University of Florida Thesis Or Dissertation Formatting Template

THE EVOLUTION AND SYSTEMATICS OF THE CALLISIA SECTION CUTHBERTIA COMPLEX (COMMELINACEAE)

IWAN EDUARD MOLGO

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2018

To my family, especially to Muriel for her love and support throughout this journey; to Isabella and Callisia, my bundle of joy who kept me going; to my parents who encouraged my education and believed in my dreams

ACKNOWLEDGMENTS

I thank my advisors Pamela S. Soltis and Douglas E. Soltis who gave me the opportunity to continue my graduate career in their lab. Both have contributed invaluable support, critical guidance, and encouragement throughout my Ph.D. program. They introduced me to my Dissertation project, which turned out to be a great learning experience in molecular and morphological phylogenetics, niche modeling, and cytogeography.

I thank my committee members Walter S. Judd and Matthew E. Smith for providing support and advice during project. I am grateful to W. Mark Whitten who has supported me tremendously and taught me different lab techniques in DNA amplification. I thank current and former members of the Soltis and Cellinese lab

(Prabha Amarasinghe, Andre Chanderbali, Michael Chester, Kurt Neubig, Ryan Folk,

Charlotte Germain-Aubrey, Matthew Gitzendanner, Lucas Majure, Evgeny Mavrodiev, Miao

Sun, Clayton Visger), for their help with methodologies and data analyses, the FLAS herbarium (Paul Corogin, Lorena Endara, Kent Perkins, Norris Williams), and the staff of the U.F. Biology Department for their assistance, friendship and encouragement.

I thank the following herbaria for access to the information on the voucher specimens of Callisia: GA, USCH, NCU, DUKE, US, AAH, FLAS, FSU, VSC, and USF. I thank the members of the Florida Native Plant Society and photographers from

Flickr.com for providing accurate locality data. I also thank all staff of the State Parks,

State Forests, National Parks, The Nature Conservancy protected areas, military reservations, and U.S. Fish and Wildlife Service protected areas in Florida, Georgia,

South Carolina, North Carolina, and Virginia for their assistance with locating and collecting plant material for this study. I thank Robert Faden for sending me plant

material from the Smithsonian collection. I am grateful to Jeff Hubbard, the Biology

Department greenhouse manager, who has always assisted in keeping the plants healthy. A special thank you goes to the family of Dr. Norman H. Giles for providing unpublished data of Callisia section Cuthbertia. I am grateful to all the undergraduate volunteers: Kylie Beauchamp, Sofia Chang, Savannah Elliot, Tess Huttenlocker, Nicolas

Kushch, Sydney Newsom, Shannon Parma, Viviana Martinez, Valeria Segui, and Emilie

Sorrel, whose contributions were invaluable to this study. I thank Andrew Walker who took me collecting in the Sandhill Game lands of North Carolina.

I also want to thank Edzard van Santen for his assistance in choosing the appropriate statistical methods for my research analyses. I would like to thank the

Organization of American States for granting me a scholarship to study at the University of Florida. In addition, the cost of lab work and fieldtrips was in part supported by graduate awards: Michael L. May Research Grant, Sigma Xi GAIR, American Society of

Plant Taxonomists, Marilyn Little Altrusa Scholarship, Florida Native Plant Society grant, and UF Biology Teaching Award. I thank the board of the Anton de Kom University of

Suriname who have supported me during my Ph.D. program.

I am grateful to John Anderson, Samuel Crothers IV, Tremaine Gregory, Roger

Lopez, and many other friends for their support and encouragement during this journey.

I thank my parents Eduard and Jacqueline Molgo for their support and encouragement throughout my life and my family in Suriname and the Netherlands for their support and encouragement. Finally, I thank my lovely wife Muriel Djaspan-Molgo and my beautiful daughters Isabella Molgo and Callisia Molgo who have been patient and very supportive throughout my studies.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 9

LIST OF FIGURES ...... 11

LIST OF ABBREVIATIONS ...... 13

ABSTRACT ...... 15

CHAPTER

1 GENERAL INTRODUCTION ...... 18

2 CYTOGEOGRAPHY OF CALLISIA SECTION CUTHBERTIA (COMMELINACEAE) ...... 25

Introduction ...... 25 Materials and Methods...... 28 Georeferencing ...... 28 Collecting of Specimens ...... 28 Chromosome Counts ...... 29 Flow Cytometry ...... 30 Results ...... 31 Georeferencing and Collecting ...... 31 Chromosome Counts ...... 31 Flow Cytometry ...... 31 Distribution Map ...... 32 Discussion ...... 33 Georeferencing ...... 33 Flow Cytometry and Genome Size ...... 34 Distribution ...... 38

3 THE EVOLUTIONARY RELATIONSHIPS AMONG THE CYTOTYPES OF CALLISIA SECTION CUTHBERTIA (COMMELINACEAE): MORPHOLOGICAL AND MOLECULAR APPROACHES ...... 55

Introduction ...... 55 Materials and Methods...... 57 Taxon Sampling ...... 57 Morphological Analysis ...... 57 Molecular Phylogenetic Analysis ...... 59 Samples ...... 59 DNA extraction ...... 59

PCR amplification and sequencing ...... 59 Phylogenetic analysis ...... 63 General Greenhouse Observations ...... 64 Results ...... 65 Morphological Analysis ...... 65 Vegetative data set ...... 65 Vegetative and reproductive data set...... 66 Molecular Phylogenetic Analysis ...... 68 Phylogenetic analysis (diploid taxa) ...... 69 Phylogenetic analysis (diploid and polyploid taxa) ...... 72 General Greenhouse Observations ...... 75 Discussion ...... 76 Morphological Analysis ...... 76 Vegetative data set ...... 76 Vegetative and reproductive data set ...... 78 Molecular Phylogenetic Analysis ...... 80 Phylogenetic analysis (diploid taxa) ...... 80 Phylogenetic analysis (diploid and polyploid taxa) ...... 82 General Greenhouse Observations ...... 87

4 ECOLOGICAL NICHE MODELS AND PAST, PRESENT, AND FUTURE GEOGRAPHIC DISTRIBUTIONS OF CALLISIA SECTION CUTHBERTIA (COMMELINACEAE) ...... 108

Introduction ...... 108 Materials and Methods...... 110 Data points ...... 110 Ecological Niche Modeling Layers ...... 111 Selection of Climatic Variables ...... 112 Ecological Niche Modeling ...... 113 Principal Component Analysis ...... 114 Results ...... 115 Predicted Suitability and Principal Component Analysis under Current Climatic Conditions ...... 115 Predicted Suitability under Past Climatic Conditions ...... 117 Predicted Suitability and Range Changes under Future Climatic Conditions . 120 Niche Overlap, Equivalence, and Breadth of ENMs of Callisia Section Cuthbertia in the Past, Present, and Future ...... 124 Discussion ...... 126 Predicted Suitability and Principal Component Analysis under Current Climatic Conditions ...... 126 Predicted Suitability under Current, Last Glacial Maximum, and Mid- Holocene Conditions ...... 129 Predicted Suitability under Future Conditions ...... 133

5 GENERAL CONCLUSIONS ...... 155

APPENDIX

A GEOREFERENCED DATA POINTS ...... 162

B VOUCHERS USED WITH GENBANK NUMBERS ...... 179

C ENM PERFORMANCE ...... 195

LIST OF REFERENCES ...... 196

BIOGRAPHICAL SKETCH ...... 213

LIST OF TABLES

Table page

2-1 Populations used in this study...... 42

2-2 Genome sizes (2C) of Callisia section Cuthbertia and their cytotypes and previously reported 2C-values...... 52

3-1 Outgroup taxa included in this study...... 89

3-2 DNA regions and their primers used in this study...... 90

3-3 Squared Mahalanobis distance and adjusted Bonferroni p-value among cytotypes in Callisia section Cuthbertia based on vegetative traits...... 91

3-4 Mean value (mm) of the vegetative traits used in the CANDISC analysis ...... 91

3-5 Canonical structure...... 91

3-6 Squared Mahalanobis distance and adjusted Bonferroni p-value among cytotypes in Callisia section Cuthbertia based on vegetative and floral traits. .... 93

3-7 Mean value (mm) of the combined vegetative and reproductive traits used in the CANDISC analysis...... 93

3-8 Data matrix and parsimony tree statistics for each sequenced region...... 96

3-9 Terminal taxa denoting duplicate sequences in diploid ITS and all cytotypes ITS tree...... 97

3-10 Terminal taxa denoting duplicate sequences in diploid chloroplast and all cytotypes chloroplast tree...... 99

3-11 Terminal taxa denoting duplicate sequences in diploid total evidence and all cytotypes total evidence tree...... 101

4-1 The estimated range suitability (km2) of Callisia section Cuthbertia in the past and current...... 142

4-2 The estimated range overlap of Callisia section Cuthbertia in the past, present and future...... 142

4-3 The estimated range suitability (km2) of Callisia section Cuthbertia in the current predictions and future projections under two climate change scenario’s (RCP2.6 and RCP8.5)...... 143

4-4 Observed niche overlap values (Schoener’s D) and results of tests of niche Identity test (ID) for Callisia section Cuthbertia...... 153

4-5 Results of niche breadth calculation for the past, current and future predictions...... 154

A-1 Georeferenced data points...... 162

C-1 The area under the curve value (AUC) and standard deviation of all ENMs in the past, present and future models...... 195

LIST OF FIGURES

Figure page

1-1 Bombyliid fly, Poecilognathus sulphurea visits tetraploid Callisia graminea...... 24

2-1 Distribution map of Callisia section Cuthbertia...... 49

2-2 Habit of Callisia section Cuthbertia...... 50

2-3 Mitotic metaphase chromosomes spreads from root tips...... 51

2-4 Histograms of fluorescence intensity (FL2-A) of propidium iodide-stained nuclei...... 53

2-5 Distribution of cytotypic variation in Callisia section Cuthbertia...... 54

3-1 Canonical discriminant ordination of Callisia section Cuthbertia based on vegetative traits...... 92

3-2 Canonical discriminant ordination of Callisia section Cuthbertia based on vegetative and reproductive traits...... 94

3-3 Canonical discriminant ordination of Callisia section Cuthbertia depicting all cytotypes...... 95

3-4 Nuclear diploid phylogeny of Callisia section Cuthbertia...... 98

3-5 Chloroplast diploid phylogeny of Callisia section Cuthbertia...... 100

3-6 Combined nuclear and chloroplast diploid phylogeny of Callisia section Cuthbertia...... 102

3-7 Nuclear phylogeny of Callisia section Cuthbertia...... 103

3-8 Chloroplast phylogeny of Callisia section Cuthbertia...... 104

3-9 Combined nuclear and chloroplast phylogeny of Callisia section Cuthbertia.. .. 105

3-10 Vivipary in Callisia section Cuthbertia...... 106

3-11 Stem thickness comparison among putative taxa in Callisia section Cuthbertia...... 107

4-1 Predicted niche suitability and statistical environmental differentiation of Callisia graminea 2x, C. graminea 4x...... 138

4-2 Predicted niche suitability and statistical environmental differentiation of Callisia ornata 2x, C. ornata 4x...... 139

4-3 Predicted niche suitability of hexaploid C. graminea and C. ornata, and statistical ecological niche divergent estimation of Callisia graminea 4x, C. ornata 2x, C. ornata 4x, C. graminea 6x and C. ornata 6x...... 140

4-4 Predicted niche suitability of C. rosea and statistical ecological niche divergent estimation of C. graminea 2x, 4x and C. rosea...... 141

4-5 A comparison of niche suitability for Callisia graminea 2x and C. graminea 4x projected in the past and present...... 144

4-6 A comparison of Niche suitability for Callisia ornata 2x and C. ornata 4x projected in the past and present...... 145

4-7 A comparison of niche suitability for Callisia rosea and the hexaploids C. graminea & C. ornata projected in the past and present...... 146

4-8 A comparison of niche suitability for Callisia graminea 2x in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5...... 147

4-9 A comparison of niche suitability for Callisia graminea 4x in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5...... 148

4-10 A comparison of niche suitability for Callisia rosea in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5...... 149

4-11 A comparison of niche suitability for Callisia ornata 2x in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5...... 150

4-12 A comparison of niche suitability for Callisia ornata 4x in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5...... 151

4-13 A comparison of niche suitability for Callisia graminea 6x and C. ornata 6x in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5...... 152

LIST OF ABBREVIATIONS bp Base pair

BS Bootstrap

Can Canonical variable

CANDISC Canonical discriminant analysis

CCSM4 Community Climate System Model, version 4

CI Consistency index

ENM Ecological niche models

FL Filament length

H Height

ILD Incongruence length difference

ITS Internal transcribed spacer (rDNA)

LGM Last Glacial Maximum

LL Leaf length

LW Leaf width

MCE Minority cytotype exclusion

MH mid-Holocene

ML Maximum likelihood

MP Maximum parsimony

PCA Principal Component Analysis

PL Petal length

PW Petal width pg Picogram

RCP Representative Concentration Pathways

rDNA Ribosomal DNA

RI Retention index

SDM Species distribution models

SL Sepal length

ST Stem thickness

SW Sepal width

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

THE EVOLUTION AND SYSTEMATICS OF THE CALLISIA SECTION CUTHBERTIA COMPLEX (COMMELINACEAE)

Iwan Eduard Molgo

May 2018

Chair: Pamela S. Soltis Major: Botany

Callisia section Cuthbertia (Commelinaceae) is one of six sections in Callisia and comprises three species of herbaceous plants endemic to the southeastern United

States. In morphological comparison with their congeners, this clade is distinguished by linear leaves and a grass-like, caespitose habit. Callisia graminea, C. ornata, and C. rosea are morphologically distinct and have a base chromosome number of x = 6. Prior to this study, polyploidy was known in C. graminea (2x, 4x, 6x). Callisia section

Cuthbertia is a polyploid complex that was taxonomically not clear, and the geographical distribution of the cytotype diversity was unknown prior to this study. The goal of this study was to sample plants of Callisia section Cuthbertia throughout its range, to investigate the genome size, ploidy, and cytogeography of these taxa, and clarify evolutionary relationships and polyploid ancestry in this complex. Morphometric data were analyzed in a canonical discriminant analysis (CANDISC) to identify the most significant variable(s) separating the taxa in Callisia section Cuthbertia. A molecular phylogenetic analysis including the three diploid species, their polyploid cytotypes, and

20 relatives in Callisia, Gibasis, and Tripogandra was conducted using nuclear and chloroplast sequence data. Using locality data from 426 specimens, niche modeling

techniques were employed to investigate the predicted suitability, niche overlap, niche breadth, and range of Callisia section Cuthbertia in the past, present, and future.

The results of these analyses confirm the presence of three ploidal levels (2x, 4x,

6x) of C. graminea, but also reveal these same three ploidal levels in C. ornata; C. rosea is exclusively diploid. The CANDISC analysis found significant differences among all cytotypes and revealed that stem thickness is the most significant variable to separate taxa within this group, followed by leaf width and sepal width. The phylogenetic results confirm the monophyly of Callisia section Cuthbertia in analyses with and without polyploid cytotypes included. Trees based on nuclear and chloroplast data for the diploids only are incongruent; in the nuclear tree, C. ornata 2x is sister to C. graminea 2x and C. rosea, while in the chloroplast tree, C. rosea is sister to C. graminea 2x and C. ornata 2x. A combined phylogeny of nuclear and chloroplast data supports a relationship of C. ornata 2x as sister to C. graminea 2x and C. rosea. These results suggest that either C. graminea 2x is an ancient homoploid hybrid or that the incongruence results from introgression. Phylogenetic analysis supports diploid C. graminea as the maternal parent of tetraploid C. graminea; tetraploid C. graminea is the maternal parent of hexaploid C. graminea and C. ornata, and diploid C. ornata is the maternal parent of tetraploid C. ornata. The environmental niche models revealed that the current distributions are similar to the predicted suitability maps. At the Last Glacial

Maximum period, the projected suitability for all entities was smaller than the current distribution. In the mid-Holocene, a range expansion was discovered for C. graminea

2x. In the future, there are signs of range expansion; however, if no actions are

undertaken against the high emissions of greenhouse gases, in 2070, a loss of suitable habitat will occur for most taxa of Callisia section Cuthbertia.

CHAPTER 1 GENERAL INTRODUCTION

Commelinaceae Mirb, also known as the dayflower or spiderwort family (Faden and Hunt 1991, Judd et al. 2016), is a well-supported clade in Commelinales (Burns et al. 2011, Evans et al. 2003) that comprises 41 genera and about 650 species (Faden

1998). The family is mainly tropical and warm temperate in distribution and is divided into two subfamilies, Cartonematoideae (Pichon) Faden ex G. C. Tucker and

Commelinoideae Faden & D. R. Hunt. Subfamily Commelinoideae has been further divided into two tribes, Tradescantieae Meisn. and Commelineae Dumort. (Faden and

Hunt 1991). Tribe Tradescantieae has been divided into seven subtribes, one of which is Tradescantiinae Rohw. Subtribe Tradescantiinae is monophyletic (Hertweck and

Pires 2014) and comprises the genera Tradescantia L., Gibasis Raf., Tripogandra Raf., and Callisia Loefling (Burns et al. 2011, Faden 1998, Hertweck and Pires 2014).

Hertweck and Pires (2014) reported that none of the genera in Tradescantiinae are monophyletic and that the subtribe consists of two clades (1) Tradescantia and Gibasis and (2) Callisia and Tripogandra.

Callisia is not monophyletic and includes ca. 20 – 23 species (Bergamo 2003,

Faden 1998) that are divided among six sections; Hadrodemas (Moore) Hunt,

Cuthbertia (Small) Hunt, Lauia Hunt, Brachyphylla Hunt, Leptocallisia Bentham &

Hooker, and Callisia (Hunt 1986, Tucker 1989). All sections are monophyletic, but due to the placement of Tripogandra within Callisia, the latter is paraphyletic (Hertweck and

Pires 2014). In order to resolve this paraphyly two options are possible. 1) Either all sections could be raised to generic rank or 2) the limits of Callisia could be expanded so

that it represents a clade. The taxa of these sections are distributed throughout the

Americas and the Caribbean (Tucker 1989).

Callisia section Cuthbertia consists of three morphologically distinct species that are endemic to the southeastern United States. Callisia graminea (Small) Tucker,

C. ornata (Small) Tucker, and C. rosea (Ventenat,) Hunt can be distinguished from

Tripogandra, Gibasis, and the rest of Callisia by grass-like, linear leaves and a caespitose habit; reduced flower bracts separate them from Tradescantia (Bergamo

2003).

In 1903, Small described Cuthbertia graminea (grassleaf roseling) and based on the type specimen locality and morphology, it was a tetraploid. Tradescantia rosea previously described by Ventenat in 1800 was in 1903 also transferred to Cuthbertia rosea (piedmont roseling). Small described in 1933 a diploid type specimen Cuthbertia ornata (Florida scrub roseling), but all C. graminea and C. ornata were placed as varieties under Tradescantia rosea in 1935 by Anderson and Woodson. In 1942,

Woodson placed Tradescantia rosea in Tripogandra as Tripogandra rosea and did not change the nomenclature of T. rosea var. graminea and T. rosea var. ornata (Woodson

1942). Woodson argued that he was not sure whether C. graminea and C. ornata were species or varieties and suggested extensive fieldwork to elucidate relationships. Even though Woodson made these changes, others, including Giles (1942, 1943), Tomlinson

(1966), and (Lakela 1972), reported evidence that concurred with Small’s delimitation of

Cuthbertia. In 1956, Rodweder placed T. rosea in Phyodina; however, Hunt (1986) transferred Tripogandra rosea to Callisia and reported that the taxonomic status of T. rosea var. graminea and T. rosea var. ornata was unclear. In 1989, Tucker transferred

both varieties into Callisia section Cuthbertia. Bergamo (2003) and Burns (2011) reported paraphyly in Callisia and questioned the current taxonomic classification.

Bergamo (2003) suggested placing taxa within Callisia section Cuthbertia back in its own genus Cuthbertia. This treatment is recommended if all sections are raised to the generic rank. This would make Callisia and Tripogandra sensu stricto monophyletic. I would recommend this treatment; however, in this study the focus is not to resolve the paraphyly of Callisia, but rather to investigate the evolution and systematics of Callisia section Cuthbertia.

Callisia graminea, C. ornata, and C. rosea range from southeastern Virginia to southern Florida (habitat: Sandhills, Flatwoods, and Scrubby-Flatwoods) and exhibit flower color variation from white to pink (Lakela 1972). All Commelinaceae flowers have deliquescent petals, are open for only a few hours during the day, and lack nectar

(Faden 1998). Callisia ornata hosts a pollen-eating bombyliid fly, Poecilognathus punctipenni (Walker) (Diptera), which also visits other taxa in Commelinaceae

(Commelina erecta L. and Tradescantia roseolens Small) (Deyrup 1988). During my extensive fieldwork, I encountered P. sulphurea (Loew) (see Figure 1-1), identified by

Gary Steck, on C. graminea. Poecilognathus punctipennis is similar but larger, darker, and with more extensive wing markings. Both species of bee flies are widespread in

Florida, but P. punctipennis seems to occur in a narrower timeframe, spring to early summer, while P. sulphurea continues into the fall months (personal communication,

Gary Steck).

Taxa in Callisia section Cuthbertia share a base chromosome number of x = 6

(Giles 1943). Giles (1942, 1943) reported three ploidal levels (2x, 4x, and 6x) for Callisia

graminea, which ranges from Virginia through central Florida. Diploid and tetraploid entities do not grow in sympatry, but hexaploids have been recorded in South Carolina and Florida in sympatry with tetraploid C. graminea (Giles 1943). A single triploid individual was encountered in North Carolina by Giles (1942). Owens (1981), who investigated self-incompatibility in Commelinaceae, reported that all genera examined in

Tradescantieae including Callisia showed a degree of self-incompatibility and that polyploidy does not affect the incompatibility reaction of those species, which agrees with self-incompatibility in diploid and tetraploid C. graminea (Kelly 1991). Nonetheless, artificial interploidal crosses among diploid and tetraploid C. graminea were successful and produced fertile seeds, which makes them not reproductively isolated in sympatry

(Kelly 1991). Lakela (1972) reported vivipary in C. graminea, but evidence of this reproductive mechanism occurring in natural habitats was lacking. Vivipary is not considered a common method of reproduction in Callisia. The only other species known to sometimes exhibit vivipary is C. warscewicziana (Hunt 1994) within Callisia section

Hadrodemas. Callisia graminea is not a popular commercial landscaping plant; however, Callisia ornata, a diploid (Giles unpublished), is endemic to central to southern

Florida and is recommended as a natural landscaping plant. Callisia rosea is a diploid

(Anderson and Sax 1936), with a distribution from North Carolina to Georgia, and these are horticulturally important plants, commercially known as “Morning Grace”.

The monophyly of section Cuthbertia was supported in molecular phylogenetic analysis by Bergamo (2003) and Hertweck and Pires (2014) and revealed that Callisia ornata is sister to C. graminea and C. rosea. However, relationships among the polyploid cytotypes and their putative progenitors are not clear.

The primary goals of this study were:

 to determine the current distribution of Callisia section Cuthbertia in the southern United States;

 to carry out flow cytometry analysis for all members of the Callisia section Cuthbertia complex, calibrated using chromosome counts of C. graminea (2x, 4x, 6x), and link these data to the distribution of the polyploid complex;

 to carry out a canonical discriminant analysis to identify morphological traits that differentiate taxa of this complex;

 to clarify the phylogenetic relationships among diploid species C. graminea, C. ornata, and C. rosea;

 to reconstruct the phylogeny of all members of the Callisia section Cuthbertia clade with an emphasis on determining the progenitors of the polyploid entities in the group;

 to carry out niche modeling analysis, to predict the distribution of Callisia section Cuthbertia in the past, present and future.

These goals are addressed in the following four chapters. In Chapter 2, I will present the current distribution of Callisia section Cuthbertia, provide flow cytometry data, genome size estimations, and the cytogeography of Callisia section Cuthbertia.

In Chapter 3, morphometrical data will be used to employ a Canonical discriminant analysis. This analysis will identify vegetative and reproductive traits that will distinguish among all diploid and polyploid cytotypes. The canonical variables will be plotted in ordination graphs to visualize the differences among taxa. I will reconstruct the phylogeny of (1) the diploid species of Callisia section Cuthbertia, and (2) both diploid and polyploid taxa based on nuclear and chloroplast data sets, and use the phylogeny to explain the origin of the polyploids in section Cuthbertia. I will provide greenhouse observations based on phenology, pest susceptibility, and asexual reproduction.

In Chapter 4, I will utilize previously gathered locality data from Chapter 2 and employ niche modeling analysis to predict the distribution of the taxa in Callisia section

Cuthbertia in the past, present, and future. With the software program MaxEnt, and the current presence data, I predict the suitable habitats for each taxon of Callisia section

Cuthbertia in the southeastern United States. The predicted suitability of the current distribution was used to project suitable areas onto abiotic environmental layers of the

Last Glacial Maximum (~ 22 ka) and mid-Holocene (~ 6 ka). The models were also projected onto the environmental layers of the future (2050 and 2070) under two hypotheses of greenhouse gas emissions and concentration pathways: RCP2.6 and

RCP8.5. In all environmental scenarios, the polyploid entities were analyzed to investigate their distributions in ecological niche space relative to those of their progenitors. In Chapter 5, I will highlight the major outcomes of this study and provide general conclusions.

Figure 1-1. Bombyliid fly, Poecilognathus sulphurea visits tetraploid Callisia graminea. A) P. sulphurea on flower B) Bombyliid fly feeding on pollen on petal C) Bombyliid fly feeding and placing front legs on stigma. August 7, 2015. Lake County, FL. Photo courtesy of author.

CHAPTER 2 CYTOGEOGRAPHY OF CALLISIA SECTION CUTHBERTIA (COMMELINACEAE)1

Introduction

Polyploidy (whole-genome duplication) is a speciation mechanism that is a major evolutionary force; in fact, all angiosperms have undergone at least one ancient polyploidy event (Amborella Genome Project 2013, Jiao et al. 2011), and polyploidy has been a key driver of angiosperm diversity (De Bodt et al. 2005, Soltis et al. 2009, Soltis and Soltis 2009, Soltis and Soltis 2016, Tank et al. 2015).

Polyploids are classified in two major categories: allopolyploids and autopolyploids. Allopolyploids are by far the more studied form and arise via hybridization between species, whereas autopolyploids originate from the multiplication of genomes within a single species. An autopolyploid is frequently considered as a cytotype within a species along with its diploid progenitor, as in Galax urceolata (Poiret)

Brummitt, (Baldwin 1941, Stebbins 1950), Chamerion angustifolium (L.) Holub (Mosquin

1967), Heuchera grossulariifolia Rydberg (Wolf et al. 1990), and Vaccinium corymbosum L. (Camp 1945, Krebs and Hancock 1989). However, autotetraploids are occasionally recognized as species distinct from their diploid parent, such as Zea perennis (Hitchcock) Reeves & Mangelsdorf (Iltis et al. 1979, Tiffin and Gaut 2001) and

Tolmiea menziesii Torrey & Gray (Judd et al. 2007). Lumping diploid progenitors with their multiple derivative cytotypes into a single species may mask evolutionary lineages and grossly underestimate biodiversity (Soltis et al. 2007).

 Reprinted with permissions from Molgo IE, Soltis DE, Soltis PS (2017) Cytogeography of Callisia section Cuthbertia (Commelinaceae). Comparative Cytogenetics 11: 553-577. doi:10.3897/compcytogen.v11i4.11984

To gain a better assessment of biodiversity and to guide conservation efforts for species of interest, data on both evolutionary and life-history characteristics are needed.

Callisia section Cuthbertia (Commelinaceae) from the southeastern U.S.A. comprises a polyploid complex, with species of conservation concern, but the extent of polyploidy and the geographic distribution of cytotype diversity are unknown.

Callisia Loefling is one of 39 genera in subfamily Commelinoideae (Burns et al.

2011) and is placed in tribe Tradescantieae subtribe Tradescantiinae. Callisia comprises approximately 23 species in six sections (Hadrodemas (Moore) Hunt,

Cuthbertia (Small) Hunt, Lauia Hunt, Brachyphylla Hunt, Leptocallisia Bentham &

Hooker, and Callisia) (Hunt 1986, Tucker 1989).

Of these sections, Cuthbertia is endemic to the U.S.A., and Brachyphylla,

Leptocallisia, and Callisia also have members that occur in the U.S.A. (Tucker 1989).

The remaining two sections (Lauia and Hadrodemas) occur in Central America, South

America, and the Caribbean. In recent phylogenetic analyses, Callisia is not monophyletic (Bergamo 2003, Burns et al. 2011), although, significantly, section

Cuthbertia is monophyletic in all analyses (Bergamo 2003, Burns et al. 2011, Hertweck and Pires 2014).

Callisia section Cuthbertia consists of three morphologically distinct species (C. graminea, C. ornata, and C. rosea) that are endemic to the southeastern U.S.A. and have a base chromosome number of x = 6 (Giles 1942, 1943). Callisia graminea (Small,

1903) Tucker, 1989, the grassleaf roseling, occurs from the southern border of Virginia through central Florida. Giles (1942, 1943) reported three ploidal levels (2x, 4x, and

6x) for this species and encountered a single triploid individual in Hoke County, NC.

Based on cytological criteria, the tetraploid was interpreted as an autopolyploid derivative of diploid C. graminea (Giles 1942, 1943). The nature of polyploidy in hexaploid C. graminea is not clear. Within C. graminea, two forms have been described:

C. graminea forma graminea has pink flowers with anthocyanin pigments, and C. graminea forma leucantha (Lakela) Tucker has white flowers and was described from two diploid cuttings (Lakela 1972). Callisia ornata (Small) Tucker (Florida scrub roseling), a diploid (Giles unpublished), is endemic to central to southern Florida.

Callisia rosea (Ventenat) Hunt (Piedmont roseling) is a diploid (Anderson and Sax

1936), with a distribution from North Carolina to Georgia.

Although earlier studies (e.g., Giles 1942, 1943) provided the general pattern of species distributions and cytotypic diversity, the extent of cytotypic variation within and among species has not been examined in detail. Additional sampling of both populations and species is required to understand the extent and distribution of cytological variation in this clade. In this study, numerous new field collections were made, and known populations of Callisia section Cuthbertia were revisited; with the use of both traditional chromosome counts and flow cytometry, the ploidy of samples spanning the entire range of Callisia section Cuthbertia was investigated. Distribution maps of cytotypes and species were generated based on the cytological data obtained here, enabling future studies of phylogeny and polyploid origins in Callisia section

Cuthbertia.

Materials and Methods

Georeferencing

To obtain locality data for Callisia graminea, C. ornata, and C. rosea, voucher specimens were examined from the following herbaria: GA, USCH, NCU, DUKE, US,

AAH, FLAS, FSU, VSC, and USF (codes follow Thiers 2016) The locality of each specimen was georeferenced by manually incorporating the label data into the web applications ACME mapper 2.1 (Poskanzer 2001) and/or GEOLocate (Rios and Bart

2010). Additional localities were obtained from the Master’s Thesis of A. Kelly (1991) and personal communications with members of the Florida Native Plant Society and photographers from Flickr.com. In all, 436 specimens were georeferenced from herbarium specimens and observation records. (See Appendix A-1: for georeferenced data points.) The data points were used to produce a distribution map using ArcGIS

10.4 (ESRI 2016) and to locate known populations and contact zones of all three species and their cytotypes.

Collecting of Specimens

The georeferenced data were used to relocate populations within the southeastern U.S.A.; additional localities were discovered by exploring similar habitats in protected areas and on private land. Collections on private land were made with permission of the land owners. Based on the georeferenced data, permits were obtained to collect in state parks, state forests, national parks, and protected areas of

The Nature Conservancy and the U.S. Fish and Wildlife Service in Florida, Georgia,

South Carolina, North Carolina, and Virginia (Table 2-1).

Mature individuals were sampled in the summers of 2012, 2013, 2014, and 2015.

Only known localities with collection years between 1970 and 2012 were visited, unless the locality was in a protected area. This approach was used to increase the chances of finding intact populations but meant that we were unable to resample all of Giles’s

(1942, 1943) locations. Voucher specimens were deposited at the University of Florida

Herbarium (FLAS); collection numbers are provided in Table 2-1.

Population localities were surveyed for individuals with different growth habit and habitat; we then collected across that diversity. Contact zones between species, based on the georeferenced localities, were more intensively surveyed by searching for distinct morphological variation (habit, leaf, and flower) to increase the probability of encountering mixed cytotypes. Two to six live plants were collected per locality. Plants were removed with 15 cm of soil circumference to increase the survival rate and placed in plastic bags. At the Department of Biology, University of Florida greenhouse, plants were then potted in a soil mixture of 1:1 sand and potting soil (Pro-Mix) and were kept under natural light. During the period from December–March, the individuals of putative diploid C. graminea and C. rosea were given a four-month dormancy treatment at 4°C to mimic their natural habitat.

Chromosome Counts

Two individuals per cytotype of C. graminea were used as a control for flow cytometry analysis by counting chromosome numbers using established methods (see below). Previous studies of members of Commelinaceae found that cell division in root tips occurs at high frequency during late morning to early afternoon (Faden and Suda

1980). After a series of hourly collections, 2:00 pm was determined to be the optimal time for collecting root tips of C. graminea, C. ornata, and C. rosea. Root tips were placed in 2 mM 8-hydroxyquinoline following Soltis (1980) for 24 hours at 4°C and then

fixed in a 3:1 absolute ethanol-glacial acetic acid solution for 24 hours. Root tips were then placed in 70% ethanol and stored until needed at 4°C. Digestion of the root tips and spreading of the chromosomes on slides were performed following the methods of

Kato et al. (2011). Chromosomes were stained with DAPI and visualized using a Zeiss

Axio Imager M2 microscope (Carl Zeiss Microscopy LLC, Thornwood, NY, U.S.A.).

Flow Cytometry

Preparation of all samples for flow cytometry followed Roberts et al. (2009), in which each sample consisted of approximately 1 cm2 of fresh leaf tissue of Callisia; 0.5 cm2 dried leaf tissue of Vicia faba L. (26.9 pg) was used as an internal standard

(Dolezel et al. 2007). Samples were finely chopped with a sharp single-edged razor blade in a petri dish for 2 min in 1 ml of cold lysis buffer (0.1 M citric acid, 0.5% v/v

Triton X−100, 1% w/v PVP−40 in distilled water) (Hanson et al. 2005, Mavrodiev et al.

2015). After 20–30 sec of incubation on a cold brick that served as a cold chopping surface, each sample was further treated and measured based on the methods of

Mavrodiev et al. (2015) on an Accuri C6 flow cytometer (BD Biosciences, San Jose, CA,

U.S.A.). In all, the ploidy of 300 samples was assessed in batches of 28 samples.

For the estimation of genome size, three plants of the same accession were analyzed using the Flow Cytometry Kaluza Analysis Software 1.3 (Beckman Coulter Life

Sciences 2016). The relative DNA content was calculated using the ratio of the mean fluorescent peak of the sample to the mean fluorescent peak of the internal standard, multiplied by the genome size of the standard, Vicia faba (Dolezel et al. 2007).

Results

Georeferencing and Collecting

All GPS points obtained here were incorporated into a map with ARCGIS 10.4

(ESRI 2016) (Figure 2-1). The results show that Callisia graminea ranges from North

Carolina to central Florida with an isolated population in southern Virginia. Callisia rosea occurs predominantly in South Carolina and Georgia, and C. ornata is found in central to southern Florida. Specimens were collected at 133 localities, of which 61 were known from the 436 georeferenced localities and 72 were newly discovered populations. A list of these localities is provided in Table 2-1, indicating the geographic origin, ploidal level with corresponding number of plants, total number of analyzed individuals, and voucher information for each sample. Illustrations of the habits of diploid C. graminea, C. ornata, and C. rosea are provided in Figure 2-2.

Chromosome Counts

Chromosome numbers were obtained for three individuals per cytotype in C. graminea, confirming the presence of 2n = 2x = 12 (diploids; Figure 2-3a), 2n = 4x = 24

(tetraploids; Figure 2-3b), and 2n = 6x = 36 (hexaploids; Figure 2-3c). The diploid and tetraploid counts were obtained for plants from known locations for which previous counts were available (Giles 1942, Kelly 1991). The hexaploids were discovered while counting spreads of putatively tetraploid C. graminea from Lake County, FL (Table 2-1).

These 2x, 4x, and 6x individuals of C. graminea were then used as references in subsequent analyses using flow cytometry.

Flow Cytometry

Ploidy was estimated via flow cytometry for 300 plants of C. graminea

(representing 96 populations), C. ornata (from 23 populations), and C. rosea (from 7

populations). The results and the number of individuals analyzed per population are given in Table 2-1. Three distinct groups of fluorescence intensities were obtained from these analyses that were congruent with chromosome counts of diploid, tetraploid, and hexaploid C. graminea. Histograms for the cytotypes of C. graminea are shown in

Figure 2-4. Results for 26 individuals (17%) of tetraploid C. graminea had a lower fluorescence intensity (suggesting a smaller genome size) than the remaining 83% of tetraploid C. graminea. The ploidy of the former plants was verified by chromosome counts, and all were tetraploid.

The relative genome size of individuals of C. rosea was similar to that of diploid

C. graminea (2n = 2x = 12) (see below), confirming that our samples of C. rosea are diploid, in agreement with the literature (Giles 1942). Most individuals of C. ornata (2n =

2x = 12) were also inferred to be diploid, as expected based on previous counts (Giles unpublished), but our analysis also revealed previously unknown tetraploid (2n = 4x=

24) and hexaploid populations (2n = 4x = 36) of C. ornata. The latter were found in

Seminole State Forest, FL, where they occur in sympatry with tetraploid individuals of C. graminea. All polyploid levels were verified with chromosome counts; chromosome spreads are depicted in Figure 2-3.

Genome size (2C-value) of cytotypes in Callisia section Cuthbertia was estimated; data are presented in Table 2-2 along with previously calculated genome sizes by Hertweck (2011) and Jones and Kenton (1984).

Distribution Map

Based on the flow cytometry data, the distribution of cytotypic variation among the 126 populations sampled [C. graminea (96 populations), C. ornata (23 populations), and C. rosea (7 populations)] was mapped (Figure 2-5). This map shows that diploid C.

graminea is restricted to two disjunct areas: one in Franklin County, VA, and the second stretching along the Fall Line from North Carolina to South Carolina. Tetraploid C. graminea has a broader distribution that runs along the coastal plain from North

Carolina to central Florida. Hexaploid C. graminea occurs in Lake and Hernando

Counties, FL, and one individual was found in Richland County, SC. In South Carolina, one hexaploid C. graminea individual was found growing sympatrically with multiple tetraploid C. graminea plants. Based on extensive collecting, our observations suggest that the tetraploid C. graminea samples from North Carolina are the largest of this species, with clumps that exhibit a diameter of over 25 cm compared to plants in South

Carolina, Georgia, and Florida, with a maximum diameter of 15 cm. Diploid C. ornata occurs in eastern Florida (from Putnam through Martin Counties), and tetraploid C. ornata occurs in western Florida (Polk, Hillsborough, Highlands, and Lake Counties).

Hexaploid C. ornata occurs in Lake and Volusia Counties in central Florida. Diploid C. rosea occurs in the piedmont of Georgia and South Carolina with some scattered populations in the coastal plain.

Discussion

Georeferencing

Callisia section Cuthbertia consists of three species native to the southeastern

U.S.A., with three ploidal levels within C. graminea and C. ornata and diploids in C. rosea. The map of the geographic distribution (Figure 2-1) of all georeferenced voucher specimens depicts all specimens of C. graminea, C. ornata, and C. rosea without ploidal levels, collected from 1894 until present. Callisia graminea is the most widely distributed of all species in the genus, ranging from Virginia to Florida. Callisia ornata is restricted to Florida; although one specimen was recorded from Charleston County, GA, C. ornata

was not found in Georgia in this study. Callisia rosea occurs mainly in Georgia and the

Carolinas, but two herbarium specimens were found from Duval County and Highlands

County, FL. The localities of these two herbarium specimens of C. rosea were vague, and C. rosea was not observed in Florida in this study.

Flow Cytometry and Genome Size

Flow cytometry analysis of ploidal levels in 300 individuals from 126 populations together with 60 additional chromosome counts confirmed the presence of diploid, tetraploid, and hexaploid cytotypes of C. graminea and C. ornata. Significantly, tetraploid and hexaploid C. ornata were previously unknown. Our analysis also confirmed that C. rosea is diploid. However, Anderson and Sax (1936) and (Ichikawa and Sparrow 1967) reported only tetraploids in C. rosea. This might be a misidentification of broad-leaved tetraploid C. graminea as C. rosea, as suggested by

Giles (1942), who only detected diploids in C. rosea. Overall, three distinct fluorescent intensity peaks were seen in the histograms among the C. graminea and C. ornata cytotypes, with peaks for the tetraploids that are approximately twice the size of those of the diploids and for the hexaploids that are approximately three times those of the diploids. This general pattern of genome size increase in polyploids is to be expected relative to their diploid progenitors (Leitch and Bennett 2004). It is interesting to note that 26 individuals (17%) of tetraploid C. graminea had a lower fluorescence intensity than the remaining 83%, suggesting a smaller genome size. The individuals with the smaller peak than that typical of other tetraploids were measured twice with the flow cytometer, and the results were consistent. The chromosome numbers of these samples were verified by chromosome counts, and all were tetraploid (2n = 4x = 24).

Reductions in genome size in polyploids are common (Leitch and Bennett 2004), and in

this study two hypotheses are possible: genome downsizing or the occurrence of multiple origins from parents having different genome sizes. Because this variation in genome size occurs among individuals within populations and because the individuals are not clustered in a single geographic area, we suggest that this variation in DNA content might be a result of genome downsizing, but this hypothesis requires further testing. Genome size can be used, with other methods, to hypothesize putative progenitors of polyploids (e.g. Eilam et al. 2010). In diploid C. graminea the estimated

2C-value is 41.75 pg; the value for tetraploid C. graminea is 78.55 pg. According to

Giles (1942), multivalent chromosome pairing was observed in tetraploid C. graminea, suggesting autopolyploidy. If tetraploid C. graminea is of autopolyploid origin, the expected DNA content would be 83.47 pg, but the observed DNA content of tetraploid

C. graminea is 4.95 pg lower than the expected 2C-value. Newly formed polyploids usually possess a DNA content equal to the sum of the 2C-values of their progenitors

(Bennett et al. 2000, Eilam et al. 2010). Over time, however, genome downsizing in polyploids relative to their progenitors is expected (Leitch and Bennett 2004), which seems to be the case in tetraploid relative to diploid C. graminea.

Due to the rarity of hexaploid C. graminea in South Carolina, we only calculated the 2C-value of hexaploids that occur in Florida. Hexaploid C. graminea may be of allo- or autopolyploid origin. If from allopolyploid origin, the expected 2C-value would be

127.06 pg, with diploid C. ornata (48.51 pg) and tetraploid C. graminea (78.55 pg) as the progenitors. The observed genome size of hexaploid C. graminea is 122.86 pg, which is lower than the expected value, again consistent with genome downsizing. In the case of an autopolyploid origin with tetraploid C. graminea (78.55 pg) as parent, we

would expect a genome size of 117.83 pg, which is approximately 5 pg less than the observed 2C-value. Genome size data do not conclusively elucidate the origins of hexaploid C. graminea; both allo- and autopolyploidy are possible, and its origin requires further testing. However, Giles (1942) noted multivalent formation, generally indicative of autpolyploidy, in hexaploid C. graminea.

