The Pennsylvania State University the Graduate School Eberly

The Pennsylvania State University

The Graduate School

Eberly College of Science

SPECIES DISTRIBUTIONS AND POPULATION STRUCTURE IN COLD SEEP

VESTIMENTIFERAN TUBEWORMS OF THE GENERA ESCARPIA AND

LAMELLIBRACHIA (POLYCHAETA, SIBOGLINIDAE)

A Dissertation in

Biology

Dominique Alexandria Cowart

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

August 2013

The dissertation of Dominique Alexandria Cowart was reviewed and approved* by the following:

Charles R. Fisher Professor of Biology Dissertation Co-Adviser

Stephen W. Schaeffer Professor of Biology Dissertation Co-Adviser

Kateryna Makova Professor of Biology Chair of Committee

Christopher House Professor of Geosciences

Douglas R. Cavener Professor of Biology Head of the Department of Biology

*Signatures are on file in the Graduate School

iii ABSTRACT

Deep-sea cold seeps support diverse biological communities that rely on the seepage of hydrocarbon gases and fluids. Due to their dependence on irregularly occurring seepage, vestimentiferan tubeworms (Family: Siboglinidae) have patchy distributions on the seafloor.

Tubeworms of the genera Escarpia and Lamellibrachia are important components of seep ecosystems, as they form long-lived aggregations that provide habitat for a variety of other deep- sea fauna.

Three described species of Escarpia, E. spicata (Gulf of California), E. laminata (Gulf of

Mexico) and E. southwardae (West African Cold Seeps), have been identified as one species through the use of mitochondrial markers Cytochrome Oxidase subunit 1 (mtCOI) and large ribosomal subunit rDNA, (mt16S), despite their geographic differences and dissimilar morphologies. Three morphologically distinct groups of Lamellibrachia are present over a bathymetric range of almost 3000m in the Gulf of Mexico; these include L. luymesi (<1000m) and two unnamed groups, L. sp. 1 and L. sp. 2 (>1000m). As with the Escarpia, mtCOI and mt16S identifies L. luymesi and L. sp. 1 as one species despite their differing depth ranges and dissimilar morphologies.

This dissertation seeks to define the geographic and bathymetric ranges of Escarpia and

Lamellibrachia by determining if geographically differentiated Escarpia and depth differentiated

Lamellibrachia are genetically distinct within each respective group. For these investigations, we make use of molecular, morphological and environmental data, all of which are necessary to ascertain population boundaries and evolutionary processes occurring in the deep sea. Here we use Exon Priming Intron Crossing (EPIC) sequencing of a nuclear gene, and develop microsatellite markers with the aid of 454 and Illumina next generation sequencing to 1)

iv determine if the described taxa are genetically differentiated and 2) identify possible population structure at the regional scale within the Gulf of Mexico and West Africa, respectively.

To identify if the three Escarpia groups are genetically distinct, we have tested the mitochondrial gene Cytochrome B (CYTB) for its utility as a phylogenetically informative marker, as well as developed and analyzed the EPIC marker Hemoglobin subunit B2 intron

(HbB2i) and 28 microsatellites in 229 Escarpia individuals collected from 12 seep sites around the world. Nine of the microsatellites were amplified across the three Escarpia taxa, and while

CYTB identifies two groups rather than three, both HbB2i and the cross-amplified microsatellites support the occurrence of three genetically distinct groups of Escarpia based on geography. At the regional scale among eight sampling sites of E. laminata (n =129) and among three sampling sites of E. southwardae (n =80), no population structure was detected.

To identify if the two Lamellibrachia groups are genetically distinct, we also tested

CYTB, HbB2i and eight microsatellites in 76 Lamellibrachia individuals collected from 13 seep sites in the Gulf of Mexico. All eight of the microsatellites were amplified across 45 L. luymesi and L. sp.1, and while both CYTB and HbB2i identify L. luymesi and L. sp. 1 as a single group, the cross-amplified microsatellites support the occurrence of two genetically distinct groups. At the regional scale among eight sampling sites of L.sp.1 (n =24) and among six sampling sites of

L. sp.2 (n =31), no population structure was detected.

Findings in this dissertation illustrate that 1) nuclear markers support original morphological descriptions of Escarpia and Lamellibrachia, suggesting that each group is constrained to their respective geographic or bathymetric region, 2) despite the patchiness and isolation of seep habitats, connectivity is high on regional scales, 3) mitochondrial markers should be used with extreme caution when attempting to detect differentiation in seep vestimentiferan groups, and lastly 4) microsatellites may be another useful tool to determine

v genetic differentiation between described species when divergence is insufficient or too recent to be detected using gene sequences alone.

vi TABLE OF CONTENTS

LIST OF FIGURES ...... ix

LIST OF TABLES...... xii

LIST OF ABBREVIATIONS...... xvi

ACKNOWLEDGEMENTS...... xix

CHAPTER1 Introduction and Motivation ...... 1

Cold seeps ...... 2 Vestimentiferan tubeworms ...... 4 The Escarpia tubeworms ...... 5 The Lamellibrachia tubeworms ...... 6 Larval biology and the influence of ocean currents on dispersal...... 8 Distributional ranges of deep sea invertebrates ...... 9 Ecological genetics of cold seep vestimentiferans...... 10

CHAPTER 2 Identification and amplification of microsattelite loci in deep-sea tubeworms of the genus Escarpia (Polychaeta, Siboglinidae)...... 13

Introduction...... 14 Materials and Methods...... 15 Results and Discussion...... 16 Acknowledgements...... 19

CHAPTER 3 Restriction to large-scale gene flow versus regional panmixia among cold seep Escarpia spp. (Polychaeta, Siboglinidae) ...... 20

Abstract ...... 21 Introduction...... 22 Escarpia vestimentiferan tubeworms...... 23 Vestimentiferan larval biology...... 26 Defining vestimentiferan population structure...... 27 Materials and Methods...... 28 Sample collection and preperation ...... 28 Sequencing and EPIC development ...... 30 Sequencing analyses...... 32 Microsattelite analyses ...... 33 Results...... 35 Mitochondrial and EPIC sequencing analysis...... 35 Genetic differentiation between Escarpia groups...... 37 Escarpia laminata population genetics...... 39 Escarpia southwardae population genetics...... 41 Discussion ...... 42

vii Barriers to gene flow across large scales ...... 42 Influence of deep currents on dispersal and recruitment of Escarpia...... 45 Conclusions...... 46 Acknowledgements...... 47

CHAPTER 4 Depth-dependent gene flow in Gulf of Mexico cold seep Lamellibrachia tubeworms (Annelida, Siboglinidae) ...... 48

Abstract ...... 49 Introduction...... 50 Cold seep Lamellibrachia tubeworms...... 51 Genetics of Gulf of Mexico Lamellibrachia spp ...... 52 Materials and Methods...... 54 Sample collection and preperation ...... 54 Marker development and sequencing analyses ...... 56 Microsattelite development and analysis ...... 57 Results...... 60 Mitochondiral and EPIC sequencing analyses...... 60 Genetic differentiation between Lamellibrachia groups...... 62 Lower slope Lamellibrachia population genetics ...... 65 Discussion ...... 66 Co-occurrence of vestimentiferan species and niche differentiation...... 66 Utility of mitochondrial genes versus microsattelite markers in identification of seep vestimentiferan species ...... 67 Influence of environmental factors on seep tubeworm species distributions...... 69 Potential for overlap of Lamellibrachia depth ranges...... 70 Low levels of genetic structure detected across geographic regions in Lamellibrachia tubeworms...... 70 Acknowledgements...... 72

CHAPTER 5 Summary and Implications ...... 73

DNA barcoding in seep vestimentiferans ...... 73 Implications of findings of dispersal and physical oceanography ...... 74 Investigations of genetically distinct vestimentiferan taxa ...... 75 Escarpia spp...... 75 Lamellibrachia spp...... 76 Use of hemoglobin molecules ...... 77 Use of microsattelite loci...... 78 Population genetics of seep vestimentiferans below 1000m...... 78 Future directions for research...... 80 Comparative genomics and laboratory breeding...... 80 Increasing loci and sample sizes ...... 80 Fine scale, temporal structure and ocean currents...... 82 Conclusions...... 82

REFERENCES ...... 83

viii APPENDIX A CHAPTER 2...... 99

APPENDIX B CHAPTER 3...... 102

Data Accessibility ...... 102

APPENDIX C CHAPTER 4...... 121

LIST OF FIGURES

Figure 1-1: Loop Current and associated eddies, Gulf of Mexico. FGBNMS: Flower Garden Banks National Marine Sanctuary...... 9

Figure 2-1: Frequency of repeat motif length for each of the three described Escarpia species...... 17

Figure 3-1: Aggregation of Escarpia southwardae, at the Worm Hole site on the West African coast (Photo complements of IFREMER, WACS campaign, dive 430)...... 24

Figure 3-2: Anterior portion, dorsal view of E. southwardae tubeworm, reprinted with permission from Andersen et al., 2004. AR: axial rod; BL: branchial lamellae; CG: ciliated groves (males); DG: dorsal groove; DVM: dorsal vestimental wings; O: obturaculum; T: trunk; V: vestimentum. NRC Research Press License # 2960880817425 ...... 25

Figure 3-3: Gulf of Mexico (GoM) cold seep sites sampled for this study. 1000 meter contours from NASA-JPL Advanced Spaceborne Thermal Emission and Reflection Radiometer (http://www.geomapapp.org)...... 30

Figure 3-4: West African cold seep sites sampled for this study. 1000 meter contours from NASA-JPL Advanced Spaceborne Thermal Emission and Reflection Radiometer (http://www.geomapapp.org)...... 31

Figure 3-5: Median-joining haplotype networks of the COI, CYTB and HbB2i genes. Colors represent Escarpia regions (described species) (grey: E. laminata from the Gulf of Mexico, white: E. southwardae from west coast of Africa and black: E. spicata from the Gulf of California. Sizes of haplotype circles and are proportional to the number of individuals possessing the same sequence and each line represents one mutational change separating two haplotypes...... 35

Figure 3-6: Top: STRUCTURE results (admixture model, three replicate runs) for three putative species of Escarpia vestimentiferan tubeworms based on nine microsatellite markers. Each vertical bar represents an individual tubeworm. The y-axis is the proportion of each individual’s genotype belonging to a distinct population cluster. Bottom: Network topologies of E. laminata (red), E. spicata (green) and E. southwardae (blue) individuals with the Shared Allele Distance (SAD) based on nine microsatellite markers. Only links with value smaller than or equal to the percolation distances are present. Nodes (circles) represent individuals. Three clusters are identified, one for each of the described Escarpia species and their respective geographic regions...... 39

Figure 4-1: Gulf of Mexico (GoM) cold seep sites sampled for this study. 1000 meter contours from NASA-JPL Advanced Spaceborne Thermal Emission and Reflection Radiometer (http://www.geomapapp.org). Lamellibrachia luymesi = dark grey circles, Lamellibrachia sp. 1 = light grey circles and Lamellibrachia sp. 2 = white circles ...... 56

Figure 4-2: Median-joining haplotype networks of the mtCYTB, HbB2i and concatenated genes. Colors represent Lamellibrachia luymesi (black) and Lamellibrachia sp. 1 (white). Sizes of

haplotype circles and are proportional to the number of individuals possessing the same sequence and each line represents one mutational change separating two haplotypes...... 61

Figure 4-3: Top: STRUCTURE results (admixture model, three replicate runs) for Lamellibrachia luymesi (black) and Lamellibrachia sp. 1 (white). Each vertical bar represents an individual tubeworm. The y-axis is the proportion of each individual’s genotype belonging to a distinct population cluster. Bottom: Network topologies of L. luymesi (blue) and L. sp. 1 (red) individuals from the Rozenfeld Distance model (RD) based on eight shared microsatellite markers. Shared Allele Distance (SAD) model results are not shown. Only links with value smaller than or equal to the percolation distances are present. Nodes (circles) represent individuals. Two clusters are identified, one for each Lamellibrachia population...... 63

Figure A2-1: Allele frequencies for each of the nine cross-amplified loci in Escarpia spicata...... 99

Figure A2-2: Allele frequencies for each of the nine cross-amplified loci in Escarpia laminata...... 100

Figure A2-3: Allele frequencies for each of the nine cross-amplified loci in Escarpia southwardae...... 101

Figure A3-1: COI maximum likelihood (ML) tree for 42 Escarpia and five Seepiophila, with fragment size of 658bp. 34 Escarpia sequences are newly isolated. Seepiophila jonesi is the outgroup taxon, and GenBank accession numbers follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions per nucleotide site ...... 102

Figure A3-2: 16S maximum likelihood (ML) tree for 45 Escarpia and five Seepiophila, with fragment size of 435bp. 38 Escarpia sequences are newly isolated. Seepiophila jonesi is the outgroup taxon, and GenBank Accession numbers follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions per nucleotide site...... 104

Figure A3-3: CYTB maximum likelihood (ML) tree for 39 Escarpia and seven Seepiophila, with fragment size of 401bp. All sequences are newly isolated. Seepiophila jonesi is the outgroup taxon, and GenBank Accession numbers follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions per nucleotide site...... 105

Figure A3-4: HbB2i maximum likelihood (ML) tree for 43 Escarpia and five Seepiophila, with fragment size of 665bp. All sequences are new isolated. Seepiophila jonesi is the outgroup taxon, and GenBank Accession number follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions per nucleotide site...... 106

Figure A4-1: mtCYTB maximum likelihood (ML) tree for 31 Lamellibrachia with a fragment size of 380bp. All Lamellibrachia sequences are newly isolated and Escarpia laminata is the outgroup taxon. GenBank accession numbers follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions...... 121

Figure A4-2: HbB2i maximum likelihood (ML) tree for 30 Lamellibrachia with a fragment size of 668bp. All Lamellibrachia sequences are newly isolated and Escarpia laminata is the outgroup taxon. GenBank accession numbers follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions per nucleotide site...... 122

Figure A4-3: Concatenation of mtCYTB and HbB2i maximum likelihood (ML) tree for 21 Lamellibrachia with a fragment size of 1048bp. Escarpia laminata is the outgroup taxon, and GenBank accession numbers follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions per nucleotide site...... 123

xii

LIST OF TABLES

Table 2-1: Summary of 454 reads for Escarpia. Number of filtered repeats: repeats where primers had lengths of 18 – 27bp, GC content >50%, Tm >50oC...... 17

Table 2-2: Characteristics of 28 microsatellite loci for E. laminata (EL) and E. southwardae (ES)...... 18

Table 3-1: Locations of seep sites sampled for this study. N refers to the number of individuals collected and region refers to location from which the collections were made: Gulf of California (GoC), Gulf of Mexico (GoM), West African Cold Seeps (WACS)...... 29

Table 3-2: Summary statistics for 28 microsatellite loci amplified in three described Escarpia species ...... 37

Table 3-3: Summary of Analysis of Molecular Variance (AMOVA) conducted for each study

under the Infinite Allele Model, FST...... 40

Table 4-1: Locations of seep sites sampled for this study. N refers to the number of individuals collected and region refers to the area in the Gulf of Mexico (GoM) from where the collections were made. Regions are defined as either west or east of the Mississippi Canyon. ... 55

Table 4-2: Characteristics of eight microsatellite loci developed for Lamellibrachia...... 59

Table 4-3: Summary statistics for eight microsatellite loci amplified in three Lamellibrachia groups...... 62

Table 4-4: Summary of Analysis of Molecular Variance (AMOVA) conducted for each study

under the Infinite Allele Model, FST...... 64

Table A3-1: Estimates of evolutionary divergence between species of Escarpia for three mitochondrial genes and one nuclear intron region. The numbers of base substitutions per site, averaging over all sequence pairs between groups are shown for each marker. Analyses were conducted using the Maximum Composite Likelihood model, and standard error estimates are shown above the diagonal...... 107

Table A3-2: Locations, number of individuals and GenBank IDs for all tubeworms tested in each of the four genes...... 108

Table A3-3: Summary statistics for Tajima’s D test of neutrality organized by gene and species. S is number of sites with substitutions; Pi is the mean number of pairwise differences. Distance method used is pairwise difference, no Gamma correction, indels not taken into account...... 109

xiii Table A3-4: Summary statistics for nine microsatellite loci amplified in three described Escarpia species...... 110

Table A3-5: Pairwise genetic differentiation matrix. FST values (below diagonal) and Jost’s D (above diagonal, calculated in GenoDive) between the three described species of Escarpia; species are significantly different at p = 0.001. Both measures quantify genetic diversity differently and both suffer from limitations, and therefore it has been suggested to use both measures in concert (Meirmans and Hedrick, 2011). FST measures the level of heterozygosity, while Jost’s D quantifies genetic diversity in terms of effective number of alleles (Jost, 2008; Ryman and Leimar, 2009. In this case, calculations of both metrics show similar relationships and patterns of differentiation between groups...... 111

Table A3-6: Primers used in this study...... 111

Table A3-7: Summary statistics for eleven microsatellite loci amplified in eight populations of Escarpia laminata (GoM). Abbreviations: number of alleles observed per locus (NA), observed (HO) and expected (HE) heterozygosity, Wright’s Inbreeding Coefficient (FIS). * bold asterisk indicates a significant difference after applying FDR correction = 0.01...... 112

Table A3-8: Pairwise genetic differentiation matrix. FST values (below diagonal) and Jost’s D (above diagonal) between the eight populations of Escarpia laminata (GoM); populations are ordered from west to east and are not significantly different (p > 0.05)...... 113

Table A3-9: Summary statistics for sixteen microsatellite loci amplified in three populations of Escarpia southwardae (WACS). Abbreviations: number of alleles observed per locus (NA), observed (HO) and expected (HE) heterozygosity, Wright’s Inbreeding Coefficient (FIS). *bold asterisk indicates a significant difference after applying FDR correction = 0.01...... 114

Table A3-10: Pairwise genetic differentiation matrix. FST values (below diagonal) and Jost’s D (above diagonal) between the three populations of Escarpia southwardae. Populations are not significantly different at p > 0.05. Both measures quantify genetic diversity differently and both suffer from limitations, and therefore it has been suggested to use both measures in concert (Meirmans and Hedrick, 2011). FST measures the level of heterozygosity, while Jost’s D quantifies genetic diversity in terms of effective number of alleles (Jost, 2008; Ryman and Leimar, 2009. In this case, calculations of both metrics show similar relationships and patterns of differentiation between groups...... 115

xiv

Table A3-13: Summary of null allele frequencies for 28 loci across three described species of Escarpia. NF refers to the null allele frequency and SE refers to the standard error. The Individual Inbreeding Model (IIM) was implemented in INEst v1.0 (Chybicki and Burczyk, 2009)...... 120

Table A4-1: Estimates of evolutionary divergence between Lamellibrachia for one mitochondrial gene, one nuclear intron and the concatenated region. The numbers of base substitutions per site, averaging over all sequence pairs between groups are shown for each marker. Analyses were conducted using the Maximum Composite Likelihood model, and standard error estimates are shown above the diagonal ...... 124

Table A4-2: Primers used in this study...... 124

Table A4-4: Summary statistics for eight microsatellite loci amplified in two Lamellibrachia morphospecies. Number of alleles observed (NA), observed (HO) and expected (HE) heterozygosity, Wright’s Inbreeding Coefficient (FIS), rarified allelic richness (AR), private allele richness (PAR), standard error (SE). Allelic and private allelic richness were rarefied over 36 samples and means are not significantly different (p > 0.05). Significant deviation from HWE after FDR correction = 0.01 is denoted by asterisk (*)...... 126

Table A4-6: Summary statistics for eight microsatellite loci amplified in two regional populations of Lamellibrachia sp1. Number of alleles observed (NA), observed (HO) and expected (HE) heterozygosity, Wright’s Inbreeding Coefficient (FIS), rarified allelic richness (AR), private allele richness (PAR), standard error (SE). Allelic and private allelic richness were rarefied over 16 samples and means are not significantly different (p > 0.05). Significant deviation from HWE after FDR correction = 0.01 is denoted by asterisk (*)...... 128

Table A4-7: Pairwise genetic differentiation matrix. FST values (below diagonal) and Jost’s D (above diagonal, calculated in GenoDive) between the Lamellibrachia, species are significantly different at p = 0.001. Both measures quantify genetic diversity differently and both suffer from limitations, and therefore it has been suggested to use both measures in concert (Meirmans and Hedrick, 2011). FST measures the level of heterozygosity, while Jost’s D quantifies genetic diversity in terms of effective number of alleles (Jost, 2008; Ryman and Leimar, 2009. In this case, calculations of both metrics show similar relationships and patterns of differentiation between groups...... 129

heterozygosity measured, S.D: standard deviation of mutation drift equilibrium distribution of heterozygosity; Prob: probability of obtaining the measured He in a sample from an equilibrium population (Piry et al., 1999). Bottom row shows Wilcoxon test results for 1000 replicates run under the TPM...... 130

Table A4-9: Null allele frequencies for all loci used in three groups of Lamellibrachia from the Gulf of Mexico. NA refers to the number of alleles, NF refers to the null allele frequency and SE refers to the standard error. The Individual Inbreeding Model (IIM) was implemented in INEst v1.0 (Chybicki and Burczyk, 2009)...... 131

Table A4-10: Escarpia and Lamellibrachia individuals used in this dissertation collected from aggregations below 1000m in the Gulf of Mexico. Each dive consisted of collections from only one aggregation. Some dives show that individuals from each genus were collected from the same aggregation...... 132

xvi

LIST OF ABBREVIATIONS

GENERAL ABBREVIATIONS

16S large ribosomal subunit rDNA AFLP Amplified Fragment Length Polymorphisms AMOVA Analysis of Molecular Variance AR Allelic Richness ARSE Allellic Richness Standard Error bp Base Pair COI Cytochrome Oxidase subunit 1 CYTB Cytochrome B DSV Deep Sea Vehicle DNA Deoxyribonucleic Acid EPIC Exon Priming Intron Crossing FDR False Discovery Rate

FIS Inbreeding Coefficient 3 FIS Inbreeding Coefficient across three populations

FST Fixation Index GoC Gulf of California GoG Gulf of Guinea GoM Gulf of Mexico HbB2i Hemoglobin subunit B2 intron He Expected Heterozygosity Ho Observed Heterozygosity HWE Hardy-Weinberg Equilibrium IFREMER Institut français de recherche pour l'exploitation de la mer

xvii

(French Research Institute for Exploitation of the Sea) LC Loop Current LLS Lower Louisiana Slope MCMC Monte Carlo Markov-Chain ML Maximum Likelihood mtDNA Mitochondrial DNA mt Mitochondrial

NA Number of alleles 3 NA Number of alleles across three populations NF Frequency of Null Alleles NGS Next Generation Sequencing NOAA National Oceanic and Atmospheric Administration N/O Navire Océanographique NSF National Science Foundation nuDNA Nuclear DNA PAR Private Allelic Richness SE PAR Private Allelic Richness Standard Error PCR Polymerase Chain Reaction RD Rozenfeld Distance ROV Remotely Operated Vehicle R/V Research Vessel SAD Shared Allele Distance SE Standard Error Tm Annealing Temperature TPM Two Phase Model TRW Topographic Rossby Waves ULS Upper Louisiana Slope USGS United States Geological Survey WACS West African Cold Seeps

xviii

Sample Locations

AC601 Alaminos Canyon 601 AC818 Alaminos Canyon 818 AT340 Atwater Valley 340 DC673 DeSoto Canyon 673 GB697 Garden Banks 697 GC184 Green Canyon 184 GC234 Green Canyon 234 GC600 Green Canyon 600 GC852 Green Canyon 852 MC294 Mississippi Canyon 294 MC344 Mississippi Canyon 344 MC751 Mississippi Canyon 751 TF4 Transform Fault 4 WR269 Walker Ridge 269 WFE West Florida Escarpment

xix

ACKNOWLEDGEMENTS

I begin by thanking my dissertation advisers Drs. Charles R. Fisher and Stephen W. Schaeffer for their support during the course of my tenure at Penn State. Under their supervision, I have greatly developed important research skills, and without their faith and foresight, I would not have experienced many wonderful opportunities that have led me to meet a whole host of other people to whom I am also grateful. I would like to thank the members of my committee, Drs. Kateryna Makova and Christopher House for their support throughout this project. I am also deeply indebted to Dr. Maria Pia Miglietta, who provided me with my first training in molecular biology bench work. It has also been a pleasure to collaborate with Dr. Kenneth Halanych (and his team), who has been supportive even before my time as a Ph.D. student, and gave me my first deep-sea dive in the Johnson SeaLink to view and collect the animals that are the focus of this dissertation. I would like to acknowledge my masters adviser, Dr. Adam Marsh; without his foresight, I would not have been as prepared to pursue my Ph.D. Also, thanks to Drs. Benjamin Cuker and James Ryan for their continued support and for allowing me to present my work to wide audiences. I am extremely appreciative to Drs. Todd LaJeunesse and Iliana Baums, as well as members of their lab groups, specifically Dr. Nicolas Polato, Dannise Ruiz, Meghann Devlin Durante, Jennifer Boulay and especially Drew Wham, who helped with methodologies important to my research and to whom I am tremendously grateful. Many thanks to my undergraduate students Andrew Mendelson and especially Chunya (Sunny) Huang, who spent countless hours processing samples and performing analyses on hundreds of tubeworms. Thank you to all of the members of Fisher and Schaeffer labs, past and present, for their assistance and advice: my good friends and great labmates Dr. Stephanie Lessard-Pilon, Pen-Yuan Hsing, Elizabeth Larcom; also to Drs. Stéphane Hourdez and Erin Becker, Arunima Sen, Elizabeth Podowski, Dr. Baptiste Faure, Matthew Porter, Max Zelenevich, Miles Saunders, Luyi Wo and Dr. Gwilym Haynes. I’d like to acknowledge Drs. Kimberlyn Nelson and Howard Fescemyer for their assistance with lab equipment; many thanks go out to other Penn State graduate students: Cheri Lee, Dr. Renee Rosier, John Parkinson, Jenny Beissinger Tennessen and Lindsey Swierk. Special thanks to Dr. Robert Vrijenhoek and his lab group at the Monterey Bay Aquarium Research

Institute (MBARI) for supplying tubeworm samples; to Dr. Karine Olu – Le Roy for her role as an outstanding chief scientist on the West African Cold Seeps cruise (WACS), and to Dr. Sophie Arnaud-Haond for her seemingly infinite patience while analyzing data and sharing her population genetics expertise with me. I also want to acknowledge Dr. Susan Carney, the team at Institut français de recherche pour l'exploitation de la mer (IFREMER) and specifically Annelise De Groote, who helped process hundreds of samples with Pen-Yuan and I. Many thanks to the captains, crews and expedition leaders of the ROVs Jason II and Victor 6000, the US deep submergence facility, DSV Alvin, R/V Atlantis, N/O Pourquoi Pas?, the NOAA vessel Ronald Brown, DSV Johnson Sea Link II, R/V Seward Johnson; their expertise was critical to the success of this work. I am beyond grateful for the support of the Department of Biology staff and members of the financial office, as well as Deb and Greg Grove, Ashley Price and the entire staff of the sequencing core facility; the Graduate School, Eberly College of Science, the Office of Graduate Educational Equity Programs (OGEEP), the Alfred P. Sloan Foundation, Alcorn Bridges students, Pallavi Eswara, Keisha Johnson, Dr. Suzanne Adair and Henry McCoullum for their support during my time at Penn State. Finally, I’d like to acknowledge and thank my true family: my grandmother, whom I would call and visit when I needed to think about things other than work; Drs. Andre Wallace and Heather Simmons who are former graduate students, current friends and fellow warriors; Dr. Shakira Nelson, Angeline Brown and Ashley Rankin, a trio of friends I could no way do without; Maria Vinca and K. Corcoran, who were available during my most difficult times, and Olivier Soubigou (my lighthouse), whom without, I would not have finished the last two years of my degree and whose patience and support has been indescribable. The work in this dissertation was supported by several funding sources including the National Science Foundation Dissertation Improvement Grant (NSF DDIG) DEB1209688 to DAC and SWS, the National Oceanographic Partnership Program (NOPP) through support from the Bureau of Ocean Energy Management (BOEM) contracts #M05PC00018 and #M08PC20038 (TDI Brooks International Prime), the National Oceanic and Atmospheric Administration’s Office of Ocean Exploration and Research (NOAA OER), Penn State Eberly College of Science, and an Alfred P. Sloan Scholarship to DAC. Despite the many trials and tribulations, I truly feel privileged to have had the opportunity to pursue my Ph.D., which would not have been possible without the support of all parties listed above.

CHAPTER 1

Introduction and Motivation

Biological communities at deep-sea hydrothermal vents and hydrocarbon seeps are unlike most communities on Earth, as they are not reliant on photosynthetic processes to produce organic matter. In the absence of light, deep dwelling chemosynthetic organisms depend upon chemicals originating from the seafloor to use as a source of energy, illustrating that life can thrive even under the most extreme conditions.

Hydrocarbon “cold” seeps support diverse and complex biological interactions that rely on the seepage of gases and fluids. Due to their dependence on irregularly occurring seepage, vestimentiferan tubeworms (Family: Siboglinidae) have patchy distributions on the seafloor; individual tubeworm aggregations can be separated by as little as one meter to hundreds of kilometers. There is currently a great deal of ambiguity regarding tubeworm biogeographic ranges and population connectivity, and much of the confusion lies in the lack of information available about mechanisms by which vestimentiferan larvae are transported to colonize these isolated habitats. On the one hand, one can measure the movement of larvae from location to location by the use of modeling aided by physiological development and hydrodynamic data (see

Marsh et al. 2001 for vents and Young et al. 2012 for seeps), but it is difficult to know when and where the tubeworms spawn in the water column, and where one must sample to document the transport of individuals. Alternatively, one can use genetic markers as an indirect measure of dispersal of individuals, which has the added benefit to infer that larvae moved between populations and were established.

2 This dissertation begins to define the geographic and bathymetric ranges of several vestimentiferan taxa by using next generation sequencing (NGS) tools to develop highly polymorphic genetic markers to infer the connectivity of tubeworm populations. These markers allow for the examination of gene flow patterns within and between populations in the Gulf of

Mexico and West African seep regions. These analyses include a combination of current and previously published morphological, molecular and environmental data to improve our understanding of how deep cold seep communities are established and maintained.

Cold Seeps

Hydrocarbon seeps fuel specialized deep-sea ecosystems along the continental margins of the Atlantic and Pacific Oceans as well as the Mediterranean Sea (Paull et al. 1984, Sibuet and

Olu 1998). Cold seep range from depths as shallow as 80m at Kagoshima Bay, Japan to over

7000m in the Japan Trench (Fujiwara et al. 2001, Miura et al. 2002, Tarasov et al. 2005).

Biological diversity is considered higher at cold seeps than at hydrothermal vents, in part due to the stability of seep habitats when compared to vents, and also due to complex interactions between environmental and biological systems that occur at colder temperatures (Sibuet and Olu

1998).

Gulf of Mexico (GoM) seep communities were some of the first to be discovered in the

1980s, are extensively studied, and have a high diversity of endemic species (Kennicutt et al.

1988, Ahron et al. 1994, McMullin et al. 2003). Seeps in the GoM develop distinctive features as a result of the sediment compaction of a thick salt layer, and subsequent creation of diapirs and faults; it is through these faults that hydrocarbon fluids and gasses migrate to surface of the seafloor (Anderson et al. 1983, Sassen 1994a, Cordes et al. 2009, Haas et al. 2010). About 175 million years ago, the Louann Salt Layer (northern GoM) and the Campeche Salt Layer (southern

3 GoM) were produced by repeated fluctuations in sea level and heavy evaporation when water levels were low (Burke 1975, Salvador 1987). Over time, river borne sediments from what is now the continental United State buried the salt layers. The accumulated weight of the hardening sediment layers (primarily sandstone and shale) deformed the underlying salt layer, creating salt pillars and diapirs. This process of “salt tectonics” created a complex network of depressions, cracks and faults in the overlying shale, which allowed for the upward migration of hydrocarbon fluids and gases (McGookey 1975, Cordes et al. 2009). Furthermore, the constant deformation of the salt layers controls the level of seepage, which supports specially adapted chemosynthetic fauna (Roberts and Aharon 1994).

West African Cold Seeps (WACS) are distributed along the Congo River canyon off of the continental margin of Gabon in the Gulf of Guinea. The first seep community at WACS was documented in 2001 on the giant pockmark Regab (measuring 800m wide and 20m deep), which was named for a local beer (Ondreas et al. 2005). Pockmarks are depressions in the seafloor that are created by over pressurized gas and fluid seepage; these depressions can be found on both passive and active margins worldwide (Olu-Le Roy et al. 2007, Cordes et al. 2009). Like the

GoM, seeps at WACS are formed by the compaction of a thick salt layer and subsequent creation of diapers and fault zones through which fluid and gas migration can occur. In contrast to the

GoM, the thick salt layer at WACS accumulated about 115 million years ago, and was gradually buried by black shale and sandstone deposits (Haas et al. 2010).

The Guaymas Basin is located in the central Gulf of California (GoC) and contains both hydrocarbon cold seeps located in the northern trough and sedimented hydrothermal vents located at the southern trough (Lonsdale and Becker 1985, Peter et al. 1990, Simoneit et al. 1990,

Kvenvolden and Simoneit1990). As is common at cold seeps, sediment thickness is high, and is attributed to the large influx of river borne sediments to the seafloor (Simoneit et al. 1990). In addition to high levels of sedimentation, the Guaymas Basin region contains several pockmark

4 depressions, located at depths >1000 m; these pockmarks are home to endemic seep organisms including the siboglind tubeworms and vesicomyid clams (Simoneit et al. 1990).

