Information to Users

INFORMATION TO USERS

This manuscripthas been reproduced from the microfilm master. UMI films the text directly from the original or copysubmitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from anytype of computerprinter.

The quality of this l'eproductioD is dependent upon the quality ef the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adverselyaffect reproduction.

In the unlikely.event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to rightin equal sectionswith small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMIdirectly to order.

UMI A Bell & Howell Information Company 300 North Zeeb Road. Ann Arbor. MI48106-1346 USA 313!761-47oo 800:521-0600

------_._.."., ... __ .. -.~-~.~'-~-~='=====~~~ ------,,--~-~.,----

A MOLECULAR PHYLOGENETIC ANALYSIS OF

REEF-BUILDING CORALS

A DISSERTATION SUBMIt lED TO THE GRADUATE DIVISION OF THE UNIVERSITY OF HAWAII IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OF

oocroaOF PHILOSOPHY

ZOOLOGY

MAY 1995

Sandra L. Romano

Dissertation Committee:

Stephen R. Palumbi, Chairperson H. Gert deCouet James E. Maragos Robert H. Richmond Rebecca L Cann

------~---_.~. _...~- ._~ .. _------~-- ---~ UMI Number: 9532623

This microform edition is protected against unauthorized copying under Title 17, United States Code.

iii

------... to my Mom (1940-1993) and my Daddy

------_._- ACKNOWLEDGMENTS

Financial support for this research has come from NSF grants to Dr. Stephen Palumbi, a NIH grant to Dr. Robert Richmond, the Dept. of Zoology, the Edmondson Research Fund, Sigma Xi, the Constance Endicott Hartt Fund through the American Associaton of University Women, and John and Veronica Romano. I gratefully acknowledge the persons who obtained coral samples for me: Dr. Robert Richmond, Dr. James Maragos (who perfected his 'coral chop' in the Solomon Islands), Dr. Bruce Carlson (who risked pens, sunglasses and clean clothes in the pruning of his pet corals) of the Waikiki Aquarium, Dr. Stephen Cairns of the National Museum of Natural History, Dr. David Krupp, Dave Gulko, and Richard Chock. I thank my advisor, friend, and mentor, Steve Palumbi, for giving freely of his time and energy through all phases of this research. He has provided me with endless enthusiasm, inspiration, and support. I thank my committee members who have provided advice, support and encouragment when I've needed it most in seeing this project through. Charlie Veron provided valuable insight into and helpful discussion about coral evolution at a crucial time during analyses. I acknowledge Cindy Hunter for her technical assistance. Robert Hixson and Chris Simon provided advice on secondary structure. George Roderick provided assistance with figures. The members of the Palumbi Lab have been instrumental in this project. They have all provided. camaraderie and technical expertise in the lab as well as helpful comments and discussion as I was trying to understand the twists and turns of my dataset. I thank them for their assistance (especially in overseeing my mini-gels), support, and encouragement. Bailey Kessing and v Andrew Martin were especially valuable, and most of the time patient, as I " made my first forays into the laboratory and the world of molecular biology. Ed Metz and Owen McMillan survived four years in the lab with me and made the ex-perience lively, educational, and fun. I thank Paul Armstrong, Tom Duda, and Frank Cipriano for helping to keep life in the lab in perspective over the last year. The administrative support of Lori Yamamura, Sally Oshiro and Audrey Shintani of the Zoology Dept., as well as of Frances Okimoto and Geri Mintomi at Kewalo Marine Lab is gratefully acknowledged. I'd like to give special thanks to the following people for helping me to maintain some sense of reality as I have been writing and to 'finish up' in as pain-free and pleasant a way as possible: Steve Palumbi, the participants in GWIZ, Cindy Hunter, Kim. del Carmen, Theresa Cabrera, Krista Ingram, Flo Thomas, Rosie Gillespie, Leanne Fernandes, Dr. Erik Meesters, Dave Hopper, Jean-Sylvain Bussac, George Roderick, Sweet William Roderick, and the members of the Delaware Expedition to Hawaii. Leanne Fernandes and Dr. Erik Meesters, while trying to be as unobtrusive as possible in our common living space so as not to disturb me, contributed enormously to maintaining my sanity and peace of mind on a daily basis in the final stages of writing this dissertation. Finally, I thank my family for their support in every way throughout my years as a graduate student.

------_._------ABSTRAcr

Review of the evolutionary history of scleractinian corals demonstrates the great deal of knowledge gained from their 240 million year fossil record, and from studies of all aspects of their biology and ecology. By contrast, forces driving the evolution of corals are not well understood due to the great variability that exists in characters used in classification and to difficulties in working with fossil specimens. Traditional hypotheses about relationships among families and suborders are tenuous and somewhat contradictory. Molecular techniques were used for the first time to evaluate relationships among corals. A segment of the mitochondrial16S ribosomal gene from 34 species in 14 families was amplified via the Polymerase Chain Reaction and sequenced. These sequences are homologous to each other and to other metazoan 165 ribosomal sequences in terms of nucleotide composition, patterns of transition and transversion substitutions, spatial patterns of substitutions, rates of divergence, and secondary structure. Parsimony analysis of the aligned sequences results in a phylogram whose major groups are supported at bootstrap values of 80% or higher. Comparisons of molecular and morphological hypotheses demonstrate that 1) this molecular topology is robust by itself, 2) it is robust in relation to traditional morphological topologies, and 3) available morphological characters are not a reliable tool for inferring family and suborder relationships. Although this molecular hypothesis supports traditional hypotheses for relationships within genera and families, clades on the molecular phylogram do not correspond to morphologically based family and suborder relationships. This molecular hypothesis suggests vii two major lineages of Scleractinia diverged from a common ancestor at least 300 mya. One of these lineages, the robust corals, tends to be heavily calcified and less architecturally complex forming predominantly massive or plate-like colonies which grow by intratentacular budding. The second lineage, the complex corals, is characterized by less heavily calcified skeletons that are architecturally complex, forming predominantly branching colonies which grow by extratentacular budding. A switch between hermaphroditism and gonochorism appears to have occurred at least 3 times. Some of the recent changes in scleractinian taxonomy are supported by this molecular hypothesis which also suggests new relationships among families to be considered.

viii

------_. _._- TABLE OF CONTENTS

ACKl\TOWLEDGMENTS v

ABSTRACT ., vii

LIST OF TABLES ...... •...•...... •...••..••...... •...•..••..•...... •.....•...... xiii LIST OFFIGURES xiv

IN'TRODUCTION 1

CHAPTER 1. The Evolutionary History of Scleractinian Corals: A

Review 0 •••••••••••••••••••••••••••••••••••••••••••••••••••••4

Introduction 4

Evolution of Scleractinians through geological time 5

The Triassic 7

The Jurassic 10

The Cretaceous : 11

The Paleogene 14

World climate and geography during the Neogene 15

The Miocene 16

The Pliocene 17

Evolution of biogegograpic provinces 18

Distribution of azooxanthe11ate corals 19

Distribution of reef-building corals 21

The Indo-Pacific. 22

Distribution of reef-building corals in the Atlantic 25

Theories of coral distribution 28

Evolution and taxonomy of scleractinian groups 31

ix Scleractinian evolution as described by Wells (1956) 33

Scleractinian evolution as described by Veron (1995) 38

Scleractinian classification according to Alloiteau (1952) 42

Scleractinian classification according to Chevalier and Beauvais (1987) 43

Scleractinian evolution based entirely on microstructural analysis 45

Summary of classifications 46

Non-morphological methods of scleractinian classification 47

Quantitative analyses of relationships among the Scleractinia 48

Biochemical studies of scleractinian evolution 49

Summary 51

CHAPTER 2. Molecular Evolution of a Portion of the Mitochondrial 16S Ribosomal Gene Region in Scleractinian Corals 63 Abstract 63

Introduction 64

Methods 66

DNA Extractions 66

Polymerase Chain Reaction 69

Cloning e 70

Sequencing 72

Alignment and secondary structure 72

Phylogenetic analyses 73

Results 74

Extractions and PCR 74 x

------Alignment and secondary structure 75 Sequence composition and size variation 76 Characteristics of robust and complex clades 77 Transitions and transversions 79 Spatial pattern of variability 79 Phylogenetic analyses 80

Discussion 82

Se-condary struct:ure 0 •••••••••••••••••••••••••••••••••••••••••••••••••••••83 Evolutionary rates of the Scleractinia 84 Patterns of molecular evolution of scleractinian sequences 85 B Sequences 86 Alternative hypotheses : 87 Evolutionary relationships among the Scleractinia 90 CHAPTER 3. Evolution Among Families and Suborders of Scleractinian Corals: J:t. New Hypothesis Based on Mitochondrial 16S Ribosomal Gene Sequences 124 Abstract 124

Introduction 0 125 Molecular analysis of scleractinian phylogeny " .127 Comparison of traditional and molecular hypotheses ., 128

Traditional hypotheses for relationships among the Scleractinia 128 A molecular hypothesis for relationships among the Scleractinia 130 Agreement of traditional and molecular hypotheses .131

------._-- Discrepancies between traditional and molecular hypotheses 132 Quantitative morphological analyses 135

Evolution of Corals from a Molecular Perspective l38

Evolutionary patterns in the Scleractinia l38 The molecular hypothesis and taxonomic classification l42 Summary of evolution in the Scleractinia 145 CONCLUSIONS 156

APPENDIX. 158

LITERATORE OTED 174

xii

...... _---- _.._---- -_._- --- •.__• 0-·- USTOFTABLES Table Page

1.1 Timeline of important events in the evolutionary history of corals 53

1.2 Classification scheme of Alloiteau (1952) for the order Scleractinia " 55

1.3 Classification scheme of Chevalier and Beauvais (1987) for the order Scleractinia 56

2.1 Species of Scleractinia sampled , 91

2.2 Base composition, substitutions and distances among all sequences and within major clades 93

2.3 Amplification of short and long sequences 94

2.4 Divergence dates, genetic distance, and divergence rates 95

3.1 Morphological characters and states l46

xiii

------LISTOF FIGURES Figure Page

1.1 Wells' (1956) hypothesis for the evolutionary history of the Scleractinia 57

1.2 Veron's (1995) hypothesis for the evolutionary history of the Scleractinia 59

1.3 Roniewicz and Morycowa's (1995) hypothesis for the evolutionary history of the Scleractinia 61

2.1 Diagram of the mitochondrial 165 ribosomal gene region amplified 96

2.2 Alignment of 37 scleractinian mitochondrial 165 ribosomal sequences and 2 cnidarian outgroups 97

2.3 Hypothesized secondary structures 108

2.4 Molecular phylogram of coral genera 116

2.5 Cladogram of relationships among cnidarians and oilier metazoans 118

2.6 Relationship between genetic distance and numbers of transition and transversion substitutions 120

2.7 Genetic distance in 60 bp windows 122

3.1A Relationships among scleractinian genera as hypothesized by Veron (1995) 149

3.1B Relationships among scleractinian genera, based on Veron (1995), sampled in this study 151

3.2A Molecular phylogram of relationships among genera of corals 153

3.28 Molecular phylogram of relationships among families of corals 155

xiv

------INTRODUcnON

The goal of evolutionary biology is to understand the patterns and processes of diversification of living things. A foundation for this understanding is a biological classifi~tion that reflects hypotheses about the evolutionary relationships of the taxa being studied, a phylogeny (Wiley 1981). Understanding the evolutionary history of a group of organisms is fundamental to the understanding of all aspects of the biology of that group. Scleractinian corals are a group of relatively simple organisms about whose biology we know a great deal. However, the evolutionary foundation for understanding the biology of these corals, based on morphological characters and the fossil record, is not a strong one. The goal of this dissertation research has been to strengthen this foundation by using methods that were not previously available, in order to better understand the e~olutionary relationsips among corals. This research has grown out of my interests in evolutionary biology and my fascination with coral reefs. Any amount of time spent underwater on a coral reef gives one the opportunity to observe the almost unending amount of morphological variability that exists among coral skeletons. These very simple animals take on myriad shapes and have played a major role in the construction of the often massive structures that are coral reefs. Their variability is not limited to the growth form of the entire colonf but can also be observed among the corallites (the cups created by individual polyps) that form the coral skeleton as well as within the structure of the corallite itself. Perhaps not surprisingly, classification of corals is based on skeletal characters because they have been, for the most part, the only characters available to coral biologists and taxonomists for comparative purposes. Many of the 1

------naturalists who first studied corals had never observed the living animal. Our knowledge of corals has greatly increased since the Possibility for observing corals in situ, due largely to the development of SCUBA. Studies of living corals in situ have greatly enhanced our understanding of morphological variability in corals and made scleractinian taxonomy somewhat tractable. As a teenager living in the Caribbean I had the opportunity to spend many happy hours exploring the coral reef environment. The fascination that I found with this diverse underwater world has always remained and is what motivated me to become a biologist. As a Masters student I studied questions about the ecology of corals and became intimately aware of the large amount of variability in corals, an awareness that was greatly increased by a long visit to the Great Barrier Reef. During this period, I was exposed to recently developed molecular techniques being used for studying-the population ecology and evolution of marine organisms. I came to realize that these techniques might provide the tool for eventually being able to understand the morphological variability inherent in corals. During a lab discussion, in response to my frustrations with coral taxonomy, Dr. Palumbi's words of solace were "There's the lab Sandra." It seemed so simple then. With intentions of simplicity and a desire to establish a framework for studying coral genetics, the goal of this dissertation research has been to use molecular techniques to reevaluate our understanding of the phylogenetic relationships among families and suborders of corals. In Chapter 1, I review the literature concerning the evolutionary history of corals. This includes a summary of 1) the evolution of the Scleractinia through geological time, 2) the biogeographical history of the group and the current theories used to

------explain this history, 3) traditional hypotheses about the evolution of relationships in the Scleractinia, and 4) the use of other than morphological characteristics in coral taxonomy. This review provides the information necessary for reevaluating our understanding of the evolution of corals. The development of molecular techniques for studying the relationships among families and among suborders of corals is described in Chapter 2 along with analyses of the data obtained with these techniques. In Chapter 3 the molecular hypothesis provided by these techniques is compared to traditional hypotheses of relationships among the Scleractinia and is used as the basis for evaluating patterns in the evolution of corals.

------.-.... CHAPTER 1. The Evolutionary History of Scleractinian Corals: A Review

Introduction

The existence of scleractinian corals as a related group of organisms (Phylum Cnidaria, Class Anthozoa, Order Scleractinia) was not recognized until the 16th century (Vaughan and Wells 1943). Originally classified as plants, the Scleractinia were not recognized as animals until the late 1700s (Vaughan and Wells 1943). Perhaps as a reflection of the confusion about the nature of corals, the name Zoophyta (Gr.zoon, animal, phyton, plant) was applied to them through the 18005 (Hyman 1940). Some of this confusion was perhaps due to the facts that these delicate polyp animals are somewhat flower-like in appearance and the calcium carbonate structures that they secrete take on a myriad of shapes. Confusion in scleractinian classification probably reached its apex in the early 1900s when Bernard abandoned the Linnaean system of nomenclature in his taxonomic work on corals (Veron and Pichon 1982), and this confusion still exists today. Skeletal characters have always been the most commonly used characters in taxonomic studies of the Scleractinia even though corals are known for their morphological plasticity (Lang 1984). Taxonomic difficulties have resulted in a lack of confidence in scleractinian systematics and as a result, the evolutionary history of the Scleractinia remains poorly understood. Here, I review the history of the Scleractinia over geological time, our understanding of the biogeography of the group, and our understanding of the evolution of the animals themselves in order to present the highlights of the evolutionary history of scleractinian corals. Recent developments in biochemical techniques and quantitative phylogenetic analysis are providing 4

---- ~------new tools for studying the evolutionary history of corals. This review will provide a framework for using these tools to synthesize a new understanding of the evolution of the Scleractinia.

Evolution of Scleractinians through geological time

The classical view of metazoan phylogeny (Hyman 1940, Hill and Wells 1956), which is also supported by recent molecular studies (Lake 1989, Erwin 1991, Lake 1991, Wainright et at. 1993), holds that the phylum Cnidaria is composed of the nearest living representatives of the ancestral stock of all living metazoans except sponges. There is fossil evidence that all three classes of the phylum were present in the Ediacaran fauna of the Vendian stage of the late Precambrian (Scrutton 1979, Jenkins 1984, Scrutton 1984) about 550 mya (Runnegar 1992). These classes are thought to have diverged from each other relatively early and there has been considerable debate over how they have evolved (Hyman 1940, Pantin 1959, Hand 1966, Werner 1973, Schmidt 1974, Scrutton 1979, Grasshoff 1984, Robson 1985,Ax 1989, Willmer 1990). Although traditionally the class Hydrozoa has been considered ancestral in the Cnidaria (Hyman 1940, Hill and Wells 1956), more recent studies indicate that the Anthozoa is probably the basal group of the phylum because of their bilateral symmetry, a relatively simple life cycle, an inability to cope well with physiologically difficult environments, the relative simplicity of their nematocysts (Werner 1973, Willmer 1990), possibilities for evolutionary change determined by soft-body construction (Grasshoff 1984), and circular

mitochondrial DNA structure (in contrast to the linear mitochondrial DNA of the other 2 classes of Cnidaria, Bridge et al. 1992).

------_ ...._....._..... _.. ------_.._- ... The ancestor of the anthozoans is thought to be a skeletonless polyp with eight mesenteries, very weakly developed musculature and eight tentacles (Hill and Wells 1956). Unlike other cnidarians, no medusoid stage occurs in anthozoans. The class is divided into three subclasses: the Octocorallla, the Hexacorallia, and the Ceriantipatharia (Brusca and Brusca 1990). The Hexacorallia includes what are commonly known as sea anemones, soft corals and stony corals. These polypoid animals may be solitary or colonial and have paired mesenteries occurring in multiples of six. In this class, the Order Scleractinia consists of the true (or stony) corals that secrete a calcium carbonate skeleton. Scleractinia are generally categorized ecologically as either hermatypic (from herma meaning reef) or ahermatypic, although the precise meaning of this distinction is not always clear because of the different ways that these terms have been used (see Schuhmacher and Zibrowius i985). As originally defined (Wells 1933) hermatypic refers to reef-building corals that harbor endosymbiotic dinoflagellates, known as zooxanthellae, while ahermatypic refers to non-reef-building corals (Wells did not state whether or not such corals could possess zooxanthellae). Hermatypic corals, in this sense, are responsible for the construction of coral reefs and are limited to the warm, shallow waters of the tropics and depths of about 100m (Rosen 1981). However there are some 'hermatypic' (i.e, they possess zooxanthellae) corals (such as Madracis, Manicina, Fungia, Diaseris, Cycloseris, and Heieropsammia) that are not major contributors to reef frameworks and occur in temperate areas or deep water (Beauvais and Beauvais 1974, Schuhmacher and Zibrowius 1985). Although ahermatypic (i.e, not possessing zooxanthellae) corals are never major contributors of shallow reef

_._- -- .__._------_. _._- frameworks, genera like Dendrophyllia, Lophelia, Desmophyllum, Solenosmilia, Oculina, and Tubastraea may occur on reefs or even form reef like structures in deep water (Beauvais and Beauvais 1974, Cairns and Stanley 1981, Schuhmacher and Zibrowius 1985). Generally, ahermatypic corals have a more extensive geographic range than hermatypic corals. They are found in all oceans, in depths from 0 to 6238 m, and at temperatures from -1.1°C to over 290C (Cairns and Stanley 1981). Approximately half of scleractinian genera are ahermatypic and most families have representatives of both groups (Wells 1956, Rosen 1981). Due to the absence of reliable data allowing the distinction of 'hermatypic' and "ahermatypic' corals in the fossil record, Beauvais and

Beauvais (1974 ) did not use these terms at all. In an effort to avoid some of the confusion associated with these terms they will not be used in this review. The terms reef-building and non-reef building will be used to denote the ecological role of particular corals in framework or bioherm building [hermatypic and constructional sensu (Schuhmacher and Zibrowius 1985)] regardless of whether they possess zooxanthellae. The terms zooxanthellate or azooxanthellate will be used to indicate the presence or absence of zooxanthellae. The Triassic The Scleractinia make their first appearance in the shallow waters of the Tethys Sea in the mid-Triassic approximately 240 mya (Stanley 1981). At the opening of the Triassic the major land masses of the world were united in the supercontinent Pangaea (Briggs 1987, Stanley 1989). The Tethys Sea was an embayment of the Panthalassa Ocean projecting into the area of equatorial Gondwanaland where the Mediterranean is found today (Briggs 1987, Stanley

------1989). The Scleractinia appear suddenly in the fossil record: eight to twelve species, both of colonial and solitary forms, in five families and three suborders can be distinguished (Vaughan and Wells 1943,Stanley 1988). They are found throughout the Tethyan region as small-size colonies inhabiting soft substrata in mostly deeper, sheltered or protected environments (Vaughan and Wells 1943,Stanley 1981,Stanley 1984, Stanley 1988). During the Triassic they contributed, along with other marine invertebrates, to the construction of low mounds no greater than three meters high on the ocean floor (Stanley 1989). By the mid-Triassic these mounds were substantial enough to be designated reefs and were composed of sc1eractinians, algae, sponges, and bryozoans (Stanley 1989). By the late Triassic these organic reefs had become much larger, some consisting of more than 20 different species of reef-building organisms (Stanley 1989). Though the role of sc1eractinians in these reefs was limited (Newell 1971,Stanley 1981), they did contribute to reef framework building (Stanley 1988). By the end of the Triassic the Scleractinia had diversified (Newell 1971, Stanley 1981): there were over 200 species (Beauvais 1982) in about 50 genera (Beauvais 1984). This was not a cosmopolitan fauna but a fairly endemic one divided into four provinces: 1) the Alpine and Near-East province, 2) the Indonesian province, 3) the Japanese province and 4) the American province (Beauvais 1982). At this time the European region appears to have been a center of diversity for the Scleractinia as they spread both to the east and the west (Beauvais 1982). It is generally thought that the first sc1eractinians were azooxanthellate and non-reef-building (Wells 1956, Cairns and Stanley 1981). The larger size of reefs at the end of the Triassic is taken as evidence for the beginning of the symbiotic relationship between reef corals and algae (Newell 1971, Stanley

------. -- 1988, Stanley 1989). G.D. Stanley (1981) and Krasnov (1984) cite evidence that at least one coral species was constructing massive reefs in the late Triassic. The presence of algal symbionts in fossil corals cannot be determined directly as the symbionts themselves are not preserved as fossils. Cowen (1988) concludes that all reef-builders in the last 550 million years harbored endosymbionts based on the observations that scleractinians from the latest Triassic and early Jurassic meet specific criteria, such as suitable skeletal structure and paleoenvironmental distribution, that can be used to infer symbiosis in fossils (Cowen 1983). Coates and Jackson (1987) found that in living corals the occurrence of zooxanthellae is correlated with morphological characteristics (growth form, corallite size, connections between corallites). They recognized these same characteristics in several fossil coral faunas and came to the conclusion that the fossil faunas of the Jurassic and Oligocene had the same characteristics as extant zooxanthellate corals. All of this evidence seems to indicate that at least one scleractinian group had acquired algal symbionts by the latest Triassic and that the acquisition of symbionts lead to subsequent scleractinian diversification and radiation (Stanley 1981). In the late Triassic, mass extinctions that affected 35% of the major animal fossil groups (Newell 1967, Stanley 1989) occurred throughout the world for unknown reasons. Twenty percent of all families of marine animals, including corals, disappeared (Sepkoski 1982). As a result of this extinction event the Jurassic opened with a relatively impoverished marine fauna (Stanley 1989). There was a period of 4-10 million years when no reefs are found (Beauvais 1984,Stanley 1988). Beginning at the end of the Triassic and then throughout the Jurassic, Pangaea slowly fragmented (Briggs 1987, Stanley 1989). As Pangaea fragmented during the Triassic, the Tethys Sea

9 extended into the growing rift between Africa and Eurasia (Tollmann and Kristan-Tollmann 1985, Stanley 1989). It continued extending westward between North and South America as these two continents separated during the Jurassic (Stanley 1989), providing a favorable environment for the spread and diversification of the Scleractinia.

The Jurassic Most of the corals of the Lower Jurassic were solitary species or small colonies scattered in the sediments (Beauvais 1984). The Scleractinia of this period are represented by what are thought to have been reef-building families, in eight suborders, that are now extinct (Beauvais 1984). At this time, representatives of the primarily non-reef-building and azooxanthellate family Caryophylliidae first appear, marking the differentiation of shallow-water and deep-water coral assemblages and establishing distinct lineages of reef building and non-reef-building corals (Wells 1956, Calms and Stanley 1981, Stanley 1981). By the mid-Jurassic, more reef-building genera had appeared, including representatives of the extant Acroporidae, Oculinidae, and Faviidae, and many species had cosmopolitan distributions (Beauvais 1984). During the Jurassic, sea level rose and marine waters spread rapidly over land (Stanley 1989). By the late Jurassic, the prevailing world climate was uniform and warm (Beauvais 1973, Beauvais 1984). Under these favorable conditions, reef accretion occurred in the tropical zone which extended considerably further north than today (Wells 1969). The Scleractinia became highly diversified and important members of reef communities (Newell 1971, Stanley 1981). The genera of the Upper Jurassic were cosmopolitan, and were mostly colonial (Beauvais 1973) although many include only one species (Beauvais 1984). These corals represent a transition fauna between the more

------ancient groups of the Mesozoic and the more modem groups of the Cenozoic (Roniewicz 1976). They are characterized by morphological features that are on the one hand unique to the Triassic and Jurassic, and on the other hand the precursors to the corals of the Cenozoic (Roniewicz 1976).The particular character of this Jurassic fauna is thought to be due to the environment inhabited by scleractinians at this time, an environment that included fairly unstable sediments in calm waters with high amounts of sedimentation (Roniewicz 1976). The most important development for the Scleractinia during the Jurassic was the evolution of diverse types of colony morphology that include a high level of integration among polyps and which led to modem framework builders (Roniewicz 1976). By the end of the Jurassic there were over 130 genera of corals, although none of these genera continued into the Tertiary (Beauvais 1984). The late Jurassic is also the time of the first appearance of coralline algae which in association with corals enhanced binding and rigidity of the reef framework (Heckel 1974). The CretaceQus By the early Cretaceous the Tethys Sea was a circumglobal tropical waterway, with a westward-flowing current (Briggs 1987,Stanley 1989).In the Early Cretaceous there was a trend of warming global temperatures (Stanley 1989) while throughout the Cretaceous there was a global rise in sea level resulting in 90% of the earth being covered by the oceans (Newell 1971, Stanley 1989). Tropical conditions prevailed over most of the world (Briggs 1987). The final breakup of Pangaea and the separation of its daughter continents during this time led to the widening of the Atlantic Ocean and the opening of the Caribbean Sea and the Gulf of Mexico (Briggs 1987, Stanley 1989).

----~-~~------Although reef-building was not very intensive in the Early Cretaceous (Beauvais and Beauvais 1974), and there is little or no fossil evidence of reef development anywhere (Wells 1956, Newell 1971), scleractinians apparently flourished and diversified throughout the increasingly favorable environment of the Tethyan Seaway (Stanley 1989). Reefs that did exist were a combination of both scleractinian corals and rudists (Newell 1971), marine bivalve mollusks that flourished during the Cretaceous in the same environments as corals, forming large aggregations and reefs. Rudists are thought to have dominated the reefs of the Early to mid-Cretaceous due to a combination of environmental changes that allowed these eurytopic organisms to flourish while the more stenotopic corals languished (Scott 1984). Rudists also grew faster and were more efficient competitors for space in the reef framework (Coates 1977). During this time it is thought that non reef-building corals living in the varied environments of the neritic and bathyal zones increased in number (Wells 1956). Early in the late Cretaceous (about 90 mya) there was an episode of extinctions that eliminated many marine species, including the reef-building rudists, and reduced the number of scleractinian genera by about 10% (Newell 1971). These extinctions are thought to be due to changes in climate to which the Cretaceous organisms were not adapted and were heaviest in tropical regions (Newell 1971,Stanley 1989). The disappearance of the rudists allowed for the subsequent success of the scleractinian corals (Newell 1971). In the late Cretaceous, the Tethyan reef-coral fauna extended from the West Indies and Mexico through the Atlantic and Indian Oceans to Malaysia. This fauna included early representatives of a few of the modem scleractinian families including the Podlloporidae and the Poritidae (Wells 1956). Rosen and Smith

------_. -- (1988) have found evidence for faunal divergence in the central Tethys region, among corals as well as echinoids, beginning in the Cretaceous. In addition non-reef-building corals were now firmly established in nearly every part of the littoral, neritic, and bathyal environments (Wells 1956). By the end of the Cretaceous there is evidence for a filter bridge in the form of an archipelago between North and South America (Briggs 1987). The influence of the newly formed deep Atlantic Ocean is evident in regional differences in coral reef communities of this period (Newell 1971) when three distinct faunal regions of Scleractinia were distinguished: 1) a northern realm of non reef-building corals found in Canada, the United States, and Northern Europe; 2) a Caribbean-Mediterranean realm of reef-building corals; and 3) an Indo-Pacific realm of non-reef-building corals found in Southern India, Madagascar, and New Zealand (Beauvais and Beauvais 1974). At the end of the Cretaceous there was a fall in sea level throughout the world such that the Tertiary began with the sea at an unusually low level (Adams 1981). During the Tertiary, the Tethyan Seaway, still providing a connection between the Atlantic and Pacific Oceans, slowly narrowed (Adams 1981). As Africa and India moved northward the Tethyan Seaway was eventually closed in this region in the Miocene, 15-19 mya (Adams 1981). Although the Atlantic and Pacific Oceans were not definitively separated by the Isthmus of Panama until the Pliocene, there were many volcanoes and other ephemeral islands in that region throughout much of the Tertiary (Adams 1981,Briggs 1987) which probably restricted exchange of marine organisms between these two oceans beginning in the Miocene (Condray and Montaggioni 1982).

._------_.- ._.- The Paleogene Coral reefs were greatly affected by the worldwide mass extinctions that occurred at the end of the Cretaceous. At this time 47% of all animal genera worldwide disappeared (Iablonski 1991). Numbers of scleractinian genera declined from approximately 90 to 35 (Newell 1971). Reef communities were not quick in recovering from these extinctions (Newell 1971, Stanley 1989); conditions for reef formation were unfavorable (Wells 1956). No reefs are seen in Paleocene (65-57 mya) rocks throughout most of the world (Newell 1971,Stanley 1989).Even during the Eocene (58-37mya), reefs continued to be sparse and poorly developed although new genera of corals did appear and the Scleractinia continued to diversify (Newell 1971). During the Paleogene (65-24 mya), scleractinian diversity was greatest in the Atlantic and Caribbean region, with rich faunas in the Mediterranean as well (Rosen 1988b). At this time both ancient and modern corals existed (Gerth 1930). By the end of the Eocene almost all of the modem families of scleractinians had emerged (Gerth 1930, Wells 1956, Rosen 1988b). The non-reef-building families were thought to be developing rapidly at this time (Wells 1956). By the end of the Eocene all 11 families that include non-reef-building and azooxanthellate forms had appeared. The Eocene ended with climatic cooling, thought to be due to polar glaciation, and a resulting global fall in sea level (Adams 1981). The beginning of the Oligocene (37-24 mya) is marked by a decrease in scleractinian diversity that is also seen in mollusks and phytoplankton around the world (Newell 1971). There was, however, widespread development of fringing and patch reefs throughout the tropics during this time (Wells 1956, Frost 1977a). By the end of the Oligocene" there were increasing differences between the faunas of 14

-----~------the Caribbean and the Mediterranean region, probably an effect of the spreading of the Atlantic Basin (Newell 1971).

World climate and geography duMS the Neogene By the early Neogene (24 mya to present) the Tethyan Seaway was

reduced to a narrow channel connecting the Indian Ocean to the proto Mediterranean, and the continents were very close to their present day positions (Adams 1981). The final closure of the Tethys occurred 15-19 mya in the mid-Miocene, as sea level decreased and Africa and Eurasia were joined, physically isolating the faunas of the Mediterranean and the Indo-Pacific, and those of the Atlantic region (Newell 1971, Chevalier 1975, Adams 1981, Veron 1985, Stanley 1989). However faunal differentiation between these two regions had been evident since the late Cretaceous (Rosen and Smith 1988). The Mediterranean at this time was a vast tropicalI sub-tropical basin where significant reef-building was occurring (Chevalier 1975). The Oligocene coral fauna of the Mediterranean was replaced by a fauna similar to the modern fauna (Chevalier 1975). However, as the European climate cooled during the Neogene (attributed in tum to the breaking up of the North Atlantic continent), the number of genera and species diminished progressively such that by the upper Miocene there were no reef-building corals left in the Mediterranean (Chevalier 1975). At the close of the Miocene (5 mya), the 'Messinian Salinity Crisis' occurred. A global fall in sea level resulting from rapid glacier formation coupled with evaporation resulted in the Mediterranean almost entirely drying up, temporarily isolating it from the Atlantic and leaving saline lakes (Adams 1981, Stanley 1989). This increase in salinity killed most marine organisms, including all of the scleractinians (Adams 1981, Coudray and Montaggioni 1982). The Straits of Gibraltar opened

------_._..... -'" ._- .. _. in the Pliocene (5-1.8 mya), refilling the Mediterranean basins from the Atlantic (Adams 1981). Throughout the Neogene, "Northern India and the Indonesian Archipelago were in a state of tectonic turmoil (Adams 1981)".

The Miocene These important geographic changes set the stage for the continued differentiation of coral faunas in the world's oceans and greatly influenced the evolution of the Scleractinia which flourished during the Neogene. Coral reefs thrived in the Miocene as algal ridges appeared for the first time (Stanley 1989). The long association of corals with coralline algae had resulted in wave-resistant coralline algal-ridge and spur-and-groove communities at the reef crest and seaward margin respectively that formed a protective barrier from powerful oceanic waves (Stanley 1989). By the Miocene, three separate coral faunas were clearly distinguished from one another 1) in the Atlantic and Eastern Pacific; 2) in the Mediterranean; and 3) in the Indian Ocean and western Pacific (Newell 1971,Chevalier 1975,Coudray and Montaggioni 1982, Veron 1985, Rosen 1988b). This continued differentiation of the Atlantic and Indo-Pacific biogeographic provinces appears to be the result of the closure of the Tethys, changing ocean circulation patterns, and the large expanses of the Atlantic and PacificOceans acting as filter barriers (Newell 1971,Coudray and Montaggioni 1982). The very favorable climatic and hydrographic conditions of the Miocene tropical seas were centered in the Caribbean and Australasian areas which appear to have been (and still are today) centers of diversity for these provinces (Newell 1971). Exchange between the coral faunas of the central Pacific and the Caribbean was practically nonexistent during this time although the Isthmus of Panama did not form until the Pliocene (Chevalier 1987). Overall diversity declined in the Atlantic/Caribbean region and rose in

16 the Indo-Pacific region due primarily to a combination of extinctions in the Atlantic and an increase in speciation in the Indo-Pacific (Rosen 1988b). Since the Miocene, no species (Coudrayand Montaggioni 1982, Chevalier 1987) and only 8 genera of reef-building corals (Acropora, Madracis, Leptoseris, Porites, Siderastrea, Favia, Scolymia, Tubastraea) are common to the two major reef provinces (Veron 1993). This would suggest that the corals in these two regions have evolved independently of each other although there is some evidence for sporadic interchange (Chevalier 1975, Coudray and Montaggioni 1982).

The PliQce~ At the beginning of the Pliocene, sea level rose as global climates warmed (Stanley 1989).The Antarctic ice sheet began to form at this time so that by the end of the Pliocene global climate was cooling. The tropical belt contracted leading to the compression of the limits of reef corals to approximately their present-day distribution (Newell 1971). The Isthmus of Panama formed at this time as well, leading to the accentuation of the pattern of differentiation mnong the coral faunas of the Caribbean and the Eastern Pacific Ocean (Rosen 1988b). During this time, the Great Barrier Reef formed off the northeastern coast of Australia (Symonds and Davies 1985).It is unclear what the effect of glaciation and resulting eustatic fluctuations of sea level during the Pleistocene had on scleractinians (Wells 1956, Newell 1971, Potts 1983,Potts 1984,Potts 1985,Veron 1986). In summary (Table 1.1), scleractinian corals arose in the late Triassic although it was not until the Jurassic that they played a significant role in reef growth. Once the rudists had gone extinct in the Cretaceous, scleractinians came to dominate the reef complexes. They became well established in the

------warmth of the circumglobal Tethyan Seaway, forming one large population with little differentiation. As the Tethys broke up, different regions started to become distinct. The Atlantic basin, associated loosely with the Eastern Pacific realm, became differentiated from the Indo-Pacific. Within the Indo-Pacific, differentiation occurred between populations in the Indian ocean and those in the Central Pacific. The Mediterranean fauna became more and more isolated as the Tethys narrowed until it disappeared after the Messinian event in the Miocene. Although the Isthmus of Panama did not definitively close until 3.5 mya, fluctuations in sea level and volcanic islands limited interchange between the Eastern Pacific and the Atlantic as early as the Miocene. The two faunas of the Indo-Pacific and Atlantic are thought to have been effectively isolated from each other for the last 25 my such that the emergence of the Isthmus only accentuated or helped to maintain a pre existing pattern (Rosen 1988b). These two realms have remained separate entities and are characterized by distinct species.

Evolution of biogegograplc provinces

Scleractinian corals today are found throughout the world's oceans. The non-reef-building and deep-water corals are the most widely distributed with many cosmopolitan genera and species (Wells 1956). These corals are not readily separated into distinct faunal provinces although this may be partly due to a lack of information (Wells 1956). In contrast, reef-building corals have a relatively limited distribution, and are found circumtropically, though in the western regions of the Atlantic and Pacific Oceans they extend somewhat into the subtropics (Rosen 1984).