Tetraploid C. ornata has a 2C-value of 87.99 pg. It could be of autopolyploid origin with diploid C. ornata (48.51 pg) as the parent given that no other extant taxa are sympatric with it. However, the expected DNA content (97.02 pg) is at least 9 pg higher than observed; in contrast, when considering tetraploid C. ornata as a possible allopolyploid with tetraploid C. graminea (78.55 pg) and diploid C. ornata (48.51 pg) as parents (based on an unreduced gamete of the latter), the results (87.79 pg) are similar to the observed DNA content. These results therefore support allopolyploidy over autopolyploidy, yet further analyses are needed to clarify the origin of this cytotype.

Hexaploid C. ornata could be of allo- or autopolyploid origin. If allopolyploid, the expected genome size would be 127.06 pg with diploid C. ornata (48.51 pg) and tetraploid C. graminea (78.55 pg) as parents. The observed DNA content is 129.73 pg, which is slightly higher than the expected 2C-value. Alternatively, it could be an allohexaploid between tetraploid C. ornata (87.99 pg) and diploid C. graminea (41.75 pg), with an expected genome size of 129.74 pg, essentially identical to the observed value. In the case of autopolyploidy, we calculated an expected 2C-value of 145.53 if the value is 3 times that of diploid C. ornata (48.51 pg), 136.5 pg if tetraploid (87.99 pg) and diploid (48.51 pg) C. ornata are considered the parents, and 131.99 pg if a reduced and unreduced gamete of tetraploid C. ornata yield the hexaploid. The latter case is

closest to the observed value, suggesting either that hexaploid C. ornata is of allopolyploid origin, or if an autopolyploid, it arose via the third possible mechanism outlined above; these hypotheses require further investigation.

Based on the Plant DNA C-values Database, http://data.kew.org/cvalues/

(Bennett and Leitch 2012), recorded species of Commelinaceae have a minimum 2C- value of 5.16 pg for Commelina erecta L.1753 and a maximum of 86.7 pg for

Tradescantia virginiana L. 1753. The DNA content of hexaploid C. graminea and hexaploid C. ornata are currently the highest within Commelinaceae and Commelinales

(Leitch et al. 2010) with 122.86 pg and 129.73 pg, respectively. Jones and Kenton

(1984) reported that the 2C-value of C. rosea is 77.3 pg, with a chromosome count of

2n = 24, consistent with tetraploidy reported by Anderson and Sax (1936) and Ichikawa and Sparrow (1967); however, as noted above, Giles (1942) only detected diploids (2n

= 12) for C. rosea, consistent with our results. The closest 2C-value to 77.3 pg is the

2C-value of tetraploid C. graminea with 78. 55 pg and 2n = 24 chromosomes; tetraploid

C. graminea plants with broad leaves may be misidentified as C. rosea (Giles 1942). A voucher specimen of C. rosea from Jones and Kenton (1984) was not reported, so we cannot assess if the plant material used for the DNA content analysis was identified correctly. A misidentification is likely since the genome size estimation of Hertweck

(2011) is close to our values. Likewise, previous tetraploid counts (Anderson and Sax

1936, Ichikawa and Sparrow 1967, Jones and Kenton 1984) may also be for tetraploid

C. graminea plants that were misidentified as C. rosea. Alternatively, there may be cryptic tetraploidy in C. rosea that we failed to detect, but given our extensive sampling, we do not believe this to be the case.

Distribution

As shown in Figure 2-5, two isolated populations of diploid C. graminea were detected. One population is in Suffolk County, VA, and the other is in North and South

Carolina. These two isolated populations may have been part of a once larger geographic range for diploid C. graminea, but due to heavy agricultural activities in this part of North Carolina, suitable habitats ranging from Johnston County to Northampton

County were transformed to farmland (personal observation). This anthropogenic influence may have caused the separation of the two isolated groups of diploid C. graminea. Tetraploid C. graminea ranges from the coastal plain of the Carolinas to central Florida, with additional populations in the Florida panhandle (Franklin County,

FL). This cytotype is clearly more abundant than diploid C. graminea; it is usually found in xeric disturbed areas and exhibits a larger growth form than diploid C. graminea.

These tetraploids were abundant in Bladen and southern Cumberland Counties, NC, which border the isolated locality of diploid C. graminea in North Carolina. These two areas (occupied by tetraploid and diploid plants, respectively) are separated by the city of Fayetteville, NC. Although diploid and tetraploid entities of C. graminea were reported to be geographically isolated (Bergamo 2003, Giles 1942, 1943, Kelly 1991), one tetraploid individual was found within a diploid population in Cheraw State Park, SC; this individual is morphologically similar to the surrounding diploid C. graminea. This finding supports Giles’s (1942) hypothesis that tetraploid C. graminea is an autotetraploid because it occurs consistently with diploid C. graminea. This hypothesis requires testing with molecular data.

The Fall Line runs essentially east-west through Georgia and from southwest to northeast in the Carolinas. Diploid C. rosea occurs on both sides of the Fall Line from

Georgia to North Carolina. In Fort Gordon (Richmond County, GA), diploid C. rosea occurs in sympatry with tetraploid C. graminea. Although these two species occur in sympatry, hybrids were not observed at the site.

Diploid C. ornata is endemic to Florida, and tetraploid individuals of C. ornata occur in western Florida. These individuals may be autopolyploid, with diploid C. ornata as their progenitor. The distribution map in Figure 5 clearly supports the assumption of autopolyploidy, because there are no other Callisia species recorded in the region of diploid and tetraploid C. ornata. Morphologically, tetraploid C. ornata individuals show an increased axillary branching pattern, which is less common in diploid individuals.

Axillary branching is a characteristic of C. graminea. Tetraploid C. graminea and diploid

C. ornata are likely parents, through the union of one reduced gamete of tetraploid C. graminea and one unreduced gamete of diploid C. ornata.

In South Carolina, one hexaploid individual of C. graminea was found growing sympatrically with multiple tetraploid individuals of C. graminea. Hexaploid C. graminea in South Carolina appeared to be rare, and in 1942 only one individual was reported by

Giles (1942). These rare hexaploid individuals may be allopolyploids, with diploid C. rosea and tetraploid C. graminea as their parents or autopolyploids with tetraploid C. graminea as their progenitor. Regarding allopolyploidy, C. rosea was not found sympatrically with tetraploid C. graminea in South Carolina; however, from the map of georeferenced specimens (Figure 2-1), there is a significant overlap of distribution between tetraploid C. graminea and diploid C. rosea in the Carolinas. With regard to autopolyploidy, individuals may have resulted through the union of one reduced and one

unreduced gamete of tetraploid C. graminea given that no other Callisia species were observed in the population.

In Lake and Hernando Counties, FL, hexaploid individuals exhibited intermediate morphological characteristics between C. graminea and C. ornata. Some populations had typical tetraploid C. graminea or diploid C. ornata characteristics (Figure 2-2). Two forms were distinguished based on habit: (1) hexaploid C. graminea and (2) hexaploid

C. ornata. Hexaploid C. graminea and one of its possible progenitors, tetraploid C. graminea, grow in sympatry at the Seminole State Forest, and hexaploid C. ornata was found growing with tetraploid C. graminea at the entrance to Brantley Branch Rd.

(Seminole State Forest). The co-occurrence of hexaploids and tetraploids suggests that the hexaploids may be of allopolyploid origin. Hexaploid C. graminea was also collected at Lake Griffin State Park, Edward Rd., Lady Lake, and Seminole State Forest, FL. In

Dunns Creek State Park and Welaka State Forest, diploid C. ornata and tetraploid C. graminea occur in sympatry; however, hexaploids were not found in these contact zones.

The rare hexaploid collected in South Carolina is most likely independently evolved from the hexaploids from Florida, and this entity from South Carolina could be either an allo- or autopolyploid. If allopolyploid, one likely parent, C. rosea, only occurs in Georgia and the Carolinas; if autopolyploid, the likely parent is tetraploid C. graminea.

The hexaploid entities of Florida might be allopolyploid due to the intermediate morphological characters, with diploid C. ornata and tetraploid C. graminea as progenitors.

Callisia graminea forma leucantha, which was reported near Tampa, FL, was not found, but one white-flowered tetraploid individual of C. graminea was encountered among pink-flowered individuals in each of the following three locations:

Sesquicentennial State Park, SC; Chesterfield Co., SC; and Tate’s Hell State Forest,

FL. One white-flowered individual of diploid C. rosea was found in Heggie’s Rock

Preserve, Appling, GA. White flowers reflect an absence of anthocyanins, which may result from mutations in any of the genes in the anthocyanin pathway or from lack of expression of potentially functional genes (Ho and Smith 2016, Rausher 2008). In

Callisia section Cuthbertia, variation in flower color is common, but there is no association between color and ploidy within or among populations. Loss of anthocyanin pigments seems to occur sporadically within this complex.

Morphological and molecular analysis is an important next step in unraveling the complex relationships among cytotypes of Callisia section Cuthbertia. This work will allow us to reveal the parentage, evolutionary history, and the evolutionary role of all cytotypes within Callisia section Cuthbertia.

Table 2-1. Populations used in this study. Geographic location, ploidy, number of plants of each ploidy, total number of analyzed individuals, and voucher information for 133 populations of Callisia graminea (G), C. ornata (O), and C. rosea (R) from the southeastern United States. * indicates a new locality with voucher specimen. Geographic coordinates Ploidy / Number of plants Population Locality State County Latitude Longitude 2x 4x 6x N Voucher #. Callisia graminea (Small) G. Tucker G-1* Gainesville Regional Airport FL Alachua 29°42.01'N 082°15.72'W 1 3 307 G-2 Jct. Tower Rd. and SW 8 Ave FL Alachua 29°38.63'N 082°25.24'W 1 4 223 G-3 Morningside Nature Center FL Alachua 29°39.56'N 082°16.45'W 1 1 234 G-4 Jct. Hwy 200 and CR. 491 FL Citrus 28°58.51'N 082°21.84'W 1 2 229 G-5* Along Rod Rd. FL Clay 30°01.52'N 081°51.95'W 1 1 225 G-6 Golden Branch Head State Park FL Clay 29°50.75'N 081°57.04'W 1 2 309 G-7 Silver Sand Lake Rd. FL Clay 29°47.49'N 081°58.32'W 1 4 311 G-8* Tate Hell State Forest along New FL Franklin 29°52.42'N 084°41.79'W 1 4 306 River G-9* Richloam State Forest/Dark Stretch FL Hernando 28°29.10'N 082°08.87'W 1 6 349 Rd. G-10* Edwards Rd., Lady Lake FL Lake 28°54.12'N 081°53.40'W 1 3 235 G-11* Lake Griffin State Park FL Lake 28°52.31'N 081°53.41'W 1 3 236 G-12* Seminole State Forest along Co. Rd. FL Lake 29°00.82'N 081°31.05'W 1 1 3 345 42 G-13* Seminole State Forest FL Lake 28°49.31'N 081°28.01'W 1 1 362 G-14* Lake Norris Rd. FL Lake 28°54.89'N 081°32.41'W 1 1 363 G-15* ATV trail at Ocala National Forest FL Marion 29°21.76'N 081°44.21'W 1 1 230 G-16 Silver River State Park FL Marion 29°12.15'N 082°02.77'W 1 4 348 G-17* Along Mason Rd. FL Putnam 29°42.50'N 082°00.77'W 1 2 224 G-18* Ordway Biological Center H1 & H2 FL Putnam 29°41.70'N 081°57.87'W 1 2 302 area G-19* Etoniah Creek State Forest FL Putnam 29°46.43'N 081°51.91'W 1 3 308 G-20 Dunns Creek State Park entrance FL Putnam 29°31.84'N 081°35.34'W 1 4 310 Sisco Rd. G-21* Welaka State Forest FL Putnam 29°28.24'N 081°39.37'W 1 2 360a

Table 2-1. Continued. Geographic coordinates Ploidy / Number of plants Population Locality State County Latitude Longitude 2x 4x 6x N Voucher #. G-22 Along State Rd. 46 GA Bulloch 32°20.94'N 081°50.57'W 1 3 242 G-23 Hwy 185 Jct. Turkey Ridge GA Charlton 30°24.76'N 082°11.70'W 1 2 317 G-24* General Coffee State Park GA Coffee 31°31.50'N 082°46.33'W 1 1 318 G-25 N. Connector Rd./206 Jct. 135 GA Coffee 31°32.27'N 082°48.75'W 1 3 319 G-26* George Smith State Park GA Emanuel 32°32.64'N 082°07.32'W 1 6 241 G-27* Ochicoo Preserve, Halls Bridge Rd. GA Emanuel 32°31.73'N 082°27.38'W 1 4 320 G-28 Fort Stewart GA Evans 32°06.92'N 081°47.10'W 1 4 243 G-29* Conway CT./Interstate Parkway GA Richmond 33°29.24'N 082°06.12'W 1 1 322 G-30 Fort Gordon GA Richmond 33°23.33'N 082°14.56'W 239 G-31* Singletary Lake State Park NC Bladen 34°35.41'N 078°26.87'W 1 3 263 G-32* Jones Lake State Park NC Bladen 34°42.11'N 078°37.22'W 1 3 268 G-33* Jones Lake State Park NC Bladen 34°42.11'N 078°37.22'W 1 269 G-34* Along NC 242 near Jones Lake State NC Bladen 34°42.00'N 078°36.35'W 1 2 270 Park G-35* Along NC 242 N. of Jones Lake State NC Bladen 34°45.40'N 078°36.56'W 1 5 271 Park G-36* White Lake, along NC 741, Barnes NC Bladen 34°39.41'N 078°30.17'W 1 5 272 Food Co. G-37* Jones Lake State Park. campsite NC Bladen 34°40.79'N 078°35.99'W 274 G-38* Along Burney Rd. underneath NC Bladen 34°44.38'N 078°43.68'W 1 4 334 powerline G-39* River Rd., underneath powerline NC Bladen 34°46.18'N 078°47.24'W 1 3 335 G-40 Bay Tree Lake State NC Bladen 34°40.22'N 078°25.66'W 1 6 261 Park/undeveloped G-41 Along Hwy 41 close to Bay Tree Lake NC Bladen 34°41.21'N 078°25.26'W 1 3 262 State Park G-42 Along Hwy 11 towards Delco under NC Bladen 34°24.61'N 078°15.60'W 1 4 266 powerline G-43 Along Jessup Pond NC Bladen 34°51.72'N 078°43.76'W 275 G-44 Lake Waccamaw State Park. NC Columbus 34°16.73'N 078°27.89'W 267

Table 2-1. Continued. Geographic coordinates Ploidy / Number of plants Population Locality State County Latitude Longitude 2x 4x 6x N Voucher #. G-45* Mack Simmons Rd. NC Cumberland 34°54.45'N 078°44.20'W 276 G-46* Along NC 210, Jct. with Sidney NC Cumberland 34°58.69'N 078°43.84'W 1 4 278 Bullard Rd. G-47* Ft. Bragg/John Mill Rd. NC Cumberland 35°10.70'N 079°05.39'W 1 3 341 G-48* Ft. Bragg/NE. training/Mc Closkey Rd. NC Cumberland 35°09.84'N 078°56.97'W 1 3 342 G-49 Cedar Creek Rd., Tatum farm NC Cumberland 34°56.32'N 078°44.58'W 1 1 277 G-50 Along Dunns Rd./NC 301 NC Cumberland 35°06.42'N 078°46.52'W 279 G-51 Open Area along NC 24 NC Harnett 35°15.61'N 079°02.47'W 1 3 284 G-52 Along Rockfish Rd. NC Hoke 34°59.32'N 079°05.82'W 1 3 286 G-53 In open area along Red Springs Rd. NC Hoke 34°52.38'N 079°12.17'W 1 4 287 G-54* Weymouth Sandhill Nature Preserve NC Moore 35°08.95'N 079°22.10'W 1 3 288 G-55 Along Riverview Dr. NC Moore 35°11.48'N 079°10.94'W 1 3 285 G-56 Along NC 11/ Hwy 53 NC Pender 34°29.72'N 078°11.49'W 1 3 264 G-57 Along NC 11/ Hwy 53 NC Pender 34°29.72'N 078°11.49'W 1 1 265 G-58* Grey Woods Rd. NC Richmond 34°57.52'N 079°38.47'W 1 3 297 G-59* Sandhills Game Land NC Richmond 35°01.83'N 079°36.70'W 1 2 336 G-60* Sandhills Game Land/442/Ledbetter NC Richmond 35°03.62'N 079°38.09'W 1 3 337 Rd. G-61* Sandhills Game Land NC Richmond 34°58.61'N 079°30.42'W 1 2 338 G-62* Sandhills Game Land SR 1331, NC Richmond 34°58.50'N 079°26.93'W 1 2 339 15/501 G-63* Sandhills Game Land, Aberdeen NC Richmond 34°59.49'N 079°26.76'W 1 3 340 Rd./Hill Creek Rd. G-64 Sandhills Game Land along NC Richmond 35°01.24'N 079°37.18'W 1 2 290 McDonald Church Rd. G-65 NC Hwy 177 NC Richmond 34°50.41'N 079°45.54'W 1 1 295 G-66 Along Saint Stevens Church Rd. NC Richmond 34°49.82'N 079°50.55'W 1 1 296 G-67 NC 242, 0.3 mi N. of Cumberland Co. NC Sampson 34°53.35'N 078°31.28'W 1 3 273 line

Table 2-1. Continued. Geographic coordinates Ploidy / Number of plants Population Locality State County Latitude Longitude 2x 4x 6x N Voucher #. G-68 Along Spiveys Corner Hwy. NC Sampson 35°10.72'N 078°28.65'W 1 2 280 G-69 Edge camp Mackall along Aberdeen NC Scotland 35°00.84'N 079°26.70'W 1 2 289 Rd. G-70 Along 1328, Hoffman Rd./Butler Rd. NC Scotland 34°59.14'N 079°31.99'W 1 2 291 G-71 Along Peach Orchard Rd. under NC Scotland 34°55.77'N 079°23.86'W 1 3 292 powerline G-72 Along US 401 and forest edge NC Scotland 34°50.49'N 079°23.98'W 1 1 293 G-73 Along forest edge of Hamlet Rd. NC Scotland 34°48.01'N 079°38.03'W 1 2 294 G-74 Along Piney Grove Church Rd. NC Wayne 35°17.32'N 077°50.92'W 1 1 281 G-75* Aiken State Park SC Aiken 33°32.55'N 081°28.92'W 1 4 324 G-76* Parcel at Jct. Hwy 283 and US SC Aiken 33°36.11'N 081°41.04'W 1 5 325 1/Columbia Hwy N G-77 Aiken Gopher Tortoise Heritage SC Aiken 33°30.00'N 081°24.52'W 1 1 231 Preserve G-78* Carolina Sandhills National Wildlife SC Chesterfield 34°31.46'N 080°13.63'W 1 3 331 Refuge G-79* Sandhill State Forest SC Chesterfield 34°33.37'N 080°03.84'W 1 3 332 G-80* H. Cooperblack Jr. Memorial SC Chesterfield 34°34.03'N 079°55.75'W 1 2 333 trail/James Rd. G-81 Along Hwy 102 SC Chesterfield 34°38.30'N 080°05.22'W 1 5 249 G-82 Teals mill Rd./Cheraw State Park SC Chesterfield 34°37.25'N 079°56.70'W 1 1 3 250 G-83 W. Old Camden Rd. SC Chesterfield 34°22.28'N 080°16.92'W 1 3 252 G-84 US 1 SC Chesterfield 34°26.17'N 080°17.44'W 1 2 253 G-85 Along Old Stagecoach Rd. SC Chesterfield 34°20.96'N 080°21.27'W 1 3 254 G-86 Along Old Georgetown Rd. E. SC Chesterfield 34°22.99'N 080°23.29'W 1 1 255 G-87 Co. Rd. S. 18-137 SC Dorchester 32°54.02'N 080°23.11'W 1 4 248 G-88 Tillman Sand Ridge Heritage SC Jasper 32°29.69'N 081°11.55'W 1 5 247 Preserve, Sandhill Rd. G-89* Along Jefferson Davis Hwy/US 1 SC Kershaw 34°18.73'N 080°32.49'W 256 G-90* Goodale State Park SC Kershaw 34°17.42'N 080°31.55'W 1 3 329

Table 2-1. Continued. Geographic coordinates Ploidy / Number of plants Population Locality State County Latitude Longitude 2x 4x 6x N Voucher #. G-91* Jefferson Davis Hwy/US 1 SC Kershaw 34°22.04'N 080°25.92'W 1 4 330 G-92* Lee State Park SC Lee 34°11.81'N 080°11.36'W 1 3 251 G-93 Shealy’s Pond Heritage Preserve SC Lexington 33°51.82'N 081°14.19'W 1 1 232 G-94 Peachtree Rock Preserve SC Lexington 33°49.71'N 081°12.11'W 1 1 233 G-95* Ft. Jackson, Area 26 B firebreak 16 SC Richland 34°00.85'N 080°47.40'W 1 2 257 G-96* Ft. Jackson, Area 34 B near Chauers SC Richland 34°02.36'N 080°43.30'W 1 3 258 Pond Rd. G-97* Ft. Jackson, Area 11 E. of Wildcat Rd. SC Richland 34°05.06'N 080°50.61'W 1 2 259 G-98 Ft. Jackson, S. edge of pond of SC Richland 33°59.96'N 080°50.03'W 1 1 260 Westons Recreation G-99 Sesquicentennial State Park SC Richland 34°05.82'N 080°54.57'W 1 3 326 G-100* Sesquicentennial State Park SC Richland 34°04.92'N 080°54.38'W 1 1 4 327 G-101 Faunas Rd. SC Richland 34°08.34'N 081°02.33'W 1 5 328 G-102* Forks of River Rd. VA Southampto 36°33.85'N 076°55.96'W 1 2 282 n G-103 Suffolk City, DCR VA Suffolk City 36°33.77'N 076°54.82'W 283 Callisia ornata (Small) G. Tucker O-1* Turkey Creek Sanctuary FL Brevard 28°01.01'N 080°36.18'W 1 1 315 O-2* Sebastian State Park FL Brevard 27°50.19'N 080°31.56'W 1 2 361 O-3 Wickham Park FL Brevard 28°09.64'N 080°39.54'W 1 1 314 O-4* Highlands State Park FL Highlands 27°28.85'N 081°31.57'W 1 4 301 O-5 Sebring Amtrak Station FL Highlands 27°29.75'N 081°26.06'W 298 O-6 Lake June in Winter Scrub State Park FL Highlands 27°17.83'N 081°25.14'W 1 2 300 O-7 Little Manatee State Park/Mustang FL Hillsborough 27°40.08'N 082°22.1'W 1 4 350 trail O-8 Little Manatee State Park/Dude trail FL Hillsborough 27°39.93'N 082°22.38'W 1 3 351 O-9* Seminole State Forest/entrance FL Lake 28°53.20'N 081°27.60'W 1 4 343 Brantley Branch Rd.

Table 2-1. Continued. Geographic coordinates Ploidy / Number of plants Population Locality State County Latitude Longitude 2x 4x 6x N Voucher #. O-10* Seminole State Forest/the Simson FL Lake 28°52.94'N 081°31.08'W 1 4 344 track O-11* Seminole State Forest/Warea tract FL Lake 28°29.99'N 081°40.03'W 1 3 346 O-12* Lake Louisa State Park/Primitive FL Lake 28°27.17'N 081°44.13'W 1 4 347 campsite O-13 Jonathan Dickinson State Park/Nature FL Martin 26°59.58'N 080°08.83'W 1 4 353 trail picnic area O-14* Tiger Creek Preserve along FL Polk 27°48.41'N 081°29.81'W 1 1 228 Pfundstein Rd. O-15* Arbuckle State Forest, School Bus FL Polk 27°39.75'N 081°23.84'W 1 3 316 Rd. O-16* Lake Kissimmee State Park, Buster FL Polk 27°55.39'N 081°21.82'W 1 2 354 Island O-17* Lake Kissimmee State Park, Catfish FL Polk 27°57.84'N 081°22.77'W 1 5 355 Creek O-18* Lake Kissimmee State Park Main FL Polk 27°57.91'N 081°28.34'W 1 5 356 entrance O-19* Welaka State Forest FL Putnam 29°28.24'N 081°39.37'W 1 1 360B O-20 Dunns Creek State Park entrance FL Putnam 29°33.34'N 081°34.94'W 1 2 312 Sisco Rd. O-21 Oscar Scherer State Park along FL Sarasota 27°10.17'N 082°27.41'W 1 5 352 Legacy trail O-22* Tiger Bay State Forest FL Volusia 29°10.22'N 081°09.56'W 1 3 313 O-23* Lake George State Forest FL Volusia 29°11.84'N 081°30.55'W 1 1 364 O-24* Deland FL Volusia 29°00.11'N 081°13.25'W 1 1 365 Callisia rosea (Vent.) D.R. Hunt R-1 Along Chert Quarry Rd. SC Allendale 33°02.28'N 081°28.26'W 1 3 245 R-2* Heggie’s Rock Preserve GA Colombia 33°32.34'N 082°15.09'W 1 3 321 R-3* Lake Russel State Park GA Elbert 34°09.60'N 082°44.42'W 1 3 237 R-4* Bobbie Brown State Park GA Elbert 33°58.35'N 082°34.64'W 1 3 238 R-5* Elijah Clarke State Park GA Lincoln 33°51.22'N 082°24.02'W 1 3 323

Table 2-1. Continued. Geographic coordinates Ploidy / Number of plants Population Locality State County Latitude Longitude 2x 4x 6x N Voucher #. R-6 Fort Gordon GA Richmond 33°23.49'N 082°14.54'W 1 3 240 R-7 Fort Stewart GA Tattnall 32°02.54'N 081°48.84'W 1 4 244

Figure 2-1. Distribution map of Callisia section Cuthbertia. Distribution of Callisia graminea, C. ornata, and C. rosea based on georeferenced data. Multiple species occurring in sympatry are designated by superimposed symbols; these locations are further indicated by black lines that highlight the symbols.

Figure 2-2. Habit of Callisia section Cuthbertia. A) diploid Callisia graminea B) diploid C. graminea flower C) diploid C. ornata D) diploid C. ornata flower E) diploid C. rosea and F) diploid C. rosea flower. Illustrations by Sofia Chang.

Figure 2-3. Mitotic metaphase chromosomes spreads from root tips. A) diploid Callisia graminea (2n = 2x = 12) B) tetraploid C. graminea (2n = 4x = 24) C) hexaploid C. graminea (2n = 6x = 36) D) diploid C. ornata (2n = 2x = 12) E) tetraploid C. ornata (2n = 4x = 24) F) hexaploid C. ornata (2n = 6x = 36) and G) diploid C. rosea (2n = 2x = 12). June 30, 2016. Gainesville, FL. Photo courtesy of author.

Table 2-2. Genome sizes (2C) of Callisia section Cuthbertia and their cytotypes and previously reported 2C-values. Voucher numbers apply only to the current study. Species Chromosomes 2C value (pg) Hertweck 2011 Jones and Kenton 1984 C. graminea 2x (IEM 342) 2n = 12 41.75 ± 0.67 C. graminea 4x (IEM 251) 2n = 24 78.55 ± 0.42 C. graminea 6x (IEM 236) 2n = 36 122.86 ± 0.8 C. ornata 2x (IEM 353) 2n = 12 48.51 ± 1.09

C. ornata 4x (IEM 352) 2n = 24 87.99 ± 0.4 C. ornata 6x (IEM 349) 2n = 36 129.73 ± 0.56 C. rosea 2x (IEM 237) 2n = 12 43.70 ± 1.78 43.52 77.3

Figure 2-4. Histograms of fluorescence intensity (FL2-A) of propidium iodide-stained nuclei. A) diploid C. graminea B) tetraploid C. graminea; and C) hexaploid C. graminea. Vicia faba was used as the internal standard.

Figure 2-5. Distribution of cytotypic variation in Callisia section Cuthbertia. Diploid C. graminea (red circles) ranges from Virginia to North and South Carolina; tetraploid C. graminea (purple circles) occurs along the coastal plain from North Carolina to central Florida; hexaploid C. graminea (black plus signs) is restricted to central Florida. Diploid C. ornata (red squares) occurs in eastern and central Florida; tetraploid C. ornata (purple squares) is restricted to central and western peninsular Florida; hexaploid C. ornata (green plus signs) is restricted to central Florida. Callisia rosea (all diploid; green diamonds) occurs along the Georgia – South Carolina border. Localities with multiple cytotypes or taxa are indicated by black lines. Note: The black plus signs are the hexaploids of C. graminea, and the green plus signs are hexaploids of C. ornata

CHAPTER 3 THE EVOLUTIONARY RELATIONSHIPS AMONG THE CYTOTYPES OF CALLISIA SECTION CUTHBERTIA (COMMELINACEAE): MORPHOLOGICAL AND MOLECULAR APPROACHES

Introduction

Callisia Loefling is one of 39 genera in subfamily Commelinoideae

(Commelinaceae) (Burns et al. 2011) and comprises approximately 23 species of perennial or rarely annual herbs distributed throughout the U.S.A., Central America,

South America, and the Caribbean (Hunt 1986, Tucker 1989). Cladistic analysis of morphological and molecular data (Evan et al 2000, Evans et al 2003), revealed that

Callisia sensu Hunt (1986) is not monophyletic. Bergamo (2003) analyzed the phylogenetic relationship among Callisia taxa with the chloroplast loci ndhF and trnL-F and concluded that Callisia is not monophyletic and that the genus Tripogandra Raf. is embedded within it. Recent phylogenetic work based on chloroplast sequence data for trnL-F and rpL16 (Hertweck and Pires 2014) confirmed the paraphyly of Callisia, but found that Callisia section Cuthbertia (Hunt 1986) is monophyletic.

Polyploidy, a key force in plant diversification (De Bodt et al. 2005, Rensing et al.

2008, Soltis and Soltis 2009, Soltis and Soltis 2016, Tank et al. 2015), is a common phenomenon throughout Callisia (Anderson and Sax 1936, Faden 1998, Giles 1942,

Jones and Jopling 1972, Leitch et al. 2010), and species limits and ploidal levels are poorly understood (Grabiele et al. 2015). Allopolyploidy involves hybridization and genome duplication and is considered more common than autopolyploidy.

Autopolyploids arise within a single species and are frequently lumped as cytotypes with their diploid progenitor (Lewis 1967), and this treatment may lead to misinterpretations of evolutionary lineages and estimation of biodiversity (Soltis et al. 2007). Due to

increasing threats to biodiversity, it is extremely important to accurately assess polyploid entities that merit recognition, because inaccurate delimitation of species may obstruct biological inferences and conservation efforts (Wiens 2007). To guide conservation efforts, it is essential to investigate and understand the genetic and evolutionary history of species (Ellstrand and Elam 1993).

The southeastern United States harbors a poorly understood endemic polyploid complex, Callisia section Cuthbertia. These species range from southeastern Virginia to southern Florida (habitat: Sandhills, Flatwoods and Scrubby-Flatwoods) and share a base chromosome number of x = 6 (Giles 1943). Giles (1942) reported three ploidal levels (2x, 4x, and 6x) in Callisia graminea, and in Chapter 2, I reported three ploidal levels (2x, 4x, and 6x) in C. ornata; polyploidy has not been recorded in C. rosea.

While it is clear that Callisia section Cuthbertia is monophyletic and that Callisia ornata is sister to C. graminea and C. rosea (Hertweck and Pires 2014), relationships among the polyploid cytotypes and their putative progenitors are not clear. Based on the current knowledge of the Callisia section Cuthbertia polyploid complex, I address the following questions: 1) Are the cytotypes morphologically differentiated? 2) If so, which morphological characters distinguish them? 3) What are the relationships among diploid and polyploid entities? 4) What taxa are the progenitors of the polyploid cytotypes?

These questions were approached by implementing a canonical discriminant analysis with morphometric data and by inferring the phylogeny of section Cuthbertia using nuclear and chloroplast sequence data for 159 accessions.

Materials and Methods

Taxon Sampling

To investigate the evolutionary relationships among the three species and cytotypes of Callisia section Cuthbertia, I analyzed morphometric data and used phylogenetic analysis of DNA sequence data from plant samples collected throughout the range of the complex. During the summers of 2012 – 2015, I collected wild mature plants in Florida, Georgia, South Carolina, and Virginia. In all, 300 plants were collected at 133 localities. Two to six live plants were collected per population (Chapter 2). All plants were grown under uniform conditions at the Department of Biology, University of

Florida greenhouse. Herbarium specimens were deposited in the University of Florida

Herbarium (FLAS).

Morphological Analysis

Data were collected on plant height (H), leaf length (LL), leaf width (LW), stem thickness (ST), sepal length (SL), sepal width (SW), petal length (PL), petal width (PW), and filament length (FL). In all, 1386 measurements were made from 61 C. graminea

2x, 133 C. graminea 4x, 11 C. graminea 6x, 20 C. ornata 2x, nine C. ornata 4x, 22 C. ornata, 6x, and 22 C. rosea individuals (approximately five measurements per plant) for the traits H, LL, LW, ST. For flower morphology (SL, SW, PL, PW and FL), 25 measurements were taken per cytotype (five flowers per plant). Height, LL, and LW were measured by ruler or measuring tape; stem thickness was measured with an iGaging IP54 Electronic Digital Caliper. Because the petals become deliquescent at noon, flowers were harvested in the morning and dissected with forceps. Petals, sepals, filaments with anther, and the remainder of the reproductive parts were carefully removed and placed on archival tape, which was attached to a 2.5” by 3.5” inch card of

archival paper. These cards were placed in a Ziploc bag with silica gel to dry to prevent bleeding of the petals. When dried, the cards were scanned with a ScanMaker 9800XL

Microtek scanner, and flower parts were measured with ImageJ 1.48v (Schneider et al.

2012).

The recorded measurements were grouped by ploidal level based on Chapter 2 and statistically analyzed with a canonical discriminant analysis (CANDISC) in SAS software (SAS Institute Inc., Cary, NC, U.S.A.) with adjusted p-values using the

Bonferroni correction (Sokal and Rohlf 1997). CANDISC is a dimension-reduction procedure that computes squared Mahalanobis distances between group means and performs both univariate and multivariate one-way analysis of variance. This technique finds a linear combination of traits that holds the highest possible multiple correlation with the group, which defines the variable that best defines separation of the species and cytotypes with Callisia section Cuthbertia. Two data sets were analyzed; 1) vegetative traits (H, LL, LW, ST) and 2) vegetative and reproductive parts (H, LL, LW,

ST, SL, SW, PL, PW, FL). The data set was separated to compensate for missing floral morphometric data from some samples. In the latter analysis, only two plants of Callisia graminea 6x were included. All samples per data set were analyzed together, however the following scatterplots were created to visualize the mean distance among specific cytotypes (based on cytogeography and cytology) within Callisia section Cuthbertia: 1)

Callisia graminea, C. ornata, and C. rosea diploids only, 2) C. graminea 2x, 4x and C. rosea, 3) C. ornata 2x, 4x and C. graminea 4x, and 4) C. ornata 2x, 6x and C. graminea

4x, 6x.

Molecular Phylogenetic Analysis

Samples

Leaf tissue was obtained from wild-collected and cultivated plants (141 samples:

108 Callisia graminea (32 diploid, 72 tetraploid, four hexaploid); 23 C. ornata (11 diploid, seven tetraploid, five hexaploid); seven C. rosea, one Tradescantia ohiensis Raf. (IEM

246) and one Commelina erecta L. (IEM 299)). Cuttings of additional outgroup taxa (13

Callisia taxa, three Gibasis taxa, two Tripogandra taxa) (Table 3-1) were provided by

Robert Faden (Smithsonian Institution) and cultivated in the Department of Biology,

University of Florida greenhouse. Vouchers were deposited in FLAS.

DNA extraction

Total genomic DNA was extracted from silica-dried material using the cetyl trimethylammonium bromide (CTAB) technique (Doyle & Doyle, 1987), modified and scaled to a 1-ml volume. For each sample, approximately 15 mg of leaf tissue was ground by bead mill and mixed in 1.2 ml CTAB 2x buffer with 50 μg of Proteinase K.

Total genomic DNA was cleaned with QIAquick PCR purification kits (Qiagen, Valencia,

California) to remove inhibitory secondary compounds and was resuspended in 100 μL of Tris-EDTA (TE) buffer.

PCR amplification and sequencing

I amplified ITS, six intergenic chloroplast spacers based on variability (trnL-F, rpL32, trnQ, trnH-K, trnS and trnG), and the chloroplast gene ycf1 (Table 3-2).

Amplification of these target loci was performed via polymerase chain reaction (PCR) in an Eppendorf Mastercycler EP Gradient S and a Whatman Biometra T3 thermocycler.

PCR amplification of ITS, trnL-F(e-f), and rpL32 was achieved using Taq polymerase

(produced in the Soltis lab from E. coli) whereas Phusion® high-fidelity DNA

polymerase (ThermoFisher Scientific, Waltham, Massachusetts) was used for the remaining loci. Phusion® high-fidelity DNA polymerase was used because of the time efficiency, high accuracy, long reads, and the ability to improve the quality of the sequence by reducing the formation of homopolymers (Fazekas et al. 2010).

The ITS region was amplified with the primers Y5 and Y4 from Hoshi et al. (2008).

PCR mixtures for 25-μL amplification for ITS reactions were as follows: 1 μL DNA template, 10.5 μ L H2O, 7 μL betaine, 2.5 μL 10x buffer, 2.5 μ L 25 mM MgCl2, 0.5 μL

2.5 mmol/L DNTPs, 0.5 μL each 10 pmol/μL primer, and 0.2 μL Taq polymerase. Due to difficulties in amplification of the ITS region, I suspected a high GC content. To achieve complete denaturation in templates with a high GC content, it is recommended to use a higher denaturation temperature and a longer incubation time (Grunenwald 2003). A five-minute denaturation step at 99°C was added before the initial PCR program. The

Taq polymerase was omitted in this additional step and was added before the remaining steps because it will become in active at 99°C (Gelfand 1989, Grunenwald 2003). The following PCR parameters were used: 99°C for 5 min, 80°C for 5 s pause to add Taq polymerase, continue at 94°C for 2 min, followed by 37 cycles (94°C for 20 s, 55°C for

20 s, 72°C for 1min) with a final extension at 72°C for 2 min.

The trnL-F region was amplified using the primers “e” and “f” (Taberlet et al. 1991).

The PCR mixtures for 25-μL amplification reactions were as follows: 1 μL DNA template, 13.5 μ L H2O, 5 μL 5x buffer, 2.5 μ L 25 mM MgCl2, 2 μL 2.5 mmol/L DNTPs,

0.5 μL each 10 pmol/μL primer, and 0.2 μL Taq polymerase. The PCR parameters were: initial denaturation at 95°C for 5 min, six -1°C touchdown cycles (94°C for 1 min,

58-53°C for 1 min, 72°C for 2.5 min), 34 regular cycles (94°C for 1 min, 52°C for 1 min,

72°C for 2.5 min), and a final extension at 72°C for 12 min. Due to the low nucleotide variability among the samples, only 73 taxa of the 149 samples were amplified.

Amplification of the rpl32 region was achieved using the primers rpL32 and trnL

(Shaw et al. 2007). PCR mixtures for 25-μL amplification reactions were as follows: 1 μL

DNA template, 17.5 μ L H2O, 2.5 μL 10x buffer, 2.5 μ L 25 mM MgCl2, 0.5 μL 2.5 mmol/L DNTPs, 0.5 μL each 10 pmol/μL primer, and 0.2 μL Taq polymerase. The PCR parameters were: initial denaturation at 94°C for 3 min, eight -1°C touchdown cycles

(94°C for 30 s, 60-52°C for 45 s, 72°C for 2 min), 30 regular cycles (94°C for 30 s, 50°C for 45 s, 72°C for 2 min), and a final extension at 72°C for 3 min.

For the following PCR reactions using Phusion® Taq, I used 20-μL reaction volumes: 1 μL DNA template, 12 μ L H2O, 4 μL 5x Phusion® HF buffer, 1 μL 50 mM

MgCl2, 0.6 μL DMSO, 0.5 μL 2.5 mmol/L DNTPs, 0.5 μL each 10 pmol/μL primer, and

0.2 μL Phusion® Taq polymerase.

The 3’ ycf1 region was amplified in two pieces with the primers F317, R1867 and

F1897, R3080. These two primer pairs yielded 2600 bp and were designed from genome skimming data from tetraploid C. graminea, obtained from shotgun sequencing (Malé et al. 2014) and assembled by BLAT (Kent 2002) against Xiphidium caeruleum Aubl.

(Haemodoraceae) and Hanguana malayana (Jack) Merr. (Hanguanaceae), both in

Commelinales. Primers were designed and tested in Geneious® 8.1.3 (Biomatters Ltd.,

Auckland, New Zealand). The PCR parameters were: initial denaturation at 98°C for 30 s, 33 cycles (98°C for 10 s, 56°C for 15 s, 72°C for 40 s), and a final extension at 72°C for 5 min. When amplification of ycf1(317-1867) for certain taxa was problematic, a touchdown approach was used with the following parameters: initial denaturation at

98°C for 2 min, six -1°C touchdown cycles (98°C for 10 s, 56-51°C for 15 s, 72°C for 40 s), 29 regular cycles (98°C for 10 s, 50°C for 15 s, 72°C for 40s), and a final extension at 72°C for 5 min.

The trnQ-5’rpS16 region was amplified using the primers trnQ(UUG) and rpS16x1

(Shaw et al. 2007). The PCR parameters were: initial denaturation at 98°C for 30 s, 33 cycles (98°C for 10 s, 55°C for 15 s, 72°C for 40 s), and a final extension at 72°C for 5 min.

Amplification of trnH-K was accomplished using the primers trnH and trnK

(Nicolosi et al. 2000). The PCR parameters were: initial denaturation at 98°C for 30 s,

31 cycles (98°C for 10 s, 66°C for 15 s, 72°C for 40 s), and a final extension at 72°C for

5 min.

The trnS- psbZ- trnG region was amplified with the primers trnS and trnG (Shaw et al. 2005). The PCR parameters were the same as those used in the amplification of trnH-K.

Amplification of the trnG- trnR region was accomplished using the forward primer trnG from Shaw et al. (2005) and the trnR reverse primer designed using the genome skimming data described above. The PCR parameters were: initial denaturation at 98°C for 30 s, 33 cycles (98°C for 10 s, 64.5°C for 15 s, 72°C for 30 s), and a final extension at 72°C for 5 min.

The PCR products were sequenced at the Interdisciplinary Center for

Biotechnology Research facility (ICBR) at the University of Florida and Eurofins

Genomics (Eurofins MWG Operon LLC, Louisville, Kentucky). Sequences were edited and assembled using Sequencher™ 5.3 (GeneCodes, Inc., Ann Arbor, Michigan) and

aligned using the program MAFFT 7.294 (Katoh and Standley 2013, Yamada et al.

2016) and manually edited in SeaView 4 (Gouy et al. 2010). The data matrices were concatenated with FASconCAT (Kück and Meusemann 2010). All sequences were deposited in GenBank. (See Appendix B)

Phylogenetic analysis

Based on cytological data of Callisia section Cuthbertia (Chapter 2, Giles 1942), multiple data sets were created: 1) diploids, ITS only, 2) diploids, chloroplast only, 3) combined chloroplast and ITS diploids only, 4) all cytotypes, ITS only, 5) all cytotypes, chloroplast only, 6) all cytotypes, combined chloroplast and ITS (total evidence). I analyzed diploids only to investigate the relationships among diploids because including polyploids in phylogenetic analysis may subvert phylogeny reconstruction (Soltis et al.

2008), and to study the origins of the polyploids, all cytotypes were included (Mavrodiev et al. 2008). The data sets were analyzed using Maximum Parsimony (MP) in PAUP* 4a

(build157) (Swofford 2002) and Maximum Likelihood (ML) in RAxML 8.2.10 (Stamatakis

2014). Phylogenetic trees were viewed and edited with MEGA 7.0.26 (Kumar et al.

