genome annotation ...... 129 Table 3.2 Porites columnaris complete mitochondrial genome annotation ...... 131 Table 4.1 Bayes Factor delimitation (BFD*) results for each analysis using path sampling (PS). The number of lineages represents the number of putative species included in each analysis. BF values are used to rank species models, relative to the species model with the lowest marginal likelihood (MLE). The model with eight lineages was supported as the best fit model...... 170 Table 5.1 PERMANOVA results calculated for three biodiversity metrics (Red Sea locality, Porites molecular clade, and Porites morphological species) for the datasets Symbiodiniaceae genus, Symbiodiniaceae majority ITS2 sequences, and Symbiodiniaceae ITS2 type profile. df = degrees of freedom...... 216 Table 5.2 Marginal test results of the DistLM analyses for each of three datasets (Symbiodiniaceae genus, Symbiodiniaceae majority ITS2

20 sequence, and Symbiodiniaceae ITS2 type profile) against the environmental factors SST, Chl- a, POC, and Salinity. SS = Sum of squares; %var = percent of variance explained by each predictor variable...... 217

1. CHAPTER 1: INTRODUCTION AND LITERATURE REVIEW

1.1 Scleractinian corals

Coral reefs are limestone structures widespread along the world’s tropical belt. Although occupying only 0.1% of the ocean floor (Spalding & Grenfell,

1997), they host an outstanding biodiversity, comparable to the one of rain forests (Connell et al., 1978). According to recent estimates, more than 25% of total marine life is supported by coral reefs, and nearly one third of total fish diversity inhabits reef ecosystems (Appeltans et al., 2012; Bowen et al.,

2013; Fisher et al., 2011). For these reasons, coral reefs are considered one of the most productive and complex ecosystems on the planet (Connell,

1978; Odum & Odum, 1955). Besides their intrinsic biological value, coral reefs are fundamental physical structures that dissipate wave energy and create sheltered environments, such as lagoons for seagrass and mangroves. At the same time, they provide costal protection, and supply food and income to millions of people on shore communities (Moberg &

Folke, 1999; Roberts, 2009). Recent estimates evaluated the asset value of coral reefs close to $1 trillion, and benefits from these ecosystems reaching at least 500 million people from over 90 countries (Gattuso et al., 2015;

Hoegh-Guldberg et al., 2015).

Nonetheless, coral reefs are also extremely unstable communities, subject to recurrent natural disturbances (e.g. tropical cyclones, outbreaks of crown of thorns starfish Acanthaster planci, El Niño events) and impact

22 by anthropogenic activities (e.g. runoff of pollutant and nutrients, overfishing, oil rigs) (Hughes et al., 1992; Hughes et al., 2003; Munday et al.,

2009), that make them one of the most endangered ecosystems on the planet.

The building blocks of coral reefs are stony corals (Cnidaria Verrill,

1865, Scleractinia Bourne, 1900), radially symmetrical invertebrates that diverged from near the base of the metazoans tree of life around 445 Ma

(Stolarski et al., 2011). Within the sub-class Hexacorallia Haeckel, 1896, which comprises five others extant orders, i.e. Corallimorpharia Carlgren,

1943, Actiniaria (not documented taxonomic authority), Antipatharia (not documented taxonomic authority), Zoantharia de Blainville, 1830, and

Ceriantharia Perrier, 1893, the order Scleractiania (from now on referred as corals) differentiates due to the ability to actively produce continuous aragonitic skeletons that create three dimensional habitats and sustain a variety of creatures. Of the approximate 1400 extant coral species, almost

60% are colonial (Cairns, 1999), but solitary corals have evolved in at least six lineages (Barbeitos et al., 2010). Coral colonies are formed by connected units, i.e. the polyps, sharing integrated physiologies (Hughes et al., 1992). A total of 865 (55.9%) reef-building Scleractinia species (i.e. zooxanthellate corals) host within their endoderm symbiotic unicellular photosynthetic dinoflagellates of the family Symbiodiniaceae (LaJeunesse et al., 2018) (https://www.coraltraits.org). This relationship restricts

23 zooxanthellate species to shallow environments in warm subtropical to tropical regions, where they constitute the major framework builders of coral reefs.

A healthy state of hermatypic corals (sensu Wells, 1933) is fundamental to support reef ecosystems. Nevertheless, in the last decades the proportion of threatened corals has increased dramatically in response to increasing natural and anthropogenic stressors (Bellwood et al., 2004;

Knowlton, 2001). According to the International Union for Conservation of

Nature (IUCN) Red List criteria, 35.3% of zooxanthellate reef coral is at elevated risk of extinction (Carpenter et al., 2008, https://www.coraltraits.org). Nevertheless, the extinction risk faced by individual coral species is still poorly understood. In an era of ongoing biodiversity loss, taxonomic works have a direct impact on the designation of biodiversity hotspots, conservation schemes, and informing politics and legislation (Agapow & Bininda-Emonds, 2004; Karl & Bowen, 1999; Mann &

Plummer, 1992). Yet, at present, we lack basic knowledge of Scleractinia taxonomy, especially at lower taxonomic levels, i.e. genus and species level. This gap of knowledge further compromises the status of coral reefs, challenging species risk extinction assessments, and rendering the designation of conservation schemes an almost impossible challenge. In this context, making accurate estimations of the cumulative effects on coral biodiversity and climate change remain unrealistic.

1.2 Traditional taxonomy and systematics in the Scleractinia

The taxonomy and systematics of stony corals have a long and complex history (Esper, 1797; Forsskål, 1775; Lamarck, 1801; Linnaeus, 1758; Pallas,

1766), yet confusion in Scleractinia systematics remains at every taxonomic level (Kitahara et al., 2016). Traditional classification of hard corals has been based largely on macro-skeletal characters (i.e. corallite level structures such as the calice, the septa, the pali, the columella, the costae) of extant and extinct taxa (see Fig. 1.1 for main coral skeletal features). These macro- features remained the main source of evidence used for delimiting species and reconstructing evolutionary relationships among corals until the late

20th Century. Starting from 1800, pioneers such as Milne-Edwards & Haime

(1857), Duncan, (1885), and Ogilvie (1896), provided classifications of stony corals based solely on marcro-skeletal features. Milne-Edwards and Haime

(1857) introduced a uniform terminology for coral morphological structures, and proposed the first comprehensive classification of stony corals.

According to their systematic scheme the order Scleractinia was subdivided into two main extant categories, the Aporosa and Perforata, and three fossil categories (Tabulosa, Tabulata, and Rugosa). Milene-

Edwards and Haime’s subdivision was based on a series of skeletal features, comprising the wall structure (i.e. compact or porose), the subdivision of the visceral chamber (completely open or subdivided), and the septal arrangement (rudimentary or well developed). In 1885, Duncan added a

25 sixth category to Milne-Edwards and Haime’s subdivision, the Fungida, basing his idea on the wall structure (either compact or porose) and adding as a distinctive morphological feature, the presence or absence of wall granulation. In the 20th Century Vaughan & Wells (1943) and Wells (1956), revised the entire order and, on the basis of macro-morphological features, developed the “traditional” hard coral classification system that has been used until modern times. According to the principle that coral evolution could be explained by animal skeletal features, Wells (1956) identified five suborders (Astrocoeniina, Fungiina, Faviina, Caryophyllina, Dendrophyllina) and 33 families (20 extant), providing the first hypothesis of evolutionary relationships among coral families (Fig. 1.2) (Budd et al., 2010). Wells’ classification of stony corals was primarily based on the septal structural framework, particularly on the trabeculae arrangement within the septa.

Alloiteau (1957) and Chevalier & Beauvais (1987), proposed alternative approaches that used micro-structural data. On the basis of differences in septal ornamentation and the structure of the trabeculae, Alloiteau recognized eight suborders within the Scleractinia, and 71 (30 extant) families. Chevalier and Bauvais (1987) followed the microstructural evidence of Alloiteau, and added three new suborders to the eight identified by Alloiteau, bringing the total to 11 suborders and 55 families (a schematic summary of the main classification systems mentioned above is provided in Table 1.1). The most recent morphological phylogeny of corals

26 was provided by Veron (1995), who recognized 13 sub-orders (6 extant) and 61 families (24 extant) within the order Scleractinia (Fig. 1.3).

The incongruence among these systematic reconstructions emphasizes the general lack of an empirical and rigorous work in traditional classification efforts. The absence of quantitative or experimental data in traditional coral taxonomy challenges the use of morphological features in providing reliable evolutionary hypothesis in corals, yet morphological traits should not be discarded but instead reconsidered under a different evolutionary perspective.

Nonetheless, skeletal morphology in corals has further limitations that compromise our understanding of scleractinian evolution. Even in 1886, in the report from the Challenger Expedition, Quelch emphasized that coral morphology varied in response to the environment. It has been experimentally demonstrated that several environmental factors can influence coral morphology, such as light, sedimentation, wave action, depth, and salinity, and the effect that these factors have on coral morphologies varies among and within taxa (Bruno & Edmunds, 1997; Budd,

1993; Miller, 1994; Randall, 1976, Todd, 2008). During the past century, the concept of phenotypic plasticity, i.e. the capacity of a single genotype to express different phenotypes in relation to different environments, has been largely regarded as environmental noise obscuring real evolutionary characteristics of organisms (Sultan, 2000). Nowadays, there is general

27 agreement that phenotypic plasticity does not only involve morphological responses, but encompasses fundamental aspects of an organism’s physiology, life-history, and behavior, that can enhance the organism’s fitness in a certain environment (Todd, 2008). Trying to reconstruct animal evolution based exclusively on their phenotype results in often meaningless and incorrect evolutionary hypotheses.

Morphological based taxonomy is subject to further limitations, such as convergent evolution and homoplasy, i.e. the independent evolution of similar features in species belonging to different evolutionary lineages, and different rates of evolution of morphological features that render making inferences about the evolution of corals following a traditional approach particularly challenging (Van Oppen et al., 2001).

Figure 1.1 Main scleractinian coral skeletal features used in traditional Scleractinia classification. Modified from Wells (1956). Treatise on Invertebrate Paleontology, P.

F336.

Figure 1.2 The classification system and evolutionary relationships among Scleractinia proposed by Wells (1956). Branches represent families, patterns represent superfamilies, and columns represent suborders. Modified from Wells (1956). Treatise on Invertebrate

Paleontology, P.363

Milne-Edwards and Suborder Distinctive morphological features Haime (1857)

Aporosa Visceral chamber: empty/subdivided

Septa: developed

Sclerenchima: compact

Perforata Visceral chamber: empty/ subdivided

Septa: developed

Sclerenchima: porose

Tabulosa (x) Visceral chamber: empty/subdivided

Septa: rudimentary

Tabulata (x) Visceral chamber: subdivided

Septa: rudimentary, exameral

Rugosa (x) Visceral chamber: subdivided

Septa: developed, tetrameral

Wells (1956)

Astrocoeniina Septa: laminar, simple spines

Trabeculae: single, simple or compound

Synapticulae: absent

Fungiina Septa: fenestrate

Trabeculae: multiple, simple or compound

Synapticulae: present

Faviina Septa: laminar, isolated spines

Trabeculae: one or more fan system of multiple simple or compound

Synapticulae: absent

Caryophyllina Septa: laminar

Trabeculae: one fan system of multiple single

Synapticulae: absent

Dendrophyllina Septa: laminar

Trabeculae: one fan system of multiple simple

Synapticulae: present

Alloiteau (1952)

Archaeocoeniidae Synapticulae: absent

Endotheca: developed

Symmetry: radial

Septa: discontinuous

Trabeaculae: few

Stylinida Synapticulae: absent

Endotheca: developed

Symmetry: radial

Septa: continuous

Trabeaculae: few

Synapticulae: present

Fungiida Endotheca: developed

Symmetry: bilateral

Septa: perforata

Trabeaculae: continuous

Synapticulae: present

Astraeoida Endotheca: developed

Symmetry: radial

Septa: continuous

Septal ornamentation: present

Trabeaculae: numerous

Synapticulae: absent

Meandriida Endotheca: developed

Symmetry: radial

Septa: continuous

Septal ornamentation: absent

Trabeaculae: numerous

Synapticulae: absent

Amphiastraeida Endotheca: developed

Symmetry: bilateral

Wall: archeotechal

Synapticulae: absent

Eupsammiida Septa: perforata

Trabeculae: discontinuous

Synapticulae: present

Caryophyllida Septa: perforata

Endotheca: rare, absent

Synapticulae: absent

Chevalier and Beauvais (1987)

Stylophyllina Septa: no plane

Trabeculae: absent

Sclerenchyma: lamellar

Radial elements: thecal origin

Pachytjecalina Septa: no plane

Trabeculae: absent

Theca: fibrous

Distichophyllina Septa: medioseptal plane with lateral axes

T Trabeculae: absent

Wall: septal or paratechal

Archaeocaeniina Septa: no plane

Trabeculae: present

Synapticulae: absent

Stylinina Trabeculae: simple

Septal granulation: not connected

Synapticulae: absent

Archeofungiina Septa: no plane

Trabeculae: present

Synapticulae: present

Fungiina Septa: perforata

Trabecule: discontinuous

Synapticulae: present

Faviina Septa: perforata

Endotheca: rare, absent

Synapticulae: absent

Caryophylliina Septal distal edge: smooth

Endotheca: absent

Trabecule: simple and compound

Dendrophylliina Trabeculae: discontinuous, horizontal and vertical sclerodermites

Figure 1.3 The classification system and evolutionary relationships among Scleractinia proposed by Veron (1995). Branches correspond to families. The width of the branches corresponds to the number of genera in each family. Modified from Veron (2000), Corals of the World, P.31

1.3 The molecular revolution

Towards the end of the 20th Century, the inadequacy of these contrasting scleractinian systematic schemes based on evolutionary insignificant morphological traits was finally confirmed with the emergence of molecular approaches to taxonomy. This led to independent hypotheses on Scleractinia evolution, prompting the production of revisions at every taxonomic level (Chen et al., 2002; Fukami et al., 2004, 2008; Goff-Vitry et al., 2004; Huang et al., 2011; Kerr, 2005; Romano & Cairns, 2000; Romano &

Palumbi, 1996). Molecular phylogenetic reconstructions revealed that extant Scleractinia fall into three major groups, i.e. the Basal, Complex, and

Robust clades, instead of the seven suborders recognized by Veron (1995)

(Fig. 1.4).

A genetic-based system for systematically cataloging metazoan biodiversity opened the era of “molecular taxonomy”. Comparing homologous DNA sequences, it became possible to identify species in many groups of organisms and assign them to higher taxonomic levels

(Tautz et al., 2003). Generally, diverging lineages acquire different mutations through time, eventually creating unique signatures that can be indicative and distinctive among different taxa (de Queiroz, 1998). In this context, Hebert & Cywinska (2003) proposed a DNA barcoding system for animal life relying upon sequence divergence in the mitochondrial cytochrome c oxidase subunit 1 (COI) gene. Mitochondrial DNA (mtDNA)

36 in animal genomes typically shows evolution rates 10 times higher than in nuclear DNA, is non-recombining, and is intronless, (Brown & George, 1979;

Brown et al., 1982). Moreover, as mtDNA is maternally inherited and haploid, it has a smaller effective population size, leading to a much faster lineage sorting (Birky et al., 1983). However, while mtDNA proved very useful for species level delineations in many animal taxa, it does not work well for

Scleractinia. As our knowledge of mtDNA increased, unique properties of mtDNA in different classes of cnidarians were identified. In the class

Anthozoa, in particular, the mtDNA is characterized by slow evolution rates, resulting in very low sequence divergence, that render it unsuitable for inferring genus- and species-level relationships (France & Hoover, 2001,

2002; Hellberg, 2006; Huang et al., 2008; McFadden et al., 2006). Notably,

Shearer & Coffroth (2008) demonstrated that intraspecific COI variation in

Scleractinia, defined as the percentage of single nucleotide mutations in the COI gene among individuals of the same species, is much lower than in most other metazoans even at the third codon position, making it impossible to discern among hard coral species on the basis of the COI gene. Moreover, other mitochondrial genes commonly used in phylogenetic studies in metazoans, such as 16S rDNA, Cytochrome b

(cytb), 12S rDNA, ATP synthase 6 (ATPs6), and NADH dehydrogenase subunits 2, 3, 4, and 6 (NAD2, NAD3, NAD4L, NAD6) are less divergent

37 between families in most anthozoans than between congeneric species in other marine invertebrates (Shearer et al., 2002).

Nonetheless, recent studies identified long highly variable non- coding regions (DNA regions that do not encode for proteins) of the mtDNA containing evolutionary informative sites. The length of these regions is highly variable among different taxa so different regions are informative for different organisms, e.g. the putative control region located between ATP8 and COI and an open reading frame located between ATP6 and NAD4 genes can provide high resolution within the pocilloporid genera

Pocillopora, Seriatopora, and Stylophora (Flot et al., 2011; Flot & Tillier, 2007;

Flot et al., 2008; Pinzón et al., 2013; Schmidt-Roach et al., 2013). Meanwhile, the mitochondrial spacer between COI and 16S rRNA is powerful in resolving species boundaries in the agariciid genera Pavona and Leptoseris

(Luck et al., 2013; Pochon et al., 2015); in the genus Pachyseris (Terraneo et al., 2014); in the merulinid genera Goniastrea, Paragoniastrea, and

Merulina (Huang et al., 2014); and in the genus Sclerophyllia (Arrigoni et al.,

2015).

Cnidarian nuclear DNA (nDNA), on the other hand, accumulates mutations at the same rate as in other animals (Hellberg 2006). The nuclear ribosomal DNA (rDNA) internal transcribed spacers ITS1 and ITS2 (ITS region) are the most commonly used markers to infer species level relationships within corals (Chen et al., 2004; Diekmann et al., 2001; Hunter et al., 1997;

Lopez & Knowlton, 1997; Medina et al., 1999; Odorico & Miller, 1997; Van

Oppen et al., 2002). rDNA constitutes a multigene family of tandem repeated units. Each unit consists of three highly conserved coding regions,

18S, 5.8S, and 28S, and two internal transcribed spacers (ITS1 and ITS2) located in-between. In many eukaryotic taxa, rDNA evolves via concerted evolution (Arnheim et al., 1980), a mechanism that homogenizes different

ITS repeated units through unequal crossing over and gene conversion

(Dover, 1982). In many cases, the rate of concerted evolution is enough to homogenize the variation among unit repeats within species, but interspecific divergence can be high (Hillis & Dixon, 1991).

Despite its widespread usage, the utility of the ITS region has long been debated in the genus Acropora where high sequence divergence

(>40% intraspecific p-distance in Acropora spp.) and the presence of pseudogenes (DNA sequences resembling genes but thought to be non- functional and non-translatable and thus subject to faster mutation rates) obscure patterns of evolution (Chen et al., 2004; Van Oppen et al., 2002;

Vollmer & Palumbi, 2004; Wei et al., 2006). Nevertheless, detailed studies have demonstrated that these features are unique in the genus Acropora, whose genome shows other peculiar features, such as the shortest ITS2 sequence in any metazoans and eukaryotes measured to date (Odorico and Miller, 1997), as well as a peculiar secondary structure, factors that lead to belief that the ITS region of Acropora substantially differs from other

Sceleractinia. The ITS region in fact, has been successfully used in phylogenetic studies for many other coral taxa, as pointed out by

Diekmann et al. (2001), Chen et al. (2004), and Benzoni et al., (2007, 2010).

In the past 20 years, a number of other nuclear genetic sequences have been used to infer coral phylogenies, such as the coding genes

Calmodulin (CalM), ATPase β (ATPs β), β-Tubulin, mini-collagen, Pax-C 46/

47 intron (PCI), and the histone cluster h2ab (Arrigoni et al., 2014; Forsman et al., 2009; Fukami et al., 2004, 2008; Hatta et al., 1999; Márquez et al., 2003;

Van Oppen et al., 2004; Vollmer & Palumbi, 2002; Wallace et al., 2007).

Overall, the most successful markers in terms of utility have bee rDNA and

β-tubulin. On the contrary, histone h2ab sequences have been used only in few species, but its utility may be extendable to other genera.

Molecular phylogenetics has helped clarifying species relationships for several taxa, yet uncertainty persists, and discrepancy among molecular markers is not uncommon. Scleractinian molecular based phylogenies may run into substantial limitations because of the potential for hybridization (Veron, 1995), i.e. the inbreeding of individuals between two different populations, and introgression, i.e. the incorporation of alleles from one species into the gene pool of another species following hybridization and/ or backcrossing (Diekmann et al., 2001; Odorico & Miller, 1997; Van

Oppen et al., 2000, 2001), and incomplete lineage sorting in recent divergent species, i.e. the retention of ancestral polymorphism (Márquez et

40 al., 2003; Queiroz, 2007). With regards to hybridization, several studies have shown that synchronized spawning among numerous species of zooxanthellate corals, i.e. mass spawning sensu Willis et al. (1985), occurs globally (Baird et al., 2009). The simultaneous release of huge quantities of sperms and eggs in a limited period of time creates an opportunity for interspecific hybridization and gene introgression (Márquez et al., 2003;

Miller, 1994; Willis et al., 1997; Willis et al., 2006).

Following in vitro fertilization trials, similar rates of hybridization and within-species fertilization have been recorded for seven species of the

Indo-Pacific genus Platygyra on the Great Barrier Reef (Miller & Babcock,

1997; Willis et al., 1997), and hybridization seems to occur in the genus

Acropora in the Indo-Pacific (Fukami et al., 2003; Van Oppen et al., 2002;

Willis et al., 1997; Wolstenholme, 2004).

Although coral hybridization seems possible to occur in in-vitro

(reviewed by Willis et al., 2006), solid species boundaries should be established a priori any work seeking to hybridize species, and in general, the role of hybridization in promoting adaptive evolution in corals is not yet clear (Wallace & Willis, 1994). Although hybridization has often been the answer to otherwise inconsistent reconstructions, the molecular techniques used in the last decades to test for hybridization cannot discriminate hybridization from introgression and incomplete lineage sorting. In this context, more conservative explanations, such as incorrect field species

41 identifications, should be considered first, and more in-depth genomic analyses added to support hybridization hypotheses.

Recent species divergence can explain low molecular diversity among morphologically distinct entities (Queiroz, 2007). For instance, Miller

& Benzie (1997) proposed that several nominal species of Platygyra were not different from a molecular point of view because they have not undergone sufficient genetic divergence and developed complete reproductive isolation. Finally, ancestral polymorphisms can be hidden by a low rate of molecular evolution, large population sizes, and/or long generation times, as well as short generation times with overlapping of generations (Van Oppen et al., 2004).

Recently, the proliferation of next-generation sequencing techniques has offered the possibility to investigate simultaneously thousands of polymorphic markers in the genome in a relative short amount of time and at a reduced cost. In particular, the development of Restriction site-Associated DNA Sequencing (RADSeq) (Baird et al., 2008) has proven powerful in resolving evolutionary, biogeographical, and phylogenomic questions not answered by traditional molecular markers (Reitzel et al.,

2013). The approach consists of the production of short sequences (30-

500bp) flanking restriction enzymes’ recognition sites, producing hundreds of thousands of loci in the genome and allowing the discovery of high- throughput single nucleotide polymorphisms (SNPs) (Andrews et al., 2016).

On the basis of the number (one or more) of restriction enzymes used and the frequency of the enzymes’ activity, several RADSeq protocols have been developed. Some RADSeq methods produce sequence data from all the cut sites of the restriction enzyme (such as the original RADSeq and

2bRAD), while more recent methods rely on analyzing sequences produced by two enzyme cut sites, separated by a chosen genomic distance (ddRAD or ezRAD for instance) (Andrews et al., 2016). Unlike other reduced-representation sequencing approaches, RADSeq does not necessarily require prior genomic information, and thus has become widely used for non-model organisms (Cruaud et al., 2014; Emerson et al., 2010;

Herrera & Shank, 2016; Herrera et al., 2015; Wagner et al., 2013). With regards to marine invertebrates, RADSeq has been used to disentangle phylogeographic patterns and population genomics of the sea anemone

Nematostella vectensis Stephenson, 1935 (Reitzel et al., 2013), and to successfully delimit species boundaries and reconstruct evolutionary relationships in the deep-sea octocoral genus Chrysogorgia (Pante et al.,

2015). RADSeq analyses have been recently used to test for introgressive hybridization in Pocillopora in the Eastern Pacific (Combosch & Vollmer,

2015), and to trying clarify evolutionary relationships in the coral Acropora

(Rosser et al., 2017), and Porites (Dimond et al., 2017; Forsman et al., 2017).

However, next generation sequencing (NGS) approaches also have their limits. In Hawaiian Porites corals for example, genomic data failed to resolve

43 between the species P. lobata Dana, 1846 and P. compressa Dana, 1846.

The lack of genetic differentiation between these two nominal species again raised the hypothesis towards hybridization and introgression, yet does not explain how under a lack of evident barriers to gene flow, species- level differences can be maintained in sympatry.

Following the unified species concept proposed by de Queiroz

(1998), an integration of different lines of evidence seems necessary to achieve a better understanding of species boundaries and evolutionary relationships in Scleractinia. According to de Queiroz, species can be considered as independent evolving metapopulation lineages (de Queiroz

1998), and the plethora of species concepts designed by evolutionary biologists should instead be regarded as “operational criteria” and integrated to provide different lines of evidence aimed at distinguishing evolutionary lineages. Each species criterion corresponds to a different property that the lineages acquire during their divergence. As lineages diverge, for example, they might acquire phenotypic differences, fixed genetic polymorphisms, or differentiate in their breeding systems or ecology.

Archaeocoenii Fungiinna aFaviin Caryophyllina aDistichophyllin Poritiina aDendrophyllin a

Figure 1.4 Phylogenetic relationships among scleractinian corals based on combined COI and CytB genes. Branch support represent Posterior Bayesian probabilities (>70%), and ML bootstrap values (>50%). Numbers in circles show connections among trees (A to D). Letter codes correspond to traditional coral families. Roman numbers indicate clade subdivisions according to the tree. Colors in the picture refer to coral suborders sensu Veron (1995) as outlined in the legend on the left side of the tree.

1.4 The use of alternative traits to clarify scleratinian evolution

The use of alternative lines of evidence can prove useful in clarifying evolutionary relationships in Scleractinia when morphology and genetic evidence fail or disagree.

Zooxanthellate corals indeed represent a synergistic complex of organisms, comprising the host coral, and all the associated symbiotic microorganisms, such as dinoflagellates, bacteria, Archaea, fungi, protists, and viruses, a system known as the “coral holobiont" (Rohwer et al. 2002,

Rosenberg et al. 2007, Bourne et al. 2016). According to the hologenome theory of evolution, natural selection would act on the organizational level represented by the coral holobiont, with the microbial symbionts playing a fundamental role in the evolution and adaptation of higher organisms

(Rosenberg et al. 2007, Gordon et al. 2013). Integrating information regarding the associations of corals and their symbiotic microorganisms is thus fundamental and might play a crucial role into understanding

Scleractinia evolution and diversity.

Particular interest is growing into better understanding the association of shallow water corals with microscopic endosymbiotic dinoflagellates of the family Symbiodiniaceae. Found in association with a range of marine invertebrates (such as Scleractinia, Zoantharia, or

Cardiida), and providing Sceractinia with up to 95% of their nutritional needs (Kawaguti et al., 1937; Falkowski et al., 1984) these photosynthetic

46 symbionts are crucial for the growth and functioning of coral reefs (Hughes et al., 2017, 2018; Sampayo et al., 2016). Symbiodiniaceae diversity in reef ecosystems is high, with hundres of “types” or “strains” being candidate species, and the specificity and variability of the associations that scleractinian-hosts form with these symbionts have proven to confer ecological advantages to corals under different ecological conditions

(Berkelmans & van Oppen, 2006; LaJeunesse et al., 2010; Rosic et al., 2015;

Hume et al., 2016). These associations influence the geographical zonation patterns of corals at large and small scales and provide the corals with different tolerance to light intensity (Baker, 2001) and temperature (Rowan

& Knowlton, 1995; Glynn et al., 2001; Berkelmans & van Oppen, 2006).

Indeed, different Symbiodiniaceae-host interactions impact the corals susceptibility to bleaching events (for a definition of bleaching see van

Oppen & Lough, 2018). Different patterns of host-symbiont associations have been documented in response to latitudinal, longitudinal, and environmental gradients, for various geographic locations and host taxa

(Oliver & Palumbi, 2009; Huang et al., 2011; Keshavmurthy et al., 2014; Hume et al., 2015, 2016; Ziegler et al., 2015, Tonk et al., 2017). Evidence derived from Symbiodiniaceae associations and different responses to bleaching events have helped clarify boundaries between two morphologically different Pocillopora corals belonging to the P. acuta species complex

(Smith et al., 2017). In fact, although molecular data did not clearly

47 differentiate between the two morphotypes, integrating symbiont data with genomic and morphological data, the authors concluded that the two morphotypes might indeed correspond to two lineages currently undergoing speciation (Smith et al., 2017). Nevertheless, our knowledge of the specificity and diversity of these associations is still poor, limiting our understanding of the ecological benefits that different associations provide

(LaJeunesse et al., 2018).

1.5 The genus Porites

The genus Porites Link, 1807 is ubiquitous on coral reefs in both the Atlantic and Indo-Pacific Oceans, as well as in the Red Sea (Veron, 2000). Colonies of Porites are one of the major building blocks in coral reefs and occur in a wide variety of reef habitats, from back reefs and lagoons, to 30 m deep or exposed walls (Frost, 1977). The genus creates laminar, encrusting and massive structures, up to 10 m high and 11 m wide and hundreds or thousands of years old (Veron, 2000, Takeuchi & Yamashiro, 2017).

Porites is morphologically distinct from other Scleractinia. The small size of its corallites, the high variability of Porites skeletal features both within individual corallum and within species, coupled with high levels of geographic variation within the large range of many species, have challenged coral taxonomists since the 19th Century, and led to the establishment of an endless repetition of ambiguous and repetitive names

(Veron & Pichon, 1982). Species identification in Porites remains one of the most difficult of all Scleractinia (Veron, 2000).

The first morphological description dates back to Pallas (1766), who grouped a number of specimens, now recognized as Porites, under the name Madrepora porites Pallas, 1766; this work was followed by Forskål

(1775), who described Madrepora solida Forskål, 1775 and Madrepora fragilia solida Forskål, 1775 from the Red Sea, (later recognized as Porites solida (Forskål, 1775) and Porites lutea Milne Edwards & Haime, 1851 from

Klunzinger (1879)). The genus Porites was established by Link (1807) with a single species, Porites polymorphus Link, 1807 (now recognized as Porites porites, the type species of the genus) and, as mentioned above, believed to form only branching colonies. No longer after, Lamark (1816) added 16 nominal species (only seven of which were further recognized as Porites) to

P. polymorphus, and provided a more rigorous description: “fixed, branching, lobate, or obtuse, with the free upper surface everywhere covered with the calices, which are regular, subcontiguous, superficial or excavated”. Nevertheless, Lamarck created confusion by including in the genus nominal species belonging to other genera, later removed by De

Blainville (1830). In 1848, Dana grouped the genus Porites with Goniopora de Blainville, 1830, in the family Poritiidae Gray, 1840. In his work, he recognized in Porites corallites the presence of 12 septa, a first circle of 5-6 pali surrounding a central point, and a second circle of granules, frequently

49 forming V-shaped pali. Milne-Edwards and Haime (1851) in their monograph, attempted to comprehensively describe the variety of different forms within Porites, classifying 27 extant and one fossil species according to the growth form, the development of the columella, and the thickness of the walls. Interestingly, it was not until Bernard (1903) that the genus specific pattern of septal fusion (Fig. 1.5) was identified, i.e. a total of

12 septa are arranged in four lateral pairs, a dorsal septum, and a ventral triplet opposed to the dorsal directive. On the basis of this arrangement, plus the number of pali and denticles, and the presence or absence of the columella, Porites species were further discriminated. The work of Bernard represents probably one of the most significant improvements in the taxonomic history of Porites, yet the extreme variability of different forms and structures of the genus forced Bernard to abandon the Linnean system of nomenclature, and rather shift to a classification approach based on geographic location. Talking about all the variety of forms in Porites the author states: “The forms of Porites are indeed like the stars in the heavens, which no man can count, but perhaps even harder to deal with than the stars, for they vary not only in position and magnitude, but also in shape and texture”.

According to the World Register of Marine Species

(https://www.marinespecies.org/, WoRMS 2019), during the taxonomic history of the genus, 190 nominal species have been described, and this

50 number does not consider hundreds more subspecies and forms for which type specimens are deposited in various museums. At present, 71 species are recognized as valid, 81 have been synonymized and 38 are either defined taxon inquirendum or nomen nudum

(https://www.marinespecies.org/, WoRMS 2019). Of these, Veron (2000) recognized a total of 52, yet no mention on the remaining nominal species present in WoRMS was made in Corals of the World (Veron, 2000). For comparison among the recognized species of Porites see Table 1.2.

In the last decade, several studies have attempted to enlighten the intricate taxonomy of Porites using an integration of morphological analyses and molecular techniques. Forsman and co-authors (2009) first evaluated the evolutionary relationships among several nominal species of

Porites from both the Pacific and the Atlantic Oceans using the rDNA ITS region, the mitochondrial COI and a mitochondrial putative control region.

The work highlighted cryptic patterns of species diversity within Porites, and widespread lack of monophyly for both Atlantic and Pacific representatives. Yet the ribosomal reconstruction discriminated the nominal species P. astreoides Lamarck, 1816, P. lichen Dana, 1846, P. bernardi

Vaughan, 1907 and P. colonensis Zlatarski, 1990 , and allowed the description of P. randalli Forsman & Birkeland, 2009 from American Samoa

(Forsman & Birkeland, 2009). Similarly, Benzoni & Stefani (2012) described P. fontanesii Benzoni & Stefani, 2012 from the southern Red Sea using an

51 integration of molecular and corallite level analyses. Using 9 single copy nuclear markers, the ITS region, and a mitochondrial control region, Prada et al. (2014) tested species boundaries among three nominal species of branching Porites in the Caribbean, but none of the genetic analyses supported P. divaricata La Sueur, 1820, P. furcata Lamarck, 1816, and P. porites as distinct entities. Hellberg et al. (2016), Forsman et al. (2017), and

Dimond et al. (2017) have made the most recent contributions to the understanding of the species delineation problem in Porites. Hellberg et al.

(2016) focused on Eastern and Central Pacific populations of P. evermanni

Vaughan, 1907 and P. lobata Dana, 1846, and using five single locus nuclear markers, the ITS region, and COI, the authors revisit species boundaries among 12 Eastern pacific Porites. Forsman et al. (2017) used

NGS techniques, i.e. RADSeq, to disentangle a species complex containing

P. lobata and P. compressa in Hawaii. These two species are generally considered to be morphologically and ecologically distinct however, molecular work suggested that they are not genetically isolated (Forsman et al., 2009). Results from 21 ezRAD libraries failed to resolve the two morphs as distinct entities, identifying more structure among geographic locations than between the putative morpho-species, underlying again the possibility for hybridization among coral species, or the lack of evolutionary meaning in the morphological traits evaluated. Finally, Dimond and co. authors (2017) re-evaluated species boundaries among the three Atlantic

52 branching Porites using RADSeq and detected diversity hidden by traditional genetic approaches.

The above works reveal deep gaps in our understanding of species boundaries and evolutionary relationships in Porites that undermine substantially the designation of conservation statuses and hence the management of biodiversity. The IUCN classifies 26.6% of Porites species as

Vulnerable or worse in the Red List of Threatened Animals (for details regarding the meaning of the categories, the criteria adopted by the IUCN council, and the list of threatened species see https://iucnredlist.org). Such designations are meaningless and conservation efforts seriously compromised if we cannot identify properly species. Clearly, there is an urgent need for integrating different approaches and identify the mechanisms shaping Scleractinia evolution and diversity.

Figure 1.5 Schematic pattern of Porites corallites main skeletal features. Modified from

Forsman (2015

Table 1.2 Porites species recognized by WoRMS Editorial Board (2019) and by Veron (2000).

