Evolutionary Genetics of Pocillopora Corals
by
David J. Combosch
Diploma Environmental Scientist, University Duisburg-Essen
A dissertation submitted to the Faculty of the College of Science of Northeastern University In partial fulfillment of the requirements for the degree of
Doctor of Philosophy
December 2, 2013
Dissertation directed by Steven V. Vollmer Associate Professor of Marine & Environmental Sciences
DEDICATION
Für Walter Combosch
ii ACKNOWLEDGEMENTS
I am grateful to my advisor Steven Vollmer for his support, guidance and friendship. I thank all my committee members, Rebeca Rosengaus, Matthew Bracken, Geoff Trussell and especially Gonzalo Giribet for their support and advice and my German mentor Helmut
Schuhmacher for many years of support.
In addition, I want to thank many colleagues and friends from the Vollmer Lab, especially Silvia Libro, Liz Hemond, Sarah Gignoux-Wolfsohn, Chris Marks and Stefan
Kaluziak as well as the Giribet Lab, especially Vanessa Gonzales and Alicia Perez-Porro and my collaborators Lauren Toth, Rich Aronson and Ian McIntyre. I benefitted greatly from the Three
Sea Program, particularly due to the help of Sal Genovese, Pete Edmunds, Ruth Gates, Brian
Helmuth, Josh Idjahdi, Jim Leichter and Bill Precht.
Numerous friends in America, Germany, Panamá, France, French Polynesia and especially in 123 River Street contributed a lot, both directly and indirectly, to this dissertation, especially my girlfriend Sarah Lemer, my brother Florian and his family Fiona, Liva, Clara und
Bruno as well as my parents Lene and Walter Combosch.
This work was supported by a German Academic Exchange Service Pre-doctoral fellowship, Teaching assistantship awards from Northeastern University and the Three Seas
Program and the National Science Foundation to Steven Vollmer.
iii ABSTRACT OF DISSERTATION
My dissertation research addresses fundamental questions about the evolution of tropical reef corals in isolated biogeographic regions. The main focus is on the effect of intra- and inter- specific gene flow within and among peripheral coral populations. Since introgression and genetic connectivity are keys factor for evolution and adaptation as well as for population resilience, recovery and persistence, the identification and quantification of past and present gene flow is of major interest, not only for basic evolutionary research but also for applied conservation.
The central question of my research is the how the widespread, but generally subordinate reef-building coral Pocillopora damicornis became the ecologically dominant reef-builder in the isolated, environmentally challenging Tropical Eastern Pacific (TEP). The dominance of TEP
Pocillopora damicornis populations is accompanied by strong differences in life-history traits, including a shift in reproductive strategy from brooding predominantly parthenogenetic larvae in the Indo-West Pacific to broadcast spawning in the TEP.
Using a combination of genetic tools ranging from multi-locus Sanger sequencing to
RAD-Seq phylogenomic analyses, I show that three genetic lineages that are found in TEP and
Central Pacific P. damicornis populations are characterized by strong genome-wide differences, which suggests that they belong to separate species. However, a subset of loci was shared among
P. damicornis colonies across lineages and across large geographic distances. The allelic composition of putative hybrid samples indicate repeated uniparental backcrossing and substantial genomic introgression among TEP Pocillopora species.
The population genetic structure of TEP P. damicornis indicates limited gene flow among and within populations, which enables effective adaptations to the environmentally
iv heterogeneous TEP environment. Predominantly sexual reproduction, significant inbreeding and small-scale spatial genetic structures over meter scales within populations further sustain divergent adaptations. This small-scale population genetic structure allows P. damicornis to adapt locally to the harsh environmental condition of the Tropical Eastern Pacific and presumably facilitated its regional dominance.
In addition, I show that Central Pacific P. damicornis display a unique mixed reproductive strategy that includes the simultaneous production of sexual and asexual larvae.
While the majority of larvae are produced parthenogenetically, most colonies also produce a subset of their larvae sexually. The proportion of sexually produced larvae decreases with colony size, suggesting that the mixed reproductive strategy is flexible and changes across the life of the coral, leading to an increasingly asexual propagation of locally successful genotypes.
Combined, I have shown that Pocillopora corals in the peripheral TEP are characterized by interspecific gene flow between P. damicornis and its congeners P. eydouxi and P. elegans and limited gene flow among and within P. damicornis populations. I showed that the morphospecies Pocillopora damicornis is composed of multiple genetic lineages, in part due to introgressive hybridization in the TEP. Pocillopora populations further differ in their reproductive strategies with central Pacific populations reproducing predominantly parthenogenetic and TEP populations reproducing mostly sexually.
v TABLE OF CONTENTS
DEDICATION II
ACKNOWLEDGEMENTS III
ABSTRACT OF DISSERTATION IV
TABLE OF CONTENTS VI
LIST OF FIGURES IX
LIST OF TABLES XI
CHAPTER 1 - GENE FLOW, POCILLOPORA CORALS AND THE TROPICAL EASTERN PACIFIC 1
Overview 1
Gene Flow 2
Pocillopora corals 7
The Tropical Eastern Pacific 10
References 14
CHAPTER 2 - POPULATION GENETICS OF AN ECOSYSTEM-DEFINING REEF CORAL - POCILLOPORA DAMICORNIS IN THE TROPICAL EASTERN PACIFIC 25
Abstract 25
Introduction 26
Materials and Methods 30
Results 35
Discussion 38
vi Conclusion 49
Acknowledgements 50
References 51
Tables 62
Figures 67
Supplementary Material 70
Chapter 3 - MIXED ASEXUAL AND SEXUAL REPRODUCTION IN THE INDO-PACIFIC REEF CORAL POCILLOPORA DAMICORNIS 71
Abstract 71
Introduction 72
Materials and Methods 74
Results 77
Discussion 78
Conclusion 85
Acknowledgements 86
References 87
Tables 93
Figures 96
CHAPTER 4 - EVOLUTIONARY GENOMICS OF POCILLOPORA CORALS 101
Abstract 101
Introduction 102
Material and Methods 105
Results 112
Discussion 116
vii Conclusions 125
Acknowledgements 126
References 127
Tables 133
Figures 136
Supplementary Material 143
FINAL CONCLUSIONS 149
viii LIST OF FIGURES
CHAPTER 2
FIGURE 1: MAP OF PANAMÁ, SHOWING THE THREE MAIN PACIFIC REGIONS AND THE SAMPLED LOCATIONS. 67
FIGURE 2: PRINCIPAL COMPONENT ANALYSIS OF POCILLOPORA DAMICORNIS POPULATIONS, CONSTRUCTED USING GENALEX 6.1. 68
FIGURE 3: SPATIAL CORRELOGRAMS OF THE SPAGEDI ANALYSIS OF SPATIAL GENETIC STRUCTURE (SGS) AMONG SPECIMEN WITHIN DISCRETE DISTANCE CLASSES. 69
FIGURE S1: RESULTS OF THE BAYESIAN CLUSTERING APPROACH IMPLEMENTED IN STRUCTURE 2.3 THAT WERE USED TO INFER THE NUMBER OF GENETIC CLUSTERS (K) IN THE MICROSATELLITE DATASET 70
CHAPTER 3
FIGURE 1: A POCILLOPORA DAMICORNIS COLONY ON THE REEF IN FRONT OF THE GUMP RESEARCH STATION IN MOOREA, FRENCH POLYNESIA 96
FIGURE 2: THE PROPORTION OF SEXUALLY PRODUCED LARVAE DECREASES AS A FUNCTION OF COLONY SIZE 97
FIGURE 3: THE PROPORTION OF SEXUALLY PRODUCED LARVAE RELEASED DECREASES DURING THE REPRODUCTIVE CYCLES OF P. DAMICORNIS COLONIES 99
CHAPTER 4
FIGURE 1A: POCILLOPORA ELEGANS COLONY IN A P. DAMICORNIS CARPET IN ISLA COIBA, GULF OF CHIRIQUÍ, PANAMÁ. 136
FIGURE 1B: PHENOTYPIC PLASTICITY OF P. DAMICORNIS/P. VERRUCOSA IN MOOREA, FRENCH POLYNESIA. 137
ix FIGURE 2: ORF AND ITS PHYLOGENY 138
FIGURE 3: RAD PHYLOGENOMIC ANALYSES 140
FIGURE 4: PRINCIPAL COORDINATE ANALYSIS OF PAIRWISE P-DISTANCES BETWEEN RAD-SEQ SAMPLES. 141
FIGURE 5: ALLELE COMPOSITION OF PUTATIVE HYBRID P. DAMICORNIS SAMPLES COMPARED TO THE EXPECTED ALLELE COMPOSITIONS OF BACKCROSSING HYBRIDS OVER SEVERAL GENERATIONS. 142
FIGURE S1: BASIC RAD LOCI CORRELATIONS 147
FIGURE S2: PAIRWISE P-DISTANCE COMPARISONS BETWEEN SAMPLES AT TRANSCRIBED VS. UNTRANSCRIBED LOCI. 148
x LIST OF TABLES
CHAPTER 2
TABLE 1: POPULATION GENETIC INDICES FOR THE SIX MICROSATELLITE LOCI. 62
TABLE 2: POPULATION GENETIC INDICES FOR THE NINE POPULATIONS. 63
TABLE 3: HIERARCHICAL AMOVA RESULTS 64
TABLE 4: PAIRWISE RST AND FST BETWEEN TEP P. DAMICORNIS POPULATIONS 65
TABLE 5: POCILLOPORA DAMICORNIS POPULATION GENETIC SURVEY (SORTED BY FST). 66
CHAPTER 3
TABLE 1: LOCI-SPECIFIC STATISTICS FOR EACH OF THE SIX SAMPLED MICROSATELLITE LOCI. 93
TABLE 2: LIST OF MATERNAL COLONIES 94
TABLE 3: RESULTS OF THE GENERALIZED LINEAR MIXED EFFECTS MODEL FOR THE MAIN LOGISTIC REGRESSION ANALYSIS 95
CHAPTER 4
TABLE 1: OVERVIEW OF THE RAD-SEQ SAMPLE COMPOSITION AND BASIC POPULATION GENETIC INDICES. 133
TABLE 2: PAIRWISE COMPARISONS BETWEEN SAMPLES USING P-DISTANCE AND IBS- PROPORTIONS. 134
TABLE 3: GENETIC MAKE-UP OF PUTATIVE HYBRID P. DAMICORNIS SAMPLES 135
TABLE S1: READ DEPTH PER SAMPLE AND PER LOCUS 143
xi TABLE S2: TRANSCRIPTOME MATCHES/ALIGNMENTS. 144
TABLE S3: IBS & NUMBER OF DIFFERENTLY FIXED LOCI 145
TABLE S4: VICE VERSA COMPARISON OF ALLELIC IDENTITY 146
xii Evolutionary Genetics of Pocillopora Corals
CHAPTER 1
GENE FLOW, POCILLOPORA CORALS AND THE TROPICAL EASTERN PACIFIC
Overview
This dissertation explores the patterns and consequences of gene flow for the evolution of reef-building corals. Each chapter highlights a different aspect of gene flow, within species, between generation and across species. Gene flow acts on many different levels and in different, sometimes contradictive ways, to shape the genetic identity and delimitations of species and populations and the genetic structures between them.
Most commonly, gene flow refers to the movement of alleles from one population to another, i.e. inter-population gene flow or population connectivity. Gene flow within populations, intra-population gene flow, is usually of little concern for mobile species but in sessile marine taxa, populations can be structured internally. Both inter- and intra-population gene flow are the focus of Chapter 2. Since evolutionary consequences of gene flow require genetic mixing, evolutionary meaningful gene flow in eukaryotic taxa is tied to reproduction, which enables the observation of individual gene flow events, from parent to offspring. Chapter
3 explores these individual steps of reproductive, vertical gene flow in a predominantly asexual
P. damicornis population from the Central Pacific Islands of French Polynesia. Interspecific gene flow via introgressive hybridization is the focus of Chapter 4.
1 Gene Flow
Gene flow is one of four evolutionary forces that change allele frequencies within and among populations over time. Mutations and genetic drift alter allele frequencies randomly across the genome by generating new alleles and eliminating or fixing alleles in populations, respectively. Natural selection acts only on select loci, favoring beneficial alleles and eliminating detrimental ones. Gene flow disperses genetic variation between populations, reintroducing alleles lost to genetic drift, spreading new mutations and fueling natural selection. However, high rates of gene flow homogenize genetic differences between populations, counter divergent selection (Slatkin 1987; Slatkin & Barton 1989; Garcia-Ramos & Kirkpatrick 1997) and prevent populations from separating into different species (Dobzhansky 1937; Mayr 1963). Gene flow thus generates interactions between mutation, drift and selection and shapes the genetic identity of populations and species, their delimitations and the genetic structure between them
(Dobzhansky 1937; Huxley 1942; Wright 1943).
In addition, gene flow is also effective over ecological time scales and has a strong impact on population demography, dynamics and diversity (Hellberg et al. 2002; Gaines et al.
2007). Demography and diversity are in turn key factors for a population’s resilience, recovery and persistence and thus ultimately drive species distribution dynamics and extinction probabilities (Hellberg et al. 2002; Jones et al. 2007; Willi & Hoffmann 2009). Gene flow data are therefore increasingly used to inform the design of conservation areas like marine protected areas (Palumbi 2003; Jones et al. 2007). Other important ecological aspects of gene flow are species invasions (Lavergne & Molofsky 2007; Dlugosch & Parker 2008) and introgressive gene flow from genetically modified organisms (e.g. Wolfenbarger & Phifer 2000; Lu & Snow 2005).
2 The identification and quantification of gene flow patterns is a main focus of conservation genetics.
Tropical reef corals are important taxa for evolutionary and population genetic research because they are the architects of the planet’s most diverse ecosystem and due to their recent declines due to a wide range of mostly anthropogenic disturbances (Wilkinson 1998, 2000,
2004), such as rising sea surface temperatures and ocean acidification (Hughes et al. 2003;
Pandolfi et al. 2003; Hoegh-Guldberg et al. 2007). Genetic research over the last 15 years has revolutionized our understanding of scleractinian evolution (Romano & Palumbi 1996; Fukami et al. 2004; Fukami et al. 2008; Budd et al. 2010). For example, DNA sequence data has revealed that the order Scleractinia consists of two basal lineages, the ‘robust’ and ‘complex’ clade, that evolve separately (Romano & Palumbi 1996; Chen et al. 2002; Fukami et al. 2008) and are inconsistent with the morphology-based taxonomy (Wells 1956; Veron 2000). However, several major issues are still controversial, for example the role of introgressive hybridization in coral evolution and speciation (Veron 1995; Vollmer & Palumbi 2002; Willis et al. 2006;
Richards et al. 2008).
Interspecific Gene Flow
Interspecific gene flow via introgressive hybridization is increasingly recognized as a major source of evolutionary novelty (Arnold 1997; Mallet 2005; Hedrick 2013). The evolutionary significance of hybridization depends on the fitness of hybrids, i.e. on their ability to contribute to subsequent generations. Backcrossing of hybrids with parental species can lead to the incorporation of genetic material from one taxon into the genome of another (Anderson
1949). Introgressed alleles can provide new evolutionary raw material (e.g. Kim & Rieseberg
3 1999; Mallet 2005; Kunte et al. 2011; Abbott et al. 2013; Hedrick 2013) that is available simultaneously at multiple loci (Kim et al. 2008; Heliconius Genome Consortium 2012).
Prominent examples for adaptive introgression include bidirectional introgression of beak traits between Darwin finch species (Grant & Grant 2008), the transition of black fur colors from dogs to wolves (Anderson et al. 2009) and the transfer of mimicry wing patterns among Heliconius butterflies (Pardo-Diaz et al. 2012; Heliconius Genome Consortium 2012). Possible marine examples include two genomic hotspots of elevated unidirectional introgression between cryptic
Cliona intestinalis species (Roux et al. 2013) and the presence of Mytilus edulis alleles at higher frequencies in distant M.’galloprovencalis populations than in neighboring populations (Gosset
& Bierne 2012). However, in both cases the adaptive significance of introgressed genes has not been determined and differential introgression might be due to intrinsic factors (Gosset & Bierne
2012; Roux et al. 2013). So far, adaptive introgression has not been documented in corals.
Importantly, introgression does usually not occur at equal rates across the genome of hybridizing species (Martinsen et al. 2001; Hohenlohe et al. 2013). Genes and surrounding gene regions that are involved in hybrid sterility and inviability (Muller–Dobzhansky genes) are unable to cross species boundaries while major parts of the genome introgress unaffected (e.g.
Ting et al. 1998; Ting et al. 2001; Barbash et al. 2003; Wu & Ting 2004; Brideau et al. 2006).
These speciation genes (sensu Wu 2001; but see Nosil & Schluter 2011 for a discussion of speciation genes) or genomic islands/continents of speciation (Turner et al. 2005; Michel et al.
2010) have so far been characterized mostly for plants (Rieseberg & Blackman 2010) and
Drosophila species (e.g. Ting et al. 2001; Barbash et al. 2003; Presgraves et al. 2003). In the marine realm, gamete adhesion molecules of free-spawning invertebrates (sea urchins & abalone) have been characterized as speciation genes of sorts since it is unclear if their
4 divergence is cause or consequence of speciation (Vacquier 1998; Palumbi 2009; Lessios 2011).
The best candidate speciation gene in corals is the Mini-Collagen locus in Caribbean Acropora corals, which is blocked from introgression into A. cervicornis while several other genes introgress from A. palmata (Vollmer & Palumbi 2002). Introgressive hybridization has been shown to occur in Pocillopora corals as well and likely plays an important role in the extraordinary success of Pocillopora corals in the remote Tropical Eastern Pacific (Combosch et al. 2008).
In chapter 4, I analyze the patterns of inter-specific gene flow among Tropical Eastern
Pacific Pocillopora species using genome-wide RAD-Seq analyses.
Inter-population Gene Flow
Our understanding of intra-specific gene flow among reefs has changed significantly as well over the last decades. In contrast to the long-held paradigm that long-distance dispersal is pervasive among marine populations (Scheltema 1971; Heck & McCoy 1978; Veron 1995), increasing evidence from coral population genetic analyses, suggest widely disparate levels of population connectivity (Warner & Cowen 2002; Ayre & Hughes 2004; Jones et al. 2009).
Several studies have recently addressed fine-scale spatial genetic structures over meter scales within coral populations to assess local gene dispersal, reproductive patterns (e.g. parentage, selfing and inbreeding) and selection pressures (Underwood et al. 2007; Miller & Ayre 2008a; van Oppen et al. 2008).
The potential for local adaptation in corals has become a recent focus of coral population genetic studies. Several studies show that populations adapted to extreme environments might contain stress-resistant genotypes (West & Salm 2003; McClanahan et al. 2007; Miller & Ayre
5 2008b). In addition, corals cope with thermal extremes by shifting between different types of their main algal symbionts (Baker et al. 2004; Rowan 2004; Jones 2008) or host differently adapted symbiont populations (Howells et al. 2011). Divergent adaptations between populations, mostly to different temperature regimes, have been documented over various spatial scales from hundreds of meters to hundreds of kilometers (e.g. D'Croz & Mate 2004; Smith-Keune & van
Oppen 2006; Ulstrup et al. 2006; Smith et al. 2007; Barshis et al. 2010). Adaptations to different environments are facilitated by limited gene flow between populations to increase the frequency of differentially beneficial alleles.
In chapter 2, I analyze patterns of gene flow within and among Tropical Eastern Pacific
Pocillopora damicornis populations and assess their potential for local population to significantly different environmental conditions.
Reproduction
In eukaryotic taxa, evolutionarily meaningful gene flow requires sexual reproduction.
Evolutionary theory implies that sex is important to decrease genetic load by exposing alleles to selection (Fisher 1941; Maynard Smith 1978; Charlesworth & Charlesworth 1987), to increase adaptation potential by generating new allele combinations (Stebbins 1957; Maynard Smith
1978; Bell 1982; Morran et al. 2009) and to escape parasitism (Lively 1987; Howard & Lively
1994). This is particularly important in unstable, heterogeneous habitats with high selection pressure and to colonize new environments (Williams 1975; Maynard Smith 1978).
Parthenogenesis, in contrast, is more efficient than sexual reproduction since it is independent of mates. It is particularly suitable for stable environments with low selection pressures and as a means to reproduce quickly to colonize open space (e.g. Williams 1975; Holman 1987; Radtkey
6 et al. 1995; Weider et al. 1999). Both reproductive strategies clearly have advantages and several taxa combine the two (Bell 1982; Hughes 1989). Instead of sex or no sex, the question is then how much sex is ideal? Theory predicts that a little sex goes a long way and the bulk of reproductive output should be asexual (Hurst & Peck 1996).
Corals predominantly reproduce via broadcast spawning of gametes that fertilize in the water column or by releasing fully developed internally brooded larvae. Both strategies had been considered to be obligatory sexual but a recent study indicated that two coral colonies
(Pocillopora damicornis) use a mixed reproductive strategy with mostly parthenogenetic and some sexually produced larvae (Yeoh & Dai 2010). However, very little is known about this unique mixed reproductive strategy and its evolutionary significance remains unclear.
In chapter 3, I explore the evolutionary significance of simultaneously sexual and asexual reproductive gene flow in P. damicornis from Moorea, French Polynesia.
Pocillopora corals
The genus Pocillopora (Lamarck 1816) belongs to the family Pocilloporidae, order
Scleractinia, in the phylum Cnidaria. Phylogenetically, Pocillopora are monophyletic and belong to the robust clade (Romano & Palumbi 1996; Chen et al. 2002).
The family Pocilloporidae (Gray 1842) is one of the youngest scleractinian families
(Wells, 1956) and appeared for the first time in the Upper Cretaceous (approximately 65mya). It is composed of three extant genera, Stylophora Schweigger 1819, Seriatopora Lamarck 1816 and Pocillopora Lamarck 1818 but recent phylogenetic evidence indicates that the genera
Madracis Milne Edwards & Haime 1849 and Stylocoeniella Ehrenberg 1834 belong into this family as well (Fukami et al. 2008). Most Pocilloporidae are phenotypically plastic and
7 phylogenetic studies have revealed major discrepancies between morphospecies and phylogenetic lineages in Pocillopora (Combosch et al. 2008; Pinzón et al. 2013), Stylophora
(Flot et al. 2011; Stefani et al. 2011), Seriatopora (Flot et al. 2008) and Madracis (Frade et al.
2010).
The genus Pocillopora is the most abundant and widespread pocilloporiid genus. Two species, P. damicornis and P. eydouxi, can be found on virtually any coral reef in the Indo-
Pacific, from South Africa and the Red Sea to the Tropical Eastern Pacific and from the southernmost coral reefs in Lord Howe Island to the northernmost reefs in Japan and Israel
(Veron 2000). While extant pocilloporids are restricted to the Indo-Pacific, fossil evidence suggests that Pocillopora cf. palmata (similar to P. eydouxi; Veron 1995) was present in the
Caribbean as recently as 82.000 years ago (Geister 1977; Pandolfi et al. 2001).
The taxonomy of Pocillopora species is complicated by ecomorphs and widespread phenotypic plasticity in response to light intensity, wave exposure, space availability, sedimentation, crab induction, grazing and cropping (Wells 1956; Pichon & Veron 1976; Veron
2000). Accordingly, well over 50 species have been described (Wells 1956; Pichon & Veron
1976; Knowlton & Jackson 1994; Veron 1995; Glynn 1999; Veron 2000) although others have suggested that all growth forms represent a single polymorphic species (Gardiner 1897; Wood-
Jones 1910). For example, Gardiner (1897 p. 942) stated “In the collections there are over 50 specimens, (...) and I am doubtful whether all these so-called species should not rather be described as varieties of one species, the characters of which would be the characters of the whole genus.” However, Pocillopora species are amongst the most intensely studied coral taxa
(Veron 1995 & 2000) and a robust, well-supported systematic resolution of this genus is missing.
8 Pocillopora damicornis is likely the most studied coral species worldwide (Veron 1995).
However, it seems like the more it is studied the more questions arise. For example, although P. damicornis is amongst the widest distributed coral species, which testifies to its dispersal potential, population genetic studies have reported significant population structure over kilometer-scales among neighboring reefs (e.g. Benzie et al. 1995; Ayre & Miller 2004; Miller &
Ayre 2004). Although the genetics of P. damicornis larvae brooding have been investigated repeatedly for 30 years, major basic questions about the asexual and/or sexual origin of its planula larvae remain (Stoddart 1983a, 1986; Ayre & Miller 2004; Sherman et al. 2006; Yeoh &
Dai 2010). Moreover, after I reported in 2008 that P. damicornis in the Tropical Eastern Pacific consists of two different genetic types (Combosch et al. 2008), others have described up to six different genetic lineages in P. damicornis populations throughout its Indo-Pacific distribution range (Souter 2010; Pinzón & LaJeunesse 2011; Pinzón et al. 2013; Schmidt-Roach et al. 2013).
One of the most intriguing characteristics of P. damicornis is its regional dominance in the isolated Tropical Eastern Pacific (TEP) while P. damicornis is elsewhere considered a minor component of the hermatypic coral taxa. Glynn et al. (1972) first described extensive, mono- specific P. damicornis stands in the TEP and suspected its inter-specific competitiveness and fast-growing ramose growth form as the main reasons for its regional success. Highsmith (1982) speculated that the local abundance in the TEP is largely due to vegetative fragmentation, i.e. clonal propagation. Richmond (1985) found differences between TEP and Indo-West Pacific populations in fecundity, reproductive allocation, growth rates and age specific mortality and interpreted them as successful adaptations of P. damicornis to the TEP. Richmond (1987) later suggested that TEP populations might represent a different species, based on these life-history differences. Glynn et al. (1991) first noticed that TEP populations reproduce via gamete-
9 spawning in contrast to Indo-West Pacific populations that brood mostly parthenogenetic larvae
(Stoddart 1983b; Stoddart 1986; Sherman et al. 2006; Yeoh & Dai 2010). In 2008, I provided genetic evidence for inter-specific hybridization among TEP Pocillopora corals and introgressive gene flow into P. damicornis but found no indications for hybridization among Indo-West
Pacific populations (Combosch et al. 2008). Pinzon and LaJeunesse (2011) recently suggested that TEP Pocillopora species, including P. damicornis, are comprised of three distinct genetic clusters (Type 1 - 3) that are independent of species identity and morphology. The evolutionary history of TEP P. damicornis population is thus controversial and nothing is known about the genetic basis of this most successful peripheral coral adaptation.
The Tropical Eastern Pacific
The TEP is best known for being one of the most isolated biogeographic regions of the world’s oceans (Darwin 1859; Chavez & Brusca 1991; Grigg & Hey 1992). It is enclosed by land and cold water and separated from the tropical central Pacific by a vast expense of deep island-free ocean, stretching over 5000km between Clipperton and the Line Islands. For continental shelf fauna, this endless open ocean constitutes a virtually “impassable barrier”
(Darwin 1859). Yet, trans-Pacific species occur in several taxa, but trans-Pacific gene flow is mostly sporadic (e.g. Rosenblatt & Waples 1986; Lessios et al. 1998; Lessios et al. 2003;
Lessios & Robertson 2006) and often too low to effectively connect TEP populations with their
Central Pacific counterparts, particularly among corals (Forsman 2003; Combosch et al. 2008;
Baums et al. 2012).
The isolation of the TEP makes its populations particularly vulnerable to population collapses and genetic pauperization. Moreover, TEP coral reefs are exposed to strong seasonal
10 upwelling and periodic El-Niño events. While seasonal upwelling generally limits coral growth
(Glynn & Stewart 1973), El-Niño events often cause massive coral mortality by increasing seawater temperature beyond tolerable thresholds for reef corals. Several major El-Niño events causing severe coral mortality have been documents over the last three decades (Glynn 1985,
1988, 1990; Glynn & Colgan 1992; Eakin & Glynn 1996; Glynn et al. 2001; reviewed in Glynn
& Enochs 2011). Moreover, suitable reef environments in the TEP are fragmented (Glynn &
Ault 2000), subject to elevated nutrient levels (Glynn & Mate 1997), increased terrestrial runoff
(Glynn 1976) and toxic algal blooms (Guzman et al. 1990). Some of these challenges are likely to become more frequent for corals worldwide due to environmental degradation and global climate change.
A recent paleontological study further showed a massive population collapse of TEP corals due to increased El-Niño variability, with subsequent recover ~2600 years ago (Toth et al.
2012) with potentially significant evolutionary consequences. Prolonged significant population bottlenecks cause a dramatic loss of genetic diversity, which eliminates evolutionary raw material and reduces the ability of populations to react to environmental change (Gilpin & Soule
1986). On the other hand, bottlenecks can also purge deleterious alleles and facilitate local selection and adaptation (Carson 1990; Keller et al. 1994). The isolation of the TEP and the fragmentation among TEP coral reefs likely enabled selection to effectively eliminate deleterious and mal-adapted alleles, creating a potentially rapidly evolving small and isolated population that was forced to adapt effectively to difficult local environmental conditions. The ability of TEP
P.’damicornis specimens to adapt to local conditions has been shown experimentally: specimens from the non-upwelling Gulf of Chiriquí are less sensitive to higher temperatures than those from the upwelling Gulf of Panamá (D'Croz & Mate 2004). No population genetic analyses of any
11 TEP coral had been done before, to assess levels of gene flow and genetic structures among TEP reefs.
In Chapter 2, I use multi-locus microsatellite data to describe the population genetic structure among and within nine Tropical Eastern Pacific P. damicornis populations across three environmentally and geographically distinct regions in Panamá. I assess the corals potential to adapt to local environmental conditions and show significant population genetic structure, both among regions and among populations indicating limited gene flow, i.e. larval dispersal.
Populations in environmentally distinct regions that are connected by only limited levels of gene flow have a significant potential for differential adaptations. Within reefs, I found high levels of genotypic diversity, significant inbreeding and local spatial genetic structures over 10m distances. These results suggest that TEP P. damicornis populations reproduce predominantly sexually, in contrast to previous speculations (Highsmith 1982), and that gene flow is severely limited within reef, even over meter scales.
In Chapter 3, I provide detailed genetic analyses of the unusual, simultaneously mixed reproductive strategy in Indo-West Pacific P. damicornis populations. Using multi-locus microsatellite data, I compare the genotypes of over 600 coral larvae with the genotypes of their brood parents. My results show that although larvae are predominantly parthenogenetic, a small subset of larvae from most colonies was produced sexually. I use logistic regression analyses to assess the impact of several external factors on the ratio of parthenogenetic and sexually produced larvae and found significantly variation in the proportion of sexual larvae with colony size, reproductive cycle day and general calendar day. Particularly interesting is the significant decrease in the proportion of sexual larvae with increasing colony size, which suggests that the
12 mixed reproductive strategy changes during the life cycle of the coral. This has important ecological and evolutionary consequences that I discuss in detail in chapter 3.
In chapter 4, I re-analyzed new and published sequence data from the nuclear rDNA
(ITS2) and mitochondrial open reading frame (ORF) with a particular focus on rooting the phylogenetic tree to understand the evolutionary trajectory among Pocillopora lineages. Both phylogenetic analyses identified previously defined genetic types that are incongruent with traditional species delimitations. The improved rooting suggests that type 3 is the most basal
Pocillopora lineage, which might indicate an Indian Ocean origin of the genus Pocillopora. I then used genome-wide Restriction-side Associated DNA sequencing (RAD-Seq) to conduct a phylogenomic comparison among three TEP Pocillopora species, two types of putative TEP hybrids and four Pacific-wide P. damicornis populations. The phylogenomic analyses revealed three distinct genomic lineages that support previous ITS and ORF-based lineages. The two hybrid pools were genetically similar to one parental lineage but also contained numerous alleles from the other parental lineage, which strongly suggest their hybrid origin and subsequent uni- parental backcrossing.