---~------Distribution of UOQxanthellate corals The azooxanthellate corals, which are mostly non-reef-building, are not found in distinct faunal provinces (Vaughan and Wells 1943, Wells 1956, Croizat 19B1). In general, many genera, and even species, are cosmopolitan in distribution although the deeper the normal habitat of a species the greater its geographic range (Vaughan and Wells 1943, Wells 1956). The appearance of the first caryophylliids in the Jurassic marks the beginning of the distinction between reef-building and azooxanthellate, generally non-reef-building corals (Wells 1956). The azooxanthellate corals became increasingly abundant in the mid-Cretaceous and by the late Cretaceous could be found in nearly every part of littoral, neritic, and bathyal environments (Wells 1956). Of the 11 families and subfamilies which include azooxanthellate forms, all had appeared by the Eocene (Wells 1956). Today most azooxanthellate corals live within the confines of the continental shelves (Wells 1956, Squires 1967). The azooxanthe1late fauna of three regions has been relatively well-studied and this work is described below. The azooxanthellate coral fauna of the Caribbean has been relatively well studied by Cairns (1979) although there are still large, poorly sampled areas. Cairns found that as with the reef-building corals, the center of species diversity in the Atlantic is the Caribbean, and that there is a sharp decrease in the number of species both to the north and south. The greatest species diversity of azooxanthellates is found between 20o-50Orn. The tropical western Atlantic Scleractinia found below 200m form a strong cosmopolitanI circumtropical group at the generic level (53% of 42 genera), with 14% amphi-Atlantic genera and 12% endemic genera. At the species level this group of azooxanthellates has a stronger endemic component (60% 19

------.---- .- --_._------of 90 species) and a much smaller cosmopolitan component. The shallow water azooxanthellates have a similar pattern where 59% of the 17 genera are

cosmopolitan, 6% are amphi-Atlantic, and 18% are endemic to the tropical western Atlantic. Likewise, at the species level, the shallow-water

azooxanthellates have a much stronger endemic component (74%) with only 1 of the 27 species (Tubastraea coccinea) being circumtropical. Cairns attributes the greater endemism at shallower depths in the tropical western Atlantic to the effectiveness of the Atlantic Ocean as a barrier to dispersal of shallow water corals. The high number of endemics among azooxanthellate corals has been explained as a reradiation following the interruption of the Tethys Sea by the emergence of the Isthmus of Panama and climatic deterioration during the Miocene and Pliocene which destroyed much of the tropical fauna of the western Atlantic (Cairns 1977). The amphi-Atlantic component of this fauna is thought to reflect a current or recent communication across the Atlantic (Cairns 1977). Cairns (1982) has also studied the azooxanthellate Scleractinia of the Antarctic and Subantarctic regions where he found 37 species, which like those of the Atlantic are not thoroughly sampled. He found that 29% of these

species are cosmopolitan and 22% are endemic to the Antarctic. He noted that the distribution of Antarctic scleractinians is unlike the pattern common to most benthic invertebrate groups which are distributed throughout the Antarctic but endemic to it. In addition, the percentage of cosmopolitan species is considerably more than that observed in other benthic invertebrate groups. Zibrowius (1980) studied the Scleractinia of the Mediterranean and north-eastern Atlantic (the coasts of Europe and Africa to Iceland, the Azores

------...•. --- .._------~~~~_._. -- and the Cape Verde Islands), an entirely azooxanthellate, non-reef-building fauna. He found a total of 81 extant species in these regions. All of the 33 extant species found in the Mediterranean were found in the northeast Atlantic except for one species believed to be recently introduced from South America and four other species with uncertain affinities. Zibrowius suggests that these four species may be Mediterranean endemics but better knowledge of the fauna of the Spanish-Moroccan Gulf and the coast of West Africa is necessary before making such a designation. The species found included 47% with an amphi-Atlantic distribution and 12% with known distributions in the Indo-Pacific.

Distribution of reef-building corals In contrast to azooxanthellate corals, the distribution of reef-building corals has been relatively well studied and is most notably characterized by a great difference in fauna and diversity between the Atlantic and the Indo Pacific provinces (Wells 1956, Porter 1972, Rosen 1981, Veron 1985, Chevalier 1987). There are 90 genera in the Indo-Pacific, 26 genera in the Atlantic with only 8 genera (Veron 1993) and no species common to the two oceans. These two main coral provinces seem to be independent of each other in terms of diversity (Wells 1969, Stehli and Wells 1971). There is one diversity focus in the Indo-West Pacific, considered the most important, and another in the Caribbean, and each is slightly north of the equator (Stehli and Wells 1971, Rosen 1984). Diversity drops off as one moves away from these foci, with a more rapid decrease in diversity eastward (Stehli and Wells 1971). These foci are also characterized by high concentrations of islands which provide suitable habitat areas for corals (Stehli and Wells 1971, Rosen 1988c).

21 In both major realms, latitudinally, temperature seems to be the most important factor controlling coral distribution, setting the overall maximum limits of reef coral diversity (Yonge 1940, Wells 1969, Rosen 1971, Dana 1975, Rosen 1981, Rosen 1984, Veron 1985, Rosen 198&), though other historical and environmental factors combine to prevent this maximum from being realized at most localities (Rosen 1981,Rosen 1984). There is a steady increase in diversity as average surface sea-water temperatures rise from 16°C to a maximum of around 25-27°C (Rosen 1984). Other factors that are thought to influence diversity gradients include ocean surface circulation patterns (Veron 1985) and area effects (Rosen 1981). It is more difficult to explain the longitudinal distribution of corals, a topic of much discussion among coral biogeographers and which will be summarized here after the following description of the coral provinces. The IndQ-Pacific The Indo-Pacific province encompasses more than one-fourth of the globe, extending from the east coast of Africa to Easter Island, including the Red Sea, the Persian Gulf, the Indian Ocean, and the Pacific Ocean (Stehli and Wells 1971, Kay 1984, Chevalier 1987). The marine fauna of the Indo-Pacific up to the Miocene was characteristic of the widespread Tethyan fauna (Kay 1984). In the Miocene, Africa and India were each joined to Eurasia, disrupting the Tethyan Seaway, and Australia arrived at its present day position relative to Asia (Adams 1981, Coudray and Montaggioni 1982, Briggs 1987). From the Miocene onwards there has been little interchange between the Indo-Pacific and the Caribbean (Coudray and Montaggioni 1982, Rosen 1984, Vermeij 1987) and so these provinces are thought to have evolved separately during this time. Tectonic events during the Miocene in the 22

.. ------_.._- --- ._ •. _------.. _- western Pacific and the Indo-Malaysian archipelago (Raven and Axelrod 1972, Coudray and Montaggioni 1982) created three major remnant faunas within· the Indo-Pacific province (in the Indian Ocean, the western Pacific, and the Pacific Ocean basin) that are thought to have since evolved independently of each other (Kay 1984), though this is not the case for the Scleractinia. The Indo-Pacific coral fauna of today appears to be derived from the Tethyan fauna and is generally homogenous at the generic level across the entire province (Wells 1954, Rosen 1971, Veron 1985). The Indo-Pacific is the largest coral province (Stehli and Wells 1971), with a reef-building, zooxanthellate coral fauna of about 700 species (Chevalier 1987) in 90 genera (Veron 1993). In the Indo-Pacific province, from the Western Indian Ocean to the Western Pacific, generic diversity of the Scleractinia is relatively constant and then progressively decreases from the Central Pacific eastwards to the Eastern Pacific (Condray and Montaggioni 1982, Veron 1985, Chevalier 1987). The Indo-Pacific has been divided into three regions: the Indian Ocean, the Central Pacific, and the Eastern Pacific (Ekman 1953). The Indian Ocean and the Central Pacific are often referred to as the Indo-West Pacific (Ekman 1953) and this is the largest coral sub-province (Veron 1993). The reef complexes of the central Indo-Pacific south to the southern Great Barrier Reef and north through eastern Indonesia to the Ryukyu Islands encompass the highest number of genera of reef-building corals in the world and are the diversity focus of the Indo-Pacific Ocean (Stehli and Wells 1971, Coudray and Montaggioni 1982, Veron 1985, Chevalier 1987, Rosen 198&, Veron 1993). The western Indian Ocean has also been considered a diversity focus of less importance than the one in the Indo-West Pacific (Rosen 1971, Coudray and Montaggioni 1982, Chevalier 1987),but recent work at the species level by

23 Sheppard (1987) indicates that there is a high degree of homogeneity throughout the Indian Ocean. The reef-building coral fauna of the Eastern Pacific is distinct from the rest of the Indo-Pacific due to its depauperate nature (Wells 1956). It consists of nine genera which are also found in the Indo-West Pacific and only three of which are found in the western Atlantic (Veron 1993). A few of the species may be endemic to the region (Guzman and Cortes, 1993).Many of the reefs are dominated by Pocillopora spp. which are often the primary frame building corals in the Eastern Pacific (Glynn et al. 1972) where environmental conditions are not ideal for coral growth (Glynn et al. 1972,Porter 1972,Dana 1975). During the Cretaceous and early Tertiary, until the Eocene, this region was populated by a Tethyan fauna that was more similar to the fauna of the tropical Atlantic region than to the Indo-West Pacific region (Dana 1975, Glynn and Wellington 1983, Newman 1986). The EastemPacific served as an effective filter barrier limiting dispersal across it from the west during the Cretaceous and Tertiary (Ekman 1953,Grigg and Hey 1992). The fossil record of the Eastern Pacific after the Eocene is well summarized by Dana (1975) and Heck and McCoy (1978) based on Durham's (1%6) data. These analyses show that after the Eocene there was an impoverishment of the Eastern Pacific and Western Atlantic coral faunas. Evidence of invasion of the Eastern Pacific from the Indo-West Pacific before the Pliocene is weak. Dana (1975) suggests that the Eastern Pacific reef-building fauna was greatly reduced following the onset of Pleistocene glaciation due to cooler temperatures and/or fluctuations in sea level. Sea-floor spreading near the close of the Pliocene was accompanied by changes in ocean circulation patterns that may have been favorable for bringing in propagules~from the Indo-West Pacific (Dana 1975).

---- _.- ..•. _...._. __ .. _------._. _.- The origin of today's Eastern Pacific fauna is not clear (Porter 1972).Dana (1975) proposed that today's Eastern Pacific fauna is the result of colonization by Indo-Pacific species during the Pleistocene and this view has been subsequently supported by other workers (Glynn 1982, Glynn and Wellington 1983,Rosen and Smith 1988). Mc.<:oy and Heck (1976) and Heck and McCoy (1978) have proposed that the Eastern Pacific fauna today is a result of the continuous existence of some coral populations which survived the apparent large-scale Eocene extinctions. They interpret the fossil record slightly ! differently from Dana (1975) and suggest that colonization from the Indo- West Pacific could not account for the modern fauna due to the limited dispersability of coral planulae. Glynn and Wellington (1983) present strong evidence against Heck and McCoy's (1978) hypothesis, arguing that the Eastern Pacificcoral fauna of today has a recent and predominantly Indo Pacific connection. Subsequent work by [okiel (1984) and-Richmond (1987) also support Glynn and Wellington's argument the former shows that corals are capable of dispersing long distances by rafting; the latter demonstrates that Pocillopora damicomis planulae are able to settle and metamorphose after 103 days. However, Glynn and Wellington (1983) do suggest the possibility originally proposed by Porter (1972) that the fauna may consist of remnants of the more extensive Tethyan fauna but in addition has been augmented by dispersal from the west. More evidence is necessary to decide which of these hypotheses or a combination thereof is most likely.

Distribution of reef-building corals in the Atlantic The history of the Scleractinia of the Atlantic is primarily a history of the Caribbean region. The Atlantic coral province today includes the Caribbean Sea bordered by the southeastern coast of the United States, the 25 eastern coast of Mexico, Central America, and northern South America, and the relatively isolated regions of Bermuda, the Brazilian coast to the south of the Amazon and the Gulf of Guinea on the western coast of Africa (Wells 1956). This province contains 54 species of reef-building corals represented by 26 genera (Cairns 1979, Cairns 1981, Veron 1993). Of these 26 genera, 18 are found only in the Atlantic and 10 of the 18 are only found in the Caribbean (Veron 1993). This fauna is thought to be an impoverished Tethyan fauna (Wells 1956, Chevalier 1975, Chevalier 1987). The greatest diversity of reef building corals is found in the Caribbean, the diversity focus of the Atlantic province (Stehli and Wells 1971, Rosen 1984). As in the Pacific, there is an eastward attenuation of species across the province (Stehli and Wells 1971, Rosen 1988c). Through the Mesozoic there was a relatively homogenous coral fauna from the eastern Mediterranean westward across the Caribbean province (Frost 1977a). Although it has been thought that a characteristic Atlantic coral fauna did not develop until the Miocene when the Mediterranean formed (Gerth 1930,Wells 1956, Frost 1972, Veron 1985), it now seems that there was a clear differentiation of this fauna sometime in the Paleogene or even possibly the Cretaceous (Rosen and Smith 1988). This initial distinctiveness of the

Caribbean province is attributed to the spreading of the Atlantic Ocean ridge system which acted as a migration barrier (Coates 1973). There were few reef building corals in the Caribbean in the Paleocene and Lower Eocene and none formed wave-resistant structures (Frost 1972). The first true reef-building faunas following the extinctions of the Cretaceous are found by the mid Eocene mostly on muddy and sandy bottoms (Frost 1972, Rosen and Turnsek 1989, Budd et al. 1992). The generic composition of Caribbean reef-coral

..... _- ._._------~_._- -- assemblages seems to have been established during the Eocene, with the exception of the family Mussidae (Budd et al, 1992).By the late Eocene most of the older elements of the reef-building coral faunas began to disappear (Frost 1972) such that early Oligocene reef-building corals are rare in the Caribbean

(Budd et ale 1992). Maximum development of Caribbean Tertiary reefs occurred in the n 1;.gocene when these assemblages were dominated by cosmopolitan genera of Tethyan affinity (Wells 1956, Frost 1972, Chevalier 1975, Frost 1977a), The reef-building corals of this time are large, abundant, and diverse (consisting of as many as 70 species) and formed prominent barrier reefs across the Caribbean (Frost and Langenheim 1974,Budd et al. 1992). At the end of the Oligocene or in the earliest Miocene, there was a massive extinction of about 75% of the cosmopolitan Tethyan reef-coral genera found in the Caribbean (Frost 1972, Frost 1977c, Budd et aI. 1992). It is not clear why this extinction occurred. Frost (1977c) suggests it was caused by the disruption of the circumtropical current system due to a combination of local tectonism in Central America and movement of major plates in the western Tethys. Thus, the fauna of the early Neogene is a transitional one between the cosmopolitan Oligocene fauna and the largely endemic fauna of the late Miocene and Pliocene (Frost 1977b). The history of the Caribbean after the Oligocene is characterized by the disappearance of many cosmopolitan genera that are still found today in the Indo-Pacific, the appearance of new groups of corals, such as the agariciids and the mussids, and proliferation of species within surviving genera (Frost 1972, Frost 1977b, Budd et ale 1992). The progressive diminishment of the Caribbean coral fauna throughout the Neogene and Quaternary has been attributed in part to the geographic

27 isolation of the region (Chevalier 1975, Rosen 198&). The tropical Atlantic province did not become entirely separate from the Indo-Pacific until the rise of the Isthmus of Panama in the Pliocene (Veron 1985). After this event important elements of the Caribbean fauna became extinct and these extinctions appear to coincide as well with the onset of global refrigeration and eustatic sea-level change in the late Pliocene (Frost 1977b). By the mid Pliocene, most of the extant coral fauna of the Caribbean was established and it was not until this time that the abundant, fast-growing SPecies of Porites and Acropora are found (Frost 1972). There is suspicion that some colonization of the Caribbean from the Eastern Pacific did occur during the last interglacial high stand of sea level (Frost 1977c).

Theories of coral distribution The explanation for the distribution patterns of corals remains a subject of much discussion. Rosen (l988c) provides an in depth review of this discussion while summarizing the different models that have been proposed to explain coral distribution. The alternative explanations include the traditional dispersalist view of centers of diversity as centers of origin of new species (Stehli and Wells 1971), and the more" recent view, taken by vicariance biogeographers, that centers of diversity are centers of accumulation of species (McCoy and Heck 1976). The most recent synthesis of all of these theories is the vortex model proposed by [okiel and Martinelli (1992). All of these theories are based on generic level analyses of reef-building corals and are briefly summarized below. The center of origin theory (or some modification of it, see Rosen 1988c) is used to explain the distribution of corals based on the major diversity foci. If patterns of coral diversity are examined at the generic level, centers of

----- ._.. _.. _. diversity also appear to be evolutionary centers where new forms evolve (Stehli and Wells 1971). New species then spread to outlying, marginal areas by dispersal of reproductive propagules (Stehli and Wells 1971). The assumption here is that modern species distributions are determined by species age and dispersal ability (Potts 1985). This model predicts that areas of high diversity will·have many young taxa while areas of low diversity will have mostly older genera with fewer numbers of young taxa, the pattern that Stehli and Wells (1971) observed in calculating the average generic age of coral genera in different regions. This same pattern has also been observed in coral-inhabiting and shore barnacles (Newman 1986). Much of the debate as to whether this is a reasonable model to explain the current distribution of corals has hinged on the dispersal ability of corals. Coral larvae have been found to remain competent for over three months (Harrigan 1972, Richmond 1987). [okiel (1984) has also shown that adult colonies are capable of long distance dispersal by rafting, and has estimated that colonies of Pocillopora found on pumice in the Indo-Pacific may have traveled a total distance of 20,000 to 40,000 km. On the other hand, Wallace et al. (1991) found that in a group of five closely related Acropora species, taxa possessing a greater number of derived characters appear to be more restricted in their distributions. This is evidence against the center-of-origin hypothesis. The vicariance theory of coral distribution was developed with consideration of plate tectonic theories and long-term climatic change. It's basic framework was laid out by McCoy and Heck (1976) in explaining the distribution patterns of corals, mangroves, seagrasses, and all inhabitants of the shallow-water habitats of the tropics. Again, there are considerable variations in this theory and these are described. in detail by Rosen (1988c).

_...._. __ ._------_.-.-- Vicariance theory holds that the worldwide distribution and diversity patterns of all three groups can best be explained ifthey are considered as remnants of a once widely distributed ancestral biota. The distribution of this ancestral biota has been modified by tectonic events and climatic changes resulting in speciation and extinction throughout the area of this original distribution (Mc:.<:oy and Heck 1976). It is argued that the reason more older groups of taxa are found in the diversity foci of today is because these are areas where many islands provide greater shoreline area for colonization and increased chances for isolation and diversification, and so provide optimum habitat where taxa have tended to accumulate. In this case, outward radiation is the result of passive dispersal over time (Mc:.<:oy and Heck 1976). All three groups have a pattern of co-occurrence dating back to the Cretaceous and this is used as evidence against dispersal from centers of diversity. In addition, Mc!:oy and Heck (1976) believe that there is no good evidence for long distance dispersal of the three groups as would be required by dispersalist theories. [okiel and Martinelli's (1992) vortex model brings together elements of both dispersalist and vicariance theories. This model explains patterns of diversity and the distribution of corals based on ocean circulation patterns that generate both an outward movement of species from a center of origin as well as a return genetic flow to the diversity focus from outlying areas. It uses the idea of the importance of long distance dispersal in moving species away from diversity centers that is critical to the center of origin model as well as the idea that allopatric speciation occurs in isolated peripheral areas that is central to vicariance models. Through computer simulations, using only an idealized ocean circulation scheme of the Pacific, [okiel and Martinelli (1992)

--_._~~~~~ .._.. _... were able to show that the total number of species would increase over time and that the distribution of species (in the Pacific) would become one where the highest diversity would be in the western equatorial region, which is the observed pattern. It was previously suggested by Veron (1985), based on his analyses of generic distributions of corals, that coral distribution patterns are most simply explained by ocean circulation patterns, as well as temperature gradients. As has been suggested by Newman (1986), it is perhaps most useful to consider dispersal and vicariance simply as facets of biogeography instead of mutually exclusive theories.

Evolution and taxonomy of scleraetinian groups

There have been many efforts to classify scleractinian corals since their recognition as animals in the late 1700s. Vaughan and Wells (1943) and Alloiteau (1952) give a complete, detailed review of the study of scleractinians up to the mid-20th century. The classification of corals has come to be based on all aspects of skeletal morphology macrostructure including the growth form of an entire colonial or solitary coral, characteristics of the corallite (the calcium carbonate cup that is the skeleton of the living polyp), features of the skeleton between corallites, septal arrangement and septal ornamentation (the septa are the internal walls that support and separate the mesenteries of the coral polyp). Since the late 1800s, much work has been done on understanding the histological structure and microstructure of the corallite. Skeletal microstructure had been developed as a tool in scleractinian taxonomy by Pratz (1882) and Ogilvie (18%) in their studies of the relations among fossil and Recent corals. These techniques have since been developed primarily by paleontologists and are especially useful in the study of fossil

31 corals. In contrast to macrostructural studies, microstructural studies are carried out on thin, transverse sections of the skeleton, with the aid of a microscope. Microstructural studies place emphasis on the elements that actually construct the corallite, especially the elements that the septa itself is composed of. The most important of these elements are the trabeculae which are the deposits around the centers of calcification and which are themselves made up of other elements. Microstructural studies take into account all of these elements including their composition, construction, and orientation as well as that of septal ornaments. Since the mid-1900s, two prominent schools of scleractinian systematics have developed. The analyses of both of these schools are based on morphological characteristics of scleractinian skeletons and neither makes use of quantitative methods of analysis. One of the schools is based on the work of Vaughan and Wells who published a revision of the order Scleractinia in 1943 that was subsequently slightly modified by Wells (1956). Vaughan and Wells based their classification scheme "primarily on the structure of the septa, but other skeletal structures and the somewhat meager knowledge of the soft parts" were taken into consideration, along with paleontological history. Their emphasis was on skeletal macrostructure although they also used microstructure to a certain degree. In the Treatise, Wells included an evolutionary hypothesis (Fig. 1.1) for the relationships among the families since the Triassic. Veron (1995) has subsequently carried out a taxonomic revision of the order based largely on Vaughan and Wells' ideas. Veron's unique contribution to the systematics of the Scleractinia has been to take into account knowledge of specimens in situ including ecotypic

_ _- .. -_ .._------variation, a characteristic that distinguishes this school from the studies of the second school of scleractinian systematics. This second school that has developed is exemplified by the 1952 work of Alloiteau in the Traite de Paleontologie. Alloiteau, a paleontologist, placed heavy emphasis on microstructural characteristics while also relying on macrostructural characteristics. Since 1952, much taxonomic work using microstructure analysis has been carried out, primarily on fossil corals, including Chevalier (1961), Chevalier (1971), Chevalier (1975), Cuif (1977), Russo (1979), Beauvais (1980), and Roniewicz and Morycowa (l993). The most recent revisions from this school are that of Roniewicz and Morycowa (1993) based entirely on microstructural data, and that of Chevalier and Beauvais (1987) who have based their classification on the work of Vaughan and Wells (1943), Alloiteau (1952, 1957),Wells (1956), and Cuif (1977) using both macroscopic and microscopic skeletal structure.

Scleractinian evolution as described by Wells (1956)

The classification of Wells (Wells 1956, Fig. 1.1) divides the entire order into 5 suborders and 33 families. Twenty of these families are extant. The families present in the oldest coral fossils of the mid-Triassic represent two major lines that subsequently gave rise to the five suborders. These two major lines, whose relationship is uncertain, are distinguished on the basis of organization and orientation of centers of calcification in the septa. One of these lines developed into the suborder Astrocoeniina and is represented in the mid-Triassic by the families Thamnasteriidae and Astrocoeniidae (both exclusively colonial and reef-building). The Astrocoeniidae developed into the main branch of the suborder which gives rise in the late Triassic to the colonial, reef-building Stylinidae. The Pocilloporidae developed from the

------main astrocoeniid line in the Mid- Cretaceous as did the Acroporidae in the late Cretaceous. The family Pocilloporidae is composed of colonial, reef building corals but also includes ahermatypic genera. The hermatypic genera are common on upper reef slopes exposed to strong wave action but are also found in deep water and in lagoons (Veron 1986). The Acroporidae, are colonial, hermatypic corals characterized by a very light, porous and rapidly growing skeleton with small polyps sometimes forming huge colonies. These characteristics are thought to have led to this group being the most successful today of all of the Scleractinia, being comprised of over 150 species (Veron 1993). Colonies of this family display all of the growth forms known for reef building corals (Veron 1986). The Thamnasteriidae have also persisted until the present but without becoming a prominent group. The Stylinidae did not give rise to any other groups and went extinct by the end of the Mesozoic. The entire suborder, Astrocoeniina, is thought to have evolved independently over the 250 million year history of scleractinians. The other major line present among the first fossil corals in the Mid Triassic was the Procyclolitidae (Fig. 1.1), a family of reef-building corals that included both solitary and colonial genera. This family eventually gave rise to the other four suborders of extant Scleractinia. In the early Jurassic the procyclolitids diverged along two lines, one of which developed into the suborder Fungiina and the other of which developed into the other 3 suborders. Along the line leading to the Fungiina, the family Calamophylliidae developed and included hermatypic corals that were both solitary and colonial. In the Jurassic, this group gave rise to two reef-building families, the Microsolenidae, which are both solitary and colonial, and the Actinacididae, which are colonial. The Microsolenidae went extinct in the late

34 Cretaceous. Although the Actinacididae were extinct by the end of the Eocene, in the late Cretaceous they had given rise to the highly successful Poritidae, colonial, reef-building corals which are still a prominent group today. In the Mid- Cretaceous the Agariciidae, a primarily reef-building, colonial group with some solitary representatives, developed from the calamophylliid line. This family is found on protected reef slopes and is common in lagoons (Veron 1986). In the late Jurassic the calamophylliid line gave rise to the extant family Siderastreidae, a group of reef-building, colonial corals. The Calamophylliidae itself was extinct by the end of the Eocene. In the mid- Jurassic, the second line of procyclolitids that eventually gave rise to the Faviina, Caryophylliina and Dendrophylliina, which developed into the Synastreidae and the Haplaraeidae (Fig. 1.1), reef-building families with solitary and colonial representatives. Although both of these families went extinct by the end of the Cretaceous, the Synastreidae gave rise to three other families, two of which are extant today. In the early Cretaceous, the Synastreidae gave rise to the Micrabaciidae, a non-reef-building, probably azooxanthellate group of free-living, solitary, cup-like corals found on sandy or silty bottoms at depths below wave base. In the late Jurassic, the Synastreidae gave rise to the Cyclolitidae, a reef-building group of both solitary and colonial corals adapted for lying loose on soft or shifting substrates and the predecessors of the Fungiidae which developed in the mid Cretaceous. The mostly reef-building fungiids include solitary and colonial forms, either fixed or free-living. The ancestral genus is a solitary reef building coral (Wells 1966). The Cyclolitidae themselves went extinct by the end of the Eocene.

35·

------_..._. _.- The suborders Faviina, Caryophylliina and Dendrophylliina are derived from the second line of procyclolitids which gave rise to the Montlivaltiidae and the Stylophyllidae in the late Triassic (Fig. 1.1). It is not

clear how these two families are related to each other and whether the montlivaltiids developed from the stylophylliids. The Stylophylliidae first appear in the Mid- Triassic as reef-building corals with both solitary and colonial representatives. This.group went extinct in the late Jurassic. The stylophylliids are thought to possibly have given rise to another group of solitary and reef-building corals, the Amphiastreidae, in the early Jurassic. This family went extinct in the late Cretaceous. The Montlivaltiidae line gave rise to the extant families in the suborders Faviina, Caryophylliina and Dendrophylliina. The Montlivaltiidae was a reef-building family of both solitary and colonial members that was first seen in the late Triassic and became highly developed by the Mid- Jurassic. In the lateJurassic the Faviidae, a mostly reef-building group with solitary and colonial members, developed from the montlivaltiids and have increased steadily ever since. The relationships within this family are not well understood though it contains the largest number of genera of any scleractinian family and is second only to the Acroporidae in number of species (Veron 1993). They occupy a great diversity of biotopes and generally have a wide geographic distribution (Veron et aI. 1977). Their success is thought to be partially due to exploitation of various modes of colony formation, with a tendency towards massive growth forms (Veron et al. 1977). This family gave rise to the Merulinidae, a reef-building, colonial group, in the late Tertiary. In the early Tertiary the main montlivaltiid line gave rise to the Mussidae. This reef building group, with both colonial and solitary genera, has heavily

_...._._------_. _._- constructed, solid skeletons and is widespread (Veron 1986). In the Mid Tertiary the Mussidae gave rise to the Pectiniidae, a reef-building group with solitary and colonial forms. In the early Cretaceous the main montllvaltiid line is thought to have given rise to the Rhizangiidae, a colonial, non-reef building, azooxanthellate family, and to the Oculinidae, a family of colonial corals that are both reef-building and non-reef-building (Veron et al. 1977). The Rhizangiidae are today widely distributed in the littoral and neritic zones. The azooxanthellate, non-reef-building family Anthemiphylliidae is a group of solitary, free-living corals, which has no paleontological history, and is thought to have developed from the rhizangiids. Late in the Jurassic, the Meandrinidae, a family composed of solitary and colonial corals that are reef building and non-reef-building, developed from the montllvaltiids. The main montlivaltiid line went extinct in the early Tertiary. The montlivaltiids gave rise to the first genus of the Caryophylliidae in the early Jurassic (Fig. 1.1). They first appeared in a neritic, non-reef environment. This is the most successful of all scleractinian families in adaptation to extremes of environment and in generic differentiation and is found world-wide (Veron and Pichon 1980). The group is primarily non-reef building and azooxanthellate (Veron and Pichon 1980). This lineage marks the beginning of the distinction between reef-building and non-reef-building corals (Vaughan and Wells 1943). Today there are approximately 50 genera that are found from the littoral zone to the bathyal zone (Cairns 1977, Cairns 1979, Cairns 1981,Cairns 1982, Cairns 1989b,Cairns 1991,Cairns and Parker 1992). This group includes solitary corals adapted to substrata affording few or very small sites for larval fixation as well as colonial, reef-building corals. In the Mid- Jurassic the caryophylliids gave rise to the Rhipidogyridae, reef-

37 building corals with solitary and colonial forms that attach to the substrate, which went extinct by the Early Cretaceous. In the early Cretaceous the Guyniidae and Flabellidae, non-reef-building, azooxanthellate families that may be free-living and attached forms, developed from the caryophylliid line. The Guyniidae went extinct in the early Tertiary. In the Mid-Cretaceous the main caryophylliid line gave rise to the Dendrophylliidae, the only family in the fifth scleractinian sub-order, the Dendrophylliina. These are mostly non reef-building corals which may be solitary or colonial.

Scleractinian evolution as described byveron (1995) In Veron's classification (1995, Fig. 1.2), the order is divided into 12 suborders with 59 families. The extant corals are found in seven suborders and 24 families. The extant suborders are the same as in Wells' (1956) classification, with some additions, although Veron does not find the relations among them so well-defined. In addition, Veron's view of the evolution of the different groups is quite different from Wells'. This is probably a reflection of the great deal of knowledge of fossil corals that has been gained since Wells did his work. . At the suborder level, the biggest difference in Veron's hypothesis is that the family Poritidae is not grouped with the Fungiina (Fig. 1.2) but is placed in its own suborder that arose in the late Jurassic, and subsequently split into two lineages in the Early Cretaceous. One of these, the Actinacidiidae, continued on through the Cretaceous until diminishing in the Paleocene and going extinct at the end of the Oligocene. The other lineage which became the family Poritidae has especially flourished since the beginning of the Eocene. Veron also moves the families Oculinidae and Meandrinidae (both in the Faviina according to Wells) to their own suborder,

... __.._------._-... the Meandriina, that originated in the Mld-Cretaceous. Included in this suborder is the Rhipidogyridae which was extinct by the late Cretaceous. Veron hypothesizes that the fossil corals of the Mid-Triassic are representative of eight suborders, three of which have extant representatives (Fig. 1.2). One of these extant suborders, Archaeocaeniina, gave rise to two lineages in the late Triassic. One lineage became the family Pamiroceriidae which went extinct in the Mid-Jurassic. The other lineage gave rise to two lines in the Mid-Jurassic. One of these, the family Tropiphylliidae, diversified and flourished in the late Jurassic and beginning Cretaceous before going extinct at the end of the Cretaceous. The second line became the main archaeocaeniid line that gave rise to the Acroporidae in the Mid-Jurassic which especially flourished and diversified from the Eocene on. The main archaeocaeniid line continued on, slowly diversifying until the Eocene when it split into the two extant families Astrocoeniidae and Pocilloporidae. Like Wells, Veron hypothesizes that this suborder evolved separately from the others during the entire evolutionary history of the Scleractinia. In Veron's hypothesis, the taxa of the Thamnasteriid line are not placed in the Fungiina (Fig. 1.2), another suborder with representatives in the Mid- Triassic. Instead, the Thamnasteriid line is considered one of the groups present in the Mid-Triassic which diversified throughout the Jurassic and Cretaceous at which time it went extinct as a family and gave rise to the extant Siderastreidae which has continued to develop since the Eocene. Veron places five other groups in this suborder with beginnings in the Early Jurassic. Two of these families, Astraeomorphidae and Cuifastraeidae, went extinct by the Mid-Jurassic while the other three, the Andemantastraeidae, the Microsolenidae, and the Procyclolitidae, continued to develop before going

39 extinct at the end of the Cretaceous. Four other lineages in this suborder arose in the Mid-Jurassic. The Dermosmiliidae diversified throughout the Jurassic, then diminished through the Cretaceous when it gave rise to the family Agariciidae before going extinct. The Agariciidae began to flourish in the Eocene and continue on today. The second lineage to appear in the Mid Jurassic, the Synastraeidae, did not start to flourish until the Cretaceous, then diversified during the Paleocene at the end of which time it went extinct. The third lineage of the Mid- Jurassic developed slowly until the Cretaceous when it started to diversify. In the late Cretaceous it gave rise to three lineages: 1) the extant, deep-sea family, Micrabaciidae; 2) a lineage that split at the beginning of the Eocene into the Funginellidae (a group of solitary corals) which went extinct at the end of the Oligocene and the Fungiacyathidae, another extant, wide-ranging coral family; and 3) the family of solitary corals, Fungiidae. The fourth lineage of the Mid-Jurassic, the Latomeandriidae, flourished throughout the Cretaceous, giving rise to a small family, the Cunnolitiidae, in the late Cretaceous. Both of these last two families went extinct at the end of the Cretaceous. Representatives of the suborder Faviina are not present in the fossil record until the Mid-Jurassic (Fig. 1.2). At this time two lineages are found in the fossil record that give rise to six of the nine extant families of this suborder. One lineage, which eventually led to the family Mussidae, remained small until the end of the Paleocene when it started to floursih. In the mid-Eocene it gave rise to the small extant family Merulinidae and in the mid-Oligocene it gave rise to the Pectiniidae. The other main lineage present in the Mid-Jurassic is the Faviidae which diversified until the end of the Cretaceous when it diminished somewhat before continuing to flourish

-"'-"- ... '--'- ._._------. --,- throughout the Cenozoic. Today the Faviidae has the most number of genera of any family and is only second in species to the Acroporidae (Veron 1993). In the early Eocene the Faviidae gave rise to the Trachyphylliidae. A very early offshoot of the Faviidae is the Montlivaltiidae, an important family throughout the Cretaceous that diminished considerably in the Paleocene, continuing with few representatives until going extinct sometime in the Oligocene. Another lineage in the suborder Faviina is first found in the fossil record in the beginning of the Cretaceous. This lineage has continued to develop and is today represented by the family Rhizangiidae and the deep water, azooxanthellate family Anthemiphylliidae. The groups in Veron's suborder Caryophylliina (Fig. 1.2) are in agreement with the Wells hypothesis (Fig. 1.1), the main difference being the elevation to family status of groups that Wells considered sub-families of the Caryophylliidae. All of these families are considered to be mostly azooxanthellate, non-reef-building corals and are found throughout the world's oceans. This is the third of Veron's suborders with representatives in the Mid-Triassic. One small group, the Remimaniphylliidae, went extinct by the end of the Triassic. Another group, the Thecocyathidae, continued on without great success until going extinct at the end of the Cretaceous. This family possibly gave rise to the Caryophylliidae at the end of the Jurassic. The Caryophylliidae have flourished throughout the Cretaceous and Cenozoic, are well represented today, and are the largest group in this suborder. Another small family, the Margarophylliidae, was present in the Mid Triassic and went extinct by the Mid-Cretaceous. Three families, the Dasmiidae, the Flabellidae, and the Guyniidae are first found in the fossil record in the beginning of the Cretaceous. The Dasmiliidae went extinct at the

------_. ----- end of the Eocene while the other two continue to flourish today. The Turbinoliidae was the last of the Caryophylliina to appear in the late Cretaceous. It has slowly grown and diversified and is still extant. The Actinaddiidae are the first representatives of the Poritiina (Fig. 1.2). This family originated in the late Jurassic and gave rise to the Poritidae in the Cretaceous. The Actinacidiidae went extinct at the end of the Oligocene but the Poritidae continue to flourish today. The Dendrophylliina are first represented by the Actraraeidae, are found in the late Jurassic and went extinct at the end of the Cretaceous (Fig. 1.2).The Dendrophylliidae first appear in the late Cretaceous, possibly as an offshoot of the Actraraeidae. This family greatly diminished in the Paleocene but has since flourished and diversifled and is extant today. Veron's suborder Meandriina was the last of the suborders to evolve and it originated in the Mid-Cretaceous, with the main lineage being the Meandrinidae (Fig. 1.2). This family is extant though it is restricted to the Atlantic region. Two other families apparently related to the Meandrinidae appear later in the Cretaceous. The Rhipidogyridae went extinct at the end of the Cretaceous while the Oculinidae are extant.

Scleractinian classification according to Alloiteau (1952) The classification established by Alloiteau (1952, Table 1.2) is distinguished by its use of microstructural data. Unlike Wells and Veron, Alloiteau did not hypothesize about the evolutionary relationships among the taxa although for most groups he did clearly layout the characters that he used to define them. He classified the Scleractinia into eight sub-orders and 65 families. Six of these suborders and 31 of the families are extant. Alloiteau's suborder Archaeocoeniida is about the same as the suborder Astrocoeniina in 42

------terms of extant taxa although like in Veron's classification, it does not include the family Thamnasteriidae. Alloiteau's families Seriatoporidae and Stylophoridae contain the genera in Wells' family Pocilloporidae. Alloiteau has placed the extant genera in Wells' family Astrocoeniidae in the suborder Astraeoida. The suborder Astraeoida is approximately equivalent to Wells' suborder Faviina, the differences being that Alloiteau has elevated two groups from sub-families in Wells' Faviidae to family level. Alloiteau's family Astrangiidae is equivalent to Wells' Rhizangiidae in this suborder. In addition Alloiteau includes taxa from Wells' orders Astrocoeniina and Fungiina in families of their own in his suborder Astraeoida. Alloiteau has erected a new suborder Meandriida which includes genera from Wells' families Meandrinidae and Oculinidae (both suborder Faviina) as well as from the Caryophylliidae in Wells' suborder Caryophylliina. Alloiteau's suborder Caryophylliida is equivalent to Veron's suborder Caryophylliina. His suborder Fungiida is very similar to Wells' suborder Fungiina. The differences include a group from Wells' Fungiidae being elevated to family status, Fungine11idae, and the removal of a genus from Wells' family Faviidae to its own family, Agathiphylliidae. Finally, Alloiteau's suborder Eupsammiida is approximately equivalent to Wells' Dendrophylliina. Alloiteau puts the genera of Wells' family Dendrophylliidae into two families, Eupsammidae and Turbinaridae.