2016) and FigTree v1.4.3 (Rambaut 2016). MP analyses were conducted on all datasets with heuristic searches with 1,000 random-addition replicates, saving 10 trees per replicate, with the tree bisection reconnection (TBR) algorithm. Support was estimated with 1,000 nonparametric bootstrap (BS) pseudo-replicates, with 10 random- addition replicates (TBR swapping) per bootstrap replicate. All nucleotide characters were weighted equally, and character states were unordered. Gaps were treated as missing data, and indels were not coded as characters. The ML analyses were conducted with RAxML using the GTR+ᴦ model of molecular evolution with 1,000 nonparametric rapid bootstrap pseudo-replicates for all data sets.

Identical or nearly identical sequences for both ITS and chloroplast loci were detected with patristic distance analysis in PAUP* 4a (build157) and the software

ElimDupes (http://hcv.lanl.gov/content/sequence/ELIMDUPES/elimdupes.html). These were removed from the analysis to improve the analysis computationally and avoid biases when interpreting the results. The partition homogeneity test (ILD) (Farris et al.

1995) was used to investigate if the chloroplast and nuclear matrices were congruent.

The ILD was performed with PAUP* 4a (build157) (Swofford 2002) by conducting a heuristic search using 100 replicates and a TBR algorithm comparing the two matrices.

Ten random-addition replicates were performed per ILD replicate, holding 10 trees and saving 10 trees per replicate. The tree topologies from the chloroplast and nuclear trees were also observed by eye for incongruence to justify combining the chloroplast and nuclear data (Johnson and Soltis 1998). Conflicting nodes with at least 70% BS support values were investigated.

General Greenhouse Observations

Collected plants were monitored in the greenhouse during the period 2012-2016 for growth patterns. Observations focused on the flowering period, the susceptibility to pests, production of deformed flowers, and asexual propagation under uniform greenhouse conditions for all cytotypes of Callisia graminea, C. ornata and C. rosea.

These observations were not analyzed statically but were noted each year within the specified period.

Results

Morphological Analysis

Vegetative data set

The CANDISC analysis for the vegetative data set is based on 1386 measurements, four traits (H, LL, LW, ST), and seven putative taxa (C. graminea 2x, 4x,

6x, C. ornata 2x, 4x, 6x, and C. rosea). The squared Mahalanobis distances and

Bonferroni adjusted p-values between putative taxon means are provided in Table 3-3.

Based on the vegetative traits, there is no significant difference between Callisia graminea 6x and C. ornata 6x (p = 0.69). All other putative taxa are significantly different

(p < 0.001). Callisia rosea is most distant from all other putative taxa, followed by C. ornata 2x. In Table 3-4, the mean values of the four vegetative traits for the putative taxa are shown. The canonical structure (Table 3-5) shows that the first canonical component (Can 1) is mainly correlated with stem thickness (ST) (0.997), and the second canonical component (Can 2) is strongly correlated with leaf width (LW) (0.565).

These results indicate that ST is the vegetative character that best discriminates all cytotypes from one another followed by LW. The third canonical variable (Can 3), LL, provides little differentiation between group means and has been omitted from these analyses. Figure 3-1 depicts the canonical discriminant ordination of Callisia section

Cuthbertia based on vegetative traits. Can 1 represents mainly ST, and Can 2 represents mainly LW with support of the other traits H and LL. The two axes explain

70% and 28% of the variation, respectively, with support of the other traits H and LL.

The mean scores are indicated with enlarged symbols in the center of the prediction ovals. Figure 3-1A depicts clear differentiation among the three diploid species, C. graminea, C. ornata, and C. rosea. Callisia graminea has the smallest ST but broader

leaves than C. ornata; C. ornata has a ST that lies between those of C. graminea and

C. rosea but has the smallest LW; C. rosea is completely separated from C. graminea and C. ornata, having the thickest stem and broadest leaves. Figure 3-1B shows that the ST of C. graminea 4x is larger than that of C. graminea 2x, but there is almost no difference in LW. Callisia rosea has a greater ST and LW. Figure 3-1C depicts that C. ornata 2x has a larger ST than C. ornata 6x, C. graminea 4x, and C. graminea 6x. The means of C. ornata 6x and C. graminea 6x are intermediate to those of C. graminea 4x and C. ornata 2x, with C. graminea 4x having the smallest ST and largest LW. Figure 3-

1D shows that C. graminea 4x is most distant from C. ornata 2x and 4x, with smaller ST and larger LW. Callisia ornata 4x has a smaller ST and a larger LW than C. ornata 2x.

Vegetative and reproductive data set

The CANDISC analysis for the vegetative and reproductive data set was based on 160 measurements for nine traits (H, LL, LW, ST, SL, SW, PL, PW, and FL) in the seven putative taxa. The squared Mahalanobis distances and Bonferroni adjusted p- values between putative taxon means are provided in Table 3-6. Based on the vegetative and reproductive traits, there is a significant difference between all putative taxa with p < 0.001. The squared Mahalanobis distance explains the distance between putative taxa, and the smallest distance is between C. ornata 4x and 6x. In Table 3-7, the mean values of the nine vegetative and reproductive traits for the putative taxa are shown. The canonical structure (Table 3-5) shows that the first canonical component

(Can 1) is strongly correlated with height (H) (0.958), followed by stem thickness (ST), and the second canonical component (Can 2) is strongly correlated with leaf width (LW)

(0.844). The third canonical component (Can 3) is strongly correlated with sepal width

(SW) (0.839). The other variable contributing to the canonical structure is leaf length

(LL) Figure 3-2 depicts the canonical discriminant ordination of Callisia section

Cuthbertia based on vegetative and reproductive traits. Can 1 mainly represents H, Can

2 mainly represents LW, and Can 3 mainly represents SW. The two axes in A, C, E, and

G explain 45% and 24% of the variation, while in B, D, F, and H, they explain 45% and

22%. The other variables contributing to the canonical structure are leaf length (LL), stem thickness (ST), sepal length (SL), petal length (PL), petal width (PW), and filament length (FL). Figure 3-2A clearly separates diploid C. graminea, C. ornata, and C. rosea into 3 different groups, explaining that C. ornata 2x is taller than C. rosea and that C. rosea is taller than C. graminea 2x. Canonical variable 2 depicts the same pattern as in

Figure 3-1A, where C. ornata 2x has the smallest LW and C. rosea has the broadest

LW, while C. graminea 2x is intermediate. In Figure 3-2B, SW separates C. graminea 2x from C. ornata 2x and C. rosea by having the smallest sepal width. The SW of C. rosea is larger than that of C. ornata 2x. Figure 3-2C depicts that C. rosea is taller than C. graminea 2x and 4x. The measurements of C. graminea 2x and 4x are overlapping but clearly show that C. graminea 2x is taller than C. graminea 4x. Canonical variable 2 explains that the LW of C. rosea is larger than C. graminea 2x and 4x. Figure 3-2D has the same relationship in H, but shows that C. graminea 2x and 4x can be separated by

SW. Figure 3-2E depicts that H is important for separating C. ornata 2x from C. ornata

6x and C. graminea 4x and 6x. Callisia ornata 2x is taller than C. graminea 4x and 6x, but it is difficult to see a clear separation between C. graminea 4x and 6x. This is also the case for Can 2. In Figure 3-2F, a strong separation is shown between C. graminea

4x and 6x based on SW. This explains that C. graminea 4x has a smaller SW than C. graminea 6x and that C. ornata 6x has a larger SW then C. ornata 2x. Figure 3-2G

shows that C. ornata 2x is taller than C. ornata 4x and C. graminea 4x, while C. ornata

4x is intermediate, and Can 2 depicts the same pattern as Figure 3-1C. Figure 3-2H explains that SW is not the most effective trait for distinguishing C. ornata 2x, C. ornata

4x, and C. graminea 4x.

In Figure 3-3, the mean scores of all putative taxa are plotted. This gives an overview of the key traits that separate all putative taxa from one another. Figure 3-3A shows that most putative taxa can be separated by stem thickness, but that C. graminea 6x and C. ornata 6x cannot be clearly differentiated. Separating C. graminea

2x and 4x by LW might be challenging due to similarities. Figure 3-3B clearly shows that

C. ornata 2x has the tallest height and the smallest leaf width and that C. rosea has the broadest LW. Callisia ornata 4x and 6x are taller than C. graminea 2x, 4x, and 6x.

Figure 3-3C depicts that SW separates C. graminea 2x, 4x, and 6x with C. graminea 6x having the broadest SW and C. graminea 2x the narrowest.

Molecular Phylogenetic Analysis

The observed sequence divergence of chloroplast and nuclear sequences is shown in Table 3-8, and the high number of parsimony-informative characters is due to the addition of 20 outgroup taxa. Nucleotide polymorphisms were observed in directly sequenced ITS products from diploid and polyploids. The nuclear and chloroplast data for diploid and polyploid taxa clearly show that Callisia section Cuthbertia is monophyletic as reported by Bergamo (2003) and Hertweck and Pires (2014). The ILD tests showed a significant difference between nuclear and combined chloroplast sequences for the diploid taxa (p = 0.03), but showed no significant difference when polyploids were added to the analysis (p = 0.35). Although significant heterogeneity was detected by the ILD test, previous studies (Barker and Lutzoni 2002, Yoder et al. 2001)

have shown that the ILD test can be overly sensitive to the numbers of parsimony- informative characters and the number of sites investigated (Darlu and Lecointre 2002).

Therefore, the chloroplast and nuclear data were combined in both the diploid and combined diploid and polyploid data sets. In all MP and ML analyses, the topologies were basically identical with the exception of reduced clade support or low resolution among clades resulting from the MP analyses. Species of Tripogandra, Gibasis, Callisia sections Brachyphylla, Callisia, and Leptocallisia, as well as Tradescantia ohiensis and

Commelina erecta, were included in the diploid data set as outgroups, but chromosome counts for these individuals are unknown. Only 50% of these taxa had usable ITS sequences data due to unreadable electropherograms with polymorphisms and bad sequence data, which could be due to hybridization, polyploidy, or mismatched primer sites.

Phylogenetic analysis (diploid taxa)

ITS diploid tree. The number of parsimony-informative characters, lengths of aligned matrices and other tree statistics for all data sets are shown in Table 3-8. The diploid ITS data set with 47 taxa was easily alignable with a length of 721 bp. The ITS matrix had 162 parsimony-informative characters, and 29 identical or nearly identical sequences of all diploid species were removed (17 Callisia graminea, 6 C. ornata, and 6

C. rosea). The terminals of the tree with asterisks (*) denote groups of identical sequences (see Table 3-9).

The ITS phylogeny of the diploid species with 18 terminals is shown in Figure 3-

4, and the monophyly of each of the three diploid species is highly supported (MP-BS

100%; ML-BS 99%). The first tree (not shown) that was obtained by these analyses showed that three C. graminea samples (IEM 250_2, IEM 280, and IEM 297) made C.

graminea 2x polyphyletic. However, these three samples were determined to be tetraploids growing in sympatry with diploid C. graminea (Chapter 2). After removing these samples, C. graminea 2x was reconstructed as monophyletic. The ITS tree topology reveals that C. ornata is sister to C. graminea and C. rosea with support of

MP-BS 90% and ML-BS 79% for the C. rosea and C. graminea clade. As previously reported by Hertweck and Pires (2014), Callisia section Brachypylla (C. navicularis

(Ortgies) D. R. Hunt and C. hintoniorum B. L.Turner) is the sister clade to Callisia section Cuthbertia with support of MP-BS 96% and ML-BS 64%. Sections Leptocallisia

(C. monandra (Sw.) Schult. & Schult. f. and C. multiflora (M. Martens & Galeotti

Standl.); MP-BS 100%; ML-BS 100%) and Callisia (C. soconuscensis Matuda and C. gentlei var. macdougalli (Miranda) D. R. Hunt; MP-BS 86%; ML-BS 85%) are also highly supported. In this tree, Gibasis venustula (Kunth) D. R. Hunt is embedded in Callisia, making the genus paraphyletic.

Chloroplast diploid tree. The chloroplast data set for the diploid and outgroup species, based on seven regions (trnL-F, rpL32, ycf1, trnH-K, trnQ, trnS and trnG), had

65 samples and 9,987 bp including gaps, with 1,220 parsimony-informative characters.

Of the seven chloroplast regions analyzed, rpL32 is the most variable region, with a length of 1,365 bp and 347 parsimony-informative characters, which is approximately

25% of its length. The 3’ ycf1 region is the longest with 2,864 bp and has 602 parsimony-informative characters (21% of its length). The trnL-F(e-f) region is the shortest and has the fewest parsimony-informative characters because the only two outgroup species (Tradescantia ohiensis and Callisia soconuscensis) were sequenced.

As in the ITS data set, identical sequences were removed (see Table 3-10) from the

matrix before analysis (21 Callisia graminea, one C. ornata, and two C. rosea). The terminals of the tree with asterisks (*) denote groups of identical sequences.

The chloroplast phylogeny for the diploid species, with 41 terminals, is shown in

Figure 3-5; just as in the ITS data set, the Callisia section Cuthbertia clade is monophyletic with high support (MP-BS 100%; ML-BS 100%). The chloroplast tree has a different topology than the ITS tree. Callisia rosea in the chloroplast tree is sister to C. ornata and C. graminea with support of MP-BS 81% and ML-BS 80%. The sister to

Callisia section Cuthbertia is Callisia section Brachypylla (C. mirantha (Torr.) D. R.

Hunt, C. navicularis and C. hintoniorum) with strong support (100% with both MP-BS and ML-BS). Within section Brachypylla, C. micrantha is sister to C. navicularis and C. hintoniorum. Section Leptocallisia is highly supported, as in the ITS tree. Tripogandra is embedded within Callisia, and Gibasis is outside the Callisia clade. Gibasis does not form a clade, which concurs with results of Hertweck and Pires (2014)

Combined ITS and chloroplast diploid tree. The combined ITS and chloroplast data sets resulted in a matrix with 53 samples and a total length of 10,711 bp. Four identical sequences were removed (see Table 3-11) from the matrix before analysis (three Callisia graminea and one C. rosea). The combined diploid ITS and chloroplast tree (Figure 3-6) also indicates that Callisia section Cuthbertia is monophyletic. The relationships of the three species favors the topology of the ITS tree where C. ornata is sister to C. rosea and C. graminea. It should be noted that the relationship of C. graminea 2x and C. rosea has a lower BS support (MP-BS 56%; ML-

BS 79%) compared to the ITS tree with MP-BS 90% and ML-BS 79%. The topology of

the other Callisia sections, including Gibasis and Tripogandra, shows no difference compared to the chloroplast tree.

Phylogenetic analysis (diploid and polyploid taxa)

ITS diploid and polyploid tree. The ITS matrix with all 35 samples has a total length of 617 characters and 51 parsimony-informative characters. As in all of the trees of diploid taxa, all duplicate sequences were removed (Table 3-9). Of the 69 duplicates that were removed, some duplicates were intercytotypic samples (C. graminea 4x (IEM

233) with C. ornata 6x (IEM 346), C. graminea 6x (IEM 235) with C. graminea 4x (IEM

268), and C. ornata 2x (IEM 312) with C. ornata 4x (IEM 228)). The ITS tree is shown in

Figure 3-7. To enhance the visualization of the tree topology and the various cytotypes, the branches of all the trees have been color coded, with the colors consistently used throughout the figures. Many branches of the polyploids have low bootstrap support

(both MP and ML BS< 50%), but the placements of certain polyploids are highly supported in the tips of the trees. These results can give implications of their parentage.

The ITS tree is consistent with the topology of the ITS for diploids only in that

Callisia section Cuthbertia is monophyletic (MP-BS 100%; ML-BS 100%) and C. ornata

2x is sister to C. rosea and C. graminea 2x. The C. ornata 2x clade is supported by MP-

BS 60% and ML-BS 62%, and C. ornata 4x is sister to C. ornata 2x. The sister relationship between C. rosea and C. graminea is supported by MP-BS 80% and ML-BS

83%, but the polyploid clade has no support from the MP analysis but 87% support from the ML analysis. In the tetraploid C. graminea clade, both C. ornata 6x and C. graminea

6x are sister taxa to C. graminea 4x (<50% both MP and ML BS). Four terminals in the tetraploid C. graminea clade are tetraploid C. ornata with low support, which might suggest relatedness among tetraploid C. ornata and C. graminea.

Chloroplast diploid and polyploid tree. The combined chloroplast matrix with all cytotypes comprised 70 samples and is 9,088 bp with 260 parsimony-informative characters. Removed duplicate sequences are shown in Table 3-10. As with the ITS matrix, 69 duplicates were removed, with some intercytotypic samples (C. graminea 2x

(IEM 286) with C. ornata 6x (IEM 343), C. graminea 4x (IEM 348) with C. graminea 6x

(IEM 235, IEM 349) and C. ornata 4x (IEM 351, IEM 352). The chloroplast tree (Figure

3-8) shows that Callisia section Cuthbertia is monophyletic. The backbone of the tree has very low support (<50% both MP and ML BS) due to the placements of the polyploids. The overall topology shows that C. rosea is sister to C. graminea 2x and C. ornata 2x, which concurs with the diploid chloroplast tree. The C. rosea clade is sister to

C. graminea 4x clade 1 and within that clade is C. graminea 6x, the only hexaploid accession (IEM 327) from South Carolina in this study. The C. ornata 2x clade 1 includes C. graminea 4x (MP-BS< 50%; ML-BS 57%) and 6x with low support (<50% both MP and ML BS). The C. ornata 4x clade clusters with C. ornata 6x with low support. In the C. ornata 2x clade 2, two accessions of C. ornata 4x are placed with high

BS support (MP-BS 92%; ML-BS 91%). Within the C. graminea 2x clade, there is strong support for the placement of some samples of C. graminea 4x (MP-BS 92%; ML-BS

95%). In the C. graminea 4x clade 4, C. graminea 6x (IEM 236) and C. ornata 6x (IEM

346) from Florida are placed as sister taxa (MP-BS 65%; ML-BS 72%), and C. ornata 4x is placed in this clade 4 as well with low support (<50% both MP and ML BS).

Combined ITS and chloroplast diploid and polyploid tree. The combined ITS and chloroplast matrix for all taxa and cytotypes comprises 100 samples and has a total length of 9,709 bp and 329 parsimony-informative characters. Removed duplicate

sequences are shown in Table 3-11. Forty-two duplicates were removed, with some intercytotypic samples (C. graminea 4x (IEM 309) with C. graminea 6x (IEM 235, IEM

349) and C. ornata 4x (IEM 298) with C. ornata (IEM 356)). The combined ITS and chloroplast tree is shown in Figure 3-9. In this tree, the monophyly of Callisia section

Cuthbertia is confirmed. The tree comprises two main clades, one with mostly diploid taxa and the second with only polyploid cytotypes. The ITS tree with all diploid taxa for

Callisia section Cuthbertia showed that C. ornata is sister to C. graminea and C. rosea.

In the total evidence tree, the relationships are shifted, and C. rosea is sister to C. ornata and C. graminea. The backbone of this clade has very low support (<50% both

MP and ML BS), but the support values for individual clades are much higher. The C. rosea clade has high support (MP-BS 99%; ML-BS 100%) and does not contain any polyploid samples. The C. ornata 2x clade is supported by MP-BS 66% and ML-BS 88% and includes samples of C. ornata 4x, suggesting parentage of C. ornata 4x. The C. graminea 2x clade is supported by MP-BS 90% and ML-BS 94% and has two samples of C. graminea 4x embedded in it. These placements suggest parentage of C. graminea

4x. The rare C. graminea 6x from South Carolina is placed as sister to the C. graminea

2x clade with very low support (<50% both MP and ML BS).

The polyploid clade, which mainly consists of C. graminea 4x, has C. graminea

6x as sister but with very low support (<50% both MP and ML BS). Callisia ornata 6x is embedded in the C. graminea clade 2 with a strong relationship between C. graminea

4x (IEM 363) and C. ornata 6x (IEM 364). The C. ornata 4x clade is sister to C. graminea 4x clade 3 with low support both (<50% both MP and ML BS), and C. ornata

4x (IEM 300) is embedded in C. graminea 4x clade 4. Callisia ornata 6x (IEM 346) and

C. graminea 6x (IEM 236) were placed in the C. graminea clade 4 as sister taxa with support of MP-BS 63% and ML-BS 66%. This relationship was also shown in the chloroplast tree with slightly higher BS support.

General Greenhouse Observations

In the period 2012-2016, 300 plants were collected and grown in the Biology

Department, University of Florida greenhouse. All plants were grown under uniform conditions except in winter. From December to March, C. rosea, C. graminea 2x, and C. graminea 4x (from GA, SC, NC, and VA) were incubated in a temperature-controlled walk-in growth chamber at 4°C. During the incubation period, plants were watered about twice a month. Plants collected in Florida (C. ornata 2x, 4x, 6x) and C. graminea 4x, 6x) were never treated during their winter dormancy.

When treated plants were taken out of the dormancy treatment in March, C. graminea started flowering in April to May and continued flowering until late August.

Due to the larger size of C. rosea, plants began flowering later, usually in May, and flowered until October. Callisia ornata did not go dormant, but began flowering in March and continued flowering until late August. Based on the examined herbarium voucher specimens and collected plants, C. graminea has a flowering span in nature from March to November with a peak in June, C. ornata has a flowering period from February until

November with a peak in May, and C. rosea has a flowering period of April to October with a peak in May.

Each year the C. ornata plants were infested with mealybugs (Pseudococcidae).

Other members of Callisia section Cuthbertia were not affected by the mealybugs, even when kept in the same room. Due to these infestations, C. ornata plants mostly produced deformed flowers in greenhouse conditions.

At the end of the flowering period (September to October), the inflorescences of the taxa (all cytotypes) in Callisia section Cuthbertia remained intact and did not wither.

Within the bracteoles attached to the peduncle, two plantlets are borne on both sides of the inflorescence (Figure 3-10). Vivipary is known to occur in Callisia, but it is not common (Hunt 1994). When the plantlets are in development, they are basically in the air, but while they develop, they get heavier, and the peduncle bends down and looks like a stolon that gives the plantlet the opportunity to root in the ground while still attached to the mother plant. This method of asexual reproduction occurs in all taxa of

Callisia section Cuthbertia in the greenhouse and in their natural habitat.

Discussion

Morphological Analysis

Vegetative data set

The vegetative dataset, based on four traits and seven putative taxa provided two traits (stem thickness (ST) and leaf width (LW)) that represent 70% and 28% of the total variation, respectively and the remainder are height (H) and leaf length (LL), with significant difference among most cytotypes. The Bonferroni-adjusted p-values showed no significant difference between C. ornata 6x and C. graminea 6x. These results might be due to the inclusion of only vegetative traits in this analysis. The discriminant variables show that both traits (ST and LW) separate the diploid species into three clusters in which C. rosea is the most distant. These results reflect the appearance of the three diploid species in their natural habitat were C. rosea is distinct in LW in combination with larger ST. Callisia graminea 2x has broader LW, but smaller ST than

C. ornata 2x.

In Figure 3-1B the mean centroid value of C. rosea is distant from the mean values of C. graminea 2x and 4x, indicating their distant relationship and it is clear that

C. graminea 4x has a larger ST than C. graminea 2x. Canonical variable 1 (ST) has the most discriminant power in this graph when the centroid mean values of C. graminea 2x and 4x are compared, but the points depicted per putative taxa overlap substantially.

This overlap suggests an autopolyploid origin of C. graminea 4x from C. graminea 2x, as suggested by Giles (1942) on the basis of morphological similarity, as autotetraploids closely resemble their diploid progenitors (Grant 1981, Soltis and Rieseberg 1986). In addition, even though the mean values of C. rosea and C. graminea 4x are significantly different, there is a fair amount of overlap between predictive ovals. Some data points of

C. graminea 4x are positioned within the predictive oval of C. rosea, which suggests that some C. graminea 4x individuals have a close morphological resemblance to C. rosea with broader leaves, thicker green stems, and a green vertical line on the sheath.

Other individuals of graminea 4x are more similar to C. graminea 2x, with prominent purple stems and vertical purple stripes on the sheath, and some are clearly intermediate between C. rosea and C. graminea 4x. In addition, the plants are tall, which is uncommon in tetraploid C. graminea. All of these plants that include traits from both C. rosea and C. graminea 4x were placed in C. graminea 4x, but these results suggest that hybridization may have occurred between C. rosea and C. graminea 4x, or perhaps that some C. graminea 4x may instead be an allopolyploid with C. rosea as the second parent, which is in agreement with the chloroplast data (Figure 3-8).

The plotted centroids of C. graminea 4x and C. ornata 2x are distant in Figure 3-

1C but with an overlap in data points. Hexaploid C. graminea and C. ornata 6x are

placed close together. Even though the cloud of data points does not distinguish a pattern, it is clear that the centroids of hexaploid C. graminea and hexaploid C. ornata are intermediate between C. ornata 2x and C. graminea 4x. These results suggest that

C. graminea 4x and C. ornata 2x might be the progenitors of both hexaploid C. graminea and C. ornata, with their intermediate morphology consistent with allopolyploidy (Soltis et al. 2014, Stebbins 1950). The ordination among C. ornata 2x, 4x and C. graminea 4x is similar to the graph in Figure 3-1C. In this case, C. ornata 2x and

4x overlap, suggesting that C. ornata 4x is autopolyploid with C. ornata 2x as the progenitor.

Vegetative and reproductive data set

The CANDISC ordination for Figure 3-2 distinguishes the taxa in Callisia section

Cuthbertia based on the canonical variables H (45%), LW (24%), and SW (22%). The remainder variables that contributed to Figure 3-2 are leaf length (LL), stem thickness

(ST), sepal length (SL), petal length (PL), petal width (PW), and filament length (FL).

The results from Figure 3-2A concur with the ordination of Figure 3-1A that the diploid taxa are distinct in morphological traits. In Figure 3-2A, C. ornata 2x is the tallest, and C. graminea 2x is the shortest. Callisia rosea is clearly separated from the other two species by its high LW. In Figure 3-2B, SW separates C. graminea 2x from C. ornata 2x and C. rosea. Figure 3-2C, which depicts the ordination of C. graminea 2x, 4x and C. rosea, illustrates that C. rosea is taller than C. graminea 2x and 4x, but that C. graminea

2x is taller than 4x. The latter does not reflect the true morphology of plants in nature because the actual mean values show that C. graminea 2x has smaller mean value

(203.80 mm) compared to C. graminea 4x (259.60 mm). The ordination results are influenced by the low sample size of H in the sampled plants (one H/plant, five H/five

plants, not 25), which does not represent the whole range of H measurements per cytotype. The contribution of the floral trait SW shows in Figure 3-2D that C. graminea

4x has larger SW than C. graminea 2x. Even though squared Mahalanobis distance

(Table 3-6) are significantly different (p<0.001) among H and SW, Can 1 and Can 2 poorly separate C. graminea 4x, 6x and C. ornata 2x. However, diploid C. ornata is clearly distinguishable from the polyploid entities. Can 3 separates the three polyploids, but the overlap in data points in Figure 3-2E-F clearly suggests that C. graminea 4x might be a progenitor of the hexaploid entities. Figure 3-2G and H clearly show intermediacy among the C. ornata 2x, 4x and C. graminea 4x. Can 3 does not show any discriminant power to separate C. graminea 4x, C. ornata 2x, 4x (Figure 3-2H). The vegetative data sets (Figure 3-1D) show a strong overlap between C. ornata 2x and 4x, suggesting autopolyploidy, but when floral traits are added (Figure 3-2G), the results instead suggest allopolyploidy, as discussed in Chapter 2. The observed DNA content was similar to that expected for an allopolyploid formed via an unreduced gamete of C. ornata 2x and a ‘normal’ gamete of C. graminea 4x.

Plotting all means of the putative taxa in Callisia section Cuthbertia (Figure 3-3) gives an overview of the squared Mahalanobis distance among diploid and polyploids entities. In all ordination plots, C. rosea is distant from the remaining putative taxa. Only when individual data points are plotted does C. graminea 4x overlap with C. rosea. The remaining taxa have interspecific and interploidal variation as discussed in detail above.

Based on the vegetative CANDISC analysis, stem thickness is the most distinctive character to identify taxa in section Cuthbertia excluding hexaploid C. graminea and C. ornata (Figure 3-11). Stem thickness increases with ploidal level in C. graminea (2x, 4x,

6x), hexaploid C. graminea and C. ornata are similar, and C. ornata 2x and 4x differ slightly. Callisia rosea has by far the largest ST, making it the most distinctive morphologically from all taxa in Callisia section Cuthbertia. These results concur with the CANDISC analysis.

Molecular Phylogenetic Analysis

Phylogenetic analysis (diploid taxa)

ITS diploid tree. The relationship among the diploid taxa generated from ITS agrees with the chloroplast tree from Hertweck and Pires (2014).based on trnL-F and rpL16. There is high BS support (MP-BS 100%; ML-BS 99%) for section Cuthbertia, and

C. ornata 2x is sister to C. rosea and C. graminea 2x. Sections Brachyphylla,

Leptocallisia, and Callisia form clades that mostly agree with Bergamo (2003) and

Hertweck and Pires (2014). There are two placements that do not concur with previous molecular work: C. gentlei var. elegans sister to C. graminea 2x (not shown) and

Gibasis venustula in Callisia. Callisia gentlei var. elegans was removed from the analysis because it made Callisia section Cuthbertia paraphyletic. This could be due to long branch attraction (LBA); a method to solve this problem is to remove one of the taxa that is causing this LBA (Siddall and Whiting 1999) or improve taxon sampling

(Soltis and Soltis 2004). For the combined ITS and chloroplast analysis, C. gentlei var. elegans was included in the analysis since the number of outgroup taxa was much higher; the topology concurs with previous studies. Previous molecular work has placed

Gibasis as sister to Callisia based on chloroplast and 5S NTS phylogenies (Burns et al.

2011). Due to low representation of the genus, homoplasy might be the possible reason for this misplacement. When combined with chloroplast data, the placement agrees with previous work (Bergamo 2003, Burns et al. 2011, Hertweck and Pires 2014).

Chloroplast diploid tree. The chloroplast phylogeny supports previous work that section Cuthbertia is monophyletic; however, the sister relationships among the three diploid species, C. graminea, C. ornata and C. rosea, disagree with the chloroplast phylogeny of Hertweck and Pires (2014). In this chloroplast phylogeny, C. rosea is sister to C. graminea 2x and C. ornata 2x; Hertweck and Pires (2014) showed that C. ornata is sister to C. graminea and C. rosea. The results of Hertweck and Pires

(2014) agree with the ITS tree in this study, however ploidy from their taxa was not investigated. The BS support for the relationship between C. ornata 2x and C. graminea

2x is MP-BS 81%; ML-BS 80%, (Figure 3.5) which is approximately equal to the BS support of the relationship of C. rosea and C. graminea 2x in the ITS tree of diploid species (Figure 3.4).

Combined ITS and chloroplast diploid tree. In the combined ITS and chloroplast phylogeny for the diploids, C. ornata 2x is sister to C. rosea and C. graminea, in contrast to my chloroplast tree and Hertweck and Pires’s (2014) chloroplast tree. Both hypotheses are possible if compared to Chapter 2 and the morphological analysis previously discussed. Callisia ornata 2x is endemic to Florida, has a few erect stems, is morphologically less grass-like than the other species, flowers almost year-round, and occurs in hotter environments than C. rosea; C. rosea and C. graminea 2x overlap in distribution and share a grass-like appearance. In contrast to C. ornata 2x, C. rosea occurs mainly in the piedmont, is most distant from all other taxa in the section Cuthbertia clade based on the CANDISC analyses, and no polyploids have been recorded for this species; polyploidy has only been reported for C. ornata and C. graminea (Chapter 2). Based on the placement of C. graminea 2x in the diploid trees, it

seems that C. graminea 2x is the taxon that moves around. In the ITS tree, C. graminea

2x is sister to C. rosea; however; in the chloroplast tree, C. graminea 2x is sister to C. ornata 2x. One possible explanation for the instability of the placement of C. graminea

2x is that it is a hybrid, formed from an ancient hybridization event of C. ornata 2x and

C. rosea. Because C. graminea is a diploid, it would be a homoploid hybrid, a rare form of hybrid (Schumer et al. 2014). In contrast, Feliner et al. (2017) noted numerous examples of homoploid hybrids in which further studies are needed. Another explanation for the incongruence in this case could be introgression of the chloroplast genome to yield a topology that is incongruent with the nuclear tree (Wendel and Doyle

1998). It is unknown whether or not diploid species of Callisia section Cuthbertia are compatible, and this issue requires further investigation. The placement of sections

Brachyphylla, Leptocallisia, and Callisia (including Gibasis and Tripogandra) in the chloroplast phylogeny and combined nuclear and chloroplast phylogeny (Figure 3-5 and

3-6) are congruent and complement the topology of Hertweck and Pires (2014).

Phylogenetic analysis (diploid and polyploid taxa)

All phylogenetic trees based on diploid and polyploid taxa support the monophyly of Callisia section Cuthbertia. These concur also with the phylogenies of nuclear and chloroplast data sets for diploid species and with previously reported phylogenies

(Bergamo 2003, Hertweck and Pires 2014). The backbone of putative taxa trees is poorly supported (<50%) due to the inclusion of polyploids. The relationships among the three diploid species agree with the separate nuclear and chloroplast trees, but not with the total evidence phylogeny. The total evidence diploid tree supported the placement of

C. ornata 2x as sister to C. rosea and C. graminea 2x, yet the total evidence phylogeny with polyploids added places C. rosea as sister to C. ornata 2x and C. graminea 2x.

This incongruence with the diploid total evidence phylogeny is poorly supported (<50%).

Within the polyploid phylogenies, the clades within chapter 3 were named to allow a better understanding on the clades being discussed and are not based on previous literature.

ITS diploid and polyploid tree. In the nuclear phylogeny of Callisia section

Cuthbertia (all putative taxa), C. graminea 2x and 4x do not form a clade, but form distinct clades in which C. rosea is sister to C. graminea 2x. These results indicate that both C. rosea and C. graminea 2x could have been parents of C. graminea 4x or both parents are C. graminea 2x. Callisia graminea 6x is sister to the C. graminea 4x clade, suggesting that C. graminea 4x is one of its progenitors. Within the C. graminea 4x clade, there are four accessions of C. ornata 4x; however, C. ornata 4x is also placed within the C. ornata 2x clade. These placements suggest multiple origins in which C. ornata 2x gives rise to C. ornata 4x, and C. graminea 4x crossed with C. ornata 2x to give rise to other accessions. Callisia ornata 6x is sister to the C. graminea 4x clade, which also suggests that C. graminea 4x may be a parent of C. ornata 6x.

Chloroplast diploid and polyploid tree. The chloroplast tree (Figure 3-8) differs from the ITS tree. Callisia graminea 4x occurs in four clades, suggesting either multiple origins from distinct diploid progenitors or hybridization and chloroplast capture after polyploid formation (or possibly both). The first C. graminea 4x clade is sister to C. rosea. This is an interesting relationship, because in the CANDISC analysis hybridization or a parent-progeny relationship was suggested between C. rosea and C. graminea 4x. In Chapter 2, I reported that approximately 26 accessions of C. graminea

4x had a smaller genome than the majority of C. graminea 4x entities. Approximately

65% of the accessions with a smaller genome are in the C. graminea 4x clade 1. This is another line of evidence that C. rosea could be one parent of C. graminea 4x and that the genome size is smaller than the average C. graminea 4x entities. If hybridization occurred between C. graminea 4x and C. rosea, assuming that C. rosea is a 2N pollen donor and interbreeding with a tetraploid (Chase 1963, Dunford 1970, Mason and Pires

2015), then the expected genome size would be higher than that typical of C. graminea

4x and not lower; however, Leitch and Bennett (2004) stated that genome reduction is common in polyploids. In this scenario, C. graminea 4x would not appear as the sister to

C. rosea in the chloroplast tree because chloroplast is maternally inherited.

Giles (1943) suggested that, due to the similarity of geological composition in the sandhills along the Fall Line, in the past, C. graminea 2x could have had a wider distribution stretching into South Carolina. In that case, C. rosea could have hybridized with C. graminea 2x, resulting in an allotetraploid entity. Artificial interploidal crosses between C. graminea 2x and 4x produce fertile seeds (Kelly 1991), which means that crosses with C. rosea could be possible, yet such tests have not been reported.

In C. graminea 4x clade 1, the rare South Carolina C. graminea 6x is placed with a group of three tetraploid C. graminea entities from South Carolina, implying that C. graminea 4x is one of the parents of the rare hexaploid C. graminea. Giles (1942) reported, based on multivalent pairing, that hexaploid C. graminea are autopolyploids, which is likely with C. graminea 2x and 4x as parents. In Kelly (1991), crosses of diploid and tetraploid C. graminea, yielded high number of viable seeds that came from diploid maternal plants with tetraploid pollen plants. The question arises as to what ploidal levels were these progeny? Unfortunately, these seedlings were not tested for ploidy; a

follow-up study on compatibility among all members of Callisia section Cuthbertia and distinguishing ploidy of the offspring is recommended to answer this question.

The C. graminea 4x clade 4 also includes C. ornata 6x, C. graminea 6x, and C. ornata 4x, which agrees with the placements in the nuclear tree. Callisia graminea 4x is most likely one of the parents of hexaploid C. graminea and C. ornata. With BS support of MP-BS 65% and ML-BS 72%, it was clear that the hexaploid entities are related by having a common parent. The most likely other parent is C. ornata 2x as evidence of sympatry between C. ornata 2x and C. graminea 4x (Chapter 2) but also morphology as was stated earlier; however, further investigation is needed to identify the second parent. Because C. ornata 4x occurs in two clades (C. ornata 2x clade 2 and C. graminea 4x clade 4), it seems that C. ornata 4x may have multiple origins: one origin with C. ornata 2x as both of its parents, suggesting autopolyploidy, and the other with C. ornata 2x and C. graminea 4x as possible parents, suggesting allopolyploidy.

Alternatively, there may have been a single origin followed by introgression to yield the second placement. Callisia graminea 4x is placed with C. graminea 2x, which supports an autopolyploid origin, as suggested by Giles (1942). Callisia graminea 4x is also placed in the C. ornata 2x clade 1 with C. graminea 6x. The placement of the hexaploid indicates that C. graminea 4x is a likely progenitor of C. graminea 6x, and the placement of C. graminea 4x and C. ornata 2x with C. graminea 6x suggests an allopolyploid origin of the hexaploid. Callisia ornata 6x is placed with C. ornata 4x, which suggests that C. ornata 4x is the maternal, and possibly the only parent. However one

C. ornata 6x cytotype is placed with C. graminea 6x, suggesting allopolyploidy for both hexaploids. The C. ornata 2x clade 2 shows the second origin of C. ornata 4x that

agrees with the ITS tree and that suggests that C. ornata 2 is the maternal parent of C. ornata 4x.

Combined ITS and chloroplast diploid and polyploid tree. Despite the incongruences among diploid entities, the placements of the polyploids in the total evidence tree (Figure 3-9) is congruent with the chloroplast tree. The results clearly show that C. graminea 4x is an autopolyploid with C. graminea 2x as its progenitor.

Callisia graminea 6x (Florida) is an allopolyploid in which C. graminea 4x is the maternal parent and C. ornata 2x is the paternal parent. Callisia graminea 6x (South Carolina) is an autopolyploid with tetraploid C. graminea as parent. Callisia ornata 4x has multiple origins in which it is an autopolyploid with C. ornata 2x as progenitor and an allopolyploid with C. graminea 4x as its maternal parent and C. ornata 2x as its paternal parent. Because C. graminea 6x and C. ornata 6x (Florida) form a clade with BS support (MP-BS 65%; ML-BS 72%) which is sister to C. graminea 4x in the chloroplast and total evidence tree, this suggests that they are allopolyploids and in both cases C. graminea 4x is the maternal parent and C. ornata 2x is the paternal parent.

Some of these incongruences between chloroplast and ITS phylogenies above, may arise from concerted evolution of ITS (Wendel 2000) a mechanism of homogenizing to a single ITS sequence. Concerted evolution after hybridization or polyploidization can mislead estimates of organismal phylogenetic relationships in lower taxonomic levels (Soltis and Soltis 1998). The electropherograms of the polyploids shows some polymorphism, indicating that more than one genome were present. This indicates that concerted evolution might be one of the reason of incongruence between the ITS and chloroplast tree in Callisia section Cuthbertia.

Species delimitation at the diploid level within the Callisia section Cuthbertia supports three monophyletic enities (C. graminea, C. ornata, and C. rosea), that fit the phylogenetic and evolutionary species concept (Baum and Donoghue 1995, Mishler and

Brandon 1987, Wiley and Mayden 2000) and that I consider to be cladospecies based on the molecular data. They are phenetically distinctive based on the morphological data and therefore reflect the diagnostic, morphological/phenetic, and apomorphic species concepts (de Queiroz 2007, de Queiroz and Good 1997, Judd et al. 2016).

Species delimitation at the tetraploid and hexaploid level is uncertain. Based on the ITS and chloroplast data, it is clear that hybridization has occurred multiple times with multiple origins of the polyploids, which poses problems in identifying monophyletic units within the tetraploids and hexaploids. Morphologically, the polyploids are not always distinct from one another, which leads to ambiguity. Due to these uncertainties, I recommend recognizing these polyploid enities simply as polyploids of the Callisia section Cuthbertia clade. A similar treatment was also applied in “Apomixis, Patterns of

Morphological Variation, and Species Concepts in subfam. Maloideae (Rosaceae)” by

Campbell and Dickinson (1990).

General Greenhouse Observations

The phenology data observed in the greenhouse concurs with the observations of Small (1933), who stated that C. ornata flowered year-round and that C. rosea and C. graminea flowered in spring and summer. Because only C. ornata was infested with mealybugs, perhaps C. ornata may have physiological properties that the other two diploid species and polyploids do not express. This could be an autapomorphy for diploid C. ornata.

Vegetative plantlets were borne in the bracts of the inflorescence. Vivipary was reported by Lakela (1972) in C. graminea f. leucantha, but the author was not sure if this form of asexual reproduction occurred in nature. Most polyploid species rely on vegetative reproduction more than their progenitors (Herben et al. 2017, Husband et al.

2013), but in this case, the diploid species also reproduce asexually. Asexual reproduction through vivipary is not common in Callisia (Hunt 1986), but it was observed in nature in late fall (personal observation). Vegetative reproduction has probably gone unnoticed, given that the plants are grass-like and inconspicuous when without flowers.

This mechanism might have been one of the reasons why this polyploid complex has been so successful.

