Assessing the Potential Effects of Seed Increase and Interspecific Hybridization on Genetic Diversity and Fitness of Cor...

MISSING IMAGE

Material Information

Title:
Assessing the Potential Effects of Seed Increase and Interspecific Hybridization on Genetic Diversity and Fitness of Coreopsis Leavenworthii
Physical Description:
1 online resource (180 p.)
Language:
english
Creator:
Smith,Sarah M
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Horticultural Science
Committee Chair:
Deng, Zhanao
Committee Co-Chair:
Clark, David G
Committee Members:
Gmitter, Frederick G
Norcini, Jeffrey G
Quesenberry, Kenneth H

Subjects

Subjects / Keywords:
asteraceae -- hymenoptera -- intraspecific
Horticultural Science -- Dissertations, Academic -- UF
Genre:
Horticultural Science thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Coreopsis leavenworthii, one of Florida?s state wildflowers, is nearly endemic to Florida and highly desirable for highway beautification and ecological restoration. Large-scale seed increase and planting of C. leavenworthii are becoming increasingly common in Florida. Previous studies have shown that potential genetic shifts and erosion may occur during seed increase, and natural hybridization can cause genetic contamination of produced seeds or planted populations. The objectives of this research were to 1) assess if phenotypic or molecular changes were occurring in C. leavenworthii over three successive generations at two locations, 2) assess the vegetative and reproductive fitness of synthetic F1 and F2 interspecific hybrids between C. leavenworthii and Coreopsis tinctoria and 3) determine the frequency and distance of pollen-mediated gene flow from C. tinctoria to C. leavenworthii. No significant differences or clear trends were detected between the increase and the original populations (G0) in 12 morphological, physiological, reproductive and disease resistance characteristics of C. leavenworthii. Molecular marker analysis revealed subtle changes in SSR marker alleles and allele frequencies, slight decreases in the total genetic diversity and slight increases in the genetic differentiation (GST) and genetic distances between the increase and the original populations. However, the original and the increase populations did not form any distinct clusters in principal component analysis, suggesting that the observed changes at the molecular level were not large enough to cause a significant genetic shift, and that the genetic diversity and integrity of the original population were maintained during seed increase. Controlled pollinations showed that C. leavenworthii and C. tinctoria were fully compatible. Different vegetative and reproductive fitness traits responded differently to interspecific hybridization: number of days to flower was found to be affected by heterosis, plant dry weight expressed heterosis followed by hybrid breakdown, pollen viability decreased likely due to chromosome mispairing and seed production decreased likely by chromosome mispairing and dilution effects. Inheritance studies indicated that the maroon spot is controlled by a single dominant gene and is homozygous in C. tinctoria, making it a reliable morphological marker to detect pollen-mediated gene flow. The highest rate of pollen-mediated gene flow from C. tinctoria to C. leavenworthii was 4.24%, which occurred when the two species were grown at a 1.5 m distance, and the observed greatest pollen dispersal distance was 61.0 m. Two Hymenoptera species were identified as pollinators for both Coreopsis species. Overall, the current seed production practices seem to be appropriate for C. leavenworthii seed increase. Pollen-mediated gene flow could result in negative effects to C. leavenworthii and should be prevented to protect the genetic diversity and integrity of this narrow endemic species.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Sarah M Smith.
Thesis:
Thesis (Ph.D.)--University of Florida, 2011.
Local:
Adviser: Deng, Zhanao.
Local:
Co-adviser: Clark, David G.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2011
System ID:
UFE0042834:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 ASSESSING THE P OTENTIAL E FFECTS OF S EED I NCREASE AND I NTERSPECIFIC H YBRIDIZATION ON G ENETIC D IVERSITY AND F ITNESS OF C OREOPSIS LEAVENWORTHII By SARAH MAGEN SMITH A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVE RSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Sarah Magen Smith

PAGE 3

3 I did it, Gram!

PAGE 4

4 ACKNOWLEDGMENTS I would like to th ank my advisor, Dr. Zhanao Deng for his guidance throughout my research project. I would also like to thank my co mmittee members, Dr. David G. Clark, Dr. Kenneth H. Quesenberry, Dr. Jeff G. Norcini and Dr. Fred G. Gmitter for advising me in their expertise of plant breeding, genetics and physiology I would like to thank Joyce Jones, Gail Bowman, James Aldrich, Li Gong, Rick Kelly, Tomas Hasing, Mary Derrick, Monica Raguckas and David Czarnecki for assisting me with my research. At last, I would like to thank the Florida Wildflower Found ation, Inc. for funding this research.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 11 ABSTRACT ................................ ................................ ................................ ................... 14 CHAPTER 1 LITERATURE REVIEW ................................ ................................ .......................... 16 Rationale ................................ ................................ ................................ ................. 16 Genetic Diversity and Integrity ................................ ................................ ................ 18 Significance of Genetic Diversity and Integrity ................................ ................. 18 Natural Selection, Local Adaptations, and Ecotypes ................................ ........ 19 Potential Changes in Genetic Diversity and Integrity during Seed Multiplication ................................ ................................ ................................ 20 Assessing Genetic Diversity and Integrity ................................ ............................... 22 Common Garden Studies ................................ ................................ ................. 22 Molecular M arkers to Assess Genetic Diversity ................................ ............... 22 Genetic Diversity Statistics ................................ ................................ ............... 25 Interspecific Hybridization ................................ ................................ ....................... 26 Occurrence of Natural Interspecific Hybridization ................................ ............. 26 Short term Effects ................................ ................................ ............................ 27 Long term Effect s ................................ ................................ ............................. 28 Pollen mediated Interspecific Gene Flow ................................ ................................ 30 Introduction ................................ ................................ ................................ ....... 30 Potential Effects ................................ ................................ ............................... 30 Effects of Pollination Vectors and Population Size on Pollen mediated Gene Flow ................................ ................................ ................................ .............. 31 Preventing Undesirable P ollen mediated Gene Flow ................................ ....... 32 An Overview of C. leavenworthii and C. tinctoria ................................ .................... 33 The Genus Coreopsis ................................ ................................ ...................... 33 C. leavenworthii and C. tinctoria ................................ ................................ ....... 33 Phenotypic and Molecular Diversity in Natural C. leavenworthii Populations ... 34 Genetic Diversity in C. leavenworthii Seed Increase Populations .................... 36 Compatibility between C. leavenworthii and C. tinctoria ................................ ... 36 Research Objectives ................................ ................................ ............................... 37 2 ASSESSING PHENOTYPIC CHANGES OF A NATURAL COREOPSIS LEAVENWORTHII POPULATION DURING SEED INCREASE ............................. 38 Justification ................................ ................................ ................................ ............. 38

PAGE 6

6 Materials and Methods ................................ ................................ ............................ 41 Seed Collection from a Natural Population ................................ ....................... 41 Seed Increase ................................ ................................ ................................ .. 42 Seed increase in central Florida ................................ ................................ 42 Seed increase in northern Flor ida ................................ .............................. 43 Assessing Phenotypic Changes ................................ ................................ ....... 44 Experimental design ................................ ................................ .................. 44 Gr owing conditions of the common garden study ................................ ...... 45 Data collection ................................ ................................ ........................... 45 Statistical analysis ................................ ................................ ...................... 46 Results ................................ ................................ ................................ .................... 47 Plant Height and Dry Weight ................................ ................................ ............ 47 Leaf Type ................................ ................................ ................................ ......... 48 Flower Characteristics ................................ ................................ ...................... 49 Days to flower ................................ ................................ ............................ 49 Disk flower size ................................ ................................ .......................... 49 Whole flower size ................................ ................................ ....................... 50 Petal lobing ................................ ................................ ................................ 51 Degree of petal overlap ................................ ................................ .............. 51 Number of ray petals pe r flower head ................................ ........................ 52 Seed Production and Germination ................................ ................................ ... 52 Powdery Mildew Severity ................................ ................................ ................. 53 Principal Component Analysis ................................ ................................ .......... 54 Discussion ................................ ................................ ................................ .............. 55 Summary ................................ ................................ ................................ .......... 57 3 ASSESSING MOLECULAR CHANGES USING SIMPLE SEQUENCE REPEAT (SSR) MARKERS OF A NATURAL COREOPSIS LEAVENWORTHII POPULATION DURING SEED INCREASE ................................ ............................ 70 Justification ................................ ................................ ................................ ............. 70 Materials and Methods ................................ ................................ ............................ 73 Plant Populations ................................ ................................ ............................. 73 Plant Tissue Collection and DNA Extraction ................................ ..................... 74 SSR Marker Analysis ................................ ................................ ....................... 74 Data Analysis ................................ ................................ ................................ ... 75 Results ................................ ................................ ................................ .................... 76 SSR Alleles and Allele Frequencies ................................ ................................ 76 Total Genetic Diversity ( H T ) within Populations ................................ ................ 78 Genetic Differentiation ( G ST ) and Distances among Populations ..................... 78 Principal Coordinate Analysis ................................ ................................ ........... 79 Discussion ................................ ................................ ................................ .............. 80 Genetic Differentiation between Seed Increase and Original Populations ....... 80 Possible Cau ses of Genetic Differentiation in C. leavenworthii Populations ..... 82 Upholding the Genetic Integrity of C. leavenworthii through Seed Increase ..... 84 Summary ................................ ................................ ................................ .......... 85

PAGE 7

7 4 INTERSPECIFIC HYBRIDIZATION BETWEEN COREOPSIS LEAVENWORTHII AND COREOPSIS TINCTORIA AND EFFECTS ON PROGENY GROWTH, DEVELOPMENT AND REPRODUCTION ......................... 92 Justification ................................ ................................ ................................ ............. 92 Materials and Methods ................................ ................................ ............................ 94 Seed Source ................................ ................................ ................................ ..... 94 Interspecific Pollinations and Hybrid Population Development ......................... 94 Intraspecific Pollination and Reference Population Development .................... 95 Crossability ................................ ................................ ................................ ....... 95 Assessing Progeny Growth, Development and Reproduction .......................... 96 Growing cond itions ................................ ................................ .................... 96 Data collection ................................ ................................ ........................... 97 Experimental design ................................ ................................ .................. 98 Statistical analysis ................................ ................................ ...................... 99 Results ................................ ................................ ................................ .................... 99 Crossability between C. leavenworthii and C. tinctoria ................................ ..... 99 Effects of Interspecific Hybridization on Plant Height ................................ ..... 100 Effects of Interspecific Hybridization on Plant Dry Weight .............................. 101 Effects of Interspecific Hybridization on Days to Flower ................................ 102 Effects of Interspecific Hybridization on Pollen Stainability ............................. 103 Effects of Interspecific Hybridization on Seed Production and Seed Germination ................................ ................................ ................................ 104 Seed production of hand pollinated F 1 populations ................................ .. 104 Seed production of open pollinated F 1 and F 2 populations ...................... 104 Seed germination of the hand pollinated F 1 population ........................... 106 Seed germination of open pollinated F 1 and F 2 populations .................... 106 Discussion ................................ ................................ ................................ ............ 106 Full Compatibility between C. leavenworth ii and C. tinctoria .......................... 106 Effects of Interspecific Hybridization ................................ ............................... 107 Heterosis ................................ ................................ ................................ .. 107 Heterosis followed by hybrid breakdown ................................ .................. 108 Chromosome mispairing and outbreeding depression ............................. 108 No effects dete cted on plant height and seed germination ...................... 109 Maternal Effects on Interspecific Hybridization ................................ ............... 109 Effects of Differences in Envi ronmental Conditions ................................ ........ 110 Differences in Pollen Stainability and Seed Production over Two Years ........ 111 Summary ................................ ................................ ................................ ........ 112 5 POLLEN MEDIATED GENE FLOW FROM COREOPSIS TINCTORIA TO COREOPSIS LEAVENWORTHII : IDENTIFYING MORPHOLOGICAL MARKERS AND DETERMINING GENE FLOW RATE AND POLLEN TRAVEL DISTANCE ................................ ................................ ................................ ............ 127 Justification ................................ ................................ ................................ ........... 127 Materials and Methods ................................ ................................ .......................... 131 Seed Source ................................ ................................ ................................ ... 131

PAGE 8

8 Identifying Morphological Differences between C. leavenworthii and C. tinctoria ................................ ................................ ................................ ........ 131 Inheritance of Morphological Traits ................................ ................................ 131 Crosses and populations ................................ ................................ ......... 131 Progeny evaluation ................................ ................................ .................. 134 Trait segregation analysis ................................ ................................ ........ 134 Assessing Pollen mediated Gene Flow ................................ .......................... 134 Setting up pollen source and trapping plots ................................ ............. 134 Collecting seed from pollen trapping plots ................................ ............... 135 Detecting gene flow events in C. leavenworthii progeny .......................... 136 Statistical analysis ................................ ................................ .................... 137 Results ................................ ................................ ................................ .................. 137 Crossability between C. leavenworthii and C. tinctoria ................................ ... 137 Morphological Differences between C. leavenworthii and C. tinctoria ............ 139 Expression and Inheritance of Trichomes ................................ ...................... 139 Expression and Inheritance of Maroon Spots ................................ ................. 140 Presence of maroon spots ................................ ................................ ....... 140 Size of maroon spots ................................ ................................ ............... 141 Expression and Inheritance of Seed Wings ................................ .................... 141 Pollen mediated Gene Flow from C. tinctoria to C. leavenworthii ................... 142 Discussion ................................ ................................ ................................ ............ 143 Crossability of C. leavenworthii and C. tinctoria ................................ ............. 143 Trichomes and Seed Wings were n ot Sui table for Detecting Gene Flow Events ................................ ................................ ................................ ......... 145 Presence of Maroon Spots is Controlled by a Single Dominant Gene and is an Applicable, Reliable Morphological Marker for Detecting Gene Flow Events ................................ ................................ ................................ ......... 146 Size of Maroon Spot is Controlled by Additional Genes ................................ 147 Pollen mediated Gene Flow Rate from C. tinctoria to C. leavenworthii was Lower than Expected ................................ ................................ .................. 147 Buffer Zones can Protect C. leavenworthii from Genetic Contamination ........ 148 The Need to Consider Seed D ispersal on Gene Flow in Coreopsis ............... 149 Hymenoptera on Coreopsis ................................ ................................ ............ 1 50 Summary ................................ ................................ ................................ ........ 151 6 CONCLUSIONS ................................ ................................ ................................ ... 162 APPENDIX: ENVIRONMENTAL COND ITIONS FOR GROWING YEARS DURING SEED PRODUCTION AND POPULATION EVALUATIONS ................................ 166 LIST OF REFERENCES ................................ ................................ ............................. 171 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 180

PAGE 9

9 LIST OF TABLES Table page 2 1 Analysis of variance of several phenotypic characteristics for seven C. leavenworthii populations produced in two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ........ 58 2 2 Comparison of several phenotypic characteristics of seven C. leavenworthii populations produced at two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ................................ ... 59 2 3 Standard errors of several phenotypic characteristics of seven C. leavenworthii populations produced at two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ........ 60 3 1 Primer sequences for 10 simple sequence repeat markers used to evaluate seven C. leavenworthii populations produced in northern and central Florida. ... 86 3 2 The number of individuals, number of alleles and percentage of polymorphic loci used to evaluate the genetic differences in seven populations of C. leavenworthii produced in northern and central Florida. ................................ ..... 87 3 3 Total genetic diversity ( H T ), G ST and genetic distances for seven populations of C. leavenworthii produced in northern and central Florida. ............................ 87 4 1 Analysis of variance for the number of seed produced per seed head and seed germination from C. leavenworthii and C. tinctoria hand pollinations to produce six populations at the Gulf Coast Research and Education Center, Wimauma, FL in 2007 and 2 008. ................................ ................................ ...... 114 4 2 Analysis of variance for several morphological and reproductive characteristics of six populations of C. leavenworthii and C. tinctoria grown at the Gulf Coast Research and Educat ion Center, Wimauma, FL in 2009 and 2010. ................................ ................................ ................................ ................ 115 4 3 Comparisons of F 1 and F 2 average values to mid parent values for each characteristic evaluated for C. leavenworthii and C. tinctoria interspec ific hybrids. ................................ ................................ ................................ ............. 116 4 4 Comparison of C. leavenworthii and C. tinctoria intraspecific and interspecific F 1 and F 2 populations. ................................ ................................ ...................... 117 4 5 Temperature and relative humidity averages throughout the growing year at the Gulf Coast and Education Center, Wimauma, FL in 2009 and 2010. ......... 118 4 6 Day length averages throughout the gro wing year at the Gulf Coast and Education Center, Wimauma, FL in 2009 and 2010. ................................ ........ 118

PAGE 10

10 5 1 Number of seed collected per seed head for each cross type made at the Gulf Coast Research and Education Center, Wimauma, FL. ........................... 153 5 2 Inheritance of the maroon spot for C. leavenworthii and C. tinctoria self, F 1 F 2 and testcross populations. ................................ ................................ ........... 154 5 3 Inheritance of seed wingedness in crosses with C. leavenworthii and C. tinctoria ................................ ................................ ................................ ............ 155 5 4 Chi square table to determine the probability of gene flow at a specific distance. ................................ ................................ ................................ ........... 155 5 5 Hymenoptera found in the field on C. leavenworthii and C. tinctoria plants at Gulf Coast Research and Education Center, Wimauma, FL. ........................... 156 A 1 Environmental conditions during seed collection period for seed increase for the genetic diversity studies at the North Florida and Gulf Coast Research and Education Centers during 2007, 2008 and 2009. ................................ ...... 166 A 2 Environmental conditions at monthly intervals during seed collection period for seed increase for the genetic diversity studies at the North Florida and Gulf Coast Research and Education Centers during 2007, 2008 and 2009. .... 167 A 3 Temperature and relative humidity averages throughout the growing year during the evaluation of the seed increase populations at the Gulf Coast Research and Education Center, Wim auma, FL in 2009 and 2010. ................. 168 A 4 Day length averages throughout the growing year during the evaluation of the seed increase populations at the Gulf Coast and Education Center, Wimauma, FL in 2009 and 2010. ................................ ................................ ...... 168 A 5 Soil and tissue nutrition components for Coreopsis plants grown in the field during the 2010 year at the Gulf Coast Research and Education Center, Wimauma, FL. ................................ ................................ ................................ .. 169

PAGE 11

11 LIST OF FIGURES Figure page 2 1 Leaf types observed in C. leavenworthii with assigned scores (Czarnecki et al., 2007). ................................ ................................ ................................ ........... 61 2 2 Distribution of leaf type scores in the natural population and six seed increase populations of C. leavenworthii produced at two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ................................ ................................ ................................ .. 62 2 3 Scores used for evaluating petal lobing of seven C. leavenworthii populations produced at two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ................................ .......... 63 2 4 Changes in petal lobing scores for seven populations of C. leavenworthii produced at two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ................................ ......... 64 2 5 Scores used for evaluating petal orientations of seven C. leavenworthii populations produced at two locations and grown at the Gulf Coast Research and Education C enter Wimauma, FL in 2009 and 2010. ................................ ... 65 2 6 Changes in the degree of petal overlap ratings for seven populations of C. leavenworthii produced at two locations and grown at the Gulf Coast Resea rch and Education Center, Wimauma, FL in 2009 and 2010. ................... 66 2 7 Differences in the number of seeds per five seed heads of seven C. leavenworthii populations produced at two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ........ 67 2 8 Distribution of powdery mildew severity among individuals for each of the seven populations of C. leavenwo rthii produced at two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ................................ ................................ ................................ ........... 68 2 9 Principal component analysis of C. leavenworthii populations gro wn in 2009 and 2010 at the Gulf Coast Research and Education Center, Wimauma, FL. .... 69 3 1 Gel image of simple sequence repeat marker bands (alleles) amplified from C. leavenworthii individuals by COR12 and detected by the LI COR 4300 DNA Analyzer. ................................ ................................ ................................ .... 88 3 2 The number of SSR marker alleles with respect to their frequencies of occurrence in the C. leavenworthii source population (G 0 ), t hree populations increased in central Florida (G 1 C, G 2 C and G 3 C) and three populations increased in northern Florida (G 1 N, G 2 N and G 3 N). ................................ ............ 88

PAGE 12

12 3 3 Regression of G ST values when compared by the di fference in the number of generations between each combination of seven C. leavenworthii populations. ................................ ................................ ................................ ........ 89 3 4 Regression of the genetic distances when compared by the difference in the number o f generations between each combination of seven C. leavenworthii populations. ................................ ................................ ................................ ........ 89 3 5 UPGMA dendrogram of the G 0 population and six seed increase populations (three from central Florida: G 1 C, G 2 C and G 3 C; and three from northern Florida: G 1 N, G 2 N and G 3 N) of C. leavenworthii populations based on their pairwise genetic distances. ................................ ................................ ................. 90 3 6 Plot of 50 individuals of the G 0 ( ) population and 299 individuals of six seed increase populations ( G 1 C, G 2 C, G 3 C, G 1 N, G 2 N, G 3 N) of C. leavenworthii based on the first two principal coordinates from 108 SSR alleles. ................................ ................................ ................................ ................ 91 4 1 Crosse s to produce populations to compare the interspecific F 1 and F 2 populations to intraspecific populations of C. leavenworthii and C. tinctoria .... 118 4 2 Differences in the amount of seed produce d per seed head by hand pollinations of six populations of C. leavenworthii and C. tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2007 and 2008. ................................ ................................ ................................ ............... 119 4 3 Di fferences in the seed germination rate of seed produced from hand pollinations of six populations of C. leavenworthii and C. tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ................................ ................................ ................................ ................ 120 4 4 Differences in plant height (cm) of six populations of C. leavenworthii and C. tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ................................ ................................ ...... 121 4 5 Differences in plant dry weight (kg) of six populations of C. leavenworthii and C. tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ................................ ................................ ...... 122 4 6 Differences in the number of days to flower of six populations of C. leavenworthii and C. tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL i n 2009 and 2010. ................................ ........ 123 4 7 Differences in the rate of pollen stainability of six populations of C. leavenworthii and C. tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ................................ ........ 124

PAGE 13

13 4 8 Differences in the amount of seed produced from five seed heads in the field of six populations of C. leavenworthii and C. tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2 010. ...... 125 4 9 Differences in the rate of seed germination of seed collected in the field of six populations of C. leavenworthii and C. tinctoria produced at the Gulf Coast Research and Educatio n Center, Wimauma, FL in 2009. ................................ 126 5 1 Morphological characteristics assessed for potential use as a morphological marker to det ect pollen mediated gene flow. ................................ .................... 157 5 2 Field design of one block out of three for the pollen mediate gene flow study. 158 5 3 The inheritance of the size of the maroon spot in F 1 hybrids from reciprocal crosses between four C. leavenworthii and four C. tinctoria parents. ............... 159 5 4 Gene flow rates occurring from C. tinctoria to C. leavenworthii plants at multiple distances in 2007 and 2008 at the Gulf Coast Research and Education Center, Wimauma, FL. ................................ ................................ ..... 160 5 5 Scatter plots of observed gene flow rates at each distance measured and logistic regression curve for the equation f itted to the rate of gene flow from C. tinctoria to C. leavenworthii at each distance over two years. ...................... 161 A 1 Differences in seed germination of seven C. leavenworthii populations produced at two locations and sown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. ................................ ........ 170

PAGE 14

14 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fu lfillment of the Requirements for the Degree of Doctor of Philosophy ASSESSING THE P OTENTIAL E FFECTS OF S EED I NCREASE AND I NTERSPECIFIC H YBRIDIZATION ON G ENETIC D IVERSITY AND F ITNESS OF COREOPSIS LEAVENWORTHII By Sarah Magen Smith August 2011 Chair: Zha nao Deng Cochair: David G. Clark Major: Horticultural Science Coreopsis leavenworthii one of s, is nearly endemic to Florida and highly desirable for highway beautification and ecological restoration. Large scale seed increase and planting of C. leavenworthii are becoming increasingly common in Florida. Previous studies have shown that potential genetic shifts and erosion may occur during seed increase and natural hybridization can cause genetic contamination of produced seeds or planted populations. The objectives of this research were to 1) assess if phenotypic or molecular changes were occurring in C. leavenworthii over three successive generations a t two locations 2) assess the vegetative and reproductive fitness of synth etic F 1 and F 2 interspecific hybrids between C. leavenworthii and C oreopsis tinctoria and 3 ) determine the frequency and distance of pollen mediated gene flow from C. tinctoria to C. leavenworthii No significant differences or clear trends were detected between the increase and the original populations (G 0 ) in 1 2 morphological physiological reproductive and disease resistance characteristics of C. leavenworthii Molecular marker analysis revealed subtle changes in SSR marker alleles and allele frequenc ies, slight decreases in the

PAGE 15

15 total genetic diversity and slight increases in the genetic differentiation ( G ST ) and genetic distances between the increase and the original populations. However, the original and the increase populations did not form any dis tinct clusters in principal component analysis, suggesting that the observed changes at the molecular level were not large enough to cause a significant genetic shift, and that the genetic diversity and integrity of the original population were maintained during seed increase. Controlled pollinations showed that C. leavenworthii and C. tinctoria were fully compatible. Different vegetative and reproductive fitness traits responded differently to interspecific hybrid ization: number of days to flower was foun d to be affected by heterosis, plant dry weight expressed heterosis followed by hybrid breakdown, pollen viability decreased likely due to chromosome mispairing and seed production decreased likely by chromosome mispairing and dilution effects Inheritanc e studies indicated that the maroon spot i s controlled by a single dominant gene and is homozygous in C. tinctoria making it a reliable morphological marker to detect pollen mediated gene flow The high est rate of pollen mediated gene flow from C. tincto ria to C. leavenworthii was 4.24%, which occurred when the two species were grown at a 1.5 m distance and the observed greatest pollen dispersal distance was 61.0 m. Two Hymenoptera species were identified as pollinators for both Coreopsis species. Over all, the current seed production practices seem to be appropriate for C. leavenworthii seed increase. Pollen mediated g ene flow could result in negative effects to C. leavenworth i i and should be prevented to protect the genetic diversity and integrity of this narrow endemic species

PAGE 16

16 CHAPTER 1 LITERATURE REVIEW Rationale has become broadly accepted by our society and is often required by federal, state or local laws or policy Exclusive native plantings have become a comm on practice in parks, forests, and natural areas, and increasingly common in roadside treatments ( Rogers and Montalvo, 2004 ) One group of native plants that have been particularly popular is native wildflowers (forbs). Many states, including Florida, Id aho, Michigan, Ohio Texas and Wisconsin have adopted n ative wildflower planting programs for highway beautification Us e of native wildflowers along roadside s has not only increased aesthetic value s but also reduced maintenance costs, enhanced wildlife habitat and biodiversity, augmented soil erosion control and suppressed noxious weeds ( Bryant and Harper Lore, 1997 ). Considering the mounting evidence of ecologically important phenotypic and genetic differences among populations within a species, n ative planting protocol s often suggest using local provenances or ecotypes as genetically appropriate plant or seed sources ( Rogers and Montalvo, 2004 ). It is expected that local ecotypes have developed adaptation, and local environments may favor local ecotyp production, soil seed bank development, seed germination and/or seedling establishment. Thus it is anticipated that local ecotypes may have better opportunities to become self sustaining with appropriate management ( Norcini et al. 200 1 2009 ). The genus Coreopsis C. leavenworthii is one of the 13 species of Coreopsis found in Florida (Gilman et al., 2007) This species is restricted to and ubiquitous in Florida ( Kabat et al., 2007 ) Natural populations have been

PAGE 17

17 documented Wunderlin and Hansen, 2004 ). The common habitats of this species include roadside ditches, wet pine flatwoods, and other mois t disturbed sites ( Kabat et al., 2007 ). Plants of this species can blo o m year round in south Florida and form a dense flower cover over the foliage. For these reasons C. leavenworthii is very desirable for use in hi ghway beautification projects. T he Florida Department of Transportation (FDOT) has used C. leavenworthii in a number of beautification projects and expressed strong interest to expand its current wildflower program and plant more local ecotype s of native wildflower s on roadsides, such as C. leavenworthii Florida wildflower growers have collected seeds from nat ural populations and increased seeds to meet the demand (Norcini and Aldrich 2004). It is anticipated that more seeds of C. leavenworthii will be produced by wildflower growers and used on more roadsides across the state. I n recent years, natural populat ions of C. leavenworthii in some areas of south Florida seem to be diminishing. This has led to a suspicion that diversity may be declining due to factors such as habitat deterioration or destruction, roadside mow ing practices and urb anization C oncerns have been also raised about th e possibility that genetic shift and erosion may occur during seed increase It has been suspected that certain genotypes of native plants may be better adapted to seed increase practices su ch as fertiliz ation, irrigation and harvest date and that these genotypes m ay produce more seeds than other genotypes (Rogers and Montalv o 2004). Additionally, C. leavenworthii is sexually compatible with C. tinctoria (Parker, 1973; Smith, 1976, 1983). Although thes e two species are sexually compatible, the lack of pollen stainability and chromosome homology supports their species status (Parker,