WoRMS (2019) Veron (200) Porites annae Crossland, 1952 Porites annae Porites aranetai Nemenzo, 1955 Porites aranetai Porites arnaudi Reyes-Bonilla & Porites arnaudi Carricart-Ganivet, 2000 Porites astreoides Lamarck, 1816 Porites astreoides Porites attenuata Nemenzo, 1955 Porites attenuate Porites australiensis Vaughan, 1918 zPorites australiensis Porites baueri Squires, 1959 Porites branneri Rathbun, 1888 Porites brannesi Porites brighami Vaughan, 1907 Porites brighami Porites cocosensis Wells, 1950 Porites cocosensis Porites colonensis Zlatarski, 1990 Porites colonensis Porites columnaris Klunzinger, 1879 Porites columnaris

Porites compressa Dana, 1846 Porites compressa Porites cumulatus Nemenzo, 1955 Porites cumulates Porites cylindrica Dana, 1846 Porites cylindrica Porites decasepta Claereboudt, 2006 Porites deformis Nemenzo, 1955 Porites deformis Porites densa Vaughan, 1918 Porites densa Porites desilveri Veron, 2000 Porites desilveri Porites divaricata Le Sueur, 1820 Porites divaricata Porites duerdeni Porites echinulata Klunzinger, 1879 Porites echinulata Porites eridani Porites evermanni Vaughan, 1907 Porites evermanni Porites exserta Pillai, 1967 Porites flavus Veron, 2000 Porites flavus Porites fontanesii Benzoni & Stefani, 2012 Porites fragosa Dana, 1846 Porites furcata Lamarck, 1816 Porites furcata Porites harrisoni Veron, 2000 Porites harrisoni Porites hawaiiensis Vaughan, 1907 Porites heronensis Veron, 1985 Porites heronensis Porites horizontalata Hoffmeister, Porites horizontalata 1925 Porites latistellata Quelch, 1886 Porites latistella Porites lichen Dana, 1846 Porites lichen Porites lobata Dana, 1846 Porites lobata Porites lutea Milne Edwards & Porites lutea Haime, 1851 Porites mannarensis Pillai, 1967 Porites mayeri Vaughan, 1918 Porites mayeri Porites minicoiensis Pillai, 1967 Porites monticulosa Dana, 1846 Porites monticulosa Porites murrayensis Vaughan, 1918 Porites murrayensis Porites myrmidonensis Veron, 1985 Porites myrmidonensis Porites napopora Veron, 2000 Porites nanopora Porites negrosensis Veron, 1990 Porites negrosensis Porites nigrescens Dana, 1848 Porites nigrescens Porites nodifera Klunzinger, 1879 Porites nodifera Porites okinawensis Veron, 1990 Porites okiwanensis Porites ornata Nemenzo, 1971 Porites ornata Porites panamensis Verrill, 1866 Porites panamensis Porites porites (Pallas, 1766) Porites porites Porites profundus Rehberg, 1892 Porites profundus Porites pukoensis Vaughan, 1907 Porites pukoensis Porites randalli Forsman & Birkeland, 2009 Porites rugosus Fenner & Veron, Porites rugosa 2000 Porites rus (Forskål, 1775) Porites rus Porites sillimaniani Nemenzo, 1976 Porites similaniana Porites solida (Forskål, 1775) Porites solida

Porites somaliensis Gravier, 1910 Porites somaliensis Porites stephensoni Crossland, 1952 Porites stephensoni Porites superfusa Gardiner, 1898 Porites sverdrupi Durham, 1947 Porites sverdrupi Porites tuberculosus Veron, 2000 Porites tuberculosa Porites vaughani Crossland, 1952 Porites vaughani

1.6 Project summary and objectives

The molecular revolution revealed that traditional taxonomy in Scleractinia was flawed at every taxonomic level; this applies also to ecologically dominant genera, such as Porites. The overall aim of this project was to use evidence from different approaches to help determine the evolutionary history of Porites from the seas around the Arabian Peninsula, and provide a framework towards a future revision of the genus and an empirical approach to taxonomy that could be applied to other challenging taxa.

As a general outline, in Chapter 1 I provided a literature review highlighting the existing gap of knowledge in the understanding of the evolution and systematic of Scleractinia, with a focus on the genus Porites. In Chapter 2 I evaluated the diversity of the genus in the Red Sea and the Gulf of Aden and I identified two species endemic to the region, Porites sp. 1 and Porites sp. 2, providing a basic molecular background and integrating micromorphological skeletal examinations of the analyzed species. In

Chapter 3 I reconstructed the complete mitochondrial genomes of two

Porites endemic to the Red Sea, Porites fontanesii Benzoni & Stefani, 2012 and Porites columnaris Klunzinger, 1879, contributing a solid framework for

56 better clarifying the phylogeny and evolution of the genus Porites in future molecular studies. In Chapter 4, I implemented the morpho-molecular results of Chapter 2, and 3 with genome-wide molecular analyses of Porites from the seas around the Arabian Peninsula, in order to provide a comprehensive hypothesis of evolution and biogeography of Porites in the

Arabinan region. Finally, in Chapter 5, I presented evidence derived from the symbiotic association of Porites with dinoflagellates of the family

Symbiodiniaceae, and in the framework of an hologenome theory of evolution, I tested the hypothesis of host-symbiont specificity.

1.7 References

Agapow, P. M., Bininda-Emonds, O. R., Crandall, K. A., Gittleman, J. L., Mace, G. M., Marshall, J. C., & Purvis, A. (2004). The impact of species concept on biodiversity studies. The Quarterly Review of Biology, 79(2), 161- 179. Alloiteau, J. (1957). Contribution à la systématique des Madréporaires fossiles (Vol. 1). Centre National de la Recherche Scientifique.

Andrews, K. R., Good, J. M., Miller, M. R., Luikart, G., & Hohenlohe, P. A. (2016). Harnessing the power of RADseq for ecological and evolutionary genomics. Nature Reviews Genetics, 17(2), 81.

Appeltans, W., Ahyong, S. T., Anderson, G., Angel, M. V., Artois, T., Bailly, N., ... & Błażewicz-Paszkowycz, M. (2012). The magnitude of global marine species diversity. Current Biology, 22(23), 2189-2202.

Arnheim, N., Krystal, M., Schmickel, R., Wilson, G., Ryder, O., & Zimmer, E. (1980). Molecular evidence for genetic exchanges among ribosomal genes on nonhomologous chromosomes in man and apes. Proceedings of the National Academy of Sciences, 77(12), 7323-7327.

Arrigoni, R., Berumen, M. L., Terraneo, T. I., Caragnano, A., Bouwmeester, J., & Benzoni, F. (2015). Forgotten in the taxonomic literature: resurrection of the scleractinian coral genus Sclerophyllia (Scleractinia, Lobophylliidae) from the Arabian Peninsula and its phylogenetic relationships. Systematics and Biodiversity, 13(2), 140-163.

Arrigoni, R., Terraneo, T. I., Galli, P., & Benzoni, F. (2014). Lobophylliidae (Cnidaria, Scleractinia) reshuffled: pervasive non-monophyly at genus level. Molecular Phylogenetics and Evolution, 73, 60-64.

Baird, A. H., Guest, J. R., & Willis, B. L. (2009). Systematic and biogeographical patterns in the reproductive biology of scleractinian corals. Annual Review of Ecology, Evolution, and Systematics, 40, 551-571.

Baird, N. A., Etter, P. D., Atwood, T. S., Currey, M. C., Shiver, A. L., Lewis, Z. A., ... & Johnson, E. A. (2008). Rapid SNP discovery and genetic mapping using sequenced RAD markers. PloS One, 3(10), e3376.

Baker, A. C. (2001). Ecosystems: reef corals bleach to survive change. Nature, 411(6839), 765.

Barbeitos, M. S., Romano, S. L., & Lasker, H. R. (2010). Repeated loss of coloniality and symbiosis in scleractinian corals. Proceedings of the National Academy of Sciences, 107(26), 11877-11882.

Berkelmans, R., & Van Oppen, M. J. (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(1599), 2305-2312.

Bellwood, D. R., Hughes, T. P., Folke, C., & Nyström, M. (2004). Confronting the coral reef crisis. Nature, 429(6994), 827.

Benzoni, F., & Stefani, F. (2012). Porites fontanesii, a new species of hard coral (Scleractinia, Poritidae) from the southern Red Sea, the Gulf of Tadjoura, and the Gulf of Aden. Zootaxa, 3447, 56–68.

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, 160(3), 421-456.

Benzoni, F., Stefani, F., Stolarski, J., Pichon, M., Mitta, G., & Galli, P. (2007). Debating phylogenetic relationships of the scleractinian Psammocora: molecular and morphological evidences. Contributions to Zoology, 76(1), 35-54.

Birky, C. W., Maruyama, T., & Fuerst, P. (1983). An approach to population and evolutionary genetic theory for genes in mitochondria and chloroplasts, and some results. Genetics, 103(3), 513-527.

Bourne, D. G., Morrow, K. M., & Webster, N. S. (2016). Insights into the coral microbiome: underpinning the health and resilience of reef ecosystems. Annual Review of Microbiology, 70, 317-340.

Bowen, B. W., Rocha, L. A., Toonen, R. J., & Karl, S. A. (2013). The origins of tropical marine biodiversity. Trends in Ecology & Evolution, 28(6), 359-366.

Brown, W. M., George, M., & Wilson, A. C. (1979). Rapid evolution of animal mitochondrial DNA. Proceedings of the National Academy of Sciences, 76(4), 1967-1971.

Brown, W. M., Prager, E. M., Wang, A., & Wilson, A. C. (1982). Mitochondrial DNA sequences of primates: tempo and mode of evolution. Journal of Molecular Evolution, 18(4), 225-239.

Bruno, J. F., & Edmunds, P. J. (1997). Clonal variation for phenotypic plasticity in the coral Madracis mirabilis. Ecology, 78(7), 2177-2190.

Budd, A. F. (1993). Variation within and among morphospecies of Montastraea. Courier Forschungsinstitut Senckenberg, 164, 241-254.

Budd, A. F., Romano, S. L., Smith, N. D., & Barbeitos, M. S. (2010). Rethinking the phylogeny of scleractinian corals: a review of morphological and molecular data.

Cairns, S. D. (1999). Species richness of recent Scleractinia. Atoll Research Bulletin, 459(459), 1–46.

Carpenter, K. E., Abrar, M., Aeby, G., Aronson, R. B., Banks, S., Bruckner, A., ... & Edgar, G. J. (2008). One-third of reef-building corals face elevated extinction risk from climate change and local impacts. Science, 321(5888), 560-563.

Chen, C. A., Chang, C., Wei, N. V., Chen, C., Lin, H., Dai, C., & Wallace, C. C. (2004). Secondary structure and phylogenetic utility of the ribosomal Internal Transcribed Spacer 2 (ITS2) in scleractinian corals, 43(4), 759–771. Zoological Studies 43 (4), 759-771.

Chen, C. A., Wallace, C. C., & Wolstenholme, J. (2002). Analysis of the mitochondrial 12S rRNA gene supports a two-clade hypothesis of the evolutionary history of scleractinian corals. Molecular Phylogenetics and Evolution, 23(2), 137-149.

Chevalier, J. P., & Beauvais, L. (1987). Ordre des scléractiniaires. Traité de Zoologie, 3(3), 403-764.

Combosch, D. J., & Vollmer, S. V. (2015). Trans-Pacific RAD-Seq population genomics confirms introgressive hybridization in Eastern Pacific Pocillopora corals. Molecular Phylogenetics and Evolution, 88, 154-162.

Connell, J. H. (1978). Diversity in tropical rain forests and coral reefs. Science, 199 (4335), 1302-1310.

Cruaud, A., Gautier, M., Galan, M., Foucaud, J., Sauné, L., Genson, G., ... & Rasplus, J. Y. (2014). Empirical assessment of RAD sequencing for interspecific phylogeny. Molecular biology and evolution, 31(5), 1272-1274.

Diekmann, O., Bak, R., Stam, W., & Olsen, J. (2001). Molecular genetic evidence for probable reticulate speciation in the coral genus Madracis from a Caribbean fringing reef slope. Marine Biology, 139(2), 221-233.

Dimond, J. L., Gamblewood, S. K., & Roberts, S. B. (2017). Genetic and epigenetic insight into morphospecies in a reef coral. Molecular ecology, 26(19), 5031-5042.

Dover, G. (1982). Molecular drive: a cohesive mode of species evolution. Nature, 299(5879), 111.

Duncan, P. M. (1885). On the structure of the ambulacra of some fossil genera and species of regular Echinoidea. Quarterly Journal of the Geological Society, 41(1-4), 419-453.

Milne-Edwards, H., & Haime, J. (1857). Histoire naturelle des Coralliaires ou Polypes proprement dits (Coraux, Gorgones, Eponges, etc.). Roret.

Emerson, K. J., Merz, C. R., Catchen, J. M., Hohenlohe, P. A., Cresko, W. A., Bradshaw, W. E., & Holzapfel, C. M. (2010). Resolving postglacial phylogeography using high-throughput sequencing. Proceedings of the national academy of sciences, 107(37), 16196-16200.

Esper, E. J. C. (1797). Fortsetzungen der Pflanzenthiere in Abbildugen nach der Natur mit Farben erleuchtet nebst Beschreibungen (Vol. 1). In der Raspeschen Buchhandlung.

Falkowski, P. G., Dubinsky, Z., Muscatine, L., & Porter, J. W. (1984). Light and the bioenergetics of a symbiotic coral. Bioscience, 34(11), 705-709.

Fisher, R., Knowlton, N., Brainard, R. E., & Caley, M. J. (2011). Differences among major taxa in the extent of ecological knowledge across four major. PLoS ONE, 6(11). doi:10.1371/journal.pone.0026556

Flot, J. F., Blanchot, J., Charpy, L., Cruaud, C., Licuanan, W. Y., Nakano, Y., ... & Tillier, S. (2011). Incongruence between morphotypes and genetically delimited species in the coral genus Stylophora: phenotypic plasticity, morphological convergence, morphological stasis or interspecific hybridization?. BMC Ecology, 11(1), 22.

Flot, J. F., Licuanan, W. Y., Nakano, Y., Payri, C., Cruaud, C., & Tillier, S. (2008). Mitochondrial sequences of Seriatopora corals show little agreement with morphology and reveal the duplication of a tRNA gene near the control region. Coral Reefs, 27(4), 789-794.

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(1-2), 80-87.

Forsman, Z. H., & Birkeland, C. (2009). Porites randalli: a new coral species (Scleractinia, Poritidae) from American Samoa. Zootaxa, 2244(1), 51-59.

Forsman, Z. H., Barshis, D. J., Hunter, C. L., & Toonen, R. J. (2009). Shape- shifting corals: molecular markers show morphology is evolutionarily plastic in Porites. BMC Evolutionary Biology, 9(1), 45.

Forsman, Z. H., Knapp, I. S. S., Tisthammer, K., Eaton, D. A. R., Belcaid, M., & Toonen, R. J. (2017). Coral hybridization or phenotypic variation? Genomic data reveal gene flow between Porites lobata and P. compressa. Molecular Phylogenetics and Evolution, 111, 132-148.

Forsskål, P. (1775). Flora Aegyptiaco-Arabica: Sive Descriptiones Plantarum, Quas Per Aegyptum Inferiorem Et Arabiam Felicem; Accedit Tabula Arabiae Felicis Geographico-Botanica. Ex Officina Mölleri, Aulae Typographi. Prostat Apud Heineck Et Faber.

France, S. C. (2001). Analysis of variation in mitochondrial DNA sequences (ND3, ND4L, MSH) among Octocorallia (= Alcyonaria) (Cnidaria: Anthozoa). Bulletin of the Bioogicall Society of Washington, 10, 110-118.

France, S. C., & Hoover, L. L. (2002). DNA sequences of the mitochondrial COI gene have low levels of divergence among deep-sea octocorals (Cnidaria: Anthozoa). Hydrobiologia, 471(1-3), 149-155.

Frost, S. H. (1977). Miocene to Holocene evolution of Caribbean province reef-building corals. In Proceedings of the 3rd International Coral Reef Symposium (Vol. 2, pp. 353-359).

Fukami, H., Budd, A. F., Paulay, G., Solé-Cava, A., Chen, C. A., Iwao, K., & Knowlton, N. (2004). Conventional taxonomy obscures deep divergence between Pacific and Atlantic corals. Nature, 427(6977), 832.

Fukami, H., Chen, C. A., Budd, A. F., Collins, A., Wallace, C., Chuang, Y. Y., ... & Knowlton, N. (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(9), e3222.

Fukami, H., Omori, M., Shimoike, K., Hayashibara, T., & Hatta, M. (2003). Ecological and genetic aspects of reproductive isolation by different spawning times in Acropora corals. Marine Biology, 142(4), 679-684.

Gattuso, J. P., Magnan, A., Billé, R., Cheung, W. W., Howes, E. L., Joos, F., ... & Hoegh-Guldberg, O. (2015). Contrasting futures for ocean and society from different anthropogenic CO2 emissions scenarios. Science, 349(6243), aac4722.

Glynn, P. W., Maté, J. L., Baker, A. C., & Calderón, M. O. (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. Bulletin of Marine Science, 69(1), 79- 109.

Le Goff-Vitry, M. C., Rogers, A. D., & Baglow, D. (2004). A deep-sea slant on the molecular phylogeny of the Scleractinia. Molecular Phylogenetics and Evolution, 30(1), 167-177.

Gordon, J., Knowlton, N., Relman, D. A., Rohwer, F., & Youle, M. (2013). Superorganisms and holobionts. Microbe, 8(4), 152-153.

Harrison, P. L. (1990). Reproduction, dispersal and recruitment of scleractian corals. Coral reefs.

Hatta, M., Fukami, H., Wang, W., Omori, M., Shimoike, K., Hayashibara, T., ... & Sugiyama, T. (1999). Reproductive and genetic evidence for a reticulate evolutionary history of mass-spawning corals. Molecular Biology and Evolution, 16(11), 1607-1613.

Hebert, P. D., Cywinska, A., Ball, S. L., & Dewaard, J. R. (2003). Biological identifications through DNA barcodes. Proceedings of the Royal Society of London. Series B: Biological Sciences, 270(1512), 313-321.

Hellberg, M. E. (2006). No variation and low synonymous substitution rates in coral mtDNA despite high nuclear variation. BMC Evolutionary Biology, 6(1), 24.

Hellberg, M. E., Prada, C., Tan, M. H., Forsman, Z. H., & Baums, I. B. (2016). Getting a grip at the edge: recolonization and introgression in eastern Pacific Porites corals. Journal of Biogeography, 43(11), 2147-2159.

Herrera, S., & Shank, T. M. (2016). RAD sequencing enables unprecedented phylogenetic resolution and objective species delimitation in recalcitrant divergent taxa. Molecular Phylogenetics and Evolution, 100, 70-79.

Herrera, S., Watanabe, H., & Shank, T. M. (2015). Evolutionary and biogeographical patterns of barnacles from deep-sea hydrothermal vents. Molecular Ecology, 24(3), 673-689.

Hillis, D. M., & Dixon, M. T. (1991). Ribosomal DNA: molecular evolution and phylogenetic inference. The Quarterly Review of Biology, 66(4), 411-453.

Hoegh-Guldberg, O. (2015). Reviving the Ocean Economy: the case for action-2015. WWF International. Gland, Switzerland, Geneva.

Huang, D., Benzoni, F., Fukami, H., Knowlton, N., Smith, N. D., & Budd, A. F. (2014). Taxonomic classification of the reef coral families Merulinidae, Montastraeidae, and Diploastraeidae (Cnidaria: Anthozoa: Scleractinia). Zoological Journal of the Linnean Society, 171(2), 277-355.

Huang, D., Licuanan, W. Y., Baird, A. H., & Fukami, H. (2011). Cleaning up the'Bigmessidae': Molecular phylogeny of scleractinian corals from Faviidae, Merulinidae, Pectiniidae and Trachyphylliidae. BMC Evolutionary Biology, 11(1), 37.

Huang, D., Meier, R., Todd, P. A., & Chou, L. M. (2008). Slow mitochondrial COI sequence evolution at the base of the metazoan tree and its implications for DNA barcoding. Journal of Molecular Evolution, 66(2), 167- 174.

Hughes, T. P., Baird, A. H., Bellwood, D. R., Card, M., Connolly, S. R., Folke, C., ... & Lough, J. M. (2003). Climate change, human impacts, and the resilience of coral reefs. Science, 301(5635), 929-933.

Hughes, T. P., Ayre, D., & Connell, J. H. (1992). The evolutionary ecology of corals. Trends in Ecology and Evolution, 7(9), 292–295. doi:10.1016/0169- 5347(92)90225-Z

Hughes, T. P., Kerry, J. T., Álvarez-Noriega, M., Álvarez-Romero, J. G., Anderson, K. D., Baird, A. H., ... & Bridge, T. C. (2017). Global warming and recurrent mass bleaching of corals. Nature, 543(7645), 373.

Hume, B. C., D'Angelo, C., Smith, E. G., Stevens, J. R., Burt, J., & Wiedenmann, J. (2015). Symbiodinium thermophilum sp. nov., a

64 thermotolerant symbiotic alga prevalent in corals of the world's hottest sea, the Persian/Arabian Gulf. Scientific Reports, 5, 8562.

Hume, B. C., Voolstra, C. R., Arif, C., D’Angelo, C., Burt, J. A., Eyal, G., ... & Wiedenmann, J. (2016). Ancestral genetic diversity associated with the rapid spread of stress-tolerant coral symbionts in response to Holocene climate change. Proceedings of the National Academy of Sciences, 113(16), 4416-4421.

Hunter, C. L. (1997). The utility of ITS sequences in assessing relationships among zooxanthellae and corals. In Proc. 8th International Coral Reef Symposium, 1997, 2, pp. 1599-1602).

Karl, S. A., & Bowen, B. W. (1999). Evolutionary significant units versus geopolitical taxonomy: molecular systematics of an endangered sea turtle (genus Chelonia). Conservation Biology, 13(5), 990-999.

Kawaguti, S. (1937) On the Physiology of Reef Corals. I. On the Oxygen Exchanges of Reef Corals, Palao Tropical Bioogical Station Studies, 1, pp. 187–198.

Kerr, A. M. (2005). Molecular and morphological supertree of stony corals (Anthozoa: Scleractinia) using matrix representation parsimony. Biological Reviews, 80(4), 543-558.

Kerr, A. M., Baird, A. H., & Hughes, T. P. (2010). Correlated evolution of sex and reproductive mode in corals (Anthozoa: Scleractinia). Proceedings of the Royal Society B: Biological Sciences, 278(1702), 75-81.

Keshavmurthy, S., Meng, P. J., Wang, J. T., Kuo, C. Y., Yang, S. Y., Hsu, C. M., ... & Chen, C. A. (2014). Can resistant coral-Symbiodinium associations enable coral communities to survive climate change? A study of a site exposed to long-term hot water input. PeerJ, 2, e327.

Kitahara, M. V., Fukami, H., Benzoni, F., & Huang, D. (2016). The new systematics of Scleractinia: integrating molecular and morphological evidence. In The Cnidaria, past, present and future (pp. 41-59). Springer, Cham.

Knowlton, N. (2001). Who are the players on coral reefs and does it matter? The importance of coral taxonomy for coral reef management. Bulletin of Marine Science, 69(2), 305-308.

LaJeunesse, T. C., Pettay, D. T., Sampayo, E. M., Phongsuwan, N., Brown, B., Obura, D. O., ... & Fitt, W. K. (2010). Long-standing environmental conditions, geographic isolation and host–symbiont specificity influence the relative ecological dominance and genetic diversification of coral endosymbionts in the genus Symbiodinium. Journal of Biogeography, 37(5), 785-800.

LaJeunesse, T. C., Parkinson, J. E., Gabrielson, P. W., Jeong, H. J., Reimer, J. D., Voolstra, C. R., & Santos, S. R. (2018). Systematic Revision of Symbiodiniaceae Highlights the Antiquity and Diversity of Coral Endosymbionts. Current Biology, 28(16), 2570-2580.

Lamarck, J. B. D. M. (1801). Systême des animaux sans vertèbres ou tableau général des classes, des ordres et des genres de ces animaux... L'auteur.

Levitan, D. R., Fukami, H., Jara, J., Kline, D., McGovern, T. M., McGhee, K. E., ... & Knowlton, N. (2004). Mechanisms of reproductive isolation among sympatric broadcast-spawning corals of the Montastraea annularis species complex. Evolution, 58(2), 308-323.

Linnaeus, C. V. (1758). Systema naturae, vol. 1. Systema Naturae.

Lopez, J. V., & Knowlton, N. (1997). Description of Montastrea coral sibling species with multiple genetic loci. In Proc. 8th Internarional Coral Reef Symposium (Vol. 2, pp. 1613-1618).

Luck, D. G., Forsman, Z. H., Toonen, R. J., Leicht, S. J., & Kahng, S. E. (2013). Polyphyly and hidden species among Hawaiʻi’s dominant mesophotic coral genera, Leptoseris and Pavona (Scleractinia: Agariciidae). PeerJ, 1, e132.

Mann, C. C., & Plummer, M. L. (1992). The butterfly problem. Atlantic Monthly, 1, 47-59.

Márquez, L. M., Miller, D. J., MacKenzie, J. B., & van Oppen, M. J. (2003). Pseudogenes contribute to the extreme diversity of nuclear ribosomal DNA in the hard coral Acropora. Molecular Biology and Evolution, 20(7), 1077- 1086.

Mayr, E. (1970). Populations, species, and evolution: an abridgment of animal species and evolution (Vol. 19). Harvard University Press.

McFadden, C. S., France, S. C., Sánchez, J. A., & Alderslade, P. (2006). A molecular phylogenetic analysis of the Octocorallia (Cnidaria: Anthozoa)

66 based on mitochondrial protein-coding sequences. Molecular Phylogenetics and Evolution, 41(3), 513-527.

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

Miller, K. J. (1994). Morphological variation in the coral genus Platygyra: environmental influences and taxonomic implications. Marine Ecology- Progress Series, 110, 19-19.

Miller, K., & Babcock, R. (1997). Conflicting morphological and reproductive species boundaries in the coral genus Platygyra. The Biological Bulletin, 192(1), 98-110.

Miller, K. J., & Benzie, J. A. H. (1997). No clear genetic distinction between morphological species within the coral genus Platygyra. Bulletin of Marine Science, 61(3), 907-917.

Moberg, F., & Folke, C. (1999). Ecological goods and services of coral reef ecosystems. Ecological Economics, 29(2), 215-233.

Munday, P. L., Leis, J. M., Lough, J. M., Paris, C. B., Kingsford, M. J., Berumen, M. L., & Lambrechts, J. (2009). Climate change and coral reef connectivity. Coral Reefs, 28(2), 379-395.

Odorico, D. M., & Miller, D. J. (1997). Variation in the ribosomal internal transcribed spacers and 5.8 S rDNA among five species of Acropora (Cnidaria; Scleractinia): patterns of variation consistent with reticulate evolution. Molecular Biology and Evolution, 14(5), 465-473.

Odum, H. T., & Odum, E. P. (1955). Trophic structure and productivity of a windward coral reef community on Eniwetok Atoll. Ecological Monographs, 25(3), 291-320.

Ogilvie, M. M. (1896). III. Microscopic and systematic study of madreporarian types of corals. Philosophical Transactions of the Royal Society of London. Series B, Containing Papers of a Biological Character, (187), 83-345.

Oliver, T. A., & Palumbi, S. R. (2009). Distributions of stress-resistant coral symbionts match environmental patterns at local but not regional scales. Marine Ecology Progress Series, 378, 93-103.

Oppen, M. V., Koolmees, E. M., & Veron, J. E. N. (2004). Patterns of evolution in the scleractinian coral genus Montipora (Acroporidae). Marine Biology, 144(1), 9-18.

Oppen, M. V., Willis, B. L., Van Rheede, T., & Miller, D. J. (2002). Spawning times, reproductive compatibilities and genetic structuring in the Acropora aspera group: evidence for natural hybridization and semi-permeable species boundaries in corals. Molecular Ecology, 11(8), 1363-1376.

Oppen, M. V., Willis, B. L., Vugt, H. V., & Miller, D. J. (2000). Examination of species boundaries in the Acropora cervicornis group (Scleractinia, Cnidaria) using nuclear DNA sequence analyses. Molecular ecology, 9(9), 1363-1373.

Oppen, M. V., & Lough, J. M. (2018). Synthesis: Coral Bleaching: patterns, processes, causes and consequences. In Coral Bleaching (pp. 343-348). Springer, Cham.

Oppen, M. V., Catmull, J., McDonald, B. J., Hislop, N. R., Hagerman, P. J., & Miller, D. J. (2002). The mitochondrial genome of Acropora tenuis (Cnidaria; Scleractinia) contains a large group I intron and a candidate control region. Journal of Molecular Evolution, 55(1), 1-13.

Oppen, M. V., McDonald, B. J., Willis, B., & Miller, D. J. (2001). The evolutionary history of the coral genus Acropora (Scleractinia, Cnidaria) based on a mitochondrial and a nuclear marker: reticulation, incomplete lineage sorting, or morphological convergence?. Molecular Biology and Evolution, 18(7), 1315-1329.

Pallas, P. S. (1766). Elenchus zoophytorum sistens generum adumbrationes generaliores et specierum cognitarum... descriptiones etc. Varrentrapp.

Pante, E., Abdelkrim, J., Viricel, A., Gey, D., France, S. C., Boisselier, M. C., & Samadi, S. (2015). Use of RAD sequencing for delimiting species. Heredity, 114(5), 450. Pinzón, J. H., Sampayo, E., Cox, E., Chauka, L. J., Chen, C. A., Voolstra, C. R., & LaJeunesse, T. C. (2013). Blind to morphology: genetics identifies several widespread ecologically common species and few endemics among Indo-Pacific cauliflower corals (Pocillopora, Scleractinia). Journal of Biogeography, 40(8), 1595-1608.

Pochon, X., Forsman, Z. H., Spalding, H. L., Padilla-Gamiño, J. L., Smith, C. M., & Gates, R. D. (2015). Depth specialization in mesophotic corals

(Leptoseris spp.) and associated algal symbionts in Hawai'i. Royal Society open science, 2(2), 140351.

Prada, C., DeBiasse, M. B., Neigel, J. E., Yednock, B., Stake, J. L., Forsman, Z. H., ... & Hellberg, M. E. (2014). Genetic species delineation among branching Caribbean Porites corals. Coral Reefs, 33(4), 1019-1030.

De Queiroz, K. (1998). The general lineage concept of species, species criteria, and the process of speciation. Endless forms: species and speciation.

De Queiroz, K. (2007). Species concepts and species delimitation. Systematic Biology, 56(6), 879-886.

Randall, R. H. (1976). Some problems in reef coral taxonomy. Micronesica, 15(1), 151-156.

Rowan, R., & Knowlton, N. (1995). Intraspecific diversity and ecological zonation in coral-algal symbiosis. Proceedings of the National Academy of Sciences, 92(7), 2850-2853.

Reitzel, A. M., Herrera, S., Layden, M. J., Martindale, M. Q., & Shank, T. M. (2013). Going where traditional markers have not gone before: utility of and promise for RAD sequencing in marine invertebrate phylogeography and population genomics. Molecular Ecology, 22(11), 2953-2970.

Rohwer, F., Seguritan, V., Azam, F., & Knowlton, N. (2002). Diversity and distribution of coral-associated bacteria. Marine Ecology Progress Series, 243, 1-10.

Romano, S. L., & Cairns, S. D. (2000). Molecular phylogenetic hypotheses for the evolution of scleractinian corals. Bulletin of Marine Science, 67(3), 1043- 1068.

Romano, S. L., & Palumbi, S. R. (1996). Evolution of scleractinian corals inferred from molecular systematics. Science, 271(5249), 640-642.

Rosenberg, E., Koren, O., Reshef, L., Efrony, R., & Zilber-Rosenberg, I. (2007). The role of microorganisms in coral health, disease and evolution. Nature Reviews Microbiology, 5(5), 355.

Rosic, N., Ling, E. Y. S., Chan, C. K. K., Lee, H. C., Kaniewska, P., Edwards, D., ... & Hoegh-Guldberg, O. (2015). Unfolding the secrets of coral–algal symbiosis. The ISME journal, 9(4), 844.

Rosser, N. L., Thomas, L., Stankowski, S., Richards, Z. T., Kennington, W. J., & Johnson, M. S. (2017). Phylogenomics provides new insight into evolutionary relationships and genealogical discordance in the reef-building coral genus Acropora. Proceedings of the Royal Society B: Biological Sciences, 284(1846), 20162182.

Sampayo, E. M., Ridgway, T., Franceschinis, L., Roff, G., Hoegh-Guldberg, O., & Dove, S. (2016). Coral symbioses under prolonged environmental change: living near tolerance range limits. Scientific Reports, 6, 36271.

Schmidt-Roach, S., Lundgren, P., Miller, K. J., Gerlach, G., Noreen, A. M., & Andreakis, N. (2013). Assessing hidden species diversity in the coral Pocillopora damicornis from Eastern Australia. Coral Reefs, 32(1), 161-172.

Shearer, T. L., & Coffroth, M. A. (2008). DNA BARCODING: Barcoding corals: limited by interspecific divergence, not intraspecific variation. Molecular Ecology Resources, 8(2), 247-255.

Shearer, T. L., Van Oppen, M. J. H., Romano, S. L., & Wörheide, G. (2002). Slow mitochondrial DNA sequence evolution in the Anthozoa (Cnidaria). Molecular Ecology, 11(12), 2475-2487.

Smith, H., Epstein, H., & Torda, G. (2017). The molecular basis of differential morphology and bleaching thresholds in two morphs of the coral Pocillopora acuta. Scientific Reports, 7(1), 10066.

Spalding, M. D., & Grenfell, A. M. (1997). New estimates of global and regional coral reef areas. Coral Reefs, 16(4), 225-230.

Stolarski, J., Kitahara, M. V., Miller, D. J., Cairns, S. D., Mazur, M., & Meibom, A. (2011). The ancient evolutionary origins of Scleractinia revealed by azooxanthellate corals. BMC Evolutionary Biology, 11(1), 316.

Sultan, S. E. (2000). Phenotypic plasticity for plant development, function and life history. Trends in Plant Science, 5(12), 537-542.

Takeuchi, I., & Yamashiro, H. (2017). Large Porites microatoll found by aerial survey at Sesoko Island, Okinawa, Japan. Coral Reefs, 36(4), 1317-1317. Tautz, D., Arctander, P., Minelli, A., Thomas, R. H., & Vogler, A. P. (2003). A plea for DNA taxonomy. Trends in Ecology & Evolution, 18(2), 70-74.

Terraneo, T. I., Berumen, M. L., Arrigoni, R., Waheed, Z., Bouwmeester, J., Caragnano, A., ... & Benzoni, F. (2014). Pachyseris inattesa sp. n. (Cnidaria,

Anthozoa, Scleractinia): a new reef coral species from the Red Sea and its phylogenetic relationships. ZooKeys, (433), 1.

Todd, P. A. (2008). Morphological plasticity in scleractinian corals. Biological Reviews, 83(3), 315-337.

Tonk, L., Sampayo, E. M., Chai, A., Schrameyer, V., & Hoegh-Guldberg, O. (2017). Symbiodinium (Dinophyceae) community patterns in invertebrate hosts from inshore marginal reefs of the southern Great Barrier Reef, Australia. Journal of Phycology, 53(3), 589-600.

Vaughan, T. W., & Wells, J. W. (1943). Revision of the suborders, families and genera of Scleractinia. Geological Society of America special Papers, 44, 1–363.

Veron, J. E. N. (1995). Corals in space and time: the biogeography and evolution of the Scleractinia. Cornell University Press.

Veron, J. E. N. (2000). Corals of the World (No. C/593.6 V4).

Veron, J. E. N. (1982). Scleractinia of eastern Australia, Part IV, Family Poritidae. Australian Institute of Marine Science, Monograph Series, 5, 159p.

Vollmer, S. V., & Palumbi, S. R. (2004). Testing the utility of internally transcribed spacer sequences in coral phylogenetics. Molecular Ecology, 13(9), 2763-2772.

Vollmer, S. V., & Palumbi, S. R. (2002). Hybridization and the evolution of reef coral diversity. Science, 296(5575), 2023-2025.

Wagner, C. E., Keller, I., Wittwer, S., Selz, O. M., Mwaiko, S., Greuter, L., ... & Seehausen, O. (2013). Genome-wide RAD sequence data provide unprecedented resolution of species boundaries and relationships in the L ake V ictoria cichlid adaptive radiation. Molecular Ecology, 22(3), 787-798.

Wallace, C. C., Chen, C. A., Fukami, H., & Muir, P. R. (2007). Recognition of separate genera within Acropora based on new morphological, reproductive and genetic evidence from Acropora togianensis, and elevation of the subgenus Isopora Studer, 1878 to genus (Scleractinia: Astrocoeniidae; Acroporidae). Coral Reefs, 26(2), 231-239.

Wallace, C. C., & Willis, B. L. (1994). Systematics of the coral genus Acropora: implications of new biological findings for species concepts. Annual Review of Ecology and Systematics, 25(1), 237-262.

Wei, N. V., Wallace, C. C., Dai, C., Ruby, K., Pillay, M., & Chen, C. A. (2006). Analyses of the ribosomal internal transcribed spacers (ITS) and the 5.8 S gene indicate that extremely high rDNA heterogeneity is a unique feature in the scleractinian coral genus Acropora (Scleractinia; Acroporidae). Zoological Studies 45, no. 3 (2006): 404-418.

Wells, J. W. (1933). Corals of the Cretaceous of the Atlantic and Gulf coastal plains and western interior of the United States (Vol. 18, No. 67). Harris co.

Wells, J. W. (1956). Scleractinia, p. F328–F444. Treatise on Invertebrate Paleontology, Pt. F, Coelenterata. Geological Society of America and University of Kansas Press, Lawrence.

Willis, B. L., Babcock, R. C., Harrison, P. L., Oliver, J. K., & Wallace, C. C. (1985). Patterns in the mass spawning of corals on the Great Barrier Reef from 1981 to 1984. 343-348.

Willis, B. L., Babcock, R. C., Harrison, P. L., & Wallace, C. C. (1997). Experimental hybridization and breeding incompatibilities within the mating systems of mass spawning reef corals. Coral Reefs, 16(1), S53-S65.

Willis, B. L., van Oppen, M. J., Miller, D. J., Vollmer, S. V., & Ayre, D. J. (2006). The role of hybridization in the evolution of reef corals. Annual Revie of Ecology, Evolution, and Systematics, 37, 489-517.

Wolstenholme, J. K. (2004). Temporal reproductive isolation and gametic compatibility are evolutionary mechanisms in the Acropora humilis species group (Cnidaria; Scleractinia). Marine Biology, 144(3), 567-582.

Ziegler, M., Roder, C. M., Büchel, C., & Voolstra, C. R. (2015). Mesophotic coral depth acclimatization is a function of host-specific symbiont physiology. Frontiers in Marine Science, 2, 4.