Table 3-1. Outgroup taxa included in this study. NMNH ID Taxon Location collected Source Voucher deposited at FLAS (IEM) 1983-197 Callisia cordifolia (Sw.) Andiers. & Woodson U.S.A., Florida, Volusia Faden, R. B. 83/37 373 1993-090 Callisia gentlei var. elegans (Alexander ex H. E. U.S.A., Louisiana Cultivated Tim Chapman 372 Moore) D. R. Hunt 1998-281 Callisia fragrans (Lindl.) Woodson Mexico, Durango Hunt 85-71 379 007-003 Callisia hintoniorum B. L.Turner Mexico, Nuevo León, Hinton, G. B. 25725 366 Aramberri 1981-074 Callisia gentlei var. macdougallii (Miranda) D. R. Hunt Mexico, Oaxaca State McDougal s.n. 359 2006-046 Callisia micrantha (Torr.) D. R. Hunt U.S.A., Texas, Patterson, T. F. s.n. 367 Cameron 1993-092 Callisia monandra (Sw.) Schult. & Schult. f. Unknown Bogner J. s.n. 1992-041 Callisia multiflora (M. Martens & Galeotti) Standl. Mexico, Chiapas Spencer, M. 92.309 377 1980-395 Callisia multiflora (M. Martens & Galeotti) Standl. Mexico, Durango Faden, R. B. 76/166A 376 1980-410 Callisia navicularis (Ortgies) D. R. Hunt Mexico, Durango Fryxell, P. A. s.n. 370 1982-291 Callisia repens (Jacq.) L. Bolivia Graf, A. B. s.n. 358 1983-193 Callisia repens (Jacq.) L. Brazil Plowman, T. C. 12852 375 1986-203 Callisia soconuscensis Matuda Ecuador Munich Botanical Garden, 374 84/3362 1982-298 Gibasis geniculata (Jacq.) Rohweder unknown Munich Botanical Garden s.n. 378 2007-030 Gibasis pellucida (M. Martens & Galeotti) D. R. Hunt U.S.A., Texas, Harris Rosen, D. J. 298 357 2003-081 Gibasis venustula (Kunth) D. R. Hunt Mexico unknown 371 1986-223 Tripogandra serrulata (Vahl) Handlos French Guiana Sastre, C. H. 8121 368 1981-080 Tripogandra serrulata (Vahl) Handlos Panama Brenner, D. 10/81 369

Table 3-2. DNA regions and their primers used in this study. Region Primer Primer sequence (5’—3’) Reference ITS Y5 TAG AGG AAG GAG AAG TAA CAA Hoshi et al. (2008) Y4 CCC GCC TGA CCT GGG GTC GC trnL-F E GGT TCA AGT CCC TCT ATC CC Taberlet et al. (1991) F ATT TGA ACT GGT GAC ACG AG rpL32 rpL32-F CAG TTC CAA AA A AAC GTA CTT C Shaw et al. (2007) trnL(UAG) CTG CTT CCT AAG AGC AGC GT ycf1 F(317) AGT TCT CAA TTC TCT ACG ACG TAT Newly designed AG R(1867) GAA TGG AAA AAC TGG TTA AAG GGT F(1897) ACC CTT TAA CCA GTT TTT CCA TTC

R(3080) GGT TTT CAA TTA TCA GAA ACA GAA TTT trnH-K trnH ACG GGA ATT GAA CCC GCG CA Nicolosi et al. (2000) trnK CCG ACT AGT TCC GGG TTC GA trnQ trnQ(UUG) GCG TGG CCA AGY GGT AAG GC Shaw et al. (2007) rpS16x1: GTT GCT TTY TAC CAC ATC GTT T trnS-G trnS(GCU) AGA TAG GGA TTC GAA CCC TCG GT Shaw et al. (2005) 3’trnG(UUC) GTA GCG GGA ATC GAA CCC GCA TC trnG-R 5’trnG2G GCG GGT ATA GTT TAG TGG TAA AA Shaw et al. (2005) trnR (UCU) GAA CCT ATA CCA AAG GTT TAG Newly designed AAG ACC

Table 3-3. Squared Mahalanobis distance and adjusted Bonferroni p-value among cytotypes in Callisia section Cuthbertia based on vegetative traits. Below diagonal: Squared Mahalanobis distance, above diagonal adjusted Bonferroni p-value. CG= C. graminea, CO= C. ornata, CR= C. rosea followed by the ploidy level. Taxa CG_2x CG_4x CG_6x CO_2x CO_4x CO_6x CR_2x CG_2x < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 CG_4x 1.69 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 CG_6x 6.75 2.02 < 0.001 < 0.001 0.69 < 0.001 CO_2x 20.78 13.06 5.72 0.01 < 0.001 < 0.001 CO_4x 15.60 8.76 2.95 0.65 < 0.001 < 0.001 CO_6x 6.70 2.50 0.29 5.67 3.09 < 0.001 CR_2x 34.35 21.59 15.85 23.11 19.07 19.33

Table 3-4. Mean value (mm) of the vegetative traits used in the CANDISC analysis Taxa H LL LW ST CG_2x 232.82 128.08 1.76 1.10 CG_4x 286.09 156.65 2.87 1.52 CG_6x 341.09 189.74 3.16 2.00 CO_2x 453.95 220.70 2.59 2.32 CO_4x 419.11 200.92 2.72 2.32 CO_6x 328.27 195.83 2.66 2.00 CR_2x 424.50 206.17 7.66 3.04

Table 3-5. Canonical structure. Canonical correlations between canonical variables and the original of 1) vegetative and 2) combined vegetative and reproductive traits. Label Vegetative traits Vegetative and reproductive traits Can1 Can2 Can3 Can1 Can2 Can3 H 0.945 -0.312 -0.093 0.958 -0.108 0.229 LL 0.915 -0.356 0.166 0.757 -0.144 0.583 LW 0.825 0.565 -0.016 0.385 0.844 0.365 ST 0.997 0.006 0.068 0.772 0.294 0.504 PL -0.174 -0.547 0.814 PW 0.158 -0.685 0.571 SL -0.435 -0.357 0.825 SW -0.423 -0.203 0.839 FL -0.462 -0.041 0.409

Figure 3-1. Canonical discriminant ordination of Callisia section Cuthbertia based on vegetative traits. Can 1 represents stem thickness and Can 2 represents leaf width. The mean scores are indicated with an enlarged symbols in the center of the prediction ovals. A) ordination of C. graminea 2x, C. ornata 2x and C. rosea, B) ordination of C. graminea 2x, C. graminea 4x and C. rosea, C) ordination of C. ornata 2x, C. ornata 6x, C. graminea 4x and C. graminea 6x, D) ordination of C. ornata 2x, C. ornata 4x and C. graminea 4x.

Table 3-6. Squared Mahalanobis distance and adjusted Bonferroni p-value among cytotypes in Callisia section Cuthbertia based on vegetative and floral traits. Below diagonal: Squared Mahalanobis distance, above diagonal adjusted Bonferroni p-value.CG= C. graminea, CO= C. ornata, CR= C. rosea followed by the ploidy level. Taxa CG_2x CG_4x CG_6x CO_2x CO_4x CO_6x CR_2x CG_2x < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 CG_4x 9.48 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 CG_6x 40.70 19.31 < 0.001 < 0.001 < 0.001 < 0.001 CO_2x 40.61 39.88 48.77 < 0.001 < 0.001 < 0.001 CO_4x 11.92 7.37 23.19 18.47 < 0.001 < 0.001 CO_6x 19.98 7.35 13.68 29.92 6.14 < 0.001 CR_2x 36.32 27.35 47.23 30.65 20.69 25.73

Table 3-7. Mean value (mm) of the combined vegetative and reproductive traits used in the CANDISC analysis. Taxa PL PW SL SW FL H LL LW ST CG_2x 9.09 9.06 4.68 2.36 5.23 203.80 113.88 1.64 1.10 CG_4x 10.44 8.76 5.55 2.99 4.91 259.60 148.68 2.96 1.53 CG_6x 13.73 11.07 7.13 3.54 6.12 332.50 183.00 3.35 2.04 CO_2x 10.71 10.37 4.97 2.52 4.77 514.40 216.00 2.44 2.28 CO_4x 10.37 9.56 5.33 2.97 5.28 357.80 183.45 2.42 2.14 CO_6x 11.77 10.65 6.01 3.38 5.11 303.40 196.26 3.42 2.04 CR_2x 9.00 8.72 4.62 2.53 5.04 419.60 191.68 7.00 2.68

Figure 3-2. Canonical discriminant ordination of Callisia section Cuthbertia based on vegetative and reproductive traits. Can 1 represents height, Can 2 represents leaf width and Can 3 represent sepal width. The mean scores are indicated with an enlarged symbol in the center of the prediction ovals. A-B) ordination of C. graminea 2x, C. ornata 2x and C. rosea, C-D) ordination of C. graminea 2x, C. graminea 4x and C. rosea, E-F) ordination of C. ornata 2x, C. ornata 6x, C. graminea 4x and C. graminea 6x, G-H) ordination of C. ornata 2x, C. ornata 4x and C. graminea 4x.

Figure 3-3. Canonical discriminant ordination of Callisia section Cuthbertia depicting all cytotypes. Only the mean scores are indicated. A) Can 1 represents stem thickness and Can 2 represents leaf width. B-C) Can 1 represents height, Can 2 represents leaf width and Can 3 represent sepal width.

Table 3-8. Data matrix and parsimony tree statistics for each sequenced region. The length (bp) includes gaps introduced during alignment. CI = consistency index, RI = retention index. Region #Terminals Length # Parsimonious. # Most CI RI (bp) informative parsimonious Characters trees ITS 111 721 162 514 0.78 0.81 trnL-F 73 411 3 336 1 1 rpL32 155 1365 347 302 0.69 0.82

ycf1(317) 151 1636 340 1738 0.77 0.85

ycf1(1897) 159 1228 262 1900 0.76 0.88 trnH-K 157 1741 99 454 0.78 0.88 trnQ 96 1529 36 1922 0.99 0.95 trnS-G 111 968 70 1572 0.88 0.93 trnG-R 145 1112 63 1948 0.85 0.95 chloroplast 159 9987 1220 70 0.76 0.84 combined All 159 10711 1346 94 0.76 0.84 combined

Table 3-9. Terminal taxa denoting duplicate sequences in diploid ITS and all cytotypes ITS tree. Duplicate: g=C. graminea, o= C. ornata, r=C. rosea followed by the ploidy and accession number. Terminal taxa Duplicate C. graminea 2x_249 g_2x_250_1, g_2x_250_3, g_2x_288, g_2x_289, g_2x_293, g_2x_296, g_2x_333, g_2x_337 C. graminea 2x_283 g_2x_284, g_2x_285, g_2x_286, g_2x_287, g_2x_290, g_2x_291, g_2x_294, g_2x_292, g_2x_295 C. graminea 4x_223 g_4x_224, g_4x_230, g_4x_239, g_4x_248, g_4x_275, g_4x_271, g_4x_278, g_4x_307, g_4x_360_1, g_4x_360_2, g_4x_360_3 C. graminea 4x_233 g_4x_234, g_4x_242, g_4x_253, g_4x_255, g_4x_257, g_4x_259, g_4x_261, g_4x_262, g_4x_273, g_4x_279, g_4x_308, g_4x_309, g_4x_311, g_4x_362, o_6x_346 C. graminea 4x_241 g_4x_247 C. graminea 4x_260 g_4x_263, g_4x_264, g_4x_265, g_4x_267, g_4x_250_2, g_4x_297 C. graminea 4x_310 g_4x_266, g_4x_270 C. graminea 4x_320 o_4x_352 C. graminea 6x_349 o_6x_344 C. graminea 6x_235 g_6x_236, o_6x_347, g_4x_268 C. ornata 2x_312 o_2x_313, o_2x_314, o_2x_315, o_2x_353, o_2x_354, o_4x_228, C. rosea 2x_240 r_2x_321, r_2x_323, r_2x_238, r_2x_237, r_2x_244, r_2x_245

Figure 3-4. Nuclear diploid phylogeny of Callisia section Cuthbertia. Maximum likelihood tree based on 1000 pseudo-replicates. Bootstrap values >50% are above branches; maximum parsimony value followed by maximum likelihood value. Branches and terminals are color coded to distinguish species and ploidy levels. Putative taxa ending with an “*” denote groups of identical sequences. “*“ = 100%, “-“ = BS< 50% and no symbol indicates no support.

Table 3-10. Terminal taxa denoting duplicate sequences in diploid chloroplast and all cytotypes chloroplast tree. Duplicate: g=C. graminea, o= C. ornata, r=C. rosea followed by the ploidy and accession number. Terminal taxa Duplicate C. graminea 2x_286 g_2x_249, g_2x_250_3, g_2x_282, g_2x_283, g_2x_284, g_2x_290, g_2x_291, g_2x_294, g_2x_332, g_2x_333, g_2x_340, g_2x_342, o_6x_343 C. graminea 2x_292 g_2x_250_1, g_2x_285, g_2x_288, g_2x_289, g_2x_293, g_2x_296, g_2x_331 C. graminea 2x_337 g_2x_336, g_2x_338 C. graminea 4x_264 g_4x_363 C. graminea 4x_348 g_4x_230, g_4x_234, g_4x_239, g_4x_248, g_4x_250, g_4x_252, g_4x_253, g_4x_254, g_4x_255, g_4x_256, g_4x_257, g_4x_258, g_4x_259, g_4x_260, g_4x_266, g_4x_268, g_4x_274, g_4x_278, g_4x_297, g_4x_306, g_4x_308, g_4x_309, g_4x_310, g_4x_311, g_4x_320, g_4x_324, g_4x_326, g_4x_328, g_4x_329, g_4x_330 g_4x_334, g_4x_335, g_6x_235, g_6x_349, o_4x_351, o_4x_352 C. graminea 4x_360 g_4x_224, g_4x_262, g_4x_272, g_4x_302, g_4x_317, g_4x_223 C. ornata 2x_361 o_2x_314, o_2x_353 C. ornata 4x_298 o_2x_356 C. rosea 2x_237 r_2x_238, r_2x_321

Figure 3-5. Chloroplast diploid phylogeny of Callisia section Cuthbertia. Maximum likelihood tree based on 1000 pseudo-replicates. Bootstrap values >50% are above branches; maximum parsimony value followed by maximum likelihood value. Branches and terminals are color coded to distinguish species and ploidy levels. Putative taxa ending with an “*” denote groups of identical sequences. “*“ = 100%, “-“ = BS< 50% and no symbol indicates no support.

100

Table 3-11. Terminal taxa denoting duplicate sequences in diploid total evidence and all cytotypes total evidence tree. Duplicate: g=C. graminea, o= C. ornata, r=C. rosea followed by the ploidy and accession number. Terminal taxa Duplicate C. graminea 2x_332 g_2x_250_3 C. graminea 2x_336 g_2x_338 C. graminea 2x_340 g_2x_342 C. graminea 4x_224 g_4x_317, g_4x_223, g_4x_302 C. graminea 4x_229 g_4x_247 C. graminea 4x_253 g_4x_256, g_4x_306, g_4x_328, g_4x_329, g_4x_324, C. graminea 4x_309 g_4x_234, g_4x_239, g_4x_252, g_4x_254, g_4x_255, g_4x_256, g_4x_257, g_4x_258, g_4x_259, g_4x_260, g_4x_269, g_4x_274, g_4x_278, g_4x_308, g_4x_311, g_4x_320, g_4x_324, g_4x_326, g_4x_328, g_4x_329, g_4x_330, g_4x_334, g_4x_335, g_4x_348, g_6x_235, g_6x_349 C. graminea 4x_360 g_4x_360_2, g_4x_360_3 C. ornata 4x_298 o_2x_356 C. rosea 2x_237 r_2x_238, r_2x_321

101

Figure 3-6. Combined nuclear and chloroplast diploid phylogeny of Callisia section Cuthbertia. Maximum likelihood tree based on 1000 pseudo-replicates. Bootstrap values >50% are above branches; maximum parsimony value followed by maximum likelihood value. Branches and terminals are color coded to distinguish species and ploidy levels. Putative taxa ending with an “*” denote groups of identical sequences. “*“ = 100%, “-“ = BS< 50% and no symbol indicates no support.

102

Figure 3-7. Nuclear phylogeny of Callisia section Cuthbertia. Maximum likelihood tree based on 1000 pseudo-replicates. Bootstrap values >50% are above branches; maximum parsimony value followed by maximum likelihood value. Branches and terminals are color coded to distinguish species and ploidy levels. Putative taxa ending with an “*” denote groups of identical sequences. “*“ = 100%, “-“ = BS< 50% and no symbol indicates no support.

103

Figure 3-8. Chloroplast phylogeny of Callisia section Cuthbertia. Maximum likelihood tree based on 1000 pseudo-replicates. Bootstrap values >50% are above branches; maximum parsimony value followed by maximum likelihood value. Branches and terminals are color coded to distinguish species and ploidy levels. Putative taxa ending with an “*” denote groups of identical sequences. “*“ = 100%, “-“ = BS< 50% and no symbol indicates no support.

104

Figure 3-9. Combined nuclear and chloroplast phylogeny of Callisia section Cuthbertia. ML tree based on 1000 pseudo-replicates. Bootstrap values >50% are above branches; MP value followed by ML value. Branches and terminals are color coded to distinguish species and ploidy levels. Putative taxa ending with an “*” denote groups of identical sequences. “*“ = 100%, “-“ = BS< 50% and no symbol indicates no support. Outgroup not shown.

105

Figure 3-10. Vivipary in Callisia section Cuthbertia: A) Peduncle with developing plantlets, B) Plantlet with flower, C) Developing plantlet on flowering inflorescence, D) Plantlet ready to set root in the ground. Notice the pedicle function as a stolon. October 31, 2016. Gainesville, FL. Photo courtesy of author

106

Figure 3-11. Stem thickness comparison among putative taxa in Callisia section Cuthbertia. Solid line within the box represents the median and the dotted line represents the mean value. The whiskers represent the 5th and 95th percentiles.

107

CHAPTER 4 ECOLOGICAL NICHE MODELS AND PAST, PRESENT, AND FUTURE GEOGRAPHIC DISTRIBUTIONS OF CALLISIA SECTION CUTHBERTIA (COMMELINACEAE)

Introduction

Polyploidy, or whole-genome duplication, is a common speciation mechanism in angiosperms (Davis and Heywood 1963, De Bodt et al. 2005, Rensing et al. 2008,

Soltis et al. 2016, Stebbins 1950, Tank et al. 2015). Two forms of polyploid speciation are recognized: allopolyploidy, or genome duplication involving hybridization between species, and autopolyploidy, or genome doubling within a species (Grant 1981,

Müntzing 1936). Polyploids are reproductively isolated from their progenitors due to post-zygotic isolating mechanisms (Grant 1981), and neopolyploids are confronted with a mating disadvantage due to the existence of only a few compatible cytotypes and an abundance of incompatible cytotypes, a process referred to as minority-cytotype exclusion (MCE, Levin 1975). A neopolyploid may persevere as an entity by outcompeting and replacing its progenitors, by coexistence due to niche segregation from its progenitor, or by finding new geographical space (Fowler and Levin 1984).

A longstanding hypothesis is that polyploids will occupy different, but also larger, geographic ranges than their diploid progenitors (Hagerup 1927, Husband et al. 2013,

Stebbins 1950). However, alternative hypotheses for the ecological niche space of polyploids compared to that of their progenitor(s) span the range of possible outcomes: polyploids have been hypothesized to have wider ecological tolerances and the ability to inhabit more extreme environments than their parental species, they may have intermediate ecological tolerances to those of their parents, they may occupy a subset of the range and/or habitat of their parents, or they may occupy a novel niche and/or

108

geographic area (see Marchant et al. 2016 for review). The ability of many polyploids to occupy disturbed areas and novel habitats suggests flexibility in ecological niche and a

MCE escape advantage for neopolyploids. This hypothesis of niche segregation in polyploid complexes based on abiotic model envelopes has been tested, but support has been inconclusive, in part due to a limited number of studies (e.g. Glennon et al.

2014, Godsoe et al. 2013, Marchant et al. 2016, Spoelhof et al. 2017, Visger et al.

2016).

Projecting environmental niches constructed for extant species onto climate maps for the Last Glacial Maximum (LGM) and mid-Holocene (MH) can elucidate past patterns of niche overlap of populations to test for niche conservatism (Peterson et al.

2008). In addition, paleo-environmental model envelopes can also be used as an alternative to identify possible niche overlap and hybridization zones among potential diploid progenitors of polyploids (Folk et al. 2017, López-Alvarez et al. 2015).

Furthermore, due to global climate change, projecting future niche suitability has become increasingly important as this approach can contribute to an understanding of possible niche shifts or niche conservatism over time and can consequently be used in conservation efforts (Guisan et al. 2014, Guisan et al. 2013).

Callisia section Cuthbertia comprises three species, C. graminea, C. rosea, and

C. ornata, that are distributed in the southeastern United States and forms a polyploid complex with both putative autopolyploids and allopolyploids. Based on our current knowledge of the distribution of Callisia section Cuthbertia, the following questions were asked: 1) How well do ecological niche models constructed from abiotic variables match the actual distributions of species of Callisia section Cuthbertia? 2) What is the niche

109

overlap and niche breadth of taxa in Callisia section Cuthbertia? 3) Are there niche shifts among the polyploids compared to their progenitors? 4) What was the projected distribution of Callisia section Cuthbertia in the Last Glacial Maximum and Mid-

Holocene time period, and was there species overlap that could have led to hybridization? and 5) What is the predicted niche suitability for species of Callisia section Cuthbertia in the future?

Ecological niche models (ENMs) or species distribution models (SDMs) have shown an increased use, due to the abilities to predict effects on patterns of biodiversity and environmental niche suitability for species (Elith and Leathwick 2009, Guisan and

Thuiller 2005). These fundamental tools are now essential for determining ecological niche shifts that are associated with speciation at different ploidal levels (e.g. Glennon et al. 2014, Godsoe et al. 2013, Hamlin et al. 2017, Marchant et al. 2016, Theodoridis et al. 2013, Visger et al. 2016). Here, using ENMs, I investigate species distribution in the past, present, and future and assess niche overlap, niche breadth, and range overlap between polyploids and their progenitors in three polyploid systems in Callisia section

Cuthbertia: 1) a putative autopolyploid complex comprising diploid and tetraploid C. graminea (including C. rosea), 2) a putative autopolyploid complex of diploid and tetraploid C. ornata, and 3) a putative allopolyploid complex of tetraploid C. graminea, diploid C. ornata, tetraploid C. ornata, and hexaploid C. graminea and C. ornata.

Materials and Methods

Data points

During the summers of 2012 – 2015, 133 locality data points for Callisia section

Cuthbertia taxa were obtained (Chapter 2) and all these data points were included in the

ENM. Although 436 specimens from herbarium vouchers were georeferenced and 755

110

specimens were mined from the Integrated Digitized Biocollections portal (iDigBio; http://www.idigbio.org), the data could not be used since the majority of these voucher specimens did not specify their ploidy. All georeferenced data points of C. rosea, which comprises only diploid entities, were used in ENM. Due to the lack of locality data for tetraploid C. ornata, vouchers were selected that represented the morphological characters of C. ornata 4x discussed in Chapter 3. Limited occurrence data points were also the case for C. graminea 6x and C. ornata 6x, each of which is confined to one or two counties. Based on Chapter 3, these allohexaploids have the same progenitors (C. graminea 4x and C. ornata 2x), and to account for the limited number of occurrence points for each, the GPS points of both hexaploids were pooled into a single group. In the ENM analysis, a minimum of 15 occurrence data points typically needed to get good predictive power (Papeş and Gaubert 2007) and in this study, 30 diploid C. graminea,

248 tetraploid C. graminea, 25 diploid C. ornata, 18 tetraploid C. ornata, 88 C. rosea, and 17 hexaploid C. graminea and C. ornata data points were used.

Ecological Niche Modeling Layers

Bioclimatic (BioClim) variables at a 2.5-minute spatial resolution were obtained from the WorldClim 1.4 database (Hijmans et al. 2005). Also, sourced from WorldClim, were climate projections for the LGM (~22ka) and MH (~6ka). Both sets of climatic layers are simulations from the Community Climate System Model, version 4 (CCSM4)

(Gent et al. 2011). For the future projections of 2050 and 2070, climatic data were used under different circumstances based on different Representative Concentration

Pathways (RCP). Both RCPs are defined by their total radiative forcing, a cumulative measure of human emissions of greenhouse gases from all sources expressed in Watts per square meter, simulated in combined assessment models to 2100

111

(http://sedac.ipcc-data.org/ddc/ar5_scenario_process/RCPs.html, accessed August

2017). RCP2.6 and RCP8.5 (van Vuuren et al. 2011a) were chosen because both signify contrasting greenhouse gas emission scenarios. RCP2.6 aims to limit the increase of global mean temperature and have a low radiative forcing (~3 Wm-2) by limiting the greenhouse gas emissions due to implementing strong mitigation scenarios with an expected temperature rise by 2100 in the range of 1.5–2°C (van Vuuren et al.

2011b). RCP8.5 corresponds to the pathway with the highest radiative forcing (~8.5

Wm-2) due to high greenhouse gas emissions (no climate-specific mitigation targets or policies set, high population growth, and modest improvements in energy-use intensity) with an expected temperature increase of 4.5°C by 2100 (Meinshausen et al. 2011,

Riahi et al. 2011). The downscaled and calibrated projections of the CCSM4 model which were used in the past projection were also use in the future projection analysis.

The altitude layer (altitude above sea level) provided by WorldClim and a geology

(lithology) raster layer were used in the current and future analysis. The geology layer was created and converted to 2.5-minute resolution in ArcGIS 10.5 (ESRI 2016) from the USGS Geologic shapefile of the United States

(https://mrdata.usgs.gov/geology/state/ accessed August 2017).

Selection of Climatic Variables

From the 19 abiotic layers with various temperature and precipitation conditions, a subset of variables was selected to reduce the number of highly correlated variables; the layers were projected and cropped in R (R Core Team 2016) to the spatial dimension of the southeastern United States. The BioClim layers were subsequently statistically tested for correlation within the southeastern UNITES STATES with a

Pearson correlation coefficient test. Highly correlated variables of r > 0.75 were

112

removed to reduce the number of climatic layers and to prevent overfitting the niche models. The final set of nine abiotic predictors used to run the ecological models consisted of altitude (alt), annual mean temperature (bio1), mean diurnal range (bio2), mean temperature of wettest quarter (bio8), mean temperature of driest quarter (bio9), annual precipitation (bio12), precipitation of driest month (bio14), precipitation of coldest quarter (bio19), and geology (geo) in the current and future analysis. For the past

ENMs, altitude and geology were removed since geology and altitude layers of the past are not available.

Ecological Niche Modeling

For the current, past (LGM and MH), and future (2050 and 2070) predictions, niche modelling was performed with maximum entropy modelling using MaxEnt v.3.3

(Phillips and Dudík 2008) with default parameters, except for number of replicates (10) and percentage of random tests (25). Extrapolating was not included in the past predictions because MaxEnt was overfitting the model. The training and testing output curves were visually compared, and the area under the curve statistic (AUC, Metz 1978) was used to evaluate model accuracy and the relative contribution of each variable to the final model. An AUC value of 0.9 and higher indicates an excellent performance of the model, but values ranging between 0.7 and 0.9 are useful as well (Swets 1988).

Overlap between niches, niche breadth, niche identity, and range overlap were implemented in ENMTools v 1.4.4 (Warren et al. 2008, 2010) for the three species and polyploids in Callisia section Cuthbertia under different climate scenarios in the past, present, and future.

The observed niche overlap, a measure of proportional similarity between two taxa, was quantified among all polyploids and species in the present, past and future

113

using the Schoener's D index, which ranges from zero to one; zero equals no niche overlap while one equals identical niches (Schoener 1968, Warren et al. 2008). Niche differentiation was assessed through the niche identity test, also known as the niche equivalence test. This test evaluates the hypothesis that the ENMs produced for two populations are equal (see Warren et al. 2010). A one-tailed t-test is employed to test whether the values of the 100 random bootstrap pseudoreplicates are significantly different from the observed niche overlap value.

Niche breadth is defined as a variety of conditions or a suite of suitable resources needed for a species to thrive (Gaston et al. 1997), which is positively correlated with the range of its geographical extent (Brown 1984). Niche breadth was quantified by applying Levin’s (1968) inverse concentration index, which ranges from zero (only one grid cell in the geographic space has a non-zero suitability) to one (all grid cells are equally suitable) (Mandle et al. 2010).

Range overlap is based on MaxEnt output data that have been converted to presence/absence predictions. The binary predictions are created by using a threshold suitability score to count as a predicted presence (Liu et al. 2005, Warren et al. 2008).

In this study, a 10-percentile species threshold was used. Everything below the threshold value is considered absent, and everything above the threshold value is present (Raes et al. 2009). The reclassification of a continuous MaxEnt raster to a binary raster was done in R (R Core Team 2016).

Principal Component Analysis

Niche divergence in multivariate space was investigated based on nine abiotic layers as variables in principal components analysis (PCA) for all entities of Callisia section Cuthbertia following Visger et al. (2016). The abiotic variable values for the

114

occurrence points were extracted using R (R Core Team 2016). The PCA analyses were performed in R using FactoMineR (Lê et al. 2008), and visualization of the variable loading plot and PCA biplot was generated in R using Factoextra (Kassambara and

Mundt 2017). Four PCA analyses were implemented for 1) C. graminea 2x, 4x, 2) C. ornata 2x, 4x, 3) C. ornata 2x, 4x, C. graminea 4x, and combined C. graminea 6x and

C. ornata 6x, and 4) C. graminea 2x, 4x and C. rosea.

Results

The ENMs for the present for all members of Callisia section Cuthbertia follow the current distribution of the three diploid and four polyploid entities. The performance for all models in the past, present, and future had AUC values > 0.90 (Table 4-1), indicating that the models accurately predicted the presence of polyploids and their progenitors.

Predicted Suitability and Principal Component Analysis under Current Climatic Conditions

Consistent with the current geographic distribution, Callisia graminea 2x and C. graminea 4x overlap in South Carolina and North Carolina (Figure 4-1A-B). The estimated range for C. graminea 2x and 4x is 14,718 km2 and 105, 299.19 km2, respectively (Table 4-1), while the estimated range overlap is 43.9% (Table 4-2). In

Figure 4-1C, a variable correlation plot is depicted, indicating the quality and importance of the 9 variables (altitude (alt), annual mean temperature (bio1), mean diurnal range

(bio2), mean temperature of wettest quarter (bio8), mean temperature of driest quarter

(bio9), annual precipitation (bio12), precipitation of driest month (bio14), precipitation of coldest quarter (bio19), and geology (geo)) in the principal components analysis. The

PCA analysis of C. graminea 2x and 4x, defined by environmental space, revealed that

115

the first principal component is annual mean temperature, which explains 40.7% of the variance of all BioClim layers and the second principal component is mean diurnal range, which explains 22.6% of the variance (Figure 4-1D). The PCA biplot indicates that tetraploid C. graminea overlaps with diploid C. graminea and has a much larger range compared to the diploid ancestor.

The predicted distributions of Florida endemics C. ornata 2x and 4x (Figure 4-2A-

B) indicate that diploid C. ornata is confined to the eastern coast of Florida and central

Florida, while tetraploid C. ornata has a high suitability in central Florida. The estimated range for C. ornata 2x and 4x is 29,831 km2 and 18,261 km2, respectively (Table 4-1), while the estimated range overlap is 45.1% (Table 4-2). The PCA analysis of diploid and tetraploid C. ornata (Figure 4-2D) shows a large distribution for C. ornata 2x and a small cluster of points of C. ornata 4x in the biplot. In Figure 4-2C, a variable correlation plot is depicted, indicating the quality and importance of the 9 variables (mentioned above) in the principal components analysis. The variables that most contributed to the PCA biplot are annual precipitation (PC1, 37.2%) and mean temperature of driest quarter (PC2,

30.9%).

The predicted suitability for hexaploid C. graminea and C. ornata (Figure 4-3A) is confined to central Florida with low suitability in Georgia. The range of both polyploids is estimated at 13,782 km2, and they have an estimated overlap of 32.3 %, 48.3%, and

16.5% with C. graminea 4x, C. ornata 2x, and 4x, respectively. The PCA analysis of tetraploid C. graminea, diploid C. ornata, tetraploid C. ornata, and pooled data points for hexaploid C. graminea and C. ornata indicated that C. graminea 4x and C. ornata 2x overlap (Figure 4-3C), and the ordination of the hexaploids is between C. graminea 4x

116

and C. ornata 2x. Callisia ornata 4x intersects all other C. graminea and C. ornata entities. In Figure 4-3C, a variable correlation plot is depicted, indicating the quality and importance of the 9 variables (mentioned above) in the principal components analysis.

Annual mean temperature (PC1) and mean diurnal range (PC2) were the variables that most contributed to the PCA biplot with 44.4% and 15.5%, respectively.

Callisia rosea has high habitat suitability in Georgia and South Carolina (Figure

4-4A) and low habitat suitability in Florida, North Carolina, and Virginia. The geographic distribution of C. rosea has an estimated range of 115,110 km2 (Table 4-1) and overlaps by 5.8% and 39.9% with the distributions of C. graminea 2x and 4x, respectively (Table

4-2). These results are similar to those of the PCA analysis in which C. rosea and C. graminea 4x overlap and C. graminea 2x intersects with both entities (Figure 4-4C). In

Figure 4-4B, a variable correlation plot is depicted, indicating the quality and importance of the 9 variables (mentioned above) in the principal components analysis. The variables that most contributed to the PCA biplot are annual mean temperature (PC1,

43.2%) and mean temperature of the driest quarter (PC1, 24.1%).

Predicted Suitability under Past Climatic Conditions

The overall suitability at the LGM for the entities of Callisia section Cuthbertia showed that there were suitable areas for C. graminea 2x and 4x, C. rosea, and C. ornata 2x and 4x, whereas no suitable niches were found for the hexaploids C. graminea and C. ornata. In the MH, all currently known species and polyploids of

Callisia section Cuthbertia had suitable niches.

The predicted area of suitability for diploid C. graminea at the LGM stretches from the panhandle of Florida to eastern Texas along the coast, with an area of 42,344 km2 (Table 4-1); this distribution is ~1.88 times larger than the current range of niche

117

suitability (Figure 4-5A). Callisia graminea 2x had an estimated range overlap of 8% with C. rosea and a 0.3% overlap with tetraploid C. ornata. In the MH, there was a shift in locality from the coastal plain of the Gulf of Mexico (LGM) to the southeastern United

States with a predicted area of 199, 903 km2 (Table 4-1), which is 12.6 times larger than present predicted suitability (Figure 4-5B). The estimated range overlap for C. graminea

2x with C. graminea 4x, the hexaploids, and C. rosea in the MH were 28%, 1.3%, and

25.8%, respectively (Table 4-2).

Low habitat suitability was noted for tetraploid C. graminea at the LGM (Figure

4.5D). These low-suitability areas were along the Florida panhandle and along the east and west coasts of Florida, extending to Grand Bahama Island. However, the limestone substrate (Buchan 2000) of this island would likely make growth of this species unlikely.

Due to the low suitability (0 – 0.215) and a 10-percentile presence logistic threshold of

0.394, the threshold raster turned out to be zero, which resulted in no suitable range for

C. graminea 4x. In the MH, there is a predicted niche suitability with a range of 97,495 km2, which is 7.4% smaller than the current distribution (Figure 4-5 E–F). Range overlap between C. graminea 4x and the combined hexaploids was estimated at 62.8%. A high suitability is noted for C. graminea 4x in Florida and along the Fall Line in Georgia and

South Carolina, and along the coastal plain of North Carolina.

Callisia ornata 2x had a projected suitability for the LGM that stretched from the eastern coast of Florida but also encompassed Grand Bahama Island (Figure 4-6A), with an estimated range of 15,304 km2 (which is 48.7% smaller than the current distribution; Table 4-1). As previously mentioned, due to a limestone substrate on Grand

Bahama Island, growth of these species would be unlikely. In the MH, the distribution of

118

C. ornata 2x increased 1.62 times its estimated LGM range (Table 4-1) and covers in the MH the whole eastern coast of Florida and central Florida (Figure 4-6). The estimated range overlap of C. ornata 2x with C. graminea 4x, C. ornata 4x, and the combined hexaploid entities is 10.3%, 46.4%, and 40.5%, respectively (Table 4-2).

The predicted range of C. ornata 4x at the LGM is limited to Florida (Figure 4-

6D), with a range of 30,561 km2, which is roughly 67.4% larger than the current distribution (Table 4-1). Even though the predicted ranges of C. ornata 2x and C. ornata

4x are both in Florida, there was no range overlap at the LGM based on my analyses. In the MH, the estimated range of C. ornata 4x was 22,172 km2 (Table 4.1) in Florida

(Figure 4-6E), a decrease of 27.5% in range compared with the LGM. Overlap in range between C. ornata 4x and C. graminea 4x was estimated at 1.1%.

At the LGM, C. rosea was confined to Florida with low suitability in Georgia and an estimated range of 68,309 km2, which is 40.7% smaller than the current predicted range (Figure 4-7A). There was an estimated range overlap of 25.5% and 93.1% with C. ornata 2x and 4x, respectively (Table 4-2); however, in MH, the range overlap changed to 0% and 4.3% respectively, because the suitable habitat of C. rosea moved to

Georgia and South Carolina (Figure 4-7B), covering an estimated range of 114,438 km2, which is similar to the current distribution (Table 4-1). In the MH, C. rosea had an estimated range overlap with C. graminea 4x, C. ornata 4x, and the hexaploid entities of

48.1%, 4.3%, and 16.1%, respectively (Table 4-2).

As previously mentioned, C. graminea 6x and C. ornata 6x had no predicted suitability in the LGM; however, in the MH, the predicted suitability for both hexaploid entities was high in Florida and low in Georgia and South Carolina (Figure 4-7E). The

119

combined hexaploid entities had an estimated range of 36,104 km2, which is 1.6 times larger than the current distribution (Figure 4-7F). An approximate range overlap of

20.4% was calculated between the hexaploid entities and C. ornata 4x.

Predicted Suitability and Range Changes under Future Climatic Conditions

All future predictions are distribution estimates in the years 2050 and 2070 under the RCP2.6 and RCP8.5 climate change scenarios, which are based on greenhouse gas emissions. The overall trend for C. graminea 2x, 4x, and C. rosea under the

RCP2.6 scenario was a suitability that moves slightly north compared to the current distribution; in contrast, under the RCP8.5 scenario, there is an extreme shift to the north relative to the current distribution. For C. ornata 2x and 4x, there are hardly any changes predicted, but an extreme expansion of suitability is noticeable for the hexaploids.

Under the current climatic parameters, C. graminea 2x has a narrow range of predicted suitability along the Fall Line in South and North Carolina and also in Virginia.

Under the RCP2.6 scenario, there is an increased suitability in 2050 and 2070 that ranges from South Carolina to Virginia with a much broader range stretching north of the Fall Line (Figure 4-8), with a total area of 67,289 km2 and 54,357 km2, in 2050 and

2070 respectively (Table 4-3). Compared to the present distribution, these ranges are

3.6 and 2.7 times larger than the current range. Under the RCP8.5 pathway in 2050 and

2070, there is an increase of nearly twice the current distribution, with an area of 27,471 km2 and 28,570 km2, respectively. North Carolina becomes less suitable because the high predicted suitability for C. graminea 2x has shifted to Virginia (Figure 4-8). Under the RCP2.6 pathway, there will be 35.4-39.1% range overlap with C. graminea 4x; however, an overlap of 41%-53.9% was noted under the RCP8.5 climatic scenario

120

(Table 4-2). The range overlap of the hexaploids under the RCP2.6 pathway with C. graminea 2x showed an increase from current 0% to 15.3% in 2050 and a decrease to

13.3% in 2070. Under the RCP8.5 pathway, an increase was noted of 31.2% in 2050 and 89.4% in 2070. Callisia rosea has a predicted range overlap with C. graminea 2x of

45.4% in 2050 and then a slight decrease of 1.8% in 2070 under the RCP2.6 scenario; however, under RCP8.5, there is only 3% overlap increase in 2050 followed by a 43% overlap increase in 2070 (Table 4.2).

Tetraploid C. graminea has the second largest area in the current distributions, with 105,299 km2, compared to the other taxa in Callisia section Cuthbertia. In Figure 4-

9, the predicted suitability in 2050 (RCP2.6 and RCP8.5) shifts to the north into Virginia, but under the RCP8.5, it is more extreme. In 2070, there is an extreme loss of niche suitability under the RCP8.5 climate change scenario. Under the RCP2.6 climate change scenario, there is a predicted range increase of 61% in 2050 followed by a decrease of 47.9% in 2070, and under RCP8.5, there is a predicted expansion of 58.9% in 2050, followed by a range contraction of 35.3% in 2070 compared to the current distribution (Table 4-3). The range overlap of the hexaploids under the RCP2.6 model with C. graminea 4x showed an increase from 32.3% for the present to 57.6% in 2050 and a decrease to 54.1% in 2070; however, under the RCP8.5 model, the range overlap indicated an increase from 32.3% in the present to 76.3% in 2050 and 83.2% in 2070.

The predicted range overlap between C. ornata 2x and C. graminea 4x contracted from

14.2% to 7.2% in 2050, followed by a slight decrease to 6.6% in 2070 under the RCP2.6 scenario; however, under RCP8.5, there is a slight increase in overlap, from 14.2% to

15.1% in 2050 followed by a 20.5% increase in 2070. Under current and future

121

predicted suitability for C. graminea 4x, there is no overlap with C. ornata 4x. Callisia rosea has a predicted range overlap with C. graminea 4x that expanded by 16.3% in

2050 and subsequently a slight increase of 2.1% in 2070 under the RCP2.6 scenario; however, under RCP8.5, there is only a 6.3% increase in 2050 followed by a 6.1% increase in 2070, which is similar to the current predicted overlap (Table 4.2).

Callisia rosea, with the largest distribution, shows the same trend as C. graminea

2x and 4x for the future, where there is a movement of niche suitability northwards, towards North Carolina and Virginia, with the RCP8.5 model having a more extreme suitability loss in 2070 (Figure 4-10). The RCP2.6 scenario shows a range expansion of

23.7% in 2050 and a decrease of 21.78% for suitable range; in contrast, RCP8.5 shows a range contraction of 8.12% and 44.16% for 2050 and 2070, respectively. Currently, there is hardly any overlap with C. rosea and the hexaploids C. ornata and C. graminea, but due to the northern shift of niche suitability for the hexaploids, there is a predicated overlap of 34.4% in 2050, 31.4% in 2070 under RCP2.6 and 50.3% in 2050, 66.3% in

2070 under the RCP8.5 climate change scenario. Callisia rosea does not have any current range overlap with C. ornata 2x and 4x and that stays unchanged in 2050 and

2070 under both RCP2.6 and RCP8.5 climate change scenarios (Table 4-2).

Callisia ornata 2x is currently confined to Florida, which might slightly change if the RCP8.5 climate change model is correct. In 2050, with RCP8.5, there is a low predicted suitability for C. ornata 2x in Georgia and South Carolina (Figure 4-11). The current range of C. ornata 2x is 29,832 km2, but under RCP2.6, the suitability changes to 43,338 km2 (45.28%) in 2050 and decreases in 2070 to 42,151 km2 (41.3%). Under

RCP8.5, the range suitability increases by 61.07% in 2050 and then decreases 20.38%

122

in 2070 (Table 4-3). Callisia ornata 2x has a range overlap with C. ornata 4x and the hexaploids in the future; with C. ornata 4x under RCP2.6 there is an overlap of 39.7% and 37.3% in 2050 and 2070, respectively, and under RCP8.5, there is an overlap of

42.9% and 38.8%, respectively. With the hexaploids, under RCP2.6, there is an overlap of 79% and 82.5% in 2050 and 2070, respectively, and under RCP8.5, there is an overlap of 98.9% and 99%, respectively (Table 4-2).

The predicted suitability for C. ornata 4x is highly similar to that of C. ornata 2x, restricted to Florida, but the suitability in Florida becomes higher in the future due to climate change. Under both RCP2.6 and RCP8.5, there is high suitability in Virginia in

2050 and 2070. The predicted suitability stays unchanged in Florida, but due to the increase of suitable habitat in Virginia, the range of C. ornata 4x increases (Figure 4-

12). Under the RCP2.6 scenario, there is an increase of 79.02% in 2050 and 138.69% in 2070, while under the RCP8.5 model, there is a 102.32% increase in 2050 and

139.86% increase in 2070 (Table 4-3).

The current predicted suitability of hexaploid C. graminea and C. ornata is restricted to Florida, with the smallest range of 13,783 km2. In the future, under the

RCP2.6 and RCP8.5 climate change scenarios, there is an extreme increase of predicted niche suitability ranging from Florida to Virginia (Figure 4-13). Under the

RCP2.6 scenario, there is an increase in range (Table 4-3), but it stays stable from 2050 to 2070 (10.4 times and 10.1 times larger than present); however, under the RCP8.5 climate change scenario, there is an increase in suitability in 2050 (16.8 times larger) followed by a 4.4x reduction of suitability in 2070 compared to 2050.