PAGE 18

18 1973) This species occurs naturally in a number of counties in Florida and has been used in highway beautification proje cts in Florida (Gilman et al., 2007; Parker, 1973; USDA, 2011b). There is real potential for C. leavenworthii and C. tinctoria to hybridize naturally and cause genetic contamination or swamping which is the replacement of local genotypes as a result of a numerical and/or fitness advantage of immigrant genotypes Thus there is a strong need to assess the effect s of seed increase, interspecific hybridization and natural gene flow on the genetic diversity and genetic integrity of C. leavenworthii Genetic D iversity and Integrity Significance of Genetic Diversity and Integrity Genetic diversity is the variation in DNA among individual s and can change due to local adaptive pressures and other processes that can influence the mating success and survival of indi viduals (Hartl, 2000 ; Rogers and Montalvo, 2004 ). Along with mutation, genetic diversity provides the variation for adapt ation to new conditions and establish ing in new habitats. Genetic diversity is also the means for a s peci es to become linked to other organisms such as pollinators, which may co evolve with the species (Futuyma, 1979) H ighly significant and positive relationship s between genetic diversity and fitness have been documented in many plants ( Reed an d Frankham, 2003 ) Genetic integrity refers to the quality and arrangement of genetic diversity in relation to natural processes that reflects changes in genetic composition caused by local natural selection and other processes th at can affect mating and surviv al of individuals Conceivably, a nything that compromises genetic diversity and integrity will likely have cascading negative effects on the species itself and associated

PAGE 19

19 organisms (Havens, 1998) As such, and in face of accelerated chan ges in climate, maintaining the quality and quantity of genetic diversity has been considered as a critical component to the long term sur vival of native plant species. However, it has been often very difficult to s ense the urgency of such action because the loss of genetic diversity is often cryptic the effects may not be obv ious or visible for many years and g enetic integrity can be severely degraded without an obv ious loss of genetic diversity. Natural Selection, Local Adaptations, and Ecotypes The gen etic diversity of a species is affected by several natural processes, including mut ation, migration, genetic drift and selection (Futuyma, 1979) Natural selection is the best known of the processes affecting genetic diversity and is the only process that directly results in populations becoming better adapted to the environment. In response to natural selection from specific local or site specific environmental conditions, such as tem perature, day length, soil type and moisture level, certain populations of a species may become better adapted to the local environment (Booth and Jones, 2001) Such locally adapted populations have been recognized as ecotypes (Turesson, 1922) Different ecotypes may also differ in phenotypic characters, such as plant heigh t, growth habits, leaf characters, earliness of maturity and reproductive habits (Grant, 1981) Genetically, different ecotypes may have different combinations of adapted genes and/or alleles. In light of the existence of ecologically important phenotypic and genetic differences among populations within a species, n ative planting protocols have urged using local ecotypes as the genetically appropriate plant or seed sources (Rogers and Montalvo, 2004). The expectation is that l ocal ecotypes are most adapte d to local conditions and most tolerant of local stresses T hus they will more likely become

PAGE 20

20 established and self sustained, which should result in reduced maintenance costs and better plant performance. Additionally, planting local ecotypes can help pre serve the genetic diversity and integrity of native plants and pres erve local pollinators, insects and other wildlife that have co evolved with the plants of the local ecotype s and depend upon them for food and shelter Potential Changes in Genetic D iversi ty and I ntegrity during S eed M ultiplication Seed or plant increase is often required for large scale planting of native species. Seed increase generally starts with collection of seed from natural populations followed by multiple generations of multiplica tion. This practice can make seed available in much larger quantities and result in lower costs of seeds In the meantime, it protect s native population s from over harvest However, every stage of the increase process fro m collection, planting location production methods, harvesting practices to seed storage can potentially cause unintentional genetic shifts in traits and erosion of genetic variation ( Knapp and Rice, 1994; Meyer and Mon sen, 1993). During collection, seed are often collected from a por tion of the natural populations. Thus only a portion of the alleles in the natural populations will be captured in the wild collection (Ellstrand and Elam, 1993; Hartl, 2000). Loss of genetic diversity and genetic integrity may occur if the sampled seeds do not represent the level and structure of genetic diversity in the natural populations S eed dormancy and seed shattering can be important adaptive traits, but they may be selected against and lost unintentionally due to artificial cultivation and harve st practices ( Cai and Morishima, 2002 ). Seed dormancy may be lost after several generations of seed increase. Harvesting seeds in a narrow time window may limit genetic variation for timing of flowering, while harvesting toward the end or beginning of

PAGE 21

21 se ed maturity can result in genetic shifts. Growth rate and the timing of flowering could also be subje ct to unintentional selection. Seed increase in a single location for multiple generations can result in adaptation to the specific production location or production practices (Burton and Burton, 2002). When seed increase is done in a habitat different from the original habit, multiple plant characteristics may change significantly and even become genetically fixed, which can be detrimental to the populati ons being increased (Antonovics, 1976). To avoid unintentional natural selection seed increase should be started from a su fficient number of individuals, using appropriate practices, and done at appropriate site s where pollinators a re present (Havens, 19 98). Seed increase in a single location for too many generations can result in local adaptation (Burton and Burton, 2002). To minimize such kind of unintentional selection during seed increase the National Resources Conservation Service ( NRCS ) allows a ma ximum of four generations of seed increase for cultivars, tested releases and selected releases. For source identified seeds, the NRCS and Association of Official Seed Certifying Agency ( AOSCA ) allow unlimi ted generations of seed increase, if no genetic sh ifts or erosion occur (Rogers and Montalvo, 2004) It has been stressed that t o maintain the genetic diversity and integrity of the natural population s being increased, the genetic diversity of the starting materials should be ass essed and used as a refer ence, and the genetic diversity and integrity of increased populations should be monitored closely. Other measures have been proposed t o counterbalance the selection for local adaptation during seed increase. One proposed measure is to allow 1 5% of gene flow

PAGE 22

22 between populations per generation (Ellstrand and Elam, 1993). The authors consider that a gene fl ow frequency of 10% or more may have a significant effect on fitness where as 1% or l ess will be of little concern. Assess ing Genetic Diversity and Inte grity Common Garden Studies Genetic diversity and integrity can be assessed at different levels and using a number of ways (Hartl 2000). Before molecular markers became available, the assessment of genetic diversity and integrity was primarily based on t he examination of phenotypic variation within populations ( Berg and Hamrick, 1997 ). Phenotypic variation is generally controlled by many genes and can be influenced by environmental factors to varying degrees. To minimize the influence of environmental c onditions on Hamrick 1997 ; Clausen et al., 1940 ). In common garden studies individuals of a population are gr own in one or more common environments (Rogers and Montalvo 2 004). This approach is necessary and still commonly used. When assessing characteristics to determine if there is a loss of genetic diversity, characteristics of the entire life cycle should be evaluated (Edmands, 2007). Molecular Mark ers to Assess Genet ic Diversity Since the mid 1960s, a number of biochemical and molecular techniques has beco me available ( Harris, 1966 ). These techniques have allowed identification of alleles at many gene loci. Molecular markers represent discrete genetic differences an d are generally insensitive to environme ntal changes, plant development and physiological changes (Hartl 2000). Thus molecular markers ha ve become a valuable approach for assessing genetic diversity and integrity ( Weising et al., 2005 ).

PAGE 23

23 Many different t ypes of molecular markers have been used to assess the genetic diversi ty within and among populations, including allozymes restriction fragment length polymorphism (RFLP), random amplified polymorphic DNA (RAPD) a mplified fragment length polymorphism (AF LP), inter simple sequence repeats (ISSR) and simple sequence repeats or microsatellites (SSRs) ( Weising et al., 2005 ). AFLP and SSR markers have been the most widely used markers in genetic diversity and integrity asse ss ments (Mueller and Wolfenbarger, 1 999). AFLP is a useful marker to determine the genetic diversity among individuals and populations (Mueller and Wolfenbarger, 1999). AFLP allows for the screening of numerous DNA regions distributed throughout the genome. It has been reported that AFLP m arkers have the widest application when analyzing genetic variation within and among populations, which can help determine conservation steps needed for a species. When producing AFLP markers, careful marker design is very important for successful PCR amp lification, especially for the adapters (Vos et al., 1995). Selectivity of the markers is acceptable with three selective nucleotides but is lost with four selective nucleotides. The number of amplified fragments may be dependent on the rare cutter enzym e and number of selective bases. There are several benefits for using AFLP markers, such as DNA samples of any origin or complexity can be used because AFLP has the ability to assess numerous loci at one time making this marker system very efficient and it reveals many polymorphic bands with a small quantity of DNA (Blears et al., 1998; Mueller and Wolfenbarger, 1999; Vos et al., 1995). This marker system is reliable, reproducible, easy, affordable and fast to generate. The number of markers produced is l imitless based on the

PAGE 24

24 different combinations of restriction enzyme combinations available and the number and combinations of selective nucleotides, allowing for minor genetic differences to be detected. There is no need of prior knowledge of the DNA of an organism before development or use and small amounts of DNA are needed that segregate in a Mendelian fashion (Mueller and Wolfenbarger, 1999). Variation is eliminated due to the high selectivity of AFLP markers. The main problem with this marker system i s that it is a dominant marker, which does not allow for homo zygous alleles to be distinguished from heterozygous alleles and can be less useful than codominant markers (Blears et al., 1998; Mueller and Wolfenbarger, 1999). Even though AFLP is known to be highly reproducible, there are some issues affecting the reproducibility of this marker system, such as the DNA used must be of high purity for complete digestion by the restriction enzymes, which can result in altering banding patterns (Blears et al., 19 98). If DNA quantities are too low due to dilution the DNA sequences flanking the restriction enzymes will not be random for some restriction fragments and the banding patterns may be altered. SSR markers are those that detect variation in the number of short repeat sequences that change in a high frequency and detect multiple alleles (Langridge and Chalmers, 2004). SSRs are developed from DNA regions that flank specific microsatellite repeats, called primer binding sites, and are amplified using PCR (Gl enn and Schable, 2005). The primer binding sites are not well conserved among distantly related species and are the most variable in the genome. SSR markers are densely distributed throughout the genome, highly variable and possibly the most powerful gen etic marker available (Goldstein and Pollock, 1997; Zane et al., 2002). SSRs show

PAGE 25

25 a large amount of polymorphism because of length variation due to the occurrence of different numbers of repeats and are widely distributed in higher plants at about one SSR for every 50 kb and are the ideal genetic markers in plants (Morgante and Olivieri, 1993). They are inherited in a Mendelian fashion, codominant and useful for population studies. Because SSR markers are not conserved, they usually have to be developed f or each species, which can be time consuming to detect, target and isolate (Glenn and Schable, 2005; Goldstein and Pollock, 1997; Zane et al., 2002). They are commonly found in noncoding regions of the DNA where nucleotide substitutions are more common. SSR makers can have a high mutation rate and can be unstable due to subtracting or adding a small number of perfect repeats and can be caused by polymerase slippage caused by slip strand mispairing errors during DNA replication and by unequal crossover or gene conversion of DNA strands during recombination (Goldstein and Pollock, 1997; Li et al., 2002). Genetic Diversity Statistics Appropriate sampling of individuals is critical for obtaining reliable estimates of genetic diversity parameters. S ample sizes that are too small can lead to underestimation of the true genetic diversity. Numerous modeling studies have shown that a sample size of at least 30 individuals should be used in molecular studies for estimating genetic diversity and integrity (Berg and Hamrick, 1997; Sinclair and Hobbs, 2009 ). The number of individuals needed to be evaluated may be different based on the species, where population structure and pollination biology can influence the movement of alleles within and among populations.

PAGE 26

26 Severa l parameters have been frequently used in molecular genetic diversity assessment, including allele frequency, expected heterozy gosity, total genetic diversity and genetic differentiation. Allele frequencies provide the base for estimating the other geneti c parameters. The most important assumption for estimating these parameters is that the populations analyzed are in Hardy Weinberg equilibri um (Berg and Hamrick, 1997). As described in population genetics, Hardy Weinberg equilibri um can be approached und er the following conditions: populations undergoing sexual reproduction, random mating, non overlapping generations, large population size, no migration, mutation or selection and equal gamete production. With Hardy Weinberg equilibrium assumed, expected h eterozygosity ( H E ) can be calculated from the observed allele frequencies (Berg and Hamrick, 1997). This parameter measures the genetic variation at the allele level and is often referred to as genetic diversity. The magnitude of H E depends on the propor tion of polymorphic loci, the number o f alleles per polymorphic locus and the distribution of allele frequencies within the population. Total heterozygosity ( H T ) is the sum of the mean genetic diversity within populations plus the genetic diversity among populations. The G ST value is an estimate of the genetic diversity that resides among populations, i.e. the genetic differentiation or divergence among populations Interspecific Hybridization Occurrence of Natural I nterspecific H ybridization Natural inte rspecific hybridization has been known to be widespread between many plant species (Arnold, 1996). C ompati bility between species is one of the main factors determining the frequency of natural hybridization. For interspecific hybridization to occur cros s pollination must occur between two species that are flowering at the

PAGE 27

27 same time and are growing close enough for a vector to transfer pollen followed by successful fertilization that develop s viable seed s that germinate and develop into a viable hybrid pl ant. Short term Effects The degree of interspecific hybridization between plants varies depending on the location and species, which also affects the level of compatibility, where few hybrids can be produced or two species can evolve into one species (Ells trand, 2003 b ). Natural hybridization within plants, where a single hybridization event can result in numerous evolutionary lineages, can be affected by the environment (Arnold, 1996). Typically, ability to survive and produce offspring due to selection that can be caused by genetic drift (Futuyma, 1979). The fertility of interspecific hybrids depends on the species and the environment where they grow (Arnold and Hodges, 1995). Some interspecific hybrids are less fit than either parent al species and can lead to nonviable F 1 progeny, F 1 hybrids that are viable but infertile or viable and fertile F 1 hybrids. Typically interspecific hybrids have a degree of sterility but are rarely completely steril e, where in a population of interspecific hybrids some hybrid genotypes have higher fitness than either parent and even those with lower fitness can have ecological consequences (Arnold et al., 1999; Ellstrand, 2003 b ; Ellstrand et al. 1999). M ost inters pecific hybrids are intermediate between the parent species and the progeny produced, typically from backcrossing, from these parents typically resemble one of the parental species (Stebbins, 1959). Even if the first generation of hybridization produces hy brids with limited fertility, later generations of hybridization may become more fertile (Arnold and Hodges, 1995). The percentage of interspecific progeny formed from highly sterile to completely sterile

PAGE 28

28 individuals may be small but can have a large effe ct biologically (Grant, 1981). Partially fertile interspecific F 1 hybrids can reproduce by selfing, sibcrossing or backcrossing that produces a second generation progeny that can continue crossing with the original plants resulting in a hybrid swarm that is an extremely variable mixture of species and interspecific hybrids. If a hybrid is less vigorous than either parent species in a certain habitat, it is less likely to be able to reproduce and disperse in that habitat (Harper et al., 1961). On the othe r hand, if the hybrids are more adapted to a habitat than the parent species, the hybrids could thrive and closely related species would evolve together. Hybrids that inbreed for further generations are more likely to become stabilized into a fertile speci es with characteristics that are intermediate between the parent species or resemble an excessive form of one of the parent species (Abbott, 1992). New species formed through introgression are expected to have reduced genetic diversity, which are expected to increase at certain loci that are very different in both parent species. Once interspecific hybridization occurs, the level of stability of the new gene complexes will depend on mechanisms preventing recombination and segregation. Long term Effects Th e effects of natural hybridization can be considered negative when rare species are lost resulting in outbreeding depression or genetic assimilation (Arnold, 1996). The negative connotation of interspecific hybridization stems from the fact that most inte rspecific hybridizations result in reduced fertility and viability. This is not true of all natural hybridizations and certain interspecific genotypes will have increased fitness from either parental species in certain environments, which could lead to ad aptation into novel habitats or the displacement of either parental species in a habitat that is a better

PAGE 29

29 fit for the hybrid genotype. Many would have a negative opinion of such a displacement, but others would argue that this could be the evolution of ce rtain species. Natural hybridization can be a creative and ongoing process in evolution of numerous groups of organisms. Natural hybridization can have a positive effect when there is genetic enrichment of an endangered form of a rare species, increased genetic variability that can allow habitat expansion and using the hybrids to integrate beneficial phenotypes and genotypes back into the parental species. Interspecific hybrids can exhibit increased fitness through heterosis. Heterosis caused by interspe cific hybridization can lead to more fit progeny due to recombination of genes from both parent species, especially if the parents are inbred, particularly in environments well suited for the hybrids (Barton, 2001). Interspecific hybrids may be more fit t han either parent species due to the hybrid population having more diverse genotypes, the ancestral lineage is reconstructed in the F 2 generation and fitness is determined by the interactions of the recombinant genes. Overall, hybrids formed from individu als of distinct environments will be less fit on average but some individuals with distinct genotypes will be more fit than either parent. Generally, more divergent populations or species will create hybrids with lower fitness (Edmands, 2007). A decline i n fitness due to outcrossing between distinct individuals or species is termed outbreeding depression and causes a significant decline in the F 1 hybrids relative to either parent or the mid parent value and is environmentally dependent (Edmands, 2007; Ells trand et al., 1999; Havens, 1998). There are two genetic mechanisms caused by outbreeding depression, dilution and hybrid breakdown (Hufford and Mazer, 2003). Dilution occurs after hybridization due to

PAGE 30

30 the divergence of locally adapted genotypes that hav e diverged because of natural selection. Hybrid breakdown occurs from hybridization of individuals from distinct environments or from the mating of individuals with distinct gene complexes. Sterility of interspecific hybrids is usually due to genic diffe rences or chromosomal rearrangements (Stebbins, 1959). Pollen mediated Interspecific Gene Flow Introduction Gene flow between closely related species occurs when species are symp atric (Harper et al. 1961) T he effectiveness of breeding barriers, the chan ces of past migrations and the similarities of habitats where species grow can affect the frequency of gene flow Animal and wind borne pollen dispersal usually follows a leptokurtic distribution wh ere a high frequency of gene flow occurs in a small rang e with little variance where pollen is rarely distributed more than 1000 m and seed is rarely distributed more than 200 m with most gene flow occurring within 20 m of the source population (Ellstrand, 1992a; Levin, 1981). There are four main variables af fecting cross pollination: the breeding system of the species, the isolation distance, plant density and the pollination vector (Bateman, 1947). Potential Effects The effects of gene flow depend on the amount and the level of success of gene flow between t wo species, which depends on the species, populations, individuals and year s (Ellstrand et al., 1999). A small amount of interspecific gene flow can have a large impact on counterbalancing evolutionary forces, such as mutation, genetic drift and selection Gene flow rates measured experimentally have typically exceeded 1% at distances of 100 m or more.

PAGE 31

31 Interspecific gene flow is most problematic when two species interbreed significantly and are relatively compatible resulting in the production of a substa ntial amount of seed (Ellstrand and Elam, 1993). Another effect of gene flow can be caused by two distinct species growing in the same habitat and hybridizing freely causing the two populations to merge into one population and thus one species idize out E ffects of Pollination Vectors and Population Size on Pollen mediated Gene Flow Most pollen collected by an insect is deposited on the next plant visited by the pollination vector, where the flight patterns of insects are random in direction (Levin, 1981). Usually there is pollen carryover for multiple plants, typically up to seven plants. The distances traveled by the vector are independent of the compatibility of the plants, seed viability and seedl ing surv ivorship. Pollen carry over and direction of flight are dependent on floral reward and can affect pollen dispersal. Pollen deposition on a particular flower head is determined by the amount of time a pollinator spends on a flower, where more time will be spent at flower heads with a higher reward and there is a positive energy gain. This leads to less pollen carry over and gene flow in nectar rich populations. Population size of both species greatly affects the potential rate of gene flow, where typicall y the gene flow rate increases as population size decreases because there are fewer targets in smaller populati ons (Ellstrand, 1992a 2003 a ; Ellstrand et al., 1999; Ellstrand and Elam, 1993). Larger source populations and smaller sink populations can resu lt in greater gene flow frequencies because the larger source populations will produce greater amounts of pollen and seed (Ellstrand et al., 1999; Ellstrand and Elam, 1993). Gene flow can disrupt local adaptation and cause extinction

PAGE 32

32 of small populations depending on the degree of isolation and the frequency of successful hybridization with compatible species with greater effects in smaller populations (Antonovics, 1976; Ellstrand, 1992a). If the population size decreases rapidly, the rate and distance of gene flow can increase rapidly moving past the boundaries of a population, which can cause swamping effects to be substantial (Antonovics, 1976). Foraging pollen vectors will spend more time in a large population compared to a small population and will f ly farther distances when plant density is low, potentially causing a greater swamping effect and a concern for plant conservation (Antonovics, 1976; Ellstrand, 1992a; Ellstrand et al., 1999; Levin, 1981). Preventing Undesirable Pollen mediated Gene Flow T o prevent unwanted gene flow into a population, pollinators and flowering times can be managed, immigrant pollen can be intercepted by planting other host populations around the population of interest or total eradication of the threatening species and int erspecific hybrids (Ellstrand, 1992a). Some species prevent interspecific hybridization through several forms of reproductive isolation, such as pollen competition (Coyne and Orr, 1998). The heterospecific pollen tubes do not grow as quickly as the consp ecific pollen tubes greatly decreasing the amount of interspecific seed produced. Generally 50 m or more is a great enough distance to prevent gene flow between two species but high outcrossing species can require about 500 m (Ellstrand and Elam, 1993). Gene flow can extend its range until it is stopped by barriers to dispersal such as climate, predators, competitors and resource availability (Slatkin, 1987). Strong barriers to prevent pollination between species can be difficult to produce if the specie s is constantly evolving (Ellstrand, 2003 a ).

PAGE 33

33 An Overview of C leavenworthii and C. tinctoria The Genus Coreopsis The genus Coreopsis is a member of Asteraceae one of the largest families of flowering plants (Heywood, 1993). Members of this family have a variety of growing habits and can be found worldwide, except the Antarctica mainland. The family has been split into subfamilies and then further into tribes. Coreopsis belongs to Asteroideae subfamily and Coreopsideae tribe (Heywood, 1993; Ryding and B remer, 1992). Th is genus consists of 8 sections ; one of the most advanced sections is Calliopsis (Smith, 1975, 1983). Coreopsis is distributed throughout the Americas, the near Pacific Islands and Africa (Smith, 1976). The genus Coreopsis contains about 55 species that are distributed across the United States, thirteen of which grow in the state of Florida, including C. lanceolata C. grandiflora C. tinctoria C. basalis and C. leavenworthii (FDOT, 1995, 2006; Gilman et al., 200 7 ; Norcini, 2002; Sherff, 1955). Eleven species of Coreopsis are considered native to Florida, including C. leavenworthii (Gilman et al., 200 7 ). C. leavenworthii and C. tinctoria Both species belong to the Calliopsis section and are known to be outcrossing species that are insect pollinated (Clewall, 1985; Wunderlin, 1998) C. leavenworthii has been reported to be growing in most counties in Florida and has been reported in only one other state, Alabama, and is considered endemic to Florida (Gilman et a l., 2007 ; Parker, 1973; USD A, 2011a). C. leavenworthii is an annual to short lived perennial with bright yellow ray flowers with a dark brown center, which are the disk flowers, that flowers from late spring in north Florida and any time in south Florida

PAGE 34

34 (Gilman et al., 2007 ). C. leavenworthii seed is solely produ ced in Florida (Norcini and Aldrich, 2007 ) C. tinctoria is an annual forb that overwinters as a rosette and germinates in late summer or fall (Whitten, 2002). The ray flowers are yellow with a red brown portion close to the disk flowers, which are dark brown. This species grows in many soil types and best in full sun. C. tinctoria has been reported to be growing in most of the United States, including six counties in Florida but is not considered native to Florida (Gilm an et al., 2007 ; Parker, 1973; USDA, 2011b). C. tinctoria is used mostly for landscape beautification that can be used in gardens, naturalized prairie or meadow plantings and along roadsid es (Whitten, 2002). Phenotypic and Molecular Diversity in Natural C leavenworthii Populations Because the nectar rewards of Asteraceae are small, plants within a population must be close in proximity and pollen dispersal will usually remain within that population because bees can probe many florets quickly and maintain a fairly high rate of energy intake (Price, 1997; Schmitt, 1980). In general Asteraceae must rely on long distance seed dispersal for the intermixing of populations and to keep genetic diversity levels high. Czarnecki et al. (2007) evaluated the phenotypi c variation in natural populations of C. leavenworthii from north, central and south Florida. Differences were found among populations in characteristics such as growth habits, leaf type, days to flower, ray flower color, ray flower diameter and plant sur vival. Together natural populations displayed a substantial level of genetic diversity. In principal component analysis, natural populations were grouped together typically based on the region of origin. Czarnecki et al. (2008) further used AFLP markers to assess the genetic diversity of natural populations of C. leavenworthii Natural populations from each region, north,

PAGE 35

35 central and south Florida were found to have a high percentage of polymorphic loci (68.6%) and a high level of genetic diversity, wi th a total genetic value from all natural populations ( H T ) of 0.309. The G ST value for the natural populations was 0.226, indicating some genetic differentiation among populations but most genetic variation remained within populations. The natural popula tions typically clustered within their geographic region of origin probably due to isolation by distance. Previously, Crawford et al. (1984) evaluated the genetic diversity of C. leavenworthii and two other sister species in the section Calliopsis using is ozymes. The total genetic diversity in C. leavenworthii and C. tinctoria was 0.187 and 0.235, respectively. It was suggested that the difference in the number of populations and individuals sampled accounted for the difference in genetic diversit y. C. t inctoria seemed to have a higher proportion of polymorphic genes, a greater number of alleles per polymorphic gene and more heterozygous loci than C. leavenworthii The higher number of alleles per polymorphic gene was caused by a greater number of low fr equency alleles at several genes in C. tinctoria When the mean genetic identities were compared within and among species, C. leavenworthii and C. tinctoria had similar values (Crawford et al., 1984) The within population genetic identities for C. leaven worthii and C. tinctoria was 0.966 and 0.958, respectively, and among the two species was 0.944. These parameters indicate that the two species we re as similar within populations as between species. Archibald et al. (2005) found an internal transcribed s pacer ( ITS ) distance of 0.110 between C. leavenworthii and C. tinctoria indicating that these two species are similar to each other.

