Louisiana State University LSU Digital Commons

LSU Doctoral Dissertations Graduate School

2014 Adaptive Divergence and Speciation across Depth in a Caribbean Candelabrum Coral Carlos Andrés Prada Montoya Louisiana State University and Agricultural and Mechanical College, [email protected]

Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_dissertations

Recommended Citation Prada Montoya, Carlos Andrés, "Adaptive Divergence and Speciation across Depth in a Caribbean Candelabrum Coral" (2014). LSU Doctoral Dissertations. 1736. https://digitalcommons.lsu.edu/gradschool_dissertations/1736

This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Doctoral Dissertations by an authorized graduate school editor of LSU Digital Commons. For more information, please [email protected]. ADAPTIVE DIVERGENCE AND SPECIATION ACROSS DEPTH IN A CARIBBEAN CANDELABRUM CORAL

A Dissertation

Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of Doctor of Philosophy

in

The Department of Biological Sciences

by Carlos Prada B.S., Universidad Jorge Tadeo Lozano, 2002 M.S., University of Puerto Rico, 2008 August 2014

To my mother, Cielo Montoya

ii Acknowledgments

I would like to thank my adviser, Mike Hellberg. Mike has been a perfect match for me. Mike allowed me the freedom to develop my own ideas, was instrumental in pointing out dead ends, and canalized my work to generate publishable material of the highest possible quality. Mike was always there to talk with me about my research and provided advice on my career plans. Mike has encouraged me to write and has been generously patience to correct all my writings, often within 24 hours. His close readings provided thoughtful comments that one by one have over the years sharpened my mind; making me a better communicator and a more independent researcher. Mike encouraged me to develop my own system and supported my ideas on the importance of long pre-reproductive selection in long-lived organisms, the role of natural selection in the formation of species, and the role of plasticity in evolution. I would also like to thank Andrew Whitehead, whose admirable scholarship has impregnated my career and increased the quality of my research. Talks with Andrew in and out of class were instrumental to my understanding of the dynamic nature of genomes, the complexity in generating phenotypes -incredibly understudied in biology- and the pervasive role of natural selection in evolution. I would like to thank Morgan Kelly, who provided detailed comments on Chapter 4 and was generous to agree to be part of my graduate committee after Bryan Carstens moved to Ohio State. I thank Robb Brumfield, who provided advice for my research and career as a whole. Robb deepened my knowledge about the neutral theory and the coalescent, especially as they apply to systematics. Having a committee member from the Museum also opened the possibility to learn about vertebrate evolution, especially of Neotropical birds.

The Hellberg lab members also provided unlimited support to my research during my stay in Baton Rouge. I want to highlight Melissa Debiasse, my lab sister, whom I shared most of my academic experiences. Melissa reviewed various aspects of my dissertation, spent unlimited hours talking with me, debunked my ideas and increased my oral communication skills. Melissa’s integrity is admirable and I had not remembered a single misunderstanding with her even during moments of extreme pressure. I also thank Ron, Alice, Karine, Mark, Cata and Maxine from whom I learned at different times during my Ph. D.

I would like to thank Bill Eberhard and Mary Jane West-Eberhard for spending many hours talking to me about my ideas, provide advice, and most of all for sharing their admirable and enviable integrative view of evolution. I feel fortunate to have my understanding of the role of plasticity in evolution –often underappreciated and misunderstood in biology- shaped first hand by Mary Jane. I am glad to have such prestigious biologists as my guides during my Ph. D.

I thank Louisiana State University, the Department of Biological Sciences and the Graduate School, which provided me with numerous opportunities, including the Huel D. Perkins and the Graduate Year fellowships, which financed my studies and provided freedom to devote my time entirely to research. I would also like to thank the academic community at the Systematics, Ecology, and Evolution division at LSU and the folks at the Museum. The wide

iii variation in species, techniques and approaches has enriched my carrier and constructed bridges for future collaborations. I specially thank Sandra Galeano and Metha Klock, who provided unforgettable talks during coffee breaks.

Funds for this project came from the National Science Foundation (OCE-0550270 to Baums and Hellberg and DEB-1311579 to Prada and Hellberg), the Society for Systematic Biologists, a Rosemary Grant Award from the Society for the Study of Evolution, a Lerner Gray Award from the American Museum of Natural Science, the Department of Marine Sciences of the university of Puerto Rico, Mayaguez through Richard Appeldorn and the NOAA project NA06NOS4780190. I would also like to thank Howard Lasker, who generously allowed me to join his team in The Bahamas. I also thank Mark Vermeij and the Caribbean Research and Management of Biodiversity (CARMABI) and Rachael Collin and The Smithsonian Tropical Research Institute for their valuable support during field collections.

Contributions from Shelby Mcllroy, Mary Alice Coffroth, Daniel Valient and Diana Beltrán helped to develop Chapter 3. Comments from Joe Niegel improved Chapter 2. I thank Paul Yoshioka, who initially inspired me to study octocorals and also provided unrestricted access to his long-term demographic data of shallow-water octocorals. I thank Matt Brown and The Socolofsky Microscopy Center at Louisiana State University that provided assistance for morphological measurements.

Last but not least, I would like to thank my family. My mother, Cielo Montoya, is largely responsible for me being here, providing the initial financial support and encouragement to pursue graduate studies. She dedicated herself to my education during my elementary years and set the foundation of my reasoning and critical thinking, which were fundamental to developing my dissertation. I thank Diana Beltran, who has stood with me all this time, has been part of most of my field expeditions, has supported most of my ideas and without whom I had simply not had been able to finished this dissertation. All my gratitude for her love and patience all this time.

iv Table of Contents

Acknowledgments……………………………………………………………………………………. iii

Abstract……………………………………………………………………………………………………..vi

Chapter 1. Introduction……………………………………………………………………………….1

Chapter 2. Long Pre-reproductive Selection and Divergence by Depth in a Caribbean Candelabrum Coral……………………………………………………………………..6

Chapter 3. Cryptic Diversity Hides Host and Habitat Specialization in a Gorgonian-algal Symbiosis……………………………………………………………………... 22

Chapter 4. Strong Natural Selection on Juveniles Maintains a Narrow Adult Hybrid Zone in a Broadcast Spawner………………………………………………………. 33

Chapter 5. Conclusions…….………………………………………………………………………..46

References……………………………………………………………………………………………..…48

Appendix A. Supplementary Information for Chapter 2……………………….……..67

Appendix B. Supplementary Information for Chapter 3………………………………74

Appendix C. Supplementary Information for Chapter 4………………………………75

Appendix D. Agreement to Reproduce Chapter 2………………………………………77

Appendix E. Agreement to Reproduce Chapter 3………………………………………78

Vita………………………………………………………………………………………………………….79

v Abstract

Since Darwin, biologists wonder how organisms cope with environmental variation, why there are so many species, and how species form. In my dissertation, I explore how species of long-lived, clonal reef organisms originate across depth gradients. In Chapter 2, I evaluate the strength of depth to isolate populations by comparing the genes and morphologies of pairs of depth-segregated populations of the candelabrum coral Eunicea flexuosa across the Caribbean. Eunicea flexuosa is a long-lived clonal cnidarian that associates with an alga of the genus Symbiodinium. Genetic analysis revealed two depth- segregated lineages, each genetically well-mixed across the Caribbean. Survivorship data, combined with estimates of selection coefficients based on transplant experiments, suggest that selection is strong enough to segregate these two lineages. Limited recruitment to reproductive age, even under weak annual selection advantage, is sufficient to generate habitat segregation because of the cumulative selection accrued during prolonged pre- reproductive growth. I then studied in detail the genetic diversity of Symbiodinium in each Shallow and Deep E. flexuosa lineage (Chapter 3). I sampled colonies of the two ecotypes across depths at three Caribbean locations. I find that each host lineage is associated with a unique Symbiodinium variant. This relationship between host and alga is maintained when host colonies are reciprocally transplanted. Even when the clades of both partners are present at intermediate depths, the specificity between host and algal lineages remained. I then test whether the Shallow/Deep adaptive divergence occurs in natural populations, by examining the frequencies of juvenile and adults. The habitat distributions of the two lineages are more distinct when inferred from adults than from juveniles. Selection coefficients from cohort data correlates with that from transplant experiments (Chapter 4). The two lineages form a narrow hybrid zone (100 m), with coincident clines of both the host coral and its algal symbiont. Effective dispersal estimates derived from the hybrid zone are small (20 m) for a broadcast spawner with a large dispersing potential (50 km). Ecological factors associated with depth act as filters generating strong barriers to gene flow despite extensive dispersal, altering morphologies, and contributing to the potential for speciation in the sea.

vi Chapter 1 Introduction

The co-occurrence of closely related marine species with high dispersal capabilities poses a challenge for evolutionary biologists. How can new species emerge without obvious geographic isolation? Caribbean coral reefs provide several examples where young endemic clades appear to have been generated within the Caribbean Basin (Taylor and Hellberg 2005). One explanation for within-basin speciation is that geological processes may have transiently divided populations long enough for new species to form (Hellberg 1998; Prada et al. 2014a). When conditions allowed, recently diverged sister species may have then expanded their ranges and come into secondary contact (Vermeij 2005).

Another possibility, albeit less explored and not mutually exclusive, is that diversification has occurred due to natural selection acting on populations occupying different habitats within the same geological region (Darwin 1859; Schluter 2009a). Ecological factors have been invoked as promoters of speciation where dispersal ability does not impede random mating (Schluter 1996). In these instances, speciation appears to have occurred due to natural selection acting on genes responsible for ecological traits. Indeed, ecological segregation is widespread among closely related marine species (Knowlton 1993). On Caribbean coral reefs, genetic differences have been detected between several ecomophs or habitat-segregated populations with overlapping ranges (Brazeau and Harvell 1994; Carlon and Budd 2002; Levitan et al. 2004; Prada et al. 2008).

One factor that commonly divides coral reef species is depth (Knowlton 1993). Depth co- varies with light, water motion, sediment transport and many other factors. Variation in the interaction of these factors over depth produces differences in resource availability, favoring combinations of traits that result in fitness differences among habitat segregated populations (Prada et al. 2008). Along with physiological changes to match the environments at different depths, depth-segregated marine broadcast-spawners often differ in the timing of spawning, which can lead to temporal reproductive isolation (Knowlton et al. 1997). Thus, depth segregation has the potential to link ecological and reproductive traits, increasing the likelihood of isolation (Van Doorn et al. 2009). Indeed, a recent genomic analysis of sequence divergence among depth-segregated urchins shows the role of natural selection in shaping such a transition (Oliver et al. 2010). Despite depth segregation being widespread in marine systems, and potentially linking ecology and reproductive isolation with known theoretical models (Tomaiuolo et al. 2007), it has received little attention.

In this thesis, I aim to examine how depth and geography interact during the early stages of speciation in Caribbean candelabrum corals. I use population genetic and coalescent-based phylogeographic analyses involving depth and geography to identify how these factors have affected divergence in Eunicea and its algal symbiont across the Caribbean. I also study how ecological speciation may proceed in long-lived sessile species across depths in coral reefs to generate and maintain biodiversity.

1 1.1. ECOLOGICAL SPECIATION

In ecological speciation, reproductive isolation is achieved by divergent natural selection acting on ecologically segregated populations even when dispersal is not an impediment to random mating (Rundle and Nosil 2005; Nosil 2012). In these instances, speciation appears to have occurred due to natural selection acting on genes responsible for ecological traits. Even when gene flow is absent during divergence, ecological speciation can accelerate the divergence process (Schluter 2009a). Similarly, because local adaptation generates alternative states in different environments, when nascent species come into contact they are less likely to fuse because both extrinsic and intrinsic factors reduce gene flow (Doebeli et al. 2005; Van Doorn et al. 2009). Ecological speciation research has provided evidence that reproductive isolation can happen rapidly in both plants and animals (Barluenga et al. 2006; Savolainen et al. 2006), producing parallel patterns across taxa and geographical regions (Østbye et al. 2005; Derome et al. 2006; Quesada et al. 2007; Schluter 2009b). Indeed, this research has provided evidence of adaptive radiations resulting from ecological speciation (Schluter 2000; Langerhans et al. 2007; Jiggins 2008). Most of these cases involve foraging behaviors in alternative habitats or predator-mediated responses (Feder et al. 1988; Schluter 1996; Lu and Bernatchez 1999; Johannesson 2001; Nosil et al. 2002; Barluenga et al. 2006; Hawthorne and Via 2001). Despite such progress, few studies have investigated the role of co-evolution across ecological gradients during ecological speciation nor explicit details of the mechanisms involved during speciation (but see (Nosil 2012). Herein I will couple population genetic analyses of both the coral host (Chapter 2) and its algal symbiont (Chapter 3) to show the degree of specificity between both partners across depth. I also show how habitat-mediated mortality generates a mechanism for assortative mating in these long-lived species (Chapter 4).

1.2. DEPTH SEGREGATION IN MARINE POPULATIONS: ECOLOGY AIDING SPECIATION

Depth is the clearest factor segregating reef biodiversity. Among sibling species in the sea reported by Knowlton (1993), over 50% involved depth as a dividing factor, even though not all comparisons included depth. Closely related anthozoan species often have different morphologies associated with differences in depth (Weil and Knowlton 1994). Indeed, on Caribbean coral reefs, genetic differences have been detected between depth-segregated populations with overlapping ranges (Brazeau and Harvell 1994; Carlon and Budd 2002; Levitan et al. 2004; Prada et al. 2008). A number of environmental variables co-vary with depth, including: light, water motion, sediment transport (both sedimentation and bed load) and the availability of food. Differences in these variables create differences in feeding strategies (Grottoli et al. 2006), in metabolic rates (Knowlton and Jackson 1994), in symbiotic relationships (Santos et al. 2004), in predator resistance (West et al. 1993) and in spawning times (Knowlton et al. 1997; Levitan et al. 2004). Despite segregation by depth being widespread in marine systems, potentially linking ecology and reproductive isolation, it has received little attention. Here, I show the importance of depth in maintaining and generating diversity on coral reefs.

2 1.3. GEOGRAPHY MAY HAVE ALSO ISOLATED POPULATIONS

In addition to depth, transient allopatry may play a role in early divergence leading to speciation. For example, the rise of the Isthmus of Panama altered ocean circulation, causing novel conditions in the Caribbean Basin that generated both widespread extinction and speciation in Caribbean mollusks, species that later expanded their distributions to the entire basin (Vermeij 2005). Geographic separation is also evident in populations of widespread distributed taxa at large (1000 km) (Baums et al. 2005) and small (<100 km) spatial scales (Gutiérrez-Rodríguez and Lasker 2004; Taylor and Hellberg 2006) in the Caribbean. Thus, new species may result from historic geographic isolation within the Caribbean. Despite the potential roles of geography and ecology in generating Caribbean biodiversity, few studies have considered both simultaneously. The genetic survey in Chapter 2 will analyze ecological divergence at multiple geographic locations to test whether population divergence is higher among depth comparisons than geographic ones.

1.4. EUNICEA FLEXUOSA AS A MODEL TO STUDY ECOLOGICAL SPECIATION

Several aspects of the biology of the candelabrum coral E unicea flexuosa makes it amenable to speciation studies: 1) it is a common (> 5 colonies/m2 (Yoshioka 2009)) Caribbean wide species (Bayer 1961), present in virtually all habitats and depths on coral reefs and rocky walls (Bayer 1961). 2) Depth-segregated populations of E. flexuosa are morphologically distinct and genetically differentiated, yet these differences are small, suggesting species have formed recently (Grajales et al. 2007; Prada et al. 2008). Assuming a within-Caribbean origin for Eunicea (Sanchez 2009) and a 30 yr generation time (Tracey et al. 2007), just 100,000 generations separate the entire radiation. 3) As a synchronous gonochoric (separate sexes) broadcast spawner with habitat-segregated populations, E. flexuosa Shallow and Deep have the potential to develop temporal reproductive isolation, thus directly linking ecology and reproductive isolation and increasing the chance of divergence in depth segregated populations (Servedio et al. 2011). 4) Because E. flexuosa is clonal, survivorship can be measured for clones experiencing different habitats. 5) First reproduction does not occur until a size of 30 cm, at least 15 years after larval settlement (Beiring and Lasker 2000). Such delayed reproduction during sedentary asexual growth allows selection to act for a long time before genes can be mixed and may increase selection in depth-adapted populations. 6) Unlike other reef taxa, fragmentation is uncommon in E. flexuosa and age can accurately be estimated from size, which facilitates comparisons across different age classes and enables experiments to understand mechanisms driving ecological speciation.

1.5. OVERVIEW OF CHAPTERS

To understand whether morphological differences between E. flexuosa colonies found at different depths were adaptive (Chapter 2), I monitored colonies for over four years and recorded survivorship per size class (in corals, size approximates age). I then estimate the probability that settling larvae would reach full maturity (about 30 years). This proportion turns out to be very low, only one in ten thousand. Second, I selected colonies from shallow

3 and deep habitats and transplanted clones of each to both depths. After three years of monitoring, I found that mean annual survivorship was highest for colonies transplanted to the same depth and lowest for reciprocally transplanted colonies. The survivorship advantage over a generation of natives over foreign colonies is 98.3% for shallow areas and 99.9% for deep. To evaluate whether this strong selection could genetically isolate populations, I sampled pairs of depth-segregated populations (189 individuals) across the Caribbean. I extracted genetic information for four molecular markers and found two distinct gene pools that correspond to depth. My results show depth is the major factor isolating the pair across the Caribbean, a pattern of divergence facilitated by the long period selection has to act between larval settlement and colony maturity. I suggest this long pre-reproductive selection may be a common mechanism enhancing adaptive divergence in other long-lived sessile species, such as trees and shrubs.

In chapter three, I evaluate the role of E. flexuosa's symbiotic algal partner in aiding the depth segregation. I first evaluate genetic diversity in both the host coral and its algal symbiont in pairs of depth segregated populations at three locations in the Caribbean. I found a tight association between the ecotype of the coral and the clade membership of its algal symbiont. To disentangle the roles of host species identity and depth, I studied populations from intermediate depths (12 – 15 m) where the E. flexuosa ecotypes co-occur. At these mid depths, the host and algal genotypes match completely, suggesting that host genotype is a better predictor of algal genotype than depth is. I then tested the stability of the host-algal relationships with a reciprocal transplant experiment. I found that coral-algal associations are mainly driven by host species identity, so that colonies in reciprocal transplants rarely switch algal types and host colonies remained true to symbiont types, even when occurring outside their usual depth. The high specificity of Symbiodinium with different E. flexuosa lineages suggests a pattern of coevolution and it is this coupling that defines the fitness of the holobiont in different environments.

To identify whether selection against maladapted migrants (immigrant inviability) drives the genetic divergence between the Shallow and Deep lineages (Chapter 4), I first compared the distribution of adults with that of juveniles at the extremes of the depth gradient. I found that while the lineages in both size classes are sorted by habitat, segregation is more pronounced in adults. I then sampled adult individuals across a depth gradient in Puerto Rico. I found that the two lineages replace each other over a depth and light gradient, producing few hybrids. Clines for morphological and molecular traits of both the coral host and its algal symbiont are coincident. The cline segregating the two lineages is remarkably small (< 100 m). I concluded that environmental selection largely maintains this hybrid zone and that the strong habitat selection segregates adult species by gradually eliminating young recruiting individuals overtime.

In summary, my dissertation research suggests that depth is a major factor isolating sister species of E. flexuosa across the Caribbean, a pattern aided by the long period that selection has to act between larval settlement and colony maturity. Selection is strong in both habitats for native over foreign species. The segregation of the sister species is accompanied by segregation of their symbiotic algae, which appear to have co-evolved with

4 each coral species. The sister corals spread gradually over the depth gradient, producing some hybrids but keeping genomic integrity. These results suggest that ecological factors - such immigrant inviability- associated with depth matter during the early stages of speciation in long-lived organisms.

5 Chapter 2 Long Pre-reproductive Selection and Divergence by Depth in a Caribbean Candelabrum Coral*

2.1. INTRODUCTION

Natural selection for adaptation to different ecological conditions has become increasingly recognized as an important factor in the formation and maintenance of species (Darwin 1859; Schluter 2009a). Among the ecological factors that may cause adaptive divergence to result in reproductive isolation is immigrant inviability (Nosil et al. 2005). Immigrant inviability results when populations in different environments are locally adapted, so that immigration to the "wrong" habitats results in increased mortality. The effects of this screening process may depend on the life-history traits of the species being studied. In species with fast generation times, such as insects and annual plants, immigrant inviability can contribute to reproductive isolation (Nosil et al. 2005). Much less is known about this process and speciation in general in long-lived organisms, which often provide the foundation of species-rich ecosystems.

On coral reefs, the most diverse marine ecosystems, such long-lived foundation species include corals, sponges, and sea fans. Like many other reef-dwelling organisms, the life cycle of these animals includes a planktonic larval stage with great dispersal potential (Hellberg 2009) and sister taxa and even more species-rich clades often share the same geographic range (Fukami et al. 2004; Taylor and Hellberg 2005). The co-occurrence of closely related species with high dispersal capabilities creates a challenge for evolutionary biologists: How can new marine species emerge without obvious geographic isolation? Caribbean coral reefs in particular provide several examples of young clades that have diverged more recently than their immediate cousins in the Eastern Pacific (Taylor and Hellberg 2005). Such species appear to have been generated by processes acting within the relatively small (∼6 million km2) confines of the Caribbean.

One explanation for within-Caribbean speciation is that geological processes (e.g. sea level fluctuations) may have transiently divided populations long enough for new species to form. Recently diverged sister species may have then expanded their ranges when conditions allowed and come into secondary contact. For example, the rise of the Isthmus of Panama altered ocean circulation, causing novel conditions in the Caribbean basin that generated both widespread extinction and speciation in Caribbean mollusks and corals (Jackson et al. 1993), species that later expanded their distributions to the entire basin. Geographic isolation is also evident between some populations of widespread taxa at large (1000 km) (Baums et al. 2005) and small (< 100 km) spatial scales (Taylor and Hellberg 2003; DeBiasse et al. 2010). Thus, new species may result from historic geographic isolation within the Caribbean.

* This Chapter was published as: Prada C, Hellberg M (2013) Long pre-reproductive selection and divergence by depth in a Caribbean candelabrum coral. Proc. Natl. Acad. Sci. USA 110: 3961-3966.

6 However, ecological segregation is also widespread on Caribbean coral reefs. Genetic differences have been detected between ecomorphs with overlapping ranges (Brazeau and Harvell 1994; Carlon and Budd 2002; Rocha et al. 2005; Duran and Rutzler 2006; Prada et al. 2008), thus habitat diversity may aid marine speciation. For example, the most common coral on Caribbean reefs was regarded as one species, but fertilization, genetic, and morphological data suggest it is a complex of three species segregated in part by depth (Knowlton et al. 1997; Fukami et al. 2004). In fact, depth segregates cryptic species on reefs more often than any other factor (Knowlton 1993).

Candelabrum corals of the genus Eunicea constitute one such group where sister species are segregated by depth (Prada et al. 2008). Eunicea constitutes the most diverse genus of anthozoans (anemones, corals and their kin) on Caribbean reefs, with 15 species plus some undescribed forms (Sanchez 2009). Eunicea species share a mutualistic relationship with algae of the genus Symbiodinum clade B (Santos et al. 2001), which restricts them to the photic zone. All Eunicea are gonochoric and reproduce sexually by spawning gametes (Beiring and Lasker 2000; Sanchez 2009).

Eunicea flexuosa is an especially suitable Species in which to test the effects of immigrant inviability associated with adaptation to depth in a wide geographical context because 1) its geographic range spans the Caribbean (Sanchez 2009), 2) it is common (> 5 colonies/m2 (Yoshioka 2009)) and present in virtually all habitats and depths on coral reefs and rocky walls (Sanchez 2009), 3) depth-segregated colonies show extensive internal and external morphological differences (Kim et al. 2004; Prada et al. 2008), 4) transplant experiments show colonies are adapted to different depths (Prada et al. 2008), and 5) first reproduction does not occur until a size of 30 cm, at least 15 years after larval dispersal (Beiring and Lasker 2000). Such delayed reproduction during sedentary asexual growth allows selection to act for a long time before genes can be mixed (as in Willliams' (1975) strawberry-coral model) and thus may increase the intensity of immigrant inviability between depth- segregated populations.