2. CHAPTER 2: MORPHOLOGY AND MOLECULES REVEAL TWO UNKNOWN SPECIES OF PORITES (SCLERACTINIA, PORITIDAE) FROM THE RED SEA AND THE GULF OF ADEN

Tullia I. Terraneoa,b*, Francesca Benzonic, Andrew H. Bairdb, Roberto

Arrigonia & Michael L. Berumena aRed Sea Research Center, Division of Biological and Environmental

Science and Engineering, King Abdullah University of Science and

Technology, Thuwal 23955–6900, Saudi Arabia

bARC Centre of Excellence for Coral Reef Studies, James Cook University,

Townsville 4811, QLD, Australia cDepartment of Biotechnologies and Bioscience, University of Milano -

Bicocca, Milan, 20126, Italy

Published in Systematics and Biodiversity

Terraneo, T.I., Benzoni, F., Baird, A.H., Arrigoni, R., Berumen, M.L. (2019). Morphology and molecules reveal two new species of Porites (Scleractinia, Poritidae) from the Red Sea and the Gulf of Aden. Systematics and Biodiversity.

* None of the results in the following chapter are intended towards a formal description of new taxa. The ICZN description of new taxa will result from a peer reviewed publication and not from this dissertation.

This chapter is edited from the original version of the paper.

2.1 Abstract

Two unknown reef coral species Porites sp. 1 and Porites sp. 2 (Scleractinia,

Poritidae) are identified from the Red Sea and the Gulf of Aden. Porites sp.

1 only occurs in the Farasan Island in the southern Red Sea, while P sp. 2 has been collected in the Yemen Hadramaut region in the Gulf of Aden. Both species presented striking in situ differences with respect to other Porites species, and were characterized by small encrusting colonies and unusual polyp colouration. In order to test the genetic distinctiveness of P. sp. 1 and

P. sp. 2 between each other and with respect to other representatives in the genus Porites, we investigated their evolutionary relationships with eight other morphologies of Porites occurring in the Red Sea and in the Gulf of

Aden. Two DNA loci, the mitochondrial putative control region and the nuclear ribosomal ITS region, were sequenced, and three species delimitation approaches based on barcoding threshold (Automated

Barcoding Gap Discovery) and coalescence theory (Poisson–Tree Process,

Generalized Mixed Yule Coalescent) were applied. Phylogenetic and species delimitation analyses were overall concordant, resolving P. sp. 1 and P. sp. 2 as two divergent but closely related lineages. Of the other morphologically defined Porites species, three were genetically differentiated, but five were genetically indistinguishable. The discovery of two regional endemics confirms the importance of the Red Sea and the

Gulf of Aden as regions of high biodiversity and suggest the need for an

74 integration of genome–wide molecular data with the re–evaluation of skeletal structures in the systematics of Porites.

Keywords: Biodiversity, Cnidaria, CR, Integrative taxonomy, rDNA, Species delimitation

2.2 Introduction

The coral genus Porites Link, 1807 (Cnidaria, Scleractinia) is ubiquitous on coral reefs in the Atlantic and the Indo–Pacific (Veron, 2000), and provides the bulk of the structural framework of coral reefs in a variety of different habitats, from exposed walls to shallow–water lagoons (Frost, 1977).

Porites can form encrusting, laminar, columnar, branching, or massive colonies, and is characterized by highly perforated walls derived from complex patterns of growth and fusion of trabeculae and synapticulae (Bernard, 1903). Corallites are small, normally ranging from 0.7 to 2 mm, and their arrangement is generally species–specific. In certain taxa, a well developed coenosteum separates the corallites, but in others corallites have fused walls and no coenosteum is formed. The typically perforated septa are formed by a regular pattern of fusing trabeculae. The innermost trabecula can be distinct from the rest of the septal structure and is referred to as a palus (Chevalier & Beauvais, 1987; Vaughan & Wells,

1943), literally a vertical pole. Porites is distinct from the other scleractinian

75 corals due to the presence of a peculiar pattern of septal fusion (Bernard,

1903), in which 12 septa are arranged into four couples of lateral pairs, a ventral triplet, and an opposite dorsal directive (Bernard, 1903: P. 13 Fig. 1;

Veron & Pichon, 1982: P. 11 Fig. 2). Porites species have been described based on corallite structures, particularly, the pattern of fusion of the ventral triplet, the presence of the columella (or lack thereof), and the number of pali and granules. Because of its small corallites, along with high morphological plasticity and geographic variability, the genus remains one of the most challenging in terms of species identification and delimitation

(Forsman et al., 2009a, 2015).

Overall, 190 nominal species of Porites have been described; of these, 71 are currently regarded as valid, 81 have been synonymized, and

38 are either defined taxon inquirendum or nomen nudum (WoRMS: http://www.marinespecies.org/), thus suggesting the need for detailed revisions or redescription, respectively.

In the last decade, several studies have addressed the intricate systematics of Porites using an integration of morphological analyses and molecular techniques. Forsman et al. (2009a) evaluated the evolutionary relationships among 22 putative species of Porites from the Pacific and the

Atlantic Oceans using the nuclear ribosomal ITS region (ITS), the mitochondrial COI gene, and the mitochondrial putative control region

(CR). They found genetic divergence among many congeneric species

76 consistent with morphospecies designations. Nevertheless, three Pacific clades were polyphyletic. These three clades were comprised of branching and massive morphologies that were genetically indistinguishable. Such a scenario could have resulted from recent speciation and incomplete lineage sorting, hybridization and introgression, or high levels of morphological variability among these lineages. Moreover, specimens matching the description of Porites lutea Milne Edwards and Haime, 1851 were split into three genetically divergent groups, showing evidence of cryptic divergence. Hellberg et al., (2016) found evidence of introgressive hybridization between Porites lobata Dana, 1846 and Porites evermanni

Vaughan, 1907 in the eastern Pacific, although none was found in Hawaiian

Porites (Forsman et al., 2017). The above works indicate a deep gap in our understanding of species boundaries and evolutionary relationships in

Porites. This can undermine our knowledge of species distribution ranges, can result in over–splitting of single evolutionary lineages, or can hide patterns of cryptic divergence. Any of these errors can substantially hinder appropriately designating the conservation status for the species in the genus.

The Red Sea and the Gulf of Aden regions were the sites of some of the first taxonomic research on Scleractinia. Several naturalists in the 18th and 19th centuries described coral species from type localities in the southern the Red Sea and Gulf of Aden (Forskål 1775; Lamarck 1816;

Ehrenberg 1834; Klunzinger 1879a; 1879b), including a number of Porites species (Forskål 1775; Esper 1797; Edwards & Haime 1860; Klunzinger 1879a,

1879b). Since then, however, coral research in the region (especially outside of the Gulf of Aqaba) has been limited until very recently, and there is still a significant gap in our knowledge of coral diversity and systematics from this area (Berumen et al., 2013; Berumen et al., 2019).

A total of 26 species of Porites have been reported from the combined area of the Red Sea and the Gulf of Aden on the basis of traditional morphology–based classifications (Sheer & Pillai, 1983; Sheppard

& Sheppard, 1991; Pichon et al., 2010; Veron, 2000). More recently, Porites decasepta Claereboudt, 2006 was described based on colony growth form and corallite morphology from southern Oman. The only morpho– molecular study of Porites from the Arabian region is from Benzoni & Stefani

(2012), which led to the description of Porites fontanesii Benzoni and Stefani,

2012 based on some samples from Yemen. Overall, the published works suggest that Porites biodiversity in the region might be higher than previously thought, and additional integrated morpho–molecular studies are needed for a better understanding of the diversity and the phylogeny of this genus. In fact, during surveys of coral diversity in the Red Sea and the

Gulf of Aden, two undescribed morphospecies of Porites were observed and collected. In the present study, we further analysed their morphology at the colony and corallite level, and tested their phylogenetic position with

78 respect to eight species of Porites from the Red Sea, the Gulf of Aden, and

Socotra to investigate their distinctiveness from other representatives of the genus. Using two molecular loci (CR and ITS) and three species delimitation approaches, we demonstrated that P. sp. 1 from the southern Red Sea and

P. sp. 2 from the Gulf of Aden are indeed two undescribed species endemic to the Red Sea and the Gulf of Aden. Micromorphological analyses of the corallites showed that these undescribed species have unique features supported by the genetic results. However, five of the other Porites morphospecies included in our analyses resolved into a single evolutionary lineage. Based on our results we suggest the integration of genome–wide studies and external lines of evidence (e.g., reproductive studies) in future works to reach a better understanding of the evolution and biogeography of Porites.

2.3 Materials and methods

2.3.1 Sampling

A total of eight Porites colonies were sampled at seven localities in the Gulf of Aden and Socotra between 2007 and 2010 within the framework of the

Total E&P – Creocean – University of Milano - Bicocca (UNIMIB) ‘Yemen

Scleractinia Biodiversity project’ (Fig. 2.1, Appendix 2.1). Twenty–eight

Porites specimens were collected between 2013 and 2017 at 14 sites along the Saudi Arabian Red Sea coast, ranging from the Gulf of Aqaba to the

Farasan Islands (Fig. 2.1, see Appendix 2.1). Sampling in Saudi Arabia occurred within the context of the biodiversity cruises carried out by the

King Abdullah University of Science and Technology (KAUST). Prior to sampling, an underwater image of each colony, a close-up of the living coral polyps if extended at daytime, and a close-up of the colony surface were taken using a Canon G11 or G15 in an Ikelite housing. Sampling was carried out while SCUBA diving with hammer and chisel. At the surface, each colony fragment was tagged and a small sample of living tissue preserved in 99% ethanol for DNA extractions. The rest of the specimens were bleached in sodium hypochlorite, rinsed in fresh water, and air dried for 24 hours for subsequent morphological examinations.

Figure 2.1 Map of the sampling localities in the Red Sea (Saudi Arabia) and north-western

Gulf of Aden (Yemen) of the Porites spp. material examined in this study.

2.3.2 Abbreviations

MNHN: Muséum National d’histoire Naturelle, Paris, France

MTQ: Museum of Tropical Queensland, Townsville, QLD, Australia

NHM: Natural History Museum London, UK

USNM: National Museum of Natural History, Washington DC, USA

Morphological analyses

Macro- and micro-morphological characters of each colony were examined using light microscopy and Scanning Electron Microscopy (SEM).

Scanning electron microscopy pictures were taken at KAUST with 600 FEG

SEM and at UNIMIB using a Vega Tescan SEM. Prior to imaging, specimen fragments were ground, mounted on stabs with silver glue, and sputter– coated with conductive gold film (UNIMIB) or Au–Pd (KAUST). Samples were morphologically identified following Klunzinger (1879a, 1879b), Sheer & Pillai

(1983), Bernard (1903), Sheppard & Sheppard (1991), Veron (2000),

Claereboudt (2006), Pichon et al., (2010), and Benzoni & Stefani (2012). The following skeletal features were considered for each coral sample: corallite diameter, wall thickness, fusion pattern of the ventral triplet, number of pali, number of denticles, presence or absence of the columella, and presence or absence of the coenosteum (see Budd et al., 2012 for glossary).

Measurements were done with the ImageJ (https://imagej.nih.gov/ij/)

Analyse tool from SEM images with scale (10 replicate/corallites for each specimen).

2.3.3 DNA extractions, amplification and sequence analyses

Coral DNA was extracted from tissues preserved in ethanol using the commercial DNeasy® Blood and Tissue kit (Qiagen Inc., Hilden, Germany) following the manufacturer’s protocol. Extracted DNA quantity and quality were assessed using a NanoDrop® 2000C spectrophotometer (Thermo

Fisher Scientific, Wilimington, USA). One mitochondrial Control Region (CR) and the nuclear ribosomal ITS region (ITS) were amplified using polymerase

82 chain reaction. The CR locus was amplified using the primers zpsRNAf (5’–

AGCAGACGCGGTGAAACTTA–3’) and zpCOX3r (5’–

GCCCAAGTAACAGTACCCCC–3’) (Terraneo et al., accepted a), while the

ITS locus was amplified using primers ITSz1 (5'–

TAAAAGTCGTAACAAGGTTTCCGTA–3') and ITSz2 (5'–

CCTCCGCTTATTGATATGCTTAAAT–3') (Forsman et al., 2009). All PCR products were purified by adding 1.5 µl Illustra ExoStar (GE Heathcare,

Buckinghamshire, UK) and incubation at 37˚ C for 60 min, followed by 85 ˚

C for 15 min, and were directly sequenced in both forward and reverse directions using an ABI 3730xl DNA Analyzer (Applied Biosystems, Carlsbad,

USA). Forward and reverse sequences were assembled and edited using

Sequencher 5.3 (Gene Codes Corp., Ann Arbor, USA). A total of 18 nuclear sequences showed intra–individual polymorphisms and were phased using

Phase (Stephens, Smith, & Donnelly, 2001, available online at http://stephenslab.uchicago.edu/ software.html) and SeqPHASE (Flot,

2010, available online at http://seqphase.mpg.de/seqphase/) when the alleles showed the same length, and using Champuru (Flot, 2007, available online at http://seqphase.mpg. de/champuru/) when the two predominant alleles were of different lengths. Alignments were performed using the E–INS–i option in MAFFT 7.130b (Katoh & Standley, 2013) and manually checked using BioEdit 7.2.5 (Hall, 1999). All new sequences generated with this study were deposited in EMBL (Appendix 2.1).

2.3.4 Phylogenetic relationships reconstruction

For each marker, the best fit substitution model was calculated using

PartitionFinder 1.1.1 (Lanfear et al., 2012) with unlinked branch lengths, the greedy search algorithm for nucleotide sequence, and a partitioning scheme comparison was performed using the corrected Akaike

Information Criterion (AIC) and the Bayesian Information Criterion (BIC).

PartitionFinder selected the evolutionary model GTR+I for CR and the

GTR+I+G for ITS. Phylogenetic relationships were reconstructed under two different criteria: Bayesian inference (BI) using MrBayes 3.2.6 (Ronquist et al.,

2012) and Maximum Likelihood (ML) using RAxML 2 (Stamatakis, 2014). The

CIPRES server (Miller et al., 2012) was used to conduct both the BI and ML analyses. BI runs were performed using four Markov Chain Monte Carlo

(MCMC) chains for 10 million generations, saving one tree every 100 generations. The tree searches were stopped when all parameters reached stationarity for effective sampling size and unimodal posterior distribution using Tracer 1.6 (Rambaut et al., 2014). The first 25% trees sampled were discarded as burn–in following indications by Tracer. ML topologies were obtained under the default parameters shown on the CIPRES server with a multiparametric bootstrap analyses of 1,000 bootstrap replicates. Intra– and inter–specific pairwise distances were calculated as p–distance with

MEGA 6 (Tamura et al., 2007).

2.3.5 Species delimitation analyses

Three independent species delimitation approaches were used to determine putative molecular species clusters in our dataset: Automatic

Barcoding Gap Discovery (ABGD) (Puillandre et al., 2012), Poisson–Tree processes (PTP) (Zhang et al., 2013), and generalized mixed Yule coalescent (GMYC) (Pons et al., 2006; Fujisawa & Barraclough, 2013). The

ABDG was used to determine groups of candidate species on the basis of pairwise genetic distances. All the alignments were imported in MEGA 6.0

(Tamura et al., 2013) to compute matrices of pairwise genetic distances using the Kimura 2–parameter (K2P) model. These matrices were further used as input files on the ABGD web–page

(http://wwwabi.snv.jussieu.fr/public/abgd/abgdweb.html) and a range of different settings were tested, showing concordant species delimitations results. Therefore, we set parameters as follows: Pmin = 0.001, Pmax = 0.1,

Steps = 10, X = 1.5, and Nb bins = 20. The Poisson–Tree process required a

Maximum Likelihood (ML) gene tree as input, thus ML trees obtained from the previous phylogenetic analyses were used to verify the robustness of the internal branches (Guindon & Gascuel, 2003). We ran PTP analyses for

400,000 MCMC generations with thinning value = 100 and burn–in = 0.25.

We visually confirmed the convergence of the MCMC chain as recommended by Zhang et al. (2013). The GMYC approach combines models of stochastic lineage growth (Yule models) with coalescence

85 theory (Pons et al., 2006; Fujisawa & Barraclough, 2013). The method requires an inferred ultrametric gene tree as input that was obtained with

BEAST 1.8.2 (Drummond et al., 2012) after removing identical haplotypes. A coalescent tree prior and the heterogeneity of the mutation rate across lineages were set under an uncorrelated lognormal relaxed clock. Analysis was run for 50 million generations, with a sampling frequency of 1,000. After checking adequate mixing and convergence of all runs with Tracer 1.6

(Rambaut et al., 2014), the first 25% of trees were discarded as burn–in and the maximum clade credibility tree was computed using TreeAnnotator

1.8.2 (Drummond et al., 2012). The resulting ultrametric tree was imported into R 3.1.3 (R Core Team, 2013) and GMYC analyses were conducted using the Splits (Ezard et al., 2009) and Ape (Paradis et al., 2004) libraries.

2.4 Results

2.4.1 Red Sea and Gulf of Aden Porites identification

Based on our morphological examinations, the 36 colonies collected were assigned to eight nominal Porites species: Porites annae Crossland, 1952,

Porites echinulata Klunzinger, 1879, P. fontanesii, Porites columnaris

Klunzinger, 1879, P. lobata, P. lutea, Porites rus Forskål, 1775, and Porites solida Forskål, 1775. Two morphologies did not match any original

86 descriptions of Porites and did not correspond to any type specimen examined, thus the morphology of P. sp. 1 and P. sp. 2 is detailed here after.

2.4.2 Taxonomy

Order Scleractinia Bourne, 1900

Family Poritidae Gray, 1840

Genus Porites Link, 1807

Porites sp. 1

(Fig. 2.2 a–d, 2.5)

REFERENCE SPECIMEN: MNHN–IK–2012–14238 Marka, Farasan Banks, Saudi

Arabia (18.220550 N, 41.324383 E), coll. F. Benzoni, 03/06/2013 (Fig. 2.2 a, c,

Fig. 2.5 a, c).

OTHER MATERIAL: KAUST SA0172 and KAUST SA0180 Marka, Farasan Banks,

Saudi Arabia (18.220550, 41.324383), coll. F. Benzoni, 03/06/2013 (Fig. 2.2 a,

Fig. 2.5 b, and q, respectively); KAUST SA1516 Aminah, Farasan Islands, Red

Sea, Saudi Arabia (17.110333, 42/067533), coll. F. Benzoni, 22/10/2014 (Fig.

2.2 c).

TYPE LOCALITY: Marka, Farasan Banks, Saudi Arabia.

MORPHOLOGICAL DETAILS: The reference specimen is a thinly encrusting colony originally partially attached to the substrate and forming a free margin on one side (Fig. 2.2 b). It has an irregularly circular outline and it is

5.8 cm long, 4.5 cm wide and 1.5 cm thick at its thickest point. The surface is smooth at the margins which were originally free, and small low–lying nodules form towards the opposite margin which was attached to the substrate in vivo. Corallites are flush with the coenosteum, or slightly sunken

(Fig. 2.5 a–b). They are separated by coenosteum which is well developed but never abundant and adjacent corallites are usually less than one corallite diameter apart (Fig. 2.6 a–b). The coenosteum is finely ornamented with spines arranged in clumps at the top of each trabecula

(Fig. 2.6 c), the clumps being 0.09 to 0.11 mm in diameter. Calice diameter ranges from 1 to 1.06 mm. The typical septa arrangement of the genus

Porites is observed with 12 septa arranged in four lateral pairs, one dorsal septum and three ventral septa (Fig. 2.6 d, 2.8 a). The ventral triplet is fused and the dorsal septum is shorter than other septa. Six obvious pali are present at the inner ends of the septa, four in front of the four lateral pairs, one in front of the ventral triplet, and one in front of the short dorsal septum

(Fig. 2.6 d, 2.8 a). The palus in front of the dorsal septum is smaller and shorter than the others. In some corallites it is not formed. The palus of the fused ventral triplet is the largest and has a triangular to three lobed outline (Fig.

2.6 d, 2.8 a). Pali of the lateral pairs have a typical heart–shaped outline

(Fig. 2.6 d, 2.8 a). The ventral and lateral pali are markedly exert from the colony surface (orange arrows in Fig. 2.6 c). Overall, the radius of the crown of pali is equal or sub–equal to the radius of the corallite. This is a main distinguishing feature of this species which gives the centre of its corallites a star–like appearance visible also to the naked eye in all the examined coralla (Fig. 2.5 a–c). Upper septal margins carry 1 large, often flattened and fan–shaped, and finely ornamented denticle (Fig. 2.2 a). A columella is present (Fig. 2.6 c, indicate by the orange arrows, d, Fig. 2.8 a) and it is typically smaller than the pali and shorter. The radii connecting the pali and the columella are visible in most corallites although irregularly developed.

The reference specimen is the largest colony observed and collected of this species, the rest of the existing specimens are smaller (Fig.

2.5 a–b). The rest of the examined material for this species has a similar growth form and outline to that of the holotype, with a tendency in some to form more pronounced knobs (Fig. 2.2 c). Overall, there is relatively little morphological variability of the corallite features among the examined specimens.

FIELD CHARACTERISTCS AND HABITAT: Living colonies of P. sp. 1 are green or greenish–brown, with the polyp tissue being typically darker than the coenosarc which is pale green or grey (Fig. 2.2 b–c). As the five larger and exert pali are usually visible through the semi–transparent polyp tissue

89 above them, their presence can also be appreciated in vivo (inset in Fig.

2.2 c). This species is rather inconspicuous due to its small colony size (Fig.

2.2 a). It was only observed in shallow (1–5 m depth) and protected reef environments.

DISTRIBUTION: Although the authors conducted an extensive sampling of

Porites all along the Saudi Arabia Red Sea coast (Terraneo et al., accepted a), this species was only observed in the Farasan Banks and Farasan Islands.

Until it is reported elsewhere, this species is thus considered a Red Sea endemic.

REMARKS: Specimen KAUST SA1516 was included in a study of the diversity and distribution in the Red Sea of the Symbiodiniaceae in Porites (Terraneo et al., accepted a). Results showed that the specimen hosted an exclusive lineage of symbionts, not present in any other of the 79 colonies analysed, and remarking the distinctness of P. sp. 1 from other congeners.

Porites sp. 2

(Fig. 2.2 e–h, 2.7)

REFERENCE SPECIMEN: MNHN–IK–2016–208 south Burum, Yemen (14.307017

N, 48.967800 E), coll. F. Benzoni, 19/03/2009, collection code UNIMIB BU070.

TYPE LOCALITY: South of the Burum Headland, Yemen.

MORPHOLOGICAL DETAILS: The reference specimen is a small (2 cm long,

1 cm wide and 4 mm thick) fragment of a very thinly encrusting colony which formed a veneer on the hard substrate (Fig. 2.2 e–h). The colony had a roughly circular outline, it was approximately 5 cm in diameter and had a smooth surface (Fig. 2.2 g). Polyps were arranged in short series in parts of the colony (Fig. 2.2 h). Corallites are separated by coenosteum and flush with its surface (Fig. 2.7 a–b). Coenosteum is well developed but never abundant and adjacent corallites are less than one corallite diameter apart (Fig. 2.7 a–b). The corallum coenosteum is finely ornamented with spines arranged in clumps at the top of each trabecula (Fig. 2.7 c–d) the clumps being 0.07 to 0.08 mm in diameter. Calices are small, their diameter ranging from 0.7 to 0.85 mm. The 12 septa are arranged in four lateral pairs, one dorsal septum and three ventral septa (Fig. 2.7 a–b, Fig. 2.8 b). The ventral triplet is fused and the dorsal septum is much shorter than other septa, sometimes reduced to a single denticle. Five pali are present at the inner ends of the septa, four in front of the four lateral pairs and one in front of the ventral triplet (Fig. 2.7 c–d, Fig. 2.8 b). The palus of the fused ventral triplet is slightly larger than the others (Fig. 2.7 c–d, Fig. 2.8 b). The pali are not exert from the colony surface. Upper septal margins carry 1 denticle which has the same size of the coenosteum ornamentations (Fig. 2.7 a–d).

A columella is absent (Fig. 2.7 a –d, Fig. 2.8 b). In some corallites, a

91 synapticula connects the pali of the dorsal lateral pairs (Fig. 2.7 c, indicated by the white arrow, Fig. 2.8 b) or the palus of the triplet with one of a lateral pair (Fig. 2.7 b). The other colonies of this species observed in situ had a similar size and shape to that of the holotype colony (Fig. 2.2 e–f).

FIELD CHARACTERISTCS AND HABITAT: Living colonies of P. sp. 2 have a distinctive colouration in vivo: the tissue around the polyp mouth is red, while the tissue above the septa is brown, and the coenosarc is white (Fig.

2.2 e–h). This species was only observed between 3 and 5 m depth.

DISTRIBUTION: One of the authors (FB) was part of several field trips along the Yemeni Gulf of Aden coast (Benzoni et al., 2012b), but this species was only observed around the Burum headland. Until it is reported elsewhere, this species is thus considered a north–west Gulf of Aden endemic.

REMARKS: Despite the attempts to sample a larger portion of it, the colony from which the reference specimen was taken crumbled in small pieces of which the examined fragment was the largest. Unfortunately, the same happened with the other observed colonies with the same morphology and colouration at different diving sites around the type locality (Fig. 2.2 e– f). The available fragment is currently gold coated and attached to a SEM stub.

Figure 2.2 Porites sp. 1 (a‒d) and Porites sp. 2 (c‒h) in situ: a, colony of KAUST SA0172 (in the dashed circle), Marka, Farasan Banks, Saudi Arabia (same as Fig. 10 b); b, colony of MNHN-

IK-2012-14238, Marka, Farasan Banks, Saudi Arabia (same as Fig. 10 a, c); c, a colony at

Aminah, Farasan Islands, Saudi Arabia, the inset shows a detail of the retracted polyps

93 through which five obvious pali (white arrows) can be observed; d, close up of the surface of the colony in b showing the typical beige-green coenosarc and brown polyp colouration; e‒f, colonies at Burum, Yemen; g, colony of MNHN-IK-2012-14238 (same as

Figs. 12 a‒d), south Burum, Yemen; h, close-up of the same colony as g showing the typical white coenosarc and red polyp colouration.

Figure 2.3 In situ images of the Porites colonies examined in this study, circular insets shows polyp detail for each colony: a, P. fontanesii, KAUST SA0438, Qita al Kirsh, Saudi Arabia; b,

P. columnaris, UNIMIB P0009, Balhaf, Yemen; c, P. rus, UNIMIB BA0109, Bir Ali, Yemen; d, P. lobata, KAUST SA0054, Farasan Banks, Saudi Arabia; e, P. lutea, UNIMIB BA0098, Sikha

Island, Yemen; f, P. echinulata, UNIMIB AD0007, Aden, Yemen; g, P. compressa, UNIMIB

BA0025, Bir Ali, Yemen; h, P. solida, UNIMIB P0011, Balhaf, Yemen.

2.4.3 Molecular results

Sequence data for the CR and ITS regions were obtained from all the 36 analysed samples (Appendix 2.1). The CR sequence of Porites porites

(Pallas, 1766) was obtained from the complete mitochondrial genome of the species deposited in GenBank

(https://www.ncbi.nlm.nih.gov/genbank/) (accession number

DQ643837.1), added to the mitochondrial dataset and used as an outgroup. Similarly, the ITS region of P. porites was downloaded from

GenBank (accession number KY867643.1) and used as outgroup in the nuclear DNA reconstruction. The CR sequence alignment consisted of 1,287 bp, with 37 polymorphic and 30 parsimony informative sites. The ITS alignment encompassed 754 bp, with 129 variable sites, 83 of which were parsimony informative. BI and ML tree topologies obtained from the two loci were congruent, recovering six highly supported molecular clades

(clades I to VI) (Fig. 2.4 a–b). Five clades were comprised of samples belonging to a single morphospecies, and their monophyly was highly supported: clade I included P. fontanesii, clade II grouped P. columnaris samples, clade III was comprised of samples of P. sp. 1, clade IV grouped samples belonging to P. rus morphology, and clade VI comprised the

96 genetic material of the holotype of P. sp. 2. All the examined specimens identified morphologically as P. annae, P. echinulata, P. lobata, P. lutea, and P. solida grouped in clade V. Mean intra– and interspecific genetic distances are summarised in Table 2.1 and Table 2.2. In particular, CR had a mean value of 0.69 % (± 0.2 standard deviation) p–distance and a p– distance of 0.73 % (± 0.2 s.d.) separated clade III and VI from the other clades, respectively. Among the other clades (I, II, IV, V) the mean p– distance was 0.79% (± 0.22 s.d.) with the smallest p–distance being 0.23% (±

0.14 s.d.) between clades II–IV and the largest p–distance being 1.74% (±

0.36 s.d.) between clades I–V. For the ITS region, a mean p–distance values of 5.5 % (± 0.91 s.d.) separated clade III from the other clades and a mean p–distance value of 5.1 % (± 0.87 s.d.) separated clade VI from the other lineages. Mean p–distance among the other clades (I, II, IV, V) was 5.65%

(± 0.85 s.d.). The smallest p–distance was 2.45% (± 0.57), between clade IV–

V, and the largest was 9.78% (± 0.12), between clades I–II.

Results from the three species delimitation analyses, i.e., ABGD, PTP, and GMYC, are summarised in Fig. 2.4 b and d. The three approaches identified partitions congruent with the phylogenetic topologies from both the CR and the ITS region. A total of six putative partitions were identified among the Porites samples. There was strong agreement between the species delimitation results and the phylogenetic clustering, recovering the following lineages: P. fontanesii (clade I), P. columnaris (clade II), P. sp. 1

(clade III), P. rus (clade IV), P. annae, P. echinulata, P. lobata, P. lutea, and

P. solida group (clade V), and P. sp. 2 (clade VI). The ABGD and PTP identified an extra clustering within clade V, encompassing samples belonging to P. annae, P. solida, and P. lobata. This split was recovered for the mitochondrial locus, but was not confirmed for the nuclear marker. The

ABGD results were consistent based on both Kimura and Jukes–Cantor distances. With a Pmax value = 0.1 the analyses showed for the seven CR partitions, and for the six ITS region putative lineages.

Figure 2.4 RAxML phylogenetic reconstructions of Red Sea and Yemen Porites at two molecular loci. a, mitochondrial Control Region, b, nuclear Internal Transcribed Spaces region. Number on the branches represent support values corresponding to Bayesian posterior probabilities (>90%), ML bootstrap values (>70%). Porites porites was selected as outgroup in both the analyses. In c, black curved lines connect the two sequences of heterozygotes individuals. Summary of putative species delimitation drawn by ABGD, PTP, and GMYC, (b, d).

Table 2.1 Genetic distance values (p-distance) within and between Porites species from the Red Sea based on the mitochondrial CR. Interspecific pairwise comparisons of genetic distance are in bold; intraspecific genetic distances are indicated along the diagonal; standard deviations are reported in brackets or on the upper right-hand portions for each set of comparisons; n.c. not calculable

Clade I Clade II Clade III Clade IV Clade V Clade VI

Clade I 0.0005 0.0035 0.0036 0.0037 0.0036 0.0036 (0.0005) Clade II 0.0154 0 0.0015 0.0014 0.0016 0.0019

Clade III 0.0162 0.0039 0 0.0018 0.0018 0.0017

Clade IV 0.0162 0.0023 0.0047 0 0.0018 0.0021

Clade V 0.0174 0.0038 0.0061 0.0046 0.0011 0.0022 (0.0005) Clade VI 0.0162 0.0047 0.0039 0.0055 0.0069 n.c.

100

Table 2.2 Genetic distance values (p-distance) within and between Porites species from the Red Sea based on the nuclear ITS region. Interspecific pairwise comparisons of genetic distance are in bold; intraspecific genetic distances are indicated along the diagonal; standard deviations are reported in brackets or on the upper right-hand portions for each set of comparisons; n.c. not calculable

Clade I Clade II Clade III Clade IV Clade V Clade VI

Clade I 0.0093 0.0128 0.0134 0.0122 0.0121 0.0122 (0.0026) Clade II 0.0978 0.0067 0.0087 0.0077 0.0062 0.0077 (0.0022) Clade III 0.1047 0.0499 0 0.0073 0.0084 0.0073

Clade IV 0.0909 0.0251 0.0398 0.0058 0.0057 0.0072 (0.0021) Clade V 0.0907 0.0302 0.0437 0.0245 0.0054 0.0081 (0.0015) Clade VI 0.1502 0.0448 0.0309 0.0362 0.0448 n.c.

2.5 Discussion

2.5.1 Comparison of morphological structures

Porites sp. 1 and P. sp. 2 are morphologically and molecularly distinct from other Porites species. Both species form small encrusting colonies and occur in shaded and turbid habitats, characteristics that might explain why they remained overlooked by previous investigations (Fig. 2.2 a–b). At the corallite level, both species present the classic pattern of septal fusion that distinguishes Porites from other Scleractinia, yet they are morphologically

101 different from each other and the other species analysed in this study for their unique combination of diagnostic characters (Fig. 2.9 a–h).

Porites sp. 1 calices range from 1 to 1.06 mm in diameter, comparable to P. columnaris, P. lutea, and P. echinulata (ranging from 1 to

1.5 mm) (Fig. 2.9 b, e, f). In comparison, P.s lobata, P. annae, and P. solida all show larger calices (ranging from 1.5 to 2 mm) (Fig. 2.9 d, g, h), while P. sp. 2, P. fontanesii and P. rus have smaller calices (ranging from 0.5 to 0.9 mm) (Fig. 2.7 c–d and Fig. 2.7 a, c). In P. sp. 1 the columella is always present

(Fig. 2.6 c–d), a feature shared with seven analysed congeners (P. columnaris – Fig. 2.9 b, P. lobata – Fig. 2.9 d, P. lutea – Fig. 2.9 e, P. echinulata

– Fig. 2.9 f, P. annae – Fig. 2.9 g, P. solida – Fig. 2.9 h) but not with P. sp. 2

(Fig. 2.7 c–d), P. fontanesii, and P. rus (Fig. 2.9 a, c). A compact and ornamented coenosteum is distinctive of P. sp. 1 (Fig. 2.6 a–b), a feature generally uncommon in Porites (Veron and Pichon, 1982) but also well developed in P. sp. 2 (Fig. 2.7 a–b) and other regional endemics like P. fontanesii and P. decasepta (Benzoni & Stefani, 2012, Clarebaudt, 2006).

Among the other species analysed in this study, P. rus also presents a well– developed coenosteum, but this species is readily told apart from the others based on the arrangement of the coenosteum into ridges. Both P. sp. 1 and P. sp. 2 have a fused ventral triplet, as P. fontanesii, P. columanris,

P. rus, P. lutea, and P. echinulata (Fig. 2.9 a, b, c, e, f), while in P. lobata, P. annae, and P. solida the three ventral septa were free (Fig. 2.9 d, g, h). The

102 most striking distinguishing feature of P. sp. 1 are five thick, exert and highly ornamented heart–shaped pali which are visible to the naked eye both on living and dry colonies, bigger than the examined congeners (Fig. 2.2 c, Fig.

2.6 a–d). Porites sp. 2 also has five pali, yet considerably thinner, not exert and less ornamented. The remaining species examined have six to eight smaller pali (Fig. 2.9 b, c, d, g), with the exception of P. fontanesii and P. lutea, and P. echinulata that present five–six pali, and P. solida that do not form pali (Fig. 2.9 a, e, f, h). Nevertheless, the characteristic heart shape of the lateral pali in P. sp. 1 (Fig. 2.6 d, Fig. 2.8 a) is unique to this species.

In vivo, P. sp. 2 might resemble the recently described P. decasepta

Claereboudt, 2006, so far only reported from Oman. The diagnostic features of P. decasepta include a bright blue colouration in the field and the presence of a modified pattern of septal fusion, in which the ventral triplet is reduced to a ventral septum facing the dorsal directive. Moreover, the septa of P. decasepta are connected to a short columella by radii. Similar to P. sp. 2, also P. decasepta presents a developed coenosteum, but SEM examination of P. sp. 2. show a different arrangement of corallite structures, with a clearly visible ventral triplet, absence of columella, and the characteristic synapticula connecting the pali of the dorsal lateral pairs.

2.5.2 Molecular considerations

103

From a molecular point of view, P. sp. 1 and P. sp. 2 are clustered by both the mitochondrial CR and the ITS region into two distinct molecular clades

(clade III, clade VI, Fig. 2.4 a and c), and the three species delimitation analyses corroborate their genetic distinctness (Fig. 2.4 b and d).

Phylogenetically, the two species are closer to each other than to the other clades (Fig. 2.4 a and c), yet their genetic distances are congruent with the interspecific genetic distances among the other Porites representatives included in our analyses (Table 2.1, Table 2.2).

2.5.3 Species complex in Porites

In this work, the phylogenetic positions of P. sp. 1 and P. sp. 2 within the genus were investigated with reference to other eight putative species of

Porites co–occurring in the Red Sea and the Gulf of Aden (Fig. 2.4 a and c).

Phylogenetic analyses were integrated with species delimitation approaches (Fig. 2.4 b and d) and skeletal SEM examinations of all species

(Fig. 2.9 a–h) to further corroborate the position of P. sp. 1 and P. sp. 2, and to investigate the distinctness among all the molecular identified clades.

Our genetic analyses corroborated the traditional morphological distinction of five lineages of Porites (P. fontanesii – clade I, P. columnaris – clade II, P. rus – clade IV, P. sp. 1 – clade III and P. sp. 2 – clade VI), yet the boundaries among five other species (P. annae, P. echinulata, P. lobata, P.