Vestimentiferan tubeworms

The family Siboglinidae consists of several species of polychaete worms, and is divided into four major groups. These groups include the Frenulates (Pogonophorans), or beard worms,

Monoliferans (“wood-eating” worms), Osedax spp. (“bone-eating worms”), and the vestimentiferan tubeworms (Webb 1969, Jones 1985, Southward 1991, Miura et al. 1997, Rouse

2001, Rouse et al. 2004, Andersen et al. 2004, Halanych 2005, Miura and Kojima 2006,

Southward et al. 2011). Siboglinds are unusual in that they lack mouths and digestive tracts, so adults rely on endosymbiotic bacteria for nutrition. With the exception of the Osedax, all siboglinids have males and females of equivalent size and mate via internal fertilization (Rouse et al. 2008).

The vestimentiferans are a diverse group that utilizes energy produced by chemosynthetic bacterial symbionts contained within the tubeworm trophosome. The trophosome is a large, well- vascularized organ that spans most of the length of the vestimentiferan body (Cavanaugh 1981,

Scott and Fisher 1995, Hourdez and Lallier 2007). Chemosynthetic bacteria are the primary producers, and use reduced chemicals seeping from the seafloor to fix carbon and produce carbohydrates for their eukaryotic hosts (Fisher 1996, Di Meo et al. 2000, Freytag et al. 2001,

Cordes et al. 2007, Cordes et al. 2009). Given the importance of the specialized bacteria in these ecosystems, the symbionts have been extensively studied to understand the role that they play in energy transduction (Southward 1982, Di Meo et al. 2000, McMullin et al. 2003, Vrijenhoek et al. 2007, Harmer et al. 2008, Duperron et al. 2009).

5 The best-known vestimentiferan is the giant red tubeworm, Riftia pachyptila (Jones

1981), which is endemic to vents in the Pacific. Their cold seep relatives belong to the genera

Lamellibrachia, Seepiophila and Escarpia, which can be found at seeps along the Pacific and

Atlantic margins (McMullin et al., 2003). Seep vestimentiferan larvae settle on exposed carbonate rock that is generated as a by-product of bacterial hydrocarbon degradation. Tubeworms grow to extend the posterior portion of their bodies (“roots”) deep into the sediment to acquire a supply of sulfide necessary to maintain their symbioses (Sassen et al. 1993, Julian et al. 1999, Freytag et al.

2001). These special adaptations may be responsible for the long life spans seen in at least two

GoM tubeworm species. Lamellibrachia luymesi and Seepiophila jonesi, have been found to live for centuries (Bergquist et al. 2000, Fisher et al. 2007).

Vestimentiferans are important components of seep habitats, as they cluster in aggregations that form “bush-like” three-dimensional structures. The size of the aggregation depends on the number of individuals as well as the stability and intensity of venting or seepage at a site (Olu-Le Roy et al. 2007). These aggregations form long-lived biogenic structures that provide food, living space and protection for a variety of other fauna (Berquist et al. 2003).

The Escarpia tubeworms

Vestimentiferans of the genus Escarpia are found at seeps below 950m and have a worldwide distribution (Olu et al. 2010). Three species have been described within the genus.

Escarpia laminata (Jones 1985) is distributed across the Lower Louisiana Slope in the GoM, and is found at depths from ~950 – 3300m (McMullin et al. 2003, Miglietta et al. 2010). Escarpia spicata (Jones 1985) is known from the western coast of North and Central America as well as the Gulf of California, and can inhabit seeps, hydrothermal vents and whale falls below 1000m

(Feldman et al. 1998, Vrijenhoek et al. 2007). Escarpia southwardae (Andersen et al. 2004)

6 occurs at pockmark depressions at West African Cold Seeps (WACS) distributed along the

Congo River Canyon at depths below 3000m (Ondréas et al. 2005, Olu-Le Roy et al. 2007).

All three Escarpia have a similar gross body plan, but differ in the details of their morphologies. Each described species differs in the shape of the axial rod, and E. spicata has a ventral ridge that is not present in E. laminata (Jones 1985). Additionally, E. southwardae is the only vestimentiferan known that does not have pinnules on the plume (Andersen et al. 2004).

Although the three described Escarpia species are geographically distant and morphologically distinct, they are considered a single taxa based on commonly used mitochondrial (mt) genetic markers Cytochrome Oxidase subunit 1 (COI) and the large ribosomal subunit rDNA (16S)

(Black et al. 1997, McMullin et al. 2003, Andersen et al. 2004, Miglietta et al. 2010). The inability of these markers to differentiate the described Escarpia species has led to suggestions that they are either a single morphologically plastic species or that there is low variability in mitochondrial genes. Low variability in mitochondrial markers may be the result of the long life spans, slow evolutionary rate, as well as efficient DNA repair systems (McMullin et al. 2003).

The large distance between populations of this single “species” would require that

Escarpia larvae travel thousands of kilometers across the Atlantic abyssal plain to maintain integrity of a single panmictic population. Given the vast distances across oceans with complex circulation patterns, this would require that larvae use an undiscovered continuum of chemosynthetic communities as “stepping stones” (Feldman et al. 1998, Cordes et al. 2007).

The Lamellibrachia tubeworms

Vestimentiferans of the genus Lamellibrachia contain at least six described species distributed along the western Pacific (L. satsuma, L. juni, L. columna), eastern Pacific (L. barhami), the Mediterranean Sea (L. anaximandri) and the Gulf of Mexico (L. luymesi) (Webb

7 1969, Southward 1991, Miura et al. 1997, McMullin et al. 2003, Miura and Kojima 2006,

Southward et al. 2011). Lamellibrachia from the GoM are some of the most extensively studied seep tubeworms and are distributed across the Louisiana Slope, from ~300m to 3300m. The most commonly studied GoM seep tubeworm, Lamellibrachi luymesi (van der Land & N∅rrevang

1975), resides on the Upper Louisiana Slope (ULS, < 1000m) and has a known range from 300 –

1000m. Two undescribed taxa, Lamellibrachia sp. 1 and Lamellibrachia sp. 2, are found on the

Lower Louisiana Slope (LLS, <1000m). L. sp. 1 ranges from ~950 – 2604m, while L. sp. 2, which is the most rarely observed of the three GoM Lamellibrachia, has a known depth range from 1175 – 3304m (Miglietta et al. 2010).

All GoM Lamellibrachia have a similar gross body plan, but differ in the details of morphology. Members of each group have gill lamellae, which are used for gas exchange and are a major feature that distinguishes Lamellibrachia from Escarpia tubeworms (Jones 1985). Each

Lamellibrachia morph differs in the details of the size and length of the vestimentum, as well as in the numbers of gill lamellae. L. luymesi has a slim vestimentum and 15 – 22 gill lamellae. L. sp. 1 has a characteristically fat and short vestimentum when compared to its body length, and both L. sp. 1 and L. sp. 2 have about 21 – 27 gill lamellae. L. sp. 2 and lacks a ventral vestimental fold that is present in L. sp. 1 (Jones 1985, McMullin et al. 2003, Miglietta et al. 2010).

Initial investigations employing mtCOI and mt16S suggested that L. luymesi and L. sp. 1 are the same species, despite depth and morphological differences. However, L.sp.2 is consistently identified as genetically distinct from the other Lamellibrachia spp by mitochondrial markers. (McMullin et al. 2003, Andersen et al. 2004, Miglietta et al. 2010). If in fact a single species, the known depth range of 2304m inhabited by L. luymesi/L. sp. 1 would suggest that this species is able to tolerate physiological constraints that accompany a range of temperatures and pressures.

8 Larval biology and the influence of ocean currents on dispersal

Adult vestimentiferans have separate sexes and the females are known to store sperm bundles. Eggs of seep tubeworms, L. luymesi and S. jonesi, are fertilized internally, and slightly buoyant embryos are released into the water column allowing the non-feeding lecithotrophic larvae to disperse on ocean currents (Young et al. 1996, Marsh et al. 2001, Vrijenhoek 2010). S. jonesi larvae reared in the laboratory survived for 21 days at 9 oC before metamorphosis, suggesting a possible dispersal range of 40 – 60km for tubeworms from < 1000m depth in the

Gulf of Mexico (Young et al. 1996, Tyler and Young 1999). In comparison, Riftia pachyptila, the giant hydrothermal vent tubeworm found on the East Pacific Rise, is estimated to have dispersal distances of about 100km along the ridge axis with a larval lifespan of five weeks (Marsh et al.

2001).

Ocean current dynamics add another dimension to the dispersal of marine groups, influencing population ranges by linking fragmented habitat patches (Palumbi 1992, Wares et al.

2001). Deep-sea currents below 1000m in the GoM exhibit complex behavior, which make it difficult to hypothesize their direction(s) and speeds of transport (Carney 2002, Cordes et al.

2007). In the GoM, the Loop Current branch of the Gulf Stream dominates the east as it flows past the Yucatan Peninsula and exits south of Florida (Figure 1-1). While the Loop Current is largely restricted to ~800 m depth, it sheds warm eddies that circulate north and west in the GoM

(Hamilton 1990). These eddies are thought to generate Rossby Waves, which penetrate deep into the ocean and have long wavelengths, small height and very slow speeds (Chelton and Shelax

1996).

Figure 1-1: Loop Current and associated eddies, Gulf of Mexico. FGBNMS: Flower Garden Banks National Marine Sanctuary: http://oceanservice.noaa.gov/facts/oceanfacts.php

Along the West African continental margin is the Congo deep-sea river fan, which

resulted from loads of sediment transported and deposited by the Congo River (Khripounoff et al.

2003, Sibuet and Vangriesheim 2009). The submarine fan covers depths from 3400 to 4800m and

contributes to the formation of cold seep habitats, providing substrate for vestimentiferan

communities (Vangriesheim et al. 2009). Deep-sea circulation below 2000m at the Congo River

fan is known to have slow moving currents that are complicated and difficult to track (Stramma

and England 1999).

Distributional ranges of deep sea invertebrates

The formation of a species is driven by complex evolutionary processes, in which natural

selection may cause isolated populations to undergo phenotypic and genotypic differentiation in

10 response to environmental conditions (Cook 1906, Schluter and Nagel 1995). There are many factors that influence speciation and distributional ranges of marine groups (see Jackson 1974); true biological species of some marine invertebrates can be difficult to identify given the lack of morphological diversity and the inability to cross and rear offspring (Palumbi 1994).

Furthermore, physical limitations of sampling in the deep sea have inhibited extensive study of deep dwelling invertebrates. Molecular genetic tools, combined with morphological and environmental data, are now universally accepted approaches for elucidating diverging evolutionary lineages (Vrijenhoek et al. 1994, Etter et al. 1999, Vrijenhoek 2010). A variety of molecular marker types, including mitochondrial, nuclear introns, and microsatellite loci, can be used to infer species boundaries (Hare 2001). Additionally, studies of biogeography and ecological adaptation not only help resolve species, but can also clarify evolutionary processes occurring among and within distinct populations.

Ecological genetics of cold seep vestimentiferans

Mitochondrial DNA generally provides powerful genetic markers for phylogeographic studies due to its moderate rate of mutation as well as its lack of introns, recombination, repetitive

DNA, or pseudogenes (Avise et al. 1987). The ease of universal primer design for PCR amplification is yet another reason why mtDNA is accepted for phylogenetic identification in many groups (Avise et al. 1987).

As noted above, several authors have attempted to infer distributional ranges and evolutionary relationships of vestimentiferans by employing mtDNA. However, because of the smaller number of copies of mtDNA when compared to an autosomal locus, mtDNA exhibits a high probability of reciprocal monophyly, and therefore the use of mtDNA can be misleading when determining the evolutionary history of species that are more recently diverged (Hudson

11 and Coyne 2002). In contrast to mtDNA, hypervariable loci, such as nuclear introns and microsatellites, can identify population differentiation previously undetected by less sensitive analyses (Shaw et al. 1999). Therefore, the analysis of many marker types is necessary to understand relationships and biogeographical distributions of vestimentiferans.

Gene flow and connectivity processes of vestimentiferans living at hydrothermal vents have been the subject of several studies because the linear arrangement of vents along a ridge crest provides a tractable model system for these studies (Tunnicliffe 1988, Mullineaux and

France 1995, Kim and Mullineaux 1998, Marsh et al. 2001). Several authors used mitochondrial markers and allozymes to uncover varying levels of gene flow in vent vestimentiferans (Black et al. 1994, Black et al. 1998, Hurtado et al. 2004, Vrijenhoek 2010). Recently, Shank and Halanych

(2007) proposed that genetic markers such as COI have under sampled genetic diversity in R. pachyptila, and investigated genetic structure using amplified fragment length polymorphisms

(AFLP) to uncover greater population subdivision than previously seen.

In contrast to vents, seep vestimentiferan connectivity is likely to be more complex as regions of active hydrocarbon seepage are patchily and haphazardly distributed. McMullin et al.

(2010) investigated genetic structure in seep populations of L. luymesi and S. jonesi, using twelve polymorphic microsatellite markers and found a pattern of isolation by distance in L. luymesi and a genetic clustering fitting age of aggregation in S. jonesi.

In this dissertation, I define the geographic ranges of Escarpia and the bathymetric ranges of Lamellibrachia vestimentiferans through the use of morphological, molecular and environmental data, all of which are necessary to ascertain boundaries of populations and evolutionary processes occurring between groups. In Chapter 2, I describe 28 polymorphic

Escarpia microsatellites that were developed with the aid of NGS tools, and test each locus’ utility as a population genetic marker.

12 In Chapter 3, I test for genetic differentiation among the three geographically separated

Escarpia taxa, as well as identify possible population structure on a regional scale within the

GoM and WACS biogeographic regions. In this chapter, I illustrate the use of Exon Priming

Intron Crossing (EPIC) sequencing and the microsatellite loci that were described in Chapter 2 to detect possible genetic differentiation between described species. In Chapter 4, I test for genetic differentiation among L. luymesi and L. sp. 1 taxa, two populations separated by depth.

Additionally, I identify possible population structure in L. sp. 1 and L. sp. 2 residing below

1000m in the GoM. In this chapter, I also apply EPIC sequencing tools, as well as microsatellite loci that were developed with the aid of a NGS transcriptome.

CHAPTER 2

Identification and amplification of microsatellite loci in deep-sea tubeworms

of the genus Escarpia (Polychaeta, Siboglinidae)

Previously published as: Cowart, DA, Huang, C, Schaeffer, SW (2012) Identification and amplification of microsatellite loci in deep-sea tubeworms of the genus Escarpia (Polychaeta, Siboglinidae). Conservation Genetics Resources 5 (2): 479-482

14 Introduction

Cold hydrocarbon seeps fuel specialized deep-sea ecosystems and are home to vestimentiferan tubeworms (Family: Siboglinidae) (McMullin et al. 2003). Vestimentiferans are important members of seep communities because they cluster in aggregations that provide living space for other fauna (Bergquist et al. 2003). Three described species of the genus Escarpia are distributed at seeps below 950m along continental margins of the Atlantic and Pacific (Olu et al.

2010). Escarpia laminata occurs across the Gulf of Mexico; Escarpia spicata is present in the

Gulf of California, and Escarpia southwardae occurs along the equatorial margin of West Africa

(Jones 1985, Andersen et al. 2004).

Seep communities are under increased environmental threat from dumping, oil and gas drilling, so we employed 454-pyrosequencing to develop microsatellites for Escarpia to understand how these communities are established and maintained. Microsatellites are often used due to high rates of polymorphism, which can confirm population differentiation undetected by less sensitive analyses, and the use of next-generation sequencing reduces time and costs associated with traditional marker development methods (Allentoft et al. 2009). Furthermore, microsatellites can be applied across species to clarify phylogenetic relationships (Primmer et al.

2005, Hausdorf and Henning 2010), and variation is expected to yield high returns for taxa within genera, with similar genome sizes and outcrossed mating systems (Barbará et al. 2007), all of which are characteristics of Escarpia. Here we report novel microsatellites that are species specific and amplify across all three Escarpia.

15 Materials and Methods

In this study, 229 Escarpia individuals were collected via submersible from twelve sites during research cruises occurring from 2003 to 2011 in the Gulf of Mexico, the Congo River

Canyon, and the Gulf of California. Tissue samples were taken from the vestimentum to avoid contamination from endosymbionts, and were preserved at either -80ºC or in 95% ethanol, then transported either to the Pennsylvania State University or to Monterey Bay Aquarium Research

Institute. Whole genomic DNA was obtained using a high salt protocol (Liao et al. 2007) with the following modifications: 3µL of RNase was added to digested samples and followed by incubation at 37ºC for one hour and 4ºC for 5 minutes before addition of 5M NaCl. The resultant

DNA pellet was allowed to air-dry overnight, then re-suspended in 50µL of sterile water and stored at -80ºC. To increase amplification yield, isolated DNA was purified using the MoBio

Power Clean DNA Clean-Up Kit; purified samples were kept at -20ºC.

One individual’s DNA from each species was submitted to the Penn State Genomics

Core Facility where three DNA libraries were constructed for Roche 454 FLX+ sequencing with

Titanium Chemistry (Roche Diagnostics). Completed sequence reads were organized using the

Galaxy workflow system (Blankenberg et al. 2010).

To identify microsatellite repeats, reads were uploaded and searched using the Tandem

Database v2.30 at http://tandem.bu.edu/cgi-bin/trdb/trdb.exe (Gelfand et al. 2006). Primers for the repeats were developed using Primer 3 v.0.4.0 (Rozen and Skaletsky 1999). To identify loci with high likelihood of amplification, primers were chosen based on lengths of 18 – 27 bp, GC content

≥50%, and melting temperature ≥50ºC. Each marker was tested for variability using the M13- tagging approach (Schuelke 2000) with DNA from several individuals.

Loci were amplified using these touchdown PCR conditions: 94 ºC (3 min); 94 ºC (30s),

59 ºC (45s), 72 ºC (30s) for 15 cycles while decreasing annealing temperature by 0.5 ºC every

16 cycle. Next, 94 ºC (30s), 52 ºC (45s), 72 ºC (30s) for 25 cycles followed by a final extension at 72

ºC (10min). Fragment analyses were run on a 3730XL DNA sequencer using GeneScan with

LIZ500 size standard (Applied Biosystems); fragments were scored manually using GeneMarker v4.0 (Soft Genetics, State College, PA). To screen for experimental artifacts and test for linkage disequilibrium, we employed MICROCHECKER v2.2.3 (van Oosterhout et al. 2004) and

GENEPOP v4.1 (Raymond and Rousset 1995). To compute observed and expected heterozygosity, Arlequin v3.5 (Schneider et al. 2000) was used, and corrections for the departure from HWE were completed using the False Discovery Rate method (Benjamini and Yekutieli

2001) in QVALUE (Storey and Tibshirani 2003) implemented in R (R Development Core Team

2011).

Results and Discussion

Pyrosequencing yielded ~250,000 reads with median read lengths between 372 and

378bp (Table 2-1). Figure 2-1 illustrates the frequency of motif length (i.e. dinucleotide, trinucleotide, etc.) for each of the three described species of Escarpia. Here we note that all three taxa have similar frequencies for each motif. Next, dinucleotides have the highest frequency in three groups. Dinucleotide repeats are common in the genomes of many eukaryotes (Tóth et al.

2000, Ellegren 2004) and therefore the high frequency of this motif seen in the Escarpia is not surprising. Nine of the E. laminata loci were amplified across all three species and contained 222 alleles. Tests for linkage disequilibrium were non significant (p > 0.05) and four of the 28 loci described here (Table 2-2) showed departure from HWE after FDR correction of 0.01. No locus expressed evidence of null alleles in every population and there was no evidence of stuttering or large allele dropout in the loci.

Table 2-1: Summary of 454 reads for Escarpia

Number of Reads with Mean Read Read Range Number of Described species Reads Repeats Length (bp) (bp) Filtered Repeats Escarpia spicata 63,572 16,823 378.2 29 - 1167 570

Escarpia laminata 72,575 18,869 372.0 29 - 955 682 Escarpia 113,204 30,807 376.1 29 - 946 1,190 southwardae Total 249,351 66,499 ------2,442

Number of filtered repeats: repeats where primers had lengths of 18 – 27bp, GC content >50%, Tm >50oC

In summary, despite the differences in the total number of reads between the three

described species of Escarpia, all three groups exhibited the same patterns with regards to motif

lengths. Each group had high frequencies of dinucleotide repeats, and low frequencies of odd

Figure 2-1: Frequency of repeat motif length for each of the three described Escarpia species.

18 number repeat motifs, suggesting possible expansion of repeats. Finally, out a screening of over

2,400- repeats, only nine polymorphic microsatellites cross-amplified in the three Escarpia taxa.

Table 2-2: Characteristics of 28 microsatellite loci for E. laminata (EL) and E. southwardae (ES) o 3 3 Locus Primers (5’ – 3’) Repeat Motif Size Range (bp) Tm ( C) NA NA HoHe FISFIS EL454_2* F: GAGACACCGTACGAATGACAGA (TGTGCGCG)19 135 - 297 56 12 0.670 0.037** R: ACTGGAACTATGGCACGGAC 31 0.720 0.031 EL454_5* F: TGCATGTATACGTGGAGTGAGA (TACATGCA)18 204 - 288 53 15 0.751 0.007 R: CCCTGCATTCATAATTGCTT 30 0.750 0.018

EL454_6 F: TGTGCTTCTCAAACCACC (CCACCAT)20 209 - 254 53 8 0.388 0.085 R: CAAAAGGCTTTCTCGTCCAC 0.415

EL454_9* F: TTATGTTCCCGTTGAAAGCC (CACCACA)13 83 - 185 54 19 0.682 0.314** R: GTGGGCGCTATGTCGTATTT 35 0.914 0.284

EL454_21* F: CTCAGCAGGAGGGTCAGTTC (GGAGGC)16 115 - 176 57 9 0.753 0.070** R: AACACACTGCTCTCTCGCAA 19 0.809 0.108

EL454_25* F: GTGCACTGTTGCATTATGGG (AGGGT)16 153 - 213 52 12 0.650 -0.021 R: CCATACATTCCATACCTTCG 12 0.608 -0.016

EL454_52* F: CGATGATGCCAACTAAATAGGG (TGTGTGGC)17 179 - 313 53 16 0.794 0.069 R: GACGACATTTAAGACAGGCG 34 0.888 0.072

EL454_54 F: CGCCCTAAGCACTGTATTCC (CACACG)14 139 - 223 56 20 0.739 0.151 R: TATAGACGCCAGGTGGAACG 0.894

EL454_64* F: AACACCAAAACTGTGTCCTCG (CATA)24 185 - 261 55 --- 0.750 --- R: ATGCGTGGAATGAAGGAGAC 20 0.835 0.039

EL454_67 F: TGCAGCAGTTTCACGAACTC (TAT)15 92 - 125 57 4 0.323 -0.143 R: ACGCTTCGTAGCACTGACCT 0.286

EL454_70* F: GTGCCGACACTTGAGCACTA (TGC)16 100 - 185 57 26 0.874 0.035 R: AGGCTAACCATGCAGTCACA 33 0.886 0.054

EL454_71* F: AATGCCAGATGTTTTCCCAA (GTT)19 115 - 199 54 4 0.150 0.518 R: TGCAATGGCACACCTGTTAC 8 0.397 0.254

ES454_8 F: CACTGGGAAAACGGTGAGTT (CATCACTA)23 125 - 198 56 9 0.530 0.096 R: CGCCTAGCTCTTTCTGAGGTT 0.589

ES454_10 F: AAGAACACGGACACGGACAC (CACACGCA)16 112 - 192 54 18 0.851 0.011 R: CGTCATCATCCTCACATTG 0.857

ES454_13 F: ATTCACCCACGCATTCTCTC (ACACACGC)21 136 - 208 56 10 0.423 -0.020 R: GAGTGGGTGAGTGCACGAG 0.415

ES454_25 F: GTGGCGTCAGATCCTCAGA (GCACAC)13 109 - 193 56 21 0.910 0.012 R: CAATGGCAGTACTTCGGGAT 0.922

ES454_28 F: ATGAATGGACTCATCCACGC (GTGCAT)24 114 - 162 53 11 0.785 -0.006 R: CATGTTTTACCCTCTTATTACGTTT 0.780

ES454_31 F: AATGAATTTCCCTCCCAAGG (ACACGC)28 131 - 207 53 22 0.794 0.116 R: TTCGATTGATGTCATGTCCG 0.900

ES454_32 F: GCGTGAGATTCCACCAGAAT (CACACG)14 108 - 168 55 5 0.164 0.302** R: CTCACGTGATGTATCGTCCG 0.240

ES454_34 F: TATCTGTCTGCCTTCCTGCC (CACGCA)23 109 - 133 55 7 0.761 -0.040 R: GACAACCGTGGCTTTTGAAT 0.729

ES454_44 F: GCCCATTTGTTTATCGTGCT (TGATGG)13 72 - 126 52 7 0.692 -0.071 R: CGGGTCTACATAAAAGCTTG 0.646

ES454_45 F: GGGATGATAACAAGAAACGC (CAGAGG)13 140 - 251 53 23 0.230 0.569** R: TGTCTGTCTGTACGTCTGTTTG 0.532

ES454_50 F: ACTGCACAAGATACGATGCG (GCGAT)20 95 - 325 55 13 0.785 0.130 R: CTAAGTCTCGTTTTCGGCGT 0.788

ES454_57 F: ACGCACTCACTCACGCAG (GTGC)28 118 - 202 56 16 0.836 0.027 R: GGGTTGGTGTTTGTGGAA 0.858

ES454_60 F: TCCTCGTCAGACAATCCAAA (TCCA)34 152 - 200 56 10 0.536 -0.083 R: CCCAACCACGGGACACTAC 0.494

ES454_64 F: CACACAAATGACATCCTAGCG (AGAC)20 104 - 137 55 8 0.473 0.010 R: GCGCATGTGTGCGTGTAT 0.475

ES454_71 F: AATGGGTACACAATACCGCC (TGTC)25 127 - 199 56 24 0.837 0.042 R: AGGCAGATGAAACGGAGTGT 0.871

ES454_95 F: CGCGTTCCTCACTATCACAA (TTA)19 124 - 148 55 8 0.739 0.016 R: CTGGTGAGCAGATGAAAGCA 0.749

*Loci amplified across three Escarpia species Tm =Annealing temperature, NA = number of observed alleles 3 NA = number of observed alleles across three species Ho = observed heterozygosity, He = expected heterozygosity 3 FIS = FIS values across three species **Significant deviation from HWE after FDR correction = 0.01

19 Acknowledgements

We would like to thank Charles Fisher, Drew Wham and Sophie Arnaud-Haond. This research was funded in part by the National Science Foundation (Award #1209688) and the

Alfred P. Sloan Scholarship. Any opinions, findings, conclusions, or recommendations expressed in this article are those of the authors and do not necessarily reflect the views of the National

Science Foundation.

CHAPTER 3

Restriction to large-scale gene flow versus regional panmixia among cold seep

Escarpia spp. (Polychaeta, Siboglinidae)

In press as: Cowart DA1, Huang C1, Arnaud-Haond S2, Carney SL3, Fisher CR1, Schaeffer SW1 (2013) Restriction to large-scale gene flow versus regional panmixia among cold seep Escarpia spp. (Polychaeta, Siboglinidae). Molecular Ecology

1Department of Biology The Pennsylvania State University 208 Erwin W. Mueller Laboratory University Park, PA 16802 - USA

2IFREMER (Institut Français de Recherche pour l'Exploitation de la MER) Unité Environnement Profond - DEEP du Département des Ressources physiques et Ecosystèmes de Fond de mer (REM) B.P. 70 – 29280 Plouzané - France

3Department of Biology Hood College 401 Rosemont Avenue Frederick, MD 21701 – USA

Abstract

The history of colonization and dispersal in fauna distributed among deep sea chemosynthetic ecosystems remains enigmatic and poorly understood because of an inability to mark and track individuals. A combination of molecular, morphological and environmental data improve understanding of spatial and temporal scales at which panmixia, disruption of gene flow or even speciation, may occur. Vestimentiferan tubeworms of the genus Escarpia are important components of deep-sea cold seep ecosystems, as they provide long-term habitat for many other taxa. Three species of Escarpia, E. spicata [Gulf of California (GoC)], E. laminata [Gulf of

Mexico (GoM)] and E. southwardae (West African Cold Seeps), have been described based on morphology, but are not discriminated through the use of mitochondrial markers (Cytochrome

Oxidase subunit 1, COI; large ribosomal subunit rDNA, 16S; Cytochrome B, CYTB. Here, we also sequenced the Exon Priming Intron Crossing (EPIC) Hemoglobin subunit B2 intron (HbB2i) and genotyped 28 microsatellites to 1) determine the level of genetic differentiation, if any, among the three geographically separated entities and 2) identify possible population structure at the regional scale within the Gulf of Mexico and West Africa. Results at the global scale support the occurrence of three genetically distinct groups. At the regional scale among eight sampling sites of E. laminata (n =129) and among three sampling sites of E. southwardae (n =80), no population structure was detected. These findings suggest that despite the patchiness and isolation of seep habitats, connectivity is high on regional scales.

Introduction

A multitude of factors influence speciation and distributional ranges of marine populations, including larval biology, life history, adult physiology, geographic and oceanic barriers, as well as temperature (Jackson 1974). In this context, life history can influence connectivity in marine habitats by either preventing or promoting isolation of populations

(Palumbi 1992). Ocean current dynamics add another dimension to the dispersal of marine groups, influencing population ranges by linking fragmented habitat patches (Johnson and Black,

1984, Wares et al. 2001, Hedgecock et al. 2007). True biological ranges of marine invertebrates in the deep sea can be especially difficult to assess given the lack of morphological diversity and the inability to cross and rear offspring (Palumbi 1994). Furthermore, physical limitations of sampling in the deep sea have inhibited extensive study of deep dwelling populations. Molecular genetic tools, combined with environmental and morphological data, are now universally accepted approaches for elucidating evolutionary lineages in deep-sea animals (Vrijenhoek et al.

1994, Etter et al. 1999, Vrijenhoek 2010). Additionally, studies of biogeography, genetic structure and ecological adaptation can provide information on the scales at which panmixia is replaced by gene flow restriction or disruption; if these processes are active over appropriate evolutionary time scales, they may lead to speciation. This question is particularly germane to the deep sea where distinct biogeographic provinces support an expectation of isolated populations in different regions, while the large scale dispersal potential of many taxa, and some population genetic data, suggest the occurrence of regular colonization events over large geographic distances.

23 Escarpia vestimentiferan tubeworms

Hydrocarbon seeps, which are located on passive ocean margins worldwide, fuel specialized deep-sea ecosystems. In the Gulf of Mexico, the continental slope has developed distinctive topography over time as a result of sediment compaction of a thick salt layer, and subsequent creation of diapirs and faults; it is through these faults that hydrocarbon fluids and gasses migrate to the seafloor (Andersen et al. 1983, Sassen et al. 1994a, Cordes et al. 2009, Haas et al, 2010). The constant deformation of the salt layer controls the level of seepage (Roberts and

Aharon 1994) which supports specially adapted chemosynthetic fauna, including the vestimentiferan tubeworms (Family: Siboglinidae), a group of polychaete worms that lack digestive tracts and depend on endosymbiotic bacteria for their nutrition (Childress and Fisher

1992). Seep vestimentiferan larvae settle on exposed carbonate rock that is generated as a by- product of bacterial hydrocarbon degradation (Sassen et al. 1993). Tubeworms extend the posterior portion of their bodies (“roots”) deep into the sediment to acquire the supply of sulfide necessary to maintain their symbioses (Julian et al. 1999, Freytag et al. 2001). These special adaptations allow individuals to have long life spans; at least two tubeworm species in the Gulf of

Mexico can live for centuries (Bergquist et al. 2000, Cordes et al., 2007a,b, Fisher et al. 2007).

Vestimentiferans are important components of seep habitats, as they cluster in aggregations that form “bush-like” three-dimensional structures (Figure 3-1), which provide living space and protection for a variety of other fauna (Bergquist et al. 2003). Furthermore, due to their dependence on hydrocarbons, habitats suitable for tubeworm communities can be separated by hundreds of kilometers (Brooks et al., 1987, Roberts et al., 2007).

Vestimentiferans of the genus Escarpia are distributed at seeps below 950m along the temperate and tropical continental margins of the Atlantic and Pacific Oceans (Olu et al. 2010).

Three species have been described within the genus.

Figure 3-1: Aggregation of Escarpia southwardae, at the Worm Hole site on the West African coast (Photo complements of IFREMER, WACS campaign, dive 430).

Escarpia laminata (Jones 1985) is distributed across the Lower Louisiana Slope in the

Gulf of Mexico (GoM), and is found at depths from ~950 – 3300m (McMullin et al. 2003,

Miglietta et al. 2010). Escarpia spicata (Jones 1985) is known from the western coast of North and Central America and the Gulf of California (GoC), and can inhabit seeps, hydrothermal vents and whale falls below 1000m (Feldman et al. 1998, Vrijenhoek et al. 2007). Escarpia southwardae (Andersen et al. 2004) occurs at pockmark depressions at West African Cold Seeps

(WACS) distributed along the Congo River Canyon at depths below 3000m (Ondréas et al. 2005,

Olu-Le Roy et al. 2007).

All three Escarpia have similar gross morphology (Figure 3-2). Members of each group have an axial rod on the top of the obturaculum, and the lack of large flare rings on their tubes contrasts with a related species, Seepiophila jonesi (Gardiner et al. 2001).