Sc1eractinian classification according to Chevalier and Beauvais (1987) Chevalier and Beauvais (1987) divided the order Scleractinia into 11 sub-orders and 55 families (Table 1.3).Six of these suborders and 29 of the families are extant. Unlike Wells (1956) and Veron (1995), Chevalier and Beauvais do not make any speculations as to the interrelationships and

------_._- evolution of families through time. This classification is very similar to that of Alloiteau (1952). The differences in their classification scheme from those of the Wells school include another suborder, a change of family from one suborder to another, and the elevation of subfamilies to family level. They use the suborder Archaeocaeniina instead of Astrocoeniina as did Alloiteau although they do include the family Astrocoeniidae as Wells and Veron. Chevalier and Beauvais's family Seriatoporidae in this suborder is the same as Pocilloporidae. The suborder Fungiina is similar to Alloiteau's Fungiida, they include Agariciidae, Siderastreidae, Poritidae, and Fungiidae. Unlike Alloiteau, they place the genera of the Funginellidae in the Fungiidae, however like Veron, they elevate the genus Fungiacyathus in the Fungiidae to a family of its own. Unlike anyone else, they establish the family Psammocoridae with taxa that Alloiteau places in Agariciidae, Wells in Thamnasteriidae, and Veron in Siderastreidae. They do not include the family Micrabaciidae in the suborder Fungiina as Alloiteau, Wells and Veron have done. Instead they place this family in the suborder Dendrophylliina. Chevalier and Beauvais's suborder Faviina is similar to Alloiteau's Astreoida. Their family Diploastreidae is approximately equivalent to Alloiteau's Agathiphylliidae and includes taxa that Wells and Veron place in the Faviidae. Like Veron, they erect the family Trachyphylliidae from taxa in the Faviidae of both Alloiteau and Wells. Their family Astrangiidae i~ equivalent to that of Alloiteau. Chevalier and Beauvais's suborder Meandriina is approximately the same as Alloiteau's Meandriida. The suborder Caryophylliina is similar to that of Alloiteau although Chevalier and Beauvais include the family Eusmilidae whose genera Alloiteau place in another family in the suborder Meandriida. Chevalier and Beauvais also

------._. -- elevate to family status a grouping that Alloiteau places in the Caryophylliidae. Finally, Chevalier and Beauvais's suborder Dendrophylliina includes the taxa in the families that Alloiteau places in the Eupsammida and adds the family Micrabaciidae which is usually placed in the Fungiina. Scleractinia» eyo1ution based entirely on migostmclural analysis At this point, it is useful to summarize the work of Roniewicz and Morycowa (1993, Fig. 1.3), although they have not carried out a complete revision of the order. They have proposed an evolutionary hypothesis for the order Scleractinia based entirely on an analysis of microstructural studies of Mesozoic and Cenozoic corals from the last 30 years. Their focus is on the relations between Triassic and Jurassic corals. The phylogeny of the Scleractinia based on these data is very different from any of the ones that have been discussed so far. Like Wells (1956) and Veron (1995), Roniewicz and Morycowa suggest polyphyletic origins of the suborder. However, their hypothesis for relations among subgroups is quite different from that of Wells. According to this scheme, the corals of the Triassic fall into four microstructural groupings that can be divided into at least seven subgroups. The extant coral fauna is derived from two of the four main microstructural groupings. Corals in the suborder Astrocoeniina from other hypotheses also group together here but they are part of the same microstructural group as some of the taxa from the suborder Fungiina. The second microstructural group includes taxa from the suborders Faviina, Fungiina, Caryophylliina and Dendrophylliina. The Caryophylliidae in the suborder Caryophylliina is the main lineage in this group but the Oculinidae (suborder Faviina) and the Dendrophylliidae (suborder Dendrophylliina) are found to be more closely related to the Caryophylliidae than the Flabellidae, a family usually also

- _ _---- ....-.. - _._-- ._------_...-. placed in the Caryophylliina. The families Mussidae, Pectiniidae, Faviidae, Merulinidae in the suborder Faviina are found to be most closely related to corals from the family Fungiidae in the suborder Fungiina. This entire group is thought to have split off from the main Caryophylliid line in the late Jurassic. The Agariciidae, from the suborder Fungiina is found to be most closely related to the Meandrinidae which has been placed in the Faviina or a suborder of its own. Summary of classifications There is general agreement among the four major treatments of extant scleractinian classification. The differences in interpretation of the relationships of extinct taxa seem to be a result of the characters used and the understanding of the evolution of these characters by each of the authors. None of these treatments are quantitative. Only considering extant taxa, the five to seven suborders in each of the classification schemes comprise roughly the same taxa. The classifications of Wells (l956) and Veron (1995) are equivalent at the suborder level except that Veron elevates the Poritidae and the Meandrinidae to suborder status. The six extant suborders of Alloiteau

(1952) and of Chevalier and Beauvais (1987) are equivalent to each other. They differ from the classifications of Wells and Veron in that they put certain families in a suborder of their own, the Meandriina. Like Wells, they consider the Poritidae related to families in the suborder Fungiina. There are differences among the classifications at the family level and below. At the family level, Alloiteau, and Chevalier and Beauvais tend to create more families and sub-families than Wells and Veron. There is general disagreement as to the taxonomic levels of groups in the Caryophylliina, which are alternatively classified. as sub-families or families. In addition,

46 there are differences in the interpretation of the relationships among families within the suborders Fungiina, Faviina and Caryophylllina.

Non-mmphologiCal methods.of scleradinian classification Although scleractinian systematics have been based on morphological characters, the extreme variability in the skeletal morphology of corals (Wijsman-Best 1974, Veron and Pichon 1976, Brakel 1977, Zibrowius 1984, Foster 1985, Willis 1985, Veron and Kelley 1988) makes these characters unreliable and has led to considerable taxonomic confusion (Randall 1976, Potts 1984). Vaughan and Wells (1943) stated that the classification of the order "into subordinal groups and categories of lesser rank is in a very unsatisfactory condition" and this is still apparent today in the classifications of Chevalier and Beauvais (1987) and Veron (1995). The delineation of species is a well-known problem in scleractinian systematics (Wallace 1974, Wijsman-Best 1974, Randall 1976, Brakel 1977, Best et al. i984, Zibrowius 1984, Chevalier and Beauvais 1987, Budd 1988, Willis 1990, Gattuso et al. 1991, Hidaka 1992). In the past 30 years various attempts have been made at finding other than morphological methods for classifying corals to try to get around the problem of variability of scleractinian skeletal morphology. These methods include reproduction, polyp morphology, behavior, ecology, and physiology and are reviewed in detail by Lang (1984). Lang has also provided an assessment of these methods and others as potential tools of scleractinian taxonomy and advocates the use of a multi-eharacter approach. None of these techniques have included molecular methods. None of these techniques have been applied to the order as a whole.

_._- --_._._- Quantitative analyses of relationships among the Scleractinia In general, the methods used to classify the Scleractinia have not included quantitative analyses. Powers (1970, Powers and Rohlf 1972) made the only attempt at a quantitative analysis of the entire order. He used 60 skeletal, polyp, and environmental characters in a phenetic analysis of 54 species of Hawaiian and Caribbean corals representative of 15 families and all five suborders. In the phenograms obtained with these data, Powers found general agreement with the classification of Vaughan and Wells (1943). Species within a genus clustered tightly together. Genera in families generally clustered together although there was some separation of groups in the Faviidae, which has a large number of genera, as well as a genus from another family clustering with this group. Taxa from the same suborders clustered together although in parts of the analysis they were not clearly distinct from one another. Powers (1972) concluded that with certain adjustments these techniques would "be a useful adjunct in scleractinian systematics" but they have not been used to attempt to further understand the evolutionary history of the order. Quantitative analyses have been more often used to study the relationships among closely related species and genera. Wallace (1974) used measurements of skeletal morphology in ranking and cluster analyses of four well characterized species of Acropora in order to better define attributes for the study of morphological variation with location. She concluded that with some modifications this method might be useful for sorting specimens into groups for further analyses but no other work has been published based on these methods. Budd (Foster 1984,Foster 1985, Budd 1988, Budd 1990, Budd and Coates 1992) has successfully used multivariate techniques to quantify 48 variation and delineate species among fossil and extant specimens within the genera Montastrea, Porites, and Siderastrea. Cairns (1984) has carried out a phylogenetic analysis of genera and sub-genera in the family Fungiidae, and Hoeksema (1989) has carried out a similar analysis of the Fungiidae at the species level, coupled with a detailed systematic study of morphological characters. The resulting phylogenetic hypotheses from these analyses differ from each other as well as from Wells' (1966) qualitative analysis of the family. Hoeksema (1989) attributed the differences in the two phylogenetic analyses to his taxonomic revision as well as to differences in characters used and interpretations of the evolution of the characters. Both authors concluded that the cladograms resulting from these analyses were most useful as working hypotheses for further study of the evolutionary relationships within the group.

Biochemical studies of sc1eractinian evolution One general category of techniques that Lang (1984) reviewed is that of biochemistry. The development of biochemical methods as tools of the taxonomist has revolutionized systematic work over the past 20 years (Wilson et al, 1985, Avise 1986, Birley and Croft 1986,Meyer and Wilson 1990). In the last 15 years workers have been developing some of these molecular techniques as tools in the study of the evolutionary history of corals. Such methods are promising as they are the first to be used that are completely independent of coral skeletal morphological variability. Using protein electrophoresis techniques, Lamberts (1979 ) found that different coral species produced different banding patterns in preliminary analyses. Ohlhorst (1984) then used this method to study the relationships of 12 species of Jamaican corals. She was able to distinguish most of these species with five

49 polymorphic enzymes (EST, ACPH, APH, and LAP) although the banding patterns did not Correlate with variability in coral colony shape or color but were often correlated with different collection sites, Ayre et al. (1991) used electrophoresis successfully to differentiate between 2 species of Acropora that are difficult to differentiate morphologically. Knowlton et al, (1992) and Van Veghel and Bak (1993) have used electrophoresis, in combination with other methods, to elucidate the relationships between three different morphotypes of Montastrea annularis. Potts et al. (1993) and Garthwaite et al. (1994) have successfully used electrophoresis to differentiate between morphologically defined species in the family Poritidae. In general, closely related species can only be distinguished by shifts in the frequencies of alleles at polymorphic loci. Workers have now begun to develop techniques using elements of genome structure and DNA sequences to study the evolutionary history of corals. Most of this work has been directed at finding ways to use nuclear sequences to study the relationship among closely related species in the family Acroporidae. Initially, McMillan et al. (1988) used DNA-DNA hybridization assays to determine the relatedness of different species in the family Acroporidae. They concluded that this technique would only be useful for indicating major differences among closely related species but would not be useful for accurately quantifying true homology or assessing relatively distant taxonomic relationships, such as comparisons of coral families or distantly related species within families. They used a repeat family as a probe in slot-blot experiments as well as partial restriction maps to examine closely related species of Acroporidae (McMillan et al. 1988, McMillan and Miller 1989, McMillan and Miller 1990). They found that these techniques

------differentiated closely related species of Acropora, and the phylogenetic groupings that these data supported were different from those suggested by traditional morphological techniques (McMillan et ale 1991). Miller et al, (1990) determined the nucleotide sequence of a histone encoding gene and subsequently used these data to study the relationship of cnidarians to other metazoans (Miller et al, 1993). Nuclear ribosomal DNA sequences have also been used to elucidate the evolutionary relationships among classes of Cnidarians and among metazoan phyla (Christen et al. 1991, Hori and Satow 1991), and included in these studies are samples from Scleractinia. Most recently, Bridge et al. (1992) used mitochondrial genome structural similarities in the Cnidaria to reconstruct class-level relationships. Although molecular techniques have been developed to look at relationships among very closely related corals and how corals are related to other groups in the phylum as well as other metazoans, none of these techniques have been developed to study the relationships of groups within the order. Work is now in progress (Romano, in prep) using mitochondrial large ribosomal subunit sequences (16S rRNA) to study the evolutionary relationship of families within the order Scleractinia.

Summary

Corals today are found throughout the tropics and have been the subject of many ecological and physiological studies from which we have gained a great deal of knowledge about all aspects of their biology. The Scleractinia have a long and rich fossil history that has enabled us to gain some understanding about the biogeography and phylogeny of the group. Due to the difficulty in finding satisfactory characters on which to base

51 scleractinian classification, all of this information is lacking a firm. evolutionary foundation on which to base a complete synthesis of our understanding of the group. The fields of molecular biology and quantitative phylogenetic analysis offer the possibility to gain a completely new perspective on the evolution of the Scleractinia. This new perspective coupled with the great deal of knowledge of corals that already exists should certainly provide the opportunity for a more complete understanding of the evolutionary history of corals.

------_. _._._------Table 1.1 Timeline of important events in the evolutionary history of corals

53 C Pleistocene 1-2 mya Worldwide glaciation T Beginning of the development of the Great Barrier Reef Closure of theIsthmus of Panama E Pliocene Smya Present day limits of reef-building corals established E Appearance of Trachyphylliidae Beginning of restricted communication across the Isthmus N of Panama R Messinian salinity crisis results in extinction of 0 Scleractinia in the Mediterranean Final closure of the Tethys T First appearance of algal ridges Z Oear distinction of Atlantic, Indian, and Pacific Ocean Miocene 24mya coral faunas I Beginning of closures and reopenings of the Tethys Collision of the Australia with the Pacific plate 0 First appearance of Pectiniidae A Mass extinctions of Caribbean Scleractinia I Oligocene 37 mya Maximum development of Caribbean Tertiary reefs R First appearance of Mussidae Eocene 58mya Unfavorable worldwide conditions for corals C Unfavorable worldwide conditions for corals y No fossil reefs Paleocene 65mya Mass extinction of Scleractinia Beginning of faunal differentiation between Indo-Pacific and Atlantic Oceans M Closure ofTethyan connection to theAtlantic Ocean Occurrence of worldwide reef-building E Extinction of rudists Widening of the Atlantic Ocean Opening of the Caribbean Seaand the Gulf of Mexico Uplift and vocanicity in the Mid-Pacific S Early representatives of the families Fungiidae, Pocilloporidae, Poritidae, Agariciidae, Oculinidae, Merulinidae, Rhizangiidae, 0 Flabellidae, Micrabaciidae, Turbinoliidae, Dendrophylliidae Cretaceous 144 mya Reefs dominated by Rudists Z Corals now important members of reef communities First appearance of coralline algae Early representatives of the families Acroporidae, 0 Faviidae, Siderastreidae, Meandrinidae, Parasmiliidae, Guyniidae Diversification of families of Scleractinia First representatives of non-reef-building, I Jurassic 208mya azooxanthellate corals, the caryophylliids Early representatives of the Astrocoeniidae Acquisition of endosymbionts byscleractinians C Triassic 245mya Appearance of Scleractinia in the Tethys Sea

_._ __.._- .. _--_._--_.- -_. __..-- .. Table 1.2 Classification scheme of Alloiteau (1952) for the order Scleractinia. Taxa with extant representatives are in boldface, extinct taxa are in plain type.

ARCHAEOCOENIIDA ASTRAEOIDA AMPHIASTRAEIDA FUNGIIDA EUPSAMMIIDA

Pinacophyllidae Montlivaltiidae Amphiastraeidae Lamellofungiidae Eupsammidae Seriatoporldae Placosmilidae Mitrodendronidae Cyclophyllopsiidae Turbinarldae Acroporldae Isastreidae Cyclastraeidae Stylophoridae Clausastraeidae ConophyllUdae Actlnastreidae Faviidae Procyclolitidae Stylophyllidae HeliaBtreidae Haplaraeidae Oulastreidae Andemantastreidae AstrangHdae Thamnasteriidae C1l Echinoporidae Agariciidae (Jl Placocaenlidae Funginellidae Columnastreldae Asteroserlidae Oculinidae Synastraeidae Anthempbylliidae Acrosmiliidae Mussidae Cunnolitidae Pectiniinae Siderastreidae Merulinidae Microsolenidae Brachyphyllidae Dermosmiliidae STYLINIDA MEANDRIIDA CARYOPHYLLIIDA Latomeandriidae Fungiidae Cyathophoridae Smilotrochiidae Thecocyathiidae Micrabaciidae Stylinidae Dendrogyrlidae Turbinollldae Agathiphylliidae Heterocaenlidae Meandriidae Caryophylliidae Poritidae Euheliidae Stylocaeniidae Guyniidae Actinacididae Hemiporitidae Parasmiliidae Flabellidae Table 1,3. Classification scheme of Chevalier and Beauvais (1987) for the order Scleractinia. Taxa with extant representatives are in boldface, extinct taxa are in plain type.

STYLOPHYLLIINA PACHYTHECALINA DISTICHOPHYLLINA ARCHEOCOENIINA ARCHAEOFUNGIINA FUNGIINA

Zardinophyllidae Pachythecalidae Coryphyllidae Tropiphyllidae Cyclastreidae Agarididae Stylophyllidae Volzeidae Distichophyllidae Astrocoeniidae Procyclolitidae Funginellidae Margarophyllidae Seriatoporidae Cyclophyllopsiidae Andemantastreidae Acroporidae Thamnasteriidae Dermosmiliidae Synastraeidae Latomeandriidae Cunnolitidae Microsolenidae Siderastreldae U1 0'.1 Psammocoridae Funglacyathldae Fungidae Poritldae

STYLININA FAVIINA MEANDRIINA CARYOPHYLLIINA DENDROPHYLLIINA

Cya thophoridae Montlivaltiidae Meandrlnidae Thecocyathiidae Dendrophylliidae Stylinidae Favlidae Galaxeldae Caryophylltdae Micrabadidae Trachyphylliidae Desmophyllidae Astraraeidae Merulinidae Dasmiidae Actinicididae Diploastraeidae Parasmiliidae Mussidae Turbinolidae Pectiniidae Eusmilidae Astrangiidae Guyniidae Oculinidae Flabellidae Anthemiphyllidae Figure 1.1 Wells' (1956) hypothesis for the evolutionary history of the Scleractinia

57 I

I 1::::::::::::4 A;:riciic:le ~ Srylophyllieae

~ F~..mgiicce f:::::.:::::.:J Faviicce ~ ~ ::,.~ Pcriticoe II,.'. l: Flabellieee ,~ _ Ceryophylliicce Astrocoeniina Fungiina Faviina Caryophylliina

58 01 U> Figure 1.2 Veron's (1995) hypothesis for the evolutionary history of the Scleractinia 09 o ." 0 t.. i ~ c: ~ ;. m c6' ~ ." iii' ell III in o III III c: 2 § § § i III o' o' en iii s i N > I\) .p. n ~I~ U1 J ~ ~ II) I~ 8 II) C - 2.CD ::i" :'~,_ ~=ae:= II) r .."$ ACROPORIDAE

,.~!!!!!!!!!~===::u SIDERASTREIDAE = AGARICIIDAE

"T1 c: ~ ae 19. iii~::::::~~~~~e~t FUNGIACYATHIDAEMICRABACIIDAE :5" II) ~ FUNGIIDAE

=ANTHEMIPHYLLIIDAE RHIZANGIIDAE PECTINIDAE _ MUSSIDAE MERULINIDAE

FAVIIDAE

~ TRACHVPHYlLlIDAE 1~.lliidae

CARYOPHYLLIIDAE

_ OCULINIDAE ,o0-?i," :ii~' ;r ;3:~ 7/11<1 I 5' '- MEANDRINIDAE '-.Lilli I I . PORITIDAE , . J.J.-I I I Adine idlidae L.!l8 "" ..,..----;;;...~ ••11 DENDROPHYLIIDAE m ...L

Figure 1.3 Roniewicz and Morycowa's (1995) hypothesis for the evolutionary history of the Scleractinia ~9 ; ;g 2 ~ -. CD c: ~ - 'i' l: ;c

II.SOlOorph. f r .c~inlstrllds "

hapiareids \---""""----aetinaeidids iu.treids '~'.~ Th,cOSlIJili/J • v. gunU/6tl / IIlontliyaltiids - -- \kobyutrllds elauu.trllds JI, IPI/Oiros"is-llyriophyllu/O group \'- aiiii _

~6Iophy"i6 TItJ/'''''6st.rll group hi!". 0 '- \) :=I\J euihstreids .::( r, I --- '-----.ndrods'o.'iii"..m.o'iiiiiiiii 1/""" ~t I \ ~ - a.s.t..r...... o,.llo..r. • · ... 1I: Ph!!.} aiero.ol.nid• ..:If'x rhipidogyrid. , i ! / stylinid. i I '!I Protolt,t,rlstu'l group! doraos ..iliids ~ i 7:I - - - '" PIIcoco,nll "dS I yoIZ.iid. '··r·~".:.~ ~: Ii ' __L "'_ -Int,rsllilil group I i i p - llu••ldae ~ --~.-L· a ._ If ::I;e.etlnllda. t. R, FaYlidae ·j .. ,. ... r} ·~I, • I. \ , ..... w.... VV~ II r- • :~~ v : ~ ••rgarophyllild. r------..J.1-- 1 Agarlellda...... • I _ rel••nlphyllilds _ / ""?". .' i~. I \\ ICaryDpftylidu ?'-.'ii ~~ .s: \ tOeullnidu ______--_-:...._~i OIlldrDphylil:l•• v V .tylophyllids \ ..; w Flab.lIldae ii .pl'I!"liid. n \ ·n c ~ ! ill£, i'•J ; 0' ::a; ---.-----~ ::a .;.FIlNI cIYonity I , ::a = ::a - - i I early Mesozoic ! Mlddl. Mesozoic i Lata Masozoic i Canozoic Shg. I I CHAPTER 2. Molecular Evolution of a Portion of the Mitochondrial16S Ribosomal Gene Region in Scleractinian Corals

Abstract

Relationships among families and suborders of scleractinian corals are poorly understood because of difficulties 1) in making inferences about the evolution of the morphological characters used in coral taxonomy and 2) in interpreting their 240 million year fossil record. Here I describe the use of molecular techniques to gain a new perspective on the evolution of relationships within the order. A segment of the mitochondrial 16S ribosomal gene from taxa of 14 families of corals has been amplified and sequenced. These sequences are homologous to other metazoan 16S ribosomal sequences and fall into two distinct clades defined by size of the amplified gene product. Comparisons of sequences from the two clades demonstrate that both sets of sequences are evolving under similar evolutionary constraints: they do not differ in nucleotide composition, numbers of transition and transversion substitutions, spatial patterns of substitutions, or in rates of divergence. The characteristics and patterns observed in these sequences as well as the secondary structures, are similar to those observed in mitochondrial ribosomal 165 DNA sequences from other taxa. The phylogenetic hypothesis generated from these sequences differs from traditional hypotheses for evolutionary relationships among the Scleractinia and suggests that a reevaluation of evolutionary affinities is needed.

------.. _- -_. Introduction

Scleractinian corals are reef-building polyp animals that first appeared in the shallow waters of the Tethys Sea in the mid-Triassic approximately 240 million years ago (mya) (Table 1.1, Stanley 1981). The biology and ecology of this order of anthozoans have been extensively studied both in the field and in the laboratory. Hypotheses about evolutionary relationships among the Scleractinia are based on the study of the calcareous exoskeleton that the polyps secrete. However, these animals are reknowned for the morphological plasticity of their skeleton, which is molded by both environmental and genetic factors (Wijsman-Best 1974, Veron and Pichon 1976, Brakel1977, Zibrowius 1984, Foster 1985,Willis 1985,Veron and Kelley 1988).The identification of species of corals has traditionally been-a serious problem for coral biologists and taxonomists (Wallace 1974, Wijsman-Best 1974, Randall 1976, BrakeI1977, Best et al. 1984, Zibrowius 1984, Chevalier and Beauvais 1987, Budd 1988, Willis 1990,Gattuso et al. 1991,Hidaka 1992). Due to this variability, the relationships between fossil corals and extant corals are difficult to interpret (Veron in press). Relationships among families and suborders, based on morphological characters and the fossil record, are still considered tenuous at best (Vaughan and Wells 1943, Veron, 1994 pers.

comm). The Scleractinia belong to the phylum Cnidaria, a group of simple, multicellular organisms that diverged from the base of the metazoan tree. It is only in the past 10 years that the molecular biology of the Cnidaria has started to be explored due to the development of rapid methods for amplifying and sequencing discreet segments of DNA. Since the 19805, 64

------McMillan, Miller and coworkers (McMillan et al. 1988, McMillan and Miller 1989, McMillan and Miller 1990, Miller et al. 1990, McMillan et al. 1991, Miller et al. 1993) have used nuclear DNA sequences to study the relationship among closely related species in the family Acroporidae. Nuclear ribosomal DNA sequences have been used to elucidate the evolutionary relationships among classes of cnidarians and among metazoan phyla (Christen et al. 1991.. Hori and Satow 1991, Wainright et al. 1993). Warrior and Gall (1985) found that members of the class Hydrozoa had linear mitochondrial DNA (mtDNA) as opposed to the circular mtDNA found in other metazoans. Bridge et al. (1992) have subsequently shown that other hydrozoans as well as members of the class Scyphozoa possess linear mtDNA whereas members of the Anthozoa possess circular mtDNA. They used these data to reconstruct class level relationships within the Cnidaria. The entire mitochondrial genome from the anthozoan Metridium senile has been sequenced and found to possess characters not seen in any other metazoans (Wolstenholme 1992). The use of these molecular techniques has enabled us to gain some knowledge of the genetics of cnidarians and they hold great promise for further developing our understanding of the evolution of these animals. In general, molecular techniques have proved to be powerful tools for the study of evolution and, in many taxa such as primates and echinoderms, for addressing evolutionary questions where other methods have failed. In the Scleractinia, these techniques have the potential for providing a new tool, in the form of genotypic characters, to study the evolutionary history of the group, that is independent of the morphological variability inherent in corals. The use of these techniques for the study of scleractinians has become feasible for two main reasons. Rapid methods for amplification of the large amounts

65 of pure DNA necessary, for sequencing require only small amounts of animal tissue via the Polymerase Chain Reaction (peR). For corals this is especially advantageous because the animal is only a thin veneer over the skeleton and so it is difficult to obtain large quantities of the animal tissue. Secondly, many corals live in association with intracellular dinoflagellate symbionts. Thus, separation of coral and algal genomes is crucial. The specificity of primer annealing in PCR has made it possible to amplify only animal DNA from intact associations of corals and their endosymbionts. I have applied these molecular techniques to examine family and suborder level relationships among the Scleractinia. The mitochondrial 16S ribosomal gene region was chosen for these analyses because it has been found to be useful in studies of relationships ranging from within species to among kingdoms (e.g., DeSalle et al, 1987, Cedergren et al. 1988, Miyamoto et al. 1989, Gatesy et al. 1992,Fang et al. 1993). The most recent classification of the Scleractinia by Veron (1995) divides extant corals into seven suborders some of which diverged from each other 240 mya based on fossil evidence. Here I describe the refinement of molecular techniques for the peR amplification of a segment of the mitochondrial large ribosomal RNA gene of corals. I have analyzed patterns of molecular evolution in the sequences from these amplified gene segments and used these sequences in phylogenetic analyses of relationships among the Scleractinia.

Methods

DNA Extractions DNA extractions were carried out on living, frozen or ethanol preserved colonies. Some specimens were collected live from Hawaiian

_._--- -_._----- waters or obtained from the Waikiki Aquarium. Some of these samples were frozen at -70°C before extraction. Specimens from Guam and the Solomon Islands were preserved in 70% ethanol. All colonies were rinsed in distilled water before freezing or extracting. Two kinds of extractions were carried out: genomic extractions and highly enriched mt DNA extractions. For genomic extractions, tissue was removed. from approximately 4 cm2 of skeleton. In some cases, living and frozen colonies were placed into a solution of 100 mM EDTA and 20 mM Tris (pH 7.5) with a drop of dimethylpolysiloxane antifoam ( "Foam Fighter", Crescent Research Chemicals, Phoenix, AZ) to decrease the formation of mucus. The coral fragments were incubated at 4°C and periodically agitated for 2 to 3 hours. This procedure served to dissociate the cells from the calcium carbonate skeleton. The slurry of cells obtained from this procedure was centrifuged at 1500 g for 10 minutes. The pellet was resuspended in 1ml of DNA isolating solution (100mM EDTA, 1% 50S, 10 mM Tris, pH 7.5). Alternatively, tissue was either scraped from the skeleton with scalpel and forceps, or in colonies with large polyps (e.g, Tubasiraea, Echinopora, Galaxea, Caulastrea) the mesenteries from 1-10 corallites were removed. These tissues were placed directly in 1ml of DNA isolating solution, to which 1J.1g of proteinase K was added. The solution was incubated at 65°C overnight, with periodic agitation. After incubation, samples were microcentrifuged at 10,000 g for 30 seconds. An equal volume of Tris-equilibrated phenol was added to the supernatant, mixed to form an emulsion, and this solution microcentrifuged at 10,000 g for 10 min to eliminate proteins. This step was repeated 2-3 times if necessary to obtain a clear supernatant. An equal volume of 1:1 phenol/chloroform was added to this supernatant, mixed to form an

-_. --_._-- --._------. -- emulsion, and microcentrifuged at 10,000 g for 5 min. This supernatant was then washed with an equal volume of chloroform to eliminate any remaining impurities and centrifuged for 5 min at 10,000 g. Whole cell genomic DNA was obtained from this supernatant by precipitation with a 1/2 volume of 7.5M sodium acetate (pH 7.5) and 3 volumes of ice-cold ethanol for at least 30 min, centrifuged at 12,000 g for 10 min, decanted, washed with 70% ethanol, dried, and resuspended in 25 -SO J1l of double distilled water or Tris EDTA (TE).

Extractions to obtain samples highly enriched in mtDNA were carried out using a modified protocol (Beckman et al. M.P. Smith pers. comm. 1993) for the ProMega Wizard Minipreps system. living corals were used in this procedure. Approximately 1 cm2 of coral tissue plus skeleton, or the mesenteries from 1-5 corallites, or 1-5 polyps, were placed in 1.5 m.l of sucrose buffer (250 mM sucrose, 10 mM EDTA, 100 mM Tris, pH 7.4) and pulverized in a glass homogenizer, on ice. This slurry was microcentrifuged at approximately 1500g for 10 minutes to pellet cells, cellular debris and skeleton. The resulting supernatant was microcentrifuged at approximately 10,000 g for 10 minutes to pellet nuclei and mitochondria. This pellet was resuspended in 200 J.Ll of Cell Resuspension Solution (50 mM Tris-HCI, pH 7.5, 10mM EDTA, 100 J.Lg/ml RNAase). Organelles were lysed by addition of 200 J.1l of Cell Lysis Solution (0.2N NaOH, 1% SDS) and then 200 J.1l Neutralization Solution (1.32M KOAc, pH 4.8) was added. This solution was microcentrifuged at 10,000 g for five minutes to pellet precipitated proteins and high molecular weight DNA. The resulting supernatant was added to 1 ml of DNA Purification Resin to bind the DNA and the solution pushed through a Wizard Minicolunm to conserve only the resin with bound DNA.

------_. -- The filter was rinsed with 2 ml of Column Wash Solution (200mM NaCl,

20mM Tris-HCI, pH 7.5, 5 mM EDTA, 55% ethanol) and microcentrifuged at 10,000 g for 30 seconds to remove any remaining traces of the wash solution. The DNA was eluted from the resin column by addition of 50 JJ.1 of TE warmed to 70°C and collected by microcentrifuging at 10,000 g for 30 seconds.

Polymerase Chain Reaction Extracted DNA was used as template for the PeR. A 1:100 dilution of genomic DNA extraction or undiluted mtDNA-enriched extraction were used. PCR products and sequences were initially obtained from the azooxanthellate coral Tubastraea coccinea in order to develop mtDNA primers that were specific for coral DNA and that could eventually be used with samples from zooxanthellate corals. Target segments of mtDNA from T. coccinea were obtained using primers designed for the 3' end of the mtDNA 165 ribosomal gene region of the azooxanthellate hydroid Hydra vulgaris. These are primer 1 (Cunningham and Buss 1993), 165g-L: 5' TCGACTGlTfACCAAAAACATAGC-3', (Fig. 2.1; this primer is the equivalent of primer 165ar-L [Palumbi et al. 1991] with an additional three nucleotides on the 3' end), and primer 1651-H (Cunningham pers. comm. 1990), 5'-TITAAAGGTCGAACAGACC-3' (Fig. 2.1; the position of this primer is five bases 5' relative to 165br-H [Palumbi et al, 1991]). PeR conditions used were 10 cycles of 30s at 94°C for denaturation, 30s at 45°C for primer annealing, and 45s at noc for extension, followed by 30 cycles of 30s at 94°C, 30s at SO°C, and 1 min at n°e. The PCR product was electrophoresed in Ix TAE on a 2% agarose gel. The PCR product (a band of 500-600 base pairs base pairs) was excised, soaked in 1 ml double distilled water three times, placed in 50 JJ.1 double distilled water, then frozen at -70°C and thawed, three times,

------_.-. -_. forcing the DNA out of the agarose slice. One J.L1 of this purified template was used in a PCR reaction. This PeR product was purified and concentrated, used as template in double strand sequencing reactions, and electrophoresed on an

8% acrylamide gel with a gradient buffer system (Palumbi et al. 1991). The T. coccinea sequence was aligned with sequence from Hydra (Cunningham and Buss 1993), from the frog Xenopus (Roe et al. 1985), and from a sea urchin, Strongylocentrotus purpuratus (lacobs et al, 1988). Primers specific for coral tissue were designed from this sequence by choosing a region that was unique to T. coccinea, the 5' primer 16Sd-L, 5'-GGTGAGACCTGCCCAATGGTI-3', and the 3' primer 16Sc-H, 5'-AACAGCGCAATAACGTITGAGAG-3' (Fig. 2.1). These coral specific primers were subsequently used in PeR reactions with all other coral samples. Reactions were carried out in 100 J.1l, with 0.5 ~ of each primer, 0.8 JlM dNTPs, 1% DMSO in PCR buffer (Palumbi et al. 1991) and 1% template. PCR conditions were 40 cycles of 30s at 94°C, 30s at 50°C, and 1 min at 72°C. Amplifications of both genomic and mtDNA-enriched samples of T. coccinea were carried out using primers provided by C. Hunter (Dept. of Botany, University of Hawaii) for the nuclear ITS region. Conditions for these amplifications were the same as described above. aoning Due to problems encountered in consistently obtaining good DNA template for sequencing, double strand PCR products were ligated into pBluescript IT KS- (Stratagene; using the modified protocol of Marchuk et al. 1991, Palumbi and Baker 1994). An 18 J.1l cocktail of 0.5 mM ATP, 0.1 pg T vector, 1 unit ECO RV and 5-10 units of ligase in ligation buffer (50 mM Tris, 10 mM MgQ2, 10 mM DTI, made as 4X working stock) was added to 2 J.L1 of

_...- ..- .... _------._. --- PCR product. This ligation was allowed to procede overnight at room temperature before being used to transform competent Escherichia coli cells (Stratagene Bacterial Strain XLI-Blue MRF'). Two ~ of ligation mixture were added to 30 ul of competent cells, and incubated for 30 minutes to attach to the outside of cells. This solution was then warmed to 42°C for 90s to allow vector to enter the cell and then cooled on ice for 1-2 minutes. The transformed cells were added to 200 J.1l DYT and allowed to incubate for one hour before being grown overnight at 370C on DYT plates containing ampicillin (100J.!g/ml), tetracycline (30 J.lg/ml), X-gal (100 J.lg/ml) and IPTG (280 ug), Colonies containing target segment inserts were identified by PCR assay of white colonies (Palumbi and Baker 1994). PCR assays were carried out as described above but in 25 J.L1 reactions using vector primers T3-long-L, 5' ATIAACCCTCACTAAAGGGAAC-3' and M13-H, 5'· . CATITfGCTGCCGGTCA·3'. Cells from white colonies 'were added directly to the reaction cocktail. PCR conditions were 30 cycles of 94°C for 305,55°C for 30s and 72°C for 30s. To obtain single stranded DNA from colonies containing the target segment, the colonies were inoculated in 3 ml of DYT with ampicillin (80 J.!g/ml) and 3x104-5pfu helper phage, allowed to grow for one hour before adding kanamycin (66 J.!g/ml) to select for cells that had taken up helper phage, and then grown overnight at 37°C in a shaking incubator. Cultures were microcentrifuged for 5 min at approximately 10,000 g to pellet cellular debris. The single stranded DNA in 1750 J.L1 of supernatant was precipitated by adding 270 J.L1 20% Polyethylene glycol/205M NaCI on ice for 30 min. Samples were microcentrifuged at 10,000 g for 5 min. The pellet was suspended in 400 III 2..5 M Ammonium Acetate. DNA was extracted by sequential addition of 400 J.L1 of buffered phenol and 400 ~ of chloroform.

------_.. - - This solution was microcentrifuged for 5 min at 10,000 g. The DNA in the supernatant was then ethanol precipitated, dried and suspended in 25 J1l of double distilled water (ddH20). This template was used in subsequent sequencing reactions. Sequencing Sequencing reactions were carried out using the Sanger dideoxy chain termination method. Primers and templates (7 J.1l of template, 1 J.IM primer in SFG sequencing buffer, Palumbi et al. 1991) were heated to 65°C in a heating block for two minutes and then allowed to anneal as the temperature dropped over 30 min to below 37°C. Labeling reactions were carried out for five minutes by adding 2 J1l ddH20,2 J1l TE, 1.0 J.1l 0.1 M OTT, 0.5 J.1l USB 5x labeling mix (dITP), 0.25 J.1l ATP_35S, and 0.25 J1l USB Sequenase Version 2.0 17 DNA Polymerase to the annealed template. Extension reactions were carried out for three minutes by adding 3.5 J1l of this cocktail to 2.5 J.1l of ddITPs of all four nucleotides. These reactions were stopped after 3 min with 4 J.1l of Stop solution (95% formamide, 20mM EDTA, 0.05% bromphenol blue, 0.05% xylene cyanol ff). Primers used for sequencing included vector primer Ml3-reverse,5'-GAATI'CAACAGCTATGACCATG-3', 16Sg,16Sc, 16Sd, and primers designed specifically for sequencing, 16Se-H,5' CGCCTITAAAAAAGTAAC-3' and 16Sf-L,5' CTACATCCAAATTGTTAGAC-3' (Fig. 21). Alignment and secondary structure Sequences were first aligned among themselves and then to outgroups. Alignments were carried out using the GCG sub-routine Pileup (Genetics Computer Group 1993). The secondary structure for these sequences was determined manually by comparison with proposed models of mitochondrial

------._.. _.- 165 ribosomal RNA secondary structure. Models hypothesized for Escherichia coli (Maly and Brimacombe 1983), Locusta migratoria (Uhlenbusch et al. 1987), Drosophila yakuba (Clary and Wolstenholme 1985), Polyamia neoyavapai (Fang et al, 1993), mouse (Glotz et al. 1981) and human (Anderson et al. 1981) were first compared and searched for conserved motifs (Hickson et al, 1994). These models were then used as templates along with guidelines suggested from previous studies of ribosomal evolution and secondary structure (Gutell et al. 1994, Hickson et al. Submitted ms) to predict the secondary structure of one robust sequence (Fungia scutaria) and one long sequence (Tubastraea coccinea) from the Scleractinia. Stems were required to be 2 base pairs or longer. Only G-C, G-T, and A-T bonds were allowed.