104 lutea, and P. solida – clade V) were molecularly indistinguishable based on the examined markers (Fig. 2.4 a–d)).

Clade V clustered samples with highly variable morphologies at both the colony and the corallite level. The majority of the specimens shared a massive colony shape (P. lobata and P. lutea representatives, Fig. 2.3 d–e), but others formed encrusting, knobby, or submassive colonies (P. echinulata, P. annae, and P. solida, Fig. 2.3 f–h). All the specimens presented different corallite morphologies derived from different combinations of diagnostic characters (Fig. 2.9 d–h). The corallite size in the clade was highly variable, ranging from 1 to 1.5 mm in P. lutea, P. echinulata, and P. annae (Fig. 2.9 e–g), up to 2 mm in P. lobata, and P. solida (Fig. 2.9 d, h). All the morphologies clustered within this clade showed a columella (Fig. 2.9 d–h), yet in P. annae might be weakly developed (Fig.

2.9 g). The presence of coenosteum is an uncommon feature in Porites, and indeed it was weakly developed in all the specimens clustered in clade V, with the exception of P. echinulata, where a developed granulated coenosteum separated corallites. The number of pali was very inconsistent among morphologies, ranging from no pali in P. solida (Fig. 2.9 h), 5–6 pali in P. lutea and P. echinulata (Fig. 2.9 e and f), to 8 pali P. lobata and P. annae (Fig. 2.9 d and g). Similarly, the triplet of ventral septa was free in P. lobata, P. annae and P. solida (Fig. 2.9 e, g, h), while forming a trident in P. lutea and P. echinulata (Fig. 2.9 e and f).

105

The outcome of our morphological observations leads us to question the resolution offered by the chosen molecular markers, and/or the examined skeletal structures and, ultimately, their value as diagnostic characters. Similar results already emerged in Forsman et al. (2009), where some Porites species were clearly resolved (for examples P. rus, Porites astreoides Lamarck, 1816, P. lichen Dana, 1846, and P. hawaiiensis

Vaughan, 1907), while some others were not molecularly differentiated, although showing very different morphologies (for example P. lobata, P. lutea, and P. cylindrica Dana, 1846). The recovery of multiple species not genetically differentiated may also be a result of hybridization, incomplete lineage sorting, or high levels of phenotypic plasticity (Veron, 1995, Todd,

2008). Further studies encompassing genome–wide sequencing or alternative lines of evidence aimed at better defining species boundaries are necessary to clarify relationships within this clade. Genome–wide approaches have proven to be useful in clarifying species boundaries within a species complex of Atlantic Porites. In fact, using nine single copy nuclear markers, the ITS region, and the mitochondrial CR, Prada et al.

(2014) tested species boundaries among three nominal species of branching Porites in the Caribbean but none of the genetic analyses supported P. porites, P. divaricata Le Sueur, 1821, and P. furcata Lamarck,

1816 as distinct entities. More recently, Dimond and co-authors (Dimond et al., 2017) re–evaluated species boundaries among these three Atlantic

106 branching Porites using a genome–wide sequencing approach (RADSeq), showing that this genomic technique was able to detect genetic diversity hidden by traditional sequencing approaches (i.e., Sanger sequencing).

2.5.4 Species biogeography and distribution

From a biogeographical perspective, the two unknown Porites present disjunct but adjacent distributions. Porites sp. 1 is only known from the southern Red Sea, i.e., the southern Farasan Banks and Farasan Islands and absent in the Gulf of Aden, while P. sp. 2 only occurs in the northwest Gulf of Aden (Fig. 2.1). Thus, the two species are here considered a Red Sea endemic, and a Gulf of Aden endemic, respectively. High levels of endemism have been previously reported for several marine taxa in the

Red Sea. For example, 12.6% of reef fishes, 12.6% of polychaetes, and 8.1% of echinoderms are endemic to this region (DiBattista et al., 2016). Estimates of hard coral endemics keep rising, with the most recent value being 6.4%

(Berumen et al., 2019). These estimates are higher than the values reported for other hotspots of endemism around the Arabian Peninsula, such as the

Arabian Gulf and the Gulf of Oman (DiBattista et al., 2016). Among the Red

Sea endemics, some species occur throughout the Red Sea (such as

Pachyseris inattesa Benzoni and Terraneo, 2013 and Sclerophyllia margariticola Klunzinger, 1879), while others do not extend south of the

Farasan Banks into the Farasan Islands, such as Acropora squarrosa

107

Ehrenberg, 1834 and Cyphastrea magna Benzoni and Arrigoni, 2017

(Berumen et al., 2019). These breaks in distribution might relate the environmental break occurring around 20º N latitude, between the Farasan

Banks and the Farasan Islands, where shallow and nutrient rich reefs systems replace the reefs found in the more oligotrophic northern Red Sea, where reefs typically have steep slopes into deeper water. For scleractinian corals,

23 Red Sea endemics occur throughout the Red Sea while 14 are only known from the central and northern parts (Berumen et al., 2019) Hence, P. sp. 1 is to date the only known coral endemic to the southern Red Sea.

Further biodiversity surveys are needed to verify its distribution. For example, until an extensive survey of coral diversity was recently conducted along the Saudi Arabia Red Sea coast, P. fontanesii was thought to be restricted to the southern Red Sea and Gulf of Aden (Benzoni & Stefani 2012).

However, the species is now known to occur throughout the Red Sea including the Gulf of Aqaba (Benzoni pers. comm; Berumen et al., 2019;

Terraneo et al., accepted a). This being said, during the exact same survey all along the Saudi Arabia Red Sea coast, P. sp. 1 was only observed in the south.

Porites sp. 2. was never observed or sampled in the Red Sea and, so far, it has only been reported from the Hadramaut region in Yemen, an area in the Gulf of Aden that has received much attention in terms of corals biodiversity in the past ten years (see for example Arrigoni et al., 2012,

108

Arrigoni et al., 2017; Arrigoni et al., 2018; Benzoni & Stefani 2012; Benzoni et al. 2012a, b; Pichon et al., 2010; Kitano et al. 2014). From an environmental point of view, corals in the Gulf of Aden are subject to extreme temperature and nutrient levels variations resulting from the monsoon driven seasonal upwelling (Pichon et al., 2010; Benzoni et al., 2012a), a factor that might limit the persistence of only well adapted species in this region.

The allopatric distribution ranges of these two species, their phylogenetic relationships, and their morphological distinctiveness indicate that P. sp. 1 and P. sp. 2 are potential vicariant species. Indeed, similar phylogeographic patterns among independent species suggest vicariant histories potentially related to periodic environmental changes.

Comparable patterns have been already reported in the region, for example, for the lobophylliids Sclerophyllia margariticola and S. maxima

Sheppard and Salm, 1988. These two species present sister phylogenetic relationships and distributions comparable to P. sp. 1 and P. sp. 2, with S. margariticola only present in the Red Sea, and S. maxima distributed in the seas around the Arabian Peninsula, with the exception of the Red Sea

(Arrigoni et al., 2015). Some works have demonstrated the importance of

Arabian Peninsula geological events in shaping the distributions and genetic clustering of fish and coral taxa (DiBattista et al., 2013, 2016;

Keshavmurthy et al., 2013). Evidence suggests that during the Pleistocene ice ages, the Red Sea connection to the Gulf of Aden was repeatedly

109 limited, isolating the Red Sea from the rest of the Indian Ocean (Rohling et al., 2009). This could have imposed selection pressure or a barrier to gene flow, resulting in genetic differentiation within a continuously distributed taxon (Saunders et al., 1986). Further sampling and a reconstruction of a dated phylogeny of Porites from this region would be necessary to test these ideas.

2.6 Conclusions

The results of this chapter provide a basic morphological and molecular background to understanding the diversity of Porites in the Red Sea and

Gulf of Aden. In particular, this work indicates that coral biodiversity in the region needs to be re–evaluated in light of a rapidly changing taxonomic framework. Indeed, finding of species unknown to taxonomy and expansion of known ranges for other species from these regions are providing a new understanding of the regional biogeography. For example, rates of endemism appear higher than previously thought, suggesting a gap of knowledge from these peripheral regions of the world’s oceans. The current study of Porites illustrates that fundamental understanding of species diversity is limited in even a key group of reef- building corals, whereas our knowledge of less-conspicuous or rare groups is probably even less complete. Finally, this work highlights the compelling need to integrate genome-wide analyses and samples from further

110 localities around the Arabian Peninsula to clarify Porites evolution in the region.

Following these considerations, in the next chapter I use next generation sequencing to provide the complete mitochondrial genomes of two Porites species endemic to the Red Sea and the Gulf of Aden, Porites fontanesii and Porites columnaris.

111

Figure 2.5 Morphology of the designated reference specimens of Porites sp. 1.: a, MNHN-

IK-2012-14238 (same specimen as in Fig. 3), on the left, and KAUST SA0180 on the right; b,

KAUST SA0172 (same specimen as in Fig. 2); c, detail of the corallites on the corallum surface of the reference specimen. Scale bars represent: Fig. a‒b, 1 cm; Fig. c, 1 mm.

112

Figure 2.6 SEM images of a fragment of the reference specimen of Porites sp. 1. MNHN-IK-

2012-14238 (same specimen as in Fig 7 b, 9 d, 10 b): a and b, corallites on different parts of the corallum surface separated by well-formed coenosteum; c, side view of two adjacent corallites showing the five exert pali (orange arrows) surrounding the columella sitting deeper in the fossa (light blue arrows); d, a corallite showing the typical septa arrangement and large pali size. Scale bars represent: a‒b, 1 mm; c‒d, 500 µm.

113

Figure 2.7 SEM images of one of the two fragments of the reference specimen of Porites sp. 2. MNHN-IK-2016-208 (same specimen as in Fig. 8): a and b, corallites on different parts of the corallum surface separated by coenosteum; c, a corallite showing the typical septal arrangement with a short dorsal directive septum and a well-formed radius uniting the two pali found in correspondence of the dorsal lateral pairs (white arrow); d, another corallite in which the radius uniting the two dorsal pali is not formed. Scale bars represent: a‒b, 1 mm; c‒d, 500 µm.

114

Figure 2.8 Schematic representation of the corallite structures observed in the reference specimens of: a, Porites sp. 1.; b, Porites sp. 2 Dashed lines indicate structures that are well formed in some corallites but absent in others within the same corallum

115

116

Figure 2.9 SEM images of the corallites of the Porites species examined in this study: a, P. fontanesii, KAUST SA0438 (same as Fig. 8 a); b, P. columnaris, UNIMIB P0009 (same as Fig. 8 b); c, P. rus, UNIMIB BA0109 (same as Fig. 18 c); d, P. lobata, KAUST SA0054 (same as Fig. 8 d); e, P. lutea, UNIMIB BA0098 (same as Fig. 8 e); f, P. echinulata, UNIMIB AD0007 (same as

Fig. 8 f); g, P. annae, UNIMIB BA0025 (same as Fig. 8 g); h, P. solida, UNIMIB P0011 (same as

Fig. 8 h).

117

2.7 References

Arrigoni, R., Stefani, F., Pichon, M., Galli, P., & Benzoni, F. (2012). Molecular phylogeny of the robust clade (Faviidae, Mussidae, Merulinidae, and Pectiniidae): an Indian Ocean perspective. Molecular Phylogenetics and Evolution, 65(1), 183–193.

Arrigoni, R., Vacherie, B., Benzoni, F., Stefani, F., Karsenti, E., Jaillon, O., ... Barbe, V. (2017). A new sequence data set of SSU rRNA gene for Scleractinia and its phylogenetic and ecological applications. Molecular Ecology Resources, 17(5), 1054–1071.

Benzoni, F., Arrigoni, R., Stefani, F., Reijnen, B. T., Montano, S., & Hoeksema, B. W. (2012a). Phylogenetic position and taxonomy of Cycloseris explanulata and C. wellsi (Scleractinia: Fungiidae): lost mushroom corals find their way home. Contributions to Zoology, 81(3).

Benzoni, F., Pichon, M., Dutrieux, E., Chaîneau, C. H., Abdulaziz, M., & Al– Thary, I. (2012b). The scleractinian fauna of Yemen: diversity and species distribution patterns. 1-6.

Bernard, H.M. (1903). The family Poritidae, I. The genus Goniopora. Catalogue of the madreporarian corals in the British Museum. Natural History, 4, 1–166.

Berumen, M.L., Arrigoni, R., Bouwmeester, J., Terraneo, T., I., & Benzoni, F. (2019). Biodiversity of Red Sea corals. Chapter 7 in Coral Reefs of the Red Sea (Voolstra, C.R., and Berumen, M.L. (eds)). Springer.

118

Berumen, M. L., Hoey, A. S., Bass, W. H., Bouwmeester, J., Catania, D., Cochran, J. E., ... Saenz–Agudelo, P. (2013). The status of coral reef ecology research in the Red Sea. Coral Reefs, 32(3), 737–748.

Bowen, B. W., Rocha, L. A., Toonen, R. J., & Karl, S. A. (2013). The origins of tropical marine biodiversity. Trends in Ecology & Evolution, 28(6), 359–366.

Budd, A. F., Fukami, H., Smith, N. D., & Knowlton, N. (2012). Taxonomic classification of the reef coral family Mussidae (Cnidaria: Anthozoa: Scleractinia). Zoological Journal of the Linnean Society, 166(3), 465–529.

Claereboudt, M. R. (2006). Reef corals and coral reefs of the Gulf of Oman. Historical Association of Oman, Oman. 344: pp.

DiBattista, J. D., Berumen, M. L., Gaither, M. R., Rocha, L. A., Eble, J. A., Choat, J. H., ... Bowen, B. W. (2013). After continents divide: comparative phylogeography of reef fishes from the Red Sea and Indian Ocean. Journal of Biogeography, 40(6), 1170–1181.

DiBattista, J. D., Roberts, M. B., Bouwmeester, J., Bowen, B. W., Coker, D. J., Lozano-Cortés, D. F., ... Kochzius, M. (2016). A review of contemporary patterns of endemism for shallow water reef fauna in the Red Sea. Journal of Biogeography, 43(3), 423–439. DiBattista, J. D., Wilcox, C., Craig, M. T., Rocha, L. A., & Bowen, B. W. (2011). Phylogeography of the Pacific Blueline Surgeonfish, Acanthurus nigroris, reveals high genetic connectivity and a cryptic endemic species in the Hawaiian Archipelago. Journal of Marine Biology.

Dimond, J. L., Gamblewood, S. K., & Roberts, S. B. (2017). Genetic and epigenetic insight into morphospecies in a reef coral. Molecular Ecology, 26(19), 5031–5042.

Drummond, A. J., Suchard, M. A., Xie, D., & Rambaut, A. (2012). Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and Evolution, 29(8), 1969–1973.

Eble, J. A., Toonen, R. J., Sorenson, L., Basch, L. V., Papastamatiou, Y. P., & Bowen, B. W. (2011). Escaping paradise: larval export from Hawaii in an Indo–Pacific reef fish, the yellow tang Zebrasoma flavescens. Marine Ecology Progress Series, 428, 245–258.

Ehrenberg C.G. (1834) Beitrage zur physiologischen Kenntniss der Corallenthiere im Allgemeinen und besunders des Rothen Meeres, nebst einem Versuche zur physiologischen Systematik der– selben.

119

Abhandlungen der Koniglichen̈ Akademie der Wissenschaften, Berlin 1832: 225–380.

Ezard, T., Fujisawa, T., & Barraclough, T. G. (2009). Splits: species’ limits by threshold statistics. R package version, 1(11), r29.

Flot, J. F. (2007). Champuru 1.0: a computer software for unraveling mixtures of two DNA sequences of unequal lengths. Molecular Ecology Notes, 7(6), 974–977

Flot, J. F. (2010). SeqPHASE: a web tool for interconverting PHASE input/output files and FASTA sequence alignments. Molecular Ecology Resources, 10(1), 162–166.

Forsskål, P. (1775). Descriptiones animalium avium: amphibiorum, piscium, insectorum, vermium; quae in itinere orientali observavit Petrus Forskål... Post mortem auctoris edidit.

Forsman, Z. H., Barshis, D. J., Hunter, C. L., & Toonen, R. J. (2009). Shape– shifting corals: molecular markers show morphology is evolutionarily plastic in Porites. BMC Evolutionary Biology, 9(1), 45.

Forsman, Z., Wellington, G. M., Fox, G. E., & Toonen, R. J. (2015). Clues to unraveling the coral species problem: distinguishing species from geographic variation in Porites across the Pacific with molecular markers and microskeletal traits. PeerJ, 3, e751.

Frost, S. H. (1977). Miocene to Holocene evolution of Caribbean province reef–building corals. In Proceedings of the 3rd International Coral Reef Symposium (Vol. 2, pp. 353–359).

Fujisawa, T., & Barraclough, T. G. (2013). Delimiting species using single– locus data and the Generalized Mixed Yule Coalescent approach: a revised method and evaluation on simulated data sets. Systematic Biology, 62(5), 707–724.

Gaither, M. R., Bowen, B. W., Bordenave, T. R., Rocha, L. A., Newman, S. J., Gomez, J. A., ... Craig, M. T. (2011). Phylogeography of the reef fish Cephalopholis argus (Epinephelidae) indicates Pleistocene isolation across

120 the Indo–Pacific Barrier with contemporary overlap in the Coral Triangle. BMC Evolutionary Biology, 11(1), 189.

Gaither, M. R., Toonen, R. J., Robertson, D. R., Planes, S., & Bowen, B. W. (2010). Genetic evaluation of marine biogeographical barriers: perspectives from two widespread Indo-Pacific snappers (Lutjanus kasmira and Lutjanus fulvus). Journal of Biogeography, 37(1), 133–147.

Giles, E. C., Saenz-Agudelo, P., Hussey, N. E., Ravasi, T., & Berumen, M. L. (2015). Exploring seascape genetics and kinship in the reef sponge Stylissa carteri in the Red Sea. Ecology and Evolution, 5(13), 2487–2502.

Guindon, S., & Gascuel, O. (2003). A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Systematic Biology, 52(5), 696–704.

Hall, T. A. (1999). BioEdit: a user–friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. In Nucleic Acids Symposium Series (Vol. 41, No. 41, pp. 95–98). [London]: Information Retrieval Ltd., c1979–c2000.

Katoh, K., & Standley, D. M. (2013). MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Molecular Biology and Evolution, 30(4), 772–780.

Keshavmurthy, S., Yang, S. Y., Alamaru, A., Chuang, Y. Y., Pichon, M., Obura, D., ... MacDonald, A. (2013). DNA barcoding reveals the coral “laboratory– rat”, Stylophora pistillata encompasses multiple identities. Scientific Reports, 3, 1520.

Klunzinger C.B. (1879a) Die Korallenthiere des Rothen Meeres, 2. Theil: Die Steinkorallen. Erster Abschnitt: Die Madreporaceen und Oculinaceen. Gutmann, Berlin, 88.

Klunzinger C.B. (1879b) Die Korallenthiere des Rothen Meeres, 3. Theil: Die Steinkorallen. Zweiter Abschnitt: Die Asteraeaceen und Fungiaceen. Gutmann, Berlin, 100.

Lamarck, J.B.P. (1816) Histoire naturelle des Animaux sans Vertebres,̀ presentant́ les caracteres̀ geń eraux́ et particuliers de ces animaux, leur

121 distribution, leurs classes, leurs familles, leurs genres, et la citation des principales especes̀ qui s'y rapportent; preć ed́ eé d'une Introduction offrant la Determinatioń des caracteres̀ essentiels de l'Animal, sa distinction du Veǵ etaĺ et des autres corps naturels, enfin, l'exposition des principes fondamentaux de la Zoologie. Detervillé & Verdiere,̀ Paris.

Lanfear, R., Calcott, B., Ho, S. Y., & Guindon, S. (2012). PartitionFinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses. Molecular Biology and Evolution, 29(6), 1695–1701.

Malay, M. C. M. D., & Paulay, G. (2010). Peripatric speciation drives diversification and distributional pattern of reef hermit crabs (Decapoda: Diogenidae: Calcinus). Evolution: International Journal of Organic Evolution, 64(3), 634–662.

Miller, M. A., Pfeiffer, W., & Schwartz, T. (2012). The CIPRES science gateway: enabling high–impact science for phylogenetics researchers with limited resources. In Proceedings of the 1st Conference of the Extreme Science and Engineering Discovery Environment: Bridging from the eXtreme to the campus and beyond (p. 39). ACM.

Nanninga, G. B., Saenz-Agudelo, P., Manica, A., & Berumen, M. L. (2014). Environmental gradients predict the genetic population structure of a coral reef fish in the Red Sea. Molecular Ecology, 23(3), 591–602.

Paradis, E., Claude, J., & Strimmer, K. (2004). APE: analyses of phylogenetics and evolution in R language. Bioinformatics, 20(2), 289–290.

Pichon, M., Benzoni, F., Chaineu, C., & Dutrieux, E. (2010). Field Guide to the hard corals of the southern coast of Yemen. Collection Parthénope, 1–256.

Pons, J., Barraclough, T. G., Gomez–Zurita, J., Cardoso, A., Duran, D. P., Hazell, S., ... Vogler, A. P. (2006). Sequence–based species delimitation for the DNA taxonomy of undescribed insects. Systematic Biology, 55(4), 595– 609.

Prada, C., M. B. DeBiasse, J. E. Neigel, B. Yednock, J. L. Stake, Z. H. Forsman, … M. E. Hellberg. (2014) Genetic species delineation among branching Caribbean Porites corals. Coral Reefs 33, no. 4:1019–1030.

Puillandre, N., Lambert, A., Brouillet, S., & Achaz, G. (2012). ABGD, Automatic Barcode Gap Discovery for primary species delimitation. Molecular Ecology, 21(8), 1864–1877.

122

Ronquist, F., Teslenko, M., Van Der Mark, P., Ayres, D. L., Darling, A., Höhna, S., ... Huelsenbeck, J. P. (2012). MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Systematic Biology, 61(3), 539–542.

Rohling, E. J., Grant, K., Bolshaw, M., Roberts, A. P., Siddall, M., Hemleben, C., & Kucera, M. (2009). Antarctic temperature and global sea level closely coupled over the past five glacial cycles. Nature Geoscience, 2, 500–504. https://doi.org/10.1038/ngeo557.

Saunders, N.C., Kessler, L.G. and Avise, J.C. (1986) Genetic variation and geographic differentiation in mitochondrial DNA of the horshoe crab, Limulus polyphemus. Genetics 112: 613–627.

Scheer, G. & Pillai, C.S.G. (1983) Report on the stony corals from the Red Sea. Zoologica, (Stuttgart) 133: 1–198.

Sheppard, C. R., & Sheppard, A. S. (1991). Corals and Coral Communities of Arabia (Vol. 12).

Skillings, D. J., Bird, C. E., & Toonen, R. J. (2011). Gateways to Hawai ‘i: genetic population structure of the tropical sea cucumber Holothuria atra. Journal of Marine Biology, 2011.

Stamatakis, A. (2014). RAxML version 8: a tool for phylogenetic analysis and post–analysis of large phylogenies. Bioinformatics, 30(9), 1312–1313.

Stephens, M., Smith, N. J., & Donnelly, P. (2001). A new statistical method for haplotype reconstruction from population data. The American Journal of Human Genetics, 68(4), 978–989.

Tamura, K., Dudley, J., Nei, M., & Kumar, S. (2007). MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Molecular Biology and Evolution, 24(8), 1596–1599.

Veron, J. E. N. (2000). Corals of the World (No. C/593.6 V4).

WoRMS Editorial Board (2019). World Register of Marine Species. Available from http://www.marinespecies.org at VLIZ. Accessed 2019–02–03. doi:10.14284/170

Zhang, J., Kapli, P., Pavlidis, P., & Stamatakis, A. (2013). A general species delimitation method with applications to phylogenetic placements. Bioinformatics, 29(22), 2869–2876.

123

3. CHAPTER 3: USING ezRAD TO RECONSTRUCT THE COMPLETE MITOCHONDRIAL GENOMES OF PORITES FONTANESII AND PORITES COLUMNARIS (CNIDARIA: SCLERACTINIA)

Tullia I. Terraneoa,b*, Roberto Arrigonia, Francesca Benzonic, Zac H. Forsmand & Michael L. Berumena aRed Sea Research Center, Division of Biological and Environmental Science and Engineering, King Abdullah University of Science and Technology, Thuwal 23955-6900, Saudi Arabia bARC Centre of Excellence for Coral Reef Studies, James Cook University, Townsville 4811, QLD, Australia cDepartment of Biotechnologies and Bioscience, University of Milano- Bicocca, Milan, 20126, Italy

dHawaii Institute of Marine Biology, Kaneohe 96744, HI, USA

Published in Mitochondrial DNA Part B, 3(1), 173-174 Terraneo, T. I., Arrigoni, R., Benzoni, F., Forsman, Z. H., & Berumen, M. L. (2018). Using ezRAD to reconstruct the complete mitochondrial genome of Porites fontanesii (Cnidaria: Scleractinia). Mitochondrial DNA Part B, 3(1), 173-174.

Published in Mitochondrial DNA Part B, 3(1), 286-287 Terraneo, T. I., Arrigoni, R., Benzoni, F., Forsman, Z. H., & Berumen, M. L. (2018). The complete mitochondrial genome of Porites harrisoni* (Cnidaria: Scleractinia) obtained using next-generation sequencing. Mitochondrial DNA Part B, 3(1), 286-287.

This chapter is edited from the original version of the paper

124

*At the time of publication P. columnaris was identified as P. harrisoni. The publication will be emended in the future but I corrected the identification for the following dissertation for consistency.

125

3.1 Abstract

Corals in the genus Porites are among the major framework builders of reef structures worldwide, yet the genus has been challenging to study due to a lack of informative molecular markers. Here, we used ezRAD to reconstruct the complete mitochondrial genomes of Porites fontanesii (GenBank accession number MG754069) and Porites columnaris (GenBank accession number MG754070), two widespread coral species endemic to the Red Sea and Gulf of Aden. The gene arrangement of P. fontanesii did not differ from other Scleractinia and consisted of 18,658 bp, organized in 13 protein-coding genes, 2 rRNA genes, and 2 tRNA genes. Genome length of P. columanris consisted of 18,630 bp, with a base composition of 25.92% A, 13.28% T, 23.06% G, and 37.73% C. Consistent with other hard corals, the P. columnaris mitogenome was arranged in 13 protein-coding genes, 2 rRNA, and 2 tRNA genes. nad5 and cox1 contained embedded Group I Introns of

11,133 bp and 965 bp respectively. These mitochondrial genomes contribute essential data to work towards a better understanding of evolutionary relationships within Porites.

Keywords: Mitochondrial genome, RAD sequencing, Scleractiania

126

3.2 Introduction

Corals of the genus Porites are among the most ecologically common corals in the tropics (Veron 2000). Porites fontanesii Benzoni & Stefani, 2012 and P. columnaris Klunzinger, 1879 are well-defined coral species belonging to the hard-coral family Poritiidae. Although only described recenlty, P. fontanesii is a common and widespread taxon in the Red Sea, with a distribution extending to the Gulf of Tadjoura, the Gulf of Aden and Socotra

(Benzoni & Stefani, 2012). Similarly, Porites columnaris is a major framework builder endemic to the Red Sea and the Gulf of Aden. The genus Porites is still taxonomically challenging in terms of species boundaries (Forsman et al., 2009, 2017; Hellberg et al., 2016), yet P. fontanesii and P. columnaris are morphologically and molecularly distinctive, presenting unique morphological features (Benzoni & Stefani, 2012).

In order to provide a better understanding of evolutionary relationships within the genus Porites, and to provide a baseline for clarifying caral endemism patterns in the seas around the Arabina Peninsula, I reconstructed the complete mitochondrial genomes of these two species endemic to the Red Sea and the Gulf of Aden.

3.3 Materials and methods

The sample of P. fontanesii was collected at Ras Qadamah Reef, in Socotra

Island, Yemen (12° 41.902 N; 53° 39.683 E), while the sample of P. columanris

127 was collected at Dhi Dahaya Reef in the southern Saudi Arabian Red Sea

(16°52.377' N; 41°26.409' E). Specimen are now deposited at King Abdullah

University of Science and Technology, Saudi Arabia (specimen voucher

SO114 and SA1704 respectively).

Genomic DNA was extracted using DNeasy® Blood and Tissue Kit

(Qiagen Inc., Hilden, Germany), quantified using Qubit 2.0 fluorometer

(Invitrogen, Carlsbad, California, USA), and digested with frequent cutting enzymes Mbol and Sau3AI (New England Biolabs, Ipswich, Massachusetts,

USA), following Toonen et al., (2013). ezRAD libraries were prepared using

Illumina TruSeq® Nano DNA kit following the manufacture’s protocol, and paired-end sequenced using HiSeq® 4000 platform in the Bioscience Core

Lab facility at King Abdullah University of Science and Technology, Saudi

Arabia. Reads were assembled to P. lobata reference mitogenome

(NC030186) using Geneious® v.10.1.3 (Biomatters Ltd. Auckland, New

Zealand), and a consensus sequence exported using 0% majority option for coverage greater then 3 X. Genes were annotated using the online platforms DOGMA (Wyman et al., 2004) and MITOS (Bernt et al., 2013), and were manually inspected. tRNA was additionally scanned with the tRNAscan-SE (Schattner et al., 2005) web server.

3.4 Results and discussion

128

The complete P. fontanesii mitogenome consisted of 18,658 bp, with the following overall base composition: A 25.81%, T 37.54%, C 13.64% and G

23.19%, in agreement with the typical mitogenome base composition (i.e.,

A+T rich) of scleractinian corals (Fukami & Knowlton 2005; Arrigoni et al.,

2016). The reconstructed genome included 13 protein-coding genes, 2 ribosomal RNA genes (rnl and rns) and 2 transfer RNA genes (trnM and trnW). Nad5 and cox1 genes were interrupted by Group I Introns. Nad5

Group I Intron consisted of 11,135 bp, comprising 10 encoding genes, while cox1 Group I Intron was 965 bp long (Table 3.1).

The complete mitogenome of P. columnaris consisted of 18,630 bp, with a base composition of 25.92% A, 13.28% T, 23.06% G, and 37.73% C. The total number of protein coding genes was 13, consistent with the typical scleractinian mitogenome organization (Medina et al., 2006; Fukami et al.,

2007). A total of 2 ribosomal RNA (rnl, nrs) and 2 transfer RNA (trnM, trnW) genes was recovered. Among the 13 protein coding genes, nad5 and cox1 were interrupted by Group I Introns, 11,133 bp and 965 bp in length respectively (Table 3.2).

A phylogenetic tree comprising the P. fontanesii and the P. columnaris mitochondrial genomes and all published mitogenomes of

Poritidae and its sister taxon, Dendrophylliidae, was reconstructed using

Bayesian inference as implemented in MrBayes 3.1.2 (Ronquist &

Huelsenbeck, 2003) for 1,000,000 generations and maximum-likelihood as

129 implemented in PhyML 3.0 (Guindon et al., 2003) (Fig. 3.1). The monophyly of the genus Porites was well supported by the mitochondrial phylogeny, and a distinctive position of P. fontanesii and P. columnaris within Porites was highlighted by the reconstruction.

The implementation of these data with other Porites mitochondrial genomes will help clarify evolution in one of most important framework builders of coral reefs.

3.5 Conclusion

Complete mitochondrial genomes have been traditionally obtained by isolation of mitochondria, shotgun sequencing, and subcloning. The approach followed in the present work has already been applied to reconstruct complete mitochondrial genomes in a variety of organisms, comprising corals, and represent a cost and labor efficient method.

Reconstructing the complete mitochondrial genomes of P. fontanesii and

P. columnaris, I provide important sources of data that will contribute towards a better understanding of the tree of life, as well as a more comprehensive understanding of Porites evolution.

Table 3.1 Porites fontanesii complete mitochondrial genome annotation

130

Region Location Length (bp) AT (%) Start Codon Stop Codon trnM 1-71 71 47.89 rnl 115-2,415 2,341 61.73 nad5 (5’) 2,574-3,292 719 64.39 GTG

Group I 3,293-3,625 11,131 63.45

Intron

(nad5) nad1 3,626-4,609 984 62.30 ATG TAA cob 4,730-5,890 1,161 62.70 ATG TAA nad2 6,460-7,557 1,098 62.11 ATG TAA nad6 7,591-8,184 594 64.14 ATA TAA atp6 8,258-8,956 699 64.81 ATG TAG nad4 9,220-10,695 1,476 62.67 ATG TAA rns 10,770- 1,029 58.70

11,800 cox3 12,033- 789 63.50 ATG TAG

12,821 cox2 12,857- 744 63.98 ATG TAG

13,600 nad4L 13,654- 300 70.33 GTG TAA

13,953 nad3 13,968- 357 64.43 GTG TAG

14,324 nad5 (3’) 14,424- 1,117 63.38 TAG

15,540 trnW 15,580- 70 50

15,649 atp8 15,682- 216 70.83 ATG TAA

15,897

131 cox1(5’) 16,021- 894 61.19 ATG

16,914

Group I 16,915- 965 63.11

Intron 17,879

(cox1) cox1 (3’) 17,880- 648 63.16 TAG

18,563

Table 3.2 Porites columnaris complete mitochondrial genome annotation

Region Location Length (bp) AT (%) Start Codon Stop Codon trnM 1-71 71 46.48 rnl 116-2,417 2,301 62.06 nad5 (5’) 2,563-3,281 719 62.26 GTG

Group I 3,282- 11,133 63.88

Intron 14,414

(nad5) nad1 3,616-4,599 984 62.40 ATG TAA cob 4,721-5,881 1,161 62.70 ATG TAA nad2 6,451-7,548 1,098 61.93 ATG TAA nad6 7,582-8,175 594 64.48 ATA TAA atp6 8,249-8,947 699 65.52 ATG TAG nad4 9,210- 1,476 63.14 ATG TAA

10,685 rns 10,761- 10,029 59.09

11,791 cox3 12,024- 789 63.75 ATG TAG

12,812

132 cox2 12,848- 744 64.65 ATG TAG

13,591 nad4L 13,645- 300 70.67 GTG TAA

13,944 nad3 13,959- 357 64.43 GTG TAG

14,315 nad5 (3’) 14,415- 1,117 63.83 TAG

15,531 trnW 15,571- 70 50

15,640 atp8 15,673- 216 69.44 ATG TAG

15,888 cox1(5’) 15,993- 894 61.41 ATG

16,886

Group I 16,887- 965 64.15

Intron 17,851

(cox1) cox1 (3’) 17,852- 648 63.60 TAG

18,535

133

columnaris

Figure 3.1 Phylogenetic reconstruction based on complete mitochondrial genomes of

Porites fontanesii, P. columnaris and other Scleractinia. Numbers at nodes represent

Bayesian posterior probabilities and maximum likelihood bootstrap values. Fungiacyathus stephanus was selected as an outgroup.

134

3.6 References

Arrigoni, R., Vacherie, B., Benzoni, F., Barbe, V. (2016) The complete mitochondrial genome of Acanthastrea maxima (Cnidaria, Scleractinia, Lobophylliidae). Mitochondrial DNA, 27:927–928. doi:10.3109/19401736.2014.926489.

Benzoni, F., & Stefani, F. (2012) Porites fontanesii, a new species of hard coral (Scleractinia, Poritiidae) from the southern Red Sea, the Gulf of Tadjoura, and the Gulf of Aden and its phylogenetic relationshios within the genus. Zootaxa, 3447:56–68.

Bernt, M., Donath, A., Jühling, F., Externbrink, F., Florentz, C., Fritzsch, G., Pütz, J., Middendorf, M., Stadler, P.F. (2013) MITOS: Improved de novo metazoan mitochondrial genome annotation. Molecular Phylogenetics and Evolution, 69:313–319. doi:10.1016/j.ympev.2012.08.023.

Forsman, Z.H, Knapp, I., Tisthammer, K. (2017) Coral hybridization or phenotypic variation? Genomic data reveal gene flow between Porites lobata and P. compressa. Molecular Phylogenetics and Evolution, 11:132– 148. doi.org/10.1016/j.ympev.2017.03.023.

Forsman, Z.H., Barshis, D.J., Hunter, C.L., Toonen, R.J. (2009) Shape-shifting corals: molecular markers show morphology is evolutionarily plastic in Porites. BMC Evolutionary Biology, 9:45. doi:10.1186/1471-2148-9-45.

Fukami, H., Knowlton, N. (2005) Analysis of complete mitochondrial DNA sequences of three members of the Montastraea annularis coral species complex (Cnidaria, Anthozoa, Scleractinia). Coral Reefs, 24:410–417. doi:10.1007/s00338-005-0023-3.

Guindon, S., Gascuel, O., Rannala, B. (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by Maximum Likelihood. Systematic Biology, 52:696–704. doi:10.1080/10635150390235520.

Hellberg, M.E., Prada, C., Tan, M.H., Forsman, Z.H., Baums, I.B.. (2016) Getting a grip at the edge: recolonization and introgression in eastern Pacific Porites corals. Journal of Biogeography, 43:2147–2159. doi:10.1111/jbi.12792.

Ronquist, F., Huelsenbeck, J.P. (2003) MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics, 19:1572–1574.

135 doi:10.1093/bioinformatics/btg180.

Schattner, P., Brooks, A.N., Lowe, T.M. (2005) The tRNAscan-SE, snoscan and snoGPS web servers for the detection of tRNAs and snoRNAs. Nucleic Acids Resources, 33:W686–W689. doi:10.1093/nar/gki366.