Each described species differs in the shape of the axial rod, and E. spicata has a ventral ridge that is not present in E. laminata (Jones 1985). Additionally, E. southwardae is the only vestimentiferan known that does not have pinnules on the plume. Pinnules are thought to be the primary site of O2 and CO2 uptake in the other species (Andersen et al. 2004).

Although the three described Escarpia species are geographically distant and morphologically distinguishable, they do not form monophyletic clades with commonly used mitochondrial markers Cytochrome Oxidase subunit 1 (COI) and the large ribosomal subunit rDNA (16S) (Black et al. 1997, McMullin et al. 2003, Andersen et al. 2004, Miglietta et al.

2010). COI has been suggested as a global standard to provide species level resolution in animals, hence its nickname, the “DNA barcoding” marker (Herbert et al. 2003, Hajibabaei et al. 2007).

However, there are limitations to using barcoding molecules (see Moritz and Cicero, 2004). The inability of this marker to differentiate the described Escarpia species has led to suggestions that

26 they are either a single morphologically plastic species or that there is low COI divergence between the Escarpia, which may be the result of the long life spans and the slow evolutionary rate seen in seep vestimentiferan mitochondrial genes (McMullin et al. 2003). The large distance between populations of a single “species” would require that Escarpia offspring travel thousands of kilometers along continental margins or across the Atlantic abyssal plain to maintain integrity of a single panmictic population, possibly using an undiscovered continuum of chemosynthetic communities as “stepping stones” (Feldman et al. 1998, Cordes et al. 2007).

Vestimentiferan larval biology

The investigation of life history traits in tubeworms is important for understanding how speciation occurs and how seep communities are established and maintained (Tyler and Young

1999). Adult vestimentiferans are sessile and have separate sexes. The females are known to store sperm bundles; eggs of the seep tubeworms from the upper Louisiana Slope of the Gulf of

Mexico, Lamellibrachia luymesi and Seepiophila jonesi, are fertilized internally, and zygotes are released into the water column allowing the lecithotrophic larvae to disperse on ocean currents

(Vrijenhoek 2010, Hilario et al. 2005). S. jonesi larvae reared in the laboratory survived for 21 days before metamorphosis, suggesting a possible dispersal range of 40 – 60km for tubeworms at

< 1000m in the Gulf of Mexico (Young et al. 1996, Tyler and Young 1999). In comparison, Riftia pachyptila, the giant hydrothermal vent tubeworm found on the East Pacific Rise, is estimated to have dispersal distances of about 100km along the ridge axis with a larval lifespan of five weeks

(Marsh et al. 2001).

27 Defining vestimentiferan population structure

Gene flow and connectivity processes of vestimentiferans living at hydrothermal vents have been the subjects of several studies because the linear arrangement of vents along a ridge crest provides a very tractable model system for these studies (Tunnicliffe 1988, Mullineaux and

France, 1995, Kim and Mullineaux, 1998, Marsh et al. 2001). Several authors used allozymes to uncover varying levels of genetic variation in vent vestimentiferans R. pachyptila, Tevnia jerichonana, and Oasisia alvinae across 4000km at the East Pacific Rise (Black et al. 1994, Black et al. 1998). Furthermore, Hurtado et al. (2004) used mtCOI to identify subdivision in R. pachyptila across the Easter Microplate Region, indicating restricted gene flow across this region.

Shank and Halanych (2007) proposed that genetic markers such as COI have under sampled genetic diversity in R. pachyptila, and investigated genetic structure using amplified fragment length polymorphisms (AFLP). With this approach, 630 polymorphic loci were uncovered to identify clustering of R. pachyptila individuals according to locality (9oN EPR,

Guaymas Basin, and Southern Eastern Pacific Rise) illustrating greater population subdivision than previously seen. Additionally, Vrijenhoek (2010) estimated within species gene flow as high to moderate in R. pachyptila and Ridgeia piscesae, which was reported as number of migrants transferred along contiguous ridge segments.

In contrast to vents, studying seep vestimentiferan connectivity is more complex as regions of active hydrocarbon seepage are patchily and haphazardly distributed, often in areas with very complex current regimes. McMullin et al. (2010) investigated genetic structure in seep populations of L. luymesi and S. jonesi, two species that often co-occur at seep sites in the northern GoM at depths < 1000m. Based on twelve polymorphic microsatellite markers, the authors identified a pattern of isolation by distance in L. luymesi and a genetic clustering fitting age of aggregation in S. jonesi. It is, however, unclear if vestimentiferan populations exhibit

28 similar population subdivisions at seeps > 1000m where the currents are even less understood.

To better understand the scale of population connectivity and the factors affecting gene flow in Escarpia, we ask 1) On a global scale, are the geographically separated Escarpia taxa as genetically similar as suggested by mitochondrial sequences, or do they form genetically distinct populations and 2) On a regional scale, is there evidence of genetic differentiation between populations at distinct isolated sites within the GoM and WACS?

To address the first question we used the classic mitochondrial DNA molecular markers mtCOI and mt16S rRNA, supplemented with another mitochondrial (mt) gene Cytochrome B

(CYTB), an Exon Priming Intron Crossing (EPIC) nuclear marker Hemoglobin subunit B2 intron

(HbB2i), as well as nine polymorphic microsatellite loci. Cytochrome B has been found to be phylogenetically informative in a wide variety of animal groups (Johns and Avise 1998). The utility of CYTB as an informative marker in deep-sea invertebrates has not been investigated in detail.

At the regional scale, the existence and pattern of effective larval dispersal versus genetic divergence was addressed by analyzing 11 microsatellites on samples collected from eight sampling sites in the GoM and 16 microsatellites on samples collected from three sampling sites at WACS. Because all E. spicata individuals were collected from a single site, a regional scale analysis was not conducted for E. spicata.

Materials and Methods

Sample collection and preparation

Tubeworms were collected via a manned submersible and remotely operated vehicles

(ROV) from seep sites during several research cruises occurring from 2003 to 2011. Escarpia

29 laminata individuals were collected from eight locations on the Lower Louisiana Slope in the

Gulf of Mexico using ROV Jason II on the National Oceanic and Atmospheric Administration

(NOAA) vessel Ronald Brown or the Deep Sea Vehicle (DSV) Alvin on the R/V Atlantis.

Escarpia southwardae individuals were retrieved from three pockmark locations near the Congo

River Canyon off of the west coast of Africa using ROV Victor 6000 on the N/O Pourquoi

Pas?, operated by IFREMER, France (Table 3-1 and Figures 3-3 and 3-4).

Provisional Seep Name Latitude Longitude Depth (m) N Region Species Escarpia spicata TF4 27.357° N 111.290° W 1784 20 GoC n = 20 AC601 26.392° N 94.514° W 2335 12 GoM

AC818 26.185° N 94.574° W 2744 8 GoM

WR269 26.677° N 91.665° W 1975 7 GoM

GC852 27.095° N 91.265° W 1437 11 GoM Escarpia laminata n = 129 AT340 27.644° N 88.365° W 2204 11 GoM

MC294 28.675° N 88.481° W 1800 8 GoM

DC673 28.516° N 87.518° W 2604 42 GoM

WFE 26.040° N 84.915° W 3304 30 GoM

Worm Hole 4.760° S 9.941° E 3189 25 WACS Escarpia southwardae Baboon 4.937° S 9.949° E 3128 28 WACS n = 80 Regab 5.798° S 9.711° E 3153 27 WACS

Escarpa spicata individuals were collected from the Transform Fault 4 (TF4) site in the Gulf of

California and provided by Robert Vrijenhoek of the Monterey Bay Aquarium Research Institute,

USA (see Vrijenhoek et al. 2007 for details on the collection localities). Sample preparation and

DNA extraction protocols were performed as described in Cowart et al. 2012. Tissue samples are

30 stored at -80°C and in 96% ethanol in Fisher Deep Sea lab at Penn State.

Figure 3-3: Gulf of Mexico (GoM) cold seep sites sampled for this study. 1000 meter contours from NASA-JPL Advanced Spaceborne Thermal Emission and Reflection Radiometer (http://www.geomapapp.org).

Sequencing and EPIC development

A 658bp fragment of mtCOI was amplified in 42 Escarpia individuals using primers

HCO and LCO (Folmer et al. 1994); a 435bp fragment of mt16S was amplified in 45 individuals using primers 16Sar and 16Sbr (Palumbi 1996), and a 401bp fragment of mtCYTB was amplified in 46 Escarpia and Seepiophila individuals using primers Cytbf and Cytbr (Boore and Brown

2000). Cytb primers contained many degenerate bases, and Escarpia specific CYTB primers were designed for this study. EPIC sequencing involves designing primers in the conserved exon regions of nuclear genes to sequence across the introns, and has been previously applied to clarify

31 population relations in both marine, freshwater and terrestrial animals (Palumbi 1996, Palumbi and Baker, 1994, Bierne et al. 2000, Schaeffer et al. 200, Li et al. 2010).

Figure 3-4: West African cold seep sites sampled for this study. 1000 meter contours from NASA-JPL Advanced Spaceborne Thermal Emission and Reflection Radiometer (http://www.geomapapp.org).

To amplify the intron region within the hemoglobin subunit B2, the HbB2i fragment was first identified in R. piscesae, by designing PCR primers based on the cDNA sequence of R. piscesae globin B2 (accession AY250083; Bailly et al. 2003). Cloning and sequencing of the resulting product revealed the 550bp intron sequence. These same primers were tested in R. pachyptila, and a similar analysis was conducted to obtain a 531bp sequence. The sequences obtained from R. piscesae and R. pachyptila were aligned, and primers were designed in the exons to flank the intron to produce a 665bp fragment in 48 Escarpia and Seepiophila

32 individuals.

PCR reactions and gel electrophoresis were performed as described in Miglietta et al.

2010, with annealing temperature for1.5min at 50 ºC (COI and 16S) or 52 ºC (CYTB and HbB2i).

Purified PCR products were submitted to the Penn State Genomics Core Facility (University

Park, PA) and run on 3730XL DNA sequencer.

Sequencing Analyses

Gene sequences were assembled and edited using Geneious Pro v5.5.5 (Biomatters Ltd.), and then aligned using ClustalW (Thompson et al. 1994) and MUSCLE (Edgar, 2004) implemented through the Geneious and MEGA 5.05 interfaces (Tamura et al. 2011). Both

ClustalW and MUSCLE gave identical alignments, and ClustalW alignments were used to generate trees. All alignments were edited and confirmed by eye to ensure that indel variation was scored consistently among taxa. Phylogenetic analyses of the alignments were conducted in

MEGA using the Maximum Likelihood (ML) (Tamura and Nei 1993) methods with 1000 bootstrap replicates. Between species distances were computed in MEGA using the Maximum

Composite Likelihood Method (Tamura et al. 2004) and are measured as the number of base substitutions per site. To test for the neutrality of the four genes, Tajima’s D (Tajima 1989) statistic was estimated using Arlequin v3.5 (Schneider and Lischer 2009). Median-joining haplotype networks were created using the program Network v4.611 (Bandelt et al., 1999, available at www.fluxus-engineering.com).

33 Microsatellite Analyses

Primers and amplification conditions for the 28 polymorphic microsatellite loci developed for E. laminata and E. southwardae are listed and described in Cowart et al. 2012 (see supporting information). Different alleles at a microsatellite locus may differ by the repeat length; however, a variety of genotyping errors can lead to single base pair offsets in allele size determinations (Pompanon et al. 2005). To correct for motif indel mutations, we calibrated the automated allele calls from fragment analysis software with actual repeat lengths by sequencing a mean of five homozygote individuals per locus to verify the length of alleles (GenBank

Accessions KC900290 - KC900365). When the apparent allele size was offset by less than one repeat, it was binned to the nearest confirmed size. This dataset was then used for all downstream analyses. When a heterozygote deficiency was observed, we tested for the presence of null alleles

(which can be particularly high in invertebrate populations, Chapuis and Estoup, 2007) using

INEst v.1.0 (Chybicki and Burczyk 2009), under the Individual Inbreeding Model.

Arlequin was used to test for significant departures from Hardy-Weinberg Equilibrium

(HWE) by computing observed and expected frequencies of heterozygosity. To adjust p-values for multiple comparisons tests, the False Discovery Rate control (FDR) (Benjamini and Yekutieli

2001) was implemented in QVALUE (Storey and Tibshirani 2003) executed in the statistical software R version 2.10.1 (R Development Core Team 2009). To estimate genetic differentiation, Wright’s F-statistics were computed with the aid of FSTAT v.2.9.3.2 (Wright

1951; Goudet 1995); these indices describe the probability that individuals from different sub- and total populations will share alleles identical by descent (Weir and Cockerham 1984). For the purposes of this study, individual refers to a single tubeworm, subpopulations are all individuals from specific seep location and total populations refer to all individuals from a particular region.

34 To compute allelic and private allele richness for each population controlled for sample size, the program HP-RARE (Kalinowski 2005) was used. To illustrate population structure within and between the three regions, the programs STRUCTURE v2.3.x (Pritchard et al. 2000,

Pearse and Crandall 2004) and EDENetworks v.2.16 (Kivelä et al. in prep) were used. For

STRUCTURE, independent allele frequencies and admixed populations were assumed; three replicate simulations were run using various values of K, with 100,000 Markov Chain Monte

Carlo (MCMC) repetitions for each cluster and burn-in of 10,000.

To assign individuals to populations while detecting the presence of hybrids as linking agents between clusters, networks based on the shared allele distance were built with

EDENetworks. EDENetworks illustrates the distribution of genetic distances (links) among agents (individuals or populations) of a system of populations, and infers the clustering of sets of individuals into distinct populations or species using percolation theory and without an a priori hypothesis based on their taxonomic identity or sampling location (Becheler et al. 2010, Moalic et al. 2011, Rozenfeld et al. 2007, Rozenfeld et al. 2008). The percolation theory allows for the splitting of a fully connected network into discrete clusters, and the critical threshold distance is known as the percolation threshold (Dpe), an inner property of the system (Stauffer and Aharony

1994, Watts, 2004). Analysis of Molecular Variance (AMOVA) (Excoffier et al. 1992) was calculated with the aid of GenoDive v.2.0b22 (Meirmans and Hedrick 2011) based on population clusters identified by STRUCTURE, if applicable. To test for signatures of bottlenecks or recent population expansion, the program BOTTLENECK (Piry et al. 1999) was run using the Two-

Phase Model and the Wilcoxon sign rank test at 1000 iterations.

35 Results

Mitochondrial and EPIC sequence analysis

The COI and 16S datasets included a total of 98 sequences, including 71 sequences of

Escarpia not previously reported, and 16 sequences from Miglietta et al. (2010). Seepiophila jonesi served as the outgroup taxon for both analyses. The complete COI dataset consisted of a

658bp fragment amplified in 42 Escarpia and six Seepiophila. The complete 16S dataset consisted of a 435bp fragment amplified in 45 Escarpia and five Seepiophila. All GenBank

Accession numbers are located in the supporting information (Table A3-2). As previously reported (McMullin et al. 2003, Miglietta et al 2010), COI and 16S phylogenies do not resolve the three described Escarpia spp. (Figures A3-1 and A3-2) due to the extremely low number of haplotypes (nine haplotypes for COI and only one for 16S). A haplotype analysis of COI is shown in Figure 3-5 and identifies the most common haplotype as shared among the three

!"#

$# !"#$% !"#&'( )*(+,

& " $

$'()*+,-

36 Escarpia (n=22), with rare haplotypes shared only within a group.

The CYTB dataset consisted of a 401bp fragment amplified in 46 Escarpia and S. jonesi.

The CYTB phylogeny identified two separate groups within Escarpia: one moderately supported clade that included E. laminata and E. spicata, from which E. southwardae is separated (Figure

A3-3). S. jonesi individuals form a highly supported cluster outside of the Escarpia group. The

CYTB haplotype network identifies a shared haplotype between E. laminata and E. spicata

(Figure 3-5), as well as three unique haplotypes for E. spicata. E. southwardae exhibited one distinct haplotype different from E. laminata/E. spicata by only one mutation.

The HbB2i dataset included 48 sequences from Escarpia and S. jonesi, and consisted of a

665bp fragment. The HbB2i phylogeny is in agreement with the separation of the described species of Escarpia, although with only moderate bootstrap support and low levels of nucleotide divergence (Figure A3-4). The HbB2i haplotype network also identifies geographically separated haplotypes (Figure 3-5). The rare haplotypes present at the three polymorphic genes are unique to a specific geographic region, with none of the rare haplotypes shared among Escarpia spp.

To provide information on the demographic history occurring among the Escarpia, we examined Tajima’s D for the three polymorphic genes tested for each group separately. None of the Tajima’s D tests showed values significantly departing from zero (Table A3-3). However, considering the low number of haplotypes and nucleotide divergence, one should cautiously accept the null hypothesis of neutrality and demographic stability given the lack of power provided by the dataset. Testing for population bottlenecks allowed us to detect of significant heterozygote excesses in all sites for E. southwardae and in one site for E. laminata (GC852) (see

Appendix B), supporting the occurrence of recent demographic expansions following establishment of these two populations.

37 Genetic differentiation between Escarpia groups

Mean summary statistics for the nine cross-species amplified microsatellite loci are detailed in Table 3-2. A total of 229 individuals were examined across all Escarpia; 222 alleles were detected across all loci with the largest number of alleles at locus EL454_9 (35), and the smallest at EL454_71 (8) (Table A3-4).

Table 3-2: Summary statistics for 28 microsatellite loci amplified in three described Escarpia species.

Escarpia taxa Summary Statistics (means across all loci) (9 loci) SE SE NA HO HE FIS AR AR PAR PAR E. spicata 8 0.690 0.780 0.194 4.750 0.376 2.71 0.447 E. laminata 13 0.681 0.753 0.149 4.870 0.490 3.24 0.668 E. southwardae 12 0.702 0.753 0.104 4.780 0.445 3.26 0.700

Escarpia laminata Summary Statistics (means across all loci) (11 loci) SE SE NA HO HE FIS AR AR PAR PAR AC601 7.3 0.615 0.686 0.171 5.120 0.747 0.260 0.102 AC818 5.1 0.602 0.693 0.191 4.460 0.560 0.300 0.104 WR269 5.7 0.658 0.725 0.182 5.260 0.688 0.230 0.094 GC852 6.7 0.626 0.682 0.151 5.100 0.598 0.400 0.132 AT340 6.6 0.680 0.720 0.083 4.990 0.594 0.330 0.108 MC388 6.4 0.670 0.773 0.179 5.160 0.629 0.540 0.175 DC673 10.8 0.645 0.702 0.186 5.250 0.664 0.380 0.106 WFE 9.7 0.612 0.702 0.194 5.250 0.644 0.420 0.126 Summary Statistics (means across all loci) Escarpia southwardae SE SE (16 loci) NA HO HE FIS AR AR PAR PAR

Regab 8.7 0.618 0.658 0.100 8.630 0.904 1.420 0.310 Wormhole 9.4 0.685 0.709 0.081 9.440 1.151 1.530 0.390 Baboon 9.2 0.636 0.664 0.149 8.930 1.101 1.680 0.520

Abbreviations: number of alleles observed across all loci (NA), mean observed (HO) and expected (HE) heterozygosity, mean Wright’s Inbreeding Coefficient (FIS), mean rarified allelic richness (AR) and SE SE standard error (AR ); private allele richness (PAR) and standard error (PAR ). Rarefied over 20 samples and means are not significantly different (P > 0.05).

38 For E. spicata, the mean number of alleles was eight, while average gene diversity spanned from 0.471 to 0.90, with a mean expected heterozygosity of 0.780. For E. laminata, the mean number of alleles was 13, while average gene diversity had a range of 0.343 to 0.921, with a mean expected heterozygosity of 0.753. E. southwardae had a mean number of alleles of 12, and gene diversity had a minimum of 0.262 and a maximum of 0.907 and a mean expected heterozygosity of 0.753. Of the 27 sampling location vs. locus combinations, three departed significantly from HWE after FDR correction at 0.01.

FIS values were calculated across all locations, as findings identified panmictic populations in each region (see results below). For E. spicata, the mean FIS value across all loci was 0.194, for E. laminata FIS is 0.149 and E. southwardae is 0.104 (Table 3-2). Mean allelic richness per site varied from 4.75 to 4.87, while private allelic richness ranged from 2.71 to 3.26.

There was no significant difference in the average allelic and private allelic richness between the three Escarpia species (p > 0.05).

Both STRUCTURE and network analyses support the occurrence of three clusters of

Escarpia (Figure 3-6), supporting the taxonomical hypothesis based on the morphological descriptions. For the network analyses, at Dpe = 1 and below, three clusters emerge with a significant clustering index (p<0.001). Just above Dpe = 1, some individual nodes are isolated, despite the connection of three clusters. Although clustering reveals strong genetic structure among the three taxa/regions, we note that below the percolation threshold, several individual genotypes are more closely related to genotypes different from the morphological and geographical groups to which they were assigned (Figure 3-6). Network analyses also identify the

GoM population as the central cluster, suggesting that colonization and diversification of the

Escarpia occurred through the GoM. FST values also identified significant differentiation between the three geographically distinct populations of Escarpia (Table A3-5). Additionally, AMOVA

39 analyses support that the three Escarpia populations are significantly different from one another

(p = 0.001, Table 3-3).

!"##

#"'#

#"&#

#"%#

#"$#

#"## E. spicata E. laminata E. southwardae

Dpe = 1.15 Dpe = 1.00 Dpe = 0.99

E. laminata E. laminata E. laminata

E. southwardae E. southwardae E. southwardae E. spicata E. spicata E. spicata

Figure 3-6: Top: STRUCTURE results (admixture model, three replicate runs) for three putative species of Escarpia vestimentiferan tubeworms based on nine microsatellite markers. Each vertical bar represents an individual tubeworm. The y-axis is the proportion of each individual’s genotype belonging to a distinct population cluster. Bottom: Network topologies of E. laminata (red), E. spicata (green) and E. southwardae (blue) individuals with the Shared Allele Distance (SAD) based on nine microsatellite markers. Clusters are also labeled. Only links with value smaller than or equal to the percolation distances are present. Nodes (circles) represent individuals. Three clusters are identified, one for each of the described Escarpia species and their respective geographic regions.

Escarpia laminata population genetics

Eleven microsatellite loci, which include three additional species-specific loci, were examined in

129 E. laminata individuals from eight sites in the GoM. A total of 167 alleles were detected across all sites where the EL454_57 locus had the greatest number of alleles at 19 (Table A3-7).

The MC294 site had the highest mean expected heterozygosity at 0.773, while WR269 exhibited the lowest at 0.658. While two sites at locus EL454_9 departed significantly from HWE, no locus

40 significantly deviated from HWE across all sites. The global FIS was 0.103, while the mean FIS values for each site ranged from 0.083 at AT340 to 0.194 at WFE.

Table 3-3: Summary of Analysis of Molecular Variance (AMOVA) conducted for each study under the Infinite Allele Model, FST. Significance (*) was tested using 1000 permutations. Results for Stepwise Mutation Model (RST) are congruent with IAM. Abbreviations: d.f, degrees of freedom; SS, sum of squares.

Variation Study Source of Variation d.f. SS p-value (%)

Within individuals 229 718.50 80.60 ---

Escarpia Question Among individuals 226 838.40 7.30 0.001* (Three species)

Among populations 2 126.80 12.10 0.001*

Within individuals 129 499.00 90.70 ---

Escarpia laminata Among individuals 121 558.01 8.90 0.001*

Among populations 7 37.12 0.05 0.053

Within individuals 80 411.50 94.80 ---

Escarpia southwardae Among individuals 77 437.29 0.49 0.002*

Among populations 2 12.82 0.03 0.149

Mean allelic richness per site spanned from 4.60 to 5.54, while private allelic richness ranged from 0.22 to 0.49 and there was no significant difference in the average allelic and private allelic richness between the eight sites (p > 0.05).

STRUCTURE analyses did not detect any genetic clustering on a regional scale among the eight sites in the GoM. Furthermore, none of the 28 pairwise population comparisons were significantly different after correction for multiple tests (p > 0.05). The global FST value was low

(0.007) with a p-value of 0.053, suggesting very limited structure on the regional scale (Table A3-

8). AMOVA testing identified a significant among individual component, for which 8.9% of the

41 total variance is explained (p = 0.001) (Table 3-3). Maximum genetic differentiation was observed between AC601 and MC294, sites that are separated by a distance of 657km, although the maximum distance between sites tested here is 980km. These findings indicate that gene flow is not correlated with geographic distance on the regional scale in the GoM.

Escarpia southwardae population genetics

Sixteen microsatellite loci were examined in 80 E. southwardae individuals from three sampling sites at WACS. A total of 223 alleles were detected across the three sites, and locus

ES454_31 exhibited the highest number of alleles at 19 (Table A3-9). The Worm Hole site exhibited the highest mean expected heterozygosity at 0.709, while Regab had the lowest at

0.658. Three sites departed from HWE at loci ES454_32 and 45. The global FIS was 0.104, while the mean FIS values for each population ranged from 0.081 at Worm Hole to 0.149 at Baboon.

Mean allelic richness per site spanned from 8.63 to 9.44, while private allelic richness ranged from 1.42 to 1.68, and there was no significant difference in the average allelic and private allelic richness among the three locations. None of the pairwise comparisons showed significant differentiation among populations; the global FST value was low (0.003) and we did not detect any difference between the three populations studied here (p = 0.149; Table A3-10).

As with the E. laminata, STRUCTURE analyses of E. southwardae did not detect any genetic clustering among the three sites at WACS. AMOVA testing identified a significant among individual component, for which as little as 0.49 % of the variance is explained (p =

0.002).

42 Discussion

Results obtained in this study show that despite the isolated and haphazard distribution of cold seep sites, there is panmixia over extensive regional scales. In contrast, we find only limited indications of gene flow among the three regions studied. These findings confirm the divergence of tubeworm populations between the three regions, and suggest the presence of mechanisms that prevent or severely limit gene flow across these spatial scales. However, in the absence of either experimental data confirming reproductive isolation or evidence of maintenance of genetic isolation in co-occurring populations, it is not possible to confirm that these differentiated populations represent biologically distinct species.

Barriers to gene flow across large scales

Although mitochondrial DNA is often used in the taxonomic identification of animals, neither the “barcoding marker” COI, nor the16S rRNA gene, resolved the relationships among the three described species of Escarpia. Furthermore, the mtCYTB gene could only resolve the

Escarpia into two groups, rather than three. The HbB2i locus, together with microsatellites, supports the existence of three differentiated groups, corresponding to the taxonomic hypothesis based on morphological and geographic criteria.

The fixed haplotypes and low diversities at mitochondrial and nuclear loci are an indication that strong demographic fluctuations occurred in the recent past and reduced the level of polymorphism. The lack of allele frequency variation in mtDNA among geographically separate populations could result from a number of factors. First, there may be insufficient time since the separation of populations. Second, as evolutionary rate is influenced by species generation time, populations with longer generation times may accumulate mutations at a slower

43 rate (Hartl and Clark, 2007). As seep vestimentiferans can live for hundreds of years (Bergquist et al. 2000), their long generation time could contribute to the lack of mutations seen in mtDNA.

Third, the low mutation rates may be the result of vestimentiferans having efficient mutation repair systems. Additionally, the exclusively maternal transmission of mtDNA reflects on the distribution of maternal lineages rather than whole genome evolution, and possible differences between male and female effective population sizes may induce different rates of evolution of mitochondrial versus nuclear genomes. Finally, selective sweeps may interfere with the usual mutation-drift balance and influence mitochondrial evolution, as organelles have smaller effective population sizes and are more prone to genetic drift (Bazin et al. 2006, Galtier et al. 2009, Moalic et al. 2012). Whether these populations are the result of recent founder events followed by genetic drift and ultimately differentiation, or from a more ancient colonization that was followed by divergence among protospecies or species remains unknown. The lack of shared rare haplotypes at the mitochondrial loci is consistent with either. Regardless, gene flow is clearly very limited between the regions inhabited by the described Escarpia spp.

Vestimentiferans are thought to have arisen around 100 million years ago (Black et al.

1997). As recently as 3.5 million years ago, deep-water flows connecting the Pacific and Atlantic oceans were disrupted by the formation of the Isthmus of Panama as the North American, South

American and Caribbean plates converged (Keigwin 1978, Knowlton 1993). Additionally, the deep water that flows from the Caribbean Sea into the GoM is trapped by geologic sills occurring below 800m in the Florida Straits and below 1900m in the Yucatan Channel, which may help to further isolate biological populations in the GoM from other parts of the Atlantic (Welsh and

Inoue 2000, Rivas et al. 2005). All three Escarpia populations experience similar habitat conditions, including the high pressure that accompanies extreme depth, as well as elevated concentrations of hydrogen sulfide required by their chemoautotrophic symbionts (Childress and

Fisher 1992). Therefore, the genetic divergence and morphological variation reported here are

44 likely not driven by habitat specialization, although, whether differences in morphology are influenced by phenotype plasticity or selective pressures remains to be investigated.

The relatively low levels of genetic differentiation reported, and the presence of rare haplotypes that are unshared between described species, supports the existence of a recent common ancestor, and possibly recent colonization events into different ocean basins. The genetic differentiation between the three Escarpia is possibly the result of a genetic drift and lack of gene exchange due to geographic and reproductive isolation. The analyses here indicate that contemporary migration between the three provinces, if occurring at all, is low. Species delimitation is difficult to demonstrate with geographically separated taxa (de Queiroz 2007). The best method to confirm the presence of reproductive barriers among potentially distinct species would be controlled crosses, which at this time are technically difficult to conduct with these deep sea taxa. However, based on the congruence of morphological differences with data from a nuclear gene and other polymorphic loci, as well as the large distances separating the described species, we suggest that the mitochondrial genetic data should not be used to suggest that the three described species are in fact a single species. We do caution that in the absence of data on reproductive isolation it is also possible that the three described species represent either separate populations of the same species with scarce and/or highly sporadic gene flow or protospecies undergoing strict divergence and on the way to speciation.

On a regional scale, the evidence for panmixia suggests that despite the occurrence of isolated seep sites, Escarpia do have the ability to connect across geographic scales of nearly

1000k in the GoM and at least 120 km at WACS. The low divergence seen across sampling locations suggests that contemporary gene flow is occurring at the regional scale for both GoM and WACS.

45 Influence of deep currents on dispersal and recruitment of Escarpia

In addition to constraints imposed by larval biology, successful dispersal of marine larvae depends on movement with ocean currents (Scheltema 1968, Johnson and Black 1984). The moderately high dispersal potential of vestimentiferan larvae, coupled with deep currents, can account for the open exchange of migrants seen within the GoM and WACS regions. Currents below 1000m are often quite complex, moving in various directions and speeds (Hamilton, 1990).

In the GoM, the Loop Current (LC) dominates the east, and while restricted to < 1000m depth, the LC sheds warm eddies that generate and influence deep Topographic Rossby Waves (TRW) at depths below 1000m (Hamilton 1990). Swirl speeds of deep eddies are estimated to range from

10 to 21 cm s-1 at 2550m, increasing speed with more complex topography (Welsh and Inoue

2000), and TRW are estimated to occasionally move at speeds of 15 – 30cm/s-1, having durations from 20 – 30 days and dominating the central and western GoM (Chelton and Shelax 1996, Oye and Lee 2002, Hamilton 2007).

Thermal tolerances of deep sea larvae also play an important role in dispersal, as larvae moving into shallower waters and thus warmer temperatures can experience changes in metabolism and feeding rates, influencing the amount of time they spend in the water column

(Young et al 1996, Young et al. 1998). McMullin et al. (2010) and Young et al. (2012) found that most larvae of the shallower living GoM tubeworm, Lamellibrachia luymesi, are retained in the same general geographic location as the adults, not dispersing > 300km from the source populations. Additionally, L. luymesi are unable to tolerate the temperatures at depths above the thermocline, further restricting the propagules to the region.

The speeds of deep currents, accompanied by prolonged larval duration, as well as the lower temperatures needed for larval development, appear to be sufficient to transport Escarpia migrants to habitats across the GoM sea floor at depths below 1000m. Additionally, there are

46 more than 50 known chemosynthetic sites distributed across the GoM (Mineral Management

Service, Gulf of Mexico 2006). This along with model data of closely related species suggests that GoM Escarpia do have the ability to colonize seeps across the regional distance described in this study.

A major feature of the West African continental margin is the Congo deep-sea river fan, which results from loads of sediment transported and deposited by the Congo River (Khripounoff et al. 2003, Sibuet and Vangriesheim 2009). The submarine fan ranges in depth from 3400 to

4800m and contributed to the formation of the cold seep habitats that are home to E. southwardae

(Vangriesheim et al. 2009). Deep-sea circulation at the Congo River fan is complex, and below

2000m, and currents have a southeastward migration path (Stramma and England 1999). At

Regab, current speeds range from 8.9 to 12.3 cm s-1, rarely exceeding 10 cm s-1 (Vangriesheim et al. 2009). Despite the relatively slow movement of currents in this region, the three sampling sites encompass a distance of less than 150 km, and if larval characteristics are comparable to other seep vestimentiferans, gene flow across these sites is realistic.

Conclusions

In this study, we used the analyses of several molecular genetic markers to support the occurrence of three genetically distinct entities that correspond to three Escarpia species described based on morphology from the Gulf of Mexico, the eastern Atlantic, and the eastern

Pacific. This data suggests limitations to large-scale dispersal by Escarpia larvae and are in agreement with a scenario of ongoing divergence after relatively recent colonization in the three regions. However, the widespread genetic homogeneity observed within regions over hundreds of kilometers also support recurrent events of moderate scale dispersal. Finally, the lack of resolution seen in three mtDNA loci contrasts with the lineage and genotype sorting revealed by

47 the nuclear datasets (one intron and several microsatellite loci). This suggests that EPIC and microsatellite markers can serve as tools to not only examine population structure within groups, but also to clarify boundaries and distributions of species with very low levels of divergence in classic bar coding genes.