Here, I use field observations to estimate the fraction of E. flexuosa that survive until the onset of reproduction, then measured survivorship advantage of native over foreign colonies at each depth by reciprocal transplants to show that native colonies have an advantage over foreigners, an advantage magnified by the cumulative effect of long pre- reproductive selection. I then evaluate morphological and genetic differentiation at two depths across the Caribbean range of E. flexuosa. If immigrant inviability operates between depths and is enhanced by delayed reproduction, then genetic divergence will be higher between sympatric populations separated by depth than among those at the same depth separated by geography. I found two depth-segregated lineages, each showing little evidence of subdivision among geographic populations across the Caribbean. Genetic lineages correlate well with morphology but some mismatched individuals are found at low frequency in both lineages. Previously observed adaptive phenotypic differences between depths (Prada et al. 2008) may thus provide both material for divergent selection and a mechanism for reproductive isolation in E. flexuosa. More generally, selection acting on

7 habitat-dependent alternative phenotypes may be especially likely to enhance phenotypic and genetic divergence in long-lived sessile clonal animals with delayed post-dispersal reproduction.

2.2. MATERIALS AND METHODS

2.2.1. Survivorship estimates per size class

To estimate survivorship, I recorded mortality of colonies per size class for over four years at Media Luna, the location used for the transplant experiment and genetic analysis. I monitored three depths per reef. At each depth, a replicated 20 m transect was placed and four 1m2 quadrats were randomly selected (total of 8m2 per depth). In total, I analyzed 161 quadrats (I excluded 31 quadrats with no or ambiguous E. flexuosa data).

2.2.2. Survivorship estimates across depths

I sampled survivorship advantage of native over foreign by reciprocally transplanting adult (>25cm) colonies from each depth to both shallow and deep areas. I selected 40 colonies from each of two depths (<3m shallow and >20m deep). Each colony was divided and the resulting fragments then transplanted to both shallow and deep areas. Survivorship was recorded yearly for two years for native and foreign colonies. I also estimated survivorship by pooling this current data with previous (Prada et al. 2008) estimates. Both transplant experiments were carried out in Parguera (Media Luna reef), Puerto Rico for the same colonies used in the genetic analysis (see below). I estimated survivorship for one generation by multiplying the annual survivorship rate by 38.88 (the estimated time to reach 70 cm, assuming a 1.8 cm/yr growth rate). I then estimated the selection advantage as the survivorship of one lineage over the total survivorship of both lineages. For example for the shallow area, the advantage of the Shallow lineage is: Shallow / (Shallow + Deep). I then assume a 1:1 settlement and estimate Hardy-Weinberg proportions.

2.2.3. Population genetic sampling

I sampled populations in the Bahamas, Panama, Puerto Rico, and Curaçao for genetic analysis. At each location, I sampled 17-38 adult colonies (> 50 cm) from each of two depths (Table S2.3, Appendix A). The sampling scheme spans regions where geographical patterns of genetic differentiation have been seen previously, including Exuma Sound and Mona Passage (Baums et al. 2005; Taylor and Hellberg 2006). I collected adult (> 50 cm) colonies at least 5 m apart to decrease the sampling of clone-mates and preserved them in 95% ethanol at -20º C. I extracted genomic DNA from these samples using the QIAGEN DNeasy Kit following the manufacturer's protocols.

2.2.4. Generation of sequence data

As a mitochondrial marker, I sequenced the MSH region, which encodes an ORF unique among animals to some anthozoans, using published primers (France and Hoover 2001). I

8 also developed three new nuclear sequence markers (Table S2.7, Appendix A) using sequences generated from a partial 454 run. Briefly, I extracted total RNA from a shallow individual from Puerto Rico using Trizol (Chomczynski and Sacchi 1987). I then produced cDNA by reverse transcription of the RNA following (Hale et al. 2009), using a combination of ClonTech SMART RT and Invitrogen SuperscriptII. The resulting PCR was verified in 1% agarose gel with fragments ranging between 500 - 4,000 bp. This PCR cDNA was fragmented by sonication and pyrosequenced at the UCLA Genomic Center. I generated a total of 27,197 reads with an average read length of 305 bp. Reads were assembled into 3685 contigs and 12,457 singletons using CAP3 (Huang and Madan 1999). Annotation with the Swiss-Prot database identified 3344 homologous sequences. I selected 20 different genes to make primer pairs. First, I aligned the selected E. flexuosa sequence against the closet homologue in the Nematostella (anemone) genome to identify the possible placement of introns. I then designed intron-spanning primer pairs anchored in coding regions using Primer 3 (Rozen and Skaletsky 2000). From the initial 20 candidates, three were selected because they amplified across samples with a desirable sequence size of 350- 700 bp. Initial screening showed these markers were single copy (≤ 2 alleles per individual) and variable within populations.

PCR amplifications were performed in a Bio-Rad T100 with the same cycling conditions (except for the annealing temperature) for all genes: an initial denaturation cycle of 3 min at 95°C, 2 min annealing at 50-56°C and 2 min extension at 72°C; followed by 38 cycles of 30 sec at 95°C, 45 sec at 50-56°C and 45 sec at 72°C and a final extension cycle at 72°C for 10 min. Amplicons were directly sequenced in both directions in an ABI 3100 using BigDye chemistry v 3.1 and the amplification primers. I resolved indel heterozygotes containing a single indel using CHAMPURU (Flot 2007). I resolved haplotypes using PHASE v2.1 (Stephens et al. 2001). I used 90% probability as cutoff, because detailed cloning studies have shown that haplotypes inferred by PHASE with a probability >70% accurately reflect haplotypes (Harrigan et al. 2008). I cloned individuals using the Invitrogen TOPO TA Cloning Kit and sequencing at least 5 clones per reaction. I deposited all sequences in Genbank (KC310499 - KC310687).

2.2.5. Analysis of genetic data

I aligned sequences with Clustal X (Larkin et al. 2007) and inferred substitution models for each marker using jModelTest (Posada 2008). Using the suggested model of evolution, I tested for intra-locus recombination using both GARD and SBP as implemented in Hy-Phy (Pond and Frost 2005; Pond et al. 2006). I also inspected for recombination using the difference of sums of squares (DSS) method, with a sliding window of 100-bp and 10-bp step size as implemented in TOPALi version 2 (Milne et al. 2004).

A parsimony haplotype network was constructed for each marker using the Templeton et al. algorithm (Templeton et al. 1992) implemented in TCS 1.21 (Clement et al. 2000). Each network was constructed with confidence level set at 95% and excluding gaps. Hierarchical genetic subdivision was analyzed using the AMOVA framework as implemented in GENODIVE (Meirmans and Van Tienderen 2004). AMOVAs were

9 performed with 10,000 permutations using conventional F-statistics. I first defined populations by depth and location. To test for geographical differentiation within lineages, I then used clustering information from the STRUCTURE analysis (see below) to classify individuals by lineage (i. e. Shallow or Deep). I calculated both AMOVA and Pairwise FST between locations within each lineage was calculated in GENODIVE (Meirmans and Van Tienderen 2004). FST values were plotted against pairwise geographical distances among populations calculated in Google Earth 6.2 using the shortest nautical distance among populations.

Sequences were recoded as frequency data (codominant) in DNAsp 4.0 (Rozas et al. 2003) and used to infer population subdivision using Bayesian clustering in STRUCTURE (Pritchard et al. 2000). I used the admixture model with a burnin of 100,000 steps followed by 20 million iterations and 10 replicates per run. I ran STRUCTURE without information of the origin of each individual, thereby reducing potential biases. I inferred population structure using a range of K's (number of inferred populations) from 1 to 8 (maximum number of populations). I then used STRUCTURE HARVESTER (Earl and vonHoldt 2011), which implements the Evanno method (Evanno et al. 2005) to estimate the most likely K based on the difference of likelihood scores. Because STRUCTURE uncovers only the deepest genetic subdivision (Evanno et al. 2005), I hierarchically ran the program within each subsequent cluster.

To test the extent to which depth, geography, and their interaction have played roles in lineage splitting and to recover the evolutionary history of all populations, I used a combination of gene trees/species tree approaches and model-based inference approaches (Johnson and Omland 2004; Pigliucci and Kaplan 2006; Anderson 2008). In STEM, I iterated the process 5 times, so that in each round gene genealogies were re-estimated, and species tree likelihood scores re-calculated. I then used the average likelihood scores across replicates to infer information theory statistics and rank all possible topologies following Anderson (2008) and Carstens and Dewey (2010). In STEM, I used clustering information from the STRUCTURE analysis to classify individuals and included only individuals with assignment probabilities > 90%; individuals with lower probability may be the result of inter-specific gene flow, which may introduce bias.

I ran IMa first using broad priors (q1, q2, qA = 20, m1, m2 = 10, t = 10) and generated more specific priors (q1, q2, qA = 3, m1 = 1, m2 = 3, t = 5) that encompassed the entire distribution of parameters in subsequent runs. I ran IMa two times starting from different seeds, each run with 20 chains, a heating scheme of -g1 0.8 -g2 0.9, and 2 - 4 million steps (2 million burn in). Once runs converged I used the L mode (load trees) to estimate likelihood scores for all nested IM models. I repeated this process at least two times for a dataset including all samples divided by lineage as suggested by STRUCTURE, habitat and morphology (I used 2 mm spindle length as a cut off). I also, run IMa for each location for each grouping (lineage, habitat and morphology). To convert IMa effective size estimates into demographic units I used the size at first reproduction (30 cm). Assuming a conservative annual growth rate of 2 cm, the generation time for E. flexuosa is 15 years.

10 Estimates based using other reported (Yoshioka and Yoshioka 1991; Beiring and Lasker 2000; Prada et al. 2008) growth rates fall within the reported 90% high posterior density range (Table S2.6, Appendix A).

I present the results of such analysis on Table S2.6 (Appendix A) and Figure S2.2 (Appendix A). Migration rates should be interpreted with caution. I present data from Bahamas, Curaçao and Panama to show that the results hold for most comparisons but recognize that the sampling design may be inappropriate to test for asymmetrical migration at these locations. In Bahamas and Curaçao I suspect the deep habitat may be deeper, while the uniqueness of Panama and the blur in the distinctness of morphologies may require a larger sampling at this locale.

I suspect that only in Puerto Rico I sampled the opposite ends of the depth gradient and thus the asymmetry in gene flow can be more rigorously tested. Puerto Rico is the most depth-segregated location, thus estimates derived from it eliminate any sampling artifact derived from the clinal distribution of both lineages. I then used model-based selection to calculate evidence ratios and rank all possible models (Anderson 2008). Because I reject a strict isolation model, I used a full model to estimate demographic parameters.

I inferred gene genealogies and estimated likelihood scores for species trees using STEM 2.0 (Kubatko et al. 2009). I manipulated the settings file to estimate the likelihood score for the best tree for each of three arrangements: 1) populations divided by geography (4 lineages), 2) populations divided by depth (2 lineages), and 3) each population as an independent lineage (8 lineages). I then used the average likelihood scores across replicates to infer information theory statistics and rank all possible topologies (Anderson 2008).

I used IMa, the isolation with migration model (Hey and Nielsen 2004), and model based inference (Anderson 2008) to evaluate all possible scenarios of divergence. My aim was to compare the support for strict isolation (zero migration) of the Shallow and Deep lineages (as delineated by STRUCTURE) to that for a model of divergence with gene flow. I also evaluated the IMa model when individuals were partitioned by habitat and morphology. I then used model-based selection to calculate evidence ratios and rank all possible models (Anderson 2008).

2.2.6. Morphological measurements

I used 5% bleach to clear the organic surface from colonies, washed the colonies twice, and dried them in ethanol. I took digital images of each colony through a Diagnostic Instrument SPOT RT Slider CCD camera attached to a Leica MZ7 stereomicroscope. To compare populations and understand the relative importance of different factors, I fit spindle length to a generalized linear model. I used a Gaussian distribution and geography, depth, and lineage as factors. I constructed all 17 possible models from these factors and used model based approaches based on AIC scores (similar to the STEM analysis) to rank each model

11 (Anderson 2008). I also used a HSD Tukey test to identify which comparisons among locations within lineages were significant. Analysis was performed with raw measurements and log transformed data with similar results.

2.3 RESULTS

2.3.1. Survivorship to reproductive size is low in shallow areas- I estimated mean annual survivorship rates (averaged over four years) for colonies divided in 3 size categories (<10 cm, 11-30 cm and >31 cm). Survivorship rates range from 69.1% for the smallest to 90.1% for the largest category (Table S2.1, Appendix A). Based on these values, the proportion of individuals (Figure 2.1) that would survive to first reproduction (30 cm) is 1% and only 0.01% would reach full maturity (70 cm).

Age (yr) 5 10 15 20 25 30 35 40 100 Full 10 maturity 1

0.1 0.01 Age at firstt Percent surviving reproductionon 0.001 10 20 30 40 50 60 70 880 Size (cm) Figure 2.1. Age/size survivorship curve for Eunicea flexuosa (log scale). Age at first reproduction (Beiring and Lasker 2000)is shown by the dotted line and full reproductive maturity with the grey bar. Age calculated based on a 2 cm/yr growth rate (Prada et al. 2008).

2.3.2. Selection is strong between habitats

I selected 40 adult colonies from shallow and 40 from deep habitats, divided them, then transplanted the resulting clones (>25 cm) to both depths. I estimated annual survivorship rates based on those observed for the two years following transplantation and by pooling data with previous estimates (Prada et al. 2008). Mean annual survivorship was highest (Table S2.2, Appendix A) for colonies transplanted to the same depth (87.1% and 94.0% for shallow and deep, respectively) and lowest for reciprocally transplanted colonies (78.5% and 63.0% for colonies transplanted to shallow and to deep, respectively). I estimated an annual increase in survivorship of native colonies over foreign ones of 9.9% in shallow areas and 33.0 % in deep areas. Over the span of a generation (about 30 years), the survivorship advantage would be 98.3% for shallow areas and 99.9% for deep. Assuming a 1:1 ratio of settlers and Hardy-Weinberg proportions, I expect 96% of the mating events in shallow habitats to be between Shallow x Shallow, with only 4% between Shallow x Deep (hybrid crosses). In deep habitats, even fewer hybrid crosses are expected, 1 in a million.

12 To compare the selection estimates with other studies, I translated them into difference in fitness between native and foreign colonies as in (Hereford 2009). There are increases of 180% and 190% in fitness of native over foreigners for shallow and deep habitats, respectively. These estimates are based on a growth rate (1.8 cm/yr) at the high end of those seen for E. flexuosa (Prada et al. 2008), the size (70 cm) at which colonies significantly contribute to reproduction (Beiring and Lasker 2000), and adults showing the highest survivorship rates. Estimates based on faster growth rates, reproduction at younger ages, and survivorships from younger colonies result in greater asymmetry in the selective advantage between habitats.

2.3.3. Two genetic lineages are segregated by depth

I sampled populations from four locations: the Bahamas, Panama, Puerto Rico, and Curaçao (Appendix A, Table S3). At each location, I sampled adult colonies (> 50 cm) at two depths, as earlier work in Puerto Rico suggested deep and shallow populations might be genetically differentiated (Prada et al. 2008).

All analyses of genetic variation indicated unambiguously that there are two lineages within E. flexuosa that segregate between deep and shallow depths. Patterns of genetic subdivision revealed by STRUCTURE showed strong support for two clusters (Figure 2.2Ab) that correlated closely with depth (Figure 2.2A). The probability of membership to either cluster was high (> 90%) for most colonies, although it dropped to 89-60% for 20 individuals (11%), perhaps as a result of hybridization between lineages. Similarly, analysis of molecular variance showed that depth explains most of the variation for both mitochondrial and nuclear markers (61% and 71%, respectively). Less variation was explained by geography (7% and 1% for mtDNA and nuclear markers, respectively) and within site variation (31% and 27% for mtDNA and nuclear markers, respectively). All fixation indices were significant (10,000 permutations) at the 0.01 level for both types of markers. Although haplotype networks for each gene did not sort completely (Figure S2.1, Appendix A), they revealed two distinct lineages that largely correlate with depth (at least for MSH and I3P). For EF1A and CT2, haplotypes sampled from different depths mixed, but clinal variation was still evident.

To test the extent to which depth, geography, and their interaction have played roles in lineage splitting and to recover the evolutionary history of all populations, I used a combined gene trees/species tree (STEM) (Kubatko et al. 2009) and model-based inference approach. The Shallow and Deep lineages were well supported by STEM. The tree with the best AICc score showed two lineages: one composed of all Shallow samples and one of all Deep (Figure 2.2C). When I further split populations by depth and geography, the best tree still joins populations by depth, but was 218 times less likely than the all Shallow and all Deep tree. Conversely, if I grouped populations by geography (i.e. combine Panama Shallow and Panama Deep), the tree with the highest likelihood score was > 1000 times worse than the best tree.

13 A B a 1 Bahamas 0.88 0.66 shallow deep 0.44 Puerto Rico 0.22 membership Probability of 0 ShallowSh ll DDeep ShallowShS ll DDeepp ShallowShS ll w DeepD ShallowSSh llow Deep D p shallow deep Puerto Rico Panama Bahamas Curaçao c b 1 Curaçao 4 0.5 3 Panama shallow deep 2 0 1 Shallow lineage only d shallow deep 0 1 Delta K (*1000) 1234 9°N 15°N 21°N 27°N K 0.5 Probability of membership 0 83°N 77°N 71°N 65°N Deeppgy lineage only C Grouped by depth Each population split Grouped by geography AICc = 1135.74 AICc = 11946.51 AICc = 12108.10

Bahamas Deep Bahamas All Deep Panama Deep Curaçao Deep Puerto Rico Deep Panama Puerto Rico Shallow Puerto Rico Bahamas Shallow All Shallow Panama Shallow Curaçao Shallow Curaçao 218 X > 1000 X

Figure 2.2. (A) Graphical summary of Bayesian clustering (a) STRUCTURE results for all samples combined. (b) Likelihood score differences (ΔK) as a function of K (number of clusters), with highest difference between K=2 and K =3. Separate runs for the (c) Shallow and (d) Deep lineages revealed in (a) show no further subdivision. (B) Proportion of individuals from shallow and deep habitats falling into the Shallow (green) and Deep (red) clusters revealed by STRUCTURE. Two individuals with ancestry probability lower than 70% were excluded. The size of the circle is proportional to the number of sampled individuals. (C) STEM results for each of three data arrangements: by depth, by depth and geography, or by geography. Corrected Akaike Information Criteria scores are indicated. Black arrows denote ratios between alternative topologies from information theory.

Depth segregation was significant (Fisher’s exact test, P < 0.0001) but the degree of segregation varied among sites (Figure 2.2B). In Puerto Rico and Bahamas it was high, with > 80% of colonies of each lineage segregated into either shallow or deep habitats. In Curaçao, the Shallow lineage dominated (100%) shallow environments, but was also common (50%) in deep areas. Thus, while the two lineages segregate over a depth gradient, the absolute depths defining the gradient vary among localities. In Puerto Rico my sampling captured the ends, but in Curaçao and the Bahamas the corresponding deep habitat may be at greater depths.

Panama was an extreme case in this regard: the Deep lineage dominated both deep and shallow environments, with the Shallow lineage present at only modest levels in shallow environments (even after adding an extra shallow collecting site). This Panamanian oddity

14 was not unexpected: particulate matter around Bocas del Toro is high, which increases food for suspension feeders and attenuates light, allowing deep-water suspension-feeders such as black corals to occur at depths as shallow as 5 m (Opresko and Sanchez 2005), whereas in other locations they are not visible until 25 m (Opresko and Sanchez 2005).

2.3.4. Little geographical differentiation within lineages

In contrast to the striking pattern of depth segregation between lineages, there was little evidence of geographical differentiation within lineages. Figures 2.2Ac and d show STRUCTURE plots for each lineage after samples were assigned to either cluster from the global analysis. No clear clusters were visible in either. A within-lineage analysis of molecular variance showed that geography significantly explains just 3% and 11% of the variation for the Shallow and Deep lineage, respectively. A pairwise FST analysis found that some within-lineage geographical comparisons were significant (Table S2.4, Appendix A), but the magnitude of these differences was not large.

2.3.5. No strict allopatry and asymmetric migration from Shallow to Deep

The rank of all Isolation with Migration models by the information theory approach suggested that migration has occurred during the divergence of the Shallow and Deep lineages (Figure 2.3). The best model had three parameters for population size and two for migration, regardless of whether I classified shallow and deep individuals by their lineage, morphology, or habitat (Table S2.5, Appendix A). Models without migration parameters (strict isolation) were 1000’s of times less likely than the best model. Migration from Shallow to Deep is higher overall and in each locality, except in Panama when samples are classified by habitat or morphology (Figure S2.2 and Table S2.6, Appendix A). This is not unexpected as in Panama morphologies are less distinct (Figure 2.4) and deep water species tend to come up shallower than elsewhere in the Caribbean (Guzmán and Guevara 1999).

The difference between the asymmetry in selection (strongest in the deep habitat) and between-lineage migration (more from Shallow to Deep) may be explained by inter-lineage differences in gamete production. In E. flexuosa, large colonies produce on average two orders of magnitude more gametes than smaller colonies (Beiring and Lasker 2000). Deep colonies are on average less than half the size of Shallow colonies and have half their polyp density (Kim et al. 2004; Prada et al. 2008). Deep lineage colonies are also less abundant than Shallow across the Caribbean; in Puerto Rico their densities are one-third those of Shallows (Yoshioka 2009). Thus by the time reproduction occurs the amount of Shallow lineage gametes may greatly outnumber those from the Deep lineage. This asymmetry in gamete production may result in asymetrical introgression, as previously shown in oaks (Lepais et al. 2009).

15

8 From Shallow-to-Deep 6 From Deep-to-Shallow

4 Probability 2

0

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Migration (m/μ)

Figure 2.3. Migration rate estimates between Shallow and Deep obtained by fitting the IMa model to all 4 loci. Estimates are scaled by the neutral mutation rate. Samples were partitioned by morphology (excluding Panama). The number of migrants between lineages per generation (Nm) is 0.4 (90% HPD interval 0.147 – 0.756) from Shallow to Deep and 0.078 (90% HPD interval 0.02 – 0.189) for Deep to Shallow.

Figure 2.4. Mean spindle length between Deep (black) and Shallow (light gray) lineages in different populations. Letters indicate significant differences for within lineage comparisons. Different letters indicate significant differences in a HSD Tukey test after Bonferroni correction. D for Deep and S for Shallow.

16 2.3.6. Lineages match phenotypes at local scales

To assess morphological divergence among populations, I measured the length of spindles, the calcium carbonate particles that provide structure to the colony. Spindle length correlates well with other morphological characters (e.g. calice size, inter-calice distance, branch thickness) and is the most divergent morphological character between deep and shallow colonies of E. flexuosa in Puerto Rico (Prada et al. 2008).

Patterns of morphological differentiation mirrored those seen genetically; lineage was a better predictor of spindle length than either geography or habitat (Table 2.1). All of the four best models had lineage as a variable, along with its interaction with either geography or habitat. When I excluded lineage or its interactions as variables, the remaining models were 100 times less likely than the best model with lineage included. Lineage fit the data better than depth, despite the two being highly correlated (Table 2.1). While morphology and genetics match well, mismatched individuals occur at low frequency (2 of 97 and 7 of 92 for Shallow and Deep, respectively).

Table 2.1. Model ranking for spindle length analysis using a general linear model.

Model AIC ∆i ML wi Evidence Ratio GDL L*G 165.56 1.00 0.61 GDL D*L G*L 167.31 1.75 0.42 0.25 2 GDL D*L G*D 170.47 4.91 0.09 0.05 12 GDL G*L G*D 171.08 5.52 0.06 0.04 16 Full model 171.08 5.52 0.06 0.04 16 GDL D*G 175.15 9.59 0.01 0.01 121 GL G*L 185.25 19.69 0 0 18864 All other models > 188 > 20 0 0 > 105

G = Geography; D = Depth; L = Lineage assignment from STRUCTURE Clustering for K = 2; ML = Maximum likelihood; AIC = Akaike information criterion inferred from thegeneralized linear model fitting; ∆i = Difference in AIC score with respect to the best model; Model Likelihoods = Relative likelihood of the model given the data; wi = Model probabilities; Evidence Ratio = Fold-difference in model probabilities against the best model.

Once lineage is accounted for, some differences between locations remained (Figure 2.4). For the Shallow lineage, colonies from Curaçao had larger spindles than colonies in Puerto Rico. For the Deep lineage, colonies in Panama had smaller spindles than for all other locations. This difference in spindle length in Panama made the morphological boundary between Shallow and Deep narrower and shifted shallower than in other locations. Thus, morphologies were diagnosable, correlated well with lineage, and segregated into shallow and deep habitats at local scales, but lineage by depth by locale interactions generated morphological overlap between taxa when populations were considered together.