123

Niche Overlap, Equivalence, and Breadth of ENMs of Callisia Section Cuthbertia in the Past, Present, and Future

Niche overlap analysis under current climatic conditions in Callisia section

Cuthbertia showed an observed niche overlap (Schoener’s D index) of D=0.304 between C. graminea 2x and C. graminea 4x while, D=0.243 between C. graminea 2x and C. rosea (Table 4-4). The observed niche overlap values between C. graminea 2x and C. ornata 2x, C. ornata 4x and the hexaploids respectively, were very low (all

D<0.1). Under current climatic conditions, C. graminea 4x has a niche overlap with C. ornata 2x, C. rosea, and the two hexaploid entities, with D values of 0.194, 0.505, and

0.366, respectively. Hexaploid C. graminea and C. ornata have, under current climatic conditions, an observed niche overlap with C. rosea, C. ornata 2x, and C. ornata 4x of

0.171, 0542, and 0.028, respectively. A Schoener’s D index of 0.35 was recorded for C. ornata 2x and C. ornata 4x, while 0.116 was recorded for C. ornata 2x and C. rosea and the niche overlap between C. rosea and C. ornata 4x is 0.278.

Because niche overlap calculations are not based on threshold raster files, an observed niche overlap could be calculated for C. graminea 4x at the LGM. The niche overlap results indicated that C. graminea 2x had a niche overlap with C. graminea 4x,

C. ornata 2x, C. ornata 4x, and C. rosea. Callisia rosea had the largest niche overlap with C. graminea 4x and the smallest with C. graminea 2x. A Pearson correlation coefficient test between range overlap and niche overlap data showed that the results of

LGM, MH, Current, 2050 RCP2.6, and 2007 RCP8.5 are strongly correlated, with a r>0.8. The correlation coefficient of 2050 RCP8.5 and 2070 RCP2.6 was <0.8. This lower correlation coefficient is due to niche overlap of C. graminea 2x with C. ornata 2x, and C. graminea 2x with C. rosea in 2050 RCP8.5. This was also the case for C. rosea

124

with C. ornata 2x and 4x in 2070 RCP2.6; however, for the other past, current or future climatic scenarios, no overlap has been detected. The equivalence test indicates that there is no significance divergence between C. graminea 4x and the hexaploids in MH,

2050 RCP2.6, 2050 RCP8.5, and 2070 RCP8.5; in addition, no significant divergence was found between the hexaploid entities and C. rosea in 2070 RCP2.6 (p>0.05). In all other tests, there is significant divergence for all species and polyploids in Callisia section Cuthbertia with p<0.01 (Table 4-4).

Niche breadth in Callisia section Cuthbertia entities in the current climatic conditions shows that C. rosea and C. graminea 4x have the highest niche breadth with

0.354 and 0.301, respectively (Table 4-5), whereas C. ornata 4x has the lowest niche breadth with a value of 0.049. At the LGM, C. rosea has the highest niche breadth, with of 0.03; however, in the MH, niche breadth increases (Table 4-5), with C. graminea 2x and C. rosea having the highest niche breadth values, both near 0.35. In the future climatic scenarios of 2050 under the RCP2.6 model, there is a predicted increase in niche breadth for all taxa of Callisia section Cuthbertia; conversely, under the RCP8.5 scenario, there is only niche breadth increase in C. graminea 4x and the hexaploid entities. The predicted value for niche breadth in 2070 under RCP2.6 indicates a slight increase for C. graminea 2x and C. ornata 4x, but the remaining taxa remain stable compared to 2050 (RCP2.6); however, under RCP8.5, all niche breadth values decrease, with C. graminea 2x, C. ornata 4x, and C. rosea shifting to values less than under current climatic conditions (Table 4-5).

125

Discussion

Predicted Suitability and Principal Component Analysis under Current Climatic Conditions

The current predicted suitability of Callisia section Cuthbertia reflects the distribution pattern of all diploid and polyploid entities discussed in Chapter 2. Callisia graminea 2x has a range of 14,718 km2 and occurs along the Fall Line with a high suitability in South Carolina and North Carolina and a low suitability in Virginia. In

Chapter 2, I suggested that, due to heavy agricultural activities in North Carolina, many previously suitable areas for C. graminea 2x were converted into farmland, extirpating numerous populations. These suggestions are supported by the ENMs, because the area ranging from Johnston County to Northampton County showed medium to high suitability for C. graminea 2x; however, no specimens have been collected in these regions. This supports the hypothesis that in the recent past, C. graminea 2x may have had a larger range stretching from North Carolina to Virginia. The predicted suitability under the current climatic envelope for the autotetraploid C. graminea is approximately

7 times larger than the predicted geographical range of its progenitor, C. graminea 2x.

This supports the hypothesis that polyploids will evolve niche differentiation and occupy different and wider geographic ranges than their diploid parents (Levin 2002, Otto and

Whitton 2000, Stebbins 1985). The range of C. graminea 4x runs from North Carolina along the coastal plain to central Florida and overlaps with its diploid progenitor in North and South Carolina, but also with C. rosea in Georgia and South Carolina. In Florida, C. graminea 4x overlaps and hybridizes with C. ornata 2x, (see Chapter 3) and form hexaploids C. graminea and C. ornata progenies. The environmental PCA analysis indicates that C. graminea 4x has evolved to a wider range, which is driven by annual

126

mean temperature and mean diurnal range. Diploid C. graminea is confined to a specific temperature regime, but tetraploid C. graminea has a tolerance of different temperature regimes. This supports Chapter 2, that C. graminea 2x and 4x are not geographically isolated, because C. graminea 4x has the capability to occur in the same habitat as C. graminea 2x with a colder temperature regime. The niche breadth of C. graminea 4x is

0.301, which is larger than its diploid progenitor C. graminea (0.106), which indicates niche expansion for C. graminea 4x based on niche categorization of Marchant et al.

(2016).

Molecular and morphological analyses (Chapter 3) suggest that C. ornata 2x is the progenitor of C. ornata 4x. Both entities are endemic to Florida, as indicated in the suitability maps. High to medium suitability was predicted along the eastern Florida coast for C. ornata 2x; however, low suitability was predicted in western and central

Florida. While conducting fieldwork, I was unable to locate C. ornata 2x in the latter areas. High suitability for tetraploid C. ornata was indicated in central Florida and along the western coast, as well as in northwestern Florida, but C. ornata 4x was not located in the latter region. Even though there is a range overlap of 14.2% between C. ornata 2x and 4x, the two entities are predominantly separated. In addition, the environmental

PCA, which is based on annual precipitation and mean temperature of the driest quarter, shows that the area of the tetraploid C. ornata is much smaller than that of the diploid progenitor. The niche breadth of C. ornata 2x was 0.104, and that of C. ornata 4x was 0.049, which is a clear example of niche contraction (Marchant et al. 2016, Parisod and Broennimann 2016). Autopolyploid cytotypes typically exhibit higher niche breadth then their diploid progenitor (McIntyre 2012, Visger et al. 2016) and a larger range

127

compared to their diploid parent(s) (Lowry and Lester 2006), although Theodoridis et al.

(2013) showed that cytotypes within Primula section Aleuritia (Primulaceae) had a smaller niche breadth and distribution range than their progenitor, as seen in C. ornata

2x and C. ornata 4x.

The hexaploids C. graminea and C. ornata included in this study were only collected in Florida. A high probability of occurrence is noted in Florida; however, a low predicted suitability is predicted in Georgia. Even though the predicted range of the hexaploids is 13,783 km2 (the smallest distribution of all entities in Callisia section

Cuthbertia), these polyploids overlap with their progenitors (C. graminea 4x and C. ornata 2x) and C. ornata 4x (Chapter 3). The ecological niche divergence estimates for

C. graminea 4x, C. ornata 2x, C. ornata 4x, C. graminea 6x, and C. ornata 6x, based on an ecological PCA analysis, indicate that the hexaploid entities are between the ordination of C. ornata 2x and C. graminea 4x. The niche breadth of the hexaploids is larger than that of C. ornata 2x (0.104) and smaller than that of C. graminea 4x (0.301), which follows Marchant et al. (2016) and Parisod and Broennimann (2016) as niche intermediacy.

Callisia rosea has the largest range of 115,110 km2, and it overlaps most with C. graminea 4x (39.9%). The distribution of C. rosea matches the distribution in Chapter 2.

Highly suitable habitat is noted in Georgia and South Carolina, covering the piedmont and coastal plain. Between these two high suitability areas, there is a low-suitability zone that has a geology of sand stone and limestone, which could explain the low suitability because most Callisia section Cuthbertia entities prefer sandy soil (personal observation). Low suitability was also shown in North Carolina, Virginia, and Florida; C.

128

rosea was also not encountered in these areas in this study. The ecological PCA analysis indicates that C. graminea 4x and C. rosea have a large area in which their

ENMs overlap based on the variables, annual temperature and mean temperature of driest quarter. A niche breadth of 0.343 for C. rosea and of 0.301 for C. graminea 4x and the large overlap in distribution support the possibility of hybridization between C. graminea 4x and C. rosea (see Chapter 3). In addition, different ploidies can coexist and hybridize in the same area, as reported by Oliver and Rejon (1980).

Predicted Suitability under Current, Last Glacial Maximum, and Mid-Holocene Conditions

Hindcasting climatic envelopes (LGH and MH) of Callisia section Cuthbertia has provided predictions of niche shifts, niche overlap, and range overlap in the past. One of the most perceptible results at the LGM was the absence of suitable areas for the hexaploids C. graminea and C. ornata and a very low suitability for C. graminea 4x. At the LGM, the predicted geographical distributions of C. graminea 2x and C. ornata 4x were much larger than the current distributions. Even though C. graminea 4x had low suitability at the LGM, in the MH it followed the same trend as the hexaploids, with a smaller reconstructed range than the current distribution. In contrast, the remaining entities all had larger geographical distributions compared to their current distributions.

The expanded high pineland and scrub habitats in the LGM are characteristic of xeric species and have been reported numerous times, as in Delcourt (1980), Owen (2002),

Watts (1971, 1980).

Similar results to those provided here were also reported by López-Alvarez et al.

(2015), who argued that range expansion in the past was based on the tolerance of the species, the availability of suitable habitat, and the existence of refugia. Lake Wales

129

Ridge in central peninsular Florida is considered a Pleistocene refugium (Lewis and

Crawford 1995) and approximates the predicted suitability for C. ornata 2x, 4x and C. rosea. At the LGM, the distribution of C. ornata 2x reached the Grand Bahama Island.

These results are unlikely, because the substrate of this island is Limestone and Callisia section Cuthbertia grows on sandy soils. High suitability for the Grand Bahama Island was recorded, because the predictor soil map was not available and omitted for the

LGM analysis, leading to questionable results. Callisia graminea 2x is the only species at the LGM that had a distribution stretching along the coast of the Gulf of Mexico, with a range from the Florida panhandle to Texas, an area also suggested to be a refugium in the Pleistocene (Walker et al. 2009). In addition, C. graminea 4x had some suitable localities along the Gulf of Mexico adjacent to those of its diploid ancestor.

Even though there is no highly suitable habitat and therefore no range overlap at the LGM calculated for C. graminea 4x, the low-suitability areas overlap with the range of C. graminea 2x, C. rosea, and C. ornata 4x. Callisia graminea 2x and C. graminea 4x had a niche overlap of 0.048 and a significant divergence between the two taxa and based on the niche breadth analysis, C. graminea 4x showed signs of niche expansion because it had a larger inverse concentration compared to its diploid ancestor. A significant divergence was obtained by the identity test in C. rosea with C. ornata 2x and

4x, and the same results were retrieved between C. ornata 2x and 4x; however, the niche breadth of tetraploid C. ornata and its diploid ancestor were slightly different, with

C. ornata 2x having a higher inverse concentration. In fact, all climatic scenarios of the past, present, and future for the comparison of diploid and tetraploid C. ornata have the same results, suggesting niche contraction in the C ornata polyploid complex. At the

130

LGM, there is no range overlap between C. ornata 2x and 4x, but in the MH, present, and future climatic scenarios, there is range overlap. This indicates perhaps that C. ornata 4x originated before or at the LGM and expanded to a different habitat than its diploid progenitor; however, range overlap takes place in the current climatic envelope and in the future. This difference could be based on the availability of suitable habitats and the tolerance of the species, per López-Alvarez et al. (2015), suggesting that suitable habitats may be limited for C. ornata 2x and 4x to expand northward, restricting the distribution and coexistence of C. ornata 2x and 4x to Florida.

The predicted suitability in the MH provided distinctive changes compared to the

LGM: 1) Callisia graminea 2x, 4x, and C. rosea have moved northwards where they are currently established; 2) the existence of suitable areas for the hexaploids C. graminea and C. ornata, which did not occur at the LGM; and 3) Callisia ornata 2x expanded into central Florida. Tetraploid C. ornata and its diploid ancestors were confined to Florida in both the LGM and MH, where there is range overlap in central Florida.

Suitable ranges in the MH increased tremendously compared to the LGM, especially in C. graminea 2x. In this time period, the southern pines were expanding their distribution (Wright 1976). These pines favored sandy soils and were adapted to frequent fires, which supports the expanding ranges in the early Holocene of Callisia section Cuthbertia as well, because all entities in section Cuthbertia favor similar environmental conditions. Giles (1942, 1943) postulated based on geological evidence that C. graminea 2x (growing on Cretaceous period deposits) was displaced by tetraploid C. graminea (growing on Pleistocene deposits). It was clear, according to

Giles (1942) that diploid ancestors grew on older sediments than the tetraploids, which

131

perhaps colonized land as it became available after the retreat of the sea in the

Pleistocene. In this study, the suitability prediction for the MH supports the hypothesis of

Giles that the range of the diploid ancestor was larger than its current range. The projections in the MH showed an estimated range that was 12.5 times larger than the current distribution. Now the questions arises: why is the current distribution of C. graminea 2x so small compared to its size in the MH? Giles (1942) suggested that the vigorous tetraploid C. graminea perhaps replaced the original range of the diploid ancestors. The results in this study are clear that C. graminea 2x and 4x have expanded tremendously in the MH (Figure 4.5 D-F) and that the niche breadth of C. graminea 2x was larger than that of tetraploid C. graminea 4x in the MH, evidence of niche contraction in C. graminea 4x due to a lower niche breadth, but simultaneously range expansion. In the current climatic scenario, C. graminea 2x shows evidence of niche contraction and range decrease the opposite of the MH, implying that C. graminea 2x is being replaced by its polyploid descendants. Along the same lines, Stebbins (1985), te

Beest et al. (2011) reported that polyploids can be invasive and be in competition with their diploid progenitors due to higher fitness (Felber and Bever 1997). The current results and these mechanisms of higher fitness of C. graminea 4x in fact support that C. graminea 4x has in part replaced its progenitor from its native habitat and expanded its range throughout the southeastern United States.

Callisia rosea had the largest range in the LGM (68.309 km2), which was 40.6% smaller than its current distribution (115,110 km2). In the MH, the projected range increased up to 114,438 km2, were it more or less stabilized to the current distribution.

Interesting was the projected overlap of C. rosea with diploid and tetraploid C. ornata at

132

the LGM that could potentially support hybridization between C. rosea and C. ornata 2x solving the incongruence in the phylogenetic trees in Chapter 3. After the MH this would not be possible since the overlap between C. rosea and C. ornata 2x and 4x, declined to zero. In the MH there was a low to medium suitability for C. rosea in Florida, which changed to low suitability compared to the current predicted niche suitability. This goes in parallel with a move northwards of C. rosea in the MH, especially as the distribution expanded in the piedmont from Georgia to South Carolina. The reason for this move is most likely associated with global climate change.

The hexaploids C. graminea and C. ornata have a range overlap with their progenitors, C. graminea 4x and C. ornata 2x, in the MH, with the largest overlap of

62.8% with C. graminea 4x; in addition, the niche overlap between the hexaploids and tetraploid C. graminea showed no significant niche divergence. As in the MH climatic analysis, the niche breadth of the hexaploid is higher than C. ornata 2x but lower than

C. graminea 4x, suggesting niche dominance (Parisod and Broennimann 2016). In the current climatic envelope, the hexaploids move from niche dominance to niche intermediacy as the identity test of C. graminea 4x and the hexaploids shows significant divergence between the hexaploids and their progenitors. This, implies gain of novel habitat (niche shift), an advantage for the hexaploids, which leads to an escape from

MCE (Fowler and Levin 1984, Visger et al. 2016).

Predicted Suitability under Future Conditions

The future predictions are based on two different greenhouse gas emissions and concentration pathways: RCP2.6 and RCP8.5. The RCP2.6 emission and concentration pathway represents full cooperation of all countries to reduce greenhouse gasses between 2010 to 2100 by 70% compared to a baseline scenario (van Vuuren et al.

133

2011b); in contrast, RCP8.5 indicates no climate change policies and the highest greenhouse gas emissions (Riahi et al. 2011).

The overall trend of the future projections showed that all species and polyploids of Callisia section Cuthbertia will have suitable areas until the end of the current century. In both RCP2.6 and RCP8.5 scenarios, C. graminea 2x will have an increased distribution; however, the greater increase will be under RCP2.6. It is also clear (Figure

4.8) that under RCP2.6 the expansion of the range will surround the current predicted suitability; in contrast, under the RCP8.5, suitable habitat will be found to the north, in

Virginia, with a loss of suitability in South Carolina and North Carolina in 2070. An increase in suitability for C. graminea 4x is observed under the RCP2.6 climatic scenario, but under RCP8.5, there is an increase in projected suitability in 2050 but a drastic decrease of 44% in suitable habitat in 2070. As seen with the diploid ancestor, suitable habitat for the tetraploid entities will be to the north in Virginia. In all scenarios, there is significant divergence between the progenitor and the polyploid. However, even if suitable habitat occurs to the north, movement of the taxa northwards could be difficult due to agriculture and fragmentation of suitable habitat.

Based on niche breadth, C. graminea 4x has a higher value than its progenitor; however, under the RCP2.6 (2050–2070), the inverse concentration had lower variation, compared to RCP8.5. These results are similar to the projected range analysis, indicating that niche breadth will increase in 2050 but will decrease in 2070. Callisia rosea also showed the same results as C. graminea 2x and 4x; under RCP2.6, there is an increase in suitability in 2050 and minimal change occurs in 2070, and under

134

RCP8.5 there is an increase in 2050 but a decrease of suitability in 2070 with a move northwards toward Virginia.

The future projections for C. ornata 2x and 4x have the same tendency as C. rosea and C. graminea 2x and 4x, but the changes are a bit different for C. ornata 4x. It seems that, under RCP2.6 and RCP8.5, there is increased suitability in North Carolina and Virginia, a locality were Callisia ornata in general does not currently occur. Based on this extensive gap, it is unlikely that C. ornata 4x will occur there in the future.

According to the analysis of variable contribution, annual mean temperature (bio1), precipitation of coldest quarter (bio19) and mean diurnal range (bio2) have the higest percent contributions to the future maps of C. ornata 4x and the extensive gap between

North Carolina and Florida.

The hexaploids are the only polyploids that have a predicted suitability that will expand throughout the whole southeastern United States. In all greenhouse emission scenarios, suitable habitat for C. graminea 6x and C. ornata 6x will move northwards, as previously seen for other members of the Callisia section Cuthbertia clade. Again, the result under RCP2.6 is more stable after expansion in 2050, but under RCP8.5, there is increased habitat suitability in 2050, but in 2070 the localities in Georgia, South

Carolina, and North Carolina become unsuitable. The identity test also showed that there is no niche divergence between C. graminea 4x and the hexaploids. Since the suitable habitat for the hexaploids is expanding towards the north with C. graminea 4x and because the hexaploids had already escaped MCE in the MH, this overlap represents coexistence of the three entities.

135

Based on the results for the future, it is clear that all entities of Callisia section

Cuthbertia will have increased predicted ranges under RCP2.6 in 2050 with no change in 2070; however, with the high greenhouse gas emission value (RCP8.5) in 2070, all distributions of all the populations are predicted to move northward. The distribution shifts of organisms moving poleward and upward in elevation are based on temperature increases that have been reported frequently (Felde et al. 2012, Lenoir et al. 2008,

Scheffers et al. 2016, Soltis 2017); in fact, climate change studies on 21 animal groups showed the same prediction (Mason et al. 2015). However, not all studies reflect the same trend, but imply that latitudinal and elevation shifts are also driven by other environmental parameters such as water availability (Fei et al. 2017, McLaughlin et al.

2017, Rapacciuolo et al. 2014, VanDerWal et al. 2012). Zhang et al. (2017) also reported that not all species will undergo a distribution shift because from the 7465 seed plant taxa studied in North America, the authors reported that some plants will lose suitable habitat, while others will experience range expansion and even extinction. The results of the future predictions imply that when policies are in place to reduce greenhouse gas emissions, the distribution of Callisia taxa will increase and stabilize over time. In contrast, with no policies in place, suitable habitats for Callisia section

Cuthbertia will decrease. The forecastings of this study also imply that, up to the end of this century, all species and polyploids in Callisia section Cuthbertia will continue to have suitable habitats. This contrasts with the situation for other species in the

Appalachians of eastern North America, which are projected to be largely extinct

(Gaynor et al. in revision, Wiens 2016) It is important to note that these results for

Callisia section Cuthbertia are projections based on prediction and unobserved data

136

and that these projections should be interpreted with caution and may deviate from the actual distributions in the future.

137

Figure 4-1. Predicted niche suitability and statistical environmental differentiation of Callisia graminea 2x, C. graminea 4x. A) Predicted niche suitability for C. graminea 2x, B) Predicted niche suitability for C. graminea 4x. C) Variable correlation plot indicating the quality and importance of the variables in the principal components analysis. D) Principal components scatterplot indicating niche expansion of C. graminea 4x. PC1 and PC2 represents bio 1 (annual mean temperature) and bio 2 (mean diurnal range) which accounted for 40.7% and 22.6% of variance respectively.

138

Figure 4-2. Predicted niche suitability and statistical environmental differentiation of Callisia ornata 2x, C. ornata 4x. A) Predicted niche suitability for C. ornata 2x, B) Predicted niche suitability for C. ornata 4x. C) Variable correlation plot indicating the quality and importance of the variables in the principal components analysis. D) Principal components scatterplot indicating that C. ornata 2x and 4x are overlapping. PC1 and PC2 represents bio 12 (annual precipitation) and bio 9 (mean temperature of driest quarter) which accounted for 37.2% and 30.9% of variance respectively.

139

Figure 4-3. Predicted niche suitability of hexaploid C. graminea and C. ornata, and statistical ecological niche divergent estimation of Callisia graminea 4x, C. ornata 2x, C. ornata 4x, C. graminea 6x and C. ornata 6x. A) Predicted niche suitability for C. graminea 6x and C. ornata 6x. B) Variable correlation plot indicating the quality and importance of the variables in the principal components analysis. C) Principal components scatterplot indicating overlap of C. ornata 2x and C. graminea 4x. Hexaploid C. graminea and C. ornata are intermediate to C. ornata 2x and C. graminea 4x, while C. ornata 4x is intermediate of C. ornata 2x and the hexaploid entities. PC1 and PC2 represents bio 1 (annual mean temperature) and bio 2 (mean diurnal range) which accounted for 44.4% and 18.5% of variance respectively.

140

Figure 4-4. Predicted niche suitability of C. rosea and statistical ecological niche divergent estimation of C. graminea 2x, 4x and C. rosea. A) Predicted niche suitability for C. rosea. B) Variable correlation plot indicating the quality and importance of the variables in the principal components analysis. C) Principal components scatterplot representing, overlap of C. graminea 2x, 4x and C. rosea. PC1 and PC2 represents bio 1 (annual mean temperature) and bio 9 (mean temperature of driest quarter) which accounted for 43.2% and 24.1% of variance respectively.

141

Table 4-1. The estimated range suitability (km2) of Callisia section Cuthbertia in the past and current. The difference is the estimated positive or negative percentage compared with the current distribution. Taxa LGM (km2) Difference (%) mid-Holocene (km2) Difference (%) Current (km2) C. graminea 2x 42,344 187.70 199,904 1258.21 14,718 C. graminea 4x 0 0 97,495 -7.41 105,299 C. graminea & C. ornata 6x 0 0 36,104 161.95 13,783 C. ornata 2x 15,304 -48.70 40,080 34.35 29,832 C. ornata 4x 30,561 67.35 22,172 21.41 18,262 C. rosea 68,309 -40.66 114,438 -0.58 115,110

Table 4-2. The estimated range overlap of Callisia section Cuthbertia in the past, present and future. Zero indicate no overlap and one indicates 100% overlap. CG2x = C. graminea 2x, CG4x = C. graminea 4x. C6x = C. graminea 6x and C. ornata 6x, CO2x = C. ornata 2x, CO4x = C. ornata 4x and CR2 = C. rosea. Taxon pair LGM mid-Holocene Current 2050 RCP2.6 2050 RCP8.5 2070 RCP2.6 2070RCP8.5 CG2x-CG4x — 0.280 0.439 0.354 0.410 0.391 0.539 CG2x-C6x — 0.013 0 0.153 0.312 0.133 0.894 CG2x-CO2x 0 0 0 0 0 0 0 CG2x-CO4x 0.003 0 0 0 0 0 0.002 CG2x-CR2x 0.083 0.258 0.058 0.512 0.081 0.494 0.512 CG4x-C6x — 0.628 0.323 0.576 0.763 0.541 0.832 CG4x-CO2x — 0.103 0.142 0.072 0.151 0.066 0.205 CG4x-CO4x — 0.011 0 0 0 0 0 CG4x-CR2x — 0.481 0.399 0.562 0.462 0.577 0.401 C6x-CO2x — 0.405 0.483 0.790 0.989 0.825 0.990 C6x-CO4x — 0.204 0.165 0.427 0.782 0.468 0.614 C6x-CR2x — 0.161 0.005 0.344 0.503 0.314 0.663 CO2x-CO4x 0 0.464 0.451 0.397 0.429 0.373 0.388 CO2x-CR2x 0.255 0 0 0 0 0 0 CO4x-CR2x 0.931 0.043 0 0 0 0 0

142

Table 4-3. The estimated range suitability (km2) of Callisia section Cuthbertia in the current predictions and future projections under two climate change scenario’s (RCP2.6 and RCP8.5). Difference is the estimated difference in percentage compared with the current distribution. Taxa Current (km2) 2050 (km2) Difference (%) 2070 (km2) Difference (%)

Climate change scenario RCP2.6 C. graminea 2x 14,718 67,289 357.18 54,357 269.32 C. graminea 4x 105,299 169,509 60.98 155,723 47.89 C. graminea 6x & C. ornata 6x 13,783 157,470 1042.51 153,765 1015.62 C. ornata 2x 29,832 43,338 45.28 42,151 41.30 C. ornata 4x 18,262 32,693 79.02 43,590 138.69 C. rosea 115,110 142,390 23.70 140,181 21.78

Climate change scenario RCP8.5 C. graminea 2x 14,718 27,471 86.65 28,570 94.11 C. graminea 4x 105,299 167,333 58.91 68,114 -35.31 C. graminea 6x & C. ornata 6x 13,783 246,512 1688.54 185,512 1245.97 C. ornata 2x 29,832 48,051 61.07 41,970 40.69 C. ornata 4x 18,262 36,947 102.32 43,803 139.86 C. rosea 115,110 105,763 -8.12 64,278 -44.16

143

Figure 4-5. A comparison of niche suitability for Callisia graminea 2x and C. graminea 4x projected in the past and present. A) C. graminea 2x projected in Last Glacial Maximum (LGM: ca. 22,000 BP), B) C. graminea 2x projected in mid-Holocene (MH: ca. 6,000 BP), C) C. graminea 2x present niche suitability, D) C. graminea 4x projected in LGM, E) C. graminea 4x projected in MH, F) C. graminea 4x present niche suitability.

144

Figure 4-6. A comparison of Niche suitability for Callisia ornata 2x and C. ornata 4x projected in the past and present. A) C. ornata 2x projected in Last Glacial Maximum (LGM: ca. 22,000 BP), B) C. ornata 2x projected in mid- Holocene (MH: ca. 6,000 BP), C) C. ornata 2x present niche suitability, D) Callisia ornata 4x projected in LGM, E) C. ornata 4x projected in MH, F) C. ornata 4x present niche suitability.

145

Figure 4-7. A comparison of niche suitability for Callisia rosea and the hexaploids C. graminea & C. ornata projected in the past and present. A) C. rosea projected in Last Glacial Maximum (LGM: ca. 22,000 BP), B) C. rosea projected in mid-Holocene (MH: ca. 6,000 BP), C) C. rosea present niche suitability, D) hexaploids in LGM, E) hexaploids projected in MH, F) hexaploids in present niche suitability.

146

Figure 4-8. A comparison of niche suitability for Callisia graminea 2x in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5. The RCP2.6 pathway (ca. 1.5 °C increase by 2100) indicates the projected niche suitability for C. graminea 2x in 2050 and 2070. The RCP8.5 pathway (ca. 4.5 °C increase by 2100) indicates the projected niche suitability for C. graminea 2x in 2050 and 2070.

147

Figure 4-9. A comparison of niche suitability for Callisia graminea 4x in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5. The RCP2.6 pathway (ca. 1.5 °C increase by 2100) indicates the projected niche suitability for C. graminea 4x in 2050 and 2070. The RCP8.5 pathway (ca. 4.5 °C increase by 2100) indicates the projected niche suitability for C. graminea 4x in 2050 and 2070.

148

Figure 4-10. A comparison of niche suitability for Callisia rosea in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5. The RCP2.6 pathway (ca. 1.5 °C increase by 2100) indicates the projected niche suitability for C. rosea in 2050 and 2070. The RCP8.5 pathway (ca. 4.5 °C increase by 2100) indicates the projected niche suitability for C. rosea in 2050 and 2070.

149

Figure 4-11. A comparison of niche suitability for Callisia ornata 2x in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5. The RCP2.6 pathway (ca. 1.5 °C increase by 2100) indicates the projected niche suitability for C. ornata 2x in 2050 and 2070. The RCP8.5 pathway (ca. 4.5 °C increase by 2100) indicates the projected niche suitability for C. ornata 2x in 2050 and 2070.

150

Figure 4-12. A comparison of niche suitability for Callisia ornata 4x in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5. The RCP2.6 pathway (ca. 1.5 °C increase by 2100) indicates the projected niche suitability for C. ornata 4x in 2050 and 2070. The RCP8.5 pathway (ca. 4.5 °C increase by 2100) indicates the projected niche suitability for C. ornata 4x in 2050 and 2070.

151

Figure 4-13. A comparison of niche suitability for Callisia graminea 6x and C. ornata 6x in the future based on estimated greenhouse gas concentration RCP2.6 and 8.5. The RCP2.6 pathway (ca. 1.5 °C increase by 2100) indicates the projected niche suitability for C. graminea 6x and C. ornata 6x in 2050 and 2070. The RCP8.5 pathway (ca. 4.5 °C increase by 2100) indicates the projected niche suitability for C. graminea 6x and C. ornata 6x in 2050 and 2070.

152

Table 4-4. Observed niche overlap values (Schoener’s D) and results of tests of niche Identity test (ID) for Callisia section Cuthbertia. Taxon pair comparisons for different scenarios indicating the overlap in Last Glacial Maximum (LGM), mid-Holocene, Current, 2050 (RCP2.6 and 8.5) and 2070 (RCP2.6 and 8.5). The niche observed niche overlap was tested with a one-tailed t test against 100 simulated niche identity values (not shown). D = significant divergence (p<0.01), NS = non-significant divergence (p>0.05). LGM mid-Holocene Current 2050 RCP2.6 2050 RCP8.5 2070 RCP2.6 2070 RCP8.5 Taxon pair Observed ID Observed ID Observed ID Observed ID Observed ID Observed ID Observed ID overlap overlap overlap overlap overlap overlap overlap CG2x-CG4x 0.048 D 0.255 D 0.304 D 0.295 D 0.259 D 0.369 D 0.326 D CG2x-CG6x — — 0.155 D 0.050 D 0.235 D 0.201 D 0.166 D 0.295 D CG2x-CO2x 0.033 D 0.069 D 0.040 D 0.043 D 0.259 D 0.080 D 0.019 D CG2x-CO4x 0.050 D 0.009 D 0.005 D 0.003 D 0.002 D 0.041 D 0.021 D CG2x-CR2x 0.092 D 0.293 D 0.243 D 0.428 D 0.257 D 0.427 D 0.431 D CG4x-CG6x — — 0.575 NS 0.366 D 0.617 NS 0.663 NS 0.043 D 0.569 NS CG4x-CO2x 0.120 D 0.249 D 0.194 D 0.180 D 0.177 D 0.015 D 0.202 D CG4x-CO4x 0.121 D 0.065 D 0.053 D 0.038 D 0.021 D 0.004 D 0.035 D CG4x-CR2x 0.219 D 0.604 D 0.505 D 0.531 D 0.484 D 0.212 D 0.450 D CG6x-CO2x — — 0.503 D 0.542 D 0.430 D 0.389 D 0.564 D 0.426 D CG6x-CO4x — — 0.231 D 0.028 D 0.185 D 0.178 D 0.329 D 0.215 D CG6x-CR2x — — 0.409 D 0.171 D 0.401 D 0.381 D 0.596 NS 0.351 D CO2x-CO4x 0.166 D 0.373 D 0.350 D 0.357 D 0.391 D 0.410 D 0.366 D CO2x-CR2x 0.307 D 0.149 D 0.116 D 0.082 D 0.066 D 0.374 D 0.042 D CO4x-CR2x 0.524 D 0.050 D 0.278 D 0.011 D 0.005 D 0.210 D 0.005 D

153

Table 4-5. Results of niche breadth calculation for the past, current and future predictions. Past Current Future 2050 Future 2070 Taxa LGM mid-Holocene Present RCP2.6 RCP8.5 RCP2.6 RCP8.5 C. graminea 2x 0.016 0.351 0.106 0.195 0.092 0.178 0.087 C. graminea 4x 0.021 0.293 0.301 0.387 0.423 0.385 0.360 C. graminea 6x 0 0.208 0.135 0.365 0.493 0.351 0.379 & C. ornata 6x C. ornata 2x 0.015 0.135 0.104 0.125 0.127 0.120 0.111 C. ornata 4x 0.013 0.051 0.049 0.077 0.049 0.090 0.076 C. rosea 0.030 0.354 0.343 0.335 0.298 0.335 0.211

154

CHAPTER 5 GENERAL CONCLUSIONS

The main purpose of this study was to investigate the distribution, the relationships, and ecological niche of the polyploid complex Callisia section Cuthbertia.

To investigate the phytogeography of cytotypes in Callisia section Cuthbertia, a total of

436 voucher specimens was georeferenced, and 133 new specimens were collected

(Chapter 2). Based on flow cytometry data, DNA content of all cytotypes in Callisia section Cuthbertia was estimated. Utilizing chromosome counts and flow cytometric analysis, cytotype distribution maps were generated. Two disjunct groups of populations of diploid Callisia graminea were discovered, and tetraploid C. graminea ranges broadly from the coastal plain of North Carolina through central Florida. One hexaploid C. graminea individual was recorded in South Carolina, and numerous individuals of hexaploid C. graminea were found in central Florida. Diploid C. ornata occurs in eastern

Florida and previously unknown tetraploid and hexaploid populations of C. ornata were discovered in western and central Florida, respectively. Callisia rosea occurs in Georgia and the Carolinas, with populations occurring on both sides of the Fall Line. The cytotype and species distributions in Callisia are complex, and these results provided hypotheses to be tested with morphological and molecular data about the origins of the polyploid cytotypes (Chapter 2); these hypotheses were investigated in Chapter 3.

Morphometric measurements were taken from collected plants in Callisia section

Cuthbertia for plant height, leaf width, leaf length, stem thickness, sepal length, sepal width, petal length, petal width, and filament length. Based on these morphological characters, a canonical discriminant analysis (CANDISC) was performed to identify the

155

most distinctive trait(s) separating the diploids and polyploids from each other. The results of the morphometric analysis showed that, based on vegetative traits only, stem thickness and leaf length were the most significant traits separating the polyploids and their putative parents. It is important to note that the CANDISC analysis with adjusted

Bonferroni p-value did not find any significant differences between hexaploid C. graminea and hexaploid C. ornata. The combined vegetative and reproductive data used in the CANDISC analysis revealed that plant height, leaf width, and sepal width had the strongest discriminant power to separate the taxa in Callisia section Cuthbertia from each other; however, due to the low variability in plant height, in the combined vegetative and reproductive data set, the depicted ordination in Figure 3-2C does not reflect results based on field observations (i.e. diploid C. graminea plants are not taller than tetraploid C. graminea in the field as found through this analysis; instead, the polyploids are larger than the diploid ancestor). The results of the CANDISC analysis also indicate that the diploids C. graminea, C. ornata and C. rosea all form separate clusters. The diploids and tetraploids in C. ornata and C. graminea tend to overlap, and the hexaploid C. graminea and C. ornata are intermediate to their putative parents, C. graminea 4x and C. ornata 2x.

Molecular phylogenetic analyses were implemented to determine the relationships among the diploid entities and to investigate the putative parents of the polyploids. The molecular analyses based on one nuclear, six chloroplast intergenic spacer regions, and one chloroplast gene were divided into two major analyses: diploids only (including 20 outgroups) and combined diploids and polyploids. The major divisions were again divided in nuclear only, chloroplast only, and a combined analysis of

156

chloroplast and nuclear data. All molecular phylogenetic analyses showed that Callisia section Cuthbertia is monophyletic, as found in previous studies. All 20 outgroups follow the same topology as Hertweck and Pires (2014). The ITS tree with diploid taxa only showed that diploid C. ornata is sister to C. rosea and C. graminea 2x, while the chloroplast tree revealed that C. rosea is sister to C. ornata 2x and C. graminea 2x.

When nuclear and chloroplast data were combined, C. ornata 2x was sister to C. graminea 2x and C. rosea.

When the polyploids were included in the analysis, the nuclear and chloroplast trees had the same topology as the diploid trees; however, the combined nuclear and chloroplast tree of diploid and polyploid entities supported a relationship of C. graminea

2x and C. ornata 2x sister to C. rosea, which is incongruent with the combined diploid tree. These results are expected to occur when polyploids and hybrids produce conflicting phylogenetic signal. Even though only the diploid trees were utilized to investigate the relationships of the diploids that does not indicate that diploids have not been affected by hybridization. Both chloroplast and nuclear trees of the diploid species reveal part of the relationship of the three species. In the nuclear tree, C. graminea 2x forms a clade with C. ornata 2x and in the chloroplast tree C. graminea 2x forms a clade with C. rosea. Hybridization and/or introgression may explain these results, and I offer just one possible scenario here. One explanation is that C. graminea 2x is a derivative of an ancient hybridization event between C. ornata 2x and C. rosea, and it maintains signatures of both parental species.

Based on the combined nuclear and chloroplast phylogeny of Callisia section

Cuthbertia, it is clear that tetraploid C. graminea is an autopolyploid with C. graminea 2x

157

as progenitor. There is also evidence that a group of tetraploid C. graminea entities hybridized with C. rosea. The rare hexaploid C. graminea from South Carolina seems to be an autopolyploid with C. graminea 4x as progenitor. Tetraploid C. ornata may be of multiple origin; at least some accessions are autotetraploid with C. ornata 2x as the diploid ancestor, whereas other accessions of C. ornata 4x form a clade with C. graminea 4x, suggesting that C. ornata 4x could also be an allopolyploid with C. graminea 4x as one of its progenitors. Hexaploid C. ornata is an allohexaploid with C. ornata 4x and C. graminea 4x as a progenitor, while hexaploid C. graminea is an allohexaploid with C. graminea 4x and C. ornata 2x as its progenitors. It is clear from my analysis that the ancestry from the narrow distributed hexaploid C. graminea from

Florida differs from the ancestry of the rare hexaploids in South Carolina. Since the rare autohexaploids from South Carolina are occasionally found among tetraploid populations it is difficult to recognize them as an established cytotype (Giles, 1943).

However the allohexaploids C. graminea in Florida are established and can be recognized as an established cytotype in Lake County, Florida.

While growing plants of Callisia section Cuthbertia in a common garden, it was noticed that plantlets were borne in the bracts of the inflorescence after the flowering season (fall). Vivipary, which is uncommon in Callisia, was also observed in the field

(personal observation). This method of asexual reproduction might have been one of the reasons why taxa in this polyploid complex have been successful.

Ecological niche modeling of Callisia section Cuthbertia entities was performed using MaxEnt, with an emphasis on evaluating the predicted habitat suitability of the three species and the polyploids in the southeastern United States (Chapter 4). In

158

addition, projections were made into the past and the future. For the past, projections were made into the Last Glacial Maximum (LGM) and mid-Holocene (MH), while for the future, projections were made for 2050 and 2070 under RCP2.6 and RCP8.5 greenhouse gas emission scenarios.

The results showed that the predicted distributions for C. graminea, C. ornata, C. rosea, and the polyploids are similar to their current distributions as reported in Chapter

2. The niche breadth of C. graminea 4x is larger than the inverse concentration of C. graminea 2x, which suggests niche expansion for C. graminea 4x. The niche breadth of

C. ornata 2x was 0.104 and of C. ornata 4x was 0.049, representing a clear example of niche contraction because the niche breadth in the diploid ancestor is larger than that of the polyploid descendants. The niche breadth of the hexaploids in the current predictions is larger than that of C. ornata 2x (0.104) and smaller than that of C. graminea 4x (0.301), resulting in niche intermediacy between the significantly divergent niches of the hexaploids and the diploid and tetraploid progenitors. These conclusions are based on the identity test only; however, a niche similarity test (or background similarity) would be preferred for future analysis to confirm the role of niche divergence in speciation in Callisia section Cuthbertia.

The projections in the LGM showed no suitable areas for hexaploids and low suitability for C. graminea 4x. The projected distribution at the LGM revealed that C. graminea 2x might have had a distribution along the coast from Florida to Texas, while the majority of C. graminea 4x, C. ornata 2x, C. ornata 4x, and C. rosea were confined to Florida. Callisia ornata 2x also had a high projected suitability for the Grand Bahama

Island, however, the limestone substrate (Buchan 2000) of this island would likely make

159

growth of this species unlikely. Range overlap at the LGM was only observed among C. ornata 2x, C. ornata 4x, and C. rosea. Overlap between C. rosea and C. ornata justifies the possible hybrid origin of C. graminea 2x. Due to the low habitat suitability for C. graminea 4x at the LGM, the range and range overlap could not be calculated.

However, in the MH, all species and polyploids had a high projected suitability in the southeastern U.S. It was clear that the projected habitat suitability for diploid and tetraploid C. graminea, diploid and tetraploid C. ornata, and the hexaploids was much larger than the current distribution. A major shift of suitable habitat occurred for C. graminea 2x, C. graminea 4x, and C. rosea, all of which shifted northward to Georgia,

South Carolina, North Carolina, and even Virginia for C. graminea 2x. The distribution of

C. graminea 2x in MH was 12.5 times larger than its current distribution which is based on projections of current predictions. An range overlap of 28% was calculated for C. graminea 2x and 4x in MH; the current distribution of C. graminea 2x has been reduced from 199,904 km2 to 12,718 km2, and C. graminea 4x has occupied the southeastern

U.S. with an increased range overlap (43.9%) with C. graminea 2x, suggesting that the distribution of C. graminea 2x was replaced by C. graminea 4x, an hypothesis that was also proposed by Giles (1943). In the LGM and MH, the distribution of diploid and tetraploid C. ornata was confined to Florida and no change was recorded in the current distribution.