Acknowledgements

We would like to thank the following people for their contribution to this project: captains, crews and expedition leaders of the ROVs Jason II and Victor 6000, the US deep submergence facility, DSV Alvin, R/V Atlantis, N/O Pourquoi Pas?, the NOAA vessel Ronald

Brown; Robert Vrijenhoek and his laboratory at the Monterey Bay Aquarium Research Institute,

Kenneth Halanych, Drew Wham, Karine Olu – Le Roy (as the chief of the WACS oceanographic cruise), Kimberlyn Nelson, Iliana Baums, Todd LaJeunesse, Ann Andersen, Stéphane Hourdez,

Pen-yuan Hsing, Annelies De Groote, Olivier Soubigou, Miles Saunders, Andrew Mendelson,

Howard Fescemyer, Nicholas Polato, Meghann Durante and Andre Wallace. This research was funded in part by the National Science Foundation (Award # 1209688), the National

Oceanographic Partnership Program (NOPP) through support from the Bureau of Ocean Energy

Management contracts #M05PC00018 and #M08PC20038 (TDI Brooks International Prime), the

National Oceanic and Atmospheric Administration’s Office of Ocean Exploration and Research

(NOAA OER), Penn State Eberly College of Science, and an Alfred P. Sloan Scholarship to

DAC. Any opinions, findings, conclusions, or recommendations expressed in this article are those of the authors and do not necessarily reflect the views of the National Science Foundation.

CHAPTER 4

Depth-dependent gene flow in Gulf of Mexico cold seep Lamellibrachia

tubeworms (Annelida, Siboglinidae)

In preparation for submission to Deep Sea Research with

Dominique A. Cowart1, Kenneth M. Halanych2, Stephen W. Schaeffer1, Charles R. Fisher1

1Department of Biology The Pennsylvania State University 208 Erwin W. Mueller Laboratory University Park, PA 16801 - USA

2Department of Biological Sciences Auburn University 101 Rouse Life Sciences Building Auburn, AL 36849 - USA

Abstract

Vestimentiferan tubeworms belonging to the genus Lamellibrachia form aggregations at hydrocarbon cold seeps across the Louisiana Slope in the Gulf of Mexico (GoM), and create biogenic structures that provide living space for a variety of other fauna. In the GoM, there are three Lamellibrachia taxa that vary in morphology and depth ranges: L. luymesi (300 to 950m), L. sp. 1 (950 – 2604m) and L. sp. 2 (1175 – 3304m). While L. sp. 2 is consistently identified as a separate species, L. luymesi and L. sp. 1 cannot be discriminated through use of commonly employed bar coding markers, mitochondrial Cytochrome Oxidase subunit 1 (COI) or large ribosomal subunit rDNA (16S). To determine if gene flow is limited between these two taxa, we employed several more faster evolving genetic markers, including mitochondrial Cytochrome B

(CYTB), Exon Priming Intron Crossing (EPIC) nuclear marker Hemoglobin subunit B2 intron

(HbB2i), and eight polymorphic microsatellite loci amplified across 45 L. luymesi and L. sp.1 individuals. Additionally, we used microsatellites to investigate population structure of L. sp. 1 and L. sp. 2 (n = 55) in the deep GoM. Despite a lack of resolution seen with CYTB and HbB2i, we find that L. luymesi and L. sp. 1 form distinct clusters that are genetically differentiated at the eight cross-amplified microsatellite loci. Furthermore, we find that both L. sp. 1 and L. sp. 2 form single populations in the Gulf of Mexico.

50 Introduction

Clarifying mechanisms that drive speciation in the marine environment can be challenging, as few absolute barriers to gene flow exist in the ocean; regions separated by hundreds of kilometers can be connected genetically, while populations in close proximity can be genetically distinct (Palumbi 1994). Abiotic factors such as light intensity, temperature, hydrostatic pressure, salinity, nutrient input, sediment type and ocean currents therefore play a significant role in the formation of species in the sea, as variations in these conditions can drive ecological differentiation in populations (Gage and Tyler 1992, Schluter and Nagel 1995).

For a species to colonize deep sea habitats, appropriate life stages must have the physiological tolerance to cope with the lack of light, low temperatures, low food density, and high pressure (Carney 2005, Aquino-Souza et al. 2008). Below 500m, ambient light is absent and salinity is fairly constant (Bruun 1957, Somero 1992a, Lalli and Parsons 1993). However, temperature can vary from about 2ºC below 2000m to 8 to 10ºC at 500m (Bruun 1957, Carney

2005) and hydrostatic pressure increases continuously with depth (Saunders and Fofonoff 1976).

Changes in temperature and pressure that are linked to increasing depth influence a variety of physiological and biochemical processes in animals; the disruption of protein structures is one of the many physiological effects associated with changes in pressure (Somero 1992b, Oliphant et al. 2011).

Due to limits in accessibility and sampling, factors influencing speciation in deep-sea benthic animals are relatively poorly known, however, there are some studies that explore this topic. Etter and Rex (1990) illustrated that regardless of developmental mode, deep dwelling snail species in the western North Atlantic occupy broader depth ranges when compared to shallow species, suggesting that biotic and physical constraints lessen with increasing depth. Howell et al.

(2004) found that three reproductively isolated morphotypes of the Atlantic seastar Zoroaster

51 fulgens live within a depth range > 3,000 m and are stratified with respect to depth. More recently, Rex and Etter (2010) suggest that many marine species defined by morphological criteria are thought to have broad bathymetric ranges, when they may actually be composed of sibling species separated by depth. Additionally, as larvae of deeper dwelling organisms may not have the capacity to tolerate warmer temperatures and lower pressure levels, this also can influence larval colonization patterns, adult distributions, and ultimately diversity and speciation in the deep sea (Tyler and Young 1998, Pradillon and Gaill 2007).

Cold seep tubeworms in the genus Lamellibrachia

Cold hydrocarbon seeps are located on deep passive oceanic margins and are dominated by endemic species that directly or indirectly obtain their nutrition from chemosynthetic primary production (Sibuet and Olu 1998). Vestimentiferan tubeworms (Siboglindae) are notable because they lack a digestive tract and are dependent on endosymbiotic bacteria for their nutrients

(Childress and Fisher 1992). At cold seeps, these tubeworms cluster in aggregations that have patchy distributions on the seafloor; these aggregations are only present when seepage is high enough to support sustained chemosynthesis (Brooks et al. 1987). Furthermore, seep tubeworm aggregations serve as foundation species that provide living space and protection for many other animals (Bergquist et al. 2003).

The vestimentiferan genus Lamellibrachia contains at least six described species that are distributed along the western Pacific (L. satsuma, L. juni, L. columna), eastern Pacific (L. barhami), the Mediterranean Sea (L. anaximandri) and the Gulf of Mexico (L. luymesi) (Webb

1969, Southward 1991, Miura et al. 1997, McMullin et al. 2003, Miura and Kojima 2006,

Southward et al. 2011). Lamellibrachia from the Gulf of Mexico (GoM) are some of the most extensively studied seep tubeworms, as these seep communities were among the first to be

52 discovered in the 1980s (Kennicutt et al. 1988, Aharon 1994, Cordes et al. 2007). Lamellibrachia in the GoM are distributed across the Louisiana Slope, which ranges from ~300m depth at the continental shelf edge to ~3,300m at the salt deformation edge of the Sigsbee Escarpment

(Cordes et al. 2009). The most commonly studied seep tubeworm, Lamellibrachia luymesi (van der Land and N∅rrevang, 1975), is found on the upper Louisiana Slope from about 300 to 950m.

Two morphologically distinct groups, including Lamellibrachia sp. 1 and the currently undescribed Lamellibrachia sp. 2, are distributed across the lower Louisiana slope (> 1000m), as well as at the base of the Florida Escarpment. L. sp. 1 ranges from ~950 – 2604m, while L. sp. 2, which is the most rarely observed of the three GoM Lamellibrachia, ranges from 1175 – 3304m

(Miglietta et al. 2010).

The three GoM Lamellibrachia taxa differ morphologically (Jones, 1985). Members of each group have gill lamellae, which are used for gas exchange and are a major feature that distinguishes Lamellibrachia from Escarpia tubeworms (Jones 1985). However, each

Lamellibrachia taxon differs in the numbers of gill lamellae as well as in the size and length of the vestimentum. L. luymesi has 15 to 22 gill lamellae, which is fewer than other GoM

Lamellibrachia, and a slimmer vestimentum. Both L. sp. 1 and L. sp. 2 have about 21 to 27 gill lamellae, and L. sp. 1 has a characteristically fat and short vestimentum when compared to its body length. Finally, L. sp. 2 and lacks a ventral vestimental fold that is present in L. sp. 1 (Jones

1985; Miglietta et al. 2010).

Genetics of Gulf of Mexico Lamellibrachia spp

Initial investigations employing two mitochondrial genes, Cytochrome Oxidase subunit 1

(COI) and large ribosomal subunit gene (16S rDNA), clearly distinguished L. sp. 2 from the other

GoM Lamellibrachia, but could not distinguish L. luymesi and L. sp. 1 (Miglietta et al. 2010).

53 Despite COI’s use as a “barcoding” molecule (Hebert et al. 2003), it as well as the 16S gene, have slow rates of change in seep vestimentiferans, and do not differentiate some other vestimentiferan taxa that are distinct species (McMullin et al. 2003; Cowart et al. 2013). Therefore, the lack of differentiation reflected in these genes may not indicate that L. luymesi and L. sp. 1 are the same species.

Biological communities living at cold seeps worldwide are often inhabited by the same families of organisms; including the siboglinid polychaetes, bathymodioline mussels and vesicomyid bivalves (Carney 1994, Sibuet and Olu 1998). At seep sites in the GoM, there is a documented species “replacement” at depths between 800 – 1000m. Although the taxa are related at the family level, most species present below ~1000m are different from those at seeps above

1000m (Carney 2005, Cordes et al. 2007). Species replacement with depth has been documented in marine ecosystems all over the world, and is often a result of differences in light or food availability, sedimentation levels, temperature, pressure and salinity, all of which play an important role in species zonation patterns throughout the ocean (MacArthur 1972; Brown 1984).

At cold seeps in the GoM, one notable exception to the documented species replacement pattern is the seep mussel Bathymodiolus childressi, which has a known range of at least 1600m (Cordes et al. 2007). To determine if depth acts as a boundary that limits gene flow between

Lamellibrachia tubeworms in the GoM, we asked whether L. luymesi and L. sp. 1 are a genetically undifferentiated population, as suggested by current genetic data, or two genetically distinct populations, as suggested by their morphology. To address this question, we employed the mitochondrial marker Cytochrome B (CYTB), Exon Priming Intron Crossing (EPIC) nuclear marker Hemoglobin subunit B2 intron (HbB2i), and eight polymorphic microsatellite loci that were amplified across 45 L. luymesi and L. sp.1 individuals. We hypothesize that L. luymesi and

L. sp. 1 will form genetically distinct clusters (“genotypic clusters”, Mallet 1995), as this follows the same general pattern of species replacement seen in other GoM seep taxa. Alternatively, if the

54 Lamellibrachia are genetically undifferentiated, this suggests that L. luymesi and L. sp.1 are a single polymorphic taxon inhabiting a depth range of >2304m in the GoM.

Finally, we provide data on population structure of Lamellibrachia species living at seeps below 1000m in the GoM. Previous studies have employed polymorphic microsatellite markers to determine if genetic structure exists within seep vestimentiferans L. luymesi, Seepiophila jonesi,

Escarpia laminata and E. southwardae living at GoM and West African cold seeps (McMullin et al. 2010, Cowart et al. 2013). Here, we used Illumina sequencing to develop eight microsatellites for L. sp. 1 and five microsatellites for L. sp. 2 to identify geographic scales of effective larval dispersal in each group.

Materials and Methods

Sample collection and preparation

Tubeworms were collected via two manned submersibles and one remotely operated vehicle (ROV) from 13 hydrocarbon seep sites during research cruises that occurred between

1995 and 2010 (Figure 4-1, Table 4-1). L. luymesi individuals were collected from four locations across the upper Louisiana Slope using the Deep Sea Vehicle (DSV) Johnson Sea Link II on the

Research Vessel (R/V) Seward Johnson, operated by Harbor Branch Oceanographic Institution.

L. sp. 1 and L. sp. 2 individuals were collected from eight and six locations, respectively, across the lower Louisiana Slope using ROV Jason II on the National Oceanic and Atmospheric

Administration (NOAA) vessel Ronald Brown and the DSV Alvin on the R/V Atlantis (Table 4-

1). A total of 76 individuals were used in this study. Sample preparation and DNA extraction protocols were performed as described in Cowart et al. (2012).

GoM Taxa Seep Name Latitude Longitude Depth (m) N Region

GC234 27.747° N 91.224° W 550 9 West Lamellibrachia GC184 27.782° N 91.508° W 540 2 West luymesi N = 21 MC751 28.063° N 89.707° W 627 8 East VK826 29.157° N 88.019° W 446 2 East GC852 27.095° N 91.265° W 1437 6 West WR269 26.677° N 91.665° W 1975 3 West AC601 26.392° N 94.514° W 2335 2 West

Lamellibrachia sp. 1 GC600 27.374° N 90.573° W 1193 4 West N = 24 GB697 27.312° N 92.638° W 1281 3 West MC294 28.675° N 88.481° W 1385 3 East MC344 28.633° N 88.169° W 1800 1 East DC673 28.516° N 87.518° W 2604 2 East GC852 27.095° N 91.265° W 1437 3 West WR269 26.677° N 91.665° W 1975 1 West

Lamellibrachia sp. 2 MC294 28.675° N 88.481° W 1385 14 East N = 31 MC344 28.633° N 88.169° W 1800 9 East DC673 28.516° N 87.518° W 2604 2 East WFE 26.040° N 84.915° W 3304 2 East

!"##"##"$$"%&'()*(

Figure 4-1: Gulf of Mexico (GoM) cold seep sites sampled for this study. 1000 meter contours from NASA-JPL Advanced Spaceborne Thermal Emission and Reflection Radiometer (http://www.geomapapp.org). Lamellibrachia luymesi = dark grey circles, Lamellibrachia sp. 1 = grey circles and Lamellibrachia sp. 2 = white circles.

Marker development and sequencing analyses

Cytochrome B primers specific to Lamellibrachia were developed by using Escarpia

CYTB primers to amplify a partial fragment in Lamellibrachia (see Cowart et al. 2013 for

Escarpia primers). From the sequences amplified and aligned, Lamellibrachia specific primers were then designed. The Hemoglobin subunit B2 intron (HbB2i) region in L. luymesi and L.sp.1 was amplified by aligning sequences obtained from Escarpia laminata (Accessions KC357398 –

KC357424); primers were designed within the exons to flank the first intron in Lamellibrachia.

All PCR reactions and gel electrophoresis were performed as described in Miglietta et al. 2010,

57 with annealing occurring for 1.5 min at 51ºC for CYTB and 54ºC HbB2i. PCR products were purified using the ExoSAP-IT® protocol (USB, Affymetrix); purified PCR products were then submitted to the Penn State Genomics Core Facility (University Park, PA) and run on 3730XL

DNA sequencer. Sequences were assembled and conflicts in the reads from the two strands were edited using DNASTAR package 8.1 (Lasergene, Madison, WI, USA) and Geneious Pro v5.5.5

(Biomatters Ltd.). Alignments were done using ClustalW (Thompson et al. 1994) and MUSCLE

(Edgar 2004) implemented through MEGA 5.05 (www.megasoftware.net, Tamura et al. 2011).

Phylogenetic analyses and between species distances were estimated with MEGA using the

Maximum Likelihood (ML) method with 1000 bootstrap replicates (Tamura and Nei 1993,

Tamura et al. 2004). To test for departures from selective neutrality of the genes, Tajima’s (1989)

D statistic was estimated using Arlequin v3.5 (Schneider and Lischer 2009). Finally, median- joining haplotype networks were created using the program Network v4.611 (Bandelt et al., 1999, available at www.fluxus-engineering.com).

Microsatellite development and analyses

Microsatellite loci used in this study were developed previously by McMullin and co- workers (2004) or with the aid of a transcriptome library produced from one adult L. luymesi.

RNA extraction was done using the RNeasy kit (Qiagen) with on-column DNase digestion, following the manufacturer's protocol. RNA was quantified with a Qubit fluorometer (Invitrogen) and evaluated on a 1% SB agarose gel. One µg of total RNA was used to synthesize first-strand cDNA using the SMART cDNA library construction kit (Clonetech) following the manufacturer's instructions except that the kit's 3' oligo was replaced with Cap-Trsa-CV oligonucleotide

(5’AAGCAGTGGTATCAACGCAGAGTCGCAGTCGGTACTTTTTTCTTTTTTV-3’) as per

Kocot et al. 2011. The Advantage 2 PCR system (Clonetech) was used to amplify full-length

58 cDNA. As few PCR cycles (usually 17-19 rounds) as possible were used during amplification.

Samples were sent to Hudson Alpha Institute for Biotechnology (Huntsville, AL) for sequencing.

Sequencing used the Illumina TrueSeq paired-end (PE) library preparation protocol (following manufacturer's protocol). Samples were run as 1/6 of a lane (i.e. - 6 samples per lane) on an

Illumina Hi-Seq 2500 using 2x100bp PE chemistry. Truseq was used for insert sizes, which were a few to several hundred base pairs in length.

To identify short tandem repeats, L. luymesi sequence reads were uploaded and searched using the Tandem Database v2.30 program (Gelfand et al. 2006); potential primer pairs for the repeats were identified using Primer 3 v.0.4.0 (Rozen and Skaletsky, 1999). Primers having lengths between 18 – 27bp, GC content ≥50%, and melting temperature ≥50ºC were chosen because they had a high likelihood of amplifying the fragment. Each marker was tested for variability using the M13-tagging approach with DNA from several individuals (Schuelke, 2000, see Cowart et al. 2012 for details on primer screening and fragment analyses). From the initial screening phase, eighteen potential primer sets were identified, and ten of these sets amplified across the target Lamellibrachia species. Additional screening found that seven of the ten primer sets were polymorphic and were used in downstream analyses (Table 4-2).

To correct for genotyping errors and calibrate automated allele calls, we used a binning process detailed in Cowart et al. (2013). The programs GENEPOP v4.1 (Raymond and Rousset

1995), MicroChecker v2.2.3 (van Oosterhout et al. 2004) and INEst v.1.0 (Chybicki and Burczyk

2009, Individual Inbreeding Model) were used to test for linkage disequilibrium and screen for the presence of null alleles. Departures from Hardy-Weinberg Equilibrium (HWE), observed and expected frequencies of heterozygosity were computed with the aid of Arlequin v3.5. We used the False Discovery Rate approach (FDR) to correct for multiple comparisons (Benjamini and

Yekutieli 2001). FDR was computed using the Q-value (Storey and Tibshirani 2003), which was implemented in the statistical software R version 2.10.1 (R Development Core Team 2009).

59 Wright’s F-statistics were estimated using the program FSTAT v.2.9.3.2 (Wright 1951, Goudet

1995) to estimate genetic differentiation among populations. For the computation of allelic

richness at each sampling location, controlled for sample size, the program HP-RARE

(Kalinowski 2005) was used.

Table 4-2: Characteristics of eight microsatellite loci developed for Lamellbrachia

o Locus Primers (5’ – 3’) Repeat Motif Size Range (bp) Tm ( C) Source b L1-2E#2 ª F:GGCAATTGTTGAGGACGTGT (CAC/AG)12 341 – 359 62 McMullin et al. 2004 R:GGAAGTGAACCAATGCTCTG b L454_6ª F:GAACGGACTCGCACGTAT (GCAC)11 128 – 248 55 This study R:ACAGCCAGGATAGGAGTCG

L454_11ª F: GGGAGGTCTTGAGAGCAGAT (GCAC)14 123 – 278 54 This study R: CTATTCCAACATTGCACACCT

L454_13ª F: CATCATCGCTACCATCATCA (CAA)20 149 – 284 54 This study R: ACCTTCCAAGCATGCGGT b L454_19ª F: CTGCCCAAGCAGACATTTAT (CAG)15 103 – 151 54 This study R: CGCCAGGTACAGGTATTGTC b L454_20ª F: GCTGATGTTGATGTTGATGC (CGA)13 93 – 117 54 This study R: AAGTTCAACTGCAGGGTCG !

L454_25ª F: ATTTCGCCGTCAACTACAAA (TTTGT)11 102 – 214 54 This study R: AAACCCAAGCTGAGAGCAC ! b L454_32ª F: CACTTTCAAGCATGCGGT (TTG)19 149 – 179 53 This study R: CATCATCGCTACCATCATCA

ªLoci amplified in Lamellibrachia sp. 1 bLoci amplified in Lamellibrachia sp. 2 Tm =Annealing temperature

We used STRUCTURE v2.3.x (Pritchard et al. 2000) and EDENetworks v.2.16 (Kivelä

et al. in prep) to determine if genetic structure was present among Lamellibrachia populations.

For STRUCTURE, independent allele frequencies and admixed populations were assumed, and

three replicate simulations were run using various values for K, or number of populations, with

100,000 Markov Chain Monte Carlo repetitions for each cluster and a burn-in of 10,000 cycles.

EDENetworks assigns individuals to population clusters while screening for the presence of

hybrids based on both the Shared Allele Distance and the Rozenfeld Distance (Rozenfeld et al.

2007, Rozenfeld et al. 2008). The Rozenfeld distance resolves ancestral polymorphism, and the

60 Shared Allele Distance illustrates recent gene flow (Moalic et al. 2011). Resulting networks define genetic distances (links) to assign individuals (nodes) to their “species” of origin without an a priori hypothesis. The percolation threshold, an inner property of EDENetworks, is the level by which a fully connected network is split into discrete clusters (Stauffer and Aharony 1994,

Watts 2004). To identify the presence of genetic structuring at the individual, subpopulation and total population scales, we calculated Analysis of Molecular Variance (AMOVA) statistics

(Excoffier et al. 1992) with the aid of GenoDive v.2.0b22 (Meirmans and Hedrick 2011).

Results

Mitochondrial and EPIC sequencing analyses

The complete CYTB dataset consisted of a 380bp fragment amplified in 31 L. luymesi/L. sp. 1 and six Escarpia individuals (Accessions KF201512 – KF201542). Escarpia laminata served as the outgroup taxon for all phylogenetic analyses in this study. Similar to previously tested mitochondrial genes COI and 16S, CYTB does not resolve L. luymesi and L. sp. 1 (Figure

A4-1; see Appendix C). A haplotype network of CYTB (Figure 4-2) identifies nine haplotypes; the two most common haplotypes are shared between both L. luymesi and L. sp. 1 and differ by only two mutational changes. This dataset also identifies several rare haplotypes, none of which are shared between the two Lamellibrachia. The 668bp HbB2i locus was amplified in 30 L. luymesi/L. sp. 1 and six Escarpia individuals (Accessions KF201543 - KF201572). As with

CYTB, the HbB2i phylogeny did not differentiate these Lamellibrachia taxa (Figure A4-2).

Within the entire HbB2i fragment, two sites identified simultaneous amplification of two different nucleotides (ambiguity). As only five individuals exhibited the ambiguity at the two sites, the nucleotides were left as ambiguous and treated as distinct alleles. The HbB2i haplotype

61 analysis identifies the most common haplotype as shared between L. luymesi and L. sp. 1 (Figure

4-2). Eight haplotypes unique to L. sp. 1 were connected to the major haplotype by only one mutation change, whereas there were only two unique haplotypes for L. luymesi identified in these analyses.

!"#

% & !"#$%& '(&)* !"#$%&+*+2+'(&)* " ./#01 $$/01 "#%/01

"'()*+,-

Figure 4-2: Median-joining haplotype networks of the mtCYTB, HbB2i and concatenated genes. Colors represent Lamellibrachia luymesi (black) and Lamellibrachia sp. 1 (white). Sizes of haplotype circles and are proportional to the number of individuals possessing the same sequence and each line represents one mutational change separating two haplotypes.

To better understand the evolutionary history of Lamellibrachia, we concatenated both

CYTB and HbB2i for 21 individuals (11 L. luymesi and 10 L. sp. 1) to produce a 1048bp fragment. While the combined dataset also fails to differentiate between L. luymesi and L. sp. 1, the haplotype network identifies higher levels of variation, with several rare haplotypes that are unique to each taxa and separated by up to seven mutation changes (Figure 4- 2).

62 Demographic history of Lamellibrachia was assessed by examining Tajima’s D in both

L. luymesi and L. sp. 1 at CYTB, HbB2i and the concatenated fragment. None of the D values

were significantly different from zero, supporting the assumption that the genes used here are

evolving in a neutral fashion. However, it must be noted that sample sizes less than the number of

loci, may limit the power of these tests (Simonsen et al. 1995).

Genetic differentiation between Lamellibrachia groups

Summary statistic means for the eight-microsatellite loci amplified across L. luymesi and

L. sp. 1 are shown in Table 4-3. All tests for linkage disequilibrium failed to reject random

association of alleles at different loci (p >0.05); mean null alleles frequencies were below 0.20 for

each population, and there was no evidence of stuttering or large allele dropout (Appendix C).

Table 4-3: Summary statistics for eight microsatellite loci amplified in three Lamellibrachia groups.

Lamellibrachia spp. Summary Statistics (means across all loci) (8 loci) SE SE NA HO HE FIS AR AR PAR PAR L. luymesi 7.37 0.64 0.63 0.00 7.16 0.96 5.52 1.26 L. sp. 1 8.62 0.57 0.64 0.11 7.83 1.86 6.19 2.28 Lamellibrachia sp. 1 N H H F AR AR P P SE (8 loci) A O E IS SE AR AR West GoM 8 0.55 0.63 0.13 3.28 1.16 2.09 0.74 East GoM 5 0.72 0.73 0.01 2.45 0.87 1.26 0.45 Lamellibrachia sp. 2 N H H F AR AR P P SE (5 loci) A O E IS SE AR AR West GoM 3.40 0.60 0.64 0.06 2.50 0.36 0.68 0.21 East GoM 6.80 0.62 0.66 0.05 2.54 0.29 0.82 0.32 Far East GoM 2.80 0.70 0.65 0.00 2.42 0.29 0.54 0.08

Abbreviations: number of alleles observed across all loci (NA), mean observed (HO) and expected (HE) heterozygosity, mean Wright’s Inbreeding Coefficient (FIS), mean rarified allelic richness (AR) and standard error SE SE (AR ); private allele richness (PAR) and standard error (PAR ). Rarefied over 16 samples and means are not significantly different (P > 0.05).

A total of 114 alleles were observed across all eight loci and 45 individuals in the dataset.

The largest number of alleles was found at the locus L454_11 for both L. luymesi (13) and L. sp.

1 (19). For L. luymesi, the mean observed and expected heterozygosity was 0.64 and 0.63, and for

L. sp. 1, the mean observed and expected heterozygosity of 0.57 and 0.64, respectively. Of the 16 population/locus combinations, only two departed from HWE after FDR correction at 0.01.

STRUCTURE and network analyses both support the occurrence of two genetically distinct groups of Lamellibrachia (Figure 4-3).

-())

)(.)

)(*)

)(/)

)(0)

)()) !"#$%&'()* !"#,"($-

!"#$%$&'()) !"#$%$*+()) !"#$%$')())

!"#,"($-

!"#$%&'()*

Figure 4-3: Top: STRUCTURE results (admixture model, three replicate runs) for Lamellibrachia luymesi (black) and Lamellibrachia sp. 1 (white). Each vertical bar represents an individual tubeworm. The y-axis is the proportion of each individual’s genotype belonging to a distinct population cluster. Bottom: Network topologies of L. luymesi and L. sp. 1 individuals from the Rozenfeld Distance model (RD) based on eight shared microsattelite markers. Shared Allele Distance (SAD) model results are not shown. Only links with value smaller than or equal to the percolation distances are present. Nodes (circles) represent individuals. Two clusters are identified, one for each Lamellibrachia population.

For the network analyses, both the Rozenfeld Distance and Shared Allele Distance models show congruent results of two well-defined clusters with one hybrid link at percolation

64 thresholds higher than 50. The link was present between an L. luymesi individual from site

GC234 and an L. sp. 1 individual from site GC600, two sites in relatively close geographic proximity and separated by about 640m in depth (Figures 4-1 and 4-3). Despite this connection between the two localities, clustering identifies strong genetic structuring by depth, which is also supported by AMOVA analyses (p = 0.001, Table 4-4).

Table 4-4: Summary of Analysis of Molecular Variance (AMOVA) conducted for each study under the Infinite Allele Model, FST. Significance (*) was tested using 1000 permutations. Results for Stepwise Mutation Model (RST) are congruent with IAM.

Source of Variation Study d.f. SS p-value Variation (%)

Within individuals 44 107.00 74.60 --- Lamellibrachia (Two taxa) Among individuals 42 116.64 5.30 0.017*

Among populations 1 31.43 20.10 0.001*

Within individuals 24 58.00 90.60 ---

Lamellibrachia Among individuals 22 63.17 8.50 0.014* sp.1

Among populations 1 3.29 0.90 0.253

Within individuals 31 48.00 95.00 ---

Lamellibrachia Among individuals 27 47.80 5.00 0.080 sp. 2

Among populations 3 4.14 0.00 0.867

Abbreviations: d.f, degrees of freedom; SS, sum of squares.

65 Lower slope Lamellibrachia population genetics

To test for the presence of genetic differentiation across the Mississippi Canyon as a result of physical, chemical and geographical barriers to dispersal associated with the canyon

(Sammarco et al. 2012), we pooled samples based on the location of sites to the west or east of the canyon (see Figure 4-1 and Cairns et al. 1993 for deep GoM faunal provinces associated with the Mississippi Canyon). Summary statistics for microsatellites amplified in west and east populations of L. sp. 1 and L. sp. 2 are detailed in Table 2. Twenty-four individuals were examined in L. sp. 1, and 71 alleles were observed across all loci. The largest number of alleles was found at locus L454_11 (18) for the West GoM population and L454_6 (9) for the East GoM population. For the West GoM, the mean observed and expected heterozygosity of 0.55 and 0.63, and for the East GoM, the mean observed and expected heterozygosity of 0.72 and 0.73. Three departures from HWE were observed at loci L454_6 and L454_14 after FDR correction at 0.01.

The L. sp. 2 dataset was divided into the West, East and Far East GoM and analyzed using five microsatellites. Thirty-one individuals were examined with 36 alleles observed across all loci. The largest number of alleles was detected at the locus L454_6 for West (6), East (15) and Far East (4) locations. For the West GoM, the mean observed and expected heterozygosity of

0.60 and 0.64; for the East GoM, the mean observed and expected heterozygosity of 0.62 and

0.66, and for the Far East, the mean observed and expected heterozygosity of 0.70 and 0.65. Two departures from HWE were observed at loci L454_6, after FDR correction at 0.01. Despite differences in sample sizes between locations, no significant difference in the average allelic or private allelic richness was detected among locations in either L. sp.1 or L. sp.2 (p > 0.05).

STRUCTURE analyses did not detect any genetic clustering among the western and eastern groups in either L. sp. 1 and L. sp. 2 datasets. AMOVA analyses found non-significant among population differentiation in either L. sp. 1 or L. sp. 2 (p > 0.05).

66 Discussion

Co-occurrence of vestimentiferan species and niche differentiation

Two clearly distinct species, L. sp. 1 and L. sp. 2, co-occur in aggregations at five of the nine deep sites investigated in this study, suggesting that co-occurrence of the two species is common at seeps in the deep GoM. Furthermore, L. sp. 1 and L. sp. 2 can occur in the same aggregations as another vestimentiferan species, E. laminata (Becker et al. 2011); all three species were collected together from aggregations at four of the nine sites investigated here. The co-occurrence of different species of vestimentiferan within a single aggregation has also been observed at seeps on the upper Louisiana Slope. Here, three distinct species, L. luymesi,

Seepiophila jonesi, and a currently undescribed Escarpia-like vestimentiferan, have all been found within the same aggregation (Bergquist et al. 2002, 2003a, 2003b).

Different species exploit specific niches in the environment, and resource partitioning in particular has implications for divergence within vestimentiferan groups (Root 1967, Becker et al.

2011). The co-existence of multiple, yet similar, vestimentiferan species within a single aggregation suggests that each species has filled a specific niche (Becker et al. 2011). On the upper Louisiana Slope, S. jonesi grows with its plume near the sediment/seawater interface

(Cordes et al. 2006, 2009). As sulfide is rarely detected over 2cm above the sediment surface, S. jonesi may be able to uptake sulfide across its plume (Julian et al. 1999, Fretytag et al. 2001).

Conversely, L. luymesi grows more upright than S. jonesi, with its plume being a meter or more above the sediment surface (Bergquist et al. 2002, Cordes et al. 2009). As a consequence, L. luymesi does not have access to significant amounts of sulfide via its plume, but instead, acquisition of sulfide occurs across the posterior “roots” (Julian et al. 1999, Freytag et al. 2001).

The orientation of L. luymesi and S. jonesi suggest each species may exhibit differences in

67 mechanisms of sulfide and O2 uptake, supporting this as a possible niche differentiation between the two.

While there are no obvious differences in the growth forms of the co-occurring L. sp. 1,

L. sp. 2 or E. laminata, Becker et al. (2011) found significant and consistent differences in the tissue stable isotope contents of co-occurring Lamellibrachia sp. and E. laminata from numerous sites around the deep GoM and propose differences in the forms or mechanisms of acquisition of inorganic nitrogen, a factor that may contribute to niche differentiation between these co- occurring species. The co-occurrence of L. sp. 1 and L. sp. 2 at sites analyzed in this study indicates that there is some as yet unknown niche differentiation occurring between these two species of Lamellibrachia.