PAGE 36

36 Genetic Diversity in C. leavenworthii Seed Increase Populations The seed production populations evaluated originated in ce ntral Florida but were used for seed increase in northern and central Florida (Czarnecki et al., 2007, 2008) The seed production populations showed less phenotypic variation for each character assessed compared to the natural populations regardless of se ed increase location, indicating that the genetic identity of the seed production populations was maintained over various generations of production. A low level of genetic differentiation was detected among the seed production populations using AFLP marke rs The percentage of polymorphic loci (63.4 to 72.5%) was comparable to the natural populations and total genetic diversity ( H T ) was 0.251. Compatibility between C. leavenworthii and C. tinctoria C. leavenworthii and C. tinctoria growing ranges approach each other in northwestern Florida, where there could be some infrequent intergradations between the two species (Smith, 1983) Twenty nine F 1 hybrids were produced and grown in the greenhouse from crosses between C. leavenworthii x C. tinctoria var. tinc toria with an average of 31% pollen stainability ( Parker, 1973; Smith, 1976). However, these F 2 progeny were stunted and did not reach maturity. The chromosome configuration in the F 1 hybrids consisted of 7 bivalents, 2 trivalents and 1 quadravalent ind icating at least two reciprocal translocations. Eleven F 1 hybrids were produced and grown from crosses between C. leavenworthii x C. tinctoria var. similis with 13% pollen stainability (Smith, 1976) Univalents, bivalents and trivalents were found at dia kinesis of pollen meiosis, also indicating several structural differences in the genomes of the two species. Smith (1983) assessed the fertility of F 1 hybrids produced in the greenhouse from crosses between these species. F 1 progeny of the cross between C. tinctoria and

PAGE 37

37 C. leavenworthii had a moderate pollen staina bility of 30 60%, which is considered high enough to a llow occasional gene exchange. Research Objectives C. leavenworthii is nearly endemic to Florida and highly d esirable for native planting in highway beautification projects. Large scale seed increase and planting of C. leavenworthii are becoming increasingly common in Florida Previous studies have shown that potential genetic shifts and erosion may occur during seed increase and natural hyb ridization can result in gene flow between species and cause genetic contamination of produced seeds or planted populations. This study seeks to assess the potential of these genetic risks during C. leavenworthii seed increase and planting and to provide guidelines for minimizing such risks and diversity and integrity. The objectives of this study were to 1) assess if there were phenotypic changes or genetic shifts in C. leavenworthii occurring over successive seed increas e generations in morphological, physiological and reproductive characteristics and determine if seed increase location has played a role in causing potential genetic shifts, 2) use SSR markers to assess potential genetic changes that might occur to C. leav enworthii populations during seed increase in central and northern Florida over three generations, 3) determine the fitness level of C. leavenworthii and C. tinctoria interspecific hybrids in the F 1 and F 2 generations by comparing them to intraspecific pop ulations of C. leavenworthii and C. tinctoria and 4) determine the inheritance of morphological markers that could potentially be used to reliably detect pollen mediated gene f low, assess the effects of planting distance on pollen mediated gene flow from C tinctoria to C. leavenworthii and identify insect pollinators that may be involved with pollen mediated gene flow.

PAGE 38

38 CHAPTER 2 ASSESSING PHENOTYPIC CHANGES OF A NATURAL COREOPSIS LEAVENWORTHII POPULATION DURING SE ED INCREASE Justification Coreopsis leaven worthii has only been recorded in Florida and two counties in Alabama. It is solely produced in Florida and commonly found along roads ide ditches and highways, where it is used by the Florida Department of Transportation (FDOT) for highway beautification and erosion control (Norcini and Aldrich 2007) Seed of C. leavenworthii is typically produced from a segment of a natural population for several generations. There are two major concerns for seed increase practices of this species 1) to uphold genetic diversity from generation to generation and 2) to assess whether producing seed and growing population s in a different location from the natural population will affect the persistence and genetic diversity of the production population Because a segment from a natural population is used for several generations of seed increase, there is the potential for a loss of characteristics such as leaf type, seed production and flowering time, if the range of characteristics is not broad enough or representing all of th e natural population to adapt to changing environmental conditions. The entire life cycle of the population should be assessed to accurately evaluate the phenotypic diversity of a population to ensure that the population was unaffected by the seed p roduction practices (Edmands, 2007). When a population is moved into a new habitat many traits are naturally selected for simultaneously, which could result in a phenotypic response that causes changes in the appearance of the individuals of a population (Antonovics, 1976). Because of this simultaneous natural selection, the fitness level of the population will depend on the correlation between the characteristics and this level of fitness can change with different habitats. The genetic diversity of the

PAGE 39

39 natural and seed production populations should be assessed to be sure that the genetic diversity is upheld from generation to generation and in different locations. Typically seed of C. leavenworthii has been produce d for several generations from the sam e natural population that ha s been used for re establishment and restoration projects. A decrease in genetic diversity may result in inadequate establish ment leading to poor plant performance and higher maintenance costs. Several factors can affect geneti c variation within and among populations, such as population size, geographic distribution, primary mode of reproduction, mating system, seed dispersal mechanism and community type of the species, wh ere inbreeding depression can result from the same plants interbreeding, a rapid decline in population size or changes in vector preferences (Hamrick, 1983). If inbreeding depression occurs, the genetic diversity of the population may decrease causing inadequacy in yield or degradation of the species over succe ssive generations. The size of the population being used for seed increase can have a large effect on inbreeding depression by hav ing a larger e ffect on larger populations rather than smaller populations because the frequency of deleterious alleles alread y declined in smaller populations making inbreeding depression less profound (Thiele et al. 2010 ; van Treuren et al. 1993 ). For example, e ven though Scabiosa co l umbaria is an outcrossing species, it experienced severe inbreeding depression that was caus ed by limited gene flow between populations and could have led to extinction (van Treuren et al. 1993). Inbreeding depression caused a decrease in biomass production, root development, adult survival and seed set.

PAGE 40

40 The most widely used approach t o assess genetic change s during seed increase or production has been conducting common garden stud ies where plants from seed increase are grown under common conditions and the ir characteristics a re evaluated side by side Simon and Kastenbauer (1979) used this de s i gn to evaluat e the m orphological traits and yield for timothy, meadow fescue and perennial rye grass that were produced in Germany and three locations in the U nited S tates After seed increase of multiple generations, seed was sent back to Germany for e valuation. In another study s eed of three cultivars of Italian ryegrass was increased over two generations at one location in Japan and four locations in the U nited S tates and the populations were evaluated for genetic stability at one location in the U n ited S tates (Rincker et al. 1982). Neither of these studies showed adverse effects from seed increase in multiple locations or over successive generations. A previous study examining the morphological characteristics of C leavenworthii natural and seed production populations revealed that there was a substantial amount of genetic diversity among the natural populations (Czarnecki et al. 2007). The populations from similar geographical regions were found to be more similar than those from different regi ons. The seed production populations originally from central Florida and increased at different sites in Florida f or one to four generations r emained genetically similar and maintained a moderate level of genetic diversity P roducing see d over successi ve generations and different locations could affect the genetic diversity of C. leavenworthii In the previous study seed production took place at several different locations and not over successive generations. It can be difficult to compare these popul ations because phenotypic differences could have occurred

PAGE 41

41 because of location, generation or both To determine if location or generation ha s caus ed morphological changes or genetic shifts, seed increase should occur at the same multiple locations over su ccessive generations. In the current study seed from a natural C. leavenworthii population in central Florida was increased simultaneously in central and northern Florida over three successive generations followed by common garden studies of the original and increased seed populations in central Florida where individuals of several populations we re grown in the same location for evaluation T he objective s of this study w ere 1) t o assess if there we re phenotypic changes or genetic shifts occurring over su ccessive seed increase generations in 12 morphological, physiological and reproductive characteristics and 2) determine if seed increase location has played a role in causing potential genetic shifts The information obtained will be very valuable for def ining C. leavenworthii pheno t yp ic diversity and integrity during seed increase Materials and Methods Seed Collection from a Natural Population Seed collected from a C. leavenworthii natural population i n Reedy Creek Mitigation Bank in Polk County, FL (Lat 27 Long 81 USDA cold hardiness zone 9a, AHS heat zone 11) was used as the seed source for subsequent increases. This source was referred to as Generation 0 ( G 0 ) The collection consis ted of 1220 seed heads from 122 individuals randomly selected out of the natural population (10 seed heads per individual). The collection was made on 1 July 2006 from 75 plants and on 24 September from another 47 plants. The soil at the growing

PAGE 42

42 site of the natural population was a mixture of ultisols, entisols and alfisols. Seed from the different individuals were combined and split into four seed lots. Seed Increase One seed lot from the natural population was maintained at the North Florida Research a nd Education Center (NFREC), Quincy, FL (Lat 30 Long 84 USDA cold hardiness zone 8b, AHS heat zone 9) in northern Florida and another was maintained at the Gulf Coast Research and Education Center (GCREC), Wimau ma, FL (Lat 27 Long 82 USDA cold hardiness zone 9b, AHS heat zone 10) in central Florida At e ach location seed was increased successively for three generations ( G 1 G 2 and G 3 ), starting in spring 2007, by growing ~100 plants from seed of a previous generation and harvesting mature seed heads from ten open pollinated flowers from each plant Harvested seed heads were dried indoor s at each site and eed storage room (10C and 50% RH ). The seed increase population was designated by its gen eration and its site of increase (C for the central Florida site and N for the northern Florida site). Seed increase in central Florida G 0 to G 1 C. Seed from G 0 was sowed on the surface of Vergro v erlite container mix A (Tampa, FL) on 16 January 2007 and ge rminated in the greenhouse (29.4/23.9C) under mist and a natural photoperiod Seedlings were transplanted in to 80 m L cell containers (Landmark Plastic Corp., Akron, OH) of Vergro v erlite c ontainer m ix A on 22 February and grown in the greenhouse (29.4/23 .9C) under a natural photoperiod for about four weeks The plants were then transplanted in to 3.2 L containers filled with Vergro v erlite c ontainer m ix A mixed with 8.6 g of 15N 3.9P 10K (5 6 mo. Osmocote, The Scotts Co., LLC, Marysville, OH) on 21 March and grown for another three weeks

PAGE 43

43 in the greenhouse until they became established Before t he established plants began flowering, they were moved out of the greenhouse onto a greenhouse bench and grown for about six weeks before seed collection under a n atural photoperiod Ten seed heads per plant were collected on 14 and 27 June from 110 plants (23.2C and 71.1% RH) (Table A 1) G 1 C to G 2 C. Seed from G 1 C was sowed o n 8 January 2008 Seedlings were transplanted on 13 February. Plants were transplanted into 3.2 L containers on 21 March Ten seed heads per plant were collected on 28 May from 82 plants, 9 June from 70 plants and 26 June from 61 plants (23.3C and 70.8% RH) (Table A 1) All other growing conditions were the same as th ose used for the incr ease of G 0 to G 1 C G 2 C to G 3 C. Seed from G 2 C was sowed on 7 January 2009 Seedlings were transplanted into 3 1 m L cell Speedling flats (128 cell Speedling flat) (Sun City, FL) on 27 January. Seedlings were transplanted in Fafard 3B soil media (Anderson, S C) into 3.2 L containers on 23 February Ten seed heads per plant were collected on 1 3 May from 120 plants, 20 May from 118 plants and 8 June from 105 plants (21.6C and 72.7% RH) (Table A 1) All other growing conditions were the same as those used for the increase of G 0 to G 1 C Seed increase in northern Florida G 0 to G 1 N. G 0 s eed was sowed on 9 January 2007 on the surface of Metro Mix 200 (Marysville, OH) in a 6 .5 L flat The flats used for seed sowing were placed on propagation mats set at 21C in a greenhouse, where the heater was set at 15.5C. T he seedlings were transplanted in 7 4 m L container s of Metro Mix 200 on 6 February and fertilized with 100 ppm of 15N 13.2P 12.4K (Miracle Gro, All Purpose Plant Food, The Scotts Co., LLC, Marysville, OH) st arting one week later The plants were

PAGE 44

44 transplanted in the field on 6 March and fertilized with 9 g/plant of 18N 2.6P 10.0K (8 9 mo. Osmocote, The Scotts Co., LLC, Marysville, OH) The field with soil type Orangeburg loamy sand, was covered in black lan dscape fabric where the plants were spaced on 30.5 cm centers and planted in a single 30.5 m row. A total of ten seed heads were collected from each of the 100 plants as the seed heads developed and matured from July through August G 1 N to G 2 N. Seed from the G 1 N p opulation was sowed on 1 January 2008 The seedlings were transplanted on 1 February The plants were transplanted into the field on 6 March. Ten seed heads per plant were collected from 14 May through June from 100 plant s. All other growing conditions were the same as th os e used for the increase of G 0 to G 1 N G 2 N to G 3 N. Seed from the G 2 N p opulation was sowed on 7 January 2009 The seedlings were transplanted on 6 February The plants were transplanted into the field on 9 March. Ten seed h ead s per plant were collected from May through June from 100 plant s. All other growing conditions were the same as those used for the increase of G 0 to G 1 N Assessing Phenotypic Changes Experimental design C ommon garden stud ies were conducted to assess po tential phenotypic changes of C. leavenworthii during seed increase. Seed f rom the different increase populations were germinated and plants w ere grown in the same environment in central Florida Plants were arranged using a randomized complete block des ign with 15 blocks and five individuals per block in 2009 and five blocks with 15 individuals per block in 2010.

PAGE 45

45 A total of 75 individuals were grown and assessed per seed increase population in the field for both year s. Growing conditions of the c ommon g arden s tudy 2009. Seed from each population was sowed on 2 July 2009 in Fafard 3B soil media into 330 m L containers in the greenhouse (29.4/23.9C). One hundred seed s w ere put into each container with five rep lications per population. S eed germination w a s counted and seedlings were transplanted on 21 July into 3 1 m L cell Speedling flats with Fafard 3B soil media. One hundred twenty eight seedlings were transplanted for each population. The plant s were transplanted in the field on 3 September in a single row. The field was a mix of Myakka, Haplaquents and St. Johns sandy soil. The beds were raised 25 cm high, 71 cm wide at the top and 81 cm wide at the base. Plants were spaced 30.5 cm apart within plots that were 92 cm apart in rows with 152 cm between rows. The plants were fertilized with a total of 169 kg ha 1 of nitrogen, 37 kg ha 1 of phosphorus and 210 kg ha 1 of potassium throughout the year through drip irrigation. 2010. Seed from each population was sowed on 20 January 2010 The number of seed lings that emerged was counted and seedlings were transplanted on 1 February. Plants were trans planted in t he field on 5 April. All other growing and field conditions were the same as 2009 except no fertilization was applied and the plants were irrigat ed through seepage irrigation. Data collection D ata collected from the field included plant height ( cm ) plant dry weight ( kg ) leaf type days to flower (DTF) disk flower size (DFS) (diameter, cm), whole flower size (WFS) ( cm ) petal lobing degree of peta l o verlap (DPO) number of ray pet als per flower head (NRP) the number of seeds produced per five seed head s seed

PAGE 46

46 germination (%) of seed collected from the field and powdery mildew severity (PMS) The tallest part of the plant was measured for plant he ight. After the plants were removed from the field, they were dried in a drying room ( 37.8 C and 20% RH) and weighed after at least three weeks of drying Data w ere collected for 2009 from 3 September 2009 to 15 January 2010 and 19 April 2010 to 14 June for 2010 Leaf types were categorized on a 1 to 7 point scale where 1 wa s a simple leaf and 7 wa s the most complex leaf (Fig ure 2 1) Leaf type data w ere taken 94 days in 2009 and 113 days in 2010 after transplanting. D FS and WFS were measured from fiv e flowers each that were approximately at the same level of maturity. Petal lobing was rated on a 1 to 6 point scale based on the majority of flowers on the plant where 1 was the most simple petal (no lobing ) and 6 was the most complex petal ( most and de epest lobing ) (Fig ure 2 3 ). The DPO was rated on a 1 to 3 point scale, where 1 had petals that were oriented in a pinwheel fashion and not touching (G) 2 was where the outside of the petals were touching but not overlapping (S) and 3 was where petals wer e overlapping (O) (Fig ure 2 5 ). P MS was rated on a 1 to 10 point scale, 1=no infection, 2=1 10%, 3=11 20%, 4=21 30%, 5=31 40%, 6=41 50%, 7= 51 60%, 8=61 70%, 9=71 80 and 10=81 1 00% infection (Hausbeck et al., 2002) P MS was taken 113 days in 2009 and 120 days in 2010 after transplanting. Seed collected in the field in 2009 was sowed on 8 10 March 2010 for seed germination tests. Statistical analysis Significant differences were found for each trait evaluated using a nalysis of v ariance (ANOVA) using PROC GLM in SAS (SAS, 1997). Data w ere transformed using the Arcsine Squ are R oot method for the seed germination tests Based on the ANOVA test, further statistical testing was performed using the Tukey W Procedure for

PAGE 47

47 mean separation analysis in SAS. Change s in genetic diversity between populations for each phenot ypic characteristic measured were described using the standard errors calculated as in Tesfaye et al. (1991) and discussed in Foote (1997) Relationships among populations were determined using pri ncipal component analysis (PCA) using NTSYS pc (NTSYSp c, version 2.2 [Rohlf, 2005]). Results Plant Height and Dry Weight There were significant differences found between year s but not among populations for plant height (Table 2 1). The mean plant height of the G 0 individuals in 2009 was 72.3 cm (Table 2 2). The mean plant height of individuals of the six increase populations was 69.1 to 73.9 cm, or 95.6 % to 102.2% of that of the G 0 The mean plant height value for the G 0 individuals in 2010 was 21.6% smal ler (56.7 cm) than that in 2009 Plant height decreased by 1 7 .4 % to 25.3% (55.1 to 61.0 cm) for the individuals of the six increase populations. The mean plant height for the increase populations in 2010 was 97.2 % to 107.6% of the mean height of G 0 whic h was similar t o the 2009 results. Standard errors can be used to describe changes in genetic diversity as in Tesfaye et al. (1991) and Foote (1997). The standard error of the plant height mean for the G 0 individuals was 0.77 and 0.73 cm in 2009 and 2 010, respectively (Table 2 3). The standard errors of the plant height means for the individuals of the six increase populations were similar to G 0 in 2009, ranging from 0.70 to 0.95 cm indicating no change in phenotypic diversity over seed increase In 201 0 there was a slight decrease in standard error (0.41 to 0.66 cm) for the plant height mean for the six seed increase

PAGE 48

48 populations indicating a slight decrease in phenotypic diversity overall from 2009 and over seed increase from G 0 in 2010 There were not significant differences between year s or among populations for plant dry weight (Table 2 1). The mean plant dry weight for the G 0 individuals was 0.110 kg per individual (Table 2 2). The mean plant dry w eight for the individuals of the six increase popu lations was 0.099 to 0.116 kg or 90.0 % to 105.5 % of that of G 0 The standard error of the plant dry weight mean for the G 0 individuals was 0.001 kg (Table 2 3) The standard errors for the plant dry weight means for the individuals of the six seed incre ase populations were similar, 0.001 to 0.002 kg indicating no change in phenotypic diversity over seed increase Leaf Type S ignificant differences were fou nd between year s but not among populations for leaf type scores (Table 2 1). The mean leaf type sco re for the G 0 individuals was 3.3 in 2009 and 3.3 to 3.4 for the six increase populations (Table 2 2 and Figure 2 2 ) The mean leaf type score for the G 0 individuals grown in 2010 was 4.2 0.9 higher than in 20 09 The mean leaf type score for the individ uals of the six seed increase populations was 3.8 to 4.2, similar to that of G 0 The standard errors for the mean leaf type scores for the individuals of the G 0 populations were 0.05 and 0.02 in 2009 and 2010, respectively, and were similar for the indivi duals of the six seed increase populations, 0.04 to 0.07 in 2009 and 0.01 to 0.03 in 2010 (Table 2 3) indicating no change in phenotypic diversity over seed increase

PAGE 49

49 Flower Characteristics Days to flower (DTF) S ignificant differences were found for DTF b etween year s and among populations in 2010 but not in 2009 (Table 2 1). In 2009 the mean number of DTF for the G 0 individuals was 109.1 days (Table 2 2). The mean number of DTF for the six increase populations was 109.9 to 113.2 days, similar to that of G 0 The mean number of DTF for the G 0 individuals in 2010 was 7.1% small er ( 101.3 days ). The mean number of DTF for the individuals of the six increase populations in 2010 was 98.9 to 103.5 days similar to the mean DTF of G 0 The significant difference detected among populations in 2010 was among G 1 C, G 3 C, G 1 N and G 2 N. The first three populations took fewer DTF (98.9 to 99.1 days ), while the last population took significantly more DTF (103.5 days). The standard error for the mean DTF was 0.84 days for the G 0 individuals in 2009 and 0.55 to 0.71 days for the individuals of the six increase populations (Table 2 3) indicating a slight decrease in phenotypic diversity over seed increase In 2010 there was a decrease in the standard error for the mean DTF which was 0.17 days for the G 0 individuals and 0.14 to 0.29 days for the individuals of the six increase populations indicating an overall decrease in phenotypic diversity from 2009 but not over seed increase within 2010 Disk flower size (DFS) There wer e significant differences for DFS between year s but not among populations (Table 2 1). The mean DFS for the G 0 individuals was 0.79 cm in 2009 (Table 2 2) The mean DFS for the individuals of the six increase populations ranged from 0.78 to 0.83 cm, simi lar to that of G 0 The mean DFS for the G 0 individuals grown

PAGE 50

50 in 2010 was 0.88 cm 1 1. 4% larg er than the mean size in 2009 DFS in creased by 3.6 % to 15.4 % (0.86 to 0.90 cm ) for all six increase populations. The mean DFS for the increase populations in 20 10 was 97. 7 % to 10 2.3 % of the mean DFS of the G 0 individuals similar to the 2009 results. The standard error for the mean DFS was 0.009 cm for the G 0 individuals in 2009 and 0.004 to 0.007 cm for the individuals of the six increase populations (Table 2 3) indicating no change in phenotypic diversity over seed increase In 2010 there was a decrease in standard error where the standard error for the mean DFS was 0.001 cm for the G 0 individuals and 0.002 to 0.007 cm for the individuals of the six increase populations indicating an overall decrease in phenotypic diversity from 2009 but not over seed increase within 2010. Whole flower size (WFS) There were significant differences for WFS between year s but not among populations (Table 2 1). The mean WFS was 3.3 cm for the G 0 individuals in 2009 (Table 2 2) Similar mean WFS (3.4 to 3.5 cm) were recorded for the individuals in the six increase populations. The mean WFS for the G 0 individuals in 2010 was 9.1% small er ( 3.0 cm) than in 2009 A similar trend ( 1 1.8 % to 17.1 % flower size decr ease) was observed for all six increase populations. Again, t he mean WFS for the six increase populations in 2010 (2.9 to 3.0 cm) was similar to that of G 0 The standard error for the mean WFS was 0.019 cm for the G 0 individu als in 2009 and 0.021 to 0.033 cm for the six increase populations (Table 2 3) indicating no change in phenotypic diversity over seed increase In 2010 there was a decrease in standard error for the mean WFS for the G 0 individuals at 0.012 and was 0.005 to 0.011 cm for

PAGE 51

51 the six seed increase populations indicating an overall decrease in phenotypic diversity from 2009 but not over seed increase within 2010 Petal lobing No significant differences were found between year s or among populations for petal lobi ng (Table 2 1). The mean petal lobing score was 4.2 for the G 0 individuals and 4.2 to 4.9 for the six increase populations (Table 2 2 and Figure 2 4 ) The standard error for the mean petal lobing score was 0.10 for the G 0 individuals and 0.01 to 0.10 for the individuals of the six increase populations (Table 2 3) indicating no change in phenotypic diversity over seed increase Degree of petal overlap (DPO) There were significant differences for the DPO between year s and among populations (Table 2 1). Th e average rating for the DPO was 2.6 for the G 0 individuals in 2009 (Table 2 2 and Figure 2 6 ) All three populations produced in central Florida had mean ratings of 2.7. The mean ratings for the populations produced in northern Florida were 2.2 (G 2 N) to 2.4 (G 1 N and G 3 N) but were not significantly different. G 2 N was significantly different from the populations produced in central Florida. The ratings decreased in 2010 compared to 2009 for all populations (Table 2 2 and Figure 2 6 ) The mean rating for DPO in 2010 was 2.4 for the G 0 individuals a 0.2 decrease from 2009 The mean rating for the populations produced in central Florida were from 2.3 to 2.4 a decrease of 0.3 to 0.4 from 2009 The mean ratings for the populations produced in northern Flo rida were from 2.0 to 2.2 a decrease of 0.2 to 0.3 from 2009 The mean rating for G 2 N w as significantly different from G 1 C, G 2 C and G 0 The standard error for the mean DPO scores for the G 0 individuals was 0.037 and 0.030 to 0.065 for the individuals of the six seed increase populations in 2009 (Table 2

PAGE 52

52 3) indicating no change in phenotypic diversity over seed increase Standard error for the mean DPO scores decreased in 2010, where the standard error was 0.017 for the G 0 individuals and ranged from 0.0 19 to 0.028 for the individual s of the six seed increase populations indicating an overall decrease in phenotypic diversity from 2009 but not over seed increase within 2010 Number of ray petals per flower head (NRP) No significant differences were detect ed between year s or among populations for the NRP per flower head (Table 2 1). The mean NRP per flower head was 8.12 for the G 0 individuals (Table 2 2) The mean NRP for the individuals of the six increase populations was 8.04 to 8.22. The standard erro r for the mean NRP for the G 0 individuals was 0.020 and 0.006 to 0.059 petals for the six increase populations (Table 2 3) indicating no change in phenotypic diversity over seed increase Seed Production and Germination There were significant differences between year s and among populations for both year s for seed production for five seed heads per plant (Table 2 1). The mean amount of seed produced for the G 0 individuals was 5 90.0 seed in 2009 ( Figure 2 7) The mean s eed production for the three central Florida increase populations ranged from 540.9 to 652.4 seed not significantly different from that of G 0 but G 2 C (540.9 seed ) w as significantly different from G 3 C (652.4 seed ) G 1 N (425.8 seed ) produced significantly less seed than G 2 N (570.5 seed ) and G 3 N (588.8 seed ) as well as G 0 and the populations from central Florida. Seed production of G 0 increased by 1.7% (599.9 seed ) in 2010 compared to 2009 (Figure 2 7) Except for one increase population (G 3 C), s eed production of the increase populations incre ased by 17.4 % to 46.3% reaching 623.0 to 737.9 seed in 2010. G 1 C

PAGE 53

53 (722.5 seed ), G 2 C (644.0 seed ) G 3 C (632.7 seed ) G 1 N (623.0 seed ) and G 3 N (691.3 seed ) did not produce significantly more seed than G 0 but G 2 N (757.9 seed ) did. Standard error for the mean number of seed produced for five seed heads was 10.31 seed for the G 0 individuals and 6.91 to 12.76 seed for the individuals in the six increase populations in 2009 (Table 2 3) indicating no change in phenotypic diversity over seed increase Standard er ror for the mean number of seed produced for five seed heads for the G 0 individuals was lower in 2010, 3.27 seed for the G 0 individuals and 4.06 to 12.47 seed for the individuals of the six increase populations indicating an overall decrease in phenotypic diversity from 2009 but not over seed increase within 2010 There were not any significant differences found among populations for seed germination i n 2009 (Table 2 1). M ean seed germination was 56.6% for the G 0 individuals and 52.1% to 61.2% for the ind ividuals of the six increase populations (Table 2 2) The standard error was 1.34% for G 0 and 1.06% to 1.48% for the six increase populations (Table 2 3) indicating no change in phenotypic diversity over seed increase Powdery Mildew Severity (PMS) S igni ficant differences were found between year s but not among populations for PMS (Table 2 1). The mean severity score for PMS for the G 0 individuals was 5.5 and 4.9 to 5.8 for the individuals in the six seed increase populations in 2009 (Table 2 2 and Figure 2 8 ) The mean PMS score decreased in 2010 for G 0 and the six increase populations compared to 2009. The mean PMS score was 3.0 for the G 0 individuals and 2.5 to 3.2 for the individuals of the six seed increase populations in 2010

PAGE 54

54 Standard error for th e mean PMS ratings in 2009 for the G 0 individuals was 0.16 and 0.10 to 0.19 for the individuals of the six seed increase populations (Table 2 3) indicating no change in phenotypic diversity over seed increase Standard error for the mean PMS ratings decr eased slightly in 2010 wh ich was 0.11 for the G 0 individuals and 0.05 to 0.10 for the individuals of the six seed increase populations indicating no change in phenotypic diversity over seed increase. Principal Component Analysis A dataset of 168 data poi nts was assembled f ro m the mean values of seven populations for 12 characteristics in two year s (7 x 12 x 2). Principal component analysis of th is dataset revealed that two principal components could account for 78.2% of the total observed variance The first principal component had an eigenvalue of 7.35 and could account for 61.2% of the total variance. The second principal component had an eigenvalue of 2.04 and could account for 1 7 0 % of the total variance. The remaining components had eigenvalues 1.04 and failed to explain much variance, so they were exclu ded from further analysis. Six characte ristics ( DFS DTF leaf type, plant height, PMS and WFS ) influenced the first pri ncipal component heavily, with loading value s 0.93. The second principal component was most influenced by the NRP plant dry weight and seed germination with loading values 0.6 4 Based on the se two principal components the populations were clustered primarily by year (Figure 2 9). Within the 2009 cluster, the northern and the central Florida populations appeared to be separated However, this clustering did not show up in 2010. The lack of consistent, clear clustering among the populations from year to year suggest s that no obvious population differentiation had occurred during the three successive generations of seed increase in either location, based on the phenotypic characteristics evaluated.