Toonen, R.J., Puritz, J.B., Forsman, Z.H., Whitney, J.L., Fernandez-Silva, I., Andrews, K.R., Bird, C.E. (2013) ezRAD: a simplified method for genomic genotyping in non-model organisms. PeerJ, 1:e203. doi:10.7717/peerj.203.

Wyman, S.K., Jansen, R.K., Boore, J.L. (2004) Automatic annotation of organellar genomes with DOGMA. Bioinformatics. 20:3252–3255. doi:10.1093/bioinformatics/bth352.

136

4. CHAPTER 4: PHYLOGENOMICS AND PHYLOGEOGRAPHY OF PORITES FROM THE ARABIAN PENINSULA

Tullia I. Terraneoa,b*, Roberto Arrigonia, Francesca Benzonic, Kiruthiga G. Mariappana, Andrew H. Bairdb, Zac H. Forsmand, Michael K. Woostera, Jessica Bouwmeestere, Alyssa Marshellf & Michael L. Berumena aRed Sea Research Centre, Division of Biological and Environmental Science and Engineering, King Abdullah University of Science and Technology, Thuwal 23955-6900, Saudi Arabia bARC Centre of Excellence for Coral Reef Studies, James Cook University, Townsville 4811, QLD, Australia cDepartment of Biotechnologies and Bioscience, University of Milano- Bicocca, Milan, 20126, Italy, dHawaii Institute of Marine Biology, Kaneohe 96744, HI, USA eDepartment of Biological and Environmental Sciences, Collage of Arts and Sciences, Qatar University, Doha, Qatar. fDepartment of Marine Science and Fisheries, College of Agricultural and Marine Sciences, Sultan Qaboos University, Muscat, Oman

Under revision in Systematic Biology

Terraneo, T.I., Arriogni, R., Benzoni, F., Mariappan, K.G., Baird, A.H., Forsman,

Z.H., Wooster, M.K., Bouwmeester, J., Marshell, J., Berumen, M.L.

Phylogenomics and phylogeography of Porites from the Arabian Peninsula

137

4.1 Abstract

The advent of high throughput sequencing technologies provides the unprecedented chance to resolve phylogenetic relationships among closely related species. By incorporating hundreds to thousands of unlinked loci and single nucleotide polymorphisms (SNPs), phylogenomic analyses have the potential to resolve species boundaries at a much higher detail, especially within recalcitrant taxa. Scleractinian corals are historically challenging to identify based on morphology and lack of an adequate set of molecular markers to investigate their phylogenies. Here, we examine the potential of Restriction-site Associated DNA sequencing (RADseq) to investigate phylogenetic relationships and species limits within the coral genus Porites. A total of 134 colonies ascribed to 13 morphological species and one unknown morphology, were collected from 16 localities in the seas surrounding the Arabian Peninsula. Reference mapping was used to retrieve and compare almost complete mitochondrial genomes, ribosomal

DNA regions, and histone regions. De novo clustering was employed to obtain loci and SNPs from the complete holobiont dataset and the coral subset that maps to coral transcriptome. A variety of species discovery methods (phylogenetic and clustering analyses) and species delimitation approaches (coalescent-based, species tree, and Bayesian Factor delimitation) was used. With some exceptions, most methods suggest the presence of eight distinct molecular lineages within the collected Porites,

138 one of which identified here for the first time. Finally, the taxonomic and biogeographic history of Porites was discussed in the context of the well- supported phylogenetic hypotheses.

Keywords: ezRAD, dDocent, mitochondrial genome, species delimitation, species tree, systematics

4.2 Introduction

Understanding of species boundaries and evolutionary relationships among organisms is a key question in biology. Recent advances in molecular and computational techniques opened new avenues to phylogenomic studies, revolutionizing our understanding of evolution and speciation also for non-model organisms (Puritz et al., 2014). Harnessing of high throughput sequencing technologies (NGS), restriction-sites- associated fragmentation of genomic DNA (RADseq) offers the opportunity to overcome problems inherit to the use of traditional sequencing approaches (Baird et al., 2009), providing genomic-wide data and a large number of homologous markers (Pante et. al., 2015). Allowing for the simultaneous discovery and typing of thousands of polymorphic loci throughout the genome, and not requiring any prior genomic resources for the study taxon (Baxter et al., 2011), RADseq is nowadays the most widely used genomic approach for high-throughput single nucleotide polymorphism (SNP) discovery and genotyping in non-model organisms

139

(Pante et al., 2014; Forsman et al., 2017). Closely related species share orthologous restriction sites, thus RADseq is broadly used to infer recent evolutionary history (Eaton & Ree, 2013; Harvey et al., 2016; Jones et al.,

2013; Laché et al., 2015). Yet, it has also been applied to clarify relatedness that transcend deep evolutionary timescales up to 63 Ma (Rubin et al.,

2012; Cariou et al., 2013; Cuard et al., 2014; Hipp et al., 2014).

Anthozoans represent a class of recalcitrant basal metazoans, for which high levels of morphological variation, phenotypic plasticity, and few orthologous conserved markers, hindered a clear understanding of evolutionary relatedness (Prada et al., 2009; Forsman et al., 2009; Paz-

García et al., 2015). Inferences for taxonomic relationships within the class have historically been informed morphologically. Yet, these animals are notoriously difficult to sample, are characterized by very simple body plans, and thus intrinsically bear few morphological informative traits. Moreover, molecular studies have uncovered widespread homoplasy and convergent evolution of morphology within the subclasses Hexacorallia and Octocorallia (Daly et al., 2003). The use of molecular barcoding also proved largely unsuccessful in species identification within the class mainly due to slow rates of mitochondrial DNA evolution (10-20 times slower compared to vertebrates) (Hellberg et al., 2006; Huang et al., 2008; Shearer et al., 2002), the presence of divergent paralogous copies in the nuclear ribosomal DNA, resulting in high levels of intraspecific variations (Odorico &

140

Miller, 1997; Sánchez & Dorado, 2008), and finally the lack of variable nuclear regions comparable among taxa (Conception et al., 2008;

McFadden et al., 2010). Yet some exception exists, such as in the order

Zaontharia, where the barcode COI region proved successful to distinguish species level boundaries for a variety of taxa (Sinniger et al., 2008). Other scenarios derived for example from incomplete lineage sorting and/or hybridization, might increase topological deviations between gene and species trees in a plethora of taxa (e.g. Arnold, 1997; Pollard et al., 2006), complicating our understanding of real evolutionary history. Recently,

RADseq has been successfully applied to untangle the phylogeny of deep- sea octocorals in the genera Chrysogorgia, Paragorgia, and Ovabunda

(Pante et al., 2014; Herrera & Shank, 2016; McFadden et al., 2017), and clarified boundaries among some species in the scleractinian genera

Pocillopora and Porites (Combosch & Vollmer, 2015; Forsman et al., 2017;

Dimond et al., 2018; Johnston et al., 2018).

The scleractinian coral genus Porites has a long history of taxonomic uncertainty. With 190 described species it remains one of the most specious and ecologically important reef dwelling corals (Veron 2000). Nevertheless, species boundaries and evolutionary relationships within Porites are to date unresolved. As for other Scleractinia, the morphological traits historically used to define species in Porites have been proven to lack evolutionary meaning, being subject to stasis and convergent evolution (Forsman et al.,

141

2015; Tisthammer et al., 2018), and informative morphological synapomorphies are yet to be identified. Multi-locus phylogenetic reconstruction has uncovered cryptic speciation and unresolved species complexes within Porites (Forsman et al., 2009; Prada et al., 2014; Hellberg et a. 2016; Terraneo et al., accepted). Patterns of introgression have also been recovered among different species in the Eastern Pacific and Hawai’i

(Hellberg et al., 2016; Forsman et al., 2017), highlighting gaps in our understanding of evolution, ecology and biogeography, and the need for an integrative approach to species delimitation within the genus.

The seas around the Arabian Peninsula, comprising the Red Sea, the

Gulf of Aden, the Gulf of Oman, and the Arabian Gulf, represent a hotspot of Porites biodiversity. High levels of biodiversity and endemism from peripheral regions such as the Red Sea, Hawai’i, and the Marquesas Islands, are expected as new studies identified these regions as potential biodiversity incubators, exporting lineages to the Indian and West Pacific

Oceans (Bowen et al., 2013; DiBattista et al., 2010, 2013, 2016; Gaither et al.,

2010, 2011). Since the 18th Century, 28 species of Porites have been reported from the Arabian Peninsula (Forskal, 1757; Lamarck, 1816;

Klunzinger, 1879; Sheer & Pillai, 1983; Sheppard & Sheepard, 1991; Veron,

2000; Pichon et al., 2010; Benzoni & Stefani, 2012; Terraneo et al., accepted a, b; Berumen et al., 2019), a value higher than other regions in the West

Pacific such as the Australian Great Barrier Reef (17 spp. sensu Veron, 2000)

142 and Japan (23 spp. sensu Veron, 2000). Recent morpho-molecular works, investigating two molecular loci and corallite level micromorphology, suggested that Porites biodiversity form the Red Sea and Gulf of Aden should be reconsidered, lumping 10 species into six genetic lineages

(Terraneo et al., accepted a, b).

In this work, we reconstructed the phylogenetic relationships from 134

RADseq libraries comprising 14 morphologies of Porites form the Arabian

Peninsula, one of which was not comparable to any described species. We reconstructed molecular phylogenies mapping complete mitochondrial genomes, nuclear ribosomal DNA, and histone regions. We compared the reconstructed trees with topologies obtained through 54,108 genome-wide

SNPs data from the holobiont, and 96,986 genome-wide SNPs from the coral data sets, and we applied genetic clustering analyses and coalescent species delimitations to explore relationships and boundaries among species. Finally, we provided a dated phylogeny of the genus and we discussed evolutionary and biogeographical patterns of Porites distributions in the Arabian Peninsula.

4.3 Materials and Methods

4.3.1 Collection and identification

143

A total of 139 Porites colonies were collected at different sites in the seas around the Arabian Peninsula, between 2013 and 2017. In particular, 22 samples from the Gulf of Tadjoura (Djibouti), 49 samples from the Gulf of

Aden (Yemen) and Socotra Island (Yemen) Socotra between 2007 and

2010 within the framework of the Total E&P – Creocean – University of Milano

- Bicocca (UNIMIB) ‘Yemen Scleractinia Biodiversity project’, 44 samples were collected along the Saudi Arabian Red Sea coast between 2013 and

2017 within the context of the biodiversity cruises carried out by the King

Abdullah University of Science and Technology (KAUST). A total of 9 samples were collected along the Omani coast in the Gulf of Oman in collaboration with the Sultan Qaboos University, and 10 samples from the Arabian Gulf

(Qatar) were collected in collaboration with the Qatar University (See

Appendix 4.1 for the pictures of the morphologies in the field, 4.2 for a list of the collected samples analysed genetically (n = 130), and 4.3 for the maps of the sampling localities). Each coral colony was imaged underwater with a Canon G15 and collected with hammer and chisel. At surface, a small piece (<1cm) of tissue from each colony was preserved in 98% ethanol or

CHAOS solution and stored for genomic analyses, while the specimens were bleached with sodium hypochlorite for 24h and air dried for morphological examination and deposit at the University of Milano-

Bicocca (UNIMIB, Italy) and at King Abdullah University of Science and

Technology (KAUST, Saudi Arabia) under unique voucher numbers.

144

Specimens were imaged with a Leica M80 microscope equipped with a

Leica IC80HD camera, and morphologically identified following Klunzinger

(1879a, 1879b), Sheer & Pillai (1983), Bernard (1903), Gravier (1910),

Sheppard & Sheppard (1991), Veron (2000), Claereboudt (2006), Pichon et al., (2010), Benzoni & Stefani (2012), Terraneo et al., accepted a, b. The following skeletal features were considered for each coral sample: corallite diameter, wall thickness, fusion pattern of the ventral triplet, number of pali, number of denticles, presence or absence of the columella, and presence or absence of the coenosteum (see Budd et al., 2012 for glossary).

4.3.2 DNA extraction and quantification

Genomic DNA was extracted using DNeasy® Blood & Tissue Kit (Qiagen,

Hilden, Germany) for the samples stored in ethanol or a phenol-chloroform- based method with a phenol extraction buffer (100 mM Tris-Cl pH 8, 10 mM

EDTA, 0.1% SDS) for the samples stored in CHAOS solution. Extracted DNA was quantified with the Qubit dsDNA High Sensitivity Assay kit (Thermo Fisher

Scientific, Waltham, MA, USA) using a Qubit® fluorometer (Thermo Fisher

Scientific, Waltham, MA, USA).

4.3.3 PCR amplification and sequencing

The mitochondrial Control Region (mtCR), was amplified for 120 Porites samples using Polymerase Chain Reaction (PCR) and the primers zpsRNSf

145

(5’ – AGC AGA CGC GGT GAA ACT TA – 3’) and zpCOX3r (5’ – GCC CAA

GTA ACA GTA CCC CC – 3’) (Terraneo et al., accepted a). Amplifications were conducted in 15 µl PCR volume, composed of 0.2 µM of each primer,

1X Multiplex PCR Master Mix (Qiagen, Hilden, Germany), and <5 ng DNA.

All the products were purified by adding 1µl Illustra ExoStar (GE Healthcare,

Buckinghamshire, UK), and incubating at 37˚ C for 60 min, followed by 85˚

C for 15 min, and directly sequenced in forward and reverse directions using an ABI 3730xl DNA Analyzer (Applied Biosystems, Carlsbad, CA, USA).

Forward and reverse sequences were assembled using Geneious® v.10.1.3

(Biomatters Ltd. Auckland, New Zealand), and multiple alignment was performed using the E-INS-i option in MAFFT v.7 (Katoh and Standley 2013) and manually checked using BioEdit v.7.2.5 (Hall 1999). All the newly produced sequences are deposited at ENA (see Appendix 4.2).

4.3.4 Restriction enzyme digestion and ezRAD libraries preparation

We followed protocols by Toonen et al., (2013) and Knapp et al., (2016) for

DNA digestion and ezRAD library preparation. In detailed, each sample was digested using frequent cutter restriction enzymes MboI and Sau3AI

(New England BioLabs, Ipswich, MA, USA) to cleave sequences at GATC cut sites (Toonen et al., 2013). Digestions were performed in a 50 µl reaction volume consisting of 43 µl dsDNA (about 1.2-1.3 µg), 5 µl of Cutsmart Buffer

(New England BioLabs, Ipswich, MA, USA), and 1 µl of each undiluted

146 restriction enzyme, under the following thermocycler profile: 37° C for 3 hours followed by 65° C for 20 minutes. Digested samples were cleaned using Agencourt AMPure XP beads (Beckmann Coulter, Danvers, MA, USA) at a 1:1.8 (DNA:beads) ratio following the standard protocol. The concentration of all cleaned digests was then checked with a Qubit®

Fluorometer 3.0 (Thermo Fisher Scientific, Waltham, MA, USA). A total amount of 200 ng of each digested DNA samples was used for the library preparation using the TruSeq® Nano DNA Library prep kit (Illumina, San

Diego, CA, USA) following the manufacture’s protocol. All libraries were size-selected at 350 bp and passed through two quality control steps, i.e. bioanalyzer and qPCR, to check size and concentration, respectively.

Finally, ezRAD libraries were normalized and combined to two pools of 65 libraries each. Each libraries pool was run in a single 150 bp paired-end lane on Illumina HiSeq 4000 System at KAUST Genomics Core Lab (Thuwal, Saudi

Arabia). The sequenced lengths, number of reads, and sample information for each library are presented in Appendix 4.4.

4.3.5 ezRAD data processing

The Illumina raw data consisted of ~443 million 150 bp reads. All samples were de-multiplexed using their unique barcode and adapter sequences, effectively removing reads that lacked identifiable barcode pairs. An average of 3 million reads per individual (N = 139) were trimmed,

147 assembled, and genotyped using the software dDocent v.2.25 (Puritz et al.,

2014) (Appendix 4.4).

Data processing was performed twice, the first time on the holobiont dataset and the second time on the coral data set. To perform the analysis as a holobiont, the protocol mentioned in Puritz et al., 2014 was followed.

In short, ~443 million reads (from 139 samples) were placed in a folder as

*.F.fq.gz and *.R.fq.gz and trimmed reads were placed as *.R1.fq.gz and

*.R2.fq.gz respectively. Reads were merged using PEAR V0.9.6 and assembled using bwa 0.7.15. SNPs were identified using FreeBayes using the settings mentioned in Forsman et al. (2017). Coral only reads were then binned. The trimmed reads were first assembled to transcriptomic reference sequences of P. lobata obtained from Forsman et al. (2017) using bowtie2

2.3.4. Later, the aligned reads were converted to bam format using samtools V1.6 and then converted to fastq using bedtools V2.26.0. These binned files were then copied to a separate folder and genotyped using dDocent v.2.25 (Puritz et al., 2014). In short, the reads were trimmed using

Trimmomatic V0.36, merged using PEAR 0.9.6 and aligned to the reference transcriptome again using bwa 0.7.15 using the setting: -t 16 -a -M -T 10 –R.

SNPs were identified using FreeBayes as mentioned in Forsman et al., 2017.

The resulting VCF files of both the holobiont and coral datasets were further filtered using VCFtools v.0.1.16 (Danecek et al., 2011). To examine the sensitivity of the phylogenetic inference to the filtering process, we

148 generated two filtered supermatrices for both the coral and the holobiont datasets. In particular, we obtained the “coral-max” and the “holobiont- max” supermatrices using the following filter options: mean depth = 3, max missing data = 50%, and minimum distance between SNPs = 5. Conversely, we generated the “coral-min” and the “holobiont-min” supermatrices using the following filter options: mean depth = 10, max missing data = 5%, and minimum distance between SNPs = 10. Haplotypes were then called and filtered for complex loci, potential paralogs, missing data, and sequencing errors using the rad_haplotyper v.1.1.8 pipeline

(https://github.com/chollenbeck/rad_haplotyper; Willis et al., 2017).

Contigs were then collapsed into genotypes for final analyses. PGDspider v.2.1.1.5 (Lischer and Excoffier 2011) was used to convert the dataset to the required file types for further analysis. At the end of the rad_haplotyper pipeline, the “coral-max” supermatrix contained 6452 loci and a total of

96986 SNPs, the “holobiont-max” supermatrix 2041 loci and a total of 54108

SNPs, the “coral-min” supermatrix 367 loci and a total of 4140 SNPs, and the

“holobiont-min” supermatrix 719 loci and a total of 10918 SNPs. Each of the resulting four concatenated loci supermatrices was analysed in RAxML-

HPC2 v8.0 (Stamatakis 2014) for maximum likelihood (ML) phylogenetic inference. All the phylogenetic analyses, we applied the GTR + GAMMA substitution model and the branch support was assessed by 1000 bootstrap

149 replicates. All ML analyses were run on the CIPRES Science Gateway (Miller et al., 2010).

4.3.6 Reference assemblies and phylogenetic analyses of mitochondrial genomes, histone, and rDNA regions

One of the main benefits of ezRAD among other RADseq techniques is that ezRAD ends up with both considerable vertical and horizontal coverages

(Toonen et al., 2013; Stobie et al., 2019). While the former coverage is used to call SNPs, the latter coverage generates very long contigs, resulting in the resolution of the complete or a large percentage of the mitochondrial genomes and other multicopy gene regions such as histones and ribosomes. Therefore, we used reference mapping against previously published reference sequences of mitochondrial genome (mitogenome hereafter), histone genes (histone hereafter), and nuclear ribosomal DNA arrays (rDNA hereafter, including the complete 18S, ITS1, 5.8S, ITS2, and 28S regions) to acquire and compare nearly complete mitogenomes, histone, and rDNA from each library. In particular, we used as references the complete mitochondrial genome of P. lobata (accession number:

NC030186, 18647 bp) and the almost complete histone (5301 bp) and rDNA

(6629 bp) sequences of P. superfusa obtained by Forsman et al. (2017). Raw de-multiplexed reads were trimmed for adapters and low-quality bps using

Trimmomatic V0.36. Trimmed reads were aligned to the reference whole

150 mitochondrial genome/rDNA/histone of P. lobata and P. superfusa using bowtie2 V2.3.4 in --fast-local mode. Aligned reads are convert to bam and indexed using samtools version 1.6, and the consensus sequences were identified using samtools mpileup V1.6 combined with Vcfutils.pl.

We aligned mitochondrial genomes, histone, and rDNA sequences using MAFFT v.7 (Katoh and Standley 2013) (all alignment data are available upon request to the corresponding author).

We determined the optimal among-gene partitioning scheme and model choice for dataset in PartitionFinder v.2 (Lanfear et al., 2012) under the Bayesian Information Criterion (BIC). The mitochondrial genomes were partitioned according to the genes and considering all intergenic regions as a single partition, with all genes were further partitioned according to the codon positions. The rDNA dataset was partitioned in five partitions, 18S,

ITS1, 5.8S, ITS2, and 28S, the histone dataset was partitioned by genes and all genes according to the codon position. Phylogenetic relationships based on these three datasets were inferred using two phylogeny reconstruction methods, i.e. bayesian inference (BI) and maximum likelihood (ML). Porites superfusa, a basal and highly divergent species from the Eastern Pacific, was selected as outgroup for all the phylogenetic analyses following Forsman et al., (2017). Bayesian phylogenetic inference was performed with MrBayes v.3.2.6 (Ronquist et al., 2012) using two independent runs of four chains. Chains were started from random trees

151 and run for 10 million generations each, being sampled every 1000 generations. Stationarity from each independent run was assessed in Tracer v1.6 (Rambaut and Drummond 2007) by checking that the effective sample sizes (ESS) of all parameters were greater than 200 and the first 10% of trees were discarded as burn-in before generating the consensus tree.

ML trees were inferred with RAxML-HPC2 v8.0 (Stamatakis 2014), using the

GTR + GAMMA model of nucleotide substitution. Node support was assessed using 1000 bootstrap replicates. Both BI and ML analyses were run on the CIPRES Science Gateway (Miller et al., 2010). Finally, the mtCR dataset was analysed following the same criteria and methods used for these three datasets.

4.3.7 Genetic clustering analyses

The VCF file including coral SNPs obtained from the dDocent v.2.25 (Puritz et al., 2014) pipeline was further filtered using VCFtools v.0.1.16 (Danecek et al., 2011) to create additional filtered datasets that contained different number of SNPs and various levels of filtering options. In particular, SNPs were filtered based on different values of mean depth (--min-meanDP as 3,

5, and 10), missing data (--max-missing-count as 5%, 20%, and 50%), and minimum distance between SNPs (--thin as 5, 10, and 300), generating a total of 27 different coral SNPs filtered datasets. The most “relaxed” dataset

(--min-meanDP 3, --max-missing-count 50%, --thin 5) included a total of

152

96986 SNPs, whereas the most “stringent” dataset (--min-meanDP 10, --max- missing-count 5%, --thin 300) contained a total of only 343 SNPs.

All these coral filtered datasets were analysed by means of clustering analyses without any a priori hypotheses about individual assignment in order to evaluate the effects of filters and to guide subsequent species delimitation analyses. First, a Principle Components Analysis (PCA) was conducted using Plink v.1.9 (Purcell et al., 2007). The main benefit of PCA is its ability to detect genetic structure without the computational burden of

Bayesian clustering algorithms and the absence of assumptions about the underlying population genetic model. Second, Admixture v.1.23

(Alexander et al., 2009) was used to detect the genetic structure among all the analysed samples. This maximum likelihood-based program implements an underlying population genetic model similar to Structure (Pritchard et al., 2000). While both programs assign individuals into clusters using population allele frequencies and ancestry proportions, Admixture has the added benefit of a fast-numerical optimization algorithm to decrease computational time while avoiding problems with MCMC convergence.

We used the cross-validation procedure to select the optimal K value

(Alexander et al., 2009), testing values ranging from K = 2 to K = 12.

4.3.8 Species delimitation and species tree inference

153

We used Bayes Factor Delimitation (BFD*) to rank species delimitation models in a multispecies coalescent framework (Leaché et al., 2014).

Briefly, BFD* consists of running SNAPP analyses (Bryant et al., 2012) on models with different numbers of species and assignments of individuals to species, estimating the marginal likelihood of each model, and ranking model fit among runs by comparing Bayes factors (BF). The BFD* approach uses path sampling to estimate the marginal likelihood (MLE) of a population divergence model directly from SNPs data (without integrating over gene trees) and has been shown to be robust to a relatively large amount of missing data (Leaché et al., 2014), being especially suited for

RADseq data. In particular, we tested the following five models: (A) one single species; (B) current taxonomy, including P. annae, P. columnaris, P. echinulata, P. sp. 1, P. fontanesii, P. sp. 2, P. harrisoni, P. lobata, P. lutea, P. monticulosa, P. somaliensis, P. solida, P. rus and P. sp. 3 (a total of 11 species and three unknown morphology); (C) assigning individuals according to the eight molecular Clades recovered in the concatenation-based phylogenies (a total of eight species); (D) lumped Clade V (P. annae, P. echinulata, P. harrisoni, P. lobata, P. lutea, P. solida), Clade VII (P. sp. 3), and

Clade VIII (P. somaliensis), split all other Clades (a total of six species); (E) assigning individuals to the partitions inferred by Admixture with the optimal

K = 5 (a total of five species). We performed the BFD* analysis using the

SNAPP package (Bryant et al., 2012) implemented in BEAST v.2.5.2

154

(Bouckaert et al., 2014). We estimated MLE of each model by running path sampling with 48 independent steps (chain length of 100000 MCMCs with a pre-burnin of 10000 steps). Model convergence was assessed by monitoring the ESS for the likelihoods of each path using Tracer v.1.6 (Rambaut and

Drummond 2007). We ranked the alternative species delimitation models by their MLE and calculated the corresponding BF to compare the models.

The strength of support from BF (2 ∗ [MLEbest – MLEalternative]) comparisons of competing models was evaluated using the framework of Kass and Raftery

(1995).

To investigate the phylogenetic relationships among the Arabian

Porites species, we used the coalescent species tree approach implemented in the SNAPP package (Bryant et al., 2012) implemented in

BEAST v.2.5.2 (Bouckaert et al., 2014). The method calculates species tree likelihoods directly from the data by estimating the probability of allele frequency change across nodes, thus bypassing the inference of individual gene trees. As species-tree methods require terminal taxa to be assigned a priori (Liu et al., 2009), we run two analyses and imposed in the first run a total of 8 taxa (following the BFD* results), including Clades I, II, III, IV, V, VI,

VII, VIII, and in the second run a total of 14 taxa, comprising Clades I, II, III,

IV, VI, VII, VIII and the six species contained in Clade V, i.e. P. annae, P. echinulata, P. harrisoni, P. lobata, P. lutea, and P. solida (following traditional taxonomy). In order to reduce the complexity in species tree

155 estimation and increase parameter convergence probability, we sampled one or two individuals and their respective haplotypes and alleles, since calculations do not benefit from adding extra individuals over number of loci (Drummond and Bouckaert 2015). Therefore, we used VCFtools v.0.1.16

(Danecek et al., 2011) to generate a supermatrix of 25 individuals and a total of 1107 unlinked biallelic SNPs with 0% missing data as required by

SNAPP. The MCMCs were run for 10 million generations with mutation rate and priors estimated during the chains and all the other settings were set as default. We monitored the traces for convergence using Tracer v.1.6

(Rambaut and Drummond 2007). We concluded the analyses when ESSs for all parameters were large (> 200) and the traces have reached stationarity, and, as suggested by the software, we discarded the first 10% of trees as burn-in. Densitree v.2.5.2 (Bouckaert et al., 2014) was used to visualize the posterior distributions of topologies as cladograms, hence allowing for a clear depiction of uncertainty in the topology.

4.3.9 Divergence time analysis

In order to provide provisional estimates of the divergence time of each of the obtained Porites molecular clades, a time-calibrated phylogenetic hypothesis was inferred based on a concatenated matrix of all protein- coding genes of the mitochondrial genome. Three mitochondrial genomes were downloaded from NCBI: the outgroup Goniopora columna

156

(accession number: NC015643), Porites panamensis (accession number:

NC024182), and Porites porites (accession number: NC008166). These two species are distributed at the opposite sides of the Isthmus of Panama.

Indeed, the former Porites species occurs along the western coasts of the

Central America (Eastern Pacific), while the latter Porites species is found along the eastern coasts of the Central America (the Caribbean). Given the sister relationships between these two species based on Terraneo et al.,

(2018 a, b), we hypothesised that the closure of the Isthmus of Panama may have caused the vicariance between P. panamensis and P. porites, and, therefore, we used the split between them as the calibration point for the timetree. Despite some researchers having proposed an early and complex emergence of the Isthmus of Panama between 23 and 7 million years ago

(Ma) (Bacon et al., 2015; Montes et al., 2015), it is now largely accepted that the Isthmus formation occurred around 3 Ma (O’Dea et al., 2016), and this is the date we used for calibration. The multiple alignment of the mitochondrial genomes was performed using MAFFT v.7 (Katoh and

Standley 2013). The concatenated matrix (15180 bp) was partitioned according to genes and to codon positions and the optimal among-gene partitioning scheme and model choice for dataset in PartitionFinder v.2

(Lanfear et al., 2012) under the Bayesian Information Criterion (BIC). The analysis was carried out under a Bayesian framework using BEAST v.2.5.2

(Bouckaert et al., 2014), under an uncorrelated (lognormal) clock model

157 and a birth-death prior process with incomplete sampling. We run two analyses of 100 million generations each, with sampling every 10000 generation. We used Tracer v.1.6 (Rambaut and Drummond 2007) to inspect the log files, ensuring that the chains have reached convergence and the ESS values for all parameters were greater than 200. We removed the first 10% of the trees from each analysis as burn-in, used LogCombiner v.2.5.2 (Bouckaert et al., 2014) to merge the files with the remaining trees, and a maximum Clade credibility chronogram with mean node heights was computed using TreeAnnotator v.2.5.2 (Bouckaert et al., 2014).

4.4 Results

4.4.1 Morphological identification of Porites

Based on our morphological examinations, the 139 colonies collected were assigned to 11 nominal Porites species: P. annae, P. columnaris, P. echinulata, P. fontanesii, P. harrisoni, P. lobata, P. lutea, P. monticulosa, P. rus, P. solida, P. somaliensis. Three morphologies did not match any original descriptions of Porites and did not correspond to any type specimen examined and are hereafter referred as P. sp. 1 (as in Chapter 2), P. sp. 2

(as in Chapter 2), and P. sp. 3. (Appendix 4.1, 4.2).

4.4.2 Mitochondrial CR, mitochondrial genomes, histone, and rDNA phylogenetic reconstructions

158

Sequence data for the mtCR was obtained for 121 samples (Appendix 4.2).

The final alignment consisted of 1287 bp, with 40 variable sites, 10 of which were singleton sites and 30 parsimony informative. Reads mapping to the

P. lobata mitochondrial genome resulted in a mean of 2737 reads, covering

88% if the reference sequence, at a mean depth of 31 ± 44 s.d. Mapping paired end reads to the P. superfusa histone and rDNA, instead resulted in a mean of 7216 and 14816 reads, covering 92% (mean depth 328 ± 321 s.d) and 97% (mean depth 617 ± 600 s.d) of the reference sequences (Appendix

4.4). The mitochondrial genome alignment (n=127) consisted of 18,647 bp, with 57 variable sites, of which 16 were singleton and 41 parsimony informative. The histone (n=127) and rDNA (n=132) alignments were 5464 bp and 6675 bp long respectively. The histone alignment contained 11 variable sites, of which five singleton sites and six parsimony informative sites, while a total of 76 variable sites, with 40 singleton and 36 parsimony informative sites were present in the rDNA. The BI and ML topologies from the four datasets were congruent, and clustered Porites samples into eight highly supported molecular clades (Clades I to VIII – Fig. 4.1, Appendix 4.5), with the exception of the mtCR were Clade VIII was not resolved (Fig. 4.1 a,

Appendix 4.5 a), and the histone reconstruction were the samples in Clade

III could not be aligned, and the clade could not be recovered (Fig. 4.1 b,

Appendix 4.5 b). Five clades were comprised of samples belonging to a single species, and their monophyly was highly supported in the four

159 reconstructions by both BI posterior probabilities and ML bootstrap values:

Clade I included P. fontanesii, Clade II grouped P. columnaris, Clade III was comprised of samples of P. sp. 1, and Clade VI comprised the genetic material of P. sp. 2, and Clade VIII clustered P. somaliensis. Specimens of P. rus and P. monticulosa were indistinguishably clustered within Clade IV, and all the examined specimens identified as P. annae, P. echinulata, P. harrisoni, P. lobata, P. lutea, and P. solida formed an unresolved species complex in Clade V, thus not adhering to the traditional morphological classification. Finally, one further clade recovered all the samples of P. sp. 3

(Clade VII). Analyses from the four regions concordantly identified Clade I as the most basal clade in the phylogeny of Porites from the Arabian

Peninsula. The mtCR, mitochondrial genome, and rDNA topologies highlighted sister relationships between Clade III and VI, and similarly the mitochondrial genome, histone and rDNA showed sister relationships between Clade IV and VII. The phylogenetic position of Clade II, V and VIII was instead inconsistent among the different reconstructions. Overall, the four topologies were congruent with results showed in Terraneo et al., accepted a, b, with the inclusion of two further clades (Clade VII and VIII).

160

a) mtCR b) Histone

0.98/ II 0.95/ VII VIII V 1/ 93 1/ 62 V 0.5/ 63 0.96/ 1/ 99 1/ - IV

0.98/ 89 0.99/ 79 VII 1/ 80 IV 1/ - 1/ 100 II 1/ 84 1/ 99 VI 0.95/ 61 I III 1/ 100 0.99/ 67 0.98/ 75 VIII I VI Porites superfusa 1/ 100 Porites superfusa 0.2 0.05

c) rDNA d) Mitogenome VI III 1/ 100 0.6/ - VI 1/ 100 III VII 1/ 100 VIII 1/ 80 VII 1/ 92 0.99/ 55 1/ 96 IV 0.78/ 100 1/ 100 IV 1/ 81 VIII 1/ 100 II 1/ 100 1/ 99

V 1/ 100 V 1/ 60 0.99/ 81

1/ 61 I 1/ - 1/ 100 II Porites superfusa I Porites superfusa 1/ 100 0.001 0.002 Porites fontanesii Porites annae-lutea-lobata-echinulata-solida-harrisoni Porites columnaris Porites sp. 2 Porites sp. 1 Porites sp. 3 Porites rus-monticulosa Porites somaliensis

Figure 4.1 Comparison of reconstructed species trees. a) coral mitochondrial CR, b) coral histone region, c) coral rDNA, d) coral mitochondrial DNA. Node values represent BI

161 posterior probabilities and ML bootstrap supports. Roman numbers from I to VIII refer to the assigned clade numbers. Colour codes are explained in the legend

4.4.3 Holobiont and coral phylogenomic analyses

Only the RAxML phylogenetic reconstructions form the “holobiont-max” and the “coral-max” are reported in Fig. 4.2, while reconstruction from the supermatrices obtained with more stringent filters are reported in Appendix

4.6.

The “holobiont-max” and “coral-max” topologies were almost identical, resulting in two well-supported trees, with high bootstrap values at every node (Fig. 4.2 a, b). The specimens were clustered within eight genetic clades in both topologies, consistently with Terraneo et al., accepted a, b, and the mtCR, mitochondrial genomes, histone, and rDNA concatenations. Porites fontanesi (Clade I), P. columanris (Clade II), P. sp.

1(Clade III), and P. sp. 2 (Clade IV), P. sp. 3 (Clade VII) and P. somaliensis

(Clade VIII) were monophyletic, while P. rus and P. monticulosa consistently merged within Clade IV, and there was no resolution in Clade V, where P. annae, P. echinulata, P. harrisoni, P. lobata, P. lutea, and P. solida were indistinguishable (Fig. 4.2 a, b). Similarly, the “holobiont-min” and the “coral- min” analyses, defined the same well supported clades, with the exception of Clade VIII, that was not supported in the reconstruction based on the

162

“coral-min” matrix, and was instead clustered within Clade IV (Appendix

4.6 b).

RAxML topologies for the “holobiont-max” and the “coral-max” datasets also showed congruent relationships among clades, with the exception of Clade II, whose phylogenetic position was conflicting between the “holobiont-max” and “coral-max” reconstructions.

Consistently, the analyses recovered sister relationships between Clade III and VI, and between Clade IV and VII. Within Clade V, the position of the six clustered species did not show any evolutionary or biogeographical pattern, and no stable evolutionary relationships were detected (Fig. 4.2 c, d, Appendix 4.6 a,b).

a) 100 III b) III 100 VI 100 VI 5 2

I I 100 100 8 7 II II 100 100 9 2

VIII VIII 7 9 100 6 3 VII VII 100 100 100

100 100 IV IV 100 100 9 9 100

3 2 9 6

V V 100 100

Goniopora sp Goniopora sp

0.05 0.03

164

Figure 4.2 Comparison of RAxML tree based on a) “holobiont-max” dataset, that allowed for 50% missing data and consisted of 54,108

SNPs, b) “coral-max” dataset, that allowed for 50% missing data and consisted of 96,986 SNPs, c) zoom-in Clade V in the “holobiont- max” dataset, d) zoom-in Clade V in the “coral-max” dataset. Values at nodes represent ML bootstrap supports. Roman numbers from

I to VIII refer to the assigned molecular clade numbers. Colour codes are explained in the legend

4.4.4 Clustering analyses

The coral dataset was further analysed by mean of two clustering methods.