Utility of mitochondrial genes versus microsatellite markers in the identification of seep vestimentiferan species

DNA “barcoding” refers to a specific region in the genome that can be used for rapid and accurate identification of species in many groups (Valentini et al. 2008). In animals, the proposed region of choice is the highly conserved COI molecule, for which universal primers have been developed. Despite its successful implementation for discriminating among species in other animal groups, COI fails to differentiate between morphologically distinct, depth-separated and/or geographically separate species of Lamellibrachia and Escarpia cold seep vestimentiferan tubeworms (Black et al. 1997, McMullin et al. 2003, Andersen et al. 2004, Miglietta et al. 2010,

Cowart et al. 2013).

In this study, we find that similar to other mitochondrial genes, CYTB does not separate

L. lumyesi and L. sp. 1. Several factors have been suggested for the low rate of mutation seen in vestimentiferan mitochondrial genes, including the possibility of young lineages where the time

68 since divergence is small, long generation times in seep tubeworms, differences between male and female effective population sizes, as well as selective sweeps which may reduce variation at this locus (McMullin et al. 2003, Bazin et al. 2006, Galtier et al. 2009, Cowart et al. 2013).

The intron regions of nuclear genes such as hemoglobin tend to be more variable then mitochondrial genes and nuclear exons, and can therefore be useful for detecting phylogenetic relationships (Sang 2002). Cowart et al. (2013) employed the Hemoglobin subunit B2 intron

(HbB2i) to detect three distinct species of Escarpia, a finding that is congruent with morphological descriptions and their geographic isolation. In this study, HbB2i did not differentiate L. luymesi and L. sp. 1. Although neither the CYTB nor HbB2i genes separate L. luymesi and L. sp. 1, the haplotype networks of both the single and concatenated genes were consistent with the hypothesis that Lamellibrachia are distinct taxa, as all rare haplotypes are unique to L. luymesi or L. sp. 1.

The eight cross-species amplified microsatellite loci group L. luymesi and L. sp. 1 into distinct genotypic clusters (Mallet 1995). The use of microsatellites to discriminate tubeworm species was previously employed within the genus Escarpia; nine microsatellites cross amplified in three taxa allowed for the identification of three distinct species of Escarpia, supporting original species descriptions based on morphology and geography (Cowart et al. 2013). Results of the current and previous studies illustrate that the usual mitochondrial genes are not variable enough to uncover finer scale vestimentiferan relationships and more variable markers, specifically microsatellites, can allow identification of closely related vestimentiferan species. We predict that analysis of additional nuclear loci will provide further support of these findings.

69 Influence of environmental factors on seep tubeworm species distributions

In the deep sea, vertical species ranges are heavily influenced by factors such as larval tolerance to temperature and pressure, orientation in the water column and swimming behavior

(Somero, 1992a, Tyler and Young 1998, 1999, Cordes et al. 2007). Lamellibrachia adults produce lecithotrophic larvae, which can persist in the water column for about three weeks and have an estimated dispersal distance of about 100km from their natal sites (Young et al. 1996,

Tyler and Young 1999, Young et al. 2012). During their time in the water column,

Lamellibrachia larvae likely experience limits to their temperature tolerance. In the GoM, temperatures ranges from 8 to 10ºC at depths shallower than 1000m, while the minimum temperature at 1600m is approximately 4ºC (Carney 2005). Young et al. (2012) have observed that L. luymesi larvae cannot tolerate temperatures found above the thermocline. If L. luymesi larvae are also unable to tolerate temperatures as low as 4ºC, they would be restricted to the upper slope, unable to survive at greater depths. Likewise, L. sp. 1 larvae may be unable to tolerate the warmer temperatures accompanied with the upper slope, thereby restricting their distribution to deeper regions.

With regards to pressure, Young and Tyler (1993) demonstrated that low pressures affect larval development mechanisms and can be as lethal to deep-sea embryos as high pressures are to shallow living embryos. As with temperature, high pressures may restrict depths at which L. luymesi and L. sp. 1 colonize. Additionally, as temperature and pressure acting in concert form complex biological interactions (Somero 1992b, Aquino-Souza et al. 2008), simultaneous changes in these parameters are likely factors controlling distributions of the Lamellibrachia tubeworms.

70 Potential for overlap of Lamellibrachia depth ranges

Contact zones between different lineages are often located in areas of narrow overlap in environmental conditions (Rocha et al. 2005). As L. luymesi and L. sp. 1 have adjacent distributions on the Louisiana Slope, they likely co-occur at the extremes of their ranges, along the margin between the upper and lower Louisiana Slope (~950m). While there are no known observations of L. luymesi and L. sp. 1 co-occurring within the same sites, this likely reflects the fact that very few sites near this depth have been sampled in the GoM (Becker et al. 2012,

Roberts et al. 2007). The adjacent ranges suggest there are likely areas where L. luymesi and L. sp. 1 do overlap in the environment. The co-existence of different tubeworm species within the same aggregation show that distinct vestimentiferan species occurring in the same area likely exploit specialized niches, with L. luymesi and L. sp. 1 being no exception.

The analyses of polymorphic microsatellite markers in this study suggest that despite possible geographic overlap, gene flow between L. luymesi and L. sp. 1 is limited. Given the data that support two distinct genotypic groups that are congruent with morphological and depth descriptions, we propose that L. luymesi and L. sp. 1 are distinct biological species. Additionally, this separation of L. luymesi and L. sp. 1 follows the “species replacement” pattern seen at cold seeps world-wide, and is likely due to parapatric speciation, which occurs in the absence of geographic or oceanic barriers, yet requires restricted gene flow or environmental barriers (Coyne and Orr 2004).

Low levels of genetic structure detected across geographic regions in Lamellibrachia tubeworms

Submarine canyons may influence gene flow if propagules are unable to cross these boundaries due to physiological or environmental constraints (Etter et al. 2005, Rex and Etter,

71 2010). In the deep GoM, analyses identify the presence of one undifferentiated population in L. sp. 1 and L. sp. 2, suggesting propagules move across the canyon. A similar pattern of lack of genetic structure has been detected in other seep vestimentiferans. McMullin et al. (2010) used five and seven microsatellite loci in L. luymesi and S. jonesi, respectively, to identify minimal genetic structure across the upper Louisiana Slope. Likewise, Cowart et al. (2013) used 11 and 16 microsatellite loci in Escarpia laminata (GoM) and Escarpia southwardae (West Africa), respectively, and found no population structure at the regional scale.

In the case of the Lamellibrachia below 1000m in the Gulf of Mexico, we find that recent gene flow occurs across geographic distances of >650km in both L. sp. 1 and L. sp. 2. ). Low temperatures encountered in the deep-sea often result in lower metabolic rates (slower utilization of energy reserves), increasing larval life and thus dispersal distances (Shilling and Manahan

1994; Young et al. 1997b). The extended dispersal period of seep vestimentiferan larvae, in addition to what is known about speeds and directions of deep currents across the GoM, and the presence of more than 50 known chemosynthetic sites distributed across the GoM, reinforce the genetic data uncovering high gene flow across sites at the regional scale in the GoM (Tyler and

Young 1999, McMullin et al. 2003, Mineral Management Services 2006, Roberts et al. 2007,

Young et al. 2012, Cowart et al. 2013).

72 Acknowledgements

We would like to thank the following people for their contribution to this project: captains, crews and expedition leaders of the DSV Johnson Sea Link II, R/V Seward Johnson

ROV Jason II, the US deep submergence facility, DSV Alvin, R/V Atlantis, the NOAA vessel

Ronald Brown; Drew Wham, Sophie Arnaud-Haond, Kevin Kocot, Kimberlyn Nelson, Iliana

Baums, Todd LaJeunesse, Chuyna Huang, Pen-yuan Hsing, Miles Saunders and Andrew

Mendelson. This research was funded in part by the National Science Foundation (Award #

1209688 and IOS-0843473 to DAC and KMH), the National Oceanographic Partnership Program

(NOPP) through support from the Bureau of Ocean Energy Management contracts

#M05PC00018 and #M08PC20038 to CRF (TDI Brooks International Prime), the National

Oceanic and Atmospheric Administration’s Office of Ocean Exploration and Research (NOAA

OER), Penn State Eberly College of Science, and an Alfred P. Sloan Scholarship to DAC. Any opinions, findings, conclusions, or recommendations expressed in this article are those of the authors and do not necessarily reflect the views of the National Science Foundation.

CHAPTER 5

Summary and Implications

DNA barcoding in seep vestimentiferans

The term “DNA barcoding” refers to the use of a specific region in the genome that can be used for rapid and accurate identification of species (Valentini et al. 2008). For animals, the proposed region of choice is the highly conserved Cytochrome C Oxidase subunit 1 (COI) molecule, for which universal primers have been developed. Additionally, the COI fragment has been purported to be nearly identical among individuals of the same species, but differs between species. While the use of this single, “standardized” fragment has been successfully implemented in many mammal groups to uncover species, COI and other mitochondrial fragments fail to differentiate between morphologically distinct, geographically distant groups of cold seep vestimentiferan tubeworms. This suggests that while the COI fragment has the potential to be the barcoding molecule for some types of animals, it cannot differentiate species in all groups of animals, and that caution is needed when applying this “universal” marker.

As DNA barcoding may enhance biodiversity inventories by being more accurate, faster, and cheaper than traditional morphological classifications (Gaston and O’Neill 2004), the search for barcoding molecules within specific groups of animals is necessary and can be aided by the use of next generation sequencing tools. In this dissertation, we use 454 genomic sequencing to uncover several short tandem repeats (microsatellites) in seep tubeworms of the genera Escarpia and Lamellibrachia. In addition to elucidating patterns of population structure within “species”, these loci support the occurrence of distinct species within Escarpia and Lamellibrachia, where

74 COI and other mitochondrial markers did not. Additionally, we find that a single locus, the intron of a nuclear gene, also resolves species within the Escarpia, but not within Lamellibrachia.

Overall, we find that while COI and other mitochondrial markers were not variable enough to separate some species of seep tubeworms, we have identified several other loci that can serve as genus-specific barcoding markers within Escarpia and Lamellibrachia. The results here also suggest that multiple markers are likely more useful when attempting to gain the most accurate identity of specific organisms when identification via a single locus fails.

Implications of findings on dispersal and physical oceanography

Previous work by Young et al. 1996 described the early developmental stages of seep vestimentiferan tubeworms Lamellibrachia luymesi and Seepiophila jonesi. Young and co- workers found that the estimated larvae survival time coupled with speed of ocean currents, suggest that seep tubeworm larvae have enormous dispersal potential. The dispersal ability of seep tubeworm larvae can be inferred from the length of larval life (duration in the water column), the direction and speeds of ocean currents by which larvae are advected, and studies of genetic similarity between disjunct populations (Tyler and Young 1999).

As seep tubeworm larvae are lecithotrophic and feed from a yolk while in the water column, the larvae tend to have greater dispersal potential than direct developers (Young et al.

1996). The estimated larval lifespan of seep tubeworms is about three weeks, and with a mean current velocity of 10cm/s at 200m above the seafloor on the Louisiana slope (MacDonald et al.

1998), McMullin et al. (2010) estimated that seep vestimentiferans living less than 1000m could have a dispersal distance of about 181km.

In this dissertation, dispersal potential of seep vestimentiferans living below 1000m at seeps in the Gulf of Mexico and along the west coast of Africa is inferred from genetic structure

75 (or lack of) of populations. Successful larval migrants leave genetic trails of their movements, which can be detected indirectly by estimating the presence of population connectivity (Hellberg et al. 2002). The genetic work in this dissertation supports previous research, confirming that seep vestimentiferans do have a vast dispersal capability, as populations are genetically homogeneous over large spatial scales, reaching distances of nearly 1000km and 150km across the Gulf of

Mexico and West African cold seeps, respectively.

Although there is strong evidence for high dispersal capabilities of tubeworms, we still find multiple species within the same geographic and bathymetric ranges. This suggests that in the presence of high dispersal, the formation of species still occurs and speciation events are likely due to factors other than physical geographic boundaries to drive genetic divergence between populations. “Invisible barriers”, such complex ocean currents driven by winds, physical boundaries and the differential heating and cooling properties of water, have important roles in the formation of species by separating populations (Palumbi 1994). Furthermore, marine populations tend to be large in size, a factor that has been shown to slow genetic divergence between populations (Hellberg et al. 2002). Finally, abrupt changes in the genomes of populations can also drive speciation; these include the occurrence of transposable elements and random genetic drift, both of which can drive genetic divergence and ultimately speciation (Palumbi

1994).

Investigations of genetically distinct vestimentiferan taxa

Escarpia spp.

Genetic analyses confirm that geographically separate Escarpia are three genetically distinct taxa that are likely proto-species undergoing strict divergence, but as to how they became

76 separate warrants further investigation. Initially, one can assume this is an explicit case of allopatric speciation, as the populations likely became divergent after the formation of the

Isthmus of Panama and the sills that border the Gulf of Mexico, which likely inhibited migrant exchange between oceanic basins. As a consequence of gene flow limitations, these separate groups have evolved morphological differences. However, whether these morphs have also evolved physiological differences is currently unknown.

All three Escarpia taxa studied here reside within similar depth ranges (McMullin et al.

2003, Andersen et al. 2004) and therefore differences in morphology are likely not driven by depth. Other environmental factors, including salinity, temperature, and level of seepage appear to be similar across the seep environments, and so morphology is also not likely driven by habitat specialization. However, dissimilarity in morphological characters such as presence or absence of pinnules and the size and shape of the axial rod may be fueled by species-specific differences in demand and utilization of nutrients at seeps, or may be possibly neutral traits.

Lamellibrachia spp.

Individuals within the genus Lamellibrachia in the GoM present a more ambiguous case regarding the question of isolation than does the Escarpia. In this study, we find that the single locus genes, both mitochondrial and nuclear, suggest that L. luymesi from the upper slope does not differ genetically from L. sp. 1 from the lower slope. However, the cross-amplified microsatellite dataset identifies the groups as being genetically distinct. The possible geographic overlap between the two taxa, specifically at the Green Canyon (GC) locations, suggest that offspring from each group may reach the upper/lower limits of the adult distributions, to settle at similar sites and mate with those in close proximity. It must also be noted that tubeworms belonging to different species often co-occur within aggregations on both the upper and lower

77 slope. For example, L. luymesi and Seepiophila jonesi (two more distantly related species) often co-occur within aggregations on the upper slope (Bergquist et al. 2002), and E. laminata, L. sp. 1 and L. sp. 2 have been found to co-occur within aggregations on the lower slope (personal observations). This suggests that morphologically distinct Lamellibrachia (L. sp. 1 and L. sp. 2) can settle in the same regions, and that depth-differentiated Lamellibrachia (L. luymesi and L. sp.

1) likely settle in the same regions as well, but have not been observed, which is likely a function of sampling abilities.

Environmental factors such as temperature and pressure vary depending on depth in the water column. Therefore, it is possible that the difference in morphology that is observed between

Lamellibrachia may be driven by habitat specialization. Gill lamellae are features that are known to aid in gas exchange, and the differences in the numbers of gill lamellae between

Lamellibrachia may be due to specific differences in demand and utilization of nutrients.

Use of hemoglobin molecules

The Hemoglobin subunit B2 intron (HbB2i) gene tested provides some of the first genetic evidence that the three Escarpia are distinct groups. The lack of differentiation seen between

Lamellibrachia at this locus may be due to incomplete divergence or just a lack of variation at this particular locus.

The hemoglobin molecule has an important role in the symbiotic lifestyle of vestimentiferans. Vestimentiferan emoglobin has a high molecular mass and is capable of binding both H2S and O2 at separate sites. Once bound, hemoglobin transports H2S and O2 through the tubeworm’s blood to the trophosome, where the symbiotic bacteria can use H2S as the energy source to fix CO2 into organic molecules (Arp et al. 1987, Fisher et al. 1988, Childress and Fisher

1992, Goffredi et al. 1997b, Zal et al. 1997). The importance of hemoglobin in the

78 vestimentiferan lifestyle provides reasoning for why this gene was selected as a candidate to detect variation in tubeworms. While the hemoglobin exon region appears to be highly conserved, the nucleotide differences within the intron suggest some minor functional differences of splicing mechanisms. The biological functions of the B2 subunit intron are unknown, so whether its purpose is to allow for alternative splicing to code for multiple proteins needs additional investigation.

Use of microsatellite loci

Microsatellites are often markers of choice in evolutionary and ecological studies due to higher rates of polymorphism, allowing for the analyses of genetic diversity previously undetected by less sensitive analyses (Queller et al.1993, Ellegren 2004). In addition to clarifying hidden population structure with species, microsatellites have been used as tools to identify cryptic taxa and delineate species (Clauss et al. 2002, Hausdorf and Henning 2010). The cross species transferability of polymorphic microsatellite markers is expected to be successful in animal groups of close evolutionary distance, such as within genera (Barbará et al. 2007). The microsatellites amplified across Escarpia and Lamellibrachia groups (respectively) are found throughout the nuclear genome and are independent loci (i.e. non-linked), which gave us the ability to assess differentiation within Escarpia spp. and Lamellibrachia spp. using more than just the HbB2i locus.

Population genetics of seep vestimentiferans below 1000m

Analyses of twenty-eight polymorphic loci developed specially for E. laminata (Gulf of

Mexico) and E. southwardae (West Africa) identify one panmictic population on a regional scale.

79 Furthermore, analyses of eight polymorphic loci amplified in L. sp. 1 and five loci in L. sp. 2 also identify one panmictic population in each group across the Gulf of Mexico (GoM).

Across sampling locations in both genera, FST values typically fell below 0.05 indicating low levels of genetic differentiation across locations (Hartl and Clark, 2007). While no significant differentiation is seen among geographic sites for either genera, it is possible that genetic structure occurs on a temporal or finer geographic scale. Localized genetic heterogeneity can result from temporal variation, as abundance and distribution of local recruits may differ genetically and one location may contain individuals from a single cohort, while another location may be a conglomeration of several cohorts (Johnson and Black 1984). Additionally, differences in ages of cohorts can account for variances in structure (Arnaud-Haond et al. 2008; McMullin et al. 2010). In the bivalve, Spisula ovalis, the temporal structure seen was possibly a part of a

“genetic mosaic” of genetically variable cohorts at each site, rather than a defined genetic entity

(David et al. 1997a).

Furthermore, contrasting patterns of genetic diversity seen at regional versus finer scale due to chaotic dispersal makes sampling strategy an important component of determining true population structure (eg. Pearl Oyster in Arnaud-Hanod et al. 2008 and sea grasses in Becheler et al. 2010). Mosaic patterns differing in age structure can be difficult to assess in many marine invertebrate populations (David et al. 1997b), and those populations in the deep sea are no different, as there are many challenges to quantitative sampling at these depth (Wilson and

Hessler 1987, Gage and Tyler 1991, Turnipseed et al., 2003). To assess finer-scale structure, more extensive spatial and temporal sampling is needed (see “Future directions for research”), however in the present research, our goals were to identify the biogeographical limits of E. laminata, E. southwardae, L. sp. 1 and L. sp. 2 restricted to geographic regions.

80 Future directions for research

Comparative genomics and laboratory breeding

Despite the evidence supporting the occurrence of distinct species within both Escarpia and Lamellibrachia, the absence of full genomic sequencing, as well as controlled crosses in attempts to produce hybrid offspring, prohibits us from confirming these are biologically species in the strictest sense (i.e. reproductively isolated). A comparative genomics project that includes genomes from multiple individuals supplemented with the 454 data collected during the course of this study would shed light on genomic regions that are similar or dissimilar between each group.

Within the nuclear genome, sequence concatenation and mega alignment would identify in which regions and/or genes does variation exists among groups.

With regards to the mitochondria, gene order is highly conserved among Annelids

(Jennings and Halanych 2005) and from the current analyses investigation of three mitochondrial genes, we suspect that most of the mitochondrial genome is identical within the Escarpia and

Lamellibrachia genera. However, the displacement loop (also known as the “D-loop” or the control region) has yet to be investigated in these animals. While portions of the D-loop are likely to be conserved, previous authors have shown that other regions are highly variable and have proven useful during the study of evolutionary history in rodents (Larizza et al. 2002). The uncovering of the D-loop within the mitochondria of Escarpia and Lamellibrachia may be useful in supporting the genetic differences detected during the course of this research.

Increasing loci and sample sizes

The Escarpia population genetics study has employed more polymorphic loci and individuals than any previous study on this group. This has allowed for more accuracy, and thus

81 confidence in the findings described herein. However, only a total of eleven sampling locations

(eight for GoM and three for WACS) were studied. In the GoM alone, there are tens of known locations that have not been investigated during the course of this project, and we cannot define gene flow within every site currently known without additional sampling. Furthermore, additional seep locations along the coast of West Africa are more recently discovered, many locations have yet to be investigated. The Lamellibrachia population genetics analyses performed in this study are initial investigations of population structure in deep GoM Lamellibrachia. To increase accuracy and thus confidence of findings, it will be necessary to obtain more individuals from other locations, and especially more polymorphic loci. Nonetheless, in the face of sampling difficulties, this Ph.D. research does provide understanding of gene flow occurring between broadly distributed seep locations in two biogeographical regions.

During the course of this research, population structure in E. spicata in the Gulf of

California was not investigated, as individuals in this study were collected from a single location/aggregation. Once additional individuals from other sampling locations have been achieved, the available 454 dataset for E. spicata could be used to identify potential polymorphic loci. Additionally, Feldman et al. 1998 compared COI sequences from E. spicata found at seeps versus those found at vents, to find that COI differs by less than 0.4% across seeps and vents.

This level of COI sequence differentiation is consistent with interspecific levels of COI divergence in vestimentiferans, generally (Black et al. 1997; Miglietta et al. 2010). As there is evidence for genetic differentiation between E. spicata residing at different chemosynthetic environments in the eastern Pacific, a larger-scale study utilizing more variable markers and more individuals from the different ecosystems would confirm the presence of genetic structure across seep, vent, and whale fall sites.

82 Fine scale, temporal structure and ocean currents

As discussed previously, additional population genetic analyses are needed to assess if there is temporal or finer genetic structure occurring in Escarpia. This would require careful sampling of various aggregations of differing ages. Finally, our understanding of oceanic currents below 1000m in the water column remains poor. As with any deep-sea collections, scientific operations at these depths are limited, and to better understand flow and direction of dispersal at seeps, an interdisciplinary and collaborative effort would likely need to be undertaken to determine the accurate speeds, directions and duration of water masses at seep sites that host biological communities.

Conclusions

As it is critical to improve our knowledge of how deep sea organisms colonize these isolated habitats, this dissertation work has clarified the dispersal boundaries of several taxa within two seep vestimentiferan genera, Escarpia and Lamellibrachia. Through the use of a holistic approach to includes combining molecular, physical and environmental data to test hypotheses regarding migration and gene flow of vestimentiferans this dissertation 1) elucidates previously undetected genetic diversity and 2) improve our knowledge of vestimentiferan movements at deep sea cold seeps. These goals are critical to enhancing our understanding of how unique cold seep communities are established and maintained.

REFERENCES

Aharon, P (1994). Geology and biology of modern and ancient submarine hydrocarbon seeps and vents: an introduction. Geo-Marine Letters 14 (2), 69-73.

Allentoft M, Schuster S, Holdaway RN, Hale ML, McLay E, Oskam C, Gilbert MTP, Spencer P, Willerslev E, Bunce M (2009) Identification of microsatellites from an extinct moa species using highthroughput (454) sequence data. Biological Techniques 46.

Andersen, AC, Hourdez, S, Marie, B, Jollivet, D, Lallier, FH, Sibuet, M (2004) Escarpia southwardae sp nov., a new species of vestimentiferan tubeworm (Annelida, Siboglinidae) from West African cold seeps. Canadian Journal of Zoology-Revue Canadienne De Zoologie, 82, 980 - 999.

Anderson, RK, Scalan, RS, Parker, PL, Behrens, EW (1983) Seep oil and gas in Gulf of Mexico slope sediments. Science, 222, 619 - 621.

Arnaud-Haond, S, Vonau, V, Rouxel, C, Bonhomme, F, Prou, J, Goyard, E, Boudry, P (2008). Genetic structure at different spatial scales in the pearl oyster (Pinctada margaritifera cumingii) in French Polynesian lagoons: beware of sampling strategy and genetic patchiness. Marine Biology, 155(2), 147-157.

Arp, AJ, Childress, JJ, Vetter, RD (1987). The sulphide-binding protein in the blood of the vestimentiferan tube-worm, Riftia pachyptila, is the extracellular haemoglobin. Journal of Experimental Biology, 128 (1), 139-158.

Aquino-Souza, R, Hawkins, SJ, Tyler, PA, (2008) Early development and larval survival of Psammechinus miliaris under deep-sea temperature and pressure conditions. Journal of the Marine Biological Association of the UK 88 (3), 453 - 461.

Avise, JC, Arnold, J, Ball, RM, Bermingham, E, Lamb, T, Neigel, JE, Reeb, CA, Saunders, NC (1987). Intraspecific phylogeography: the mitochondrial DNA bridge between population genetics and systematics. Annual review of ecology and systematics, 18, 489-522.

Bailly, X, Leroy, R, Carney, S, Collin, O, Zal, F, Toulmond, A, and Jollivet, D. (2003) The loss of the hemoglobin H2S-binding function in annelids from sulfide-free habitats reveals molecular adaptation driven by Darwinian positive selection. Proceedings of the National Academy of Sciences, USA, 100: 5885 - 5890.

Barbará T, Palma-Silva, C, Paggi, GM, Bered, F, Fay, MF, Lexer, C (2007). Cross-species transfer of nuclear microsatellite markers: potential and limitations. Molecular Ecology 16, 3759 - 3767.

Bandelt, HJ, Forster, P, Röhl, A (1999) Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution, 16:37 - 48.

Bazin, E, Glémin, S, Galtier, N (2006) Population size does not influence mitochondrial genetic diversity in animals. Science, 312 (5773), 570 - 572.

84 Becheler, R, Diekmann, O, Hily, C, Moalic, Y, Arnaud-Haond, S (2010) The concept of population in clonal organisms: mosaics of temporally colonized patches are forming highly diverse meadows of Zostera marina in Brittany. Molecular Ecology, 19, 2394 – 2407

Becker, EL, Macko, SA, Lee, RW, Fisher, CR (2011) Stable isotopes provide new insights into vestimentiferan physiological ecology at Gulf of Mexico cold seeps. Naturwissenschaften, 98 (2), 169-174.

Benjamini, Y, Yekutieli, D (2001) The control of the false discovery rate in multiple testing under dependency. Annals of statistics, 29, 1165 - 1188.

Bergquist, DC, Williams, FM, Fisher, CR (2000) Longevity record for deep-sea invertebrate. Nature, 403, 499 - 500.

Bergquist, DC, Urcuyo, IA, Fisher, CR (2002) Establishment and persistence of seep vestimentiferan aggregations on the upper Louisiana slope of the Gulf of Mexico. Marine Ecology Progress Series, 241, 89–98

Bergquist, D, Ward, T, Cordes, EE, McNelis, T, Howlett, S, Kosoff, R, Hourdez, S, Carney, R, Fisher, CR (2003a) Community structure of vestimentiferan-generated habitat islands from Gulf of Mexico cold seeps. Journal of Experimental Marine Biology and Ecology, 289 (2) 197 - 222.

Bergquist, DC, Andras, JP, McNelis, T, Howlett, S, Van Horn, MJ, Fisher, CR (2003b) Succession in Gulf of Mexico cold seep vestimentiferan aggregations: the importance of spatial variability. Marine Ecology, 24 (1), 31-44.

Bierne, N, Lehnert, SA, Bedier, E, Bonhomme, F, Moore, SS (2000) Screening for intron-length polymorphisms in penaeid shrimps using exon-primed intron-crossing (EPIC)-PCR. Molecular Ecology, 9(2), 233 - 235.

Black, MB, Halanych, KM, Maas, PAY, Hoeh, WR, Hashimoto, J, Desbruyères, D, Lutz, RA, Vrijenhoek, RC (1997) Molecular systematics of vestimentiferan tubeworms from hydrothermal vents and cold-water seeps. Marine Biology, 130, 141 - 149.

Black, MB, Lutz, RA, Vrijenhoek, RC (1994) Gene flow among vestimentiferan tube worm (Riftia pachyptila) populations from hydrothermal vents of the eastern Pacific. Marine Biology, 120(1), 33 - 39.

Black, MB, Trivedi, A, Maas, PA, Lutz, RA, Vrijenhoek, RC (1998) Population genetics and biogeography of vestimentiferan tube worms. Deep-Sea Research Part II, 45(1-3), 365 - 382.

Blankenberg D, Kuster GV, Coraor N, Ananda G, Lazarus R, Mangan M, Nekrutenko A, Taylor J (2010) Galaxy: A Web-Based Genome Analysis Tool for Experimentalists. Current Protocols in Molecular Biology 19.10, 19.10.11 - 19.10.21.

Boore, JL, Brown, WM (2000) Mitochondrial Genomes of Galathealinum, Helobdella, and Platynereis: Sequence and Gene Arrangement Comparisons Indicate that Pogonophora Is Not a Phylum and Annelida and Arthropoda Are Not Sister Taxa. Molecular Biology and Evolution, 17, 87 - 106.

85 Brooks JM, Kennicutt MC, Fisher CR, Macko, SA, Cole, K, Childress, JJ, Bidigare, RR, Vetter, RD (1987) Deep-sea hydrocarbon seep communities - evidence for energy and nutritional carbon sources. Science, 238, 1138-1142.

Brown JH (1984) On the relationships between abundance and distribution of species. The American Naturalist 124 (2), 255 - 279

Bruun, A, (1957) Deep sea and abyssal depths. Treatise on marine ecology and paleoecology 1, 641 - 672.

Cairns, SD, Opresko, DM, Hopkins, TS, Schroeder, WW (1993) New records of deep-water Cnidaria (Scleractinia & Antipatharia) from the Gulf of Mexico. Northeast Gulf Science 13 (1), 1 - 11

Carney, R, (1994) Consideration of the oasis analogy for chemosynthetic communities at Gulf of Mexico hydrocarbon vents. Geo-Marine Letters 14, 149 - 159.

Carney R (2002) Currents in a Cul-de-Sac. NOAA Ocean Explorer.

Carney, RS, (2005) Zonation of deep biota on continental margins. Oceanography and Marine Biology: An Annual Review 43, 211-278, Taylor and Francis

Cavanaugh, CM, Gardiner, SL, Jones, ML, Jannasch, HW, Waterbury, JB (1981). Prokaryotic Cells in the Hydrothermal Vent Tube Worm Riftia pachyptila Jones: Possible Chemoautotrophic Symbionts. Science 213, 340-342.

Chapuis, M, Estoup, A (2007) Microsatellite null allele and estimation of population differentiation. Molecular Biology and Evolution, 24, 621 - 631.

Chelton, DS, Schlax, MG (1996) Global observations of oceanic Rossby waves. Science, 272, 234 - 238.

Childress, JJ, Fisher, CR (1992) The biology of hydrothermal vent animals - physiology, biochemistry and autotrophic symbioses. Oceanography and Marine Biology 30, 337 - 441.

Chybicki, I, Burczyk, J (2009) Simultaneous estimation of null alleles and inbreeding coefficients. Journal of Heredity, 100, 106 - 113.

Clauss MJ, Cobban, H., Mitchell-Olds, T. (2002) Cross-species microsatellite markers for elucidating population genetic structure in Arabidopsis and Arabis (Brassecaeae). Molecular Ecology 11, 591 - 601.

Cordes, EE, Bergquist, DC, Predmore, BL, Jones, C, Deines, P, Telesnicki, G, Fisher, CR (2006) Alternate unstable states: convergent paths of succession in hydrocarbon-seep tubeworm- associated communities. Journal of Experimental Marine Biology and Ecology, 339 (2), 159-176.

Cordes, EE, Carney, SL, Hourdez, S, Carney, RS, Brooks, JM, Fisher, CR (2007) Cold seeps of the deep Gulf of Mexico: Community structure and biogeographic comparisons to Atlantic equatorial belt seep communities. Deep-Sea Research Part I-Oceanographic Research Papers, 54, 637 - 653.

86 Cordes EE, Bergquist DC, Redding MR, Fisher CR. 2007. Patterns of growth in cold seep vestimentiferans includingSeepiophila jonesi: a second species of long-lived vestimentiferan. Marine Ecology, 28:160–68

Cordes, EE, Bergquist, DC, Fisher, CR (2009) Macro-Ecology of Gulf of Mexico Cold Seeps. Annual Review of Marine Science, 1, 143 - 168.

Cowart, DA, Huang, C, Schaeffer, SW. Identification and amplification of microsatellite loci in deep-sea tubeworms of the genus Escarpia (Polychaeta, Siboglinidae). Conservation Genetics Resources, 5 (2): 479-482

Cowart, D., Huang, C, Arnaud-Haond, S, Carney, SL, Fisher, CR, Schaeffer, SW, (2013) Restriction to large-scale gene flow versus regional panmixia among cold seep Escarpia spp. (Polychaeta, Siboglinidae). Molecular Ecology

Coyne, JA, Orr, HA, (2004) Speciation. Sinauer Associates.

David, P, Berthou, P, Noel, P, and Jarne, P (1997a) Patchy recruitment patterns in marine invertebrates: a spatial test of the density-dependent hypothesis in the bivalve Spisula ovalis. Oecologia, 111(3), 331-340.