Phylogenetic analyses Two different phylogenetic analyses of aligned sequences were carried out using PAUP (Swofford 1993). In the first analysis, two scleractinian sequences were aligned with the published sequences of Hydra vulgaris (Cunningham and Buss 1993), Strongylocentrotus purpuratus (Jacobs et a1. 1988), Locusta migratoria (Uhlenbusch et al. 1987), and human (Anderson et al. 1981). In addition, other cnidarian sequences provided by Diane Bridge (1994) were used in the analysis. These included two sequences from another order in the class Anthozoa, the octocorals Renilla and Leptogorgia; two sequences from the class Scyphozoa, Aurelia and Pelagia; and two sequences from the class Hydrozoa, Liriope and Physalia. The most parsimonious reconstruction of the relationship among these 12 sequences was found with a branch and bound search in PAUP using only variable positions and designating the human sequence as the outgroup. The significance of these groupings was measured with 1000bootstrap replicates.

_._ .. _....._..- .._----- In the second phylogenetic analysis, all scleractinian sequences were aligned with Hydra vulgaris and Renilla. The most parsimonious hypothesis for the relationship among these taxa was generated with a heuristic search using a random addition sequence in PAUP, using only variable nucleotide positions, ignoring insertion/deletions, and with Hydra vulgaris designated as the outgroup. One hundred bootstrap replicates were carried out on these data. A neighbor joining analysis was also carried out on these sequences using PHYLIP (Felsenstein 1993).

Results

Extractions and peR Genomic extractions of scleractinian samples yielded between 0.75 and 250 J.l.g DNA. Sequences were obtained from 34 species of Scleractinia, representing 29 genera in 14 families (Table 21) from all seven extant suborders identified in the latest revision of the group (Veron 1995). Thirty three of these species were reef-building corals from the Indo-Pacific and one was a non-reef-building, deep sea species collected at 4100 m from the eastern

Pacific Ocean. peR with scleractinian samples resulted in the amplification of DNA segments ranging in size from 406-565 base pairs, corresponding to positions 5085-5528 of the mtDNA sequence of Stongylocentrotus purpuratus (Jacobs et al, 1988) and positions 13355-13010 of the mtDNA sequence of Drosophila yakuba (Clary and Wolstenholme 1985). When aligned with Renilla and Hydra, these sequences comprise 577 sites (Fig. 2.2), of which 262 are constant, 210 informative and 105 uninformative.

------._- _.... Alignment and secondary structure Alignment of all cnidarian sequences was straightforward (Fig. 2.2). There are 16 insertion/deletion events (Fig. 2.2) among the scleractinian sequences although they were not used in phylogenetic analyses. The secondary structures hypothesized for both short and long sequences are somewhat different from each other but both are similar to models proposed for other organisms (Fig. 23). The majority of stems and loops are conserved among both sequences. There are two major differences between the short and long sequences. First, the region in the long sequence consisting of six stem/loop regions (area 1 in Fig. 2.3,) is missing in the short sequence and is replaced by a region consisting of a single loop. Second, the stem loop structure in the long sequence (area 2 in Fig. 2.3) which consists of a stem with two branches consists of a single stem (area 2) in the short sequence. These two areas also differ in Escherichia coli, and Locusta migratoria. In area 1 in E. coli there are eight stem loop structures that are associated with the long principal stem while in L. migratoria there are only three stem loop structures associated with the homologous stem. In area 2, the stem loop structure in E. coli is branched while the homologous stem loop structure in L. migratoria is not. Substitutions among sequences occur in both stems and loops. When substitutions occur in stems (e.g., between Pocillopora and Fungia in the robust clade) they tend to preserve base pairing more than expected by chance alone. Out of changes in nine paired bases, six preserve pairings of adjacent strands whereas only 2.3 are expected (on average) by the rules developed by Dixon and Hillis (l993). In two sections, so many multiple changes occurred between Pocillopora and Pungia that new secondary structures need to be

.._.. _..-.-.- ._------drawn. Despite this, the proportion of paired bases in redrawn sections is about the same (53% in Pocillopora, 64% in Fungia). The region of the long sequence that is deleted in the short sequence is also a region of relatively high sequence variation among taxa of the complex clade. There are four small insertion/deletion events in this region ranging from 1-19 base pairs. Substitution rate also appears high. Acropora and Tubasiraea differ at 16 out of 98 bases in this region although they differ at only 10% of the positions over all in the 16S sequence. The other major region of sequence heterogeneity (approximately positions 420-540) also shows some differences in length, There are seven insertion/deletion events here, ranging up to 16 base pairs, and secondary structure diagrams differ greatly between long sequences, short sequences, and Pocillopora. Sequence differences in this region are so high between short and long sequences that alignment was difficult. Even among sequences of the same size, there are marked sequence differences. For example, Acropora and Tubastraea differ at 15 of 106 homologous positions in this region. Sequence compoSIOf Ion and SIZesize varivarIarIOn The mean base composition of the scleractinian sequences is 22.0% G, 33.5% A, 31.0% T and 13.0% C (Table 2.2). The mean difference (Jukes-Cantor corrected genetic distance) among all sequences is 18.2%, ranging from 0 to 31.4% (Table 2.2). All of the sequences are on average 46.1 % different from the two anthozoan outgroups. There is a mean of 58.9% transition substitutions and 41.1% transversion substitutions among all sequences (Table 2.2). In the most parsimonious reconstruction of the relationships among all of the scleractinian sequences, with Renilla and Hydra as outgroups, the scleractinian sequences fall into two major clades (Fig. 2.4) which are on

76 average 29.4% (range 20.3-32.2) different from each other. These two major clades can be defined by size as a result of insertion/deletion events. One clade is characterized by shorter segments that range in size from 406-425 bp and differ from each other by a mean of 7.6%(Table 2.2). This clade is representative of robust corals (which will be described in detail in Chapter 3) and will subsequently be referred to as the robust clade. The other clade is characterized by longer segments that range in size from 536-565bp and that are a mean of 7.5% different from each other (Table 2:2). It is representative of complex corals (which will be described in detail in Chapter 3) and will subsequently be referred to as the complex clade. Conservation of segment size within genera and families was confirmed by amplification of mtDNA-enriched samples extracted using ProMega Wizard Minipreps from 31 species of corals, including 10 species of corals that had not previously been sampled and that were not sequenced (Table 2.1). These 10 taxa included two genera from the Pocilloporidae, six different species of Acropora, and two different species of Porites (Table 2.1). Amplification of all mtDNA-enriched samples generated segments of the same size as generated from other members of the family or genus that had been amplified and sequenced. Characteristics of robust and complex clades The segments from the robust and complex clades were examined in a variety of ways that suggest that they are 1) mitochondrial in origin and 2) homologous to each other. One sequence from each of these clades was used in a phylogenetic analysis with 165 mitochondrial sequences from cnidarians (Fig. 2.5). In this analysis both of the sequences are more closely related to each other than to any other taxa (the bootstrap value for this grouping is 95%; Fig.

------25). The base composition of each clade is similar to the other and does not differ from all of the scleractinian sequences taken together (Table 2.2). To examine the possibility that sequences from either the robust or complex clade are not of mitochondrial origin (e.g., a nuclear pseudogene, Gel1issen and Michaelis 1987), segment size was compared from amplifications of both genomic DNA and mtDNA-enriched fractions from 17 species. In 14 of these, amplifications from mtDNA-enriched samples generated the same size segments as was generated from total genomic DNA. Amplification of a mtDNA-enriched sample from one of these templates (T. coccinea) with nuclear ribosomal ITS primers failed, although amplification of genomic DNA was successful with these same primers. In three species, Turbinaria peltata, Euphyllia ancora, and Echinopora lamellosa, different size segments were generated from different extractions (Table 23). One of each of these sequences has been given the designation lS, and these B sequences will be discussed in detail below. Repeated DNA extractions gave similar, multiple amplification products (Table 2.3) showing that contamination was an unlikely source of B sequences. Amplification of the expected size fragment tended to occur from mtDNA-enriched samples (10 of 12 observations) whereas amplifications from total genomic DNA were more likely to give the 'wrong' size product (9 of 12 observations). These results suggest that both the short and the long segments are generated from mtDNA but that is sequences may represent nuclear pseudogenes, jumping peR products, or other odd gremlins. Further comparisons of patterns of evolution in the robust and complex clades, described below, also support the hypothesis that sequences in both clades are homologous to each other.

78 Transitions and transversions The percentage of transition and transversion substitutions between the clades was similar (Table 22). Overall, transitions accumulate more quickly than transversions in both clades (Fig. 26). This pattern was especially true in comparisons of closely related taxa. For species within genera, transitions outnumbered transversions 9:1. For closely related genera (with sequence divergence <1%) transitions outnumbered transversions 3:1, except in the clade that includes the genera in the Faviina and Caryophylliina (Fig. 2.4). In this clade, seven of.the eight observed changes are transversions. Closer inspection reveals that five of these eight changes are at positions 236 and 369, which closely flank the major deletion observed in these genera (positions 237-366 in Fig. 22). Thus these transversions may be caused by molecular mechanisms different from those operating .elsewhere in the gene.

Spatial pattern of variabiHty Spatial patterns of nucleotide substitution across the sequenced region in the two clades show substantial similarities. The genetic distance between two sequences was calculated for 60 bp windows along the sequences. This was repeated by shifting the 60 bp window three base pairs downstream along the entire sequence. The results of this analysis give a visual image of variability along the length of two sequences. Comparisons were carried out between two of the most different sequences within each of the major clades (Figs. 2.7a,b), between two of the most different sequences among the two major clades (Fig. 2.7c), and between two sequences from two different species of sea urchins (Fig. 2.7d). In all of these 'comparisons the segment of 16S analyzed is relatively conserved over the first 200 nucleotides, does not align very well over the next 50-100 nucleotides, is conserved in the following 50

------_.- nucleotides, and then variable again for the next 100 nucleotides before the final nucleotides that are fairly well aligned The degree of variability in the poorly aligned regions is least when comparing two long sequences, higher among the two sequences from robust corals, and greatest when comparing a sequence from a robust and a complex coral. When comparing sequences from robust and complex corals, there is also a poorly aligned region in the first 100 nucleotides. The first gap in this graph (Fig. 27c) indicates a region where there is an insertion/deletion between the two sequences. The second gap in the graph is an area where there is so much variability that the sequences are poorly aligned. Comparison of the same 16s mtDNA gene region from two genera of sea urchins [from two different families thought to have diverged approximately 60 mya (Smith et ale 1992)] results in a similar trimodal pattern. When comparing distantly related sequences of the robust (Fig. 2.7a) or complex (Fig. 27b) clades, the same areas tend to show high sequence variability. These regions of sequence variability also correspond to regions of variability in secondary structure. It is these regions where major differences in secondary structure exist between the two clades and more minor differences exist within clades. These similar patterns suggest that both sets of sequences appear to be evolving under similar evolutionary constraints, and that these constraints are similar to those in the ribosomal RNA of other phyla.

Phylogenetic analyses Parsimony analysis verified that the coral sequences reported here clustered with other Cnidaria, and that the Cnidaria as a whole form a monophyletic group. In the topology generated in this analysis (Fig. 2.5), the 80 Scleractinian sequences group together, and this grouping is supported by a bootstrap value of 95%. They are most closely related to two other anthozoan taxa from the order Octocorallia, and these four taxa form a monophyletic group supported by a bootstrap value of 59%. Taxa from the other cnidarian classes group together and all of the cnidarians form a monophyletic clade supported by a bootstrap value of 86%.These results show that the sequences reported here are derived from coral" mitochondrial genes, not from zooxanthellae or coral nuclear genes (but see Discussion of alternative hypotheses). The evolutionary hypothesis for the relationships among the Scleractinia and in relation to two cnidarian outgroups, generated by parsimony analysis, is presented in Fig. 2.4. All of the branches in the topology are supported by bootstrap values of 80% or higher. The same topology was generated by a neighbor-joining analysis. The Scleractinia sequences fall into two major clades that differ from each other by a mean of 29.4%. These two clades are also characterized by an insertion/deletion that occurs just after position 250 (Fig. 2.2). The taxa in the robust clade, shown on the upper branch of the topology, fall into three major groups. One group, consisting of two species in the family Pocilloporidae, is rather distinct (14%) from other members of the robust clade. The other two groups in the robust clade are more closely related to each other. The complex clade, shown on the lower branch of the topology, is composed of four groups that are approximately equally different from each other. All species and genera placed in the same family by morphological criteria grouped together (groupings supported at bootstrap values of >80%),

81 with one exception. This exception was one genus (Leptastrea) in the Faviidae that grouped with the family Siderastreidae in the molecular topology. Sequence length is conserved within genera and families. The same size segment was always amplified from the 6 genera for which more than one species was assayed (four congeneric pairs, three species of Porites and eight species of Acropora), including ten species that were not sequenced (Table 21). Except for the three species from which both segments were amplified, the same size segment was also always amplified from genera in the same family, including two genera that were not sequenced (Table 2.1, Fig. 24). Although generic and family designations are broadly supported by Fig. 24, relationships above this taxonomic level are unexpected.

Discussion

Evidence from amplification of genomic DNA and mtDNA-enriched samples, phylogenetic analysis, and comparisons with other animals supports the hypothesis that all scleractinian sequences reported here are most likely 165 mtDNA sequences from the coral animal. Sequences amplified from both azooxanthellate and zooxanthellate scleractinians appear to be homologous to 165 sequences from other metazoans (Fig. 2.5). In addition, the Scleractinia form a monophyletic group within the Cnidaria and, as expected, are most closely related to other members of the class Anthozoa (Fig. 2.5). Taxa from other cnidarians group together as classes and all of the' cnidarian sequences form a clade that is monophyletic (Fig. 25). The nucleotide composition of these sequences from both major scleractinian clades does not differ greatly from each other or from the nucleotide composition of other marine invertebrates like sea urchins (Table

82 22). Size variation in this region of the mitochondrial 165 gene is known from insects with divergence times of up to 250 mya (Derr et al. 1992,Pashley and Ke 1992,Fang et al. 1993). This size variation appears to be largely the result of an insertion/deletion event that took place early on in the evolutionary history of the Scleractinia.

Secondary structure The secondary structures hypothesized for sequences from robust and complex corals are similar to secondary structure models for Escherichia coli, insects and mammals (Fig. 2.3). They differ from other models in that apparently conserved motifs observed in the sequences from these other organisms are not all present in the scleractinian sequences. These motifs were also not found in the secondary structure of nematodes (Gutell et al. 1992). Absence of conserved motifs may be a consequence of the limited number of taxa (of the metazoa, mostly insects and mammals) from which mitochondrial 165 DNA sequences are available with which comparisons can be made. Comparison of mitochondrial 125 ribosomal RNA sequences from 175 invertebrate and vertebrate taxa has provided a much clearer view of the secondary structure of this molecule (Hickson et al, Submitted ms), A similar study of available 165 ribosomal DNA sequences would greatly enhance secondary structure hypotheses for this molecule as well. In spite of the similarity of coral secondary structure models to those for other organisms, the secondary structures for the short and long sequences differ from each other as a result of the insertion/deletion event between them. Variation is known in three-dimensional rRNA structure (Gutell et al. 1994). It is possible that differences in structure affect the functioning of the ribosomes, but little is understood about functional correlations within

83 ribosomal secondary structure (Gerbi 1985). For example, it has been demonstrated that protein synthesis is not affected by deletion of a large stem loop region of the 165 rRNA of EscherichiJZ coli (Gravel et ale 1989). Because so little is known about ribosomal tertiary structure and function, there is no reason to conclude that the large deletion in the robust clade could not be functional. In every other respect, both short and long sequences are similar to other functional ribosomal sequences. 5tudy of the entire mitochondrial 165 ribosomal DNA from these organisms would also help in evaluating the significance of the different structures in corals.

Evolutionaty rates of the Scleractinia An attempt to estimate rates of evolution of these ribosomal sequences was made using relatively closely related genera and families with the same first date of appearance in the fossil record (Table2.4). We estimate the rate to be about 0.10% divergence/million years (myr) or less. This rate is similar to the divergence rates of mitochondrial 165 sequences observed among insects (DeSalle et ale 1987) though much lower than the rate of approximately 0.5 to 1% per myr observed in bovids (Gatesy et ale 1992). The estimate of this rate could be improved using taxa whose divergence times are known precisely. Further studies of mitochondrial 165 DNA sequences from a much wider variety of taxa will be necessary to determine whether different taxa are evolving at different rates. De5alle et ale (1987) suggested that insect 165 sequences were evolving slowly because of high constraints imposed by a high AT content. Coral sequences do not have such a highly skewed nucleotide bias, and the reasons for the slow evolutionary rate are unknown.

84 P..Attems of molecular evolution of scleractinian sequences A comparison of patterns of molecular evolution in sequences belonging to the two major scleractinian clades demonstrates that sequences

from both clades are evolving under evolutionary constraints similar to those observed in other metazoan mitochondrial 165 ribosomal DNA sequences. The approximately equal number of transition substitutions and transversion substitutions among both clades of scleractinian sequences is also observed in sea urchins (Table 22) and Drosophila (DeSalle et al. 1987). There are slightly more transitions than transversions among the scleractinian sequences in contrast to the sea urchin and insect sequences. The sea urchins are represented by two genera whose genetic distance is 22% (Table 2.2). Comparisons of 165 sequences from more closely related species show a higher transition transversion bias of about 2:1 (Kessing 1991). High transition biases are typical of mitochondrial DNA evolution, and are thought to reflect lack of efficient mis-match repair mechanisms in mitochondria. However, transition biases are known to be lower in the 165 gene regions than at silent sites of protein coding regions (DeSalle et al. 1987). In insects, low transitions may be related to the high AT bias shown by mtDNA in these groups that is thought to be the result of functional constraints (DeSalle 1992?). The low number of transitions along with the relatively low transition transversion bias in corals may be the result of functional constraints in the ribosomal DNA. 'The spatial pattern of nucleotide substitutions is also similar in each clade (Figs. 2.7a,b). In comparisons of sequences from the robust and complex clades, the region where a gap occurs between the sequences and the region where the sequences align poorly are the regions of highest variability in 85

------_ •. - ._. ------._._------_._- ---- comparisons of sequences within each clade (Fig. 2.7>. Homologous regions are also highly variable when different species of sea urchins are compared (Fig. 2.7d). The regions of highest variability are also the regions where differences in secondary structure are observed in both clades, as well as in E. coli and L. migratoria (Fig. 2.3). These comparisons are further evidence that evolutionary constraints on sequence evolution of coral 165 sequences are similar to those acting in other taxa. BSequences The amplification of two different sequences from three species is unexpected. For each species, one sequence falls clearly in the same phylogenetic position as other members of its family. The second sequence (8) is most similar to corals in another suborder. One possible hypothesis for these 8 sequences is that they are the result of heteroplasmy occuring in these taxa, possibly due to a hybridization event in the past. Pacific corals live in close association with each other. They are often broadcast spawners (Richmond and Hunter 1990) that spawn once a year in synchrony (Harrison et al. 1984, Harrison and Wallace 1990, Richmond and Hunter 1990)and so a natural opportunity for hybridization exists. In fertilization experiments, hybridization has been observed between closely related and distantly related species of the genera Acropora, Montipora and Platygyra (Willis et ale 1992). Hybridization was also observed among favids and acroporids, families from both major clades. The hybrids within genera were able to complete development and continue growing for at least three years. Development of hybrids between families across clades has not yet been observed. It is difficult to imagine that hybridization alone could account for these observations if mitochondrial DNA is maternally transmitted in corals

------_. as is the case for most organisms studied (Wilson et al. 1985, Avise 1986). However, unusual modes of mtDNA inheritance have been observed in other animals (Zouros et al. in press). It is possible that mitochondrial transmission in corals is different from what is known for other animals, or that paternal transmission during hybridization occurs as is hypthesized for mussels (Zouros et al. in press). In this study, one colony from one location was sampled for the majority of species. In all cases sequences of the same size were generated from species in the same genus and genera in the same family. It is possible that further sampling of more individuals in each species from a wider variety of locations and of more species and genera will demonstrate that IS sequences are more common than observed here. The sequence from Leptastrea that was the only example of a genus not grouping with other family members (although it grouped in the same major clade) may be another example of one of these IS sequences. Further investigation will be necessary to more completely characterize this phenomenon.

Alternatiye hypotheses The evidence presented here strongly supports the hypothesis that all of the scleractinian sequences are from mitochondrial16S ribosomal DNA and that there was an early split in the ancient corals that led to robust and complex clades. However, major differences in secondary structure along with the large amount of genetic distance among sequences from the robust and complex clades require that alternative hypotheses be considered. Short and long sequences do not appear to be the result of contamination of peR reactions. All results were consistent and repeatable: any given sample always produced amplification products of the same size.

------_. --.-_._------Negative controls included in sets of PCR reactions were always blank. In the three species where both size segments were amplified (Table 23), repeated extractions gave similar, multiple amplification products. One alternative hypothesis is that either the short or the long sequences are actually derived from a pseudogene. Although mitochondrial pseudogenes have been observed in the nuclear genome of other organisms (Iacobs et ale 1983, Tsuzuki et ale 1983, Fukuda et al, 1985, Gellissen and Michaelis 1987, Zullo et ale 1991, Smith et ale 1992, Lopez et ale 1994) they have characteristics that are generally very different from those observed in functional mtDNA sequences. Pseudogenes are not thought to have any evolutionary constraints (Li et ale 1985). For this reason, no patterns in spatial distribution of substitutions would be expected after a great length of time (e.g., over 200 million years for corals).. and rates of evolution are expected to be relatively high (Li and Graur 1991). As described above, none of these predictions are supported by examination and comparison of the scleractinian sequences. Another alternative hypothesis is that either the short or the long sequences are mitochondrial pseudogenes that exist in the mitochondrial or nuclear genome. The predicted secondary structure for the short sequences, which appears modified relative to the long sequences and other secondary structure models, may be evidence for this hypothesis. However, it was possible to amplify both short and long sequences from only three taxa. In 41 of 44 cases, only a single amplification product was observed. In 19 cases, this single product was short whereas 22 cases showed only a long amplification. In none of these cases was there evidence of a second amplification product. If the short sequence represents a pseudogene in the mitochondria, does this

------_. imply these taxa have no copy of the long sequence? Because lack of a functional 16S gene is unlikely, I conclude that both short and long versions of the mtDNA amplification are functional. It is also possible that one of these sequences is from a duplication of the mitochondrial 16S ribosomal gene within the mitochondrial genome. Duplication of large segments of the mitochondrial genome, including ribosomal genes, has been observed in lizards (Moritz and Brown 1986, Moritz and Brown 1987). Because of the large amount of divergence between the short and long sequences, this duplication would have had to occur over 300 mya. Subsequently, different families might have lost one or the other of the copies independently, shedding either the robust clade or complex clade copy at some point in time. It seems unlikely that two copies of the same functional ribosomal genes would be maintained in the mitochondrial genome over evolutionary time as the mitochondrial genome is known for its genetic economy (Attardi 1985). In addition, families might appear in different clades, but they should still group into suborders within clades correctly. That is.families of, say, the suborder Faviina that retained the short sequence should cluster together as should families in this suborder that retained the long sequence. In contrast to the prediction, the topologies of the two clades are clearly different from one another. In mussels of the genus Mytilu5, two mitochondrial genotypes occur that differ by about 20% (Zouros et, al, in press). MtDNA inheritance is sex specific in these organisms. Females have one mtDNA type whereas males have the other mtDNA type but may beheteroplasmic for both types. Because most corals are hermaphrodites, such a gender specific system is unlikely in the present case.

------._- The hypothesis best supported by the data gathered in this study is that both the short and the long sequences are from the mitochondrial16S ribosomal RNA gene region and are both functional. Divergence between these two clades represents divergence of two lineages of coral species not the duplication of two loci.

Eyolutionary relationships among the Scleractinia The evolutionary hypothesis for the relationships among the Scleractinia (Fig. 2.4) generated from these data is robust. All groupings shown in Fig. 2.4 are supported by bootstrap values of 82% or higher while the two major clades on the tree are supported by bootstrap values of 100%. Genera within families, as determined by morphological taxonomy, group together on the molecular topology. By contrast, relationships among families and suborders hypothesized from morphological data are not supported by the molecular data. The approximate rate of divergence of 0.10% or less suggests that the divergence time of the two major clades of Scleractinia is greater than 300 mya. This date is conservatively calculated and is earlier than estimates of 240 mya as the first appearance of the Scleractinia in the fossil record (Stanley 1981). As a result, it is possible that the divergence of the two major clades of the Scleractinia occurred before the acquisition of a calcium carbonate skeleton which was subsequently invented twice, but more precise rate estimates will be nec~ssary to determine this. A more comprehensive examination of coral evolutionary history is presented in Chapter 3.

------TABLE 2.1 Speciesof Scleractinia sampled. Speciesare in phylogenetic order based on Veron (1995). For certain genera it was not possible to identify the species and these are indicated by sp. Source indicates geographic origin of coral samples. 'Aquar' indicates that these samples were obtained from the Waikiki Aquarium. Extraction describes what kind of extraction was carried out (total genomic VB. mtDNA-enriched) and whether samples were living, frozen, or preserved in ethanol. ..indicates samples that were amplified but not sequenced.

SEGMENT SUBORDER FAMILY SPECIES SOURCE EXTRACTION SIZE

Archaeocoenilna Pocilloporldae Pocillopora damicornis Hawaii Genomic/frozen/Wizard Short Pocillopora meandrina Hawaii Genomic/frozen/Wizard Short Seriatopora hystrix" Fiji/Aquar Wizard/living Short Stylophora pistillata" Fiji/Aquar Wizard/living Short Acroporidae Montipora uerrucosa Hawaii Genomic/frozen/Wizard Long Montipora digitata Palau/Aquar Wizard/living long Anacropora sp. Palau/Aquar Wizard/living long Acropora cytherea Guam Genomic/alcohol long Acropora humilis Guam Genomic/alcohol Long Acropora formosa" Fiji/Aquar Wizard/living long \0 ..- Acropora austera" Fiji/Aquar Wizard/living long Acropora nana" Fiji/Aquar Wizard/living Long Acropora micropthalma" Fiji/Aquar Wizard/living Long Acropora elseyi" Fiji/Aquar Wizard/livlng Long Acropora sp." Guam/Aquar Wizard/living Long

Fungiina Siderastreidae Psammacora stellata Hawaii Genomic/frozen/Wizard Short Coscinaraea sp. SolomonIslands Genomic/alcohol Short Agarididae Pavona varians Hawaii Genomic/frozen/Wizard Long Leptoseris incrustans Hawaii Wizard/living Long Fungiidae Fungia (Lobactis) scutaria Hawaii Genomic/frozen/Wizard . Short Fungia (Cycloseris) fragilis Hawaii Genomic/living Short Fungia (Cycloseris) vaughani Hawaii Genomic/living Short Zoopilus echinatus Fiji/Aquar Genomic/frozen Short Fungiacyathidae Fungiacyathus marenzellerl Eastern Pacific Genomic/alcohol Long Table 2.1 (continued) Speciesof Scleractinia sampled

SEGMENf SUBORDER FAMILY SPECIES SOURCE EXTRACTION SIZE

Faviina Pectiniidae Pectinia alcicornis Palau/Aquar Genomic/living/Wizard Short Mussidae Lobophyllia hemprichii Palau/Aquar Genomic/living/Wizard Short MeruUnidae Hydnophora rigida Palau/Aquar Genomic/frozen Short Merulina scabricula Fijl/Aquar Wizard/living Short Faviidae Caulastrea furcata Flji/Aquar Genomic/frozen/Wizard Short Cyphastrea ocellina Hawaii Genomic/frozen/Wizard Short Echinopora lamellosa Fiji/Aquar Genomic/frozen/Wizard Both Lepiastrea bottae Hawaii Wizard/living/Wizard Short Leptoria phrygia Guam Genomic/alcohol Short

Caryophylliina Caryophyllldae Catalaphyllia jardinei Indo-Pac/Aquar Genomic/living Short Euphyllia ancora Palau/Aquar Genomic/living Both \0 Wizard/living N Meandriina Oculinidae Galaxea [ascicularis Guam Genomic/alcohol/Wizard Loog Achrelia horrescens Fiji/Aquar Genomic/frozen Long

Porltllna Porltldae Porites compressa Hawaii Genomic/frozen/Wizard Loog Porites lobata" Hawaii Wizard/living Long Porites cylindrica" Fijl/Aquar Wizard/living Long Gonipora stokesii Palau/Aquar Genomic/frozen Long Goniopora &p. Palau/Aquar Genomic/living Long

Dendrophyllilna Dendrophylllidae Turbinaria peltata Indo-Pac/Aquar Genomic/frozen/Wizard Both Tubastraea coccinea Hawaii Genomic/living/Wizard Long Table 2.2 Base composition, substitutions and distance among all sequences and within major clades.

Among all Among all Among all Among short sequences long sequences sequences sea urchinsa

G (%) 19.9 24.1 22.0 22.0

A (%) 35.3 31.9 33.5 33.0

T (%) 32.7 29.9 31.0 26.2 \0 w C (%) 12.1 14.1 .13.0 18.8

Transition substitutions (%) 54.8 61.9 58.9 45.2

Transversion substitutions (%) 45.2 38.1 41.1 54.8

Distanceb (mean %) 7.65 7.5 18.2 22.0

a Stongylocentrotus purpuratus and Paracentrotus lividus, genera from two families thought to have diverged from each other 60 mya (Smith et a1. 1992) b Jukes Cantor corrected genetic distance (Li and Graur 1991) I'

Table 2.3 Amplification of short and long sequences from total genomic and mtDNA-enriched extractions of three species of corals.

Amplification ofShort Amplification of long sequences sequences Clade of other froriitolal- - -IrommlDNA:-----fiomtotal fronimTDNA- Coral Species family members genomic DNA enriched genomic DNA enriched

Turbinaria peltata long 2* o 2 4

~ Euphyllia ancora short o 3 4" 2"

Echinopora lamellosa short 1 3 3" o

.. amplification of incorrect fragment size as determined from comparison to other genera in the same families Table 2.4 Divergence dates, genetic distance, and divergence rates calculated from mitochondrial 165 sequences from various taxa.

Genetic Divergence Divergence distance rate Taxa compared dates (mya) (%) (%/myd Coralss Genera 30-60 0.7-2.5 0.02-0.04 Families 67-100 3.9-11.6 0.06-0.12 Insectsb \0 Drosophila species 2-60 0.4-3 0.05-0.2 Ut Aedes albopictus vs, Drosophila species 200 14.0 0.07 Bovidsv Genera: Damaliscus dorcas vs. D. hunteri 2.5 6.6 2.6 Tribes: Bos taurus vs. Tragelaphus angasi U.S 15.5 1.3

a Coral genera compared: Acropora vs. Montipora (40-50mya), Pavona vs. Leptoseris, (30-40 mya), Porites vs. Goniopora (40-60 mya), Caulastrea V5. Leptoria (40-60 mya); Coral families compared: Poritidae (100 mya), Dendrophylliidae (100 mya), Agariciidae (67 mya), Oculinidae (67-100 mya), Acroporidae (67 mya) b Drosophila species compared: sproati, pilimana, disjuncta, affinidisjuncta, mimica, stigma, clavisetae, [unebris, yakuba; data from DeSalle et al. 1987. C Data from Gatesy et al. 1992 D", F ... ~E ~C 5' I -t 179 bp -t 199 bp I 159 bp H 3 ' ~ .., ~ Hydra primer 1 (G) Hydra primer L

Figure 2.1. Diagram of mitochondrial 165 ribosomal gene region amplified showing placement and orientation of primers. \0 -...J Figure 2.2 Alignment of 37 scleractinian mitochondrial 165 ribosomal sequences and. 2 cnidarian outgroups, Gaps are indicated by '-' and '.' indicates a nucleotide matching the first nucleotide in that column. Missing data are indicated by'?'. Outgroup sequences were provided by D. Bridge. II I

10 20 30 40 50 60 Pocillopora damicornis GAATCTTAAT ATGTTTGTTC CGCTTA-CTA ATAAAGACAA TTAAATGGCC GCGGTAACAC Pocillopora meandrina Cycloseris fragilis .T ••••A:.A .AA •••••• TT ••••• -T •• Cycloseris vaughani •T ••••A ••A .AA••••••TT ••••• -T•• Fungia scutaria .T ••••A ••A .AA ••••••TT •••••-T •• zoopilus echinatus .T ....A ••A .AA .•.... TT ..... -T .. Coscinarea sp. .G••••A ••A .AA•••••• TT •••••-T•• Psarnmocora stellata .T ••••A ••A .AA •••••• TT •••••-T•• Leptastrea bottae .G....A ..A .AA •.•.•. TT •.••• -T .. Euphyllia ancora •T ••••A ••A•A •••••••T T.....-T •• • ••••••••GC ••••••••• Catalaphyllia jardinei •T ••••A ••A .A••••••• TT ••••• -T •• • ••••••••GC ••••••••• Turbinaria peltata ~ •G ••••A ••A .A••••••• TT ••••• -T •• • ••••••••GC ••••••••• Merulina scabricula •G ••••A ••A .A••••••• TT •••••-T •• • ••••••••GC ••••••••• Hydnophora sp. •G ••••A ••A .A•••••••TT ••••• -T •• • ••••••••GC ••••••••• Echinopora lamellosa •G ••••A ••A .A ••••••• TT ••••• -T •• • ••••••••GC ••••••••• Caulastraea furcata .G ••••A ••A .A••••••• TT ••••• -T•• • ••••••••GC ••••••••• Cyphastrea ocellina •G ••••A ••A .A••••••• TT ••••• -T •• • ••••••••GC ••••••••• \0 Leptoria phrygia .CA..A .A T T -T .. • •••••• ••G c . 00 Pectinia alcicornis • G. • ••A ••A•A ••••••• TT •••••-T •• • •••••• -I • G c . Lobophyllia hemprichii .G••••A ••A .A•••••••TT •••••-T •• • ••••••••T Achrelia horrescens •T ••••A ••A .G•••G ••AG A ••CA.CT •• ••T ••••••• Galaxea fascicularis .T••••A ••A .G•••G ••AG A ••CA.CT •• ••T ••••••• Euphyllia ancora B •T •••••••A .G•••G ••AG A ••CA.CT •• ••T ••••••• Acropora cytherea AT •••••••A .G•••A ••AG G •••A.CT •• • •••••• T •• Acropora humilis AT •••••••A .G •••A ••AG G •••A.CT •• • •••••• T •• Anacropora sp. AT •••••••A .G•••A ••AG G •••A.CT •• • •••••• T •• Montipora verrucosa AT •••••••A .G•••A ••AG G •••A.CT •• • •••••• T •• Montipora digitata AT •••••••A .G'•••A ••AG G •••A.CT •• • ••••••T •• Turbinaria peltat3 .T ••••A ••A GG•••G ••AG A ••CAGCT.T Tubastraea coccinea .T••••A ••A GG•••G ••AG A ••CAGCT.T Echinopora lamellosa ~ .T••••A ••A GG•••G ••AG A •• CA.CT.T Goniopora stokesii •T ••••A ••A GG•••G ••AA A •••AGC •• T Goniopora sp. .T ••••A ••A GG•••G ••AA A •••AGC •• T Porites compressa .T ••••A ••A GG•••G ••AA A ••CA.CT.T •••••••••• Fungiacyathus marenzelleri .T ••••A ••A GG•••G ••AG A ••CAGCT.T •••••••••• •• T ••••••• Pavona varians •T ••••••GA •G •••G ••AA G •• GA • TT.. • ••••••••• • .T •••••• T Leptoseris incrustans .T •••••••A .G•••G ••AA G ••GA.TT •••••••••••• • .T •••••• T Renilla T.TAG.G••AC ••AGGAC.A ACG.CTAAAG CG •••CC ••• C.G ••••••• •• •• •• ••T. Hydra -.------AAT T •••• T .AT. • •••••••AT ••A •••••T. 70 80 90 100 110 120 Pocillopora darnicornis TGACTGTGAT AATGTAGCGT AATCAATTGT CAATTAATTG TTGACCGGTA TGAATGGTAT Pocillopora meandrina Cycloserls fragl1ls .A •••••••A • •••••••A. • ••A •••••• • ••••TA ••• Cycloseris vaughani .A•••••••A • •••••••A. • ••A•••••• • •••• TA ••• Fungia scutaria .A•••••••A • •••••••A. · ••AG••••. • •••• T •••• Zoopilus echinatus .A•••••••A • •••••••A. • ••A •••••• • ••••T •••• Coscinarea sp. .A•••••••A • •••••••A. • ••A •••••• • •••• T •••• Psammocora stellata •A •••••••A • •••••••A. • ••A •••••• • ••••T •••• Leptastrea bottae .A•••••••A • •••••••A. • ••A •••••• • •••• T •••• Euphyllia ancora .A•••••••A T ••••••••• • •••••••G. Catalaphyllia jardinei .A•••••••A T ••••••••• • •••••••G. Turbinaria peltata n .A•••••••A T ••••••••• • •••••••G. Merulina scabricula .A•••••••A T ••••••••• • •••••••G. Hydnophora sp. .A •••••••A T ••••••••• • •••••••G. Echlnopora lamellosa .A .•••••.A T ••••••••• • •••••••G. Caulastraea furcata .A •••••••A T ••••••••• • •••••••G. Cyphastrea ocellina .A •••••••A T ••••••••• • •••••••G. \0 Leptoria phrygia .A••.....A T ••••••••• • •••••••G. \0 Pectinia alcicornis .A•••••••A T ••••••••• • •••••••G. Lobophyllia hemprichii .A •••••••A T ••••••••• • •••••••G.