The PCA results for the “coral-max” data set are showed in Fig. 4.3. The first two principal component axes (PC1 and PC2) clearly distinguished Clade

I, Clade II, Clade III, Clade V and Clade VI, while Clade IV, VII, and VIII are nor clearly separated (Fig. 4.3 a). The results from PC3 and PC4 similarly isolated Clade II, Clade III, Clade V, and Clade VI, as well as Clade IV and

VII. In this case Clade I and Clade VIII were closer to each other compared to the results from PC1 and PC2 (Fig 4.3 b). Results from 27 combinations of filtering option yielded similar results, and are reported in Appendix 4.7.

The Admixture analyses of 96986 unlinked SNPs supported K = 5 as the optimal model (CV error = 0.24, Fig. 4.4 a, Appendix 4.8, 4.9). With K = 5,

Clade I, Clade II, Clade V, and Clade VIII were distinguished. Clade VI resulted from admixture of Clade I, II, and VIII; Clade III was not distinct from

Clade VIII, and Clade VII was composed of admixed individuals of Clade

IV and VIII (Fig. 4.4 a). The second most likely clustering was K = 4. In this case, Clade I, Clade II, Clade IV and Clade V were distinguished, while

Clade III, VI, and VIII resulted from admixture of different individuals, and VII was not distinguished from Clade IV (Fig. 4.4 b). The third most likely clustering was K = 8. In this case, a clustering concordant to the phylogenetic reconstructions was expected, yet only five clusters were well separated (Clade I, Clade II, Clade IV, Clade VII, Clade VIII). Clade III was 166 not distinguished form Clade VIII, and Clade V was separated but comprised of admixed individuals (Fig. 4.4 c). Cv Errors calculations from 27 filtering combinations of the coral SNPs mean depth (cv), missing data (ms), and minimum distance (md) and K values ranging from 2 to 12 are reported in Appendix 4.9.

167

Porites fontanesii CLADE I Porites annae-lutea-lobata-echinulata-solida-harrisoni CLADE V Porites columnaris CLADE II Porites sp. 2 CLADE VI Porites sp. 1 CLADE III Porites sp. 3 CLADE VII Porites rus-monticulosa CLADE IV Porites somaliensis CLADE VIII a) b)

0.0 0.0 PC2 PC4 −0.1 −0.2

−0.2

−0.4

0.0 0.1 0.2 0.3 −0.1 0.0 0.1 0.2 0.3 PC3 PC1

Figure 4.3 Principal Component Analysis (PCA) from the “coral-max” dataset that allowed for 50% missing data, and consisted of 96,986 SNPs. a) PCA results based on PC1 (x axis) and PC2 (y axis), b) PC3 (x axis) and PC4 (y axis). Colors in the picture are explained in the legend

168

CLADE I Porites annae-lutea-lobata Porites fontanesii -echinulata-solida-harrisoni CLADE V CLADE II Porites columnaris Porites sp. 2 CLADE VI CLADE III Porites sp. 1 Porites sp. 3 CLADE VII Porites rus-monticulosa CLADE IV Porites somaliensis CLADE VIII CLADE III CLADE I CLADE VI CLADE II CLADE VII CLADE VIII CLADE IV CLADE V a) 1.00

0.75

0.50 K=5

0.25

0.00

1.00 b)

0.75

K=4 0.50

0.25

0.00

1.00 c)

0.75

0.50 K=8

0.25

0.00

169

Figure 4.4 Admixture plots from the “coral-max” dataset that allowed for 50% missing data, and consisted of 96,986 SNPs. From top to bottom: a) K=5, best suggested model, b) K=4, second most probable model, K=8, c) third most probable model. Colour codes are explained in the legend. Clade from I to VIII refer to the assigned molecular clades

4.4.5 Species delimitation analyses

In our analyses, we tested five species delimitation models (models A to E) to identify the number of species within Arabian Porites. The BFD* best supported model followed the phylogenomic results, splitting the specimens into eight species corresponding to eight molecular clades recovered by the phylogenomic recostructions (model (C), MLE = -6426.17,

BF = 0). The second model supported agreed with the current taxonomy recovering 14 species among our samples (model (B) (MLE = 6444.46, BF =

40.58)). Model (D) (six species) was ranked third (MLE = -7060.83, BF =

636.66). The admixture clustering results (model E, five species), was only the fourth best supported model (MLE = -7473.54, BF = 2098.74). Finally, model

(A) lumping all the specimens as one single species, had the lowest MLE value (MLE = 9484.78, BF = 6121.22) (Table 4.1).

The species tree coalescent approach based on the BFD* highest ranking (eight lineages), recovered eight species (Clades I to VIII) consistent with the phylogenomic reconstructions from the mitochondrial and nuclear alignments, and from the “holobiont-max” and the “coral-max” SNPs datasets (Fig. 4.5). All the clades show high node support values, with the

170 exception of Clade II – P. columnaris, whose relationships remain unclear within the phylogeny (Fig. 4.5), as already evidenced by discordance between the “holobiont-max” and “coral-max” SNPs RAxML analysis (Fig.

4.2 a, b). The remaining relationships among clades confirmed the previous reconstructions. Results from the species tree derived from the current taxonomy and 14 lineages, recovered again a total of eight species among the samples, clustering in one single species P. rus and P. monticulosa, and P. annae, P. echinulata, P. harrisoni, P. lobata, P. lutea, and P. solida (see Appendix 4.10 for the species tree reconstructed on the basis of 14 lineages).

Table 4.1 Bayes Factor delimitation (BFD*) results for each analysis using path sampling (PS).

The number of lineages represents the number of putative species included in each analysis. BF values are used to rank species models, relative to the species model with the lowest marginal likelihood (MLE). The model with eight lineages was supported as the best fit model.

171

Model Model Number MLE BF Rank name specifications of lineages A One single 1 -9484.78 6121.22 5 species B Current 14 -6444.46 40.58 2 taxonomy C Molecular 8 -6424.17 - 1 clades D Lumped 6 -7060.83 636.66 3 Clades V, VII, VIII and all other clades E Optimal K 5 -7473.54 2098.74 4 according to Admixture

MLE = Marginal likelihood (loge), BF = Bayes factor (2 * [MLEbest – MLEalternative])

172

CLADE I

CLADE VI

1 1

CLADE III

CLADE II

CLADE VII

0.54 1

CLADE IV

CLADE VIII

CLADE V

Porites fontanesii Porites annae-lutea-lobata-echinulata-solida-harrisoni Porites columnaris Porites sp. 2 Porites sp. 1 Porites sp. 3 Porites rus-monticulosa Porites somaliensis

Figure 4.5 Species tree of Porites. The claudogram represents the posterior distribution of species trees inferred with SNAPP, based on a priori imposition of xx taxa, 1107 unlinked biallelic SNPs and 0% missing data. High colour densities are representative of high topology agreement in the species tree. Values on branches represent posterior probabilities. Colour codes are explained in the legend

173

4.4.6 Divergence time analysis

The time calibrated phylogeny inferred eight strongly supported molecular clades (Clades I to VIII) (posterior probabilities = 1, with exception of one node with posterior probability = 0.99) (Fig. 4.6). Clade I – P. fontanesii was distantly related to all the other Arabian Porites. Indeed, it diverged in the mid Miocene, ~10.5 Ma. The other lineages had more recent origins during the Pliocene and Pleistocene. The Arabian endemic Clades III – P. sp. 1 and

VI – P. sp. 2 split from the other Porites, ~3.3 Ma at the end of the Pliocene, yet the two lineages diverged ~2.2 Ma in the Pleistocene. Clade II – P. columnaris, also originated during the Pliocene (3.1 Ma). The lineage originating P. rus – Clade IV, and P. sp – Clade VII, split ~2.5 Ma, yet the divergence between clade IV and VII is more recent (~1.2 Ma). Finally, the massive species complex in Clade V diverged from P. somaliensis ~2.3 Ma, while Clade V was relatively young, and diverged only ~500 Ka.

175

Figure 4.6 Species tree calibrated chronogram. Purple bars represent 95% highest posterior densities (HPD). Node symbols represent posterior probabilities as explained in the legend. Values at nodes represent estimate time of node divergence. The scale bar represents millions of years and is based on the ICS International Chronostratigraphic Chart

4.5 Discussion

This work used RADseq to clarify evolutionary relationships among 13 morphospecies and one unknown morphology of Porites from the Arabian

Peninsula. We reconstructed topologies from one mitochondrial marker

(mtCR), almost complete mitochondrial genomes, rDNA, and histone regions, for 139 Porites libraries, and we compared SNPs from the entire holobiont and coral datasets. The holobiont and the coral SNPs transcriptomic dataset were highly concordant, highlighting that non- coding regions or other components in the coral holobiont do not mask the phylogenetic signal of the coral protein coding genes. Clustering analyses and species delimitation analyses confirmed that P. fontanesii (Clade I), P. columanris (Clade II), P. sp. 1 (Clade III), P. sp. 2 (Clade VI), and P. somaliensis (Clade VIII) are separate evolving entities as currently described. P. rus and P. monticulosa (Clade IV), and P. annae, P. echinulata, P. harrisoni, P. lobata, P. lutea, and P. solida (Clade V) are instead molecularly indistinguishable and clustered into two species complexes. Finally, our data confirmed that the unknown morphology in

Clade VII (P. sp) from the Gulf of Aden and Oman is molecularly distinct.

4.5.1 Diversity of Porites from the Arabian Peninsula

From the 28 species of Porites that have been reported from the Arabian region in the past 200 years, we identified and collected 13. Moreover, we 177 collected one separate morphology that we could not trace back to any described material. This reflects either insufficient sampling effort and/or inaccurate previous biodiversity reports. Coalescent methods and species delimitation data suggested that Porites diversity from the seas around the

Arabian Peninsula should be reconsidered. The 14 morphospecies that we collected were genetically lumped into eight genetic lineages, one of which identified for the first time in this study. Disagreement between morphology-based taxonomy and phylogenetic systematics is common in corals, and often the result of morphological variability, phenotypic plasticity, incomplete lineage sorting and/ or introgressive hybridization

(Swofford, 1991). Phylogenomic reconstructions from mitochondrial genomes, rDNA, histone regions, and SNPs, concordantly clustered P. rus and P. monticulosa into one single lineage (Clade IV, Fig. 4.1, Fig. 4.2), and these results are corroborated by clustering and species delimitation analyses (Fig. 4.3, 4.4, 4.5). Similar insights are also highlighted by Terraneo et al., accepted a, were one mitochondrial and one nuclear locus, together with symbionts data, failed to separate the two species in the Red

Sea, clustering them into a single genetic lineage. Morphologically, P. rus and P. monticulosa differ at the colony level, being characterized by different growth forms (Appendix 4.1 d, e), i.e. “P. rus normally forms branches whereas P. monticulosa tents to be massive or form plates”

(Veron 2000, vol. 3, pg 314). Nevertheless, colony growth-form has been

178 largely discredited as an informative trait to taxonomy in corals (Romano and Palumbi 1996; Fukami et al., 2004, 2008; Huang et al., 2011), being subject to variability and environmental induced plasticity. For instance, light induced morphological plasticity has been documented in field transplantation experiments in Porites sillimaniani in Japan (Muko et al.,

2000), where different light intensity along the reef slope drove branching or mounding growth forms of the colonies. Our analyses suggested that P. monticulosa should be synonymized with the more ancient P. rus.

Nevertheless, we strongly encourage the inclusion of specimens from additional localities in the Indian and Pacific Oceans, as well as further traits, such as the reproductive biology of the two morphologies, before any formal taxonomic reconsiderations.

Similarly, our genomic data concluded without contrasting signals that Clade V is complex of five morphospecies: P. annae, P. echinulata, P. harrisoni, P. lobata, P. lutea, and P. solida. Morphological variability can often result in apparent genetic polyphyly trough erroneous assignment of alternative morphologies to different species. From a morphological point of view, colonies of P. lobata, P. lutea, and P. solida share a massive growth form, while P. annae and P. harrisoni have a more digitate-like or columnar morphology, and P. echinulata forms mainly encrusting colonies. These six species present corallite level differences (as summarized in Terraneo et al., accepted b), that indeed have been the source of the original description

179 of the distinct morphospecies. Van Oppen et al., 2002 suggested that multiple morphologies may belong to a single randomly mating population, and might represent the result of different interactions among genotypes in different environments. In this scenario multiple morphologies erroneously assigned to the species level, could generate genetic polyphyly as in the case of our species complex in Clade V. An alternative hypothesis is that incomplete lineage sorting and weak genetic drift led to a misleading phylogenetic reconstruction (Queiroz, 1998, 2007). In this scenario, the polyphyly in Clade V could be explained by rapid diversification or recent speciation of the clustered lineages. Furthermore, phylogenetic signals can be hidden by the transfer of genes among divergent lineages undergoing hybridization and introgression (Diekmann et al., 2001; Odorico & Miller,

1997; Van Oppen et al., 2000, 2001). In vitro trials suggested that hybridization in corals may be common (Willis et al., 2006), and has been reported in many genera such as Platygyra, Leptoria and Acropora

(Richards et al., 2008). In the Caribbean, hybridization has been reported between the species A. cervicornis and A. palmata, and backcrossing of the hybrid A. prolifera with the parental species seams to occur at low frequencies too (van Oppen et al., 2000). Combosch et al. (2008) first reported hybridization among Pocilloopra damicornis, P. eydouxi and P. elegans in the Eastern Pacific, and RADseq recently confirmed one-way introgression among these species (Combosch and Vollmer, 2015). Our

180 genome-wide data suggest that the six lineages in Clade V, are a single species, nevertheless, further analyses are necessary to exclude signatures consistent with introgressive hybridization or incomplete lineage sorting.

Species complexes in corals are very common, yet our understanding of these is still vague. Further analyses including wider geographic sampling and all the morphotypes ascribed to this molecular lineage might allow to better valuate these hypotheses.

4.5.2 Biogeography of Porites in the Arabian Peninsula

Understanding how species biogeographic diversity is evolutionary shaped remains a central topic in biology. Porites molecular lineages show a peculiar phylogeographical distribution in the seas around the Arabian

Peninsula, with some lineages widespread around the region, and others whose distribution appears restricted (see Appendix 4.3 in supplementary material).

Among our samples, five lineages are considered Arabian endemics, each with a distinct distribution (Clade I, P. fontanesii; Clade II, P. columnaris; Clade III, P. sp. 1; Clade VI, P. sp. 2; Clade VII, P. sp. 3). Of these, two (Clade I, P. fontanesii; Clade II, P. columnaris) are widely distributed in the Red Sea and Gulf of Aden, but not in the Arabian Gulf; Porites sp. 2

(Clade VI) and P. sp. 3 are restricted to the Gulf of Aden, and P. sp. 1 is so far the only southern Red Sea endemic coral (Terraneo et al., accepted b).

181

High rates of endemism are expected in the Arabian Peninsula. This region has been recognized as an endemism hotspot in the Indian Ocean

(Sheppaed and Sheppard, 1991; Obura 2012; DiBattista 2016), and estimates suggest that 11% of the corals in the basins around the Arabian

Peninsula are indeed endemic (Berumen el al., 2019) (this value does not take into consideration two endemic species described by Terraneo et al., accepted b – Terraneo pers. comm). Although the evolutionary processes that led to the origin of endemism hotspots in peripheral areas remain elusive, the diversity of the habitats and environments, and the complex geological and paleoclimatic history of the seas around the Arabian

Peninsula might have played a key role in shaping the current biodiversity patterns (Sheppard 1992; Bosworth et al., 2005). The Bab Al Mandeb strait is the only connection between the Red Sea and the Gulf of Aden. Limited water exchange driven by the Indian Ocean monsoon system occurs trough this shallow and narrow channel, that provides a potential barrier to genetic exchange between the Red Sea and the rest of the Indian Ocean.

Moreover, a monsoon-driven upwelling system causes major fluctuations in water temperature and nutrients in the Gulf of Aden, limiting reef development in this region as opposed to the oligotrophic biodiverse waters of the Red Sea, and limiting the persistence of only some well adapted species in this region. The Arabian endemic lineages recovered, diverged between 10.5 Ma and 1.2 Ma, consistent with the existence of an

182

Indian Ocean center of origin, during the Miocene in the western and northern Indian Ocean (Bowen et al., 2013; Obura, 2016). From a geological perspective the Red Sea reef system originated around 5 Ma

(Siddal et al., 2003), and hypothesis exist that extreme environmental conditions might have caused mass extirpation from the Red Sea during the Pleistocene ice ages (Sheppard et al., 1992) until as early as 20.000 years ago. As shown in Chapter 2, P. sp. 1 and P. sp. 2 present a disjunct but adjacent distribution in the southern Red Sea and Gulf of Aden. In the present work we date their divergence around 2.2 Ma. Their biogeography and phylogenetic position indicate that they are vicariant species, that probably originated during the Pleistocene glaciation events, from fragmentation of a previously widespread lineage.

Only one out of eight Clades of Porites is widespread throughout the

Arabian Peninsula (Clade V). According to our analyses, this lineage is relatively young and diverged 500 Ka, and is also the most diverse in terms of morphology. This might indicate the presence of species of recent divergence in the molecular lineage, that would still need to acquire fixed genomic signatures.

Only 16% of the corals that inhabit the Arabian basins are found in the Arabian Gulf, where water temperature oscillations as high as 20ºC occur during the year, and high nutrient input from the Gulf of Oman make its reef systems some of the most extreme for corals survival in the tropics.

183

The absence of the majority of the lineages from the Arabian Gulf can be moreover attributed to the recent origin of this shallow sea, dated only 14 kya (Lamberg, 1996), thus is not surprising that only one lineage out of eight is actually found in the Arabian Gulf.

To conclude, Clade V, together with Clade IV (P. rus - P. monticulosa) and Clade VIII (P. somaliensis) have been reported to occur at other localities in the Indo-Pacific (Veron, 2000), yet their distribution needs to be re-evaluated in light of a similar morpho-molecular approach.

4.6 Conclusions

This work harnesses on the power of NGS coupled with phylogenomics, clustering, and species delimitation methods, to clarify the diversity and evolutionary relationships of the coral genus Porites in the seas around the

Arabian Peninsula. Our results demonstrate the utility of reduced representation sequencing in resolving species boundaries in recalcitrant taxa, highlighting a deep gap in the understanding of coral biodiversity patterns in the Arabian region.

Major gaps remain in our understanding of biodiversity, biogeography, evolution, and ecology of Porites, and the present work calls for an urgent taxonomic revision of one of the most important reef building corals. Further studies encompassing ecological, symbiont association, and reproductive data will be necessary to determine the

184 presence of functional differences and reproductive isolation mechanisms among the morphologies lumped in the species complexes in Clade IV and

185

4.7 References

Alexander, D. H., Novembre, J., & Lange, K. (2009). Fast model-based estimation of ancestry in unrelated individuals. Genome Research, 19(9), 1655-1664.

Arnold, M. L. (1997). Natural Hybridization and Evolution. Oxford University Press on Demand.

Bacon, C. D., Silvestro, D., Jaramillo, C., Smith, B. T., Chakrabarty, P., & Antonelli, A. (2015). Biological evidence supports an early and complex emergence of the Isthmus of Panama. Proceedings of the National Academy of Sciences, 112(19), 6110-6115.

Baxter, S. W., Davey, J. W., Johnston, J. S., Shelton, A. M., Heckel, D. G., Jiggins, C. D., & Blaxter, M. L. (2011). Linkage mapping and comparative genomics using next-generation RAD sequencing of a non-model organism. PloS One, 6(4), e19315.

Bernard, H.M. (1903). The family Poritidae, I. The genus Goniopora. Catalogue of the madreporarian corals in the British Museum. Natural History 4, 1–166.

Bosworth, W., Huchon, P., & McClay, K. (2005). The Red Sea and Gulf of Aden basins. Journal of African Earth Sciences, 43(1-3), 334-378.

Bouckaert, R., Heled, J., Kühnert, D., Vaughan, T., Wu, C. H., Xie, D., ... & Drummond, A. J. (2014). BEAST 2: a software platform for Bayesian evolutionary analysis. PLoS Computational Biology, 10(4), e1003537.

186

Bowen, B. W., Rocha, L. A., Toonen, R. J., & Karl, S. A. (2013). The origins of tropical marine biodiversity. Trends in Ecology & Evolution, 28(6), 359-366.

Bryant, D., Bouckaert, R., Felsenstein, J., Rosenberg, N. A., & RoyChoudhury, A. (2012). Inferring species trees directly from biallelic genetic markers: bypassing gene trees in a full coalescent analysis. Molecular Biology and Evolution, 29(8), 1917-1932.

Cariou, M., Duret, L., & Charlat, S. (2013). Is RAD-seq suitable for phylogenetic inference? An in-silico assessment and optimization. Ecology and Evolution, 3(4), 846-852.

Claereboudt, M. R. (2006). Reef corals and coral reefs of the Gulf of Oman. Historical Association of Oman, Oman. 344: pp.

Combosch, D. J., Guzman, H. M., Schuhmacher, H., & Vollmer, S. V. (2008). Interspecific hybridization and restricted trans-Pacific gene flow in the Tropical Eastern Pacific Pocillopora. Molecular Ecology, 17(5), 1304-1312.

Combosch, D. J., & Vollmer, S. V. (2015). Molecular Phylogenetics and Evolution Trans-Pacific RAD-Seq population genomics confirms introgressive hybridization in Eastern Pacific Pocillopora corals. Molecular Phylogenetics and Evolution, 88, 154–162. doi:10.1016/j.ympev.2015.03.022

Concepcion, G. T., Crepeau, M. W., Wagner, D., Kahng, S. E., & Toonen, R. J. (2008). An alternative to ITS, a hypervariable, single-copy nuclear intron in corals, and its use in detecting cryptic species within the octocoral genus Carijoa. Coral Reefs, 27(2), 323-336.

Daly, M., Fautin, D. G., & Cappola, V. A. (2003). Systematics of the hexacorallia (Cnidaria: Anthozoa). Zoological Journal of the Linnean Society, 139(3), 419-437.

Danecek, P., Auton, A., Abecasis, G., Albers, C. A., Banks, E., DePristo, M. A., ... & McVean, G. (2011). The variant call format and VCFtools. Bioinformatics, 27(15), 2156-2158.

DiBattista, J. D., Wilcox, C., Craig, M. T., Rocha, L. A., & Bowen, B. W. (2011). Phylogeography of the Pacific Blueline Surgeonfish, Acanthurus nigroris, reveals high genetic connectivity and a cryptic endemic species in the Hawaiian Archipelago. Journal of Marine Biology, 2011, 17.

DiBattista, J. D., Berumen, M. L., Gaither, M. R., Rocha, L. A., Eble, J. A., Choat, J. H., ... Bowen, B. W. (2013). After continents divide: comparative

187 phylogeography of reef fishes from the Red Sea and Indian Ocean. Journal of Biogeography, 40(6), 1170–1181.

Diekmann, O. E., Bak, R. P. M., Stam, W. T., & Olsen, J. L. (2001). Molecular genetic evidence for probable reticulate speciation in the coral genus Madracis from a Caribbean fringing reef slope. Marine Biology, 139(2), 221– 233. doi:10.1007/s002270100584.

Dimond, J., Gamblewood, S., & Roberts, S. (2017). Genetic and epigenetic insight into morphospecies in a reef coral. Molecular Ecology. 26(19), 5031- 5042.

Drummond, A. J., & Bouckaert, R. R. (2015). Bayesian evolutionary analysis with BEAST. Cambridge University Press.

Forsskål, P. (1775). Descriptiones animalium avium: amphibiorum, piscium, insectorum, vermium; quae in itinere orientali observavit Petrus Forskål... Post mortem auctoris edidit.

Forsman, Z., Knapp, I., & Tisthammer, K. (2017). Coral hybridization or phenotypic variation? Genomic data reveal gene flow between Porites lobata and P. compressa. Molecular Phylogenetics.

Gaither, M. R., Toonen, R. J., Robertson, D. R., Planes, S., & Bowen, B. W. (2010). Genetic evaluation of marine biogeographical barriers:

188 perspectives from two widespread Indo-Pacific snappers (Lutjanus kasmira and Lutjanus fulvus). Journal of Biogeography, 37(1), 133-147.

Gaither, M. R., Jones, S. A., Kelley, C., Newman, S. J., Sorenson, L., & Bowen, B. W. (2011). High connectivity in the deepwater snapper Pristipomoides filamentosus (Lutjanidae) across the Indo-Pacific with isolation of the Hawaiian Archipelago. PLoS One, 6(12), e28913.

Gravier, C. (1910). Sur quelques formes nouvelles de Madréporaires de la baie de Tadjourah. Bulletin du Muséum National d’Histoire Naturelle, 16, 273-276.

Hellberg, M. E. (2006). No variation and low synonymous substitution rates in coral mtDNA despite high nuclear variation. BMC Evolutionary Biology, 6, 24. doi:10.1186/1471-2148-6-24.

Hipp, A. L., Eaton, D. A., Cavender-Bares, J., Fitzek, E., Nipper, R., & Manos, P. S. (2014). A framework phylogeny of the American oak clade based on sequenced RAD data. PloS One, 9(4), e93975.

Huang, D., Meier, R., Todd, P. a., & Chou, L. M. (2008). Slow mitochondrial COI sequence evolution at the base of the metazoan tree and its implications for DNA barcoding. Journal of Molecular Evolution, 66(2), 167– 174. doi:10.1007/s00239-008-9069-5.

Johnston, E. C., Forsman, Z. H., Flot, J. F., Schmidt-Roach, S., Pinzón, J. H., Knapp, I. S., & Toonen, R. J. (2017). A genomic glance through the fog of plasticity and diversification in Pocillopora. Scientific Reports, 7(1), 5991.

189

Kass, R. E., & Raftery, A. E. (1995). Bayes factors. Journal of the American Statistical Association, 90(430), 773-795.

Katoh, K., & Standley, D. M. (2013). MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Molecular Biology and Evolution, 30(4), 772–780.

Klunzinger C.B. (1879a) Die Korallenthiere des Rothen Meeres, 2. Theil: Die Steinkorallen. Erster Abschnitt: Die Madreporaceen und Oculinaceen. Gutmann, Berlin, 88pp.

Klunzinger C.B. (1879b) Die Korallenthiere des Rothen Meeres, 3. Theil: Die Steinkorallen. Zweiter Abschnitt: Die Asteraeaceen und Fungiaceen. Gutmann, Berlin, 100pp.

Knapp, I., Puritz, J., Bird, C., Whitney, J., Sudek, M., Forsman, Z., & Toonen, R. (2016). ezRAD—an accessible next-generation RAD sequencing protocol suitable for non-model organisms_v3. 1. Protocols. io Life Sciences Protocol Repository.

Lamarck, J.B.P. (1816) Histoire naturelle des Animaux sans Vertebres,̀ presentant́ les caracteres̀ geń eraux́ et particuliers de ces animaux, leur distribution, leurs classes, leurs familles, leurs genres, et la citation des principales especes̀ qui s'y rapportent; preć ed́ eé d'une Introduction offrant la Determinatioń des caracteres̀ essentiels de l'Animal, sa distinction du Veǵ etaĺ et des autres corps naturels, enfin, l'exposition des principes fondamentaux de la Zoologie. Detervillé & Verdiere,̀ Paris.

Leaché, A. D., Fujita, M. K., Minin, V. N., & Bouckaert, R. R. (2014). Species delimitation using genome-wide SNP data. Systematic Biology, 63(4), 534-5.

Lischer, H. E., & Excoffier, L. (2011). PGDSpider: an automated data conversion tool for connecting population genetics and genomics programs. Bioinformatics, 28(2), 298-299.

Liu, L., Yu, L., Pearl, D. K., & Edwards, S. V. (2009). Estimating species phylogenies using coalescence times among sequences. Systematic Biology, 58(5), 468-477.

Márquez, L., Miller, D., & MacKenzie, J. (2003). Pseudogenes contribute to

190 the extreme diversity of nuclear ribosomal DNA in the hard coral Acropora. Molecular Biology. 20(7), 1077-1086.

Miller, M. A., Pfeiffer, W., & Schwartz, T. (2010, November). Creating the CIPRES Science Gateway for inference of large phylogenetic trees. In 2010 Gateway Computing Environments Workshop (GCE) (pp. 1-8).

McFadden, C. S., Sánchez, J. A., & France, S. C. (2010). Molecular phylogenetic insights into the evolution of Octocorallia: a review. Integrative and Coparative Biology, 50(3), 389-410.

McFadden, C. S., Haverkort-Yeh, R., Reynolds, A. M., Halàsz, A., Quattrini, A. M., Forsman, Z. H., ... & Toonen, R. J. (2017). Species boundaries in the absence of morphological, ecological or geographical differentiation in the Red Sea octocoral genus Ovabunda (Alcyonacea: Xeniidae). Molecular Phylogenetics and Evolution, 112, 174-184.

Montes, C., Cardona, A., McFadden, R., Morón, S. E., Silva, C. A., Restrepo- Moreno, S., ... & Bayona, G. A. (2012). Evidence for middle Eocene and younger land emergence in central Panama: implications for Isthmus closure. The Geological Society Bulletin, 124(5-6), 780-799.

Muko, S., Kawasaki, K., Sakai, K., Takasu, F., & Shigesada, N. (2000). Morphological plasticity in the coral Porites sillimaniani and its adaptive significance. Bulletin of Marine Science, 66(1), 225-239.

Obura, D. (2012). The diversity and biogeography of Western Indian Ocean reef-building corals. PloS One, 7(9), e45013.

Obura, D. (2016). An Indian Ocean centre of origin revisited: Palaeogene and Neogene influences defining a biogeographic realm. Journal of Biogeography, 43(2), 229-242.

O’Dea, A., Lessios, H. A., Coates, A. G., Eytan, R. I., Restrepo-Moreno, S. A., Cione, A. L., ... & Stallard, R. F. (2016). Formation of the Isthmus of Panama. Science Advances, 2(8): e1600883.

Odorico, D., & Miller, D. (1997). Variation in the ribosomal internal transcribed spacers and 5.8 S rDNA among five species of Acropora (Cnidaria; Scleractinia): patterns of variation consistent with. Molecular Biology and Evolution. 14(5), 465-473.

191

Pante, E., Abdelkrim, J., Viricel, A., Gey, D., France, S. C., Boisselier, M. C., & Samadi, S. (2015). Use of RAD sequencing for delimiting species. Heredity, 114(5), 450.

Paz-García, D. A., Hellberg, M. E., García-de-León, F. J., & Balart, E. F. (2015). Switch between morphospecies of Pocillopora corals. The American Naturalist, 186(3), 434-440.

Pichon, M., Benzoni, F., Chaineu, C., & Dutrieux, E. (2010). Field Guide to the hard corals of the southern coast of Yemen. Collection Parthénope, 1-256.

Pollard, D. A., Iyer, V. N., Moses, A. M., & Eisen, M. B. (2006). Widespread discordance of gene trees with species tree in Drosophila: evidence for incomplete lineage sorting. PLoS Genetics, 2(10), e173.

Purcell, S., Neale, B., Todd-Brown, K., Thomas, L., Ferreira, M. A., Bender, D., ... & Sham, P. C. (2007). PLINK: a tool set for whole-genome association and population-based linkage analyses. The American Journal of Human Genetics, 81(3), 559-575.

Puritz, J. B., Hollenbeck, C. M., & Gold, J. R. (2014). dDocent: a RADseq, variant-calling pipeline designed for population genomics of non-model organisms. PeerJ, 2, e431.

Queiroz, K. De. (1998). The General Lineage Concept of Species, Species Chteria, and the Process of Speciation A Conceptual Unification and Terminological Recommendations.

Queiroz, K. De. (2007). Species concepts and species delimitation. Systematic Biology.

Richards, Z. T., van Oppen, M. J., Wallace, C. C., Willis, B. L., & Miller, D. J. (2008). Some rare Indo-Pacific coral species are probable hybrids. PLoS One, 3(9), e3240.

Romano, S. L., & Palumbi, S. R. (1996). Evolution of scleractinian corals inferred from molecular systematics. Science, 271(5249), 640-642.

Rambaut, A., & Drummond, A. J. (2007). Tracer v1. 6 http://beast. bio. ed. ac. uk.

192

Ronquist, F., Teslenko, M., Van Der Mark, P., Ayres, D. L., Darling, A., Höhna, S., ... & Huelsenbeck, J. P. (2012). MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Systematic biology, 61(3), 539-542.

Rubin, B. E., Ree, R. H., & Moreau, C. S. (2012). Inferring phylogenies from RAD sequence data. PloS one, 7(4), e33394.

Sánchez, J. A., & Dorado, D. (2008, July). Intragenomic ITS2 variation in Caribbean seafans. In Proc 11th International Coral Reef Symposium (pp. 1383-1387).

Scheer, G. & Pillai, C.S.G. (1983) Report on the stony corals from the Red Sea. Zoologica, (Stuttgart) 133, 1-198.

Shearer, T., & Coffroth, M. (2008). DNA BARCODING: Barcoding corals: limited by interspecific divergence, not intraspecific variation. Molecular Ecology Resources. 8(2), 247-255.

Sheppard, C. R., & Sheppard, A. S. (1991). Corals and Coral Communities of Arabia (Vol. 12).

Sheppard, C., Price, A., & Roberts, C. (1992). Marine ecology of the Arabian region: patterns and processes in extreme tropical environment.

Sinniger, F., Reimer, J. D., & Pawlowski, J. (2008). Potential of DNA sequences to identify zoanthids (Cnidaria: Zoantharia). Zoological Science, 25(12), 1253-1261.

Stamatakis, A. (2014). RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics, 30(9), 1312-1313.

Stobie, C. S., Cunningham, M. J., Oosthuizen, C. J., & Bloomer, P. (2018). Finding stories in noise: Mitochondrial portraits from RAD data. Molecular Ecology Resources. 19(1), 191-205.

Terraneo, T. I., Arrigoni, R., Benzoni, F., Forsman, Z. H., & Berumen, M. L. (2018a). Using ezRAD to reconstruct the complete mitochondrial genome of Porites fontanesii (Cnidaria: Scleractinia). Mitochondrial DNA Part B, 3(1), 173-174.

Terraneo, T. I., Arrigoni, R., Benzoni, F., Forsman, Z. H., & Berumen, M. L. (2018b). The complete mitochondrial genome of Porites harrisoni (Cnidaria:

193

Scleractinia) obtained using next-generation sequencing. Mitochondrial DNA Part B, 3(1), 286-287.

Terraneo, T.I., Fusi, M., Hume, C.C., Arriogni, R., Voolstra, C.R., Benzoni, F., Forsman Z.H., Berumen, M.L. (a) Environmental latitudinal gradients and host specificity shape Symbiodiniaceae distribution in Red Sea Porites corals. Accepted in Journal of Biogeography.

Terraneo, T.I., Benzoni, F., Baird, A.H., Arrigoni, R., Berumen, M.L. (b) Morphology and molecules reveal two new species of Porites (Scleractinia, Poritidae) from the Red Sea and the Gulf of Aden. Accepted in Systematics and Biodiversity.

Tisthammer, K. H., & Richmond, R. H. (2018). Corallite skeletal morphological variation in Hawaiian Porites lobata. Coral Reefs, 37(2), 445-456.

Toonen, R. J., Puritz, J. B., Forsman, Z. H., Whitney, J. L., Fernandez-Silva, I., Andrews, K. R., & Bird, C. E. (2013). ezRAD: a simplified method for genomic genotyping in non-model organisms. PeerJ, 1, e203.

Veron, J. E. N. (2000). Corals of the World (No. C/593.6 V4).

Van Oppen, M. V., Willis, B. L., Vugt, H. V., & Miller, D. J. (2000). Examination of species boundaries in the Acropora cervicornis group (Scleractinia, Cnidaria) using nuclear DNA sequence analyses. Molecular Ecology, 9(9), 1363-1373.

Van Oppen, M. J., Willis, B. L., Van Rheede, T., & Miller, D. J. (2002). Spawning times, reproductive compatibilities and genetic structuring in the Acropora aspera group: evidence for natural hybridization and semi-permeable species boundaries in corals. Molecular Ecology, 11(8), 1363-1376.

Willis, B. L., van Oppen, M. J., Miller, D. J., Vollmer, S. V., & Ayre, D. J. (2006). The role of hybridization in the evolution of reef corals. Annual Review of Ecology, Evolution and Systematics, 37, 489-517.

Willis, S. C., Hollenbeck, C. M., Puritz, J. B., Gold, J. R., & Portnoy, D. S. (2017). Haplotyping RAD loci: an efficient method to filter paralogs and account for physical linkage. Molecular Ecology Resources, 17(5), 955-965.

194

5. CHAPTER 5: ENVIRONMENTAL LATITUDINAL GRADIENTS AND HOST SPECIFICITY SHAPE SYMBIODINIACEAE DISTRIBUTION IN RED SEA PORITES CORALS

Tullia I. Terraneoa,b*, Marco Fusia, Benjamin C. C. Humea, Roberto Arrigonia,

Christian R. Voolstraa, Francesca Benzonic, Zac H. Forsmand, Michael L.

Berumena aRed Sea Research Center, Division of Biological and Environmental

Science and Engineering, King Abdullah University of Science and

Technology, Thuwal 23955-6900, Saudi Arabia

bARC Centre of Excellence for Coral Reef Studies, James Cook University,

Townsville 4811, QLD, Australia

cDepartment of Biotechnologies and Bioscience, University of Milano-

Bicocca, Milan, 20126, Italy; dHawaii Institute of Marine Biology, Kaneohe

96744, HI, USA

Published in Journal of Biogeography

Terraneo, T.I., Fusi, M., Hume, C.C., Arriogni, R., Voolstra, C.R., Benzoni, F., Forsman Z.H., Berumen, M.L. (2019). Environmental latitudinal gradients and host specificity shape Symbiodiniaceae distribution in Red Sea Porites corals.