17 2.4 DISCUSSION

I found that 1) selection is strong in both habitats for native over foreign colonies, 2) colonies of the candelabrum coral Eunicea flexuosa occurring in shallow and deep habitat are segregated into two genetically and morphologically divergent lineages, 3) genetic divergence is weak among geographic samples within each lineage, 4) the Shallow and Deep lineages have exchanged genes since their initial divergence, and 5) migration between them has been higher from Shallow to Deep. Reciprocal transplantation suggests divergence between colonies found at different depths is adaptive and selection is strong enough to maintain the separation of the two lineages at the ends of the depth gradient. This divergent selection may by itself result in reproductive isolation via immigrant inviability (poor performance and viability of larvae settling at the wrong depth) and temporal isolation (spawning at different times) (Levitan et al. 2004). I suggest divergence between Shallow and Deep lineages of E. flexuosa is driven by adaptation to different conditions at different depths (Schluter 2009a), facilitated by developmental plasticity (West-Eberhard 2003; Fitzpatrick 2012) and prolonged opportunity for depth-specific selection prior to reproduction.

2.4.1. Ecological gradients, adaptive divergence and speciation

Divergent selection along depth gradients is generated by associated factors such as the quantity and quality of light, the force of waves and currents, sediment load, the distribution of predators and mutualists, and the availability and composition of food. In response to this variation in environmental conditions, closely-related species or populations found at different depths have been found to differ in their morphologies (Fukami et al. 2004; Prada et al. 2008), feeding strategies (Lasker et al. 1983; Lesser et al. 2010), metabolic rates (Lesser et al. 2010), symbiotic relationships (Kirk et al. 2009), resistance to predators (West et al. 1993), and spawning times (Knowlton et al. 1997; Levitan et al. 2004). In E. flexuosa's close relative Plexaura homomalla (Eunicea and Plexaura are paraphyletic), depth-segregated sister species increase heterotrophy and decrease photosynthesis as depth increases (Lasker et al. 1983).

Such depth-related divergence seems to be correlated with the differentiation of closely related species, as in bryozoans (Jackson and McKinney 1990). Depth gradients in light are hypothesized as a cause of speciation by sensory drive in cichlids (Seehausen et al. 2008) and factors associated with depth appear to have driven speciation in rockfish (Ingram 2011). Among marine sibling species reported by Knowlton (1993), over half involved depth as a dividing factor, even though not all comparisons included depth. Thus, adaptive divergence by depth is consistent with a role for depth in maintaining and generating species differences in marine and aquatic settings. These patterns associated with depth are analogous to those seen for other environmental gradients, such as altitude, salinity, or temperature. Such gradients generate adaptive divergence in vertebrates (Whitehead et al. 2011), invertebrates (Sanford and Kelly 2011) and plants (Hodges and Arnold 1994).

18 Because such abiotic factors generate adaptive divergence across multiple species, they also generate gradients in biotic interactions (such predator-prey interactions) (Freeman and Byers 2006), creating additional axes of complexity.

2.4.2. High gene flow within species

Divergence between depths at small scales (as little 76 m separate deep and shallow populations I sampled) in E. flexuosa contrasts with the high connectivity seen among populations within lineages across 1000s of km. Such high connectivity is not uncommon to marine species with planktonic larval dispersal (Hellberg 2009), ranging to extremes in which gene flow spans the broadest expanses of the Pacific Ocean (Lessios and Robertson 2006). Thus adaptive divergence occurs here despite the potential for high levels of gene flow (Whitehead et al. 2011). Such adaptive divergence under high gene flow generates a pattern where neutral genetic divergence is primarily partitioned by habitats, with little geographic structure. Ecotypes are connected across vast geographical spans, thus producing a pattern of single ecological segregation as seen in E. flexuosa and shared by other broadcast spawners such as the star coral Montastraea annularis (Fukami et al. 2004), some coral reef fishes (Rocha et al. 2005), and the Hawaiian limpets Cellana (Bird et al. 2011).

In contrast, under low gene flow, genetic divergence first occurs geographically and then ecologically. The pattern has been detailed in the marine snail Littorina saxatillis (Johannesson 2001) and is similar to those in other marine brooders where ecology has been proposed to drive speciation, such as the corals Favia fragum (Carlon and Budd 2002) and Seriatopora hystrix (van Oppen et al. 2011). Clonal brooders also tend to reproduce at younger ages (smaller sizes) (Szmant 1986), increasing the role of geography in local adaptation. This difference between brooders and broadcasters has parallels in winged and wingless phytophagous insects. In pea aphids, segregation by host plant occurs at large geographical scales with little geographical differentiation (Ferrari et al. 2012). Conversely, in walking sticks, pairs of habitat-segregated ecotypes have occurred multiple times (Nosil et al. 2005). Thus the partition of genetic diversity during ecological speciation depends on the dispersal biology of the studied species.

2.4.3. Divergence between habitat-segregated populations with prolonged pre- reproductive selection

Ecological filtering can help maintain species differences and contribute to reproductive isolation of populations occurring in different habitats (Nosil et al. 2005). In E. flexuosa, adult colonies native to their depth survive far better than do foreign colonies, an advantage enhanced by the cumulative effect of many years of pre-reproductive selection. For a settling larva to reach adulthood, it must pass a series of ecological challenges until it reaches a safe size of at least 30 cm (> 15 years) and begins reproductive activity. Significant genetic contributions to the population likely take far longer, however, not until attaining a colony size of 70 cm (> 30 years), because such large colonies contribute 98% of all eggs in E. flexuosa (Beiring and Lasker 2000).

19 The difference in fitness between native and foreign colonies that I found here is three- times higher than any reported for similar transplant experiments (Hereford 2009). It also exceeds values used in mathematical models of ecological speciation in which adaptive divergence is sufficient to generate reproductive isolation in discrete habitats (Giraud et al. 2010; Thibert-Plante and Hendry 2011) or along clines (Clarke 1966; Endler 1977). The long period between when a larva settles and when it becomes a colony that contributes genetically to the next generation thus decreases gene flow between depth-adapted lineages and enhance the opportunity for habitat-specific selection and divergent adaptation. Such immigrant inviability may be common in slowly maturing sessile species with broad dispersal (like some scrubs and trees (Savolainen et al. 2006; Alberto et al. 2010)). In contrast, when adults are the primary dispersers, immigrant inviability would have to be exceptionally strong to have such an effect, as migrants can reproduce as soon as they settle, allowing locally adapted genotypes to mix with maladapted foreign ones.

Strong divergent selection between habitats could also stifle the early stages of divergence if dispersal is limited and a narrow range of phenotypes limit opportunities to colonize new habitats (Fitzpatrick 2012). In E. flexuosa, however, larval dispersal is high (Figure 2.2) and colonies phenotypically plastic (Prada et al. 2008). Temporal isolation may also emerge from adaptation to different depths and serve to further isolate depth-segregated populations. As for flowering plants segregated by altitude (Hodges and Arnold 1994; Hall and Willis 2006), marine broadcasters segregated by depth often differ in their time of spawning, decreasing the probability of cross-fertilization between depth-segregated colonies (Knowlton et al. 1997; Levitan et al. 2004). Before spawning events, colonies in deep areas of E. flexuosa lack mature eggs compared to shallow ones (Kim et al. 2004). One hour difference in the timing of spawning is sufficient for sperm to dilute to the point where crossbreeding is unlikely in other corals (Knowlton et al. 1997; Levitan et al. 2004). While the exact traits that affect ecological fitness are unknown, population segregation by depth thus has the potential to increase reproductive isolation by pleiotropic interactions (Servedio et al. 2011; Fitzpatrick 2012) between fitness traits and those that control timing of reproduction , increasing the chance of genetic divergence in depth-segregated populations adapted to different ecological conditions.

2.4.4. Speciation by prolonged selection on adaptively plastic habitat-specific contrasting forms: Divergence in the sea

Data on E. flexuosa show how a developmentally plastic organism with ecologically variable phenotypes may diverge genetically to produce different species without geographic isolation, even in groups with widely dispersing propagules. Habitat-specific alternative phenotypes, whether discrete or at the ends of a cline, represent different habitat-mediated sets of expressed genes (Moczek et al. 2011). To the degree that they are independently expressed, they are independently exposed to habitat-specific selection and can, given genetic variation in their form (Waddington 1942; Moczek et al. 2011), diverge genetically without reproductive isolation (West-Eberhard 2003). Due to genetic accommodation or assimilation of the contrasting phenotypes (Waddington 1942; West-Eberhard 2003), they may diverge genetically such that migrants between habitats become effectively inviable,

20 as suggested by my data and as hypothesized for speciation between alternative phenotypes (West-Eberhard 2003; Pfennig et al. 2010) and in clines (Clarke 1966; Endler 1977). A broader analysis of the evolution of plasticity in the genus Eunicea would provide a framework to further test this hypothesis.

As I discuss here, habitat-related divergence may apply especially in organisms with a long period of pre-reproductive selection, which affords an increased “opportunity for selection” (Crow 1961), and in organisms that are resident for a prolonged period in a single place. In sessile clonal animals, independent selection on condition- (depth-) sensitive phenotypes is enhanced because immotile colonies cannot escape the selective factors of a given stable habitat. Selection is intense in juveniles and decreases exponentially with colony size, so that clonal organisms delay (for a couple of decades in E. flexuosa) sexual reproduction and allocate all resources to vegetative growth (Williams 1975). Even once these younger colonies begin to reproduce, most gametes in a population will still come from the large, old (over half a century), well-adapted colonies that dominate sexual reproduction. These conditions characterize E. flexuosa and other sessile clonal organisms, such as anthozoans, sponges, long-lived trees (e.g., oaks and pines), and shrubs (e.g. Ceanothus) (Connell 1973; Williams 1975), which invest heavily in vegetative growth prior to reproduction (Connell 1973; Williams 1975). This process of habitat-related divergence without reproductive isolation may therefore be more common than usually appreciated. It can occur whether the contrasting phenotypes are initially genetically cued (Clarke 1966; Endler 1977) or environmentally induced (West-Eberhard 2003; Prada et al. 2008; Scoville and Pfrender 2010), for it is the contrast between the different phenotypes and the underlying sets of expressed genes that, under selection, drives divergence - not the cues (genetic or environmental) that govern their expression.

Taken together, the combination of habitat-responsive plasticity, long pre-reproductive selection, sessile residence in distinctive habitats, and the potential for temporal separation of reproduction in different habitats, exemplified in E. flexuosa, may help solve the long- standing question of how speciation can occur in marine populations with widely dispersing larval forms.

21 Chapter 3 Cryptic Diversity Hides Host and Habitat Specialization in a Gorgonian-algal Symbiosis*

3.1 INTRODUCTION

An obligate association between cnidarians and unicellular algae of the genus Symbiodinium forms the foundation of coral reefs, the most diverse marine ecosystem. In these nutrient-poor environments, algae provide organic compounds and energy to their coral hosts, while the coral provides recycled inorganic nutrients and access to a stable light environment for the alga (Muscatine et al. 1981; Muscatine et al. 1984; Colombo- Pallotta et al. 2010). This endosymbiosis is physiologically controlled by both partners (Muscatine et al. 1983; Davy et al. 2012). At the genomic level, the coral’s functional genome is uniquely shaped by different Symbiodinium species (DeSalvo et al. 2010). At the extreme, incompatible symbionts activate immune responses in their hosts, even in coral species that associate with multiple distantly related phylotypes (Voolstra et al. 2009). The synchronized efforts of both partners affect their growth and thermal tolerance (Little et al. 2004; Abrego et al. 2008), so that Symbiodinium can facilitate coral acclimation to heterogeneous environments (Chen et al. 2005; Sampayo et al. 2007), and even to warmer oceans (Rowan et al. 1997; Baker 1999; Rowan 2004; Berkelmans and van Oppen 2006; Sampayo et al. 2008).

Given the functional and phylogenetic diversity within Symbiodinium, corals may acclimate through symbiont acquisition. Once known as a single species, nine major clades of Symbiodinium are now recognized, with inter-cladal genetic distances at rDNA markers that far exceed those among bird families (Rowan and Powers 1992; Rowan and Knowlton 1995; Coffroth and Santos 2005; LaJeunesse et al. 2010; Pochon and Gates 2010). Lineages within these clades are often segregated by habitat (usually depth) (Rowan and Knowlton 1995; Sampayo et al. 2007) and geography (LaJeunesse et al. 2010). Corals can exploit functional differences among these Symbiodinium lineages either via facultative associations that may vary over a host's lifetime or species-specific coupling, which results in coevolution of corals and dinoflagellates.

At deep evolutionary scales (i.e. family level), corals may associate with more than one highly divergent lineage of dinoflagellates in the genus Symbiodinium, with some coral species harboring multiple symbiont phylotypes within a single colony (Rowan and Knowlton 1995; Little et al. 2004). At first glance, then, corals may seem flexible in their associations with Symbiodinium and able to interact with any symbiont type. The uptake of multiple genotypes of Symbiodinium is notorious early in coral development, when many Symbiodinium species can be found within a developing polyp (Poland et al. 2013).

* Prada, C., McIlroy, S. E., Beltrán, D. M., Valint, D. J., Ford, S. A., Hellberg, M. E. and Coffroth, M. A. (2014), Cryptic diversity hides host and habitat specialization in a gorgonian-algal symbiosis. Molecular Ecology. doi: 10.1111/mec.12808

22

However, while background or transient symbiont phylotypes may exist (Mieog et al. 2007; LaJeunesse et al. 2009; Silverstein et al. 2012), most adult anthozoan colonies host a single dominant algal type (Goulet and Coffroth 2003; Thornhill et al. 2009; Finney et al. 2010; Pettay et al. 2011), opening the possibility for coevolution between partners (LaJeunesse et al. 2004; LaJeunesse 2005; Finney et al. 2010; Thornhill et al. 2013; Thornhill et al. 2014). At micro-evolutionary scales, symbiont types may associate with individual anthozoan species and even intraspecific phylogeographic lineages (Santos et al. 2003b; LaJeunesse 2005; Goulet 2006; Finney et al. 2010; Pinzon and LaJeunesse 2011). These close evolutionary associations are particularly pronounced in shallow water Caribbean gorgonians, which associate almost exclusively with Symbiodinium in clade B (LaJeunesse 2002; Santos et al. 2004; Goulet et al. 2008). Recent work has verified that Clade B (as other clades) contains many cryptic species, with a high degree of host-symbiont specificity (Finney et al. 2010).

Fine scale genetic markers, such microsatellites, provide the ability to test for co-evolution at the population level (Santos et al. 2003b; Santos et al. 2004). Genetic variation and structure in host populations has been correlated with host reproductive traits (brooding vs. spawning, (Kahng et al. 2011)), while the major factors that affect Symbiodinium distribution include limited dispersal (Santos et al. 2003b; Magalon et al. 2006; Thornhill et al. 2009; LaJeunesse et al. 2010) and, in particular, light environment (Rowan et al. 1997; Finney et al, 2010). Selection pressures that affect the distribution of one or both partners can shape co-evolution, however reciprocal evolutionary change can only be detected when the genetic structure of both partners are examined across relevant spatial scales.

Eunicea flexuosa is a broadcast spawning gorgonian with larvae that acquire their symbionts horizontally from the water column (Beiring and Lasker 2000; Coffroth et al. 2001; Kim et al. 2004). Thus, relative to brooding corals with vertical transmission, E. flexuosa maintains the potential to shuffle its symbionts across generations. A microsatellite-based analysis of E. flexuosa and Symbiodinium population structure by Wirshing et al. (2013) found that, contrary to previous studies in Caribbean gorgonians (Santos et al. 2003b; Coffroth and Santos 2005; Andras et al. 2009; Kirk et al. 2009), the genetic structure of host and symbiont populations were not correlated. However, the sampling design of this study did not consider depth as a potential structuring force. Cryptic host diversity and habitat (depth) thus offer unexplored alternative explanations for the genetic variation of Symbiodinium found in E. flexuosa.

Here, I elucidate variation in Symbiodinium associated with the recently diverged Shallow and Deep lineages of E. flexuosa using four different types of molecular markers. E. flexuosa provides an excellent system for disentangling the roles of host specificity, habitat, and geography in partitioning Symbiodinium diversity because: 1) E. flexuosa is composed of a sympatric pair of recently diverged Shallow and Deep lineages (Prada et al. 2008; Prada and Hellberg 2013), 2) these host lineages co-occur at intermediate depths (Prada and Hellberg 2013), and 3) E. flexuosa lineages are widespread across the Caribbean and the Shallow/Deep distribution varies geographically (Bayer 1961; Prada and Hellberg 2013).

23 I first analyzed Symbiodinium variation at higher taxonomic levels using ribosomal ITS2 and chloroplast sequences, along with nuclear sequences to elucidate within clade B variation. I then used microsatellites to resolve population subdivision within Symbiodinium lineages at geographical scales. I also tested the stability of the host- symbiont relationships with a reciprocal transplant experiment. If coral-algal associations are mainly driven by host species identity (Finney et al. 2010; LaJeunesse et al. 2010), then colonies in reciprocal transplants will rarely switch symbiont types and host colonies will remain true to symbiont types, even when occurring outside their usual depth.

3.2 MATERIALS AND METHODS

3.2.1. Sampling and Sequence Data

Colonies were sampled at three locations: Puerto Rico, Bahamas, and Curaçao [for full descriptions of these locations see (Prada et al. 2010; Prada and Hellberg 2013)]. These locations span regions where the distributions of depth-segregated cryptic species within E. flexuosa have been previously investigated (Prada and Hellberg 2013). At each location, I sampled adult colonies (> 50 cm in height) at two depths, as earlier work suggested deep (> 20 m) and shallow (< 5 m) populations are genetically differentiated across the Caribbean (Prada et al. 2008; Prada and Hellberg 2013). At each depth, I collected tissue from between 16 and 38 colonies. To avoid sampling clones, collected colonies were at least 5 m apart. Multilocus genotypes confirmed that all colonies came from unique genets.

To disentangle the roles of host species identity and depth, I also sampled 40 colonies from intermediate depths (12 – 15 m) where the E. flexuosa lineages co-occur. In Puerto Rico I also reciprocally transplanted coral colonies to native and foreign depths (see above; 38 per depth, 152 total fragments) and re-sampled them after 18 months. Fitness across depths for these colonies was assessed previously (Prada and Hellberg 2013). I preserved all samples in 95% ethanol at -20ºC. I extracted genomic DNA from these samples using either a 2X CTAB method (Coffroth et al. 1992) or a QIAGEN DNeasy Kit following the manufacture's protocols.

Host genotypes were scored for one mtDNA marker, msh, and three nuclear markers: inositol-3-phosphatase (i3p), elongation factor 1 alpha (ef1a), and calcium transporter 2 (ct2). These multi-locus genotypes have been used previously to discern the two depth based cryptic E. flexuosa lineages (Prada and Hellberg 2013). To determine which Symbiodinium clade was found in E. flexuosa, I amplified the ribosomal ITS2 and the hypervariable region of the chloroplast 23s-rDNA following earlier protocols [(LaJeunesse 2002) for ITS2; (Santos et al. 2003a) for cp23s-rDNA]. I sequenced 224 bases of the ITS2 (see below). The cp23s-rDNA amplicons were sized in polyacrylamide gels using a metric ruler with DNA ladders as size references and alleles were named based on established clade and fragment length nomenclature (Santos et al. 2003a).

To identify cryptic lineages within Symbiodinium B1/184, the predominant symbiont type, I used DNA sequences from two chloroplast genes and one nuclear locus. I amplified and

24 sequenced samples for the two chloroplast markers following previous protocols [(Moore et al. 2003) for psbA; (Santos et al. 2002) for cp23s-rDNA]. I also sequenced a subset of samples (42) from all populations (three locations and two depths) using the flanking region of the microsatellite B7SYM15, following previous procedures (Pettay and LaJeunesse 2007). Amplicons were directly sequenced in both directions (except B7SYM15, forward only) in an ABI 3100 using BigDye chemistry v 3.1 and both amplification primers. Symbiodinium is haploid so no allele phasing was required. I edited and assembled all sequences using Geneious v4.5.5 (Drummond AJ 2009).

To further characterize genetic variation across sampling location in Symbiodinium B1/184, I genotyped all colonies at four microsatellite markers: B7SYM15, B7SYM34, B7SYM36 (Pettay and LaJeunesse 2007) and GV2 100 (Andras et al. 2009). To test for host- symbiont associations following a change in environment, I genotyped post-transplanted colonies at the B7SYM15, B7SYM34 and B7SYM36 loci. Primers and PCR conditions for these loci were done according to conditions in previous studies (Pettay and LaJeunesse 2007; Andras et al. 2009). Fragments were amplified with fluorescently labeled primers and visualized on a 7% polyacrylamide gel in a LI-COR NEN® Global IR2 DNA Sequencer (Santos et al. 2003). Microsatellites were scored by eye relative to 50-350 bp size standards (LI-COR Biotechnology Division).

3.2.2. Data Analysis

To identify the patterns of associations between Symbiodinium within-clade variants and E. flexuosa, I built a phylogeny based on the ribosomal ITS2. To root the genealogy, I included sequence data from a recent description for Symbiodinium B1 (LaJeunesse et al. 2012) (AF333510, AF333511 and AF333512) and also from the widely used culture 704 (JN558059). I estimated the model of molecular evolution using MrAIC v. 1.4.5 (Nylander 2004). I used Bayesian analyses to estimate each gene tree (Mr. Bayes v. 3.1) and estimated nodal support using posterior probabilities (Huelsenbeck and Ronquist 2001). I also constructed parsimony haplotype networks for the chloroplast markers cp23s and psbA and the nuclear B7SYM15 using the Templeton et al. (Templeton et al. 1992)algorithm implemented in TCS 1.21 (Clement et al. 2000). Each network was constructed with a confidence level set at 95% and excluding gaps.

Once genealogical patterns had been uncovered, I analyzed hierarchical genetic subdivision within the two major Symbiodinium lineages using microsatellite data. To apportion genetic subdivision among host genotype, depth, and geography, I used Analysis of Molecular Variance (AMOVA) as implemented in GENODIVE (Meirmans and Van Tienderen 2004). I performed 100,000 permutations using standard F-statistics. I defined populations by host’s genotype, depth, and geography. To further characterize geographical differentiation within the Symbiodinium group designated B1/B184, I used a Bayesian clustering approach implemented by STRUCTURE (Pritchard et al. 2000).

I used the admixture model in STRUCTURE without information of the origin of each individual. For each STRUCTURE analysis I set up a burn-in of one million steps followed by

25 5 million iterations and 10 replicates per run. I ran STRUCTURE using a variety of K’s (number of inferred populations) from the minimum (1) to the maximum (6) and then used the Evanno method (Evanno et al. 2005) as implemented in STRUCTURE HARVESTER (Earl and vonHoldt 2011) to infer the number of populations (K’s) present in the dataset. I processed replicates with best K (either 2 or 5) in CLUMMP (Jakobsson and Rosenberg 2007) using the default parameters. To generate figures I used Distruct 1.1 (Rosenberg 2004). To test for host-symbiont associations between depths, I ran STRUCTURE assuming the most likely K (= 2) for samples collected in Puerto Rico.

3.3 RESULTS

3.3.1. Eunicea flexuosa Shallow and Deep associate with Symbiodinium B1/184

Both ITS2 and chloroplast data suggest that all E. flexuosa colonies harbor Symbiodinium phylotype B1/B184, regardless of the host’s genotype, depth, or geography (Figure 3.1). Of the 183 colonies sampled, no colony contained any Symbiodinium from other clades or any other clade B types. I recovered a single ITS2 haplotype for all individuals, but multiple haplotypes for cp23s, psbA, and SYM15 (Figure 3.2). Genetic variation within these three loci revealed differences in algal composition between host colonies sampled in shallow and deep habitats (Figure 3.2). Three haplotypes were recovered from analysis of SYM15. These divided into two distinct lineages, which sorted by depth and matched the Shallow and Deep host lineages. psbA also had three haplotypes: one found solely in Shallow hosts, another solely in Deep hosts, with a third found mainly in Deep hosts but also in shallow colonies from the Bahamas. cp23s contained four haplotypes: two of high frequency that sorted by depth and host, and two endemic ones restricted to Bahamas Shallow and Curaçao Deep, respectively (Figure 3.2B).

Puerto Rico Shallow Puerto Rico Deep Curacao Shallow Curacao Deep Bahamas Shallow Bahamas Deep S. minutum (B1) S. minutum (B1) culture 704 S. psygmophilum (B2) S. muscatinei 0.01

Figure 3.1. Symbiodinium Clade B phylogeny based on ITS2 from colonies of Eunicea flexuosa sampled across depths and locations across the Caribbean. Black circles indicate posterior probability > 0.9.