Achrella horrescens • •••C ••••• • ••••••A •• • ••G •••••• • ••••• 0 • G. Galaxea fasclcularis • •••C ••••• • ••••••A •• • ••G •••••• • •••••••G. Euphyllia ancora n • •••C ••••A • ••••••A •• • ••G •••••• • •••••••G. Acropora cytherea .A•• C •••.A • •••. .•A .• • ••G •••••• • •••••••G. Acropora hurnilis .A•• C ••••A • ••••••A •• • •• G •••••• Anacropora sp. •A ••C ••••A • ••••••A •• • ••G•••••• • •••••••G. Montlpora verrucosa .A ••C ••••A • ••••••A •• • ••G •••••• • •••••••G. Montipora digitata .A••C ••••A • ••••••A •• • ••G •••••• • •••••••G. Turbinaria peltata • •••C ••••• • ••••••A •• • ••G •••••• • •••••••G. Tubastraea coccinea • •••C ••••• • ••••••A •• • ••G •••••• • •••••••G. Echinopora lamellosa n • •••C ••••• • ••••••A •• • ••G •••••• • •••••••G. Goniopora stokesii • •••C ••••• • ••••••A •• • ••G •••••• • •••••••G. Goniopora sp. • •••C •••• • ••••••A •• • •• G •••••• • •••••••G. Porites compressa • •••C ••••• • ••••••A •• • •• G •••••• • •••••••G. Fungiacyathus marenzelleri • •••C ••••• • ••••••A •• • •• G •••••• • •••••••G. Pavona varians • •••C ••••A • ••••••••A • ••A •••A •• • •••••••G. Leptoseris incrustans • •••C ••••A • ••••••••A • ••A •••A •• • •••••••G. Renl11a • ••• C ••••• · AC ..•A•..A.C •••••••• •• • ••G.G- ••• · A.C Hydra • ••••••A •• • .A•••••A. • ••T.G•••• ••T .•••••• A ••GAGA.A. • •••••••TA 130 140 150 160 170 180 Pocillopora damicornis CCCGAAAGTT TTTCTGTCTT AAAAAAATAT TCAATGAAAT TAAATCTGTA GTGAAGATGC pocillopora meandrina ...... ·.. '" ...... · ...... ·...... ·...... Cycloseris fragilis .A•••••••• ·...... • ••••••••C•T •••••••• .G•••T •••• ·...... Cycloseris vaughani •A •••••••• ·...... • ••••••••C•T •••••••• •G••• T •••• ·...... Fungia scutaria •A •••••••• ·...... • ••••••••C•T •••••••• .G••• T •••• ·...... Zoopilus echinatus •A •••••••• ·...... · ...... e .T .•••.••• .G .•. T "' ••• ...... Coscinarea sp. •A •••••••• ...... • ••••••••C•T •••••••• .G•••T •••• ·...... Psammocora stellata •A •••••••• ·...... • ••••••••C•T •••••••• .G •••T •••• ·...... Leptastrea bottae •A •••••••• ·...... • ••••••••C•T •••••••• .G•••T •••• ·...... Euphyllia ancora .A •••••••• .c ...... • •••••••• C•T •••••••• .G••• T •••• ·...... Catalaphyllia jardinei •A •••••••• .C•••••••• •••••••••C .T•• ; ••••• .G••• T •••• ·...... Turbinaria peltata B .A•.•••••• •C •••••••• • ••••••••C•T •••••••• .G••• T •••• ·...... Merulina scabricula •A •••••••• .c ...... • ••••••••C•T •••••••• .G••• T •••• ·...... Hydnophora sp. •A •••••••• •C •••••••• • ••••••••C•T •••••••• .G ••• T •••• ·...... Echinopora lamellosa .A •••••••• •C •••••••• • ••••••••C•T •••••••• •G ••• T •••• ·...... Caulastraea furcata .A •••••••• •C •••••••• • •••••••• C•T •••••••• .G •..T.••. ·...... Cyphastrea ocellina .A •••.•••• .c ...... • •••••• ••C•T •••••••• .G••• T •••• ·...... 0 Leptoria phrygia •A •••••••• •C •••••••• • ••••••••C•T •••••••• .G••• T •••• -0 ·...... Pectinia alcicornis •A •••••••• .c ...... • •••••• ••C• T •••••••• .G ••. T••.. ·...... Loboph}'llia hemprichii •A •••••••• •C •••••••• • ••••••••C•T •••••••• .G•••T •••• ·......

Achrella horrescens •A ••••G ••C .CA••••••• • .GG ••••T. C ••G •••••• •G ••• T •••• • •••• 0 •••• Galaxea fascicularis .A•••• G••C .CA••••••• • .GG ••••T. C •• G•••••• .G•.• T •••• ·...... Euphyllia ancora B •A •••GG••C .CA••••••• • .GG ••••T. C ••G •••••• .G•••T •••• ·...... Acropora cytherea .A•••GG••C .CA••••••• • .GG ••••C. C ••G •••••• .G...T ••.. ·...... Acropora humilis .A•••GG••C .CA••••••• • .GG ••••C. C •• G•••••• .G.•• T •••. ·...... Anacropora sp. •A ••• GG••C .CA••••••• • .GG•...c. C•.G•••••• .G•••T •••• ·...... Montlpora verrucosa .A •••GG••C .CA••••••• ••GG••••C. C •• G •••••• •G••• T •••• ·...... Montipora digitata .A •••GG••C .CA••••••• • .GG ••••C. C •• G •••••• .G ••• T •••• ·...... Turbinaria peltata •A ••• GG••C .CA••••••• • .GG ••••T. C •• G •••••• .G••• T •••• ·...... Tubastraea coccinea .A•••GG••C .CA••••••• • .GG ••••T. C •• G •••••• .G••• T •••• ·...... Echinopora lamellosa B .A •••GG••C .CA••••••• • .GG ••••T. C •• G •••••• .G••• T •••• ·...... Goniopora stokesii .A•••GG••C .CG••••••• • .G•••••G. C •• G •••••• .G••• T •••• ·...... Goniopora sp. .A••• GG••C .CA••••••• • .G•••••G. C •• G•••••• •G •••T •••• ·...... Porites compres8a .A •••GG••C .CA••••••• • .G•••••G. C •• G•••••• .G•••T •••• ·...... Fungiacyathus marenzelleri .A•••GG••C .CA ••••••• • .G•••••T. C ••G •••••• •G ••• T •••• ...... Pavona varians .A•••GG••C .CG••••••• ••GG••G.T. C •• G•••••• •G••CT •••• ·...... Leptoseris incrustans .A•••GG••C .CG••••••• ••GG••G.T. C •• G •••••• •G ••CT •••• ·...... Renilla .A•••GG•• C ••A ••••••C ••G.G••A.G C ••••••••• ••T .AT •••• • •••••••A• Hydra .A•••• TT •• •CA••••••• • .-••••A •• .TTT.A•••• •G ••ATAA•• ••T ••••••• 190 200 210 220 230 240 Pocillopora darnicornis TACATTTTAA TTGTTAGACG AGAAGTCCCC ATGGAGCTTT ACTGTAAGCT TATATATATA Pocillopora meandrina ...... ·...... ·...... ·...... ·...... ·...... Cycloseris fragilis ••••• •A ••• • •••A ••••• ...... • •••A ••A •• • .AGA----- Cycloseris vaughani • •••••A ••• • •••A ••••• ...... ·...... • •••A ••A •• • .AGA-· --- Fungia scutaria • •••••A ••• ••••A ••••• ...... ·...... • •••A ••A •• • .AGA----- Zoopilus echinatus ... II •••A .... · .. .. .A ...... ·...... • •••A ••A •• • .AGA----- Coscinarea sp. • •••••A ••• • •••A ••••• ·...... ·...... • •••A ••A •• • .AGA----- Psammocora stellata ••••••A ••• • •••A ••••• ·...... ·...... • •••A ••A •• • .AGA----- Leptastrea bottae • •••••A ••• ••• •A ••••• ...... • ••••• 0 .... • •••A ••A •• • .AGA----- Euphyllia ancora • •••••AA •• ••••A ••••• ·...... ·...... • •••AGG ••• • .-GA.---- Catalaphyllia jardinei • •••••AA •• • •••A ••••• ·...... ·...... ••••AGG ••• • .-GA.---- Turbinaria peltata B ••••••AA •• • •••A ••••• ...... ·...... • •••AGG ••• • .-GA.---- Merulina scabricula • •••••AA •• • •••A ••••• ...... ·...... • •••AGG ••• •• -GAT---- Hydnophora sp. • •••••AA •• • •••A ••••• ·...... ••• •AGG ••• • .-GAT---- Echinopora lamellosa ••••••AA •• ... ".A...... ·...... ·...... AGG .... •• -GA.---- Caulastraea furcata ••••••AA •• • •••A ••••• ·...... ·.... " ..... •• • •AGG ••• • .-GA.---- Cyphastrea ocellina • •••••AA •• • •••A ••••• ·...... ••••AGG••• •• -GA------0 Leptoria phrygia ••••••AA •• ••••A ••••• ...... ·...... ••••AGG••• ••-GAC---- Pectinia alcicornis • •••••AA •• ••••A ••••• ·...... ·...... ••••AGG••• ••-GA.---- - •• ••AGG ••• Lobophyllia hemprichii • •••••AA •• • .. ••A ...... ·...... • •• ••A •••• ·.GGGG---- Achrelia horrescens •••• •CCA •• • ••••••••A •••••A •••• • ••A.A•••• •• ••G •• A •• •• •G.GGC.- Galaxea fascicularis ••• • •CCA •• • ••••••••A •••••A •••• • ••A.A•••• • •••G ••A •• • ••G.GGC.- Euphyllia ancora B •••••CCA •• • ••••••••A •••••A •••• • ••A.A •••• • •••G ••A •• •••G.GGC.- ACrOpora cytherea • ••••CCA •• ·...... • ••••A •••• • •. •.A.••• • •••G ••A •• • ••G.GGC.- Acropora hum11is • ••••CCA •• ·...... • ••••A •••• • ••••A •••• ••••G ••A •• • ••G.GGCC- Anacropora sp. • ••••CCA •• ·...... • ••••A •••• • ••••A •••• • •••G ••A •• •• • G.CGC.- Montipora verrucosa •••• •CCA •• ·...... •••••A •••• •••••A •••• • •••G ••A •• •• •G.CGC.- Montipora digitata • ••••CCA •• ·...... • ••••A •••• • ••••A •••• ••••G ••A •• •• • G.CGC.- Turbinaria peltata •••• •CCA •• • •••••••• A •••••••••• ....•A .•.. ••••G ••A •• •••G.GGC.T Tubastraea coccinea •••••CCA •• · ...•.. .•A •...••.•.. • ••••A •••• • •••G ••A •• •• •G.GGC.T Echinopora lamellosa B • ••••CCA •• • ••••••••A •••••••••• •••• •A •••• • •••G ••A •• •••G.GGC.T Goniopora stokesii • ••••CCA •• • ••••••••A •••••••••• •••• •A •••• • •••G ••A •• • ••G.GGC.T Goniopora sp. •••••CCA •• • ••••••••A •••••••••• • ••••A •••• • •••G ••A •• • ••G.GGC.T Porites cornpressa •••••CCA •• • ••••••••A •••••••••• • ••••A •••• • •••G ••A •• •••G.GGC.T Fungiacyathus marenzelleri • ••••CCA •• · ...... A .•..•...•• • ••••A •••• • •••G.GA•• • •• GAGGC. T Pavona varians • ••••C.A•• • ••••••••A •••••••••• • ••••A •••• ••••G.GA•• •• •G.GGC.- Leptoseris incrustans •••••C.A•• • ••••••••A •••••••••• • ••••A •••• • •••G.GA•• •• • G.GGC.- Renilla •••••AAAC. • •••A ••••• •A •••A ••• T ••C ••••••• ••••G.TA.C G •••ACGT •• Hydra • .TT.AAA-. • •••A ••••• •AG ••A ••• T ••A ••••••• • ••A •••A •• .T.T.C.TA• 250 260 270 280 290 300 Pocillopora darnicornis ATTTTTTTT------Pocillopora meandrina · .------Cycloseris fragilis Cycloseris vaughani Fungia scutaria Zoopilus echinatus Coscinarea sp. Psammocora stellata Leptastrea bottae Euphyllia ancora Catalaphyllia jardinei Turbinaria peltata 8 Merulina scabricula Hydnophora sp. Echinopora lamellosa Caulastraea furcata Cyphastrea ocellina o Leptoria phrygia -tv Pectinia alcicornis Lobophyllia hemprichii Achrelia horrescens ------GA CTTAAATAGT TTAAAAATGT GGTG------CGGGTCA Galaxea fascicularis ------GA CTTAAATAGT TTAAAAATGT GGTG------CGGGTCA Euphyllia ancora 8 ------GA CTTAAATAGT TTAAAAATGT GGTG------CGGGTCA Acropora cytherea ------GA CTTAAATAGT TTTAAAAGGT GGTG------CGGATTA Acropora humllis ------GA CTTAAATAGT TTTAAAAGGT GGTG------CGGATTA Anacropora sp. ------GA CTTAAATAGT TTTAAAATGT GGTG------CGGATTA Montlpora verrucosa ------GA CTTAAATAGT TTTAAAATGT GGTG------CGGATTA Montipora digitata ------GA CTTAAATAGT TTTAAAATGT GGTG------CGGATTA Turblnaria peltata .AA •• AA •• T ATTTCTTACA AATAAATAGT TTTTAAATGT GGTGTTAACC TCTCGGATTA Tubastraea coccinea · AA •.AA ..T ATTTCTTACA AATAAATAGT TTTTAAATGT GGTGTTAACC TCTCGGATTA Echinopora lamellosa 8 .AA ..AA ..T ATTTCTTACA AATAAATAGT TTTTAAATGT GGTGTTAACC TCTCGGATTA Goniopora stokesii · AA •.AA •• T ATTTCTTACA AATAAGTAGT TTTAAAATGT GGTGTTAGCC CCTTGGATAA Goniopora sp. .AA •• AA •• T ATTTCTTACA AATAAGTAGT TTTAAAATGT GGTGTTAATC CCTTGGATTA Porites compressa .AA •• AA .•T ATTTCTTACA AATAAGTAGT TTTAAAATGT GGTGTTAACC TCTTGGATTA Fungiacyathus marenzelleri .AA ..AA ..T ATTTCTTACA AATAAATAGT TTTAAAATGT GGTGTTAACC TCTCGGGTAA Pavona varians ------GA CTTAAATAGT TTTAAAATGT GGTG------TGACTCA Leptoseris incrustans ------GA CTTAAATAGT TTTAAAATGT GGTG------TGACTCA Renl11a Hydra 310 320 330 340 350 360 Pocillopora damicornis Pocillopora meandrina Cycloseris fragilis Cycloseris vaughani Fungia scutaria zoopilus echinatus Coscinarea sp. Psarnmocora stellata Leptastrea bottae EuphyIIia ancora Catalaphyllia jardinei Turbinaria peltata 8 Merul.ina scabricula Hydnophora sp. Echinopora lamellosa Caulastraea furcata ..... Cyphastrea ocellina o Leptoria phrygia ~ Pectinia alcicornis Lobophyllia hemprichii Achrelia horrescens ACTAAGGCTA GAAAAGCGCA CTTTTTTATT TGAAGAAAGG CAACTCACGA ACTTATGTTT Galaxea fascicularis ACTAAGGCTA GAAAAGCGCA CTTTTTTATT TGAAGAAAGG CAACTCACGA ACTTATGTTT EuphyIIia ancora 8 ACTAAGGCTA GAAAAGCGCA CTTTTTTATT TGAAGAAAGG CAACTCACGA ACTTATGTTT Acropora cytherea ATTAGGGTTA -AAAAGTTCA CTTTTTTATT TGAAGAAAGG CAACTCAAAA ACTTATATCT Acropora humilis ATTAGGGTTA -AAAAGTTCA CTTTTTTATT TGAAGAAAGG CAACTCAAAA ACTTATATCT Anacropora sp. ATTAGGGTTA -AAAAGCTCA CTTTTTTATT TGAAGAAAGG CAACTCAAAA ACTTATGTCT Montipora verrucosa ATTAGGGTTA -AAAAGCTCA CTTTTTTATT TGAAGAAAGG CAACTCAAAA ACTTATGTCT Montipora digitata ATTAGGGTTA -AAAAGCTCA CTTTTTTATT TGAAGAAAGG CAACTCAAAA ACTTATGTCT Turbinaria peltata ACGAGGGTTA GAAAAGCTCA CTCTTTTATT TTAAGAAAGC CAACTCAAAA ATTTATGTCT Tubastraea coccinea ACGAGGGTTA GAAAAGCTCA CTCTTTTATT TTAAGAAAGC CAACTCAAAA TTTTATGTCT Echinopora lameIIosa 8 ACGAGGGTTA GAAAAGCTCA CTCTTTTATT TTAAGAAAGC CAACTCAAAA ATTTATGTCT Goniopora stokesii ATTGGGGTTA GAAAAGCTCA CTCTTTTATT TTAAGAAAGC CAACTCAAAA ATTTATGTCT Goniopora sp. ATTGGGGTTA GAAAAGCTCA CTCTTTTATT TTAAGAAAGC CAACTCAAAA ATTTATGTCT Porites cornpressa ATTGGGGTTA GAAAAGCTCA CTCTTTTATT TTAAGAAAGC CAACTCAAAA ACTTATGTCT Fungiacyathus rnarenzelleri CCTAGGGTTA GAAAAGCTCA CTCTTTTATT TAAAGAAAGC CAACTCAAAG AATTATGTCT Pavona varians ACTAGGGTTA GAAAAGCTCA C-TTTTTATT TTAAGAAAGG CAACTTAAAA ATATATGTTT Leptoseris incrustans ACTAGGGTTA GAAAAGCTCA C-TTTTTATT TTAAGAAAGG CAACTTAAAA ATATATGTTT Renilla ------CTTAGAAT GATCGATACT AAGTACCCTA ATTCTAAGTT Hydra ------TTT------370 380 390 400 410 420 Pocillopora damicornis ------TTTT ATAAGTAAGA CAGTTTTGTT GGGGCGACAG TTTTTTAAAA AGTAACGAAA

Pocillopora meandrina ·...... • •••• 0 •••• ·...... ·...... ·...... Cycloseris fragilis ------••C. • ••••••G •• ·...... ·...... ·...... ·...... Cycloseris vaughani ------•• C. • ••••••G •• ·...... ·...... ·...... ·...... Fungia scutaria ------..c. • ••••••G •• ·...... ·...... ·...... ·...... Zoopilus echinatus ------••C. • ••••••G •• ·...... ·...... ·...... ·...... Coscinarea sp. ------..c. •••••••G •• ...... ·...... ·...... ·...... Psammocora stellata ------..c . •• •••••G •• ·...... ·...... ·...... ·...... Leptastrea bottae ------..c. • •••••• G •• ...... ·...... ·...... ·...... Euphyllia ancora ------••A. •••••eGG •• ·...... ·...... ·...... • •••••A ••• Catalaphyllia jardinei ------••A. •••••CGG•• ·...... ·...... ·...... • •••••A ••• Turbinaria peltata B ------•• C• ••• • • CGG•• ·...... ·...... ·...... • •••••A ••• Merulina scabricula ------••C• •••• • CGG•• ·...... ·...... ·...... • •••••A ••• Hydnophora sp. ------••C. • ••••CGG•• ·...... ·...... ·...... • •••••A ••• Echinopora lamellosa ------•• C• ••• • • CGG•• ·...... ·...... ·...... • .•• ••A •.. Caulastraea furcata ------••C. • ••••CGG•• ·...... ·...... ·...... • •••••A ••• 1-0 Cyphastrea ocellina ------••A. •••••CGG•• ·...... • •••• 0 •••• • •••••A ••• o ·...... • ••••CGG•• • ••••• 0 ••• ~ Leptoria phrygia ------... . ·...... ·...... • ••• ••A •.• Pectinia alcicornis ------•• C• •••• • CGG•• ·...... ·...... ·...... • ••• ••A ••• Lobophyllia hemprichii ------..c. • •• • •eGG •• ·...... ·...... ·...... • ••• ••A ••• Achrelia horrescens TTGGGG •••G • •••••GG•• • ••• ••G••• · ...... c. ec...- .... • ••••••••G Galaxea fascicularis TTGGGG •••G • •••••GG•• • •••••G ••• • •••••••C. CC •••A •••• • ••••••••G Euphyllia ancora B TTGGGG •••G ••••••GG•• • •••••G ••• • ••••A ••C. C:C •••A •••• • ••••••••G Acropora cytherea TTGGGA •••G • •••••GG•• ••••••G ••• • ••••••TC. CC •••A •••• • ••••••••G Acropora humilis TTGGGA •••G • •••••GG•• • •••••G ••• • •••••• TC. CC •••A •••• • ••••••••G Anacropora sp. TTGGGAC •• G •••• • • GG•• • ••• ••G••• •...... TC. CC •••A •••• • ••••••••G Montipora verrucosa TTGGGA •••G ••••• • GG•• • ••• ••G••• • ••.. • . TC. CC •••A •••• • ••••••••G Montipora digitata TTGGGGC •• G ••••• • GG•• • •••••G ••• • •••••• TC. CC •••A •••• • ••••••••G Turbinaria peltata TTGGGA ••AG • •••••GG•• • •••••G ••• · ...... c. CC •••A •••• • ••••••••G Tubastraea coccinea TTGGGA ••AG • •••••GG•• • •••••G ••• • •••••••C. CC •••A •••• • ••••••••G Echinopora lamellosa ~ TTGGGA ••GG • •••••GG•• • •••••G ••• • •••••••C. CC •••A •••• • ••••••••G Goniopora stokesii TTGGGG.G.G • •••••GG•• • •••••G ••• · ...... c. CC •••A •••• • ••••••••G Goniopora sp. TTGGGG.G.G • •••••GG•• • ••••• G ••• • •••••••C. CC •••A •••• • ••••••••G Porites compressa TTGGGA •• GG • •••••GG•• • •••••G ••• · ...... c. CC •••A •••• • •••••••• G Fungiacyathus marenzelleri TTGGGA •••G • •••••GG•• • •••••G ••• • •••••••C. CC •••A •••• • ••••••••G Pavona varians TTGGGA ••GG • •••••G ••• ••••••G ••• • •••••••C. CC •••A •••• • ••••••••G Leptoseris incrustans TTGGGA •• GG • ••••• G••• • ••••• G ••• · ...... c. CC •••A •••• • ••••••••G Renilla AATAAA •••• G.TG •••••• ••••••A ••• • ••••••• TA CC •••G ••T. • .A••••••G Hydra ------.G T •••••A ••• •• •••••• T. ·...... TA.------430 440 450 460 470 480 pocillopora damicornis ATGAGCTATG ACGCAT-GTT TAACTTTGAA ATTTTTTTAT TGATGAGAC------pocillopora meandrina ...... ------Cycloseris Eragilis ....A..... GTAA.AA.C .•C••.C....•GA •...--..A.GAGACA------Cycloseris vaughani · A. . . .. GTAA.AA. C. . C•..C. . .. . GA .•••--. .A.GAGACA------Fungia scutaria ·A•.... GTAA.AA.C. .C C.... .GA --. .A.GAGACA------Zoopilus echinatus ..•.A..•.. GTAA.AA.C •.CC....•GA --..A.GAGACA------Coscinarea sp. . ...A..•.. GTAA.AA.C ..C•••C...•.GA •...--••A.GAGACA------Psarnmocora stellata · ..•A GTAA.AA.C. .C .•.C. • .. .GA .•••--. .A.GAGACA------Leptastrea bottae ..•.A GTAA.AA.C ..C•..C.....GA ....--..A.GAGACA------Euphyllia ancora .C ..A.•... G--.T.A.C..C..CC ....•AA .•..--•.A.G.GACA------ Catalaphyllia jardinei .C ..A..... G--.T.A.C..C..CC .....M •...--.•A.G.GACA------Turbinaria peltata 8 .C .•A••... G--. T.A.C. .C .•CC. . •. .M••••--. .A.G.GACA------ Merulina scabricula .C ..A••.•. G--.T.A.C•.C•.CC .....AA ••••--.•A.G.GACA------ Hydnophora sp. .C •.A..... G--.T.A.C. .C ..CC. . .. .AA ....--. .A.G.GACA------ Echinopora lamellosa .C ..A....• G--.T.A.C. .C ..CC. . .. .AA •..•--. .A.G.GACA------ Caulastraea Eurcata ·C..A. . . .. G--. T.A.C. . C•.CC. . •• .AA ••••--. .A.G. GACA------.... Cyphastrea ocellina .C ..A..... G--.GAA.C •.C•.CC ..••.AA ••••--••A.G.GACA------o Leptoria phrygia .C ..A..•.. G--. T.A.C. .C ..CC. . •. .AA••••--. .A.G.GACA------ Ut Pectinia alcicornis .C •.A..••• G--.T.A.C.•C..CC •...•AA •••.--••A.G.GACA------Lobophyllia hemprichii .C ..A G--.G.A.C. .C .•CC. . .. .AAC ••.--. .A.G.GACA------Achrelia horrescens GC.G TAA GATACATA .....T.G ..TC TGAC.GCC.G G.GGAGAC.C A------Galaxea Eascicularis GC.G•..TAA GATACATA ...••T.G •. TC TGAC.GCC.G G.GGAGAC.C A------ Euphyllia ancora 8 GC.G•.•TAA GATACATA .• C•• T.G .•TC TGAC.GCC.G G.GGAGAC.C A------ Acropora cytherea GC.G.•.TM GATATATA .•.GGT.A .• TC TGAC.GCC.G G.GGA.AC.T A------ Acropora humilis GC.G...TAA GATATATA ••.GGT.A •.TC TGAC.GCC.G G.GGA.AC.T A------ Anacropora sp. GC.G•.•TAA GATATATA ••..GT.A•. TC TGAC.GCC.G G.GGA.AC.T A------ Montipora verrucosa GC.G..•TAA GATATATA ..••GT.A•. TC TGAC.GCC.G G.GGA.AC.T A------Montipora digitata GC.G TAA GATATATA ..•.GT.A..TC TGAC.GCC.G G.GGA.AC.T A------Turbinaria peltata GC.G TAA GATATATT ....G..A..TC TGAC.GC ..A G.GG.GAT.C T------Tubastraea coccinea GC.G••• TAA GATATATT •••.GT.A.•TC TGAC.GCC.A G.GG .•AT.C T------ Echinopora larnellosa 8 GC.G..•TAA GATATATT ••.•GT.A•. TC TGAC.GC •• A G.GG.GAT.C T------ Goniopora stokesii GC.G••. TAA GATATATT •••.GT.A.• TC TGAC.GCG.A G.GG.GAC.C T------ Goniopora sp, GC.G•..TAA GATATATT ••..GT.A•. TC TGAC.GCG.A G.GG.GAC.C T------ Porites compressa GC.G•••TAA GATATATT ...•GT.A..TC TGAC.GCG.A G.GG •• AT.C T------Fungiacyathu8 rnarenzelleri GC.G.••TAA GATACATA GTGA ..TC TGAC.GCC.A G.GGAGAT.C T------Pavona varians GC.GA ..TAA GATGTAT GT.A..TC TGAC.GC •• A G.GGAGAC.C A------Leptoseris incrustans GC.GA .•TAA GATGTAT .•••.GT.A•• TC TGAC.GC •• A G.GGAGAC.C A------Renilla GC ..•..TAT GGTAT.AA.A A..TC.CATT .GCC.CAC.G G•.ACAC CCCTAGCTGA Hydra -CA.AAATAA G.AATATAA. A•.T..ATTT •..G.A.A A.AC.ATTT AACAA----- 490 500 510 520 530 540 Pocillopora darnicornis ------A TTTTTGGTGT GTTTTTTGAT CC--GTTATT TTGAATGAAA Pocillopora meandrina ------. ·...... ·...... ·.--...... ·...... Cycloseris fragilis ------T .....A.... • ...•A.... ·•--...G.. .G ...... Cycloseris vaughani ------T ..•..A.... •••••A •••• • .--••• G •• .G ...... Fungia scutaria ------T .....A.•.. • ....A••.. ·. --•••G•• .G ....•..• Zoopilus echinatus ------T .....A.... • ....A.... ·. --...G.. ·G...... • Coscinarea sp. ------T .....A.... • ....A•... ·. --..•G•. .G...•...• Psarnmocora stellata ------T .....A.... · ....A.... ·. --...G.. ·G...... Leptastrea bottae ------T .....A.... · ....A.... ·. --...G.. .G...... Euphyllia ancora ------...... •••••A •••• · .--...G.. .G ...... •. Catalaphyllia jardinei ------·...... ••.A •... • • --•••G•• •G •••••••• Turbinaria peltata 8 ------·...... · •..•A•..• ·•--...G•• •G.••.••.. Merulina scabricula ------...... •••••A •••• • .--•••G•• .G•••••••• Hydnophora sp. ------·...... · ..••A.... ·. --.•.G.• .G ...... •. Echinopora lamellosa ------·...... • ....A•... ·. --...G•. •G.•.•..•• Caulastraea furcata ------·...... • •.•.A•... ·. --••.G.• •G.•.••.•• Cyphastrea ocellina ------·...... • •• ••A •••• • • --•••G •• .G •.•.•••. -0 Leptoria phrygia ------·...... • .• ••A•••• • • --•••G•• .G .....••. 0\ Pectinia alcicornis ------·...... • •• ••A •••• • • --•••G•• •G•••••••• LobophylHa hemprichli ------·...... •••••A•••A •• --•••G•• •G•••••••• Achrelia horrescens ---GAGCAGA TACTTATCTT .GGAC •...G •.••AAGTGA ••CGT.A••• •A.GG ••••• Galaxea fascicularis ---GAGCAGA CACTTATCTT .GGAC •...G •.••AAGTGA ••CGT.A••• •A.GG ••••• Euphyllia ancora 8 ---GAGCAGA CACTTATCTT .AGA .....G ..••AAGTGA ••CGT.A.•. •AAGG ..•.. Acropora cytherea ---GAGCAGA CACTTATCTT .GGA ••...G •...AAGTGA ••CGT.A.•. .A .GG••••. Acropora humilis ---GAGCAGA CACTTATCTT .GGA •••..G ••••AAGTGA .•CGT.A••. .A.GG••••. Anacropora sp. ---GAGCAGA CACTTATCTT .GGA •••••G ••.•AAGTGA ••CGT.A•.. .A.GG•••• < Montipora verrucosa ---GAGCAGA CACTTATCTT .GGA .....G .•..AAGTGA •.CGT.G.•• .A.GG ••••• Montipora digitata ---GAGCAGA CACTTATCTT .GGA ..•••G .•..AAGTGA ••CGT.A••• •A.GG.•••• Turbinaria peltata ---GAGCAGA CACTTA--TT ..•.A..•.G ••••AAGTGA .•CGT.A••• •A.GG ••••• Tubastraea coccinea ---GAGCAGG CACTTA--CT ....A..•.G ••••AAGTGA ••CGT.A•.• •A.GG.••.. Echinopora lamellos 8 ---GAGCAGA CACTTA--TT ..•.A.••.G ••••AAGTGA ••CGT.A••• •A.GG.•••• Goniopora stokesii ---GAGCAGA CACTTATTTT ••..A••..G .•..AAGTGA ••CGT.A••• •A.GG••••• Goniopora sp; ---GAGCAGA CACTTA-TTT .•..A....G ....AAGTGA .•CGT.A••. .A.GG..... Porites compressa ---GAGCAGG CACTTA---T ..•.A•...G ••.•AAGTGA ••CGT.A••• •A.GG•••.. Fungiacyathus marenzelleri ---GAGCAGA CACTTT--CT ...•A....G •...AAGTGA ••CGT.A••. .A.GG••.•• Pavona varians ---GAGCAGA CACTTGTCTT .GGA .•...G ....AAGTGA ..CGT.AGC. CA •.G..... Leptoseris incrustans ---GAGCAGA CACTTGTCTT .GGA .....G ....AAGTGA ..CGT.AGC. CA •.G••••• Renilla CACAAGGACC TCGCTAGGGT .CC.AT ...G .CGA.AGTGA ..CGA.AT.A ..ATCCA ... Hydra ------TT AC.A.A ..AG .C.A.AATGA ..CGT.AT.A .AAT.AAT .. 550 560 570 577 Pocillopora damicornis AAAATAACGA -AAACAAATA AAAGTTACCC TGGGGAT pocillopora meandrina ·...... -·...... ·...... ·...... Cycloseris fragilis ·...... -·...... ·...... ·...... Cycloseris vaughani ·...... -·...... ·...... ·...... Fungia scutaria ...... -·...... ·...... ·...... Zoopilus echinatus ·...... -·...... ·...... ·...... Cosc.inarea sp; ·...... -·...... ·...... ·...... Psammocora stellata ·...... -·...... ·...... ·...... Leptastrea bottae ·...... -·...... ·...... Euphyllia ancora • •••C. •• •• - ••••••••• ·...... ·...... Catalaphyllia jardinei ·. . .c ..... - ...... ·...... ·...... Turbinaria peltata 8 ••••C ••••• - ••••••••• ·...... ·...... Merulina scabricula • •••C. •• •• - ••••••••• • • 0 ••••••• ·...... Hydnophora sp. • ••• C ••••• - ••••••••• ...... ·...... Echinopora lamellosa • ••• C ••••• -·...... ·...... ·...... Caulastraea furcata · ...c. . . .. - ...... ·...... ·...... Cyphastrea ocellina ....c..... - ...... ·...... ·...... Leptoria phrygia ·. . .c. . . .. - ...... ·...... ·...... 0 Pectinia alcicornis · ...c. . . .. - ...... ·...... ·...... -...,J Lobophyllia hemprichii ·...... -·...... ·...... • Q Cit •••• Achrelia horrescens T.GT •••••• T •••••••?? ??11111111 1111111 Galaxea fascicularis T. GT. ••• •• T ••••••••• • ••••••• T• ·...... Euphyllia ancora 8 T. GT. ••• •• T ••••••••• • •••••••T• ·...... Acropora cytherea T .GT •••••• T ••••••••• ·...... ·...... Acropora humilis T .GT •••••• T ••••••••• ·...... ·...... Anacropora sp: T .GT •••••• T ••••••••• ·...... ·...... Montipora verrucosa T.GT •••••• TG •••••••• ·...... ·...... Montipora digitata T .GT. ••• •• T ••••••••• ·...... ·...... Turbinaria peltata T .GT •.••••• T •• G•••••• ·...... ·...... Tubastraea coccinea T .GT •••••• T •• G•••••• ·...... ·...... Echinopora lamellosa 8 T .GT •••••• T •• G •••••• ·...... ·...... Goniopora stokesii T .GT •••••• TT •••••••• ·...... ·...... Goniopora sp. T. GT. ••• •• TT •••••••• ·...... ·...... Porites compressa T. GT. ••• •• T ••••••••• • •••••••• 0 ·...... Fungiacyathus marenzelleri T .GT •••••• T ••••••••• ·...... ·...... Pavona varians GGGT•••••• T ••••••••• ·...... ·...... Leptoseris incrustans GGGT•••••• T ••••••••• ·...... ·...... Renilla G•••••T ••• - ••G.GG ••• • •••C ••••G .A••••• Hydra T. •••••• •• TT ••TTG ••• ••• •C ••••T .A••••• Figure 2.3 Hypothesized secondary structure for (A) the short sequence of Pungia scutaria and (B) the long sequence of Tubastraea coccinea along with proposed models for (C) L. migraioria and (0) E. coli for comparison.

108

------A.Fungia scutaria, Section IV G AA AC AA T-A A-T A-T AA T-A A A T-A A-T AT-A A-T TeGT TA-T C-G C-G T T T T G-CC GG A ~ ACT TTGTAT GCTA CAC G GGTAT ACGAT GTA T C TGGT TAAA G CA AA A C CT GT-A G-CA A-T TeGA T-A CAG C-GA G G A-T CT-A A-T A-T A-T GeT TGeT TT-G GeT GeTC G-C T-A 5' T AT-AT GAT T C T TAG A AGeTA A TeG A G-C A A-T A T-A A G-C ATACTTAATGAAATTGAATTT-ATTA 3' TO SECT. V

109 A. Fungia scutaria, Section V TGAAAA CC AT TT T-A T-A C-G GeT Area 1. A-T G-C GeT T-A A-TA CA C-GT C-GA A GTC-G A '1'0 SECT. IV A AG G A 5 1 TA TTGTA AcGA C~TTTTGT 3 1 GAGAGTTTGC- GAAAAT AACAA AGCA AG CGGGGT A-T AAA C A-T A AT-AC T GT A C AA TG CA .T C C ATG T-AG A T G T A TG T-A TG T C G AT T-A ATT G G AAT T-A AA A G GT A-T AA C AATAG AT ~ AC-GA AG G-CA G TA A-T TTC G AT TG A GG GTT A-T T-A T-A T-A T-A Area 2. TT-AG ~-C A-T AA-TC GA TC CT 110 B. Tubastraea coccinea, Section IV

AA AGAC T A A A TT TT CT-AA AA GA-TA A A C-G T-G C-G TC-GGTTGGTAGAG-C C CT GAC GTTGTA GCTACA GTA A GCGATGT C TGGT T CGAA C ACT A-TA G-C G-C TeG T-A C-GA ~-A C-GG GG-TA A-T CC-G G-C GG-CT A-T T-A G-C T-A T-A GeT G-C TT GeT TAA 5' G T C T T A AG A AG_~ G GeT G G-C A A-T A T-A A G-C TTTCCAGTGAAATTGAATTT-ATCC 3' TO SECT. V

111 B. Tubastraea coccinea, Section V TGG TTAAA G·T CT AT'G Box 1 GT A"T GA A-T T-A T-A T CT GeT «TTT-A C....GC TA-TT «T C~G GA G TT-A TA A~~ G CA-T Area 1 A T~TT A-T GGG-C 'AG A-T A-A A G-C T CT-AC A G"T A. TG-CT C-G T-A T T-A G-C C-G AT-A AA-TC TA-T » T-A T t Box 1 -r-rA AG-CA T-A » 0 3 ' AA-TC AC-GA AT GA-TA GT GTeGA AT A-T G CC-G T C-G T AGAAGTC-G T CA AG G ATTGTTAGA FACf1''F7' G CAATAAGC-GAAAATAAACGAATAGCAATTG G CGGG T A-T A AGT A-T ATTTAGGT G AAAT • III • fll' "CC AT-AC ATTG CCAGT TTTG G Gdl'GC A Cc GAA GT G-C CG" C ./ GC T GeT G :r C G A-T A T G G Area 2. T-A iTA A AG T-A G A CG TeG C A AA A T-A A A ATCCTG GC A~GCACTT~CTT-A AT GA AT I ,. I 1111 I·T I AGGG G GAACCGTCAGTCTA-T T-A GeT A-T TTT

112 C. Locusta migratoria, Section IV

A TAA A T AT TT T-A GT AT A-T CA TGeTT AC_GA TC-GT T-G T-A C-G T-A GeT AC T AAAAACATGTC-GTCTAACCT C GCGGTATTATG G C T TATTTGT C AGGTTGG ACG ATGGTAA G CG-C TA AT AC C-G A-TA C-GA G-C T-A AT-A T-A ~eT T-A GeT A-T T-A A-T C-G A-T A-T AT-A GeT TA C GeT A-T G-C C-G GeT G-T T T G-C TAA 5' T C T T A A T A A T A T TTTTAGAATTTAACTTTTAAGTTA 3' TO SEC'r.V

113

----- _.- ... _..._.._.. _------_ .. -_. -_. C. Locusta migratoria, Section V "-':: , T ATTAT / ) T-A TT AA // A-T TT-AT // T-GTTAA Area 1. T-G TT A-T // T-ATTTT GGG A-T // A-T T. G T-A // C-G GT'AGT T-A // A-TTTT.... ~C TAAT A-T // T TGAAGAA AT "II I' A-T // T-ATT ACTTCTCAA TTA A-T // T-A TA TTATAAATCATTAATTTATGT T ~,-,,'\ "". ,,'\'- \,' ~-\ / / / C-G GAATAT TAGTGTTTAA TACC T \. ~./ G-T TAG T 5' TO SECT. IV A-T A G-C A A-T A TA-T G C-G G C-G C A G AAG A C-GATAA T CG CA AGTT TAAATAT TTCTTGGGA CGTAATCATTTTTTAG CT ,1'1' \ '\'1""1 1\·1"1' ,111\ TATTTGAATAAGAATCCTA GCGTTAG G AAAATG AT T T A GCT A C C T AGCCAGGTCGGTTT C I" ",1.',,' A GACGG CCAGTTGAG T T AC C, T A TG TCTG-T AI C G AT A-T G.... G T G 'T T-A TA A CAT 'c T-A T T GT C A G A-T A' T G T A A-T GG A T C T-A A C C A-T 3' T T A-T T AA A T'l' C-G Area 2. T-G C-G AG-TG TT GA

114

------_._--- - _. D. Escherichia coli, Sections IV and V

115

.:.' Figure 2.4 Molecular phylogram (50% majority rule consensus of 112 most parsimonious reconstructions generated with a heuristic search using 10 random addition sequences in PAUP) of relationships among genera of 34 species of corals based on sequences from the mitochondrial 165 ribosomal gene region.