David, P, Perdieu, MA, Pernot, AF, and Jarne, P (1997b) Fine-grained spatial and temporal population genetic structure in the marine bivalve Spisula ovalis. Evolution, 51 (4), 1318-1322.

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

Edgar, RC (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research, 32 (5), 1792-1797.

Di Meo, CA, Wilbur, AE, Holben, WE, Feldman, RA, Vrijenhoek, RC, Cary, SC (2000) Genetic variation among endosymbionts of widely distributed vestimentiferan tubeworms. Applied and Environmental Microbiology, 66 (2), 651-658.

Duperron, S, De Beer, D, Zbinden, M, Boetius, A, Schipani, V, Kahil, N, and Gaill, F (2009). Molecular characterization of bacteria associated with the trophosome and the tube of Lamellibrachia sp., a siboglinid annelid from cold seeps in the eastern Mediterranean. FEMS Microbiology Ecology, 69 (3), 395-409.

Edgar, RC (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research, 32 (5), 1792-1797.

Ellegren H (2004) Microsatellites: simple sequences with complex evolution. Nature Reviews: Genetics 5, 435 - 445.

Etter, RJ, Rex, MA, (1990) Population differentiation decreases with depth in deep-sea gastropods. Deep Sea Research Part A. Oceanographic Research Papers 37 (8), 1251 - 1261.

Etter, RR, Rex, MA, Chase, MC, Quattro, JM (1999) A genetic dimension to deep-sea biodiversity. Deep- Sea Research Part I, 46, 1095 - 1099.

Excoffier, L, Smouse, P, Quattro, J (1992) Analysis of molecular variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA restriction data. Genetics, 131, 479 - 491.

Feldman, RS, Shank, TM, Black, MB, Baco, AR, Smith, CR, Vrijenhoek, RC (1998) Vestimentiferan on a Whale Fall. Biological Bulletin, 194, 116 - 119.

Fisher, CR, Childress, JJ, and Sanders, NK (1988) The role of vestimentiferan hemoglobin in providing an environment suitable for chemoautotrophic sulfide-oxidizing endosymbionts. Symbiosis, 5(3), 229-246.

Fisher C (1996) Ecophysiology of primary production at deep-sea vents and seeps. Biosystematics and Ecology series 11, 313-336.

Fisher, C, Roberts, H, Cordes, E, Bernard, B (2007) Cold Seeps and Associated Communities of the Gulf of Mexico. Oceanography, 20, 118 - 129.

Folmer, O, Black, M, Hoeh, W, Lutz, R, Vrijenhoek, R (1994) DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molecular Marine Biology and Biotechnology, 3, 294 - 299.

Freytag, JK, Girguis, PR, Bergquist, DC, Andras, JP, Childress, JJ, Fisher, CR (2001) A paradox resolved: Sulfide acquisition by roots of seep tubeworms sustains net chemoautotrophy. Proceedings of the National Academy of Sciences, 98, 13408 - 13413.

Fujiwara, Y, Kato, C, Masui, N, Fujikura, K, Kojima, S (2001) Dual symbiosis in the cold-seep thyasirid clam Maorithyas hadalis from the hadal zone in the Japan Trench, western Pacific. Marine Ecology Progress Series, 214, 151-159.

Gage, JD, Tyler, PA, (1992) Deep-sea biology: a natural history of organisms at the deep-sea floor. Cambridge University Press.

Galtier, N, Nabholz, N, Glémin, S, Hurst GDD (2009) Mitochondrial DNA as a marker of molecular diversity: a reappraisal. Molecular Ecology 18, 4541 - 4550.

Gardiner, SL, McMullin, E, Fisher, CR (2001) Seepiophila jonesi, a new genus and species of vestimentiferan tube worm (Annelida : Pogonophora) from hydrocarbon seep communities in the Gulf of Mexico. Proceedings of the Biological Society of Washington, 114, 694 - 707.

Gaston, KJ, O'Neill, MA (2004) Automated species identification: why not? Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 359 (1444), 655-667.

Gelfand Y, Rodriguez A, Benson G (2006) TRDB - The Tandem Repeat Database. Nucleic Acids Research 35, D80 - D87.

Goffredi, SK, Childress, JJ, Desaulniers, NT, and Lallier, FJ (1997). Sulfide acquisition by the vent worm Riftia pachyptila appears to be via uptake of HS-, rather than H2S. Journal of Experimental Biology, 200(20), 2609-2616.

Goudet, J (1995) FSTAT (Version 1.2): A computer program to calculate F-Statistics. The Journal of Heredity, 86, 485 - 486.

Haas, A, Peckmann, J, Elvert, M, Sahling, H, Bohrmann, G (2010) Patterns of carbonate authigenesis at the Kouilou pockmarks on the Congo deep-sea fan. Marine Geology, 268, 129 - 136.

Hajibabaei, M, Singer, GAC, Hebert, PDN, Hickey, DA (2007) DNA barcoding: how it complements taxonomy, molecular phylogenetics and population genetics. Trends in Genetics, 23, 167 - 172.

Halanych, KM (2005) Molecular phylogeny of siboglinid annelids (aka pogonophorans): a review. In Morphology, molecules, evolution and phylogeny in Polychaeta and related taxa (pp. 297-307). Springer Netherlands.

Hamilton, P (1990) Deep currents in the Gulf of Mexico. Journal of Physical Oceanography 20, 1180 - 1207.

Hamilton, P (2007) Deep-Current Variability near the Sigsbee Escarpment in the Gulf of Mexico. Journal of Physical Oceanography, 37, 708 - 726.

Hare M (2001) Prospects for nuclear gene phylogeography. Trends in Ecology and Evolution 16, 700 - 706.

Harmer TL, Rotjan RD, Nussbaumer, AD, Bright, M, Ng, AW, DeChaine, EG, Cavanaugh, CM (2008) Free-Living Tube Worm Endosymbionts Found at Deep-Sea Vents. Applied and Environmental Microbiology 3895 - 3898.

Hartl, DL, Clark, AG (2007) Principals of population genetics, fourth edition. Sinauer Associates, Inc. Publishers, Sunderland, Massachusetts

Hausdorf B, Hennig, C (2010) Species delimitation using dominant and codominant multilocus markers. Systematic Biology 59, 491 - 503.

Hebert, PDN, Ratnasingham, S, de Waard, JR (2003) Barcoding animal of life: cytochrome c oxidase subunit 1 divergences among closely related species. Proceedings of the Royal Society of London B, 270, S96 - S99.

Hedgecock D, Barber PH, Edmands S (2007) Genetic approaches to measuring connectivity. Oceanography-Washington DC-Oceanography Society, 20, 70 – 79.

Hellberg, ME, Burton, RS, Neigel, JE, Palumbi, SR (2002) Genetic assessment of connectivity among marine populations. Bulletin of Marine Science, 70 (Supplement 1), 273-290.

Hilario, AY, Young, CM, Tyler, PA (2005) Sperm Storage, Internal Fertilization, and Embryonic Dispersal in Vent and Seep Tubeworms (Polychaeta:Siboglinidae: Vestimentifera). Biological Bulletin, 208, 20 - 28.

Hourdez S, Lallier, FH (2007). Adaptations to hypoxia in hydrothermal vent and cold seep invertebrates. Reviews in Environmental Biotechnology 6, 143 - 159.

Hudson R, Coyne, JA (2002) Mathematical consequences of the genealogical species concept. Evolution 36, 1557 - 1565.

Hurtado, LA, Lutz, RA, Vrijenhoek, RC (2004) Distinct patterns of genetic differentiation among annelids of eastern Pacific hydrothermal vents. Molecular Ecology, 13(9), 2603 - 2615.

Howell, K, Rogers, A, Tyler, P, Billett, DM, (2004) Reproductive isolation among morphotypes of the Atlantic seastar species Zoroaster fulgens (Asteroidea: Echinodermata). Marine Biology 144 (5), 977 - 984.

Jackson, J (1974) Biogeogrpahic consequences of eurytopy and stenotopy among marine bivalves and their evolutionary significance. The American Naturalist, 108, 541 - 560.

Jennings RM, Halanych KM (2005) Mitochondrial genomes of Clymenella torquata (Maldanidae) and Riftia pachyptila (Siboglinidae): Evidence for conserved gene order in Annelida. Molecular Biology and Evolution 22, 210-222.

Johns, GC, Avise, JC (1998) A Comparative Summary of Genetic Distances in the Vertebrates from the Mitochondrial Cytochrome b Gene. Molecular Biology and Evolution, 15, 1481 - 1490.

Jones M (1981) Riftia pachyptila Jones: Observations on the Vestimentiferan Worm from the Galapagos Rift. Science, 213, 333-336.

Jones, M (1985) On the Vestimentifera, new phylum: six new species, and other taxa, from hydrothermal vents and elsewhere. Bulletin of the Biological Society of Washington, (6), 117 - 158.

Johnson, MS, Black, R (1984) Pattern Beneath the Chaos: The Effect of Recruitment on Genetic Patchiness in an Intertidal Limpet. Evolution 38(6), 1371 - 1383.

Julian, D, Gaill, F, Wood, E, Arp, AJ, Fisher, CR (1999) Roots as a site of hydrogen sulfide uptake in the hydrocarbon seep vestimentiferan. Lamellibrachia sp. Journal of Experimental Biology, 202, 2245 - 2257.

Kalinowski, S (2005) HP-RARE 1.0: a computer program for performing rarefaction on measures of allelic richness. Molecular Ecology Resources, 5, 187 - 189.

Keigwin, L (1978) Pliocene closing of the Isthmus of Panama, based on biostratigraphic evidence from nearby Pacific Ocean and Caribbean Sea cores. Geology, 6, 630 - 634.

Kennicutt MC, Brooks, JM, Bidigare, RR, Denoux, GJ (1988). Gulf of Mexico hydrocarbon seep communities—I. Regional distribution of hydrocarbon seepage and associated fauna. Deep Sea Research Part A. Oceanographic Research Papers 35, 1639 - 1651.

Khripounoff, A, Vangriesheim, A, Babonneau, N, Crassous, P, Dennielou, B, Savoye, B (2003) Direct observation of intense turbidity current activity in the Zaire submarine valley at 4000 m water depth. Marine Geology, 194, 151 - 158.

Kim, SL, Mullineaux, LS (1998) Distribution and near-bottom transport of larvae and other plankton at

90 hydrothermal vents. Deep Sea Research II, 45.1 – 3, 423 - 440.

Kivelä, M, Arnaud-Haond, S, Saramäki, J. EDENetwork: Ecological and Evolutionary Networks. in prep.

Knowlton, N (1993) Sibling species in the sea. Annual Review in Ecological Systematics, 24, 189 - 216.

Kocot, KM, Cannon, JT, Todt, C, Cirarella, MR, Kohn AB, Meyer, A, Santos, SR, Schander, C, Moroz, LL, Leib, B, Halanych, KM (2011). Phylogenetics reveals deep molluscan relationships. Nature 477, 452 - 456

Kvenvolden, KA, Simoneit, BR (1990) Hydrothermally Derived Petroleum: Examples from Guaymas Basin, Gulf of California, and Escanaba Trough, Northeast Pacific Ocean (1). AAPG Bulletin, 74 (3), 223-237.

Lalli, CM, Parsons, TR, (1993) Biological oceanography: an introduction. Butterworth-Heinemann.

Li, C, Riethoven, J JM, Ma, L (2010) Exon-primed intron-crossing (EPIC) markers for non-model teleost fishes. BMC evolutionary biology, 10(1), 90.

Liao X, Wang D, Yu X, Li W, Cheng L, Wang J, Tong J (2007) Characterization of novel microsatellite loci in rare minnow (Gobiocypris rarus) and amplification in closely related species in Gobioninae. Conservation Genetics 8, 1003 - 1007.

Lonsdale, P, Becker, K (1985) Hydrothermal plumes, hot springs, and conductive heat flow in the Southern Trough of Guaymas Basin. Earth and Planetary Science Letters, 73 (2), 211-225.

MacDonald, IR (1998) Stability and change in Gulf of Mexico chemosynthetic communities: interim report. Prepared for the Department of the Interior, Minerals Management Service, Gulf of Mexico OCS Region.

Mallet, J. (1995) A species definition for the modern synthesis. Trends in Ecology & Evolution, 10(7), 294-299.

Marsh, AM, Mullineaux, LS, Young, CM, Manahan, DT (2001) Larval dispersal potential of the tubeworm Riftia pachyptila at deep-sea hydrothermal vents. Nature, 411 (6838), 77 - 80.

McGookey DP (1975). Gulf coast Cenozoic sediments and structure: An excellent example of extracontinental sedimentation. Gulf Coast Association of Geological Societies Transactions 25, 104 - 120.

McMullin, ER, Hourdez, S, Schaeffer, SW, Fisher, CR (2003) Phylogeny and biogeography of deep sea vestimentiferan tubeworms and their bacterial symbionts. Symbiosis, 34, 1 - 41.

McMullin, ER, Wood, J, Hamm, S, (2004) Twelve microsatellites for two deep sea polychaete tubeworm species, Lamellibrachia luymesi and Seepiophila jonesi, from the Gulf of Mexico. Molecular Ecology Notes, 4 (1), 1-4.

91 McMullin, E, Nelson, K, Fisher, CR, Schaeffer, SW (2010) Population structure of two deep sea tubeworms, Lamellibrachia luymesi and Seepiophila jonesi, from the hydrocarbon seeps of the Gulf of Mexico. Deep-Sea Research Part I, 57, 1499 - 1509.

Meirmans, P, Hedrick, PW (2011) Assessing population structure: FST and related measures. Molecular Ecology Resources, 11, 5 - 18.

Miglietta, M, Hourdez, S, Cowart, DA, Schaeffer, SW, Fisher, CR (2010) Species boundaries of Gulf of Mexico vestimentiferans (Polychaeta, Siboglinidae) inferred from mitochondrial genes. Deep-Sea Research Part II, 57, 1916 - 1925.

Minerals Management Service Gulf of Mexico (2006) Gulf of Mexico OCS Oil and Gas Lease Sales: 2007-2012. Western Planning Area Sales 204, 207, 210, 215, and 218. Central Planning Area Sales 205, 206, 208, 213, 216, and 222. Draft Environmental Impact Statement. Volume II, pp. 1 - 170, Figure 173-179, page 115. U.S. Department of the Interior.

Miura T, Tsukahara, J, Hashimoto, J (1997) Lamellibrachia Satsuma, A new species of vestimentiferan, Worms Bay, Japan. Proceedings of the Biological Society of Washington 110, 447 - 456.

Miura, T, Nedachi, M, and Hashimoto, J (2002) Sulphur sources for chemoautotrophic nutrition of shallow water vestimentiferan tubeworms in Kagoshima Bay. Journal of the Marine Biological Association of the UK, 82(4), 537-540.

Miura T, Kojima, S (2006) Two new species of vestimentiferan tubeworm (Polychaeta: Siboglinidae aka Pogonophora) from the Brothers Caldera, Kermadec Arc, South Pacific Ocean. Species Diversity 11, 209 - 224.

Moalic, Y, Arnaud-Haond, S, Perrin, C, Pearson, GA, Serrao, EA (2011) Travelling in time with networks: Revealing present day hybridization versus ancestral polymorphism between two species of brown algae, Fucus vesiculosus and F. spiralis. BMC Evolutionary Biology, 11(1), 33.

Moalic, Y, Desbruyères, D, Duarte, CM, Rozenfeld, AF, Bachraty, C, Arnaud-Haond, S (2012) Biogeography Revisited with Network Theory: Retracing the History of Hydrothermal Vent Communities. Systematic Biology, 61(1), 127 - 137.

Moritz C, Cicero C (2004) DNA Barcoding: Promise and Pitfalls. PLoS Biology 2(10), e354, 1529 – 1531.

Mullineaux, LS, France, SC (1995) Dispersal mechanisms of deep-sea hydrothermal vent fauna. Geophysical Monograph Series, 91, 408 - 424.

Oey, LY, Lee, HC (2002) Deep eddy energy and topographic Rossby Waves in the Gulf of Mexico. Journal of Physical Oceanography, 32(12). 3527 - 3499.

Oliphant, A., Thatje, S., Brown, A., Morini, M., Ravaux, J., & Shillito, B. (2011). Pressure tolerance of the shallow-water caridean shrimp Palaemonetes varians across its thermal tolerance window. The Journal of Experimental Biology, 214 (7), 1109-1117.

92 Olu-Le Roy, K, Caprais, JC, Fifis, A, Fabri, MC, Galeron, J, Budzinsky, H, Le Menach, K, Khripounoff, A, Ondréas, H, Sibuet, M (2007) Cold-seep assemblages on a giant pockmark off West Africa: spatial patterns and environmental control. Marine Ecology, 28, 115 - 130.

Olu, KC, Cordes, EE, Fisher, CR, Brooks, JM, Sibuet, M, Desbruyères, D (2010) Biogeography and Potential Exchanges Among the Atlantic Equatorial Belt Cold-Seep Faunas. Public Library of Science One (PLoS), 5, 1 - 11.

Ondréas, H, Olu, K, Fouquet, Y, Charlou, JL, Gay, A, Dennielou, B, Donval, JP, Fifis, A, Nadalig, T, Cochonat, P, Cauquil, E, Bourillet, JF, Le Moigne, M, Sibuet, M (2005) ROV study of a giant pockmark on the Gabon continental margin. Geo-Marine Letters, 25, 281 - 292.

Palumbi, S (1992) Marine speciation on a small planet. Trends in Ecology and Evolution, 7, 114 - 118.

Palumbi, S (1994) Genetic divergence, reproductive isolation, and marine speciation. Annual Review in Ecological Systematics, 25, 547 - 572.

Palumbi SR (1996) Nuclei Acids II: The Polymerase Chain Reaction. Molecular Systematics (ed. Hillis DM, Moritz, C., Mable, B.K.). Sinauer Associates, Inc. Sunderland, Massachusetts

Palumbi, SR, Baker, CS (1994) Contrasting population structure from nuclear intron sequences and mtDNA of humpback whales. Molecular Biology and Evolution, 11(3), 426 - 435.

Paull, CK., Hecker, B, Commeau, R, Freeman-Lynde, RP, Neumann, C, Corso, WP, Golubic, S, Hook, JE, Sikes, E Curray, J (1984). Biological communities at the Florida Escarpment resemble hydrothermal vent taxa. Science 226: 4677 (1984): 965-967.

Pearse, DC, Crandall, KA (2004) Beyond Fst: Analysis of population genetic data for conservation. Conservation Genetics, 5, 585 - 602.

Peter, JM, Simoneit, BRT, Kawka, OE, Scott, SD (1990) Liquid hydrocarbon-bearing inclusions in modern hydrothermal chimmeys and mounds from the southern trough of Guaymas Basin, Gulf of California. Applied Geochemistry, 5 (1), 51-63.

Piry, S, Luikart, G, Cornuet, JM (1999) Computer note. BOTTLENECK: a computer program for detecting recent reductions in the effective size using allele frequency data. Journal of Heredity, 90 (4), 502 - 503.

Pradillon, F, Gaill, F, (2007) Pressure and life: some biological strategies. Life in Extreme Environments, 341 - 355.

Pompanon, F, Bonin, A, Bellemain, E, Taberlet, P (2005) Genotyping errors: causes, consequences and solutions. Nature Reviews: Genetics, 6, 847 - 860.

Primmer C, Painter JN, Koskinen MT, Palo JU, Merilä J (2005) Factors affecting avian cross-species microsatellite amplification. Journal of Avian Biology 36, 348 - 360.

Pritchard, JS, Stephens, M, Donnelly, P (2000) Inference of Population Structure Using Multilocus Genotype Data. Genetics, 155, 945 - 959.

R Development Core Team (2009) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org.

Raymond M, Rousset F (1995) GENEPOP (Version 1.2): Population genetics software for exact tests and ecumenicism. Journal of Heredity 86.

Rex, MA, Etter, RJ, (2010) Deep-sea biodiversity: pattern and scale. Harvard University Press.

Rocha, LA, Robertson, DR, Roman, J, Bowen, BW, (2005) Ecological speciation in tropical reef fishes. Proceedings of the Royal Society B: Biological Sciences 272 (1563), 573 - 579.

Rivas DB, Badan, A, Ochoa, J (2005) The ventilation of the deep Gulf of Mexico. Journal of Physical Oceanography, 35, 1763 - 1781.

Roberts H, Aharon, P (1994) Hydrocarbon-derived carbonate buildups of the northern Gulf of Mexico continental slope: a review of submersible investigations. Geo-Marine Letters, 14, 135 - 148.

Roberts HH, Carney R, Kupchik M, Fisher CR, Nelson K, Becker E, Goehring L, Lessard-Pilon S, Telesnicki, G, Bernard B, Brooks JM, Bright M, Cordes E, Hourdez S, Hunt J, Shedd W, Boland G, Joye S, Samarkin V, Bernier M, Bowles M, MacDonald I, Niemann H, Petersen C, Morrison C, Potter, J (2007) ALVIN explores the deep northern Gulf of Mexico slope. Eos Transactions American Geophysical Union, 88:341–42

Rocha, LA, Robertson, DR, Roman, J, Bowen, BW, (2005) Ecological speciation in tropical reef fishes. Proceedings of the Royal Society B: Biological Sciences 272 (1563), 573 - 579.

Root, RB (1967) The niche exploitation pattern of the blue-gray gnatcatcher. Ecological monographs, 37(4), 317-350.

Rouse GW (2001). A cladistic analysis of Siboglinidae Caullery, 1914 (Polychaeta, Annelida): formerly the phyla Pogonophora and Vestimentifera. Zoological Journal of the Linnean Society 132, 55- 80.

Rouse, GW, Goffredi, SK, Vrijenhoek, RC (2004) Osedax: bone-eating marine worms with dwarf males. Science, 305(5684), 668-671.

Rouse, GW, Worsaae, K, Johnson, SB, Jones, WJ, Vrijenhoek, RC (2008) Acquisition of dwarf male “harems” by recently settled females of Osedax roseus n. sp. (Siboglinidae; Annelida). The Biological Bulletin, 214(1), 67-82.

Rozen S, Skaletsky H (1999) Primer3 on the WWW for general users and for biologist programmers. Bioinformatics Methods and Protocols 132, 365 - 386.

Rozenfeld, AF, Arnaud-Haond, S, Hernández-García, E, Eguíluz, VM, Matías, M A, Serrão, E, Duarte, CM (2007) Spectrum of genetic diversity and networks of clonal organisms. Journal of the Royal Society Interface, 4(17), 1093-1102.

94 Rozenfeld, AF, Arnaud-Haond, S, Hernández-García, E, Eguíluz, VM, Serrão, E A, Duarte, CM (2008) Network analysis identifies weak and strong links in a metapopulation system. Proceedings of the National Academy of Sciences, 105 (48), 18824 - 18829.

Queller, DC, Strassmann, JE, Hughes, CR (1993) Microsatellites and kinship. Trends in Ecology & Evolution, 8(8), 285-288.

Sammarco, PW, Brazeau, DA, Sinclair, J (2012) Genetic connectivity in scleractinian corals across the northern Gulf of Mexico: oil/gas platforms, and relationship to the Flower Garden Banks. PLoS One, 7(4), e30144.

Sang, T, (2002) Utility of low-copy nuclear gene sequences in plant phylogenetics. Critical Reviews in Biochemistry and Molecular Biology 37 (3), 121-147.

Sassen, R, Roberts, HH, Aharon, P, Larkin, J, Chinn, EW, Carney, R (1993) Chemosynthetic bacterial mats at cold hydrocarbon seeps, Gulf of Mexico continental slope. Organic Geochemistry, 20, 77 - 89.

Sassen, R, MacDonald, IR, Requejo, AG, Guinasso, NL, Kennicutt, MC, Sweet, ST, Brooks, JM (1994a) Organic geochemisty of sediments from chemosynthetic communities, Gulf of Mexico slope. Geo-Marine Letters, 14, 110 - 119.

Saunders, PM, Fofonoff, NP, (1976) Conversion of pressure to depth in the ocean. Deep-Sea Research 23, 109 – 111.

Schaeffer SW, Goetting-Minesky SWP, Graybill JL, Miller JM, Kim K, Miller JG, Anderson WW (2003) Evolutionary genomics of inversions in Drosophila pseudoobscura: Evidence for epistasis. Proceedings of the National Academy of Sciences USA 100: 8319-8324.

Scheltema, R (1968) Dispersal of larvae by equatorial ocean currents and its importance to the zoogeography of shoal-water tropical species. Nature, 217, 1159 - 1162.

Schluter, DN, Nagel, LM, (1995) Parallel speciation by natural selection. The American Naturalist 146 (2), 292 - 301.

Schneider, S, Lischer, H (2009) ARLEQUIN, version 3.5. An integrated software package for population genetics data analysis. University of Berne, Berne.

Schuelke M (2000) An economic method for the fluorescent labeling of PCR fragments. Nature Biotechnology 18, 233 - 234.

Scott KM, Fisher CR (1995) Physiological ecology of sulfide metabolism in hydrothermal vent and cold seep vesicomyid clams and vestimentiferan tube worms. American Zoologist 35, 102-111.

Shank, TH, Halanych, KM (2007) Toward a mechanistic understanding of larval dispersal: insights from genomic fingerprinting of the deep-sea hydrothermal vent tubeworm Riftia pachyptila. Marine Ecology, 28, 25 - 35.

95 Shaw PW, Pierce, GJ, Boyle, PR (1999) Subtle population structuring within a highly vagile marine invertebrate, the veined squid Loligo forbesi, demonstrated with microsatellite DNA markers. Molecular Ecology 8, 407 - 417.

Shilling, FM, Manahan, DT (1994) Energy metabolism and amino acid transport during early development of Antarctic and temperate echinoderms. The Biological Bulletin 187 (3), 398 - 407.

Sibuet M, Olu, K (1998) Biogeography, biodiversity and fluid dependence of deep-sea cold-seep communities at active and passive margins. Deep-Sea Research Part II 45, 517 - 567.

Sibuet, M, Vangriesheim, A (2009) Deep-sea environment and biodiversity of the West African Equatorial margin. Deep-Sea Research Part II: Topical Studies in Oceanography, 56, 2156 - 2168.

Simoneit, BRT, Lonsdale, PF, Edmond, JM, Shanks, WC (1990) Deep-water hydrocarbon seeps in Guaymas Basin, Gulf of California. Applied Geochemistry, 5(1), 41-49.

Simonsen KL, Churchill GA, Aquadro CF (1995) Properties of statistical tests of neutrality for DNA polymorphism data. Genetics 141: 413-429

Somero, GN, (1992a) Biochemical ecology of deep-sea animals. Cellular and Molecular Life Sciences 48 (6), 537 - 543.

Somero, GN, (1992b) Adaptations to high hydrostatic pressure. Annual review of physiology 54 (1), 557 - 577.

Southward EC (1982) Bacterial Symbionts in Pogonophora. Journal of the Marine Biological Association of the United Kingdom 62, 889 - 906.

Southward EC (1991) Three new species of Pogonophora, including two vestimentiferans, from hydrothermal sites in the Lau Back-arc Basin (Southwest Pacific Ocean). Journal of Natural History 25, 859 - 881.

Southward EC, Andersen, AC, Hourdez, S (2011) Lamellibrachia anaximandri n. sp., a new vestimentiferan tubeworm (Annelida) from the Mediterranean, with notes on frenulate tubeworms from the same habitat. Zoosystema 33, 245 - 279.

Stauffer D, Aharony A (1994) Introduction to Percolation Theory, Taylor & Francis, CRC Press, London

Storey, JD, Tibshirani, R (2003) Statistical significance for genomewide studies. Proceedings of the National Academy of Sciences, 100 (16), 9440 - 9445.

Stramma, L, England, M (1999) On the water masses and mean circulation of the South Atlantic Ocean. Journal of Geophysical Research, 104, 20.

Tajima, F (1989) Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics, 123 (3) 585 - 595.

96 Tamura, K, Nei, M. (1993) Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Molecular Biology and Evolution, 10, 512 - 526.

Tamura, K, Nei, M, Kumar, S (2004) Prospects for inferring very large phylogenies by using the neighbor-joining method. Proceedings of the National Academy of Sciences of the United States of America, 101, 11030 - 11035.

Tamura, K, Peterson, D, Perterson, N, Stecher, G, Nei, M, Kumar, S (2011) MEGA5: Molecular Evolutionary Genetics Analysis using Maximum Likelihood, Evolutionary Distance, and Maximum Parsimony Methods. Molecular Biology and Evolution, 28, 2731 - 2739.

Tarasov, VG, Gebruk, AV, Mironov, AN, Moskalev, LI (2005) Deep-sea and shallow-water hydrothermal vent communities: two different phenomena? Chemical Geology, 224(1), 5-39.

Templeton, AR, Crandall, KA, Sing, CF, (1992) A Cladistic Analysis of Phenotypic Associations With Haplotypes Inferred From Restriction Endonuclease Mapping and DNA Sequence Data. III: Cladogram Estimation. The Genetics Society of America.

Thompson, JD, Higgins, DG, Gibson, TJ (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Research, 22, 1.

Tóth, G, Gáspári, Z, Jurka, J (2000) Microsatellites in different eukaryotic genomes: survey and analysis. Genome Research, 10 (7), 967-981.

Tunnicliffe, V (1988) Biogeography and evolution of hydrothermal-vent fauna in the eastern Pacific Ocean. Proceedings of the Royal Society of London B, 233, 347 - 366.

Turnipseed, M, Knick, KE, Lipcius, RN, Dreyer, J, and Van Dover, CL (2003) Diversity in mussel beds at deep‐sea hydrothermal vents and cold seeps. Ecology Letters 6 (6), 518-523.

Turnipseed M, Jenkins C, Van Dover C (2004) Community structure in Florida Escarpment seep and Snake Pit (Mid-Atlantic Ridge) vent mussel beds. Marine Biology 145, 121-132.

Tyler, PA, Young, CM, (1998) Temperature and pressure tolerances in dispersal stages of the genus Echinus (Echinodermata: Echinoidea): prerequisites for deep-sea invasion and speciation. Deep- Sea Research Part II 45 (1 - 3), 253 - 277.

Tyler, PY, Young, CM (1999) Reproduction and dispersal at vents and cold seeps. Journal of Marine Biological Association of the U.K. 79, 193 - 208.

Valentini, A, Pompanon, F, Taberlet, P (2009) DNA barcoding for ecologists. Trends in Ecology & Evolution, 24 (2), 110-117.

Van der Land, J, NØrrevang, A (1975) The systematic position of Lamellibrachia [sic] (Annelida, Vestimentifera). Journal of Zoological Systematics EvolForsch (Sonderheft), 86 - 101.

97 van Oosterhout, C, Hutchinson, WF, Willis, DRM, Shipley, P (2004) MICROCHECKER: software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology 4, 535 - 538.

Vangriesheim, A, Khripounoff, A, Crassous, P (2009) Turbidity events observed in situ along the Congo submarine channel. Deep-sea Research II, 56, 2208 - 2222.

Vrijenhoek, RC (2010) Genetic diversity and connectivity of deep-sea hydrothermal vent metapopulations. Molecular Ecology, 19, 4391 - 4411.

Vrijenhoek, RC, Duhaime, M, Jones, WJ (2007) Subtype Variation Among Bacterial Endosymbionts of Tubeworms (Annelida: Siboglinidae) from the Gulf of California. Biological Bulletin, 212, 180 - 184.

Vrijenhoek, RS, Schutz, SJ, Gustafson, RG, Lutz, RA (1994) Cryptic species of deep-sea clams (Mollusca: Bivalvia: Vesicomyidae) from hydrothermal vent and cold-water seep environments. Deep-Sea Research I, 41, 1171 - 1189.

Wares, JG, Gaines, SD, Cunningham, CW (2001) A comparative study of asymmetric migration events across a marine biogeographic boundary. Evolution, 55, 295 - 306.

Watts DJ (2004) The "new" science of networks. Annual Review of Sociology, 30, 243-270.

Webb M (1969). Lamellibrachia barhami, gen. nov., sp. nov. (Pogonophora), from the northeast Pacific. Bulletin of Marine Science 19, 18 - 47.

Weir, B, Cockerham, CC (1984) Estimating F-statistics for the analysis of population structure. Evolution, 38, 1358 - 1370.

Welsh, S, Inoue, M (2000) Loop Current rings and the deep circulation in the Gulf of Mexico. Journal of Geophysical Research, 105, 951 - 959.

Wilson, GD, Hessler, RR (1987) Speciation in the deep sea. Annual Review of Ecology and Systematics, 185-207.

Wright, S (1951) The genetic structure of populations. Annual Eugeneics, 15, 323 - 354.

Young, CM, Tyler, PA, (1993) Embryos of the deep-sea echinoid Echinus affinis require high pressure for development. Limnology and oceanography, 178 - 181.

Young, CM, Vazquez, E, Metaxas, A, Tyler, PA (1996) Embryology of vestimentiferan tubeworms from deep-sea methane/sulphide seeps. Nature, 381, 514 - 516.

Young, C, Tyler, PA, Fenaux, L, (1997) Potential for deep sea invasion by Mediterranean shallow water echinoids: pressure and temperature as stage-specific dispersal barriers. Marine Ecology Progress Series 154 (197 - 209).

98 Young, CM, Ekaratne, SUK, Cameron, JL. 1998. Thermal tolerances of embryos and planktotrophic larvae of Archaeopneustes hystrix (Spatangoidea) and Stylocidaris lineata (Cidaroidea), bathyal echinoids from the Bahamian Slope. Journal of Experimental Marine Biology and Ecology, 223,65 - 76.