The future projections revealed different suitability situations based on RCP2.6 and RCP8.5. Under RCP2.6, there is an increase in suitability in 2050, and the range stabilizes in 2070; however, under RCP8.5, there is an increase in suitability but in

2070, suitable habitat for all taxa within Callisia section Cuthbertia decreases and

160

moves northward. Tetraploid C. ornata seems to have suitable areas in North Carolina and Virginia, but due to the large gap between Florida and North Carolina, the plants would not be able to move northward. Although the projected suitability might decrease or increase in the future for Callisia section Cuthbertia, the results suggest that the existence of taxa in section Cuthbertia is guaranteed due to a range of suitable areas in the future with or without global warming policies in place. These future projections are only based on one global climate model, the Community Climate System Model, version

4 (CCSM4). It would be advised to incorporate different climate models in future studies, to investigate if results of CCSM4 projections are supported. It is therefore important to interpret these results with caution, because they are based on prediction and unobserved data which can deviate from the actual distributions in the future.

161

APPENDIX A GEOREFERENCED DATA POINTS

Table A-1. Georeferenced data points. Examined and georeferenced voucher specimens of Callisia graminea (CG), C. ornata (CO) and C. rosea (CR) with the geographic location, collector, collection number and date collected. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected Callisia graminea (Small) Tucker CG FL Alachua Gainesville 29°51.77'N 082°07.87'W E. West et al s. n. May 25, 1935 CG FL Alachua Gainesville, Cedar Ridge 29°38.63'N 082°25.26'W D. Tilley 2152 August 13, 2001 area CG FL Alachua Morningside Nature Center. 29°39.26'N 082°16.72'W C. Kabat et al 40 May 4, 2001 CG FL Alachua Paynes Prairie State Park 29°31.69'N 082°17.14'W C. Easley 298 May 31, 1986 CG FL Alachua Paynes Prairie State Park 29°31.14'N 082°17.87'W W. Faircloth et al 627 May 3, 1987 CG FL Citrus Fort Cooper State Park 28°48.74'N 082°18.70'W R. Hattaway FC0114 May 12, 2000 CG FL Citrus N. end of Lake Tsala 28°58.38'N 082°22.30'W J. Beckner 630 October 16, Apopka 1964 CG FL Citrus NE. of Holder 28°58.53'N 082°21.78'W L. Baltzell 8252 April 11, 1976 CG FL Clay Bayard Point 29°56.45'N 081°37.82'W C. Slaughter et al 12510 May 21, 2001 CG FL Clay Black Creek Ravines 30°03.39'N 081°50.76'W L. Anderson 18702 June 11, 1999 Conservation Area CG FL Clay Blue Pond 29°52.22'N 082°01.00'W G. Fleming et al 8385 August 3, 1972 CG FL Clay Gold Head Branch State 29°50.90'N 081°57.74'W E. West et al s. n. May 12, 1939 Park CG FL Clay Gold Head Branch State 29°50.82'N 081°57.64'W O. Baynard s. n. April 1, 1944 Park CG FL Clay Gold Head Branch State 29°52.35'N 081°55.93'W S. Bergamo 50 July 11, 1981 Park CG FL Clay Jennings State Forest 30°09.74'N 081°57.23'W S. Orzell et al 20105 July 9, 1992 CG FL Clay On Fla. 225, 4 mi SE. of 30°00.17'N 082°03.15'W J. Beckner 1315 June 2, 1966 Lawtey CG FL Clay SE. of Lawtey 30°04.89'N 082°02.60'W J. Abbott et al 13635 June 16, 2000 CG FL Clay Silver Sands Lake Drive 29°47.58'N 081°58.53'W J. Kunzer 504 April 10, 2004 CG FL Clay Silver Sands Lake Drive 29°47.58'N 081°58.53'W J. Kunzer 537 May 17, 2004 CG FL Clay Silver Sands Lake Drive 29°47.58'N 081°58.53'W J. Kunzer 504A April 10, 2004 CG FL Clay Silver Sands Lake Drive 29°47.58'N 081°58.53'W J. Kunzer 543A May 17, 2004

162

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CG FL Dixie Ca. 3 air mi SE. of 29°38.08'N 083°21.30'W L. Anderson 12046 June 2, 1989 Steinhatchee CG FL Duval 1 mi E. of US 17, N. side of 30°25.90'N 081°37.61'W C. Becker et al 1234 April 15, 1959 Jacksonville CG FL Duval Fort Caroline National 30°22.22'N 081°29.15'W D. Giannasi et al 1551 July 1, 2005 Monument CG FL Duval Julington-Durbin Peninsula; 30°07.93'N 081°32.35'W S. Orzell et al 20011 July 3, 1992 Bayard CG FL Franklin Along Ridge Rd. Eastpoint 29°45.26'N 084°52.15'W L. Anderson 10874 September 2, 1987 CG FL Franklin Apalachicola 29°48.77'N 084°50.60'W J. Bozeman 11216 August 8, 1967 CG FL Franklin E. of Tidal Creek on N. side 29°55.26'N 084°30.77'W L. Anderson 24534 May 29, 2009 of Hwy 9 CG FL Franklin Panacea 29°58.71'N 084°22.99'W N. Henderson 67-1474 August 22, 1967 CG FL Franklin Parking Area of St. James 29°56.96'N 084°30.21'W L. Anderson 7088 April 30, 1984 fire tower CG FL Franklin Saint Teresa 29°56.22'N 084°30.36'W S. Orzell et al 12168 September 14, 1989 CG FL Lake 2 mi E. of Jct. with Fla 561 28°45.63'N 081°43.17'W J. Beckner 2373 April 13, 1970 W. of Lake Jem CG FL Lake E. of Lady Lake 28°55.54'N 081°52.74'W L. Baltzell 11104 May 18, 1980 CG FL Lake Eustis 28°50.49'N 081°41.62'W J. Ray s. n. July, 1894 CG FL Lake Mt. Plymouth. 28°48.48'N 081°31.99'W R. Daubenmire et s. n. June 16, 1977 al CG FL Lake N. of Alexander Springs 29°05.39'N 081°33.76'W D. White et al s. n. May 30, 1984 CG FL Lake N. of Leesburg 28°51.67'N 081°54.24'W L. Baltzell 3851 April 19, 1972 CG FL Lake Ocala National Forest, 29°04.81'N 081°34.14'W R. Wilbur et al 2659 August 17, 1950 CG FL Lake Silver Lake 28°50.51'N 081°47.90'W W. Tisdale s. n. April 19, 1935 CG FL Leon St. Marks National Wildlife 30°09.09'N 084°08.85'W M. Kuck 60 s. d. Refuge CG FL Levy 3 mi S. of Williston 29°20.41'N 082°26.78'W H. Hume s. n. March 26, 1937 CG FL Levy S. of Morriston 29°15.18'N 082°26.49'W R. Kral 4515 April 7, 1957 CG FL Marion 4 mi N. of Ocala 29°14.50'N 082°09.16'W W. Porter s. n. July 7, 1938 CG FL Marion 5.6 mi E. of Jct. US 301, 29°26.88'N 082°2.19'W R. Smith et al 300 July 2, 1961 along Fla 318

163

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CG FL Marion E. of Dunnellon 29°03.21'N 082°17.78'W L. Baltzell 10293 May 14, 1978 CG FL Marion Ocala National Forest 29°20.71'N 081°44.17'W B. Lambert 113 April 15, 1949 CG FL Marion Ocala National Forest 29°22.15'N 081°49.22'W W. Judd et al 2652 May 13, 1980 CG FL Marion Orange Springs 29°28.26'N 081°58.10'W E. West et al s. n. May 27, 1941 CG FL Marion Orange Springs 29°28.26'N 081°58.10'W E. West et al s. n. May 28, 1941 CG FL Marion Salt Springs Island, near 29°20.27'N 081°44.76'W G. Guala II et al 1066 September 24, Salt Springs 1988 CG FL Marion Silver River State Park. 29°12.02'N 082°02.73'W J. Hubbard 497 May 21, 2006 CG FL Nassau Ralph E. Simmons 30°46.76'N 081°57.24'W C. Slaughter 12722 August 15, 2001 Memorial State Forest CG FL Nassau Ralph E. Simmons 30°46.77'N 081°57.25'W L. Anderson 19443 September 16, Memorial State Forest 2000 CG FL Putnam Dunns Creek State Park 29°33.30'N 081°35.09'W M. Lelong et al 1536 June 26, 2002 CG FL Putnam Melrose 29°42.27'N 082°01.77'W D. Ward 1845 April 18, 1960 CG FL St Johns N. of Adams Acres Rd., St. 29°52.40'N 081°23.29'W L. Anderson 14927 July 12, 2006 Augustine CG FL Sumter ENE. of the Withlacoochee 28°52.55'N 082°08.00'W M. Strong 2330 May 22, 2000 River CG FL Sumter N. side of FL 44, 1 mi 28°53.15'N 082°05.99'W R. Wunderlin et al 9637 June 28, 1984 WNW of I-75 CG FL Sumter W. of Wildwood 28°52.81'N 082°07.52'W M. Strong 3640 August 26, 2006 CG FL Wakulla 1.7 mi N. of Panacea 30°04.85'N 084°23.70'W L. Anderson 5261 June 5, 1981 CG FL Wakulla Beside Rd. 404 at gate on 30°01.26'N 084°26.58'W L. Anderson 23974 June 20, 2008 W. side Rte. 372. CG FL Wakulla St. Marks National Wildlife 30°04.09'N 084°23.33'W R. Godfrey 83216 May 12, 1989 Refuge CG FL Wakulla St. Marks National Wildlife 30°01.75'N 084°24.51'W L. Anderson 22280 August 25, 2006 Refuge CG GA Ben Hill Fitzgerald 31°52.06'N 083°15.21'W J. Hardin et al 16166 May 18, 1953 CG GA Bulloch Cypress Lake Rd. 32°22.39'N 081°53.07'W A. Kelly 106 June 5, 1989 CG GA Bulloch Lower Lotts Creek Church 32°21.33'N 081°51.31'W C. O Neal 35 June 21, 1965 CG GA Burke Girard 33°09.61'N 081°45.88'W J. Bozeman et al 8967 April 22, 1967 CG GA Candler Canoochee River 32°25.40'N 082°09.07'W R. Bow et al 161 May 13, 1975

164

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CG GA Candler Metter 32°25.88'N 082°09.51'W W. Walker et al 139 May 13, 1975 CG GA Charlton GA 185 30°26.37'N 082°11.92'W G. Norsworthy et al s. n. May 28, 1972 CG GA Charlton GA 185, & 2 mi S. of GA 94 30°25.16'N 082°11.98'W A. Kelly 109 June 6, 1989 CG GA Charlton Moniac 30°26.28'N 082°11.89'W G. Norsworthy et al s. n. May 28, 1972 CG GA Coffee E. side Seventeen Mile 31°31.75'N 082°48.85'W M. Hopkins 723 April 22, 1965 River CG GA Coffee NE. of Douglas 31°31.48'N 082°49.16'W R. Carter et al 4179 June 28, 1985 CG GA Coffee Along E. side of Seventeen 31°30.21'N 082°44.42'W R. Carter et al 6550 April 29, 1988 Mile River CG GA Coffee NE. of Douglas 31°32.17'N 082°48.68'W A. Kelly 110 June 7, 1989 CG GA Emanuel N. side of GA 46 about 1.7 32°22.39'N 082°21.1'W R. Crook 300 July 1, 1993 mi E. of US 1 CG GA Evans Fort Stewart Military 31°51.68'N 081°37.34'W T. Zebryk 160 April 22, 1992 Reservation CG GA Johnson Kite 32°41.65'N 082°30.02'W S. Jones 23295 June 15, 1979 CG GA McIntosh Everett 31°29.97'N 081°36.51'W J. Bozeman 767 July 8, 1962 CG GA McIntosh Ft. Barrington Rd. 31°30.12'N 081°35.36'W J. Bozeman 2194 October 7, 1962 CG GA Pierce Blackshear 31°21.41'N 082°14.25'W J. Bozeman 9385 June 9, 1967 CG GA Richmond Bennoch Mill Rd. 33°18.67'N 081°57.06'W W. Duncan 3472 June 20, 1941 CG GA Tattnall Bluff over Ohoopee River 32°04.84'N 082°10.14'W A. Kelly 107 June 5, 1989 CG GA Tattnall Cobbtown 32°17.19'N 082°13.32'W J. Bozeman 10609 June 27, 1967 CG GA Tattnall Reidsville 32°06.66'N 082°9.63'W A. Cronquist 5333 June 16, 1948 CG GA Tattnall Reidsville 32°07.44'N 082°10.04'W H. Ahles 54167 June 11, 1961 CG GA Telfair GA 149, 9.4 mi N. of GA 31°58.80'N 082°50.65'W A. Kelly 108 June 5, 1989 117 CG GA Telfair N. side of Turnpike Creek 31°59.52'N 082°55.09'W R. Harper et al 16920 August 1, 1953 CG GA Toombs Lyons 32°14.91'N 082°16.83'W J. Bozeman 10635 June 27, 1967 CG GA Toombs Lyons 32°10.90'N 082°17.62'W R. Wilbur 3099 April 26, 1953 CG GA Ware Ware County 31°15.71'N 082°17.59'W J. Bozeman et al 3792 June 7, 1966 CG GA Wilkinson 1.6 mi E. on dirt Rd. from 32°50.25'N 083°04.32'W A. Kelly 111 June 7, 1989 GA 112 CG GA Wilkinson Sandy Creek 31°51.55'N 083°13.30'W J. Pyron et al 3179 June 13, 1938

165

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CG GA Bacon Along E. bank of Little 31°25.91'N 082°25.98'W R. Carter et al 7339 August 18, 1988 Hurricane Creek CG GA Camden Kingsland 30°46.35'N 081°40.12'W R. Carter et al 17643 June 22, 2007 CG GA Charlton 3 airmiles of Moniac 30°28.99'N 082°11.95'W R. Carter et al 16432 April 7, 2006 CG GA Irwin Alapaha River 31°32.16'N 083°23.83'W W. Faircloth 2863 August 24, 1965 CG NC Bladen Ammon 34°45.68'N 078°36.27'W R. Wilbur 48776 July 6, 1988 CG NC Bladen Ammon 34°44.79'N 078°36.40'W R. Wilbur 48950 July 21, 1988 CG NC Bladen Ammon 34°45.68'N 078°36.27'W R. Wilbur 59467 August 21, 1991 CG NC Bladen Ammon 34°50.49'N 078°33.70'W D. Zobel s.n July 22, 1966 CG NC Bladen Ammon 34°45.52'N 078°36.41'W R. Wilbur 5903 June 5, 1957 CG NC Bladen Bay Tree Lake State Park 34°40.55'N 078°23.41'W R. Wilbur 48697 July 6, 1988 CG NC Bladen Bay Tree Lake State Park 34°39.37'N 078°23.35'W R. Carter 7220 July 30, 1988 CG NC Bladen Elizabethtown 34°41.83'N 078°36.14'W A. Kelly N/A April 3, 1988 CG NC Bladen Garland-White Lake 34°43.21'N 078°26.72'W J. Kelly s. n. November 6, 1967 CG NC Bladen Jerome 34°46.21'N 078°46.62'W R. Wilbur 62343 June 1, 1994 CG NC Bladen Lagoon 34°34.68'N 078°23.29'W R. Wilbur 49312 July 29, 1988 CG NC Bladen Smith’s Pond 34°43.67'N 078°25.57'W R. Wilbur 49353 July 29, 1988 CG NC Bladen Tobermory 34°48.29'N 078°53.03'W A. Kelly 113 June 18, 1989 CG NC Bladen White Lake 34°34.22'N 078°26.27'W R. Weaver 811 May 19, 1967 CG NC Bladen White Lake 34°38.83'N 078°33.18'W D. Stone 2674 May 29, 1968 CG NC Bladen White Lake 34°39.22'N 078°30.44'W D. Stone 2676 May 29, 1968 CG NC Bladen White Lake 34°39.22'N 078°30.44'W D. Stone et al 2742 May 28, 1970 CG NC Bladen White Lake 34°39.30'N 078°29.87'W H. Blomquist 5747 May 15, 1932 CG NC Bladen White Lake 34°40.28'N 078°29.41'W R. Wilbur 10360 May 28, 1968 CG NC Bladen White Lake 34°38.88'N 078°34.33'W R. Wilbur 10362 May 28, 1968 CG NC Bladen White Oak 34°44.05'N 078°43.11'W A. Kelly N/A September, 1987 CG NC Brunswick Along Rt. 211 34°03.04'N 078°17.52'W G. Moore 3655 May 3, 1998 CG NC Brunswick NE. of Seaside towards 33°54.04'N 078°26.97'W R. Wilbur 5671 May 11, 1957 Ocean Isle Beach

166

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CG NC Brunswick Shallotte 33°58.64'N 078°21.31'W R. Wilbur 48383 June 23, 1988 CG NC Brunswick Shallotte 33°55.68'N 078°16.34'W R. Wilbur 48407 June 23, 1988 CG NC Brunswick Shallotte 33°55.89'N 078°17.64'W R. Wilbur 61042 June 18, 1992 CG NC Columbus Lake Waccamaw 34°16.77'N 078°28.03'W R. Wilbur 49855 August 12, 1988 CG NC Cumberland Alongside Cedar Creek Rd. 34°51.81'N 078°43.82'W A. Kelly N/A September, 1987 CG NC Cumberland Bladen county line 34°54.53'N 078°44.03'W R. Wilbur 50101 August 17, 1988 CG NC Cumberland Cedar creek 34°54.75'N 078°44.03'W R. Wilbur 56596 August 16, 1990 CG NC Cumberland Cedar creek 34°56.14'N 078°44.45'W R. Wilbur 56659 August 16, 1990 CG NC Cumberland Fayetteville 34°59.05'N 078°45.78'W R. Wilbur 5587 May 10, 1957 CG NC Cumberland Fayetteville 35°02.47'N 078°54.82'W R. Wilbur 5815 May 17, 1957 CG NC Cumberland Fayetteville 34°58.72'N 078°43.94'W R. Wilbur 49467 July 29, 1988 CG NC Cumberland Jessup Pond 34°51.72'N 078°43.78'W R. Wilbur et al 43146 July 22, 1987 CG NC Cumberland Wade 35°06.77'N 078°46.28'W A. Kelly 112 June 18, 1989 CG NC Harnett Cameron pond 35°15.31'N 078°57.63'W R. Wilbur 54971 May 19, 1990 CG NC Harnett Olivia 35°21.25'N 079°06.54'W R. Wilbur 9116 June 29, 1967 CG NC Harnett Spout Springs 35°14.34'N 079°01.29'W R. Wilbur 55650 June 13, 1990 CG NC Harnett Spring Lake 35°15.35'N 079°02.27'W A. Kelly N/A September, 1987 CG NC Hoke Aberdeen 35°04.09'N 079°21.53'W R. Wilbur 61172 July 8, 1992 CG NC Hoke Dundarrach 34°59.31'N 079°05.82'W R. Wilbur 59170 June 8, 1991 CG NC Hoke McCain 35°03.40'N 079°20.89'W A. Kelly N/A July 10, 1988 CG NC Hoke Montrose 35°02.69'N 079°20.22'W D. Stone et al 2739 May 28, 1970 CG NC Hoke Raeford 35°00.95'N 079°18.72'W R. Wilbur 61105 July 8, 1992 CG NC Hoke Raeford 34°51.98'N 079°12.05'W A. Kelly N/A April 3, 1988 CG NC Moore Aberdeen 35°06.52'N 079°22.90'W H. Hespenheide 326 May 12, 1964 CG NC Moore Aberdeen 35°06.52'N 079°25.36'W D. Stone et al 2738 May 28, 1970 CG NC Moore Aberdeen 35°06.87'N 079°23.21'W R. Wilbur 55514 June 6, 1990 CG NC Moore Aberdeen 35°09.54'N 079°28.55'W A. Kelly N/A June 25, 1988 CG NC Moore Carthage 35°20.92'N 079°21.88'W D. Correll 7010 October 24, 1936

167

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CG NC Moore Lakeview 35°14.52'N 079°22.35'W D. Stone 2671 May 29, 1968 CG NC Moore Lakeview 35°14.69'N 079°20.62'W D. Stone 2737 May 28, 1970 CG NC Moore Lakeview 35°13.85'N 079°20.53'W D. Stone 2670 May 29, 1968 CG NC Moore Mt. Pleasant 35°11.59'N 079°10.45'W R. Wilbur 55698 June 13, 1990 CG NC Moore Pinehurst 35°14.67'N 079°26.72'W J. Luteyn 2626 July 24, 1971 CG NC Moore Pinehurst 35°13.87'N 079°23.61'W R. Wilbur 10405 May 28, 1968 CG NC Moore Pinehurst 35°11.73'N 079°23.73'W H. Oosting 34739 October 7, 1934 CG NC Moore Pinehurst 35°14.13'N 079°26.89'W R. Wilbur 41746 May 29, 1987 CG NC Moore West End 35°14.13'N 079°34.39'W D. Stone 2672 May 29, 1968 CG NC New Wilmington 34°19.63'N 077°59.85'W E. Myers 2 July 11, 1970 Hanover CG NC New Wilmington 34°19.77'N 077°59.95'W R. Wilbur 6147 June 12, 1957 Hanover CG NC Pender Atkinson 34°29.79'N 078°11.54'W A. Kelly 98 May 5, 1988 CG NC Pender Holly Shelter Game land 34°24.61'N 077°39.82'W B. Sorroe 8415 June 8, 1995 CG NC Pender Moores Creek National 34°27.21'N 078°07.22'W D. Sieren 3444 June 6, 1983 Battlefield CG NC Richmond Camp Mackall 35°01.65'N 079°30.5'W A. Radford 14373 July 24, 1956 CG NC Richmond Hamlet 34°50.40'N 079°45.54'W A. Kelly N/A June 26, 1988 CG NC Richmond Hampstead 34°49.81'N 079°50.50'W A. Kelly N/A June 25, 1988 CG NC Richmond Hoffman 35°03.52'N 079°29.71'W R. Wilbur 62435 June 1, 1994 CG NC Richmond Hoffman 35°01.21'N 079°37.23'W A. Kelly N/A April 27, 1989 CG NC Richmond Rockingham 34°58.00'N 079°39.09'W D. Correll 1066 June 17, 1935 CG NC Richmond Rockingham 34°57.09'N 079°44.20'W R. Wilbur 41719 May 29, 1987 CG NC Robeson Lumberton 34°37.47'N 079°01.02'W D. Correll 7393 June 3, 1937 CG NC Robeson Red Springs 34°47.00'N 079°11.75'W H. Ahles et al 28914A June 21, 1957 CG NC Robeson St. Pauls 34°47.85'N 078°55.87'W D. Stone 2741 May 28, 1970 CG NC Sampson Clinton 34°59.09'N 078°24.64'W R. Godfrey 4516 June 11, 1938 CG NC Sampson Ingold 34°46.03'N 078°21.49'W H. Ahles et al 30151 June 28, 1957 CG NC Sampson Newton Grove 35°10.78'N 078°28.64'W A. Kelly N/A April 17, 1989 CG NC Sampson Roseboro 34°56.05'N 078°30.97'W R. Wilbur 17823 June 6, 1974

168

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CG NC Sampson Roseboro 34°52.32'N 078°31.98'W R. Wilbur 49062 July 21, 1988 CG NC Sampson Roseboro 34°53.02'N 078°31.36'W A. Kelly N/A April 17, 1989 CG NC Sampson White Lake 34°40.32'N 078°26.23'W D. Correll 1252 June 20, 1935 CG NC Scotland Camp Mackall 35°00.84'N 079°26.71'W R. Wilbur 55631 June 12, 1990 CG NC Scotland Gibson 34°48.08'N 079°38.09'W A. Kelly N/A June 26, 1988 CG NC Scotland Hoffman 35°01.54'N 079°33.06'W R. Wilbur 62093 May 19, 1995 CG NC Scotland Hoffman 35°01.54'N 079°33.06'W R. Wilbur 62103 May 19, 1994 CG NC Scotland Hoke-Scotland county line 34°59.70'N 079°26.8'W H. Ahles et al 28569 June 20, 1957 on US 15-501 CG NC Scotland Sandhills Gameland 35°00.18'N 079°32.28'W A. Kelly N/A May 19, 1988 CG NC Scotland Sandhills Gameland 34°59.89'N 079°32.67'W A. Kelly N/A June 2, 1988 CG NC Scotland Sandhills Gameland 34°53.74'N 079°27.03'W R. Wilbur 46818 May 20, 1988 Management Area CG NC Scotland Sandhills Gameland 34°55.64'N 079°30.12'W R. Wilbur 62112 May 19, 1994 CG NC Scotland Wagram 34°55.25'N 079°23.13'W A. Kelly N/A June 2, 1988 CG NC Wayne Dudley 35°17.45'N 078°01.04'W J. Duke 1045 June 11, 1958 CG NC Wayne Goldsboro 35°17.24'N 077°50.96'W A. Radford 22002 May 3, 1957 CG NC Wayne Mt. Olive 35°18.71'N 078°02.17'W E. Myers 1 July 11, 1970 CG SC Aiken Bishop Gravatt Center 33°44.25'N 081°35.12'W D. Johnson 533 May 16, 1995 CG SC Aiken Foley farm 33°44.45'N 081°36.17'W T. Callahan 681404- April 14, 1968 1 CG SC Aiken New Ellenton 33°23.45'N 081°34.12'W J. Duke 620 May 16, 1958 CG SC Allendale Smith’s lake 33°02.28'N 081°28.29'W C. Aulbach Smith 2325 July 9, 1982 CG SC Allendale SR 102, 0.85 mi S. of Jct. 33°03.68'N 081°28.82'W A. Kelly 105 June 4, 1989 with SR 125 CG SC Bamberg Bamberg 33°19.03'N 080°59.21'W C. Horn 4412 May 28, 1991 CG SC Bamberg Denmark 33°18.74'N 081°11.44'W H. Ahles et al 25977 May 26, 1957 CG SC Barnwell Kline 33°09.55'N 081°15.55'W H. Ahles et al 57077 June 10, 1962 CG SC Barnwell Martin 33°04.49'N 081°28.70'W W. Kelly et al s. n. July 11, 1952 CG SC Calhoun Lone star 33°37.88'N 080°33.60'W J. Nelson et al 3352 May 16, 1984 CG SC Chesterfield Bethune 34°26.16'N 080°17.57'W A. Kelly N/A July 10, 1988

169

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CG SC Chesterfield Cheraw 34°37.27'N 079°56.67'W C. Horn et al 2583 May 13, 1988 CG SC Chesterfield Cheraw 34°38.41'N 079°57.66'W A. Kelly N/A July 9, 1988 CG SC Chesterfield Cheraw State Park 34°37.26'N 079°56.70'W J. Bozeman et al 9182 May 25, 1967 CG SC Chesterfield Chesterfield 34°42.25'N 080°14.99'W A. Radford 12358 June 5, 1956 CG SC Chesterfield Hartsville 34°27.66'N 080°2.09'W A. Kelly N/A June 26, 1988 CG SC Chesterfield Patrick 34°34.90'N 080°01.93'W A. Kelly N/A July 10, 1988 CG SC Chesterfield Patrick 34°38.19'N 080°05.22'W A. Kelly N/A July 10, 1988 CG SC Chesterfield Sand Hills State Forest 34°29.89'N 080°01.24'W J. Castrale 144 June 19, 1976 CG SC Clarendon St. Paul 33°31.60'N 080°25.57'W A. Radford 24489 June 11, 1957 CG SC Darlington Black Creek 34°24.74'N 079°59.50'W C. Horn 2748 June 7, 1988 CG SC Darlington Darlington 34°18.59'N 079°50.28'W B. Smith 310 June 10, 1940 CG SC Darlington Hartsville 34°22.87'N 080°06.63'W B. Smith 304 June 10, 1940 CG SC Darlington Hartsville 34°21.31'N 080°07.27'W J. Bozeman et al 9183 May 25, 1967 CG SC Darlington Hartsville 34°22.64'N 080°05.91'W J. Norton s. n. May 15, 1920 CG SC Darlington Patrick 34°27.65'N 080°02.07'W J. Duke 1836 August 3, 1958 CG SC Dillon Little Rock 34°33.19'N 079°27.11'W H. Ahles et al 32211 July 25, 1957 CG SC Dorchester St. George 33°05.43'N 080°29.50'W H. Ahles et al 26352 May 27, 1957 CG SC Georgetown Plantersville 33°30.60'N 079°15.14'W A. Kelly 104 June 3, 1989 CG SC Georgetown Sandy Island 33°35.20'N 079°07.59'W J. Nelson et al 21829 April 14, 2001 CG SC Georgetown Tom Yawkey Wildlife 33°15.14'N 079°15.42'W J. Nelson 9452 July 26, 1990 Center CG SC Georgetown Tom Yawkey Wildlife 33°17.50'N 079°10.36'W J. Nelson et al 9931 October 24, Center 1990 CG SC Horry Cotton patch bay 33°47.18'N 078°56.43'W W. Batson et al s. n. June 15, 1987 CG SC Horry Sandhill Lake 34°07.02'N 079°13.23'W A. Pittman et al 5199304 May 19, 1993 CG NC Horry Wampee 33°49.71'N 078°45.14'W R. Wilbur 73779 June 8, 2001 CG SC Horry Watts Bay 33°49.42'N 078°50.45'W J. Nelson et al 7946 June 12, 1989 CG SC Jaspar Tillman Sand Ridge 32°29.71'N 081°11.59'W J. Nelson 20649 July 23, 1999 Heritage CG SC Kershaw Camden 34°24.19'N 080°22.97'W C. Horn 2775 June 22, 1988 CG SC Kershaw Camden 34°10.29'N 080°47.32'W A. Kelly N/A July 10, 1988

170

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CG SC Kershaw Sanders Creek 34°19.99'N 080°33.40'W A. Pittman et al 6259814 June 6, 1998 CG SC Kershaw Savage Bay Heritage 34°19.63'N 080°31.34'W C. Wood 33 August 3, 2000 Preserve CG SC Kershaw SR 15, about 17.5 mi NE. 34°12.37'N 080°16.03'W A. Kelly 114 July 13, 1989 of US 1 CG SC Lancaster Kershaw 34°34.48'N 080°32.26'W H. Ahles et al 27421 June 6, 1957 CG SC Lee Bishopville 34°15.52'N 080°10.05'W A. Radford 24296 June 6, 1957 CG SC Lexington Batesburg 33°53.82'N 081°32.23'W J. Duke 721 May 16, 1958 CG SC Lexington Edmund 33°50.67'N 081°13.87'W S. Hutto 206 August 19, 1986 CG SC Lexington Highway 602 33°56.65'N 081°08.64'W J. Barry s. n. April 30, 1967 CG SC Lexington Pelion 33°43.64'N 081°12.98'W C. Leland Rodgers 227 May 20, 1972 CG SC Lexington US 321 33°55.23'N 081°03.86'W S. Hutto et al 14 August 5, 1982 CG SC Marion Brittons Neck 33°45.73'N 079°15.73'W J. Nelson et al 25853 May 17, 2006 CG SC Marlboro Bennettsville 34°47.72'N 079°42.5'W C. Horn et al 2678 June 1, 1988 CG SC Marlboro Clio 34°43.43'N 079°35.88'W A. Kelly N/A July 10, 1988 CG SC Marlboro Drake 34°29.56'N 079°38.45'W A. Radford 12543 June 10, 1956 CG SC Orangeburg Orangeburg 33°30.02'N 080°56.29'W H. Ahles 25254 May 10, 1957 CG SC Richland Ft. Jackson 34°04.35'N 080°49.19'W J. Nelson et al 12697 June 1, 1992 CG SC Richland Sesquicentennial State 34°05.93'N 080°54.49'W M. Moody 2 June 10, 1991 Park CG SC Richland Sesquicentennial State 34°05.77'N 080°54.52'W W. Cely 67 May 22, 2007 Park CG SC Richmond Faunus Road 34°08.24'N 081°02.30'W J. Nelson 22756 June 11, 2002 CG SC Richmond SC Army National Guard 33°59.90'N 080°49.80'W J. Nelson et al 12447 May 13, 1992 HQ CG SC Saluda Ridge Spring 33°48.83'N 081°39.08'W A. Radford 23153 May 26, 1957 CG SC Sumter Poinsett State Park 33°48.22'N 080°30.91'W S. Tandon s. n. May 4, 1964 CG SC Sumter Sumter 33°50.39'N 080°22.77'W A. Radford 24050 June 5, 1957 CG VA Southampton Franklin 36°34.74'N 076°56.40'W D. Ware et al 7005 May 13, 1978 CG VA Southampton Franklin 36°35.05'N 076°56.22'W A. Harvill 17571 September 28, 1967 CG VA Southampton Franklin 36°34.55'N 076°56.56'W L. Smith et al s. n. June 25, 1940

171

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CG VA Suffolk City Suffolk City 36°35.29'N 076°54.42'W H. Ahles et al 58184 June 22, 1963 Callisia ornata (Small) Tucker CO FL Brevard Eau Gallie 28°10.22'N 080°38.98'W A. Rhoads s. n. March 16, 1937 CO FL Brevard Melbourne 28°04.97'N 080°38.54'W S. Hood s. n. June 16, 1951 CO FL Brevard Melbourne Beach 28°03.16'N 080°33.09'W S. Hopkins s. n. July 14, 1896 CO FL Brevard Wickham Park 28°09.47'N 080°39.64'W T. MacClendon et 3 April 4, 2002 al CO FL Broward Carpenter Scrub Natural 26°09.74'N 080°10.66'W P. Howell 687 June 20, 2003 area CO FL Broward Ft. Lauderdale 26°08.85'N 080°11.13'W W. Coker s. n. February 18, 1946 CO FL Charlotte Charlotte Harbor Preserve 26°50.40'N 082°13.26'W E. Gandy ch0022 October 18, State Park 2007 CO FL Charlotte Prairie/Shell Creek 26°59.28'N 081°57.76'W A. Franck 2002 May 6, 2010 CO FL Citrus Fort Cooper State Park 28°48.93'N 082°18.31'W F. Hattaway 114 May 12, 2000 CO FL Collier Bonita Springs 26°19.69'N 081°47.47'W R. Huck et al 3958a March 30, 1986 CO FL Collier Bonita Springs 26°18.68'N 081°47.72'W R. Huck et al 3958b March 30, 1986 CO FL Collier Bonita Springs 26°19.51'N 081°46.72'W R. Huck et al 3958c March 30, 1986 CO FL Collier Marco Island 25°56.37'N 081°42.66'W O. Lakela 30830 July 1, 1967 CO FL Collier Marco Island 25°55.91'N 081°41.73'W E. West et al s. n. April 15, 1954 CO FL Desoto Deep Creek Preserve 27°03.47'N 082°00.39'W A. Franck 761 July 31, 2008 CO FL Desoto Deep Creek Preserve 27°03.36'N 082°00.70'W A. Franck 2158 May 27, 2010 CO FL Hardee Cattle Range Station 27°23.78'N 081°56.41'W B. Kirk s. n. July 1, 1942 CO FL Hardee Paynes Creek Historic 27°37.36'N 081°48.52'W E. Gandy PC0142 July 5, 2007 State Park CO FL Hardee S. of Fort Green Springs 27°35.13'N 081°56.31'W B. Hansen et al 12442 November 9, 1993 CO FL Hernando Dade City 28°23.24'N 082°11.65'W H. Davis 15628 May 5, 1971 CO FL Highlands Amtrak Station 27°29.75'N 081°26.06'W D. Hall et al 1383 May 16, 1985 CO FL Highlands Archbold Biological Station 27°11.05'N 081°21.00'W S. Bergamo 99-205 June 10, 1999 CO FL Highlands Avon Park 27°33.98'N 081°24.68'W K. Delaney 1630 April 25, 1988 CO FL Highlands Avon Park Lakes 27°36.60'N 081°33.31'W S. Christman 1742 May 16, 1987

172

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CO FL Highlands Desoto Rd., E. of Sebring 27°28.31'N 081°23.63'W S. Christman 413 April 22, 1986 CO FL Highlands Highlands Hammock State 27°28.26'N 081°31.14'W A. Bishop HH0165 March 17, 2003 Park CO FL Highlands Highlands Park Estates 27°18.25'N 081°19.55'W S. Christman 1655 April 5, 1987 CO FL Highlands Lake June in Winter State 27°17.82'N 081°25.15'W D. Stark LJ64 April 26, 2001 Park CO FL Highlands Sebring 27°28.39'N 081°31.87'W J. Mcfarlin 8358 August 6, 1934 CO FL Highlands Sebring 27°29.71'N 081°26.47'W J. Mcfarlin 10137 February 6, 1935 CO FL Highlands Sebring 27°29.84'N 081°29.37'W C. Bryson 17801 April 5, 2000 CO FL Highlands Venus ESE Scrub 27°04.47'N 081°21.33'W S. Christman 1630 April 4, 1987 CO FL Hillsborough Balm Boyette Tract 27°46.75'N 082°14.83'W S. Landry et al s. n. September 23, 1994 CO FL Hillsborough Little Manatee River State 27°39.88'N 082°22.59'W J. Myers 30 May 20, 1998 Park CO FL Hillsborough Little Manatee River State 27°39.66'N 082°22.17'W J. Myers 128 September 9, Park 1998 CO FL Hillsborough Little Manatee River State 27°39.67'N 082°22.28'W J. Myers 393 June 10, 1999 Park CO FL Hillsborough Tampa 27°56.83'N 082°27.52'W A. Garber s. n. May, 1876 CO FL Hillsborough Tarpon Rd. 28°08.56'N 082°36.83'W V. Ducey 16 June 17, 1962 CO FL Hillsborough University South Florida 28°01.36'N 082°32.34'W O. Lakela 30811 June 29, 1967 Campus CO FL Indian River St. Sebastian River State 27°49.61'N 080°33.72'W J. Scanlon 414 May 13, 2002 Buffer Preserve CO FL Lake 0.8 mi W. of U.S. 441, 6 mi 28°53.78'N 081°53.85'W L. Baltzell 7172 April 13, 1975 N. of Leesburg. CO FL Lake E. of Grassy Lake, 1.5 mi 28°35.74'N 081°43.89'W D. Ward 8346 May 27, 1972 NE. of Minneola CO FL Lake Guerrant park 28°54.99'N 081°39.87'W W. Murrill s. n. May 7, 1941 CO FL Lake Teardrop Lake 28°37.10'N 081°43.48'W S. Christman 1519 March 17, 1987 CO FL Lee Bonita Beach 26°19.88'N 081°46.54'W E. Brown s. n. May 14, 1985 CO FL Manatee Bradenton 27°29.43'N 082°35.55'W A. Cuthbert s. n. June 17, 1917 CO FL Manatee Bradenton 27°29.43'N 082°35.55'W A. Cuthbert s. n. June 17, 1916

173

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CO FL Manatee Emerson Point Preserve 27°31.28'N 082°34.34'W G. Weber 758 May 7, 1932 CO FL Manatee South Fork State Park 27°36.75'N 082°14.31'W R. Owens SF0112 June 13, 1992 CO FL Manatee Wingate Creek State Park 27°27.11'N 082°08.88'W J. Weber WC0057 April 11, 1991 CO FL Manatee Wingate Creek State Park 27°27.41'N 082°08.64'W C. Becker WC0223 August 13, 1997 CO FL Marion Florida Greenways and 28°59.45'N 082°11.07'W S. Carr 3648 September 8, Trails 2001 CO FL Martin Jonathan Dickinson State 26°59.53'N 080°08.84'W D. Correll et al 49882 June 17, 1978 Park CO FL Martin Jonathan Dickinson State 26°59.56'N 080°08.94'W R. Woodbury et al s. n. April 1, 1990 Park CO FL Orange 6 mi N. of Orlando 28°55.00'N 081°39.93'W W. Murrill s. n. May 7, 1941 CO FL Orange Conway 28°30.16'N 081°19.84'W E. West et al s. n. July 23, 1929 CO FL Osceola Along Poinciana Blvd 28°09.53'N 081°26.71'W B. Hansen et al 11746 June 16, 1988 CO FL Osceola Buck Lake South Scrub 28°11.23'N 081°08.85'W S. Christman 356 April 9, 1986 CO FL Osceola Bull Creek Wildlife 28°02.42'N 080°56.86'W C. Lotspeich 445 April 25, 1978 management Area CO FL Osceola E. of FL 530 28°20.33'N 081°35.05'W D. Ward 1876 May 4, 1960 CO FL Osceola Jane Green Creek on 28°03.71'N 080°53.55'W R. Huck 46 June 29, 1976 Kempher Rd. Deer Park CO FL Osceola Yeehaw 27°41.68'N 080°54.02'W S. Hood 4311 June 20, 1951 CO FL Palm Beach Palm Beach Gardens 26°50.70'N 080°05.50'W P. Cassen 516 August 21, 1969 CO FL Pasco Eagle Point Park/Pasco 28°13.49'N 082°44.99'W S. Hood 3512 May 17, 1950 Palms Preserve CO FL Pasco Starkey Wilderness Park 28°15.04'N 082°36.54'W E. Ferguson 75 February 5, 2003 CO FL Pasco Starkey Wilderness Park 28°14.66'N 082°36.45'W E. Ferguson 138 June 17, 2003 CO FL Pinellas Brooker Creek Preserve 28°08.70'N 082°39.19'W B. Hansen et al 12841 June 7, 1994 CO FL Pinellas Clearwater, Tampa 27°59.69'N 082°44.50'W P. Genelle et al 316 September 7, 1970 CO FL Pinellas Clearwater, Tampa 28°00.47'N 082°47.39'W M. Lelong 5349 April 4, 1970 CO FL Pinellas Curlew Rd. (County 28°02.80'N 082°45.26'W B. Lambert s. n. June 15, 2005 preserve area) CO FL Pinellas Safety Harbor Fire station 27°59.68'N 082°41.27'W G. Fleming 4041 May 7, 1988 #2

174

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CO FL Pinellas Safety Harbor Fire Station 27°58.78'N 082°42.45'W G. Fleming 4114 August 7, 1988 #2 CO FL Polk Davenport 28°09.82'N 081°35.35'W S. Hood 3532 May 22, 1950 CO FL Polk Lakeland 27°45.68'N 081°59.57'W W. Faircloth 7312 May 18, 1973 CO FL Polk Saddle Blanket Scrub 27°39.24'N 081°34.07'W P. Corogin SB288 April 8, 2007 Preserve CO FL Polk Scrub at Arnold Hammock 27°42.80'N 081°23.16'W C. van Hoek et al 590 June 29, 1996 CO FL Polk Tiger Creek Preserve 27°48.73'N 081°29.33'W P. Corogin TC144 July 8, 2005 CO FL Polk 1.3 mi N. of Jct. Fla 630, on 27°45.87'N 081°59.51'W W. Faircloth 7312 May 18, 1973 Fla 37 CO FL Putnam Dunns Creek State Park 29°33.30'N 081°35.09'W B. Herring et al 1536 June 26, 2002 CO FL Sarasota Brohard Park 27°04.09'N 082°26.83'W M. Ferguson et al 89-5 April 14, 1989 CO FL Sarasota Dee prairie Creek Preserve 27°07.36'N 082°14.71'W L. Birch et al 238 June 12, 2009 CO FL Sarasota Old Miakka Preserve 27°20.02'N 082°15.75'W A. Franck 891 September 11, 2008 CO FL Sarasota Oscar Scherer State Park 27°10.56'N 082°27.60'W J. Weber et al OS0067 July 7, 1992 CO FL Sarasota Oscar Scherer State Park 27°10.87'N 082°27.45'W D. Griffin OS0550 April 14, 2006 CO FL Seminole Altamonte Springs 28°40.14'N 081°26.77'W L. Baltzell 1062 May 2, 1969 CO FL Seminole Sanford 28°47.44'N 081°16.44'W A. Bitting s. n. July 20, 1894 CO FL St. Lucie Savannas State Reserve. 27°20.34'N 080°18.68'W M. Garland et al 950 September 23, 1992 CO FL Volusia West of Daytona 29°14.55'N 081°13.75'W L. Baltzell 4873 April 3, 1973 CO GA Charlton 6 mi N. of intersection with 30°26.48'N 082°11.89'W G. Norsworthy et al s. n. May 28, 1972 Hwy 23-121 Callisia rosea (Vent.) Hunt CR FL Duval Near Jacksonville 30°19.91'N 081°39.35'W A. Curtiss 2998 May CR FL Highlands Brighton 27°13.53'N 081°05.72'W A. Moldenke et al 29752 May 16, 1975 CR GA Candler Upper Lotts Creek 32°29.32'N 081°58.72'W N. Coile et al 2727 May 6, 1982 CR GA Chatham Pooler 32°03.22'N 081°21.16'W W. Duncan et al 17332 Augustus 24, 1953 CR GA Chatham Savannah City Hall 31°59.75'N 081°16.17'W W. Duncan 20900 June 10, 1958 CR GA Coffee Douglas 31°31.73'N 082°49.23'W R. Carter et al 4179 June 28, 1985 CR GA Columbia Appling 33°33.32'N 082°14.98'W J. Pyron et al 1682 May 23, 1937