PAGE 55

55 Discussion The seed germination tests in the current study did not take dormant seed into account by performing seed viability tests of non germinated seed. Seed of C. leavenworthii has been reported as exhibiting after ripening and the viability declined with longer storage (Norcini and Aldrich, 2007). Kabat et al. (2007) reported that a small percentage of fresh seed were nond ormant and most were viable, dormant seed. Most of the seed gradually became nondormant and a high percentage of viable seed would germinate when expos e d to warm, moist conditions under a wide range of temperatures commonly found in Florida Based on ge rmination and tetrazolium staining tests, seed viability of C. leavenworthii seeds has been estimated at 57.0% (Norcini and Aldrich, 2008). The time of year of seed production and collection can affect seed quality, where seed that we re harvested from May through July we re of the best quality ( Norcini et al., 2006) Seed viability was about 75% for this time period with fertilizer enhancing seed production. Unintentional selection was minimized by transplanting the first 128 randomly selected seedlings to produce the next generation of seed increase. The most vigorous or first germinated seedlings were not transplanted over weaker seedlings. The only type of unintentional selection that could have occurred was the exclusion of seedlings from dormant seed Dormant seed need specific environmental conditions to germinate that were not used. This could have led to fewer seedlings in later generations that did not exhibit seed dormancy. Although C. leavenworthii ha s been documented as an outcrossing species it can still have effects from inbreeding depression through biparental inbreeding, as reported in Gaillardia pulchella (Heywood, 1993). Biparental inbreeding can be caused by

PAGE 56

56 outcrossing between genetic relatives, where there can be a decline in fitnes s in progeny of related individuals than nonrelated individuals. This can also cause a reduction in seed production in species that exhibit self incompatibility, where C. leavenworthii has been reported as sporophytically self incompatible. The increase p opulations had similar mean values with G 0 over both year s for nine out of the 12 characteristics evaluated including four characteristics (leaf type, petal lobing, DPO and PMS ) evaluated on rating scales. There were significant differen ces among populat ions for DTF in 2010 but not in 2009 DPO over both year s and seed production over both year s These differences were not consistent by generation or location, indicating that no genetic shift occurred by the seed increase practices. In PCA, the north er n Florida populations appeared to be clustered away from the central Florida populations in 2009, but that clustering disappeared in 2010. There were not consistent changes in phenotypic diversity to indicate that location had an effect on seed increase o ver successive generations. Considering these results, it can be conclude d that three generations of successive seed increase in central or northern Florida will not likely to cause significant population differentiation or genetic shift in C. leavenworth ii seed production on plant growth and development, leaf morphology, flowering and flower morphology, seed production and powdery mildew resistance. Simon and Kastenbauer (1979) found similar results when meadow fescue, timothy and perennial ryegrass were multiplied over two generations. These forage species were relatively unaffected by multiplication over two generations. Rincker et al. (1982) found no consistent shifts in plant growth characteristics evaluated for Italian ryegrass in two and three gene rations evaluated.

PAGE 57

57 There were yearly effects found for the seed increase populations at both locations, possibly due to fertilization, irrigation, humidity, day length and differences in the time of year of the growing seasons It is difficult to pred ict which factors affected phenotypic changes between year s and because there were not any consistent changes in populations over successive generations, these factors did not seem to have a detrimental effect on phenotypic diversity. Each year will be di fferent when producing seed over successive generations and multiple locations. This illustrates the importance of using a phenotypically and genetically diverse starting population and maintaining it in each successive generation to have the ability to a djust to changing environmental conditions. When the population by year interaction (gen otype x environment) was tested, there was only one significant interaction found, which was for seed production. Because the year s were significantly different and m ost of the population by year interactions were not significant, the year s were treated separately. This indicate d that t he environmental conditions had an effect on the phenotype of most of the traits evaluated, but the effects were consistent within eac h year Summary After evaluating these popul ations over two year s, there were no clear trends of genetic shift in C. leavenworthii using the current commercial seed production practices due to generation or location The standard error values indicated no change in variation of the traits evaluated It is apparent that environmental conditions can cause changes in some of the phenotypic characteristics of this species but did not affect the phenotypic diversity in the current study Even though there wer e differences between year s, the results within each year did not chan ge in generation or location due to seed increase practices

PAGE 58

58 Table 2 1 Analysis of variance of several phenotypic characteristics for seven C leavenworthii populations produced in two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. Between Year s Year s*Populations Among Populations z Plant Characteristic Year F value Probability F value Probability F value Probability Plant height 2009 95.26 0.0001 0.43 0.859 0 0.98 0.4422 2010 1.93 0.1162 Plant dry weight Combined 0.00 0.9531 1.27 0.2782 1.38 0.2265 Leaf type 2009 34.06 0.0001 0.64 0.6951 0.44 0.8473 2010 2.6 0 0.0439 Days to flower 2009 96.35 0.0001 0.76 0.606 0 1.18 0 .3236 2010 4.02 0.0063 Disk flower diameter 2009 19.04 0.0001 0.96 0.459 0 1.73 0.124 0 2010 1.22 0.3299 Whole flower diameter 2009 162.93 0.0001 0.56 0.7591 1.25 0.2876 2010 1.89 0.1233 Petal lobing Combined 6.44 0.9861 0.61 0.7226 1.81 0.1028 Petal overlap 2009 0 .00 0.0125 0.25 0.9585 4.33 0.0008 2010 3.98 0.0067 Ray petals (No.) Combined 0.98 0.3254 0.3 0 0.9365 0.97 0.4463 Seed production 2009 23.53 0.0001 3.08 0.0079 8.5 0 0.0001 2010 3.21 0.0186 Seed germination 2009 N A NA NA NA 1.44 0.2099 Powdery m ildew severity 2009 70.7 0.0001 0.48 0.825 0.99 0.4353 2010 1.63 0.1832 z Sample size among blocks was 5 in 2009 and 15 in 2010 and 75 among populations for both year s.

PAGE 59

59 Table 2 2 Comparison of several phenotypic ch aracteristics of seven C leavenworthii populations produced at two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. Plant Characteristic Year G 0 G 1 C G 2 C G 3 C G 1 N G 2 N G 3 N Plant height (cm) 2009 72.3 ns 73.1 73.0 73.8 69.1 71.8 73.9 2010 56.7 ns 58.9 56.4 55.1 56.6 57.2 61.0 Plant dry weight (kg) Combined 0.110 ns 0.116 0.114 0.114 0.103 0.099 0.105 Leaf type 2009 3.3 ns 3.4 3.4 3.4 3.3 3.3 3.3 2010 4.2 ns 4.1 4.0 3.9 3.8 4.2 4.0 Days to flower (days) 2009 109.1 ns 112.2 112.3 112.6 109.9 112.2 113.2 2010 101.3ab z 99.1b 100.2ab 98.9b 99.0b 103.5a 100.8ab Disk flower size (cm) 2009 0.79 ns 0.83 0.78 0.78 0.79 0.82 0.80 2010 0.88 ns 0.86 0.87 0.90 0.88 0.90 0.88 Whole flower size (cm) 2009 3.3 ns 3.4 3.5 3.4 3.4 3.5 3.5 2010 3.0 ns 2.9 2.9 3.0 2.9 3.0 3.0 Petal lobing Combined 4.2 ns 4.6 4.4 4.4 4.2 4.9 4.5 Petal o verlap 2009 2.6ab z 2.7a 2.7a 2.7a 2.4ab 2.2b 2.4ab 2010 2.4a z 2.4a 2.4a 2.3ab 2.2ab 2.0b 2.1ab Ray petals (n o.) Combined 8.12 ns 8.05 8.08 8.04 8.22 8.07 8.10 Seed germination (%) 2009 56.6 ns 52.1 61.2 60.2 58.3 55.6 59.7 Powdery mildew severity 2009 5.5 ns 4.9 5.3 5.5 5.3 5.4 5.8 2010 3.0 ns 3.1 3.0 3.0 3.0 2.5 3.2 ns Means z

PAGE 60

60 Table 2 3. Standard errors of several phenotypic characteristics of seven C leavenworthii populations produced at two locations and grown at the Gulf Coast Researc h and Education Center, Wimauma, FL in 2009 and 2010. Plant Characteristic Year G 0 G 1 C G 2 C G 3 C G 1 N G 2 N G 3 N Plant height (cm) 2009 0.77 0.95 0.72 0.75 0.75 0.90 0.70 2010 0.73 0.41 0.61 0.46 0.50 0.66 0.42 Plant dry weight (kg) Combined 0.001 0.002 0.00 2 0.002 0.002 0.002 0.001 Leaf type 2009 0.05 0.05 0.05 0.06 0.07 0.04 0.04 2010 0.02 0.03 0.02 0.03 0.03 0.01 0.01 Days to flower (no.) 2009 0.84 0.58 0.71 0.55 0.69 0.70 0.61 2010 0.17 0.27 0.14 0.29 0.16 0.23 0.20 Disk flower size (cm) 2009 0.009 0.004 0.007 0.007 0.006 0.005 0.005 2010 0.001 0.003 0.004 0.007 0.003 0.004 0.002 Whole flower size (cm) 2009 0.019 0.021 0.030 0.025 0.029 0.033 0.028 2010 0.012 0.006 0.009 0.011 0.010 0.005 0.010 Petal lobing Combined 0.10 0.07 0.10 0.07 0.01 0. 06 0.08 Petal o verlap 2009 0.037 0.032 0.031 0.030 0.065 0.047 0.050 2010 0.017 0.020 0.028 0.024 0.022 0.019 0.020 Ray petals (n o.) Combined 0.020 0.006 0.011 0.006 0.059 0.013 0.012 Seed production (no.) 2009 10.31 7.73 12.76 11.35 11.38 12.46 6.91 2010 3.27 5.61 11.84 4.06 9.78 12.47 4.94 Seed germination (%) 2009 1.34 1.06 1.21 1.21 1.48 1.35 1.41 Powdery mildew severity 2009 0.16 0.14 0.17 0.19 0.12 0.17 0.10 2010 0.11 0.07 0.10 0.08 0.07 0.08 0.05

PAGE 61

61 Figure 2 1. Leaf types observed in C leavenwo rthii with assigned scores (Czarnecki et al., 200 7 ).

PAGE 62

62 Figure 2 2 Distribution of leaf type scores in the natural population and six seed increase populations of C leavenworthii produced at two locations and grown at the Gulf Coast Research and E ducation Center, Wimauma, FL in 2009 and 2010. The Y axis indicates the frequency of each score occurring for each generation at each location.

PAGE 63

63 Figure 2 3 Scores used for evaluating petal lobing of seven C leavenworthii populations produced at two lo cations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010.

PAGE 64

64 Figure 2 4 Changes in petal lobing scores for seven populations of C leavenworthii produced at two locations and grown at the Gulf Coast Research and Educ ation Center, Wimauma, FL in 2009 and 2010. The Y axis indicates the frequency of each score occurring for each generation at each location.

PAGE 65

65 Figure 2 5 Scores used for evaluating petal orientations of seven C leavenworthii populations produced at two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. (G=1, S=2, O=3)

PAGE 66

66 Figure 2 6 Changes in the degree of petal overlap ratings for seven populations of C leavenworthii produced at two locations and grown a t the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. The Y axis indicates the frequency of each rating occurring for each generation at each location.

PAGE 67

67 Figure 2 7 Differences in the number of seeds per five seed heads of seven C leavenworthii populations produced at two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. z Means within cultivar not followed by the p within each year Means for each year were statistically analyzed separately and cannot be compared to each other.

PAGE 68

68 Figure 2 8 Distribution of powdery mildew severity among individuals for each of the seven populations of C leavenworthii produc ed at two locations and grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. The Y axis indicates the frequency of each severity score occurring for each generation at each location.

PAGE 69

69 Figure 2 9 Principal component analys is of C leavenworthii populations grown in 2009 and 2010 at the Gulf Coast Research and Education Center, Wimauma, FL.

PAGE 70

70 CHAPTER 3 ASSESSING MOLECULAR CHANGES USING SIMPLE SEQUENCE REPEAT (SSR ) MARKERS OF A NATURAL CO R EOPSIS LEAVENWORTHII POPULATION DURIN G SEED INCREASE Justification Coreopsis leavenworthii is endemic to Florida, solely produced in Florida and commonly planted along roadsides in highway beautification projects ( Kabat et al., 2007 ; Norcini and Aldrich, 2007) Current C. leavenworthii comme rcial seed production practices start with a natural population and use a segment of that population to produce seed over numerous generations Seed produced are required to not only be of high quality but also preser ve the level of composition of genetic diversity of the original, naturally occurring population ( Hartl, 2000 ). P roduction practices fertilization, soil moisture and soil type, can change and/or cause loss of genetic diversity in the produced seeds and affect plant performance and survival. For endemic species, these negative effects could be dramatic and impact the survival of the species. In a previous study (Chapter 2) the effects of seed increase on the genetic diversity and integrity of C. leavenworthii populations w ere assessed based o n phenotypic characterization. The information gained wa s very valuable, but the number of characteristics available for analysis was limited. In addition, the assessed characteristics were mostly quantitative which we re known to be prone to influence s by many factors, such as growth conditions and plant developmental stages. Consequently subtle genetic changes (or genetic shifts) might have been masked by environmental effects and not identified. Therefore, there has been a strong need for the use of more powerful tools to assess and monitor potential genetic shifts during seed increase based on the practices used by C. leavenworthii seed producers, where

PAGE 71

71 seed is collected from a natural population and increased for several generations Molecular mark ers, especially those based on DNA polymorphisms can be desirable tools for such assessment and monitoring because they can not be affected by environmental changes, plant developmental stages or physiological status, and can be analyzed early and reliably A number of molecular m arker systems have been used to assess the genetic diversity and composition of plant population s or species including allozymes, RAPDs, RFLPs, AFLP s and SSR s Among them AFLP and SSR marker s have been the most commonly used syst ems for genetic diversity and relationship analysis (Mueller and Wolfenbarger, 1999; Weising et al., 2005) For AFLP analysis prior knowledge of the DNA sequence has not been required a great number of markers have been produced and they have been repor ted as reliable, reproducible, affordable and fast to generate (Blears et al., 1998; Mueller and Wolfenbarger, 1999; Vos et al., 1995). AFLP markers have been reported as dominant and the DNA must be of high purity for complete digestion by the restrictio n enzymes On the other hand, SSR markers have been reported as codominant, detect ed a high level of polymorphism in genetic diversity studies and we re widely distributed in higher plants ( Blears et al., 1998; Mu eller and Wolfenbarger, 1999 ). SSR marker s we re species specific, thus they typica lly must be designed for species to be analyzed which has been time consuming and costly (Glenn and Schable, 2005; Goldstein and Pollock, 1997; Zane et al., 2002). For example AFLP markers have been used to assess genetic changes in awned slender wheatgrass over three generations (Ferdinandez et al. 2005). T h e authors revealed a significant reduction in marker variation in the first and second generations

PAGE 72

72 and a large portion of AFLP markers changed frequenc ies T he change in band frequency indicated a genetic shift in two generations of seed increase with the possibility of further losses in later generations. In another study AFLP markers were used to assess the genetic diversity of a composite of four blue gram a seed sources over three generations (Fu et al. 2004). The AFLP markers detected some genetic change over three generations but no t enough to indicate a genetic shift. The composite of seed was able to maintain a high level of genetic diversity over th ree generations of seed production. SSR markers were used to determine if the genetic integrity of wheat was being upheld for up to 24 multiplications (Borner et al., 2000). There was no contam ination found and a high level of genetic identity was upheld in the wheat multiplication populations. The genetic identity of rye was assessed using SSR markers over 7 13 generations of multiplication where there was significantly different allele frequencies detected in four out of six accessions (Chebotar et al ., 2003) T hese studies illustrate d the usefulness of molecular markers in evaluating the genetic diversity and integrity over generations of seed increase. Czarnecki et al. (2008) detected a high level of total g e neti c diversity ( H T =0.309) in natural C leavenworthii populations from north, central and south Florida using AFLP markers Their studies also showed that s eed production populations that originated in central Florida and were increased in northern and central Florida for a various number of generations (G 1 G 4 ) maintained a relatively high level of genetic diversity (0.251). It wa s not known if producing seed from a starting natural population for several generations or in different locations, especially those produced in different locations than the origin of the natural population, affects the genetic diversity of C. leavenworthii

PAGE 73

73 Although t he phenotypic work on these populatio ns over two year s (Chapter 2) indicated that the phenotyp ic diversity was not affected by generation or location a molecular study wa s needed to compliment these findings to determine if genetic changes or shifts occurred over several generation s of seed increase at two locations The objective s of the current study w ere to use SSR markers to assess potential geneti c changes that might occur to C. leavenworthii populations during seed increase in central and northern Florida over three generations to determine if an acceptable level of genetic diversity and integrity wa s maintained over seed increase and, if changes were occurring, quantify the nature and rate of changes Materials and Methods Plant Populations Seed collected from a C. leavenworthii natural population in Reedy Creek Mitigation Bank in Polk County, FL hardines s zone 9a, AHS heat zone 11) was used as the seed source for subsequent increases. This source was referred to as Generation 0 (G 0 ) The collection consisted of 1220 seed heads from 122 individuals randomly selected out of the natural population ( 10 seed heads per individual). One seed lot from the natural population was sent to the North Florida Research and Education Center (NFREC), Quincy, FL in northern Florida and another was kept at the Gulf Coast Research and Education Center (GCREC), Wimau ma, FL zone 9b, AHS heat zone 10) in central Florida At e ach location seed was increased successively for three generations ( G 1 G 2 and G 3 ), starting in spring 2007, by growing ~100 plants from seed of a previous generation and harvesting mature seed heads from

PAGE 74

74 ten open pollinated flowers from each plant Harvested seed heads were dried indoors at each site and then stored at the The seed increase population was designated by its generation and its site of increase (C for the central Florida site and N for the northern Florida site). Therefore, the generations produced in central Flori da were labeled G 1 C, G 2 C and G 3 C and those produced in northern Florida were labeled G 1 N, G 2 N and G 3 N. Additional information on the production of these populations can be found in Chapter 2. Plant Tissue Collection and DNA Extraction Approximately 50 100 mg of young leaf tissue was collected from 55 individual s of each population and dried in a dark container with silic a gel beads for about three days. DNA was extracted using the m icroprep p rotocol of Fulton et al. (1995) and dissolved in 1 x TE buffer ( pH 8.0, 10 mM Tris Cl and 1 mM EDTA) DNA concentrations were determined using the Nanodrop Spectrophotometer ND 1000 ( NanoDrop Technologies, Inc., Wilmington, DE ) a nd adjusted to reach a n 8 10 ng/ L concentration. SSR Marker Analysis Polymerase chain rea ctions (PCRs) were carried out on an Eppendorf Vapo.protect Mastercycler Pro 384 (Eppendorf, Westbury, NY) (Table 3 1) Each reaction was done in a 10 L volume containing 1 x PCR buffer (New England Biolabs, Ipswich, MA, USA) 1. 5 mM of MgCl 2 2 mM dNTP s 0.25 pmol of forward primer with an 2.25 pmol of IRD700 or IRD800 infared dye labeled M13 tail primer ( Eurofins MWG Operon Huntsville, AL USA) and 0.2 5 units of Taq DNA p olymerase with 8 ng of genomic DNA. The forward and reverse primers for each SSR marker were designed from Coreopsis genomic DNA sequences enriched with the GA motif (L. Gong and Z. Deng, unpublished). Depending on the

PAGE 75

75 SSR markers, one of two thermal cycling programs was used for performing PCRs. The first cycling pro g ram started with initial denaturation at 95 C for 45 s followed by 10 cycles of 95 C for 45 s, 55 C for 45 s and 72 C for 60 s followed by 30 cycles of 95 C for 45 s, 45 C for 45 s, and 72 C for 60 s with a final extension of 72 C for 5 min. The second cycling program started with initial denaturation at 95 C for 45 s followed by 7 cycles of 95 C for 45 s, 68 C for 45 s and 72 C for 60 s followed by 31 cycles of 95 C for 45 s, 50 C for 45 s and 72 C for 30 s with a final extension of 72 C for 5 min. PCR amplified DNA fragments were denatured at 95 C for 5 min and then separated in 6.5% denaturing polyacrylamide gels (18 cm long 1 x TBE buffer ) on a LI COR 4300 DNA A naly zer (LI COR biosciences, Lincoln, NE, USA) The DNA Analyzer was pre run for 25 min at 1500 V, 40mA current and 45 C. After the pre run cycle, about 0.8 L PCR reaction w as loaded to each well, and the DNA Analyzer was r u n for about 1.5 h at 1500 V, 40 mA and 45 C. After the DNA samples were fi nished running on the gel, a picture of the gel was stored for later scoring. The bands were scored manually in an absent (0) or present (1) fashion (Figure 3 1) Data Analysis Population genetics computer programs were used to calculate the genetic diver sity within populations, the genetic differentiation among populations and genetic distances among populations. Total genetic diversity ( H T ) within and among populations was estimated, and the proportion of total genetic diversity residing among populatio ns ( G ST ) was calculated in POPGENE (Yeh et al. 1997) where G ST = ( H T H S )/ H T (Nei, 1987) The genetic relationships among populations were assessed based on the allele frequencies within populations and the pairwise genetic distances among populations th at were calculated in POPGENE The

PAGE 76

76 G ST values and genetic distances calculated were compared to the difference in generation s between populations using PROC REG in SAS (SAS, 1997). The output from matrices w as then used in the MEGA program (Kumar et al. 2004) to generate an unweighted pair grouping method with arithmetic means (UPGMA) dendrogram of populations The genetic relationships among populations were also assessed based on the SSR phenoty pes and pairwise Apostol (or simple match) genetic distances. From the Apostol genetic distance matrix, a principal coordinate analysis (PCoA) was performed using DCENTER and EIGEN values provided in NYSYSpc (Rohlf, 2005). The genetic relationships among individuals were displayed in a two dimensional plot using MXPLOT in NTSYSpc Results SSR Alleles and Allele Frequencies The 10 SSR primer pairs amplified a total of 104 alleles in 50 G 0 individuals and the number of alleles amplified per primer pair rang ed from 4 to 17 with an average of 10.4 (Table 3 2 ), indicating a high level of genetic polymorphism at these SSR loci in the population. These primer pairs detected 97, 91 and 87 alleles in the G 1 C, G 2 C and G 3 C individuals, respectively, and 97, 94 and 9 5 alleles in the G 1 N, G 2 N and G 3 N individuals, respectively, although a similar number (46 to 52) of individuals was surveyed in each of the above populations. Compared to the G 0 population, the seed increase populations h ad 6.7% to 16.3% fewer alleles. T he reduction in allele number appeared to be more evident with certai n SSR markers (e.g. COR7, COR10 and COR12) and in advanced generations (e.g. G 3 C and G 3 N) and COR4 in G 3 C (Table 3 2) In some occasions (population G 3 C and marker COR12), up to 35% of t he G 0

PAGE 77

77 Marker COR4 detected 30% fewer alleles in both G 2 C and G 3 C, but did not show such a loss in G 1 N to G 3 N. Except for this marker, no obvious differences were observed between the two seed increa se sites in allele loss. The number of alleles for the re maining six markers was relatively stable from G 0 through G 3 and between the two seed increase sites. There were four alleles (COR18 1, COR8 5, COR4 11 and COR12 16) that were not detected in G 0 but appeared in one or more of the increased populations T here were t wo alleles that were present in G 0 but not observed in any of the increased populations. The lost alleles (COR7 13 and COR7 14) were both amplified by the same SSR marker, COR7. The allel e frequency in the G 0 population ranged from 0.010 to 0.853 per allele with an average of 0.151. A similar range of allele frequency (0.010 to 0.860) was observed in each of the increase populations. The average allele frequency in the increase populatio ns fluctuated slightly, between 0.155 and 0.131, or between 102.7% and 86.8% of that of G 0 To further assess potential changes in allele frequency, SSR alleles were separated into 11 groups based on their observed frequencies : (1) <0.05, (2) 0.05 to <0. 15, (3) 0.15 to <0.25, 0.25 to <0.35, (5) 0.35 to <0.45, (6) 0.45 to <0.55, (7) 0.55 to <0.65, 0.65 to <0.75, (9) 0.75 to <0.85, (10) 0.85 to <0.95, and (11) 0.95, and the distribution of alleles in these categor ies w ere compared among po pulations (Figure 3 2) The most obvious changes appeared to be in Groups 1, 2 and 3. The G 0 population had 35 alleles in Group 1, 42 alleles in Group 2 and 15 alleles in Group 3 (Figure 3 2 ). The G 1 C population had a similar number of alleles in Group 1

PAGE 78

78 (35) and Group 2 (43), but fewer alleles in Group 3 (11). In G 2 C and G 3 C, more alleles fell into Group 1 (48 49), but fewer alleles in Group 2 (34 25) and Group 3 (8 12). A similar trend of change was evident in G 1 N through G 3 N where more alleles had t heir frequencies <0.05 and fewer alleles had their frequencies between 0.05 and 0.25. Total Genetic Diversity ( H T ) within Populations The H T for the G 0 population was 0.1736 (Table 3 3 ). The H T for the G 1 C, G 2 C and G 3 C populations was 0.1666, 0.1558 and 0 .1593, respec tively, corresponding to 9 6 .0 %, 89.7% and 91.8% of the G 0 H T The H T for the G 1 N, G 2 N and G 3 N populations was 0.1706, 0.1590 and 0.1541, respectiv ely, equivalent to 98.3%, 91.6% and 88.8% of that of the G 0 population. Thus, thes e data indicate 1.7% to 11.2% decrease in H T within each increase population. The decrease appeared to be stabilized at around 90% of G 0 H T The change was similar between the two seed increase sites, or the decrease in H T is independent of seed increa se location. Genetic Differentiation ( G ST ) and Distances among Populations G ST values were calculated to determine the proportion of genetic diversity residing among populations. The remaining proportion of genetic diversity resides within populations (Be rg and Hamrick, 1997). The G ST value between G 0 and G 1 C, G 2 C or G 3 C was 0.0244, 0.0394 and 0.0513, respectively, indicating a slight increase in G ST or population differentiation with each successive generation (Table 3 3 ). The G ST value between G 0 and G 1 N, G 2 N or G 3 N was 0.0238, 0.0282 and 0.0399, respectively indicating a similar increase in the G ST value and population differentiation between the two seed increase sites. G ST increase d faster when the populations were increased in central Florida than in northern Florida. G enetic differentiation increase d as the number of generations between each pair of populations increase d (Figure 3 3).

PAGE 79

79 The pairwise genetic distance (Nei, 1978) between G 0 and G 1 C, G 2 C or G 3 C was 0.0080, 0.0141 and 0.0196, respectiv ely (Table 3 3) The pairwise genetic distance between G 0 and G 1 N, G 2 N or G 3 N was 0.0079, 0.0092 and 0.0142, respectively. These values again indicate a slight increase in population differentiation with each successive generation The genetic distances between seed increase populations ranged from 0.00 29 to 0.01 33 with an average of 0.0082. The genetic distance between each population pair increase d as the number of generations between each pair increase d (Figure 3 4). An UPGMA dendrogram was construct ed from the matrix of genetic distances among these populations (Tab le 3 3 ). It clustered G 1 C and G 1 N with G 0 and the two G 2 populations with the two G 3 populations (Figure 3 5 ). This pattern of clustering suggests that a subtle but consistent genetic di fferentiation had occurred as seed increase progressed successively. T he clustering of seed increase populations was mainly by generation rather than by seed increase site. Principal Coordinate Analysis Based on a total of 10 8 x 349 data points a matrix o f Apostol (s imple match) distances among 349 individuals from the seven populations were calculated in NTSYS pc and used to conduct a PCoA. Individuals of the G 0 and six increase populations were widely scattered, and no distinct grouping of any of the pop ulations was obser ved in the PCoA plot (Figure 3 6 ). This distribution suggests that although it had occurred and was increasing over successive generations, the differentiation among the populations was still relatively weak.