26

3.3.2. Symbiodinium diversity corresponds to host lineage irrespective of depth

The AMOVA suggested that all factors (host lineage, depth, and geography) significantly influenced symbiont type, with host and depth accounting for a larger proportion of the variance than geography. In a hierarchical analysis with geography, host lineage explained 93% of the variation while depth accounted for only 64% (Table 3.1), with contributions from geography becoming insignificant. Genetic variation thus is primarily divided by host lineage.

psbA A SYM15 cp23S

n = 42 n = 183 n = 183

B Shallow

Host genotype Deep

Shallow Habitat

intermediate

deep

Figure 3.2. Genetic diversity in Symbiodinium clade B1/B184 within E. flexuosa across the Caribbean. A) Haplotype networks for the flanking region of the microsatellite SYM15 and the chloroplast psbA and cp23S. B) Distribution of haplotypes by host genotype or depth across the Caribbean for each locus. Circles are drawn proportional to the number of individuals contained and colors signify the different haplotypes.

27 Table 3.1. Analysis of molecular variance partitioning genetic subdivision within the two Symbiodinium lineages among host lineage, depths, and geography. P < 0.05 shown underlined. psbA cp23S Overall

FST % Variance FST % Variance FST % Variance Host lineage 0.611 61.1 0.692 69.2 0.652 65.2 Depth 0.497 49.7 0.58 58 0.539 53.9 Geography 0.512 51.2 0.293 29.3 0.41 41

Host lineage - Geography Host lineage 0.663 66.3 0.929 92.9 0.786 78.6 Geography 0.297 29.7 -0.014 -1.4 0.153 15.3 Depth - Geography Depth 0.399 39.9 0.64 64 0.51 51 Geography 0.353 35.3 0.033 3.3 0.205 20.5

To further disentangle the role of host lineage and depth in predicting the symbiont lineage, I sampled colonies at an intermediate depth in Puerto Rico, where the two host lineages co-occur. The results suggest that the influence of host lineage is stronger than depth: for the 40 colonies I sampled from intermediate depth, I found an absolute host- algal lineage association (Figure 3.3). In addition, two colonies mismatched to habitats for host lineage (i.e., E. flexuosa Shallow corals found in deep environments) were also mismatched for their symbiont lineage. In Curaçao for example, where about half of the colonies in deep environments are E. flexuosa Shallow, these depth-mismatched colonies hosted Symbiodinium B1/B184 Shallow symbiont. This host-symbiont association was also evident in Puerto Rico.

1' ' 0.8' ' 0.6' Coral host ' 0.4' ' 0.2'

1' ' 0.8' ' 1 2 3 5 0.6' ' Algae 0.4' ' 0.2' Probability*of*membership* shallow''''''''''''' deep' intermediate''

1 2 3 5

Figure 3.3. STRUCTURE summary results for host and Symbiodinium sampled at shallow, deep and intermediate sites in Puerto Rico.

28 3.3.3. Symbiodinium B1/184 host-depth associated lineages segregate further geographically

Bayesian clustering of microsatellite variation suggests two major gentic clusters of symbionts that largely correspond to the host Shallow and Deep lineages, respectively (Figure 3.4). Both the short and the long datasets provided similar results; I present results from the full dataset in the main text and results from the short dataset as supplementary material (Figure S3.1). When K was increased to five (K with highest likelihood), geographical differentiation within each Shallow/Deep symbiont clusters was revealed (Figure 3.4). This geographical structure within Symbiodinium contrasts with that of the host, where both Deep and Shallow lineages are apparently connected across the Caribbean (Prada and Hellberg 2013).

Figure 3.4. Bayesian clustering for all microsatellites and all samples combined for Symbiodinium B1/184. Top panel shows K = 2 as suggested by the Evanno method and lower panel shows K = 5, K with the highest likelihood.

3.3.4. Host-Symbiodinium associations were temporally stable

Of the 26 colonies for which pre- and post-transplantation Symbiodinium genotypic data were available, three (of 15) switched Deep to Shallow B1/B184 (p = 0.22, Fisher exact test two tails) and two (of 11) from Shallow to Deep (p = 0.24, Fisher exact test two tails). I also observed three (out of 26) switching to a different symbiont in the same lineage. Symbiodinium genotypes thus are generally temporally stable over a time scale of 18 months, yet not absolutely so. This within lineage symbiont change may be due to within colony variation in symbiont types, resulting from sampling different parts of the colony before and after transplantation, or a change in the frequencies of types already present in the colony prior to transplantation.

3.4 DISCUSSION

I found two closely related, yet distinct lineages of Symbiodinium clade B that associate tightly with distinct lineages of their host E. flexuosa. Although both host coral and their

29 corresponding symbiont types are associated with different depths, surveys of hosts at intermediate depths and transplant experiments demonstrate that the effect of host identity is stronger than that of depth. Divergence within each Symbiodinium lineage largely follows a geographical pattern of segregation. The tight association between octocorals and their Symbiodinium species seen here only becomes evident when cryptic host species are recognized, revealing a holobiont (coral + symbiotic algae) unit critical for understanding adaptation at each depth or geographical locale.

3.4.1. Cryptic species and coevolution

The abundance of cryptic species among both host and symbionts has led to the recognition that “ecological specialists” may dominate coral-Symbiodinium symbioses (Finney et al 2010). The discovery of correlated cryptic diversity in this octocoral and Symbiodinium association echoes coevolutionary studies in terrestrial systems, in which the use of populations of both partners, along with multiple molecular markers analyzed within a phylogenetic framework, have been critical to delineating symbiotic relationships.

For example, in cases in which the coevolutionary status of fig-wasp interactions have been questioned, finer scale genetic variation has been critical to understand the one-to-one interaction of fig trees and their pollinating wasps (Kobmoo et al. 2010). Similarly, ecotypic morphological variation in yucca plants drove the evolution of two species of pollinator moths, so that once plant cryptic variation is accounted for, the pattern of coevolution is clear in an otherwise widespread and promiscuous plant species (Godsoe et al. 2008). Perhaps more relevant, given the tight physiological interaction reflected in coral-symbiont associations, are yeast-termite (Prillinger et al. 1996) and legume-rhizobium (Heath et al. 2012) mutualisms. In both cases, revealing cryptic variation in the yeast and bacterial symbionts has exposed dozens of species that specialize on a single host and in some cases even particular habitats within those hosts (Prillinger et al. 1996; Heath et al. 2012). In rhizobium, the genomes of both partners work in concert and genetic variation within either species generate fitness differences to the plant (Heath et al. 2012). In a similar way, we are starting to understand how the functional genome of corals changes in predictable ways depending on the types of Symbiodinium they host (DeSalvo et al. 2010). Cryptic diversity present in the lineages of E. flexuosa and Symbiodinium B1/B184 may be an ecologically relevant association, with fitness costs to both for mismatching and genomic coordination that has evolved over many generations.

Failing to recognize cryptic species in either host or symbiont may lead to an underestimation of diversity, an overestimation of the ecological and physiological ranges of individual species (i.e., one generalist instead of many specialists), and a mistaken view of how species acclimate or adapt to changing conditions. Wirshing et al. (2013), for example, did not consider cryptic species within E. flexuosa and Symbiodinium, leading to their conclusion of a flexible association between E. flexuosa and Symbiodinium B1/B184 and geographic variation within E. flexuosa. They made no note of the depth of their sampled colonies, opening the possibility that geography could be confounded with cryptic species identity. This seems likely in Panama (their most divergent population), where

30 both lineages of E. flexuosa co-occur in shallow areas (Prada and Hellberg 2013). Bocas del Toro (Panama) is an atypical Caribbean location in which local sedimentation often draws organisms typical of deep water into shallow depths. Black corals, for example, can be observed as shallow as 5 m, whereas they occur > 20 m elsewhere in the Caribbean (Opresko and Sanchez 2005). Wirshing et al. (2013) also note null alleles for all loci (except Plf67) in their Panama samples. Such a pattern would be expected if the microsatellites amplify well for the shallow lineage (for which the primers were developed), but fail for colonies of the divergent deep lineage, which are at high frequency in shallow areas in Panama. Thus, when reconsidered in the light of cryptic host species, E. flexuosa's symbiont associations are like those of other Caribbean gorgonians, in which different Symbiodinium phylotypes associate with different octocoral species (Santos et al. 2004).

Coevolutionary patterns of divergence are not restricted to octocorals. Recent work in several brooding corals, including Agaricia (Bongaerts et al. 2013), Seriatopora (Bongaerts et al. 2010) and Madracis (Frade et al. 2008), suggest species or genetically differentiated populations have co-evolved with their algal symbionts. Specific coral-symbiont associations seem more prevalent in brooding scleractinians in which Symbiodinium are transmitted vertically. The Symbiodinium - E. flexuosa tight association is remarkable and similar to that in Orbicella species (Thornhill et al. 2014) because specialization proceeded in a broadcasting species in the absence of direct vertical transmission and in a system in which larvae can travel large distances (> 10 km) (Prada and Hellberg 2013; Wirshing et al. 2013) and acquire symbionts after metamorphosed into a first polyp (Poland et al. 2013).

The coral-symbiont association need not be viewed as a strict dichotomy of flexibility or species-specificity. While specificity is common in octocorals (Santos et al. 2004; Goulet 2006), some scleractinian corals acquire various Symbiodinium phylotypes over ecological scales (Rowan et al. 1997; LaJeunesse et al. 2009). However, such apparent flexibility contains finer specificity (Finney et al. 2010). Orbicella faveolata for example, associates with Symbiodinium from clades A, B, C, and D, but only with specific phylotypes within those four clades, specially within clade C (Thornhill et al. 2014). Such mixed patterns of flexibility at macroevolutionary timescales and specificity at microevolutionary scales beg for more studies to understand the evolution of symbiosis in anthozoans.

3.4.2. Depth, photobiology and ecological speciation

Diversity in Symbiodinium has often been associated with habitat heterogeneity. In a landmark study, Rowan et al. (1997) showed how subtle changes in light availability within single coral colonies can segregate Symbiodinium genetic variation. More recent work suggests depth, often correlated with light, underlies Symbiodinium variation across habitats (Sampayo et al. 2008; Kirk et al. 2009; Finney et al. 2010; Lesser et al. 2010). In these heterogeneous environments, different host-symbiont combinations cause differences in survivorship and growth for the holobiont (Berkelmans and van Oppen 2006; Sampayo et al. 2007). Segregation in Symbiodinium types in E. flexuosa is associated with a transition in depth, raising the likelihood that different host-symbiont combinations perform better at their native depths (Sampayo et al. 2007).

31 Studies of the photobiology of Symbiodinium suggest different species have different capacities and perform better at different light conditions, such those that vary with depth (Iglesias-Prieto and Trench 1994). B1 species (as in E. flexuosa), for example, show different biophysical properties, such active fluorescence and photosystem-specific signatures (Robison and Warner 2006; Hennige et al. 2008; Hennige et al. 2011). Moreover, Symbiodinium species segregated by depth have different light optima and enhanced capacity (i.e., photosystem pressure) at native depths, even if these depth-segregated species are reciprocally transplanted (Iglesias-Prieto et al. 2004). It is thus likely that the enhanced survivorship of E. flexuosa to each habitat is partly due to its tight association with a specific Symbiodinium type.

In E. flexuosa, the formation and maintenance of the two lineages appears to have been promoted by immigrant inviability (Prada and Hellberg 2013), which occurs when locally adapted populations migrate to suboptimal environments, increasing mortality of these maladapted individuals (Nosil et al. 2005). For both E. flexuosa and Symbiodinium, their partner becomes another “environment”. Immigrant inviability can thus act in two ways. First, it can act against individuals with the “wrong” host-symbiont coupling. Next, even for pairs with the "right" host-symbiont coupling, inviability will sort holobionts by depth. Unlike brooders, in which most host-symbiont specialization has been observed, E. flexuosa acquires its Symbiodinium horizontally (from the surrounding environment upon settlement, rather than via its parent), increasing the potential for immigrant inviability.

3.4.3. Geographic variation and phylogeography

Symbiodinium in E. flexuosa shows patterns of geographical differentiation within each Shallow and Deep lineage, along with the segregation by host species and depth. Geographical differentiation is pervasive in Symbiodinium and most studies have reported geographical differentiation often (as here) at scales smaller than their host (Andras et al. 2011). For example, Andras et al. (2011) found that populations of Symbiodinium are similar between Puerto Rico and the Bahamas but different from those in Curaçao. I found the same pattern for Symbiodinium associated with E. flexuosa Deep, but a full segregation across localities for the Shallow lineage. The higher partition across geography by the Shallow lineage suggests that, even at small depth contours physical and biological processes vary in the Caribbean, uniquely altering connectivity patterns of populations at different depths.

32 Chapter 4 Strong Natural Selection on Juveniles Maintains a Narrow Adult Hybrid Zone in a Broadcast Spawner

4.1 INTRODUCTION

Natural selection via immigrant inviability occurs when the propagules of locally adapted populations suffer high mortality upon migrating to unsuitable habitats (Nosil et al. 2005; Rundle and Nosil 2005). Immigrant inviability can be pronounced in organisms where the interval between recruitment and sexual reproduction is long, as in corals and trees (Prada and Hellberg 2013; Lindtke et al. in press). While the effects of selection may be small over any given year, cumulative effects over many years before reproduction begins can generate a strong ecological filter against immigrants. Immigrant inviability can act across environmental gradients, generating clines or hybrid zones. If reproductive isolation occurs as a by-product of immigrant inviability, new species can arise by natural selection (Darwin 1859; Nosil et al. 2005; Rundle and Nosil 2005; Schluter 2009a), often across environmental gradients where they generate hybrid zones (Endler 1977).

The segregation of adults of different species into different habitats is pronounced in many long-lived, sessile species with large dispersal, such as wind-pollinated trees (Paine et al. 2009; Baldeck et al. 2013) and reef corals that disperse via planktonic larvae (Goreau 1959; Kinzie 1973), but whether such segregation arises before or after recruitment remains unknown. Reciprocal transplant experiments show that adults suffer higher mortality in their non-native habitat (Lowry et al. 2008; Marshall et al. 2010; Prada and Hellberg 2013). These studies show adaptive divergence occurs and selection can be strong, but the extent to which they reveal anything about the role of immigrant inviability in speciation depends on pre-recruitment processes and whether reproductive isolation is complete.

Pre-recruitment process, such as larval habitat choice (Connell 1972; Grosberg 1982), may cause a pattern of habitat segregation in adult populations without immigrant inviability. Larval habitat choice is pronounced in many sessile marine organisms (Carlon 2002; Baird et al. 2003), particularly in intertidal habitats (Grosberg 1982). Pre- and post-recruitment processes are not mutually exclusive, however, and while larvae may prefer a particular habitat, subsequent immigrant inviability may still operate to more completely segregate adults. Indeed, the two processes are interrelated, and post-recruitment mortality should favor habitat choice by larvae over evolutionary time (Grosberg 1981). If larvae choose their habitat and subsequent mortality is independent of habitat, then the distribution of species across habitats will be similar across age classes. In contrast, immigrant inviability will result in different distributions between age or size classes, as selection weeds out maladapted recruits over time. The distribution of different size-age classes across habitats can thus provide evidence as to whether pre- or post- recruitment mortality drives adult habitat-segregation and whether and to what degree immigrant inviability works. Immigrant inviability should also generate narrow hybrid zones because selection is acting across environmental gradients (Endler 1975; Slatkin 1975; Endler 1977). Habitat

33 mediated selection may be acute in organisms in which dispersal potential exceeds the scale of environmental change (Endler 1979), such as wind-pollinated plants and marine broadcasters with pelagic larvae. A long period (10's of years) between dispersal and the onset of reproduction should enhance the opportunity for selection to sort individuals by habitat before they reproduce (Prada and Hellberg 2013), thereby reducing the formation of hybrids.

The candelabrum coral Eunicea flexuosa is an especially suitable species in which to evaluate the role of immigrant inviability in the segregation of lineages across habitats (depths) because: 1) different selective regimes associated with variation in depth occur at far smaller scales than dispersal, which appears to approximate panmixia across the Caribbean (Prada and Hellberg 2013), 2) first reproduction does not occur until over 30 years of age (Beiring and Lasker 2000), and 3) unlike other reef taxa, fragmentation is uncommon and age can accurately be estimated from size, making comparisons across age classes easier. Previously, I suggested immigrant inviability might operate on habitat- segregated lineages of E. flexuosa because reciprocally transplanted adult colonies revealed adaptive differences between genetically differentiated shallow and deep populations (Prada et al. 2008; Prada and Hellberg 2013). Here, I compare the distribution of these lineages in adults and juveniles at the extremes of a depth gradient in Puerto Rico. I find that while the lineages in both size classes are sorted by habitat, segregation is more pronounced in adults. I also found that the two lineages replace each other over a depth and light gradient, producing few hybrids. Clines for three morphological traits and six genetic markers are coincident. The width of the cline is extremely narrow (< 100 m), to my knowledge the narrowest known when scaled by dispersal potential. I conclude that environmental selection largely maintains this hybrid zone and that the strong habitat selection segregates adult lineages by gradually eliminating young recruiting individuals over time. Such immigrant inviability helps to explain the abundance of depth-segregated sibling species (Knowlton 1993).

4.2 METHODS

4.2.1. Biology of Eunicea flexuosa

Eunicea flexuosa is a colonial marine cnidarian that lives in association with algae of the genus Symbiodinium. Bush-like colonies 1 - 1.5 m tall form by the asexual budding of polyps (Bayer 1961). Colonies acquire their algal symbionts from the environment at the single polyp stage shortly after larval metamorphosis (Coffroth et al. 2001; Prada et al. 2014b). E. flexuosa comprises two depth-segregated lineages, each adapted to its own depth as inferred from reciprocal transplants (Prada et al. 2008; Prada and Hellberg 2013). Each lineage maintains its own algal symbiont phylotype, even when both lineages meet at intermediate depths (Prada et al. accepted). Colonies of E. flexuosa have separate sexes, and sexual reproduction occurs through spawning of gametes that are fertilized in the water column (Beiring and Lasker 2000). Sperm and eggs are positively buoyant and, after release, rise to the surface, where fertilization more readily occurs (Prada per obs.). As in other octocorals, most fertilization takes place within 30 minutes of gamete release (Alino

34 and Coll 1989). In the closely related broadcast spawner, Plexaura kuna, most mating occurs between neighboring colonies but at least 8% of fertilization events involves colonies over 20 m apart (Coffroth and Lasker 1998). Gametes remain viable for at least 2 hours after release in other free spawning Caribbean anthozoans (Lasker et al. 1996; Levitan 2004). In Obicella annularis, which co-occurs with E. flexuosa, eggs move on average over 500 m, and not less than 100 m, in 100 minutes (Levitan et al. 2004). Thus, in less than two hours, eggs commonly disperse distances greater than the transition separating Shallow and Deep forms (100 m).

Larvae develop planktonically and larval dispersal appears to be high; widely separated (> 1000 km) locations form panmictic populations (Prada and Hellberg 2013). Isolation with Migration (IMa) analyses suggest introgression between the Shallow and Deep lineages (Prada and Hellberg 2013), but whether migration is ongoing was heretofore unknown.

4.2.2. Sampling and Sequence Data

To estimate variation in frequency of the two lineages among juveniles, I sampled 511 candelabrum coral colonies < 3 cm at Media Luna Reef, Puerto Rico (N17° 56’ 04.1”; W67° 02’ 52.5”). This same population was used to estimate the strength of selection between depth ecomorphs by replicated transplants (Prada et al. 2008; Prada and Hellberg 2013) and multi-year survivorship per size class (Prada and Hellberg 2013). Samples were collected at least 5 m apart to avoid sampling of clone mates. I genetically identified 437 as E. flexuosa juveniles (276 in shallow and 161 in deep habitats) using both the mitochondrial msh and the nuclear i3p (see below for details). To estimate the frequencies in adults (>50 cm), I pooled data from my previous studies and added some new individuals. I sampled 99 adult individuals from shallow and 103 from the deep habitat. The adults come from this and previous studies: 40 from (Prada et al. 2008); 38 from (Prada and Hellberg 2013) and 21 new for shallow and 44 from (Prada et al. 2008); 38 from (Prada and Hellberg 2013) and 21 new for deep. I used Fisher exact tests to detect frequency differences between age classes in both shallow and deep habitats.

To inspect for a possible hybrid zone between shallow and deep habitats, I sampled 159 adult colonies (> 50 cm) of E. flexuosa at seven locations across the depth gradient in the same focal population in Puerto Rico. The linear transect was ~200 m and varied in depth from 3 to 22 m. At each location I collected tissue from between 14 and 38 colonies. I collected colonies at least 5 m apart to decrease the chance of sampling of clone-mates. Previous multi-locus genotyping of well over 300 colonies has not revealed any non-unique multi-locus genotypes (Prada et al. 2008; Prada and Hellberg 2013). Samples were preserved in 95% (vol./vol.) ethanol and stored at -20ºC. I extracted genomic DNA from these samples using QIAGEN DNeasy Kit following the manufacture's protocols. For each individual from this depth transect, I sequenced five nuclear markers and one mtDNA marker for the host and one chloroplast locus for the algal symbiont. I sequenced the mitochondrial msh region, which encodes an ORF unique to some anthozoans, using published primers (France and Hoover 2001). I also genotyped individuals at three

35 previously described nuclear markers (Prada and Hellberg 2013): Inositol 3- phosphatase (i3p), Elongation Factor 1 alpha (ef1) and Calcium transporter 2 (ct2). In addition, I sequenced individuals at two new nuclear markers developed from sequences generated from a partial 454 run (Prada and Hellberg 2013). One of the two regions is anonymous (locus 564) and the other is an exon-crossing intron of the Glyceraldehyde-3-phosphate dehydrogenase (g3pd) homologue of the Nematostella vectensis genome. From these initial sequences I generated the following primers: (Gly3p_For1) 5’-GAATGGAAAGTTGACT GGAATGG-3’; (Gly3p_Rev1) 5’- GAAGCAATGTACTTGATGAGATCC-3’ for the Glyceraldehyde- 3-phosphate dehydrogenase and (EF_IB564F1) 5’-TAGCTCATTT CTCTCACTGTCACG-3’; (EF_IB_CONT564R1) 5’AAGTGATCCAACCTCTCATCTCG-3’ for the anonymous marker. I also amplified the chloroplast coding area of the psbA for all colonies across the hybrid zone, as I earlier found that it can differentiate among Symbiodinium phylotypes in depth-segregated species of E. flexuosa (Prada et al. accepted). I amplified samples for the psbA following previous protocols (Moore et al. 2003).

PCR amplifications were performed in a Bio-Rad T100 with the same cycling conditions (except for the annealing temperature) for all genes: an initial denaturation cycle of 3 min at 95°C, 2 min annealing at 50-56°C and 2 min extension at 72°C; followed by 38 cycles of 30 sec at 95°C, 45 sec at 50-56°C and 45 sec at 72°C and a final extension cycle at 72°C for 10 min. Amplicons were directly sequenced in both directions in an ABI 3100 using BigDye chemistry v 3.1 and both amplification primers.

I resolved heterozygotes containing a single indel using CHAMPURU (Flot 2007) and cloned individuals with multiple indel heterozygous using the Invitrogen TOPO TA Cloning Kit and sequencing at least 5 clones per reaction. I edited and assembled all sequences using Geneious v4.5.5 (Drummond AJ 2009). For sequences with multiple heterozygous sites, I resolved haplotypes using PHASE v2.1 (Stephens et al. 2001). I used a 90% probability as cutoff as haplotypes inferred by PHASE with a probability >70% accurately reflect haplotypes (Harrigan et al. 2008). I corroborated inferred haplotypes by directly cloning at least 10% of the sequences.

To assess morphological divergence among populations, I measured spindle length, calice diameter, and branch thickness. These three characters are the most divergent morphological characters between deep and shallow colonies of E. flexuosa in Puerto Rico and each character responds to different axes of variation as revealed by PCA analysis (Prada et al. 2008). To quantify morphological variation, I bleached colonies to clear the organic material, washed twice with distilled water, and dried in ethanol. For each colony I collected digital images using a Diagnostic Instrument SPOT RT Slider CCD camera attached to a Leica MZ7 stereomicroscope at the LSU Socolofsky Microscopy Center. To estimate morphological variation across the cline, I converted continuous morphological measurements into frequency data. I used the Bernoulli function as implemented in the hybrid zone R package HZAR (Derryberry et al. 2014). For each location, I estimate the mid-point among all values and conducted Bernoulli trials to classify values below or above such mid-point and then calculate frequency at each location.

36 4.2.3. Hybrid Zone Analyses

To understand the structure of the hybrid zone across the depth gradient I carried out three analyses. I first estimated clines for each marker to characterize variation across depth and test whether clines (both molecular and morphological) were coincident. I then used genetic markers to estimate a hybrid index, the inbreeding coefficient, and a measure of LD. Once I estimated the width of the hybrid zone for each marker and identified areas in which LD was highest (e. g. the center of the cline), I used the LD measure to estimate realized dispersal.