175

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CR GA Elbert Gregg Shoals area 34°12.84'N 082°44.76'W N. Coile et al 2254 May 23, 1981 CR GA Elbert Savannah River 34°03.38'N 082°38.05'W W. Credle et al 813 June 5, 1979 CR GA Elbert Savannah River 34°03.36'N 082°37.60'W W. Credle et al 828 June 5, 1979 CR GA Elbert Tallow Hill 34°06.73'N 083°01.83'W N. Coile et al 1807 May 27, 1978 CR GA Evans Fort Stewart 32°06.73'N 081°47.71'W T. Zebryk et al 430 June 25, 1992 CR GA Greene White plains 31°31.73'N 082°49.23'W J. Allison 2734 May 2, 1987 CR GA Hancock Sparta 33°13.52'N 082°58.61'W W. Duncan et al 30879 June 14, 1983 CR GA Hart Savannah River 34°18.15'N 082°47.70'W W. Credle et al 776 May 29, 1979 CR GA Houston Oaky Woods WMA 32°29.44'N 083°32.50'W P. Lynch et al 258 May 14, 2009 CR GA Jones Union Hill Church 33°10.37'N 083°31.68'W R. Campbell 12 July 11, 1970 CR GA Laurens Ben Hall's Lake 32°38.66'N 082°51.56'W C. Lodge et al s. n. May 9, 1977 CR GA Liberty Fort Stewart 31°51.91'N 081°33.56'W T. Zebryk et al 549 August 6, 1992 CR GA Morgan Hard Labor Creek State 33°39.26'N 083°36.23'W J. Hill et al 1359 May 25, 1980 Park CR GA Oglethorpe Lexington 33°58.31'N 083°00.38'W F. Montgomery et 203 July 14, 1965 al CR GA Putnam Oconee river 33°17.77'N 083°13.75'W J. Allison 2153 May 12, 1984 CR GA Richmond Fort Gordon 33°23.46'N 082°14.42'W M. Moore et al 1490 May 7, 1992 CR GA Twiggs Griswoldville, Butts Tract 32°52.20'N 083°26.83'W T. Govus 762 June 5, 1996 CR GA Twiggs Ocmulgee W.M.A. 32°34.46'N 083°28.61'W P. Lynch et al 428 July 18, 2009 CR GA Washington Tennille 32°57.05'N 082°49.57'W W. Duncan 13509 May 11, 1952 CR GA Wilkes Danburg 33°52.76'N 082°38.11'W J. Allison 2778 September 21, 1986 CR NC Cleveland Boiling Springs 35°12.01'N 081°39.94'W W. Fox 3929 July 12, 1950 CR NC Cleveland Broad River 35°12.08'N 081°39.83'W O. Freeman 56268 May 31, 1956 CR NC Cumberland Fayetteville 34°52.86'N 078°50.85'W J. Duke 1214 June 21, 1958 CR NC New Wilmington 33°53.49'N 078°1.48'W D. Sieren 1274 May 21, 1975 Hanover CR SC Abbeville Russell Dam 34°03.49'N 082°37.20'W C. Douglas et al 1424 May 23, 1978 CR SC Aiken Barnwell 33°17.72'N 081°43.93'W N. Coile et al 5256 April 30, 1989 CR SC Aiken Hamburg 33°28.95'N 081°56.89'W A. Radford 548 June 20, 1940

176

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CR SC Aiken North Augusta 33°28.82'N 081°57.04'W H. Ahles et al 56826 June 9, 1962 CR SC Allendale Barton 32°52.04'N 081°24.44'W C. Bell 4031 June 30, 1956 CR SC Allendale Brier Creek 32°57.53'N 081°27.78'W C. Bell 2704 May 13, 1956 CR SC Allendale Luray 32°49.50'N 081°17.08'W J. Nelson et al 5745 July 9, 1987 CR SC Allendale Smith Lake 33°01.37'N 081°30.07'W A. Pittman et al 5309706 May 30, 1997 CR SC Anderson Russell Dam 34°14.96'N 082°44.08'W W. Credle et al 759 May 25, 1979 CR SC Bamberg Olar 33°06.49'N 081°09.40'W H. Ahles et al 30560 July 4, 1957 CR SC Bamberg SC 70 33°18.64'N 081°11.57'W D. Rayner 2229 May 16, 1985 CR SC Barnwell Williston 33°23.85'N 081°27.52'W H. Ahles et al 56977 June 10, 1962 CR SC Beaufort Yemassee 32°41.14'N 080°49.70'W C. Bell 3743 June 27, 1956 CR SC Beaufort Yemassee 32°40.36'N 080°46.78'W J. Nelson et al 28347 May 10, 2010 CR SC Berkeley Lake Moultrie 33°20.68'N 080°08.58'W H. Ahles et al 26623 May 28, 1957 CR SC Berkeley Moncks Corner 33°10.81'N 080°11.98'W H. Ahles et al 30695 July 5, 1957 CR SC Calhoun Cameron 33°35.94'N 080°34.38'W H. Ahles et al 25515 May 19, 1957 CR SC Charleston Parkers Ferry 32°43.63'N 080°22.77'W H. Ahles et al 25771 May 20, 1957 CR SC Cherokee Gaffney 35°07.93'N 081°35.38'W H. Ahles et al 26811 June 4, 1957 CR SC Chester Broad River 34°35.70'N 081°25.39'W C. Bell 9532 June 27, 1957 CR SC Chester Carlisle 34°35.70'N 081°25.39'W O. Freeman 56259 May 26, 1956 CR SC Colleton Combahee River 32°41.00'N 080°39.63'W C. Bell 2325 May 8, 1956 CR SC Dorchester Leveston Bluff 32°54.16'N 080°23.04'W S. Hill 19325 May 14, 1988 CR SC Edgefield Sumter National Forest 33°39.06'N 082°04.38'W J. Nelson 18173 May 4, 1997 CR SC Fairfield Between Shelton and 34°32.61'N 081°17.95'W C. Bell 9310 June 26, 1957 Woodward CR SC Fairfield Jenkinsville 34°17.50'N 081°18.17'W J. Nelson et al 22688 May 31, 2002 CR SC Fairfield S 225 34°23.60'N 081°14.73'W S. Wooten et al 834 June 22, 1983 CR SC Georgetown Baruch plantation 33°20.56'N 079°14.10'W J. Barry 52 June 20, 1967 CR SC Georgetown Georgetown 33°25.70'N 079°08.83'W W. Coker s. n. June 16, 1941 CR SC Greenwood Hodges 34°17.49'N 082°11.72'W A. Radford 22907 May 25, 1957 CR SC Greenwood Ninety Six 34°07.91'N 082°01.52'W A. Radford 26619 July 7, 1957

177

Table A-1. Continued. Taxa State County Locality Latitude Longitude Collector Coll. # Date collected CR SC Hampton SC 119 32°37.22'N 081°14.68'W C. Aulbach- Smith 1522 July 15, 1981 et al CR SC Hampton Shirley 32°39.10'N 081°19.8'W C. Aulbach- Smith 2299 June 28, 1982 et al CR SC Hampton Yemassee 32°43.14'N 080°53.54'W C. Bell 2527 May 11, 1956 CR SC Horry Myrtle Beach 33°38.97'N 078°55.79'W W. Coker s. n. June 23, 1931 CR SC Horry North of Myrtle Beach 32°52.14'N 079°46.82'W C. Bell 7689 May 1, 1957 CR SC Jasper County Rd. 21 32°35.18'N 080°44.66'W H. Ahles et al 12273 May 9, 1956 CR SC Kershaw Sanders Creek 34°21.05'N 080°32.41'W A. Pittman et al 5229816 May 22, 1998 CR SC Lancaster Rocky Creek 34°32.85'N 080°50.44'W H. Ahles 27368 June 6, 1957 CR SC Laurens Reedy River Dam 34°27.39'N 082°11.69'W C. Bell 8095 June 3, 1957 CR SC Marion Brittons Neck 30°19.91'N 081°39.35'W C. Bell 7816 May 26, 1957 CR SC McCormick Clarks Hill 27°13.53'N 081°05.72'W A. Radford 34760 June 8, 1958 CR SC Newberry Mills Creek 32°29.32'N 081°58.72'W C. Bell 7016 May 11, 1957 CR SC Orangeburg Branchville 32°03.22'N 081°21.16'W H. Ahles et al 25405 May 18, 1957 CR SC Orangeburg McDougal 31°59.75'N 081°16.17'W A. Darr 1160 August 25, 2001 CR SC Orangeburg Rowesville 31°31.73'N 082°49.23'W J. Fairey 357 August 22, 1967 CR SC Richland Harbison 33°33.32'N 082°14.98'W K. Peterson 37 June 22, 1999 CR SC Richland Heathwood Columbia 34°12.84'N 082°44.76'W J. Chopwan s. n. May 12, 1939 CR SC Richland Wildcat Rd. 34°03.38'N 082°38.05'W J. Nelson 10983 August 7, 1991 CR SC Spartanburg Croft State Park 34°03.36'N 082°37.60'W C. Bell 8280 June 4, 1957 CR SC Union Buffalo 34°06.73'N 083°1.83'W J. Nelson 6021 August 26, 1987 CR SC Union Herbert 32°06.73'N 081°47.71'W C. Bell 8454 June 5, 1957 CR SC Union Sumter National Forest 31°31.73'N 082°49.23'W C. Horn 6816 June 9, 1993 CR SC Union Union 33°13.52'N 082°58.61'W J. Allison 3812 May 21, 1989 CR SC Williamsburg Black River 34°18.15'N 082°47.70'W J. Nelson 7676 May 19, 1989 CR SC Williamsburg Kingstree 32°29.44'N 083°32.50'W A. Radford 24872 June 12, 1957

178

APPENDIX B VOUCHERS USED WITH GENBANK NUMBERS

List of taxa sampled with taxonomic authorities followed by botanical garden number with collector cultivated and/or ploidy, locality, collector, collection number, herbarium acronym, and GenBank accession numbers for nrITS, trnL-F, rpL32, ycf1(317)-

(1867), ycf1(1897)-(3080), trnH-K, trnQ, trnS-trnG, trnG-trnR sequences generated in Chapter

3. Missing data for given region is listed as: —.

Callisia cordifolia (Sw.) Andiers. & Woodson, NMNH 1983-197, Faden, R. B.

83/37 Cult., —; U.S.A.: Florida, Volusia Co., Molgo 373 (FLAS), —, —, MG835009,

MH166415, MH166566, MG834853, —, MG835160, —; Callisia fragrans (Lindl.)

Woodson, NMNH 1998-281, Hunt 85-71, —, Mexico: Durango, Molgo 379 (FLAS), —,

—, MG835012, MH166417, MH166569, MG834856, —, MG835162, —; Callisia gentlei var. elegans (Alexander ex H. E. Moore) D. R. Hunt, NMNH 1993-090, Cult.

Tim Chapman, —; U.S.A.: Louisiana, Molgo 372 (FLAS), MG639916, —, MG835010, —

, MH166567, MG834854, —, MG835161, —; Callisia gentlei var. macdougallii

(Miranda) D. R. Hunt, NMNH 1981-074, McDougal s.n., —; Mexico: Oaxaca State,

Molgo 359 (FLAS), MG639919, —, MG835014, MH166419, MH166571, MG834858, —,

MG835164, MG834710; Callisia graminea (Small) G.C.Tucker, diploid: 2n = 2x = 12,

U.S.A.: South Carolina, Chesterfield Co., Molgo 249 (FLAS), MG639925, —,

MG835021, MH166427, MH166580, MG834867, MG834617, MG835173, MG834713;

C. graminea (individual 1), diploid: 2n = 2x = 12, U.S.A.: South Carolina, Chesterfield

Co., Molgo 250 (FLAS), MG639926, MG834543, MG835022, MH166428, MH166581,

MG834868, —, MG835174, MG834714; C. graminea (individual 2), tetraploid: 2n = 4x

= 24, U.S.A.: South Carolina, Chesterfield Co., Molgo 250 (FLAS), MG639927, —,

179

MG835023, MH166429, MH166582, MG834869, —, MG835175, MG834715; C. graminea (individual 3), diploid: 2n = 2x = 12, U.S.A.: South Carolina, Chesterfield Co.,

Molgo 250 (FLAS), MG639928, —, MG835024, MH166430, MH166583, MG834870, —,

MG835176, MG834716; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina,

Sampson Co., Molgo 280 (FLAS), MG639929, —, MG835025, MH166431, MH166584,

MG834871, MG834618, —, MG834717; C. graminea, diploid: 2n = 2x = 12, U.S.A.:

North Carolina, Wayne Co., Molgo 281 (FLAS), MG639930, MG834544, MG835026,

MH166432, MH166585, MG834872, MG834619, —, MG834718; C. graminea, diploid:

2n = 2x = 12, U.S.A.: Virginia, Southampton Co., Molgo 282 (FLAS), MG639931,

MG834545, MG835027, MH166433, MH166586, MG834873, MG834620, MG835177,

MG834719; C. graminea, diploid: 2n = 2x = 12, U.S.A.: Virginia, Suffolk City Co., Molgo

283 (FLAS), MG639932, MG834546, MG835028, MH166434, MH166587, MG834874,

MG834621, MG835178, MG834720; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North

Carolina, Harnett Co., Molgo 284 (FLAS), MG639933, MG834547, MG835029,

MH166435, MH166588, MG834875, MG834622, —, MG834721; C. graminea, diploid:

2n = 2x = 12, U.S.A.: North Carolina, Moore Co., Molgo 285 (FLAS), MG639934,

MG834548, MG835030, MH166436, MH166589, MG834876, MG834623, —,

MG834722; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North Carolina, Hoke Co.,

Molgo 286 (FLAS), MG639935, MG834549, MG835031, —, MH166590, MG834877,

MG834624, —, MG834723; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North Carolina,

Hoke Co., Molgo 287 (FLAS), MG639936, MG834550, MG835032, MH166437,

MH166591, MG834878, MH166410, —, MG834724; C. graminea, diploid: 2n = 2x =

12, U.S.A.: North Carolina, Moore Co., Molgo 288 (FLAS), MG639937, MG834551,

180

MG835033, MH166438, MH166592, MG834879, MG834625, —, MG834725; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North Carolina, Scotland Co., Molgo 289

(FLAS), MG639938, MG834552, MG835034, MH166439, MH166593, MG834880,

MG834626, —, MG834726; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North Carolina,

Richmond Co., Molgo 290 (FLAS), MG639939, MG834553, MG835035, MH166440,

MH166594, MG834881, MG834627, —, MG834727; C. graminea, diploid: 2n = 2x =

12, U.S.A.: North Carolina, Scotland Co., Molgo 291 (FLAS), MG639940, —,

MG835036, MH166441, MH166595, MG834882, MG834628, —, MG834728; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North Carolina, Scotland Co., Molgo 292

(FLAS), MG639941, MG834554, MG835037, MH166442, MH166596, MG834883,

MG834629, MG835179, MG834729; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North

Carolina, Scotland Co., Molgo 293 (FLAS), MG639942, MG834555, MG835038,

MH166443, MH166597, MG834884, MG834630, —, MG834730; C. graminea, diploid:

2n = 2x = 12, U.S.A.: North Carolina, Scotland Co., Molgo 294 (FLAS), MG639943,

MG834556, MG835039, MH166444, MH166598, MG834885, MG834631, —,

MG834731; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North Carolina, Richmond Co.,

Molgo 295 (FLAS), MG639944, MG834557, MG835040, MH166445, MH166599,

MG834886, MG834632, —, MG834732; C. graminea, diploid: 2n = 2x = 12, U.S.A.:

North Carolina, Richmond Co., Molgo 296 (FLAS), MG639945, MG834558, MG835041,

MH166446, MH166600, MG834887, MG834633, —, MG834733; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina, Richmond Co., Molgo 297 (FLAS),

MG639946, —, MG835042, MH166447, MH166601, MG834888, MG834634, —,

MG834734; C. graminea, diploid: 2n = 2x = 12, U.S.A.: South Carolina, Chesterfield

181

Co., Molgo 331 (FLAS), —, —, MG835043, MH166448, MH166602, MG834889,

MG834635, —, MG834735; C. graminea, diploid: 2n = 2x = 12, U.S.A.: South Carolina,

Chesterfield Co., Molgo 332 (FLAS), —, —, MG835044, MH166449, MH166603,

MG834890, —, —, MG834736; C. graminea, diploid: 2n = 2x = 12, U.S.A.: South

Carolina, Chesterfield Co., Molgo 333 (FLAS), MG639947, MG834559, MG835045,

MH166450, MH166604, MG834891, —, —, MG834737; C. graminea, diploid: 2n = 2x =

12, U.S.A.: North Carolina, Richmond Co., Molgo 336 (FLAS), MG835046, MH166451,

MH166605, MG834892, MG834636, MG835180, MG834738; C. graminea, diploid: 2n

= 2x = 12, U.S.A.: North Carolina, Richmond Co., Molgo 337 (FLAS), MG639948,

MG834560, MG835047, MH166452, MH166606, MG834893, MG834637, MG835181,

MG834739; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North Carolina, Richmond Co.,

Molgo 338 (FLAS), —, —, MG835048, MH166453, MH166607, MG834894, MG834638,

MG835182, MG834740; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North Carolina,

Richmond Co., Molgo 339 (FLAS), —, —, MG835049, MH166454, MH166608,

MG834895, MG834639, MG835183, MG834741; C. graminea, diploid: 2n = 2x = 12,

U.S.A.: North Carolina, Richmond Co., Molgo 340 (FLAS), —, —, MG835050,

MH166455, MH166609, MG834896, MG834640, MG835184, MG834742; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North Carolina, Cumberland Co., Molgo 341 (FLAS), —,

—, MG835051, MH166456, MH166610, MG834897, —, MG835185, MG834743; C. graminea, diploid: 2n = 2x = 12, U.S.A.: North Carolina, Cumberland Co., Molgo 342

(FLAS), —, —, MG835052, MH166457, MH166611, MG834898, MG834641,

MG835186, MG834744; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Alachua

Co., Molgo 223 (FLAS), MG639949, —, MG835053, MH166458, MH166612,

182

MG834899, MG834642, MG835187, MG834745; C. graminea, tetraploid: 2n = 4x = 24,

U.S.A.: Florida, Putnam Co., Molgo 224 (FLAS), MG639950, —, MG835054,

MH166459, MH166613, MG834900, —, MG835188, MG834746; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Clay Co., Molgo 225 (FLAS), MG639951, —,

MG835055, MH166460, MH166614, MG834901, —, MG835189, MG834747; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Citrus Co., Molgo 229 (FLAS), —,

—, MG835056, MH166461, MH166615, MG834902, —, MG835190, MG834748; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Marion Co., Molgo 230 (FLAS),

MG639952, MG834561, MG835057, MH166462, MH166616, MG834903, —,

MG835191, MG834749; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina,

Aiken Co., Molgo 231 (FLAS), MG639953, —, MH156548, MH166463, MH166617,

MG834904, —, MG835192, MG834750; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.:

South Carolina, Lexington Co., Molgo 232 (FLAS), —, —, MG835058, MH166464,

MH166618, MG834905, MG834643, MG835193, MG834751; C. graminea, tetraploid:

2n = 4x = 24, U.S.A.: South Carolina, Lexington Co., Molgo 233 (FLAS), MG639954,

MG834562, MG835059, MH166465, MH166619, MG834906, —, MG835194,

MG834752; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Alachua Co., Molgo

234 (FLAS), MG639955, MG834563, MG835060, MH166466, MH166620, MG834907,

—, MG835195, MG834753; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Georgia,

Richmond Co., Molgo 239 (FLAS), MG639956, MG834564, MG835061, MH166467,

MH166621, MG834908, MG834644, MG835196, MG834754; C. graminea, tetraploid:

2n = 4x = 24, U.S.A.: Georgia, Emanuel Co., Molgo 241 (FLAS), MG639957,

MG834565, MG835062, MH166468, MH166622, MG834909, MG834645, MG835197,

183

MG834755; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Georgia, Bulloch Co., Molgo

242 (FLAS), MG639958, —, MG835063, MH166469, MH166623, MG834910, —,

MG835198, MG834756; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Georgia, Evans

Co., Molgo 243 (FLAS), MG639959, MG834566, MG835064, MH166470, MH166624,

MG834911, —, MG835199, MG834757; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.:

South Carolina, Jasper Co., Molgo 247 (FLAS), MG639960, MG834567, MG835065,

MH166471, MH166625, MG834912, MG834646, MG835200, MG834758; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina, Dorchester Co., Molgo 248 (FLAS),

MG639961, MG834568, MG835066, MH166472, MH166626, MG834913, —,

MG835201, MG834759; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina,

Lee Co., Molgo 251 (FLAS), MG639962, MG834569, MG835067, —, MH166627,

MG834914, —, MG835202, MG834760; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.:

South Carolina, Chesterfield Co., Molgo 252 (FLAS), —, —, MG835068, MH166473,

MH166628, MG834915, —, MG835203, MG834761; C. graminea, tetraploid: 2n = 4x =

24, U.S.A.: South Carolina, Chesterfield Co., Molgo 253 (FLAS), MG639963, —,

MG835069, MH166474, MH166629, MG834916, MG834647, MG835204, MG834762;

C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina, Chesterfield Co., Molgo

254 (FLAS), MG639964, MG834570, MG835070, MH166475, MH166630, MG834917,

—, MG835205, MG834763; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South

Carolina, Chesterfield Co., Molgo 255 (FLAS), —, MG834571, MG835071, MH166476,

MH166631, MG834918, MG834648, MG835206, MG834764; C. graminea, tetraploid:

2n = 4x = 24, U.S.A.: South Carolina, Kershaw Co., Molgo 256 (FLAS), —, —,

MG835072, MH166477, MH166632, MG834919, —, MG835207, MG834765; C.

184

graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina, Richland Co., Molgo 257

(FLAS), MG639965, MG834572, MG835073, MH166478, MH166633, MG834920, —,

MG835208, MG834766; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina,

Richland Co., Molgo 258 (FLAS), —, —, MH156549, MH166479, MH166634, —, —,

MG835209, MG834767; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina,

Richland Co., Molgo 259 (FLAS), MG639966, —, MG835074, MH166480, MH166635,

MG834921, —, MG835210, MG834768; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.:

South Carolina, Richland Co., Molgo 260 (FLAS), MG639967, —, MG835075,

MH166481, MH166636, MG834922, —, MG835211, MG834769; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina, Bladen Co., Molgo 261 (FLAS),

MG639968, MG834573, MG835076, MH166482, MH166637, MG834923, —,

MG835212, MG834770; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina,

Bladen Co., Molgo 262 (FLAS), MG639969, —, MG835077, MH166483, MH166638,

MG834924, —, MG835213, MG834771; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.:

North Carolina, Bladen Co., Molgo 263 (FLAS), MG639970, MG834574, —, MH166484,

MH166639, MG834925, —, MG835214, MG834772; C. graminea, tetraploid: 2n = 4x =

24, U.S.A.: North Carolina, Pender Co., Molgo 264 (FLAS), MG639971, MG834575,

MG835078, MH166485, MH166640, MG834926, MG834649, MG835215, MG834773;

C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina, Pender Co., Molgo 265

(FLAS), MG639972, MG834576, MG835079, MH166486, MH166641, MG834927, —,

MG835216, MG834774; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina,

Bladen Co., Molgo 266 (FLAS), MG639973, —, MG835080, MH166487, MH166642,

MG834928, MG834650, MG835217, MG834775; C. graminea, tetraploid: 2n = 4x = 24,

185

U.S.A.: North Carolina, Columbus Co., Molgo 267 (FLAS), MG639974, MG834577,

MG835081, MH166488, MH166643, MG834929, MG834651, MG835218, MG834776;

C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina, Bladen Co., Molgo 268

(FLAS), MG639975, MG834578, MG835082, —, MH166644, MG834930, —,

MG835219, MG834777; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina,

Bladen Co., Molgo 269 (FLAS), —, —, MG835083, MH166489, MH166645, MG834931,

—, MG835220, MG834778; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North

Carolina, Bladen Co., Molgo 270 (FLAS), MG639976, —, MG835084, MH166490,

MH166646, MG834932, —, MG835221, MG834779; C. graminea, tetraploid: 2n = 4x =

24, U.S.A.: North Carolina, Bladen Co., Molgo 271 (FLAS), MG639977, MG834579,

MG835085, MH166491, MH166647, MG834933, —, MG835222, MG834780; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina, Bladen Co., Molgo 272

(FLAS), MG639978, —, MG835086, MH166492, MH166648, MG834934, —,

MG835223, MG834781; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina,

Sampson Co., Molgo 273 (FLAS), MG639979, MG834580, MG835087, MH166493,

MH166649, MG834935, MG834652, MG835224, MG834782; C. graminea, tetraploid:

2n = 4x = 24, U.S.A.: North Carolina, Bladen Co., Molgo 274 (FLAS), —, —,

MG835088, MH166494, MH166650, MG834936, —, MG835225, MG834783; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina, Bladen Co., Molgo 275

(FLAS), MG639980, MG834581, MG835089, MH166495, MH166651, MG834937,

MG834653, MG835226, MG834784; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.:

North Carolina, Cumberland Co., Molgo 276 (FLAS), —, —, MG835090, MH166496,

MH166652, MG834938, —, MG835227, MG834785; C. graminea, tetraploid: 2n = 4x =

186

24, U.S.A.: North Carolina, Cumberland Co., Molgo 277 (FLAS), —, —, MG835091,

MH166497, MH166653, MG834939, —, MG835228, MG834786; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina, Cumberland Co., Molgo 278 (FLAS),

MG639981, MG834582, MG835092, MH166498, MH166654, MG834940, —,

MG835229, MG834787; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina,

Cumberland Co., Molgo 279 (FLAS), MG639982, —, MG835093, MH166499,

MH166655, MG834941, MG834654, —, MG834788; C. graminea, tetraploid: 2n = 4x =

24, U.S.A.: Florida, Putnam Co., Molgo 302 (FLAS), MG639983, —, MG835094,

MH166500, MH166656, MG834942, MG834655, —, MG834789; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Franklin Co., Molgo 306 (FLAS), MG639984, —

, MH156550, MH166501, MH166657, MG834943, MG834656, —, MG834790; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Alachua Co., Molgo 307 (FLAS),

MG639985, —, MG835095, MH166502, MH166658, MG834944, MG834657, —,

MG834791; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Putnam Co., Molgo

308 (FLAS), MG639986, —, MG835096, MH166503, MH166659, MG834945,

MG834658, —, MG834792; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Clay

Co., Molgo 309 (FLAS), MG639987, —, MG835097, MH166504, MH166660,

MG834946, MG834659, —, MG834793; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.:

Florida, Putnam Co., Molgo 310 (FLAS), MG639988, —, MG835098, MH166505,

MH166661, MG834947, MG834660, —, MG834794; C. graminea, tetraploid: 2n = 4x =

24, U.S.A.: Florida, Clay Co., Molgo 311 (FLAS), MG639989, —, MG835099,

MH166506, MH166662, MG834948, MG834661, —, MG834795; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Georgia, Charlton Co., Molgo 317 (FLAS), —, —,

187

MG835100, MH166507, MH166663, MG834949, —, —, MG834796; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Georgia, Coffee Co., Molgo 318 (FLAS), —, —,

MG835101, MH166508, MH166664, MG834950, MH166411, —, MG834797; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Georgia, Coffee Co., Molgo 319 (FLAS), —,

—, MG835102, MH166509, MH166665, MG834951, MG834662, —, MG834798; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Georgia, Emanuel Co., Molgo 320 (FLAS),

MG639990, MG834583, MG835103, MH166510, MH166666, MG834952, MG834663,

—, MG834799; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Georgia, Richmond Co.,

Molgo 322 (FLAS), —, —, MG835104, MH166511, MH166667, MG834953, MG834664,

—, MG834800; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina, Aiken

Co., Molgo 324 (FLAS), —, —, MG835105, MH166512, MH166668, MG834954, —, —,

MG834801; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina, Aiken Co.,

Molgo 325 (FLAS), —, —, MG835106, MH166513, MH166669, MG834955, MG834665,

—, MG834802; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina, Richland

Co., Molgo 326 (FLAS), —, —, MG835107, MH166514, MH166670, MG834956, —, —,

MG834803; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina, Richland

Co., Molgo 327 (FLAS), MG639991, MG834584, MG835108, MH166515, MH166671,

MG834957, MG834666, MG835230, MG834804; C. graminea, tetraploid: 2n = 4x = 24,

U.S.A.: South Carolina, Richland Co., Molgo 328 (FLAS), —, —, MG835109,

MH166517, MH166673, MG834959, MG834668, —, MG834806; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina, Kershaw Co., Molgo 329 (FLAS), —, —

, MG835110, MH166518, MH166674, MG834960, MG834669, —, MG834807; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: South Carolina, Kershaw Co., Molgo 330

188

(FLAS), —, —, MG835111, MH166519, MH166675, MG834961, MG834670, —,

MG834808; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina, Bladen Co.,

Molgo 334 (FLAS), —, —, MG835112, MH166520, MH166676, MG834962, MG834671,

—, MG834809; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: North Carolina, Bladen

Co., Molgo 335 (FLAS), —, —, MG835113, MH166521, MH166677, MG834963, —, —,

MG834810; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Marion Co., Molgo

348 (FLAS), —, —, MG835114, MH166522, MH166678, MG834964, MG834672,

MG835232, MG834811; C. graminea (individual 1), tetraploid: 2n = 4x = 24, U.S.A.:

Florida, Welaka Co., Molgo 360 (FLAS), MG639992, MG834585, MG835115,

MH166523, MH166679, MG834965, MG834673, MG835233, MG834812; C. graminea

(individual 2), tetraploid: 2n = 4x = 24, U.S.A.: Florida, Welaka Co., Molgo 360 (FLAS),

MG639993, MG834586, MG835116, MH166524, MH166680, MG834966, MG834674,

MG835234, MG834813;

C. graminea (individual 3), tetraploid: 2n = 4x = 24, U.S.A.: Florida, Welaka Co., Molgo

360 (FLAS), MG639994, MG834587, MG835117, MH166525, MH166681, MG834967,

—, MG835235, MG834814; C. graminea (individual 4), tetraploid: 2n = 4x = 24, U.S.A.:

Florida, Welaka Co., Molgo 360 (FLAS), —, —, MG835118, MH166526, MH166682,

MG834968, MG834675, MG835236, MG834815; C. graminea, tetraploid: 2n = 4x = 24,

U.S.A.: Florida, Lake Co., Molgo 362 (FLAS), MG639995, MG834588, MG835119,

MH166527, MH166683, MG834969, MG834676, MG835237, MG834816; C. graminea, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Lake Co., Molgo 363 (FLAS), —, —,

MG835120, MH166528, MH166684, MG834970, MG834677, MG835238, MG834817;

C. graminea, hexaploid: 2n = 6x = 36, U.S.A.: Florida, Lake Co., Molgo 235 (FLAS),

189

MG639996, MG834589, MG835121, —, MH166685, MG834971, —, MG835239,

MG834818; C. graminea, hexaploid: 2n = 6x = 36, U.S.A.: Florida, Lake Co., Molgo

236 (FLAS), MG639997, MG834590, MG835122, MH166529, MH166686, MG834972,

MG834678, MG835240, MG834819; C. graminea, hexaploid: 2n = 6x = 36, U.S.A.:

South Carolina, Richland Co., Molgo 327 (FLAS), —, —, MH166516, MH166672,

MG834958, MG834667, MG835231, MG834805; C. graminea, hexaploid: 2n = 6x = 36,

U.S.A.: Florida, Lake Co., Molgo 345 (FLAS), —, —, MG835123, MH166530,

MH166687, MG834973, MG834679, MG835241, MG834820; C. graminea, hexaploid:

2n = 6x = 36, U.S.A.: Florida, Hernando Co., Molgo 349 (FLAS), MG639998,

MG834591, MG835124, MH166531, MH166688, MG834974, MG834680, MG835242,

MG834821; Callisia hintoniorum B. L.Turner, NMNH 007-003, Hinton 25725, —,

Mexico: Nuevo León, Aramberri Co., Molgo 366 (FLAS), MG639918, —, MG835013,

MH166418, MH166570, MG834857, MG834616, MG835163, MG834709; Callisia micrantha (Torr.) D. R. Hunt, NMNH 2006-046, Patterson s.n., —, U.S.A.: Texas,

Cameron Co., Molgo 367 (FLAS), —, —, MG835015, MH166420, MH166572,

MG834859, —, MG835165, MG834711; Callisia monandra (Sw.) Schult. & Schult. f.,

NMNH 1993- 092, Bogner s.n., —; —;Bogner s.n. (NMNH), MG639920, —, —, —,

MH166573, MG834860, —, MG835166, —; Callisia multiflora (M. Martens & Galeotti)

Standl., NMNH 1992-041, Spencer 92.309, —, Mexico: Chiapas, Molgo 377 (FLAS),

MG639921, —, —, MH166421, MH166574, MG834861, —, MG835167, —; C. multiflora, NMNH 1980-395, Faden 76/166A, —; Mexico: Durango, Molgo 376 (FLAS),

MG639922, —, MG835016, MH166422, MH166575, MG834862, —, MG835168, —;

Callisia navicularis (Ortgies) D. R. Hunt, NMNH 1980-410, Fryxell s.n., —, Mexico:

190

Durango, Molgo 370 (FLAS), MG639923, —, MG835017, MH166423, MH166576,

MG834863, —, MG835169, MG834712; Callisia ornata (Small) G.C.Tucker, diploid: 2n

= 2x = 12, U.S.A.: Florida, Putnam Co., Molgo 312 (FLAS), MG640000, —, MG835128,

—, MH166692, MG834978, MH166412, —, MG834822; C. ornata, diploid: 2n = 2x =

12, U.S.A.: Florida, Volusia Co., Molgo 313 (FLAS), MG640001, —, MG835129,

MH166535, MH166693, MG834979, MG834682, —, MG834823; C. ornata, diploid: 2n

= 2x = 12, U.S.A.: Florida, Brevard Co., Molgo 314 (FLAS), MG640002, MG834592,

MG835130, MH166536, MH166694, MG834980, MG834683, MG835246, MG834824;

C. ornata, diploid: 2n = 2x = 12, U.S.A.: Florida, Brevard Co., Molgo 315 (FLAS),

MG640003, —, MG835131, MH166537, MH166695, MG834981, MG834684, —,

MG834825; C. ornata, diploid: 2n = 2x = 12, U.S.A.: Florida, Polk Co., Molgo 316

(FLAS), MG640004, MG834593, MG835132, MH166538, MH166696, MG834982,

MG834685, MG835247, MG834826; C. ornata, diploid: 2n = 2x = 12, U.S.A.: Florida,

Martin Co., Molgo 353 (FLAS), MG640005, MG834594, MG835133, MH166539,

MH166697, —, MG834686, MG835248, MG834827; C. ornata, diploid: 2n = 2x = 12,

U.S.A.: Florida, Polk Co., Molgo 354 (FLAS), MG640006, —, MG835134, MH166540,

MH166698, MG834983, MG834687, MG835249, MG834828; C. ornata, diploid: 2n =

2x = 12, U.S.A.: Florida, Polk Co., Molgo 355 (FLAS), MG640007, MG834595,

MG835135, MH166541, MH166699, MG834984, MG834688, MG835250, MG834829;

C. ornata, diploid: 2n = 2x = 12, U.S.A.: Florida, Polk Co., Molgo 356 (FLAS), —, —,

MG835136, MH166542, MH166700, MH166413, MG834689, MG835251, MG834830;

C. ornata, diploid: 2n = 2x = 12, U.S.A.: Florida, Brevard Co., Molgo 361 (FLAS),

MG640008, MG834596, MG835137, MH166543, MH166701, MG834985, MG834690,

191

MG835252, MG834831; C. ornata, diploid: 2n = 2x = 12, U.S.A.: Florida, Volusia Co.,

Molgo 365 (FLAS), —, —, MG835138, MH166544, MH166702, MG834986, MG834691,

MG835253, MG834832; C. ornata, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Polk Co.,

Molgo 228 (FLAS), MG640009, MG834597, MG835139, MH166545, MH166703,

MG834987, —, MG835254, MG834833; C. ornata, tetraploid: 2n = 4x = 24, U.S.A.:

Florida, Highlands Co., Molgo 298 (FLAS), MG640010, MG834598, MG835140,

MH166546, MH166704, MG834988, MG834692, —, MG834834; C. ornata, tetraploid:

2n = 4x = 24, U.S.A.: Florida, Highlands Co., Molgo 300 (FLAS), MG640011,

MG834599, MG835141, MH166547, MH166705, MG834989, MG834693, —,

MG834835; C. ornata, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Highlands Co., Molgo

301 (FLAS), MG640012, MG834600, MG835142, MH166548, MH166706, MG834990,

MG834694, —, MG834836; C. ornata, tetraploid: 2n = 4x = 24, U.S.A.: Florida,

Hillsborough Co., Molgo 350 (FLAS), MG640013, MG834601, MG835143, MH166549,

MH166707, MG834991, MG834695, MG835255, MG834837; C. ornata, tetraploid: 2n =

4x = 24, U.S.A.: Florida, Hillsborough Co., Molgo 351 (FLAS), MG640014, MG834602,

MG835144, MH166550, MH166708, MG834992, MG834696, MG835256, MG834838;

C. ornata, tetraploid: 2n = 4x = 24, U.S.A.: Florida, Sarasota Co., Molgo 352 (FLAS),

MG640015, MG834603, MG835145, MH166551, MH166709, MG834993, MG834697,

MG835257, MG834839; C. ornata, hexaploid: 2n = 6x = 36, U.S.A.: Florida, Lake Co.,

Molgo 343 (FLAS), —, —, MG835146, —, MH166710, MG834994, —, MG835258,

MG834840; C. ornata, hexaploid: 2n = 6x = 36, U.S.A.: Florida, Lake Co., Molgo 344

(FLAS), MG640016, MG834604, MG835147, MH166552, MH166711, MG834995,

MG834698, MG835259, MG834841; C. ornata, hexaploid: 2n = 6x = 36, U.S.A.:

192

Florida, Lake Co., Molgo 346 (FLAS), MG640017, MG834605, MH156551, MH166553,

MH166712, MG834996, MG834699, MG835260, MG834842; C. ornata, hexaploid: 2n

= 6x = 36, U.S.A.: Florida, Lake Co., Molgo 347 (FLAS), MG640018, MG834606,

MG835148, MH166554, MH166713, MG834997, MG834700, MG835261, MG834843;

C. ornata, hexaploid: 2n = 6x = 36, U.S.A.: Florida, Volusia Co., Molgo 364 (FLAS),

MG640019, MG834607, MG835149, MH166555, MH166714, MG834998, MG834701,

MG835262, MG834844; Callisia repens (Jacq.) L., NMNH 1982-291, Cult. Graf s.n.,

Bolivia, Molgo 375 (FLAS), —, —, MG835018, MH166424, MH166577, MG834864, —,

MG835170, —; C. repens, NMNH 1983-193, Cult. Plowman 12852, Brazil, Molgo 358

(FLAS), —, —, MG835019, MH166425, MH166578, MG834865, —, MG835171, —;

Callisia rosea (Vent.) D.R.Hunt, diploid: 2n = 2x = 12, U.S.A.: Georgia, Elbert Co.,

Molgo 237 (FLAS), MG640020, MG834608, MG835150, MH166556, MH166715,

MG834999, MG834702, MG835263, MG834845; C. rosea, diploid: 2n = 2x = 12,

U.S.A.: Georgia, Elbert Co., Molgo 238 (FLAS), MG640021, MG834609, MG835151,

MH166557, MH166716, MG835000, —, MG835264, MG834846; C. rosea, diploid: 2n =

2x = 12, U.S.A.: Georgia, Richmond Co., Molgo 240 (FLAS), MG640022, MG834610,

MG835152, MH166558, MH166717, MG835001, MG834703, MG835265, MG834847;

C. rosea, diploid: 2n = 2x = 12, U.S.A.: Georgia, Tattnall Co., Molgo 244 (FLAS),

MG640023, MG834611, MG835153, MH166559, MH166718, MG835002, MG834704,

MG835266, MG834848; C. rosea, diploid: 2n = 2x = 12, U.S.A.: South Carolina,

Allendale Co., Molgo 245 (FLAS), MG640024, MG834612, MG835154, MH166560,

MH166719, MG835003, MG834705, MG835267, MG834849; C. rosea, diploid: 2n = 2x

= 12, U.S.A.: Georgia, Columbia Co., Molgo 321 (FLAS), MG640025, MG834613,

193

MG835155, MH166561, MH166720, MG835004, MG834706, —, MG834850; C. rosea, diploid: 2n = 2x = 12, U.S.A.: Georgia, Lincoln Co., Molgo 323 (FLAS), MG640026, —,

MG835156, MH166562, MH166721, MG835005, MG834707, —, MG834851; Callisia soconuscensis Matuda, NMNH 1986-203, Munich Botanical Garden, 84/3362, —,

Ecuador, Molgo 374 (FLAS), MG639924, MG834542, MG835020, MH166426,

MH166579, MG834866, —, MG835172, —; Commelina erecta L., —; Molgo 299

(FLAS), MG639917, —, MG835011, MH166416, MH166568, MG834855, MG834615,

—, MG834708; Gibasis geniculata (Jacq.) Rohweder, NMNH 1982-298, Munich

Botanical Garden s.n., —, —, Molgo 378 (FLAS), —, —, MG835125, MH166532,

MH166689, MG834975, —, MG835243, —; Gibasis pellucida (M. Martens & Galeotti)

D. R. Hunt, NMNH 2007-030, Rosen, 298, —; U.S.A.: Texas, Harris Co., Molgo 357

(FLAS), —, —, MG835126, MH166533, MH166690, MG834976, MG834681,

MG835244, —; Gibasis venustula (Kunth) D. R. Hunt, NMNH 2003-081, —, —,

Mexico, Molgo 371, MG639999, —, MG835127, MH166534, MH166691, MG834977, —

, MG835245, —; Tradescantia ohiensis Raf., —; U.S.A.: South Carolina, Hampton

Co., Molgo 246 (FLAS), —, MG834614, MG835157, MH166563, MH166722,

MG835006, —, MG835268, MG834852; Tripogandra serrulata (Vahl) Handlos, NMNH

1986-223,Sastre, 8121, —; French Guiana, Molgo 369 (FLAS), — , —, MG835158,

MH166564, MH166723, MG835007, —, MG835269, —; T. serrulata, NMNH 1981-080,

Brenner, 10/81, —; Panama, Molgo 368 (FLAS), —, —, MG835159, MH166565,

MH166724, MG835008, —, MG835270, —.