Young, CM, He, R, Emlet, RB, Li, Y, Qian, H, Arellano, SM, Gaest, AV, Bennett, KC, Wolf, M, Smart, TI, Rice, ME (2012) Dispersal of deep-sea larvae from the Intra-American seas: simulations of trajectories using ocean models. Integrative and Comparative Biology, 52(4), 483 - 496.

Zal, F, Suzuki, T, Kawasaki, Y, Childress, JJ, Lallier, FH, and Toulmond, A (1997) Primary structure of the common polypeptide chain b from the multi‐hemoglobin system of the hydrothermal vent tube worm Riftia pachyptila: An insight on the sulfide binding‐site. Proteins: Structure, Function, and Bioinformatics 29 (4), 562-574.

APPENDIX A

CHAPTER 2

Figure A2-1: Allele frequencies for each of the nine cross amplified loci in Escarpia spicata

100

Figure A2-2: Allele frequencies for each of the nine cross amplified loci in Escarpia laminata

101

!"#$%&'$(")*+,-$%.$/ #"!!!

!"+!!

!"*!!

!")!!

!"(!!

!"'!! !"#$%#&'( !"&!!

!"%!!

!"$!!

!"#!!

!"!!!

)**#*# #' # #%' #(& #*% #+( $!* $$! $+) #*$ $!( $#( $$* $&* $** #$* #%+ #&+ #') #)# #*' #%% #&( #(% #'% #)* $!% #+% $#) $%% $'# $)% $+# %#% #+% $$# $&# $(# #$# #$+ #%( #&' #'% #(% #$& #+!

,6'%7 +,-.-/0 +,-.-/. +,-.-/1 +,-.-/02 +,-.-/0. +,-.-/.0 +,-.-/3- +,-.-/45 +,-.-/42

Figure A2-3: Allele frequencies for each of the nine cross amplified loci in Escarpia southwardae

APPENDIX B

CHAPTER 3

Data Accessibility

The data produced during the course of this project is in the format of DNA sequences, microsatellite genotype calls, figures, maps, photos and tables, which include organized raw data and statistical output. Additionally, there is metadata for every siboglinid tubeworm used for this study. To store and disseminate this data to the public, we used three resources other than the supporting information of the manuscript: NCBI GenBank, TreeBase and Dryad. NCBI GenBank (http://www.ncbi.nlm.nih.gov/genbank/) is a public database funded by the NIH, to allow DNA/genetic sequences data to be uploaded and accessed by the scientific community. A total of 238 new DNA sequences were analyzed in this study have been uploaded onto the GenBank data repository. The accession numbers are KC357321-KC357424, KC870942-KC871002, KC900290 - KC900365. TreeBase is a data repository for phylogenetic data (www.treebase.org). All phylogenetic trees and alignments reside under Tree Base submission number 14129. Lastly, the data repository Dryad (http://www.datadryad.org/) is used for scientific data published in peer- reviewed articles. Unlike other repositories, Dryad allows all types of data to be submitted for public viewing, including methodologies and geographical information associated with samples. We have uploaded spreadsheets containing the raw microsatellite genotypes and metadata (cryovial ID, sample ID, date, region and site of collection, deep sea vehicle, dive number, coordinates, depth, collection method, morphotype, tissue sample and preservation method), under Dryad doi:10.5061/dryad.586b5. These resources are currently available to the international public in addition to the published manuscript, and experimental protocols will be available upon request.

103

ESCARPIA_SOUTHWARDAE_KC870954 ESCARPIA_SOUTHWARDAE_KC870953 ESCARPIA_SOUTHWARDAE_KC357343 ESCARPIA_SOUTHWARDAE_KC357342 ESCARPIA_SOUTHWARDAE_KC357338 ESCARPIA_SOUTHWARDAE_KC870955 ESCARPIA_SOUTHWARDAE_KC357337 ESCARPIA_SOUTHWARDAE_KC870956 ESCARPIA_SOUTHWARDAE_KC357335 ESCARPIA_SOUTHWARDAE_KC357334 ESCARPIA_SOUTHWARDAE_KC870957 ESCARPIA_LAMINATA_GU059207 ESCARPIA_SOUTHWARDAE_KC357336 ESCARPIA_SOUTHWARDAE_KC870958 62 ESCARPIA_SOUTHWARDAE_KC357341 ESCARPIA_LAMINATA_GU059199 ESCARPIA_LAMINATA_GU059193

19 ESCARPIA_LAMINATA_GU059163 ESCARPIA_LAMINATA_GU059189 ESCARPIA_LAMINATA_GU059197 ESCARPIA_LAMINATA_GU059203 ESCARPIA_LAMINATA_GU059211 ESCARPIA_LAMINATA_KC357321 ESCARPIA_LAMINATA_KC357322 ESCARPIA_LAMINATA_KC357323 ESCARPIA_LAMINATA_KC357324 ESCARPIA_LAMINATA_KC357325 25 ESCARPIA_LAMINATA_KC357326 ESCARPIA_LAMINATA_KC357327 ESCARPIA_LAMINATA_KC357328 ESCARPIA_SPICATA_KC357329 ESCARPIA_SPICATA_KC357330 ESCARPIA_SPICATA_KC357331 38 ESCARPIA_SPICATA_KC357332 ESCARPIA_SPICATA_KC357333 ESCARPIA_SOUTHWARDAE_KC357344 ESCARPIA_SPICATA_KC870959 ESCARPIA_SPICATA_KC870960 52 ESCARPIA_SPICATA_KC870961 ESCARPIA_LAMINATA_GU059213 ESCARPIA_SOUTHWARDAE_KC357339 ESCARPIA_SOUTHWARDAE_KC357340 SEEPIOPHILA_JONESI_GU059186 SEEPIOPHILA_JONESI_GU059187 SEEPIOPHILA_JONESI_GU059191 99 SEEPIOPHILA_JONESI_GU059188 SEEPIOPHILA_JONESI_GU059181 SEEPIOPHILA_JONESI_GU059184

0.01 Figure A3-1: COI maximum likelihood (ML) tree for 42 Escarpia and five Seepiophila, with fragment size of 658bp. 34 Escarpia sequences are newly isolated. Seepiophila jonesi is the outgroup taxon, and GenBank accession numbers follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions per nucleotide site.

104

ESCARPIA_LAMINATA_GU068183 ESCARPIA_LAMINATA_GU068195 ESCARPIA_LAMINATA_GU068193 ESCARPIA_LAMINATA_GU068187 ESCARPIA_LAMINATA_GU068181 ESCARPIA_LAMINATA_GU068179 ESCARPIA_LAMINATA_GU068177 ESCARPIA_LAMINATA_KC357346 ESCARPIA_LAMINATA_KC357353 ESCARPIA_LAMINATA_KC357352 ESCARPIA_LAMINATA_KC357351 ESCARPIA_LAMINATA_KC357350 ESCARPIA_LAMINATA_KC357349 ESCARPIA_LAMINATA_KC357348 ESCARPIA_LAMINATA_KC357347 ESCARPIA_SPICATA_KC357368 ESCARPIA_SPICATA_KC357366 ESCARPIA_LAMINATA_KC357345 ESCARPIA_SPICATA_KC357364 ESCARPIA_SPICATA_KC357365 ESCARPIA_SPICATA_KC357367 ESCARPIA_SPICATA_KC357369 99 ESCARPIA_SPICATA_KC357370 ESCARPIA_SPICATA_KC357371 ESCARPIA_SPICATA_KC870942 ESCARPIA_SPICATA_KC870943 ESCARPIA_SOUTHWARDAE_KC870944 ESCARPIA_SOUTHWARDAE_KC357363 ESCARPIA_SOUTHWARDAE_KC870945 ESCARPIA_SOUTHWARDAE_KC357362 ESCARPIA_SOUTHWARDAE_KC870946 ESCARPIA_SOUTHWARDAE_KC870947 ESCARPIA_SOUTHWARDAE_KC870948 ESCARPIA_SOUTHWARDAE_KC870949 ESCARPIA_SOUTHWARDAE_KC870950 ESCARPIA_SOUTHWARDAE_KC870951 ESCARPIA_SOUTHWARDAE_KC870952 ESCARPIA_SOUTHWARDAE_KC357357 ESCARPIA_SOUTHWARDAE_KC357358 ESCARPIA_SOUTHWARDAE_KC357361 ESCARPIA_SOUTHWARDAE_KC357354 ESCARPIA_SOUTHWARDAE_KC357356 ESCARPIA_SOUTHWARDAE_KC357360 ESCARPIA_SOUTHWARDAE_KC357359 ESCARPIA_SOUTHWARDAE_KC357355 SEEPIOPHILA_JONESI_GU068287 SEEPIOPHILA_JONESI_GU068282 SEEPIOPHILA_JONESI_GU068275 SEEPIOPHILA_JONESI_GU068285 SEEPIOPHILA_JONESI_GU068273

0.002

105

ESCARPIA_LAMINATA_KC870962 ESCARPIA_LAMIANTA_KC357395 ESCARPIA_LAMINATA_KC357397 ESCARPIA_LAMINATA_KC357396 ESCARPIA_LAMIANTA_KC357394 ESCARPIA_LAMINATA_KC870963 ESCARPIA_LAMINATA_KC870964 ESCARPIA_LAMINATA_KC870965 ESCARPIA_LAMINATA_KC870966 ESCARPIA_LAMIANTA_KC357393 ESCARPIA_LAMINATA_KC870967

43 ESCARPIA_LAMIANTA_KC357392 ESCARPIA_LAMIANTA_KC357391 ESCARPIA_LAMINATA_KC870968 ESCARPIA_LAMINATA_KC870969

62 ESCARPIA_SPICATA_KC357379 ESCARPIA_SPICATA_KC357382 7 ESCARPIA_SPICATA_KC357378 59 63 ESCARPIA_SPICATA_KC870970 ESCARPIA_LAMINATA_KC870971 ESCARPIA_LAMIANTA_KC357390 ESCARPIA_SPICATA_KC357380 ESCARPIA_SPICATA_KC870972 ESCARPIA_SPICATA_KC870973 ESCARPIA_SPICATA_KC357381 ESCARPIA_SPICATA_KC357377 55 ESCARPIA_SPICATA_KC870974 ESCARPIA_SOUTHWARDAE_KC357386 ESCARPIA_SOUTHWARDAE_KC357384 ESCARPIA_SOUTHWARDAE_KC357388 ESCARPIA_SOUTHWARDAE_KC357389 ESCARPIA_SOUTHWARDAE_KC357383 ESCARPIA_SOUTHWARDAE_KC357385 ESCARPIA_SOUTHWARDAE_KC357387 ESCARPIA_SOUTHWARDAE_KC870975 ESCARPIA_SOUTHWARDAE_KC870976 ESCARPIA_SOUTHWARDAE_KC870977 ESCARPIA_SOUTHWARDAE_KC870978 ESCARPIA_SOUTHWARDAE_KC870979 SEEPIOPHILA_JONESI_KC357374 SEEPIOPHILA_JONESI_KC357373 SEEPIOPHILA_JONESI_KC357372 SEEPIOPHILA_JONESI_KC357376 100 SEEPIOPHILA_JONESI_KC357375 SEEPIOPHILA_JONESI_KC870980 SEEPIOPHILA_JONESI_KC870981

0.01 Figure A3-3: CYTB maximum likelihood (ML) tree for 39 Escarpia and seven Seepiophila, with fragment size of 401bp. All sequences are newly isolated. Seepiophila jonesi is the outgroup taxon, and GenBank Accession numbers follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions per nucleotide site.

106

ESCARPIA_LAMINATA_KC357414 ESCARPIA_LAMINATA_KC357418 ESCARPIA_LAMINATA_KC357413 ESCARPIA_LAMINATA_KC357419 ESCARPIA_LAMINATA_KC357416 ESCARPIA_LAMINATA_KC357417 ESCARPIA_LAMINATA_KC357415 ESCARPIA_LAMINATA_KC357420 63 ESCARPIA_LAMINATA_KC870982 ESCARPIA_LAMINATA_KC870983 ESCARPIA_LAMINATA_KC870984 ESCARPIA_LAMINATA_KC870985 ESCARPIA_LAMINATA_KC870986 ESCARPIA_LAMINATA_KC870987 ESCARPIA_LAMINATA_KC870988 ESCARPIA_LAMINATA_KC870989 ESCARPIA_LAMINATA_KC870990 ESCARPIA_SOUTHWARDAE_KC870991 ESCARPIA_SOUTHWARDAE_KC870992 ESCARPIA_SOUTHWARDAE_KC870993 64 ESCARPIA_SOUTHWARDAE_KC357410 ESCARPIA_SOUTHWARDAE_KC357408 ESCARPIA_SOUTHWARDAE_KC357409 ESCARPIA_SOUTHWARDAE_KC870994 ESCARPIA_SOUTHWARDAE_KC870995 ESCARPIA_SOUTHWARDAE_KC870996 ESCARPIA_SOUTHWARDAE_KC357406 ESCARPIA_SOUTHWARDAE_KC357412 ESCARPIA_SOUTHWARDAE_KC357404 ESCARPIA_SOUTHWARDAE_KC357405 ESCARPIA_SOUTHWARDAE_KC357407 ESCARPIA_SOUTHWARDAE_KC357411 ESCARPIA_SOUTHWARDAE_KC357403 ESCARPIA_SPICATA_KC357401 ESCARPIA_SPICATA_KC357400 ESCARPIA_SPICATA_KC357399 ESCARPIA_SPICATA_KC870997 ESCARPIA_SPICATA_KC870998 ESCARPIA_SPICATA_KC870999 ESCARPIA_SPICATA_KC871000 ESCARPIA_SPICATA_KC357398 ESCARPIA_SPICATA_KC357402 66 ESCARPIA_SPICATA_KC871001 SEEPIOPHILA_JONESI_KC357424 SEEPIOPHILA_JONESI_KC357421 SEEPIOPHILA_JONESI_KC357422 99 SEEPIOPHILA_JONESI_KC357423 SEEPIOPHILA_JONESI_KC871002

0.001 Figure A3-4: HbB2i maximum likelihood (ML) tree for 43 Escarpia and five Seepiophila, with fragment size of 665bp. All sequences are new isolated. Seepiophila jonesi is the outgroup taxon, and GenBank Accession number follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions per nucleotide site.

Gene: COI n = 42 Gene: CYTB n = 39

E.laminata E.southwardae E.spicata S.jonesi E.laminata E.southwardae E.spicata S.jonesi E.laminata 0.000 0.000 0.033 E.laminata 0.003 0.001 0.632

E.southwardae 0.001 0.000 0.033 E.southwardae 0.005 0.003 0.687 E.spicata 0.001 0.001 0.033 E.spicata 0.002 0.007 3.246 S.jonesi 0.108 0.108 0.108 S.jonesi 0.145 0.127 0.134

Gene: 16S n = 45 Gene: HbB2i n = 43

E.laminata E.southwardae E.spicata S.jonesi E.laminata E.southwardae E.spicata S.jonesi

E.laminata 0.000 0.000 0.007 E.laminata 0.001 0.002 0.006

E.southwardae 0.000 0.000 0.007 E.southwardae 0.002 0.002 0.005 E.spicata 0.000 0.000 0.007 E.spicata 0.004 0.002 0.005 S.jonesi 0.021 0.021 0.021 S.jonesi 0.016 0.014 0.013

Table A3-2: Locations, number of individuals and GenBank IDs for all tubeworms tested in each of the four genes.

Gene: COI n = 42 Gene: CYTB n = 39

# of # of Population Location GenBank IDs Population Location GenBank IDs individuals individuals

KC357377 - KC357382 Transform KC357329 -KC357333, Transform E. spicata 8 E. spicata 10 KC870970, KC870972, Fault KC870959-KC870961 Fault KC870974 KC357384 - KC357387 DC673 7 KC357322 - KC357328 DC673 8 KC357391 - KC357393 KC870968-KC870969 AT340 2 GU059203, GU059207 AC601 2

KC870964, KC870967 AC601 3 GU059211, KC357321 WR269 2

E. laminata E. laminata KC870963, KC870965 WR269 2 GU059199, GU059197 GC852 2

KC870966, KC870971 GC852 1 GU059193 AC818 2

KC870962, KC870973 AC818 2 GU059163, GU059189 MC294 1

KC357334 -KC357336, Regab 4 KC870957 Regab 5 KC357385 - KC357389

KC357337 -KC357343, KC357384, E. Worm Worm 3 KC870955, KC870956, E. southwardae 4 KC870975-KC870977 southwardae Hole KC870958 Hole

KC357344, KC357383, Baboon 10 KC870953 -KC870954 Baboon 3 KC870978-KC870979

Gene: 16S n = 45 Gene: HbB2i n = 43 # of # of Population Location GenBank IDs Population Location GenBank IDs individuals individuals Transform KC357364 - KC357371 Transform KC57398 - KC57402, E. spicata 10 E. spicata 10 Fault KC870942 - KC870943 Fault KC870997- KC871001

DC673 7 KC357347 - KC357351 DC673 8 KC57413 - KC57420 KC870983, KC870984 AT340 2 GU068183, GU068187 AT340 2

KC870985, KC870986 AC601 2 GU068195, GU068193 AC601 2 E. laminata E. laminata GU068181, GU068179, WR269 3 WR269 1 KC870987 KC357346 GC852 1 GU068177 AC818 1 KC870989 KC870982, KC870990 MC294 1 KC357345 MC294 3

KC357354 - KC357356, Regab 7 KC357358, KC357362, Regab 5 KC57408 - KC57412 KC870944, KC870949

E. Worm Worm KC57403 - KC57407, 1 KC357357 E. southwardae 3 southwardae Hole Hole KC870991- KC870993 KC357359 - KC357361, KC870994 - KC870997 KC357363, Baboon 11 Baboon 8 KC870945 -KC870948, KC870950 - KC870952

109

Gene: COI E_laminata E.southwardae E.spicata S.jonesi Mean s.d. Sample Size 8 12 5 4 7.250 3.594 S 1 3 0 1 1.250 1.260

Pi 1.071 0.636 0.000 40.000 10.426 19.720

Tajima’s D 1.166 -1.180 0.000 1.633 0.405 1.259

p-value 0.926 0.108 1.000 0.960 0.748 0.428

Gene: 16S E_laminata E.southwardae E.spicata S.jonesi Mean s.d.

Sample Size 18 20 16 10 16 4.320

S 0 1 0 0 0.250 0.500 Pi 0.000 0.100 0.000 0.000 0.0250 0.0500 Tajima’s D 0.000 -1.164 0.000 0.000 -0.291 -0.582

p-value 1.000 0.144 1.000 1.000 0.786 0.428

Gene: CYTB E_laminata E.southwardae E.spicata S.jonesi Mean s.d.

Sample Size 8 7 6 5 6.500 1.291

S 0 0 3 0 0.750 1.500 Pi 0.000 0.000 1.400 0.000 0.350 0.700 Tajima’s D 0.000 0.000 0.338 0.000 0.084 0.169 p-value 1.000 1.000 0.655 1.000 0.913 0.175 Gene: hbB2i E_laminata E.southwardae E.spicata S.jonesi Mean s.d. Sample Size 16 20 10 8 13.5 5.507 S 0 0 2 0 0.500 1.000 Pi 0.000 0.000 1.088 0.000 0.435 0.871 Tajima’s D 0.000 0.000 1.742 0.000 0.435 0.871

Table A3-4: Summary statistics for nine microsatellite loci amplified in three described Escarpia species

Population EL454_2 EL454_5 EL454_9 EL454_21 EL454_25 EL454_52 EL454_64 EL454_70 EL454_71 Mean

E. spicata NA 9 7 7 6 8 9 12 11 3 8

N = 20 HO 0.900 0.700 0.842 0.300 0.850 0.667 0.684 0.667 0.600 0.690

HE 0.830 0.765 0.816 0.680 0.827 0.860 0.900 0.871 0.476 0.780

FIS -0.083 0.085 0.040 *0.560 -0.028 0.317 0.282 0.325 0.308 0.194

E. laminata NA 12 15 19 9 12 16 14 18 4 13

N = 129 HO 0.700 0.730 0.631 0.752 0.610 0.821 0.853 0.872 0.162 0.681

HE 0.725 0.735 0.920 0.800 0.593 0.886 0,862 0.915 0.342 0.753

FIS 0.039 0.010 *0.396 0.060 -0.009 0.072 0.067 0.081 0.751 0.149

E. southwardae NA 15 11 10 9 7 18 17 10 4 12

N = 80 HO 0.825 0.862 0.605 0.710 0.740 0.873 0.720 0.712 0.276 0.702

HE 0.870 0.876 0.874 0.780 0.741 0.906 0.746 0.726 0.262 0.753

FIS 0.049 0.015 *0.350 0.154 0.005 0.051 0.127 0.019 0.210 0.104

Abbreviations: number of alleles observed per locus (NA), observed (HO) and expected (HE) heterozygosity, Wright’s Inbreeding Coefficient (FIS). *bold asterisk indicates a significant difference after applying FDR correction = 0.01

111

Escarpia laminata Escarpia spicata Escarpia southwardae

Escarpia laminata --- 0.422 0.455

Escarpia spicata 0.115 --- 0.501 Escarpia southwardae 0.130 0.134 ---

Table A3-6: Primers used in this study Gene Primers Forward (5’ – 3’) Reverse (5’ – 3’) Reference COI LCO GGT CAA CAA ATC ATA AAG TAA ACT TCA GGG TGA Folmer et al., 1994 HCO ATA TT GG CCA AAA AA TCA

16S 16sar CGC CTG TTT ATC AAA AAA ACG TGA TCT GAG TTC AGA Palumbi, 1996 16sbr CAT CCG G Cytb CytbF GGW TAY GTW YTW CCW GCR TAW GCR AAW ARR Boore and Brown, CytbR TGR GGW CAR AT AAR TAY CAY TCW GG 2000 CYTB CYTBF AAA TCC TCC CTC ACA GTC CCA GCC CCG ATG AAT CTT This study CYTBR ATT TC hbB2 intron hbB2F CAG TTC ACT GGA CGT CGT CTA TGC AGG CAG ACC S. Carney, 2005 (Riftia/Ridgeia) hbB2R CT ATC G (dissertation) hbB2i (Escarpia) hbB2i_F TCC ATC GCC CCA GGC TGT GCC TTG AAT TCG TTG CTG This study hbB2i_R CTT C TT

112

Table A3-7: Summary statistics for eleven microsatellite loci amplified in eight populations of Escarpia laminata (GoM). Abbreviations: number of alleles observed per locus (N ), observed (H ) and expected (H ) heterozygosity, Wright’s Inbreeding Coefficient (F ). * bold asterisk indicates a significant difference A O E IS after applying FDR correction = 0.01

Population EL454_2 EL454_5 EL454_6 EL454_9 EL454_21 EL454_25 EL454_52 EL454_54 EL454_67 EL454_70 EL454_71 Mean

AC601 NA 6 8 3 12 5 5 11 11 2 13 2 7.3 N = 12 HO 0.583 0.833 0.454 0.750 0.750 0.420 0.750 0.920 0.111 0.920 0.000 0.615 HE 0.710 0.793 0.627 0.931 0.822 0.492 0.884 0.891 0.111 0.920 0.160 0.686 FIS 0.179 -0.050 0.392 0.195 0.088 0.154 0.152 -0.028 0.816 0.000 1.000 0.171 AC818 NA 5 5 3 8 6 4 9 6 2 4 2 5.1 N =8 H 0.625 0.750 0.500 0.750 0.625 0.625 0.750 0.500 0.375 0.750 0.143 0.602 O HE 0.742 0.792 0.425 0.883 0.820 0.575 0.883 0.818 0.325 0.708 0.494 0.693 FIS 0.157 0.053 -0.176 0.151 0.235 -0 0.151 0.559 -0.154 -0.059 0.795 0.191 WR269 NA 5 6 4 8 4 4 9 6 2 10 3 5.7 N =7 HO 0.571 0.857 0.333 1.000 0.714 1.000 0.571 0.800 0.285 1.000 0.167 0.658 HE 0.593 0.791 0.454 0.924 0.703 0.692 0.912 0.890 0.264 0.945 0.621 0.725 FIS 0.037 -0.083 0.519 0.082 0.016 -0.444 0.373 0.350 -0.083 -0.058 0.800 0.182 GC852 N 7 8 3 6 7 5 9 10 3 10 3 6.7 A N =11 HO 0.545 0.818 0.400 0.500 0.730 0.600 1.000 0.700 0.222 1.000 0.100 0.626

HE 0.714 0.814 0.360 0.803 0.840 0.600 0.900 0.890 0.216 0.895 0.280 0.682 FIS 0.152 0.062 0.229 0.247 0.067 0.160 -0.009 0.160 0.171 0.171 0.250 0.151 AT340 NA 6 7 5 11 5 4 6 8 4 10 2 6.6 N =11 HO 0.730 0.636 0.545 0.890 0.818 0.636 0.730 0.800 0.454 0.730 0.300 0.680 H 0.684 0.805 0.467 0.930 0.775 0.671 0.818 0.895 0.400 0.880 0.395 0.720 E FIS -0.018 0.071 0.067 0.071 0.087 0.164 0.087 0.079 0.152 0.25 -0.004 0.083 MC294 NA 6 6 1 8 7 6 8 7 3 8 3 6.4

N =8 HO 0.875 0.750 --- 0.500 0.875 0.750 0.875 0.625 0.428 0.857 0.125 0.670 HE 0.833 0.620 --- 0.910 0.883 0.683 0.920 0.883 0.384 0.890 0.610 0.773 FIS -0.050 -0.216 1 0.450 0.009 -0.098 0.045 0.292 0.286 0.174 0.795 0.179 DC673 N 11 11 7 14 8 8 11 17 3 18 3 10.8 A N =42 HO 0.833 0.762 0.440 0.641 0.714 0.571 0.890 0.750 0.187 0.880 0.286 0.645

HE 0.736 0.747 0.520 0.920 0.791 0.600 0.880 0.902 0.180 0.920 0.304 0.702 FIS 0.244 0.246 0.133 *0.151 0.127 0.223 0.202 0.169 0.197 0.158 0.105 0.186 WFE NA 9 15 6 13 7 8 12 13 3 17 3 9.7 N =30 HO 0.600 0.600 0.433 0.423 0.800 0.600 0.800 0.821 0.520 0.862 0.083 0.612 HE 0.724 0.639 0.450 0.910 0.821 0.580 0.883 0.910 0.427 0.930 0.230 0.702 FIS 0.248 0.249 0.189 *0.146 0.283 0.117 0.213 0.119 0.177 0.113 0.324 0.194

AC601 AC818 WR269 GC852 AT340 MC388 DC673 WFE

AC601 0.054 0.019 0.030 0.017 0.079 0.001 0.029 --- AC818 0.014 0.000 0.058 0.032 0.045 0.057 0.086 --- WR269 0.007 0.007 0.000 0.000 0.000 0.000 0.018 --- GC852 0.016 0.006 0.000 0.009 0.048 0.003 0.014 --- AT340 0.020 0.018 0.000 0.000 0.010 0.005 0.019 --- MC388 0.048 0.036 0.027 0.024 0.021 0.037 0.020 ---

DC673 0.007 0.016 0.000 0.000 0.003 0.020 0.000 --- WFE 0.018 0.024 0.000 0.005 0.011 0.007 0.000 ---

114

NA HO HE FIS NA HO HE FIS NA HO HE FIS ES454_8 6 0.480 0.601 0.201 6 0.555 0.606 0.154 6 0.555 0.556 0.001 ES454_10 11 0.916 0.845 -0.027 15 0.821 0.848 0.031 13 0.815 0.877 0.072 ES454_13 9 0.400 0.420 0.049 7 0.444 0.422 0.073 8 0.423 0.408 0.086 ES454_25 18 0.880 0.938 0.063 15 0.926 0.902 0.017 16 0.923 0.924 0.043 ES454_28 7 0.791 0.791 0.058 7 0.785 0.777 -0.010 8 0.777 0.773 -0.006 ES454_31 19 0.800 0.918 0.129 14 0.852 0.901 0.095 11 0.730 0.872 0.201 ES454_32 4 0.160 0.405 *0.606 2 0.035 0.035 0.000 5 0.296 0.273 -0.084 ES454_34 6 0.791 0.720 -0.025 5 0.714 0.732 0.025 7 0.777 0.737 -0.054 ES454_44 5 0.680 0.627 -0.083 5 0.740 0.643 -0.070 6 0.654 0.670 0.093 ES454_45 10 0.200 0.452 *0.558 14 0.296 0.660 *0.582 11 0.192 0.466 0.633 ES454_50 11 0.904 0.859 0.135 7 0.571 0.685 0.167 9 0.880 0.820 0.032 ES454_57 9 0.840 0.841 0.001 13 0.852 0.880 0.075 10 0.815 0.852 0.044 ES454_60 9 0.720 0.624 -0.153 6 0.481 0.441 0.032 7 0.407 0.420 0.028 ES454_64 6 0.682 0.644 0.159 6 0.428 0.443 0.032 4 0.308 0.337 0.230 ES454_71 16 0.913 0.902 0.080 16 0.958 0.861 0.066 15 0.640 0.846 0.314 ES454_95 6 0.800 0.754 -0.061 7 0.714 0.784 1.00 5 0.704 0.708 0.006 Mean 9.4 0.685 0.709 0.081 9.2 0.636 0.664 0.149 8.7 0.618 0.658 0.100

Worm Hole --- 0.014 0.008

Baboon 0.007 --- 0.001

Regab 0.004 0.000 ---

Table A3-11: BOTTLENECK summary statistics for E. laminata sites. T.P.M: Two Phase Model (Wilcoxon test, 1000 replicates). N: sample size, ko: observed number of alleles, He: Hardy-Weinberg heterozygosity, Heq: expected heterozygosity measured, S.D: standard deviation of mutation drift equilibrium distribution of heterozygosity; Prob: probability of obtaining the measured He in a sample from an equilibrium population (Piry et al., 1999). Population: AC601 Under the T.P.M. locus N ko He Heq S.D. DH/sd Prob EL454_2 24 6 0.71 0.759 0.067 -0.736 0.185 EL454_5 24 8 0.793 0.84 0.042 -1.098 0.126 EL454_6 22 3 0.628 0.503 0.138 0.899 0.188 EL454_21 24 5 0.822 0.698 0.085 1.46 0.007 EL454_25 24 5 0.493 0.7 0.081 -2.559 0.031 EL454_52 24 11 0.884 0.904 0.022 -0.918 0.16 EL454_70 24 13 0.917 0.93 0.017 -0.789 0.167 EL454_54 24 11 0.891 0.905 0.023 -0.578 0.231 EL454_67 18 2 0.111 0.319 0.153 -1.358 0.237 EL454_71 24 2 0.159 0.288 0.155 -0.831 0.347 EL454_9 24 12 0.931 0.916 0.025 0.596 0.286 Population: AC818 Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob EL454_2 16 5 0.742 0.743 0.072 -0.023 0.392 EL454_5 16 5 0.792 0.741 0.068 0.739 0.297 EL454_6 16 3 0.425 0.53 0.124 -0.848 0.232 EL454_21 16 6 0.817 0.801 0.053 0.293 0.51 EL454_25 16 4 0.575 0.649 0.098 -0.762 0.202

116

EL454_52 16 9 0.883 0.905 0.023 -0.946 0.169 EL454_70 16 4 0.708 0.661 0.091 0.52 0.385 EL454_54 12 6 0.818 0.842 0.04 -0.588 0.314 EL454_67 16 2 0.325 0.316 0.151 0.061 0.556 EL454_71 14 2 0.495 0.324 0.15 1.133 0.259 EL454_9 16 8 0.883 0.879 0.03 0.146 0.581 Population: WR269 Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob EL454_2 14 5 0.593 0.761 0.064 -2.616 0.042 EL454_5 14 6 0.791 0.818 0.049 -0.549 0.278 EL454_6 12 4 0.455 0.696 0.082 -2.934 0.036 EL454_21 14 4 0.703 0.673 0.084 0.36 0.467 EL454_25 14 4 0.692 0.675 0.088 0.202 0.586 EL454_52 14 9 0.912 0.922 0.019 -0.506 0.335 EL454_70 14 10 0.945 0.943 0.013 0.152 0.779 EL454_54 10 6 0.889 0.872 0.031 0.544 0.596 EL454_67 14 2 0.264 0.328 0.146 -0.442 0.474 EL454_71 12 3 0.621 0.562 0.118 0.502 0.466 EL454_9 12 8 0.924 0.918 0.02 0.317 0.69 Population: GC852 Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob EL454_2 22 7 0.714 0.809 0.052 -1.827 0.059 EL454_5 22 8 0.814 0.846 0.041 -0.776 0.173 EL454_6 20 3 0.358 0.51 0.133 -1.148 0.185 EL454_21 22 7 0.84 0.812 0.048 0.577 0.34 EL454_25 20 5 0.6 0.718 0.078 -1.517 0.087 EL454_52 22 9 0.896 0.873 0.032 0.746 0.252 EL454_70 20 10 0.895 0.903 0.023 -0.38 0.316 EL454_54 20 10 0.889 0.9 0.029 -0.36 0.264 EL454_67 18 3 0.216 0.522 0.129 -2.366 0.055 EL454_71 20 3 0.279 0.506 0.133 -1.7 0.117 EL454_9 12 6 0.803 0.841 0.042 -0.915 0.157 Population: AT340 Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob EL454_2 22 6 0.684 0.768 0.062 -1.363 0.107 EL454_5 22 7 0.805 0.811 0.051 -0.107 0.389 EL454_6 22 5 0.468 0.709 0.082 -2.943 0.017 EL454_21 22 5 0.775 0.701 0.088 0.835 0.201 EL454_25 22 4 0.671 0.626 0.104 0.434 0.427 EL454_52 22 6 0.818 0.769 0.064 0.769 0.226 EL454_70 22 10 0.879 0.893 0.026 -0.57 0.257 EL454_54 20 8 0.895 0.855 0.038 1.063 0.101 EL454_67 22 4 0.398 0.628 0.099 -2.308 0.036