PAGE 80

80 Discussion Genetic Different iation between Seed Increase and Original Populations In this study, the highest G ST values (0.0513 and 0.0399) were observed between G 0 and the two advanced generations (G 3 C and G 3 N). These values were close to the G ST value (0.046) detected among three other seed increase populations (SP1, SP2, and SP3) of C. leavenworthii using the AFLP marker system (Czarneck i et al., 2008). The three SP populations originated from a natural population in another county in central Florida (Orange County) and were incr eased under different climatic zones in Florida (Alachua, Gadsden or Pasco County) for one, two, or four generations. When the present and the previous studies were considered together, a total of nine seed increase populations of C. leavenworthii had bee n examined. The two studies suggest that there was only a very low level of genetic differentiation in C. leavenworthii seed production populations and that molecular markers we re extremely powerful in revealing genetic differentiation before it bec a me ev ident at the phenotypic level. Among the major concerns in native forb seed increase are the potential of multiple seed increase locations, production practices and increase generations to cause significant differentiation in increase populations In the current study, G 0 seed s w ere collected from central Florida, which was in USDA cold hardiness zone 9 a and AHS heat zone 1 1 ; the northern Florida increase site was located in USDA cold hardiness zone 8b AHS heat zone 9 while the central Florida increase s ite was in USDA cold hardiness zone 9 b AHS heat zone 1 0 Nevertheless, the G ST values of northern Florida increased populations were not greater than those of the central Florida increased populations. These results suggest that C. le a venworthii seed m a y be increased either i n central or northern Florida.

PAGE 81

81 Slightly higher G ST values were detected between the central Florida increase populations (G 1 C, G 2 C, and G 3 C). This was somewhat unexpected, because the central Florida seed increase site and the site where the G 0 seeds were collected shared a higher level of similarity in environmental conditions, especially in climatic conditions, including AHS heat zoning and USDA cold hardiness zoning. As described in the materials and methods, C. leavenworthii pla nts were grown in the ground beds at the northern Florida increase site while at the central Florida increase site the plants were grown in containers filled with artificial commercial potting substrate. This difference in production system between the tw o seed increase sites might have led to further difference s in other aspects such as fertility and soil moisture levels. Conceivably the container production system would be less similar than the ground bed production system to the soil conditions where the natural population was grown, and higher levels of grow ing condition dissimilarity might result in greater population differentiation or higher G ST values. It remains to be determined if this was the actual cause of the higher G ST values between centr al Florida increased populations and the G 0 population. The effect of increase generation on population differentiation was obvious (R 2 = 0.6299, P = < 0.0001 ) in C. leavenworthii as a linear relationship was detected between population pairwise G ST values and the generational differential between populations (Figure s 3 3 ) The correlation between generational differential and genetic distances had the same effects (R 2 = 0.6 428 P = < 0.0001) (Figure 3 4) If the same rate of change is assumed for extended generations, the regression equation predicts that G ST will remain below 0.10 even after eight consecutive generations of seed increase. It is

PAGE 82

82 unknown whether or not this linear relationship and rate of G ST change will continue for how many generations i n C. leavenworthii seed increase. It seems plausible to assume that as seed increase approaches some unknown generations, the rate of G ST change may slow down and the G ST value will reach a plateau because there will be less genetic diversity in later gen erations causing less changes from generation to generation Possible Causes of Genetic Differentiation in C. leavenworthii Populations Previously Czarnecki et al. (2008) reported a genetic shift in an introduced C. leavenworthii population (R1). The R1 p opulation originated from central Florida (Orange County) and introduced to northern Florida using seeds increased for one generation in central Florida (SP2). When it was analyzed after growing for one year in northern Florida, R1 became more similar to northern Florida natural populations than to central Florida natural populations. It was suspected that gene flow from northern Florida local populations was the main cause of such genetic shift, because C. leavenworthii is an outcrossing, self incompatib le species and there were local natural populations in close proximity. In the present study, there were four alleles that were not detected in G 0 but appeared in one or more of the increase populations, which suggest s that gene flow from outside of the in crease populations might have occurred. At both seed increase sites during the 3 year seed increase, no local natural populations were present close by, and residual C. leavenworthii and other Coreopsis plants from the soil seed bank were removed before t hey came into flowering Thus the possibility of gene flow from outside sources could be ruled out. T wo other possibilities seem to be more likely for the appearing of the four alleles. The G 0 individuals randomly selected for microsatellite

PAGE 83

83 marker anal ysis were not carrying all the alleles present in the G 0 individuals used for seed increase. Evaluation of l arge populations is needed to accurately assess rare alleles. The second possibility i s random genetic drift, which is known to occur frequently f rom generation to generation espec ially when population sizes a re small Smaller populations have greater var iation in gene frequency, where random changes in gene frequency can occur (Futuym a, 1979) It is likely that both factors might have contribute d to the appearance of these alleles in seed increase populations because only a portion of the seed was used to produce each generation and was used for molecular analysis Mutation is another possibility for the appearance of these different alleles. Co mpared to the G 0 population, the increase populations had 6.7% to 16.3% fewer alleles present in the populations and more alleles in the populations with frequencies below 0.05. Further, there were two alleles amplified by the SSR marker COR7 that were pr esent in the G 0 population but not in any of the increase populations. These results seem to suggest that allele loss and allele frequency change indeed occurred during seed increase. Several factors might have contributed to this observed allele loss an d/or allele frequency change, inc luding the founder effect ( small population size typically due to a bottleneck ), genetic drift, unintended selection non random mating etc (Futumya, 1979) It is not known which factors and how much each factor affected this change in the alleles present in the seed increase populations. Because only a portion of the seed produced is used to produce the next generation, the founder effect and genetic drift is likely to occur. Unintended selection could have occurred

PAGE 84

84 bas ed on seed collection practices (early or late in the season ) or seed germination practices that did not break dormancy of some seed. Upholding the Genetic Integrity of C leavenworthii through Seed Increase In the PCoA plot the individuals from G 0 and the six increase populations were intermixed and did not form any definitive clustering by generation or location. This indicates that although some genetic changes (or shift s ) had occurred, the increase populations remained very similar genetically to the o riginal population. Overall the current molecular marker study confirms that the genetic integrity of C. leavenworthii was maintained during seed increase for up to three generations using the current production practices. It is likely that more generati ons could be produced from the original, natural popul ation as indicated by the linear regression analysis. T he linear regression equation also suggest s that l ater generations should be assessed by common garden studies or molecular analysis to ensure tha t satisfactory level of genetic integrity is preserved Two similar studies assessed the genetic integrity of wheat and rye. Borner et al. (2000) determined that a high degree of genetic identity was upheld in up to 24 generations of multiplication in whe at using SSR markers Chebotar et al. (2003) reported that the genetic integrity of two of the six rye accession s was upheld for 6 12 generations of multiplication using SSR markers. The other four accessions had significantly different allele frequencie s after bein g multiplied f o r 7 13 times which could be due to sample size or the lack of detection of rare alleles. These studies illustrate that the number of generations a species can uphold its genetic integrity depends on the species and population s tructure It is important to use a large starting population

PAGE 85

85 that represents a broad range of the alleles present in the population as well as large populations for producing each generation of seed increase. Summary SSR marker s were highly polymorphic an d detected an average of 10.4 alleles per primer pair in the G 0 population of C. leavenworthii SSR markers also prove d to be very powerful in revealing changes in alleles and allele frequencies. Total genetic diversity, G ST and genetic distance values changed slightly during seed increase at both locations when compared to the G 0 population The largest differences were observed in the later generations (G 3 ) from both locations with the greatest G ST value of 0.0513 and genetic distance of 0.0196 Gene tic differentiation ( G ST ) and genetic distance between populations increased linearly with the increase in generation There did not appear to be a significant e ffect from producing seed in a different location from that of the natural population (G 0 ). T he observed genetic changes did not seem to cause obvious genetic differentiation between the increase populations and the original population which is congruent with the phenotypic data of the same C. leavenworthii populations These results indicate th at the overall genetic integrity of C. leavenworthii populations was upheld during three generations of seed increase Current seed production practices can continue for C. leavenworthii but later seed production populations should be evaluated to confir m the genetic integrity of C. leavenworthii seed increase populations.

PAGE 86

86 Table 3 1 Primer sequences for 10 simple sequence repeat markers used to evaluate seven C leavenworthii populations produced in northern and central Florida SSR markers Primers Prim er sequences (5' 3') Annealing temperature ( C) Cor4 Forward ACCCAATCCAATCCCTTCTC 50 Reverse TCATCGTTCGTGTGACACATT Cor7 Forward GAGAGAACGGGGGAAAGAG 50 Reverse TTCCAATCCTAAATACCTAGAAACC Cor8 Forward GTTCTTTGGGAGGGTGTTATCG 50 Reverse CATGGCATCACAA GCAGGTT Cor10 Forward GAAGCCCAAAAGCCTAATTG 45 Reverse TTTCTCCTAGCTTTCCTGCTG Cor12 Forward CTCACCCGTGATGTCGAGTT 45 Reverse ACATCTCACCCTCCCCTGAC Cor18 Forward AAGCACACATAACCGCTCCT 50 Reverse TGCTCTCTGCCATGAATCAC Cor19 Forward GGATCTCCTTCTTGCCTC CT 50 Reverse AGCCATAAACCCCAGATCCT Cor21 Forward GAAAATGAGAAGACGAGGAA 45 Reverse AAGATGCTTTAACTTGACAGATT Cor23 Forward CAGCTGGCCCATATCCTTCT 45 Reverse AGCTCGTCACAAAGGTTGAGG Cor65 Forward GGCCACGTCTCCTTTTTACA 45 Reverse TTGAAATTGAAATGGGGTATGA G

PAGE 87

87 Table 3 2 The number of individuals, number of alleles and percentage of polymorphic loci used to evaluate the genetic differences in seven populations of C leavenworthii produced in northern and central Florida Populations G 0 G 1 C G 2 C G 3 C G 1 N G 2 N G 3 N Combined Individuals surveyed (no.) 50 51 49 50 52 46 51 349 Alleles detected COR4 11 10 7 7 11 9 11 12 COR7 14 11 11 11 9 11 10 14 COR8 8 8 7 6 6 8 9 9 COR10 6 6 5 4 6 4 5 6 COR12 17 16 14 11 18 16 13 18 COR18 4 5 4 4 4 4 4 5 COR19 13 11 1 3 13 13 13 13 13 COR21 13 13 12 13 13 11 12 13 COR23 14 13 14 14 13 14 14 14 COR65 4 4 4 4 4 4 4 4 Total 104 97 91 87 97 94 95 10 8 Pp (%) 96.1 88.2 83.3 78.4 88.6 88.7 88.2 Table 3 3 Total genetic diversity ( H T ) G ST and genetic distances for se ven populations of C leavenworthii produced in northern and central Florida H T G 0 G 1 C G 2 C G 3 C G 1 N G 2 N G 3 N G 0 0.1736 0.0244 z 0.0394 0.0513 0.0238 0.0282 0.0399 G 1 C 0.1666 0.008 0 z 0.0382 0.0367 0.0127 0.0231 0.0336 G 2 C 0.1558 0.0141 0.0133 0.0183 0 .0354 0.0195 0.013 0 G 3 C 0.1593 0.0196 0.0128 0.005 0 0.0342 0.0346 0.0143 G 1 N 0.1706 0.0079 0.003 0 0.0123 0.0119 0.024 0 0.0292 G 2 N 0.159 0 0.0092 0.0069 0.0052 0.0114 0.0075 0.0244 G 3 N 0.1541 0.0142 0.0113 0.0029 0.0034 0.0096 0.0072 z G ST values ar e above the diagonal and pairwise genetic distances (Nei, 1978) between populations are below the diagonal.

PAGE 88

88 Figure 3 1. Gel image of simple sequence repeat marker bands (alleles) amplified from C. leavenworthii individuals by COR12 and detected by the L I COR 4300 DNA Analyzer. Figure 3 2. The number of SSR marker alleles with respect to their frequencies of occurrence in the C. leavenworthii source population (G 0 ), three populations increased in central Florida (G 1 C, G 2 C and G 3 C) and three population s increased in northern Florida (G 1 N, G 2 N and G 3 N).

PAGE 89

89 Figure 3 3. Regression of G ST values when compared by the difference in the number of generations between each combination of seven C. leavenworthii populations. Figure 3 4. Regression of the geneti c distances when compared by the difference in the number of generations between each combination of seven C. leavenworthii populations.

PAGE 90

90 Figure 3 5 UPGMA d endrogram of the G 0 population and six seed increase populations (three from central Florida: G 1 C G 2 C and G 3 C; and three from northern Florida: G 1 N, G 2 N and G 3 N) of C leavenworthii populations based on their pairwise genetic distances.

PAGE 91

91 Figure 3 6 Plot of 50 individuals of the G 0 ( ) population and 299 individuals of six seed increase populations ( G 1 C, G 2 C, G 3 C, G 1 N, G 2 N, G 3 N) of C leavenworthii based on the first two principal coordinates from 10 8 SSR allele s

PAGE 92

92 CHAPTER 4 INTERSPECIFIC HYBRID IZATION BETWEEN COREOPSIS LEAVEN WORTHII AND COREOPSIS TINCTORIA AND EFFECTS ON PROGE NY GROWTH, DEVELOPME NT AND REPRODUCTION Justification Coreopsis C. leavenworthii grow throughout the state (USDA, 2011a) and along roadsides in hig hway beautification projects by the Florida Department of Transportation ( FDOT ) Because C. leavenworthii seed is solely produced in Florida and many other species of Coreopsis grow in Florida, it is important to prevent contamination from other Coreopsis species into natural and seed production C. leavenworthii populations Most species are not compatible with C. leavenworthii but C. tinctoria has been found to set interspecific hybrid seed with C. leavenworthii and the interspecific F 1 hybrids have 22% pollen stainability (Smith, 1976) C. tinctoria has been reported as growing in the northern part of Florida (Smith, 1983) and in some counties in Florida (USDA, 2011b). FDOT has also reported using C. tinctoria in highway beautification projects. Becau se it is known that both species are growing in Florida (USDA, 2011a, 2011b) and they are cross compatible, information on the level of compatibility and the effects of interspecific hybridization is needed. This will help determine the fitness level of i nterspecific hybrids compared to each species and methods of prevention of gene flow and interspecific hybridization between these two species Many studies have assessed the effects of interspecific hybrid ization by producing synthetic hybrids and evalua ting the vegetative and reproductive characteristics of these interspe ci fic hybrids. Frequently the F 1 generation does not exhibit all of the potential consequences of interspecific hybridization, so it is necessary to produce F 2 or later and backcross

PAGE 93

93 ge nerations if possible. The backcross population can be very useful to predict the effects of interspecific hybridization because in natural settings most of the next generation will result from backcross pollinati ons (Hufford and Mazer, 2003). Arriola and Ellstrand (1997) found that synthetically produced interspecific F 1 hybrids between S orgh u m halepense and S. bicolor were not significantly different for the seven reproductive or vegetative characteristics evaluated compared to both parent al species Sy nthetic interspecific F 1 Oryza rufipogon and O. sativa hybrids were evaluated for eleven reproductive and vegetative characteristics during the entire life cycle of the crop and found that the hybrids had the same fitness level as the parent al species (Son g et al. 2004). Th e hybrids were slightly inferi or in reproductive characteristics with lower seed set and pollen viability but showed greater fitness in vegetative characteristics indicating that the interspecific hybrids we re viable enough to persist in later generations allowing interspecific hybridization to continue. S everal synthetic interspecific hybrid populations were produced using Eucalyptus gunnii as the maternal parent and E. cordata E. dalrympleana E. viminalis E. macarthurii E. nitens E. globulus and E. ovata as the pollen sources (Potts et al. 1987). Although wide intra and interspecific crosses we re more likely to set seed in Eucalyptus rather than selfing and close intraspecific crosses p lants that we re produced from selfing or close intraspecific crosses we re more likely to survive than wide intra and interspecific crosses. Plants that did survive from interspecific crosses were significantly taller and more vigorous than those from intraspecific crosses. These studies illus trate that interspecific crosse s can produce vigorous and viable interspecific hybrids that can

PAGE 94

94 persist in later generations, making the evaluation of C. leavenworthii C. tinctoria hybrids important and critical. T he information known is limited but inter specific hybridization could greatly affect natural and seed production C. leavenworthii populations by contamination for several generations T he objective of this study was to determine the fitness level of C. leavenworthii C. tinctoria interspecific hy brids in the F 1 and F 2 generations by comparing them to intraspecific populations of C. leavenworthii and C. tinctoria This will help determine if these interspecific hybrid s are likely to persist in nature and affect pure populations of either species. Materials and Methods Seed Source C. leavenworthii seed was collected by Nan cy Bissett from a natural population at the Reedy Creek Mitigation Bank in Polk County, FL. C. tinctoria seed was purchased from Wildseed Farms in Fredericksburg, TX. Interspecifi c Pollinations and Hybrid Population Development Synthetic i nterspecific crosses were made in June, July and August 2007 to produce two F 1 hybrid population s one (COLE F 1 ) using C. leavenworthii as the maternal parent and the other one (COTI F 1 ) using C. tinctoria as the maternal parent ( Figure 4 1) Forty C. leavenworthii and thirty five C. tinctoria plants were used to make each F 1 population, where flower buds on the seed parent plants, ten for COLE F 1 and five for COTI F 1 were bagged prior to pollina tion. Once the disk flowers of the maternal plants were mature on the flower head and anthe sis was occurring, the pollen was removed from the flower head. The pollen source was collect ed by removi ng pollen from one flower head of each plant of the approp riate population into a plastic container

PAGE 95

95 T he flower head of the maternal plant was dipped into the container with the appropriate pollen source. The flower head was rebagged and not removed until the seed head matured. After all of the seed for each s eed parent was mature, it was collected and bulked per maternal parent and cross type which will be referred to as a line from this point forward. The F 1 population s w ere self pollinated to produce two synthetic F 2 population s (COLE F 2 and COTI F 2 ) using the same pollination technique Selfing was done between 3 and 25 November 2008 Parental plants used for both cross types were grown in 3 2 L containers filled with Vergrow verlite container mix A (Tampa, FL) mixed with 8.6 g of 15 N 3.9 P 1 0K ( 5 6 mo. O smocote, The Scotts Co., LLC, Marysville, OH) per container in the greenhouse (29.4/23.9C). Supplemental lighting was provided from 12:00 AM to 4:00 AM every morning until all seed was mature and collected from the plants. Intraspecific Pollination and R eference Population Development Synthetic C leavenworthii intraspecific crosses (COLE REF ) were made between 27 July and 5 August 2007 to produce the self population ( Figure 4 1) C tinctoria intraspecific crosses (COTI REF ) were made between 7 and 10 A ugust 2007 to produce the self population The same C. leavenworthii and C. tinctoria plants were used t o produce the intraspecific populations as the interspecific F 1 populations with five flower buds bagged f or each cross type using the same pollination technique previously described Crossability The number of seed produced per seed head was counted for three seed heads per plant per cross type from the hand pollinations completed in the greenhouse. Seed

PAGE 96

96 germination tests were completed for both year s by counting the number of seedlings that emerged for each population. Assessing Progeny Growth, Development and Reproduction Common garden studies were conducted in 2009 and 2010 to evaluate the effects of interspecific hybridization on the fitness of C. l eavenworthii In each year individuals from the various generations were grown under the same conditions in the greenhouse and subsequently in the field, and they were assessed on a number of vegetative and reproductive characteristics, including plant he ight, plant dry weight, days to flower, pollen stain ability seed production and seed germination Growing conditions 2009. Seed w ere so wed on 19 February 2 009 into 14 8 m L c ell Speedling flats (32 cell flats) (Sun City, FL) in Fafard 3B soil ( Anderson, SC ) media. The seed were placed on the surface of the soil and germinated under mist irrigation in the greenhouse (29.4/23.9C). The seedlings were transplanted in 3 1 m L cell Speedling flats (128 cell flats) in Fafard 3B soil media on 14, 15 and 16 March. The plants were then transplanted in the field on 4 May The field was a mix of Myakka, Haplaquents and St. Johns sandy soil. The beds were 25 cm high, 71 cm wide at the t op and 81 cm wide at the base. Plants were spaced 55.9 cm apart in rows with two r ows per bed. The rows in each bed were spaced 40 cm apart. The beds were fumigated with 67% methyl bromide: 33% chloropicrin at 197 kg ha 1 The plants were fertilized with a total of 169 kg ha 1 of nitrogen, 37 kg ha 1 of phosphorus and 210 kg ha 1 of potassium throughout the year through drip irrigation. 2010 Seed w ere s owed on 20 January 2010 into 14 8 m L cel l Speedling flats in Fafard 3B soil media. The seedlings were transplanted in 3 1 m L cell Speedling flats in

PAGE 97

97 Fafard 3B soil media on 15 and 16 F ebruary. Plants were trans planted in the field i n a single row 30.5 cm apart on 5 April. All field conditions were the same as the previous year except that no fertilization was applied and the plants were irrigat ed through seepage irrigation. Data colle ction Plant height and dry weight. P lant height was quantified in centimeters by measuring the tallest point for each plant growing in its respective year on 13 July 2009 and 2 June 2010 In 2009 plants were harvested from the field on 15 21 July, 72 78 d ays after field planting, and dried between 21.1 34.4C. In 2010 plant s were harvested from the field on 15 17 June 71 73 days after field planting, and dried at 37.8C and 20% relative humidity for 3 4 weeks before weight was recorded Days to flower. T he number of days to flower was recorded from 23 April to 31 July 2009 and 28 April to 7 June 2010 The date was recorded for each plant in the field when the first flower open ed and the number of days was counted from seed sowing. Field s eed production a nd seed germination Five s eed heads were collected from each plant in the field on 8 July 2009 and 2 and 3 June 2010 and the number of seed was counted The seed collected in the field from 2009 were sowed for seed germination tests where 100 seed colle cted from each plant in the field were sowed in 14 8 m L cell Speedling flats in Fafard 3B soil media for each block separately and the number of emerged seedlings was counted three weeks after sowing All other seed germination conditions were the same as previously described. Pollen stainability. For 2009 flower heads were collected in September and October. A total of t en flower heads were collected for each line (maternal parent and

PAGE 98

98 cross type ) prior to anth esis and bulke d together for pollen staining. Flower heads were soaked in about 200 L of la ctophenol cotton blue stain ( ENG Sci entific Inc., Clifton, NJ) for about 3 h at room temperature. About 10 L of stain was pipetted on the slide with a few ray flowers and a cover slip was placed on top of th e stain. Three slides per line were prepared and pictures of four fields per slide were taken from the microscope. An average of 818 pollen grains per line was counted. Pollen grains were considered stainable if they were stained a deep, uniform blue an d were very developed (large and plump). Flowers heads were collected on 3 June 2010 and stained in lactophenol cotton blue stain ( Fluka Ana lytical, Buchs, Switzerland ) overnight at room temperature O ne flower head per line (maternal parent and cross typ e) from each block was collected and bulked, resulting in eight flower heads per line Three slides per line were prepared and pictures of three fields per slide were taken. An average of 258 pollen grains per line was examin ed. All other procedures for pollen staining were th e same as the previous year Experimental design Thirty seed per line w ere split into three replications of ten seeds per replication. Each replication was sowed into an individual cell of a 14 8 m L cell Speedling flat. There were 25 lines for each population in 2009 and 13 to 25 lines for each population in 2010. The plants were grown in a randomized complete block design in the field with 15 blocks in 2009 and 8 blocks in 2010 with 1 plant/line /block. A total of 375 plants in 20 09 and 104 375 in 2010 were planted per cross type. All plants from all six cross types were randomized within each bl ock.

PAGE 99

99 Statistical analysis The data points for each characteristic measured for each line were averaged by block. The average for each bl ock was then used for statistical analysis. Significant differences were found for each data point evaluated by Analysis of Variance (ANOVA) using PROC GLM in SAS (SAS, 1997) Data w ere transformed using the Arcsine Square R oot method for seed germinatio n tests and pollen stainability Based on the ANOVA test, further statistical analysis was performed using the Tukey W Procedure for mean separation analysis in SAS. Results Crossability between C. leavenworthii and C. tinctoria On average the crosses amo ng 40 C. leavenworthii individuals ( COLE ) produced 50.2 seed per seed head and the crosses among 35 C. tinctoria individuals ( COTI ) produced 14.9 seed per seed head (Figure 4 2) Thus, COLE produced 236.9% more seed per seed head th an COTI indicating a r em arkable difference between the two species in seed produ ction per seed head (Table 4 1 ) This difference seems to be largely due to the difference between the two species in the number of female florets that each flower head bears. On average the inters pecific c rosses with C. leavenworth ii being the seed parent (COLE x COTI) produced 61.9 seed per seed head, and the interspecific crosses with C. tinctoria being the seed parent (COTI x COLE) produced 25.0 seed per seed head (Figure 4 2). Thus, the COLE x COTI crosses produced significantly more seed (23.3%) than the COLE crosses, and the COTI x COLE crosses produced significantly more seed (67.8%) than the COTI crosses, indicating full crossability between the two species.