194

APPENDIX C ENM PERFORMANCE

Table C-1. The area under the curve value (AUC) and standard deviation of all ENMs in the past, present and future models. CG2x = C. graminea 2x, CG4x = C. graminea 4x. CG6x = C. graminea 6x and C. ornata 6x, CO2x = C. ornata 2x, CO4x = C. ornata 4x and CR2 = C. rosea. LGM mid-Holocene Current 2050 RCP2.6 2050 RCP8.5 2070 RCP2.6 2070 RCP8.5 Taxa AUC sd AUC sd AUC sd AUC sd AUC sd AUC sd AUC sd CG2x 0.984 0.008 0.989 0.005 0.977 0.015 0.968 0.019 0.975 0.015 0.974 0.017 0.971 0.019 CG4x 0.939 0.012 0.938 0.012 0.909 0.019 0.903 0.020 0.903 0.020 0.908 0.019 0.906 0.019 CG6x 0.993 0.398 0.993 0.399 0.986 0.397 0.986 0.398 0.989 0.396 0.988 0.397 0.985 0.397 CO2x 0.977 0.007 0.975 0.006 0.973 0.006 0.970 0.007 0.969 0.012 0.972 0.008 0.973 0.006 CO4x 0.993 0.599 0.993 0.599 0.986 0.599 0.989 0.598 0.988 0.599 0.987 0.598 0.989 0.597 CR2x 0.939 0.016 0.939 0.017 0.905 0.032 0.904 0.027 0.904 0.032 0.903 0.031 0.904 0.031

195

LIST OF REFERENCES

Amborella Genome Project (2013) The Amborella genome and the evolution of flowering plants. Science 342: 1467. https://doi.org/10.1126/science.1241089

Anderson E, Sax K (1936) A cytological monograph of the American species of Tradescantia. Botanical Gazette 97: 433–476. http://www.jstor.org/stable/2471708 .

Baldwin JT (1941) Galax: the genus and its chromosomes. Journal of Heredity 32: 249– 254. https://doi.org/10.1093/oxfordjournals.jhered.a105053

Barker FK, Lutzoni FM (2002) The utility of the incongruence length difference test. Systematic Biology 51: 625-637. http://www.jstor.org/stable/3070944

Baum DA, Donoghue MJ (1995) Choosing among alternative "Phylogenetic" species concepts. Systematic Botany 20: 560-573. https://doi.org/10.2307/2419810

Bennett M, Leitch I (2012) Plant DNA C-values database (release 6.0, Dec. 2012). http://data.kew.org/cvalues/ [accessed 19 May 2017]

Bennett MD, Bhandol P, Leitch IJ (2000) Nuclear DNA amounts in angiosperms and their modern uses—807 new estimates. Annals of Botany 86: 859-909. https://doi.org/10.1006/anbo.2000.1253

Bergamo S (2003) A phylogenetic evaluation of Callisia Loefl. (Commelinaceae) based on molecular data. Ph.D. Dissertation, Athens, Georgia, U.S.A.: University of Georgia, 160 pp.

Brown JH (1984) On the relationship between abundance and distribution of species. The American Naturalist 124: 255-279. https://doi.org/10.1086/284267

Buchan KC (2000) The Bahamas. Marine Pollution Bulletin 41: 94-111. https://doi.org/10.1016/S0025-326X(00)00104-1

Burns JH, Faden RB, Steppan SJ (2011) Phylogenetic studies in the Commelinaceae subfamily Commelinoideae inferred from nuclear ribosomal and chloroplast DNA sequences. Systematic Botany 36: 268–276. https://doi.org/10.1600/036364411x569471

Camp WH (1945) The north American blueberries with notes on other groups of Vacciniaceae. Brittonia 5: 203–275. https://doi.org/10.2307/2804880

Campbell CS, Dickinson TA (1990) Apomixis, patterns of morphological variation, and species concepts in subfam. Maloideae (Rosaceae). Systematic Botany: 124- 135. https://doi.org/10.2307/2419022

196

Chase SS (1963) Analytic breeding in Solanum tuberosum L.–a scheme utilizing parthenotes and other diploid stocks. Canadian Journal of Genetics and Cytology 5: 359-363. https://doi.org/10.1139/g63-049

Darlu P, Lecointre G (2002) When does the incongruence length difference test fail? Molecular Biology and Evolution 19: 432-437. https://doi.org/10.1093/oxfordjournals.molbev.a004098

Davis PH, Heywood VH (1963) Principles of angiosperm taxonomy. Van Nostrand, Princeton New Jersey, 556 pp.

De Bodt S, Maere S, Van de Peer Y (2005) Genome duplication and the origin of angiosperms. Trends in Ecology & Evolution 20: 591-597. https://doi.org/10.1016/j.tree.2005.07.008 de Queiroz K (2007) Species concepts and species delimitation. Systematic Biology 56: 879-886. https://doi.org/10.1080/10635150701701083 de Queiroz K, Good DA (1997) Phenetic clustering in biology: a critique. The Quarterly Review of Biology 72: 3-30. http://www.jstor.org/stable/3036809

Delcourt PA (1980) Goshen Springs: late Quaternary vegetation record for southern Alabama. Ecology 61: 371-386. https://doi.org/10.2307/1935195

Deyrup MA (1988) Pollen-Feeding in Poecilognathus punctipennis (Diptera: Bombyliidae). The Florida Entomologist 71: 597-605. https://doi.org/10.2307/3495019

Dolezel J, Greilhuber J, Suda J (2007) Estimation of nuclear DNA content in plants using flow cytometry. Nature Protocols 2: 2233–2244. https://doi.org/10.1139/g63-04910.1038/nprot.2007.310

Dunford MP (1970) Triploid and tetraploid hybrids from diploid x tetraploid crosses in Grindelia (Compositae). American Journal of Botany 57: 856-860. https://doi.org/10.2307/2441344

Eilam T, Anikster Y, Millet E, Manisterski J, Feldman M (2010) Genome size in diploids, allopolyploids, and autopolyploids of Mediterranean Triticeae. Journal of Botany 2010: 12. http://dx.doi.org/10.1155/2010/341380

Elith J, Leathwick JR (2009) Species distribution models: ecological explanation and prediction across space and time. Annual Review of Ecology, Evolution, and Systematics 40: 677-697. https://doi.org/10.1146/annurev.ecolsys.110308.120159

197

Ellstrand NC, Elam DR (1993) Population genetic consequences of small population size: implications for plant conservation. Annual Review of Ecology and Systematics 24: 217-242. https://doi.org/10.2307/2097178

ESRI (2016) ArcGIS Desktop. release 10.5. Redlands, CA: Environmental Systems Research Institute. Available from http://www.esri.com/software/arcgis, [accessed August 2016]

Evans TM, Faden RB, Simpson MG, Sytsma KJ (2000) Phylogenetic relationships in the Commelinaceae: I. a cladistic analysis of morphological data. Systematic Botany 25: 668-691. https://doi.org/10.2307/2666727

Evans TM, Sytsma KJ, Faden RB, Givnish TJ (2003) Phylogenetic relationships in the Commelinaceae: II. a cladistic analysis of rbcL sequences and morphology. Systematic Botany 28: 270-292. https://doi.org/10.1043/0363-6445-28.2.270

Faden RB (1998) Commelinaceae. In: Kubitzki K (Ed) Flowering Plants · Monocotyledons: Alismatanae and Commelinanae (except Gramineae). Springer Berlin Heidelberg, Berlin, Heidelberg, 109-128.

Faden RB, Hunt DR (1991) The classification of the Commelinaceae. Taxon 40: 19-31. https://doi.org/10.2307/1222918

Faden RB, Suda Y (1980) Cytotaxonomy of Commelinaceae: chromosome numbers of some African and Asiatic species. Botanical Journal of the Linnean Society 81: 301–325. https://doi.org/10.1111/j.1095-8339.1980.tb01681.x

Farris JS, Kallersjo M, Kluge AG, Bult C (1995) Constructing a significance test for incongruence. Systematic Biology 44: 570-572. https://doi.org/10.2307/2413663

Fazekas AJ, Steeves R, Newmaster SG (2010) Improving sequencing quality from PCR products containing long mononucleotide repeats. Biotechniques 48: 277-285. https://doi.org/10.2144/000113369

Fei S, Desprez JM, Potter KM, Jo I, Knott JA, Oswalt CM (2017) Divergence of species responses to climate change. Science Advances 3: e1603055. https://doi.org/10.1126/sciadv.1603055

Felber F, Bever JD (1997) Effect of triploid fitness on the coexistence of diploids and tetraploids. Biological Journal of the Linnean Society 60: 95-106. https://doi.org/10.1006/bijl.1996.0090

Felde VA, Kapfer J, Grytnes JA (2012) Upward shift in elevational plant species ranges in Sikkilsdalen, central Norway. Ecography 35: 922-932. https://doi.org/10.1111/j.1600-0587.2011.07057.x

198

Feliner GN, Álvarez I, Fuertes-Aguilar J, Heuertz M, Marques I, Moharrek F, Piñeiro R, Riina R, Rosselló J, Soltis P (2017) Is homoploid hybrid speciation that rare? An empiricist’s view. Heredity 118: 513-516. https://doi.org/10.1038/hdy.2017.7

Folk RA, Visger CJ, Soltis PS, Soltis DE, Guralnick RP (2017) Geographic range dynamics drove ancient hybridization in a lineage of angiosperms. bioRxiv 129189. https://doi.org/10.1101/129189

Fowler NL, Levin DA (1984) Ecological constraints on the establishment of a novel polyploid in competition with its diploid progenitor. American Naturalist 124: 703- 711. https://doi.org/10.1086/284307

Gaston KJ, Blackburn TM, Lawton JH (1997) Interspecific abundance-range size relationships: an appraisal of mechanisms. Journal of Animal Ecology 66: 579- 601. https://doi.org/10.2307/5951

Gaynor ML, Blaine MD, Soltis DE, Soltis PS (in revision) Absence of niche divergence among ploidal levels in the classic autopolyploid system, Galax urceolata (Diapensiaceae). American Journal of Botany.

Gelfand DH (1989) Taq DNA polymerase. In: Erlich HA (Ed) PCR technology. Palgrave Macmillan, London, 17-22.

Gent PR, Danabasoglu G, Donner LJ, Holland MM, Hunke EC, Jayne SR, Lawrence DM, Neale RB, Rasch PJ, Vertenstein M, Worley PH, Yang Z-L, Zhang M (2011) The community climate system model version 4. Journal of Climate 24: 4973- 4991. https://doi.org/10.1175/2011jcli4083.1

Giles NH (1942) Autopolyploidy and geographical distribution in Cuthbertia graminea Small. American Journal of Botany: 637–645. https://doi.org/10.1002/j.1537- 2197.1942.tb10259.x

Giles NH (1943) Studies of natural populations of Cuthbertia graminea in the Carolinas. Journal of the Elisha Mitchell Scientific Society 59: 73–79. http://dc.lib.unc.edu/cdm/ref/collection/jncas/id/1866

Giles NH (unpublished) The evolution of a natural polyploid complex – The genus Cuthbertia. 12 pp.

Glennon KL, Ritchie ME, Segraves KA (2014) Evidence for shared broad-scale climatic niches of diploid and polyploid plants. Ecology Letters 17: 574-582. https://doi.org/10.1111/ele.12259

Godsoe W, Larson MA, Glennon KL, Segraves KA (2013) Polyploidization in Heuchera cylindrica (Saxifragaceae) did not result in a shift in climatic requirements. American Journal of Botany 100: 496-508. https://doi.org/10.3732/ajb.1200275

199

Gouy M, Guindon S, Gascuel O (2010) SeaView Version 4: a multiplatform graphical user interface for sequence alignment and phylogenetic tree building. Molecular Biology and Evolution 27: 221-224. https://doi.org/10.1093/molbev/msp259

Grabiele M, Daviña JR, Honfi AI (2015) Cytogenetic analyses as clarifying tools for taxonomy of the genus Callisia Loefl.(Commelinaceae)/Análisis citogenéticos como herramienta para esclarecer la taxonomía del género Callisia Loefl.(Commelinaceae). Gayana Botanica 72: 34. https://doi.org/10.4067/S0717- 66432015000100005

Grant V (1981) Plant speciation. Columbia University Press, New York, 563 pp.

Grunenwald H (2003) Optimization of polymerase chain reactions. In: Barlett JMS, Stirling D (Eds) Methods in molecular biology, PCR protocols. Humana Press, Totowa, New Jersey, 89-99.

Guisan A, Petitpierre B, Broennimann O, Daehler C, Kueffer C (2014) Unifying niche shift studies: insights from biological invasions. Trends in Ecology & Evolution 29: 260-269. https://doi.org/10.1016/j.tree.2014.02.009

Guisan A, Thuiller W (2005) Predicting species distribution: offering more than simple habitat models. Ecology Letters 8: 993-1009. https://doi.org/10.1111/j.1461- 0248.2005.00792.x

Guisan A, Tingley R, Baumgartner JB, Naujokaitis‐Lewis I, Sutcliffe PR, Tulloch AI, Regan TJ, Brotons L, McDonald‐Madden E, Mantyka‐Pringle C (2013) Predicting species distributions for conservation decisions. Ecology Letters 16: 1424-1435. https://doi.org/10.1111/ele.12189

Hagerup O (1927) Empetrum hermaphroditum (Lge) Hagerup: A new tetraploid, bisexual species. Dansk botanisk arkiv 5: 1-17.

Hamlin JAP, Simmonds TJ, Arnold ML (2017) Niche conservatism for ecological preference in the Louisiana iris species complex. Biological Journal of the Linnean Society 120: 144-154. https://doi.org/10.1111/bij.12884

Hanson L, Boyd A, Johnson MAT, Bennett MD (2005) First Nuclear DNA C-values for 18 Eudicot families. Annals of Botany 96: 1315–1320. https://doi.org/10.1093/aob/mci283

Herben T, Suda J, Klimešová J (2017) Polyploid species rely on vegetative reproduction more than diploids: a re-examination of the old hypothesis. Annals of Botany 120: 341-349. https://doi.org/10.1093/aob/mcx009

200

Hertweck KL (2011) Genome evolution in monocots. Ph.D. Dissertation, Columbia, U.S.A.: University of Missouri, 153 pp.

Hertweck KL, Pires JC (2014) Systematics and evolution of inflorescence structure in the Tradescantia alliance (Commelinaceae). Systematic Botany 39: 105–116. https://doi.org/10.1600/036364414X677991

Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A (2005) Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology 25: 1965-1978. https://doi.org/10.1002/joc.1276

Ho WW, Smith SD (2016) Molecular evolution of anthocyanin pigmentation genes following losses of flower color. BioMed Central Evolutionary Biology 16: 98. https://doi.org/10.1186/s12862-016-0675-3

Hoshi Y, Shirakawa J, Hasebe M, Fukushima K, Kondo K (2008) Tandem repeat rDNA sequences derived from parents were stably maintained in hexaploids of Drosera spathulata complex (Droseraceae). Cytologia 73: 313-325. https://doi.org/10.1508/cytologia.73.313

Hunt DR (1986) Amplification of Callisia Loefl.: American Commelinaceae: XV. Kew Bulletin 41: 407–412. https://doi.org/10.2307/4102950

Hunt DR (1994) Commelinaceae. In: Davidse G, Sousa SM, Chater AO (Eds) Flora Mesoamerica. Instituto de Biologia, Universidad Nacional Autonoma de Mexico (UNAM) Mexico City, Mexico 157-173.

Husband BC, Baldwin SJ, Suda J (2013) The incidence of polyploidy in natural plant populations: major patterns and evolutionary processes. In: Greilhuber J, Dolezel J, Wendel JF (Eds) Plant genome diversity Volume 2: physical structure, behaviour and evolution of plant genomes. Springer Vienna, Vienna, 255-276.

Ichikawa S, Sparrow AH (1967) Radiation-induced loss of the reproductive integrity in the stamen hairs of a polyploid series of Tradescantia species. Radiation Botany 7: 429- 441. https://doi.org/10.1016/0033-7560(67)90022-1

Iltis HH, Doebley JF, Guzmán RM, Pazy B (1979) Zea diploperennis (Gramineae): a new Teosinte from Mexico. Science 203: 186–188. https://doi.org/10.1126/science.203.4376.186

Jiao Y, Wickett NJ, Ayyampalayam S, Chanderbali AS, Landherr L, Ralph PE, Tomsho LP, Hu Y, Liang H, Soltis PS, Soltis DE, Clifton SW, Schlarbaum SE, Schuster SC, Ma H, Leebens-Mack J, dePamphilis CW (2011) Ancestral polyploidy in seed plants and angiosperms. Nature 473: 97–100. https://doi.org/10.1038/nature09916

201

Johnson LA, Soltis DE (1998) Assessing congruence: empirical examples from molecular data. In: Soltis DE, Soltis PS, Doyle JJ (Eds) Molecular systematics of plants II: DNA sequencing. Kluwer Academic Publishers, Boston, 297-348.

Jones K, Jopling C (1972) Chromosomes and the classification of the Commelinaceae. Botanical Journal of the Linnean Society 65: 129-162. https://doi.org/10.1111/j.1095-8339.1972.tb00929.x

Jones K, Kenton A (1984) Mechanism of chromosome change in the evolution of the tribe Tradescantieae (Commelinaceae). In: Sharma A, Sharma AK (Eds) Chromosomes in Evolution of Eukaryotic Groups. CRC Press, Boca Raton, Florida, 143-168.

Judd WS, Campbell CS, Kellogg EA, Stevens PF, Donoghue MJ (2016) Plant Systematics: a phylogenetic approach. Sinauer Associates, Inc., Sunderland, MA, 677 pp.

Judd WS, Soltis DE, Soltis PS, Ionta G (2007) Tolmiea diplomenziesii: a new species from the Pacific Northwest and the diploid sister taxon of the autotetraploid T. menziesii (Saxifragaceae). Brittonia 59: 217–225. https://doi.org/10.1663/0007- 196x(2007)59[217:tdansf]2.0.co;2

Kassambara A, Mundt F (2017) Extract and visualize the results of multivariate data analyses. version 1.0.5. website: http://www.sthda.com/english/rpkgs/factoextra

Kato A, Lamb JC, Albert PS, Danilova T, Han F, Gao Z, Findley S, Birchler JA (2011) Chromosome painting for plant biotechnology. In: Birchler JA (Ed) Plant chromosome engineering: methods and protocols, methods in molecular biology. Humana Press, Totowa, New York, U.S.A., 67–96.

Katoh K, Standley DM (2013) MAFFT Multiple sequence alignment software version 7: improvements in performance and usability. Molecular Biology and Evolution 30: 772-780. https://doi.org/10.1093/molbev/mst010

Kelly AW (1991) Distribution and crossing relationships within and among diploid and tetraploid populations of Cuthbertia graminea Small. M.Sc. Thesis, Durham, North Carolina, U.S.A.: Duke University, 95 pp.

Kent WJ (2002) BLAT—The BLAST-like alignment tool. Genome Research 12: 656- 664. https://doi.org/10.1101/gr.229202

Krebs SL, Hancock JF (1989) Tetrasomic inheritance of isoenzyme markers in the highbush blueberry, Vaccinium corymbosum L. Heredity 63: 11–18. https://doi.org/10.1038/hdy.1989.70

202

Kück P, Meusemann K (2010) FASconCAT: Convenient handling of data matrices. Molecular Phylogenetics and Evolution 56: 1115-1118. https://doi.org/10.1016/j.ympev.2010.04.024

Kumar S, Stecher G, Tamura K (2016) MEGA7: molecular evolutionary genetics analysis version 7.0 for bigger datasets. Molecular Biology and Evolution 33: 1870-1874. https://doi.org/10.1093/molbev/msw054

Lakela O (1972) Field observations of Cuthbertia (Commelinaceae) with description of a new form. SIDA 5: 26–32. http://www.jstor.org/stable/41967342

Lê S, Josse J, Husson F (2008) FactoMineR: an R package for multivariate analysis. Journal of statistical software 25: 1-18. https://hal.archives-ouvertes.fr/hal- 00359835

Leitch IJ, Beaulieu JM, Chase MW, Leitch AR, Fay MF (2010) Genome size dynamics and evolution in monocots. Journal of Botany 2010: 18. https://doi.org/10.1155/2010/862516

Leitch IJ, Bennett MD (2004) Genome downsizing in polyploid plants. Biological Journal of the Linnean Society 82: 651–663. https://doi.org/10.1111/j.1095- 8312.2004.00349.x

Lenoir J, Gégout JC, Marquet PA, de Ruffray P, Brisse H (2008) A significant upward shift in plant species optimum elevation during the 20th century. Science 320: 1768-1771. https://doi.org/10.1126/science.1156831

Levin DA (1975) Minority cytotype exclusion in local plant populations. Taxon 24: 35-43. https://doi.org/10.2307/1218997

Levin DA (2002) The role of chromosomal change in plant evolution. Oxford University Press, 240 pp.

Levin R (1968) Evolution in changing environments: some theoretical explorations. Princeton University Press, Princeton, New York, 132 pp.

Lewis H (1967) The taxonomic significance of autopolyploidy. Taxon 16: 267-271. https://doi.org/10.2307/1216373

Lewis PO, Crawford DJ (1995) Pleistocene refugium endemics exhibit greater allozymic diversity than widespread congeners in the genus Polygonella (Polygonaceae). American Journal of Botany 82: 141-149. https://doi.org/10.2307/2445522

Liu C, Berry PM, Dawson TP, Pearson RG (2005) Selecting thresholds of occurrence in the prediction of species distributions. Ecography 28: 385-393. https://doi.org/10.1111/j.0906-7590.2005.03957.x

203

López-Alvarez D, Manzaneda AJ, Rey PJ, Giraldo P, Benavente E, Allainguillaume J, Mur L, Caicedo AL, Hazen SP, Breiman A, Ezrati S, Catalán P (2015) Environmental niche variation and evolutionary diversification of the Brachypodium distachyon grass complex species in their native circum- Mediterranean range. American Journal of Botany 102: 1073-1088. https://doi.org/10.3732/ajb.1500128

Lowry E, Lester SE (2006) The biogeography of plant reproduction: potential determinants of species’ range sizes. Journal of Biogeography 33: 1975-1982. https://doi.org/10.1111/j.1365-2699.2006.01562.x

Malé P-JG, Bardon L, Besnard G, Coissac E, Delsuc F, Engel J, Lhuillier E, Scotti- Saintagne C, Tinaut A, Chave J (2014) Genome skimming by shotgun sequencing helps resolve the phylogeny of a pantropical tree family. Molecular Ecology Resources 14: 966-975. https://doi.org/10.1111/1755-0998.12246

Mandle L, Warren DL, Hoffmann MH, Peterson AT, Schmitt J, von Wettberg EJ (2010) Conclusions about niche expansion in introduced Impatiens walleriana populations depend on method of analysis. PLoS ONE 5: e15297. https://doi.org/10.1371/journal.pone.0015297

Marchant DB, Soltis DE, Soltis PS (2016) Patterns of abiotic niche shifts in allopolyploids relative to their progenitors. New Phytologist 212: 708-718. https://doi.org/10.1111/nph.14069

Mason AS, Pires JC (2015) Unreduced gametes: meiotic mishap or evolutionary mechanism? Trends in Genetics 31: 5-10. https://doi.org/10.1016/j.tig.2014.09.011

Mason SC, Palmer G, Fox R, Gillings S, Hill JK, Thomas CD, Oliver TH (2015) Geographical range margins of many taxonomic groups continue to shift polewards. Biological Journal of the Linnean Society 115: 586-597. https://doi.org/10.1111/bij.12574

Mavrodiev EV, Chester M, Suárez-Santiago VN, Visger CJ, Rodriguez R, Susanna A, Baldini RM, Soltis PS, Soltis DE (2015) Multiple origins and chromosomal novelty in the allotetraploid Tragopogon castellanus (Asteraceae). New Phytologist 206: 1172–1183. https://doi.org/10.1111/nph.13227

Mavrodiev EV, Soltis PS, Soltis DE (2008) Putative parentage of six Old World polyploids in Tragopogon L. (Asteraceae: Scorzonerinae) based on ITS, ETS, and plastid sequence data. Taxon 57: 1215-1215. http://www.jstor.org/stable/27756775

204

McIntyre PJ (2012) Polyploidy associated with altered and broader ecological niches in the Claytonia perfoliata (Portulacaceae) species complex. American Journal of Botany 99: 655-662. https://doi.org/10.3732/ajb.1100466

McLaughlin BC, Ackerly DD, Klos PZ, Natali J, Dawson TE, Thompson SE (2017) Hydrologic refugia, plants, and climate change. Global Change Biology 23: 2941- 2961. https://doi.org/10.1111/gcb.13629

Meinshausen M, Smith SJ, Calvin K, Daniel JS, Kainuma MLT, Lamarque J-F, Matsumoto K, Montzka SA, Raper SCB, Riahi K, Thomson A, Velders GJM, van Vuuren DPP (2011) The RCP greenhouse gas concentrations and their extensions from 1765 to 2300. Climatic Change 109: 213. https://doi.org/10.1007/s10584-011-0156-z

Metz CE (1978) Basic principles of ROC analysis. Semin Nucl Med 8: 283-298. https://doi.org/10.1016/S0001-2998(78)80014-2

Mishler BD, Brandon RN (1987) Individuality, pluralism, and the phylogenetic species concept. Biology and Philosophy 2: 397-414. https://doi.org/10.1007/bf00127698

Mosquin T (1967) Evidence for autopolyploidy in Epilobium angustifolium (Onagraceae). Evolution 21: 713–719. https://doi.org/10.2307/2406768

Müntzing A (1936) The evolutionary significance of autopolyploidy. Hereditas 21: 363- 378. https://doi.org/10.1111/j.1601-5223.1936.tb03204.x

Nicolosi E, Deng ZN, Gentile A, La Malfa S, Continella G, Tribulato E (2000) Citrus phylogeny and genetic origin of important species as investigated by molecular markers. Theoretical and Applied Genetics 100: 1155-1166. https://doi.org/10.1007/s001220051419

Oliver J, Rejon MR (1980) The relation between isozymes and ploidy level. Its application to biogeographical studies of Muscari atlanticum (Liliaceae). Taxon: 27-32. https://doi.org/10.2307/1219593

Otto SP, Whitton J (2000) Polyploid incidence and evolution. Annual Review of Genetics 34: 401-437. https://doi.org/10.1146/annurev.genet.34.1.401

Owen W (2002) The history of native plant communities in the south. Southern forest resource assessment Edited by DN Wear and JG Greis USDA For Serv Gen Tech Rep GTR-SRS-53: 47-61. https://www.srs.fs.usda.gov/sustain/report/pdf/chapter_02e.pdf

Owens S (1981) Self-incompatibility in the Commelinaceae. Annals of Botany 47: 567- 581. https://doi.org/10.1093/oxfordjournals.aob.a086054

205

Papeş M, Gaubert P (2007) Modelling ecological niches from low numbers of occurrences: assessment of the conservation status of poorly known viverrids (Mammalia, Carnivora) across two continents. Diversity and Distributions 13: 890-902. https://doi.org/10.1111/j.1472-4642.2007.00392.x

Parisod C, Broennimann O (2016) Towards unified hypotheses of the impact of polyploidy on ecological niches. New Phytologist 212: 540-542. https://doi.org/10.1111/nph.14133

Peterson AT, Nyári ÁS, Crandall K (2008) Ecological niche conservatism and pleistocene refugia in the thrush-like mourner, Schiffornis sp., in the neotropics. Evolution 62: 173-183. https://doi.org/10.1111/j.1558-5646.2007.00258.x

Phillips SJ, Dudík M (2008) Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography 31: 161-175. https://doi.org/10.1111/j.0906-7590.2008.5203.x

Poskanzer J (2001) ACME mapper. ACME Laboratories, version 2.1. A High-precision general purpose mapping application, website: http://mapper.acme.com/ [accessed March 2013]

R Core Team (2016) R: A language and environment for statistical computing. . R Foundation for Statistical Computing. website: URL http://www.R-project.org/

Raes N, Roos MC, Slik JWF, Van Loon EE, ter Steege H (2009) Botanical richness and endemicity patterns of Borneo derived from species distribution models. Ecography 32: 180-192. https://doi.org/10.1111/j.1600-0587.2009.05800.x

Rambaut A (2016) FigTree. version 1.4.3. website: http://tree.bio.ed.ac.uk/software/figtree/ [accessed January, 2017]

Rapacciuolo G, Maher SP, Schneider AC, Hammond TT, Jabis MD, Walsh RE, Iknayan KJ, Walden GK, Oldfather MF, Ackerly DD, Beissinger SR (2014) Beyond a warming fingerprint: individualistic biogeographic responses to heterogeneous climate change in California. Global Change Biology 20: 2841-2855. https://doi.org/10.1111/gcb.12638

Rausher MD (2008) Evolutionary transitions in floral color. International Journal of Plant Sciences 169: 7–21. https://doi.org/10.1086/523358

Rensing SA, Lang D, Zimmer AD, Terry A, Salamov A, Shapiro H, Nishiyama T, Perroud P-F, Lindquist EA, Kamisugi Y (2008) The Physcomitrella genome reveals evolutionary insights into the conquest of land by plants. Science 319: 64-69. https://doi.org/10.1126/science.1150646

206

Riahi K, Rao S, Krey V, Cho C, Chirkov V, Fischer G, Kindermann G, Nakicenovic N, Rafaj P (2011) RCP 8.5—A scenario of comparatively high greenhouse gas emissions. Climatic Change 109: 33. https://doi.org/10.1007/s10584-011-0149-y

Rios NE, Bart HL (2010) GEOLocate. Tulane University Museum of Natural History, version 3.22. Georeferencing software for Natural History Collections, website: http://www.museum.tulane.edu/geolocate/ [accessed March 2013]

Roberts AV, Gladis T, Brumme H (2009) DNA amounts of roses (Rosa L.) and their use in attributing ploidy levels. Plant Cell Reports 28: 61–71. https://doi.org/10.1007/s00299-008-0615-9

Rohweder O (1956) Die Farinosae in der Vegetation von El Salvador Cram, de Gruyter & Co. Hamburg. Abhandlungen aus dem Gebiet der Auslandskunde Reihe C Naturwissenschaften 18: ii-xvi.

Scheffers BR, De Meester L, Bridge TC, Hoffmann AA, Pandolfi JM, Corlett RT, Butchart SH, Pearce-Kelly P, Kovacs KM, Dudgeon D (2016) The broad footprint of climate change from genes to biomes to people. Science 354: aaf7671. https://doi.org/10.1126/science.aaf7671

Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nature methods 9: 671-675. https://doi.org/10.1038/nmeth.2089

Schoener TW (1968) The Anolis lizards of Bimini: resource partitioning in a complex fauna. Ecology 49: 704-726. https://doi.org/10.2307/1935534

Schumer M, Rosenthal GG, Andolfatto P (2014) How common is homoploid hybrid speciation? Evolution 68: 1553-1560. http://dx.doi.org/10.1111/evo.12399

Shaw J, Lickey EB, Beck JT, Farmer SB, Liu W, Miller J, Siripun KC, Winder CT, Schilling EE, Small RL (2005) The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. American Journal of Botany 92: 142-166. https://doi.org/10.3732/ajb.92.1.142

Shaw J, Lickey EB, Schilling EE, Small RL (2007) Comparison of whole chloroplast genome sequences to choose noncoding regions for phylogenetic studies in angiosperms: the tortoise and the hare III. American Journal of Botany 94: 275- 288. https://doi.org/10.3732/ajb.94.3.275

Siddall ME, Whiting MF (1999) Long-branch abstractions. Cladistics 15: 9-24. https://doi.org/10.1111/j.1096-0031.1999.tb00391.x

207

Small JK (1933) Manual of the southeastern flora :being descriptions of the seed plants growing naturally in Florida, Alabama, Mississippi, eastern Louisiana, Tennessee, North Carolina, South Carolina and Georgia. The University of North Carolina Press, Chapel Hill, NC, 1222 pp.

Small JK, Rydberg PA (1903) Flora of the southeastern United States; being descriptions of the seed-plants, ferns and fern-allies growing naturally in North Carolina, South Carolina, Georgia, Florida, Tennessee, Alabama, Mississippi, Arkansas, Louisiana and the Indian territory and in Oklahoma and Texas east of the one-hundredth meridian. The author, New York, 1279 pp.

Sokal RR, Rohlf FJ (1997) Biometry: the principles and practice of statistics in biological research. W. H. Freeman and Company, 887 pp.

Soltis DE (1980) Karyotypic relationships among species of Boykinia, Heuchera, Mitella, Sullivantia, Tiarella, and Tolmiea (Saxifragaceae). Systematic Botany 5: 17–29. https://doi.org/10.2307/2418732

Soltis DE, Albert VA, Leebens-Mack J, Bell CD, Paterson AH, Zheng C, Sankoff D, dePamphilis CW, Wall PK, Soltis PS (2009) Polyploidy and angiosperm diversification. American Journal of Botany 96: 336–348. https://doi.org/10.3732/ajb.0800079

Soltis DE, Mavrodiev EV, Doyle JJ, Rauscher J, Soltis PS (2008) Featured Paper—ITS and ETS sequence data and phylogeny reconstruction in allopolyploids and hybrids. Systematic Botany 33: 7-20. https://doi.org/10.1600/036364408783887401

Soltis DE, Rieseberg LH (1986) Autopolyploidy in Tolmiea menziesii (Saxifragaceae): genetic Insights from enzyme electrophoresis. American Journal of Botany 73: 310-318. https://doi.org/10.2307/2444186

Soltis DE, Soltis PS (1998) Choosing an approach and an appropriate gene for phylogenetic analysis. In: Soltis DE, Soltis PS, Doyle JJ (Eds) Molecular systematics of plants II. Kluwer Academic Publishers, Norwell Massachusetts, 1- 42.

Soltis DE, Soltis PS (2004) Amborella not a “basal angiosperm”? Not so fast. American Journal of Botany 91: 997-1001. https://doi.org/10.3732/ajb.91.6.997

Soltis DE, Soltis PS, Schemske DW, Hancock JF, Thompson JN, Husband BC, Judd WS (2007) Autopolyploidy in angiosperms: have we grossly underestimated the number of species? Taxon 56: 13–30. http://www.ingentaconnect.com/content/iapt/tax/2007/00000056/00000001/art00 003

208

Soltis DE, Visger CJ, Marchant DB, Soltis PS (2016) Polyploidy: pitfalls and paths to a paradigm. American Journal of Botany. https://doi.org/10.3732/ajb.1500501

Soltis PS (2017) Digitization of herbaria enables novel research. American Journal of Botany 104: 1281-1284. https://doi.org/10.3732/ajb.1700281

Soltis PS, Liu X, Marchant DB, Visger CJ, Soltis DE (2014) Polyploidy and novelty: Gottlieb's legacy. Philosophical Transactions of the Royal Society B: Biological Sciences 369. https://doi.org/10.1098/rstb.2013.0351

Soltis PS, Soltis DE (2009) The role of hybridization in plant speciation. Annual Review of Plant Biology 60: 561–588. https://doi.org/10.1146/annurev.arplant.043008.092039

Soltis PS, Soltis DE (2016) Ancient WGD events as drivers of key innovations in angiosperms. Current Opinion in Plant Biology 30: 159–165. https://doi.org/10.1016/j.pbi.2016.03.015

Spoelhof JP, Soltis PS, Soltis DE (2017) Pure polyploidy: Closing the gaps in autopolyploid research. Journal of Systematics and Evolution 55: 340-352. https://doi.org/10.1111/jse.12253

Stamatakis A (2014) RAxML version 8: a tool for phylogenetic analysis and post- analysis of large phylogenies. Bioinformatics 30: 1312-1313. https://doi.org/10.1093/bioinformatics/btu033

Stebbins GL (1950) Variation and evolution in plants. Columbia University Press, New York, U.S.A., 643 pp.

Stebbins GL (1985) Polyploidy, hybridization, and the invasion of new habitats. Annals of the Missouri Botanical Garden 72: 824-832. https://doi.org/10.2307/2399224

Swets JA (1988) Measuring the accuracy of diagnostic systems. Science 240: 1285- 1293. https://doi.org/10.1126/science.3287615

Swofford DL (2002) PAUP*: Phylogenetic analysis using parsimony (*and other methods). Sinauer Associates, version 4.0b10.

Taberlet P, Gielly L, Pautou G, Bouvet J (1991) Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105- 1109. https://doi.org/10.1007/bf00037152

Tank DC, Eastman JM, Pennell MW, Soltis PS, Soltis DE, Hinchliff CE, Brown JW, Sessa EB, Harmon LJ (2015) Nested radiations and the pulse of angiosperm diversification: increased diversification rates often follow whole genome duplications. New Phytologist 207: 454–467. https://doi.org/10.1111/nph.13491

209

te Beest M, Le Roux JJ, Richardson DM, Brysting AK, Suda J, Kubešová M, Pyšek P (2011) The more the better? The role of polyploidy in facilitating plant invasions. Annals of Botany 109: 19-45. https://doi.org/10.1093/aob/mcr277

Theodoridis S, Randin C, Broennimann O, Patsiou T, Conti E (2013) Divergent and narrower climatic niches characterize polyploid species of European primroses in Primula sect. Aleuritia. Journal of Biogeography 40: 1278-1289. https://doi.org/10.1111/jbi.12085

Index Herbariorum: A global directory of public herbaria and associated staff. New York Botanical Garden. http://sweetgum.nybg.org/science/ih/ [accessed August.2012]

Tiffin P, Gaut BS (2001) Sequence diversity in the tetraploid Zea perennis and the closely related diploid Z. diploperennis: insights from four nuclear loci. Genetics 158: 401-412. http://www.genetics.org/content/158/1/401.short

Tomlinson PB (1966) Anatomical data in the classification of Commelinaceae. Journal of the Linnean Society of London, Botany 59: 371-395. https://doi.org/10.1111/j.1095-8339.1966.tb00069.x

Tucker GC (1989) The genera of Commelinaceae in the southeastern United States. J Arnold Arbor 70: 97–130. http://www.jstor.org/stable/43782223 van Vuuren DP, Edmonds J, Kainuma M, Riahi K, Thomson A, Hibbard K, Hurtt GC, Kram T, Krey V, Lamarque J-F, Masui T, Meinshausen M, Nakicenovic N, Smith SJ, Rose SK (2011a) The representative concentration pathways: an overview. Climatic Change 109: 5. https://doi.org/10.1007/s10584-011-0148-z van Vuuren DP, Stehfest E, den Elzen MGJ, Kram T, van Vliet J, Deetman S, Isaac M, Klein Goldewijk K, Hof A, Mendoza Beltran A, Oostenrijk R, van Ruijven B (2011b) RCP2.6: exploring the possibility to keep global mean temperature increase below 2°C. Climatic Change 109: 95. https://doi.org/10.1007/s10584- 011-0152-3

VanDerWal J, Murphy HT, Kutt AS, Perkins GC, Bateman BL, Perry JJ, Reside AE (2012) Focus on poleward shifts in species; distribution underestimates the fingerprint of climate change. Nature Climate Change 3: 239. https://doi.org/10.1038/nclimate1688

Ventenat ÉP (1800) Description des plantes nouvelles et peu connues, cultivees dans le jardin de JM Cels. avec figures par EP Ventenat. De l'imprimerie de Crapelet, an 8, Paris, 426 pp.

210

Visger CJ, Germain-Aubrey CC, Patel M, Sessa EB, Soltis PS, Soltis DE (2016) Niche divergence between diploid and autotetraploid Tolmiea. American Journal of Botany 103: 1396-1406. https://doi.org/10.3732/ajb.1600130

Walker MJ, Stockman AK, Marek PE, Bond JE (2009) Pleistocene glacial refugia across the Appalachian Mountains and coastal plain in the millipede genus Narceus: evidence from population genetic, phylogeographic, and paleoclimatic data. Bmc Evolutionary Biology 9: 25. https://doi.org/10.1186/1471-2148-9-25

Warren DL, Glor RE, Turelli M (2008) Environmental niche equivalency versus conservatism: quantitative approaches to niche evolution. Evolution 62: 2868- 2883. https://doi.org/10.1111/j.1558-5646.2008.00482.x

Warren DL, Glor RE, Turelli M (2010) ENMTools: a toolbox for comparative studies of environmental niche models. Ecography 33: 607-611. https://doi.org/10.1111/j.1600-0587.2009.06142.x

Watts WA (1971) Postglacial and interglacial vegetation history of southern Georgia and central Florida. Ecology 52: 676-690. https://doi.org/10.2307/1934159

Watts WA (1980) The late Quaternary vegetation history of the southeastern United States. Annual Review of Ecology and Systematics 11: 387-409. https://doi.org/10.1146/annurev.es.11.110180.002131

Wendel JF (2000) Genome evolution in polyploids. Plant Molecular Biology 42: 225- 249. http://dx.doi.org/10.1023/A:1006392424384

Wendel JF, Doyle JJ (1998) Phylogenetic incongruence: window into genome history and molecular evolution. In: Soltis DE, Soltis P, Doyle JJ (Eds) Molecular systematics of plants II: DNA sequencing. Kluwer Academics Publisher, Norwell, Massachusettes, U.S.A., 265-296.

Wiens JJ (2007) Species delimitation: new approaches for discovering diversity. Systematic Biology 56: 875-878. https://doi.org/10.1080/10635150701748506

Wiens JJ (2016) Climate-related local extinctions are already widespread among plant and animal species. Plos Biology 14: e2001104. https://doi.org/10.1371/journal.pbio.2001104

Wiley E, Mayden RL (2000) The evolutionary species concept. Species concepts and phylogenetic theory: a debate: 70-89.

Wolf PG, Soltis DE, Soltis PS (1990) Chloroplast-DNA and allozymic variation in diploid and autotetraploid Heuchera grossulariifolia (Saxifragaceae). American Journal of Botany 77: 232–244. https://doi.org/10.2307/2444645

211

Woodson RE (1942) Commentary on North America genera of Commelinaceae. Annals of the Missouri Botanical Garden 29: 141-154. https://doi.org/10.2307/2394315

Wright HE (1976) The dynamic nature of Holocene vegetation: A problem in paleoclimatology, biogeography, and stratigraphic nomenclature. Quaternary Research 6: 581-596. https://doi.org/10.1016/0033-5894(76)90028-4

Yamada KD, Tomii K, Katoh K (2016) Application of the MAFFT sequence alignment program to large data—reexamination of the usefulness of chained guide trees. Bioinformatics 32: 3246-3251. https://doi.org/10.1093/bioinformatics/btw412

Yoder AD, Irwin JA, Payseur BA, Wiens J (2001) Failure of the ILD to determine data combinability for slow loris phylogeny. Systematic Biology 50: 408-424. https://doi.org/10.1080/10635150116801

Zhang J, Nielsen SE, Chen Y, Georges D, Qin Y, Wang S-S, Svenning J-C, Thuiller W (2017) Extinction risk of North American seed plants elevated by climate and land-use change. Journal of Applied Ecology 54: 303-312. https://doi.org/10.1111/1365-2664.12701

212

BIOGRAPHICAL SKETCH

Iwan Eduard Molgo was born in Paramaribo, Suriname, to Eduard Stuart Molgo and Jacqueline Theresia Molgo Huiswoud in 1974. He graduated from the Henri

Dalhberg High School in 1993 and started his undergraduate career in 1994 at the

Anton de Kom University of Suriname (AdeKUS). In 2002, he graduated from AdeKUS with a Bachelor of Science degree in agriculture and later that year he started working at the Foundation of Nature Conservation of Suriname as Research Coordinator and

Wildlife Supervisor. In 2006, he was hired as Assistant Researcher at the National

Herbarium of Suriname and was responsible for all orchid related research and the orchid collections. After working for three years, Iwan decided to begin his graduate career at the University of Florida under the supervision of Norris Williams where he studied the evolutionary relationships of Dendrophylax porrectus. Iwan graduated with a

M.S. in botany in August 2011 and continued his graduate career under the supervision of Pamela and Douglas Soltis at the University of Florida. He focused on elucidating the evolutionary relationships of the Callisia section Cuthbertia complex, endemic to the southeastern United States. Iwan graduated with a Ph.D. in botany in May 2018 and will return to his country where he will be recognized as the first Surinamese with a doctorate degree in Botany.

213