13 References
Abbott R., D. Albach, S. Ansell, et al. 2013 Hybridization and speciation. Journal of Evolutionary Biology 26: 229-246.
Anderson E. 1949 Introgressive hybridization John Wiley and Sons, New York.
Anderson T.M., B.M. vonHoldt, S.I. Candille, et al. 2009 Molecular and Evolutionary History of Melanism in North American Gray Wolves. Science 323: 1339-1343.
Arnold M.L. 1997 Natural Hybridization and Evolution Oxford University Press, Oxford.
Ayre D. & T. Hughes 2004 Climate change, genotypic diversity and gene flow in reef-building corals. Ecology Letters 7: 273-278.
Ayre D. & K. Miller 2004 Where do clonal coral larvae go? Adult genotypic diversity conflicts with reproductive effort in the brooding coral Pocillopora damicornis. Marine Ecology Progress Series 277: 95-105.
Baker A., C. Starger, T. McClanahan & P. Glynn 2004 Coral reefs: corals' adaptive response to climate change. Nature 430: 741.
Barbash D.A., D.F. Siino, A.M. Tarone & J. Roote 2003 A rapidly evolving MYB-related protein causes species isolation in Drosophila. Proceedings of the National Academy of Sciences 100: 5302-5307.
Barshis D., J. Stillman, R. Gates, et al. 2010 Protein expression and genetic structure of the coral Porites lobata in an environmentally extreme Samoan back reef: does host genotype limit phenotypic plasticity? Molecular Ecology 19: 1705-1720.
Baums I.B., J.N. Boulay, N.R. Polato & M.E. Hellberg 2012 No gene flow across the Eastern Pacific Barrier in the reef-building coral Porites lobata. Molecular Ecology 21: 5418- 5433.
Bell G. 1982 The masterpiece of nature: the evolution and genetics of sexuality Kluwer Academic Publishers.
Benzie J., A. Haskell & H. Lehman 1995 Variation in the genetic composition of coral (Pocillopora damicornis and Acropora palifera) populations from different reef habitats. Marine Biology 121: 731-739.
Brideau N.J., H.A. Flores, J. Wang, et al. 2006 Two Dobzhansky-Muller Genes Interact to Cause Hybrid Lethality in Drosophila. Science 314: 1292-1295.
Carson H.L. 1990 Increased genetic variance after a population bottleneck. Trends in Ecology & Evolution 5: 228-230.
14 Charlesworth D. & B. Charlesworth 1987 Inbreeding depression and its evolutionary consequences. Annual Review of Ecology and Systematics 18: 237-268.
Chavez F. & R. Brusca 1991 The Galápagos Islands and their relation to oceanographic processes in the tropical Pacific. Galápagos Marine Invertebrates: Taxonomy, Biogeography, and Evolution in Darwinís Islands: 9-33.
Chen C.A., C.C. Wallace & J. Wolstenholme 2002 Analysis of the mitochondrial 12S rRNA gene supports a two- clade hypothesis of the evolutionary history of scleractinian corals. Molecular Phylogenetics and Evolution 23: 137-149.
Combosch D., H. Guzman, H. Schuhmacher & S. Vollmer 2008 Interspecific hybridization and restricted trans-Pacific gene flow in the Tropical Eastern Pacific Pocillopora. Molecular Ecology 17: 1304-1312.
D'Croz L. & J. Mate 2004 Experimental responses to elevated water temperature in genotypes of the reef coral Pocillopora damicornis from upwelling and non-upwelling environments in Panamá. Coral Reefs 23: 473-483.
Darwin C. 1859 On the origins of species by means of natural selection or the preservation of favoured races in the struggle for life John Murray, London.
Dlugosch K.M. & I.M. Parker 2008 Founding events in species invasions: genetic variation, adaptive evolution, and the role of multiple introductions. Molecular Ecology 17: 431- 449.
Dobzhansky T. 1937 Genetics and the Origin of Species Columbia University Press, New York.
Eakin C. & P. Glynn 1996 Low tidal exposures and reef mortalities in the eastern Pacific. Coral reefs 15: 73-80.
Fisher R.A. 1941 Average excess and average effect of a gene substitution. Annals of eugenics 11: 53-63.
Flot J.-F., J. Blanchot, L. Charpy, et al. 2011 Incongruence between morphotypes and genetically delimited species in the coral genus Stylophora: phenotypic plasticity, morphological convergence, morphological stasis or interspecific hybridization? BMC Ecology 11: 22.
Flot J.-F., W.Y. Licuanan, Y. Nakano, et al. 2008 Mitochondrial sequences of Seriatopora corals show little agreement with morphology and reveal the duplication of a tRNA gene near the control region. Coral Reefs 27: 789-794.
Forsman Z. 2003 Phylogeny and phylogeography of Porites & Siderastrea (Scleractinia: Cnidaria) species in the caribbean and eastern pacific; based on the nuclear ribosomal ITS region, University of Houston.
15 Frade P.R., M.C. Reyes-Nivia, J. Faria, et al. 2010 Semi-permeable species boundaries in the coral genus Madracis: introgression in a brooding coral system. Molecular Phylogenetics and Evolution 57: 1072-1090.
Fukami H., A. Budd, G. Paulay, et al. 2004 Conventional taxonomy obscures deep divergence between Pacific and Atlantic corals. Nature 427: 832-835.
Fukami H., C. Chen, A. Budd, et al. 2008 Mitochondrial and nuclear genes suggest that stony corals are monophyletic but most families of stony corals are not (Order Scleractinia, Class Anthozoa, Phylum Cnidaria). PLoS One 3.
Gaines S., B. Gaylord, L. Gerber, A. Hastings & B. Kinlan 2007 Connecting places: the ecological consequences of dispersal in the sea. Oceanography 20: 90.
Garcia-Ramos G. & M. Kirkpatrick 1997 Genetic models of adaptation and gene flow in peripheral populations. Evolution 51: 21-28.
Gardiner J.S. 1897 On some Collections of Corals of the Family Pocilloporidae from the SW Pacific Ocean. Proceedings of the Zoological Society of London 65: 941-953.
Geister J. 1977 Occurrence of Pocillopora in late Pleistocene Caribbean coral reefs Geologisches Institut, Universitaet Bern.
Gilpin M.E. & M.E. Soule 1986 Minimum viable populations: processes of species extinction. In: Conservation Biology: The Science of Scarcity and Diversity (ed. Soule ME), pp. 19– 34. Sinauer, Sunderland, USA.
Glynn P. 1985 El Nino-associated disturbance to coral reefs and post-disturbance mortality by Acanthaster planci. Marine Ecology Progress Series 26: 4-13.
Glynn P. 1988 El Nino warming, coral mortality and reef framework destruction by echinoid bioerosion in the eastern Pacific. Galaxea 7: 129-160.
Glynn P. 1990 Coral mortality and disturbances to coral reefs in the tropical eastern Pacific. Elsevier Oceanography Series 52: 55-126.
Glynn P.& J. Ault 2000 A biogeographic analysis and review of the far eastern Pacific coral reef region. Coral Reefs 19: 1-23.
Glynn P.& M. Colgan 1992 Sporadic disturbances in fluctuating coral reef environments: El Nino and coral reef development in the eastern Pacific. Integrative and Comparative Biology 32: 707-718.
Glynn P., N. Gassman, C. Eakin, et al. 1991 Reef coral reproduction in the eastern Pacific: Costa Rica, Panamá, and Galápagos Islands (Ecuador). I. Pocilloporidae. Marine Biology 109: 355-368.
16 Glynn P. & J. Mate 1997 Field guide to the Pacific coral reefs of Panamá. Proceedings of the 8th International Coral Reef Symposium 1, 145-166.
Glynn P., J. Mate, A. Baker & M. Calderon 2001 Coral bleaching and mortality in Panamá and Ecuador during the 1997/98 El Nino Southern Oscillation Event: spatial/temporal patterns and comparisons with the 1982/83 event. Bulletin of Marine Science 69: 79-109.
Glynn P. & R. Stewart 1973 Distribution of coral reefs in the Pearl Islands (Gulf of Panamá) in relation to thermal conditions. Limnology and Oceanography 18: 367-379.
Glynn P., R. Stewart & J. McCosker 1972 Pacific coral reefs of Panamá: structure, distribution and predators. International Journal of Earth Sciences 61: 483-519.
Glynn P.W. 1976 Some physical and biological determinants of coral community structure in the eastern Pacific. Ecological Monographs 46: 431-456.
Glynn P.W. 1999 Pocillopora inflata, a new species of scleractinian coral (Cnidaria: Anthozoa) from the tropical eastern Pacific. Pocillopora inflata, una nueva especie de coral escleractinio (Cnidaria: Anthozoa) del Pacífico oriental tropical. Pacific Science 53: 168- 180.
Glynn P.W. & I.C. Enochs 2011 Invertebrates and their roles in coral reef ecosystems. Coral Reefs: An Ecosystem in Transition. pp. 273-325.
Gosset C.C. & N. Bierne 2012 Differential introgression from a sister species explains high FST- outlier loci within a mussel species. Journal for Evolutionary Biology 26: 14-26.
Grant B.R. & P.R. Grant 2008 Fission and fusion of Darwin's finches populations. Philosophical Transactions of the Royal Society 363: 2821-2829.
Grigg R. & R. Hey 1992 Paleoceanography of the tropical eastern Pacific Ocean. Science 255: 172.
Guzman H., J. Cortes, P. Glynn & R. Richmond 1990 Coral mortality associated with dinoflagellate blooms in the eastern Pacific (Costa Rica and Panamá). Marine Ecology Progress Series 60: 299-303.
Heck K. & E. McCoy 1978 Long-distance dispersal and the reef-building corals of the eastern Pacific. Marine Biology 48: 349-356.
Hedrick P.W. 2013 Adaptive introgression in animals: examples and comparison to new mutation and standing variation as sources of adaptive variation. Molecular Ecology 22: 4606–4618.
Heliconius Genome Consortium 2012 Butterfly genome reveals promiscuous exchange of mimicry adaptations among species. Nature: 487: 94-98
17 Hellberg M., R. Burton, J. Neigel & S. Palumbi 2002 Genetic assessment of connectivity among marine populations. Bulletin of Marine Science 70: 273-290.
Highsmith R.C. 1982 Reproduction by fragmentation in corals. Marine Ecology Progress Series 7: 207-226.
Hoegh-Guldberg O., P. Mumby, A. Hooten, et al. 2007 Coral reefs under rapid climate change and ocean acidification. Science 318: 1737.
Hohenlohe P.A., M.D. Day, S.J. Amish, et al. 2013 Genomic patterns of introgression in rainbow and westslope cutthroat trout illuminated by overlapping paired-end RAD sequencing. Molecular Ecology 22: 3002–3013.
Holman E.W. 1987 Recognizability of sexual and asexual species of rotifers. Systematic Biology 36: 381–386.
Howard R. & C. Lively 1994 Parasitism, mutation accumulation and the maintenance of sex. Nature 367: 554-557.
Howells E.J., V.H. Beltran, N.W. Larsen, et al. 2011 Coral thermal tolerance shaped by local adaptation of photosymbionts. Nature Climate Change 2: 116-120.
Hughes R. 1989 A functional biology of clonal animals Chapman and Hall.
Hughes T.P., A.H. Baird, D.R. Bellwood, et al. 2003 Climate change, human impacts, and the resilience of coral reefs. Science 301: 929-933.
Hurst L.D. & J.R. Peck 1996 Recent advances in understanding of the evolution and maintenance of sex. Trends in Ecology & Evolution 11: 46-52.
Huxley J. 1942 Evolution, the modern synthesis G. Allen & Unwin ltd, London,.
Jones G., G. Almany, G. Russ, et al. 2009 Larval retention and connectivity among populations of corals and reef fishes: history, advances and challenges. Coral Reefs 28: 307-325.
Jones G., M. Srinivasan& G. Almany 2007 Population Connectivity and Conservation of Marine Biodiversity. Oceanography 20: 100-111.
Jones R.J. 2008 Coral bleaching, bleaching-induced mortality, and the adaptive significance of the bleaching response. Marine Biology 154: 65-80.
Keller L.F., P. Arcese, J.N.M. Smith, W.M. Hochachka & S.C. Stearns 1994 Selection against inbred song sparrows during a natural population bottleneck. Nature 372: 356-357.
Kim M., M.-L. Cui, P. Cubas, et al. 2008 Regulatory Genes Control a Key Morphological and Ecological Trait Transferred Between Species. Science 322: 1116-1119.
Kim S.-C. & L. Rieseberg 1999 Genetic architecture of species differences in annual sunflowers: implications for adaptive trait introgression. Genetics 153: 965-977.
18 Knowlton N. & J.B.C. Jackson 1994 New taxonomy and niche partitioning on coral reefs: jack of all trades or master of some? Trends in Ecology & Evolution 9: 7-9.
Kunte K., C. Shea, M.L. Aardema, et al. 2011 Sex Chromosome Mosaicism and Hybrid Speciation among Tiger Swallowtail Butterflies. PLoS Genet 7: e1002274.
Lavergne S. & J. Molofsky 2007 Increased genetic variation and evolutionary potential drive the success of an invasive grass. Proceedings of the National Academy of Sciences 104: 3883-3888.
Lessios H., J. Kane & D. Robertson 2003 Phylogeography of the pantropical sea urchin Tripneustes: Contrasting patterns of population structure between oceans. Evolution 57: 2026-2036.
Lessios H., B. Kessing & D. Robertson 1998 Massive gene flow across the world's most potent marine biogeographic barrier. Proceedings of the Royal Society of London. Series B: Biological Sciences 265: 583-588.
Lessios H.& D. Robertson 2006 Crossing the impassable: genetic connections in 20 reef fishes across the eastern Pacific barrier. Proceedings of the Royal Society B: Biological Sciences 273: 2201-2209.
Lessios H.A. 2011 Speciation Genes in Free-Spawning Marine Invertebrates. Integrative and Comparative Biology 51: 456-465.
Lively C. 1987 Evidence from a New Zealand snail for the maintenance of sex by parasitism. Nature 328: 519-521.
Lu B.-R. & A. Snow 2005 Gene flow from genetically modified rice and its environmental consequences. Bioscience 55: 669-678.
Mallet J. 2005 Hybridization as an invasion of the genome. Trends in Ecology & Evolution 20: 229-237.
Martinsen G.D., T.G. Whitham, R.J. Turek & P. Keim 2001 Hybrid populations selectively filter gene introgression between species. Evolution 55: 1325-1335.
Maynard Smith J. 1978 The evolution of sex. Cambridge University Press, Cambridge.
Mayr E. 1963 Animal Species and Evolution Belknap Press, Cambridge.
McClanahan T.R., M. Ateweberhan, C.A. Muhando, J. Maina & M.S. Mohammed 2007 Effects of climate and seawater temperature variation on coral bleaching and mortality. Ecological Monographs 77: 503-525.
Michel A.P., S. Sim, T.H.Q. Powell, et al. 2010 Widespread genomic divergence during sympatric speciation. Proceedings of the National Academy of Sciences 107: 9724.
19 Miller K. & D. Ayre 2004 The role of sexual and asexual reproduction in structuring high latitude populations of the reef coral Pocillopora damicornis. Heredity 92: 557-568.
Miller K. & D. Ayre 2008a Population structure is not a simple function of reproductive mode and larval type: insights from tropical corals. Journal of Animal Ecology 77: 713-724.
Miller K. & D. Ayre 2008b Protection of genetic diversity and maintenance of connectivity among reef corals within marine protected areas. Conservation Biology 22: 1245-1254.
Morran L.T., M.D. Parmenter & P.C. Phillips 2009 Mutation load and rapid adaptation favour outcrossing over self-fertilization. Nature 462: 350-352.
Nosil P. & D. Schluter 2011 The genes underlying the process of speciation. Trends in Ecology & Evolution 26: 160-167.
Palumbi S. 2003 Population genetics, demographic connectivity, and the design of marine reserves. Ecological Applications 13: 146-158.
Palumbi S.R. 2009 Speciation and the evolution of gamete recognition genes: pattern and process. Heredity 102: 66-76.
Pandolfi J., R. Bradbury, E. Sala, et al. 2003 Global trajectories of the long-term decline of coral reef ecosystems. Science 301: 955.
Pandolfi J.M., J.B.C. Jackson & J. Geister 2001 Geologically Sudden Extinction of Two Widespread Late Pleistocene Caribbean Reef Corals. Evolutionary patterns: growth, form, and tempo in the fossil record in honor of Allan Cheetham: 120.
Pardo-Diaz C., C. Salazar, S.W. Baxter, et al. 2012 Adaptive Introgression across Species Boundaries in Heliconius Butterflies. PLoS Genet 8: e1002752.
Pichon M. & J. Veron 1976 Scleractinia of eastern Australia part I Families Thamnasteriidae, Astrocoeniidae, Pocilloporidae Australian Govment Publication Service.
Pinzón J.H. & T. LaJeunesse 2011 Species delimitation of common reef corals in the genus Pocillopora using nucleotide sequence phylogenies, population genetics and symbiosis ecology. Molecular Ecology 20: 311-325.
Pinzón J.H., E. Sampayo, E. Cox, et al. 2013 Blind to morphology: genetics identifies several widespread ecologically common species and few endemics among Indo-Pacific cauliflower corals ( Pocillopora, Scleractinia). Journal of Biogeography 40: 1595–1608.
Presgraves D.C., L. Balagopalan, S.M. Abmayr & H.A. Orr 2003 Adaptive evolution drives divergence of a hybrid inviability gene between two species of Drosophila. Nature 423: 715-719.
20 Radtkey R.R., S.C. Donnellan, R.N. Fisher, et al. 1995 When species collide: the origin and spread of an asexual species of gecko. Proceedings of the Royal Society B: Biological Sciences 259: 145-152.
Richards Z., M. van Oppen, C. Wallace, B. Willis & D. Miller 2008 Some rare Indo-Pacific coral species are probable hybrids. PLoS One. e3240
Richmond R. 1985 Variations in the population biology of Pocillopora damicornis across the Pacific 6, 101-106.
Richmond R. 1987 Energetic relationships and biogeographical differences among fecundity, growth and reproduction in the reef coral Pocillopora damicornis. Bulletin of Marine Science 41: 594-604.
Rieseberg L.H. & B.K. Blackman 2010 Speciation genes in plants. Annals of Botany 106: 439- 455.
Romano S. & S. Palumbi 1996 Evolution of scleractinian corals inferred from molecular systematics. Science 271: 640.
Rosenblatt R. & R. Waples 1986 A genetic comparison of allopatric populations of shore fish species from the Eastern and Central Pacific Ocean: dispersal or vicariance? Copeia 1986: 275-284.
Roux C., G. Tsagkogeorga, N. Bierne & N. Galtier 2013 Crossing the Species Barrier: Genomic Hotspots of Introgression between Two Highly Divergent Ciona intestinalis Species. Molecular Biology and Evolution 30: 1574-1587.
Rowan R. 2004 Coral bleaching: thermal adaptation in reef coral symbionts. Nature 430: 742.
Scheltema R.S. 1971 The dispersal of the larvae of shoal-water benthic invertebrate species over long distances by ocean currents. Proceedings of the 4th European Marine Biology Symposium 1: 7-28.
Schmidt-Roach S., P. Lundgren, K.J. Miller, et al. 2013 Assessing hidden species diversity in the coral Pocillopora damicornis from Eastern Australia. Coral Reefs 32: 161-172.
Sherman C., D. Ayre & K. Miller 2006 Asexual reproduction does not produce clonal populations of the brooding coral Pocillopora damicornis on the Great Barrier Reef, Australia. Coral Reefs 25: 7-18.
Slatkin M. 1987 Gene flow and the geographic structure of natural populations. Science 236: 787.
Slatkin M. & N. Barton 1989 A comparison of three indirect methods for estimating average levels of gene flow. Evolution 43: 1349-1368.
21 Smith L., D. Barshis & C. Birkeland 2007 Phenotypic plasticity for skeletal growth, density and calcification of Porites lobata in response to habitat type. Coral Reefs 26: 559-567.
Smith-Keune C. & M. van Oppen 2006 Genetic structure of a reef-building coral from thermally distinct environments on the Great Barrier Reef. Coral Reefs 25: 493-502.
Souter P. 2010 Hidden genetic diversity in a key model species of coral. Marine Biology 157: 875-885.
Stebbins G.L. 1957 Self fertilization and population variability in the higher plants. The American Naturalist 91: 337-354.
Stefani F., F. Benzoni, S.-Y. Yang, et al. 2011 Comparison of morphological and genetic analyses reveals cryptic divergence and morphological plasticity in Stylophora (Cnidaria, Scleractinia). Coral Reefs: 1-17.
Stoddart J. 1983a A genotypic diversity measure. Journal of Heredity 74: 489.
Stoddart J. 1983b Asexual production of planulae in the coral Pocillopora damicornis. Marine Biology 76: 279-284
Stoddart J. 1986 Biochemical genetics of Pocillopora damicornis in Kaneohe Bay, Oahu, Hawaii. Results of the 1983 Summer Research Training Program at the Hawaii Institute of Marine Biology: 133-142.
Ting C.T., A. Takahashi & C.I. Wu 2001 Incipient speciation by sexual isolation in Drosophila: Concurrent evolution at multiple loci. Proceedings of the National Academy of Sciences 98: 6709-6713.
Ting C.T., S.C. Tsaur, M.L. Wu & C.I. Wu 1998 A rapidly evolving homeobox at the site of a hybrid sterility gene. Science 282: 1501-1504.
Toth L.T., R.B. Aronson, S.V. Vollmer, et al. 2012 ENSO Drove 2500-Year Collapse of Eastern Pacific Coral Reefs. Science 337: 81-84.
Turner T.L., M.W. Hahn & S.V. Nuzhdin 2005 Genomic Islands of Speciation in Anopheles gambiae. PLoS Biology 3: e285.
Ulstrup K., R. Berkelmans, P. Ralph & M. Van Oppen 2006 Variation in bleaching sensitivity of two coral species across a latitudinal gradient on the Great Barrier Reef: the role of zooxanthellae. Marine Ecology Progress Series 314: 135-148.
Underwood J.N., L.D. Smith, M.J.H. Van Oppen & J.P. Gilmour 2007 Multiple scales of genetic connectivity in a brooding coral on isolated reefs following catastrophic bleaching. Molecular Ecology 16: 771-784.
Vacquier V. 1998 Evolution of gamete recognition proteins. Science 281: 1995.
22 van Oppen M., A. Lutz, G. De'ath, L. Peplow & S. Kininmonth 2008 Genetic traces of recent long-distance dispersal in a predominantly self-recruiting coral. PLoS One 3: e3401.
Veron J.E.N. 1995 Corals in space and time: the biogeography and evolution of the scleractinia Cornell University Press, Ithaca.
Veron J.E.N. 2000 Corals of the World AIMS, Townsville, Australia.
Vollmer S.V. & S.R. Palumbi 2002 Hybridization and the evolution of reef coral diversity. Science 296: 2023-2025.
Warner R. & R. Cowen 2002 Local retention of production in marine populations: evidence, mechanisms, and consequences. Bulletin of Marine Science 70: 245-249.
Weider L.J., A. Hobek, P.D.N. Hebert & T.J. Crease 1999 Holarctic phylogeography of an asexual species complex- II. Allozymic variation and clonal structure in Arctic Daphnia. Molecular Ecology 8: 1-13.
Wells J.W. 1956 Scleractinia. In: Treatise on Invertebrate Paleontology, Volume F, Coelenterata (ed. Moore RC), pp. 353– 367. Geological Society of America and University of Kansas Press, Lawrence, KS.
West J.M. & R.V. Salm 2003 Resistance and resilience to coral bleaching: implications for coral reef conservation and management. Conservation Biology 17: 956-967.
Wilkinson C. 1998 Status of coral reefs of the world: 1998. Australian Institute of Marine Science, Townsville; UNEP, Nairobi; Global Coral Reef Monitoring Network; International Coral Reef Intiative, Paris.
Wilkinson C. 2000 Status of coral reefs of the world: 2000. Australian Institute of Marine Science, Townsville; UNEP, Nairobi; Global Coral Reef Monitoring Network; International Coral Reef Intiative, Paris.
Wilkinson C. 2004 Status of coral reefs of the world: 2004 Australian Institute of Marine Science, Townsville; UNEP, Nairobi; Global Coral Reef Monitoring Network; International Coral Reef Intiative, Paris.
Willi Y. & A. Hoffmann 2009 Demographic factors and genetic variation influence population persistence under environmental change. Journal of Evolutionary Biology 22: 124-133.
Williams G. 1975 Sex and evolution Princeton University Press, Princeton, NJ.
Willis B., M. van Oppen, D. Miller, S. Vollmer & D. Ayre 2006 The role of hybridization in the evolution of reef corals. Annual Review of Ecology Evolution and Systematics 37: 489- 517.
Wolfenbarger L. & P. Phifer 2000 The ecological risks and benefits of genetically engineered plants. Science 290: 2088-2093.
23 Wood-Jones F. 1910 Coral and atolls Lovell, Reeve and Co, London, UK.
Wright S. 1943 Isolation by distance. Genetics 28: 114.
Wu C.I. 2001 The genic view of the process of speciation. Journal of Evolutionary Biology 14: 851-865.
Wu C.I. & C.T. Ting 2004 Genes and speciation. Nature Reviews Genetics 5: 114-122.
Yeoh S. & C. Dai 2010 The production of sexual and asexual larvae within single broods of the scleractinian coral, Pocillopora damicornis. Marine Biology 157: 351-359.
24 CHAPTER 2
POPULATION GENETICS OF AN ECOSYSTEM-DEFINING REEF CORAL -
POCILLOPORA DAMICORNIS IN THE TROPICAL EASTERN PACIFIC1
Abstract
Background: Coral reefs in the Tropical Eastern Pacific (TEP) are amongst the most peripheral and geographically isolated in the world. This isolation has shaped the biology of TEP organisms and lead to the formation of numerous endemic species. For example, the coral Pocillopora damicornis is a minor reef-builder elsewhere in the Indo-West Pacific, but is the dominant reef- building coral in the TEP, where it forms large, mono-specific stands, covering many hectares of reef. Moreover, TEP P. damicornis reproduces by broadcast spawning, while it broods mostly parthenogenetic larvae throughout the rest of the Indo-West Pacific. Population genetic surveys for P. damicornis from across its Indo-Pacific range indicate that gene flow (i.e. larval dispersal) is generally limited over hundreds of kilometers or less. Little is known about the population genetic structure and the dispersal potential of P. damicornis in the TEP.
Methodology: Using multilocus microsatellite data, we analyzed the population structure of TEP
P. damicornis among and within nine reefs and test for significant genetic structure across three geographically and ecologically distinct regions in Panamá.
Principal Findings/Conclusions: We detected significant levels of population genetic structure
(global RST = 0.162), indicating restricted gene flow (i.e. larvae dispersal), both among the three
1 Published as: Combosch, D.J. & S.V. Vollmer 2011. Population genetics of an ecosystem- defining reef coral - Pocillopora damicornis in the Tropical Eastern Pacific. PLoS One 6(8):
25 regions (RRT = 0.081) as well as within regions (RSR = 0.089). Limited gene flow across a distinct environmental cline, like the regional upwelling gradient in Panamá, indicates a significant potential for differential adaptation and population differentiation. Individual reefs were characterized by unexpectedly high genet diversity (avg. 94%), relatively high inbreeding coefficients (global FIS = 0.183), and localized spatial genetic structure among individuals (i.e. unique genets) over 10m intervals. These findings suggest that gene flow is limited in TEP P. damicornis populations, particularly among regions, but even over meter scales within populations.
Introduction
Populations at the periphery of species’ distribution ranges often exist in suboptimal and/or unstable conditions at the organism’s physiological or ecological limits (Soule 1973).
They tend to be small as well as geographically and genetically isolated from central core populations (Brown et al. 1996; Vucetich & Waite 2003). Their small population sizes and genetic isolation increase genetic drift and inbreeding and limit genetic diversity and cohesiveness among populations (Wright 1931; Slatkin 1987; Garcia-Ramos & Kirkpatrick
1997). As a result, peripheral populations tend to have reduced genetic diversity and are often heavily reliant on local sources of recruitment rather than long-distance recruitment from core populations (e.g. Arnaud-Haond et al. 2006; Bergl & Vigilant 2007; Beatty et al. 2008; Noreen et al. 2009). Their isolation and low levels of genetic diversity may also hamper their ability to cope with environmental changes (e.g. global climate change) and to recover from disturbances, making them more vulnerable to extinction (Lawton 1993; Hoffmann & Blows 1994). However, small population sizes in genetic isolation can also facilitate local selection and adaptation, give
26 rise to evolutionary innovation (Budd & Pandolfi 2010) and allow peripheral populations to exploit new ecological niches (Templeton 1980; Carson & Templeton 1984).
One of the most isolated and peripheral biogeographic regions of the world’s oceans is the Tropical Eastern Pacific (TEP). It is separated from the Indo-West Pacific by 5000km of open-ocean that Darwin (1859) described as “impassable” for shallow water marine species
(Ekman 1935; Mayr 1954). While not completely impassable, the marine fauna of the TEP consists of a limited subset of trans-Pacific species (i.e. species distributed across the Pacific), which are typically interconnected by low levels of gene flow (e.g. Lessios et al. 1999;
McCartney et al. 2000; Colborn et al. 2001; Lessios et al. 2001; Lessios & Robertson 2006;
Combosch et al. 2008; Duda & Lessios 2009), as well as numerous endemic species (Glynn &
Ault 2000; Colborn et al. 2001; Hughes et al. 2002). For example, 34 species of reef-building corals are recognized in the TEP (compared to 581 in the Western Pacific), of which 27 are trans-Pacific (80%) and seven (20%) are regional endemics (Veron 2000). Major environmental challenges for reef corals in the TEP include significant freshwater discharges and seasonal upwelling, leading to high turbidity, elevated nutrient levels and reduced seawater temperatures.
Reef habitats in the TEP are confined mostly to a narrow continental shelf and often fragmented by long stretches of estuaries, mangroves and sandy beaches (Briggs 1974; Glynn & Ault 2000).
The isolation of the TEP from the Indo-West Pacific and major differences in environmental factors and species compositions between these regions enabled significant ecological and life-history shifts in many TEP species. For example, the scleractinian coral
Pocillopora damicornis is a minor reef-building coral in the Indo-West Pacific, but the primary reef-builder in the TEP (Glynn & Macintyre 1977; Glynn & Wellington 1983; Guzman & Cortes
1993). Pocillopora damicornis in the TEP forms large, mono-specific carpets measuring tens to
27 hundreds of hectares (Glynn et al. 1996) that are thought to be maintained predominantly by asexual (i.e. vegetative) fragmentation (Highsmith 1982; Richmond 1987). Moreover, while
Indo-West Pacific P. damicornis populations brood mostly asexual (parthenogenetic) larvae
(Stoddart 1983; Yeoh & Dai 2010), populations in the TEP reproduce by broadcast-spawning its gametes (Glynn et al. 1991). This reproductive shift could have important genetic consequences since a broadcast spawning coral should not be able to produce parthenogenetic larvae (Harrison
& Wallace 1990; Richmond & Hunter 1990) and because broadcast spawning corals are often considered to have higher dispersal potentials than brooding corals (e.g. Hellberg 1995; Hellberg
1996; Ayre & Hughes 2000; Nishikawa et al. 2003; Nishikawa 2008), particularly over small spatial scales (Stoddart 1988b; McFadden & Aydin 1996; Miller 1998; Costantini et al. 2007 but see Underwood et al. 2007; Miller & Ayre 2008a). Hermaphroditic broadcast-spawning corals might also have a higher potential for selfing and inbreeding resulting from external fertilization
(Heyward & Babcock 1986; Hedgecock 1994; Willis et al. 1997).