To determine the shape of each empirical cline, I used the three mathematical descriptions of Szymura and Barton (1991) and (Szymura and Barton 1986) (Appendix 1). These equations describe a sigmoidal curve with a center (maximum slope) and two tails (one in each side) modeled as an exponential decay. The shape of the sigmoid curve is determined by the width of the cline and the position of its center. The tails of the curve are dependent on the distance between the center of the curve and the end of the tail and the slope of the tail (Gay et al. 2008). From this description, I tested five different models. The cline could have: no tails, symmetrical tails, asymmetrical tails, left tail only, or right tail only. I fit molecular and morphological clines to test these five models using a Metropolis-Hasting algorithm as implemented in the hybrid zone R package HZAR using the functions hzar.fitRequest and hzar.doFit (Maley 2012; Derryberry et al. 2014).

Because the shape of the cline can vary depending on how it is scaled, I included scaling differences in the models. Clines were scaled by the minimum (Pmin) and maximum (Pmax) frequencies in the cline. I divided scaling in three ways: no scale (Pmin =0 and Pmax = 1), empirical estimates (Pmin= minimum observed frequency, Pmax = maximum observed frequency) or best fit (Pmin and Pmax). By using combinations of cline shapes (five possibilities) and scaling (three possibilities), I tested 15 different models for each cline. For each model and for each cline, I specified three independent chains for the MCMC search, each 7 X 105 steps long. To estimate parameters and generate the posterior distribution, I combined the chains, removed 7 x 104 as burn in, and then estimated the cline parameters. To ensure convergence of the parameter estimates, I repeated the procedure above at least three times. I recorded the AICc for each model within each cline and chose the one with the lowest score to infer cline width and center (Table S4.1). To test whether clines from different characters show concordance, I compared their confidence intervals, which were defined as two log-likelihood scores above and below the mean (Edwards 1992).

To calculate a hybrid index, I used loci that are fixed (msh, i3p, ct2, g3pd) between lineages and I converted Shallow and Deep alleles into a 1 and 0 system then scaled by dividing by 8, so a pure Shallow individual will have a score of 1 (all alleles for four fixed loci) and a pure Deep would have 0. To visualize the distribution of hybrids, I used the program NewHybrids (Anderson and Thompson 2002). I estimated the probability of first and second-generation hybrids from genotypic data from all six loci across the hybrid zone. I ran NewHybrids at least four times and averaged estimates across replicates.

37 Once I had cline and hybrid estimates, I calculated linkage disequilibrium between loci across the hybrid zone. Of the seven molecular markers, five (host msh, i3p, ct2, g3pd and the algal psbA) are fixed between the Shallow and Deep lineages. Barton and Gale (1993) suggest dispersal (σ) can be estimated as: √(Drw1w2), where D is the LD between the two markers, w1 and w2 refer to the cline width for each marker and r is the recombination rate. I assumed no association between markers (r = 0.5). where D is the LD between the two markers, w1 and w2 refer to the cline width for each marker and r is the recombination rate. I estimated LD in ANALYZE (Barton and Baird 1998) as Rij , a pairwise LD measure standardized by allele frequencies �!" = �!"/√(�!�!�!�!)). I also estimated inbreeding coefficients (FIS) across the hybrid zone via maximum likelihood as implemented in ANALYZE. I then combined width and LD estimates at the cline center and estimated realized dispersal assuming a tension zone model. I assumed no association between markers (r = 0.5).

4.2.4. Environmental variables

Coincidence of molecular and morphological clines with variation in environmental features may suggest hybrid zones maintained by habitat-based selection (Endler 1977; Moore and Price 1993). I estimated environmental variation across the hybrid zone by quantifying variation in depth and light. I used a Suunto Gekko Gauge Computer Console and a submersible irradiance meter (Li-Cor Biosciences LI-1400) to estimate depth and light, respectively. I graphed downwelling irradiance profiles to document light variation across depth. I corrected values by light measurements in air with an external collector for cloud corrections. Each measure was taken at least four times during 2007 to capture the variability in rainy and sunny seasons.

4.2.5. Selection and micro-geographic adaptation analysis

To estimate selection against genotypes in mismatched depths, I measured the frequency of juveniles and adults of the two ecotypes in both shallow and deep environments. I estimated the change in frequency relative to the change in frequency of the most common genotype in a given depth and subtracted the value from 1 (Hartl 2000). For example, in shallow areas the fitness of the Shallow lineage is 1 and the fitness of the Deep lineage is the ratio of the frequency change between the Deep and the Shallow minus 1.

To estimate the spatial scale of the adaptation (d) between the Shallow and the Deep lineage, given the dispersal potential of the species, I used the widest cline width estimate (78 m) and divided by the dispersal neighborhood distance of the species. I estimated the dispersal neighborhood distance (σ) using Rosset’s (1997) approach:

1 � = 4�!�

38 where De is the effective density and m is the slope between geographical distance and FST/(1−FST). To estimate m, I averaged FST across four loci for populations across the Caribbean (Prada and Hellberg 2013) and estimated geographical distance among populations in Google Earth 6.2 using the shortest nautical distance. To estimate De I used earlier estimates of the effective population size (Ne) for E. flexuosa (Prada and Hellberg 2013) derived from Isolation with Migration analysis (Hey and Nielsen 2004). I then estimated the number of effective individuals per area using the total Caribbean coral reef area 26,000 km2 (Burke and Maidens 2004) as a proxy for the area occupied. Because this underestimates effective density (De), I also estimated De assuming the species only uses 10% of reef habitat. This approach renders an effective density (De) that ranges between 7.69 and 76.90 individuals per square kilometer. I additionally estimated effective density (De) using my direct field estimate of the highest number of colonies per meter. Decreasing area occupation and assuming a high density across the species range, bias estimates to the lowest possible dispersal for a species that has Caribbean wide connectivity.

4.3 RESULTS

4.3.1. Adults show stronger habitat segregation than juveniles in both shallow and deep area

I collected genetic data from 437 juveniles (276 in shallow and 161 in deep habitats) and 202 adults (99 in shallow and 103 in deep habitats) of E. flexuosa. Segregation was evident in both juvenile and adults but was stronger in adults (Figure 4.1) (P < 0.001 for Shallow and P < 0.01 for deep habitats).

Deep(lineage(( shallow((((((((deep((

Shallow((lineage((

recruits(

n(=(276((( n(=(161(

adults(

n(=(99((( n(=(103(((

P(=(0.001( P(=(0.025(

Figure 4.1. Frequencies of Shallow (grey) and Deep (black) lineages of E. flexuosa at the extremes of the depth gradient. I used Fisher exact test to estimate differences between juvenile and adult frequencies.

39 4.3.2. Clines are narrow, sigmoidal, and coincident across molecular and morphological markers

I collected morphological and genetic data from 159 adults over a depth gradient. I found a transition in morphological and allele frequencies of both coral and algae (Figure 4.2). The hybrid zone between the Shallow and Deep forms of E. flexuosa occurred over less than 77 m (Table 4.1). The maximum likelihood fitted cline for half of the markers was a sigmodal model without tails and fixed frequencies (Figure 4.2). A left-tail model with fixed frequencies was a better fit for locus psbA (Table S4.1). All clines - molecular and morphological – for both host and algal symbiont coincided at the cline center (Table 4.1). The 2Log-Likelihood intervals of all cline centers overlapped, ranging from 116 - 130 m along the linear transect.

4.3.3. Linkage disequilibrium is highest at the hybrid zone center

Estimates of inbreeding (FIS) and scaled linkage disequilibrium (Rij) increased at the center of the hybrid zone. The ML estimate of FIS showed a significant heterozygote deficit at the center of the hybrid zone: while FIS values were near zero at the transect extremes, they increase at 116 and 125 m for all loci (Figure 4.3A). Linkage disequilibrium also increased substantially at the center of the hybrid zone (Figure 4.3B), peaking at 116 and 125 m and dropped to 0.35 at both shallow and deep ends. The center of the zone is consistent with selection against hybrids across the ecotone (Endler 1977; Szymura and Barton 1986; Singhal and Moritz 2012).

4.3.4. Low effective dispersal across habitats despite large dispersal potential: Selection across depths

Mean estimates of effective dispersal base on linkage disequilibrium between pairs of loci at the center of the hybrid zone (125 m) ranged from 4.9 to 10.0 meters per generation, with a mean value of 7.0 m per gen. The dispersal potential from parent to offspring estimated from pairwise FST data is 55.52 km using the entire Caribbean area, 17.56 km if I assume the species occupies only 10% of the area, and a conservative 2.39 km if I assume the high density of 4.38 colonies per m2 across the entire range. Based on this conservative estimate, the effective dispersal across depth is at least 30X smaller than the dispersal potential of E. flexuosa (d = 0.03). While the juveniles I sampled were already segregated to some degree (Figure 4.1), selection against mismatched genotypes relative to native ones is high: 0.8 for the Deep ecotype in shallow and 0.5 for Shallow in deep habitats.

4.3.5. Environmental factors correlate with the hybrid zone center

Clines of molecular and morphological characters coincide with variation in depth and light conditions (Figure 4.2, bottom panel). The zone produces a sharp gradient that separates the Shallow and Deep forms of E. flexuosa at about 116 - 130 m along the transect, which occurs at depths of 14 - 16 m, right at the reef slope and coincident with the break from

40 shallow fore reef to a deep fore reef habitat (Ballantine et al. 2008). In Puerto Rico, this ecological transition is accompanied by a rapid change in light (Figure 4.2, bottom panel).

msh (mitochondrial) gly3p ef1α ! ! Frequency!of!!most!common!allele!

ct2 i3p 564

type of shallow Frequency Spindle length Branch thickness Calice diameter

0

0.6

5 ) -2 m -1

) 10

m 0.4 , umols d 15 Depth ( Depth 0.2 20 (E Light psbA (algal symbiont) 25 0 0 50 100 150 0 50 100 150

Distance along the transect (m)

Figure 4.2. Maximum likelihood cline shape estimates for the molecular (six), morphological characters (three) of the coral and the chloroplast of its algal symbiont (one). The grey area represents the 95% confidence intervals around the cline fit. Spindle length and calice diameter were estimated in µm and branch thickness in mm. Bottom graphs show variation in depth and light across locations. Genetic clines were estimated by using the frequency of the most common allele.

41 Table 4.1. Maximum likelihood estimates for cline center and confidence intervals (range estimated as low and high 2 Log-likelihood units). Selection coefficients for nine characters (*mtDNA, ** nDNA, # morphology and &cpDNA from algal symbiont) as inferred using non- standardized (D) and standardize linkage disequilibrium (Rij) using the program ANALYZE.

Character Center (m) Width (m) msh* 118.90 (107.33 - 127.64) 40.58 (19.32 - 84.24) gly3p** 117.58 (104.37 - 126.05) 55.62 (23.81 - 110.68) 1633** 125.90 (110.07 - 137.09) 46.07 (9.85 - 120.28) I3P** 116.96 (100.99 - 124.56) 42.08 (17.60 - 122.18) EF1A** 127.73 (121.61 - 139.04) 33.34 (16.76 - 75.87) 564** 119.80 (26.43 - 198.36) 76.58 (0.09 - 229.79) Spindle length# 128.62 (121.34 - 135.76) 58.40 (42.20 - 84.76) Calice diameter# 109.88 (95.69 - 120.49) 64.53 (32.85 - 118.16) Branch thickness# 123.26 (109.62 - 133.38) 57.21 (18.11 - 121.18) psbA& 121.78 (116.05 - 124.17) 29.70 (23.38 - 62.02)

)))))A) 1) 0.8)

0.6) IS) ) F 0.4)

0.2)

0) )))))))))B) 0) 50) 100) 150)

0.8! ) ij ) R 0.5!

)))))))))C) 0.2! 0! 50! 100! 150!

+ !!1! 1.0 ! + + !

0.8!0.8 ! ! + +

0.6!0.6 !

Frequency !

0.4!0.4 ! !

0.2!0.2 ) ) Mean)hybrid)index) )

+ +

!!0!!!!!!!!!!!!!!!!!!!!!!50!!!!!!!!!!!!!!!!!!!!100!!!!!!!!!!!!!!!!!!!150!0 50 100 150 ! Distance Distance)along)the)transect)(m)))

Figure 4.3. (A) Inbreeding coefficient (FIS), (B) standardized linkage disequilibrium (Rij) and (C) mean hybrid index along the depth gradient. Gray shadow indicates the center of the hybrid zone. Linkage disequilibrium is scaled by allele frequencies.

42 4.3.6. No first generation hybrids - 15 of the 159 (9%) individuals surveyed were identified by NewHybrids as hybrids. No hybrids were seen at the shallowest site (3 - 4 m deep). The proportions of hybrids at other sites ranged from 4.5% to 21.0 % but were not statistically different from each other (P = 0.6). Most hybrids were admixed individuals, with no first generation hybrids, only one F2 individual, and all others later backcrosses.

4.4 DISCUSSION

The two genetically differentiated but incompletely isolated forms of Eunicea flexuosa meet in a narrow hybrid zone (< 100 m lateral distance) that follows light and depth gradients, with clines coincident across morphological and molecular markers. The center of this hybrid zone is characterized by a deficit of heterozygotes, with no first generation hybrids and most hybrids being later generation backcrosses. The narrow hybrid zone is maintained by strong selection (> 0.5) against juveniles mismatched to the habitat they settled in. The high selection inferred from colonies at different ages/sizes is consistent with earlier estimates derived from field transplant experiments of adults (Prada et al. 2008; Prada and Hellberg 2013). Even under conservative assumptions about larval dispersal potential, the spatial scale of adaptation to depth in E. flexuosa is at least 30X smaller than the dispersal potential of the species.

4.4.1. The Eunicea flexuosa hybrid zone is unusually narrow for a marine broadcaster

The hybrid zone of E. flexuosa stands among the narrowest hybrid zones characterized, comparable with those of terrestrial snails (Stankowski 2013) and grasshoppers (Kawakami et al. 2009), animals with notably meager dispersal potentials. The narrowness of hybrid zone is unusual for marine broadcasters, which spawn their gametes and produce larvae capable of dispersing 10’s of kilometers. This degree of dispersal has few terrestrial parallels, with the best being pollen producing trees (Schueler and Schlünzen 2006) and pines (Williams 2010) or birds (Pyne 1992; Carling et al. 2011). In these, however, hybrid zones exceed the dispersal ability of the hybridizing taxa (Carling and Brumfield 2009; Carling et al. 2011; Hamilton et al. 2013).

In contrast, marine broadcasters commonly generate hybrid zones narrower than their dispersal distances and often follow environmental gradients in salinity, temperature, and habitat (Hilbish 1985; Duggins et al. 1995; Gardner 1997). Dispersal potential per se is thus a rather weak predictor of realized gene flow, and factors such as ecological filtering (Marshall et al. 2010) or genomic incompatibilities in hybrids must play a role in shaping the distribution of genetic variation (Burton et al. 2006). This is evidenced by geographic variation in the width of hybrid zones shown in some species. Blue mussels – the most studied hybrid system in the ocean - form multiple hybrid zones across Europe and North America that vary by 100’s of kilometers in width (Hilbish 1985; Gardner 1997; Riginos and Cunningham 2005). Similarly, killifish (Fundulus) generate clines that extend 500 km near New Jersey (McKenzie 2013) but just 27 km in Florida (Duggins et al. 1995).

43 4.4.2. Mechanisms maintaining the narrow E. flexuosa hybrid zone

Previously, I used reciprocal transplants of adults to show that immigrant inviability might drive diversification across depths in E. flexuosa (Prada and Hellberg 2013). Differences between juveniles and adults in the inferred lineage frequencies (Figure 4.1) suggest that post-recruitment processes play a role in the adult habitat differentiation. Immigrant inviability prevents gene flow and generates the tight hybrid zone we see in E. flexuosa. Because I sampled colonies < 3 cm tall (about 2 yrs old), the selection coefficient may be an underestimate, as I missed effects on younger stages. Even under strong immigrant inviability, larval habitat choice may still operate and bias the proportion of initial settlers (Grosberg 1982; Baird et al. 2003). I suspect that strong habitat selection in corals species across depths may drive the evolution of habitat choice, as seen in other marine invertebrates (Grosberg 1981).

Immigrant inviability across ecological gradients also predicts that clines in molecular and morphological variation are scaled by the strength of selection, even under high gene flow, and will coincide with environmental gradients (Endler 1977; Moore and Price 1993). The hybrid zone described here (Figure 4.2) tightly correlates with a change in depth and light and matches closely with the transition from a shallow gorgonian-dominated habitat to a deep platy-coral environment, separated by a transitional fore-reef slope dominated by massive (mostly Orbicella) corals (Ballantine et al. 2008). The direct evidence linking the hybrid zone with environmental variation, the habitat driven mortality at different age-size classes, and the fitness differential across depths highlights the importance of adaptation to depth in generating and maintaining coral reef diversity.

Ecological factors associated with depth that could cause adaptive divergence include light, food availability, predators, water motion and sediment transport and composition (Prada and Hellberg 2013). Studies of E. flexuosa and close relatives have shown that depth differences may lead to shifts in symbiotic algae (Prada et al. accepted), morphology (Kim et al. 2004; Prada et al. 2008), and feeding strategies (Lasker et al. 1983). Adaptation to depth thus likely entails a multi-faceted response, affecting various genomic regions simultaneously. Such coupling between multiple genomic regions shaped by environmental selection suggests that genomic and environmental incompatibilities in hybrids are interconnected and act in concert (Nurnberger et al. 1995).

In addition, adaptation to depth can cause differences in timing of spawning and assortative mating, which in turn can also reinforce the Shallow/Deep hybrid zone as in other systems (Jiggins and Mallet 2000). Differential timing of spawning have been reported in other broadcasting corals (Lessios 1984; Knowlton et al. 1997; Levitan et al. 2004) and some preliminary evidence suggest E. flexuosa develop eggs at different times at different depths (Kim et al. 2004). Marine broadcasters also often have fast coevolving sperm-egg interacting proteins that are highly specific and often driven to diverge interspecifically by positive selection (Swanson and Vacquier 2002; Hellberg et al. 2012).

44 I suggest that habitat selection may initially generate divergence, then genetic incompatibilities and assortative mating build through time and reinforce the depth segregation that we see today in the Shallow/Deep lineages of E. flexuosa.

4.4.3. Strong parallel environmental gradients plus long pre-reproductive selection generates narrow hybrid zone despite large dispersal potential

Two aspects of the biology of E. flexuosa and the environment in which the diverging lineages meet may help generate its narrow cline: 1) sharp and parallel multi-faceted environmental gradients and 2) strong selection by cumulative post-settlement, pre- reproductive selection.

To generate a narrow stable hybrid zone in a broad disperser, an environmental gradient strong enough to generate local adaptation between populations must be present (Clarke 1966; Slatkin 1975; Endler 1977; Moore 1977). It has been shown empirically and theoretically that sharp clines are generated if the magnitude of selection is strong (> 0.1), independent of the amount of gene flow (Endler 1975, 1977). In marine systems salinity can generate steep (10’s of km) clines in both vertebrates (Whitehead et al. 2011) and invertebrates (Gardner 1997; Riginos and Cunningham 2005), while temperature generates broader (100’s of km) clines (Gardner 1997; Whitehead and Crawford 2006) in the same taxa. In both cases, hybrid zones or locus-specific clines are tightly related to environmental variation and loci directly under environmental selection haven been identified (Hilbish and Koehn 1984; Hilbish and Koehn 1985; Schmidt and Rand 2001; Whitehead et al. 2011).

But perhaps the feature that most promotes the development of a narrow zone is the long post-settlement juvenile period of E. flexuosa. Habitat selection is exceptionally strong in sedentary long-lived organisms like trees and corals, in which a long time elapses between dispersal of propagules and the onset of reproduction in sessile adults. Such long time (> 20 years) generates an extended opportunity for selection to act and sort individuals by habitat before they reproduce (Prada and Hellberg 2013). Narrow hybrid zones between long-lived species with high dispersal along steep environmental gradients may be unduly found in long-lived sessile marine broadcasters such as corals, sea fans, and sponges, but also across elevational gradients in long-lived plants like oaks and pines. For example, in Japanese sub-alpine coniferous forest, Pinus pumila and P. parviflora form a hybrid zone just 325 m wide (125 m in elevation) (Watano et al. 2004), a pattern echoed by pines in China (Li et al. 2010) and the Rocky Mountains (Rehfeldt 1988).

Long-lived species with high dispersal potential can thus generate narrow hybrid zones along steep environmental gradients, a pattern that the shift in juvenile frequencies I document here suggests can arise from immigrant inviability. The strong diversifying selection across these gradients may explain the large number of sibling species that segregate by depth (Knowlton 1993) and is consistent with divergence occurring at spatial scales far smaller than previously considered.

45 Chapter 5 Conclusions

Speciation research seeks to understand how species acquire reproductive isolation. In my dissertation, I first evaluated the roles of ecology and geography during speciation. My work suggests that selection counteracts large amounts of gene flow and plays a major role in lineage splitting in candelabrum corals of the genus Eunicea. Depth often segregates closely related coral reef species, suggesting it may provide a mechanism for how species form in these remarkably species-rich ecosystems.

In Chapter 2, I evaluated the strength of one ecological factor (depth) to isolate populations by comparing the genes and morphologies of pairs of depth-segregated populations of the candelabrum coral Eunicea flexuosa across the Caribbean. Eunicea is endemic to the Caribbean and all sister species co-occur. Eunicea flexuosa is widespread both geographically and across reef habitats. The genetic analysis revealed two segregated lineages: one Shallow, the other Deep. Field survivorship data, combined with estimates of selection coefficients based on transplant experiments, suggest that selection is strong enough to segregate these two lineages. Genetic exchange between the Shallow and Deep lineages occurred either immediately after divergence or the two have diverged with gene flow. Migration occurs asymmetrically from the Shallow to Deep lineage. Limited recruitment to reproductive age, even under weak annual selection advantage, is sufficient to generate habitat segregation because of the cumulative prolonged pre-reproductive selection.

In Chapter 3, I found that the two ecotypes of E. flexuosa, a species with a larvae free of symbionts, each associate with a unique Symbiodinium B1/184 variant. This relationship between host and alga is maintained when host colonies are reciprocally transplanted, although a few cases of algal switching were also observed. Even when the clades of both partners are present at intermediate depths, the specificity between host and algal lineages remained absolute. This tight association suggests coevolution of the two partners and this coupling that defines the fitness of the holobiont (coral + algae) in different environments. These species interactions may influence evolutionary patterns and, when coupled with ecological factors, may play a role in speciation. I further suggest that unrecognized cryptic diversity may mask host-algal specificity and change the inference of evolutionary processes in mutualistic associations.

In Chapter 4, I examined the frequencies of juveniles and adults of the two depth- segregated lineages, each with larvae that disperse geographically over spatial scales orders of magnitude greater than the distances separating the co-occurring types. I found that distributions of the two lineages in both shallow and deep environments are more distinct when inferred from adults than from juveniles. I also found an extremely narrow hybrid zone (< 100 m) between the two ecotypes, with coincident clines for molecular and morphological characters of both the host coral and its algal symbiont. Effective dispersal

46 estimates derived from the hybrid zone are remarkably small (< 20 m) for a broadcast spawner. Selection coefficients against mismatched genotypes derived from cohort data correlates well with selection coefficients from field transplant experiments.

As a whole, my dissertation work suggests that natural selection against maladapted migrants generate and maintain diversity in coral reefs and that adaptive divergence across depths acting on populations of long-lived organisms is particularly strong because the long time that selection acts before sexual reproduction. Symbiont specificity is also likely to contribute to this ecological divergence. During ecological speciation, species interactions may reinforce prezygotic isolation by preventing the establishment of maladapted coral-alga-depth combinations, each of these combinations being a potential link in which immigrant inviability can operate. Strong habitat selection across depth gradients may also provide the mechanism for the evolution of habitat choice, which in turn decreases gene flow across pairs of depth-segregated species. Knowlton (1993) highlighted the vast diversity of sibling species in the sea. We are now beginning to understand how marine diversity segregated along ecological factors like depth creates complexes of co-diversifying species, increasing diversity in the sea. Long pre-reproductive selection in sessile species adapted to distinct habitats, with the potential for temporal separation of reproduction, exemplified in E. flexuosa, may help solve the long-standing question of how speciation can occur in marine populations with widely dispersing larval forms.