116

----- _.- ... _.... _.- ._ .._------_..._- -_. Archaeocoeniina iii Fungiina • Faviina • Caryophylliina • Meandriina II Poritiina • Dendrophylliina • 100 ROBUST CORALS 100

100

Acropora spp. 1001----«. Anacropora COMPLEX CORALS Montipora spp. Echinopora 8 Turbinaria Tubastraea Goniopora spp. Porites Fungiacyathus s-__~~ Pavona 100 Leptoseris t------Renilla ...... ------Hydra I 117 5% divergence

._------_._- - Figure 2.5 Cladogram (50% majority rule consensus of 2 most parsimonious reconstructions generated with a branch and bound search using PAUP) of relationships among 2 scleractinians, 7 cnidarians (courtesy of D. Bridge), sea urchin (Jacobs et al. 1988), and human (Anderson et al. 1981)

118

------._- --- Tubastraea 95 Scleractinia

Cyphastrea 59 I Anthozoa RenlIIa 100

Leptogorgia

Hydra

67 Liriope Hydrozoa 86

Physalia

81 Aurelia 88 Scyphozoa Pelagia I Locust

Human

Sea Urchin

119 Figure 2.6 Relationship between genetic distance and numbers of transition and transversion substitutions for all possible pairwise comparisons of (A) short sequences (slope for transitions =2.3, transversions» 1.4; ratio of slopes = 1.6) and (B) long sequences (slope for transitions = 2.9, transversions = 2.0; ratio of slopes = 1.4).

120

------_._- 50 A. Short sequences

~ 40 o ;: ::s ~- 30 .a ::s en -...o 20 CD ..t::! E ::s 10 • Transitions z o Transversions

o 2 4 6 8 10 12 14 Genetic distance (%)

50 B. Long sequences .- cen 40 ••I! _ o ;: •• ::s ;;- 30 .a ::s o '0 ...CD .c 0000 E ooS'aoo • Transitions ::s o Transversions z OOClllDQ)O 0000 0

o 2 4 6 8 10 12 14 Genetic distance (%)

121

------_._. Figure 2.7 Genetic distance in 60 bp windows shifted every 3 nucleotide positions for (A) two short sequences (Pocillopora meandrina vs. Catalaphyllia jardinei), (B) two long sequences tAcropora humilis vs, Paoona uarians), (C) a short and a long sequence (Fungia scutaria vs. Tubasiraea coccinea), and (0) two sea urchin sequences (Paracentrotus [Cantatore et ale 1989] vs. Strongylocentrotus [Jacobs et ale 1988]).

122

-- ,,---- '---- ._------0.6 A. Two short sequences 0.5 0.4 0.3 0.2 0.1 0-1------.-----.-----..----.....--- 100 200 300 400 0.6 B. Two long sequences 0.5 0.4 0.3 0.2 -?fl. 0.1 -w O-f--...... -...... ,.....----....---....----...... ,.....----....-- o 100 200 . 300 400 500 z i5 tJ) 0.6 C. A short and a long sequence -C 0.5 r------I I ~~~~--\ 0.4 I : GAP 0.3 ALIGNM~NT I 0.2 I I I 0.1 I I O~-...... -...... ,.....----....---L--..,.....:.---..&.,.....----....-- 100 200 300 400 500

0'6~ D. Two urchin sequences I ---\ 0.5 1\::, , 0.4 :: II 0.3 ... : : 02 . VPGOR:A~GN~ENTV 0.1 :: I• O-l--_-_----...... L...-.-L-_---__~_.c==. 100 200 300 400 500

NUCLEOTIDE POSITION 123

------CHAPTER 3. Evolution Among Families and Suborders of Scleradinian Corals: A New Hypothesis Based on Mitochondrial16S Ribosomal Gene Sequences

Abstract

The forces driving natural selection in the Scleractinia and the evolutionary history of the order remain poorly understood despite the great deal of knowledge about many aspects of the biology of these animals. Comparisons of molecular and morphological hypotheses for the evolution of the Scleractinia demonstrate that 1) the molecular topology presented here based on 165 ribosomal gene sequences is robust, 2) morphological characters are not a reliable tool for discerning among family and suborder relationships, and 3) this molecular topology explains the morphological data almost as well as the morphological data. Although this molecular hypothesis agrees with traditional classifications of species and genera, there is little concordance with relationships hypothesized among families and suborders. The molecular hypothesis presented here suggests that the Scleractinia are represented today by two major lineages of hard corals that diverged from each other at least 300 mya. One of these lineages, the robust corals, tends to be heavily calcified with skeletal elements that solidly fit together, and has predominantly massive or plate-like growth forms. The second lineage, the complex corals, is represented by taxa that are less heavily calcified, with skeletal elements that are loosely constructed and architecturally complex. Complex corals are predominantly branching but also have a wide variety of growth forms. Predominant mode of budding, which is thought to influence

124

------growth forms, are different in the two groups: robust corals grow primarily by intratentacular budding while complex corals grow primarily by extratentacular budding. These different architectural strategies may represent slightly different adaptive solutions to living in a space-limited coral reef environment. The evolution of sex is similar in each of the major lineages in that hermaphroditic and gonochoric species form separate groups in each lineage- a switch between these two kinds of gonad arrangement appears to have occurred at least three times. This molecular hypothesis supports some traditional hypotheses and some of the most recent changes in scleractinian taxonomy but also suggests new relationships among families to be considered. Although there are exceptions to the major trends observed in the two clades, the two clades clearly represent distinct suites of characters that are two successful experiments in the evolution of external calcium carbonate skeletons. These different architectural strategies provide for the first time a phylogenetic hypothesis based on the living animal for studies of the evolution and ecology of corals.

Introduction

The order Scleractinia is a highly successful and diverse group of marine invertebrates found throughout the world's oceans (Chapter 1). These organisms flourish in the tropics where they are the key components of the coral reef ecosystem. The coral reefs constructed by these animals have been described as the rain forests of the oceans. Besides being the primary framework builders of coral reefs, the Scleractinia also provide food and habitat for the high diversity of organisms that comprise the coral reef

125

------.-. ecosystem. Corals are relatively simple animals that secrete a calcium carbonate skeleton that takes on a complex array of shapes and forms. This skeleton is responsible for the extensive scleractinian fossil record which starts in the mid-Triassic (Chapter 1), and which has provided us with a great deal of knowledge about past marine environments. Scleractinia have also been the focus of studies of marine biogeography (Dana 1975,McCoy and Heck 1976, Rosen 1984, McManus 1985, Newman 1986, Richmond 1987, Rosen 1988a, Richmond 1990,Grigg and Hey 1992, Knowlton et al. 1993), and of global and regional climate change (refs. in Knowlton et al. 1992). As for most organisms, classification of the Scleractinia has been based on morphological characters. However, corals have been described as exhibiting "unending homoplasy' (Veron 1995) which, along with the great deal of variability in morphological characters (Wijsman-Best 1974, Veron and Pichon 1976, Brakel1977, Lang 1984, Foster 1985, Willis 1985, Veron 1995), has resulted in considerable taxonomic confusion (Randall 1976, Potts 1984). Until very recently (Veron, 1994 pers. comm.), identification of coral species has been difficult at best. By contrast, the identification of genera and their families is considered relatively straightforward (Veron 1994, pers. comm.). At higher taxonomic levels, family and suborder relationships based on morphology are tenuous (Wells 1956). The hypotheses for evolutionary relationships that have been proposed until now (Alloiteau 1952, Wells 1956, Chevalier and Beauvais 1987, Roniewicz and Morycowa 1993, Veron 1995) have been based on a combination of inferences about morphological characters and occurrence of genera and families in the fossil record. However, identification of coral species in the fossil record is considered unreliable due to lack of knowledge of intra-specific variation within fossil

126

------species and poor preservation of the morphological characters used in species identification (Veron 1995). Due to these weaknesses, Veron (1995) suggests only suborder groupings but does not propose relationships among suborders, and makes very few suggestions for relationships among families in the most recent hypothesis for relationships among the Scleractinia, (Fig. 3.1a). The forces driving natural selection in the Scleractinia and the evolutionary history of the order remain poorly understood despite the great deal of knowledge about many aspects of the biology of these animals.

Molecular analysis of scleractinian phylogeny The development of molecular techniques in the last ten years has provided a highly useful tool in phylogenetic studies (Hillis 1987, Patterson 1987, Novacek 1994), often leading to the resolution of phylogenies where other methods have failed (reviewed in Wilson et ala 1985, Avise 1986, Birley and Croft 1986). Molecular techniques provide a large number of characters, with a genetic basis, that are largely free of nonheritable variation (Hillis 1987). These techniques hold great promise for resolving evolutionary relationships among the Scleractinia for two reasons: 1) they provide genotypic characters, in the form of DNA sequences, that are in contrast to the tremendous phenotypic variability of skeletal characters and 2) they bypass a fossil record that is difficult to interpret. I have developed molecular techniques for obtaining DNA sequences from scleractinian corals (Chapter 2). In this chapter, I present the results of using these sequences to test traditional hypotheses, based on morphological characters (Fig. 3.1), for the relationships among families and suborders of reef-building scleractinian corals. A comparison of traditional and molecular hypotheses demonstrates that the molecular hypothesis presented here is

127

_-=..__0 ••• 0 _ robust relative to traditional hypotheses and provides us with a new perspective on the evolution of the order. This molecular hypothesis demands that we reconsider our thinking about the relationships of families and suborders of corals, makes it possible to consider the forces that have been driving evolution in the group and provides a framework for reevaluating relationships among families and suborders.

Comparison of traditional and molecular hypotheses

Traditional hypotheses for relationships among the Scleractinia Despite a great deal of taxonomic work on corals since the mid-19th century, Vaughan and Wells (1943) were the first to propose a hypothesis for the relationships among all families and suborders of both extinct and extant scleractinians. A modified version of this hypothesis proposed by Wells (1956, Fig. 1.1) has since been used as the framework for understanding the 240 million year evolutionary history of the order. This hypothesis is based on Wells' understanding of morphological characters used in classification of fossil and extant corals, his inferences about how these characters have evolved, and information from the fossil record. As described in detail in Chapter 1, Wells proposed five suborders and 20 families of extant Scleractinia. He considered the order polyphyletic in that the families present in the oldest known coral fossils of the mid Triassic represent two major lines. In his hypothesis, one of these lines evolved separately over the entire evolutionary history of the order and is represented by one suborder today, the Astrocoeniina. The second line gave rise to the other four extant suborders. The Fungiina, Faviina, and Caryophylliina arose from this second

128

------.-. -_. line in the Jurassic, and the Dendrophylliina were derived from the Caryophylliina in the Cretaceous. Since the work of Wells, our knowledge of coral classification has been greatly enriched, mostly due to the research of Veron and co-workers (Veron and Pichon 1976, Veron et al. 1977, Veron and Pichon 1980, Veron and Pichon 1982, Veron and Wallace 1984,Veron 1986, Veron 1993). Their research has included the study of corals in situ and extensive work on intraspecific variation. Veron (1995) has revised the Wells (1956) hypothesis (Fig. 3.1)based largely on this taxonomic work, along with reevaluations of and new findings in the fossil record by workers since Wells. In this revised hypothesis, the order is divided into seven extant suborders and 24 extant families. As described in Chapter 1, 16 of these families include reef-building members. Although Veron groups families into suborders, he does not hypothesize relationships among the suborders as did Wells (1956). Within suborders he proposes few relationships among families. His suborder groupings are largely those of Wells with a few exceptions. His suborder Archaeocoeniina, originating in the Triassic, is the equivalent of Wells' Astrocoeniina. Veron's suborder Fungiina, originating in the Jurassic, is the equivalent of Wells' with two exceptions. The family Poritidae is removed from the Fungiina to its own suborder, the Poritiina, and the genus Fungiacyathus is removed from the family Fungiidae to its own family, the Fungiacyathidae. Veron has also modified the suborder Faviina, for which he proposes a Jurassic origin. He has removed the families Oculinidae and Meandrinidae to their own suborder, the Meandriina with a Cretaceous origin. In addition the genera Anthemiphyllia and Trachyphyllia are each elevated to families of their own. Veron also hypothesizes that the

129 Merulinidae are more closely related to the Pectinidae and the Mussidae than to the Faviidae as had been hypothesized by Wells. Veron's suborder Caryophylliina, with a Jurassic origin, is the same as Wells' except that three subfamilies in the Caryophylliidae are elevated to family status. Veron's suborder Dendrophylliina, with a Jurassic origin, is the same as Wells'.

A molecular mothesis for relationships among the Scleractinia The molecular hypothesis for evolution of the Scleractinia presented here is based on DNA sequences from 14 of the 16 families of reef-building corals, representative of all seven suborders (Table 2.1). The collection and analysis of these data is described in Chapter 2. The phylogram generated from these data using parsimony analysis is presented in Fig. 3.2 A phylogram generated by neighbor-joining analysis results in the same topology. All of the major groupings in this phylogram are robust: they are supported at bootstrap values of 80% or higher. The scleractinian sequences fall into two major clades that show an average sequence divergence of 29.4%, suggesting a divergence early in the history of the Scleractinia, more than 300 mya (Table 24). Clades are supported in 100%ofbootstrap replicates, and species in each of the major clades are more closely related to each other than to species in the other clade. One clade, the robust corals (see Chapter 2) consists of three groups. One of these groups, the Pocilloporidae, is quite distinct from the others (13% genetic divergence). The other two groups, one with genera from three different families and the other with genera from five different families, are 6.9% different from each other. The second major clade, the complex corals, is composed of four lineages with an average divergence of 10%. Parsimony or neighbor-joining analysis cannot discern the branching order of these four lineages.

130

------"_. -- Agreement of traditional and molecular hJpotbeses The molecular data support traditional hypotheses for relationships within genera and families of the Scleractinia. In all five genera from which two or more species were analyzed, congeners formed a monophyletic cluster and were each others' closest relatives. In nine of the 10 families from which more than one genus was sampled, genera grouped with other genera from the same morphological family (e.g., Turbinaria and Tubastraea). Leptastrea was the only genus of the five sampled from the Faviidae that did not group with other genera from that family. This family has the greatest number of genera of all of the scleractinian families as well as the next to the highest number of species, and Leptastrea is not considered very closely related to other genera in the family (Veron 1995). It is also possible that this is a B sequence (see Chapter 2). Despite this single exception, the molecular data strongly support the traditional identifications of coral genera and families. Genera and families appear to be robust evolutionary units. Some major features of the traditional tree are supported by the molecular data. The Fungiidae and the Siderastreidae in the suborder Fungiina group together. The grouping of these two families traditionally is based on similar septal structures and the presence of synapticulae (an element contributing to the construction of the corallite wall). All of the families in the suborder Faviina group together (Merulinidae, Faviidae, Pectinidae and Mussidae). These families are grouped together morphologically due to their similar septal sb'ucture and the presence of synapticulae.

131 Di,crepancies between traditional and molecular hypotheses Many of the groupings predicted by the molecular hypothesis presented here do not agree with relationships among families suggested by Wells and Veron, and do not agree at all with relationships among suborders suggested by Wells. This molecular hypothesis supports groupings of families from different morphological suborders. In some cases, families from different morphological suborders (e.g., the Dendrophylllidae, Poritidae and Fungiacyathidae) appear to be more closely related to each other than to families from the same morphological suborder. Families representative of distantly related suborders (Poritidae and Dendrophyliidae, Acroporidae and Oculinidae, Faviidae and Caryophylllidae) group together. Families considered to be related morphologically, based on septal and corallite wall structure, are among some of the most different groups on the molecular phylogram (Agariciidae and Fungiidae, 5iderastreidae and Fungiidae, Acroporidae and Pocilloporidae). One family from the Archaeocoeniina is represented in each of the two major clades of this molecular hypothesis. These two families are grouped together in the Archaeocoeniina based on their similar septal structures, small corallites, similar polyp structure and development yet their 165 sequences differ by 30%. The four families representing the suborder Fungiina, as hypothesized by Veron, are not all closely related: two group together in one major clade and two are found in the other major clade. The Poritidae, which was moved from the Fungiina by Veron, appears to be related to two other families from the Fungiina as well as to the Dendrophylliidae, Acroporidae, and Oculinidae. Four of the five families in the Faviina, as hypothesized by Wells, are found in the robust clade but these families appear all very closely related to each other in contrast 132

------to the traditional hypothesis. The fifth family, the Oculinidae, is found in the complex clade and is a family that was moved to another suborder by Veron. The Caryophylliidae appear to be closely related to the four families in the Faviina and not at all related to the Dendrophylliidae which is found in the complex clade. This molecular hypothesis also suggests a very different perspective from traditional hypotheses on the early evolutionary history of the group. It suggests that there are two major clades of extant Scleractinia derived from a common ancestor that existed at least 300 mya, if not longer (Table 2.4). This is reminiscent of the two lineages hypothesized by Vaughan and Wells as being ancestral to the corals when they appear 240 mya, but the taxa in the two lineages in the molecular phylogram do not correspond to those in the two lineages hypothesized by Vaughan and Wells (l943). Veron's hypothesis (1995; Fig. 3.1a ) also suggests that two extant suborders ate derived from Triassic lineages but does not show any relationships between these two groups and the five other scleractinian suborders that have appeared in the fossil record after the Triassic. Paleontologists who have focused on the microstructural characteristics of scleractinian skeletons have also hypothesized that the extant taxa in the order are derived from two groups (Roniewicz and Morycowa 1993, Fig. 1.3). Again, the two lineages suggested from the microstructural analysis are different from the molecular analysis, but they also differ from the traditional analysis. One group includes representatives of the Archaeocoeniina along with genera from the Siderastreidae and the Poritidae. The second group includes members of the Faviina, Caryophylliina and Dendrophylliina as well as the Agariciidae and the genus Pungia (from the Fungiina). Only one of the four major groupings

133 of extant taxa, that of the Mussidae, Pectiniidae, Faviidae, Merulinidae, and the Fungia group, is supported by this molecular hypothesis. All of these hypotheses seem to support the existence of two scleractinian lineages in the Triassic. However, even the traditional hypotheses do not agree on how modern Scleractinia have evolved from these two lineages. This is perhaps not surprising given that these hypotheses about the relationships of scleractinian families are based on studies of fossils which are often problematic. The scleractinian skeleton is laid down as aragonite which is converted to calcite during diagenesis, a process that often alters skeletal microstructure (Oliver 1980), perhaps affecting recognition of fossil ancestors of extant corals. The nature of fossil samples has also made it difficult to take into account morphological variability in identifying fossil taxa. In addition, paleontologists tend to work with samples from limited time periods and there appears to often be a lack of communication between researchers studying samples from different time periods and with taxonomists of extant Scleractinia. It seems that all of these factors, along with those described earlier concerning morphological variability in corals, have contributed to the confusion about the evolution of families and suborders in the Scleractinia. Although the molecular hypothesis presented here supports the concept of the descent of modern Scleractinia from two lineages in the Triassic, its novel suggestions about the relationships of families seems to add to the evidence that our understanding of the evolution of modern taxa from fossil specimens needs to be reconsidered. Further comparisons of traditional and molecular hypotheses, described below, demonstrate that the molecular hypothesis presented here provides a framework for the reevaluation of these family and suborder relationships.

134

_ .•._. __ .. ._. _-0. Quantitative morphological analyses To compare the reliability of the traditional and molecular hypotheses I evaluated the ability of both hypotheses to explain morphological and molecular characters using MacClade (Maddison and Maddison 1992). I generated a morphological dataset (see Appendix A) by determining forty eight skeletal and biological characters (Table 3.1) from the taxonomic literature (Alloiteau 1952, Wells 1956, Veron and Pichon 1976, Veron et al. 1977,Veron and Pichon 1980, Cairns 1982, Veron and Pichon 1982, Veron and Wallace 1984, Veron 1986, Wells 1986, Chevalier 1987, Chevalier and Beauvais 1987, Hoeksema 1989, Veron 1993, Veron 1995) for all taxa in the molecular analysis. Morphological characters were equally weighted. Both sets of characters were mapped onto the traditional topology (Fig. 3.1b) and the molecular topology (Fig. 3.2). The number of steps (based on parsimony analysis) required to fit both sets of characters on both topologies was determined. The lengths of these trees were compared to the lengths of 10,000 random trees generated by PAUP (Swofford 1993) from each of the datasets. Within the Scleractinia, the morphological characters require 235 steps to fit on the morphological topology and 274 steps to fit the molecular topology. The molecular topology in this case is 16.6% longer than the morphological topology. Both sets of data have about the same amount of homoplasy on the morphological tree, the consistency index for the morphological data is 0.61 and for the molecular data 0.53. A random sample of 10,000 trees generated with the morphological data is significantly skewed to the left (gt =-0.523, p<0.05; Hillis and Huelsenbeck 1992). The shortest random tree is 417 steps indicating that both the traditional and molecular topologies explain the morphological data better than any random tree.

135

_..... -_...• ------When only the Scleractinia are considered, the molecular characters require 319 steps to fit the molecular topology and 554 steps to fit the traditional topology. The traditional topology is 74% longer than the molecular topology when explaining evolution of the molecular characters. There appears to be more homoplasy in the morphological dataset (consistency index = 0.45) than in the molecular dataset (consistency index = 0.77). A random sample of 10,000 trees generated with the molecular data is significantly skewed to the left (g1= -0.274, p<0.05). The shortest random tree is 1364 steps which indicates that both topologies explain the molecular data better than any random tree. Both the morphological and the molecular topologies are phylogenetically informative for both the morphological and the molecular data. However, the comparisons of the fit of morphological and molecular characters to either hypothesis demonstrate that this molecular hypothesis fits the morphological characters much better than the morphological hypothesis fits the molecular characters. While this molecular hypothesis requires only 17% more steps than the morphological hypothesis to fit morphological characters, the morphological hypothesis requires 74% more steps than this molecular hypothesis to fit the molecular characters. In addition, the consistency index for the morphological characters on the morphological topology (0.61) is lower than the consistency index for the molecular characters on the molecular topology (0.77) which suggests that the morphological characters have a greater amount of homoplasy. The majority of morphological characters have at least one character state that is found in more than one suborder and one character state that varies in at least one family. As a result of this homoplasy, there are few morphological characters

136

------that support among family and among suborder groupings in the morphological hypothesis and so it is not a robust one. In comparison, the molecular characters fit to this molecuar topology are less homoplasious and the molecular topology is very robust as evidenced by high bootstrap values and similar results based on different tree-building methods. Differences between the robustness of these morphological and molecular hypotheses must also be due to the greater number of molecular characters available (210 informative sites vs, 47 morphological characters). The strength of the morphological dataset might be improved by giving more weight to certain characters. However, few hypotheses exist for the evolution of morphological characters of corals and little is understood about the development of these characters. These comparisons of the molecular and morphological hypotheses demonstrate that 1) the molecular topology presented here based on 165 ribosomal gene sequences is robust, 2) these morphological characters are not a reliable tool for discerning among family and suborder relationships (as has been suggested by Veron [1995] and is described earlier in this chapter), and 3) this molecular topology explains the morphological data almost as well as the morphological data. It is precisely relationships among families and suborders that have long been considered problematical (Wells 1956). Despite the fact that a great deal of knowledge about scleractinian skeletal morphology has accumulated for over a century, few recent authors (Vaughan and Wells 1943, Alloiteau 1952, Wells 1956, Alloiteau 1957,Cuif 1977, Roniewicz and Morycowa 1993,Veron 1995) have made hypotheses about these relationships. The molecular data presented here provide for the first time a robust hypothesis for the evolution of among family and among suborder

137

------_.__ . __.- -- _._----_.------relationships of the Scleractinia. The hypothesis based on these data provides us with the chance to explore the forces that have been influencing the evolution of corals and to develop testable hypotheses about the evolutionary biology of these animals.

Evolution of Corals from a Molecular Perspective

Evolutionary patterns in the Scleractinia The most striking feature of the molecular phylogram is that it suggests that the Scleractinia have been evolving along two major lines throughout most of their evolutionary history. This is also suggested by the fossil record, as described above, in which two major lineages have been thought to be distinct as early as the mid-Triassic (although the traditional view of which coral families are the descendants of these two lineages is not supported by the molecular data). Do the biological attributes of the species represented in the two major clades of the molecular topology reflect their separate genetic lineages? I have attempted to address this question by looking for patterns in the biological attributes of the taxa used in our molecular analysis. Although there is a wide range of variability observed among genera and families in terms of these attributes, there are some general patterns that emerge from which we can generate hypotheses about the forces driving natural selection in the Scleractinia. The most visible difference among the taxa in the two major clades is related to their skeletal construction and growth form or architecture. Most of the taxa in the robust clade tend to be less architecturally complex than the taxa in the complex clade. These families have relatively solid, heavily calcified skeletons and they have mostly massive, incrusting or plate-like

138

_..._._- .. _._------growth forms or are free-living with one exception, the Pocilloporidae which forms more architecturally complex, branching colonies. The skeletons of the species in the complex clade, except the Oculinidae, are not so heavily calcified. The septa and corallite walls are perforate (except for the Oculinids), and the corallite architecture itself is relatively loose and complex. Corals in this clade exhibit a wider variety of skeletal forms including fingers, spires, columns, small bushes, branches, large thickets, lamellae, plates, and tables. This group includes the genus Acropora which has the greatest number of species of all scleractinian genera and exhibits the widest range of architectures of any coral species (Veron 1995). This genus has been described as an "overwhelming evolutionary success" due to small corallites that permit fine detail in skeletal development, to having two different kinds of corallites which allows for deterministic growth forms, and to having porous skeletal microstructure, allowing for maximum strength "in supporting the weight of architecturally complex structures (Veron 1995). It may be that the lack of heavy calcification among species in the complex clade has been accompanied by the ability of these species to take on a wide variety of growth forms that allow them to exploit a certain set of microzones in the coral reef environment as they tend to colonize and monopolize substrate more quickly. The more massive growth forms of the species in the robust clade are coincident with heavy calcification that has perhaps permitted these species to exploit a different set of reef microzones these taxa better survive natural catastrophes such as storms and large waves. These tendencies towards architectural complexity are not necessarily reflected in degree of colony integration (sensu Coates and Jackson 1985). Although the most architecturally complex groups demonstrate high levels

139 of colony integration (e.g., Acropora), groups that have primarily massive growth forms may also demonstrate a high degree of colony integration (e.g., Faviidae) as has been described by Coates and Oliver (1973) and demonstrate stability over time in the face of natural forces. Another apparent difference between the two clades is that the species in the robust clade grow primarily by intratentacular budding, with some exceptions (Pocilloporidae, some Faviidae, solitary corals), while all of the species in the complex clade, with the exception of the Agariciidae and the solitary Fungiacyathus, grow by extratentacular budding. Intratentacular budding is the development of two or more polyp mouths by invagination of the oral disk inside the tentacular ring and results in interconnected polyps (Wells 1956). Extratentacular budding is the development of two polyp mouths by invagination of the edge zone or coenosarc outside the tentacular ring and results in polyps with minimal connections to each other (Wells 1956). These different modes of budding largely determine the growth form of colonies (Wells 1956). This molecular hypothesis suggests that extratentacular budding is more amenable to the formation of architecturally complex skeletons than intratentacular budding. The Pocilloporidae seems to be distinct from other members of the robust clade both molecularly and biologically. They share the small polyps, complex and loose skeletal structure, and extratentacular budding seen in the complex clade. This may indicate that these traits are ancestral for both coral clades (relatively lightly calcified, architecturally complex, with extratentacular budding) but does not exclude the possibility that these traits are independently derived.

140 The two major clades of this molecular phylogram also seem to reflect trends in the evolution of sex in the Scleractinia. Patterns of sexual reproduction in the Scleractinia can be viewed in terms of gonad arrangement (hermaphroditic vs. gonochoric) and mode of reproduction (broadcast spawning vs, brooding) (Harrison and Wallace 1990, Richmond and Hunter 1990). Mode of reproduction is highly variable within families and genera (Harrison 1985), and there is no consistent pattern of mode of reproduction on the molecular phylogram (or on the morphological one). However, species with similar patterns of gonad arrangement group together on the molecular phylogram. In both major clades, hermaphroditic and gonochoric species form separate groups. In the robust clade; the Fu..ngildae and Siderastreidae have separate sexes, whereas the clade including the Merulinids, Favids, Pectinids, and Mussids includes mostly hermaphrodites. The exceptions to this pattern are the Caryophylliids which have separate sexes and Leptastrea (possibly represented by a 15 sequence) which is a hermaphrodite. In the complex clade, the Oculinidae and the Acroporidae are hermaphrodites whereas the Dendrophylliids, Poritids, and Agaricids have separate sexes. Again there is an exception in that the genus Porites has been reported to have both hermaphroditic and gonochoric members (Harrison and Wallace 1990). Gonad arrangement in Fungiacyathus is unknown. The exceptions to the patterns appear to be taxa that have variable patterns of gonad arrangement in general. Mixed (hermaphroditic and gonochoric) gonad arrangements have been observed in six species of corals (Harrison and Wallace 1990). A population of one species of Caryophyllia has been found to include both gonochoric and hermaphroditic colonies (Harrison 1985). One species of Porites has also been observed to have mixed gonad arrangements.

141 Both gonochorism and hermaphroditism have been suggested as ancestral states for the Scleractinia. Szmant (1986) suggested that hermaphroditism was ancestral because of the numerical dominance of hermaphroditic species. Harrison (1990) suggested that gonochorism is ancestral based on sperm morphology. Harrison (1985) has also hypothesized that hermaphroditic species represent the ancestral condition in families with variable sexuality. The predominant patterns of gonad arrangement in the robust clade of the molecular cladogram suggest that hermaphroditism is ancestral and that separate sexes evolved once or perhaps twice, but no pattern is apparent in the complex clade. The molecu!ar hypothesis and taxonomic classificatiQn Although for the most part the classification of coral genera into families has been straightforward, there are some genera whose affiliations have been uncertain. Although modern coral taxonomists have not considered morphological affinities among families to be strong, in some cases hypotheses have been made about these relationships. Some of these hypotheses are also based on speculations about the evolution of certain unusual characters found in few taxa. The molecular topology provides a means for evaluating some of these hypotheses. The work of Veron and coworkers over the past 20 years has resulted in the movement of genera to different families and the elevation of genera to family status. These changes are the result of their detailed studies of morphological variation in corals and are largely supported by the molecular hypothesis presented here. For example, the genus Psammocora was traditionally included in the family Thamnasteriidae, suborder Astrocoeniina by Wells (1956) because of its septal structure and tentacles being similar to

142

------~~----- _.- those of some genera in the Pocilloporidae (another family in the Astrocoeniina), Veron (1986) placed this genus in the Siderastreidae (suborder Fungiina) because of its similarities to Coscinarea. He determined that Psammocora's wall structure, septal fusion, and other skeletal characteristics are typical of genera in the Agariciidae and Siderastreidae (Veron and Pichon 1976). The molecular tree is in agreement with these changes. This suggests that wall structure along with other shared derived characteristics are perhaps more useful for establishing phylogenetic relationships than septal structure. There is no basis for raising the genus Psammocora to family status as was done by Chevalier and Beauvais (1987) for reasons not stated. Alloiteau's (1952) hypothesis that Euphyllia belonged in the same family as Acrhelia and Galaxea (without explanation) is not supported by the molecular data. Veron (1986) removed Hydnophora from the Faviidae and placed it with the Merulinidae because of similarities in branch tip structure and between extended polyps within genera in the Merulinidae. The agreement of the molcular data with this shift suggests that soft-tissue polyp structure may be useful in determining phylogeny. Although all of the genera in the Acroporidae appear to be closely related, Montipora and Anacropora have more closely related 165 sequences than either has to Acropora. Perhaps this supports the description of Anacropora as a branching Montipora (Veron 1986) although this may also be due to a lack of resolution in this dataset. Axial polyps found in Acropora and some species of Cyphastrea appear to have evolved twice in the Scleractinia as both genera are found in different major clades of the molecular phylogram. In the family Poritidae, Porites and

Goniopora have been called Ifvery distant" genera (Veron 1986), and are thought to have diverged in the Eocene. The molecular data show these to be

143

------'-_. _.- as different as some families .. and they are the most divergent genera (1.9%) within a family. These examples again suggest that shared, derived charactheristics are useful for establishing coral relationships. Although Fungiacyathus has previously been placed in the Fungiidae, most recently (Chevalier and Beauvais 1987) this genus has been elevated to family status because of its unique wall stucture among the Fungiina. This hypothesis is supported by the grouping of Fungiacyathus apart from the Fungiidae and the Siderastreidae, which are in the robust clade, and from the Agariciidae in the complex clade. It also indicates again the value of wall structure in determining scleractinian phylogeny This molecular hypothesis also somewhat supports changes in family and suborder relationships suggested in Veron's (1995) recent revision of the order. A relationship between the Pectinidae, and the Merulinidae, hypothesized from skeletal morphology, is also suggested by the molecular data. However, both of these families appear to be also closely related to the Faviidae and the Caryophylliidae. All four of these families are more closely related to each other than to the Mussidae, a family that Veron hypothesized to be closely related to the Pectinidae and Merulinidae. Wells' (1956) former, grouping of the Merulinidae with the Faviidae based on septal structure is supported by this molecular hypothesis. The Poritidae were moved to a suborder by themselves because of their unique 'microstructural characteristics. This change is supported by the molecular phylogram presented here which groups this family by themselves, although related to the Dendrophylliidae and the Fungiacyathidae. This suggests that microstructural data is a useful family level characteristic. Veron has also placed the Oculinidae in a new suborder, the Meandriina, suggesting this

144 family is not closely related to families in the Faviina. A distant relationship between the Oculinidae and other families, is also suggested by the molecular data. The derivation of the Dendrophylliidae from the Caryophylllidae based on modifications of septal structure and the corallite wall, as suggested by Wells (1956), is not supported by the molecular data. Summary of evolution in the Scleractinia The perspective of the evolution of the Scleractinia based on molecular data is an entirely new one. This molecular hypothesis suggests that the Scleractinia may have evolved from a skeletonless ancestor 300 mya. If this is the case, the scleractinian skeletal template has evolved twice and is today represented by two major lineages of hard corals. One of these lineages tends to be more robust yet less architecturally complex while the other lineage is represented by taxa with less robust skeletons that are more architecturally complex. Although these characteristics hold for most of -the taxa in each clade, they can only be regarded as general trends because in every case (at the family and generic levels) there are exceptions. It is possible that these somewhat subtle differences have led to the maximum exploitation of the wide variety of microzones found on a coral reef and so to the great diversity of corals that exist today. The majority of hypotheses for the evolutionary history of the Scleractinia have all been based on an evolutionary analysis of macro- and micro-structural characters of the animal skeleton with only a limited consideration of the living animal itself (mostly due to the difficulties of studying the living animal either in the lab or in situ). These molecular data give us a new way to generate hypotheses about the evolutionary biology of the animals themselves.