Population genetic surveys among larvae-brooding P. damicornis populations across the
Indo-West Pacific indicate that larval dispersal (i.e. gene flow) is generally limited over hundreds of kilometers (e.g. Ayre & Hughes 2004; Miller & Ayre 2004, 2008b) and in some cases even among neighboring reefs (e.g. Benzie et al. 1995; Ayre & Miller 2004; Miller & Ayre
2004). Some of the highest levels of genetic structure have been described at the margins of P. damicornis Indo-Pacific range (Stoddart 1984; Ayre & Hughes 2004; Miller & Ayre 2004,
2008b). For example, populations in Southeast Australia are more genetically distinct over
700km (FST = 0.32; Miller & Ayre 2008b) than more central Great Barrier Reef populations over 1200km (FST = 0.06; Ayre et al. 1997). Most studied populations of P. damicornis in the
Indo-West Pacific also showed significant levels of local inbreeding and contained some level of
28 clonality (Ayre et al. 1997; Adjeroud & Tsuchiya 1999; Sherman et al. 2006; Whitaker 2006;
Souter et al. 2009).
Little is known about the population structure of broadcast-spawning P. damicornis populations in the TEP. Combosch et al. (2008) documented evidence for restricted transpacific dispersal between Central and Eastern Pacific populations of P. damicornis, using ribosomal
DNA sequence data (FST = 0.419, p < 0.001). Using multilocus genetic data, Pinzon &
LaJeunesse (2011) examined the genetic structure among multiple TEP Pocillopora species, including 7 well-recognized Indo-Pacific species (P. damicornis, P. eydouxi, P. meandrina, P. verrucosa, P. capitata, P. elegans, P. woodjonesi). They identified three genetic clusters or types
(Type I, II & III) in their multi-species TEP sample of Pocillopora. All three of these types were found in P. damicornis from across the TEP. Type I was found throughout the TEP, including
Panamá, Type II was found only in Clipperton Atoll and Type III was found in Panamá and
Galápagos. Yet, no genetic structure was detected within the broadly distributed Type I group over 3500km of the TEP.
One of the best-studied areas in the TEP is the Pacific coast of Panamá, which is dominated by two major gulfs, the Gulf of Chiriquí in the west and the Gulf of Panamá in the east. The two gulfs differ in several important ecological factors (Glynn et al. 1972; Glynn &
Stewart 1973; Glynn 1976, 1977; Highsmith 1980; Glynn & Leyte Morales 1997; Glynn & Ault
2000), most notably in the strength of seasonal upwelling, which is strong in the Gulf of Panamá, but virtually absent in the Gulf of Chiriquí (Glynn & Mate 1997; Glynn & Ault 2000). The
Azuero Peninsula, which separates the two gulfs, is a transition zone with moderate upwelling
(D'Croz & Mate 2004). Corals from the two gulfs differ significantly in their thermal tolerances
29 (Glynn & D'Croz 1990; D'Croz & Mate 2004) and ability to recover from major disturbances like El-Niño induced bleaching events (Glynn 2001).
Using multilocus microsatellite data, we examined the population genetic structure of
TEP P. damicornis within and among nine Panamanian populations across the three geographically and ecologically distinct regions (Figure 1). Based on the strong geographic and ecologic differences between the three regions, we predicted significant regional differences in
P. damicornis population genetic structure. Within reefs, we analyzed spatial patterns of genetic diversity to estimate the relative contributions of clones versus unique genets to localized spatial genetic structure (SGS). We predict that high rates of vegetative fragmentation should lead to significant SGS due to clonal aggregations. In contrast, high rates of sexual reproduction by broadcast spawning should inhibit SGS due to increased outcrossing and the mixing of genotypes at the sea surface during larval development.
Materials and Methods
Collection of coral samples for this project was approved by the Autoridad Nacional del
Ambiente and was conducted under permit number ANAM SEA7108.
Sampling Sites
Pocillopora damicornis was sampled from nine reefs out of three regions (Gulf of
Chiriquí, Azuero Peninsula, Gulf of Panamá) across a well-documented upwelling gradient (e.g.
Glynn & Mate 1997) along 1400km of coastline in Panamá (Figure 1). Four reefs (Uva Island,
Granito de Oro, Canal de Afuera Island and Bahia de Damas on Coiba Island) were sampled in the Gulf of Chiriquí in the west, where upwelling is absent and sea surface temperatures rarely
30 drop below 25°C (Glynn & Mate 1997). Coastal coral reefs are poorly developed in the Gulf of
Chiriquí due to extensive freshwater runoff, but substantial reef development has occurred around its numerous islands, including the largest (136ha), thickest (13m) and oldest (5600a)
TEP reefs (Glynn & Macintyre 1977; Glynn & Mate 1997). Two reefs (Achotines Bay and
Iguana Island) were sampled in the centrally located Azuero Peninsula where occasional moderate upwelling occurs (D'Croz & Mate 2004). Coastal reefs are more common in the
Azuero Peninsula than in the gulfs, most likely due to reduced freshwater runoff. Three reefs
(Saboga and Contadora Islands in the Perlas Archipelago and Taboga Island) were sampled in the eastern Gulf of Panamá where strong seasonal upwelling leads to declines in sea surface temperature to ~15°C (Glynn & Mate 1997). Coastal reefs are uncommon in the Gulf of Panamá and reef development occurs primarily around near-shore islands and in the Perlas Archipelago.
Sampling and Genetic analyses
On each reef, 18 to 36 nubbins (2cm) of P. damicornis were sampled every ten meters along reef sites composed of dense stands of P. damicornis to analyze its small-scale spatial genetic structure. Samples were preserved in Guanidinium-Isothiocyanate (GITC) DNA buffer for genetic analysis (Fukami et al. 2004). DNA was extracted using standard Phenol-Chloroform extraction and ethanol precipitation.
Ten microsatellite loci were amplified using primers developed by Magalon et al. (2004)
(Pv2 & Pv6) and Starger et al. (2007) (Pd2, Pd4, Pd5, Pd6, Pd7, Pd8, Pd9 & Pd10). Six of these loci (Pd2, Pd4, Pd5, Pd6, Pv2 & Pv6) amplified consistently using modified PCR protocols. PCR amplifications were carried out separately for each locus, following a nested PCR protocol for labeling, including forward primers with M13 tails and M13 primers with IR-labels (Schuelke
31 2000). Each PCR mix contained 2pmol forward and 8pmol of each reverse and m13 primer in a
10µl reaction with 1µl 10x PCR buffer, 0.2mM dNTPs, 0.25U AmpliTaq DNA polymerase
(®Applied Biosystems) and 0.5µl template DNA. PCR profiles varied slightly by locus: For Pd2,
Pd4, Pd5 and Pd6, the PCR profile consisted of 30 cycles at 94°C for 45s, 55°C for 35s and 72°C for 35s followed by 8 cycles with a lower annealing temperature (53°C) for the M13 labeling.
For Pv2 and Pv6, the profile was modified to 30 cycles at 94°C for 45s, 56°C for 45s and 72°C for 30s followed by 8 labeling cycles with 53°C annealing temperature. Both protocols included an initial 3min denaturing step at 94°C and a final extension step at 72°C for 7min. The PCR products with incorporated infrared dyes were then sized and scored on a LICOR 4300 Genetic
Analyzer.
Population genetics
Population genetic statistics, including allele frequencies, clonality, probability of identity, F-statistics and Analyses of Molecular Variance (AMOVA) were calculated using
GenAlEx 6.1 (Peakall & Smouse 2006). The program Micro-Checker 2.3 (van Oosterhout et al.
2004) was used to test for stutter bands, large allele dropout and null alleles. Linkage disequilibrium among loci was assessed using a Markov chain method (Guo & Thompson 1992) in GENEPOP 4.0.10 (Raymond & Rousset 1995), followed by sequential Bonferroni correction to account for multiple comparisons (Rice 1989). Clones were identified in GenAlEx and because they introduce biases into population genetic analyses, only unique genets were used unless otherwise noted.
Population genetic structure was calculated as FST (Wright 1951), which assumes an infinite allele model (Kimura & Crow 1964), and RST (Slatkin 1995), which is based on a
32 stepwise mutation model of allele evolution (Kimura & Ohta 1978). Hierarchical AMOVA were used to estimate genetic structure among regions, among populations and among populations within regions (Excoffier et al. 1992). Statistical significance of RST and FST is based on 999 permutations. RST recovered more genetic structure, especially among regions, whereas FST had a lower resolution and higher intra-regional variation (FRT = 15% of FST compared to RRT = 50% of RST). As a result, we preferentially focus on RST (e.g. Tables 3 & 4). Principal Coordinate
Analyses (Orloci 1978), implemented in GenAlEx, were used to detect and visualize the major patterns of pairwise RST population comparisons. Isolation by distance among populations was tested using a Mantel test, implemented in GenAlEx 6.1, to test for an association between pairwise population genetic differences and geographic distances.
The Bayesian clustering method implemented in STRUCTURE 2.3 (Pritchard et al.
2000) was used to infer the likely number of genetic clusters (K) in the dataset. STRUCTURE estimates the posterior probability P (X | K) that the data fit the hypothesis of K clusters using a
Markov chain Monte Carlo approach to optimize genotypic equilibrium (linkage equilibrium and
HWE) within each cluster. Five independent runs for each K from 2 to 19 were conducted using the admixture model and independent allele frequencies with a burn-in period of 105 steps and
106 Markov chain Monte Carlo replications for data collection (Supplementary Figure S1).
STRUCTURE runs were aligned using Clumpp1.1 (Jakobsson & Rosenberg 2007) and the bar plot was generated with Distruct1.1 (Rosenberg 2004). To verify the optimal value for K, we compared the P (X | K) results with ∆K, the second order rate of change in the likelihood of K
(Evanno et al. 2005), which corresponds to the strength of genetic subdivision among clusters in the data (Figure S1b).
33 Fine-scale Spatial Genetic Structure (SGS) within populations was analyzed using the program SPAGeDi 1.2 (Hardy & Vekemans 2002). To distinguish between spatial genetic structure (SGS) due to clonal aggregations and SGS among unique genets, we analyzed and compared two different dataset: one dataset consisted of all 207 samples including clones (i.e. the ramet dataset) and one dataset contained only the 194 unique genotypes excluding clones (i.e. the genet dataset). SGS was estimated for each of the three regions separately and for all populations combined using the pairwise kinship coefficient Fij (Loiselle et al. 1995; Vekemans
& Hardy 2004). Other kinship coefficients including Ritland (1996) and Moran’s I (Hardy &
Vekemans 1999) produced similar results (not shown). 95% confidence intervals and standard errors were estimated by 10,000 permutations of the genetic and the spatial datasets. Kinship values outside the 95% confidence intervals were interpreted as significant SGS at the spatial distance. Three parameters were used to describe and compare SGS among regions and studies:
F(1), genetic patch size and Sp. F(1) is the average kinship among individuals within the smallest distance interval (10m). Its statistical significance (p) was obtained by comparing the slope b(F) of F(r) on ln(r) to 9999 random permutations of individuals among locations within populations using a Mantel test. The genetic patch size is the distance that corresponds to the first x-intercept of the kinship correlogram (cf. Sokal & Wartenberg 1983). Within the genetic patch, individuals are more closely related than the population average, i.e. the association between genotypes is positive, while individuals outside of the patch are genetically independent, i.e. their correlation is negative. The Sp statistic (Vekemans & Hardy 2004) is based on F(1) and the decrease of SGS with distance (b(F)).
34 Results
Genetic diversity, heterozygosity and clonality
Multilocus microsatellite data (n ≥ 5 loci) was obtained for 207 Pocillopora damicornis colonies from nine reef populations (Tables 1 & 2). The six amplified loci yielded between four and ten alleles each (mean NA = 7). Three pairs of loci showed indications of significant linkage disequilibria, but none showed consistent significant linkage disequilibria in more than two populations. Consequently, all six loci were considered as unlinked. Significant heterozygote deficits were detected in five loci (Table 1), but only Pv2 showed significant deficits in all populations, which is indicative of null alleles. The program Micro-Checker confirmed the presence of null alleles in Pv2, which was then excluded from further analyses.
Using the five remaining loci, 194 unique genets were identified with a high probability of identity (p < 3x10-5; Table 2) among the 207 genotyped samples. All nine populations were comprised predominantly of unique, i.e. sexually derived genets (93.8 ± 5.5%). Clonal genotypes were detected in all but two populations (Canal & Iguana), but only Saboga (Gulf of Panamá) contained more than two clones (n = 4, i.e. 18%). No clonal genotypes were shared among populations, in contrast to larvae-brooding Indo-West Pacific populations (Souter et al. 2009;
Starger et al. 2010). Since clones introduce biases into population genetic analyses, only unique genets were used in subsequent analyses, unless otherwise noted.
Using the genet only dataset, significant heterozygote deficits were detected in six out of the nine populations (Table 2), indicating widespread local inbreeding. Inbreeding was significant in the Gulf of Panamá (FIS = 0.255 ± 0.122) and in the Gulf of Chiriquí (FIS = 0.196 ±
0.171), but not in the Azuero Peninsula (FIS = 0.048 ± 0.135) due to an excess of heterozygotes in Achotines. Each population contained on average 4.7 alleles per locus. Slightly more alleles
35 were detected in the Gulf of Chiriquí (avg. NA = 5.1 ± 0.26) than in the Azuero Peninsula (4.5 ±
0.20) and in the Gulf of Panamá (4.3 ± 0.31). Interestingly, private alleles were only found in
Gulf of Chiriquí populations, which contained the only regional-specific allele (Pd2, 223bp) as well.
Genetic Structure among reefs and regions
Hierarchical AMOVA revealed significant levels of population genetic structure (p <
0.005) among populations, among regions, and among populations within regions (Table 3).
Genetic structure among populations was highly significant for both global RST (0.162; p <
0.001) and global FST (0.053; p < 0.001). For the R-statistics, population structure was equivalent among regions (RRT = 0.081) and among populations within regions (RSR = 0.089). All three regions were significantly different from each other (p < 0.001), but the two gulf regions were most distinct (pairwise RST = 0.138), whereas the Azuero Peninsula, which lies between the two gulfs, showed lower pairwise genetic structure (Gulf of Chiriquí RST = 0.070; Gulf of Panamá
RST = 0.101). For the F-statistics, population structure was predominantly due to differences among populations within regions (FSR = 0.045; Table 3).
Fourteen out of 36 pairwise RST comparisons between populations were significant after sequential Bonferroni adjustments (Table 4). Twelve of the significant pairwise RSTs were between populations from different regions, whereas only two comparisons within regions were significant, both involving the Gulf of Chiriquí population Granito. These strong regional differences can be seen as well in the principal coordinate analyses, where populations cluster into distinct regional groupings (Figure 2). However, no significant isolation-by-distance pattern was observed among populations (p = 0.013).
36 Bayesian clustering implemented in STRUCTURE detected five genetic clusters in the dataset (Figure S1). All five clusters were found in each population, but their distribution differed significantly among populations and particularly among regions (Figure 1). The genetic clusters 1 and 2 were most common in the Gulf of Chiriquí (28% & 25%, respectively), clusters
4 and 5 were most prevalent in the Gulf of Panamá (22% & 39%) and cluster 3 was dominant in the Azuero Peninsula (32%). Cluster 1 and 2 were also common in the Azuero Peninsula (23% and 24%), which indicates that its P. damicornis populations are more similar to the Gulf of
Chiriquí than to the Gulf of Panamá - as indicated by the pairwise RST values.
Fine-scale spatial genetic structure (SGS) on reefs
In addition to the genetic structure among regions and populations, strong fine-scale spatial genetic structure (SGS) was detected among individuals within reefs. SGS was significant over 10m intervals in both the ramet dataset (i.e. including clones; Framets = 0.102; p <0.001) and the genet dataset (i.e. excluding clones; Fgenets = 0.086, p <0.001). This indicates that corals 10 meters apart are significantly more closely related to each other than to the rest of the population.
Kinship among individuals (Fij) was only slightly elevated when clones were included, reflecting the low frequency of clones in the dataset (Figure 3). While SGS is to be expected in coral populations due to clonal aggregations as a consequence of asexual (vegetative) fragmentation, significant SGS in the genet dataset indicates that the observed structure is due to non-random spatial relatedness among distinct, i.e. sexually produced, genets. Across all nine populations, the average genetic patch size (i.e. the distance at which individuals on average are as related as across the entire population) was 50m and the Sp statistic was 0.055 (i.e. rather high compared to plant literature data; Vekemans & Hardy 2004).
37 Significant levels of SGS out to 10m were detected in all three regions (p < 0.05; Figure
3b-d). Average F10m-kinship in the Azuero Peninsula and Gulf of Panamá was similarly high
(F10m ~ 0.109), almost equivalent to a first-cousin relatedness (FIJ = 0.125) whereas F10m-kinship in the Gulf of Chiriquí was two times lower (0.059; p < 0.05). The genetic patch size was almost twice as large in the Gulf of Panamá (70m) than in the Gulf of Chiriquí and Azuero Peninsula
(40m). These differences in the strength and extent of SGS are reflected in the Sp statistic, which indicates that SGS is strongest in the western Gulf of Panamá (0.062), intermediate in the central
Azuero Peninsula (0.059) and lowest in the eastern Gulf of Chiriquí (0.049).
Discussion
The Tropical Eastern Pacific (TEP) is one of the most geographically isolated biogeographic regions in the world’s oceans (Darwin 1859). Its isolation has profound consequences for the diversity and ecology of TEP ecosystems, most notably coral reefs (Glynn
& Ault 2000). TEP coral reefs are remarkable since they are built predominantly by Pocillopora damicornis (Glynn & Macintyre 1977; Glynn & Wellington 1983; Guzman & Cortes 1993). TEP
P. damicornis also distinguish by its broadcast gamete-spawning reproductive strategy.
Multilocus microsatellite data from nine Panamanian populations demonstrate that TEP P. damicornis populations from Panamá are characterized by strong population genetic structure
(RST = 0.162), within and particularly among its three major regions, the Gulf of Chiriquí, the
Azuero Peninsula and the Gulf of Panamá. Moreover, the large and mono-specific stands of P. damicornis that characterize TEP reefs are not dominated by a few, highly clonal genotypes as previously thought (Highsmith 1982; Richmond 1987), but instead are comprised of numerous distinct genets, indicating frequent sexual reproduction. Significant fine-scale spatial genetic
38 structure (SGS) among individual genets and highly significant levels of heterozygote deficits further indicate widespread inbreeding and limited gene dispersal over meter scales, which is unusual for free-spawners. These tiered layers of significant population genetic structure indicate that gene flow (i.e. larval dispersal) is restricted among and within reefs, particularly between the environmentally distinct regions, which indicates a significant potential for regional adaptations to the different ecological conditions in each region.
Population Structure
As hypothesized, we detected strong genetic differences among the three geographically and ecologically distinct regions in Panamanian P. damicornis populations (RRT = 0.081).
Panamá’s two major gulfs – the Gulf of Chiriquí and the Gulf of Panamá - which are geographically most distant, were most genetically distinct (RST = 0.138). These genetic differences correspond with pronounced ecological differences between these two regions, e.g. in species compositions and environmental factors (e.g. Glynn 1976; Glynn et al. 1991; Glynn &
Mate 1997). The most significant environmental difference is the pattern of seasonal upwelling, which is strong in the Gulf of Panamá dropping temperatures down to 15°C, but is virtually absent in the Gulf of Chiriquí (Glynn & Mate 1997; Glynn & Ault 2000). The Azuero Peninsula, which bisects the two regions, is a transition zone with intermediate upwelling conditions
(D'Croz & Mate 2004). It also has intermediate levels of genetic structure compared to the other regions (RST = 0.070 & 0.101).
Upwelling delivers cool, nutrient-rich waters to the near-shore environment. In tropical habitats like the TEP, they have significant impacts on nutrient and temperature-sensitive reef fauna (Glynn 1977; Glynn & Leyte Morales 1997). In the TEP, variations in the severity of
39 upwelling restrict the regional and local distributions of coral reefs (Glynn & Stewart 1973;
Glynn & Ault 2000) and of major reef taxa, including corals (Glynn & Ault 2000) and key corallivores like the sea star Acanthaster planci (Glynn 1976). Lower temperatures associated with seasonal upwelling reduce reef development (Glynn et al. 1972), coral growth rates (Glynn
1977) and coral reproduction (Glynn et al. 1991), while elevated nutrients increase competition with algae (Glynn & Stewart 1973; Birkeland 1977), bioerosion (Highsmith 1980) and water turbidity (Glynn & Ault 2000). Corals from regions with different upwelling regimes often display different thermal tolerances, indicating local adaptations. For example, P. damicornis from the non-upwelling Gulf of Chiriquí are less sensitive to higher temperatures than those from the upwelling Gulf of Panamá (Glynn & D'Croz 1990; D'Croz & Mate 2004). Similar differences in thermal adaptation among P. damicornis populations have also been reported from the Great
Barrier Reef (Ulstrup et al. 2006). Yet, upwelling can also buffer against critically high sea surface temperatures and reduce coral bleaching and mortality. For example, during the 1997/98
El-Niño, seasonal upwelling in the Gulf of Panamá reduced rising seawater temperatures (mean
27°C), preventing significant coral bleaching and mortality. In contrast, at the same time in the non-upwelling Gulf of Chiriquí, mean seawater temperature rose to 30°C causing severe coral bleaching and mortality (Glynn 2001). While reefs in upwelling environments may thus be better able to weather predicted increases in sea surface temperatures and El Niño frequencies resulting from global climate change (e.g. Trenberth & Hoar 1997), coral populations in non-upwelling habitats seem to be better adapted to elevated temperatures (Glynn & D'Croz 1990; D'Croz &
Mate 2004). The genetic structure of Panamanian P. damicornis indicates that even among broadcast-spawning coral populations, adaptations to different thermal regimes can occur over distances of ~100km.
40 Limited inter-regional gene flow in Panamanian P. damicornis may also explain prolonged differences in post-disturbance reef recovery between the Gulf of Chiriquí and the
Gulf of Panamá. For example, the 1982/83 El Niño caused mass coral mortalities and reduced average coral cover on most Panamanian reefs by 80%, but even though mortality rates were similar in both gulfs, reefs in the Gulf of Panamá had recovered by 1992, whereas reefs in the
Gulf of Chiriquí recovered only 3-6% of their populations by 1997 (Glynn 2001). The significant population genetic structure between regions indicates that larval dispersal from recovered Gulf of Panamá populations to the still depauperate Gulf of Chiriquí populations was very limited and this may have prolonged the recovery of reefs in the Gulf of Chiriquí. Moreover, due to the limited gene flow (i.e. larval dispersal) between individual reefs, damaged populations depended strongly on surviving colonies for repopulation. Acanthaster predation on surviving colonies in the Gulf of Chiriquí potentially hampered initial recovery significantly, while corallivorous sea stars are absent in the Gulf of Panamá. Another important factor for the differences in P. damicornis recovery between Gulfs may be geographic differences in Symbiodinium identity (or type). Baker et al. (2004) showed that in the Gulf of Panamá, Pocillopora corals with clade C bleached severely in 1997 while colonies with clade D were unaffected by bleaching, which lead to a relative increase in colonies with clade D from 43% in 1995 to 63% in 2001. Higher levels of clonality in the Gulf of Panamá (e.g. Saboga) further indicate that asexual fragmentation might have played a more prominent role in the recovery of Gulf of Panamá populations compared to their Gulf of Chiriquí counterparts.
Using multiple genetic markers, Pinzon & LaJeunesse (2011) recently suggested that
TEP Pocillopora are comprised of three distinct genetic types (Type I - III), independent of species identity and morphology. This is controversial because they included at least seven well-
41 recognized Indo-Pacific species that must then either be reclassified or synonymized by their genetic types. Their dataset included a limited number of P. damicornis samples from the Gulf of
Panamá (n = 26), which were comprised of two of the three genetic types (Type I & II). Using a similar set of microsatellites, our STRUCTURE analyses do not support two distinct genetic types, but instead identified five genetic clusters (Figure 1 & Supplementary Figure S1). In addition, Pinzon & LaJeunesse (2011) failed to detect significant genetic structures within their
Pocillopora type I over 3500km from Mexico to Panamá and the Galápagos. Given that we observe strong genetic structure in P. damicornis within Panamá, it seems likely that the absence of geographic structure in Pinzon & LaJeunesse’s (2011) analyses are an artifact of pooling multiple species (i.e. gene pools) for population genetic analyses.
Comparisons with brooding Indo-West Pacific P. damicornis populations
Similar levels of population genetic structure over scales of hundreds of kilometers have also been observed in other brooding and broadcast-spawning coral species in the Indo-Pacific and the Caribbean (van Oppen & Gates 2006; Vollmer & Palumbi 2007). In P.’damicornis, allozyme studies from across the Indo-West Pacific indicate that populations exhibit a range of significant population genetic structures with FST values ranging from 0.06 to as high as 0.39
(Table 5). Some of the highest levels of population structure have been detected in isolated or peripheral populations. For example, FST values among high-latitude P. damicornis populations in southeast Australia averaged 0.37 over 700km (Miller & Ayre 2008b) compared to 0.06 among Great Barrier Reef populations over 1200km (Ayre et al. 1997). Similarly, FST values among the southernmost P. damicornis populations around Lord Howe Island (Miller & Ayre
42 2004) were twice as high as among similarly spaced populations on the Great Barrier Reef (FST =
0.102 vs. FST = 0.044; Ayre & Miller 2004).
Direct comparisons of population genetic data are possible with two P. damicornis microsatellite datasets from peripheral populations in East-African (Souter et al. 2009) and core populations in the Indonesian Archipelago (Starger et al. 2010) since they used the same or similar subsets of the microsatellite loci used here. Both studies found highly significant population genetic structure. Levels of genetic differentiation among the peripheral East-African populations were lower (FST = 0.02 over 860km) than among TEP populations (FST = 0.05 over
400km). Indonesian populations show the same degree of genetic differentiation as TEP populations (FST = 0.05) over a 10-fold wider geographic distance (3000km vs. 400km). We observed a much lower allelic diversity in the TEP (NA/locus = 7) compared to P. damicornis in
Indonesia and East-Africa (NA/locus = 14.0 & 13.4, respectively). Lower allelic diversities are expected for peripheral populations and the lower diversity in TEP populations, compared to the similarly peripheral East-African populations, is likely due to the extreme isolation of the TEP and very limited transpacific gene flow (FST = 0.419; Combosch et al. 2008). As expected, inbreeding was found to be highly significant in both peripheral populations, in the TEP and in
Eastern Africa (Souter et al. 2009). Comparisons among P. damicornis allozyme studies confirm all three population genetic trends in peripheral populations: 1) higher genetic structure (e.g.
Ayre et al. 1997 vs. Miller & Ayre 2008b) 2) decreased allelic diversity (Miller & Ayre 2008b) and 3) higher inbreeding (Miller & Ayre 2008b). These patterns have also been observed in peripheral populations of other coral species (Stoddart 1988a; Ayre & Hughes 2004; Noreen et al. 2009; Nunes et al. 2009).
43 Our results indicate lower genetic diversity and higher levels of population structure among broadcast-spawning populations (TEP) than among larvae brooding populations (Indo-
West Pacific), which may reflect intrinsic differences between the two reproductive strategies. In
Indo-West Pacific populations, parthenogenetic larvae brooding preserves genotypes, which reduces the rate of genetic drift and allelic loss, while allelic diversity in TEP populations are likely reduced due to reproductive sweepstakes, promoted by its broadcast-spawning reproductive strategy (Hedgecock 1994; Ruzzante et al. 1996; Christie et al. 2010). Contrary to our results, inter-specific comparisons mostly indicate that brooding corals have reduced dispersal potentials compared to broadcast spawners (Hellberg 1996; Carlon 1999; Ayre &
Hughes 2000; Nishikawa et al. 2003). This indicates that population structure in reef corals is not predominantly driven by reproductive strategies. Other factors are likely as important, for example effective population sizes and species-specific larval behaviors (e.g. Palumbi 1995;
Miller & Ayre 2008a).
Fine-scale Population Genetic Structure
The observed fine-scale spatial genetic structure (SGS) and the high inbreeding coefficients (average FIS = 0.18) indicate that gene dispersal is limited in Panamanian P. damicornis, not only between reefs, but within reefs as well. Fine-scale SGS arises from a variety of historic, demographic, and evolutionary processes (Loveless & Hamrick 1984;
Epperson 2003; Vekemans & Hardy 2004). Over small spatial scales, SGS is most often the result of limited gene dispersal and can be influenced by a broad array of life-history and reproductive traits including clonality, selfing, inbreeding, fertilization, fecundity, larval life history, and recruitment (Epperson 2003; Vekemans & Hardy 2004; Born et al. 2008). Clonality
44 can be an important determinant of SGS because clonal aggregations increase the average kinship over short distance intervals (e.g. Stoddart 1988b). We observed fine-scale SGS out to
10m in both, the genet-only and the genet-plus-clone dataset, but since clones were rare (6.2% of samples), clonality had only a small effect on the observed SGS (Figure 3). Instead, the fine- scale SGS observed in TEP P. damicornis was predominantly driven by significant non-random relatedness among individual genets out to 10 meters, which is unusual for free-spawning benthic marine populations.
The two most likely explanations for the observed SGS among genets in TEP P. damicornis are cohort recruitment (i.e. SGS among larval recruits, independent of parents) or proximity recruitment (i.e. SGS among parents and recruits). In brooding coral species, proximity recruitment is facilitated by the internal (i.e. stationary) larval development and the frequent settlement of larvae within meters of the maternal colony (Gerrodette 1981; Best &
Resing 1987; Carlon & Olson 1993). Although broadcast spawning was never directly observed, several independent studies provide convincing evidence that TEP P. damicornis do not release brooded larvae but instead free-spawn gametes (Richmond 1985; Glynn et al. 1991; Carpizo-
Ituarte et al. 2010). In broadcast-spawning corals, like TEP P. damicornis, larval development occurs externally and larval settlement after gamete release takes at least 1-3 days (Miller &
Mundy 2003). Although corals generally spawn during times of minimal water movement
(Babcock et al. 1986; van Woesik 2009), it is unlikely that drifting coral larvae would stay so close together (cohort recruitment) or remain within meters of the maternal colony (inhibiting proximity recruitment) over several days during their dispersive phase to create such a fine-scale spatial genetic structure (10m).
45 One potential scenario that could significantly reduce passive dispersal would be if gametes and/or larvae sink down and escape the flow of the water column. Although coral gametes typically float on the sea surface during fertilization and development (Willis & Oliver
1990), gametes of two Pocillopora (P. eydouxi & P. verrucosa; Kinzie 1993) and several other coral species (e.g. Fungia scutaria, F. fungites, Goniastrea favulus) are known to be negatively buoyant (Krupp 1983; Harrison & Wallace 1990; Gilmour 2002; Miller & Ayre 2008a).
Negatively buoyant gametes sink down after spawning and fertilize close to the substrate.
Fertilization at the substrate also favors mating among nearby colonies, which limits gene dispersal, and should facilitate SGS. Even if fertilization occurs at the sea surface but the zygotes sink down promptly, larval dispersal would still be reduced considerably. A minority of propagules could then remain in the water column to account for long-distance dispersal.