PAGE 100

100 The seed germination rates for s eed produced by hand pollinations were not significantly different between year s (Table 4 1 ). The mean seed germination rates for COLE was 67.7% and for the COTI was 67.9%, which was not significantly different (Figure 4 3) The mean seed germination rat es for the seed produced from crosses COLE x COTI was 69.7% and from crosses COTI x COLE was 71.7%, which were not significantly different from each other or the self pollinations. These results show ed again that there were no biological factors preventin g these interspecific seed from germination and tha t the two species we re fully compatible. Effects of Interspecific Hybridization on Plant Height There were significant differences between year s and among populations for plant heig ht (Table 4 2 ). The mea n plant height of the various populations in 2010 was significantly less than that in 2009: 10.4 14.2% reduction for the parental populations, 17.6 18.0% reduction for the F 1 populations and 9.7 11.1% reduction for the F 2 populations. In 2009 the mean pla nt height of the COLE REF and COTI REF population s was 69.7 and 95.6 cm, respectively (Figure 4 4). In 2010, the mean plant height of the COLE REF and COTI REF populations were 59.8 and 85.6 cm, respectively. Thus the plants of the COTI REF population w ere on average 25.9 cm (in 2009) or 25.8 cm (in 2010) taller than those of the COLE REF population. In 2009, the mean plant height of the COLE F 1 population was 82.7 cm (Figure 4 4) similar to the mid parent value (82.6 cm) (Table 4 3 ) The same trend wa s observed in 2010, where the mean plant height of COLE F 1 (67.8 cm) was similar to the mid parent value (72.7 cm). The mean plant height for COLE F 1 was significantly greater than that of the COLE REF population and significantly less than that of the CO TI REF

PAGE 101

101 population in both year s. The mean plant height for the COLE F 2 population was 77.6 cm in 2009 and 70.1 cm in 2010, similar to the mid parent values and the values of the COLE F 1 population in both year s. Overall, similar results w ere observed for the COTI F 1 and COTI F 2 populations as compared to the mid parent values, except for that in 2009 the COTI F 1 s value see med to be greater than the mid parent value (Table 4 3 ) indicating no effect due to interspecific hybridization (Table 4 4 ) In 2009, COTI F 1 and COTI F 2 populations had a mean plant height of 91.8 and 89.4 cm, respectively, and in 2010, they had a mean plant height of 75.6 and 79.5 cm, respectively (Figure 4 4) When the two F 1 and F 2 populations were compared, the mean plant heights o f the COTI F 1 s were 11.0% (in 2009) to 11.5% (in 2010) taller than the COLE F 1 s; the COTI F 2 s were 15.2% (in 2009) to 13.4% (in 2010) taller than the COLE F 2 s. These differences were consistent for the F 1 to F 2 populations and indicate a maternal effe ct from C. tinctoria on the plant height of its hybrids with C. leavenworthii Effects of Interspecific Hybridization on Plant Dry Weight There were no significant differences between year s for plant dry weight so the data from the two year s were combined for determining the effects of interspecific hybridization on plant dry weight (Table 4 2 ). The mean p lant dry weight for COLE REF (0.083 kg) was significantly less th an COTI REF (0.156 kg) (Figure 4 5). Thus on average the plant dry weight of the COTI plants were 88.0% greater than that of the COLE plants impos ing a maternal effect on the F 1 and even F 2 generations The plant weight of COTI F 1 was 21.3% greater than COLE F 1 and the plant weight of COTI F 2 was 21.7% greater than COLE F 2 maternal effect seemed to be consistent from F 1 to F 2

PAGE 102

102 A heterotic effect (Table 4 4 ) was present i n F 1 populations, particularly in the COTI F 1 population. The plant dry weight of COLE F 1 was 0.136 kg (Figure 4 5) 4.6% greater than the mid parent value (0.130 kg per plant ) (Table 4 3 ) Whereas the plant weight of COTI F 1 was 0.165 kg, which is 26.9% greater than the mid parent value, even 5.8% greater than the high parent value, which was COTI REF (0.156 kg). The mean plant dry weight for COLE F 2 was 0. 115 kg (Figure 4 5) This is only 84.5 % of the mid parent value (Table 4 3 ) and 84.6% of the COLE F 1 a breakdown in plant dry weight in the F 2 generation. A similar breakdown seemed to be present in the COTI F 2 population (Table 4 4 ) The COTI F 2 (0.140 kg ) was only 84.8% of its F 1 2 greater than the mid parent value. Effects of Interspecific Hybridization on Days to Flower There were significant differences among p opulations and between year s for the num ber of days to flower (Table 4 2 ). The mean number of days to flower of the various populations in 2010 was significantly greater than that in 2009: 6.0 8.3% increase for the parental populations, 0.3 4.7% increase for the F 1 populations and 6.3 7.8% increase for the F 2 populations. In 2009 and 2010 COLE REF (107.3 and 113.7 days respectively ) took significantly fewer days to flower than COTI REF (114.2 and 123.7 days respectively ) ( Figure 4 6) T he plants of the COTI REF population took on average 7.0 (in 2009) and 10.0 (in 2010) days longer to flower than th ose of the COLE REF population. There seemed to be an interaction between year and the effect of interspecific hybridization. In 2010 when the plants were gr own in spring to summer, the mean number of days to flower for the COLE F 1 COLE F 2 COTI F 1 and COTI F 2 populations

PAGE 103

103 were 117.4, 119.9, 120.5, 121.9 days, respectively (Figure 4 6), all similar to the mid parent value (117.6 days) (Table 4 3 ) and not showi ng any heterotic, hybrid breakdown or maternal effects. However, when the plants were grown in 2009, the two F 1 populations (COLE F 1 and COTI F 1 ) and one F 2 population (COTI F 2 ) had their mean number of days to flower between 117.1 and 114.2 days, greater than their mid parent value (110.8 days). The only exception was the COLE F 2 population, whose number of days to flower was 111.2 days. Thus in 2009, the F 1 and F 2 populations show ed a heterotic effect (Table 4 4 ) Effects of Interspecific Hybridization on Pollen Stain ability There were significant differences found bet ween year s and among populations in 2009 for pollen staina b i lity (Table 4 2 ). In 2009 the mean pollen stainability for COLE REF (20.1%) and COTI REF (16.7%) were not significantly differe nt ( Figure 4 7) The mean pollen stainability for COLE F 1 was 3.8% and for COTI F 1 was 6.9% Thus, the interspecific hybrids had pollen stainabilities significantly less than COLE REF and COTI REF and the mid parent value (18.4%) (Table 4 3 ) The mean pollen stainability for COLE F 2 was 6.1% and for COTI F 2 was 7.8% which w as less than the mid parent value significantly lo wer than COLE REF and COTI REF but similar to COLE F 1 and COTI F 1 The significant decrease for th e mean pollen stainability for the COLE F 1 and F 2 and COTI F 1 and F 2 populations from the COLE REF and COTI REF populations indicate d a reduction possibly due to chromosome mispairing in the interspecific progeny of C. leavenworthii and C. tinctoria (Table 4 4 ) The mean pollen stain ability of the parental populations in 2010 was significantly less than that in 2009 (17.4 32.3%). In 2010 none of the populations were significantly

PAGE 104

104 different for pollen stainability (Table 4 2 ) The mean p ollen stainability for COLE REF and COTI REF was 3.5 % and 5.4%, respectively (Figure 4 7) The mean pollen stainability for COLE F 1 and COTI F 1 was 5.5% and 4.0%, respectively. The mean pollen stainability for COLE F 2 and COTI F 2 was 4.5% and 2.7%, respectively. The F 1 an d F 2 populations from both species were similar to the mid parent value (4.5%) (Table 4 3 ). Effects of Interspecific Hybridization on Seed Production and Seed Germination Seed production of hand pollinated F 1 populations The mean number of seed produced pe r seed head for the COLE x COTI F 1 pollinations was 35.3 seed (Figure 4 2), which wa s 70.3% of that of the COLE pollinations and 57.0% of that of the COLE x COTI pollinations. The COLE x COTI F 1 pollinations produced significantly less seed per seed head than the COLE and COLE x COTI pollinations. The mean number of seed produced by the COTI x COLE poll inations was 31.9 seed, which wa s 214.1% of that of the COTI pollinations and 127.6% of that of the COTI x COLE pollinations. The COTI x COLE F 1 pollinati ons produced significantly more seed than the COTI pollinations and not significantly different amounts of seed than the COTI x COLE pollinations. Seed p roduction of o p en pollinated F 1 and F 2 p opulations There were significant differences for seed producti on from open pollinations in the field between year s and among the population s (Table 4 2 ). The mean seed production from open pollinations in the field of the various populations in 2010 was significantly greater than that in 2009: 31.9 44.3% increase fo r the parental populations, 51.0 78.8% increase for the F 1 populations and 41.8 77.1% increase for the F 2 populations.

PAGE 105

105 In 2009 COLE REF produced significantly more seed (476.0 seed) than COTI REF (416.2 seed) ( Figure 4 8) In 2010 COLE REF again produced more seed (627.9 seed) than COTI REF (600.4 seed) but the difference (27.5 seed) w as not signifi cant The mean seed production for COLE F 1 was 294.0 seed in 2009 and 525.8 seed in 2010 (Figure 4 8) which was less than the mid parent value (446.1 and 614. 1 seed respectively ) (Table 4 3 ) The mean seed production for COLE F 1 was significantly less than COLE REF and COTI REF in 2009 and was significantly lower than COLE REF and not significantly different than COTI REF in 2010 The mean seed production fo r COLE F 2 was 309.7 seed in 2009 and 548.6 seed in 2010, which was less than the mid parent value for both year s. The mean seed production for COLE F 2 was not significantly different from COLE F 1 but was significantly less than COTI F 1 in 2009 and was not significantly different than COLE F 1 and COTI F 1 in 2010. The mean seed production for COTI F 1 was 336.9 seed in 2009 and 508.7 seed in 2010 (Figure 4 8) which was significantly less than COLE REF and COTI REF for both year s and less than the mid parent value (Table 4 3 ) for both year s The mean seed production for COTI F 2 was 378.8 seed in 2009 and 537.2 seed in 2010 which was less than the mid parent value for both year s The mean seed production for COTI F 2 was significantly greater than COLE F 1 and not significantly different from COTI F 1 in 2009 and not significantly different than COLE F 1 and COTI F 1 in 2010 Although seed production was greater in 2010, both year s were similar where the parental populations were greater than the F 1 and F 2 popula tions and the F 2 populations increased slightly from the F 1 populations suggest ing chromosome mispairing and

PAGE 106

106 dilution (Table 4 4 ) in the F 1 and F 2 populations with slight recover y in the F 2 populations Seed germination of the hand pollinated F 1 populatio n The mean seed germination rate for the seed produced by COLE x COTI pollinations was 79.1%, which was significantly greater than the COLE and COLE x COTI hand pollinations (Figure 4 3). The mean seed germination rate for the seed produced by hand in the COTI x COLE F 1 pollinations was 83.0%, which was significantly greater than the COTI and COTI x COLE hand pollinations. Seed germination of open pollinated F 1 and F 2 p opulations There were sig nificant differences among populations for seed germination rat es from the seed collected from open pollinations in the field in 2009 (Table 4 2 ). The mean seed germination rate for COLE REF (42.9%) was s ignificantly l ower than COTI REF (55.7%) ( Figure 4 9) On average the seed germination rate for COLE F 1 w as 49.7% and COTI F 1 was 50.9% not significantly different than COLE REF or COTI REF the mid parent value (49.3%) (Table 4 3 ) The mean seed germination rate for COLE F 2 was 46.4% and for COTI F 2 was 49.3% both similar to the mi d parent value and C OLE F 1 and COTI F 1 The F 1 and F 2 populations were not significantly different from either parental population indicating that interspecific hybridization had no obvious effect on seed germination (Table 4 4) Discussion Fu ll Compatibility between C. leavenworthii and C. tinctoria Both interspecific hand pollination crosses (COLE x COTI and COTI x COLE) produced more seed per seed head than the COLE and COTI hand pollinations when comparing the maternal parents. C. leavenwo rthii plants produced more seed per seed

PAGE 107

107 head from the hand pollinations compared to C. tinctoria possibly because there we re more disk flowers per flower head in C. leavenworthii than C. tinctoria Because of this, there appeared to be a maternal effect for seed production for the synthetic hand pollinations. As previous work by Smith (1976) has shown, C. leavenworthii and C. tinctoria are highly compatible. When Parker (1973) made interspecific hybrid crosses between C. leavenworthii and C. tinctoria 29 F 1 plants were recovered from one seed head and in the current study about 30.5 F 1 plants were recovered per seed head. Effects of Interspecific Hybridization In general there are two effects of interspecific hybridization, heterosis and outbreeding dep ression (Hufford and Mazer, 2003). Heterosis occurs when one generation of hybridization result s in increased fitness in the F 1 generation due to increased heterozygosity but will slightly decline in later generations due to increased homozygosity This was detected for number of days to flower (2009) in the current study (Table 4 4 ) Outbreeding depression is the reduction in mean population fitness between genetically distinct populations relative to parental populations. There are two types of outbre eding depression, hybrid breakdown and dilution. When hybrid breakdown occurs, there is a reduction in fitness due to the disruption of co adapted gene complexes. The F 1 generations appeared to be exhibit ing heterosis in the F 1 generation and hybrid brea kdown in the F 2 generations for plant dry weight. Dilution is the reduction in fitness caused by the expression of only one half of the alleles of each parent which appeared to affect seed production Heterosis D ays to flower (in 2009) appeared to be und ergoing heterosis because the values of both characteristics were greater than the mid parent value. Arriola and Ellstrand

PAGE 108

108 (1997) found no difference in the mean number of days to flower between interspecific hybrid and parental plants when S. bicolor and S. halepense were crossed, where there w ere not significant differen ces for either parental species or the interspecific hybrids. These studies indicate that the effects on plant development of interspecific hybrids can change depending on the species st udied. Heterosis followed by hybrid breakdown Plant dry weight appeared to be undergoing heterosis followed by hybrid breakdown because there was an increase in the F 1 generation and then a decrease in the F 2 generation. Similarly, Arriola and Ellstrand ( 1997) found no difference in aboveground biomass between the interspecific hybrids and parental populations for crosses between S. bicolor and S. halepense C hromosome mispairing and outbreeding depression Pollen stainability reduce d drastically in the F 1 and F 2 generations in 2009. In 2010 the percentage of stainable pollen was very similar among the populations. Seed production in the field was undergoing outbreeding depression because of r eduction s in both year s. Both of these reductions were due at l east partially to chromosome mispairing producing non viable pollen and a reduced number of seed ( Parker, 1973; Smith 1976 ). The reduction in seed production was partially due to dilution because of the difference in number of disk flowers between C. leav enworthii and C. tinctoria and their interspecific hybrids. Arriola and Ellstrand (1997) found much variation in pollen stainability in interspecific hybrid (0 98%) and parental populations (0 92%), indicating no effect on interspecific hybridization on po llen stainability. The same wide range was found for seed production per panicle between the interspecific hybrid and parental populations

PAGE 109

109 a gain indicating n o effect from interspecific hybrid i zation. Pollen viability, determined by pollen germination te sts, and seed set was significantly lower for the interspecific hybrids than either parental species, where O. sativa had significantly higher pollen viability and seed set than O. rufipogon when these two species were crossed (Song et al., 2004). These s tudies indicate that the effects on pollen viability and seed production of interspecific hybrids can change depending on the species studied. No effects detected on plant height and seed germination Plant height was not affected by interspecific hybridiza tion because the averages for the F 1 and F 2 populations were similar to the mid parent values for both years. Seed germination did not appear to be affected by interspecific hybridization compared to the parental populations A ll of the F 1 and F 2 populat ions had seed germination rates between those of the parental populations. Maternal Effects on Interspecific Hybridization The vegetative characteristics, plant height and dry weight, and days to flower w ere a ffected by the maternal species. When C. leave nworthii was used as the maternal parent, plant height was consistently lower compared to the crosses where C. tinctoria was the maternal parent. This occurred for all three cross types ( intraspecific F 1 and F 2 ) over both year s. T he C leavenworthii pop ulations were consistently lower than the C. tinctoria populations for plant dry weight indicating maternal effects When C. leavenworthii was used as the maternal parent, the individuals of these populations on average took fewer days to flower compared to C. tinctoria as the maternal parent in 2009 indicating maternal effects. In 2010 both the F 1 and F 2 populations had similar values to the mid parent value, not showing any maternal effects. The only exception to

PAGE 110

110 this was for COLE F 1 and COTI F 1 Th e other charac teristics evaluated did not indicate maternal effects. Effects of Differences in Environme ntal Conditions The characteristics showed similar trends over both year s but the values were quite different for each characteristic most likely due t o differences in the environmental conditions over both year s. The fertilization rate and irrigation methods used between the two year s were very different. In 2009 drip irrigation with fertilizer was applied to the plants. In 2010 supplemental fertiliz ation was not used and the plants were irrigated by seepage irrigation. The growth and development of the plants in 2009 could have resulted from competition for light, nutrients and water because the plants were grown in two rows per bed instead of a sin gle row as in 2010. There was not much difference in temperature or humidity over both year s although both conditions were slightly higher in 2009 (Table 4 5 ). This most likely did not affect the growth and development of the plants. The other major effe ct on the growth and development of the plants is the differences in day length between the two year s. It took the plants a fewer number of days to flower in 2009 compared to 2 010 most likely due to the fact that on average the plants were growing under l onger day length conditions in 2009 compare d to 2010 (Table 4 6 ). Because Coreopsis is a photoperiod sensitive crop, the number of hours of darkness affects the number of days to flower. Coreopsis plants flower under long day (short night) conditions, so the plants flower when night length drops below a certain amount of time. When the population by year (gen otype by environment) interactions were tested, the interactions for all of the traits evaluated were significant (Table 4 2) This indicated

PAGE 111

111 that t he environment affected the phenotype of the individuals and this was different for each year indicating that environment had a strong affect on the traits evaluated. Differences in Pollen Stainability and Seed Production over Two Year s High temperatures and relative humidities at anthesis can have a detrimental effect on Asteraceae pollen vitality (Hoekstra and Bruinsma, 1975). Smith (1976) reported C. leavenworthii and C. tinctoria interspecific hybrids had a n average pollen stainability of 22%. The hy brids were grown in the greenhouse when pollen was collected at anthesis whereas the plants in the current study were grown in the field where many environmental factors in the field are different and can have a detrimental e ffect on pollen viability and can change from season to season. Pollen viability appears to be higher in the greenhouse (personal observation) because the environmental factors, such as temperature and humidity, are easier to control and stabilize. The differences in the environmenta l conditions between the two year s in the current study caused the differences in pollen viability between the two year s including the time of year that the pollen was collected Pollen was collected in late September and early October in 2009 and early J une in 2010. Day length, temperature and humidity can be very different in these months. Although on average the environmental conditions were similar, it is difficult to determine the time frame of changes in the environmental conditions that could have affected pollen viability. The individuals in 2010 had dehiscent pollen and many bees visiting the flowers. In 2009 the pollen on the flower heads o f the individuals did not noticeably dehisce. Because of this difference it is possible that the insect 2010 before flower head collection for the pollen stainability tests. It has been reported that cotton blue can overestimate pollen stainability (Archibald et al. 2005). However, it

PAGE 112

112 seems to be an accurate stain for Coreopsis because stained pollen grains normally appear dark, plump and much larger in size than unstained pollen grain s that are small and shrunken. Although the re was a lower frequency of stainable pollen in 2010, more seed per seed head was p roduced in the field in 2010 compared to 2009 The plants had more flowers and a long er flowering period in 2010 comp a red to 2009 (personal observation). There appeared to be more insects, especially H ymenoptera, in the field during 2010 compared to 2009 This could be due to the differences in crops planted around the Coreopsis plants. The main difference between the two year s is in 2009 there were many crops for insects to extract nectar and transfer pollen. In 2010 there were not many crops around t he Coreopsis plants, so the insect s most likely fed off the C. leavenworthii plants Summary Because COLE x COTI and COTI x COLE produced more seed than COLE and COTI respectively, these two species are full y compatible This indicates that interspecific hybridization could occur under natural conditions and persist over multiple generations. N umber of days to flower had heterotic effects in the C. leavenworthii C. tinctoria interspecific hybrids Plant dry weight of the C. leavenworthii C. tinctoria in terspecific hybrids appeared to be affected by heterosis in the F 1 generation followed by hybrid breakdown in the F 2 generation. P ollen stainability and seed production were reduced in the F 1 and F 2 generations compared to the parental populations possibl y due to chromosome mispairing and dilution for seed production This also indicates that the C. leavenworthii C. tinctoria interspecific hybrids are vigorous and may increase in viability over several generations, which would allow them to persist in lat er generations in a natural setting. Finally the environmental conditions under which the plants a re

PAGE 113

113 grown affected the growth and development of the plants, indicating that growth and development differences can occur between seasons and years resulting in different effects from interspecific hybridization This helped determine the fitness level of interspecific hybrids compared to intraspecific populations of both parental species. Because there was some outbreeding depression found, gene flow betwee n these two species should be prevented. More generations could be created to assess the growth and development as well as backcross generations C. leavenworthii C. tinctoria interspecific h ybrids made in a natural setting should be assessed for growth and development and persistence with natural, pure populations of both species.

PAGE 114

114 Table 4 1 Analysis of variance for the number of seed produced per seed head and seed germination from C leavenworthii and C tinctoria hand pollinations to produce six pop ulations at the Gulf Coast Research and Education Center, Wimauma, FL in 2007 and 2008. Among populations Between year s Plant characteristic Sample size F value Probability F value Probability Seed production 75 46.36 0.0001 NA NA Seed germination 375 9.78 0.0001 0 0.9732

PAGE 115

115 Table 4 2 Analysis of variance for several morphological and reproductive characteristics of six populations of C leavenworthii and C tinctoria grown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2 010. Among populations Between year s Population* year Plant characteristic Sample size F value Probability F value Probability F value Probability Plant height 375/200 z 50.83/28.08 0.0001/0.0001 165.54 0.0001 2.78 0.0184 Plant dry weight 35.78 y 0.0001 0.05 0.82 20 3.99 0.0017 Days to flower 23.93/18.18 0.0001/0.0001 153.68 0.0001 8.73 0.0001 Pollen stainability 9.26/1.11 0.0001/0.3558 37.10 0.0001 5.88 0.0001 Seed production 32.64/6.63 0.0001/0.0001 453.73 0.0001 3.28 0.0069 Seed germination 3.0 8 x 0.0119 NA NA NA NA z Value for 2009 year listed before the slash and 2010 after (2009/ 2010) y 2009 and 2010 years were not significantly different, so values for both years were combined. x Data taken for only 2009 year.

PAGE 116

116 Table 4 3 Comparisons of F 1 an d F 2 average values to mid parent values for each characteristic evaluated for C. leavenworthii and C. tinctoria interspecific hybrids. 2009 2010 Plant c haracteristic Mid parent F 1 F 2 Mid parent F 1 F 2 Plant height (cm) 82.6 87.3 83.5 72.7 71.7 75.1 Pl ant dry weight (kg) 0.14 0.16 0.12 0.12 0.14 0.14 Days to flower 110.8 116.1 112.9 117.6 118.9 120.9 Pollen stainability (%) 18.4 5.4 7.0 4.5 4.8 3.6 Seed production (field) (No.) 446.1 315.5 344.3 614.1 517.2 542.9 Seed germination (field) (%) 49.3 50 .3 47.8 NA NA NA

PAGE 117

117 Table 4 4 Comparison of C. leavenworthii and C. tinctoria intraspecific and interspecific F 1 and F 2 populations. Plant characteristic COLE vs. COTI F 1 F 2 Maternal effect Plant height COTI 25.8 25.9 cm taller Mid parent value Mid parent value Yes Plant dry weight COTI 88.0% heavier Heterosis Mid parent value Yes Days to flower COTI took 7 10 days longer to flower Heterosis in 2009 Heterosis in 2009 Yes 2009/No 2010 Pollen stainability Similar Chromosome m ispairing Chromosome mispairing No Seed production COLE produced 27 60 more seed per 5 seed heads Dilution Dilution No Seed germination COLE ~13% reduced seed germination Mid parent value Mid parent value No

PAGE 118

118 Table 4 5. Temperature and relative humid ity averages throughout the growing year at the Gulf Coast and Education Center, Wimauma, FL in 2009 and 2010. Year AVG Temp ( C ) AVG min ( C ) AVG max ( C ) Relative Humidity (%) 2009 25.9 20.8 32.4 80.29 2010 24.4 18.2 31.3 76.51 Table 4 6. Day length averages throughout the growing year at the Gulf Coast and Education Center, Wimauma, FL in 2009 and 2010. Year Overall (h ) Prior to 1st flower (h) Flowering period (h) 1 mo. Prior to and including flowering period (h) 2009 13.11 12.18 13.69 13.44 2010 1 2.35 11.84 13.59 13.23 Figure 4 1. Crosses to produce populations to compare the interspecific F 1 and F 2 populations to intraspecific populations of C. leavenworthii and C. tinctoria

PAGE 119

119 Figure 4 2 Differences in the amount of seed produced per seed h ead by hand pollinations of six populations of C leavenworthii and C tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2007 and 2008. z Means not followed by same letter are significantly different by

PAGE 120

120 Figure 4 3 Differences in the seed germination rate of seed produced from hand pollinations of six populations of C leavenworthii and C tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. z Me ans not followed by same letter are significantly different by

PAGE 121

121 Figure 4 4 Differences in plant height (cm) of six populations of C leavenworthii and C tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. z Means within year not followed by

PAGE 122

122 Figure 4 5 Differences in plant dry weight (kg) of six populations of C leavenworthii and C tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. z Means not followed by same letter are

PAGE 123

123 Figure 4 6 Differences in the number of days to flower of six populations of C leavenworthii and C tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. z Means within year not

PAGE 124

124 Figure 4 7 Diff erences in the rate of pollen stainability of six populations of C leavenworthii and C tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. z Means within year not followed by same letter are significantly dif

PAGE 125

125 Figure 4 8 Differences in the amount of seed produced from five seed heads in the field of six populations of C leavenworthii and C tinctoria produced at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. z Means within year

PAGE 126

126 Figure 4 9 Differ ences in the rate of seed germination of seed collected in the field of six populations of C leavenwo rthii and C tinctoria produced at the Gulf Coast Research and Education Cent er, Wimauma, FL in 2009 z Means not followed

PAGE 127

127 CHAPTER 5 P OLLEN MEDIATED GENE FLOW F ROM COREOPSIS TINC TORIA TO COREOPSIS LEAVENWORTHII : IDENTIFYING MORPHO LOGICAL MARKERS AND DETERMINING GENE FLO W RATE AND POLLEN TR AVEL DISTANCE Justification When two compatible species are in close proximity and have overlapping flowering periods, interspecific pollen medi ated gene flow can occur (Ellstrand et al. 1999). Many fact ors can affect gene flow rate, such as plant species and density, fertility, flowering synchrony, number of plants flowering at the same time, vector type, nectar supply, topography, wind speed a nd direction, temperature and humidity ( Bateman, 1947; Beckie and Hall, 2008). The lower the density of plants in a population, the greater the swamping effect can be from gene flow and little gene flow can be evolutionarily significant for a population o r species (Antonovics, 1976; Ellstrand, 1992b). A larger pollen source population can cause a greater rate of gene flow (Ellstrand et al. 1989). Because gene flow can be affected by so many variables, the magnitude of gene flow can var y between seasons, years, species and individuals (Ellstrand et al. 1999). The pollination biology of a species is a major determining factor of pollen mediated gene flow. Gene flow typically occurs at greater distances for outcrossing species than inbreeding species. A species can be wind, water, insect or bird pollinated, which can greatly affect the rate and distance of gene flow. The genus Coreopsis is typically pollinated by insects. For insect pollinated species gene flow is greatly dependent on the type of polli nator that forages on a crop because it can be a generalist or a specialist and can determine flight distance due to reward and energy requirements (Schmitt, 1980). Bees have been recorded as moving between near

PAGE 128

128 neighbor plants and move short distances if the rewards are suitable, which could restrict gene flow. Bees typically visit more flower heads per plant and more plants per foraging session compared to other insects. Bees can use pollen and nectar to feed larvae instead of visiting flowers for necta r collecting purposes that result in pollination of the flower heads (Muller and Kuhlmann, 2008). Plants have evolved various mechanisms to alter the nutritional quality and toxicity of the pollen to limit pollen loss as found in Asteraceae In general, insect pollinators forage on plants based on the energy requirements to get to the plants and the energy rewards that the nectar will give the insect (Schmitt, 1980). While different insects have different foraging patterns, insect foraging can have a gr eat affect on gene flow. If the energy and time spent traveling to a population is too costly, the insects will restrict the distance traveled (Osborne et al. 2008). Insects continue feeding on a flower until the nectar of that flower is depleted (Levin 1981). Insect pollinated plants typically receive pollen from many plants (Chapman et al. 2005). If two species are growing within the range of vector travel, a plant could receive pollen from either or both species. In some cases pollination with het erospecific pollen has reduced pollen germination on the stigma, slower pollen tube growth in the style and decreased fertilization of the ovules, which can result in fewer interspecific seed forming, compared to pollination with conspecific pollen. The ra te of pollen mediated gene flow between species can be estimated by growing both species in experimental fields. Two types of experimental arrangements have been commonly used to determine pollen mediated gene flow between species ( Beckie and Hall, 2008). In a continuous experimental design, the pollen trap species is

PAGE 129

129 planted around the centralized pollen donor species. In the discontinuous design, the pollen trap species is several distances apart from t he pollen donor species Spatial arrangements and densities of the pollen donor and trap populations need to be carefully considered as they can have an effect on gene flow rates (Bateman, 1947). Pollen mediated gene flow typically follows a leptokurtic distribution, where most gene flow occurs between p lants that are closer together and decreases at greater distances with most gene flow occurring within 20 m of the source population (Arriola and Ellstrand, 1996; Ellstrand, 1992a; Levin, 1981 ). Animal and insect vectors rarely carry pollen more than 1000 m and generally carry pollen farther distances when plant density is low. In many experimental designs, gene flow rates are underestimated at distances greater than 30 m because it is difficult to accurately assess all of the interactions between the ind ividuals, populations and insects carrying pollen (Levin, 1981). Arriola and Ellstrand (1996) conducted an experiment where interspecific hybridization was assessed by planting Sorghum halepense around a field of S bicolor up to 100 m away which took pla ce at two locations over two years. Pollen mediated g ene flow was greater at distances closer to the pollen source ( S. halepense ) location at about a rate of 10% gene flow at 0 m. At the 5 m distance the gene flow rate decreased to about 3%. Gene flow d id occur at the 100 m distance at a rate as high as 2% at one location one year and was highly variable depending on the year and location of the study. S. bicolor is an outcrossing species, whereas S. halepanse is typically a self pollinated species but will outcross. S. bicolor and S. halepense are both wind pollinated and were reported to cross pollinate spontaneously in nature and

PAGE 130

130 interspecific hybridization occurred successfully by hand pollinations. Both species of Sorghum grow in the same areas a nd have overlapping flowering periods. As in the study with Sorghum C oreopsis leavenworthii and C. tinctoria have both b een reported as growing in similar locations, where both species have been reported as growing in Florida. C. leavenworthii has been r eported in most counties in Florida and two counties in Alabama and is considered endemic to Florida. C. tinctoria has been reported in most of the United States, including some counties in Florida where C. leavenworthii also exists Because it has been reported that C. leavenworthii and C. tinctoria ar e cross compatible by Smith (1976) and confirmed by the synthetic F 1 and F 2 populations produced in Chapter 4 it is possible that these two species could cross pollinate when grown at a certain distance fr om each other. The possible distance and frequency of gene flow is not known. The distance of pollen mediated gene flow that c ould occur between these two species needs to be assessed to help prevent natural interspecific hybridization. Previous results from Chapter 4 indicate that the F 1 hybrids are relatively vigo rous and are reproductively viable, although there was a decrease in v i ability in the interspecific F 1 and F 2 hybrids. These results indicate that the interspecific hybrids could compete and intercross with intraspecific populations. The objectives of this study were to determine the inheritance of morphological marker s that could potentially be used to reliably detect pollen mediated gene flow, to assess the effects of planting distance on p ollen mediated gene flow from C. tinctoria to C. leavenworthii and to identify insect pollinators that may be involved with pollen mediated gene flow.