145

------_.. __.- _._------~---~.-. -_. Table 3.1. Morphological characters and their states for Scleractinia sampled, determined from the literature (see text) 1. Trabeculae number 9. Lateral septal ornamentation 0: none 0: none 1: few 1: granulated folds 2: numerous 2: pinnules 3: fultures 2. Trabecular structure 4: spinose 0: none 1: trabecular spines 10. Septal arrangement 2: fenestrate, porous 0: none 3: laminar, compact 1: octoradial 4: retiform, amplexoide 2: hexaradial 3: altered hexaradial 3. Trabecular composition 4: hexaradial with fusion 0: none 5: pourtales plan 1: simple 6: thamnasteroid 2: compound 11 . Septal wall 4. Trabeculae arrangement 0: absent 0: absent 1: synapticulothecal 1: 1 divergent fan system 2: septotheca 2: .2 divergent fan systems 3: many divergent fan systems 12. Non-septal wall 0: absent 5. Septal margins 1: trabecular 0: none 2: parathecalperitheca 1: smooth, acute 2: dentate, lacerate 13. Endotheca 3: minute 0: none-rare 4: granular, beadlike 1: dissepimental, vesicular 5: lobulate, acute 2: dissepimental, cellular 3: tabular 6. Septa 4: both 0: none 1: not exsert 14. Coenosteum 2: exsert 0: poorly developed to none 3: poorly developed 1: dissepimental 2: spinulose 7. Simple synapticulae 3: tabulo-columnar 0: absent 4: reticulated 1: present 5: compact

8. Compound synapticulae 0: absent 1: present

146 15. Coenosteum surface 24. Corallite size 0: none 0: none 1: spinose 1: 1-3 mm 2: granulate. echinulated 2: >2mm 3: vesicular 3: poorly delimited 4: costate 5: reticulated 25. Corallite profile 6: spongy 0: none 1: not immersed. not exsert 16. Pali 2: immersed 0: absent 3: exsert 1: present 4: highly exsert

17. Paliform lobes 26. Axial polyp 0: absent 0: absent 1: present 1: present

18. Trabecular columella 27. Verrucae 0: absent 0: absent 1: parietal 1: present 2: spongy 28. Stomodaeal ridge 19. Fascicular columella 0: absent 0: absent 1: present 1: present. styliform 29. Polyp development 20. Lamellar columella 0: 2 cycles or less 0: absent 1: >2 cycles 1: present 30. Budding pattern 21. Costae 0: solitary-none 0: absent 1: extratentacular 1: poorly developed 2: poly-circumoral 2: well-developed 3: poly-intramural 3: pseudo 4: poly-intramural-circumoral 5: distomodaeal 22. Confluent costae 6: tristomodaeal 0: absent 7: poly-circummural 1: present 31. Autotomy 23. Costal ornamentation 0: absent 0: absent 1: present 1: toothed 2: granulated 32. Transverse fission 0: absent 1: present 147 33. Colony division 39. Attached 0: absent 0: attached 1: present 1: free-living 2: polyp bail out 2: both

34. Gonad arrangement 40. Collines 0: hermaphroditic 0: absent 1: gonochoristic 1: present

35. Sexual mode 41. Stomodaea 0: spawns 0: absent 1: planulates 1: present

36. Colony formation 42. Mesoglea 0: solitary 0: reduced 1: hydnophoroid 1: developed 2: dendroid 3: phaceloid 43. Symmetry 4: plocoid 0: bilateral 5: cerioid 1: radial 6: flabellate 7: meandroid 44. Tentacle number 8: reptoid, thamnasteroid 0: hexamerous 1: octamerous 37. Corallum shape 2: polymerous 0: none 1: discoidal 45. Mesenterial nematocysts 2: incrusting 0: absent 3: massive 1: present 4: ramose 5: foliaceous, reptoid 46. Entodermal gonads 6: laminar 0: present. 1: absent 38. Zooxanthellate 0: no 47. Oral end 1: yes 0: disk 1: hydranth

48. Tentacles 0: pinnate 1: simple

148

------._. Figure 3.1A. The evolutionary hypothesis for genera and suborder relationships among the Scleractinia as hypothesized by Veron (1995).

149

-- -. -_. ------OSI C> ~ "U 0 ~ ~ IDiii" m CD= 3:_. "t1 IlJ aI VJ o ~ VJ c: 8 2 § = ('). VJ iliiili ~ U'I sq 81 I~I ~ '3 "~hylida ! I! - ~ASTROCOENIIOAE --- D"mim~Ariidaf POCILLOPOAIDAE ACROPORIDAE

'. ;...... "... :. SIDERASTREIDAE

3:::f:!_••~ AGARICIIDAE

"c: ae ~. iii~~:::~~~~~~i~rFUNGIACYATHIDAEM I CRAB AC II DAE 5' ~ ID .. '._ FUNGIIDAE

~ANTHEMIPHYWIDAE RHIZANGIIDAE PECTINIDAE MUSSIDAE MERULINIDAE

FAVIIDAE

II ~ TRACHVPHYLLIIDAE ~··i MontlMaltiidae

C> III -

-+~""----4-"".-~ FLABELLIDAE

,00 . ~ \ ~ OCULINIDAE De. '?;Q,,~ ~ '- MEANDRINIDAE 11cr'O,o.~~~. ::-rID~:;_iiiiiiiiii;:t:=~=~~ r.<>Rm DAE ~ci ~ii;~.~i:Jarae "OIillllllr.--.... DENDROPHYLIIDAElidae Figure 3.1B. Relationships among scleractinian genera, based on Veron (1995), sampled in this study. Colors correspond to morphological suborders, the names of which are listed in the legend.

151

------_._-- ._- --.I Pocillopora spp.

Archaeocoenii na I Montipora spp. ~ -. Anacropora

I I Acropora spp.

Cyc/oseris spp. rP Fungia Zoopilus Fungiacyathidae Fungiina I Pavona I Leptoseris

I Psammocora "l Coscinarea Pectinia - Lobophyllia I Hydnophora B Faviina Merulina Leptoria Echinopora Caulastrea Cyphastrea Leptastrea

Caryophylliina I Euphyllia I Catalaphyllia

Meandriina I Acrhelia ~ Gaiaxea

Poritiina Porites 4 Goniopora spp. Turbinaria Dendrophylliina I I Tubastraea 152

------._------~~~~- Figure 3.2 Molecular phylogram (50% majority rule consensus of 112 most parsimonious reconstructions with a heuristic search using PAUP) based on the mitochondrial 16s ribosomal gene region from 34 species of corals. Numbers on branches represent values from 100 bootstrap replicates. All branches shown are supported by 70% or greater of bootstrap replicates. Full species names are listed in Table 2.1. Colors correspond to morphological suborders. A. Relationships among genera. B. Relationships among families.

153

------A.

Archaeocoeniina III Cycloseris spp, Fungiina • 60 Fungia Faviina Zoopilus • Coscinaraea Caryophylliina .. Psammocora Meandriina II Leptastrea Poritiina Euphyllia • Catalaphyllia Dendrophylliina .. 100 Turbinaria B ROBUST Merulina CORALS Hydnophora 100 Echinopora Caulastrea 99 Cyphastrea Leptoria Pectinia Lobophyllia L-_-I Pocillopora Spp. 100 Acrhelia Galaxea 97 Euphyllia B Acropora spp. 100&----1. Anacropora COMPLEX CORALS Montipora spp. Echinopora B Turbinaria Tubastraea Goniopora spp. Porites Fungiacyathus L- ~ Pavona 100 Leptoseris 1------Renilla 11...------Hydra 154 5% divergence B.

Fungiidae Archaeocoeniina • Fungiina • Siderastreidae Faviina • Faviidae Caryophylliina .. Caryophylliidae Meandriina iii ROBUST Turbinaria B Poritiina • CORALS 100 Dendrophylliina • 10 Merulinidae Faviidae

100

Oculinidae 97 Euphyllia B

10011---_~ Acroporidae COMPLEX CORALS Echinopora B Dendrophylliidae

Poritidae

Fungiacyathidae ~---l""'O~O Agariciidae

1------Renil/a .....------Hydra

155 5% divergence Conclusions The work described herein shows thatphylogenetic analysis of165 ribosomal DNA sequences provides a new tool for evaluating the evolutionary history of corals. These molecular data have provided for the first time a robust framework useful for testing hypotheses about the evolution and ecology of corals. Mitochondrial16S ribosomal sequences from 14 families of Scleractinia fall into two clades that are 29.4% different from each other. Sequences from both clades appear tobe evolving under similar evolutionary constraints: they do not differ in nucleotide composition, numbers of transition and transversion substitutions, spatial patterns of substitutions, or in rates of divergence. The characteristics and patterns observed in these sequences are similar to those observed in mitochondrial ribosomal16S DNA sequences from other taxa as are their hypothesized secondary structures. The phylogram generated from these sequences based on parsimony analysis is a robustone. Comparisons of this molecular hypothesis with traditional hypotheses based on morphology demonstrate that itis also robust relative to traditional hypotheses and that morphological characters are poor tools for evaluating higher level taxonomic relationships among corals. The phylogenetic hypothesis presented here supports traditional hypotheses for the relationships among species and genera. Itdoes not supporttraditional hypotheses for relationships among families and among suborders. These molecular data suggest that two major lineages of scleractinians have existed for at least 300 million years. These lineages represent two different patterns ofgrowth in corals. One lineage is represented by taxa that are relatively robust, relatively heavily calcified and form mostly massive or plate-like

156

------._... - colonies. These corals tend to grow byintratentacular budding which is thought to influence growth form. The second lineage is represented by taxa that are less robust butmore architecturally complex. Corals from this second lineage exhibit a wide range of growth forms and the architecture ofindividual corallites is relatively loose and complex. They grow by extratentacular budding for the most part which suggests that this mode ofbudding may playa role in the formation of these architecturally complex skeletons. These differences ingrowth strategies may have enabled scleractinians to maximally exploit the microzones of the coral reef environment. Patterns of sexuality in these two lineages are similar. Each major clade is divided into a group of taxa that is mostly hermaphroditic and a group that is mostly gonochoric. These patterns should serve as hypotheses to be tested in further exploring the forces that have played a role in the evolution of the wide variety of shapes that corals take on.

This molecular hypothesis does lend support to some relationships suggested by morphology. The examples of this seem to occur where shared derived characters have been used in classification. The paucity of these examples is another indication of our lack of understanding of the evolution of the coral skeleton. A detailed analysis ofscleractinian morphological characters using the molecular hypothesis as a framework may contribute to a more complete understanding of the great deal of variability observed in the coral skeleton and so to a better understanding of how these animals have evolved.

157

------._. APPENDIX Morphological dataset using character states listed in table 3.1 for scleractinia used in this molecular study Character: trabeculae number trabecular structure trabecular composition P. damicomis few retiform-amplexoide simple P. meandrina few retiform-amplexoide simple S.hystrix few trabecular spines, acanthines simple S. pistillata few trabecular spines, acanthines simple A. cytherea few laminar, compact simple A. humilis few laminar, compact simple M. verrucosa few trabec-ilar spines, acanthines simple M. digitata few trabecular spines, acanthines simple Anacropora spp. few laminar, compact simple P. compressa numerous fenestrate, porous simple Goniopora sp, numerous fenestrate, porous simple G. siokesii numerous fenestrate, porous simple P. varians numerous fenestrate, porous simple L. incrustans numerous fenestrate, porous simple P. stellata few fenestrate, porous simple Coscinarea sp. numerous fenestrate, porous compound C.fragilis numerous fenestrate, porous compound C. vaughani numerous fenestrate, porous compound z. echinatus numerous fenestrate, porous compound F. scuiaria numerous fenestrate, porous compound F. manzelleri few fenestrate, porous compound A. horrescens numerous laminar, compact simple G. [ascicularis numerous laminar, compact simple P. aleicornis numerous laminar, compact compound L. hemprichii numerous laminar, compact simple C.ocellina numerous laminar, compact simple L. bottae numerous laminar, compact simple C·furcata numerous laminar, compact simple L. phrygia numerous laminar, compact simple E. lamellosa numerous laminar, compact simple M. scabricula numerous laminar, compact compound H. rigida numerous laminar, compact simple C.jardinei numerous laminar, compact compound E.aneora numerous laminar, compact simple T. peltata numerous laminar, compact simple T. coccinea numerous laminar, compact simple Renilla none none none Hydra none none none 158 Character: trabeculae arrangement septal margins P. damicomis absent smooth-acute P. meandrina absent smooth-acute S. hystrix absent dentate-lacerate S. pistillata absent smooth-acute A. cytherea absent smooth-acute A. humilis absent smooth-acute M.verrucasa absent smooth-acute M. digitata absent smooth-acute Anacropora spp. absent dentate-lacerate P. compressa divergent minute&granular-beadlike Goniopora sp, divergent dentate-Iacerate&minute G. stokesii divergent dentate-Iacerate&minute P. uarians divergent granular-beadlike L. incrustans divergent granular-beadlike P. stellata divergent granular-beadlike Caseinarea sp. divergent granular-beadlike C·fragilis divergent dentate-lacerate C. vaughani divergent dentate-lacerate Z. eehinatus divergent dentate-Iacerate&lobulate-acute F. scuiaria divergent dentate-lacerate F. manzelleri divergent lobulate-acute A. horrescens divergent smooth-acute G. faseicularis divergent smooth-acute&granular-beadlike P. alcicomis divergent smooth-acute&denlate-Iacerate L. hemprichii many divergent systems dentate-lacerate&lobulate-acute C. ocellina vergent&2 divergent systei granular-beadlike L. bottae vergent&2 divergent syster dentate-Iacerate&minute C·furcata 2 divergent systems dentate-lacerate L.phrygia 2 divergent systems dentate-Iacerate&minute E. lamellosa 2 divergent systems minute M. scabricula divergent :ientate-Iacerate&granular-beadlike H. rigida 2 divergent systems minute C.jardinei divergent dentate-Iacerate&minute E. ancora divergent minute T.peltata divergent smooth-acute T. coeeinea divergent smooth-acute Renilla absent none Hydra absent none

159 Character: septa simple synapticulae compound syapticulae P. damicomis poorly developed absent absent P. meandrina poorly developed absent absent S.hystrix poorly developed absent absent S. pistillata not exsert absent absent A. cytherea not exsert absent absent A. humilis notexsert absent absent M.verrucosa not exsert absent absent M. digitata notexsert absent absent Anacropora spp. notexsert absent absent P. compressa notexsert present absent Goniopora sp. not exsert present absent G. stokesii notexsert present absent Pi trarians not exsert present present L. incrustans not exsert present present P. stellata not exsert present present Caseinarea sp. notexsert present present C·fragilis exsert absent present C.vaughani exsert absent present z.echinatus exsert absent present F. scutaria exsert absent present F. manzelleri notexsert absent present A. horrescens exsert absent absent G. [ascicularis exsert absent absent P. aleicornis notexsert absent absent L. hemprichii exsert absent absent Ci ocellina exsert absent absent L. bottae exsert absent absent C·fureata exsert absent absent L.phrygia exsert absent absent E. lamellosa exsert absent absent M. scabricula . exsert absent absent H. rigida notexsert absent absent C.jardinei exsert absent absent E. ancora exsert absent absent T.peltata notexsert present absent T. coccinea notexsert present absent Rettilla none absent absent Hydra none absent absent

160 Character: lateral septal ornamentation septal arrangement P. damicomis none hexaradial P. meandrina none hexaradial S.hystrix none hexaradial S. pistillata granulated folds hexaradial A. cytherea ? hexaradial A.humilis ? hexaradial M. verrucosa none hexaradial M. digitata none hexaradial Anacropora spp. none hexaradial P. compressa none hexaradial with fusion Goniopora sp, none hexaradial with fusion G. stokesii none hexaradial with fusion P. uarians granulated folds&pinnules thamnasteroid L. incrustans pinnules thamnasteroid P. stellata granulated folds&pinnules hexaradial with fusion Coseinarea sp. granulated folds&pinnules hexaradial with fusion C·fragilis granulated folds&fultures hexaradial C.vaughani granulated folds&fultures hexaradial Z. echinatus granulated folds&fultures hexaradial F. scutaria granulated folds&fultures hexaradial with fusion F. manzelleri granulated folds hexaradial with fusion A. horrescens granulated folds hexaradial G. faseicularis granulated folds hexaradial P. alcicomis granulated folds hexaradial L. hemprichii granulated folds altered hexaradial C. ocellina granulated folds hexaradial L. bottae granulated folds hexaradial C·furcata granulated folds hexaradial L.phrygia granulated folds hexaradial E. lamellosa granulated folds hexaradial M. scabricula granulated folds hexaradial with fusion H.rigida spinose altered hexaradial C.jardinei granulated folds hexaradial E. aneora granulated folds hexaradial T.peltata granulated folds pourtales plan T. eoecinea ? pourtales plan Renilla none none Hydra none none

161

••..•__·c ~~== Character: septal wall non-septal wall endotheca P. damicomis absent absent tabular P. meandrina absent absent tabular S. hystrix absent absent tabular S. pistillata absent absent tabular A. cytherea synapticulothecal absent none-rare A.humilis synapticulothecal absent none-rare M. uerrucosa synapticulothecal absent none-rare M. digitata synapticulothecal absent none-rare Anacropora spp. synapticulothecal absent none-rare P. compressa synapticulothecal absent tabular Goniopora sp. synapticulothecal absent tabular G. stokesii synapticulothecal absent tabular P.varians absent&synapticulothecal absent none-rare L. incrustans absent&synapticulothecal absent none-rare P. stellata synapticulothecal absent none-rare Coseinarea sp. synapticulothecal absent none-rare C·fragilis synapticulothecal absent none-rare C. vaughani synapticulothecal absent none-rare Z. echinatus synapticulothecal absent none-rare F. scutaria synapticulothecal absent none-rare F. manzelleri septotheca absent none-rare A. horrescens septotheca absent none-rare G. fascicularis septotheca absent none-rare P. alcicomis absent trabecular ssepimental-vesicu L. hemprichii septotheca absent ssepimental-vesicu C.ocellina septotheca absent both L. bottae septotheca absent issepimental-celluh C·furcata septotheca absent both L.phrygia septotheca absent ssepimental-vesicu E. lamellosa septotheca absent none-rare M. scabricula absent absent none-rare H. rigida septotheca absent ssepimental-vesicu C.jardinei absent absent ssepimental-vesicu E. ancora septotheca absent ssepimental-vesicu T.peltata absent laratheca-perithec none-rare T. eoecinea synapticulothecal absent none-rare Renilla absent absent none-rare Hydra absent absent none-rare

162 Character: coenosteum coenosteum surface pall P. damicomis compact granulate-echinulated absent P. meandrina compact granulate-echinulated absent S.hystrix compact spinose absent S.pistillata compact spinose absent A. cytherea tabulo-columnaire granulate-echinulated absent A. humilis spinulose costate&reticulated absent M.verrucosa reticulated granulate-echinulated&spongy absent M. digitata reticulated granulate-echinulated&Spongy absent Anacropora spp. reticulated spinose absent P. compressa poorly developed to none none present Goniopora sp. poorly developed to none none present G. stokesii poorly developed to none none present P. uarians compact none absent L. incrusians compact none absent P. stellata poorly developed to none reticulated absent Coscinarea sp. poorly developed to none none absent C·fragilis poorly developed to none none absent C.vaughani poorly developed to none none absent z.echinatus poorly developed to none none absent F. scutaria poorly developed to none none absent F. manzelleri poorly developed to none none absent A. horrescens compact granulate-echinulated present G. fascicularis dissepimental granulate-echinulated present P. alcicornis compact none absent L. hemprichii poorly developed to none vesicular absent C.ocellina dissepimental spinose&granulate-echinulated absent L. bottae compact granulate-echinulated absent C·fuTcata poorly developed to none none absent L.phrygia poorly developed to none none absent E. lamellosa compact spinose absent M. scabricula poorly developed to none none absent H. rigida dissepimental none absent C.jardinei poorly developed to none costate present E. ancora dissepimental none absent T.peltata reticulated spongy absent T. coccinea reticulated reticulated absent Renilla poorly developed to none none absent Hydra poorly developed to none none absent

163 Character: paliform trabecular columella fascicular columella P. damicornis absent absent absent P. meandrina absent absent absent S.hystrix absent absent present-styliform S. pistillata absent absent present-styliform A. cytherea absent absent absent A. humilis absent absent absent M.verrtlcosa absent absent absent M. digitata absent absent absent Anacropora spp, absent absent absent P. compressa absent parietal&spongy absent Goniopora sp. absent parietal absent G. siokesii absent parietal absent P. varians absent absent present-stylifonn L. incrustans absent parietal absent P. stellaia absent parietal absent Coscinarea sp, absent spongy absent C·fragilis lobes spongy absent C. vaughani lobes spongy absent Z. echinaius lobes spongy -absent F. scuiaria lobes spongy absent F. manzelleri absent&lobes spongy absent A. horrescens absent absent absent G. [ascicularis absent absent absent P. alcicomis lobes spongy absent L. hemprichii absent spongy absent C.ocellina lobes spongy absent L. bottae lobes parietal absent C·furcata lobes spongy absent L.phrygia lobes absent absent E. lamellosa lobes spongy absent M.scabricula lobes parietal&spongy absent H. rigida lobes absent absent Cijardinei absent absent absent E. ancora absent absent absent T.peltata absent absent&spongy absent T. coccinea absent absent&spongy absent Renilla absent absent absent Hydra absent absent absent

164

----~-~--~.__._-_.- Character: lamellar columella costae confluent costae P. damicornis absent poorly developed absent P. meandrina absent poorly developed absent S.hystrix absent poorly developed 'absent S. pistillata absent poorly developed absent A. cytherea absent pseudo absent A. humilis absent pseudo absent M. uerrucosa absent absent absent M. digitata absent absent absent Anacropora spp. absent pseudo absent P. compressa absent poorly developed absent Goniopora sp, absent poorly developed absent G. stokesii absent poorly developed absent Pi uarians absent well-developed present L. incrustans absent well-developed present P. stellata absent poorly developed present Cosdnarea sp. absent poorly developed present C.fragilis absent well-developed absent C. uaughani absent well-developed absent Z. echinatus absent well-developed absent F. scuiaria absent well-developed absent F. manzelleri absent well-developed absent A. horrescens absent poorly developed absent G. [ascicularis absent well-developed absent P. alcicomis absent well-developed present L. hemprichii absent poorly developed absent C. ocellina absent poorly developed absent L. bottae absent poorly developed absent c.furcata absent well-developed absent L.phrygia present poorly developed present E. lamellosa absent y developed&well-deve absent M. scabricula absent absent absent H. rigida present poorly developed absent C.jardinei absent&present poorly developed absent E.ancora absent yo deveioped&well-deve absent Ti peliaia absent absent absent T. coccinea absent poorly developed absent Renilla absent absent absent Hydra absent absent absent

165 Character. costal ornamentation corallite size corallite profile P. damicomis absent >2mm immersed P. meandrina absent >2mm immersed S.hystrix granulated >2mm immersed S. pistillata granulated >2mm immersed A. cytherea ? 1-3mm highly exsert A. humilis ? 1-3mm highlyexsert M. uerrucosa absent >2mm immersed M. digitata absent >2mm immersed Anacropora spp, absent >2mm immersed&exsert P. compressa absent 1-3mm immersed Goniopora sp, absent >2mm immersed G. siokesii absent >2mm not immersed-not exsert P. uarians absent poorly delimited not immersed-not exsert L. incrustans absent poorly delimited not immersed-not exsert P. stellata absent poorly delimited not immersed-not exsert Coscinarea sp. granulated poorly delimited not immersed-not exsert C.fragilis granulated >2mm none C. vaug!.zani granulated >2mm none z.echinaius granulated >2mm none F. scutaria granulated >2mm none F. manzelleri toothed >2mm none A. horrescens granulated >2mm highly exsert G. fascicularis granulated >2mm highly exsert P. alcicomis absent poorly delimited not immersed-not exsert L. hemprichii toothed >2mm not immersed-not exsert C.ocellina toothed >2mm exsert L. bottae absent >2mm exsert C·furcata absent >2mm highlyexsert L.phrygia granulated poorly delimited not immersed-not exsert E. lamellosa granulated >2mm exsert M. scabricula absent >2mm not immersed-not exsert H. rigida absent >2mm not immersed-not exsert C.jardinei absent >2mm not immersed-not exsert E. ancora absent >2mm not immersed-not exsert T.peltata absent >2mm immersed&exsert Ti cocdnea absent >2mm highlyexsert Renilla absent none none Hydra absent none none

166

------_._. - - . Character. axial polyp verrucae stomodaeal ridge polyp development P. damicomis absent absent absent 2 cyclesor less P. meandrina absent present absent 2 cyclesor less S.hystrix absent absent absent 2 cyclesor less S. pistillata absent absent absent 2 cyclesor less A. cytherea present absent absent 2 cyclesor less A. humilis present absent absent 2 cyclesor less M. verrucosa absent absent absent 2 cyclesor less M. digitata absent absent absent 2 cyclesor less Anacropora spp, absent absent absent 2 cyclesor.less P. compressa absent absent present 2 cyclesor less Gonioporasp. absent absent present > 2 cycles G.stokesii absent absent present > 2 cycles P. varians absent absent present > 2 cycles L. incrustans absent absent present > 2 cycles P.stellata absent absent present > 2 cycles Coscinarea sp. absent absent present > 2 cycles C·fragilis absent absent present > 2 cycles C. vaughani absent absent present > 2cydes z.echinatus absent absent present > 2 cycles F. scutaria absent absent present > 2cydes F. manzelleri absent absent present > 2 cycles A. horrescens absent absent present. > 2 cycles G. [ascicularis absent absent present > 2 cycles P. alcicomis absent absent present > 2 cycles L. hemprichii absent absent present > 2 cycles Ci ocellina absent absent present > 2 cycles L. bottae absent absent present > 2 cycles C.furcata absent absent present > 2 cycles L.phrygia absent absent present > 2 cycles E. lamellosa absent absent present > 2 cycles M. scabricula absent absent present 2 cyclesor less H. rigida absent absent present > 2'cycles C.jardinei absent absent present > 2 cycles E. ancora absent absent present > 2 cycles T.peltata absent absent present > 2 cycles T. coccinea absent absent present >2 cycles Renilla absent absent present 2 cyclesor less Hydra absent absent present 2 cyclesor less

167 Character: budding pattern Autotomy Transverse fission P. damicomis extratentacular absent absent P. meandrina extratentacular absent absent S.hystrix extratentacular absent absent S. pistillata extratentacular absent absent A. cytherea extratentacular absent absent A. humilis extratentacular absent absent M.verrucosa extratentacular absent absent M. digitata extratentacular absent absent Anacroporaspp, extratentacular absent absent P. compressa extratentacular absent absent Goniopora sp. extratentacular absent absent G. stokesii extratentacular absent absent P. varians distomodaeal&tristomodaeal absent absent L. incrusians distomodaeal&tristomodaeal absent absent P. stellata . distomodaeal&tristomodaeal absent absent Coscinarea sp. distomodaeal&tristomodaeal absent absent C·fragilis solitary-none present present C. vaughani solitary-none present present Z. echinatus poly-circumoral-poly-circumoral absent present F. scutaria solitary-none absent present F. manzelleri solitary-none absent present A. horrescens extratentacular absent absent G. [ascicularis extratentacular absent absent P. alcicomis poly-intramural-circumoral absent absent L. hemprichii poly-intramural-circumoral absent absent C. ocellina extratentacular absent absent L. bottae extratentacular absent absent C.furcata distomodaeal absent absent L.phrygia poly-intramural absent absent E. lamellosa extratentacular absent absent M. scabricula poly-intramural absent absent H. rigida poly-eircummural absent absent Cijardinei poly-intramural-circumoral absent absent E. ancora ntramural&distomodaeal&tristomc absent absent T.peltata extratentacular absent absent T. coccinea extratentacular absent absent Renilla extratentacular absent absent Hydra extratentacular absent absent

168 Character: colony division gonad arrangement sexual mode P. damicornis polyp bail out hermaphroditic planulates P. meandrina absent hermaphroditic planulates S.hystrix polyp bail out hermaphroditic planulates S. pistillata absent gonochoristic planulates A. cytherea absent hermaphroditic planulates A. humilis absent hermaphroditic spawns M. verrucosa absent hermaphroditic spawns M. digitata absent hermaphroditic spawns Anacropora spp. absent hermaphroditic spawns P. compressa absent gonochoristic spawns Goniopora sp. absent gonochoristic spawns G. stokesii present gonochoristic spawns P. varians absent gonochoristic spawns L. incrustans absent ? ? P. stellata absent gonochoristic spawns Coscinarea sp. absent gonochoristic spawns C.fragilis absent gonochoristic spawns C.vaughani absent gonochoristic spawns z.echinatus absent gonochoristic spawns F. scutaria absent gonochoristic spawns F. manzelleri absent ? ? A. horrescens absent hermaphroditic spawns G. fascicularis absent hermaphroditic spawns P. alcicornis absent hermaphroditic spawns L. hemprichii absent hermaphroditic spawns C.ocellina absent hermaphroditic planulates L. bottae absent hermaphroditic planulates C·furcata absent hermaphroditic planulates L.phrygia absent hermaphroditic planulates E. lamellosa absent hermaphroditic planulates M. scabricula absent hermaphroditic planulates H. rigida absent hermaphroditic planulates C.jardinei absent gonochoristic spawns E.ancora absent gonochoristic awns&planula T.peltata absent gonochoristic spawns T. coccinea absent gonochoristic planulates Renilla absent ? ? . Hydra absent aaphroditiczegonochor planulates

169

------_. -- Character: colony formation corallum shape P. damicomis plocoid ramose P. meandrina plocoid ramose S.hystrix plocoid ramose S. pistillata plocoid ramose A. cytherea plocoid ramose A. humilis plocoid ramose M. uerrucosa plocoid massive&ramose&foliaceous-reptoid M. digitata plocoid ramose Anacropora spp. plocoid ramose P. compressa cerioid ramose Goniopora sp, cerioid massive G. stokesii cerioid massive P. tiarians cerioid incrusting&massive L. incrustans reptoid-thamnasteroid incrusting P.stellata plocoid ramose Coscinarea sp. cerioid massive C·fragilis solitary discoidal C. vaughani solitary discoidal Z. eehinatus cerioid discoidal F. scutaria solitary discoidal F. manzelleri solitary discoidal A. horrescens dendroid ramose G. fascicularis plocoid massive P. alcicornis meandroid foliaceous-reptoid L. hempriehii iceloideflabellatesemeandr massive C.ocellina plocoid incrusting L. bottae plocoid massive C·furcata phaceloid ramose L. phrygia meandroid massive E. lamellosa plocoid foliaceous-reptoid&laminar M.seabrieula meandroid foliaceous-reptoid&laminar H. rigida hydnophoroid ramose C.jardinei flabellate&meandroid foliaceous-reptoid E. ancora flabellate&meandroid foliaceous-reptoid T.peltata plocoid foliaceous-reptoid&laminar T. coccinea plocoid massive Renilla plocoid none Hydra solitary none

170

. . ._.__ 0._ Character. zooxanthellate attached collines stomodaea mesoglea P. damicomis yes attached absent present developed P. meandrina yes attached absent present developed S.hystrix yes attached absent present developed S. pistillata yes attached absent present developed A. cytherea yes attached absent present developed A. humilis yes attached absent present developed M.verrucosa yes attached absent present developed M. digitata yes attached absent present developed Anacropora spp, yes attached absent present dev.eloped P. compressa yes attached absent present developed Goniopora sp, yes attached absent present developed G. stokesii yes both absent present developed P. varians yes attached present present developed L. incrustans yes attached absent present developed P.stellata yes attached present present developed Coscinarea sp. yes attached absent present developed C·fragilis yes free-living absent present developed C. vaugh ani yes free-living absent present developed Z. echinaius yes free-living absent present developed F.scutaria yes free-living absent present developed F. manzelleri no free-living absent present developed A. horrescens yes attached absent present developed G. [ascicularis yes attached absent present developed P. alcicomis yes attached absent present developed L. hemprichii yes attached absent present developed C.ocellina yes attached absent present developed L. bottae yes attached absent present developed C.furcata yes attached absent present developed L.phrygia yes attached present present developed E. lamellosa yes attached absent present developed M.scabricula yes attached present present developed H. rigida yes attached present present developed C.jardinei yes both absent present developed E. ancora yes attached absent present developed T.peltata yes attached absent present developed T. coccinea no attached absent present developed Renilla no free-living absent pr~sent developed Hydra no free-living absent absent reduced

171 Character: symmetry tentacle # mesenterial nematocysts P. damicomis bilateral hexamerous present P. meandrina bilateral hexamerous present S.hystrix bilateral hexamerous present S. pisiillaia bilateral hexamerous present A. cytherea bilateral hexamerous present A. humilis bilateral hexamerous present M. tierrucosa bilateral hexamerous present M. digitata bilateral hexamerous present Anacropora spp, bilateral hexamerous present P. compressa bilateral hexamerous present Goniopora sp. bilateral hexamerous present G. siokesii bilateral hexamerous present P. uatians bilateral hexamerous present L. incrustans bilateral hexamerous present P. stellata bilateral hexamerous present Coscinarea sp. bilateral hexamerous present C. fragilis bilateral hexamerous present C. vaughani bilateral hexamerous present Z. echinaius bilateral hexamerous present F. scuiaria bilateral hexamerous present F. manzelleri bilateral hexamerous present A. horrescens bilateral hexamerous present G. fascicularis bilateral hexamerous present P. alcicomis bilateral hexamerous present L.hemprichii bilateral hexamerous present C.ocellina bilateral hexamerous present L. bottae bilateral hexamerous present C·furcata bilateral hexamerous present L.phrygia bilateral hexamerous present E. lamellosa bilateral hexamerous present M. scabricula bilateral hexamerous present H. rigida bilateral hexamerous present C.jardinei bilateral hexamerous present E. ancora ' bilateral hexamerous present T.peltata bilateral hexamerous present T. coccinea bilateral hexamerous present Renilla bilateral octamerous present Hydra radial polymerous absent

172 Character: entodermal gonads oral end tentacles P. damicomis present disk simple P. meandrina present disk simple S.hystrix present disk simple S.pistillata present disk simple A. cytherea present disk simple A. humilis present disk simple M. uerrucosa present disk simple M. digitata present disk simple Anacropora spp. present disk simple P. compressa present disk simple Goniopora sp. present disk simple G. stokesii present disk simple P. tiarians present disk simple L. incrustans present disk simple P. stellata present disk simple Coscinarea sp. present disk simple C. fragilis present disk simple C. tiaughani present disk simple Z. echinatus present disk simple F. scutaria present disk simple F. manzelleri present disk simple A. horrescens present disk simple G. [ascicularis present disk simple P. alcicomis present disk simple L. hemprichii present disk simple C.ocellina present disk simple L. bottae present disk simple C.furcata present disk simple L.phrygia present disk simple E. lamellosa present disk simple M. scabricula present disk simple H. rigida present disk simple C.jardinei present disk simple E. ancora present disk simple T.peltata present disk simple T. coccinea present disk simple Renilla present disk pinnate Hydra absent hydranth simple

173

------.- -.- _.... _-_._. LITERATURE CITED Adams, C. G. 1981. An outline of Tertiary palaeogeography. Eds. Greenwood, P. H. and Cocks, L. R M. Chance. change and challenge: the evolving earth. Cambridge: Brit. Mus. (Nat.Hist) and Cambridge University

Press. pp. 221-235. Alloiteau, J. 1952. Madreporaires Post-Paleozoiques. Ed. Piveteau, J. Traite de PaIeontologie. Paris: Masson et Compagnie. pp. 539-684. Alloiteau, J. 1957. Contribution ~ la systematique des madreporaire fossiles. Paris: Centre National de la Recherche Scientifique. 461 pages. Anderson,S., Bankier, A. T., Barrell, B. G., de Bruijn, M.. H. L., Collison, A. R, Drouin, J., Eperon, I. C., Nierlich, D. P., Roe, B. A., Sanger, F., Schrier, P.

H., Smith, A. J. H, Staden, R and Young, I. G. 1981. Sequence and organization of the human mitochondrial genome. Nature 290: 457 465. Attardi, G. 1985. Animal mitochondrial DNA: An extreme example of genetic economy. Int. Rev. Cyt. 93: 93-145. Avise, J. C. 1986. Mitochondrial DNA and the evolutionary genetics of higher animals. Philosophical Transactions of the Royal Society, London B

312: 325-342. Ax, P. 1989. Basic phylogenetic systemization of the Metazoa. Eds. Fernholm, B., Bremer, K. and Jomvall, H. The Hierarchy of Life. Amsterdam: Elsevier Science Publishers H.V. pp. 229-245. Ayre, D. J., Veron, J. E. N. and Dufty, S. L. 1991. The corals Acropora palifera and Acropora cuneata are genetically and ecologically distinct. Coral Reefs 10: 13-18.

174 Beauvais, L. 1973. Upper Jurassic hermatypic corals. Ed. Hallam, A. Atlas of Palaeobiogeography. Amsterdam: Elsevier. pp. 317-328. Beauvais, L. 1980. Sur la taxinomie des madreporaires mesozotques, Acta Palaeontol. Polonica 25: 345-360. Beauvais, L. 1982. Paleobiogeographie des madreporaires du Trias. Bulletin Societe Geologique du France 24: 963-970. Beauvais, L. 1984. Evolution and diversification of Jurassic Scleractinia. Palaeontogr. Amer. 54: 219-224. Beauvais, L. and Beauvais, M. 1974. Studies on the world distribution of the upp~r Cretacous corals. Proceedings of the Second International Coral Reef Symposium, Brisbane. 475-494. Beckman, K. B., Smith, M. F. and Orrego, C. 1993. Purification of mitochondrial DNA with Wizard™ Minipreps DNA purification system. Promega Notes 43: 10-13. Best, M. B., Boekschoten, G. J. and Oosterbaan, A. 1984. Species concept and ecomorph variation in living and fossil Scleractinia. Palaeontogr. Amer. 54: 70-79. Birley, A. J. and Croft, J. H. 1986. Mitochondrial DNA and phylogenetic relationships. Ed. Dutta, S. K. DNA Systematics. Florida: CRC Press. pp. 107-137. Brakel, W. H. 1977. Corallite variation in Porites and the species problem in corals. Proc, 3rd Int. Symp. Coral Reefs, Miami 2: 459-462.

Bridge, D. 1994. PhD Dissertation. Yale University. Bridge, D., Cunningham, C. W., Schierwater, B., DeSalle, R. and Buss, L. W. 1992. Class-level relationships in the phylum Cnidaria: Evidence from mitochondrial genome structure. PNAS 89: 8750-8753.

175 Briggs, J. C. 1987. Biogeography and Plate Tectonics. Amsterdam: Elsevier. 204 pp. Brusca, R C. and Brusca, G. J. 1990. Invertebrates. Sunderland, MA: Sinauer. pp. Budd, A. F. 1988. Large-scale evolutionary patterns in the reef-coral Montastraea: the role of phenotypic plasticity. Proceedings of the 6th International Coral Reef Symposium, Australia. 3:393-398. Budd, A. F. 1990. Longterm patterns of morphological variation within and among species of reef-eoraIs and their relationship to sexual reproduction. Syst. Bot. 15: 150-165. Budd, A. F. and Coates, A. G. 1992 Nonprogressive evolution in a clade of Cretaceous Montastrea-like corals. Paleobiology 18: 425-446. Budd, A. F., Stemann, T. A. and Stewart, R H 1992. Eocene Caribbean reef corals: A unique fauna from the Gatuncillo formation of Panama. J. Paleontol. 66: 570-594. Cairns, S.D. 1977. Stony Corals I. Caryophylliina and Dendrophylliina (Anthozoa: Scleractinia). Memoirs of the Hourglass Cruises 3: 1-27. Cairns, S. D. 1979. The deep-water Scleractinia of the Caribbean Sea and adjacent waters. Studies on the Fauna of Curacao and Other Caribbean Islands 180: 1-333. Cairns, S. D. 1981. Marine flora and fauna of the Northeastern United States. Scleractinia. NOAA Technical Report NMFS 438: 1-14. Cairns, S. D. 1982. Antarctic and subantarctic Scleractinia. Ed. Kornicker, 1. S. Biology of the Antarctic Seas XI. Washington, D.C.: American Geophysical Union. pp. 1-74.