Whether TEP P. damicornis gametes or larvae are positively or negatively buoyant is currently unknown, since gamete spawning has not yet been observed directly.2
A second indication for limited gene dispersal within populations is the significant inbreeding coefficients. Inbreeding is a consequence of non-random mating among a limited number of individuals, which limits gene dispersal and promotes SGS (Zhao et al. 2009). The broadcast-spawning reproductive strategy of TEP P. damicornis might facilitate inbreeding in several ways. Firstly, the external fertilization of broadcast-spawners lacks the potential parental control of internal fertilization in brooders (Heyward & Babcock 1986; Willis et al. 1997).
Secondly, the high fecundity of broadcast-spawners leads to reproductive sweepstakes, which can promote inbreeding (Heyward & Babcock 1986; Willis et al. 1997). However, heterozygote deficits were as pronounced in larvae-brooding East-African populations (FIS = 0.26), and
2 P. damicornis has meanwhile been observed to release negatively buoyant eggs at the GBR by Schmidt et al. (2012)
46 significant inbreeding is common in P. damicornis populations throughout the Indo-Pacific, regardless of reproductive strategy, local environment, population history, and analyzed genetic marker (Table 5). Inbreeding, and potentially SGS, may thus be inherent features of
P.’damicornis.
While SGS is common in plants (reviews by Heywood 1991; Hardy & Vekemans 1999;
Vekemans & Hardy 2004), it has only been documented in four other coral species. Underwood et al. (2007) described a similar pattern of SGS in the brooding coral Seriatopora hystrix, including a significantly elevated kinship among genets in the smallest distance interval (F1 =
20m) and a roughly comparable genetic patch size (80m). Among broadcast-spawning corals,
Stoddart (1988b) and Miller and Ayre (2008a) found no significant genet SGS in Acropora digitifera and Goniastrea favulus, respectively. However, Miller and Ayre (2008a) found significant genet SGS in the broadcast-spawning Platygyra daedalea, including significantly positive kinship in the smallest distance interval (F1 = 5m) and a Genetic Patch Size of only 17m.
This is the only other documented broadcast-spawning coral with genet SGS. To account for this pattern, Miller and Ayre (2008a) suspected “an unknown element of larval behavior or development” - e.g. that larvae might be negatively buoyant.
Small-scale SGS over tens of meters has been documented for at least eight other sessile marine invertebrates, including two sponges (Calderon et al. 2007; Blanquer et al. 2009), a temperate soft coral (McFadden & Aydin 1996), a black coral (Miller 1998), a red coral
(Costantini et al. 2007), a bryozoan (Pemberton et al. 2007) and two tunicates (Yund & O'Neil
2000; David et al. 2010). For six out of these eight taxa, SGS seems to be driven by limited dispersal of brooded, philopatric larvae. Both broadcast-spawners with significant SGS, the
47 Anthipatharian coral Antipathes fiordensis (Miller 1998) and the solitary tunicate Styela plicata
(David et al. 2010) have negatively buoyant larvae (A. fiordensis) or eggs (S. plicata).
Clonal structure
Because of its large mono-specific stands and little to no detectable larvae recruitment
(Richmond 1985), TEP P. damicornis reefs were long thought to be highly clonal, persisting locally through vegetative asexual fragmentation (Highsmith 1982; Richmond 1987). Our data indicate that only 6% of corals sampled every 10m apart were clones, i.e. big clones (>10m) were uncommon. In contrast, 94% of samples consist of unique, sexually-derived genets. This indicates that sexual recruitment is sufficient to maintain high levels of genet diversity in TEP P. damicornis. Simulation models by Neigel & Avise (1983) indicate that in populations of the
Caribbean coral Acropora cervicornis, which depends heavily on vegetative fragmentation, 5% sexual recruitment (i.e. 95% asexual) is enough to generate high genotypic diversities (~60%), independent of initial genet diversity.
Compared to other P. damicornis studies, levels of clonality in the TEP (6%) seem low.
However, direct comparisons among studies are biased by different sampling strategies and genetic markers. Lower levels of clonality tend to be detected with microsatellites compared to allozyme markers (avg. 22% vs. 43%; Table 5) and with large sampling intervals compared to exhaustive samplings (e.g. Souter et al. 2009, Starger et al. 2010 and the present study vs. Yeoh
& Dai 2010). Comparisons among studies using the same marker and similar sampling indicate higher levels of clonality in peripheral populations (TEP, East Africa & Lord Howe Island) compared to more central populations (Indonesia & Great Barrier Reef; Table 5). Higher clonality in peripheral populations has been observed in other corals (Baums et al. 2006; Noreen
48 et al. 2009) and terrestrial plants (Eckert & Barrett 1993; Beatty et al. 2008). Clonal reproduction is considered to enable persistence in the commonly adverse environmental conditions towards range margins, at the expense of sexual reproduction benefits (Herlihy & Eckert 2002). One example of reduced sexual benefits due to asexual reproduction is limited gene flow among and within populations as described here for the peripheral TEP P. damicornis populations.
Conclusion
Regional structure and localized SGS detected in Panamanian Pocillopora damicornis has important consequences for the conservation management of TEP coral reefs. Recent devastating mass mortalities, for example due to El Niño, demonstrate the fragility of TEP coral reefs (Glynn 1985, 1990; Glynn & Colgan 1992; Eakin & Glynn 1996; LaJeunesse et al. 2007).
However, regional populations reacted very differently to these large-scale disturbances (Glynn
2001) and are further differently affected by regional disturbances including Acanthaster predation (Glynn 1973, 1988), toxic algal blooms (Guzman et al. 1990) and seasonal upwelling
(Glynn & Stewart 1973). In combination with the significant genetic differences among regions, these variations highlight the importance of environmentally, ecologically and genetically defined regional management and conservation units in Panamá and possibly throughout the
TEP. Limited gene flow and significant genetic differences between reefs further indicate that local population dynamics on individual reefs are vital for reef resilience and adaptation in
Panamá. For example, although the limited gene flow potentially hampers population recovery, it also allows for local and regional adaptation, which should be preserved and accounted for in conservation management.
49 Acknowledgements
Field and logistic support was provided by Rich Aronson, Ian McIntyre and Carmen
Schloeder. Thank you to Sarah Lemer, Liz Hemond, Sean Kent, Silvia Libro, Flavia Nunes and two anonymous reviewers for their comments on the manuscript.
50 References
Adjeroud, M. & M. Tsuchiya 1999 Genetic variation and clonal structure in the scleractinian coral Pocillopora damicornis in the Ryukyu Archipelago, southern Japan. Marine Biology 134(4): 753-760.
Arnaud-Haond, S., S. Teixeira, S. Massa, C. Billot, P. Saenger, G. Coupland, C.M. Duarte & E. Serrao 2006 Genetic structure at range edge: low diversity and high inbreeding in Southeast Asian mangrove (Avicennia marina) populations. Molecular Ecology 15(12): 3515-3525.
Ayre, D. & T. Hughes 2004 Climate change, genotypic diversity and gene flow in reef-building corals. Ecology Letters 7: 273-278.
Ayre, D., T. Hughes & R. Standish 1997 Genetic differentiation, reproductive mode, and gene flow in the brooding coral Pocillopora damicornis along the Great Barrier Reef, Australia. Marine Ecology Progress Series 159: 175-187.
Ayre, D. & K. Miller 2004 Where do clonal coral larvae go? Adult genotypic diversity conflicts with reproductive effort in the brooding coral Pocillopora damicornis. Marine Ecology Progress Series 277: 95-105.
Ayre, D.J. & T.P. Hughes 2000 Genotypic diversity and gene flow in brooding and spawning corals along the Great Barrier Reef, Australia. Evolution 54(5): 1590-1605.
Babcock, R., G. Bull, P. Harrison, A. Heyward, J. Oliver, C. Wallace & B. Willis 1986 Synchronous Spawnings of 105 Scleractinian Coral Species on the Great-Barrier-Reef. Marine Biology 90(3): 379-394.
Baker, A., C. Starger, T. McClanahan & P. Glynn 2004 Corals' adaptive response to climate change. Nature 430(7001): 741.
Baums, I.B., M.W. Miller & M.E. Hellberg 2006 Geographic variation in clonal structure in a reef-building Caribbean coral, Acropora palmata. Ecological Monographs 76(4): 503- 519.
Beatty, G., P. McEvoy, O. Sweeney & J. Provan 2008 Range-edge effects promote clonal growth in peripheral populations of the one-sided wintergreen Orthilia secunda. Diversity and Distributions 14(3): 546-555.
Benzie, J., A. Haskell & H. Lehman 1995 Variation in the genetic composition of coral (Pocillopora damicornis and Acropora palifera) populations from different reef habitats. Marine Biology 121(4): 731-739.
Bergl, R. & L. Vigilant 2007 Genetic analysis reveals population structure and recent migration within the highly fragmented range of the Cross River gorilla (Gorilla gorilla diehli). Molecular Ecology 16(3): 501-516.
51 Best, B. & J. Resing 1987 Active settlement of coral planulae: the effect of flow and swimming behavior on local recruitment. American Zoologist 127:103A.
Birkeland, C. (1977) The importance of rate of biomass accumulation in early successional stages of benthic communities to the survival of coral recruits. Proceedings of the 3rd International Coral Reef Symposium 1: 16-21.
Blanquer, A., M. Uriz & J. Caujape-Castells 2009 Small-scale spatial genetic structure in Scopalina lophyropoda, an encrusting sponge with philopatric larval dispersal and frequent fission and fusion events. Marine Ecology Progress Series 380:95-102.
Born, C., O. Hardy, H. Chevallier, S. Ossari, C. Atteke, E. Wickings & M. Hossaert-Mckey 2008 Small-scale spatial genetic structure in the Central African rainforest tree species Aucoumea klaineana: a stepwise approach to infer the impact of limited gene dispersal, population history and habitat fragmentation. Molecular Ecology 17(8): 2041-2050.
Briggs, J.C. (1974) Marine Zoogeography McGraw-Hill, New York.
Brown, J., G. Stevens & D. Kaufman 1996 The geographic range: size, shape, boundaries, and internal structure. Annual Review of Ecology and Systematics 27(1): 597-623.
Budd, A.F. & J.M. Pandolfi 2010 Evolutionary novelty is concentrated at the edge of coral species distributions. Science 328(5985): 1558-1561.
Calderon, I., N. Ortega, S. Duran, M. Becerro, M. Pascual & X. Turon 2007 Finding the relevant scale: clonality and genetic structure in a marine invertebrate (Crambe crambe, Porifera). Molecular Ecology 16(9): 1799-1810.
Carlon, D. & R. Olson 1993 Larval dispersal distance as an explanation for adult spatial pattern in two Caribbean reef corals. Journal of Experimental Marine Biology and Ecology 173:247-263.
Carlon, D.B. 1999 The evolution of mating systems in tropical reef corals. Trends in Ecology & Evolution 14(12): 491-495.
Carpizo-Ituarte, E., V. Vizcaino-Ochoa, G. Chi-Barragan, O. Tapia-Vazquez, A. Cupul-Magana & P. Medina-Rosas 2010 Evidence of sexual reproduction in the hermatypic corals Pocillopora damicornis, Porites panamensis, and Pavona gigantea in Banderas Bay, Mexican Pacific. Ciencias Marinas 37(1): 97-112.
Carson, H. & A. Templeton 1984 Genetic revolutions in relation to speciation phenomena: the founding of new populations. Annual Review of Ecology and Systematics 15(1): 97-132.
Christie, M., D. Johnson, C. Stallings & M. Hixon 2010 Self-recruitment and sweepstakes reproduction amid extensive gene flow in a coral-reef fish. Molecular Ecology 19: 1042– 1057.
52 Colborn, J., R. Crabtree, J. Shaklee, E. Pfeiler & B. Bowen 2001 The evolutionary enigma of bonefishes (Albula spp.): cryptic species and ancient separations in a globally distributed shorefish. Evolution 55(4): 807-820.
Combosch, D., H. Guzman, H. Schuhmacher & S. Vollmer 2008 Interspecific hybridization and restricted trans-Pacific gene flow in the Tropical Eastern Pacific Pocillopora. Molecular Ecology 17(5): 1304-1312.
Costantini, F., C. Fauvelot & M. Abbiati 2007 Fine-scale genetic structuring in Corallium rubrum: evidence of inbreeding and limited effective larval dispersal. Marine Ecology Progress Series 340: 109-119.
D'Croz, L. & J. Mate 2004 Experimental responses to elevated water temperature in genotypes of the reef coral Pocillopora damicornis from upwelling and non-upwelling environments in Panamá. Coral Reefs 23(4): 473-483.
Darwin, C. (1859) On the origins of species by means of natural selection or the preservation of favoured races in the struggle for life John Murray, London.
David, G., D. Marshall & C. Riginos 2010 Latitudinal variability in spatial genetic structure in the invasive ascidian, Styela plicata. Marine Biology 157(9): 1955-1965.
Duda, T. & H. Lessios 2009 Connectivity of populations within and between major biogeographic regions of the tropical Pacific in Conus ebraeus, a widespread marine gastropod. Coral Reefs 28(3): 651-659.
Eakin, C. & P. Glynn 1996 Low tidal exposures and reef mortalities in the eastern Pacific. Coral Reefs 15(2): 120.
Eckert, C. & S. Barrett 1993 Clonal reproduction and patterns of genotypic diversity in Decodon verticillatus. American Journal of Botany 80(10): 1175-1182.
Ekman, S. (1935) Tiergeographie des Meeres Akademische Verlagsgesellschaft mbh.
Epperson, B. (2003) Geographical genetics Princeton University Press.
Evanno, G., S. Regnaut & J. Goudet 2005 Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology 14(8): 2611-2620.
Excoffier, L., P.E. Smouse & J.M. Quattro 1992 Analysis of Molecular Variance Inferred from Metric Distances among DNA Haplotypes - Application to Human Mitochondrial-DNA Restriction Data. Genetics 131(2): 479-491.
Fukami, H., A. Budd, D. Levitan, J. Jara, R. Kersanach & N. Knowlton 2004 Geographic differences in species boundaries among members of the Montastraea annularis complex based on molecular and morphological markers. Evolution 58(2): 324-337.
53 Garcia-Ramos, G. & M. Kirkpatrick 1997 Genetic models of adaptation and gene flow in peripheral populations. Evolution 51(1): 21-28.
Gerrodette, T. 1981 Dispersal of the solitary coral Balanophyllia elegans by demersal planular larvae. Ecology 62(3): 611-619.
Gilmour, J. 2002 Substantial asexual recruitment of mushroom corals contributes little to population genetics of adults in conditions of chronic sedimentation. Marine Ecology Progress Series 235: 81-91.
Glynn, P. 1973 Acanthaster: effect on coral reef growth in Panamá. Science 180(4085): 504.
Glynn, P. 1976 Some physical and biological determinants of coral community structure in the eastern Pacific. Ecological Monographs 46(4): 431-456.
Glynn, P. 1977 Coral growth in upwelling and nonupwelling areas off the Pacific coast of Panamá. Journal of Marine Research 35(567-585.
Glynn, P. 1985 El Nino-associated disturbance to coral reefs and post-disturbance mortality by Acanthaster planci. Marine Ecology Progress Series 26(4): 295-300.
Glynn, P. (1988) Predation on coral reefs: some key process, concepts and research directions Proceedings of the 6th International Coral Reef Symposium 1: 51-62.
Glynn, P. 1990 Coral mortality and disturbances to coral reefs in the tropical eastern Pacific. Elsevier Oceanography Series 52: 55-126.
Glynn, P. (2001) History of significant coral bleaching events and insights regarding amelioration Coral Bleaching and Marine Protected Areas, 36-39.
Glynn, P. & J. Ault 2000 A biogeographic analysis and review of the far eastern Pacific coral reef region. Coral Reefs 19(1): 1-23.
Glynn, P. & M. Colgan 1992 Sporadic disturbances in fluctuating coral reef environments: El Nino and coral reef development in the eastern Pacific. Integrative and Comparative Biology 32(6): 707-718.
Glynn, P. & L. D'Croz 1990 Experimental evidence for high temperature stress as the cause of El Nino-coincident coral mortality. Coral Reefs 8(4): 181-191.
Glynn, P., N. Gassman, C. Eakin, J. Cortes, D. Smith & H. Guzman 1991 Reef coral reproduction in the eastern Pacific: Costa Rica, Panamá, and Galápagos Islands (Ecuador). I. Pocilloporidae. Marine Biology 109(3): 355-368.
Glynn, P. & G. Leyte Morales 1997 Coral reefs of Huatulco, West Mexico: reef development in upwelling Gulf of Tehuantepec. Revista de BiologÌa Tropical 45: 1033-1048.
54 Glynn, P. & I. Macintyre (1977) Growth rate and age of coral reefs on the Pacific coast of Panamá. Proceedings of the 3rd International Coral Reef Symposium 2, 251-259.
Glynn, P. & J. Mate (1997) Field guide to the Pacific coral reefs of Panamá 1: 145-166.
Glynn, P. & R. Stewart 1973 Distribution of coral reefs in the Pearl Islands (Gulf of Panamá) in relation to thermal conditions. Limnology and Oceanography 18(3): 367-379.
Glynn, P., R. Stewart & J. McCosker 1972 Pacific coral reefs of Panamá: structure, distribution and predators. International Journal of Earth Sciences 61(2): 483-519.
Glynn, P., J. Veron & G. Wellington 1996 Clipperton Atoll (eastern Pacific): oceanography, geomorphology, reef-building coral ecology and biogeography. Coral Reefs 15(2): 71-99.
Glynn, P. & G. Wellington 1983 Corals and Coral Reefs of the Galpagos Islands (with an annotated list of the scleractinian corals of the Galpagos by JW Wells). University of California, Berkeley, California, USA.
Guo, S.W. & E.A. Thompson 1992 A Monte-Carlo Method for Combined Segregation and Linkage Analysis. American Journal of Human Genetics 51(5): 1111-1126.
Guzman, H. & J. Cortes 1993 Arrecifes coralinos del Pacifico Oriental Tropical: revision y perspectivas - Coral reefs of the Tropical Eastern Pacific: review and perspectives. Revista de BiologÌa Tropical 41(3A): 535-557.
Guzman, H., J. Cortes, P. Glynn & R. Richmond 1990 Coral mortality associated with dinoflagellate blooms in the eastern Pacific (Costa Rica and Panamá). Marine Ecology Progress Series 60(3): 299-303.
Hardy, O. & X. Vekemans 1999 Isolation by distance in a continuous population: reconciliation between spatial autocorrelation analysis and population genetics models. Heredity 83(2): 145-154.
Hardy, O.J. & X. Vekemans 2002 SPAGeDi: a versatile computer program to analyse spatial genetic structure at the individual or population levels. Molecular Ecology Notes 2: 618- 620.
Harrison, P. & C. Wallace (1990) Reproduction, dispersal, and recruitment of scleractinian corals. In: Ecosystems of the world: Coral reefs (ed. Dubinsky Z), pp. 133-207. Elseiver.
Hedgecock, D. (1994) Does variance in reproductive success limit effective population sizes of marine organisms. In: Genetics and evolution of aquatic organisms (ed. Beaumont A), pp. 122-134. Chapman & Hall.
Hellberg, M. 1996 Dependence of gene flow on geographic distance in two solitary corals with different larval dispersal capabilities. Evolution 50(3): 1167-1175.
55 Hellberg, M.E. 1995 Stepping-Stone Gene Flow in the Solitary Coral Balanophyllia elegans - Equilibrium and Nonequilibrium at Different Spatial Scales. Marine Biology 123(3): 573- 581.
Herlihy, C. & C. Eckert 2002 Genetic cost of reproductive assurance in a self-fertilizing plant. Nature 416(6878): 320-323.
Heyward, A. & R. Babcock 1986 Self-and cross-fertilization in scleractinian corals. Marine Biology 90(2): 191-195.
Heywood, J. 1991 Spatial analysis of genetic variation in plant populations. Annual Review of Ecology and Systematics 22(1): 335-355.
Highsmith, R. 1980 Geographic patterns of coral bioerosion: a productivity hypothesis. Journal of Experimental Marine Biology and Ecology 46(2): 177-196.
Highsmith, R.C. 1982 Reproduction by fragmentation in corals. Marine Ecology Progress Series 7: 207-226.
Hoffmann, A. & M. Blows 1994 Species borders: ecological and evolutionary perspectives. Trends in Ecology & Evolution 9(6): 223-227.
Hughes, T.P., D.R. Bellwood & S.R. Connolly 2002 Biodiversity hotspots, centres of endemicity, and the conservation of coral reefs. Ecology Letters 5(6): 775-784.
Jakobsson, M. & N.A. Rosenberg 2007 CLUMPP1: CLUster Matching and Permutation Program Version 1.1.
Kimura, M. & J. Crow 1964 The number of alleles that can be maintained in a finite population. Genetics 49(4): 725.
Kimura, M. & T. Ohta 1978 Stepwise mutation model and distribution of allelic frequencies in a finite population. Proceedings of the National Academy of Sciences 75(6): 2868.
Kinzie, R. (1993) Spawning in the reef corals Pocillopora verrucosa and Pocillopora eydouxi at Sesoko Island, Okinawa. Galaxea, 11: 93-105.
Krupp, D. 1983 Sexual reproduction and early development of the solitary coral Fungia scutaria (Anthozoa: Scleractinia). Coral Reefs 2(3): 159-164.
LaJeunesse, T., H. Reyes-Bonilla & M. Warner 2007 Spring ‘bleaching’ among Pocillopora in the Sea of Cortez, Eastern Pacific. Coral Reefs 26(2): 265-270.
Lawton, J. 1993 Range, population abundance and conservation. Trends in Ecology & Evolution 8(11): 409-413.
56 Lessios, H. & D. Robertson 2006 Crossing the impassable: genetic connections in 20 reef fishes across the eastern Pacific barrier. Proceedings of the Royal Society B: Biological Sciences 273(1598): 2201-2209.
Lessios, H.A., B.D. Kessing & J.S. Pearse 2001 Population structure and speciation in tropical seas: Global phylogeography of the sea urchin Diadema. Evolution 55(5): 955-975.
Lessios, H.A., B.D. Kessing, D.R. Robertson & G. Paulay 1999 Phylogeography of the pantropical sea urchin Eucidaris in relation to land barriers and ocean currents. Evolution 53(3): 806-817.
Loiselle, B., V. Sork, J. Nason & C. Graham 1995 Spatial genetic structure of a tropical understory shrub, Psychotria officinalis (Rubiaceae). American Journal of Botany 82(11): 1420-1425.
Loveless, M. & J. Hamrick 1984 Ecological determinants of genetic structure in plant populations. Annual Review of Ecology and Systematics 15(1): 65-95.
Magalon, H., S. Samadi, M. Richard, M. Adjeroud & M. Veuille 2004 Development of coral and zooxanthella-specific microsatellites in three species of Pocillopora (Cnidaria, Scleractinia) from French Polynesia. Molecular Ecology Notes 4(2): 206-208.
Mayr, E. 1954 Geographic speciation in tropical echinoids. Evolution: 1-18.
McCartney, M., G. Keller & H. Lessios 2000 Dispersal barriers in tropical oceans and speciation in Atlantic and eastern Pacific sea urchins of the genus Echinometra. Molecular Ecology 9(9): 1391-1400.
McFadden, C.S. & K.Y. Aydin 1996 Spatial autocorrelation analysis of small-scale genetic structure in a clonal soft coral with limited larval dispersal. Marine Biology 126(2): 215- 224.
Miller, K. 1998 Short-distance dispersal of black coral larvae: inference from spatial analysis of colony genotypes. Marine Ecology Progress Series 163: 225-233.
Miller, K. & D. Ayre 2004 The role of sexual and asexual reproduction in structuring high latitude populations of the reef coral Pocillopora damicornis. Heredity 92(6): 557-568.
Miller, K. & D. Ayre 2008a Population structure is not a simple function of reproductive mode and larval type: insights from tropical corals. Journal of Animal Ecology 77(4): 713-724.
Miller, K. & D. Ayre 2008b Protection of genetic diversity and maintenance of connectivity among reef corals within marine protected areas. Conservation Biology 22(5): 1245- 1254.
Miller, K. & C. Mundy 2003 Rapid settlement in broadcast spawning corals: implications for larval dispersal. Coral Reefs 22(2): 99-106.
57 Neigel, J. & J. Avise 1983 Clonal diversity and population structure in a reef-building coral, Acropora cervicornis: self-recognition analysis and demographic interpretation. Evolution 37(3): 437-453.
Nishikawa, A. 2008 Degree and Pattern of Gene Flow in Several Scleractinian Corals in the Ryukyu Archipelago, Southern Japan. Pacific Science 62(3): 413-421.
Nishikawa, A., M. Katoh & K. Sakai 2003 Larval settlement rates and gene flow of broadcast- spawning (Acropora tenuis) and planula-brooding (Stylophora pistillata) corals. Marine Ecology Progress Series 256: 87-97.
Noreen, A., P. Harrison & M. Van Oppen 2009 Genetic diversity and connectivity in a brooding reef coral at the limit of its distribution. Proceedings of the Royal Society B: Biological Sciences 276(1675): 3927.
Nunes, F., R. Norris & N. Knowlton 2009 Implications of isolation and low genetic diversity in peripheral populations of an amphi-Atlantic coral. Molecular Ecology 18(20): 4283- 4297.
Orloci, L. 1978 Multivariante analysis in vegetation research. 451p. Junk, Den Haag
Palumbi, S. (1995) Using genetics as an indirect estimator of larval dispersal. In: Ecology of marine invertebrate larvae (ed. McEdward L), pp. 369-387. CRC Press, Boca Raton, FL.
Peakall, R. & P.E. Smouse 2006 GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. Molecular Ecology Notes 6(1): 288-295.
Pemberton, A., L. Hansson, S. Craig, R. Hughes & J. Bishop 2007 Microscale genetic differentiation in a sessile invertebrate with cloned larvae: investigating the role of polyembryony. Marine Biology 153(1): 71-82.
Pinzón, J.H. & T. LaJeunesse 2011 Species delimitation of common reef corals in the genus Pocillopora using nucleotide sequence phylogenies, population genetics and symbiosis ecology. Molecular Ecology 20(2): 311-325.
Pritchard, J.K., M. Stephens & P. Donnelly 2000 Inference of population structure using multilocus genotype data. Genetics 155: 945-959.
Raymond, M. & F. Rousset 1995 Genepop (Version-1.2) - Population-Genetics Software for Exact Tests and Ecumenicism. Journal of Heredity 86(3): 248-249.
Rice, W. 1989 Analyzing tables of statistical tests. Evolution 43(1): 223-225.
Richmond, R. 1985 Variations in the population biology of Pocillopora damicornis across the Pacific. Marine Biology 6: 101-106.
58 Richmond, R. 1987 Energetic relationships and biogeographical differences among fecundity, growth and reproduction in the reef coral Pocillopora damicornis. Bulletin of Marine Science 41(2): 594-604.
Richmond, R. & C. Hunter 1990 Reproduction and recruitment of corals: Comparisons among the Caribbean, the Tropical Pacific, and the Red Sea. Marine Ecology Progress Series 60(1): 185-203.
Ritland, K. 1996 Estimators for pairwise relatedness and individual inbreeding coefficients. Genetical Research 67: 175–185.
Rosenberg, N.A. 2004 DISTRUCT: a program for the graphical display of population structure. Molecular Ecology Notes 4(1): 137-138.
Ruzzante, D., C. Taggart & D. Cook 1996 Spatial and temporal variation in the genetic composition of a larval cod (Gadus morhua) aggregation: cohort contribution and genetic stability. Canadian Journal of Fisheries and Aquatic Sciences 53(12): 2695-2705.
Schmidt-Roach, S., K.J. Miller, E. Woolsey, G. Gerlach & A. Baird 2012 Broadcast spawning by Pocillopora species on the Great Barrier Reef. PLoS ONE 7(12): e50847.
Schuelke, M. 2000 An economic method for the fluorescent labeling of PCR fragments. Nature Biotechnology 18(2): 233-234.
Sherman, C., D. Ayre & K. Miller 2006 Asexual reproduction does not produce clonal populations of the brooding coral Pocillopora damicornis on the Great Barrier Reef, Australia. Coral Reefs 25(1): 7-18.
Slatkin, M. 1987 Gene flow and the geographic structure of natural populations. Science 236(4803): 787.
Slatkin, M. 1995 A measure of population subdivision based on microsatellite allele frequencies. Genetics 139(1): 457.
Sokal, R. & D. Wartenberg 1983 A test of spatial autocorrelation analysis using an isolation-by- distance model. Genetics 105(1): 219.
Soule, M. 1973 The epistasis cycle: a theory of marginal populations. Annual Review of Ecology and Systematics 4: 165-187.
Souter, P., O. Henriksson, N. Olsson & M. Grahn 2009 Patterns of genetic structuring in the coral Pocillopora damicornis on reefs in East Africa. BMC Ecology 9(1): 19.
Starger, C., P. Barber, Ambariyanto & A. Baker 2010 The recovery of coral genetic diversity in the Sunda Strait following the 1883 eruption of Krakatau. Coral Reefs: 1-19.
59 Starger, C.J., S.S.R. Yeoh, C.F. Dai, A.C. Baker & R.O.B. Desalle 2007 Ten polymorphic STR loci in the cosmopolitan reef coral, Pocillopora damicornis. Molecular Ecology Notes 8(3): 619-621
Stoddart, J. 1984 Genetic differentiation amongst populations of the coral Pocillopora damicornis off Southwest Australia. Coral Reefs 3(3): 149-156.
Stoddart, J. 1988a Coral populations fringing islands - larval connections. Australian Journal of Marine and Freshwater Research 39(1): 109-115.
Stoddart, J. (1988b) Historecognition and fine-scale spatial genetic structure in sessile benthic invertebrates. In: Invertebrate Historecognition (eds. Grosberg R, Hedgecock D, Nelson K), pp. 111-125. Plenum Publishing Corporation, New York.
Stoddart, J.A. 1983 Asexual production of planulae in the coral Pocillopora damicornis. Marine Biology 76:279-284.
Templeton, A. 1980 The theory of speciation via the founder principle. Genetics 94(4): 1011- 1020.
Trenberth, K. & T. Hoar 1997 El Nino and climate change. Geophysical Research Letters 24(23): 3057-3060.
Ulstrup, K., R. Berkelmans, P. Ralph & M. Van Oppen 2006 Variation in bleaching sensitivity of two coral species across a latitudinal gradient on the Great Barrier Reef: the role of zooxanthellae. Marine Ecology Progress Series 314: 135-148.
Underwood, J.N., L.D. Smith, M.J.H. Van Oppen & J.P. Gilmour 2007 Multiple scales of genetic connectivity in a brooding coral on isolated reefs following catastrophic bleaching. Molecular Ecology 16(4): 771-784. van Oosterhout, C., W. Hutchinson, D. Willis & P. Shipley 2004 MICRO-CHECKER: software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology Notes 4(3): 535-538. van Oppen, M. & R. Gates 2006 Conservation genetics and the resilience of reef-building corals. Molecular Ecology 15(13): 3863-3883. van Woesik, R. 2009 Calm before the spawn: global coral spawning patterns are explained by regional wind fields. Proceedings of the Royal Society B: Biological Sciences 277(1682): 715-722.
Vekemans, X. & O. Hardy 2004 New insights from fine-scale spatial genetic structure analyses in plant populations. Molecular Ecology 13(4): 921-935.
Veron, J.E.N. (2000) Corals of the World AIMS, Townsville, Australia.
60 Vollmer, S.V. & S.R. Palumbi 2007 Restricted gene flow in the Caribbean staghorn coral Acropora cervicomis: Implications for the recovery of endangered reefs. Journal of Heredity 98(1): 40-50.