195

This chapter is edited from the original version of the paper

196

5.1 Abstract

To assess the diversity of algal symbionts of the family Symbiodiniaceae associated with the coral genus Porites in the Red Sea, and to test for host- specificity and environmental variables driving biogeographical patterns of algal symbiont distribution, eighty Porites coral specimens were collected along the Saudi Arabian Red Sea coast. Species boundaries were assessed morphologically and genetically (putative Control Region, ITS region).

Community composition of symbiotic dinoflagellates of the family

Symbiodiniaceae was also assessed. Using the ITS2 marker with the

SymPortal framework, Symbiodiniaceae data at the genus, majority ITS2 sequence, and ITS2 type profile were used to assess symbiont diversity and distribution patterns. These were analyzed in relation to coral host diversity, geographic location, and environmental variables.

Among the 80 Porites samples, 10 morphologies were identified. These corals were clustered into five lineages (clades I to V) by each of the markers independently. Clades I, II, and III each comprised of a single

Porites morphology, while clades IV and V contained up to five distinct morphologies. The diversity of Symbiodiniaceae associated with Porites was high and latitudinal differentiation was observed. In particular, a shift from a Cladocopium- to a Durusdinium-dominated community was found along the north-south gradient. Symbiont diversity showed patterns of geographic-specific association at Symbiodiniaceae genus, majority ITS2

197 sequence, and ITS2 type profile level. Specific associations with host- genotypes (but not morphological species) were also recovered when considering Symbiodiniaceae majority ITS2 sequence and ITS2 type profiles.

This study provides the first large scale molecular characterization of

Symbiodiniaceae communities associated with Porites corals from the

Saudi Arabian Red Sea. The use of intragenomic diversity data enabled the resolution of host-symbiont specificity and biogeographical patterns of distribution, previously unachievable with the ITS2 marker alone. Finally, correlation among symbiont diversity and Red Sea environmental gradients was documented.

Keywords: Latitudinal gradient, ITS2, Next-Generation Sequencing,

Scleractinia, symbiosis, SymPortal

5.2 Introduction

Shallow water tropical and subtropical corals rely on their association with microscopic endosymbiotic dinoflagellates of the family Symbiodiniaceae.

Providing the corals with up to 95% of their nutritional needs (Falkowski et al., 1984), these photosynthetic symbionts are crucial for the growth and functioning of coral reefs (Hughes et al., 2017, 2018; Sampayo et al., 2016).

Symbiodiniaceae diversity in reef ecosystems is high and the specificity and variability of the associations that scleractinian-hosts form

198 with these symbionts have proven to confer ecological advantages to corals under different ecological conditions (Berkelmans & van Oppen,

2006; LaJeunesse et al., 2010; Rosic et al., 2015; Hume et al., 2016). These associations influence the geographical zonation patterns of corals at large and small scales and provide the corals with different tolerance to light intensity (Baker, 2001) and temperature (Rowan & Knowlton, 1995; Glynn et al., 2001; Berkelmans & van Oppen, 2006). Indeed, different

Symbiodiniaceae-host interactions impact the corals’ susceptibility to bleaching events (for a definition of bleaching see van Oppen & Lough,

2018). The potential of corals to “shuffle” (i.e., replacement of dominant population by a background resident population) or “switch” (i.e., the exogenous uptake of a different population from the environment) their symbiont communities towards more tolerant ones could play a major role towards ecological resilience of coral reefs under predicted climate change scenarios (Baker et al., 2004; Sampayo et al., 2008; Kemp et al.,

2014). Different patterns of host-symbiont associations have been documented in response to latitudinal, longitudinal, and environmental gradients, for various geographic locations and host taxa (Oliver & Palumbi,

2009; Huang et al., 2011; Keshavmurthy et al., 2014; Hume et al., 2015, 2016;

Ziegler et al., 2015; Tonk et al., 2017). Nevertheless, our knowledge of the specificity and diversity of these associations is still poor, limiting our understanding of the ecological benefits that different associations provide

199

(LaJeunesse et al., 2018). Despite the lack of a single commonly accepted molecular marker for Symbiodiniaceae diversity typing, the Internal

Transcribed Spacer II (ITS2) region is currently the most widely used barcode locus within the family (Hume et al., 2016, 2018; Smith et al., 2017). The multicopy nature of this marker means that both intragenomic and intergenomic sequence variants may be present within any single coral

Symbiodiniaceae sample. Distinguishing between these sequence variant sources is difficult and therefore the majority of commonly used analytical approaches aim to collapse the confounding intragenomic diversity (Arif et al., 2014; Cunning et al., 2017). However, the intragenomic diversity harboured within every Symbiodiniaceae genome may be taxonomically informative. Gel-based techniques have made use of this diversity to improve taxonomic resolution for more than 15 years (LaJeunesse, 2002).

Most recently, the SymPortal analytical framework (symportal.org; github.com/SymPortal/SymPortal_framework) has been developed to make use of this intragenomic diversity for resolving genetic delineations using next generation sequencing (hereafter “NGS”) ITS2 data. By leveraging the informative nature of Symbiodiniaceae intragenomic diversity, finer-scale resolutions of genetic delineations are now possible; these delineations far surpass what were previously achievable with the ITS2 marker. As one of the hottest and most saline regions of the ocean, the Red

Sea represents an ideal setting to explore Symbiodiniaceae diversity in a

200 system where natural conditions are already exceeding the thresholds typical for Scleractinia persistence elsewhere in the world. Moreover, due to limited freshwater inflow, low circulatory exchange with the Indian

Ocean, and high evaporation rates (~2m year-1), the Red Sea displays extreme latitudinal environmental gradients (Trommer et al., 2009; Raitos et al., 2013). In particular, the sea surface temperature (SST) maxima ranges from 26°C (± 1.1°C) in the north to 31.3°C (± 1.1°C) in the south (Osman et al., 2018), the primary productivity increases from northern oligotrophic waters to southern nutrient rich waters (Raitos et al., 2013), and the salinity drops from 41 in the north to 36 in the south (Ngugi et al., 2012). The Red

Sea is also recognized as a marine biodiversity hotspot, harbouring more than 2000 species of fish and 50 genera of corals (Briggs & Bowen, 2012;

Berumen et al., 2013; DiBattista et al., 2016a, b). Hermatypic corals of the cosmopolitan genus Porites are among the most abundant, widespread, and diverse zooxanthellate scleractinians in the Red Sea (Sheppard &

Sheppard, 1991). Up to 15 species of Porites have been reported from the region following traditional morphology-based classifications (Sheer & Pillai,

1983; Sheppard & Sheppard, 1991; Veron, 2000), however, species boundaries within the genus remain unresolved (Forsman et al., 2009, 2015,

2017; Prada et al., 2014; Dimond et al., 2017). Besides their role as fundamental reef builders, Porites corals are also among the most resistant corals to increasing water temperatures (LaJeunesse et al., 2003). During

201 both the 2010 and the 2016 bleaching events, Porites was among the least affected coral genera in the Red Sea, with less than 40% of the resident population showing signs of bleaching (Furby et al., 2013; Monroe et al.,

2018).

In this work, we applied NGS to explore the diversity of the

Symbiodiniaceae community associated with Porites in the Saudi Arabian

Red Sea along a 12° latitudinal gradient. We tested for host-specificity as well as geographic and environmental variables driving biogeographical patterns of algal symbiont distribution, with the aim of providing a better understanding of the ecological resilience of Red Sea coral reefs.

5.3 Material and Methods

5.3.1 Sampling and identification

A total of 80 Porites coral colonies were collected at seven coastal localities along the Saudi Arabian Red Sea between 2013 and 2017 (Fig. 5.1,

Appendix 5.1). Each coral colony was imaged in the field using a Canon

G15 camera while SCUBA diving. A fragment of approximately 10cm3 was sampled from each scleractinian colony using hammer and chisel. Once in the laboratory, a subsample of 1cm3 was taken from each specimen and preserved in 99% ethanol for further molecular analyses. The rest of the coral was bleached in sodium hypochlorite for 48 hours to remove fresh tissue

202 and was air dried for further morphological observations. Dried skeletons were imaged using a Canon G15 and were used for traditional morphology-based species identification.

Specimens were identified based on skeletal morphology features of the corallum and corallites following Sheer & Pillai (1983), Sheppard &

Sheppard (1991), and Veron (2000).

Figure 5.1 Sampling location and specimen collection overview (a) Red Sea map with sampling localities. (b) to (j) Porites morphologies encountered among our 80 collected samples. (b) Porites fontanesii, (c) Porites columnaris, (d) Porites sp. 1, (e) Porites rus, (f)

Porites monticulosa, (g) Porites lutea, (h) Porites lobata, (i) Porites annae, (j) Porites echinulata, (k) Porites solida.

203

5.3.2 Coral host DNA extraction and PCR amplification

Genomic DNA was extracted from the coral samples using the DNeasy®

Blood & Tissue Kit (Qiagen Inc., Hilden, Germany). Extracted DNA was quantified with AccuBlue High Sensitivity dsDNA quantitation kit (Biotium,

Inc.) using a Qubit® fluorometer (ThermoFisher Scientific, Inc, Wilmington,

USA). Details on PCR amplifications and primers are provided in Appendix

5.2.

Forward and reverse sequences were assembled and edited using

Sequencher 5.3 (Gene Codes Corp., Ann Arbor, USA). Nuclear sequences that showed intra-individual polymorphisms were phased using Phase

(Stephens et al., 2001; available online at http://stephenslab.uchicago.edu/ software.html) and SeqPHASE (Flot,

2010; available online at http://seqphase. mpg.de/seqphase/) when the alleles showed the same length, and using Champuru (Flot, 2007; available online at http://seqphase.mpg. de/champuru/) when the two predominant alleles were of different lengths. Alignments were performed using the E-INS-i option in MAFFT 7.130b (Katoh & Standley, 2013) and manually checked using BioEdit 7.2.5 (Hall, 1999). All the produced sequences are deposited at GenBank (see Appendix 5.1).

5.3.3 Coral host phylogeny reconstructions

204

For each marker, the best fit substitution model was calculated using

Information Criterion (AIC) and the Bayesian Information Criterion (BIC). For the mtCR, PartitionFinder selected the evolutionary model GTR+I, while for the ITS region the GTR+I+G was the most suitable model. Phylogenetic relationships were reconstructed under two different criteria: Bayesian inference (BI) using MrBayes 3.2.6 (Ronquist et al., 2012) and Maximum

Likelihood (ML) using RAxML 2 (Stamatakis, 2014). The CIPRES server (Miller et al., 2012) was used to run both the BI and ML analyses. BI runs were performed using four Markov Chain Monte Carlo (MCMC) chains for 10 million generations, saving one tree every 100 generations. The tree searches were stopped when all parameters reached stationarity for effective sampling size and unimodal posterior distribution using Tracer 1.6

(Rambaut et al., 2014). The first 25% trees sampled were discarded as burn- in following indications by Tracer. ML topologies were obtained under the default parameters shown on the CIPRES server with a multiparametric bootstrap analyses of 1,000 bootstrap replicates.

5.3.4 Symbiodiniaceae MiSeq sequencing library preparation

205

Symbiodiniaceae types were characterized using PCR amplification of the

ITS2 region (ITS2) for the Illumina MiSeq platform in the KAUST Bioscience

Core Laboratory (Thuwal, Saudi Arabia). Details on PCR amplifications, library preparation, and sequencing are provided in Appendix 5.2.

5.3.5 Symbiodiniaceae MiSeq data processing

Symbiodiniaceae NGS ITS2 data were analysed using the SymPortal framework (symportal.org; Hume et al., 2018) by submitting paired fastq.gz files directly to the framework. A standardized quality control (QC) of sequences was conducted as part of the submission. Briefly, the standard

SymPortal QC is conducted using mothur 1.39.5 (Schloss et al., 2009), the

BLAST+ suite of executables (Camacho et al., 2009), and minimum entropy decomposition (MED; Eren et al., 2015). The SymPortal scripts for data submission and QC are available at symportal.org. The Symbiodiniaceae analyses conducted in this study were conducted on the standard

SymPortal output. This output contains two count tables: one provides sequence abundances listed by sample (see Appendix 5.3), and the second provides ITS2 type profile abundances listed by sample (see

Appendix 5.3). In the second count table, alongside the ITS2 type profile abundances, different hierarchical levels are given for each of the identified Symbiodiniaceae genotypes. Specifically, for this analysis, the clade (genus), majority ITS2 sequence (most abundant ITS2 sequence), and

206

ITS2 type profile (representative of putative taxa) were used.

5.3.6 Environmental data

A total of four environmental variables were considered in this study, namely chlorophyll-a (Chl-a), sea surface temperature (SST), particulate organic carbon (POC), and salinity. The first three variables were gathered for each of the seven sampling localities of the Red Sea from 2010 to 2017 from the National Aeronautic and Space Administration (NASA) Giovanni website (Acker & Leptoukh, 2007), developed and maintained by the NASA

Goddard Earth Sciences Data and Information Services Center

(https://giovanni.gsfc.nasa.gov/giovanni/). In particular, monthly average

Chl-a, SST, and POC data were derived from 4 km resolution data from the

Moderate Resolution Imaging Spectroradiometer (MODIS) Aqua database.

For each sampling location, annual averages were then calculated.

Salinity for each sampling site was calculated from March 2010 (winter) data in Ngugi et al., (2012). All environmental data are listed in Appendix

5.3.

5.3.7 Statistical analyses

Statistical analyses were performed using PRIMER 6.1.15 (Primer-E, Plymouth,

UK) with the add-on PERMANOVA+ package (Anderson et al., 2008).

Permutational multivariate analysis of the variance (PERMANOVA) was

207 performed on Bray-Curtis distance matrices to test for significant differences in Symbiodiniaceae community composition along the Red Sea latitudinal gradient, and to test for morphological and molecular host-specificity with

Porites corals. In particular, the three Symbiodiniaceae input datasets

(genus level, ITS2 majority sequence, ITS2 type profile) were tested for compositional differences for three biodiversity metrics: Red Sea locality

(seven levels: Gulf of Aqaba, Duba, Al Wajh, Yanbu, Thuwal, Farasan Banks,

Farasan Islands), Porites molecular clade (five levels: clades I, II, III, IV, V), and Porites morphological species (10 levels: P. annae, P. echinulata, P. fontanesii, P. columnaris, P. lobata, P. lutea, P. monticulosa, P. rus, P. solida,

P. spp). The factors were fixed and orthogonal. Prior to running the

PERMANOVA, we verified the homogeneity of the dispersions of the categorical variables using PERMDISP.

Canonical Analysis of Principal coordinates (CAP) for each single factor was performed as a validation, effectively testing how well CAP can correctly re-allocate the samples to their respective groups (Anderson &

Willis, 2003).

A marginal test in distance-based linear modelling (DistLM) was used to explore the correlation between Symbiodiniaceae diversity (genus level, majority ITS2 sequence, ITS2 type profile) and the four environmental variables (salinity, Chl-a, DOM, and SST) (Anderson et al., 2008). A Bray-

Curtis dissimilarity matrix was built with the inclusion of a dummy variable

208

(+1) to accommodate for zeros in the biological data. Bi-plots of the CAP ordination with Bray-Curtis distance were computed using R and the package “vegan” (Oksanen et al., 2013) to visualize the relationship between environmental variables (i.e., Chl-a, SST, POC, and salinity) and each biodiversity metrics (i.e., locality, Porites molecular clade, Porites morphological species) and Symbiodiniaceae at the genus level, ITS2 majority sequence, and ITS2 type profiles.

5.4 Results

5.4.1 Traditional and molecular identification of Porites

Among the 80 Porites colonies collected, a total of nine nominal species were identified based on morphological characters: Porites annae, P. echinulata, P. fontanesii, P. columnaris, P. lobata, P. lutea, P. monticulosa,

P. rus, and P. solida. One sample did not match any of the existing original descriptions of Porites species in the literature and, therefore, was referred to as Porites sp1.

Sequence data for the mtCR and ITS region were obtained from all

80 analysed samples (see Appendix 5.1). One sequence of Goniopora sp. was added to the two datasets and used as outgroup in both the reconstructions following Kitano et al. (2014). The mtCR sequence alignment consisted of 1,292 bp, with 68 polymorphic and 31 parsimony

209 informative sites. The ITS alignment encompassed 795 bp with 103 variable sites, 88 of which were parsimony informative. BI and ML tree topologies obtained from the two regions were largely congruent, recovering five highly-supported molecular clades in our samples (Clades I to V) (Fig. 5.2).

Three clades were comprised of a single morphospecies, and their monophyly was highly supported: clade I included all P. fontanesii material, clade II grouped P. columnaris, and clade III was comprised solely of a sample from P. sp1. Conversely, clades IV and V grouped more than one morphospecies of Porites. Namely, clade IV included P. rus and P. monticulosa, while clade V comprised five different morphologies of

Porites, i.e., P. annae, P. echinulata, P. lobata, P. lutea, and P. solida.

Figure 5.2 RAxML phylogeny reconstructions of Red Sea Porites at two molecular loci. (a) mitochondrial Control Region (b) nuclear

Internal Transcribed Spaces region. Number on the branches represent support values corresponding to Bayesian posterior probabilities (>90%), ML bootstrap values (>7%). Goniopora was selected as outgroup in both the analyses. In (b) black curved lines connect the two sequences of heterozygotes individual

5.4.2 Symbiodiniaceae community structure

A total of 8,159,993 sequences were generated using Illumina MiSeq. After filtering, 5,506,746 sequences were analysed with the SymPortal framework, and 156 Symbiodiniaceae-defining ITS2 intra-genomic variants were recorded within our 80 Porites samples. A total of 77 ITS2 type profiles were recovered that were represented by 38 distinct ITS2 sequences found to be the most abundant in any one of the ITS2 type profiles (see Appendix 5.3).

Symbiodiniaceae community composition at the genus level, the majority

ITS2 sequence level (most abundant ITS2 sequence for a given ITS2 type profile), and the ITS2 type profile level was visualized using stacked bar charts to compare the relative abundance for three factors, namely, locality, Porites molecular clade, and Porites morphological species (Fig.

5.3).

Overall, the most abundant genus of Symbiodiniaceae associated with Porites in the Red Sea was Durusdinium (51%), followed by

Cladocopium (46%); only 3% of the sequences belonged to the genus

Symbiodinium (Fig. 5.3). PERMANOVA analysis identified a strong geographical structure in the Symbiodiniaceae genera distribution along the latitudinal gradient of the Saudi Arabian Red Sea (F73 = 6.42, p = 0.001,

Table 5.1), with the cross-validation analysis re-assigning 42.5% of the sequences to the correct geographic location (Table 5.1). In particular, in the northern Red Sea (i.e. Gulf of Aqaba and Duba), Porites colonies 212 exclusively harboured Cladocopium. From north to south, the

Symbiodiniaceae community shifted gradually from Cladocopium towards

Durusdinium. Indeed, Durusdinium represented 80% of the community in the southern Red Sea (i.e., Farasan Islands). Finally, the genus Symbiodinium only appeared in the Farasan Islands and in low abundance (8%) (Fig. 5.3).

No significant correlations between Symbiodiniaceae genera and Porites morphological species (F68 = 0.73, p = 0.74) or Porites molecular lineages (F75

= 1.14, p = 0.30 - Table 5.1) were recorded.

The majority of ITS2 sequences for a given ITS2 type profile was C15

(accounting for 28% of the entire majority ITS2 sequences), followed by D1

(18% of the total diversity) (Fig. 5.3). A strong latitudinal structure of

Symbiodiniaceae community emerged by analysing the distribution of the majority ITS2 sequence for the ITS2 type profiles predicted (F73 = 2.27, p =

0.001 - cross-validation analysis reassigned 38.75% of the majority ITS2 sequences to the correct location – Table 5.1). In the north, from the Gulf of

Aqaba to Al Wajh, the C15 sequence was the most abundant majority ITS2 sequence, contributing to the entire diversity in the Gulf of Aqaba (100%) and to more than 40% of the diversity in Duba (58%) as well as in Al Wajh

(42%). In Yanbu, D1 (34%) and C15 (38%) were the most abundant majority

ITS2 sequences, while in Thuwal C15 was the most abundant majority ITS2 sequence (39%) together with D1 or D4 (23%) and C15 or C60a (23%).

Similarly, in the Farasan Banks C15, D1, and D4 sequences were the most

213 abundant majority ITS2 sequences (31%, 21%, and 19%, respectively).

Finally, in the Farasan Islands, D1 or D6 and D1 or D4 sequences mostly contributed to the majority ITS2 sequences of the community (36% and 21% respectively) (Fig. 5.3). Symbiodiniaceae host specificity was recovered when considering the majority of ITS2 sequences in relation to Porites molecular clade (F75 = 1.71, p = 0.01; cross-validation reassigned 48.75% of the majority ITS2 sequences to correct Porites molecular clade – Table 5.1).

In Porites clades I, II, III, and IV, the majority ITS2 sequences of the ITS2 types profiles was represented by one or two sequences. In particular, in clade I:

C15 sequence (42%) and D1 or D4 or D1n (32%); clade II: C15 sequence

(45%) and D1 sequence (41%); clade III: A1 sequence (100%). clade IV: D1 or D4 (37%) and D1 (16%). Finally, in Porites clade V, three sequences of

Symbiodiniaceae made up more than 50% of the diversity: C15 (25%), D1 or D6, and D1 (Fig. 5.3). No correlation between Porites morphological species and Symbiodiniaceae ITS2 majority was identified (F68 = 1.19, p = 0.1

– Table 5.2).

A total of 77 Symbiodiniaceae ITS2 type profiles were recovered by the SymPortal analytical framework, distributed as follows: 38 type profiles belonged to the genus Cladocopium, 26 to Durusdinium, and 13 to

Symbiodinium (see Appendix 5.3). Locality, Porites molecular clade, and

Porites morphological species were strongly correlated with the

Symbiodiniaceae type profiles recovered (F73 = 1.49, p = 0.0004; F75 = 1.79,

214

p = 0.001; F68 = 1.31, p = 0.006, respectively – Fig. 5.3), but the cross-validation analyses only reassigned correctly ITS2 type profiles to the geographic location (42.5%), and the Porites molecular clade (55%), while only 10% of them were regrouped into correct morphological species (Table 5.1).

Figure 5.3 Stacked bar charts (100%) comparing Symbiodiniaceae diversity at the (a) (b) (c) genus level, (d) (e) (f) majority ITS2 sequence level (the 20 most abundant shown), (g) (h) (i) ITS2 type profile level (the 20 most abundant shown). For each of these three datasets, bar charts were plotted for three different factors: (a) (d) (g) Porites morphological species, (b) (e) (h)

Porites molecular clade, (c) (f) (j) Red Sea localities

Table 5.1 PERMANOVA results calculated for three biodiversity metrics (Red Sea locality,

Porites molecular clade, and Porites morphological species) for the datasets

Symbiodiniaceae genus, Symbiodiniaceae majority ITS2 sequences, and

Symbiodiniaceae ITS2 type profile. df = degrees of freedom.

Symbiodiniaceae Symbiodiniaceae Symbiodiniaceae genus majority ITS2 sequences ITS2 type profiles

df F p F p F p

Porites molecular clade 4 1.1427 0.307 1.7108 0.01 1.7959 0.001

Res 75

CAP 48.75% 55%

Porites morphological species 11 0.7351 0.74 1.1925 0.108 1.3184 0.006

Res 68

CAP 10%

Red Sea locality 6 6.4268 0.001 2.2736 0.001 1.4954 0.004

Res 73 CAP 42.5% 38.75% 42.5%

Table 5.2 Marginal test results of the DistLM analyses for each of three datasets

(Symbiodiniaceae genus, Symbiodiniaceae majority ITS2 sequence, and

Symbiodiniaceae ITS2 type profile) against the environmental factors SST, Chl- a, POC, and

Salinity. SS = Sum of squares; %var = percent of variance explained by each predictor variable.

Environmental F p % var factor Symbiodiniaceae SST 20.289 0.001 20.64 genus

Chl- a 34.536 0.001 30.68

POC 35.656 0.001 29.18

Salinity 35.656 0.001 31.37

Symbiodiniaceae SST 5.3512 0.001 6.42 majority ITS2 sequence Chl- a 9.3287 0.001 10.68

POC 8.8295 0.001 10.16

Salinity 9.639 0.001 10.99

Symbiodiniaceae SST 2.5913 0.002 3.22 ITS2 type profile Chl- a 3.8551 0.001 4.71

POC 3.6152 0.001 4.43

Salinity 4.0011 0.001 4.88

218

5.4.3 Symbiodiniaceae diversity in relation to environmental variables

DistLM showed that the four environmental variables included in the analyses (i.e. salinity, Chl-a, POC, and SST) were statistically significantly explaining the variation of Symbiodiniaceae diversity at the

Symbiodiniaceae genus level, majority ITS2 sequences, and the ITS2 type profile (P ≤ 0.002, Table 5.2, Appendix 5.4). Interestingly, in all the datasets analysed, salinity was the most influential variable, which alone explained

31.3%, 10.9%, and 4.9% of the entire variability (Symbiodiniaceae genus, majority ITS2 sequence, and ITS2 type profile, respectively), followed by Chl- a, POC, and SST at all the levels (Table 5.2, Appendix 5.4).

5.5 Discussion

5.5.1 Porites morpho-molecular diversity

Porites is one of the most speciose zooxanthellate coral genera in the world, accounting for over 160 described species, 62 of which are currently recognized as valid (WoRMS, 2018). Almost a quarter of these have been previously recorded in sympatry in the Red Sea (Sheer & Pillai, 1983;

Sheppard & Sheppard, 1991; Veron, 2000), rendering the region a biodiversity hotspot for Porites corals. In our study, we characterized 10 Red

Sea Porites morphospecies at two molecular loci. Only three out of the 10 morphospecies proved to be monophyletic (i.e. P. fontanesii, P. columnaris,

219 and P. sp1) and showed concordant morpho-molecular species boundaries. The remaining seven species (P. annae, P. echinulata, P. lobata, P. lutea, P. solida, P. monticulosa, and P. rus) clustered into two evolutionary lineages (clades IV and V). Our molecular data demonstrate that the identification of Porites species based exclusively on morphological features does not match with the molecular lineages occurring in the Red Sea, at least based on the markers used herein (Fig.

5.2). As such, a deep gap in our understanding of species boundaries and evolutionary relationships in the genus is confirmed (Forsman et al., 2009,

2015, 2017). The use of a multi-disciplinary taxonomic approach has proven useful to fill this gap for several coral genera (Benzoni et al., 2010; Schmidt-

Roach et al., 2014; Kitahara et al., 2016). Indeed, recent studies showed that coupling multi-locus molecular evidence with new morphological evidence (e.g., micromorphology and microstructure), reproductive biology data, and symbiosis insights could provide us with a new understanding of Scleractinia systematics and evolution.

5.5.2 Symbiodiniaceae biogeography across Red Sea gradients

The Red Sea environmental gradients correlated with the Symbiodiniaceae biogeographical patterns observed. Sequence data showed a shift from the genus Cladocopium to Durusdinium going from the north (Gulf of

Aqaba and Duba) to the south (Farasan Islands) of the Red Sea, with corals

220 in the central part of the Red Sea (Al Wajh, Yanbu, Thuwal, Farasan Banks) harbouring both Cladocopium and Durusdinium (Fig. 5.1, 5.3).

Symbiodinium only appears in the Farasan Islands at relative low abundance. Similar biogeographical patterns highlighting a community break south of the Gulf of Aqaba have been previously found in coral symbiont communities. For example, Sawall et al. (2014) found discontinuity in the symbiont communities associated with the coral Pocillopora verrucosa between the Gulf of Aqaba and the rest of the Red Sea. Similar to our findings, P. verrucosa colonies from the Gulf of Aqaba were characterized by the unique association with the genus Cladocopium, while only Symbiodinium was found in the remaining sites of the central and southern Red Sea. Similarly, Arrigoni et al. (2016) investigated the symbiont community of Red Sea Stylophora. All samples from the Gulf of Aqaba harboured Cladocopium, while outside coral host colonies also associated with Symbiodinium. This break has been proposed to be mainly driven by cooler temperature at the entrance of the Gulf of Aqaba and, in particular, has been proposed that Cladocopium might have colder water preference (Ulstrup et al., 2006, Sawall et al., 2014). However, whilst this cooler water preference hypothesis may be an effective explanation for individual Symbiodiniaceae taxa, it is important to note that this trend should not be extrapolated to include the whole of the genus

221

Cladocopium, especially when exceptionally thermally tolerant members exist, i.e. C. thermophilum. (D’Angelo et al., 2015).

Genetic breaks have also been documented among population between the central Red Sea and the Farasan Islands in the south. For example, genetic breaks have been reported in different fish populations, (Froukh &

Kochzius, 2007; Nanninga et al., 2014; Saenz-Agudelo et al., 2015), sponges

(Giles et al., 2015), and mussels (Shefer et al., 2004). From these studies, the genetic brakes matched with environmental transitions occurring around

16˚- 20˚ N in the Red Sea. In contrast to the rest of the Red Sea, the Farasan

Islands region is a shallow reef system, characterized by less saline but warm, eutrophic, and turbid waters (Sheppard & Sheppard, 1991). Our data show a predominance of Durusdinium sequences in this environment, with high proportions of D. trenchii (D1-D4) expected. This distribution fits into theories that D. trenchii is a stress-resilient taxon within the Symbiodiniaceae, and thus found in association with warm and turbid environments or in response to stressful events (Baker, 2001; LaJeunesse et al., 2008). Recent studies showed association of P. lutea, P. lobata, and P. harrisoni with C. thermophilum symbionts in the southern Arabian Gulf where SST (36° C) and salinity (42) levels exceed the ones from the Red Sea (D’Angelo et al., 2015;

Hume et al., 2016, 2018). D’Angelo et al. (2015) showed a community transition along the temperature and salinity gradient occurring between the southern Arabian Gulf and the Gulf of Oman: C. thermophilum was

222 associated with 100% of Porites in the southern Arabian Gulf; in the transition zone between the Arabian Gulf and the Gulf of Oman Porites associated with C. thermophilum as well as C15 lineages and D. trenchii; finally, in the

Gulf of Oman, where environmental conditions more resemble those present elsewhere in the tropical Indo-Pacific belt, Porites associated mainly with C15 lineages, as elsewhere in Porites hosts. In the Red Sea, we find a similar pattern. In the north and central Red Sea, Porites is found mainly in association with C15 radiation, the most common symbiont association within Porites. In the south, where the thermal stressors are higher (although not reaching levels as extreme as in the Arabian Gulf)

Porites associates with the more resilient D. trenchii.

5.5.3 Host-symbiont specificity

By identifying previously overlooked Symbiodiniaceae ITS2 sequences, we were able to identify host-specific association patterns. Indeed, our data showed specific coupling between Porites genotypes (clade I to V) and

Symbiodiniaceae ITS2 type profiles and the majority sequences that represent them. C15 radiation sequences were shared among four of the five genetic lineages of Porites, and made up the most of the ITS2 sequences in Clade I, II and V, confirming that the C15 radiation is commonly associated with Porites (LaJeunesse 2004; Franklin et al., 2012;

Keshavmurthy et al., 2014). Nevertheless, association is not exclusive in our

223 data and might vary depending on the host and the environment. Clade

III was the only clade associated with A1 sequences, although small sample size prevents us from drawing any firm conclusion about this association.

No specific pattern of association among morphologically- described nominal species of Porites and Symbiodiniaceae was recovered.

This is an informative result in an era of coral taxonomic revolution, supporting the evidence that coral skeletal morphology alone can lead to misleading classifications. Evolutionary relationships in corals are being revised at every taxonomic level by combining genomic evidence with other lines of evidence. Our results highlighted that such evidence can come from detailed analyses of the associated symbiont community, a trait that has been so far only considered in a limited number of studies.

5.6 Conclusions

This chapter provides an overview of Symbiodiniaceae diversity associated with Porites corals in the Saudi Arabian Red Sea, one of the hottest and most saline environments in the world. This study was able to define zooxanthellae diversity at a new level through the use of an analytical approach leveraging the taxonomically-informative intragenomic sequence diversity harboured in every Symbiodiniaceae genome.

Biogeographical patterns of symbiont distribution were distinguished along

224

2000 km of Red Sea coast, and a strong correlation among symbiont diversity and Red Sea environmental gradients was documented.

Data from this chapter evidenced a correlation among

Symbiodiniaceae diversity and Porites molecular clades (but not morphological species), corroborating the results from chapters 1, 2 and 3.

Finally, this chapter highlights that evidence derived from different approaches, such as the investigation of the symbiont community, can prove useful in understanding coral biodiversity in challenging taxa.

225

5.7 References

Acker, J. G., & Leptoukh, G. (2007). Online analysis enhances use of NASA earth science data. Eos, Transactions American Geophysical Union, 88(2), 14-17.

Anderson, M. J., & Willis, T. J. (2003). Canonical analysis of principal coordinates: a useful method of constrained ordination for ecology. Ecology, 84(2), 511-525.

Anderson, M. J, Gorley, R. N., & Clarke, R. K. (2008). Permanova+ for Primer: Guide to Software and Statistical Methods. Primer-E Limited.

Arif, C., Daniels, C., Bayer, T., Banguera-Hinestroza, E., Barbrook, A., Howe, C. J., ... & Voolstra, C. R. (2014). Assessing Symbiodinium diversity in scleractinian corals via next-generation sequencing-based genotyping of the ITS2 rDNA region. Molecular Ecology, 23(17), 4418-4433.

Arrigoni, R., Benzoni, F., Terraneo, T. I., Caragnano, A., & Berumen, M. L. (2016). Recent origin and semi-permeable species boundaries in the scleractinian coral genus Stylophora from the Red Sea. Scientific Reports, 6, 34612. Baker, A. C. (2001). Ecosystems: reef corals bleach to survive change. Nature, 411(6839), 765.

Baker, A. C. (2004). Symbiont diversity on coral reefs and its relationship to bleaching resistance and resilience. In Coral health and disease (pp. 177- 194). Springer, Berlin, Heidelberg.

Berumen, M. L., Hoey, A. S., Bass, W. H., Bouwmeester, J., Catania, D., Cochran, J. E., ... & Saenz-Agudelo, P. (2013). The status of coral reef ecology research in the Red Sea. Coral Reefs, 32(3), 737-748.

226

Briggs, J. C., & Bowen, B. W. (2012). A realignment of marine biogeographic provinces with particular reference to fish distributions. Journal of Biogeography, 39(1), 12-30.

Camacho, C., Coulouris, G., Avagyan, V., Ma, N., Papadopoulos, J., Bealer, K., & Madden, T. L. (2009). BLAST+: architecture and applications. BMC Bioinformatics, 10(1), 421.

Castresana, J. (2000). Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Molecular Biology and Evolution, 17(4), 540-552.

Cunning, R., Gates, R. D., & Edmunds, P. J. (2017). Using high-throughput sequencing of ITS2 to describe Symbiodinium metacommunities in St. John, US Virgin Islands. PeerJ, 5, e3472.

D'angelo, C., Hume, B. C., Burt, J., Smith, E. G., Achterberg, E. P., & Wiedenmann, J. (2015). Local adaptation constrains the distribution potential of heat-tolerant Symbiodinium from the Persian/Arabian Gulf. The ISME Journal, 9(12), 2551.

DiBattista, J. D., Howard Choat, J., Gaither, M. R., Hobbs, J. P. A., Lozano- Cortés, D. F., Myers, R. F., ... & Berumen, M. L. (2016a). On the origin of endemic species in the Red Sea. Journal of Biogeography, 43(1), 13-30.

DiBattista, J. D., Roberts, M. B., Bouwmeester, J., Bowen, B. W., Coker, D. J., Lozano-Cortés, D. F., ... & Kochzius, M. (2016b). A review of contemporary patterns of endemism for shallow water reef fauna in the Red Sea. Journal of Biogeography, 43(3), 423-439.

Dimond, J. L., Gamblewood, S. K., & Roberts, S. B. (2017). Genetic and epigenetic insight into morphospecies in a reef coral. Molecular ecology, 26(19), 5031-5042.

Eren, A. M., Morrison, H. G., Lescault, P. J., Reveillaud, J., Vineis, J. H., & Sogin, M. L. (2015). Minimum entropy decomposition: unsupervised oligotyping for sensitive partitioning of high-throughput marker gene sequences. The ISME journal, 9(4), 968.

Falkowski, P. G., Dubinsky, Z., Muscatine, L., & Porter, J. W. (1984). Light and the bioenergetics of a symbiotic coral. Bioscience, 34(11), 705-709.

227

Flot, J. F. (2007). Champuru 1.0: a computer software for unraveling mixtures of two DNA sequences of unequal lengths. Molecular Ecology Notes, 7(6), 974-977.

Flot, J. F. (2010). SeqPHASE: a web tool for interconverting PHASE input/output files and FASTA sequence alignments. Molecular Ecology Resources, 10(1), 162-166.

Forsman, Z., Wellington, G.M., Fox, G.E., & Toonen, R.J. (2015). Clues to unraveling the coral species problem: distinguishing species from geographic variation in Porites across the Pacific with molecular markers and microskeletal traits. PeerJ, 3, e751.

Forsman, Z.H., Knapp, I.S.S., Tisthammer, K., Eaton, D.A.R., Belcaid, M., & Toonen, R.J. (2017). Coral hybridization or phenotypic variation? Genomic data reveal gene flow between Porites lobata and P. compressa. Molecular Phylogenetics and Evolution, 111, 132-148.