5.1 FUTURE WORK

Speciation research seeks to understand both how species acquire reproductive isolation and how they become phenotypically differentiated (Mayr 1942). During my dissertation, I studied the mechanisms that may have lead to reproductive isolation between depth- segregated species of corals. I aim now to understand how phenotypes of recently diverged species form. Long-lived clonal organisms are extraordinarily morphologically and physiologically plastic. I have previously documented the extent and limits of some of this morphological plasticity in the two depth-segregated lineages of E. flexuosa. I am now studying how plasticity in gene regulation couples with morphological plasticity, providing a mechanistic understanding on how new phenotypes of recently diverged species form. During ecological speciation, a phenotypically plastic population may become constitutively fixed (genetically assimilated) by selection for one phenotype in a particular environment (Waddington 1942; West-Eberhard 2003). Thus the genomic variation that facilitates plasticity over ecological scales may provide the substrate for natural selection to act on and generate adaptive divergence and species-specific morphologies. I am currently studying gene expression variation in identical genetic backgrounds in both Shallow and Deep lineages in both shallow and deep habitats to test if plasticity in gene expression within lineages is correlated with divergence between lineages. For clonal taxa (such as sponges, corals, fungi and plants) where phenotypic plasticity is common, adaptive divergence between species may originate from plasticity differences in acclimation responses to environmental variation.

47 References

Abrego, D., K. E. Ulstrup, B. L. Willis, and M. J. H. van Oppen. 2008. Species-specific interactions between algal endosymbionts and coral hosts define their bleaching response to heat and light stress. Proc. R. Soc. Lond., Ser. B: Biol. Sci. 275:2273-2282.

Alberto, F., N. J, J. Derory, O. Lepais, R. Vitalis, D. Galop, and A. Kremer. 2010. Population differentiation of sessile oak at the altitudinal front of migration in the French Pyrenees. Mol. Ecol. 19:2626–2639.

Alino, P. and J. Coll. 1989. Observations of the synchronized mass spawning and postsettlement activity of octocorals on the Great Barrier Reef , Australia : Biological Aspects. Bull. Mar. Sci. 45:697-707.

Anderson, D. R. 2008. Model Based Inference in the Life Sciences. Springer, New York.

Anderson, E. and E. Thompson. 2002. A model-based method for identifying species hybrids by using multilocus genetic data. Genetics 160:1217–1229.

Andras, J. P., N. L. Kirk, M. A. Coffroth, and C. D. Harvell. 2009. Isolation and characterization of microsatellite loci in Symbiodinium B1/B184, the dinoflagellate symbiont of the Caribbean sea fan coral, Gorgonia ventalina. Mol. Ecol. Notes 9:989–993.

Andras, J. P., N. L. Kirk, and C. D. Harvell. 2011. Range-wide population genetic structure of Symbiodinium associated with the Caribbean Sea fan coral, Gorgonia ventalina. Mol. Ecol. 20:2525–2542.

Baird, A., R. Babcock, and C. Mundy. 2003. Habitat selection by larvae influences the depth distribution of six common coral species. Mar. Ecol. Prog. Ser. 252:289-293.

Baker, A. 1999. Reef corals bleach to survive change. Nature 411:765-766.

Baldeck, C. A., K. E. Harms, J. B. Yavitt, R. John, B. L. Turner, R. Valencia, H. Navarrete, S. Bunyavejchewin, S. Kiratiprayoon, A. Yaacob, M. N. N. Supardi, S. J. Davies, S. P. Hubbell, G. B. Chuyong, D. Kenfack, D. W. Thomas, and J. W. Dalling. 2013. Habitat filtering across tree life stages in tropical forest communities. Proc. R. Soc. Lond., Ser. B: Biol. Sci. 280:9.

Ballantine, L., R. Appeldoorn, and P. Yoshioka. 2008. Biology and ecology of puerto rican coral reefs. Pp. 375–406 in B. Riegl, and R. Dodge, eds. Coral Reefs of the USA.

Barluenga, M., K. Stolting, W. Salzburger, M. Muschick, and A. Meyer. 2006. Sympatric speciation in Nicaraguan crater lake cichlid fish. Nature 439:719–723.

Barton, N. H. and S. Baird. 1998. Analyse 1.10, http://helios.bto.ed.ac.uk/evolgen/Mac/Analyse/.

48 Barton, N. H. and K. S. Gale. 1993. Genetic analysis of hybrid zones in R. G. Harrison, ed. Hybrid zones and the evolutionary process. Oxford University press, New York.

Baums, I. B., M. W. Miller, and M. E. Hellberg. 2005. Regionally isolated populations of an imperiled Caribbean coral, Acropora palmata. Mol. Ecol. 14:1377-1390.

Bayer, F. M. 1961. The Shallow-Water Octocorallia of the West Indian Region. A Manual for marine biologists. Martinus Nijhoff, The Hague.

Beiring, E. A. and H. R. Lasker. 2000. Egg production by colonies of a gorgonian coral. Mar. Ecol. Prog. Ser. 196:169-177.

Berkelmans, R. and M. J. H. van Oppen. 2006. The role of zooxanthellae in the thermal tolerance of corals: a ‘nugget of hope’ for coral reefs in an era of climate change. Proc. R. Soc. Lond., Ser. B: Biol. Sci. 273:2305–2312.

Bird, C., B. Holland, B. Bowen, and R. Toonen. 2011. Diversification of sympatric broadcast- spawning limpets (Cellana spp.) within the Hawaiian archipelago. Mol. Ecol. 20:2128-2141.

Bongaerts, P., P. R. Frade, J. J. Ogier, K. B. Hay, J. van Bleijswijk, N. Englebert, M. J. Vermeij, R. P. Bak, P. M. Visser, and O. Hoegh-Guldberg. 2013. Sharing the slope: depth partitioning of agariciid corals and associated Symbiodinium across shallow and mesophotic habitats (2-60 m) on a Caribbean reef. BMC Evol. Biol. 13:205.

Bongaerts, P., C. Riginos, T. Ridgway, E. Sampayo, M. Van Oppen, N. Englebert, F. Vermeulen, and O. Hoegh-Guldberg. 2010. Genetic divergence across habitats in the widespread coral Seriatopora hystrix and its associated Symbiodinium. PLoS ONE 5: 10871- 10871.

Brazeau, D. A. and C. D. Harvell. 1994. Genetic structure of local populations and divergence between growth forms in a clonal invertebrate, the Caribbean octocoral Briareum asbestinum. Mar. Biol. 119:53-60.

Burke, L. and J. Maidens. 2004. Reefs at Risk. World Resources Institute, Washington, D.C. (USA).

Burton, R. S., C. K. Ellison, and J. S. Harrison. 2006. The sorry state of F2 hybrids: Consequences of rapid mitochondrial DNA evolution in allopatric populations. Am. Nat. 168:14-24.

Carling, M. D. and R. T. Brumfield. 2009. Speciation in Passerina buntings: introgression patterns of sex-linked loci identify a candidate gene region for reproductive isolation. Mol. Ecol. 18:834–847.

49 Carling, M. D., L. G. Serene, and I. J. Lovette. 2011. Using historical DNA to characterize hybridization between Baltimore (Icterus galbula) and Bullock’s (Icterus bullockii) orioles. Auk 128: 61-68.

Carlon, D. B. 2002. Production and supply of larvae as determinants of zonation in a brooding tropical coral. J. Exp. Mar. Biol. Ecol. 268:33-46.

Carlon, D. B. and A. F. Budd. 2002. Incipient speciation across a depth gradient in a scleractinian coral? Evolution 56:2227-2242.

Carstens, B. C. and T. Dewey. 2010. Species delimitation using a combined coalescent and information-theoretic approach: An example from North American Myotis bats. Syst. Biol. 59:400-414.

Chen, C., A. Wang, L. Fang, and Y. Yang. 2005. Fluctuating algal symbiont communities in Acropora palifera (Scleractinia : Acroporidae) from Taiwan. Mar. Ecol. Prog. Ser. 295:113–121.

Chomczynski, P. and N. Sacchi. 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate- phenol-chloroform extraction. . Anal. Biochem. 162:156– 159.

Clarke, B. C. 1966. The evolution of morpho-ratio clines. Am. Nat. 100:389-402.

Clement, M., D. Posada, and K. Crandall. 2000. TCS: a computer program to estimate gene genealogies. Mol. Ecol. 9(10):1657-1660.

Coffroth, M. A. and H. R. Lasker. 1998. Larval paternity and male reproductive success of a broadcast-spawning gorgonian, Plexaura kuna. Mar. Biol. 131:329±337.

Coffroth, M. A., H. R. Lasker, M. E. Diamond, J. A. Bruenn, and E. Bermingham. 1992. DNA fingerprints of a gorgonian coral: a method for detecting clonal structure in a vegetative species. Mar. Biol. 114:317-325.

Coffroth, M. A. and S. R. Santos. 2005. Genetic diversity in Symbiodinium. Protist 156:19−34.

Coffroth, M. A., S. R. Santos, and T. L. Goulet. 2001. Early ontogenetic expression of specificity in a cnidarian-algal symbiosis. Mar. Ecol. Prog. Ser. 222:85–96.

Colombo-Pallotta, M. F., A. Rodríguez-Román, and R. Iglesias-Prieto. 2010. Calcification in bleached and unbleached Montastraea faveolata: evaluating the role of oxygen and glycerol. Coral Reefs 29:899-907.

Connell, J. 1973. Population ecology of reef- building corals. Pp. 205-245 in A. Jones, and R. Endeam, eds. Biology and Geology of Coral Reefs. Academic Press, London.

50 Connell, J. H. 1972. Community Interactions on Marine Rocky Intertidal Shores. Annu. Rev. Ecol. Syst. 3:169-192.

Crow, J. 1961. Mechanisms and trends in human evolution. Daedalus 90:416–431.

Darwin, C. 1859. On the Origin of Species by Means of Natural Selection or the Preservation of Favored Races in the Struggle for Life. J. Murray, London.

Davy, S. K., D. Allemand, and V. M. Weis. 2012. Cell biology of cnidarian-dinoflagellate symbiosis. Microbiol. Mol. Biol. Rev. 76:229–261.

DeBiasse, M. B., V.P. Richards, and M. S. Shivji. 2010. Genetic assessment of connectivity in the common reef sponge, Callyspongia vaginalis (Demospongiae: Haplosclerida) reveals high population structure along the Florida reef tract. Coral Reefs 29:47-55.

Derome, N., P. Duchesne, and L. Bernatchez. 2006. Parallelism in gene transcription among sympatric lake whitefish (Coregonus clupeaformis Mitchill) ecotypes. Mol. Ecol. 15:1239–1249.

Derryberry, E. P., G. E. Derryberry, J. M. Maley, and R. T. Brumfield. 2014. HZAR: Hybrid Zone Analysis using an R software package. Mol. Ecol. Res. 14:652–663.

DeSalvo, M. K., S. Sunagawa, P. L. Fisher, C. R. Voolstra, R. Iglesias-Prieto, and M. Medina. 2010. Coral host transcriptomic states are correlated with Symbiodinium genotypes. Mol. Ecol. 19:1174-1186.

Doebeli, M., U. Dieckmann, J. A. J. Metz, and D. Tautz. 2005. What we have also learned: adaptive speciation is theoretically plausible. Evolution 59:691-695.

Drummond AJ, A. B., Cheung M, Heled J, Kearse M, Moir R, Stones-Havas S, Thierer T, Wilson A. 2009. Geneious v4.5.4. Available at: http://www.geneious.com.

Duggins, J. C., A. Karlin, T. Mousseau, and K. Relyea. 1995. Analysis of a hybrid zone in Fundulus majalis in a northeastern Florida ecotone. Heredity 74:117–128.

Duran, S. and K. Rutzler. 2006. Ecological speciation in a Caribbean marine sponge. Mol. Phylogen. Evol. 40:292-297.

Earl, D. and E. vonHoldt. 2011. STRUCTURE HARVESTER: a website and program for visualizing STRUCTURE output and implementing the Evanno method. Cons. Gen. Res. 4:359-361.

Endler, J. A. 1975. Gene flow and population differentiation. Science 179:243-250.

Endler, J. A. 1977. Geographic variation, speciation, and clines. Princeton University Press, Princeton, NJ.

51 Endler, J. A. 1979. Gene flow and life history patterns. Genetics 93: 263–284.

Evanno, G., S. Regnaut, and J. Goudet. 2005. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Mol. Ecol. 14:2611–2620.

Feder, J. L., C. A. Chilcote, and G. L. Bush. 1988. Genetic differentiation between sympatric host races of the apple maggot fly Rhagoletis pomonella. Nature 336:61–64.

Ferrari, J., J. West, S. Via, and H. Godfray. 2012. Population genetic structure and secondary symbionts in host-associated populations of the pea aphid complex. Evolution 66: 375–390.

Finney, J. C., T. Pettay, E. M. Sampayo, M. E. Warner, H. A. Oxenford, and T. C. LaJeunesse. 2010. The relative significance of host–habitat, depth, and geography on the ecology, endemism, and speciation of coral endosymbionts in the genus Symbiodinium. Microb. Ecol. 60:250–263.

Fitzpatrick, B. M. 2012. Underappreciated consequences of phenotypic plasticity for ecological speciation. International Journal of Ecology 256017.

Flot, J.-F. 2007. Champuru 1.0: a computer software for unraveling mixtures of two DNA sequences of unequal lengths. Mol. Ecol. Notes 6:974-977.

Frade, P. R., F. De Jongh, F. Vermeulen, J. Van Bleijswijk, and R. P. M. Bak. 2008. Variation in symbiont distribution between closely related coral species over large depth ranges. Mol. Ecol. 17:691–703.

France, S. C. and L. L. Hoover. 2001. Analysis of variation in mitochondrial DNA sequences (ND3, ND4L, MSH) among Octocorallia (=Alcyonaria) (Cnidaria: Anthozoa). Bull. Biol. Soc. Wash. 10:110-118.

Freeman, A. and J. Byers. 2006. Divergent induced responses to an invasive predator in marine mussel populations. Science 313:831833.

Fukami, H., A. Budd, D. Levitan, J. Jara, R. Kersanach, and N. Knowlton. 2004. Geographical differences in species boundaries among members of the Montastraea annularis complex based on molecular and morphological markers. Evolution 58:324-337.

Gardner, J. P. A. 1997. Hybridization in the sea. Pp. 1–78 in J. H. S. Blaxter, and A. J. Southward, eds. Adv. Mar. Biol. Academic Press.

Giraud, T., P. Gladieux, and S. Gavrilets. 2010. Linking emergence of fungal plant diseases and ecological speciation. Trends Ecol. Evol. 25:387-395.

Godsoe, W., J. B. Yoder, C. P. Smith, and O. Pellmyr. 2008. Coevolution and diversification in the Joshua tree yucca moth mutualism. Am. Nat. 171:816-823.

52 Goreau, T. F. 1959. The ecology of Jamaican coral reefs I. Species composition and zonation. Ecology 40:67-90.

Goulet, T. L. 2006. Most corals may not change their symbionts. Mar. Ecol. Prog. Ser. 321:1- 7.

Goulet, T. L. and M. A. Coffroth. 2003. Genetic composition of zooxanthellae between and within colonies of the octocoral Plexaura kuna, based on small subunit rDNA and multilocus DNA fingerprinting. Mar. Biol. 142:233–239.

Goulet, T. L., C. Simmons, and D. Goulet. 2008. Worldwide biogeography of Symbiodinium in tropical octocorals. Mar. Ecol. Prog. Ser. 355:45-58.

Grajales, A., C. Aguilar, and J. A. Sánchez. 2007. Phylogenetic reconstruction using secondary structures of Internal Transcribed Spacer 2 (ITS2, rDNA): finding the molecular and morphological gap in Caribbean gorgonian corals. BMC Evol. Biol. 7:90.

Grosberg, R. K. 1981. Competitive ability influences habitat choice in marine invertebrates. Natue 290:700-702.

Grosberg, R. K. 1982. Intertidal zonation of barnacles: the influence of planktonic zonation of larvae on the vertical distribution of adults. Ecology 63:894-899.

Grottoli, A., L. Rodrigues, and J. Palardy. 2006. Heterotrophic plasticity and resilience in bleached corals. Nature 440:1186-1189.

Gutiérrez-Rodríguez, C. and H. R. Lasker. 2004. Microsatellite variation reveals high levels of genetic variability and population structure in the gorgonian coral Pseudopterogorgia elisabethae across the Bahamas. Mol. Ecol. 13:2211-2221.

Guzmán, H. M. and C. A. Guevara. 1999. Coral reefs of Bocas del Toro, Panama: III. Distribution, structure, diversity and conservation status of reefs in Pastores, Cristobal, Popa and Cayo Agua islands. Rev. Biol. Trop. 47:659-676.

Hale, M., C. McCormick, J. Jackson, and J. DeWoody. 2009. Next-generation pyrosequencing of gonad transcriptomes in the polyploid lake sturgeon (Acipenser fulvescens): the relative merits of normalization and rarefaction in gene discovery. BMC Genomics 10:11.

Hall, M. C. and J. H. Willis. 2006. Divergent selection on flowering time contributes to local adaptation in Mimulus guttatus populations. Evolution 60:2466–2477.

Hamilton, J. A., C. Lexer, and S. N. Aitken. 2013. Differential introgression reveals candidate genes for selection across a spruce (Picea sitchensis × P. glauca) hybrid zone. New Phytol. 197:927-938.

53 Harrigan, R., M. Mazza, and M. Sorenson. 2008. Computation vs. cloning: evaluation of two methods for haplotype determination. Mol. Ecol. Res. 8:1239–1248.

Hartl, D. L. 2000. A Primer of Population Genetics. Sinauer Associates, Sunderland, Massachusetts.

Hawthorne, D. J. and S. Via. 2001. Genetic linkage of ecological specialization and reproductive isolation in pea aphids. Nature 412:904-907.

Heath, K. D., P. V. Burke, and J. R. Stinchcombe. 2012. Coevolutionary genetic variation in the legume-rhizobium transcriptome. Mol. Ecol. 21:4735-4747.

Hellberg, M. 2009. Gene flow and isolation among populations of marine animals. Annu. Rev. Ecol., Evol. Syst. 40:291-310.

Hellberg, M., A. Dennis, P. Arbour-Reily, J. Aagaard, and W. Swanson. 2012. The Tegula tango: a co-evolutionary dance of interacting, positively-selected sperm and egg proteins. Evolution 66:1681-1694.

Hellberg, M. E. 1998. Sympatric sea shells along the sea's shore: the geography of speciation in the marine gastropod Tegula. Evolution 52:1311-1324.

Hennige, S. J., M. McGinley, A. G. Grottoli, and M. E. Warner. 2011. Photoinhibition of Symbiodinium spp. within the reef corals Montastraea faveolata and Porites astreoides: implications for coral bleaching. Mar. Biol. 158:2515-2526.

Hennige, S. J., D. J. Suggett, M. E. Warner, K. E. McDougall, and D. J. Smith. 2008. Unravelling coral photoacclimation: Symbiodinium strategy and host modification. Pp. 7-11. Proceedings of the 11th International Coral Reef Symposium, Ft. Lauderdale, Florida.

Hereford, J. 2009. A quantitative survey of local adaptation and fitness trade-offs. Am. Nat. 173:579–588.

Hey, J. and R. Nielsen. 2004. Multilocus methods for estimating population sizes, migration rates and divergence times, with applications to the divergence of Drosophila pseudoobscura and D. persimilis. Genetics 167:747–760.

Hilbish, J. T. and R. K. Koehn. 1984. Dominance in Physiological Phenotypes and Fitness at an Enzyme Locus. Science 229:52-54.

Hilbish, T. J. 1985. Demographic and temporal structure of an allele frequency cline in the mussel Mytilus edulis. Mar. Biol. 86:163-171.

Hilbish, T. J. and R. K. Koehn. 1985. The physiological basis of natural selection at the LAP locus. Evolution 39:1302–1317.

54 Hodges, S. and M. Arnold. 1994. Floral and ecological isolation between Aquilegia formosa and A. pubescens. Proc. Natl. Acad. Sci. USA 91:2493-2496.

Huang, X. and A. Madan. 1999. CAP3: A DNA sequence assembly program. Genome Res. 9:868-877.

Huelsenbeck, J. P. and F. Ronquist. 2001. MrBayes: Bayesian inference of phylogenetic trees. Bioinformatics 17:754–755.

Iglesias-Prieto, R., V. H. Beltrán, T. C. LaJeunesse, H. Reyes-Bonilla, and P. E. Thomé. 2004. Different algal symbionts explain the vertical distribution of dominant reef corals in the eastern Pacific. Proc. R. Soc. Lond., Ser. B: Biol. Sci. 271:1757-1763.

Iglesias-Prieto, R. and R. K. Trench. 1994. Adaptation and acclimation to irradiance in symbiotic dinoflagellates I. Response of the photosynthetic unit to changes in photon flux density. Mar. Ecol. Prog. Ser. 113:163-175.

Ingram, T. 2011. Speciation along a depth gradient in a marine adaptive radiation. Proc. R. Soc. Lond., Ser. B: Biol. Sci. 278:613–618.

Jackson, J. and F. McKinney. 1990. Ecological processes and progressive macroevolution of marine clonal benthos. Pp. 173–209 in R. Ross, and W. Allmon, eds. Causes of evolution University of Chicago Press, Chicago.

Jackson, J. B. C., P. Jung, A. G. Coates, and L. S. Collins. 1993. Diversity and extinction of tropical American mollusks and emergence of the Isthmus of Panama. Science 260:1624-1626.

Jakobsson, M. and N. Rosenberg. 2007. CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinformatics 23:1801.

Jiggins, C. and J. Mallet. 2000. Bimodal hybrid zones and speciation. Trends Ecol. Evol. 15:250-255.

Jiggins, C. D. 2008. Ecological speciation in mimetic butterflies. Bioscience 58:541–548.

Johannesson, K. 2001. Parallel speciation: a key to sympatric divergence. Trends Ecol. Evol. 16:148–153.

Johnson, J. and K. Omland. 2004. Model selection in ecology and evolution. Trends Ecol. Evol. 19:101-108.

Kahng, S., B. Benayahu, and H. R. Lasker. 2011. Sexual reproduction in octocoral. Mar. Ecol. Prog. Ser. 443:265-283.

55 Kawakami, T., R. Butlin, M. Adams, D. Paull, and S. Cooper. 2009. Genetic analysis of a chromosomal hybrid zone in the Australian grasshoppers (Vandiemenella, viatica species group). Evolution 63:139–152.

Kim, E., H. R. Lasker, M. A. Coffroth, and K. Kim. 2004. Morphological and genetic variation across reef habitats in a broadcast-spawning octocoral. Hydrobiologia 530/531:423–432.

Kinzie, R. A. 1973. The zonation of West Indian gorgonians. Bull. Mar. Sci. 23:93-155.

Kirk, N. L., J. P. Andras, C. D. Harvell, S. R. Santos, and M. A. Coffroth. 2009. Population structure of Symbiodinium sp. associated with the common sea fan, Gorgonia ventalina, in the Florida Keys across distance, depth, and time. Mar. Biol. 156:1609- 1623.

Knowlton, N. 1993. Sibling species in the sea. Annu. Rev. Ecol., Evol. Syst. 24:189-216.

Knowlton, N. and J. B. C. Jackson. 1994. New taxonomy and niche partitioning on coral reefs: Jack of all trades or master of some? Trends Ecol. Evol. 9:7-9.

Knowlton, N., J. L. Mate, H. M. Guzman, R. Rowan, and J. Jara. 1997. Direct evidence for reproductive isolation among the three species of the Montastraea annularis complex in Central America (Panama and Honduras). Mar. Biol. 127:705-711.

Kobmoo, M., M. Hossert-McKey, J. Y. Rasplus, and F. Kjellberg. 2010. Ficus racemosa is pollinated by a single population of a single agaonid wasp species in continental South-East Asia. Mol. Ecol. 19:2700–2712.

Kubatko, L., B. Carstens, and L. Knowles. 2009. STEM: species tree estimation using maximum likelihood for gene trees under coalescence. Bioinformatics 25:971–973.

LaJeunesse, T. C. 2002. Diversity and community structure of symbiotic dinoflagellates from Caribbean coral reefs. Mar. Biol. 141:387-400.

LaJeunesse, T. C. 2005. Species" radiations of symbiotic dinoflagellates in the Atlantic and Indo-Pacific since the Miocene-Pliocene transition. Mol. Biol. Evol. 22:570–581.

LaJeunesse, T. C., J. E. Parkinson, and J. D. Reimer. 2012. A genetics-based description of Symbiodinium minutum sp. nov. and S. psygmophilum sp. nov. (Dinophyceae), two dinoflagellates symbiotic with Cnidaria. J. Phycol. 48:1380-1391.

LaJeunesse, T. C., D. T. Pettay, E. M. Sampayo, N. Phongsuwan, B. E. Brown, D. O. Obura, O. Hoegh-Guldberg, and W. K. Fitt. 2010. Long-standing environmental conditions, geographic isolation and host−symbiont specificity influence the relative ecological dominance and genetic diversification of coral endosymbionts in the genus Symbiodinium. J. Biogeogr. 37:785−800.