Vucetich, J. & T. Waite 2003 Spatial patterns of demography and genetic processes across the species' range: null hypotheses for landscape conservation genetics. Conservation Genetics 4(5): 639-645.
Whitaker, K. 2006 Genetic evidence for mixed modes of reproduction in the coral Pocillopora damicornis and its effect on population structure. Marine Ecology Progress Series 306: 115-124.
Willis, B. & J. Oliver 1990 Direct tracking of coral larvae - implications for dispersal studies of plaktonic larvae in topographically complex environments. Ophelia 32(1-2): 145-162.
Willis, B.L., R.C. Babcock, P.L. Harrison & C.C. Wallace 1997 Experimental hybridization and breeding incompatibilities within the mating systems of mass spawning reef corals. Coral Reefs 16: S53-S65.
Wright, S. 1931 Evolution in Mendelian populations. Genetics 16: 97-159.
Wright, S. 1951 The genetical structure of populations. Annuals Eugenetics 15: 323-354.
Yeoh, S. & C. Dai 2010 The production of sexual and asexual larvae within single broods of the scleractinian coral, Pocillopora damicornis. Marine Biology 157(2): 351-359.
Yund, P.O. & P.G. O'Neil 2000 Microgeographic genetic differentiation in a colonial ascidian (Botryllus schlosseri) population. Marine Biology 137(4): 583-588.
Zhao, R., H. Xia & L. Bao-Rong 2009 Fine-scale genetic structure enhances biparental inbreeding by promoting mating events between more related individuals in wild soybean (Glycine soja; Fabaceae) populations. American Journal of Botany 96(6): 1138-1147.
61 Tables
Table 1: Population genetic indices for the six microsatellite loci.
Indices are based on genets, i.e. unique genotypes, per locus and over all loci (= Total).
N = Number of samples; NA = Number of alleles per locus; Ho = Heterozygosity observed; He =
Heterozygosity expected; FIS = (He-Ho)/He = Inbreeding coefficient; SE = Standard Error; *p<0.05, **p<0.005 per loci
Marker N NA Ho ± SE He ± SE FIS Pd2 183 8 0.551 ± 0.035 0.633 ± 0.019 0.239** Pd4 161 5 0.292 ± 0.056 0.489 ± 0.044 0.606** Pd5 188 9 0.509 ± 0.031 0.641 ± 0.014 0.267** Pd6 187 6 0.625 ± 0.077 0.629 ± 0.021 0.077 Pv2 173 4 0.109 ± 0.033 0.426 ± 0.062 0.831** Pv6 172 10 0.597 ± 0.057 0.738 ± 0.020 0.298** Total 194 42 0.447 ± 0.032 0.593 ± 0.020 0.367**
62 Table 2: Population genetic indices for the nine populations.
Indices are based on genets, i.e. unique genotypes, except for number of samples. Pop = Population: U = Uva, G = Granito, C = Canal de Afuera, B = Bahia de Damas, A = Achotines, I = Iguana, S = Saboga, Co = Contadora, T = Taboga;
N = Number of samples; NG = Number of unique genotypes; NG/N = Genet ration; NA = Average number of alleles per population; SE = Standard Error; Ho = Heterozygosity observed; He =
Heterozygosity expected; FIS = (He-Ho)/He = Inbreeding coefficient; *p<0.005 per pop
Region Pop N NG/N NA ± SE Ho ± SE He ± SE FIS ± SE U 23 0.91 5.2 ± 0.8 0.589 ± 0.069 0.624 ± 0.049 0.062 ± 0.065 Gulf of G 21 0.95 5.4 ± 0.9 0.531 ± 0.110 0.687 ± 0.018 0.236 ± 0.154* Chiriqui C 18 1.00 5.0 ± 0.9 0.378 ± 0.081 0.640 ± 0.039 0.421 ± 0.103* B 30 0.93 4.8 ± 0.5 0.580 ± 0.081 0.613 ± 0.075 0.065 ± 0.026 Azurero A 23 0.91 4.4 ± 0.7 0.615 ± 0.079 0.607 ± 0.044 -0.048 ± 0.168 Peninsula I 24 1.00 4.6 ± 1.0 0.544 ± 0.120 0.605 ± 0.080 0.143 ± 0.159* S 22 0.82 4.0 ± 0.3 0.390 ± 0.072 0.575 ± 0.059 0.336 ± 0.087* Gulf of Co 25 0.96 4.6 ± 0.7 0.445 ± 0.093 0.648 ± 0.047 0.314 ± 0.128* Panama T 21 0.95 4.2 ± 0.4 0.562 ± 0.057 0.634 ± 0.019 0.114 ± 0.084* Total 207 0.94 4.7 ± 0.2 0.515 ± 0.029 0.626 ± 0.016 0.183 ± 0.042*
63 Table 3: Hierarchical AMOVA results
Hierarchical AMOVA results showing levels of genetic structure among regions (RRT/FRT), among populations within regions (RSR/FSR) and among populations (RST/FST).
Source of Variation df Variance Variation Differentiation p R-statistics
Among Regions 2 1331 8.1% RRT = 0.081 < 0.001
Among Populations 6 1336 8.1% RSR = 0.089 < 0.001
Within Populations 379 13755 83.8% RST = 0.162 < 0.001 F-statistics
Among Regions 2 0.014 0.8% FRT = 0.008 0.004
Among Populations 6 0.081 4.5% FSR = 0.045 < 0.001 Within Populations 379 1.700 94.7% < 0.001 FST = 0.053
64 Table 4: Pairwise RST and FST between TEP P. damicornis populations
Pairwise RST (above diagonal) and FST (below diagonal) between TEP P. damicornis populations.
*p<0.05; bold = significant after Sequential Bonferroni correction GC = Gulf of Chiriquí, AP = Azuero Peninsula, GP = Gulf of Panamá; U = Uva, G = Granito, C = Canal de Afuera, B = Bahia de Damas, A = Achotines, I = Iguana, S = Saboga, Co = Contadora, T = Taboga;
RST Gulf of Chiriqui Azuero Peninsula Gulf of Panama
FST U G C B A I S Co T GC U 0.114* 0.025 0.025 0.086* 0.092* 0.053* 0.102* 0.309* G 0.035* 0.229* 0.124* 0.000 0.000 0.065* 0.085* 0.281* C 0.033* 0.229* 0.029 0.156* 0.188* 0.080* 0.129* 0.316* B 0.065* 0.073* 0.047* 0.086* 0.116* 0.014 0.065* 0.243* AP A 0.015 0.001 0.037* 0.069* 0.013* 0.046* 0.072* 0.222* I 0.039* 0.048* 0.025* 0.045* 0.029* 0.039* 0.037* 0.227* GP S 0.019 0.061* 0.015 0.053* 0.040* 0.022* 0.000 0.133* Co 0.025* 0.039* 0.040* 0.060* 0.031* 0.033* 0.000 0.091* T 0.087* 0.072* 0.064* 0.065* 0.071* 0.069* 0.046* 0.030*
65 Table 5: Pocillopora damicornis population genetic survey (sorted by FST).
1= consistent heterozygote deficits, reported as differences between He & Ho, i.e. inconvertible; 2= two different types of P. damicornis; * = significant; - = not significant;
Locations Scale (km) FST FIS Clones Marker Reference GBR (One Tree Island) 5 0.04 * 0.23 18% Allozymes (7 loci) Ayre & Miller 2004 GBR (One Tree Island) 1 0.05 * 0.03 - 13% Allozymes (5 loci) Benzie et al. 1995 GBR (One Tree Island) 5 0.06 * 0.11 * 17% Allozymes (8 loci) Sherman et al. 2006 GBR 1200 0.06 * 0.28 * 20% Allozymes (7 loci) Ayre et al. 1997 Japan (Okinawa) 650 0.06 * 0.18 * 39% Allozymes (7 loci) Adjeroud & Tsuchiya 1999 Northwest Australia 900 0.08 * 0.18 85% Allozymes (6 loci) Whitaker 2006 Lord Howe Island 10 0.10 * 0.17 28% Allozymes (7 loci) Miller & Ayre 2004 Lord Howe Island to GBR 2500 0.15 * N/A N/A Allozymes (7 loci) Ayre & Hughes 2004 Southeast Australia to GBR 1200 0.24 * 0.44 49% Allozymes (8 loci) Miller & Ayre 2008b Southeast Australia 700 0.32 * 0.47 42% Allozymes (8 loci) Miller & Ayre 2008b Southwest Australia 400 0.39 * 1 N/A 82% Allozymes (4 loci) Stoddart 1984 Hawaii (Kaneohe Bay) 10 N/A N/A 71% Allozymes (4 loci) Stoddart 1986
East Africa 860 0.02 * 0.26 * 13% Microsats (6 loci) Souter et al. 2009 Indonesia 3300 0.05 N/A 2% Microsats (9 loci) Starger et al. 2010 Tropical Eastern Pacific 400 0.05 * 0.18 * 6% Microsats (5 loci) THIS STUDY Taiwan (Nanwan Bay) 0.5 N/A N/A 66% Microsats (7 loci) Yeoh & Dai 2010
TEP to Central-West Pacific 8000 0.42 * ITS2-5.8S Combosch et al. 2008
66 Figures
Figure 1: Map of Panamá, showing the three main Pacific regions and the sampled locations.
The pie charts show the distribution of the five genetic clusters identified by STRUCTURE 2.3 in the three regional and nine reef populations. The bar plot shows each individual sample as a vertical bar with colors indicating the relative proportions of each genetic cluster. Map created on www.aquarius.ifm-geomar.de.
U = Uva Island; C = Canal de Afuera Island; G = Granito de Oro Island; B = Bahia de Damas, Coiba Island; A = Achotines Bay; I = Iguana Island; S = Saboga Island; Co = Contadora Island; T = Taboga Island.
67
Figure 2: Principal component analysis of Pocillopora damicornis populations, constructed using GenAlEx 6.1.
Distances between populations were analyzed as RST pairwise genetic distances. The two axes explained 90.4% of the total variation. Blue circles = Populations in the Gulf of Chiriquí; Red triangles = Azuero Peninsula; Green squares = Gulf of Panamá.
68
Figure 3: Spatial correlograms of the SpaGeDi analysis of Spatial Genetic Structure (SGS) among specimen within discrete distance classes.
Figure 3a) Spatial correlogram of the average pairwise kinship Fij as a function of distance over all populations.
Figure 3b-d) Spatial correlograms over the populations in each of the three regions, b) the Gulf of Chiriquí, c) the Azuero Peninsula and d) the Gulf of Panamá.
Black lines represent the genet dataset, excluding clones and grey lines show the results for the ramet dataset, including clones. Dashed lines are the permuted 95% confidence intervals. Error bars are 95% bootstrap errors of the genet datasets.
69 Supplementary Material
Figure S1: Results of the Bayesian clustering approach implemented in STRUCTURE 2.3 that were used to infer the number of genetic clusters (K) in the microsatellite dataset
The graph shows the mean log likelihood L(K) for K = 1-19 (±SD) and ∆K, the second order rate of change of L(K) for K = 2–18. The most likely number of cluster is K = 5, since it has the highest likelihood value (L(K) = -2059) and a local maximum for ∆K.
70 Chapter 3
MIXED ASEXUAL AND SEXUAL REPRODUCTION IN THE INDO-PACIFIC REEF
CORAL POCILLOPORA DAMICORNIS3
Abstract
Pocillopora damicornis is one of the best-studied reef-building corals, yet its somewhat unique reproductive strategy remains poorly understood. Genetic studies indicate that P. damicornis larvae are produced almost exclusively parthenogenetically, and yet population genetic surveys suggest frequent sexual reproduction. Using microsatellite data from over 580 larvae from 13 colonies, we demonstrate that P. damicornis displays a mixed reproductive strategy where sexual and asexual larvae are produced simultaneously within the same colony.
The majority of larvae were parthenogenetic (94%), but most colonies (10 out of the 13) produced a subset of their larvae sexually. Logistic regression indicates that the proportion of sexual larvae varied significantly with colony size, cycle day and calendar day. In particular, the decrease in sexual larvae with colony size suggests that the mixed reproductive strategy changes across the life of the coral. This unique shift in reproductive strategy leads to increasingly asexual replications of successful genotypes, which (in contrast to exclusive parthenogens) have already contributed to the recombinant gene pool.
3 Published as: Combosch D.J. & S.V. Vollmer 2013. Mixed asexual and sexual reproduction in the Indo-Pacific reef coral Pocillopora damicornis. Ecology and Evolution 3(10): 3379-3387
71 Introduction
Evolutionary theory predicts that sex is beneficial because recombination generates new, potentially advantageous gene combinations (Weismann 1886; Fisher 1930; Muller 1932) and helps to purge deleterious alleles (Muller 1964; Kondrashov 1988) by exposing them to natural selection (Hill & Robertson 1966; Felsenstein 1974). In contrast, asexual reproduction propagates successful genotypes and does not require mates (e.g. Weismann 1886; Williams
1975; Holman 1987; Radtkey et al. 1995; Weider et al. 1999). Most asexual lineages depend on occasional sexual recombination to add variation to their genomes, purge deleterious mutations, and adapt to changing environments. A mixed strategy of asexual propagation with low levels of sex combines the best of both strategies (Hurst & Peck 1996). For example, sperm-dependent parthenogenesis allows for recombination through occasional leakage of parental DNA from con- or hetero-specific sexual progenitors (Beukeboom & Vrijenhoek 1998). Sexual and parthenogenetic reproduction are usually well separated by time (e.g. in cyclic parthenogens) or population (geographic parthenogens) or produce different offspring (e.g. in many insects).
Most reef-building corals reproduce sexually and asexually. Sexual reproduction is achieved by either free spawning eggs and sperm, or internally brooding larvae inside the coral polyp (Harrison 2011). The resulting planula larvae then disperse, settle, and metamorphose to form a new coral polyp. In colonial corals, the coral colony then grows by asexual fission of individual polyps. Some corals can reproduce asexually via vegetative fragmentation (Highsmith
1982). In a handful of coral species, genetically identical coral colonies can also be formed through the production of parthenogenetic larvae (Harrison 2011). Parthenogenesis was first documented in the Indo-Pacific coral Pocillopora damicornis (Stoddart 1983), where it has been confirmed repeatedly using allozyme data (Stoddart 1986; Ayre & Miller 2004; Sherman et al.
72 2006). None of these studies detected any sign of sexual reproduction, even though population genetic surveys suggest frequent sexual recombination amidst predominantly asexual reproduction (e.g. Ayre & Miller 2004). Recent microsatellite analyses indicate that two P. damicornis colonies in Taiwan produced a subset (6 and 29%) of their larvae sexually (Yeoh &
Dai 2010), suggesting that P. damicornis can simultaneously produce asexual and sexual larvae.
The exact origin of the larvae remains unclear. Several authors hypothesized that asexual larvae originate via budding, analogous to polyp budding (Muir 1984; Stoddart & Black 1985; Ward
1992), however, histological observations of gametogenesis and embryogenesis suggest that parthenogenetic larvae are the result of vegetative embryogenesis in diploid oocytes (Harriott
1983; Martin-Chavez 1986; Diah-Permata et al. 2000).
The evolutionary significance of parthenogenetic reproduction in P. damicornis remains unknown. The extent and patterns of parthenogenetic larvae production among colonies are poorly defined and nothing is known about the potentially different roles and fates of sexual vs. parthenogenetic larvae. Using microsatellite genotyping, we quantified the extent of sexual and asexual reproduction across multiple colonies of P. damicornis. We examined whether colony genotype, size or habitat are associated with the production of parthenogenetic and/or sexual larvae. In addition, we asked if parthenogenetic and/or sexual larvae are preferentially released at particular times during the lunar and/or reproductive cycle, to characterize potential differences
(e.g. dispersion vs. population maintenance - sensu Stoddart 1988).
73 Materials and Methods
Organisms and larvae collection methods
In January 2010, 23 P. damicornis colonies (Figure 1) were collected from fringe and back reef sites on the north shore of Moorea, French Polynesia. Colonies were transported to the
Gump Research Station and kept in large seawater tanks (~5m3). A small nubbin (1cm) was sampled from each colony and preserved in GITC-buffer for genetic analysis. Three days before the new moon (January 15th and February 13th), colonies were transferred to individual containers with continuous seawater inflow. The outflow of each container was diverted into a short PVC tube with a mesh filter (100µm) where the larvae accumulated during planulation
(Figure 1). Larvae were collected using a 1000µl micropipette and preserved individually in
200µl Guanidinium-Isothiocyanate (GITC) buffer (Fukami et al. 2004). Between January 17 and
20, colonies were monitored throughout the night from 1900h to 0600h and larvae were collected and preserved as frequently as possible (i.e. every two to three hours). In February 2010, planulation lasted from February 14 to 26, i.e. until two nights before the full moon, and larvae were collected only in the morning.
Parent colonies
Pocillopora damicornis were collected from fringing and back reef habitats to assess potential environmental influences on reproductive strategies. Pocillopora damicornis is very abundant on fringe reefs around Moorea while on back reefs it is restricted to wave-protected grooves and crevices.
74 Colony size is often related to colony age and fecundity (Harrison 2011). Its impact on reproduction was measured in two ways. Colony size (as plane surface area, based on the longest and shortest colony diameter) was included in the main logistic regression analyses as a continuous variable. In addition, a separate comparison of larvae production between large and small colonies was conducted. Small colonies (n=7) were defined as having a maximum diameter of less than 10cm and a surface area under 75cm2, while large colonies (n=4) had minimum diameter over 10cm and a surface area over 130cm2 (Table 2).
Eighteen out of the 23 P. damicornis colonies collected released each between 1 and over
1000 larvae during our observations following the new moon in January and February 2010. In order to ensure reliable characterization of each colony’s reproductive patterns, only colonies that released more than 10 larvae (n = 13) were included in the analyses presented here.
Genetic analyses
DNA was extracted with Agencourt DNAdvance® isolation kits (Beckman Coulter Inc.).
Ten microsatellite loci were amplified with primers developed by Magalon et al. (2004) and
Starger et al. (2007). PCR products included infrared labels on forward primers, and were manually sized and scored on a LICOR 4300 Genetic Analyzer. Six loci amplified consistently using modified PCR protocols (Combosch & Vollmer 2011) and were used in all subsequent analyses (Table 1). Only larvae that were genotyped successfully at five or more loci were included in the analyses. Linkage disequilibrium among loci was assessed using a Markov Chain
Monte Carlo resampling method (Guo & Thompson 1992) in GENEPOP (Raymond & Rousset
1995), followed by sequential Bonferroni correction to account for multiple comparisons.
75 Probability of Identity (PI) and the inbreeding coefficient FIS were calculated using GenAlEx
6.4.1 (Peakall & Smouse 2006).
Larvae with genotypes identical to their parent were considered parthenogenetically produced. Larvae with genotypes that differed from the maternal genotype were considered sexually produced. Sexually produced larval genotypes were examined for signs of selfing (i.e. absence of alleles not already present in the maternal genotype).
Data analyses
Logistic regression analyses were used because the data are dichotomous (i.e., sexual vs. asexual) and binomially distributed (Wilson & Hardy 2002; Warton & Hui 2011). The main logistic regression analysis was conducted using a generalized linear mixed effects model with the proportion of sexually produced larvae per colony per day as the response variable. Six different predictor variables were included as fixed effects in the model. Three factors were colony-related, time-independent factors: colony genotype, colony size, and colony habitat.
Three factors were time-related, colony-independent factors: calendar day, lunar day, and release day. Lunar day and release day were used to assess daily patterns of larval release across reproductive cycles (i.e. January and February 2010). Lunar day refers to days after the new moon while release day refers to the day after a specific colony started to release larvae (i.e. days into the larval-release cycle of a colony). “Colony ID” was treated as a random effect to determine whether the results were consistent across colonies.
A second logistic regression analysis was used to compare the proportion of sexually produced larvae between small and large colonies. The same type of generalized linear mixed effects model was used with the two size categories as predictor variables. All statistical analyses
76 were conducted in the program R version 2.14.0 (R Core Team 2012) using the glmer function in the lme4 package (Bates et al. 2009).
Results
Over 3000 larvae were released by the 18 Pocillopora damicornis colonies. One colony
(P10), released more than 1000 larvae over just three nights in January 2010. In total, 960 larvae were collected and preserved in GITC buffer for microsatellite genotyping; 597 larvae were successfully genotyped at 5 to 6 microsatellite loci (Table 1). Data analyses were conducted on
583 larvae that were released by 13 colonies, which each released 10 or more larvae (Table 2).
The six microsatellite loci had between 2 and 6 alleles (Table 1). No significant linkage disequilibria (LD) between loci were detected among colony genotypes (p > 0.05 after
Sequential Bonferroni correction). The probability of identity (PI) to identify unique genotypes using all six loci was 3.9x10-4 (Table 1).
Out of the 583 genotyped larvae, 546 larvae (93.7%) had identical genotypes to their parent and were scored as parthenogenetic larvae. The genotypes of the remaining 37 larvae
(6.3%) differed from the parental genotype at one allele over one or more loci and were scored as sexual larvae. Eight of the 37 sexual larvae could have been produced via meiotic selfing; seven of these larvae had identical siblings, but differed from their brood parent’s genotypes.
Ten out of the 13 genotyped P. damicornis colonies released both sexually and asexually produced larvae (Table 2). All ten produced mostly parthenogenetic larvae (>70%). The other three colonies produced only parthenogenetic larvae (see Table 2). Two out of these three coral colonies had identical genotypes (P8 & P9). However, a third colony had the identical genotype
77 as well (P6) and produced some larvae sexually. This suggests that parthenogenesis is influenced by environmental, demographic and genetic factors.
Reproduction changes with colony size and over time
Logistic regression analyses revealed significant effects of Colony Size, Cycle Day and
Calendar Day on the proportions of sexually produced larvae (Table 3). Colony Genotype,
Colony Habitat, and Lunar Day were not statistically significant (p > 0.05).
The proportion of sexually produced larvae declined significantly with colony size (LRT, p = 0.029; Figure 2). Moreover, small colonies produced significantly more larvae sexually
(~10.5%) than large colonies (~2%; Table 2 and Figure 2; single factor LRT: p = 0.02). The proportion of sexually produced larvae also declined significantly over the course of each colony’s larval release cycle (Cycle Day; Figure 3; LRT, p = 0.004). However, the proportion of sexually produced larvae increased significantly over the two month monitoring period
(Calendar Day; LRT, p = 0.040), e.g. from January (7/223; 3.1%) to February (30/360; 8.3%).
Discussion
Microsatellite data from 583 coral larvae and their 13 brood-parents indicate that
Pocillopora damicornis in Moorea use a simultaneously mixed sexual and asexual reproductive strategy. Over two reproductive cycles, ten out of the thirteen surveyed P. damicornis colonies produced both sexual and parthenogenetic larvae. All colonies produced on average 87% of their larvae parthenogenetically; three colonies produced only parthenogenetic larvae. The proportion of sexual reproduction decreased significantly with increasing colony size and during each colony’s individual larval release cycle. Yet, over the bimonthly monitoring period, sexual
78 reproduction increased. Parental genotype, habitat and lunar day did not have significant effects on the proportion of sexual larvae. While the exact mechanisms of this strategy remain unknown, it appears that P. damicornis applies a minimal sex strategy, dominated by asexual parthenogenesis.
Previous genetic analyses of P. damicornis did not detect any sexually produced larvae
(Stoddart 1983; Stoddart 1986; Ayre & Miller 2004; Sherman et al. 2006). Sexually produced larvae might have been missed due to their low frequency (~6%), or mis-identified as parthenogenetic due to the low resolution of allozyme markers. In contrast, allozyme-based population genetic surveys of P. damicornis provided evidence for sexual recombination (i.e. high genotypic diversity) in P. damicornis populations across the Central-West Pacific (Western
Australia: Stoddart 1984a, b and Whitaker 2006; Great Barrier Reef: Benzie et al. 1995, Ayre et al. 1997, Ayre & Miller 2004 and Sherman et al. 2006; Lord Howe Island: Miller & Ayre 2004;
Okinawa: Adjeroud & Tsuchiya 1999; and potentially in Hawaii: Stoddart 1986). Histological studies of P. damicornis in Taiwan and Hawaii also suggested regular embryogenesis and found no indications of asexual budding (Martin-Chavez 1986; Diah-Permata et al. 2000). Recently,
Yeoh and Dai (2010) used multi-locus microsatellite markers to compare larvae of two P. damicornis colonies in Taiwan and documented asexual and sexual larvae release. Both histologic and indirect and direct genetic studies thus indicate that P. damicornis reproduces mixed, sexually and asexually, throughout the Central-West Pacific.
Mixed reproduction
One of the most intriguing characteristics of the mixed reproductive strategy of P. damicornis is the observed simultaneous release of parthenogenetic and sexual larvae.
79 Simultaneous mixed reproduction is rare in animals. In most animals that display mixed reproduction, sexual and asexual reproduction are separated by different conditions. In cyclical parthenogens such as monogonont rotifers, aphids and Daphnia, sexual reproduction is limited to particular seasons, environmental conditions and/or population densities (e.g. Hebert 1974;
Gómez & Carvalho 2001; Simon et al. 2002). In geographic parthenogens like squamate lizards and grasshoppers, asexual reproduction occurs in isolated populations, at particular high altitudes and latitudes or at the margins of species ranges (Vrijenhoek & Parker 2009). Animals that reproduce simultaneously sexually and asexually generally produce parthenogenetic offspring that are either haploid (e.g. hymenoptera) or predominantly homozygous (e.g. cockroaches and termites). Pocillopora damicornis are somewhat distinct in that sexual and apomictic parthenogenetic reproduction occurs simultaneously within the same coral colony.
Recent histological examinations clearly indicate that P. damicornis larvae result from oocytes following regular embryogenesis (Diah-Permata et al. 2000). Since parthenogenetic larvae have identical genotypes, including several heterozygotic loci, they must stem from apomictic diploid eggs. But haploid eggs are required for sexually produced larvae. This suggests that both haploid and diploid eggs would have to be produced; yet no indications of distinct types of oocytes and/or eggs have ever been reported in P. damicornis.
Alternatively, eggs might be generally diploid and the maternal chromosome constitution is retained or restored post-fertilization in parthenogenetic larvae. For example, in pseudogametic taxa, parthenogenetic reproduction requires sperm to stimulate embryogenesis without incorporating the parental DNA. Frequently, however, sperm nuclei evade complete degeneration and chromosome fragments, complete chromosome or entire sperm genomes ‘leak’ into the parthenogenetic eggs (Schartl et al. 1995; D'Souza et al. 2006; D’Souza & Michiels
80 2009). The paternal chromosomes are then either added to the genome (chromosome addition) or the maternal ploidy level is restored by removing a maternal chromosome set. Thus, a predominantly apomictic parthenogenetic clutch would contain a few genetically diverse propagules with small to moderate genotypic differences (D’Souza & Michiels 2009), which is exactly what we observed in P. damicornis.
Pseudogamy is rare, but occurs in a wide variety of taxa. It is more common among hybrid lineages, but non-hybrid pseudogametic taxa are also known (e.g. platyhelminthes,
Beukeboom & Vrijenhoek 1998). Since parental leakage occurs regularly in pseudogametic taxa, it has been termed “minimal sex” (Beukeboom & Vrijenhoek 1998) or “cryptic sex” (Schlupp
2005).
There are several striking similarities between the genetic patterns of mixed reproduction in P. damicornis and in the pseudogametic turbellarian Schmidtea polychroa (Platyhelminthes).
Two microsatellite studies of S. polychroa (D'Souza et al. 2004; D'Souza et al. 2006) report similar proportions of sexual offspring per parent (2-25% compared to 2-30% here) and across parents (6% each). Additionally, the number of loci affected by parental leakage (1-4 out of 4) and the relative prevalence of parental leakage among fertile parents (6/11) is very similar to our results (1-5 out of 6 loci and 10/13 fertile parents). Pseudogamy also explains the remarkable synchrony of asexual and sexual reproduction.
A common consequence of parental leakage in pseudogametic taxa is temporary polyploidy (Beukeboom & Vrijenhoek 1998). In P. damicornis larvae, polyploid allozyme genotypes have been described previously (Stoddart 1986) and several sexually produced larvae had multiple alleles at individual loci in this study (i.e. two maternal and one paternal). However, in other cases sexual reproduction led to a decrease in the number of observed alleles, from
81 heterozygous in the colony to homozygous in the larvae. The reproductive patterns observed in
P. damicornis thus match patterns described in pseudogametic taxa. Because most genotypic distances between maternal and larval genotypes were small (one or two loci), it is possible that somatic mutations and/or chimeric colonies (reviewed in van Oppen et al. 2011) may have contributed to the signature of sexual reproduction.
Evolutionary significance
Mixed reproduction is often described as a best-of-both-worlds scenario (Hurst & Peck
1996; Peck & Waxman 2000). It enables effective propagation of successful genotypes via parthenogenesis while sexual recombination constantly generates offspring with new allele combinations for selection. The extent of these benefits depends on the frequency of sex. In many cases a predominantly asexual strategy with a low frequency of sex seems to be most effective (Hurst & Peck 1996; Peck & Waxman 2000).
Mixed reproduction conveys many evolutionary advantages. In organisms using mixed reproductive strategies, natural selection acts not only on the level of the entire genome, as in self-replicating asexual taxa, but also on the level of individual genes, which are exposed to natural selection due to occasional sexual recombination. Moreover, the extent of asexual and sexual reproduction could be fine-tuned among genotypes so that successful genotypes are replicated mostly intact while less successful genotypes are more involved in sexual recombination.
Mixed reproduction can help organisms adapt to changing environments. The extent of asexual and sexual reproduction may vary across environments. If environmental conditions deteriorate and organisms become stressed, mixed reproducers might be capable of increasing
82 the proportion of sexual reproduction. Increased sexual recombination would generate a higher proportion of novel genotypes and a more diverse cohort of offspring to deal with the new condition.
Another benefit of mixed reproduction is reproductive assurance. Reproductive failure because of sperm limitation is a serious problem for many broadcast-sperm dependent marine invertebrates (Yund 2000). Parthenogenesis assures reproduction independent of fertilization success. Moreover, if unfertilized eggs of pseudogametic taxa are capable of eventually developing into parthenogenetic larvae without pseudo-fertilization, a second tier of parthenogenetic larvae would start to develop and consequently be released slightly delayed.
This is consistent with the observed decrease in the proportion of sexual larvae during the reproductive cycle of individual colonies.
Sexual reproduction decreases with increasing colony size
Among the most remarkable characteristics of P. damicornis mixed reproduction is that bigger colonies reproduce more asexually than smaller colonies (Figure 2). Because the proportion of sexual reproduction was genotype and habitat independent, the decrease of sexual reproduction with colony size appears to be a general life-history strategy. Colony size is an important proxy for a colony’s fecundity and evolutionary fitness (Hughes et al. 1992).
Correlations between colony size and reproductive strategy therefore have important evolutionary consequences. Large colonies affirm the success of their genotypes in local environments. The shift towards increasingly asexual reproduction by larger colonies could thus lead to increased recruitment and survival of these successful genotypes in larval cohorts.
However, in contrast to fully parthenogenetic taxa, successful brood parents will have already
83 contributed “good alleles” to the gene pool when they were small. Over time, this reproductive strategy should lead to a highly inbred population structure with moderately high genotypic diversity (due to ongoing sexual recombination). The genotype distribution should consist of a few successful genotypes, present in multiple colonies (but particularly dominant among big colonies) and many subordinate genotypes. This is exactly what has been observed in population genetic and demographic studies, not only in French Polynesia (Marks et al., in prep), but also in other P. damicornis populations across the Central-West Pacific (e.g. Stoddart 1984b; Whitaker
2006).