117

Population: AC601 Under the T.P.M. locus N ko He Heq S.D. DH/sd Prob EL454_2 24 6 0.71 0.759 0.067 -0.736 0.185 EL454_5 24 8 0.793 0.84 0.042 -1.098 0.126 EL454_6 22 3 0.628 0.503 0.138 0.899 0.188 EL454_21 24 5 0.822 0.698 0.085 1.46 0.007 EL454_25 24 5 0.493 0.7 0.081 -2.559 0.031 EL454_52 24 11 0.884 0.904 0.022 -0.918 0.16 EL454_70 24 13 0.917 0.93 0.017 -0.789 0.167 EL454_54 24 11 0.891 0.905 0.023 -0.578 0.231 EL454_67 18 2 0.111 0.319 0.153 -1.358 0.237 EL454_71 24 2 0.159 0.288 0.155 -0.831 0.347 EL454_9 24 12 0.931 0.916 0.025 0.596 0.286 Population: AC818 Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob EL454_2 16 5 0.742 0.743 0.072 -0.023 0.392 EL454_5 16 5 0.792 0.741 0.068 0.739 0.297 EL454_6 16 3 0.425 0.53 0.124 -0.848 0.232 EL454_21 16 6 0.817 0.801 0.053 0.293 0.51 EL454_25 16 4 0.575 0.649 0.098 -0.762 0.202 EL454_52 16 9 0.883 0.905 0.023 -0.946 0.169 EL454_70 16 4 0.708 0.661 0.091 0.52 0.385 EL454_54 12 6 0.818 0.842 0.04 -0.588 0.314 EL454_67 16 2 0.325 0.316 0.151 0.061 0.556 EL454_71 14 2 0.495 0.324 0.15 1.133 0.259 EL454_9 16 8 0.883 0.879 0.03 0.146 0.581 Population: WR269 Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob EL454_2 14 5 0.593 0.761 0.064 -2.616 0.042 EL454_5 14 6 0.791 0.818 0.049 -0.549 0.278 EL454_6 12 4 0.455 0.696 0.082 -2.934 0.036 EL454_21 14 4 0.703 0.673 0.084 0.36 0.467 EL454_25 14 4 0.692 0.675 0.088 0.202 0.586 EL454_52 14 9 0.912 0.922 0.019 -0.506 0.335 EL454_70 14 10 0.945 0.943 0.013 0.152 0.779 EL454_54 10 6 0.889 0.872 0.031 0.544 0.596 EL454_67 14 2 0.264 0.328 0.146 -0.442 0.474 EL454_71 12 3 0.621 0.562 0.118 0.502 0.466 EL454_9 12 8 0.924 0.918 0.02 0.317 0.69 Population: GC852 Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob EL454_2 22 7 0.714 0.809 0.052 -1.827 0.059 EL454_5 22 8 0.814 0.846 0.041 -0.776 0.173 EL454_6 20 3 0.358 0.51 0.133 -1.148 0.185 EL454_21 22 7 0.84 0.812 0.048 0.577 0.34 EL454_25 20 5 0.6 0.718 0.078 -1.517 0.087 EL454_52 22 9 0.896 0.873 0.032 0.746 0.252 EL454_70 20 10 0.895 0.903 0.023 -0.38 0.316 EL454_54 20 10 0.889 0.9 0.029 -0.36 0.264 118

Table A3-12: BOTTLENECK summary statistics for E. southwardae sites Regab, Wormhole and Baboon. T.P.M: Two Phase Model (Wilcoxon test, 1000 replicates). N: sample size, ko: observed number of alleles, He: Hardy-Weinberg heterozygosity, Heq: expected heterozygosity measured, S.D: standard deviation of mutation drift equilibrium distribution of heterozygosity; Prob: probability of obtaining the measured He in a sample from an equilibrium population (Piry et al., 1999). REGAB Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob ES454_8 54 6 0.556 0.7 0.085 -1.696 0.064 ES454_10 54 13 0.878 0.882 0.028 -0.171 0.346 ES454_13 52 7 0.403 0.753 0.067 -5.226 0.001 ES454_25 52 16 0.924 0.91 0.031 0.443 0.337 ES454_28 54 8 0.773 0.785 0.06 -0.202 0.359 ES454_31 52 11 0.873 0.855 0.037 0.469 0.382 ES454_32 54 5 0.273 0.638 0.103 -3.558 0.006 ES454_34 54 7 0.738 0.75 0.073 -0.166 0.338 ES454_44 52 6 0.67 0.7 0.086 -0.345 0.293 ES454_45 52 11 0.467 0.855 0.037 -10.564 0 ES454_50 50 9 0.82 0.816 0.05 0.069 0.445 ES454_57 54 10 0.853 0.834 0.044 0.431 0.386 ES454_60 54 7 0.419 0.745 0.073 -4.488 0 ES454_64 52 4 0.337 0.556 0.129 -1.699 0.082 ES454_71 50 15 0.847 0.908 0.019 -3.242 0.01 ES454_95 54 5 0.708 0.64 0.111 0.608 0.303 WORMHOLE Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob ES454_8 50 6 0.601 0.704 0.086 -1.209 0.108 ES454_10 48 11 0.846 0.86 0.033 -0.44 0.262 ES454_13 50 8 0.42 0.784 0.063 -5.815 0.001 ES454_25 50 18 0.939 0.93 0.015 0.614 0.331 ES454_28 48 7 0.791 0.756 0.067 0.521 0.364 ES454_31 50 19 0.918 0.936 0.013 -1.347 0.099 ES454_32 50 4 0.406 0.559 0.13 -1.177 0.14 ES454_34 48 6 0.72 0.71 0.082 0.116 0.464 ES454_44 50 5 0.628 0.65 0.102 -0.215 0.329 ES454_45 50 10 0.452 0.838 0.043 -8.976 0 ES454_50 42 11 0.859 0.867 0.032 -0.233 0.324 ES454_57 50 9 0.841 0.817 0.046 0.511 0.371 ES454_60 50 9 0.624 0.816 0.048 -3.988 0.005 ES454_64 44 6 0.645 0.716 0.08 -0.894 0.16 ES454_71 46 16 0.902 0.919 0.019 -0.837 0.148 ES454_95 50 6 0.754 0.707 0.082 0.58 0.323 BABOON Under the T.P.M. Locus n ko He Heq S.D. DH/sd Prob ES454_8 54 6 0.606 0.705 0.086 -1.146 0.128

119

REGAB Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob ES454_8 54 6 0.556 0.7 0.085 -1.696 0.064 ES454_10 54 13 0.878 0.882 0.028 -0.171 0.346 ES454_13 52 7 0.403 0.753 0.067 -5.226 0.001 ES454_25 52 16 0.924 0.91 0.031 0.443 0.337 ES454_28 54 8 0.773 0.785 0.06 -0.202 0.359 ES454_31 52 11 0.873 0.855 0.037 0.469 0.382 ES454_32 54 5 0.273 0.638 0.103 -3.558 0.006 ES454_34 54 7 0.738 0.75 0.073 -0.166 0.338 ES454_44 52 6 0.67 0.7 0.086 -0.345 0.293 ES454_45 52 11 0.467 0.855 0.037 -10.564 0 ES454_50 50 9 0.82 0.816 0.05 0.069 0.445 ES454_57 54 10 0.853 0.834 0.044 0.431 0.386 ES454_60 54 7 0.419 0.745 0.073 -4.488 0 ES454_64 52 4 0.337 0.556 0.129 -1.699 0.082 ES454_71 50 15 0.847 0.908 0.019 -3.242 0.01 ES454_95 54 5 0.708 0.64 0.111 0.608 0.303 WORMHOLE Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob ES454_8 50 6 0.601 0.704 0.086 -1.209 0.108 ES454_10 48 11 0.846 0.86 0.033 -0.44 0.262 ES454_13 50 8 0.42 0.784 0.063 -5.815 0.001 ES454_25 50 18 0.939 0.93 0.015 0.614 0.331 ES454_28 48 7 0.791 0.756 0.067 0.521 0.364 ES454_31 50 19 0.918 0.936 0.013 -1.347 0.099 ES454_32 50 4 0.406 0.559 0.13 -1.177 0.14 ES454_34 48 6 0.72 0.71 0.082 0.116 0.464 ES454_44 50 5 0.628 0.65 0.102 -0.215 0.329

ES454_45 50 10 0.452 0.838 0.043 -8.976 0 ES454_50 42 11 0.859 0.867 0.032 -0.233 0.324 ES454_57 50 9 0.841 0.817 0.046 0.511 0.371 ES454_60 50 9 0.624 0.816 0.048 -3.988 0.005 ES454_64 44 6 0.645 0.716 0.08 -0.894 0.16 ES454_71 46 16 0.902 0.919 0.019 -0.837 0.148 ES454_95 50 6 0.754 0.707 0.082 0.58 0.323 BABOON Under the T.P.M. Locus n ko He Heq S.D. DH/sd Prob ES454_8 54 6 0.606 0.705 0.086 -1.146 0.128 ES454_10 56 15 0.848 0.9 0.024 -2.205 0.045 ES454_13 54 9 0.422 0.812 0.049 -8.013 0 ES454_25 54 15 0.902 0.904 0.021 -0.11 0.379

ES454_28 56 7 0.778 0.75 0.066 0.418 0.408 ES454_31 54 14 0.901 0.895 0.024 0.274 0.487

ES454_32 56 2 0.036 0.241 0.166 -1.236 0.164 ES454_34 56 5 0.732 0.636 0.105 0.919 0.157 ES454_44 54 5 0.643 0.642 0.11 0.012 0.414 ES454_45 54 14 0.66 0.893 0.028 -8.48 0.002 ES454_50 56 7 0.686 0.745 0.073 -0.818 0.181 ES454_57 54 13 0.88 0.883 0.028 -0.099 0.376 ES454_60 54 6 0.441 0.701 0.089 -2.922 0.02 ES454_64 56 6 0.443 0.702 0.087 -2.997 0.015 ES454_71 48 16 0.861 0.916 0.019 -2.985 0.01 120

Summary Statistics for null alleles Escarpia taxa (means across all loci) (9 loci) NA NF SE

E. spicata 8 0.123 0.067

E. laminata 13 0.086 0.022

E. southwardae 12 0.058 0.028 Summary Statistics for null alleles Escarpia laminata (means across all loci) (11 loci) NA NF SE

AC601 7.3 0.153 0.082

AC818 5.1 0.179 0.103

WR269 5.7 0.203 0.111

GC852 6.7 0.132 0.076

AT340 6.6 0.093 0.059

MC294 6.4 0.173 0.101

DC673 10.8 0.151 0.041

WFE 9.7 0.117 0.046 Summary Statistics for null alleles Escarpia southwardae (means across all loci) (16 loci) NA NF SE

Regab 8.7 0.085 0.016

Wormhole 9.4 0.076 0.017

Baboon 9.2 0.100 0.016

121

APPENDIX C

CHAPTER 4

S03-271_LAMELLIBRACHIA_SP1 S10-101_LAMELLIBRACHIA_SP1 3-5_LAMELLIBRACHIA_SP1 S03-55_LAMELLIBRACHIA_SP1 S03-54_LAMELLIBRACHIA_SP1 S03-97_LAMELLIBRACHIA_LUYMESI S03-98_LAMELLIBRACHIA_LUYMESI S03-99_LAMELLIBRACHIA_LUYMESI S03-273_LAMELLIBRACHIA_LUYMESI S03-100_LAMELLIBRACHIA_LUYMESI S10-5_LAMELLIBRACHIA_SP1 S09-58_LAMELLIBRACHIA_LUYMESI 65 S09-210_LAMELLIBRACHIA_LUYMESI S03-63_LAMELLIBRACHIA_SP1 66 S03-166_LAMELLIBRACHIA_SP1 S03-189_LAMELLIBRACHIA_SP1 S03-108_LAMELLIBRACHIA_LUYMESI S03-201_LAMELLIBRACHIA_SP1 1-10_LAMELLIBRACHIA_LUYMESI S09-198_LAMELLIBRACHIA_LUYMESI S03-103_LAMELLIBRACHIA_LUYMESI 55 S09-59_LAMELLIBRACHIA_LUYMESI S03-349_LAMELLIBRACHIA_LUYMESI S09-212_LAMELLIBRACHIA_LUYMESI S03-96_LAMELLIBRACHIA_LUYMESI 48 S03-350_LAMELLIBRACHIA_LUYMESI S03-53_LAMELLIBRACHIA_SP1 S03-165_LAMELLIBRACHIA_SP1 S03-173_LAMELLIBRACHIA_SP1 S10-13_LAMELLIBRACHIA_SP1 9-3_LAMELLIBRACHIA_SP1 ESCARPIA_LAMINATA_KC357394 ESCARPIA_LAMINATA_KC357395 ESCARPIA_LAMINATA_KC357396 100 ESCARPIA_LAMINATA_KC357397 ESCARPIA_LAMINATA_KC870962 ESCARPIA_LAMINATA_KC870963

0.1 Figure A4-1: mtCYTB maximum likelihood (ML) tree for 31 Lamellibrachia with a fragment size of 380bp. All Lamellibrachia sequences are newly isolated and Escarpia laminata is the outgroup taxon. GenBank accession numbers follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions per nucleotide site. ML and NJ produced congruent trees for all genes, and only the ML trees are reported here.

122

S03-168_LAMELLIBRACHIA_SP1 S03-165_LAMELLIBRACHIA_SP1 S03-174_LAMELLIBRACHIA_SP1 S03-189_LAMELLIBRACHIA_SP1 S10-5_LAMELLIBRACHIA_SP1 S10-13_LAMELLIBRACHIA_SP1 S03-271_LAMELLIBRACHIA_SP1 S10-101_LAMELLIBRACHIA_SP1 S03-96_LAMELLIBRACHIA_LUYMESI S03-101_LAMELLIBRACHIA_LUYMESI S03-351_LAMELLIBRACHIA_LUYMESI S03-98_LAMELLIBRACHIA_LUYMESI S03-273_LAMELLIBRACHIA_LUYMESI S03-103_LAMELLIBRACHIA_LUYMESI S03-100_LAMELLIBRACHIA_LUYMESI S03-99_LAMELLIBRACHIA_LUYMESI S03-275_LAMELLIBRACHIA_LUYMESI S09-212_LAMELLIBRACHIA_LUYMESI S09-211_LAMELLIBRACHIA_LUYMESI S09-210_LAMELLIBRACHIA_LUYMESI S03-108_LAMELLIBRACHIA_LUYMESI 65 S03-97_LAMELLIBRACHIA_LUYMESI S10-92_LAMELLIBRACHIA_SP1 S03-55_LAMELLIBRACHIA_SP1 S03-53_LAMELLIBRACHIA_SP1 S03-201_LAMELLIBRACHIA_SP1 60 S03-192_LAMELLIBRACHIA_SP1 S03-178_LAMELLIBRACHIA_SP1 S03-173_LAMELLIBRACHIA_SP1 S09-198_LAMELLIBRACHIA_LUYMESI ESCARPIA_LAMINATA_KC870987 ESCARPIA_LAMINATA_KC870986 ESCARPIA_LAMINATA_KC870985 100 ESCARPIA_LAMINATA_KC870984 ESCARPIA_LAMINATA_KC870983 ESCARPIA_LAMINATA_KC870982

0.1 Figure A4-2: HbB2i maximum likelihood (ML) tree for 30 Lamellibrachia with a fragment size of 668bp. All Lamellibrachia sequences are newly isolated and Escarpia laminata is the outgroup taxon. GenBank accession numbers follow the species name. Bootstrap support (1000 replicates, Tamura-Nei model) is located either above or below the node. Scale is measured as number of substitutions per nucleotide site.

123

S03-98_LAMELLIBRACHIA_LUYMESI S03-271_LAMELLIBRACHIA_SP1 S03-99_LAMELLIBRACHIA_LUYMESI S03-100_LAMELLIBRACHIA_LUYMESI 25 S03-273_LAMELLIBRACHIA_LUYMESI S10-101_LAMELLIBRACHIA_SP1 S10-5_LAMELLIBRACHIA_SP1 40 63 S03-189_LAMELLIBRACHIA_SP1 S03-55_LAMELLIBRACHIA_SP1 36 S09-210_LAMELLIBRACHIA_LUYMESI S03-97_LAMELLIBRACHIA_LUYMESI S03-108_LAMELLIBRACHIA_LUYMESI

42 S03-53_LAMELLIBRACHIA_SP1

19 S03-165_LAMELLIBRACHIA_SP1 S09-198_LAMELLIBRACHIA_LUYMESI S10-13_LAMELLIBRACHIA_SP1 S03-103_LAMELLIBRACHIA_LUYMESI S09-212_LAMELLIBRACHIA_LUYMESI S03-96_LAMELLIBRACHIA_LUYMESI S03-201_LAMELLIBRACHIA_SP1 S03-173_LAMELLIBRACHIA_SP1 ESCARPIA_LAMINATA_KC870969/KC870986 ESCARPIA_LAMINATA_KC357392/KC357417 ESCARPIA_LAMINATA_KC357394/KC357418 99 ESCARPIA_LAMINATA_KC357395/KC357419 ESCARPIA_LAMINATA_KC357390/KC357416

0.1

124

Gene: mtCYTB L. luymesi L. sp. 1 E. laminata L. luymesi 0.002 66.885 L. sp. 1 0.006 66.849 E. laminata 8.038 8.050 Gene: HbB2i L. luymesi 0.001 8.707 L. sp. 1 0.001 8.708 E. laminata 1.789 1.783 Genes: mtCYTB HbB2i L. luymesi 0.001 13.702 L. sp. 1 0.003 13.703 E. laminata 1.105 1.102

Table A4-2: Primers used in this study

Gene Primers Forward (5’ – 3’) Reverse (5’ – 3’) CYTB CYTB_Lamelli_F GGG GWC AGA TAA GTT AGC GAA AAG GYA AGT ATC (Lamellibrachia) CYTB_Lamelli_R TYT GAG GAG ART CAG G

hbB2i hbB2i_Lamelli_F TCG CCC CCA GGC TGT CTT CTT GAA TTC GTT GCT GTT (Lamellibrachia) hbB2i_Lamelli_R C GAC G

125

Table A4-3: Summary statistics for Tajima’s D test of neutrality organized by gene and species. S is number of sites with substitutions; Pi is the mean number of pairwise differences. Distance method used is pairwise difference, no Gamma correction, indels not taken into account. Gene: mtCYTB L. luymesi L. sp. 1 Mean s.d. Sample Size 32 30 31 1.141 S 5 9 7 2.830 Pi 1.812 2.206 2.010 0.279 Tajima’s D 1.234 -0.088 0.572 0.935 p-value 0.892 0.524 0.708 0.260 Gene: HbB2i L. luymesi L. sp. 1 Mean s.d. Sample Size 30 30 30 0.000 S 3 5 4 1.414 Pi 0.588 1.312 0.950 0.512 Tajima’s D -0.524 0.109 -0.207 0.448 p-value 0.338 0.589 0.463 0.177 Genes: CYTB & L. luymesi L. sp. 1 Mean s.d. HbB2iSample Size 22 20 21 1.313 S 8 9 8.5 0.707 Pi 2.550 3.326 2.940 0.550 Tajima’s D 0.531 1.070 0.800 0.381 p-value 0.742 0.883 0.812 0.100

Morphospecies L1-2E#2 L454_6 L454_11 L454_13 L454_19 L454_20 L454_25 L454_32 Mean ± SE NA 6 7 13 8 3 6 8 8 7.37 HO 0.53 0.75 0.80 0.65 0.16 0.75 0.83 0.66 0.64 L. luymesi HE 0.57 0.66 0.89 0.69 0.15 0.63 0.80 0.69 0.63 N = 20 FIS 0.08 0.00 0.11 0.06 0.00 0.00 0.00 0.04 0.00 AR 5.84 6.68 12.48 7.60 2.95 5.70 8.00 8.00 7.16 ± 0.96 PAR 3.84 6.68 12.48 7.60 0.95 3.80 6.12 2.66 5.52 ± 1.26 NA 4 16 19 10 7 3 3 7 8.62 HO 0.82 0.71 0.79 0.23 0.76 0.66 0.16 0.43 0.57 L. sp 1. HE 0.62 0.89 0.94 0.84 0.73 0.52 0.16 0.45 0.64 N = 24 FIS 0.00 0.21 0.16* 0.73* 0.00 0.00 0.00 0.03 0.11 AR 3.78 13.72 17.07 9.54 6.57 2.75 2.88 6.30 7.83 ± 1.86 PAR 1.78 13.72 17.07 9.54 4.57 0.85 1.00 0.96 6.19 ± 2.28

127

Table A4-5: Summary statistics for eight microsatellite loci amplified in two regional populations of Lamellibrachia sp1. Number of alleles observed (NA), observed (HO) and expected (HE) heterozygosity, Wright’s Inbreeding Coefficient (FIS), rarified allelic richness (AR), private allele richness (PAR), standard error (SE). Allelic and private allelic richness were rarefied over 16 samples and means are not significantly different (p > 0.05). Significant deviation from HWE after FDR correction = 0.01 is denoted by asterisk (*). Lamellibrachia sp.1 L1-2E#2 L454_6 L454_11 L454_13 L454_19 L454_20 L454_25 L454_32 Mean ± SE NA 3 13 18 9 7 5 3 6 8 HO 0.78 0.66 0.78 0.19 0.70 0.72 0.17 0.39 0.55 H 0.61 0.87 0.96 0.83 0.73 0.54 0.16 0.35 0.63 West GoM E FIS 0.00 0.24* 0.19 0.78* 0.04 0.00 0.00 0.00 0.13 AR 3.00 8.50 11.73 6.75 5.21 3.33 2.14 3.61 3.28 ± 1.16 PAR 0.00 3.58 6.41 2.51 2.34 0.44 0.70 1.33 2.09 ± 0.74 NA 4 9 8 6 3 4 2 4 5 HO 1.00 0.83 0.83 0.33 1.00 1.00 0.17 0.60 0.72 H 0.71 0.91 0.89 0.89 0.75 0.79 0.17 0.73 0.73 East GoM E FIS 0.00 0.09 0.07 0.65* 0.00 0.00 0.00 0.20 0.01 AR 4.00 9.00 8.00 6.00 3.00 4.00 2.00 4.00 2.45 ± 0.87 PAR 1.00 4.08 2.68 1.76 0.14 1.11 0.55 1.71 1.26 ± 0.45

Lamellibrachia sp. 2 L12E#2 L454_6 L454_19 L454_20 L454_32 Mean ± SE

NA 2 6 4 3 2 3.40

HO 0.50 0.75 1.00 0.50 0.25 0.60 West GoM HE 0.50 0.93 0.82 0.68 0.25 0.64 (WR269 and GC852) FIS 0.00 0.22 0.00 0.29 0.00 0.06 AR 2.00 3.57 3.00 2.41 1.50 2.50 ± 0.36

PAR 1.00 1.33 0.12 0.49 0.45 0.68 ± 0.21

NA 5 15 5 4 5 6.80

HO 0.81 0.65 0.64 0.56 0.43 0.62 East GoM HE 0.67 0.90 0.76 0.57 0.38 0.66 (MC294 and MC344) FIS 0.00 0.28* 0.15 0.00 0.00 0.05 AR 2.53 3.46 2.80 2.10 1.81 2.54 ± 0.29

PAR 1.06 1.91 0.40 0.06 0.65 0.82 ± 0.32

NA 2 4 3 3 2 2.80

HO 1.00 0.25 1.00 1.00 0.25 0.70 Far East GoM HE 0.60 0.82 0.83 0.73 0.25 0.65 (DC673 and FIS 0.00 0.73* 0.00 0.00 0.00 0.00 WFE) AR 2.00 3.00 3.00 2.60 1.50 2.42 ± 0.29

PAR 0.49 0.68 0.36 0.78 0.38 0.54 ± 0.08

129

Lamellibrachia spp. L. luymesi L. sp. 1 L. luymesi --- 0.482 L. sp. 1 0.213 ---

Lamellibrachia sp. 1 West GoM East GoM West GoM --- 0.006 East GoM 0.003 ---

Lamellibrachia sp. 2

West GoM East GoM Far East GoM West GoM --- -0.059 0.032 East GoM -0.033 --- 0.022 Far East GoM 0.008 0.014 ---

Table A4-8: BOTTLENECK summary statistics for L. luymesi and L. sp. 1 across all sites, respectively. I.A.M: Infinite Alleles Model; T.P.M: Two Phase Model. N: sample size, ko: observed number of alleles, He: Hardy-Weinberg heterozygosity, Heq: expected heterozygosity measured, S.D: standard deviation of mutation drift equilibrium distribution of heterozygosity; Prob: probability of obtaining the measured He in a sample from an equilibrium population (Piry et al., 1999). Bottom row shows Wilcoxon test results for 1000 replicates run under the TPM.

L. luymesi Under the I.A.M. Under the T.P.M.

locus n ko He Heq S.D. DH/sd Prob Heq S.D. DH/sd Prob L12E#2 38 6 0.575 0.664 0.114 -0.786 0.192 0.731 0.072 -2.188 0.034 L454_6 40 7 0.667 0.709 0.097 -0.435 0.269 0.765 0.066 -1.501 0.081 L454_11 40 13 0.894 0.872 0.039 0.559 0.333 0.898 0.022 -0.199 0.344 L454_13 40 8 0.688 0.747 0.085 -0.682 0.216 0.801 0.054 -2.085 0.045 L454_19 38 3 0.152 0.391 0.168 -1.426 0.144 0.455 0.146 -2.078 0.054 L454_20 40 6 0.633 0.653 0.115 -0.174 0.347 0.719 0.08 -1.071 0.134 L454_25 36 8 0.806 0.763 0.081 0.541 0.355 0.81 0.051 -0.064 0.388 L454_32 36 8 0.694 0.755 0.082 -0.745 0.204 0.809 0.048 -2.381 0.039 L. sp.1 Under the I.A.M. Under the T.P.M. locus n ko He Heq S.D. DH/sd Prob Heq S.D. DH/sd Prob

L12E#2 46 4 0.622 0.496 0.157 0.804 0.246 0.564 0.128 0.456 0.404 L454_6 48 16 0.894 0.894 0.032 -0.004 0.408 0.917 0.018 -1.3 0.106

L454_11 48 19 0.938 0.921 0.023 0.727 0.256 0.937 0.013 0.091 0.602 L454_13 44 10 0.844 0.799 0.068 0.649 0.268 0.845 0.038 -0.037 0.397 L454_19 42 7 0.727 0.705 0.098 0.227 0.497 0.761 0.071 -0.484 0.223 L454_20 48 3 0.526 0.38 0.172 0.846 0.263 0.449 0.151 0.511 0.398 L454_25 48 3 0.16 0.381 0.173 -1.281 0.174 0.444 0.152 -1.871 0.066 L454_32 46 7 0.448 0.686 0.105 -2.254 0.039 0.759 0.065 -4.811 0.003

Lamellibrachia luymesi Locus NA NF SE L1-2E#2 6 0.148 0.077 L454_6 7 0.048 0.041 L454_11 13 0.067 0.046 L454_13 8 0.065 0.051 L454_19 3 0.178 0.101 L454_20 6 0.046 0.039 L454_25 8 0.073 0.062 L454_32 8 0.113 0.081 Mean 7 0.099 0.068 Lamellibrachia sp. 1 Locus NA NF SE L1-2E#2 4 0.051 0.043 L454_6 16 0.096 0.052 L454_11 19 0.083 0.045 L454_13 10 0.394 0.075 L454_19 7 0.080 0.070 L454_20 3 0.049 0.040 L454_25 3 0.092 0.070 L454_32 7 0.134 0.072 Mean 8.8 0.193 0.064 Lamellibrachia sp. 2 Locus NA NF SE L1-2E#2 6 0.041 0.039 L454_6 16 0.133 0.054 L454_19 5 0.101 0.069 L454_20 4 0.083 0.053 L454_32 5 0.050 0.043 Mean 8.2 0.167 0.066

132

Seep name Dive Latitude Longitude Depth (m) Number of individuals E. laminata (42) DC673 J2-539 28.516° N 87.518° W 2604 L. sp. 1 (2) L. sp. 2 (2) E. laminata (6) 4662 28.675° N 88.481° W 1800 L. sp. 1 (3) MC294 L. sp. 2 (14) J2-541 28.675° N 88.481° W 1800 E. laminata (2) E. laminata (30) WFE 3916 26.040° N 84.915° W 3304 L. sp. 2 (2) 4180 27.644° N 88.365° W 2204 E. laminata (6) AT340 4179 27.386° N 88.219° W 2218 E. laminata (5) E. laminata (5) 4187 27.095° N 91.265° W 1437 L. sp. 1 (4) L. sp. 2 (1) E. laminata (4) GC852 4186 27.057° N 91.099° W 1420 L. sp. 1 (2) E. laminata (2) J2-273 27.057° N 91.099° W 1633 L. sp. 2 (1) 4177 27.095° N 91.265° W 1437 L. sp. 2 (1) E. laminata (1) J2-275 26.406° N 91.396° W 1964 L. sp. 1 (1) WR269 E. laminata (6) 4191 26.677° N 91.665° W 1975 L. sp. 1 (2) L. sp. 2 (1) 4192 26.110° N 94.344° W 2785 E. laminata (6) AC818 J2-282 26.185° N 94.574° W 2744 E. laminata (2) 4193 26.236° N 94.308° W 2339 E. laminata (7) AC601 E. laminata (5) 4196 26.392° N 94.514° W 2235 L. sp. 1 (2) GC600 4174 27.374° N 90.573° W 1193 L. sp. 1 (4) GB697 J2-274 27.312° N 92.638° W 1281 L. sp. 1 (3) L. sp. 1 (1) MC344 4663 28.633° N 88.169° W 1800 L. sp. 2 (9)

VITA

Dominique Alexandria Cowart

Education Ph.D. Biology – August 2013 The Pennsylvania State University, University Park, PA M.S. Marine Biosciences – May 2008 University of Delaware – Newark, DE B.S. Marine Sciences – May 2005 Texas A&M University – College Station, TX

Research experience August 2008 – August 2013 The Pennsylvania State University, University Park, PA Dissertation research: Species distributions and population structure in cold seep vestimentiferan tubeworms of the genera Escarpia and Lamellibrachia (Polychaeta, Siboglinidae) Supervisors: Charles R. Fisher and Dr. Stephen W. Schaeffer August 2006 – June 2008 University of Delaware, College of Earth, Ocean and Environment, Lewes, DE Thesis research: Salinity sensitivity of early embryos of the Antarctic sea urchin, Sterechinus neumayeri Supervisor: Adam G. Marsh August 2005 – November 2005 United States Antarctic Program Grant# 0238281 Research Assistant for Antarctic Field Team Primary Investigator Adam G. Marsh

Publications Cowart, DA, Huang, C, Arnaud-Haond, S, Carney, SL, Fisher, CR, Schaeffer, SW (2013) “Restriction to large scale gene flow versus regional panmixia among cold seep Escarpia spp. (Polychaeta, Siboglinidae).” in press in Molecular Ecology Cowart, DA, Huang, C, Schaeffer, SW (2012) “Identification and amplification of microsatellite loci in deep- sea tubeworms of the genus Escarpia (Polychaeta, Siboglinidae).” Conservation Genetics Resources. 5 (2): 479 - 482 Thiel, V, Hügler, M, Blümel, M, Baumann, HI, Gärtner, A, Schmaljohann, R, Strauss, H, Garbe-Schönberg, D, Petersen, S, Cowart, DA, Fisher, CR, Imhoff, JF (2012) “Widespread occurrence of two carbon fixation pathways in tubeworm endosymbionts: lessons from hydrothermal vent associated tubeworms from the Mediterranean Sea.” Frontiers in Microbiology. 3: 423 Cowart, DA, Guida, SM, Shah, IH, Marsh, AG (2011). “Effects of Ag nanoparticles on survival and oxygen consumption of zebra fish embryos, Danio rerio.” Journal of Environmental Science and Health, Part A: Toxic/Hazardous Substance & Environmental Engineering. 46 (10): 1122 – 1128 Miglietta, MP, Hourdez, S, Cowart, DA, Schaeffer, SW, Fisher, CR (2010) “Species boundaries of Gulf of Mexico vestimentiferans (Polychaeta, Siboglinidae) inferred from mitochondrial genes”. Deep Sea Research II: Topical Studies in Oceanography. 57 (21): 1916 – 1925 Cowart, DA, Ulrich, PN, Miller, DC, Marsh, AG (2009) “Salinity sensitivity of early embryos of the Antarctic sea urchin, Sterechinus neumayeri”. Polar Biology. 32: 435–441

Grants and Fellowships March 2012 – 2014 NSF Doctoral Dissertation Improvement Award (DDIG) Award # 1209688 May 2009 – August 2013 Alfred P. Sloan Fellowship August 2008 – May 2010 Braddock Graduate Fellowship, The Pennsylvania State University August 2006 – May 2008 University of Delaware Graduate Scholars Fellowship August 2003 – May 2005 William Paul Ricker Memorial Scholarship

Teaching experience Teaching Assistant, Department of Biology, The Pennsylvania State University, 2009 – 2012 Biology 110 – Basic Concepts and Biodiversity – Fall 2009, 2010, 2012 Biology 406 – Symbiosis – Spring 2012