PAGE 131

131 Materials and Methods Seed Source The C. leavenworthii (COLE) plants used in this study were from seed co llected by Nancy Bissett from a natural C. leavenworthii population at the Reedy Creek Mitigation Bank in Polk County, FL. The C. tinctoria (COTI) plants were produced from seed that w as purchased from Wildseed Farms in Fredericksburg, TX. Identifying Mor phological Differences between C. leavenworthii and C. tinctoria Plants of the two species were grown side by side and closely examined during the course of study t o find morp hological characteristics that we re present in one species but absent in the othe r species. About one hundred plants were assessed for morphological differences in the flowers, leaves, stems and seed to determine the inheritance pattern of these traits for possible use as a morphological marker to detect pollen mediated gene flow. Inh eritance of Morphological Traits Crosses and p opulation s Five types of crosses were made to investigate the inheritance of three morphological traits (Figure 5 1) in C. leavenworthii and C. tinctoria including intraspecific crosses for each species, recip rocal interspecific crosses between the two species, selfing and s ibling of interspecific hybrids and testcrosses. Hand pollinations were carried out on plants at the University of Florida, Gulf Coast Research and Education Center, Wimauma, FL. Parental plants for all cross types were grown in 3 2 L plastic containers filled with Vergrow verlite container mix A (Tampa, FL) mixed with 8.6 g of 15N 3.9P 10K ( 5 6 mo. Osmocote, The Scotts Co., LLC, Marysville, OH) per container in the greenhouse (29.4/23.9C) T he only exception was the parental plants

PAGE 132

132 for the testcross pollinations were grown in 2 .1 L containers and the soil media was mixed with 5.6 g of 15N 3.9P 10K per container. For all types of crosses, the flower heads were bagged at the bud stage. Po llen was collected by tapping the whole flower head in a plastic container, where the pollen was deposited. To pollinate flower heads of the maternal plant, they were dipped into the container and rubbed against the available pollen in the container. The pollinated flower head was rebagged and the bag was kept on until the seed head was matured and harvested. Seed from all crosses w ere sowed on the surface of Vergrow verlite container mix A in 532 mL containers in the greenhouse (29.4/23.9C) and kept und er mist irrigation S eedling s were transplanted into 14 8 mL cell Speedling flats (32 cell flats) (Sun City, FL) of Vergrow verlite container mix A mixed with 0.4 g of 15N 3.9P 10K per cell The only exception was an initial group of interspecific seed lin gs w ere transplanted in 3 1 mL cell S peedling flats (128 cell flat s) Intraspecific crosses and populations Synthetic C. leavenworthii (COLE) intraspecific crosses were made between 27 July and 5 August 2007 and C tinctoria (COTI) intraspecific crosses we re made between 7 and 10 August 2007 Seed w ere sowed on 11 January 2008 Seedlings were transplanted on 2 February and fertilized with 300 ppm of 15N 13P 12.5K (Plantex, Plant Products CO. Ltd., Brampton Ontario, Canada) Interspecific crosses and F 1 po pulations Four COLE plants (L1 to L4) and four COTI plants (T1 to T4) were randomly chosen and cross pollinated in 32 combinations between 14 June and 30 July 2007 Sixteen of the crosses ha d the COLE plants as the maternal parent (COLE x COTI) and sixte en ha d COTI as the maternal parent (COTI x

PAGE 133

133 COLE). A n initial portion of seed from eight of the populations w as sow ed on 23 October 2007 to begin to assess the inheritance patterns of the traits of the F 1 interspecific hybrids Seedlings were transplanted on 15 November and fertilized with 20 N 2.2P 25K The field was a mix of Myakka, Haplaquents, and St. Johns sandy soil. The beds were raised 20 cm high, 61 cm wide at the top and 71 cm wide at t he base The beds were fumigated with 67% methyl bromide: 33% chloropicrin at 197 kg ha 1 The plants were transplanted in the field on 30 and 31 January 2008 and spaced 30.5 cm apart within plots and 1 2 2 cm in rows. The plants were fertilized with a to tal of 153 kg h a 1 of nitrogen and potassium throughout the season through drip irrigation. Phosphorus was added throughout the season as needed. To obtain additional trait inheritance data on the remaining interspecific F 1 population s and those already assessed more interspecific s eed were sowed on 27 June 2008 Seedlings were transplanted on 14 July and fertilized with 300 ppm of 20 N 8.7P 16.7K Marysville, OH) F 1 selfing and sibling crosses and F 2 populations Two type s of F 2 populations were produced by 1) selfing ( seve n populations ) or 2) sibling F 1 hybrids ( five populations ) T he seed from selfing and sibling crosses w ere sowed i n August 2008 All seedlings from both types of crosses were transplanted on 21 August and fertilized with 300 ppm of 20N 8.7P 16.7K. Testcross es and populations The testcross populations were produced from 1 July to 6 August 2007, where eight populations were produced. To produce these populations, COLE x COTI interspecific hybrids were b ackcrossed with COLE The

PAGE 134

134 seed w ere sowed on 27 October. Seedlings were transplanted on 18 November and fertilized with 300 ppm of 20N 8.7P 16.7K. Seed production a nalysis Prior to sowing the seed, the number of seed from each cross was counted. Seed p roduction of each cross was stati sti cally analyzed using the ANOVA using PROC GLM Progeny e valuation Trichomes. When F 1 progeny were about 80 days old, they w ere assessed for the presence or absence of trichomes on the leaf petiole. Seed wings. The seed from the intraspecific COLE intraspecific COTI F 1 hybrid and the F 2 sibling crosses were evaluated for the presence or absence of the wingedness of the achene. Maroon spot. The presence or absence of the maroon s pot was recorded for each plant of both intraspecific, F 1 F 2 and testcross populations The size of the maroon spot was eval uated by measur ing three ray petals f rom three flower heads (cm) for each plant of the 32 F 1 hybrid populations. T he measurements from the t hree flowers per plant were averaged, which were then averaged for each population. Trait segregation a nalysis The inheritance patterns were statistically tested using the chi square goodness of fit test using a one gene model Assessing Pollen m ediated Gene Flow Setting up pollen source and trapping plots C. leavenworthii plants were planted in the field in plots at 1.5, 3.0, 7. 6, 15.2, 30.5, 45.7, 61.0, 76.2 and 91.4 m away from C. tinctoria plants during 2007 and 2008 at the University of Flor ida, Gulf Coast Research and Educa tion Center, Wimauma, FL ( Figure

PAGE 135

135 5 2 ). The plots were planted in a disc ontinuous design, where C. leavenworthii plots were in the same bed but plots were not in contact with each other. For both year s, there were three b locks with 16 plants per plot and t he blocks were planted at least 91.4 m away from each other. Once the progeny of each plot was grown out, data was recorded on the presence or absence of the maroon spot for each plant as they bloomed. About 5,700 and 5 ,600 plants were count ed for 2007 and 2008, respectively. 2007. C. leavenworthii and C. tinctoria seed w ere sowed on 16 January 2007 into 532 mL containers on the surface of Vergrow ver l ite container mix A media in the greenhouse and put under mist. The s eedlings were transplanted about three weeks later in 80 m L cells (Landmark Plastic Corp., Akron, OH) of Vergrow ver l ite container mix A media The field was a mix of Myakka, Haplaquents and St. Johns sandy soil. The beds were raised 25 cm high, 71 cm wi de at the top and 81 cm wide at the base. The beds were fumigated with 67% methyl bromide: 33% chloropicrin at 197 kg h a 1 Plants were spaced 30.5 cm apart within plots that were 92 cm apart in rows with 152 cm between rows. The plants were transplante d in the field on 20 April and e ach bed contained two rows of plants. The plants were fertilized with a total of 169 kg ha 1 of nitrogen, 37 kg ha 1 of phosphorus and 210 kg ha 1 of potassium through drip irrigation. 2008. C. leavenworthii and C. tinctor ia seed w ere sowed on 9 January 2008 and transplanted into 80 m L cells of Vergrow ver lite container mix A media on 13 February. The plants were transplanted into the field on 26 March. All other growing conditions were the same as the previous year C oll ecting seed from pollen trapping plots Seed heads were harvested from the plants grown in 2007 on 31 May to 4 June, 5 July and 11 to 14 July 2007 and from the plants grown in 2008 on 28 May, 25 June and

PAGE 136

136 17 and 18 July 2008 for harvests 1, 2 and 3, respecti vely Plants were in full bloom around the end of April for both year s and t en seed heads were harvested per plant for all harvests Detecting gene flow events in C. leavenworthii progeny 2007. Seed collected from the second and third harvests in 2007 wer e sow ed on 7 September 2007 in a 2 .1 L plastic container of Fafard soil mix 1P (Anderson, SC) at the University of Florida, Gainesville, FL and placed under mist. The seedlings were transplanted on 12 October into 80 m L c ell trays of Fafard soil mix 2P (A nderson, SC) and fertilized with 150 ppm of 20 N 8.7P 16.7K Resets were completed on 19 and 25 October. Seed collected from the sec ond and third harvests in 2007 were sowed on 10 and 11 January 2008 at the University of Florida, Gulf Coast Research and E ducation Center, Wimauma, FL. The seedlings were transplanted on 15 and 18 February into 3 1 m L Speedling trays of Vergrow verlite container mix A The beds were raised 20 cm high, 61 cm wide at the top and 71 cm wide at the base. The plants were spaced 30.5 cm apart within plots and 122 cm in rows. The beds were fumigated with 67% methyl bromide: 33% chloropicrin at 197 kg ha 1 The plants were transplanted in the field on 2 April. The plants were fertilized with a total of 153 kg ha 1 of nitroge n and potassium through drip irr igation. Phosphorus was added as needed. The data for the plants grown in Gainesville, FL and Wimauma, FL was combined for 2007. 2008. Seed collected from the first and second harvests in 2008 w ere sowed on 26 June 2008 in a 2 .1 L plastic container of on the surface of Vergrow verlite container mix A and put under mist. The seedlings were transplanted into 3 1 m L cell S peedling flat s of Vergrow verlite container mix A on 6 8 August and were fertilized with 300 ppm of 20N 8.7P 16. 7K. Resets were done on 12, 13, 14, 20 and 27 August. The plants

PAGE 137

137 were planted in the field on 12 and 15 September. All other growing conditions were the same as those at the Gulf Coast Research and Education Center in 2007 Statistical analysis For each year the percentage of plants with the maroon spot present out of the total number of plants evaluated was calculated for each block and then combined. The gene flow rate for each distance and year was fit to a logistic regression curve using PROC GENMO D for the Logistics procedure in SAS (SAS, 1997). The L ogistics procedure can be used to analyze binary response, where an individual can take one of two possible values. In the case of the current study it was used to analyze the maroon spot, where it was either present or absent on the ray flower. Data w ere fit to the following equation: p=1/(1+e 3.2387+0.1673D 0.00124D^ 2 ) Where p= the probability at a specific distance that gene flow will occur, e=the inverse of the natural logarithm and D=the distanc e at which gene flow was being measured. Results Crossa bility between C. leavenworthii and C. tinctoria On average 46.4 seed per seed head were produced from five C. leavenworthii (COLE) maternal in dividuals that were crossed usi ng the pollen from 40 C. le avenworthii individuals (Table 5 1) On average 18.5 seed per seed head were produced from five C. tinctoria (COTI) maternal individuals that were cross ed usi ng the pollen from 35 individuals. The COLE intraspecific crosses produced 1 50.8% more seed per seed head than the COTI intraspecific crosses which is congruent with the

PAGE 138

138 results in Chapter 4 with C. leavenworthii having more disk flowers per flower head than C. tinctoria On average COLE x COTI interspecific crosses produced 53.9 seed per seed head which was similar to that found in Chapter 4 where 61.9 seed per seed head were produced COTI x COLE interspecific crosses produced on average 27.7 seed per seed head which was similar to that found in Chapter 4 where 25.0 seed per seed head were pr oduced, (Table 5 1) The COLE x COTI crosses produced 16.2% more seed than the COLE intraspecific crosses and the COTI x COLE crosses produced 49.7% more seed than the COTI intraspecific crosses. Because b oth types of F 1 interspecific crosses produced mo re seed than the intraspecific crosses in the previous (Chapter 4) and current studies these two species appear to be fully compatible. On average the F 1 self crosses produced 2.8 seed per seed head and the F 1 sibling crosses produced 17.8 seed per seed h ead (Table 5 1) which wa s a 535.7% increase for the F 1 sibling crosses. The F 1 self crosses produced 5.2% and 10.1% of the seed per seed head that the COLE x COTI and COTI x COLE crosses produced, respectively. The F 1 sibling crosses produced 33.0% and 64.3% of the seed per seed head that the COLE x COTI and COTI x COLE crosses produced, respectively. On average the testcrosses produced 18.4 seed per seed head, whic h wa s a 3.4% increase from the F 1 sibling and a 557.1% increase from the F 1 selfing cross es in seed per seed heads. The testcrosses produced 34.1% and 66.4% of the seed per seed head that the COLE x COTI and COTI x COLE crosses produced, respectively. Because there was a reduction in seed produced per seed head for both types of F 1 crosses a nd the testcrosses, hybrid breakdown appears to be occurring.

PAGE 139

139 Morphological Differences between C. leavenworthii and C. tinctoria Three morphological character differences were ev aluated, which were trichomes, s eed wingedness and the maroon spot on ray flo wers (Figure 5 1) Trichomes were found to be ab sent in 86 C. leavenworthii p lants and present in 35 C. tinctoria plants. Seed wingedness was found to be present in 5722 C. leavenworthii seed but absent in 870 C. tinctoria seed The maroon spot was foun d to be absent in 86 C. leavenworthii p lants and present in 35 C. tinctoria plants. Intra and interspecific crosses were made between the two species to evaluate the inheritance patterns for these characteristics. Expression and Inheritance of Trichomes Trichomes were present o n a ll F 1 hybrid plants (about 1 000 plants evaluated) COLE x COTI and COTI x COLE in a 1:0 genetic ratio indicating that trichomes we re dominantly inherited from C. tinctoria However, the size of trichomes and number varied amon g F 1 plants but the size and number were consistent within each plant The trichomes were 1 or 2 per leaf petiole to as many as 20 per leaf petiole and could be long (up to 5 mm) or short (about 2 mm) in size. All combinations were found on the F 1 hybrid plants. When interspecific hybrids were assessed for trichomes in a previous study at the North Florida Research Education Center (Quincy, FL), 38 interspecific hybrids did not have any trichomes visible to the naked eye, 21 had few, visible trichomes an d 21 had numerous, visible trichomes (personal communication). This pattern of segregation suggests that the presence of trichomes on C. tinctoria leaf petioles might be controlled by more than one locus and the loci were not in homozygosity. Because of this nature, trichomes would not be the ideal morphological marker for tracking pollen mediated gene flow from C. tinctoria to C. leavenworthii

PAGE 140

140 Expression and Inheritance of Maroon Spots Presence of maroon spots I ntraspecific crosses from five COLE seed p arents using 40 COLE pollen plants resulted in 86 progeny and th ey all lack ed the maroon spot. I ntraspecific crosses from five COTI seed parents were made using 35 COTI pollen plants Thirty five progeny were produced that all expressed the maroon spot on the ray flowers ( Table 5 2 ) Sixteen COTI x COLE crosses using four COLE and four COTI plants produced 493 progeny and all expressed the maroon spot, indicating that the presence of the maroon spot is dominant o ver no spo t. The maroon spot was express ed in an additional 525 p rogeny produced from reciprocal COLE x COTI crosses, suggesting that the trait is under nuclear gene control Selfing F 1 plants produced only a limited number of F 2 progeny (32 individuals) Of them, 25 expressed the maroon spot a nd seven lacked the maroon spot, thus the presence of the maroon spots segregat ed into a 3:1 ratio 2 = 0 1.33 and P =0.25 1.00) as expected for a single dominant gene controlled trait S ibling F 1 plants produced more progeny ( 306 ). Among these F 2 progeny 227 expressed the maroon spot and 79 plants did not. Again the trait segregat ed into a 3:1 ratio 2 =0.005 1.45 and P =0.23 0.94) When F 1 interspecific (COLE x COTI) individual s were back crossed with COLE, 163 out of 370 progeny expressed the maroon spot and 207 did not The segregation 2 =0.12 2.60 and P =0.11 0.73). These results show that the presence of the maroon spot wa s controlled by a single dominant homozygous nuclear gene in C tinctoria and was easily identified in the interspecific hybrids making it a reliable morphological marker to identify pollen mediated gene flow from C. tinctoria to C. leavenworthii

PAGE 141

141 Size of maroon spots This size of the maroon spot for C. tinctoria ind ividual s ( T1 T4 ) was larger for T2 and T3 and smaller for T1 and T4. These individuals were used as the maternal parents in interspecific crosses (COTI x COLE) and they produced 384 progeny. When T1, T2, T3 and T4 were used as the maternal parent, the av erage maroon spot size of the progeny was 0.43 (120 individuals) 0.50 (103 individuals) 0.55 (80 individuals) and 0.21 (81 individuals) cm respectively (Figure 5 3) When C. tinctoria was used as the maternal parent and crossed with any of the four C. leavenworthii parent s the size of the maroon spot of the progeny was largest for T3, T2, T1 and T4 respectively To determine if there was a maternal effect on the size of the maroon spot re ciprocal crosses were made with C. leavenworthii (L1 L4) as the maternal parent and C. tinctoria as th e paternal parent (COLE x COTI), and these crosses produced 35 5 progeny. When T1, T2, T3 and T4 were used as the paternal parent, the average maroon spot size of the progeny was 0.41 (102 individuals) 0.49 (90 indiv iduals) 0.63 (89 individuals) and 0.19 (74 individuals) cm respectively (Figure 5 3) When C. tinctoria was used as the paternal parent, independent of the C. leavenworthii maternal parent, the size of the maroon spot of the progeny was largest for T3 T2, T1 and T4 respectively Because the same pattern was observed when C. tinctoria was used as the maternal and paternal parent, there is not a maternal or reciprocal e ffect and the maroon spot size wa s solely dependent on the maroon spot size of the C. tinctoria parent. Expression and Inheritance of Seed Wings All COLE intraspecific crosses produced winged seed (5722 seed evaluated) and all COTI intraspecific crosses produced non winged seed (807 seed evaluated) ( Table

PAGE 142

142 5 3 ). When COLE plants were used as the seed parents and crossed with COTI pollen (COLE x COTI), all of the seed (7433 ) were winged. When COTI plants were used as the seed parents and crossed with COLE pollen (COTI x COLE), all seed were non winged for 26 of the crosses. However, the tr ait segregated in four other COTI x COLE crosses, 52 winged seed: 89 non winged seed. When interspecific F 1 plants (COLE x COTI) were self pollinated, two c rosses produced winged seed (23 ), three crosses produced winged seed (48) and non winged seed (375) and one cross produced only non winged seed (100). The segregation of seed wingedness wa s possibly due to the phase of seed development that the C. tinctoria plants were undergoing when the crosses were completed or the seed source (Parker, 1973). C. ti nctoria has a winged and wingless phase and segregation in seed wingedness in C. leavenworthii and C. tinctoria interspecific hybrids was found as well indicating that the inheritance of seed wingedness wa s incompletely dominant because it was not comple tely dependent on the winged phenotype of the C. tinctoria parent Pollen mediated Gene Flow from C. tinctoria to C. leavenworthii For each pollen trap plot 581 681 progeny were evaluated in 2007 and another 593 654 progeny were evaluated in 2008 to ident ify the progeny from the C. leavenworthii individu als expressing C. tinctoria above described inheritance study, these individuals would indicate poll en mediated gene flow from C. tinctoria into C. leavenworthii The iden tified individuals were then used to calculate gene flow rate s There was no significant difference in gene flow rate among block s in 2007 but in 2008 the gene flow rate in B lock 3 seemed to be much lower than in B locks 1 and 2 and these blocks w ere thus evaluated separately. For both year s the 1.5 m plot had the

PAGE 143

143 highest rate of gene flow, 3.89% (25 out of 646 individuals ) in 2007 (averaged from three blocks) 4.24% (17 out of 401 individuals) in 2008 for B locks 1 and 2 and 0.93% (2 out of 213 individual s) in 2008 for B lock 3 (Figure 5 4) In 2007 and 2008 the farthest distance that gene flow occurred was 61.0 m (1 out of 649 individual s ) and 15.2 m (1 out of 430 individual s ) respe ctively. At these distances, the observed gene flow rates were 0.13% and 0.23%, respectively. ANOVA results indicated that t he gene flow rates were not significantly different between the two year s (F= 1.85 P =0. 1829 ) so the gene flow rate data from the two years were pooled and a regression curve was fit to the observed data using the Logistics procedure (Table 5 4 ). The curve sh ows that the closer the plants we re grown, the more likely gene flow can occur where p=1/(1+e 3.2387+0.1673D 0.00124D^ 2 ) (Figure 5 5 ). The equation was used to estimate the probability of gene flow e vents at a particular distance. Based on th is equation separating the species by 28 m or 60 m would lower the gene flow rate to about 0.10% or 0.01% respectively Disc ussion Crossa bility of C. leavenworthii and C. tinctoria Based on the amount of seed p roduced from the interspecific crosses, the two species we re compatible and could produce a large number of seed when hand pollinated which wa s congruent with the results in Chapter 4 and a previous report by Smith (1976). The amount of seed per seed hea d produced by the COLE x COTI crosses (53.9 seed per seed head) was similar to that by COLE intraspecific crosses ( 46.4 seed per seed head ) The COTI x COLE crosses produced 27.7 seed per seed head similar to th at of COTI intraspecific crosses (18.5 seed per seed head) The amount of seed produced per seed head by the COLE x COTI and COTI x COLE

PAGE 144

144 crosses in the current study was similar to the results from the previous study (61.9 seed per seed head for COLE x COTI crosses and 25.0 seed per seed head for COTI x COLE crosses in Chapter 4) In both studies, more seed were produced per seed head when COLE was the maternal parent than when COTI w as the maternal parent. T he F 1 sibling and test cross es pro duced an amount of seed per see d head comparable to the C OTI intraspecific po llin ations. The F 1 sibling crosses produced an average of 17.8 seed per seed head and the test crosses produced an average of 18.4 seed per seed head, which we re lower than the mid parent value ( 32.5 seed per seed head ) Fewer seed wer e produced from the F 1 sibling crosses compared to the COLE x COTI and COTI x COLE crosses which wa s likely due to outbreeding depression T he F 1 individuals continued to produce seed when cross pollinated with other F 1 plants The F 1 sibling crosses pro duced 1 7. 8 seed per seed head, which was similar to that found in the previous study (Chapter 4) where 35.3 seed per seed head were produced. This suggests that the interspecific hybrids and later generations w ould continue to outcross, which w ould allow the interspecific plants to continue spreading its pollen to eith er parental species or interspecific plants Outcrossing s pecies can be at a greater risk of gene flow and possible extinction (Ellstrand, 1992 b ). The F 1 self crosses produced only about th ree seed per seed head, much fewer than any other type of cross in this study (Table 5 1 ), illustrating strong self incompat ibility in these C leavenworthii C. tinctoria hybrids The genus Coreopsis as well as Asteraceae, is k nown to be sporophytically self incompatible (Brewbaker, 1957). A self incompatibility study with C. tinctoria had suggested that there is a minimum of seven alleles controlling sporophytic self incompatibility (Sharma, 1971). Two other

PAGE 145

145 Coreopsis species, C. californica and C. big elovii in this same study were found to have at least five alleles controlling the sporophytic self incompatibility mechanism. The sel f incompatibility of the F 1 self crosses show ed that these alleles we re inherited in interspecific crosses and w ould pro mote outcrossing, which c ould result in more outcrossing between interspecific hybrids and backcrossing with C. leavenworthii and C. tinctoria Trichomes and Seed Wings w e re not Suitable for Detecting Gene Flow Events Presence of trichomes in C. tinctoria appeared to be a dominant trait, as in the F 1 interspecific hybrids. But the expression of this trait varied considerably among progeny. Further, it appeared to be segregating in the F 1 progeny of C. tinctoria and C. leavenworthii (Jeff Norcini, personal communication) but could also be due to the lack of visibility without the use of magnification These results suggest that the trait may be controlled by two or more loci and some of the loci may be heterozygous. Thus, this trait may allow detection o f some pollen mediate d gene flow from C. tinctoria in t o C. leavenworthii but it is very likely that the gene flow rates based on such a marker will be underestimated and inaccurate. The wingedness of the achene was not a reliable marker because it did not follow the inheritance pattern of a homozygous dominantly controlled gene in the F 1 or F 2 ge nerations. Smith (1983) reported C. tinctoria seed ranging in seed wingedness from absent to fully winged. When C. tinctoria and C. cardaminefolia were cross pol linated, all of the F 1 progeny produced seed with partial wings and the F 2 p rogeny produced seed ranging from non winged to fully winged (Smith and Parker, 1971). It was concluded that seed wingedness was dominantly controlled by one or two loci. Parker (1973) reported that C. tinctoria had a winged and wingless phase and differe d in

PAGE 146

146 wingedness depending on the source of the seed which determined the wingedness of the seed produced. The wingedness of C. tinctoria was found to be incompletely dominant by Parker (1973) wh ich wa s further supported when crossed with C. leavenworthii because some seed were winged and others non winged. Because of the range in seed wingedness of C. tinctoria reported, different ecotypes of this species may have different gen otypes for seed wingedness, making it an unreliable morphological marker for gene flow detection. Presence of Maroon Spots i s Controlled by a Single Dominant Gene and i s an Applicable, Reliable Morphological Marker for Detecting Gene Flow Events T he maroon spot of C. tinctoria segregated as a single, dominant gene in all of the population s examined Additionally, the character was observed in all 1 018 plants used for inheritance and gene flow studies. Should the next C. tinctoria individual appear non spo tted, the frequency of this recessive phenotype would be 0.000981 ( one out of 1 019 F 1 individuals) the recessive allele (non spotted) would occur at a frequency of 0.031 327 and the dominant allele (spotted) would be pr esent at a frequency of 0.968673 T hus in case this scen ario does occur, the maroon spot will allow the detection of nearly 97% of pollen mediate gene flow from C. tinctoria into C. leavenworthii Previously, Smith and Parker (1971) crossed C. tinctoria with C. cardaminefolia (non spotted) and observed that the maroon spot was expressed in a ll F 1 progeny and segregated among the F 2 progeny. T wo individuals lacked the maroon spot and 25 individuals expressed the maroon spot The authors were not able to d etermine the inheritance of the maro on spot based on th e available data It seems that t he reported segregation does not fit a 3:1 ratio expected for a single dominant gene controlled trait

PAGE 147

147 2 =4.46, P =0.03). It will remain to be determined whether this deviation is due to small size of th e F 2 population or the differences between C. cardaminefolia and C. leavenworthii Size of Maroon Spot i s Controlled by Additional Genes Based on the consistency of the results of the size of the maroon spot of the C. leavenworthii C. tinctoria hybrids, th e size of the maroon spot appear ed to be dependent on the size of the spot of the C. tinctoria parent and to be controlled by multiple alleles with no reciprocal effects (Figure 5 3 ). Smith and Parker (1971) ob served variation in the maroon spot size amon g progeny of controlled crosses between C. tinctoria and C. cardaminefolia and hypothesized that the spot size was controlled by two or more loci with additive effects Pollen mediated Gene Flow Rate from C. tinctoria to C. leavenworthii wa s Lower than Exp ected Because of the high level of crossability found between C. leavenworthii and C. tinctoria observed in hand pollinations in the greenhouse it was expected that there would be a relative ly high frequency of pollen mediated gene flow from C. tinctoria to C. leavenworthii in the field as insect s foraged on Coreopsis flowers Plants of both species produced large numbers of f lowers in the field, and their flowering time overlap ped throughout the year thus there should be plenty of opportunity for C. tin ctoria to cross pollinate C. leavenworthii Nevertheless, the highest rate of pollen mediated gene flow from C. tinctoria into C. leavenworthii observed in two years was 4.24% when plants of the two species were planted 1.5 m apart. No gene flow events w ere detected when plants of the two species were planted 76.2 m and 91.4 m apart.