Sexual reproduction decreases during colonies’ reproductive cycle
Sexually produced larvae were released significantly earlier during the reproductive cycle of individual colonies than asexually produced larvae. One possible reason is that un-fertilized larvae eventually develop into parthenogenetic larvae, as outlined above (reproductive assurance hypothesis). Alternatively, sexually produced larvae might simply develop faster than their asexual counterparts, as is commonly observed in other taxa, for example in termites (Matsuura
2010) and cockroaches (Corley & Moore 1999). Parental factors might stimulate the development of sexual offspring, as in Drosophila melanogaster (Yasuda et al. 1995) and
Caenorhabditis elegans (Browning & Strome 1996). Faster development of sexual offspring could also be mediated genetically, e.g. via heterosis. Ultimately, sexually produced larvae might simply be preferred by the mother, as a means to allocate resources in a way that maximizes overall offspring fitness (Michiels et al. 1999).
84 Poly-embryos
Three different colonies (P3, P6 & P17) released sexually produced, but genetically identical siblings (two pairs of twins and one triplet). One pair of genetically identical, sexually produced P. damicornis larvae had also been observed in Taiwan (Yeoh & Dai 2010).
Genetically identical, sexually produced twins result from a single zygote that eventually divides into two or more clonal poly-embryos, akin to monozygotic twins in humans. This form of limited poly-embryony is considered accidental and does not seem to bear much evolutionary significance (Craig et al. 1997). While most P. damicornis polyps tend to produce only one larva at the time, multiple larvae per polyp are not uncommon (e.g. Stoddart & Black 1985; Martin-
Chavez 1986). Moreover, viable settlement-competent poly-embryos are known from other coral species, where free-swimming embryos break apart and multiply (Heyward & Negri 2012).
Conclusion
The results of this study in combination with Yeoh and Dai (2010), and several population genetics studies throughout the Central-West Pacific, showed that asexual and sexual larvae brooding is the predominant reproductive strategy of P. damicornis populations in the
Central-West Pacific. Populations at the margins of the species distribution range reproduce by broadcast spawning, e.g. at the Southern GBR (Schmidt-Roach et al. 2012), in the Tropical
Eastern Pacific (Glynn et al. 1991; Chávez-Romo & Reyes-Bonilla 2007) and in Western
Australia (Stoddart & Black 1985; Ward 1992), although taxonomic uncertainties exist for some of these peripheral populations (e.g. Schmidt-Roach et al. 2013). Multiple P. damicornis types with reproductive differences have been confirmed in Hawaii (Richmond & Jokiel 1984) and inferred in Western Australia (Ward 1995) and the Western Indian Ocean (Souter 2010). In
85 addition, inter-specific and inter-generic hybridization was inferred for peripheral P. damicornis populations in the Tropical Eastern Pacific (Combosch et al. 2008) and at Lord Howe Island
(Miller & Ayre 2004). In light of the confusing, sometimes contradictory results of allozyme- based studies, it is now clear that more high-resolution data (microsatellites, SNPs) are needed to assess the spatial and temporal variability of sexual reproduction and the impact of environmental variation on the reproductive strategy of P. damicornis. Eighty years after the first description of the reproductive strategy of P. damicornis (Marshall & Stephenson 1933) a lot of basic questions remain to be answered.
Acknowledgements
Field and logistics support was provided by Pete Edmunds, Gonzalo Giribet, Sal
Genovese and the Three Seas Program. Statistical advice was provided by Sean Kent and Tarik
Gouhier. DJC was supported by a Three Seas Teaching Fellowship and a German Academic
Exchange Service (DAAD) Pre-doctoral Research Fellowship. Thank you to Sarah Lemer, Sarah
Gignoux-Wolfsohn, Christopher Marks, and two anonymous reviewers for their comments on the manuscript.
86 References
Adjeroud M. & M. Tsuchiya 1999. Genetic variation and clonal structure in the scleractinian coral Pocillopora damicornis in the Ryukyu Archipelago, southern Japan. Marine Biology 134: 753-760.
Ayre D., T. Hughes & R. Standish 1997 Genetic differentiation, reproductive mode, and gene flow in the brooding coral Pocillopora damicornis along the Great Barrier Reef, Australia. Marine Ecology Progress Series 159: 175-187.
Ayre D. & K. Miller 2004 Where do clonal coral larvae go? Adult genotypic diversity conflicts with reproductive effort in the brooding coral Pocillopora damicornis. Marine Ecology Progress Series 277: 95-105.
Bates D., M. Maechler & B. Dai 2009 lme4: linear mixed- effects models using S4 classes. R package version 0.999375-31. http://cran.r-project.org/web/packages/lme4i.
Benzie J.A.H. & S.T. Williams 1995 Gene Flow among Giant Clam (Tridacna gigas) Populations in Pacific Does Not Parallel Ocean Circulation. Marine Biology 123: 781- 787.
Beukeboom L. & R. Vrijenhoek 1998 Evolutionary genetics and ecology of sperm-dependent parthenogenesis. Journal of Evolutionary Biology 11: 755-782.
Browning H. & S. Strome 1996 A sperm-supplied factor required for embryogenesis in C. elegans. Development 122: 391-404.
Chávez-Romo H. & H. Reyes-Bonilla 2007 Reproducción sexual del coral Pocillopora damicornis al sur del Golfo de California, México. Ciencias Marinas, 33: 495-501.
Combosch D.J., H. Guzman, H. Schuhmacher, S. Vollmer 2008 Interspecific hybridization and restricted trans-Pacific gene flow in the Tropical Eastern Pacific Pocillopora. Molecular Ecology 17: 1304-1312.
Combosch D.J. & S.V. Vollmer 2011 Population Genetics of an Ecosystem-Defining Reef Coral Pocillopora damicornis in the Tropical Eastern Pacific. PLoS ONE 6, e21200.
Corley L.S. A.J. Moore 1999 Fitness of alternative modes of reproduction: developmental constraints and the evolutionary maintenance of sex. Proceedings of the Royal Society of London. Series B: Biological Sciences 266: 471-476.
Craig S., L. Slobodkin, G. Wray & C. Biermann 1997 The ‘paradox’of polyembryony: a review of the cases and a hypothesis for its evolution. Evolutionary Ecology 11: 127-143.
D'Souza T.G., R.D. Schulte, H. Schulenburg & N.K. Michiels 2006 Paternal inheritance in parthenogenetic forms of the planarian Schmidtea polychroa. Heredity 97: 97-101.
87 D'Souza T.G., M. Storhas, H. Schulenburg, L.W. Beukeboom & N.K. Michiels 2004 Occasional sex in an 'asexual' polyploid hermaphrodite. Proceedings of the Royal Society of London. Series B: Biological Sciences 271: 1001-1007.
D’Souza T.G. & N.K. Michiels 2009 Sex in Parthenogenetic Planarians: Phylogenetic Relic or Evolutionary Resurrection? In: Lost Sex - The Evolutionary Biology of Parthenogenesis (eds. Schön I, Martens K, Dijk Pv), pp. 377-397. Springer Dordrecht, Heidelberg.
Diah-Permata W.D., R.A. Kinzie & M. Hidaka 2000 Histological studies on the origin of planulae of the coral Pocillopora damicornis. Marine Ecology Progress Series 200: 191- 200.
Felsenstein J. 1974 The evolutionary advantage of recombination. Genetics 78: 737-756.
Fisher R.A. 1930 The genetical theory of natural selection. Oxford University Press, Oxford.
Fukami H., A. Budd, D. Levitan, et al. 2004 Geographic differences in species boundaries among members of the Montastraea annularis complex based on molecular and morphological markers. Evolution 58: 324-337.
Glynn P., N. Gassman, C. Eakin, et al. 1991 Reef coral reproduction in the eastern Pacific: Costa Rica, Panamá, and Galápagos Islands (Ecuador). I. Pocilloporidae. Marine Biology 109: 355-368.
Gómez A. & G. Carvalho 2001 Sex, parthenogenesis and genetic structure of rotifers: microsatellite analysis of contemporary and resting egg bank populations. Molecular Ecology 9: 203-214.
Guo S.W. & E.A. Thompson 1992 A Monte-Carlo Method for Combined Segregation and Linkage Analysis. American Journal of Human Genetics 51: 1111-1126.
Harriott V.J. 1983 Reproductive seasonality, settlement, and post-settlement mortality of Pocillopora damicornis (Linnaeus), at Lizard Island, Great Barrier Reef. Coral Reefs 2: 151-157.
Harrison P.L. 2011 Sexual Reproduction of Scleractinian Corals. In: Coral Reefs: An Ecosystem in Transition (eds. Dubinsky Z, Stambler N), pp. 59-85. Springer Science + Business Media B.V.
Hebert P.D. 1974 Enzyme variability in natural populations of Daphnia magna II. Genotypic frequencies in permanent populations. Genetics 77: 323-334.
Heyward A. & A.P. Negri 2012 Turbulence, Cleavage, and the Naked Embryo: A Case for Coral Clones. Science 335: 1064.
Highsmith R.C. 1982 Reproduction by fragmentation in corals. Marine Ecology Progress Series 7: 207-226.
88 Hill W. & A. Robertson 1966 The effect of linkage on limits to artificial selection. Genetical Research 8: 269-294.
Holman E.W. 1987 Recognizability of sexual and asexual species of rotifers. Systematic Biology 36: 381-386.
Hughes T.P., D. Ayre & J.H. Connell 1992 The Evolutionary Ecology of Corals. Trends in Ecology and Evolution 7: 292-295.
Hurst L.D. & J.R. Peck 1996 Recent advances in understanding of the evolution and maintenance of sex. Trends in Ecology and Evolution 11: 46-52.
Kondrashov A.S. 1988 Deleterious mutations and the evolution of sexual reproduction. Nature 336: 435-440.
Magalon H., S. Samadi, M. Richard, M. Adjeroud & M. Veuille 2004 Development of coral and zooxanthella-specific microsatellites in three species of Pocillopora (Cnidaria, Scleractinia) from French Polynesia. Molecular Ecology Notes 4: 206-208.
Marshall S. & T. Stephenson 1933 The breeding of reef animals. Part l. The corals. In: Scientific Reports of the Great Barrier Reef Expedition (1928–29), pp. 219–245.
Martin-Chavez E. 1986 Gametogenesis and origin of planulae in the hermatypic coral Pocillopora damicornis. In Results of the 1983 Summer Research Training Program at the Hawaii Institute of Marine Biology, 193-206.
Matsuura K. 2010 Sexual and Asexual Reproduction in Termites. In: Biology of Termites: a Modern Synthesis (eds. Bignell DE, Roisin Y, Lo N), pp. 255-277. Springer Science + Business Media B.V.
Michiels N., L. Beukeboom, J. Greeff & A. Pemberton 1999 Individual control over reproduction: an underestimated element in the maintenance of sex? Journal of Evolutionary Biology 12: 1036-1039.
Miller K. & D. Ayre 2004 The role of sexual and asexual reproduction in structuring high latitude populations of the reef coral Pocillopora damicornis. Heredity 92: 557-568.
Muir P. 1984 Periodicity and asexual planulae production in Pocillopora damicornis (Linnaeus) at Magnetic Island BSc Honours thesis, James Cook University of North Queensland, Townsville.
Muller H.J. 1932 Some genetic aspects of sex. The American Naturalist 66: 118-138.
Muller H.J. 1964 The relation of recombination to mutational advance. Mutational Research 1: 2-9.
Peakall R. & P.E. Smouse 2006 GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. Molecular Ecology Notes 6: 288-295.
89 Peck J. & D. Waxman 2000 What's wrong with a little sex? Journal of Evolutionary Biology 13: 63-69.
R Core Team 2012 R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.
Radtkey R.R., S.C. Donnellan, R.N. Fisher, et al. 1995 When species collide: the origin and spread of an asexual species of gecko. Proceedings of the Royal Society of London. Series B: Biological Sciences 259: 145-152.
Raymond M. & F. Rousset 1995 Genepop (Version 1.2) - Population-Genetics Software for Exact Tests and Ecumenicism. Journal of Heredity 86: 248-249.
Richmond R. & P. Jokiel 1984 Lunar periodicity in larva release in the reef coral Pocillopora damicornis at Enewetak and Hawaii. Bulletin of Marine Science 34: 280-287.
Schartl M., I. Nanda, I. Schlupp, et al. 1995 Incorporation of subgenomic amounts of DNA as compensation for mutational load in a gynogenetic fish. Nature 373: 68-71.
Schlupp I. 2005 The evolutionary ecology of gynogenesis. Annual Review of Ecology Evolution and Systematics 36: 399-417.
Schmidt-Roach S., K.J. Miller, E. Woolsey, G. Gerlach & A.H. Baird 2012 Broadcast Spawning by Pocillopora Species on the Great Barrier Reef. PLoS One 7, e50847.
Schmidt-Roach S., P.Lundgren, K.J. Miller, et al. 2013 Assessing hidden species diversity in the coral Pocillopora damicornis from Eastern Australia. Coral Reefs 32: 161-172
Sherman C., D. Ayre, K. Miller 2006 Asexual reproduction does not produce clonal populations of the brooding coral Pocillopora damicornis on the Great Barrier Reef, Australia. Coral Reefs 25: 7-18.
Simon J.C., C. Rispe, P. Sunnucks 2002 Ecology and evolution of sex in aphids. Trends in Ecology and Evolution 17: 34-39.
Souter P. 2010 Hidden genetic diversity in a key model species of coral. Marine Biology 157: 875-885.
Starger C., S. Yeoh, C. Dai, A. Baker & R. Desalle 2007 Ten polymorphic STR loci in the cosmopolitan reef coral, Pocillopora damicornis. Molecular Ecology Notes 8: 619-621.
Stoddart J. 1984a Genetic differentiation amongst populations of the coral Pocillopora damicornis off Southwest Australia. Coral Reefs 3: 149-156.
Stoddart J. 1984b Genetical structure within populations of the coral Pocillopora damicornis. Marine Biology 81: 19-30.
90 Stoddart J. 1988 Coral populations fringing islands - larval connections. Australian Journal of Marine and Freshwater Research 39: 109-115.
Stoddart J. & R. Black 1985 Cycles of gametogenesis and planulation in the coral Pocillopora damicornis. Molecolar Ecology Progress Series 23: 153-164.
Stoddart J.A. 1983 Asexual production of planulae in the coral Pocillopora damicornis. Marine Biology 76: 279-284.
Stoddart J.A. 1986 Biochemical genetics of Pocillopora damicornis in Kaneohe Bay, Oahu, Hawaii. In: Results of the 1983 Summer Research Training Program at the Hawaii Institute of Marine Biology, pp. 133-148.
Vrijenhoek R.C. & E.D. Parker 2009 Geographical parthenogenesis: general purpose genotypes and frozen niche variation. In: Lost Sex (eds. Schön I, Martens K, Dijk P), pp. 99-131. Springer Dordrecht. van Oppen M.J.H., P. Souter, E.J. Howells, A. Heyward & R. Berkelmans 2011 Novel Genetic Diversity Through Somatic Mutations: Fuel for Adaptation of Reef Corals? Diversity 3: 405-423.
Ward S. 1992 Evidence for broadcast spawning as well as brooding in the scleractinian coral Pocillopora damicornis. Marine Biology 112: 641-646.
Ward S. 1995 Two patterns of energy allocation for growth, reproduction and lipid storage in the scleractinian coral Pocillopora damicornis. Coral Reefs 14: 87-90.
Warton D.I. & F.K.C Hui 2011 The arcsine is asinine: the analysis of proportions in ecology. Ecology 92: 3-10.
Weider L.J., A. Hobaek, P.D. Hebert & T. Crease 1999 Holarctic phylogeography of an asexual species complex - II. Allozymic variation and clonal structure in Arctic Daphnia. Molecular Ecology 8: 1-13.
Weismann A. 1886 Die Bedeutung der sexuellen Fortpflanzung für die Selektions-Theorie Gustav Fischer Verlag, Jena, Germany.
Whitaker K. 2006 Genetic evidence for mixed modes of reproduction in the coral Pocillopora damicornis and its effect on population structure. Marine Ecology Progress Series 306: 115-124.
Williams G. 1975 Sex and evolution Princeton University Press, Princeton, NJ.
Wilson K. & I. Hardy 2002 Statistical analysis of sex ratios: an introduction. In: Sex Ratios: concepts and research methods (ed. Hardy I), pp. 48-92. Cambridge University Press, New York, USA.
91 Yasuda G.K., G. Schubiger & B.T. Wakimoto 1995 Genetic characterization of ms (3) K81, a paternal effect gene of Drosophila melanogaster. Genetics 140: 219-229.
Yeoh S. & C. Dai 2010 The production of sexual and asexual larvae within single broods of the scleractinian coral, Pocillopora damicornis. Marine Biology 157: 351-359.
Yund P.O. 2000 How severe is sperm limitation in natural populations of marine free-spawners? Trends in Ecology and Evolution 15: 10-13.
92 Tables
Table 1: Loci-specific statistics for each of the six sampled microsatellite loci. n = Number of samples; N(A) = Number of Alleles; PI = Probability of Identity; n, N(A) and PI are based on the full dataset; Sex = Number of times the locus varied in sexual larvae; FIS = Inbreeding coefficient among unique colony genotypes; * = significant deviation from HWE (p < 0.05);
Locus n N(A) PI Sex FIS Pd1 585 2 0.38 5 -0.54 * Pd2 591 3 0.21 14 -0.42 * Pd4 558 3 0.20 12 -0.29 * Pd5 592 3 0.34 9 -0.36
Pd6 593 3 0.47 16 0.10
Pv6 545 6 0.15 11 -0.16 Total 596 20 3.9 x 10-4 67 -0.28
93 Table 2: List of maternal colonies
List of maternal colonies, including colony size, the total number of genotyped larvae per colony
(# Larvae), the number and the percentage of sexually produced larvae per colony (% and # Sex, respectively). Colonies with identical genotypes have the same grey background (P3 & P5; P6,
P8 & P9).
Parent Size [cm2] # Larvae # Sex % Sex P3 67 27 4 14.8% P5 53 23 1 4.3% P6 74 10 2 20.0% P8 139 55 0 0% P9 69 45 0 0% P10 163 165 4 2.4% P11 130 71 2 2.8% P12 72 56 1 1.8% P17 60 46 12 26.1% P18 25 41 6 14.6% P20 174 14 0 0% P23 n/a 10 3 30.0% P28 n/a 20 2 10.0% Total 60 583 37 6.3%
94 Table 3: Results of the generalized linear mixed effects model for the main logistic regression analysis
Significance was determined using a Likelihood Ratio Test (LRT) χ ** = p < 0.005 = highly significant effects; * = p< 0.05 = significant effect.
Predictor variable χ2 df p Cycle Day 8.36 1 0.004 ** Colony Size 4.74 1 0.029 * Calendar Day 4.21 1 0.040 * Colony Habitat 0.73 1 0.392 Colony Genotype 0.44 1 0.505 Lunar Day 0.03 1 0.873
95 Figures
Figure 1: A Pocillopora damicornis colony on the reef in front of the Gump Research
Station in Moorea, French Polynesia
The small insert shows a P. damicornis larvae on a mesh collection filter (100µm).
96
Figure 2: The proportion of sexually produced larvae decreases as a function of colony size
The proportion of sexually produced larvae decreases as a function of colony size (solid line).
The average proportions of sexually produced larvae per colony are shown as black diamonds
(), the average proportions per colony per cycle day are shown as white circles (o). Two different colony size classes are distinguished in this graph. Small colonies have a planar surface area under 75cm2 and a maximum diameter of less than 10cm (n=7). Large colonies have a surface area over 130cm2 (n=4) a minimum diameter of more than 10cm. The cross on the left
97 shows the average and the standard deviation for sexually produced larvae (and colony size) by small colonies. The cross on the right shows the significantly different results for big colonies
(single factor LRT: p = 0.02).
98 %
Sex 0% 25% 50% 75% 100% 0123456789 Cycle Day
Figure 3: The proportion of sexually produced larvae released decreases during the reproductive cycles of P. damicornis colonies
The proportion of sexually produced larvae released decreases during the reproductive cycles of
P. damicornis colonies (solid line). The average proportions of sexually produced larvae per cycle day are shown as black diamonds (w), the average proportions per cycle day per colony are shown as white circles (o). The cumulative percentage of sexual larvae across colonies is indicated by a dashed line (---).
99
100 CHAPTER 4
EVOLUTIONARY GENOMICS OF POCILLOPORA CORALS
Abstract
Background: Discrepancies between morphology-based taxonomy and phylogenetic systematics are common in corals. In the reef coral genus Pocillopora, nine recently described genetic lineages disagree fundamentally with the 16 currently recognized species. A coherent systematic classification, however, depends on an evolutionary understanding of the observed phenomenon.
Method: To clarify the phylogeographic relationships within this genus, we re-analyzed new and published sequence data from the rDNA marker ITS2 and the mitochondrial open reading frame
(ORF). We then used genome-wide Restriction-side Associated DNA sequencing (RAD-Seq) to conduct a phylogenomic comparison among three Tropical Eastern Pacific (TEP) Pocillopora species, three genetic lineages (types 1, 3 & 5) and two putative TEP hybrids.
Results/Conclusion: Our phylogenetic analyses indicated four distinct groups of genotypes.
Phylogenomic analyses revealed that genetic lineages from three of these groups are consistent with significant genome-wide differences, which suggests that they belong to separate species although they were all derived from colonies of the morphospecies Pocillopora damicornis.
Putative hybrids were composed of a mix of alleles from both parental lineages and allelic ratios indicate repeated uni-parental backcrossing and genome-wide introgression. However, a subset of loci distinguished P. damicornis colonies from colonies with other morphologies, across
101 lineages and across large geographic distances, which indicates a genetic component of species- specific colony morphology.
Introduction
The primary goal of taxonomy is to arrange taxa in hierarchical categories that reflect their systematic or evolutionary relationships. Evolutionary relationships are commonly described using phylogenetic data while taxonomic classifications are still largely based on morphology. Disagreements between morphology-based taxonomic descriptions and phylogenetic systematics can result from a variety of factors ranging from phenotypic plasticity to introgressive hybridization. Phylogenetic data often provides better resolution than morphological characters and can be used to identify cryptic species that are morphologically indistinguishable. Cryptic species may be the result of morphological stasis or unexpected convergent evolution of morphological traits (i.e. homoplasy). Incomplete lineage sorting and introgressive hybridization are the most common sources for misleading molecular phylogenies.
Disagreements between morphology- and genetics-based classifications are common in scleractinian corals (e.g. Fukami et al. 2008). Morphological characters traditionally used to delineate coral taxonomic groups have been shown to be relatively poor predictors of evolutionary relationships, even at the family level (Romano & Palumbi 1996; Fukami et al.
2004b; Fukami et al. 2008), and molecular data revealed cryptic species in many genera (Stobart
& Benzie 1994; Fukami et al. 2004a; Keshavmurthy et al. 2013). Commonly used genetic markers evolve slowly in corals, limiting their phylogenetic resolution (e.g. van Oppen et al.
1999; Shearer et al. 2002; Hellberg 2006), and genetic mixing via introgressive hybridization appears to be common in many coral genera (reviewed by Willis et al. 2006), which is
102 particularly problematic for analyses that rely on relatively few molecular markers, as is often the case for corals (reviewed by Fukami 2008),
Introgressive hybridization confounds species identities by passing genetic material from one taxon into the genome of another. Rates of introgression (i.e. genetic mixing) are not equal across the genomes of hybridizing species (e.g. Martinsen et al. 2001; Hohenlohe et al. 2013).
Selection can limit interspecific gene flow of genes involved in hybrid inviability and sterility while allowing introgression in other parts of the genome (e.g. Martinsen et al. 2001; Ting et al.
2001; Barbash et al. 2003; Wu & Ting 2004; Brideau et al. 2006). These “speciation” genes define species differences (i.e. genic view of speciation, Wu 2001) and can be used to study the evolutionary forces that shaped the species (e.g. Rieseberg et al. 1999; Presgraves et al. 2003;
Rieseberg et al. 2003). For example, regulatory elements that diverged under positive selection are frequently implicated to cause hybrid incompatibilities among closely related Drosophila species (e.g. Ting et al. 1998; Wu & Ting 2004; Brideau et al. 2006; Ortíz-Barrientos et al.
2007). In other cases, introgression can provide evolutionary novelty by passing potentially adaptive genetic variation between species (e.g. Grant & Grant 1994; Kim & Rieseberg 1999;
Mallet 2005; Whitney et al. 2010; Kunte et al. 2011; Hedrick 2013).
Conflicts between molecular phylogenetics and species descriptions based on morphology are common in the ecological important and widespread coral genus Pocillopora.
Lack of genetic monophyly in Pocillopora species was first described in the Tropical Eastern
Pacific (TEP) and attributed to introgressive hybridization (Combosch et al. 2008). Most TEP
Pocillopora species share the same rDNA and mtDNA lineages (type 1 after Pinzón et al.
103 2013)4. However, the most common TEP coral, P. damicornis, contains two distinct lineages, the shared type 1 and the private type 3, which in the TEP is exclusive to P. damicornis (Combosch et al. 2008; Pinzón & LaJeunesse 2011)5. In addition, several P. damicornis colonies contain alleles from both lineages, i.e. they are heterozygous for type 1 and 3 (1/3). This indicates a pattern of introgressive hybridization between P. damicornis and its TEP congenerics P. eydouxi and P. elegans that allows for inter-specific gene flow of type 1 alleles into P.’damicornis damicornis (Combosch et al. 2008). Subsequent broad-scale phylogenetic analyses of
Pocillopora sp. by Pinzón et al. (2013) indicate that widespread and well-defined Pocillopora morphospecies often contain multiple distinct genetic lineages that are shared by other morphospecies across broad swaths of their Indo-Pacific ranges. In particular the widely distributed genetic lineages type 1 and 3 are shared among several morphospecies while other types (2, 4, 5, 6, 7 and 8) seem to be mono-specific.
In order to assess the evolutionary history of the coral genus Pocillopora, we first re- analyze the phylogenetic relationships among Pocillopora lineages using new and published sequence data from nuclear internal transcribed spacer 2 (ITS) and the mitochondrial open reading frame (ORF). We then use genome-wide Restriction-site Associated DNA sequencing
(RAD-Seq) to conduct a phylogenomic comparison among three TEP Pocillopora species
(P.’damicornis, P. eydouxi and P. elegans), four trans-Pacific P. damicornis populations (from
Japan, French Polynesia and Panamá) and two types of putative TEP hybrids (sensu Combosch et al. 2008).
4 This is the same lineage that had been termed Type III by Combosch et al. (2008) and species A by Flot et al. (2010). To avoid further confusion, I am subsequently adopting the names of Pinzon et al. (2013), which includes most described lineages. 5 based on taxonomic re-assessments of Figure 5 in Pinzon & LaJeunesse (2011).
104 Material and Methods
Sample collection & DNA extraction
Coral nubbins were collected from Pocillopora damicornis, P. eydouxi and P. elegans in the Gulf of Chiriquí and the Gulf of Panamá (both Panamá) as well as from P.’damicornis populations in Okinawa (Japan) and the Tuamotus (French Polynesia). Samples were preserved in Guanidinium-Isothiocyanate (GITC) DNA buffer and genomic DNA was extracted using standard Phenol-Chloroform and Agencourt DNAdvance isolation kits (Beckman Coulter Inc.).
In total, more than 300 coral nubbins were collected and processed. ITS2 and ORF markers were sequenced in 231 and 165 specimens, respectively and 147 specimens were RAD sequenced in eight pooled samples (Table 1).
ITS & ORF genotyping and phylogenetics
The ribosomal internal transcribed spacer 2 (ITS) and a mitochondrial open reading frame (ORF) were PCR amplified and sequenced using published primers (Flot & Tillier 2007;
Combosch et al. 2008). Additional sequence data for major Pocillopora lineages (Pinzón et al.
2013) was obtained from GeneBank (accession numbers HQ378559–HQ378561, JX994072–
JX994088, HQ378552–HQ378557 and KC015015–KC015037). ORF outgroup sequences were obtained from GeneBank for the pocilloporiids Seriatopora hystrix (ORF: HQ878586; ITS:
AY722794) and Stylophora sp. 6 (ORF: JN558909, JN558908, JN558882; ITS: JN559111,
JN559109, JN559080). Sequences were aligned with MUSCLE as implemented in Geneious® and adjusted manually.
6 Stylophora lineages A, B and C (Flot et al. 2011)
105 ITS sequence data from 38 samples contained double peaks, indicating two distinct and heterozygote sequences. Heterozygote sequences were resolved manually using Clark’s method
(Clark 1990) and the program Indelligent (Dmitriev & Rakitov 2008) for length-variant heterozygotes. ITS sequence data were aligned as described above, but a complex microsatellite region in ITS was not alignable in a reliable, evolutionarily consistent way (Figure S4a) and between 15 and 34bp were therefore removed from each sequence.
Maximum Likelihood phylogenetic analyses were conducted with RAxML 7.4.4
(Stamatakis 2006; Stamatakis et al. 2008) on the CIPRES Web Portal (Miller et al. 2010). We used the GTR-GAMMA model, which includes a parameter for site heterogeneity among sites, and 1000 rapid bootstrap replicates to estimate clade confidence. Additional Maximum
Parsimony analyses were conducted with MEGA (Tamura et al. 2011), including 1000 Bootstrap replicates.
RAD-Seq
Genomic DNA quality and quantity was checked on agarose gels after calibration with a subset of samples with DNA 1000 chips on an Agilent BioAnalyzer 2100 (Agilent Technologies).
Approximately 50ng of genomic DNA from each specimen was pooled to yield 1µg of pooled genomic DNA per sample. The RAD-Seq pools were composed as described in Table 1. Briefly, samples Pdam_Ja and Pdam_Tu contained P. damicornis specimens from Japan and the
Tuamotus Islands in French Polynesia, respectively, with ITS type 5 genotypes. The samples
Pdam_GC and Pdam_GP contained P. damicornis specimens from populations in the Gulf of
Chiriquí and Gulf of Panamá, respectively, with ITS type 3 genotypes. The samples Peyd and
Pele contained P. eydouxi and P. elegans specimens from Panamá with ITS type 1 genotypes.
106 The sample Pdam_T1 contained P. damicornis specimens from Panamá with ITS type 1 genotypes and the sample Pdam_T1/3 contained P. damicornis specimens from Panamá with heterozygous ITS type 1 and type 3 genotypes.
RAD-Seq libraries were prepared according to modified protocols (Etter et al. 2012).
Genomic DNA was digested, using a high-fidelity Pst1 enzyme (New England Biolabs). The resulting fragments were ligated to P1 adapters that consist of the corresponding restriction-site overhang, and a sample-specific 4bp barcode sequence. Resulting DNA constructs were randomly fragmented to an average size of 500bp, using a Covaris S220 Ultrasonicator (Covaris
Inc.), and 300-500bp fragments were size-selected using agarose gel electrophoresis. DNA fragments were then ligated to a second P2 adapter, with the reverse amplification primer sequence and a divergent Y end that enables PCR amplification only if the corresponding P1 adapter is present within the same DNA fragment. PCR amplifications were carried out with
10µl 5x Phusion Buffer, 0.5µl dNTPs (25mM), 2.5µl of forward and reverse primer, 0.5µl
Phusion Polymerase, 5µl DNA template and water to 50µl (all NEB). The PCR profile consisted of 14-18 cycles at 98°C for 10s, 65°C for 30s and 72°C for 30s with an initial denaturation step at 98°C for 30s and a final extension step at 72°C for 5min. Resulting libraries were size-selected again (350-550bp) to remove remaining adapters and primers and verified electrophoretically on an Agilent BioAnalyzer 2100 using DNA High Sensitivity chips (Agilent Technologies) and fluorometrically using a Qubit Fluorometer (Invitrogen). Multiplexed samples were single-end sequenced (1x 50bp, 4 samples per lane) on an Illumina HiSeq 2000 platform (Illumina, Inc, San
Diego, California, USA) at the FAS Center for Systems Biology (Harvard University).