Froukh, T., & Kochzius, M. (2007). Genetic population structure of the endemic fourline wrasse (Larabicus quadrilineatus) suggests limited larval dispersal distances in the Red Sea. Molecular Ecology, 16(7), 1359-1367.

Franklin, E. C., Stat, M., Pochon, X., Putnam, H. M., & Gates, R. D. (2012). GeoSymbio: a hybrid, cloud-based web application of global geospatial bioinformatics and ecoinformatics for Symbiodinium–host symbioses. Molecular Ecology Resources, 12(2), 369-373.

Furby, K. A., Bouwmeester, J., & Berumen, M. L. (2013). Susceptibility of central Red Sea corals during a major bleaching event. Coral Reefs, 32(2), 505-513.

228 comparisons with the 1982–1983 event. Bulletin of Marine Science, 69(1), 79- 109.

Hall, T. A. (1999, January). BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. In Nucleic acids symposium series (Vol. 41, No. 41, pp. 95-98). [London]: Information Retrieval Ltd., c1979-c2000.

Huang, H., Dong, Z., Huang, L., Yang, J., Di, B., Li, Y., ... & Zhang, C. (2011). Latitudinal variation in algal symbionts within the scleractinian coral Galaxea fascicularis in the South China Sea. Marine Biology Research, 7(2), 208-211.

Hughes, T. P., Anderson, K. D., Connolly, S. R., Heron, S. F., Kerry, J. T., Lough, J. M., ... & Claar, D. C. (2018). Spatial and temporal patterns of mass bleaching of corals in the Anthropocene. Science, 359(6371), 80-83.

Hume, B. C., D'Angelo, C., Smith, E. G., Stevens, J. R., Burt, J., & Wiedenmann, J. (2015). Symbiodinium thermophilum sp. nov., a thermotolerant symbiotic alga prevalent in corals of the world's hottest sea, the Persian/Arabian Gulf. Scientific Reports, 5, 8562.

Hume, B. C., D'Angelo, C., Burt, J. A., & Wiedenmann, J. (2018). Fine-scale biogeographical boundary delineation and sub-population resolution in the Symbiodinium thermophilum coral symbiont group from the Persian/Arabian Gulf and Gulf of Oman. Frontiers in Marine Science.

Katoh, K., & Standley, D. M. (2013). MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Molecular Biology and Evolution, 30(4), 772-780.

Kemp, D. W., Hernandez-Pech, X., Iglesias-Prieto, R., Fitt, W. K., & Schmidt, G. W. (2014). Community dynamics and physiology of Symbiodinium spp.

229 before, during, and after a coral bleaching event. Limnology and Oceanography, 59(3), 788-797.

Kitano, Y. F., Benzoni, F., Arrigoni, R., Shirayama, Y., Wallace, C. C., & Fukami, H. (2014). A phylogeny of the family Poritidae (Cnidaria, Scleractinia) based on molecular and morphological analyses. PLoS One, 9(5), e98406.

Kitahara, M. V., Fukami, H., Benzoni, F., & Huang, D. (2016). The new systematics of Scleractinia: integrating molecular and morphological evidence. The Cnidaria, Past, Present and Future (pp. 41-59). Springer, Cham.

LaJeunesse, T. C (2002). Diversity and community structure of symbiotic dinoflagellates from Caribbean coral reefs. Marine Biology, 141(2), 387-400.

LaJeunesse, T. C. (2004). “Species” radiations of symbiotic dinoflagellates in the Atlantic and Indo-Pacific since the Miocene-Pliocene transition. Molecular Biology and Evolution, 22(3), 570-581.

LaJeunesse, T. C., Loh, W. K., Van Woesik, R., Hoegh-Guldberg, O., Schmidt, G. W., & Fitt, W. K. (2003). Low symbiont diversity in southern Great Barrier Reef corals, relative to those of the Caribbean. Limnology and Oceanography, 48(5), 2046-2054.

LaJeunesse, T. C., & Pinzon, J. H. (2007). Screening intragenomic rDNA for dominant variants can provide a consistent retrieval of evolutionarily persistent ITS (rDNA) sequences. Molecular Phylogenetics and Evolution, 45(1), 417-422.

LaJeunesse, T. C., Bonilla, H. R., Warner, M. E., Wills, M., Schmidt, G. W., & Fitt, W. K. (2008). Specificity and stability in high latitude eastern Pacific coral- algal symbioses. Limnology and Oceanography, 53(2), 719-727.

230

LaJeunesse, T. C., & Thornhill, D. J. (2011). Improved resolution of reef-coral endosymbiont (Symbiodinium) species diversity, ecology, and evolution through psbA non-coding region genotyping. PLoS One, 6(12), e29013.

Miller, M. A., Pfeiffer, W., & Schwartz, T. (2012, July). The CIPRES science gateway: enabling high-impact science for phylogenetics researchers with limited resources. In Proceedings of the 1st Conference of the Extreme Science and Engineering Discovery Environment: Bridging from the eXtreme to the campus and beyond (p. 39). ACM.

Monroe, A. A., Ziegler, M., Roik, A., Röthig, T., Hardenstine, R. S., Emms, M. A., ... & Berumen, M. L. (2018). In situ observations of coral bleaching in the central Saudi Arabian Red Sea during the 2015/2016 global coral bleaching event. PloS one, 13(4), e0195814.

Ngugi, D. K., Antunes, A., Brune, A., & Stingl, U. (2012). Biogeography of pelagic bacterioplankton across an antagonistic temperature–salinity gradient in the Red Sea. Molecular Ecology, 21(2), 388-405.

Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P. R., O’hara, R. B., ... & Oksanen, M. J. (2013). Package ‘vegan’. Community Ecology Package, 2(9).

Oliver, T. A., & Palumbi, S. R. (2009). Distributions of stress-resistant coral symbionts match environmental patterns at local but not regional scales. Marine Ecology Progress Series, 378, 93-103.

Osman, E. O., Smith, D. J., Ziegler, M., Kürten, B., Conrad, C., El-Haddad, K. M., ... & Suggett, D. J. (2018). Thermal refugia against coral bleaching

231 throughout the northern Red Sea. Global Change Biology, 24(2), e474- e484.

Raitsos, D. E., Pradhan, Y., Brewin, R. J., Stenchikov, G., & Hoteit, I. (2013). Remote sensing the phytoplankton seasonal succession of the Red Sea. PloS one, 8(6), e64909.

Rambaut, A., Drummond, A. J., & Suchard, M. (2014). Tracer v1. 6 http://beast. bio. ed. ac. uk. Tracer.

Rosic, N., Ling, E. Y. S., Chan, C. K. K., Lee, H. C., Kaniewska, P., Edwards, D., ... & Hoegh-Guldberg, O. (2015). Unfolding the secrets of coral–algal symbiosis. The ISME Journal, 9(4), 844.

Rowan, R., & Knowlton, N. (1995). Intraspecific diversity and ecological zonation in coral-algal symbiosis. Proceedings of the National Academy of Sciences, 92(7), 2850-2853.

Saenz-Agudelo, P., Dibattista, J. D., Piatek, M. J., Gaither, M. R., Harrison, H. B., Nanninga, G. B., & Berumen, M. L. (2015). Seascape genetics along environmental gradients in the Arabian Peninsula: insights from ddRAD sequencing of anemonefishes. Molecular Ecology, 24(24), 6241-6255.

Sampayo, E. M., Ridgway, T., Bongaerts, P., & Hoegh-Guldberg, O. (2008). Bleaching susceptibility and mortality of corals are determined by fine- scale differences in symbiont type. Proceedings of the National Academy of Sciences Sampayo, E. M., Ridgway, T., Franceschinis, L., Roff, G., Hoegh-Guldberg, O., & Dove, S. (2016). Coral symbioses under prolonged environmental change: living near tolerance range limits. Scientific Reports, 6, 36271.

Sawall, Y., Al-Sofyani, A., Banguera-Hinestroza, E., & Voolstra, C. R. (2014). Spatio-temporal analyses of Symbiodinium physiology of the coral Pocillopora verrucosa along large-scale nutrient and temperature gradients in the Red Sea. PloS one, 9(8), e103179.

232

Scheer, G., & Pillai, C. S. (1983). Report on the stony corals from the Red Sea.

Schloss, P. D., Westcott, S. L., Ryabin, T., Hall, J. R., Hartmann, M., Hollister, E. B., ... & Sahl, J. W. (2009). Introducing mothur: open-source, platform- independent, community-supported software for describing and comparing microbial communities. Applied and Environmental Microbiology, 75(23), 7537-7541.

Schmidt-Roach, S., Miller, K. J., Lundgren, P., & Andreakis, N. (2014). With eyes wide open: a revision of species within and closely related to the Pocillopora damicornis species complex (Scleractinia; Pocilloporidae) using morphology and genetics. Zoological Journal of the Linnean Society, 170(1), 1-33.

Shefer, S., Abelson, A., Mokady, O., & Geffen, E. L. I. (2004). Red to Mediterranean Sea bioinvasion: natural drift through the Suez Canal, or anthropogenic transport?. Molecular Ecology, 13(8), 2333-2343.

Sheppard, C. R., & Sheppard, A. S. (1991). Corals and coral communities of Arabia (Vol. 12).

Smith, E. G., Ketchum, R. N., & Burt, J. A. (2017). Host specificity of Symbiodinium variants revealed by an ITS2 metahaplotype approach. The ISME Journal, 11(6), 1500.

Stamatakis, A. (2014). RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics, 30(9), 1312-1313.

Stephens, M., Smith, N. J., & Donnelly, P. (2001). A new statistical method for haplotype reconstruction from population data. The American Journal of Human Genetics, 68(4), 978-989.

Tonk, L., Sampayo, E. M., Chai, A., Schrameyer, V., & Hoegh-Guldberg, O. (2017). Symbiodinium (Dinophyceae) community patterns in invertebrate hosts from inshore marginal reefs of the Southern Great Barrier Reef, Australia. Journal of Phycology, 53(3), 589-600.

Trommer, G., Siccha, M., van der Meer, M. T., Schouten, S., Damsté, J. S. S., Schulz, H., ... & Kucera, M. (2009). Distribution of Crenarchaeota tetraether membrane lipids in surface sediments from the Red Sea. Organic Geochemistry, 40(6), 724-731.

233

Ulstrup, K. E., Berkelmans, R., Ralph, P. J., & Van Oppen, M. J. H. (2006). Variation in bleaching sensitivity of two coral species across a latitudinal gradient on the Great Barrier Reef: the role of zooxanthellae. Marine Ecology Progress Series, 314, 135-148. van Oppen, M. J. H., & Lough, J. M. (2018). Synthesis: coral bleaching: patterns, processes, causes and consequences. Coral Bleaching (pp. 343- 348). Springer, Cham.

Veron, J. E. N. (2000). Corals of the World (No. C/593.6 V4).

WoRMS Editorial Board, 2018. World Register of Marine Species. Available from http://www.marinespecies.org at VLIZ. (accessed 2016-06-04).

Ziegler, M., Roder, C. M., Büchel, C., & Voolstra, C. R. (2015). Mesophotic coral depth acclimatization is a function of host-specific symbiont physiology. Frontiers in Marine Science, 2, 4.

234

6. CHAPTER 6: CONCLUSIONS

6.1 General conclusions

This dissertation represents a substantial contribution towards the formulation of a reliable hypothesis on coral evolutionary history. The use of traditional molecular taxonomy with the integration of micromorphological data, revolutionized the understanding of Scleractinia taxonomy, systematics, and evolution during the past 15 years (Fukami et al., 2004,

2008; Wallace et al., 2007; Benzoni et al., 2007, 2010, 2012; Richards et al.,

2008, 2013; Forsman et al., 2009, 2017; Budd and Stolarski, 2009, 2011;

Kitahara et al., 2016; Stolarski et al., 2011; Huang et al., 2011, 2014; Budd et al., 2012; Pinzon et al., 2013; Schmidt-Roach et al., 2013a, 2013b, 2014;

Terraneo et al., 2014, 2016, 2017; Kitano et al., 2014; Arrigoni et al., 2014,

2016a, b, 2017, 2018). Nevertheless, this approach is still far from providing a clear and definitive hypothesis of Scleractinia evolutionary history, especially when some of the most important and specious genera such as

Acropora, Montipora, and Porites still await revision. Several limitations inherited with traditional morphological and molecular data, together with the incongruence among morphological and molecular reconstructions, as well as nuclear and mitochondrial phylogenies, urge towards the use of alternative line of evidence to clarify species boundaries and evolution for these recalcitrant groups (Todd, 2008; Van Oppen et al., 2001; Fukami et al.

235

2004). Clarifiying evolutionary relationships in Scleractinia can provide background information that could be applied to a variety of fields, from ecological, biological, and paleontological studies, to biodiversity assessments, and finally conservation.

To better understand evolutionary relationships in Porites, in this work, different lines of evidence derived from morphology data, genome-wide molecular data, and symbiont association data, were used in synergy following the unified concept of species proposed by de Quiroz (1998).

According to de Queiroz, species can be considered as independent evolving metapopulation lineages (Queiroz, 1998), and the plethora of species concepts designed by evolutionary biologists should be integrated to provide different lines of evidence aimed to distinguish evolutionary lineages. The results from this thesis show that this approach can be helpful in clarifying the taxonomy and evolutionary relationships of one of the most challenging scleractinian taxa in terms of species identification. The approach employed here could be applied to explore relationships within other problematic coral genera, and provide the basis for a new framework to future coral taxonomy.

This thesis comprises the largest number of samples of Porites morpho- molecularly assessed to date from the seas around the Arabian Peninsula.

Before this dissertation, only six sequences of Porites from the Arabian region were analyzed using a combined morpho-molecular approach and

236 deposit to GenBank (https://www.ncbi.nlm.nih.gov/nucleotide/) (Benzoni

& Stefani, 2012). With over 200 sequences, two complete mitochondrial genomes, and almost complete mitochondrial genomes, rDNAs, histone regions and 96,990 SNPs for around 140 samples, this work represent the most comprehensive molecular database of Porites so far. Moreover,

5,506,746 ITS2 sequences, representing 156 Symbiodiniaceae-defining ITS2 intra-genomic variants and 77 ITS2 type profiles were also produced, providing a substantial contribution to the SymPortal online repository

(http://www.symportal.org/).

This thesis provides the first phylogenetic reconstruction of Porites form the Arabian region using traditional molecular markers, genome-wide data, species delimitation analyses and micromorphology. Moreover, in this dissertation, I identified three species of Porites not corresponding to any

Porites species in the literature and endemic to the Arabian region. Overall,

I prove that Porites diversity from the Arabian Peninsula as previously assessed is not representative of real biodiversity and biogeography patterns, and needs to be re–evaluated in light of a rapidly changing taxonomic framework. Descriptions of new species, expansion of ranges for other species, and phylogenetic works at several taxonomic levels, are in fact providing a new understanding of regional biogeography and evolution for several taxa in the Arabian Peninsula (Bouwmeester et al.,

2015; Arrigoni et al., 2016a, 2016b; Terraneo et al., 2014, 2016, 2017; Berumen

237 et al., 2019). For example, high endemism patterns in the western and northern Indian Ocean, comprising the Arabian region, coupled with phylogenetics, and paleooceanography support the” Indian Ocean

Center of Origin” hypothesis for shallow marine organisms (Obura et al.,

2016). Endemism, by definition, encompasses species with restricted geographical distribution, that are carachterized by either ancient origin

(derived from an ancestral widespread species) (Willis, 1922) or are instead young and newly established (Stebbins & Major, 1965). Neutral models of speciation, comprising allopatric speciation, parapatric and peripatric speciation, can thus contribute towards the establishment of these endemism patterns. Moreover, the presence of specific ecological conditions, can also constrain the distribution of endemic species

(ecological speciation trough local adaptation) (Stebbins, 1980).

According to the” Indian Ocean Center of Origin” theory, current high endemis areas in the Indian Ocean, such as the Red Sea, the Arabinan Sea, and the Mascarene Islands, were center of marine biodiversity origin during the Oligocene. For corals, this pattern is supported by endemism and deep divergence within several coral genera, comprising Acropora,

Coscinaerea, Stylophora, and Siderastrea (Obura et al., 2016), and the high tectonic activities of the above-mentioned areas during the Neogene

(Bosworth et al., 2005). The results of this thesis, show deep divergence of at least five endemic Porites lineages around the Arabian Peninsula,

238 suggesting an Indian Ocean origin between 10.5 Ma and 1.2 Ma. Providing a better understanding of endemic patterns and how endemism originated in this region, can finally provide new insights into the origin of marine biodiversity in the Indian Ocean.

Following the hologenome theory of evolution, I further corroborated my findings by integrating symbiotic data. By identifying previously overlooked Symbiodiniaceae ITS2 sequences, I was able to identify host- specific association patterns. Indeed, our data showed specific coupling between Porites genotypes and Symbiodiniaceae ITS2 type profiles and the majority sequences that represent them. No specific pattern of association among morphologically-described nominal species of Porites and

Symbiodiniaceae was recovered. This is an informative result in an era of scleractinian taxonomic revolution, supporting the evidence that coral skeletal morphology alone can lead to misleading classifications. Our results highlighted that better evidence can come from detailed analyses of the associated symbiont community, a trait that has been so far only considered in a limited number of studies.

Overall, this dissertation contributes important basic knowledge that could be applied to future conservation strategies for reef corals diversity.

Elucidating where species boundaries lay, and providing evolutionary information on how species are related is essential knowledge for an efficient protection and preservation of biological diversity. Indeed, it has

239 been suggested that in the future, the loss of phylogenetic diversity might be higher than the loss of species, because threatened species seem to have a non-random distribution in the phylogeny (Forest et al., 2015; Veron et al., 2015). Nevertheless, assessments of biodiversity are still mainly based on species counts, yet the intrinsic problems associated with the high variability of coral morphologies render these assessments inaccurate, a situation that could be improved by incorporating genetic and phylogenetic data.

The results from this work provide extensive background data for future works and provide a significant contribution to the knowledge of evolution of Porites.

6.2 Future perspectives

During the time of this PhD, I started other projects aimed at a better understanding of Porites taxonomy and evolution. These works are currently focusing on a. the inclusion of further sampling localities in the Indo-Pacific in the phylogeny proposed in Chapter 2, and 4, b. the inclusion of further species of Porites in the phylogeny proposed in Chapter 2 and 4, c. the formal description of species from understudied localities in the Indo-

Pacific, d. unravelling of cryptic divergence within widespread morphological species of Porites, and finally, e. integrating further lines of evidence to taxonomy in Porites based on the study of the species

240 reproductive traits and conducting hybridization trials between genetically indistinguishable species.

As a result of this unified approach to taxonomy a formal revision of

Porites will hopefully be carried out in the future.

241

6.3 References

Arrigoni, R., Benzoni, F., Terraneo, T. I., Caragnano, A., & Berumen, M. L. (2016a). Recent origin and semi-permeable species boundaries in the scleractinian coral genus Stylophora from the Red Sea. Scientific Reports, 6, 34612.

Arrigoni, R., Berumen, M. L., Chen, C. A., Terraneo, T. I., Baird, A. H., Payri, C., & Benzoni, F. (2016b). Species delimitation in the reef coral genera Echinophyllia and Oxypora (Scleractinia, Lobophylliidae) with a description of two new species. Molecular Phylogenetics and Evolution, 105, 146-159.

Arrigoni, R., Berumen, M. L., Huang, D., Terraneo, T. I., & Benzoni, F. (2017). Cyphastrea (Cnidaria: Scleractinia: Merulinidae) in the Red Sea: phylogeny and a new reef coral species. Invertebrate Systematics, 31(2), 141-156.

Arrigoni, R., Berumen, M. L., Stolarski, J., Terraneo, T. I., & Benzoni, F. (2018). Uncovering hidden coral diversity: a new cryptic lobophylliid scleractinian from the Indian Ocean. Cladistics.

Arrigoni, R., Terraneo, T. I., Galli, P., & Benzoni, F. (2014). Lobophylliidae (Cnidaria, Scleractinia) reshuffled: Pervasive non-monophyly at genus level. Molecular Phylogenetics and Evolution, 73(1). doi:10.1016/j.ympev.2014.01.010.

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, 160(3), 421–456. doi:10.1111/j.1096-3642.2010.00622.

Benzoni, F., Stefani, F., Stolarski, J., Pichon, M., Mitta, G., & Galli, P. (2007). Debating phylogenetic relationships of the scleractinian Psammocora:\nmolecular and morphological evidences. Contributions to Zoology, 76(1), 35–54.

242

Bosworth, W., Huchon, P., & McClay, K. (2005). The red sea and gulf of aden basins. Journal of African Earth Sciences, 43(1-3), 334-378.

Bouwmeester, J., Benzoni, F., Baird, A. H., & Berumen, M. L. (2015). Cyphastrea kausti sp. n.(Cnidaria, Anthozoa, Scleractinia), a new species of reef coral from the Red Sea. ZooKeys, (496), 1.

Budd, A. F., & 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(2), 142-165.

Budd, A. F., & Stolarski, J. (2011). Corallite wall and septal microstructure in scleractinian reef corals: comparison of molecular clades within the family Faviidae. Journal of Morphology, 272(1), 66-88.

Forest, F., Crandall, K. A., Chase, M. W., & Faith, D. P. (2015). Phylogeny, extinction and conservation: embracing uncertainties in a time of urgency. Philosophical transactions of the Royal Society of London. Series B, Biological Sciences, 370(1662), 20140002-20140002.

Forsman, Z. H., Knapp, I., & Tisthammer, K. (2017). Coral hybridization or phenotypic variation? Genomic data reveal gene flow between Porites lobata and P. compressa. Molecular Phylogenetics, 111 (132-148).

Fukami, H., Budd, A. F., Paulay, G., Solé-Cava, A., Allen Chen, C., Iwao, K., & Knowlton, N. (2004). Conventional taxonomy obscures deep divergence between Pacific and Atlantic corals. Nature, 427(6977), 832–835. doi:10.1038/nature02321.1.

Fukami, H., Chen, C. A., Budd, A. F., Collins, A., Wallace, C., Chuang, Y. Y., … Knowlton, N. (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(9), e3222. doi:10.1371/journal.pone.0003222.

243

Huang, D., Licuanan, W. Y., Baird, A. H., & Fukami, H. (2011). Cleaning up the “Bigmessidae”: molecular phylogeny of scleractinian corals from Faviidae, Merulinidae, Pectiniidae and Trachyphylliidae. BMC Evolutionary Biology, 11(1), 37. doi:10.1186/1471-2148-11-37.

Kitahara, M., Fukami, H., Benzoni, F., & Huang, D. (2016). The new systematics of Scleractinia: integrating molecular and morphological evidence. The Cnidaria, Past, Present.

Obura, D. (2016). An Indian Ocean centre of origin revisited: Palaeogene and Neogene influences defining a biogeographic realm. Journal of Biogeography, 43(2), 229-242.

Pinzón, J. H., Sampayo, E., Cox, E., Chauka, L. J., Chen, C. A., Voolstra, C. R., & Lajeunesse, T. C. (2013). Blind to morphology: Genetics identifies several widespread ecologically common species and few endemics among Indo-Pacific cauliflower corals (Pocillopora, Scleractinia). Journal of Biogeography, 40(8), 1595–1608. doi:10.1111/jbi.12110.

Queiroz, De K. (1998). The general lineage concept of species, species criteria, and the process of speciation. In Endless forms: Species and Speciation.

Richards, Z. T., Miller, D. J., & Wallace, C. C. (2013). Molecular phylogenetics of geographically restricted Acropora species: implications for threatened species conservation. Molecular Phylogenetics and Evolution, 69(3), 837- 851.

Richards, Z. T., van Oppen, M. J., Wallace, C. C., Willis, B. L., & Miller, D. J. (2008). Some rare Indo-Pacific coral species are probable hybrids. PLoS One, 3(9), e3240.

Schmidt-Roach, S., Lundgren, P., Miller, K. J., Gerlach, G., Noreen, A. M. E., & Andreakis, N. (2013a). Assessing hidden species diversity in the coral Pocillopora damicornis from Eastern Australia. Coral Reefs, 32(1), 161–172. doi:10.1007/s00338-012-0959-z.

244

Schmidt-Roach, S., Miller, K. J., & Andreakis, N. (2013b). Pocillopora aliciae: a new species of scleractinian coral (Scleractinia, Pocilloporidae) from subtropical Eastern Australia. Zootaxa, 3626(4), 576-582.

Stebbins, G., L. (1980). Rarity of plant species: A synthetic viewpoint. Rho- dora, 82(829), 77–86.

Stebbins, G., L., & Major, J. (1965). Endemism and speciation in the Cali- fornia flora. Ecological Monographs, 35(1), 1–35. https://doi.org/10. 2307/1942216

Terraneo, T. I., Berumen, M. L., Arrigoni, R., Waheed, Z., Bouwmeester, J., Caragnano, A., … Benzoni, F. (2014). Pachyseris inattesa sp. n. (Cnidaria, Anthozoa, Scleractinia): A new reef coral species from the Red Sea and its phylogenetic relationships. ZooKeys, (433). doi:10.3897/zookeys.433.8036.

Todd, P. A. (2008). Morphological plasticity in scleractinian corals. Biological Reviews, 83(3), 315-337. van Oppen, M. J., McDonald, B. J., Willis, B., & Miller, D. J. (2001). The evolutionary history of the coral genus Acropora (Scleractinia, Cnidaria) based on a mitochondrial and a nuclear marker: reticulation, incomplete lineage sorting, or morphological convergence?. Molecular Biology and Evolution, 18(7), 1315-1329.

245

Veron, J., Stafford-Smith, M., DeVantier, L., & Turak, E. (2015). Overview of distribution patterns of zooxanthellate Scleractinia. Frontiers in Marine Science, 1, 81.

Wallace, C. C., Chen, C. a., Fukami, H., & Muir, P. R. (2007). Recognition of separate genera within Acropora based on new morphological, reproductive and genetic evidence from Acropora togianensis, and elevation of the subgenus Isopora Studer, 1878 to genus (Scleractinia: Astrocoeniidae; Acroporidae). Coral Reefs, 26(2), 231–239. doi:10.1007/s00338-007-0203-4.

Willis, J. C. (1922). Age and Area, a Study in Geographical Distribution and Origin of Species, by JC Willis,... with Chapters by Hugo De Vries, HB Guppy, Mrs. EM Reid, James Small... The University Press.

246

LIST OF PUBLICATIONS

Peer reviewed papers resulting from this research

Terraneo, T. I., Arrigoni, R., Benzoni, F., Forsman, Z. H., & Berumen, M. L. (2018). Using ezRAD to reconstruct the complete mitochondrial genome of Porites fontanesii (Cnidaria: Scleractinia). Mitochondrial DNA Part B, 3(1), 173-174. Contribution: I designed the study, performed genomic labwork, performed reads assembly, phylogenomic analyses, genome annotation and wrote the paper.

Terraneo, T. I., Arrigoni, R., Benzoni, F., Forsman, Z. H., & Berumen, M. L. (2018). The complete mitochondrial genome of Porites harrisoni (Cnidaria: Scleractinia) obtained using next-generation sequencing. Mitochondrial DNA Part B, 3(1), 286-287. Contribution: I designed the study, performed genomic labwork, performed reads assembly, phylogenomic analyses, genome annotation and wrote the paper.

Terraneo, T. I., Fusi, M., Hume, C. C., Arriogni, R., Voolstra, C. R., Benzoni, F., Forsman Z. H., Berumen, M. L. Environmental latitudinal gradients and host specificity shape Symbiodiniaceae distribution in Red Sea Porites corals. (2019). Journal of Biogeography. Contribution: I designed the study, collected part of the samples, performed genetic analyses, prepared MiSeq libraries, performed data analyses with contribution of M. F and wrote the paper

Terraneo, T. I., Benzoni, F., Baird, A. H., Arrigoni, R., Berumen, M. L. Morphology and molecules reveal two new species of Porites (Scleractinia, Poritidae) from the Red Sea and the Gulf of Aden. (2019). Systematics and Biodiversity. Contribution: I designed the study, collected part of the samples, performed genetic labwork and analyses, performed species delimitation analyses, performed imaging with Scanning Electron Microscopy, contributed with the formal species description and wrote the paper.

Peer reviewed papers resulting from this research (Under revision)

Terraneo, T. I., Arriogni, R., Benzoni, F., Mariappan, K. G., Baird, A. H., Forsman, Z. H., Wooster, M. K., Bouwmeester, J., Marshell, J., Berumen, M. L. Phylogenomics and phylogeography of Porites from the Arabian Peninsula. Under revision in Systematic Biology.

247

Contribution: I designed the study, collected part of the samples, performed genetic labwork and analyses, performed genomic library preparation, performed phylogenomic analyses and wrote the paper.

Peer reviewed papers not resulting from this research

Coker D. J., DiBattista J. D., Stat, M., Arrigoni, R., Reimer J., Terraneo, T. I., Villalobos Vazquez de la Parra, R., Nowicki J. P., Bunce, M., Berumen, M. L. Metabarcoding approaches increase dietary resolution for benthic feeding butterflyfishes. Submitted to Scientific Reports. Contribution: I contributed with the study design, performed analyses on coral samples and helped with the ms writing. I contributed with presence/ absence data for a variety of coral genera.

Berumen, M. L., Arrigoni, R., Bouwmeester, J., Terraneo, T. I., & Benzoni, F. (2019). Biodiversity of Red Sea corals. Chapter 7 in Coral Reefs of the Red Sea (Voolstra, C.R., and Berumen, M.L. (eds)). Springer. Contribution: I was in charge of the “Climate change and Red Sea corals” paragraph writing and proofreading of the remaining paragraphs of the book chapter.

Arrigoni, R., Maggioni, D., Montano, S., Hoeksema, B. W., Seveso, D., Shlesinger, T., Terraneo, T. I., Tiethbol, M. D., Berumen, M. L. (2018). An integrated morpho-molecular approach to delineate species boundaries of Millepora from the Red Sea. Coral Reefs, 37(4), 967-984. Contribution: I helped collecting samples along the Saudi Arabian Red Sea and contributed for the molecular lab work and molecular analyses. I finally contributed with the ms writing.

Arrigoni, R., Berumen, M. L., Stolarski, J., Terraneo, T. I., & Benzoni, F. (2018). Uncovering hidden coral diversity: a new cryptic lobophylliid scleractinian from the Indian Ocean. Cladistics. Contribution: I contributed with the molecular lab work and molecular analyses. I finally contributed with the ms writing.

Terraneo, T. I., Arrigoni, R., Benzoni, F., Tietbohl, M. D., & Berumen, M. L. (2017). Exploring the genetic diversity of shallow-water Agariciidae (Cnidaria: Anthozoa) from the Saudi Arabian Red Sea. Marine Biodiversity, 47(4), 1065-1078. Contribution: I designed the study, collected part of the samples, performed genetic labwork and analyses, and wrote the paper.

248

Contribution: I helped collecting samples and took part in the molecular lab work and molecular analyses. I helped imaging the samples with Scanning Electron Microscopy and contributed with the ms writing.

Arrigoni, R., Berumen, M. L., Chen, C. A., Terraneo, T. I., Baird, A. H., Payri, C., & Benzoni, F. (2016). Species delimitation in the reef coral genera Echinophyllia and Oxypora (Scleractinia, Lobophylliidae) with a description of two new species. Molecular Phylogenetics and Evolution, 105, 146-159. Contribution: I took part in the molecular lab work and molecular analyses and contributed with the ms writing.

Arrigoni, R., Benzoni, F., Terraneo, T. I., Caragnano, A., & Berumen, M. L. (2016). Recent origin and semi-permeable species boundaries in the scleractinian coral genus Stylophora from the Red Sea. Scientific Reports, 6, 34612. Contribution: I helped collecting samples and took part in the molecular lab work and molecular analyses. I helped imaging the samples with Scanning Electron Microscopy and contributed with the ms writing.

Terraneo, T. I., Benzoni, F., Arrigoni, R., & Berumen, M. L. (2016). Species delimitation in the coral genus Goniopora (Scleractinia, Poritidae) from the Saudi Arabian Red Sea. Molecular Phylogenetics and Evolution, 102, 278- 294. Contribution: I designed the study, collected the samples, performed genetic labwork and analyses, performed species delimitation analyses, performed imaging with Scanning Electron Microscopy, performed imaging with Optical microscopy, performed morphometric analyses and wrote the paper.

Arrigoni, R., Benzoni, F., Huang, D., Fukami, H., Chen, C. A., Berumen, M. L., ...Terraneo, T.I.,… & Zayasu, Y. (2016). When forms meet genes: revision of the scleractinian genera Micromussa and Homophyllia (Lobophylliidae) with a description of two new species and one new genus. Contributions to Zoology, 85(4). Contribution: I helped collecting samples and took part in the molecular lab work and molecular analyses. I finally contributed with the ms writing.

249

LIST OF APPENDICES

Appendix 2.1 List coral specimens examined in the study. For each sample, voucher code, morphological identified species, sampling locality, and

GenBank accession numbers at each analyzed locus are listed. SA stands for Saudi Arabian Red Sea where samples were collected.

Appendix 4.1 In vivo colonies of Porites morphologies collected for this study. a) Porites fontanesii, b) Porites columnaris, c) Porites sp. 1, d) Porites rus, e) Porites monticulosa, f) Porites lutea, g) Porites lobata, h) Porites annae, i) Porites echinulata, j) Porites solida, k) Porites harrisoni, l) Porites sp.

2, m) Porites sp. 3, n) Porites somaliensis.

Appendix 4.2 List of coral specimens successfully examined in the study in at least one locus. For each sample, voucher code, morphological identified species, sampling locality, and Ena accession numbers at each analysed locus are listed.

Appendix 4.3 Distribution maps and sampling localities at 16 sites around the Arabian Peninsula of Porites molecular lineages recovered in this study. a) P. fontanesii – clade I, b) P. columnaris – clade II, c) Porites sp. 1 – clade

III, d) P. rus, P. monticulosa – cladeIV, e) P. annae, P. echinulata, P. harrisoni,

P. lobata, P. lutea, and P. solida – clade V, f) Porites sp. 2– clade VI, g) P. sp.

3 – clade VIII, h) P. somaliensis – clade VIII. Numbers from 1 to 16 refer to the sampling localities. 1 to 8 Red Sea; 9 to 14 Gulf of Aden and Socotra Island;

250

15 Gulf of Oman; 16 Arabian Gulf. Lighter filled dots correspond to published morpho-molecular data regarding the species occurrence.

Appendix 4.4 Samples voucher codes, number of reads after trimming, and percentage of reads mapped to reference sequence. T_reads = trimmed reads; acov = mean coverage; sdcov = standard deviation of coverage; refseq = percentage of reference sequence covered.

Appendix 4.5 Comparison of reconstructed species trees. a) coral mitochondrial CR, b) coral histone region, c) coral rDNA, d) coral mitochondrial DNA. Node values represent BI posterior probabilities and ML bootstrap supports. Roman numbers from I to VIII refer to the assigned clade numbers.

Appendix 4.6 Comparison of RAxML tree based on a) “holobiont-min” dataset, that allowed for 5% missing data and consisted of 10918 SNPs, b)

“coral-min” dataset, that allowed for 5% missing data and consisted of 4140

SNPs. Values at nodes represent ML bootstrap supports. Roman numbers from I to VIII refer to the assigned molecular clade numbers.

Appendix 4.7 Comparison of Principal Component Analyses (PCA) obtained from 27 combinations of the coral SNPs mean depth – cv = 3, 5,

10, missing data – ms = 50%, 25%, 5%, minimum distance – md = 5, 10, 300.

For each panel from a) to iii) PCA results based on (left) PC1 (x axis) and

PC2 (y axis), (right) PC3 (x axis) and PC4 (y axis). Color codes are explained in the legends.

251

Appendix 4.8 Comparison of Admixture plots from the coral dataset obtained from 27 combinations of SNPs filters. In particular: mean depth – cv = 3, 5, 10, missing data – ms = 50%, 25%, 5%, minimum distance – md = 5,

10, 300. Colour codes are explained in the legends. In each figure from a) to iii) from top to bottom: K=5, best suggested model, K=4, second most probable model, K=8, third most probable model. Colour codes are explained in the legend. Clade from I to VIII refer to the assigned molecular clades.

Appendix 4.9 CV Error calculations obtained during Admixture runs, from 27 combinations of the coral SNPs filters. In particular: mean depth – cv = 3, 5,

10, missing data – ms = 50%, 25%, 5%, minimum distance – md = 5, 10, 300.

Appendix 4.10 Species tree of Porites. The claudogram represents the posterior distribution of species trees inferred with SNAPP, based on a priori imposition of 14 taxa, 1107 unlinked biallelic SNPs and 0% missing data. High colour densities are representative of high topology agreement in the species tree. Values on branches represent posterior probabilities.

Appendix 5.1 List of coral specimens examined in the study. For each sample, voucher code, morphological identified species, sampling locality, and GenBank accession numbers at each analyzed locus are listed. SA stands for Saudi Arabian Red Sea where the samples were collected.

Appendix 5.2 Details on molecular methods.

252

Appendix 5.3 SymPortal outputs and environmental data

(https://drive.google.com/open?id=1TAdZe_WGXWxlb5e8WmKBWXKIgYE-

82aM).

Appendix 5.4 Bi-plot CAP ordination of Porites (divided for location, molecular clades and morphological species) and Symbiodiniaceae clade (a-c), majority ITS2 sequence (d-f) and ITS2 type profiles (g-i) in relation with the four environmental variables studied

(https://drive.google.com/open?id=1TAdZe_WGXWxlb5e8WmKBWXKIgYE-

82aM).