176 Cairns, S. D. 1984. An application of phylogenetic analysis to the Scleractinia: Family Fungiidae. Palaeontogr. Amer. 54: 49-57. Cairns, S. D. 1989b. A Revision of the Ahermatypic Scleractinia of the Philippine Islands and Adjacent Waters, Part 1: Fungiacyathidae, Micrabaciidae, Turbinoliinae, Guyniidae, and Flabellidae. Smithsonian Contrib. Zool. No. 486: 1-136. Cairns, S. D. 1991. A Revision of the Ahermatypic Scleractinia of the Galapagos and Cocos Islands. Smithsonian Contrib. Zool. No. 504: 1-44. Cairns, S. D. and Parker, S. A. 1992. Review of the Recent Scleractinia (stony corals) of South Australia, Victoria, and Tasmania. Rec, S. Aust, Mus., Monogr. Ser. No.3: 1-82. Cairns, S. D. and Stanley, G. D. J. 1981. Ahermatypic coral banks: Living and fossil counterparts. Proceedings of the Fourth International Coral Reef Symposium, Manila. 1:611-618. Cantatore, P., Roberti, M., Rainaldi, G., Gadaleta, M. N. and Saccone, C. 1989. The complete nucleotide sequence, gene organization, and genetic code of the mitochondrial genome of Paracentrotus lioidus. J. BioI. Chem. 264:10965-10975. Cedergren, R, Gray, M. W., Abel, Y. and Sankoff, D. 1988. The evolutionary relationships among known life forms. J. Mol. Evol. 28: 98-112. Chevalier, J.-P. 1987.Ordre des scIeractiniaires. Ed. Grasse, P.-P. Traite de Zoologie. Anatomie, Systematique, Biologie. Paris: Masson. pp. 403

764.

Chevalier, J.-P. and Beauvais, L. 1987. Ordre des scIeractiniaires: Systematique. Ed. Grasse, P.-P. Traite de Zoologie. Anatomie, Systematique, Biologie. Paris: Masson. pp. 679-753.

177 Chevalier, J. P. 1961. Recherches sur les Madreporaires et les formations recifales miocenes de la Mediteranee Occidentale. Memoires de la Societe geologique de France, Nouvelle sene 40: 1-562. Chevalier, J. P. 1971. Les ScIeractiniaires de la Melanesie Francaise, Expedition francaise sur les recifs coralliens de la Nouvelle-Caledonie 5(I-re): 1-307. Chevalier, J. P. 1975a. Apercu sur la faune corallienne recifale du Neogene. Second Symposium international sur les coraux et recifs coralliens fossiles, 359-366. Chevalier, J. P. 1975b. Les ScIeractiniaires de la Melanesie Francaise. Expedition francaise sur les redfs coralliens de la Nouvelle-Caledonia 7(IT-me): 1-407. Christen, R, Ratto, A., Baroin, A., Perasso, R, Grell, K G. and Adoutte, A. 1991. An analysis of the origin of metazoans, using comparisons of partial sequences of the 28s RNA, reveals an early emergence of triploblasts. EMBO 10: 499-503. Clary, D. O. and Wolstenholme, D. R. 1985. The ribosomal RNA genes of Drosophila mitochondrial DNA. Nuc. Acids Res. 13: 4029-4045. Coates, A. C. and Jackson, J. B. C. 1987. Clonal growth, algal symbiosis, and reef formation by corals. Paleobiology 13: 363-378. Coates, A. G. 1973. Cretaceous Tethyan coral-rudist biogeography related to the evolution of the Atlantic Ocean. Ed. Hughes, N. F. Organisms and Continents Through Time. pp.169-174. Coates, A G. 1977. Jamaican Coral-Rudist frameworks and their geologic setting. Eels. Frost, S. H., Weiss, M. P. and Saunders, J. B. Reefs and Related. Carbonates- Ecology and Sedimentology. Tulsa: The American Association of Petroleum Geologists. pp. 83-91.

178

>----_.__._. _.....- .__ .- Coates, A. G. and Jackson, J. B. C. 1985. Morphological themes in the evolution of donal and adona! marine invertebrates. Eels. Jackson, J. B. C., Buss, L W. and Cook, R. E. Population Biology and Evolution of Clonal Organisms. New Haven: Yale University Press. pp. 67-106. Coates, A. G. and Oliver, W. A. J. 1973. Coloniality in zoantharian corals. Eds. Boardman, R. S., Cheetham, A. H. and Oliver, W. A. J. Animal Colonies: Development and Function Through Time. Stroudsburg: Dowden, Hutchinson, and Ross, Inc. pp. 3-28. Coudray, J. and Montaggioni, L 1982. Coraux et recifs coralliens de la province indo-pacifique: repartition geographique et altitudinale en relation avec la tectonique globale. Bull. Soc. geol. France 24: 981-993. Cowen, R. 1983. Algal symbiosis and its recognition in the fossil record. Eds. Tevesz, M. J. S. and McCall, P. L. Biotic Interactions in Recent and Fossil Benthic Communities. New York: Plenum Press. pp. 431-478. Cowen, R. 1988. The role of algal symbiosis in reefs through time. Palaios 3: 221-227. Croizat, L 1981. Biogeography: Past, Present, and Future. Eels. Nelson, G. and Rosen, D. E. Vicariance Biogeography: A critique. New York: Columbia University Press. pp. 501-523. Cuif, J.-P. 1977. Arguments pour une relation phyletique entre les madreporaires paleozoiques et ceux du trias: Implications systematiques de l'analyse microstructurale des Madreporaires triasiques. Memoires de la Societe Geologique de France, Nouvelle sene 129: 1-54.

179 Cunningham, C. W. and Buss, L. W. 1993. Molecular evidence for multiple episodes of paedomorphosis in the family Hydractiniidae. Biochemical Systematics and Ecology 21: 57-69. Dana, T. F. 1975. Development of contemporary Eastern Pacific coral reefs. Mar. BioI. 33: 355-374. Derr, J. N., Davis, S. K.,. Woolley, J. B. and Wharton, R. A. 1992 Variation and the phylogenetic utility of the large ribosomal subunit of mitochondrial DNA from the insect order Hymenoptera. Molec. Phylog.Evol. 1:1~147. DeSalle, R. 1992. The phylogenetic relationships of flies in the family Drosophilidae deduced from mtDNA sequences. Molec. Phylog. Evol. 1: 31-40. DeSalle, R., Freedman, T., Prager, E. M. and Wilson, A. C. 1987. Tempo and mode of sequence evolution in mitochondrial DNA of Hawaiian Drosophila. J. Mol. Evol. 26: 157-164. Dixon, M. T. and Hillis, D. M. 1993. Ribosomal RNA secondary structure: Compensatory mutations and implications for phylogenetic analysis. Mol. BioI. Evol. 10: 256-267. Durham, J. W. 1966. Coelenterates, especially stony corals, from the Galapagos and Cocos Islands. Ed. Bowman, R. I. The Galapagos. Berkeley: University of California Press. pp. 123-135. Ekman, S. 1953. Zoogeography of the Sea. London: Sidgwick and Jackson, Ltd. 417pp. Erwin, D. H. 1991. Metazoan phylogeny and the Cambrian radiation. TREE 6: 5-8.

180

----- _._ _ - .- Fang, Q., IV, W. C. B., Blocker, H. D. and Whitcomb, R. F. 1993. A phylogeny of new world Deltocephalus-like leafhopper genera based on mitochondrial 165 ribosomal DNA sequences. Molec. Phylog. Evol. 2: 119-131. Felsenstein, J. 1993. Phylogeny inference package (PHYLIP), version 5.3c. Seattle, Dept. of Genetics, University of Washington. Foster, A. B. 1984. The species concept in fossil hermatypic corals: A statistical approach. Palaeontogr. Amer. 54: 58-69. Foster, A. B. 1985. Variation within coral colonies and its importance for interpreting fossil species. J. Paleontol. 59: 1359-138l. Frost, S. H. 1972. Evolution of Cenozoic Caribbean coral faunas. Carib. Geol. Conf. Trans. 6: 461-464. Frost, S. H. 1977a. Oligocene reef coral biogeography Caribbean and western Tethys. Mem. Bur. Rech. geol. miti1r. 89: 342-352.· Frost, S. H. 1977b. Miocene to Holocene evolution of Caribbean province reef building corals. Proceedings of the 3rd International Coral Reef Symposium, Miami. 353-359. Frost, S. H. 1977c. Cenozoic reef systems of Caribbean- prospects for paleoecologic synthesis. Eds. Frost, S. H., Weiss, M. P. and Saunders, J. B. Reefs and Related Carbonates- Ecology and Sedimentology. Tulsa: The American Association of Petroleum Geologists. pp. 93-110. Frost, S. H. and Langenheim, R. L. 1974. Cenozoic Reef Biofacies. Dekalb: Northern Illinois University Press. 388 pp. Fukuda, M., Wakasugi, 5., Tsuzuki, T., Homiyama, H., Shimada, K. and Miyata, T. 1985. Mitochondrial DNA-like sequences in the human nuclear genome. J. Mol. BioI. 186: 257-266.

181

----- .._---- .------_._------Garthwaite, R, Potts, D., Veron, J. E. N. and Done, T. J. 1994. Electrophoretic identification of poritid species (Anthozoa: Scleractinia). Coral Reefs 13: 49-56. Gatesy, J., Yelon, D., DeSalle, R. and Vrba, E. S. 1992. Phylogeny of the Bovidae (Artiodactyla, Mammalia), based on mitochondrial ribosomal DNA sequences. Mol. BioI. EvoI. 9: 433-446. Gattuso, J., Pichon, M. and [aubert, J. 1991. Physiology and taxonomy of scleractinian corals: a case study in the genus Stylophora. Coral Reefs 9: 173-182 Gellissen, G. and Michaelis, G. 1987. Gene transfer: Mitochondria to Nucleus. Eds. Lee, J. J. and Frederick, J. F. Endocytobiology m. Annals of the New York Academy ofSciences. pp. 391-401. Genetics Computer Group. 1993. The GCG Package. Madison. Gerbi, S. A. 1985. Evolution of Ribosomal DNA. Ed. MacIntyre, R. J. Molecular Evolutionary Genetics. New York: Plenum Press. pp. 419-517. Gerth, H. 1930. The evolution of reefcorals during the Cenozoic Period. Proceedings of the 4th Pacific Science Congress, 2A:333-350. Glotz, c., Zwieb, c., Brimacombe, R., Edwards, K. and Kossel, H. 1981. Secondary structure of the large subunit ribosomal RNA from Escherichia coli, Zea mays chloroplast, and human and mouse mitochondrial ribosomes. Nuc, Acids Res. 9: 3287-3306. Glynn, P. W. 1982. Coral communities and their modifications relative to past and prospective Central American Seaways. Adv. Mar. Biol, 19: 91-132. Glynn, P. W., Stewart, H. H. and McCosker, J. E. 1972. Pacific coral reefs of Panama: Structure, distribution and predators. GeoI. Rundschau 61: 481-519.

182

------Glynn, P. W. and Wellington, G. M. 1983. Galapagos coral communities: Relationships and regional comparisons. Ed. G.M.Wellington, P. W. G. a. Corals and Coral Reefs of the Galapagos Islands. Berkeley: Univ. of Calif. Press. pp. 161-209. Grasshoff, M. 1984. Cnidarian phylogeny- A biomechanical approach. Palaeontogr. Amer. 54: 127-135. Gravel, M., Leclerc, D., Melancon, P. and Brakier-Gingras, L. 1989. The conserved 900 stem/loop region in Escherichia coli 165 ribosomal RNA is not required for protein synthesis. Nuc. Acids Res. 17: 2723-2732. Grigg, R. W. and Hey, R. 1992 Paleoceanography of the Tropical Eastern '" Pacific Ocean. Science 255: 172-178. Gutell, R. R., Larsen, N. and Woese, C. R. 1994. Lessons from an evolving rRNA: 165 and 235 rRNA structures from a comparative perspective. Micro. Rev. 58: 10-26. Gutell, R. R., 5chnare, M. N. and Gray, M. W. 1992. A compilation of large subunit (235 and 235-like) ribosomal RNA structures. Nuc, Acids Res. 20, Supplement: 2095-2109. Guzman, H. M. and Cortes, J. 1993. Arrecifes coralinos del Pacifico Oriental tropical: Revision y perspectivas. Rev. BioI. Trop. 41: 535-557. Hand, C. 1966. On the evolution of the Actiniaria. Ed. Rees, W. J. The Cnidaria and Their Evolution. 5ymp. zoo. Soc. Lond. 16: 135-146. Harrigan, J. F. 1972. The planula larva of Pocillopora damicornis; lunar periodicity of swarming and substratum selection behavior. PhD Dissertation. Dept. of Zoology, University of Hawaii.

183

....- ._- ._._------_.- Harrison, P. L. 1985. Sexual characteristics of scleractinian corals: Systematic and evolutionary implications. Proceedings of the 5th International Coral Reef Congress 4: 337-342 Harrison, P. L., Babcock, R C., Bull, G. D., Oliver, J. K., Wallace, C. C. and Willis, B. L. 1984. Mass spawning in tropical reef corals. Science 223: 1184-1189. Harrison, P. L. and Wallace, C. C. 1990. Reproduction, dispersal and recruitment of scleractinian corals. Ed. Dubinsky, z. Coral Reefs. New York: Elsevier. pp. 133-207. Heck, K. 1. J. and McCoy, E. D. 1978. Long-distance dispersal and the reef building corals of the Eastern Pacific. Mar. BioI. 48: 349--356. Heckel, P. H. 1974. Carbonate buildups in the geologic record: A review. Soc. Econ. Paleontol. Mineral. Spec. Pub. 18: 90-154. Hickson, R E., Simon, C., Cooper, A" Spicer, G., Sullivan, J. and Penny, D. Submitted. Refinement of a secondary structure model for the third domain of animal 125 rRNA, with a comparison of alignment programs. Molec. BioI. Evol. : Hidaka, M. 1992. Use of nematocyst morphology for taxonomy of some related species of scleractinian corals. Galaxea 11: 21-28. Hill, D. and Wells, J. W. 1956. Cnidaria- General Features. Ed. Moore, R. C. Treatise on Invertebrate Paleontology: Part F Coelenterata. Geol. Soc. of Amer. and Univ. of Kansas Press. pp. F5-F9. Hillis, D. M. 1987. Molecular versus morphological approaches to systematics. Ann. Rev. Ecol.Syst. 18:~ Hillis, D. M. and Huelsenbeck, J. P. 1992. Signal, noise and reliability in molecular phylogenetic analyses. J. Hered. 83: 189-195.

184

---- _.-. - .. __ .._------Hoeksema, B. W. 1989. Taxonomy, phylogeny and biogeography of mushroom corals (Scleractinia: Fungiidae). Zoologische Verhandelingen 254: 1-295. Hori, H. and Satow, Y. 1991. Dead-end evolution of the Cnidaria as deduced from 55 ribosomal RNA sequences. Hydrobiol. 216/217: 505-508. Hyman, L. 1940. The Invertebrates. I. Protozoa through Ctenophora. New York: Mct:;raw-Hill. 726 pp. Jablonski, D. 1991. Extinctions: A paleontological perspective. Science 253: 754 757. Jacobs, H. T., Elliott, D. J., Math, V. B. and Farquaharson, A. 1988. Nucleotide sequence and gene organization of sea urchin mitochondrial DNA. J. Mol. BioI. 202: 185-217. Jacobs, H. T., Posakony, J. W., Grula, J. W., Roberts, J. W., Xin, J.-H., Britten, R. J. and Davidson, E. H. 1983. Mitochondrial DNA sequences in the nuclear genome of Strongylocentrotus purpuratus. J. Mol. BioI. 165: 609-632.

Jenkins, R. J. F. 1984. Interpreting the oldest fossil cnidarians. Palaeontogr. Amer. 54: 95-104. [okiel, P. 1984.Long distance dispersal of reef corals by rafting. Coral Reefs 3: 113-116. [okiel, P. and Martinelli, F. J. 1992. The vortex model of coral reef biogeography. J. Biogeogr. 19:449-458. Kay, E. A. 1984. Patterns of speciation in the Indo-West Pacific. Spec. Pubis. B.P. Bishop Museum 72: 15-31.

185

------._- Kessing, B. D. 1991. Strongylocentrotid sea urchin mitochondrial DNA: Phylogenetic relationships and patterns of molecular evolution. Masters Thesis. Dept. of Zoology, University of Hawaii. Knowlton, N., Weigt, L. A., Solorzano, L. A., Mills, D. K. and Bermingham, E. 1993. Divergence in proteins, mitochondrial DNA, and reproductive compatibility across the Isthmus of Panama. Science 260: 1629-1632 Knowlton, N., Weil, E., Weigt, L. A. and Guzman, H. M. 1992 Sibling species in Montastrea annularis, coral bleaching, and the coral climate record. Science 255: 330-333. Krasnov, E. V. 1984. Evolutionary trends of scleractinians and coral communities in Post-Paleozoic time (Abstract). Palaeontogr. Amer. 54: 225. Lake, J. A. 1989. Origin of the multicellular animals. Eds. Fernholm, B., Bremer, K and Jomvall, H. The Hierarchy of Life.- Amsterdam: Elsevier Science Publishers B.V. pp. 273-278. Lake, J. A. 1991. Tracing origins with molecular sequences: metazoan and eukaryotic beginnings. TIBS 18: 46-50. Lamberts, A. E. 1979. Reef coral identification using protein electrophoresis: A preliminary report. Proceedings of the International Symposium on Marine Biogeography and Evolution in the Southern Hemisphere, New Zealand. 2:671-677. Lang, J. C. 1984. Whatever works: The variable importance of skeletal and of non-skeletal characters in scleractinina taxonomy. Palaeontogr. Amer. 54: 18-44. Li, W.-H. and Graur, D. 1991. Fundamentals of Molecular Evolution. Sunderland: Sinauer Associates, Inc. 284 pp.

186

------Li, W.-H, Luo, c-c, and Wu, C. I. 1985. Evolution of DNA sequences. Ed. MacIntyre, R. J. Molecular Evolutionary Genetics. New York: Plenum Press. pp. 1-94. Lopez, J. V., Yuhki, N., Masuda, R., Modi, W. and O'Brien, S. J. 1994. Numt, a recent transfer and tandem amplification of mitochondrial DNA to the nuclear genome of the domestic cat. J. Mol. Evo!. 39: 174-190. Maddison, W. P. and Maddison, D. R. 1992. MacClade: Analysis of phylogeny and character evolution. Sunderland, Massachusetts: Sinauer Associates. Maly, P. and Brimacombe, R. 1983. Refmed secondary structure models for the 16S and 23S ribosomal RNA of Escherichia coli. Nuc, Acids Res. 11: 7263-7286. Marchuk, D., Drumm, M., Saulino, A. and Collins, F. S. 1991. Construction of T-vectors, a rapid and general system for direct cloning of unmodified PeR products. Nuc. Acids Res. 19: 1154. McCoy, E. D. and Heck, K. L. 1976. Biogeography of corals, seagrasses, and mangroves: An alternative to the center of origin concept. Syst. Zool. 25: 201-210. McManus, J. W. 1985. Marine speciation, tectonics and sea-level changes in Southeast Asia. Proceedings of the 5th International Coral Reef Congress 4: 133-138. McMillan, J., Mahony, T., Veron, J. E. N. and Miller, D. J. 1991. Nucleotide sequencing of highly repetitive DNA from seven species in the coral genus Acropora (Cnidaria: Scleractinia) implies a division contrary to morphological criteria. Mar. BioI. 110: 323-327.

187

----~------. -.- McMillan, J. and Miller, D. J. 1989. Nucleotide sequences of highly repetitive DNA from scleractinian corals. Gene 1989: 185-186. McMillan, J. and Miller, D. J. 1990. Highly repeated DNA sequences in the scleractinian coral genus Acropora: Evaluation of cloned repeats as taxonomic probes. Mar. BioI. 104: 483-487. McMillan, J., Yellowlees, D., Heyward, A., Harrison, P. and Miller, D. J. 1988. Preparation of high molecular weight DNA from hermatypic corals and its use for DNA hybridization and cloning. Mar. Biol. 98: 271-276. Meyer, A. and Wilson, A. C. 1990. Origin of tetrapods inferred from their mitochondrial DNA affiliation to lungfish. J. Mol. Evol. 31: 359-364. Miller, D. J., Harrison, P. L, Mahony, R. J., McMillan, J. P., Odorico, D. M. and tenLohuis, M. R. 1993. Nucleotide sequence of the histone gene cluster in the coral Acropora formosa (Cnidaria; Scleractinia): Features of histone gene structure and organization are comrrion to diploblastic and triploblastic metazoans. J. Mol. Evol. 37: 245-253. Miller, D. J., McMillan, J., Miles, A., tenLohuis, M. and Mahony, T. 1990. Nucleotide sequence of the histone H3-encoding gene from the scleractinian coral Acropora formosa (Cnidaria: Scleractinia). Gene 93: 319-320. Miyamoto, M. M., Tanhauser, S. M. and Laipis, P. J. 1989. Systematic relationships in the artiodactyl tribe Bovini (family Bovidae), as determined from mitochondrial DNA sequences. Syst. Zool. 38: 342 349. Moritz, C. and Brown, W. M. 1986. Tandem duplication of D-loop and ribosomal RNA sequences in lizard mitochondrial DNA. Science 233: 1425-1427.

188 Moritz, C. and Brown, W. M. 1987. Tandem duplications in animal mitochondrial DNAs: Variation in incidence and gene content among lizards. PNAS 84: 7183-7187. Newell, N. D. 1967. Revolutions in the history of life. Geol, Soc. Amer. Spec. Paper 89: 63-91. Newell, N. D. 1971. An outline history of tropical organic reefs. Amer. Mus. Nov. 2465: 1-37. Newman, W. A. 1986. Origin of the Hawaiian marine fauna: Dispersal and vicariance as indicated by barnacles and other organisms. Eds. Gore, R H. and Heck, K. L. Crustacean Biqgeography. pp.21-49. Novacek, M. J. 1994. Morphological and molecular inroads to phylogeny. Eds. Grande, L. and Rieppel, O. Inter,preting the Hierarchy of Nature. San Diego: Academic Press. pp. 85-131. Ogilivie, M. M. 1896. Microscopic and systematic study of madreporarian types

of corals. Phil. Trans. r. Soc. London, (B) 187: 83-345. Ohlhorst, S. L. 1984. The use of polyacrylamide gel electrophoresis in coral taxonomy. Palaeontogr. Amer. 54: 45-48. Oliver, W. A. J. 1980. The relationship of the scleractinian corals to the rugose corals. Paleobiology 6: 146-160. Palumbi, S. R. and Baker, C. S. 1994. Contrasting population structure from nuclear intron sequences and mtDNA of humpback whales. Mol. Biol, Evol. 11: 426-435. Palumbi, S. R., Martin, A. P., Romano, S. L., McMillan, W.O., Stice, L. and Grabowski, G. 1991. The Simple Fool's Guide to PCR. Honolulu: Dept. of Zoology, University of Hawaii. 44 pp. Pantin, C. F. A. 1959. Diploblastic animals. Proc, linn. Soc. Lond. 171: 1-18.

189

----~~~~~ ------Pashley, D. P. and Ke, L. D. 1992. Sequence evolution in mitochondrial ribosomal and ND-l genes in Lepidoptera: Implications for phylogenetic analyses. Mol. BioI. Evol. 9: 1061-1075. Patterson, C. 1987. Introduction. Ed. Patterson, C. Molecules and morphology in evolution: conflict or compromise? Cambridge: Cambridge University Press. pp. 1-22- Porter, J. W. 1972. Ecology and species diversity of coral reefs on opposite sides of the Isthmus of Panama. Bull. BioL Soc. Wash. 2: 89-116. Potts, D. C. 1983. Evolutionary disequilibrium among Indo-Pacific corals. Geol. Soc. Amer. Bull. 33: 619-632. Potts, D. C. 1984. Generation times and the Quaternary evolution of reef building corals. Paleobiology 10: 48-58. Potts, D. C. 1985. Sea-level fluctuations and speciation in Scleractinia. Proceedings of the 5th International Coral Reef Congress, Tahiti. 4:127 132 Potts, D. C., Budd, A. F. and Garthwaite, R. L. 1993. Soft tissue vs, skeletal approaches to species recognition and phylogeny reconstruction in corals. Cour. Forsch.-Inst Senckenberg 164: 221-231. Powers, D. 1970. A numerical taxonomic study of Hawaiian reef corals. Pac,

Sci. 24: 180-186. Powers, D. A. and Rohlf, F. J. 1972. A numerical taxonomic study of Caribbean and Hawaiian reef corals. Syst. Zool. 21: 53-64. Pratz, E. 1882. Ueber die verwandtschaftlichen Beziehungen einiger Korallegattungen mit hauptsachlicher Beriicksichtigung ihrer Septal Struktur. Paleontographica XXIX: 81-122.

190

------Randall, R. H. 1976. Some problems in reef coral taxonomy. Micronesica 12:

151-156.

Raven, P. H. and Axelrod, D. I. 1972. Plate tectonics and Australasian

paleobiogeography. Science 176: 1379-1386. Richmond, R. H. 1987. Energetic relationships and biogeographical differences among fecundity, growth and reproduction in the reef coralPocillopora damicornis. Geol. Soc. Amer. Bull. 41: 594-604. Richmond, R. H 1987. Energetics, competency, and long-distance dispersal of planula larvae of the coral Pocillopora damicornis. Mar. Biol, 93: 527

533. Richmond, R. H. 1990. Relationships among reproductive mode, biogeographic distribution patterns and evolution in scleractinian corals. Eds. Hoshi, M. and Yamashita, O. Advances in Invertebrate

Reproduction. Amsterdam: Elsevier Science Publishers B.V. pp. 317 322.

Richmond, R. H. and Hunter, C. L. 1990. Reproduction and recruitment of corals: Comparisons among the Caribbean, the Tropical Pacific, and the

Red Sea. Mar. Ecol. Prog, Ser. 60: 185-203. Robson, E. A. 1985. Speculations on coelenterates. Eds. George, J. D., Gibson, R. and Platt, H. M. The Origins and Relationships of Loser Invertebrates.

Oxford: Clarendon Press. pp. 397. Roe, B. A., Ma, D.-P., Wilson, R K. and Wong, J. F.-H. 1985. The complete nucleotide sequence of the Xenopus laeois mitochondrial genome. J. BioI. Chern. 260: 9759-9774.

Roniewicz, E. 1976. Les scleractiniaires du jurassique superieur de la Dobrogea Centrale, Roumanie. Palaeon. Pol. 34: 2-22.

191

------_._- -- Roniewicz, E. and Morycowa, E. 1993. Evolution of the Scleractinia in the light of microstructural data. Cour. Forsch.-Inst. Senckenberg 164: 233 240. Rosen, B. R. 1971. The distribution of reef coral genera in the Indian Ocean. Symp. zool. Soc. Lond. 28: 263-299. Rosen, B. R. 1981. The tropical high diversity enigma- the corals' eye view. Ed. Forey, P. L. The Evolving Biosphere. London: Cambridge University Press. pp. 103-129. Rosen, B. R. 1984. Reef coral biogeography and climate through the Late Cainozoic: Just islands in the sun or a critical pattern of islands? Ed. Brenchley, P. Fossils and Climate. Chichester: John Wiley and Sons, Ltd. pp. 201-262- Rosen, B. R. 1988a. Biogeographic patterns: a perceptual overview. Eds. Myers, A. A. and Giller, P. S. Analytical Biogeography. New York: Chapman and Hall. pp. 23-55. Rosen, B. R. 1988b. From fossils to earth history: applied historical biogeography. Ed. Giller, A. A. M. a. P. S. Analytical Biogeography. New York: Chapman and Hall. pp. 437-481. Rosen, B. R. 1988c. Progress, problems and patterns in the biogeography of reef corals and other tropical marine organisms. Helg. Meer. 42: 269-301. Rosen, B. R. and Smith, A. B. 1988. Tectonics from fossils? Analysis of reef coral and sea-urchin distributions from late Cretaceous to Recent, using a new method. Eds. Audley-Charles, M. G. and Hallam, A. Gondwana and Tethys. pp.275-306.

192 Rosen, B. R. and Turnsek, D. 1989. Extinction patterns and biogeography of scleractinian corals across the Cretaceous/Tertiary boundary. Mem. Ass. Australas. Palaeontols 8: 355-370. Runnegar, B. 1992. Evolution of the earliest animals. Ed. Schopf, J. W. Major Events in the History of Life. Boston: Jones and Bartlett Publishers. pp. 65-93. Russo, A. 1979. Studio monografico sui coralli dell' Eocene di Possagno (Treviso, Italy). Contr. Paleont. Univ. Modena 182: 1-87. Schmidt, H. 1974. On evolution in the Anthozoa. Proceedings of the Second International Coral Reef Symposium, Brisbane. 1:533-560. Schuhmacher, H. and Zibrowius, H. 1985. What is hermatypic? A redefinition of ecological groups in corals and other organisms. Coral Reefs 4: 1-9. Scott, R. W. 1984. Evolution of early Cretaceous reefs in the Gulf of Mexico. Palaeontogr. Amer. 54: 406-412. Scrutton, C. T. 1979. Early Fossil Cnidarians. Ed. House, M. R. The Origin of Major Invertebrate Groups. London: Academic Press. pp. 161-207. Scrutton, C. T. 1984. Late Precambrian and early Paleozoic history of cnidarians. Palaeontogr. Amer. 54: 93-94. Sepkoski, J. J. J. 1982. Mass extinctions in the Phanerozoic oceans: A review. Eds. Silver, L. T. and Schultz, P. H. Geological Implications of Large Asteroids and Comets on the Earth. pp. 283-289. Sheppard, C. R. C. 1987. Coral species of the Indian Ocean and adjacent seas: A synonymized compilation and some regional distributional patterns. Atoll Res. Bull. 307: 1-32 Smith, A. B., Lafay, B. and Christen, R. 1992. Comparative variation of morphological and molecular evolution through geologic time: 285

193

------._ _-_ ..- ribosomal RNA versus morphology in echinoids. Phil. Trans. r. Soc. London, (B) 338: 365-382. Smith, M. F., Thomas, W. K and Patton, J. L. 1992. Mitochondrial DNA-like sequence in the nuclear genome of an akodontine rodent. Mol. Biol. Evol. 9: 204-215. Squires, D. F. 1967. The evolution of the deep-sea coral family Micrabaciidae. Studies in Tropical Oceanography 5: 502-510. Stanley, G. D. J. 1981. Early history of scleractinian corals and its geological consequences. Geology 9: 507-511. Stanley, G. D. J. 1988. The history of early Mesozoic reef communities: A three-step process. Palaios 3: 170-183. Stanley, S. M. 1984. Temperature and biotic crises in the marine realm. Geology 12: 205-208. Stanley, S. M. 1989. Earth and Life Through Time. New York: Freeman and

Co. 689pp. Stehli, F. G. and Wells, J. W. 1971. Diversity and age patterns in hermatypic corals. Syst. Zool. 20: 115-126. Swofford, D. L. 1993. Phylogenetic Analysis Using Parsimony (PAUP), version 3.1.1. Champaign: illinois Natural History Survey. Symonds, P. A. and Davies, P. J. 1985. Origin of the Great Barrier Reef (Abstract). Proceedings of the 5th International Coral Reef Congress, Tahiti. 3:371. Szmant, A. 1986. Reproductive ecology of Caribbean reef corals. Coral Reefs 5: 43-54. Tollmann, A. and Kristan-Tollmann, E. 1985. Paleogeography of the

European Tethys from Paleozoic to Mesozoic and the Triassic relations

194

----- .._.._ __ _------_ _.. _ . of the eastern part of Tethys and Panthalassa. Eds. Nakazawa, K. and Dickins, J. M. The Tethys: Her Paleo,geograph). and Paleobiogeography from Paleozoic to Mesozoic. Tokyo: Tokai University Press. pp. 3-22. Tsuzuki, T., Nomiyama, H., Setoyama, C., Maeda, S. and Shimada, K. 1983. Presence of mitochondrial-DNA-like sequences in the human nuclear

DNA. Gene 25: 223-229. Uhlenbusch, I., McCracken, A. and Gellissen, G. 1987. The gene for the large (16S) ribosomal RNA from the Locusta migratoria mitochondrial genome. Curro Genet. 11: 631-638. Van Veghel, M. and Bak, R 1993. Intraspecific variation of a dominant Caribbean reef building coral, Montastrea annularis: genetic, behavioral and morphometric aspects. Mar. Ecol. Prog. Ser. 92: 255-265. Vaughan, T. W. and Wells, J. W. 1943. Revision of the suborders, families and genera of the Scleractinia. Geological Society of America Special Papers 44: 1-363. Vermeij, G. J. 1987. The dispersal barrier in the tropical Pacific: Implications for molluscan speciation and extinction. Evolution 41: 1046-1058. Veron, J. E. N. 1985. Aspects of the biogeography of hermatypic corals. Proceedings of the 5th International Coral Reef Congress, Tahiti. 83-88. Veron, J. E. N. 1986. Corals of Australia and the Indo-Pacific. Honolulu: Australian Institute of Marine Science and the University of Hawaii

Press. 644 pp. Veron, J. E. N. 1993. Biodiversity of reef corals: Is there a problem in the Indo Pacific centre of diversity? Proceedings of the Colloquium on Global Aspects of Coral Reefs. Health, Hazards, and History., University of Miami. 365-370.

195

------Veron, J. E. N. 1993. A Biogeographic Database of Hermatypic Corals. Townsville: Australian Institute of Marine Science. 433 pp. Veron, J. E. N. in press. Corals in Space and Time: The Biogeography and Evolution of the Scleractinia. University of New South Wales Press. pp.Pages. Veron, J. E. N. and Kelley, R. 1988. Species stability in reef corals of Papua New Guinea and the Indo-Pacific. Mem. Ass. Australas. Palaeontols 6: 1-69. Veron, J. E. N. and Pichon, M. 1976. Scleractinia of eastern Australia, Part 1. Australian Institute of Marine Science, Monograph Series 1: 1-86. Veron, J. E. N. and Pichon, M. 1980. Scleractinia of Eastern Australia, Part m. Canberra: Australian National University Press. 433 pp. Veron, J. E. N. and Pichon, M. 1982. Scleractinia of Eastern Australia, Part IV. Canberra: Australian National University Press. 159 pp. Veron, J. E. N., Pichon, M. and Wijsman-Best, M. 1977. Scleractinia of Australia Part n. Canberra: Australian Government Publishing Service. Z:J3pp. Veron, J. E. N. and Wallace, C. C. 1984. Scleractinia of Australia, Part V. Canberra: Australian National University Press. 485 pp. Wainright, P.O., Hinkle, G., Sogin, M. L. and Stickel, S. K. 1993. Monophyletic origins of the Metazoa: An evolutionary link with Fungi. Science 260: 340-342-

Wallace, C. C. 1974. A numerical study of a small group of Acropora specimens (Scleractinia: Acroporidae). Mem. Qld. Mus. 17: 55-61. Wallace, C. c, Pandolfi, J. M., Young, A. and Wolstenholme, J. 1991. Indo Pacific coral biogeography. Aust. Syst. Bot. 4: 199-210.

196 Warrior, R. and Gall, J. 1985. The mitochondrial DNA of Hydra attenuata and Hydra littoralis consists of two linear molecules. Arch. Sc, Geneve 38: 439-445. Wells, J. W. 1933. Corals of the Cretaceous of the Atlantic and Gulf coastal plains and western interior of the United States. Bull. Am. Paleontol. 18:85-288. Wells, J. W. 1954. Bikini and nearby atolls, Marshall Islands: Recent corals of the Marshall Islands. Prof. pap. U.S. geol Surv. 260-1: 385-486. Wells, J. W. 1956. Scleractinia. Ed. Moore, R. C. Treatise on Invertebrate Paleontology: Coelenerata. Kansas: Geological Society of America and University of Kansas Press. pp. F328-F443. Wells, J. W. 1966. Evolutionary development in the scleractinian family Fungiidae. Symposia of the Zoological Society of London 16: 223-246. Wells, J. W. 1969. Aspects of Pacific coral reefs. Microneska 5: 317-322. Wells, J. W. 1986. A list of scleractinian generic and subgeneric taxa, 1758-1985. Fossil Cnidaria 15: 1-69. Werner, B. 1973. New investigations on systematics and evolution of the class Scyphozoa and the phylum Cnidaria. Publs, Seto. mar. bioI. Lab 20: 35 61. Wijsman-Best, M. 1974. Habitat-induced modification of reef corals (Faviidae) and its consequences for taxonomy. Proc, 2nd Int. Coral Reef Symp. 2: 217-228. Wiley, E. O. 1981. Phyl~enetics: The theory and practice of phylogenetic systematics. New York: Wiley and Sons. 439 pp.

197

------Willis, B. L. 1985. Phenotypic plasticity versus phenotypic stability in the reef corals Turbinaria mesenierina and Pavona cactus. Proceedings of the 5th International Coral Reef Congress: 107-112- Willis, B. L. 1990. Species concepts in extant scleractinian corals: Considerations based on reproductive biology and genotypic population structures. Syst. Bot. 15: 136-149. Willis, B. L., Babcock, R. C., Harrison, P. L and Wallace, C. C. 1992. Experimental evidence of hybridization in reef corals involved in mass spawning events (Abstract). Proceedings of the 7th International Coral Reef Symposium, Guam. 1:504. Willmer, P. 1990. Invertebrate Relationships: Patterns in Animal Evolution. New York: Cambridge University Press. 400 pp. Wilson, A. c., Cann, R. L., Carr, S. M., George, M., Gyllensten, U. B., Helm Bychowski, K M., Higuchi, R. G., Palumbi, S. R., Prager, E. M., Sage, R. D. and Stoneking, M. 1985. Mitochondrial DNA and two perspectives on evolutionary genetics. BioI. J. Linn. Soc. 26: 375-400. Wolstenholme, D. R. 1992. Animal mitochondrial DNA: Structure and evolution. Int. Rev. Cyt. 141: 173-216. Yonge, C. M. 1940. The biology of reef-building corals. Scientific Reports of the Great Barrier Reef Expedition 1: 353-385. Zibrowius, H. 1980. Les Scleractiniaires de la Mediterranee et de l'Atlantique nord-oriental. Mem. Instit. Ocean. 11: 1-227. Zibrowius, H. 1984. Taxonomy in ahermatypic scleractinian corals. Palaeontogr. Amer. 54: 80-85. Zouros, E., Ball, A. 0., Saavedra, C. and Freeman, K. R. in press. A new type of mitochondrial DNA inheritance in the blue mussel Mytilus. PNAS.

198

---- _._ .....- .._-_. __._------_. _._ .. zene.s, Sieu, L. C., Slightom, J. L., Hadler, H. I. and Eisenstadt, J. M. 1991. Mitochondrial D-loop sequences are integrated in the rat nuclear genome. J. Mol. BioI. 221: 1223-1235.

199

------