Quality filtering and loci assembly was conducted with the STACKS software pipeline
(version 0.999; Catchen et al. 2011; 2013). Samples were de-multiplexed and raw reads were
107 quality trimmed to remove low-quality reads (Phred score <10) and reads without complete barcode or restriction cut site. Barcode sequences were removed and remaining reads (47bp) were assembled into ‘stacks’, the RAD equivalent of alleles, if five or more identical reads were found within a sample. These invariable stacks were then compiled into sample-specific loci if they were less than 2 nucleotides (= 5%) different from each other. Sample-specific loci were then assembled into homologous loci if they were less than 3 nucleotides (= 7%) different between samples. A range of assembly parameters was tested against the P. damicornis transcriptome (-m: 3-20; -M: 1-3; -n: 1-3) and the final parameter setting (-m 3; -M 2; -n 3) resulted in the highest number and the highest proportion of loci aligning to unique locations within the TEP P. damicornis transcriptome (Table S2).
Maximum likelihood models, implemented in STACKS, were used to identify sequence polymorphisms and filter out sequencing errors. Three restrictive filters were applied to increase the proportion of unique genomic loci: 1) all loci with more than 1 SNP were removed from both datasets, 2) the STACKS deleveraging algorithm was used to identify and remove significantly highly repetitive stacks that likely represent sequencing errors and/or repetitive genomic regions, and 3) the STACKS populations program was used to filter out loci with invalid SNP calls. The remaining loci consist of 47bp of sequence data per locus, either with zero or one variable nucleotide position within and across samples.
RNA-Seq
A transcriptome was assembled from RNA-Seq data in order to identify the proportion of
RAD loci that were derived from the coral versus its algal symbionts by mapping the RAD loci against the transcriptome. Total RNA was extracted from P. damicornis samples collected at
108 Taboga Island (Panamá Province) and in Achotines Bay (Los Santos Province) in the
Panamanian Pacific. Coral tissues were preserved in TriReagent (Molecular Research Center,
Inc.), flash frozen in liquid nitrogen and stored at -80°C. Colonies were genotyped as described above and the transcriptomes of four hybrid colonies were sequenced, two P. damicornis specimen heterozygous for type 1 and type 3 and two specimens homozygous for type 1. Total
RNA was extracted following the manufacturer's protocol, RNA quality was assessed on an
Agilent Bioanalyzer 2100 and extractions with RIN values of 6 or higher were used for cDNA library preparations. RNA sequencing was performed using a modified RNA-Seq protocol (Libro et al., in press) and multiplexed samples were sequenced as described above (1x 50bp, 4 samples per lane, Illumina HiSeq 2000 at Harvard University). Barcoded samples were de-multiplexed and raw sequencing reads were quality trimmed to remove sequences with a Phred score of less than 30 and a read length of less than 15bp using custom Perl Scripts in the FASTX-Toolkit
(http://hannonlab.cshl.edu). De novo transcriptome assembly was carried out in Trinity
(Grabherr et al. 2011) using the combined reads of all 4 samples.
For each RAD locus, strict majority consensus sequences were generated and aligned to the transcriptome using Bowtie version 0.12.9 (Langmead et al. 2009) with up to one mismatch over the 47bp of RAD locus length. In addition, RAD loci consensus sequences were aligned to a publicly available P. damicornis transcriptome (Traylor-Knowles et al. 2011) and transcriptomes of two different Symbiodinium types that are commonly found in Pocillopora species (Ladner et al. 2012). By mapping RAD loci against these transcriptomes, we are able to identify transcribed symbiont and coral loci and estimate the overall proportion of coral versus symbiont loci.
Transcriptome-mapped RAD loci were also used to annotate protein-coding RAD loci.
Putative gene identities were identified using homology searches with BLASTx (Altschul et al.
109 1990) against the Swiss-Prot protein database. Blast matches with an e-value of less than 10-5 were deemed homologous and their GO terms and gene functions were obtained with Blast2Go
(Conesa et al. 2005).
RAD-Seq Analyses:
Assembled RAD loci were organized in two different datasets: The main dataset contained all loci that were recovered from at least two samples (referred to as the ‘P2’ dataset) and was used for all analyses, unless stated otherwise. A second dataset, containing loci found across all 8 samples (‘P8’ dataset) was used to calculate pairwise population genetic metrics (e.g. heterozygosity, private alleles, pairwise distances). For the phylogenomic analyses, heterozygous
SNPs were coded following IUPAC guidelines and a single consensus sequence was generated for each pooled sample by concatenating variable RAD loci. Phylogenomic analyses were conducted with two complementary methods for P2, P8 and two reduced P2/P8 subsets that contained only differently fixed loci (i.e. homozygous within samples but polymorphic between).
Maximum Likelihood analyses were conducted with RAxML 7.4.4 as described above.
Maximum parsimony (MP) analyses are well suited for phylogenetic analyses of SNP data (e.g.
Decker et al. 2009; Yoder et al. 2013). MP trees were identified with the program dnapars, as included in the Phylip package (Version 3.695; Felsenstein 1989, 2005). Parsimony tree searches were conducted with 111 randomized sample input orders (using the “jumble” option) and the default settings (‘More thorough search’; ‘Ordinary Parsimony’; ‘retaining 10,000 trees while searching’). Bootstrap support was assessed with 100 (P2) and 1000 (P8) bootstrap replicate datasets, respectively, that were generated with the program seqboot (Phylip 3.695).
110 Heterozygote loci, private alleles and pairwise comparisons were calculated using parsing scripts in R (R Core Team 2012). Pairwise genetic distances between samples (Table 3) were assessed in three different ways. The most conservative approach (P-distance) is based on the proportion of differentially fixed loci over all fixed loci in a specific pairwise comparison between samples (Nei & Kumar 2000). To include heterozygous loci, two approaches developed for microsatellite (Bowcock et al. 1994) and SNP data (Isolation-by-state, IbS) were tested. The
Bowcock method is based on the proportion of shared alleles in each pairwise comparison (e.g.
A/A vs. A/C = ½). The IbS method is based on the proportion of identical comparisons (i.e. A/A vs. A/C = ¾). Since both methods resulted in virtually identical pairwise distances, only the IbS results are presented here. Both P-distances and IbS-distances are reported as proportional differences (Table 3) and as the total number of differently fixed loci/different comparisons
(Table S3). In addition, a Principal Coordinate Analysis (PCoA) was conducted based on pairwise P-distances using GenAlEx (Version 6.4, Peakall & Smouse 2006). To assess genome- wide levels of selection among samples, pairwise distances between samples were calculated separately for transcribed and un-transcribed loci (Figure S2).
Since P. damicornis is known to share alleles with its Tropical Eastern Pacific congeners
P. eydouxi and P. elegans, patterns of allele sharing in the putative hybrid samples (Pdam_T1/3
& Pdam_T1) were analyzed at 4333 loci segregating between the type 1 samples P.’eydouxi and
P. elegans versus the type 3 sample P. damicornis from the Gulf of Chiriquí to determine the degree of genetic admixture in the two putative hybrid samples.
111 Results
ITS & ORF Phylogenies
Phylogenetic analyses of mitochondrial open reading frame (ORF) sequence data recovered four distinct genetic groups among the eight previously identified types (Figure 2).
The new outgroups indicated that type 3 sequences are basal and closely related to type 7. The remaining three genetic groups cluster types 1 and 8 sequences, types 2 and 6 sequences, and types 4 and 5 sequences. This phylogenetic order separates genetic types found only in larvae- brooding P. damicornis (types 4 & 5) from types that have been found in broadcast spawning
Pocillopora species (types 1 and 3). Types 2 and 3, which were only found in locally endemic
P.’effusus (type 2, from Clipperton & Galápagos) and P. ligulata (type 6, from Clipperton), seem to be closer related to the larvae-brooding P.’damicornis clade (types 4 & 5).
The ITS phylogeny was largely consistent with the ORF phylogeny (Figure 2) but the bootstrap support for the four major genetic groups was low. Type 1, 3 and 5 were clearly separated on the ITS tree and Type 1 sequences differed from most other Pocillopora lineages by seven fixed nucleotide substitutions. The two local endemics, types 2 and 6, shared some of these fixed differences, but type’1 genotypes were still separated by at least 3 unique fixed nucleotide changes (Figure S3).
RAD-Seq loci
In total, over 154 million high-quality RAD-Seq reads were obtained with an average of
12.4 million reads per pooled sample. STACKS assembled 452,367 RAD loci; 207,228 of which were shared between two or more samples and 33,710 (7.5%) occurred in all eight. Further quality filters left 112,644 RAD loci, from which 39,581 single nucleotide polymorphisms
112 (SNPs) were recovered from at least two samples (P2) with an average cover of 89 reads per locus. 15,511 loci with 6,769 SNPs were recovered from all 8 samples (P8) with an average locus cover of 133 reads (Table S1).
RAD phylogenomic analyses consistently returned three well-supported genetic clades, regardless of the applied phylogenetic method and analyzed data subset (Figure 3). The three clades matched the three major genetic groups identified with ITS and ORF and were labeled accordingly as clade 1, 3 and 5 based on their match to the existing ORF and ITS genetic types.
Clade 1 contained all four TEP samples with type 1 genotypes, i.e. the two morphospecies P. eydouxi (Peyd) and P. elegans (Pele) and both hybrid pools (Pdam_T1 & Pdam_T1/3). Clade 3 contained the TEP P. damicornis populations with type 3 genotypes (Pdam_GC & Pdam_GP).
Clade 5 contained the central-west Pacific P. damicornis populations from Japan (Pdam_Ja) and the Tuamotus (Pdam_Tu) with type 5 genotypes. Pocillopora damicornis populations in the central-west Pacific (clade 5) and in the TEP (clade 3) were clearly separated by significant phylogenetic distances with 100% bootstrap support (Figures 3 & S2). Samples within clade 1 were not well separated.
Pairwise comparisons supported the large genetic differences among the three major clades (Figure 4, Tables 2 & S3). For example, less than 5% of all fixed SNPs differed within clades while over 20% usually differed between clades. Both pairwise comparisons, however, indicated that P. damicornis from the Gulf of Panamá (Pdam_GP) shares more alleles with type’1 samples than does its P. damicornis counterpart from the Gulf of Chiriquí (Pdam_GC;
Figure’4). Pairwise comparisons indicated significant levels of population differentiation between type 5 populations over a large geographic distance (10,000km). TEP type 3 populations showed comparable levels of population structure over a much shorter geographic
113 distance (~400km), which is congruent with population genetic analyses of TEP P.’damicornis populations (Combosch & Vollmer 2011). However, the intermediate position of the Pdam_GP sample (Figure 4) indicates that pairwise comparisons were affected by its unexpectedly mixed composition, which cannot be accurately separated from the population structure. Pairwise comparisons showed little evidence for genetic differentiation within the type 1 genetic lineage and the mixed P.’damicornis sample. The morphospecies P. eydouxi and P. elegans differed by
12 fixed differences (P-distance: 0.003; the largest within this group) and the two hybrid pools did not differ by a single fixed locus. Remarkably, the heterozygous P.’damicornis sample
(Pdam_T1/3) was less different from all four type 3 and type 5 samples than any other type 1 sample (followed by Pdam_T1; Figure’4, Table 2).
Evidence for Introgressive Hybridization
The genetic make-up of the two putative P. damicornis hybrid pools – heterozygous ITS type 1 and type 3 P. damicornis (Pdam_T1/3) and homozygous type 1 P. damicornis (Pdam_T1)
- was analyzed in detail to detect signatures of genetic admixture at the 4,333 segregating loci between the parental lineages type 3 (Pdam_GC) and type 1 (Peyd & Pele; Table 3). Both P. damicornis hybrid types were dominated by type 1 alleles shared with P. eydouxi and P. elegans.
The P. damicornis hybrids that were heterozygous for ITS type 1 and 3 had 87% type 1 alleles and 13% type 3 alleles. Type 1 alleles were fixed at 78% of all segregating loci, both alleles were present at 18% and fixed type 3 alleles were found at 4% of segregating loci. The P.’damicornis hybrids that were homozygous at ITS for type 1 (the P.’eydouxi and P. elegans type) had 94% type 1 alleles and 6% type 3 alleles at segregating sites. They were fixed for type 1 alleles at
89% of all segregating loci, had both allele types at 10% and fixed type 3 alleles at 1% of these
114 loci. Remarkably, 21 out of the 34 loci that were fixed for type 3 alleles in hybrids with ITS type’1 (Pdam_T1) were also fixed for type 3 alleles in the other hybrid (62%). This indicates that these fixed loci are not a random sample of the 4,333 segregating loci but are selected.
Across all six TEP samples, 105 loci segregated between samples with P. damicornis morphologies and samples with P. eydouxi and/or P. elegans morphologies7. A trans-Pacific comparison of these loci with the P. damicornis samples from Japan and the Tuamotus identified
33 loci that were fixed for the P. damicornis allele (if present). The other 72 segregating loci were either absent (n = 56) or fixed for the Peyd/Pele alleles (n = 16); only 5% were mixed within or between samples (compared to 35% genome-wide). These loci are suitable candidates to further explore genetic differences between morpho-species, within and across genetic lineages.
RAD-Seq transcriptome matches
Approximately one quarter of all RAD P8 loci (3,831 out of 15,511) mapped to either our newly assembled transcriptome or the published P. damicornis transcriptome (Traylor-Knowles et al. 2011) and 10% of transcriptome contigs were directly sampled by a RAD locus. In contrast, less than 0.001% of all RAD loci (35/42 out of 112,644) mapped to the publicly available transcriptomes of Symbiodinium type C and type D (Ladner et al. 2012) that are commonly found in TEP and central-west Pacific Pocillopora species and lineages (e.g. Cunning et al. 2013). This indicates that the RAD-Seq data set is dominated by genetic variation in the coral host and not its algal symbionts.
7 Only loci are considered that were present in one type 3 sample (Pdam_GC and/or Pdam_GP), one hybrid sample (Pdam_T1 and/or Pdam_T1/3) and P. eydouxi or P. elegans.
115
Discussion
The goal of this study was to describe the evolutionary relationships in the coral genus
Pocillopora by re-examining published and new sequence data and to analyze genome-wide allelic patterns among common morphospecies (P. damicornis, P. eydouxi & P. elegans) and their putative TEP hybrids using RAD-Seq. We confirmed the presence of several distinct genetic groups in Pocillopora using mitochondrial ORF sequence data and re-rooted the phylogenetic tree using new outgroups. We found that mtDNA and rDNA-based genetic types are consistent with genome-wide differences in the RAD-Seq data, which suggests that some of the described genetic lineages correspond to different species, despite their incongruence with taxonomically recognized morphospecies. Allelic comparisons between putative hybrid and pure-breed samples confirmed introgressive hybridization between genetic lineages in Panamá
(Combosch et al. 2008), but identified a subset of species-specific RAD loci that were shared among all P. damicornis specimens regardless of their introgression history and their geographic origin.
Phylogenetic analyzes of nuclear rDNA (ITS), mitochondrial ORF, and now genome- wide RAD-Seq data demonstrate that well-recognized Pocillopora morphospecies are not monophyletic, but instead are comprised of multiple distinct genetic lineages that are shared with other morphospecies. This disagreement between morphospecies and genetics lineages in
Pocillopora corals was first reported as multiple shared ITS genotypes in Tropical Eastern
Pacific (TEP) species due to introgressive hybridization (Combosch et al. 2008). Pocillopora species with multiple genotypes have subsequently been reported from elsewhere in the TEP
(Flot et al. 2010; Pinzón & LaJeunesse 2011), East Africa (Souter 2010) and the Great Barrier
Reef (Schmidt-Roach et al. 2012). A recent Indo-Pacific wide phylogeographic study identified
116 nine different genetic types, including the three types that were used in the present study (types
1, 3 & 5 sensu Pinzón et al. 2013). The disagreement between morphology and phylogenetics has caused some to argue for a lineage-based description of Pocillopora species by genetic type
(e.g. type 1 versus type 3; Pinzón et al. 2013). However, this does little to resolve the conflict since these genetic types cannot be distinguished visually in the field and because typing does not reconcile how or why the disagreement between morphology and genetics exists.
Phylogenetic analyses
Our re-analyses of published ORF data with the inclusion of new data from 231
P.’damicornis samples from Japan (n = 17), French Polynesia (23) and Panamá (191) and a focus on outgroups to root the tree identified three main genetic groups in Pocillopora. The basal
Pocillopora type 3 lineage is most genetically diverse (11 ORF genotypes), speciose
(7’morphospecies) and geographically widespread (TEP to the Red Sea; Figure 2; Pinzón et al.
2013). Even though type 3 contains multiple species, they all appear to be broadcast spawning.
The center of type 3 diversity is in the Indian Ocean and the most basal genotypes, 3e and 3h, were found only in the western Indian Ocean and the Red Sea (Pinzón et al. 2013). This might indicate an Indian Ocean origin of the genus, like its closest sister genus Stylophora (Flot et al.
2011; Keshavmurthy et al. 2013). The other two Pocillopora lineages that we sequenced with
RAD-Seq (types 1 and 5) were significantly different from the type 3 lineage (Figure 2). Type 1 genotypes are geographically widespread, have been found across multiple morphospecies and appear to be broadcast spawners (like type 3). Type 5 genotypes have so far only been found in
P. damicornis colonies in the Indo-West Pacific and are notably absent from the TEP. Type 5 and type 4 lineages include the brooding Pocillopora corals, for example the predominantly
117 asexually brooding P. damicornis (Combosch & Vollmer 2013; Torda et al. 2013). All lineages that are closely related to type 5 (i.e. types 2, 4 and 6) appear to be restricted to single morphospecies (P. effusus, P. damicornis & P. ligulata, respectively).
Even though the three lineages are not monophyletic with respect to the current species classification scheme in Pocillopora, the newly rooted tree indicates that reproductive strategy may be a key character in the group with the basal genetic groups containing types 3 being broadcast spawners and all brooding corals containing type 4 or type 5 genotypes. Brooding and broadcast spawning were thought to be plastic within Pocillopora species, while the phylogeny suggest they are a key evolutionary trait, with brooding probably originating only once within this predominantly broadcast-spawning genus. This result disagrees with the published phylogeny by Pinzon et al. (2013) that suggested that brooding was ancestral to broadcast spawning because of their rooting scheme.
Our RAD-based phylogenomic analyses included almost 40,000 variable SNPs from three Pocillopora morphospecies and three genetic lineages (types 1, 3 and 5). The resulting
RAD phylogeny (Figure 3) is consistent with the phylogenetic patterns observed with rDNA ITS and mitochondrial ORF trees (Figure 2). The two sympatric, but genetically distinct types of TEP
P. damicornis, described based on ITS (Combosch et al. 2008), ORF and microsatellite loci
(Pinzon & LaJeunesse 2011), were confirmed by the RAD-Seq data. Genome-wide genetic differences between TEP P. damicornis types far exceeded levels of population differentiation within types, even over large biogeographic distances (e.g. over 10,000km between Japan and the Tuamotus; Table 2). Pocillopora eydouxi and P. elegans from the TEP belonged to the same
RAD-Seq clade, shared the same ITS2 and ORF haplotypes (Combosch et al. 2008) and were only 0.3% different based on pairwise comparisons (P-distance, Table 2), indicating that they are
118 closely related and or possibly interbreeding. The RAD-Seq phylogeny further showed that
P.’damicornis from the central-west Pacific, which had previously been described as a distinct type based on ITS sequence data (Type I in Combosch et al. 2008), indeed belongs to a different genomic lineage.
Hybridization in the TEP
Introgressive hybridization between Tropical Eastern Pacific (TEP) Pocillopora had previously been inferred based on the presence of heterozygous ITS type 1/type 3 genotypes in
Panamanian P. damicornis specimens (Combosch et al. 2008), based on admixtures signatures in multi-locus microsatellite data (Pinzón & LaJeunesse 2011; Pinzón et al. 2013) and due to mismatching ITS type 3 and ORF type 1 genotypes (Combosch, unpublished results; Pinzón et al. 2013)8. RAD-Seq phylogenomic analyzes confirm this pattern of admixture across the genomes of these species in the TEP. In the RAD-Seq data, P. damicornis with mixed type 1 and
3 ITS lineages (Pdam_T1/3) was genetically much closer to the type 1 lineages of their TEP congeners P. elegans and P. eydouxi than to the pure P. damicornis type 3 sample (Figure 4).
This is expected if specimens heterozygous for ITS type 1/3 are the result of hybridization between types 1 and 3, but cannot be explained as a result of ITS incomplete lineage sorting in type 3. The heterozygous P.’damicornis type 1 and 3 sample (Pdam_T1/3) contained a mix of type 1 (87%) and type 3 alleles (13%) in both homozygous and heterozygous states (Table 3).
This suggests that these heterozygous Pdam_T1/3 specimens were at least a few generations removed from their most recent type 3 ancestors (Figure 5). Hybridization and backcrossing
8 Pinzon et al. (2013) label these ITS types 1D and 1F but they differ by 3 fixed SNPs from type 1 while they are only 1 SNP different from type 3.
119 seem limited to the type 1 maternal lineage, i.e. type 1 eggs9, because virtually all heterozygotes had mitochondrial type 1 genotypes. Since these colonies have the morphology of P.’damicornis, colony morphology is likely dictated by the paternal lineage even though their genomes are largely the other type. The allele composition of the Pdam_T1 sample (94% type 1 vs. 6% type
3) indicates that P. damicornis specimens homozygous for ITS type 1 are further removed from their most recent type 3 ancestors (Figure 5), as indicated by their ITS genotype.
The dominant species of TEP coral reefs, P. damicornis, is the central component of the local hybridization system. It is the only species in the TEP that frequently contains both genetic lineages and its TEP population is the only one that contains genotypes heterozygous for ITS type 1 and 3. Based on population genetic surveys in Panamá, 14% of P. damicornis specimens are heterozygous for ITS type 1 and 3 (n = 26 out of 191 genotyped specimens), 46% are homozygous for type 1 and 40% are homozygous for type 3. Thus, so far, we have sampled more specimens with a hybrid history than pure breed P. damicornis (114 vs. 77). Despite population genetic evidence for significant regional differentiation of P. damicornis populations between the upwelling Gulf of Chiriquí and the non-upwelling Gulf of Panamá (Combosch & Vollmer 2011), we detected no significant difference in the distribution of hybrids versus pure-breeds. This might indicate a mosaic hybrid zone (sensu Rand & Harrison 1989) between type 1 and type 3 specimens with a patchy distribution of hybrid and pure breed P. damicornis in Panamá.
Despite extensive genotyping (>500 genotyped samples) elsewhere in the TEP, not a single ITS type 3 genotype has so far been found in Mexico (Pinzón & LaJeunesse 2011). Since type 3 genotypes are widespread throughout the Indo-Pacific, selection likely accounts for the absence of type 3 genotypes in Mexico. Since none of the surveyed Mexican reefs is located in
9 Reports of type 3 genotypes in P. capitata and P. elegans are likely due to taxonomic mis- identification (e.g. in Figure 5b Pinzon & LaJeunesse 2011) or indicate backcrossing.
120 an upwelling zone, the more heterogeneous environment in Panamá may benefit the sympatric coexistence of both types. Moreover, differential thermal adaptations of local populations in two gulfs have been shown experimentally (D'Croz & Mate 2004) and can generate adaptation clines that can stabilize or trap hybrid zones locally (Bierne et al. 2011), which might contribute to the occurrence of local hybridization among Pocillopora types in Panamá. Alternatively, type 3 genotypes might have arrived in Panamá only ‘recently’, e.g. with human facilitation or during a strong El-Niño event (Glynn & Ault 2000), or type 1 is replacing type 3 specimens in Panamá just like it might have in Mexico.
In the TEP, 105 loci segregated between samples with P. damicornis morphologies
(Pdam_GC, Pdam_GP, Pdam_T1 and/or Pdam_T1/3) and samples with P. eydouxi and/or
P.’elegans morphologies. Preliminary analyses against our unpublished transcriptome indicate that 10 out of these 105 loci map against contigs in the transcriptome and three contigs have well-defined protein annotations (e-value 10-5). All three corresponding proteins are involved in gene regulation, two are involved in transcription regulation, and one is a RNA-binding protein.
Regulatory genes are prime candidates for hybrid sterility and inviability, i.e. speciation genes
(e.g. Ortíz-Barrientos et al. 2007) and several speciation genes in other taxa were found to have regulatory functions (e.g. Ods & Hmr; Ting et al. 1998; Brideau et al. 2006). Other segregating
RAD loci are possibly linked to additional candidate speciation genes as indicated by their low heterozygosity (5% compared to 23% genome-wide) and high fixation within lineages (90%). In particular the 33 loci that segregate for P.’damicornis samples across all three genetic lineages and over 20,000 km of open ocean are suitable candidates for further investigations.
121 Hybridization outside the TEP
Indications for hybridization among Pocillopora species and genetic lineages have been reported from many other locations throughout the Indo-Pacific (e.g. Miller & Ayre 2004; Souter
2010; Schmidt-Roach et al. 2013). For example, Pinzon et al. (2013) identified admixture between types (presumably 3 and 5) in Hawaii using multi-locus microsatellite data and found one specimen with a heterozygous mitochondrial ORF type 5 and 6. This is interesting because it would involve hybridization among presumed broadcasting and brooding Pocillopora lineages.
In addition, we re-analyzed published sequence data from Hawaii (Flot et al. 2008) and found heterozygote ITS type 3 and 5 genotypes in 9 out of 16 P. damicornis specimen, mismatching
ITS type 3 and ORF type 5 genotypes in two additional P. damicornis specimen and one heterozygote ITS type 3 and 5 with ORF type 4, i.e. three distinct genetic lineages. Moreover, mismatching ORF type 1 with ITS type 3 (mislabeled as 1E akin type 1D and 1F discussed above) were detected in 7 P. meandrina and one P. eydouxi (Flot et al. 2008). These mixed typed individuals suggest that introgressive hybridization is more common among Pocillopora types than previously thought.
Colony morphology, phylogenetics and the evolution of Pocillopora corals
The main result of this study, that rDNA- and mtDNA-based genetic types are consistent with genome-wide RAD-Seq lineages, strongly supports the hypothesis that some of the described genetic types correspond to interbreeding and reproductively isolated units, i.e. biological species (sensu Mayr 1963). Additional morphological, ecological and reproductive studies will now be necessary to identify and delimitate valid Pocillopora species within and among these distinct genetic lineages. Importantly, the analyzed types 1, 3 and 5 correspond to
122 the three main Pocillopora lineages (Figure 2). Most other Pocillopora types are in fact poorly delimitated and either share the same ITS type (3 and 7), share both ITS and ORF types (1 & 9), are based on less than a handful of samples (type 8) or can simultaneously be shared with two other types (type 4).
If biological species in Pocillopora are defined solely by genetic lineages, it is unexpected that species share multiple distinct colony morphologies, often all across their distribution range. Colony morphologies can be shared because they are ancestral (i.e. reticulate evolution), plastic (i.e. taxonomy is based on unsuitable characters), as a result of independent convergent evolution or because the underlying morphology genes have been exchanged via introgressive hybridization. Highly variable, plastic morphologies are widespread among
Pocillopora corals and have lead to numerous new species descriptions (e.g. Pichon & Veron
1976; Figure 1b). However, it is unlikely that phenotypic plasticity alone accounts for the full range of discrepancies observed here, including eight distinct, recurrent colony morphologies in two distinct lineages. First of all, studies have shown consistent links between Pocillopora morphospecies and specific genotypes (e.g. Schmidt-Roach et al. 2013). In addition, phenotypic plasticity depends on environmental differences to drive morphologic variation (West-Eberhard
1989) and many of these morphospecies occur sympatrically together on reefs. Within the homogeneous TEP P. damicornis carpets (Figure 1a), it is hard to imagine that environmental differences could account for the stark morphological differences between morphospecies.
Divergent positive selection can generate and/or maintain genetically determined recurrent colony morphologies within distinct genetic lineages via convergent evolution, morphological stasis or introgressive hybridization. Convergent morphological evolution could be present to some extent, but it seems very unlikely that it can account for all eight observed
123 morphospecies shared between type 1 and type 3. The exchange of morphology-driving genes via introgressive hybridization cannot fully account for the observed number of shared morphologies either because the number of distinct morphologies (at least 8) is much bigger then the number of genetic lineages that share them (2; types 1 & 3). The presence of shared ancestral morphologies is somewhat supported by the observation that the phylogenetically basal types 1 and 3 shared most morphospecies while more derived lineages are monomorphic (e.g. types 2, 4,
5 & 6), which might indicate a loss of ancestral morphologies, e.g. due to founder effects.
Importantly, none of these different scenarios are mutually exclusive and some of them are in fact more likely to occur together than separately, for example due to mutually favorable conditions (e.g. strong selection). It is therefore most likely that a combination of multiple factors generates the observed discrepancies in Pocillopora.
In the TEP, we have shown that in some cases disagreements can be explained by introgression hybridization. It is unlikely, however, that introgressive hybridization explains all disagreements between morphological taxonomy and phylogenetic lineages through the Indo-
Pacific. It is a first step, however, and we have shown that the isolated Tropical Eastern Pacific with its reduced complexity of Pocillopora corals is a good place to continue disentangling this evolutionary puzzle. Alternatively, Hawaii potentially offers the unique opportunity to study the evolution and interaction of both major coral reproductive strategies. Since all Hawaiian
P.’damicornis colonies that were heterozygous for ITS type 3 and 5 contained either ORF type 4 or 5, broadcast-spawned sperm from type’3 might be capable of fertilizing the eggs of larvae- brooding type 4 and 5 colonies internally. This type of hybrid fertilization seems to be unique to
Hawaii and would allow the identification of genes involved in different coral reproductive strategies and possibly in the origin of some of these different Pocillopora lineages.
124
Significant discrepancies between morphology-based taxonomy and phylogenetic systematics are common in pocilloporiid genera and have been described in Stylophora (Flot et al. 2011; Stefani et al. 2011), Seriatopora (Flot et al. 2008) and Madracis (Frade et al. 2010).
Different evolutionary scenarios and combinations of mechanisms have been proposed for each genus. For example, while introgressive hybridization has been documented in Madracis (Frade et al. 2010) and Pocillopora (this study), it has been ruled out in Stylophora based on the congruence of nuclear and mitochondrial markers (Flot et al. 2011). However, studies in all four genera have benefitted from common transferable technical advances. For example the description of the mitochondrial genome in Pocillopora (Flot & Tillier 2007) enabled the development of mitochondrial sequence markers in Seriatopora (Flot et al. 2008) and Stylophora
(Flot et al. 2011). We hope that his study testifies to the usefulness and power of genome-wide
RAD sequencing to disentangle the complex evolutionary history of corals in general and pocilloporiids in particular.
Conclusions
In summary, we conclude that genome-wide differences between genetically defined lineages likely correspond to interbreeding, reproductively isolated units, and hence different species. To define, delimit and recognize these new species, additional genetic, morphological, ecological, and reproductive studies will be necessary (e.g. Combosch & Vollmer 2013; Torda et al. 2013). In addition, we strongly support the re-analysis of the morphology-based taxonomic descriptions of corals in general and Pocillopora in particular in light of recent genetic findings.
Several recent publications have made significant progress (e.g. Stefani et al. 2008; Benzoni et al.