56 LaJeunesse, T. C., R. T. Smith, J. C. Finney, and H. A. Oxenford. 2009. Outbreak and persistence of opportunistic symbiotic dinoflagellates during the 2005 Caribbean mass coral ‘bleaching’ event. Proc. R. Soc. Lond., Ser. B: Biol. Sci. 276:4139–4148.

LaJeunesse, T. C., D. J. Thornhill, E. F. Cox, F. G. Stanton, W. K. Fitt, and G. W. Schmidt. 2004. High diversity and host specificity observed among symbiotic dinoflagellates in reef coral communities from Hawaii. Coral Reefs 23:596-603.

Langerhans, R. B., M. E. Gifford, and E. O. Joseph. 2007. Ecological speciation in Gambusia fishes. Evolution 61: 2056–2074.

Larkin, M., G. Blackshields, N. Brown, R. Chenna, P. McGettigan, H. McWilliam, F. Valentin, I. Wallace, A. Wilm, R. Lopez, J. Thompson, T. Gibson, and D. Higgins. 2007. Clustal W and Clustal X version 2.0. Bioinformatics 23:2947-2948.

Lasker, H., D. Brazeau, J. Calderon, M. Coffroth, R. Coma, and K. Kim. 1996. In situ rates of fertilization among broadcast spawning gorgonians. Biol. Bull. 190:45–55.

Lasker, H. R., M. D. Gottfried, and M. A. Coffroth. 1983. Effects of depth on the feeding capabilities of two octocorals. Mar. Biol. 73:73–78.

Lepais, O., R. J. Petit, E. Guichoux, J. E. Lavabre, F. Alberto, A. Kremer, and S. Gerber. 2009. Species relative abundance and direction of introgression in oaks. Mol. Ecol. 18:2228-2242.

Lesser, M. P., M. Ojimi, R. Gates, M. Stat, M. Slattery, and A. G. Grottoli. 2010. Photoacclimatization by the Coral Montastraea cavernosa in the Mesophotic Zone: Light, Food, and Genetics. Ecology 91:990-1003.

Lessios, H. and D. Robertson. 2006. Crossing the impassable: genetic connections in 20 reef fishes across the eastern Pacific barrier. Proc. R. Soc. Lond., Ser. B: Biol. Sci. 273:2201–2208.

Lessios, H. A. 1984. Possible prezygotic reproductive isolation in sea urchins separated by the Isthmus of Panama. Evolution 35:1144- 1148.

Levitan, D. R. 2004. Density-dependent sexual selection in external fertilizers: variances in male and female fertilization success along the continuum from sperm limitation to sexual conflict in the sea urchin Strongylocentrotus franciscanus. Am. Nat. 164:298– 309.

Levitan, D. R., H. Fukami, J. Jara, D. Kline, T. A. McGovern, K. M. McGhee, C. A. Swanson, and N. Knowlton. 2004. Mechanisms of reproductive isolation among sympatric broadcast-spawning corals of the Montastraea annularis complex. Evolution 58:308- 323.

57 Li, S., Y. Chen, H. Gao, and T. Yin. 2010. Potential chromosomal introgression barriers revealed by linkage analysis in a hybrid of Pinus massoniana and P. hwangshanensis. BMC Plant Biol. 10:37.

Lindtke, D., Z. Gompert, C. Lexer, and B. C. A. in press. Unexpected ancestry of Populus seedlings from a hybrid zone implies a large role for postzygotic selection in the maintenance of species. Mol. Ecol.:10.1111/mec.12759.

Little, A. F., M. J. H. van Oppen, and B. F. Willis. 2004. Flexibility in algal endosymbioses shapes growth in reef corals. Science 304:1492– 1494.

Lowry, D., J. Modliszewski, K. Wrigh, C. Wu, and J. Willis. 2008. The strength and genetic basis of reproductive isolating barriers in flowering plants. Phil. Trans. R. Soc. B 363:3009-3021.

Lu, G. and L. Bernatchez. 1999. Correlated trophic specialization and genetic divergence in sympatric lake whitefish ecotypes (Coregonus clupeaformis): support for the ecological speciation hypothesis. Evolution 53:1491–1505.

Magalon, H., E. Baudry, A. Huste, M. Adjeroud, and M. Veuille. 2006. High genetic diversity of the symbiotic dinoflagellates in the coral Pocillopora meandrina from the South Pacific. Mar. Biol. 148:913 – 922.

Maley, J. M. 2012. Ecological Speciation of King Rails (Rallus elegans) and Clapper Rails (Rallus longirostris). Biological Sciences. Louisiana State University, Baton Rouge, LA.

Marshall, D., K. Monro, M. Bode, M. Keough, and S. Swearer. 2010. Phenotype – environment mismatches reduce connectivity in the sea. Ecol. Lett. 13:128–140.

Mayr, E. 1942. Systematics and the Origin of Species. Columbia Univ. Press, New York.

McKenzie, J. L. 2013. Genetic characterization of a hybrid zone in killifish (Fundulus heteroclitus) : evidence for assortative mating or selection against hybrids. Zoology. University of British Columbia, British Columbia.

Meirmans, P. G. and P. H. Van Tienderen. 2004. GENOTYPE and GENODIVE: two programs for the analysis of genetic diversity of asexual organisms. Mol. Ecol. Notes 4:792- 794.

Mieog, J., M. van Oppen, N. Cantin, W. Stam, and J. Olsen. 2007. Real-time PCR reveals a high incidence of clade D at low levels in four scleractinian corals across the Great Barrier Reef: implications for symbiont shuffling. Coral Reefs 26:449-457.

58 Milne, I., F. Wright, G. Rowe, D. Marshall, D. Husmeier, and G. McGuire. 2004. TOPALi: software for automatic identification of recombinant sequences within DNA multiple alignments. Bioinformatics 20:1806– 1807.

Moczek, A., S. Sultan, S. Foster, C. Ledón-Rettig, I. Dworkin, H. Nijhout, E. Abouheif, and D. Pfenning. 2011. The role of developmental plasticity in evolutionary innovation. Proc. R. Soc. Lond., Ser. B: Biol. Sci. 276:2705–2713.

Moore, R. B., K. M. Ferguson, W. K. W. Loh, O. Hoegh-Guldberg, and D. A. Carter. 2003. Highly organized structure in the non-coding region of the psbA minicircle from clade C Symbiodinium. Int. J. Syst. Evol. Microbiol. 53:1725–1734.

Moore, W. S. 1977. An evaluation of narrow hybrid zones in vertebrates. Q. Rev. Biol. 52:263-278.

Moore, W. S. and J. T. Price. 1993. Nature of selection in the Northern Flicker Hybrid Zone and its implications for speciation theory. Pp. 196–225 in R. G. Harrison, ed. Hybrid Zones and the Evolutionary Process. Oxford University Press, Oxford.

Muscatine, L., P. G. Falkowski, and Z. Dubinsky. 1983. Carbon budgets in symbiotic associations. Pp. 649–673 in H. E. A. Schenk, and W. Schwemmler, eds. Endocytobiology II. Intracellular Space as Oligogenetic Ecosystem. Walter de Gruyter, Berlin and New York.

Muscatine, L., P. G. Falkowski, J. W. Porter, and Z. Dubinsky. 1984. Fate of photosynthetic fixed carbon in light- and shade-adapted colonies of the symbiotic coral Stylophora pistillata. Proc. R. Soc. Lond., Ser. B: Biol. Sci. 222:181–202.

Muscatine, L., L. R. McCloskey, and R. E. Marian. 1981. Estimating the daily contribution of carbon from zooxanthellae to coral animal respiration. Limnol. Oceanogr. 26:601– 611.

Nosil, P. 2012. Ecological speciation. Oxford University Press Oxford.

Nosil, P., B. J. Crespi, and C. P. Sandoval. 2002. Host-plant adaptation drives the parallel evolution of reproductive isolation. Nature 417:440-443.

Nosil, P., T. Vines, and D. Funk. 2005. Perspective: reproductive isolation caused by natural selection against immigrants from divergent habitats. Evolution 59:705–719.

Nurnberger, B., N. H. Barton, C. J. MacCallum, J. Gilchrist, and M. Appleby. 1995. Natural selection on quantitative traits in the Bombina hybrid zone. Evolution 49:1224– 1238.

Nylander, J. A. A. 2004. MrAIC. Program distributed by the author, Evolutionary Biology Centre, Uppsala University.

59 Opresko, D. M. and J. A. Sanchez. 2005. Caribbean shallow-water black corals (Cnidaria: Anthozoa: Antipatharia). Caribb. J. Sci. 41:492–507.

Østbye, K., T. F. Næsje, L. Bernatchez, O. T. Sandlund, and K. Hindar. 2005. Morphological divergence and origin of sympatric populations of European whitefish (Coregonus lavaretus L.) in Lake Femund, Norway. J. Evol. Biol. 18:683–702.

Paine, C. E. T., K. E. Harms, and J. Ramos. 2009. Supplemental irrigation increases seedling performance and diversity in a tropical forest. J. Trop. Ecol. 25:171-180.

Pettay, D. T. and T. C. LaJeunesse. 2007. Microsatellites from clade B Symbiodinium spp. specialized for Caribbean corals in the genus Madracis. . Mol. Ecol. Notes 7:1271– 1274.

Pettay, D. T., D. C. Wham, J. H. Pinzón, and T. C. LaJeunesse. 2011. Genotypic diversity and spatial–temporal distribution of Symbiodinium clones in an abundant reef coral. Mol. Ecol. 20:5197-5212.

Pfennig, D. W., M. A. Wund, E. C. Snell-Rood, T. Cruickshank, C. D. Schlichting, and A. P. Moczek. 2010. Phenotypic plasticity's impacts on diversification and speciation. Trends Ecol. Evol. 25:459-467.

Pigliucci, M. and J. Kaplan. 2006. Making Sense of Evolution: The Conceptual Foundations of Evolutionary Biology. Chicago University Press, Chicago.

Pinzon, J. H. and T. C. LaJeunesse. 2011. Species delimitation of common reef corals in the genus Pocillopora using nucleotide sequence phylogenies, population genetics, and symbiosis ecology. Mol. Ecol. 20:311-325.

Pochon, X. and R. D. Gates. 2010. A new Symbiodinium clade (Dinophyceae) from soritid foraminifera in Hawai’i. Mol. Phylogen. Evol. 56:492−497.

Poland, D. M., J. M. Mansfield, N. L. Kirk, T. L. Shearer, S. Connelly, C. L. Lewis, and M. A. Coffroth. 2013. Annual and habitat variation in Symbiodinium acquisition by juvenile Briareum asbestinum (Cnidaria: Octocorallia). Mar. Ecol. Prog. Ser. 476:23-37.

Pond, S. and S. Frost. 2005. Datamonkey: rapid detection of selective pressure on individual sites of codon alignments. Bioinformatics 21:2531-2533.

Pond, S., D. Posada, M. Gravenor, C. Woelk, and S. Frost. 2006. Automated Phylogenetic Detection of Recombination Using a Genetic Algorithm. Mol. Biol. Evol. 23:1891- 1901.

Posada, D. 2008. jModelTest: Phylogenetic Model Averaging. Mol. Biol. Evol. 25:1253-1256.

60 Prada, C. and M. E. Hellberg. 2013. Long pre-reproductive selection and divergence by depth in a Caribbean candelabrum coral. Proc. Natl. Acad. Sci. USA 119:53 – 60

Prada, C., M. B. DeBiasse, J. E. Neigel, B. Yednock, J. L. Stake, Z. H. Forsman, I. B. Baums, and M. E. Hellberg. 2014a. Genetic species delineation among branching Caribbean Porites corals. Coral Reefs 10.1007/s00338-014-1179-5.

Prada, C., S. E. McIlroy, D. M. Beltrán, D. J. Valint, S. A. Ford, M. E. Hellberg, and M. A. Coffroth. 2014b. Cryptic diversity hides host and habitat specialization In a gorgonian-algal symbiosis. Mol. Ecol. 23:3330–3340.

Prada, C., N. V. Schizas, and P. M. Yoshioka. 2008. Phenotypic plasticity or speciation? A case from a clonal marine organism. BMC Evol. Biol. 8:47.

Prada, C., E. Weil, and P. M. Yoshioka. 2010. Octocoral bleaching during unusual thermal stress. Coral Reefs 29:41–45.

Prillinger, H., R. Messner, H. König, R. Bauer, K. Lopandic, O. Molnar, P. Dangel, F. Weigang, T. Kirisits, T. Nakase, and L. Sigler. 1996. Yeasts associated with termites: A phenotypic and genotypic characterization and use of coevolution for dating evolutionary radiations in asco- and basidiomycetes. Syst. Appl. Microbiol. 19:265- 283.

Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference of population structure using multilocus genotype data. Genetics 155:945–959.

Pyne, R. B. 1992. Indigo bunting in A. Poole, P. Stettenheim, and F. Gill, eds. The Birds of North America. The Academy of Natural Sciences, Washington, DC; The American Ornithologists’ Union, Philadelphia.

Quesada, H., D. Posada, A. Caballero, P. Moran, and E. Rolan-Alvarez. 2007. Phylogenetic evidence for multiple sympatric ecological diversification in a marine snail. Evolution 61:1600–1612.

Rehfeldt, G. E. 1988. Ecological genetics of Pinus contorta from the Rocky Mountains (USA): A synthesis. Silvae Genet. 37:131-135.

Riginos, C. and C. W. Cunningham. 2005. Invited Review: Local Adaptation and Species Segregation in Two Mussel (Mytilus edulis x M. trossulus) Hybrid Zones Mol. Ecol. 14:381-400.

Robison, J. D. and M. E. Warner. 2006. Differential impacts of photoacclimation and thermal stress on the photobiology of four different phylotypes of Symbiodinium (pyrrhophyta) 1. J. Phycol. 42:568–579.

61 Rocha, L., D. Robertson, J. Roman, and B. Bowen. 2005. Ecological speciation in tropical reef fishes. Proc. R. Soc. Lond., Ser. B: Biol. Sci. 272:573–579.

Rosenberg, N. A. 2004. DISTRUCT: a program for the graphical display of population structure. Mol. Ecol. Notes 4: :137–138.

Rousset, F. 1997. Genetic differentiation and estimation of gene flow from F-statistics under isolation by distance. Genetics 145:1219–1228.

Rowan, R. 2004. Coral bleaching: Thermal adaptation in reef coral symbionts. Natue 430:742.

Rowan, R. and N. Knowlton. 1995. Intraspecific diversity and ecological zonation in coral- algal symbiosis. Proc. Natl. Acad. Sci. USA 92:2850–2853.

Rowan, R., N. Knowlton, A. Baker, and J. Jara. 1997. Landscape ecology of algal symbionts creates variation in episodes of coral bleaching. Nature 388:265-269.

Rowan, R. and D. A. Powers. 1992. Ribosomal RNA sequences and the diversity of symbiotic dinoflagellates (zooxanthellae) Proc. Natl. Acad. Sci. USA 89:3639-3643.

Rozas, J., J. C. Sandez-DelBurrio, X. Messeguer, and R. Rozas. 2003. DnaSP: DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19:2496-2497.

Rozen, S. and J. Skaletsky. 2000. Primer3 on the WWW for general users and for biologist programmers. Pp. 365-386 in S. Krawetz, and S. Misener, eds. Bioinformatics Methods and Protocols: Methods in Molecular Biology. Humana Press, Totowa, NJ.

Rundle, H. and P. Nosil. 2005. Ecological speciation. Ecol. Lett. 8:336–352.

Sampayo, E., L. Franceschinis, O. Hoegh-Guldberg, and S. Dove. 2007. Niche partitioning of closely related symbiotic dinoflagellates. Mol. Ecol. 16:3721–3733.

Sampayo, E. M., T. Ridgway, P. Bongaerts, and O. Hoegh-Guldberg. 2008. Bleaching susceptibility and mortality of corals are determined by fine-scale differences in symbiont type. Proc. Natl. Acad. Sci. USA 105:10444–10449.

Sanchez, J. 2009. Systematics of the candelabrum gorgonian corals (Eunicea Lamouroux; Plexauridae; Octocorallia; Cnidaria). Zool. J. Linn. Soc. 157:237–263.

Sanford, E. and M. Kelly. 2011. Local adaptation in marine invertebrates. Ann. Rev. Mar. Sci. 3:509–535.

Santos, S., C. Gutierrez-Rodriguez, and M. Coffroth. 2003a. Phylogenetic identification of symbiotic dinoflagellates via length heteroplasmy in domain V of chloroplast large subunit (cp23s)-rDNA sequences. Mar. Biotechnol. 5:130-140.

62 Santos, S. R., C. Gutierrez-Rodrıguez, H. R. Lasker, and M. A. Coffroth. 2003b. Symbiodinium sp. associations in the gorgonian Pseudopterogorgia elisabethae in the Bahamas: high levels of genetic variability and population structure in symbiotic dinoflagellates. Mar. Biol. 143:111–120.

Santos, S. R., T.L. Shearer, A.R. Hannes, and M. A. Coffroth. 2004. Fine-scale diversity and specificity in the most prevalent lineage of symbiotic dinoflagellates (Symbiodinium, Dinophyceae) of the Caribbean. Mol. Ecol. 13:459-469.

Santos, S. R., D. J. Taylor, and M. A. Coffroth. 2001. Genetic comparisons of freshly isolated vs. cultured zooxanthellae: Implications for extrapolating to the intact symbiosis. J. Phycol. 37:900-912.

Santos, S. R., D. J. Taylor, R. A. Kinzie, M. Hidaka, K. Sakai, and M. A. Coffroth. 2002. Molecular phylogeny of symbiotic dinoflagellates inferred from partial chloroplast large subunit 23s-rDNA sequences. Mol. Phylogen. Evol. 23:97–111.

Savolainen, V., M. C. Anstett, C. Lexer, I. Hutton, J. J. Clarkson, M. V. Norup, M. P. Powell, D. Springate, N. Salamin, and W. J. Baker. 2006. Sympatric speciation in palms on an oceanic island. Nature 441:210–213.

Schluter, D. 1996. Ecological speciation in postglacial fishes. Philos. Trans. R. Soc. Lond., Ser. B: Biol. Sci. 351:807–814.

Schluter, D. 2000. The Ecology of Adaptive Radiation. Oxford University Press, Oxford.

Schluter, D. 2009a. Evidence for Ecological Speciation and Its Alternative. Science 323:737- 741.

Schluter, D. 2009b. Genetics and ecological speciation. Proc. Natl. Acad. Sci. USA 106:9955– 9962.

Schmidt, P. S. and D. M. Rand. 2001. Adaptive maintenance of genetic polymorphism in an intertidal barnacle: Habitat and life history stage-specific survivorship of Mpi genotypes. Evolution 55:1336-1344.

Schueler, S. and K. H. Schlünzen. 2006. Modeling of oak pollen dispersal on the landscape level with a mesoscale amospheric model. Environ. Model Assess 11:179 - 194.

Scoville, A. G. and M. E. Pfrender. 2010. Phenotypic plasticity facilitates rapid parallel adaptation to an introduced predator. Proc. Natl. Acad. Sci. USA 278:2705-2713.

Seehausen, O., Y., I. S. Terai, K. L. Magalhaes, H. D. J. Carleton, R. Mrosso, I. v. d. S. Miyagi, M. V. Schneider, M. E. Maan, H. H. Tachida, Imai, and N. Okada. 2008. Speciation through sensory drive in cichlid fish. Nature 455:620–626.

63 Servedio, M., G. van Doorn, M. Kopp, A. Frame, and P. Nosil. 2011. Magic traits in speciation: ‘magic’ but not rare? . Trends Ecol. Evol. 26:389–397.

Silverstein, R. N., A. Correa, and A. Baker. 2012. Specificity is rarely absolute in coral-algal symbiosis: implications for coral response to climate change. Proc. R. Soc. Lond., Ser. B: Biol. Sci. 279:2609-2618.

Singhal, S. and C. Moritz. 2012. Strong selection maintains a narrow hybrid zone between morpho-logically cryptic lineages in a rainforest lizard. Evolution 66:1474-1489.

Slatkin, M. 1975. Gene flow and selection on a two-locus system. Genetics 81:787–802.

Stankowski, S. 2013. Ecological speciation in an island snail: evidence for the parallel evolution of a novel ecotype and maintenance by ecologically dependent postzygotic isolation. Mol. Ecol. 22:2726-2741.

Stephens, M., N. Smith, and P. Donnelly. 2001. A new statistical method for haplotype reconstruction from population data. Am. J. Hum. Genet. 68:978--989.

Swanson, W. J. and V. D. Vacquier. 2002. The rapid evolution of reproductive proteins. Nature Review in Genetics 3:137–144.

Szmant, A. M. 1986. Reproductive ecology of Caribbean reef corals. Coral Reefs 5:43-54.

Szymura, J. and N. H. Barton. 1986. Genetic analysis of a hybrid zone between the fire- bellied toads, Bombina bombina and B. variegata, near Cracow in souhern Poland. Evolution 40:1141- 1159.

Szymura, J. and N. H. Barton. 1991. The genetic structure of the hybrid zone between the firebellied toads Bombina bombina and B. variegata: comparisons between transects and between loci. Evolution 45:237-261.

Taylor, M. S. and M. E. Hellberg. 2003. Genetic evidence for local retention of pelagic larvae in a Caribbean reef fish. Science 299:107-109.

Taylor, M. S. and M. E. Hellberg. 2005. Marine radiations at small geographic scales: speciation in Neotropical reef gobies (Elacatinus). Evolution 59:374-385.

Taylor, M. S. and M. E. Hellberg. 2006. Comparative phylogeography of a genus of coral reef fishes: biogeographical and genetical concordance in the Caribbean. Mol. Ecol. 15:695-707.

Templeton, A., K. A. Crandall, and C. F. Sing. 1992. A cladistic analysis of phenotypic associations with haplotypes inferred from restriction endonuclease mapping and DNA sequence data. III. Cladogram estimation. Genetics 132: 619-633.

64 Thibert-Plante, X. and A. P. Hendry. 2011. The consequences of phenotypic plasticity for ecological speciation. J. Evol. Biol. 24:326–342.

Thornhill, D. J., A. M. Lewis, D. C. Wham, and T. C. LaJeunesse. 2014. Host-specialist lineages dominate the adaptive radiation of reef coral endosymbionts. Evolution 68:352– 367.

Thornhill, D. J., Y. Xiang, W. K. Fitt, and S. R. Santos. 2009. Reef endemism, host specificity and temporal stability in populations of symbiotic dinoflagellates from two ecologically dominant Caribbean corals. PLoS ONE 4:e6262.

Thornhill, D. J., Y. Xiang, P. M. Pettay, M. Zhong, and S. R. Santos. 2013. Population genetic data of a model symbiotic cnidarian system reveal remarkable symbiotic specificity and vectored introductions across ocean basins. Mol. Ecol. 22:4499-4515.

Tomaiuolo, M., T. F. Hansen, and D. R. Levitan. 2007. A theoretical investigation of sympatric evolution of temporal reproductive isolation as illustrated by marine broadcast spawners. Evolution 61:2584-2595.

Tracey, D., H. Neil, P. Marriott, A. Andrews, G. Cailliet, and J. Sánchez. 2007. Age and growth of two genera of deep-sea bamboo corals (Family Isididae) in New Zealand waters. Bull. Mar. Sci. 81:393–408.

Van Doorn, G. S., P. Edelaar, and F. J. Weissing. 2009. On the origin of species by natural and sexual selection. Science 326:1704–1707. van Oppen, M., P. Bongaerts, J. Underwood, L. Peplow, and T. Cooper. 2011. The role of deep reefs in shallow reef recovery: an assessment of vertical connectivity in a brooding coral from west and east Australia. Mol. Ecol. 20:1647-1660.

Vermeij, G. J. 2005. One-way traffic in the western Atlantic: causes and consequences of Miocene to early Pleistocene molluscan invasions in Florida and the Caribbean. Paleobiology 31:624–642.

Voolstra, C. R., J. A. Schwarz, J. Schnetzer, S. Sunagawa, M. K. DeSalvo, A. M. Szmant, M. A. Coffroth, and M. Medina. 2009. The host transcriptome remains unaltered during the establishment of coral-algal symbioses. Mol. Ecol. 18:1823-1833.

Waddington, C. H. 1942. Canalization of development and the inheritance of acquired characters. Nature 150:563-565.

Watano, Y., A. Kanai, and N. Tani. 2004. Genetic structure of hybrid zones between Pinus pumila and P. parviflora var. pentaphylla (Pinaceae) revealed by molecular hybrid index analysis. Am. J. Bot. 91:65–72.

65 Weil, E. and N. Knowlton. 1994. A multi-character analysis of the Caribbean coral Montastraea annularis (Ellis and Solander, 1786) and its two sibling species, M. faveolata (Ellis and Solander, 1786) and M. franksi (Gregory, 1895). Bull. Mar. Sci. 55:151-175.