125 2010; Arrigoni et al. 2012; Schmidt-Roach et al. 2013). However, coral phylogeneticists must provide taxonomists with sound evolutionary explanations in addition to their list of cryptic genetic types. The evolution of many coral taxa is certainly complex and multi-facetted and genomic approaches might be necessary to resolve their evolutionary histories.
Acknowledgements
Support for fieldwork, laboratory work, bioinformatics and genetic analyses was provided by Sarah Lemer, Rich Aronson, Pete Edmunds, Rosa Fernández, Sal Genovese,
Vanessa Gonzales, Gonzalo Giribet, Stefan Kaluziak, Ian McIntyre, Alicia Pérez-Porro, Carmen
Schloeder and Prashant Sharma.
126 References
Altschul S.F., W. Gish, W. Miller, E.W. Myers & D.J. Lipman 1990 Basic Local Alignment Search Tool. Journal of Molecular Biology 215: 403-410.
Arrigoni R., F. Stefani, M. Pichon, P. Galli & F. Benzoni 2012 Molecular phylogeny of the Robust clade (Faviidae, Mussidae, Merulinidae, and Pectiniidae): An Indian Ocean perspective. Molecular Phylogenetics and Evolution 65: 1-11.
Barbash D.A., D.F. Siino, A.M. Tarone & J. Roote 2003 A rapidly evolving MYB-related protein causes species isolation in Drosophila. Proceedings of the National Academy of Sciences 100: 5302-5307.
Benzoni F., F. Stefani, M. Pichon & P. Galli 2010 The name game: morpho-molecular species boundaries in the genus Psammocora (Cnidaria, Scleractinia). Zoological Journal of the Linnean Society 160: 421-456.
Bierne N., J. Welch, E. Loire, F. Bonhomme & P. David 2011 The coupling hypothesis: why genome scans may fail to map local adaptation genes. Molecular Ecology 20: 2044–2072.
Bowcock A., A. Ruiz-Linares, J. Tomfohrde, et al. 1994 High resolution of human evolutionary trees with polymorphic microsatellites. Nature 368: 455-457.
Brideau N.J., H.A. Flores, J. Wang, et al. 2006 Two Dobzhansky-Muller Genes Interact to Cause Hybrid Lethality in Drosophila. Science 314: 1292-1295.
Catchen J.M., A. Amores, P. Hohenlohe, W. Cresko & J.H. Postlethwait 2011 Stacks: building and genotyping loci de novo from short-read sequences. G3: Genes, Genomes, Genetics 1: 171-182.
Catchen J.M., P.A. Hohenlohe, S. Bassham, A. Amores & W.A. Cresko 2013 Stacks: an analysis tool set for population genomics. Molecular Ecology 22: 3124-3140.
Clark A. 1990 Inference of haplotypes from PCR-amplified samples of diploid populations. Molecular Biology and Evolution 7: 111-122.
Combosch D., H. Guzman, H. Schuhmacher & S. Vollmer 2008 Interspecific hybridization and restricted trans-Pacific gene flow in the Tropical Eastern Pacific Pocillopora. Molecular Ecology 17: 1304-1312.
Combosch D.J. & S.V. Vollmer 2011 Population Genetics of an Ecosystem-Defining Reef Coral Pocillopora damicornis in the Tropical Eastern Pacific. PLoS One 6: e21200.
Combosch D.J. & S.V. Vollmer 2013 Mixed asexual and sexual reproduction in the Indo-Pacific reef coral Pocillopora damicornis. Ecology & Evolution 3: 3379–3387.
127 Conesa A., S. Götz, J.M. García-Gómez, et al. 2005 Blast2GO: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics 21: 3674- 3676.
Cunning R., P.W. Glynn & A.C. Baker 2013 Flexible associations between Pocillopora corals and Symbiodinium limit utility of symbiosis ecology in defining species. Coral Reefs 32: 795-801.
D'Croz L. & J. Mate 2004 Experimental responses to elevated water temperature in genotypes of the reef coral Pocillopora damicornis from upwelling and non-upwelling environments in Panamá. Coral Reefs 23: 473-483.
Decker J.E., J.C. Pires, G.C. Conant, et al. 2009 Resolving the evolution of extant and extinct ruminants with high-throughput phylogenomics. Proceedings of the National Academy of Sciences 106: 18644-1856.
Dmitriev D.A. & R.A. Rakitov 2008 Decoding of Superimposed Traces Produced by Direct Sequencing of Heterozygous Indels. PLoS Computational Biology 4: e1000113.
Etter P.D., S. Bassham, P.A. Hohenlohe, E.A. Johnson & W.A. Cresko 2012 SNP discovery and genotyping for evolutionary genetics using RAD sequencing. Molecular Methods for Evolutionary Genetics 772: 157-178.
Felsenstein J. 1989 PHYLIP 3.2 manual. University of California Herbarium, Berkeley, California.
Felsenstein J. 2005 Accuracy of Coalescent Likelihood Estimates: Do We Need More Sites, More Sequences, or More Loci? Molecular Biology and Evolution 23: 691-700.
Flot J.-F. & S. Tillier 2007 The mitochondrial genome of Pocillopora (Cnidaria: Scleractinia) contains two variable regions: The putative D-loop and a novel ORF of unknown function. Gene 401: 80-87.
Flot J.-F., H. Magalon, C. Cruaud, A. Couloux & S. Tillier 2008 Patterns of genetic structure among Hawaiian corals of the genus Pocillopora yield clusters of individuals that are compatible with morphology. Comptes Rendus Biologies 331: 239-247.
Flot J.-F., J. Blanchot, L. Charpy, et al. 2011 Incongruence between morphotypes and genetically delimited species in the coral genus Stylophora: phenotypic plasticity, morphological convergence, morphological stasis or interspecific hybridization? BMC Ecology 11: 22-40.
Flot J.-F., A. Couloux & S. Tillier 2010 Haplowebs as a graphical tool for delimiting species: a revival of Doyle's "field for recombination" approach and its application to the coral genus Pocillopora in Clipperton. BMC Evolutionary Biology 10: 372.
128 Frade P.R., M.C. Reyes-Nivia, J. Faria, et al. 2010 Semi-permeable species boundaries in the coral genus Madracis: introgression in a brooding coral system. Molecular Phylogenetics and Evolution 57: 1072-1090.
Fukami H. 2008 Short review: molecular phylogenetic analyses of reef corals. Galaxea 10: 47- 55.
Fukami H., A. Budd, D. Levitan, et al. 2004a Geographic differences in species boundaries among members of the Montastraea annularis complex based on molecular and morphological markers. Evolution 58: 324-337.
Fukami H., A. Budd, G. Paulay, et al. 2004b Conventional taxonomy obscures deep divergence between Pacific and Atlantic corals. Nature 427: 832-835.
Fukami H., C. Chen, A. Budd, et al. 2008 Mitochondrial and nuclear genes suggest that stony corals are monophyletic but most families of stony corals are not (Order Scleractinia, Class Anthozoa, Phylum Cnidaria). PLoS One 3: e3222.
Glynn P. & J. Ault 2000 A biogeographic analysis and review of the far eastern Pacific coral reef region. Coral Reefs 19: 1-23.
Grabherr M.G., B.J. Haas, M. Yassour, et al. 2011 Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nature Biotechnology 29: 644-652.
Grant P.R. & B.R. Grant 1994 Phenotypic and Genetic Effects of Hybridization in Darwin's Finches. Evolution 48: 297-316.
Hedrick P.W. 2013 Adaptive introgression in animals: examples and comparison to new mutation and standing variation as sources of adaptive variation. Molecular Ecology 22: 4606–4618.
Hellberg M.E. 2006 No variation and low synonymous substitution rates in coral mtDNA despite high nuclear variation. BMC Evolutionary Biology 6: 24.
Hohenlohe P.A., M.D. Day, S.J. Amish, et al. 2013 Genomic patterns of introgression in rainbow and westslope cutthroat trout illuminated by overlapping paired-end RAD sequencing. Molecular Ecology 22: 3002–3013.
Keshavmurthy S., S.-Y. Yang, A. Alamaru, et al. 2013 DNA barcoding reveals the coral “laboratory-rat”, Stylophora pistillata encompasses multiple identities. Scientific Reports 3: 1-7.
Kim S.-C. & L. Rieseberg 1999 Genetic architecture of species differences in annual sunflowers: implications for adaptive trait introgression. Genetics 153: 965-977.
Kunte K., C. Shea, M.L. Aardema, et al. 2011 Sex Chromosome Mosaicism and Hybrid Speciation among Tiger Swallowtail Butterflies. PLoS Genet 7: e1002274.
129 Ladner J.T., D.J. Barshis & S.R. Palumbi 2012 Protein evolution in two co-occurring types of Symbiodinium: an exploration into the genetic basis of thermal tolerance in Symbiodinium clade D. BMC Evolutionary Biology 12: 1-1.
Langmead B., C. Trapnell, M. Pop & S.L. Salzberg 2009 Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biology 10: R25.
Mallet J. 2005 Hybridization as an invasion of the genome. Trends in Ecology & Evolution 20: 229-237.
Martinsen G.D., T.G. Whitham, R.J. Turek & P. Keim 2001 Hybrid populations selectively filter gene introgression between species. Evolution 55: 1325-1335.
Mayr E. 1963 Animal Species and Evolution Belknap Press, Cambridge.
Miller K. & D. Ayre 2004 The role of sexual and asexual reproduction in structuring high latitude populations of the reef coral Pocillopora damicornis. Heredity 92: 557-568.
Miller M.A., W. Pfeiffer & T. Schwartz 2010 Creating the CIPRES Science Gateway for inference of large phylogenetic trees 5: 1-8.
Nei M. & S. Kumar 2000 Molecular evolution and phylogenetics Oxford University Press, USA.
Ortíz-Barrientos D., B.A. Counterman & M.A.F. Noor 2007 Gene expression divergence and the origin of hybrid dysfunctions. Genetica 129: 71-81.
Peakall R. & P.E. Smouse 2006 GENALEX 6: Genetic analysis in Excel. Population genetic software for teaching and research. Molecular Ecology Notes 6: 288-295.
Pichon M. & J. Veron 1976 Scleractinia of eastern Australia part I Families Thamnasteriidae, Astrocoeniidae, Pocilloporidae Australian Government Publication Service.
Pinzon J.H. & T. LaJeunesse 2011 Species delimitation of common reef corals in the genus Pocillopora using nucleotide sequence phylogenies, population genetics and symbiosis ecology. Molecular Ecology 20: 311-325.
Pinzon J.H., E. Sampayo, E. Cox, et al. 2013 Blind to morphology: genetics identifies several widespread ecologically common species and few endemics among Indo-Pacific cauliflower corals (Pocillopora, Scleractinia). Journal of Biogeography 40: 1595–1608.
Presgraves D.C., L. Balagopalan, S.M. Abmayr & H.A. Orr 2003 Adaptive evolution drives divergence of a hybrid inviability gene between two species of Drosophila. Nature 423: 715-719.
R Core Team 2012 R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.
130 Rand D. & R. Harrison 1989 Ecological genetics of a mosaic hybrid zone: mitochondrial, nuclear, and reproductive differentiation of crickets by soil type. Evolution 43: 432-449.
Rieseberg L.H., O. Raymond, D.M. Rosenthal, et al. 2003 Major ecological transitions in wild sunflowers facilitated by hybridization. Science 301: 1211-1216.
Rieseberg L.H., J. Whitton & K. Gardner 1999 Hybrid zones and the genetic architecture of a barrier to gene flow between two sunflower species. Genetics 152: 713-727.
Romano S. & S. Palumbi 1996 Evolution of scleractinian corals inferred from molecular systematics. Science 271: 640.
Schmidt-Roach S., P. Lundgren, K.J. Miller, et al. 2013 Assessing hidden species diversity in the coral Pocillopora damicornis from Eastern Australia. Coral Reefs 32: 161-172.
Schmidt-Roach S., K.J. Miller, E. Woolsey, G. Gerlach & A.H. Baird 2012 Broadcast Spawning by Pocillopora Species on the Great Barrier Reef. PLoS One 7: e50847.
Shearer T.L., M.J.H. van Oppen, S.L. Romano & G. Wörheide 2002 Slow mitochondrial DNA sequence evolution in the Anthozoa (Cnidaria). Molecular Ecology 11: 2475-2487.
Souter P. 2010 Hidden genetic diversity in a key model species of coral. Marine Biology 157: 875-885.
Stamatakis A. 2006 RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22: 2688-2690.
Stamatakis A., P. Hoover & J. Rougemont 2008 A Rapid Bootstrap Algorithm for the RAxML Web Servers. Systematic Biology 57: 758-771.
Stefani F., F. Benzoni, M. Pichon, G. Mitta & P. Galli 2008 Genetic and morphometric evidence for unresolved species boundaries in the coral genus Psammocora (Cnidaria; Scleractinia). Hydrobiologia 596: 153-172.
Stefani F., F. Benzoni, S.-Y. Yang, et al. 2011 Comparison of morphological and genetic analyses reveals cryptic divergence and morphological plasticity in Stylophora (Cnidaria, Scleractinia). Coral Reefs 30: 1033-1049.
Stobart B. & J.A.H. Benzie 1994 Allozyme Electrophoresis demonstrates that the Scleractinian Coral Montipora digitata is two Species. Marine Biology 118: 183-190.
Tamura K., D. Peterson, N. Peterson, et al. 2011 MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Molecular Biology and Evolution 28: 2731-2739.
Ting C.T., A. Takahashi & C.I. Wu 2001 Incipient speciation by sexual isolation in Drosophila: Concurrent evolution at multiple loci. Proceedings of the National Academy of Sciences 98: 6709-6713.
131 Ting C.T., S.C. Tsaur, M.L. Wu & C.I. Wu 1998 A rapidly evolving homeobox at the site of a hybrid sterility gene. Science 282: 1501-1504.
Torda G., P. Lundgren, B.L. Willis & M.J.H. van Oppen 2013 Genetic assignment of recruits reveals short- and long-distance larval dispersal in Pocillopora damicornis on the Great Barrier Reef. Molecular Ecology: doi: 10.1111/mec.12539.
Traylor-Knowles N., B.R. Granger, T. Lubinski, et al. 2011 Production of a reference transcriptome and a transcriptomic database (Pocillopora-Base) for the cauliflower coral, Pocillopora damicornis. BMC Genomics 12: 585. van Oppen M.J., B.L. Willis & D.J. Miller 1999 Atypically low rate of cytochrome b evolution in the scleractinian coral genus Acropora. Proceedings of the Royal Society Series B 266: 179-183.
West-Eberhard M. 1989 Phenotypic plasticity and the origins of diversity. Annual Review of Ecology and Systematics 20: 249-278.
Whitney K.D., E.J. Baack, J.L. Hamrick, et al. 2010 A role for nonadaptive processes in plant genome size evolution? Evolution 64: 2097-2109.
Willis B., M. van Oppen, D. Miller, S. Vollmer & D. Ayre 2006 The role of hybridization in the evolution of reef corals. Annual Review of Ecology Evolution and Systematics 37: 489- 517.
Wu C.I. 2001 The genic view of the process of speciation. Journal of Evolutionary Biology 14: 851-865.
Wu C.I. & C.T. Ting 2004 Genes and speciation. Nature Reviews Genetics 5: 114-122.
Yoder J.B., R. Briskine, J. Mudge, et al. 2013 Phylogenetic Signal Variation in the Genomes of Medicago (Fabaceae). Systematic Biology 62: 424-438.
132 Tables
Table 1: Overview of the RAD-Seq sample composition and basic population genetic indices.
Central-west Pacific P. damicornis populations were characterized by high levels of genetic polymorphism (9-12%), many private alleles (>800) and high proportions of fixed private alleles
(5-10%). Most TEP samples (Peyd, Pele, Pdam_T1 & Pdam_T1/3) had low levels of polymorphisms (7-8%), few private alleles (<250) and low proportions of fixed private alleles
(<2%). No significant correlations were found between the number of raw reads and the number of pooled individuals per sample (Figure S1) with the number of heterozygous loci and/or the number of private alleles.
Region: CWP = central-west Pacific; TEP = Tropical Eastern Pacific; Location: Tuamotus =
French Polynesia; Gulf of Chiriquí & Gulf of Panamá = Panamá; Npool = Number of pooled specimens; 1PA = Private Alleles; Reads [M] = number of raw reads [in million reads] after quality control; Hets = heterozygous loci = loci that are variable within each sample.
Genotypes Reads Variable 1 Sample Species Region Location N Hets PA ITS2 ORF pool [M] loci
Pdam_Ja Japan Type Type 17 3.5 18,350 1,405 842 CWP Pdam_Tu Tuamotus 5 5 23 9.3 26,210 1,793 993 Gulf of Pdam_GC 21 17.6 27,037 1,151 539 Pocillopora Chiriquí Type Type Gulf of 3 3 Pdam_GP damicornis 18 7.6 21,533 1,561 240 Panamá Type Pdam_T1/3 21 26.3 31,586 1,217 135 TEP Gulf of 1/3 Chiriquí Pdam_T1 Type 13 9.5 15,275 1,201 116 & Type 1 Peyd P. eydouxi Gulf of 16 22.5 33,301 1,224 225 1 Panamá Pele P. elegans 18 2.8 29,293 1,074 167
147 99.1 39,581 6,128 3,242
133 Table 2: Pairwise comparisons between samples using P-distance and IbS-proportions.
The average amount of differently fixed SNPs between sample pairs was 588; since 4,187 loci were on average fixed between samples, the average P-Distance was 0.14. Allele-specific pairwise comparisons (IBS) indicated on average 7,073 alleles differences in 27,076 allelic comparisons, which correspond to an average IbS of 1,780 and an average IbS-proportion of
0.26. The average sample pair was thus fixed for the same allele at 12,901 P8 loci, for different alleles at 588 loci and contained 2,000 heterozygous loci.
Above: IbS-proportions = Proportion of allele comparisons that are different;
Average = 0.261 ±’0.094
Below: P-distance = Proportion of fixed loci that are different; Average = 0.140 (±’0.084)
Intra-clade comparisons are boxed in
Grey background highlights the increase in pairwise distances between type 1 samples and
P.‘damicornis samples with ITS type 3 and type 5 (Pdam_T1/3 < Pdam_T1 < Pele/Peyd).
Pdam Pdam Pdam Pdam Pdam Pdam Pele Peyd Ja Tu GC GP T1/3 T1 Pdam_Ja 0.227 0.313 0.324 0.333 0.336 0.334 0.341 Pdam_Tu 0.045 0.315 0.327 0.335 0.338 0.337 0.344 Pdam_GC 0.206 0.183 0.196 0.302 0.306 0.305 0.312 Pdam_GP 0.198 0.176 0.046 0.237 0.239 0.238 0.244 Pdam_T1/3 0.233 0.212 0.204 0.091 0.116 0.120 0.127 Pdam_T1 0.238 0.218 0.210 0.094 0.000 0.117 0.124 Pele 0.244 0.224 0.214 0.100 0.001 0.001 0.125 Peyd 0.246 0.227 0.218 0.099 <0.001 0.001 0.003
134 Table 3: Genetic make-up of putative hybrid P. damicornis samples
Genetic make-up of putative hybrid P. damicornis samples (Pdam_T1/3 & Pdam_T1) at 4333 loci that are segregating between type 3 (Pdam_GC) and type 1 (Peyd/Pele). Percentages are calculated based on the number of alleles/loci present in each sample. 581 segregating loci were not recovered from Pdam_T1/3, 882 loci were not recovered from Pdam_T1 and 1065 were not recovered from either one (i.e. “both”).
To evaluate levels of random ‘background’ variation of SNPs between lineages, the type 1 samples Peyd and Pele were analyzed vice versa at all 3,723 loci segregating between type 3
(Pdam_GC) and Pdam_T1-Pdam_T1/3 (Table S4).
Alleles Loci Type 3 Type 1 Type 3 Type 1/3 Type 1
(Pdam_GC) (Peyd/Pele) (Pdam_GC) Mixed (Peyd/Pele) Type 3 8666 100% 4333 100% (Pdam_GC) Pdam_T1/3 967 13% 6630 87% 134 4% 699 18% 2962 78% Pdam_T1 403 6% 6494 94% 34 1% 335 10% 3077 89% Both 298 5% 5850 90% 21 1% 230 7% 2638 81% Type 1 8666 100% 4333 100% (Peyd/Pele)
135 Figures
Figure 1a: Pocillopora elegans colony in a P. damicornis carpet in Isla Coiba, Gulf of
Chiriquí, Panamá.
136
Figure 1b: Phenotypic plasticity of P. damicornis/P. verrucosa in Moorea, French Polynesia.
137
Figure 2: ORF and ITS Phylogeny
Phylogenetic tree and 1000 Bootstrap replicates from RAxML. The final ORF alignment was
632 bp long and contained on average 565 bp of Pocillopora sequence data. The final ITS alignment was 643 bp long and contained on average 389 bp of Pocillopora sequence data.
138 Outgroup sequences were obtained from GeneBank for pocilloporiid corals Seriatopora sp. and
Stylophora sp. A, B & C (sensu Flot et al. 2011).
139
Figure 3: RAD phylogenomic analyses
Tree topology and branch lengths [substitutions/site] are based on Maximum likelihood reconstructions (RAxML). The tree is rooted according to the ORF phylogeny (Figure 2).
Bootstrap support (in bold) is based on 1000 pseudo-replicates for RAxML and 100 pseudo- replicates for Maximum Parsimony (DNApars) and reported for analyses based on all loci and fixed SNPs loci only (RAxML / fix RAxML / DNApars / fix DNApars). The tree topology was identical among all four analyses with the exception of one node (intra type 1) in the fixed
Maximum Parsimony analyses. The tree topology for P8 loci was identical and the bootstrap support was virtually identical (results not shown).
140
P-Distance Principal Coordinate Analysis (PCoA)
Pdam_GC
Pdam_GP
Pdam_T1/3 Pdam_T1 Pele
Pdam_Tu Peyd 2(26%) Coordinate Pdam_Ja Coordinate 1 (73%)
Figure 4: Principal Coordinate Analysis of pairwise P-distances between RAD-Seq samples.
PCA is drawn proportionally (i.e. according to percentages per axis). The first principal component explains 73% of the variance (53% for IBS, Figure S3) and separated the pure P. damicornis samples from the hybridizing samples in clade 1. The second principal component explained 26% of the genetic variance and separates the clade 3 from the others.
Among type 1 sample, the two hybrids (Pdam_T1/3 & Pdam_T1) are more similar to other P. damicornis samples than the other two type 1 samples (Peyd & Pele). The unexpected intermediate identity of Pdam_GP is also evident.
141 Hybrid composition 100% F1
75%
50% F2 Heterozygosity
25% F3
Pdam_T1/3 F4
Pdam_T1 F5 0% 50% 40% 30% 20% 10% 0% Pdam alleles
Backcross Generation F1 F2 F3 Pdam_T1/3 F4 Pdam_T1 F5 Heterozygous alleles 100% 50% 25% 18% 13% 10% 6% Type 3 alleles 50% 25% 13% 13% 6% 6% 3%
Figure 5: Allele composition of putative hybrid P. damicornis samples compared to the expected allele compositions of backcrossing hybrids over several generations.
The heterozygote sample Pdam_T1/3 is on average probably only 3-4 generations removed from the initial hybridization event while Pdam_T1 is on average probably 4-5 generations removed.
142 Supplementary Material
Table S1: Read Depth per sample and per locus
Table S1a shows the minimum (Min), Average (Avg) and Maximum (Max) read depth per sample across loci and the total number of reads per sample.
Reads per sample Pdam Pdam Pdam Pdam Pdam Pdam Peyd Pele Total Ja Tu GC GP T1/3 T1 Min 5 5 5 5 5 5 10 5 5 Avg 37 54 164 77 175 25 220 64 133 Max 381 866 1264 609 1129 230 1220 394 1264
Sum 255,555 375,134 1,134,319 531,066 1,211,264 175,617 1,517,396 442,122 8,288,661
143 Table S2: Transcriptome matches/alignments.
RAD loci are considered to match transcriptome contigs if their consensus sequence aligns with
0 or 1 mismatch over the 47bp sequence (2%).
Matches #: The number of RAD loci consensus sequences that match each transcriptome.
Matches %: The proportion of RAD loci that match to a specific transcriptome.
Multiple matches: Percentage of loci that matched to multiple sites within a specific
transcriptome (mostly due to multiple perfect alignments across the 47bp sequences).
Transcriptome coverage: Percentage of transcriptome contigs that had at least one RAD locus
Transcriptome contigs: The number of contigs in each of the 4 transcriptomes.
1New = new transcriptome for TEP Pdam_T1 and Pdam_T1/3 (Combosch, unpublished);
2Public = Traylor-Knowles et al. 2012 = likely type 3 and/or type 5.
3Combined = loci that match to either transcriptomes were considered to be transcribed loci.
4Symbiodinium Clade C & D = Ladner et al. 2012
Matches Multiple Transcriptome # % matches coverage contigs P. damicornis 1New P8 3,383 21.8% 1.7% 3.7% 73,522 P2 10,673 9.5% 0.8% 10.1%
2 Public P8 1,961 12.6% 2.9% 2.4% 70,786 P2 6,518 5.8% 1.2% 7.2%
3 Combined P8 3,831 24.7% P2 12,772 11.3%
4Symbiodinium Clade C P8 1 <0.1% 0.0% 0.0% 26,947 P2 35 <0.1% 0.0% 0.0% Clade D P8 0 0.0% 0.0% 0.0% 23,657 P2 42 <0.1% 0.0% 0.0%
144 Table S3: IbS & Number of differently fixed loci
Above: = IbS distances between sample pairs; Average = 1775 (+/- 574)
Below: = Differently fixed loci among sample pairs; Average = 588 (+/- 395)
Intra-clade comparisons are boxed in
Grey background highlights the increase in pairwise distances between type 1 samples and
P.‘damicornis samples with ITS type 3 and type 5 (Pdam_T1/3 < Pdam_T1 < Pele/Peyd).
Pdam Pdam Pdam Pdam Pdam Pdam Pele Peyd Ja Tu GC GP T1/3 T1 Pdam_Ja 1544 2126 2199 2262 2282 2273 2321 Pdam_Tu 181 2139 2223 2280 2299 2289 2339 Pdam_GC 887 722 1332 2055 2083 2070 2124 Pdam_GP 778 633 206 1609 1622 1621 1660 Pdam_T1/3 983 820 917 397 792 817 862 Pdam_T1 1009 845 947 408 0 798 844 Pele 1063 895 988 443 5 6 852 Peyd 1037 874 979 427 2 3 14
145 Table S4: Vice versa comparison of allelic identity
Vice versa comparison of allelic identity revealed minor allele sharing between type 3 and
Peyd/Pele, most likely due to backcrossing. To evaluate the level of random ‘background’ variation of SNPs between lineages, the type 1 samples Peyd and Pele were analyzed vice versa at all 3,723 loci segregating between type 3 (Pdam_GC) and Pdam_T1-Pdam_T1/3 (Table S4a).
Vice-versa comparisons revealed minor background allele sharing between type 3 and Peyd/Pele.
Interestingly Peyd contained more type 3 alleles than Pele, which could be due to backcrossing and introgression of type 3 alleles into Peyd and would indicates preferred backcrossing of hybrids with P. eydouxi.
Hybrids comparison- Type 3 (Pdam_GC) vs. Type 1(Peyd &\ Pele): 4333 segregating loci
Type 3 fix Hets Peyd-Pele fixed Both hybrids 21 0.5% 230 5% 2,638 61% Pdam_T1 34 0.8% 335 8% 3,076 71% Pdam_T1/3 134 3.1% 699 16% 2,961 68% Pdam_GP 1,153 44% 897 35% 543 21%
Vice Versa comparison – Type 3 (Pdam_GC) vs. (Pdam_T1/Pdam_T1/3): 3723 segregating loci
Type 3 fix Hets T1-T1/3 fixed Peyd & Pele 2 0.1% 13 0% 1,572 42% Pele 7 0.2% 47 1% 1,685 45% Peyd 29 0.8% 150 4% 3,022 81% Pdam_GP 858 23% 869 23% 493 13%
146 Polymorphic P8 loci
#reads [M] Read Depth #specimen #polym/Raw reads #polym/N pool #polym/read depth 60 R² = 0.07098 R² = 0.273 R² = 0.10592 250
50 200
40 150 30 100 20
50 10
0 0 1,000 1,200 1,400 1,600 1,800 # Polymorphic loci
Private Alleles (PA)
# reads [M] # specimen Read Depth PA/Raw reads PA/N pool PA/Read depth 60 R² = 0.10061 R² = 0.25495 R² = 0.11488 250
200 45
150 30 100
15 50
0 0 0 200 400 600 800 1000 # Private Alleles
Figure S1: Basic RAD loci correlations
No significant correlation between the number of heterozygous loci and the number of specimens per sampling pool (R2 = 0.27), raw reads per sample (R2 = 0.07) and the number of reads per locus (R2 = 0.11) nor between private alleles and the number of raw reads per sample
2 2 (R2 = 0.10), the number of pool specimen (R = 0.25) and the number of raw reads (R = 0.11).
147 0.25
TU_T1/2
0.20
0.15
y = 0.9497x - 0.0019 R² = 0.985 0.10
GC<>GP
P-Distance for Transcribed loci loci P-DistanceTranscribed for JA<>TU 0.05
0.00 0.00 0.05 0.10 0.15 0.20 0.25 P-Distance for Untranscribed loci Figure S2: Pairwise P-distance comparisons between samples at transcribed vs. untranscribed loci.
Green line: expected neutral 1:1 ratio. Dotted black line: average trend among samples.
Average All = 0.140; Average Transcribed = 0.133; Average Untranscribed = 0.142;
Pairwise comparisons over transcribed versus un-transcribed loci indicated higher levels of differentiation among un-transcribed loci in 24 out of 28 pairwise comparisons, indicating genome-wide mild balancing selection. Further indications for balancing selection are the increasing gap between transcribed vs. transcribed loci with larger genetic differences between samples and higher proportions of private alleles at untranscribed loci (not shown).
Three pairwise comparisons (in red) are more differentiated at transcribed loci, which could indicate positive selection between the two geographic P. damicornis populations pairs Gulf of
Chiriquí vs. Gulf of Panamá & Tuamotus vs. Japan.
148 FINAL CONCLUSIONS
I have shown that even in stable, oceanic environments TEP P. damicornis populations do have sex, which is well organized and guarantees that every genotype contributes into the common gene pool while successful genotypes are allowed to produce more offspring on their own, i.e. parthenogenetically (Chapter 3). In contrast, TEP P. damicornis populations have more sex, are genetically fine-structured (Chapter 2) and exchange large parts of their genome with local congenerics (Chapter 4). This combination facilitates adaptations and is most likely the key to the extraordinary success of P. damicornis in the environmentally challenging TEP. In addition, I showed that distinct genetic lineages within TEP Pocillopora damicornis correspond to genome-wide differences but colonies with P. damicornis morphologies share a core set of genes (candidate speciation genes) across genomic lineages (Chapter 4). Moreover, regional P. damicornis populations are composed of different genomic lineages and exchange genes between lineages and with local congenerics via introgressive hybridization (Chapter 4).
However, it feels like with every question I answered, five new questions arose: What are the molecular adaptations that enable corals to survive in the harsh TEP environments and how do they differ among environments? What are the ecological functions of different larval types and what are the molecular mechanisms that enable their simultaneous production. What evolutionary mechanisms keep the distinct genomic lineages separate within morphospecies and why is the same morphospecies composed of different genomic lineages in different locations?
But I will leave that for my next research project…
149