West, J. M., C. D. Harvell, and A. M. Walls. 1993. Morphological plasticity in a gorgonian coral (Briareum asbestinum) over a depth cline. Mar. Ecol. Prog. Ser. 94:61-69.

West-Eberhard, M. J. 2003. Developmental Plasticity and Evolution. Oxford University Press, New York.

Whitehead, A. and D. Crawford. 2006. Neutral and adaptive variation in gene expression. Proc. Natl. Acad. Sci. USA 103:5425-5430.

Whitehead, A., J. Roach, S. Zhang, and F. Galvez. 2011. Genomic mechanisms of evolved physiological plasticity in killifish distributed along an environmental salinity gradient. Proc. Natl. Acad. Sci. USA 108:6193-6198.

Williams, C. G. 2010. Long-distance pine pollen still germinates after meso-scale dispersal. Am. J. Bot. 97:846–855.

Williams, G. 1975. Sex and evolution. Princeton University Press., Princeton, New Jersey.

Wirshing, H., K. A. Feldhein, and A. W. Baker. 2013. Vectored dispersal of Symbiodinium by larvae of a Caribbean gorgonian octocoral. Mol. Ecol. 22:4413–4432.

Yoshioka, P. 2009. Sediment transport and the distribution of shallow-water gorgonians. Caribb. J. Sci. 45:254-259.

Yoshioka, P. M. and B. B. Yoshioka. 1991. A comparison of the survivorship and growth of shallow-water gorgonian species of Puerto Rico. Mar. Ecol. Prog. Ser. 69:253-260.

66 Appendix A Supplementary Information for Chapter 2

Figure 2.1S. Parsimony haplotype networks for each locus. The size of the circles is proportional to haplotypes frequency. Nuclear markers have twice as many haplotypes.

Mismatch Inositol 3 Elongation factor Calcium repair gene phosphatase 1 alpha (nuclear) transporter 2 (mtDNA) (nuclear) (nuclear)

Shallow habitats

Deep habitats

67 Figure 2.2S. Migration rate estimates between Shallow and Deep lineages per location obtained by fitting the IMa model to all 4 loci. Samples were divided by habitat (top), lineage (middle), and morphology (bottom). Solid lines represent migration from Shallow to Deep, dotted lines from Deep to Shallow. Note that migration estimates by lineage are lower than for habitat or morphology, but that the probabilities of the estimates are higher.

2 Puerto Rico By habitat Bahamas Curaçao Panama 1

0 0 1 2 3 4 5 6 6 5 By lineage 4 3 2 Probability Probability 1 0 0 0.5 1 4

3 By morphology

2

Probability Probability 1 Probability

0 0 1 2 3 4 5 6

Migration (m/μ)

68 Table S2.1. Annual survivorship per size class for E. flexuosa across reefs and depths over four years.

Size Survivorship n category (cm)

< 10 0.691 151 11 - 30 0.823 81 > 31 0.901 24

All 0.805 256

Table S2.2. Annual survivorship for reciprocally transplanted colonies for two independent experiments.

Survivorship rate per year (n) Previous Current Treatment (2006 -2007)* (2010 – 2012) Average deep to shallow 0.844 (30) 0.725 (40) 0.785 (70) shallow to shallow 0.867 (15) 0.875 (40) 0.871 (55) deep to deep 0.956 (15) 0.925 (40) 0.940 (55) shallow to deep 0.622 (30) 0.638 (40) 0.630 (70) * Extracted from (Prada et al. 2008).

Table S2.3. Localities of the collections and number of colonies sampled per depth (N).

Location Coordinates Reef Depth N

Panama 9.231 N, 82.137 W Crawl Key < 5m 13 - Crawl Key > 15m 13 9.346 N, 82.212 W Bastimentos < 5m 12 Bahamas 23.468 N; 76.062 W Lee Stocking Island < 5m 21 25.570 N, 77.210 W Cross Harbor Ridge > 25m 20 Curaçao 12.083 N, 68.897 W Aquarium < 5m 17 12.134N, 68.985 W Dive Inn > 25m 17 Puerto Rico 17.560 N, 67.029 W Media Luna < 5m 38 - Media Luna 20m 38

Table S2.4. Within-lineage pairwise Fst values. Shallow and Deep lineage comparisons are below and above the diagonal, respectively. Significant values after Bonferroni correction are underlined.

Puerto Rico Panama Bahamas Curaçao Puerto Rico 0.101 0.140 0.117 Panama 0.007 0.105 0.060 Bahamas 0.032 0.005 0.138 Curaçao 0.029 0.002 0.062

69 Table S2.5. Information theoretic statistics for each of the IMa models for each partition (lineage, habitat and morphology).

Lineage Evidence Model k - log(P) AIC i Likelihoods wi Δ Ratio FULL 5 3.300 3.400 0.000 1.000 0.624 AAADE 3 0.954 7.908 4.508 0.105 0.065 9.528 ABC0D 4 0.001 8.000 4.600 0.100 0.063 9.974 ABADE 4 0.001 8.000 4.600 0.100 0.063 9.974 ABBDE 4 0.001 8.000 4.600 0.100 0.063 9.974 AACDE 4 0.051 8.103 4.703 0.095 0.059 10.499 ABCDD 4 0.826 9.652 6.252 0.044 0.027 22.783 ABBDD 3 2.040 10.079 6.679 0.035 0.022 28.205 ABADD 3 2.672 11.344 7.944 0.019 0.012 53.096 AACDD 3 4.281 14.561 11.161 0.004 0.002 265.257 AAADD 2 8.078 20.155 16.755 < 1-15 < 1-15 > 1015 ABCD0 4 51.743 111.486 108.086 < 1-18 < 1-18 > 1017 ABC00 3 87.159 180.318 176.918 < 1-18 < 1-18 > 1017 ABB00 2 89.281 182.561 179.161 < 1-18 < 1-18 > 1017 ABA00 2 91.660 187.320 183.920 < 1-18 < 1-18 > 1017 AAC00 2 101.821 207.642 204.242 < 1-18 < 1-18 > 1017 AAA00 1 105.853 213.706 210.306 < 1-18 < 1-18 > 1017 Habitat FULL 5 3.300 3.400 0.000 1.000 0.900 AACDD 3 2.444 10.887 7.487 0.024 0.021 42.246 ABBDD 3 2.608 11.215 7.815 0.020 0.018 49.784 ABBDE 4 1.711 11.421 8.021 0.018 0.016 55.185 ABCDD 4 1.806 11.612 8.212 0.016 0.015 60.697 AACDE 4 2.417 12.834 9.434 0.009 0.008 111.855 ABADD 3 3.585 13.170 9.770 0.008 0.007 132.317 AAADE 3 3.773 13.545 10.145 0.006 0.006 159.605 ABADE 4 2.982 13.964 10.564 0.005 0.005 196.743 AAADD 2 5.058 14.117 10.717 0.005 0.004 212.364 ABC0D 4 192.035 392.070 388.670 < 1-15 < 1-15 > 1015 ABCD0 4 305.298 618.597 615.197 < 1-18 < 1-18 > 1017 AAA00 1 460.517 923.034 919.634 < 1-18 < 1-18 > 1017 AAC00 2 460.517 925.034 921.634 < 1-18 < 1-18 > 1017 ABA00 2 460.517 925.034 921.634 < 1-18 < 1-18 > 1017 ABB00 2 460.517 925.034 921.634 < 1-18 < 1-18 > 1017 ABC00 3 460.517 927.034 923.634 < 1-18 < 1-18 > 1017 Morphology FULL 5 3.300 3.400 0.000 1.000 0.764 ABBDD 3 0.627 7.254 3.854 0.146 0.111 6.870 ABCDD 4 0.505 9.010 5.610 0.060 0.046 16.530 ABBDE 4 0.620 9.240 5.840 0.054 0.041 18.541 ABADD 3 2.347 10.693 7.293 0.026 0.020 38.344 ABADE 4 2.239 12.479 9.079 0.011 0.008 93.635 AACDE 4 2.730 13.460 10.060 0.007 0.005 152.918 AACDD 3 3.904 13.807 10.407 0.005 0.004 181.908 AAADE 3 9.049 24.098 20.698 0.000 0.000 31228.924 AAADD 2 10.992 25.984 22.584 0.000 0.000 80185.662 ABCD0 4 84.222 176.443 173.043 < 1-15 < 1-15 > 1015 ABC0D 4 100.622 209.243 205.843 < 1-18 < 1-18 > 1017 ABB00 2 299.551 603.101 599.701 < 1-18 < 1-18 > 1017 ABC00 3 298.786 603.573 600.173 < 1-18 < 1-18 > 1017 ABA00 2 306.990 617.980 614.580 < 1-18 < 1-18 > 1017 AAC00 2 323.727 651.454 648.054 < 1-18 < 1-18 > 1017 AAA00 1 326.608 655.216 651.816 < 1-18 < 1-18 > 1017

The first three letters of the models represent the populations sizes for the three lineages (Deep, Shallow and ancestral). The next two letters (or 0) indicate the migration from Deep to Shallow and from Shallow to Deep. Lineages with the same letter indicate same population size and zero indicaticates no migration. AIC = Akaike information criterion inferred from the generalized linear model fitting. ∆i = Difference in AIC score with respect to the best model. Model Likelihoods = Relative likelihood of the model given the data. wi = Model probabilities. Evidence Ratio = Difference in model probabilities between the proposed model and the best model (ABC0D).

70 Table S2.6. Parameter estimates and boundaries of the 90% highest posterior density (HPD) for the Isolation with Migration model between the Shallow and Deep lineages. The geometric mean of the mutation rate per year was 4.96 x 10-7. The posterior distribution for the time parameter was flat. 90 HPD was estimated as the widest range across the three runs. Estimates from Bahamas, Curaçao and Panama should be interpreted with caution as the shallow and deep habitats may represent intermediate areas and not the strict opposite sides of the depth gradient.

θ1 θ2 θa m1/µ m2/µ t/µ N1 N2 N1m1 N2m2 Habitat All samples1 6.69 4.17 3.29 1.09 1.61 1.14 224684 140036 1.82 1.68 All samples2 6.70 4.22 3.53 0.97 1.75 1.09 224916 141753 1.63 1.84 Average 6.69 4.19 3.41 1.03 1.68 1.11 224800 140895 1.73 1.76 90HPD lower 4.83 3.08 1.37 0.46 1.02 0.31 162387 103619 0.56 0.79 90HPD Upper 8.37 5.33 5.66 1.48 2.47 1.81 281161 179091 3.09 3.30 Bahamas1 3.40 2.25 4.92 0.10 0.49 4.81 114126 75512 0.08 0.28 Bahamas2 3.38 2.27 4.86 0.10 0.49 4.86 113508 76214 0.09 0.28 Bahamas3 3.42 2.32 3.92 0.10 0.45 3.15 115002 77802 0.09 0.26 Curaçao1 2.29 2.74 4.29 0.84 2.33 3.43 76872 91933 0.48 1.60 Curaçao2 2.27 2.56 4.77 0.98 3.26 5.64 76204 85921 0.56 2.09 Curaçao3 2.25 2.74 4.21 0.90 2.31 3.47 75666 92128 0.51 1.58 Panama1 6.29 0.88 4.51 2.52 1.48 2.58 211151 29427 3.95 0.32 Panama2 6.05 0.86 4.71 2.51 1.50 3.09 203130 28994 3.79 0.32 Panama3 6.80 0.82 4.81 3.91 2.37 3.13 228325 27419 6.64 0.48 Puerto Rico1 3.80 2.56 4.88 0.05 0.66 3.78 127726 85850 0.04 0.42 Puerto Rico2 3.88 2.59 4.79 0.05 0.66 3.39 130198 87144 0.05 0.43 Puerto Rico3 3.81 2.52 4.91 0.05 0.67 3.81 127948 84729 0.04 0.42 Lineage All samples1 6.20 4.06 3.42 0.05 0.40 1.44 208279 136338 0.08 0.40 All samples2 6.35 3.79 3.77 0.05 0.41 1.71 213425 127215 0.08 0.38 Average 6.28 3.92 3.60 0.05 0.40 1.57 210852 131777 0.08 0.39 90HPD lower 4.76 2.88 0.60 0.00 0.20 0.81 159911 96814 0.00 0.14 90HPD Upper 7.47 4.87 6.20 0.11 0.57 3.00 250850 163623 0.20 0.69 Bahamas1 3.67 2.10 3.39 0.10 0.25 2.66 123222 70510 0.09 0.13 Bahamas2 3.71 2.09 3.46 0.09 0.26 2.70 124616 70252 0.09 0.14 Curaçao1 2.96 1.12 4.04 0.13 1.19 3.43 99279 37680 0.10 0.33 Curaçao2 2.93 1.12 4.18 0.14 1.19 3.58 98453 37522 0.10 0.33 Panama1 2.86 1.24 3.89 0.18 0.41 3.35 96095 41661 0.13 0.13 Panama2 2.59 1.16 1.18 0.10 0.30 2.13 86915 38842 0.07 0.09 Puerto Rico1 3.85 2.36 4.57 0.05 0.41 3.64 129395 79270 0.04 0.24 Puerto Rico2 3.79 2.35 4.61 0.04 0.40 3.55 127299 79076 0.04 0.24 Puerto Rico3 3.91 2.47 4.22 0.05 0.38 3.46 131246 83116 0.05 0.23 Morph All samples1 6.91 3.50 3.49 0.62 0.67 1.43 231993 117686 1.06 0.59 All samples2 6.95 3.54 3.75 0.62 0.70 1.30 233293 119003 1.08 0.62 Average 6.93 3.52 3.62 0.62 0.68 1.37 232643 118345 1.07 0.60 90HPD lower 5.07 2.51 1.00 0.31 0.33 0.58 170428 84440 0.39 0.21 90HPD Upper 8.79 4.47 5.79 0.93 1.01 2.27 295390 150016 2.03 1.13 Bahamas1 3.66 2.08 3.30 0.10 0.26 2.78 122842 69976 0.09 0.14 Bahamas2 3.66 2.07 3.37 0.09 0.27 2.81 122916 69469 0.09 0.14 Curaçao1 3.19 1.10 4.09 0.30 1.81 3.30 106991 36961 0.24 0.50 Curaçao2 3.15 1.09 4.20 0.30 1.82 3.60 105856 36592 0.24 0.50 Panama1 6.49 0.90 4.68 2.59 1.67 3.04 218074 30193 4.21 0.38 Panama2 6.41 0.89 4.68 2.58 1.68 3.11 215212 29988 4.13 0.38 Puerto Rico1 3.84 2.36 4.53 0.04 0.41 3.67 128989 79196 0.04 0.24 Puerto Rico2 3.85 2.38 4.38 0.04 0.39 3.50 129412 79858 0.04 0.23

*Estimates inferred from Puerto Rico Only. θ1 = 4N1u The population mutation rate Shallow. θ2 = 4N2u The population mutation rate for Deep. m1/µ = The migration rate, per mutation, from Deep to Shallow. m2/µ = The migration rate, per mutation, from Shallow to Deep. t/µ = Divergence time scaled by mutation rate. Estimates should be interpreted carefully as the distribution was flat. N1 = Effective size of Shallow. N2 = Effective size of Deep. NA = Ancestral effective population size. N1m1 = Migrants per generation from Deep to Shallow. N2m2 = Migrants per generation from Shallow to Deep.

71

Table S2.7. Genetic markers used in this study with closest blastx match. *Primers from France and Hoover (2001).

Code Blastx match Primer sequence Size (bp) XP001638916.1 5’TGGAAGAAGCTAAAAGGAGTGATAC3’ Inositol 3 phosphatase I3P 369-370 Nematostella vectensis 5’TTGTTCTCCTGCTTTGAAGTTACC3’ BAB63212.1 5’GTGTCGAGACAGGAGTTCTCAAG3’ Elongation factor 1 alpha EF1a 401 Ciona intestinalis 5’GCAATATGAGCAGTGTGACAATC3’ 5’ATTCTGACACAGCCTACAGAGTC3’ Calcium transporter 2 CT2 XP001641230.1 263 5’TATATGTTCCACATAGCAGCTTGG3’ ABV26565.1 *ND42599F: 5'GCCATTATGGTTAACTATTAC 3' Mismatch repair gene MSH 631 Nematostella vectensis Mut- 3458R: 5' TSGAGCAAAAGCCACTCC 3'

72 Table S2.8. Summary of genetic variation within localities. All measurements after excluding gaps.

1 2 3 4 5 6 Locus Population N S h Hd π *100 θW *100 MSH PA_S 10 7 4 0.71 0.27 0.40 CU_S 25 3 4 0.58 0.17 0.13 BA_S 22 3 4 0.52 0.11 0.13 PR_S 40 5 6 0.60 0.11 0.19 PA_D 28 5 3 0.66 0.31 0.21 CU_D 9 7 4 0.58 0.29 0.41 BA_D 19 9 6 0.60 0.44 0.41 PR_D 36 8 5 0.61 0.18 0.31 All_S 95 13 10 0.60 0.12 0.41 All_D 94 12 9 0.76 0.35 0.38 All 189 19 16 0.78 0.44 0.52

I3P PA_S 20 8 9 0.83 0.43 0.62 CU_S 50 11 14 0.74 0.40 0.73 BA_S 44 10 11 0.72 0.45 0.69 PR_S 80 12 12 0.58 0.30 0.71 PA_D 56 19 10 0.68 0.44 1.12 CU_D 18 16 6 0.72 1.09 1.26 BA_D 38 13 10 0.65 0.34 0.84 PR_D 72 22 13 0.61 0.51 1.23 All_S 194 24 26 0.69 0.37 1.16 All_D 184 34 24 0.69 0.53 1.60 All 378 52 47 0.84 1.72 2.21

EF1A PA_S 20 10 8 0.87 0.64 0.71 CU_S 50 10 10 0.82 0.58 0.56 BA_S 44 13 16 0.88 0.78 0.75 PR_S 80 14 17 0.92 0.74 0.71 PA_D 56 3 4 0.34 0.09 0.16 CU_D 18 5 4 0.31 0.21 0.36 BA_D 38 2 3 0.24 0.06 0.12 PR_D 72 13 8 0.26 0.14 0.67 All_S 194 18 24 0.90 0.72 0.82 All_D 184 18 13 0.29 0.12 0.78 All 378 31 32 0.77 0.60 1.23

CT2 PA_S 20 6 6 0.76 0.63 0.64 CU_S 50 5 5 0.60 0.47 0.42 BA_S 44 6 7 0.73 0.52 0.52 PR_S 80 9 11 0.75 0.54 0.69 PA_D 56 6 6 0.37 0.24 0.58 CU_D 18 6 5 0.55 0.42 0.66 BA_D 38 7 6 0.45 0.39 0.63 PR_D 72 8 7 0.58 0.38 0.63 All_S 194 13 15 0.72 0.54 0.85 All_D 184 12 13 0.50 0.35 0.85 All 378 23 26 0.80 0.65 1.40

1 Number of chromosomes sampled per population. Number of individuals for mtDNA. 2 Number of segregating sites. 3 Number of haplotypes per population. 4 Haplotypic diversity (maximum = 1). 5 Nucleotide diversity. 6 Watterson’s estimator of genetic diversity.

73 Appendix B Supplementary Information for Chapter 3

Figure S3.1. Bayesian clustering for the short dataset for all microsatellites for Symbiodinium B1/184. K = 2 as suggested by the Evanno method.

Puerto'Rico' Puerto'Rico' Puerto'Rico' Curaçao' Bahamas' Curaçao' Bahamas' shallow' deep' mid_depth' 'deep' deep' shallow' shallow' 1" " 0.8" " 0.6" " 0.4" " 0.2" Probability*

of*membership*

1 2 3 4 5 6 7

74 Appendix C Supplementary Information for Chapter 4

The cline shape was modeled using the formulas described by (Szymura and Barton 1986, 1991) as implemented in HZAR (Derryberry et al. 2014). The three expressions are:

! !(!!!) (a) � = 1 + ���ℎ ! !

!(!! !!! ) ! ! (b) � = ��� ! ! !

!!(!! !!! ) ! ! (c) � = 1 − ��� ! ! !

Where, c is the cline center (measured in meters from the shallowest site), w is the cline width, and x is the distance in meters along the transect. zL and zR is the distance left or right from the cline center (c) and θL and θR are the exponential decay left or the right of those distances from the cline center (tail slope). When, z and θ reach zero and one, equations b and c for either left or right approximate equation a. The cline center (c) is the distance at which the modeled character changes most rapidly and the cline width (w) approximates the spatial scale in which the change occurs.

75 Table S4.1. Corrected Akaike information criteria for each model for each marker. The model can have either no tails, left or right, symmetrical (mirror) or asymmetrical tails. Frequencies can be modeled as fixed minimum and maximum value from the data, as a free parameter or as none in which case they vary from 0 to 1.

Spindle Calice Branch model msh gly3p ct2 i3p ef1 564 psbA length diameter thickness free, none 12.63 10.32 10.91 16.24 12.88 9.19 16.63 12.38 8.84 10.21 free, left 16.98 13.05 14.46 20.04 16.99 13.34 16.70 16.68 13.05 13.62 free, right 16.99 14.46 15.01 20.35 17.01 13.33 20.79 16.65 13.06 14.59 free, both 21.14 17.21 18.20 24.08 21.20 17.54 21.28 21.10 17.54 17.91 free, mirror 16.84 14.36 14.67 20.29 16.96 13.37 20.63 16.58 13.09 13.77 fixed, none 8.41 7.52 8.31 14.42 29.61 5.16 14.53 11.21 5.02 6.48 fixed, left 12.09 8.88 9.31 15.60 13.38 9.12 11.35 11.88 8.75 9.13 fixed, right 12.59 11.61 12.40 18.51 33.70 9.25 18.62 15.39 9.20 10.65 fixed, both 16.84 13.18 13.56 19.89 17.67 13.29 15.56 16.19 13.13 13.21 fixed, mirror 12.59 11.56 12.23 17.24 15.92 9.21 18.43 11.89 9.20 10.52 none, none 24.61 21.30 30.19 24.95 40.57 50.02 12.06 10.46 9.45 12.87 none, left 13.09 9.75 21.61 16.20 14.73 9.44 12.70 13.01 9.02 10.02 none, right 28.79 25.40 25.43 29.03 44.67 15.89 16.15 14.18 13.64 17.05 none, both 17.41 14.20 28.33 19.82 17.97 13.59 16.19 16.55 13.25 13.67 none, mirror 28.79 25.40 25.42 29.02 13.13 9.45 14.92 12.19 10.02 11.40

76 Appendix D Agreement to Reproduce Chapter 2

Below is the authorization from PNAS to reproduce Chapter 2.

Reprint permission request, manuscript number 201208931.

Thank you for your message. Authors do not need to obtain permission for the following uses of material they have published in PNAS: (1) to use their original figures or tables in their future works; (2) to make copies of their papers for their classroom teaching; or (3) to include their papers as part of their dissertations.

Please note that citation to the original source should be included (full journal reference). Because this material published after 2008, a copyright notice is not needed. Feel free to contact us with any additional questions you might have.

Best regards, Kay McLaughlin for Diane Sullenberger Executive Editor PNAS

77 Appendix E Agreement to Reproduce Chapter 3

Below is the authorization from Molecular Ecology to reproduce Chapter 3.

78 Vita

Carlos was born in Bogotá, Colombia, he did his undergraduate degree in Marine Biology at the Jorge Tadeo Lozano University, where he developed a special interest in corals. He graduated with honors, having completed a senior project looking at the effects of bleaching on coral growth rates. He then moved to Puerto Rico, where under the supervision of Paul Yoshioka and Nick Schizas he graduated with a Master of Science. His thesis looked at variation in morphological characters across depths in octocorals. His time in Puerto Rico transformed his career, as Carlos was able to immerse himself in coral reef ecology, providing the foundations for his latter work as doctoral student. In Puerto Rico, he also developed an interest in educating the wider public and participated in numerous outreach activities, such bringing elementary kids in touch with marine life, participating in forums to develop a management plan for La Parguera reserve, and giving informal talks about marine conservation to K-12 teachers. He then moved to Baton Rouge, where he initiated his doctoral studies under the guidance of Mike Hellberg. Carlos developed a project on the role of natural selection in speciation. While in Baton Rouge, Carlos strengthened his ties with tropical biology and as such was able to spend time in Curaçao, The Bahamas, Panama, Puerto Rico and Colombia. He also continued his contact with the public and annually participated in Ocean Commotion, an activity that brings kids from Baton Rouge to the excitement of marine life. Carlos will soon join the Medina lab at Pennsylvania State University to work on the genomics of coral adaptation and its application to coral reef restoration.

79