PAGE 148

148 Arias and Rieseberg (1994) reported that crop and weed hybridization of sunflower occurred at a frequency of gene flow of 0.27 at 3 m and 0.02 at 1 000 m. Although these freq uencies are much lower than those found in the current study, they can still be biologically significant. From a stat istical standpoint the amount of gene flow that could occur in sunflower and in the current study with Coreopsis there wa s a low chance t hat an insect would carry pollen from one species to another. From a biological sta ndpoint, a few pollinations between the species could result in a large amount of interspecific seed produced (Grant, 1981) Based on the interspecific seed produced by ha nd in this study, about 30 50 seed per seed head could be produced from each C. leavenworthii C. tinctoria interspecific pollination, which could turn into many interspecific plants in a natural setting. T he vigor of the interspecific hybrids and later ge nerat ions should be assessed in a mixed population in a natural setting Buffer Zones can Protect C. leavenworthii from Genetic Contamination The observed gene flow from C. tinctoria to C. leavenworthii follows a leptokurtic curve where the great est frequ ency of gene flow occurred at the 1.5 m plot and the gene flow rate consecutively decreased for each plot afterwards The farthest distance that gene flow occurred was 61.0 m with no gene flow occurring at the 76.2 and 91.4 m plots. T he highest frequency of gene flow occurs when the two species are plante d closest together typically wi thin a few meters of the pollen donor population (Beckie and Hall, 2008; Ellstrand, 1992b ; Levin, 1981). By growing the species farther apart, gene flow will be less likely Even though most contamination can occur at closer distances, gene flow occurring at farther distances can affect a population also. It only takes one individual per generation to be evolutionarily important. Therefore, the tail end of the leptokurtic tail should not be disregarded (Ellstrand, 1992b). Most gene flow occurred

PAGE 149

149 from the 1.5 15.2 m plots in both year s, but some gene flow did occur at the 30.5 61.0 m plots in 2007. Based on these results and the regression curve fit for the data, these tw o species should be separated by about 60 m to minimize the possibility of gene flow. Arias and Rieseberg (1994) reported that crop and weed hybridization of sunflower occurred from 3 1 000 m, suggesting that gene flow can occur at great distances and shou ld be considered when creating buffer zones between two species. A trap crop may be also utilized to protect C. leavenworthii This entails growing a narrow strip of plants at the edge of the pollen trap field and remov ing these plants before seed harvest A pollen trap crop can be the same species as the pollen receptor crop. The rows of the pollen trap crop the pollen load from the other species thus protecting the natural or seed production population s The Need to Consider Seed Dispersal on Gene Flow in C oreopsis In addition to pollen mediated gene flow, s eed dispersal may be another important pathway leading to gene flow in Coreopsis This seems especially true consider ing that a number of Coreopsis s pecies, including C. l eavenworthii have winged or spiked achene s that can be carried far in gusts of wind or storms (Tadesse et al. 1995). Thus, the possibility of l ong range seed dispersal exists in C. leavenworthii and other Coreopsis species ( Cain et al. 2000; Levin, 198 1; Rukini, 2008 ) As shown in this and previous studies ( Chapter 4 and Smith, 1976 1983 ), when C. leavenworthii and C. tinctoria grow in close proximity they can cross pollinate and produce viable hybrids. These hybrids can become a rich source of poll en and seed for further gene flow, interspecific hybridization, genetic swamping and gene pool contamination.

PAGE 150

150 Hymenoptera on Coreopsis Hymenoptera has been reported as pollinating members of the Asteraceae, but specific in sects have not been reported as po llinating Coreopsis ( Arias and Rieseberg, 1994; Levin and Kerster, 1969; Muller and Kuhlmann, 2008). Insects were collected from C. tinctoria and C. leavenworthii plants o n 9 June and 25 September 2008 and identified by the Florida Department of Agricultu re and Consumer Services, Division of Plant Industry in Gainesville, FL. For both collections, all of the insects collected were Hymenoptera (Table 5 5 ). Six H ymenoptera species were found visiting C. leavenworthii and four species visit i ng C. tinctoria Three species Halictus poeyi Scolia nobilitata and Philanthus ventilabris were visiting both species of Coreopsis Halictus poeyi and Scolia nobilatata are known to be plant pollinators (Triplehorn and Johnson, 2005). It remains to be known whether the other Hymenoptera species pollinate Coreopsis plants. It has been reported that some Hymenoptera species collect po llen to feed larvae or look for other insects for f eeding Th ese kind s of insects are generally not hairy for carrying pollen. Some me mbers of Asteraceae have been shown to contain toxic compounds, have low protein content, lack essential nutrients and have pollen grains that are difficult to degrade by insects (Muller and Kuhlmann, 2008). This has caused some members of Hymenoptera to become specialized to members of Asteraceae that are able to overcome these barriers to use pollen for food In general larger insects travel farther distances than smaller insects to forage (Greenleaf et al. 2007). Pollinators respond to different stimu li in the environment and environmental conditions can alter foraging patterns (Levin and Kerster, 1969). It has been documented that o nce insect s find a food source, they remain loyal to it and forage over short distances. When the nectar is depleted from one population,

PAGE 151

151 the insects, especially bee s, will find the closest population of the same species to forage. Because the nectar rewards of Asteraceae are small, plants within a population must be close in proximity and pollen dispersal w ill usually remain within that population because bees can probe many florets quickly and maintain a fairly high rate of energy intake (Price, 1997 ; Schmitt, 1980 ). In the field the insect s seemed to work flowers in a continuous manner (personal observati on). They visited a flower and then found another flower in close proximity to visit next. Usually the insects depending on the species, did not move to a flower that wa s located several meters away. The types of crops surrounding the Coreopsis plants c ould explain some of the differences in frequency of gene flow between these two Coreopsis species T he Coreopsis plants were surrounded by many different vegetable crops and ornamentals in 2007 and they did not seem to affect the rate of gene flow betwe en each block. In 2008 the Coreopsis plants were surrounded mostly by ornamentals (Blocks 1 and 2), such as caladium ( Caladium hortulanum ) and Lantana camera The plants of Block 3 were next to a tomato field ( Solanum lycopersicum ) where there w as a m uch lower rate of gene flow in 2008 Because the gene flow rates were similar between 2007 and Blocks 1 and 2 of 2008, it does not appear that the crops surrounding the Coreopsis plants affected the gene flow rates, but because the gene flow rates were mu ch lower in Block 3 of 2008, it is possible that the insects preferred tomato over Coreopsis The greater diversity of crops may have encouraged the greater distances that gene flow occurred in 2007. Summary Hand pollination results showed that C. leavenw orthii and C. tinctor ia we re highly compatible and their interspecific crosses produced similar numbers of seed as their

PAGE 152

152 intraspecific crosses did Inheritance studies revealed that t he maroon spot in C. tinctoria wa s controlled by a single dominant gene and C. tinctoria was homozygous at this locus, thus this spot could serve as a reliable morphological marker to reveal gene flow events from C. tinctoria to C. leavenworthii and quantify the rate of such directional gene flow Following a discontinuous de sign, field gene flow studies were conducted in replicated blocks and repeated over two years. Results showed that pollen mediated gene flow occurred at the highest frequency when the two species were planted 1.5 m apart and continuously occurred as far a s the two species were 61.0 m apart. Natural gene flow from C. tinctoria to C. leavenworthii follow ed a le ptokurtic curve. If validated, this leptokurtic equation will provide a very valuable model for estimating the potential gene flow from C. tinctoria to C. leavenworthii and planning buffer zones for minimizing such gene flow. Hymenoptera was identified as the main insect pollinators of Coreopsis

PAGE 153

153 Table 5 1 Number of seed collected per seed head for each cross type made at the Gulf Coast Research an d Education Center, Wimauma, FL. Cross type Seed parent Pollen parent Seed head examined (no.) Seed examined (no.) Mean Intraspecific COLE COLE 25 696 46.4ab y Intraspecific COTI COTI 25 278 18.5cd Interspecific COLE COTI 48 2586 53.9a Interspecific COT I COLE 48 1328 27.7bc F 1 self F 1 z F 1 21 59 2.8d F 1 sibling F 1 F 1 18 321 17.8cd Testcross F 1 COLE 24 424 18.4cd y Significant differences with Cross type being significantly different with an F value of 9.91 a nd z F 1 interspecific hybrids were produce by crossing COLE x COTI.

PAGE 154

154 Table 5 2. Inheritance of the maroon spot for C. leavenworthii and C. tinctoria self, F 1 F 2 and testcross populations. Cross type Seed parent Pollen Parent Crosses (no.) Spotte d Non spotted Expected Ratio 2 value P value Intraspecific COLE COLE 5 0 86 0:1 0 1 Intraspecific COTI COTI 5 35 0 1:0 0 1 Interspecific F 1 COLE COTI 16 525 0 1:0 0 1 Interspecific F 1 COTI COLE 16 493 0 1:0 0 1 Interspecific F 2 self F 1 z F 1 7 25 7 3:1 0 1.33 0.25 1.00 Interspecific F 2 sib F 1 F 1 5 227 79 3:1 0.005 1.45 0.23 0.94 Testcross F 1 COLE 8 163 207 1:1 0.12 2.60 0.11 0.73 z F 1 interspecific hybrids were produce by crossing COLE x COTI.

PAGE 155

155 Table 5 3 Inheritance of seed wingedness in crosses wit h C. leavenworthii and C. tinctoria Cross type Seed parent y Pollen parent Crosses ( n o.) Winged Wingless Intraspecific COLE COLE 40 5722 0 Intraspecific COTI COTI 30 0 870 Interspecific COLE COTI 40 7433 0 Interspecific COTI COLE 26 0 1989 4 52 89 Interspecific F 1 F 1 z F 1 2 23 0 3 48 375 1 0 100 y Seed counted for this study was collected from the seed parent to produce seed for cross type listed. z F 1 interspecific hybrids were produce by crossing COLE x COTI. Table 5 4 Chi square tabl e to determine the probability of gene flow at a specific distance. Parameter DF Estimate Standard error Wald Chi Square Intercept 1 3.2387 0.1569 426.1609*** Distance 1 0.1673 0.0295 32.2638*** Distance*Distance 1 0.00124 0.000398 9.7724**

PAGE 156

156 Table 5 5 Hymenoptera found in the field on C. leavenworthii and C. tinctoria plants at G ulf C oast R esearch and E ducation C enter Wimauma, FL. Family Genus Species Common Name C. leavenworthii C. tinctoria Colletiae Colletes sp. Plasterer bee 1 Vespidae Euody nerus sp. Potter wasp 1 Halictidae Halictus poeyi Halictid bee 6 12 Scoliidae Scolia nobilitata S coliid wasp 2 1 Sphecidae Philanthus ventilabris Digger wasp 1 2 Tiphiidae Myzinum sp. Tiphiid wasp 2 Vespidae Polistes bahamensis Paper wasp 1 Insect species (no.) 6 4

PAGE 157

157 Figure 5 1. Morphological characteristics assessed for potential use as a morphological marker to detect pollen mediated gene flow A) Trichomes on the leaf petiole of C. tinctoria B) Seed wings on C. leavenworthii se ed, C) Non winged seed of C. tinctoria D) Differences in seed wingedness of COLE x COTI seed, E) Maroon spotted flower of C. tinctoria (top) and non spotted flower of C. leavenworthii (bottom) and F) Differences in maroon spot size of C. tinctoria

PAGE 158

158 Fig ure 5 2 Field design of one block out of three for the pollen mediate gene flow study.

PAGE 159

159 Figure 5 3 The inheritance of the size of the maroon spot in F 1 hybrids from reciprocal crosses between four C. leavenworthii and four C. tinctoria parents Figur es on the left side indicated the maroon spot size when C. tinctoria was used as the maternal parent and those on the right side are when C. tinctoria was used as the paternal parent.

PAGE 160

160 Figure 5 4 Gene flow rate s occurring from C. tinctoria to C. leavenw orthii plants at multiple distances in 2007 and 2008 at the Gulf Coast Research and Education Center, Wimauma, FL.

PAGE 161

161 Figure 5 5 Scatter plots of observed gene flow rates at each distance measured and logistic regression curve for the equation fitted to t he rate of gene flow from C. tinctoria to C. leavenworthii at each distance over two year s.

PAGE 162

162 CHAPTER 6 CONCLUSIONS Coreopsis C. leavenworthii is one of the 13 species of Coreopsis found in Florida. C. leavenworthii has been reported as growing in most counties in Florida and only two counties in Alabama. The Florida Department of Transportation (FDOT) grows C. leavenworthii for highway beautification and erosion control along roadsides and ditches. Wildflower growers h ave collected seed from C. leavenworthii natural populations and increased seed to meet the demand. Concerns have been raised about the possibility that genetic shift and erosion may occur during seed increase. C. tinctoria has been reported as growing i n most of the United States, including some counties in Florida but is not native to Florida. It has been reported that these two species are cross compatible and can produce viable interspecific hybrids. The genetic diversity of C. leavenworthii had not been assessed at the phenotypic or molecular levels when seed is increased over successive generations or multiple locations. The effects of interspecific hybridization on the fitness of C. leavenworthii C. tinctoria hybrids were not known. Because both species are growing in Florida, information on the distance and frequency of gene flow was needed. There were no significant differences between the seed increase populations and the original population (G 0 ) for the means of most traits evaluated, includi ng plant height, plant dry weight, leaf type, number of days to flower, disk flower size, whole flower size, petal lobing, number of ray petals, seed germination and powdery mildew severity. The mean values for two traits, petal overlap and number of seed produced per five seed heads, did show significant differences between increase populations and G 0 but did not show a consistent trend that indicated a genetic shift. It was concluded

PAGE 163

163 that seed increase over three generations in two climatic zones did n ot seem to cause any genetic shift or loss of genetic variation in vegetative, physiological and reproductive characteristics of C leavenworthii SSR markers were highly polymorphic in C. leavenworthii and detected an average of 10.4 alleles per primer pa ir in G 0 SSR marker analysis revealed subtle changes in alleles and allele frequencies during seed increase. Some of the changes seemed to be due to random genetic drift, as some alleles were present in G 0 but not in the increase populations and some ot hers were not in G 0 but in the increase populations. The total genetic diversity was 0.1736 in G 0 and it slightly decreased in the six increase populations with the lowest values in the G 3 populations (0.1541 to 0.1706). C ompared with G 0 genetic differ entiation ( G ST ) and genetic distance increased slightly as seed increase advanced to later generations ( 0.0 513 and 0.0196, respectively) PCoA did not reveal any distinct clustering of G 0 and the increase populations, suggesting that a lthough there w ere s ome genetic changes occurring in the seed increase populations, they were not large enough to cause a significant genetic shift Therefore, the genetic diversity and integrity of the original populations were maintained during seed increase Based on thes e results, the current seed production practices seem to be appropriate for C. leavenworthii and increasing seed in northern or central Florida for three generations did not seem to cause any significant negative effects on C. leavenworthii genetic diversi ty and integrity. Nevertheless, subtle allele and allele frequency changes did occur at the molecular level. Should seed increase be advanced to later generations, it will be important to monitor these changes at phenotypic and

PAGE 164

164 molecular levels and to en sure that C. leavenworthii upheld. Interspecific crosses between C. leavenworthii and C. tinctoria produced more seed than their respective intraspecific crosses and interspecific seed germinated well, indicating that C. leavenworthii and C. tinctoria are fully compatible. N umber of days to flower appeared to have heterotic effects in the C. leavenworthii C. tinctoria hybrids. Plant dry weight appeared to be affected by heterosis in the F 1 generation and hybrid breakd own in the F 2 generation. There were significant decreases in the F 1 and F 2 generations for seed production and pollen stainability compared to the intraspecific populations. Pollen stainability was likely affected by chromosome mispairing, while seed pr oduction was likely affected by chromosome mispairing and dilution. Inheritance studies indicated that the maroon spot is controlled by a single dominant gene that is homozygous in C. tinctoria making it a reliable morphological marker for detecting and q uantifying pollen mediated gene flow from C. tinctoria to C. leavenworthii The highest observed rate of pollen mediated gene flow from C. tinctoria to C. leavenworthii was 4.24% when plants of the two species were grown 1.5 m away. The farthest distance that gene flow was observed was 61.0 m. Several Hymenoptera insects were found on both species of Coreopsis but only two are likely pollinators of both Coreopsis species. Based on these results, it appears that C. leavenworthii and C. tinctoria are full y compatible and will produce seed naturally by insect pollination. Overall, it appeared that fitness of the interspecific hybrids was not affected in the vegetative traits, but reproductive fitness was affected in the F 1 and F 2 generations. Backcross po pulations

PAGE 165

165 could additionally be assessed in the field to better understand the effects of interspecific hybridization on fitness. Because of the level of compatibility found, gene flow rates were lower than expected, possibly due to pollen competition or population model These estimates do not take seed dispersal into account, and it is unknown how seed dispersal will affect gene flow. Regression analysis predicts that 0.10% gene flow may occur when the two species are 28 m away from each other, and 0.0 1% gene flow may occur when they are 60 m away. Finally because of the negative effects found in the C. leavenworthii C. tinctoria hybrids, pollen mediated gene flow from C. tinctoria to C. leavenworthii should be prevented to protect the genetic diversi ty and integrity of C. leavenworthii

PAGE 166

166 APPENDIX: ENVIRONMENTAL CONDIT IONS FOR GROWING YEAR S DURING SEED PRODUCTION AND POPULATION EVALUATIO NS Table A 1 Environmental condi tions during seed collection period for seed increase for the genetic diversity st udies at the North Florida and Gulf Coast Research and Education Centers during 2007, 2008 and 2009. Year Location Temperature ( C ) Relative Humidity (%) Rainfall (in.) Wind (mph) 2007 GCREC 23.2 71.08 0.09 8.04 NFREC 23.1 70.15 0.07 5.40 2008 GCREC 23 .3 70.75 0.11 7.00 NFREC 20.9 71.19 0.09 5.32 2009 GCREC 21.6 72.72 0.10 7.63 NFREC 21.6 76.88 0.25 5.42

PAGE 167

167 Table A 2. Environmental conditions at monthly intervals during seed collection period for seed increase for the genetic diversity studies at t he North Florida and Gulf Coast Research and Education Centers during 2007, 2008 and 2009. Year Location Month Temperature ( C ) Relative Humidity (%) Rainfall (in.) Wind (mph) 2007 GCREC Mar 19.4 69 0.62 8.79 Apr 20.7 68 3.65 7.98 May 24.0 67 0.0 0 8 .69 Jun 25.7 78 5.07 6.87 NFREC Mar 16.3 68 1.71 4.98 Apr 17.6 64 1.01 6.09 May 23.1 59 0.99 6.06 Jun 25.8 72 3.12 5.18 Jul 26.5 77 4.63 4.79 Aug 27.5 79 3.28 5.49 2008 GCREC Mar 18.9 71 3.42 9.17 Apr 20.6 65 0.03 7.46 May 24.7 7 0 2.81 7.83 Jun 26.1 77 5.85 5.52 NFREC Mar 14.4 68 3.66 6.53 Apr 18.1 68 0.17 5.25 May 23.0 70 1.48 5.8 0 Jun 25.8 77 4.5 0 4.04 2009 GCREC Mar 19.0 69 1.2 0 7.84 Apr 21.5 70 0.0 0 8.66 May 24.3 78 6.3 0 6.83 Jun 26.9 79 4.8 0 6.07 NFR EC Mar 15.3 73 7.1 0 5.99 Apr 18.5 72 9.85 6.04 May 22.9 81 5.11 5.63 Jun 26.8 79 6.01 4.25

PAGE 168

168 Tabl e A 3 Temperature and relative humidity averages throughout the growing year during the evaluation of the seed increase populations at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. AVG temp ( C ) MIN temp ( C ) MAX temp ( C ) Relative humidity (%) 2009 21.0 15.8 27.4 80.5 2010 24.3 18.1 31.2 76.2 Table A 4 Day length averages throughout the growing year during the eval uation of the seed increase populations at the Gulf Coast and Education Center, Wimauma, FL in 2009 and 2010. Overall (h ) Prior to 1st fl (h) Flowering pd (h) 1 mo. prior to and including fl pd (h) 2009 11.9 13.4 11.44 11.86 2010 12.43 11.71 13.38 12.9 9

PAGE 169

169 Table A 5. Soil and tissue nutrition components for Coreopsis plants grown in the field during the 2010 year at the Gulf Coast Research and Education Center, Wimauma, FL. P K Ca Mg Zn Mn Cu pH Fe TKN Soil sample 72.88 48.40 530.80 49.68 13.12 7.36 12.14 5.00 Plant tissue sample 3482 24360 16286 2929 127.70 103.60 21.40 105.80 45520

PAGE 170

170 Figure A 1 Differences in seed germination of seven C leavenworthii populations produced at two locations and sown at the Gulf Coast Research and Education Center, Wimauma, FL in 2009 and 2010. z Means within cultivar method. Means for each year were statistically analyzed separately and cannot be co mpared to each other.

PAGE 171

171 LIST OF REFERENCES Abbott, R.J. 1992. Plant invasions, interspecific hybridization and the evolution of new taxa. TREE. 7(12):401 405. Antonovics, J. 1976. The nature of limits to natural selection. Ann. Missouri Bot. Gard. 63:224 2 47. Archibald, J.K., M.E. Mort, D.J. Crawford, and J.K. Kelly. 2005. Life history affects the evolution of reproductive isolation among species of Coreopsis (Asteraceae). Evolution. 59(11):2362 2369. Arnold, M.A. and S.A. Hodges. 1995. Are natural hybrids fit or unfit relative to the parents? TREE. 10(2):67 71. Arnold, M.L. 1996. Natural Hybridization and Evolution. Oxford University Press, New York. Arnold, M.L., M.R. Bulger, J.M. Burke, A.L. Hempel, and J.H. Williams. 1999. Natural hybridization: How long can you go and still be important? Ecology. 80(2):371 381. Arias, D.M. and L.H. Ries e berg. 1994. Gene flow between cultivated and wild sunflowers. Theor. Appl. Genet. 89:655 660. Arriola, P.E. and N.C. Ellstrand. 1996. Crop to weed gene flow in the genus Sorghum (Poeaceae): Spontaneous interspecific hybridization between johnsongrass, Sorhum halepense and crop sorghum, S. bicolor Amer. J. Bot. 83(9):1153 1160. Arriola, P.E. and N.C. Ellstrand. 1997. Fitness of interspecific hybrids in the genus Sorghum : Persistence of crop genes in wild populations. Ecol. Appl 7(2):512 518. Barton, N.H. 2001. The role of hybri dization in evolution. Mol. Ecol 10:551 568. Bateman, A.J. 1947. Contami nation of seed crops. J. Genet 48:257 275. Beckie, H.J. and L.M. Hall. 20 08. Simple to complex: Modeling crop pollen m ediated gene flow. Plant Sci 175:615 628. Berg, E.E. and J.L. Hamrick. 1997. Quantification of genetic diversity at allozyme loci. Can. J. For. Res. 27:415 424. Blears, M. J., S.A. d e Grandis, H. Lee, an d J.T. T revors. 1998. Amplified fragment length polymorphism (AFLP): A review of the procedu re and its applications. J. Ind. Microbiol. Biot 21:99 114.

PAGE 172

172 Booth, D.T. and T.A. Jones. 2001. Plants for ecological restoration: A foundation and a philosophy for the futu re. Native Plants J. 2(1):12 20. Borner, A., S. Chebotar and V. Korzun. 2000. Molecular characterization of the genetic integrity of wh eat ( Triticum aestivum L.) germ plasm after long term maintenance Theor. Appl. Genet. 100:494 497. Brewbaker, J.L. 1957. Pollen cytology and self incompatibili ty systems in plants. J. Hered 48:271 277. Bryant, B. and B.L. Harper Lore. 1997. Where wildflowers bloom, so does hope. US Dept of Transportation, Federal Highway Administration. Public Roads Online. Vol. 61 (3). Bur ton, P.J. and C.M. Burton. 2002. Promoting genetic diversity in the production of large quantities of native plant seed. Ecol Restor. 20:117 123. Cai H.W. and H. Morishima. 2002. QTL clusters reflect character associati ons in wild and cultivated rice. Th eor. Appl. Genet. 104 :1217 1228. Cain, M.L., B.G. Milligan, and A.E. Strand. 2000. Long distance seed dispersal in plant populations. Amer. J. Bot. 87(9):1217 1227. Chapman, M.A., D.G. Forbes, and R.J. Abbott. 2005. Pollen competition among two species of Senecio (Asteraceae) that form a hybrid zone on Mt. Etna, Sicily. Amer. J. Bot. 92(4):730 735. Chebotar, S., M.S. Roder, V. Korzun, B. Saal, W.E. Weber and A. Borner. 2003. Molecular studies on genetic integrity of open pollinating species rye ( Secalse cer eal L.) after long term gene bank maintenance Theor. Appl. Genet. 107:1469 1476. Clausen J., D.D. Keck, and W.M. Hiesey. 1940. Experimental studies on the nature of species. I. Effect of varied environments on western North American plants. Washington (DC ): Carnegie Institute. Publication 250. Clewell, A.F. 1985. Guide to the vascular plants of the Florida panhandle. Univ. Press of Florida, Tallahassee, Fla. Coyne, J.A. and H.A. O rr. 1998. The evolutionary genetics of speciation. Phil. Trans. R. Soc. Land. 353:287 305. Crawford, D.J., E.B. Smith, and R.E. Pilatowski. 1984. Isozymes of Coreopsis section Calliopsis (Compositae): Genetic variation within and divergence among the species. Brittonia. 36(4):375 381. Czar necki, D.M., J.G. Norcini, and Z Deng. 200 7. Phenotypic diversity of Coreopsis leavenworthii To rr. and Gray (Asteraceae). Nativ e Plants J. 8(1):45 57.

PAGE 173

173 Czarnecki, D.M., M.N. Rao, J.G. Norcini, F.G. Gmitter, and Z. Deng. 2008. Genetic diversity and differentiation among natural, production, and intr oduced populations of narrowly endemic sp e cies Coreopsis leavenworthii (Asteraceae). J. Amer. Soc. Hort. Sci. 133:234 241. Edmands, S. 2007. Between a rock and a hard place: Evaluating the relative risks of inbreeding and outbreeding for conservation and m anagement. M ol. E col 16:463 475. Ellstrand, N.C. 1992 a Gene flow amo ng seed populations. New Forest 6:241 256. Ellstrand, N.C. 1992b. Gene flow by pollen: Implications for plant conservation genetics. OIKOS. 63:77 86. Ellstrand, N.C. 2003 a Current know ledge of gene flow in plants: Implications for transgene flow. Phil. Trans. R. Soc. Lond. 358:1163 1170. Ellstrand, N.C. 2003 b Dangerous liaisons? When cultivated plants mate with their wild relatives. Johns Hopkins University Press, Baltimore. Ellstrand, N.C., B. Devlin, and D.L. Marshall. 1989. Gene flow by pollen into small populations: Data from experimental and natural stands of wild radish. Proc. Natl. Aced. Sci. 86:9044 9047. Ellstrand, N.C. and D.R. Elam. 1993. Population genetic consequences of sm all population size: Implication s for plant conservation. Annu. Rev. of Ecol. Syst 24: 217 242. Ellstrand, N.C., H.C. Prentice, and J.F. Hancock. 1999. Gene flow and introgression from domesticated plants into thei r wild relatives. Annu. Rev. Ecol. Syst. 3 0:539 563. FDOT. 1995. Fl orida highway landscape guide. Florida Department of Transportation, Tallahassee, FL. FDOT. 2006. 2006 FDOT design standards permanent erosion control. Florida Department of Transportation, Tallahassee, FL. Ferdinandez, Y.S.N., B.E Coulman, and Y.B. Fu. 2005. Detecting genetic changes over two generations of seed increase in an awned slender w h eatgrass population using AFLP markers. Crop Sci. 45:1064 1068. Foote, M. 1997. The evolution of morphological diversity. Ann. Rev Ecol Sy st 28:129 152. Fu, Y.B., Y.S.B. Ferdinandez, A.T. Phan, B. Coulman, and K.W. Richards. 2004. AFLP variation in four blue grama seed sources. Crop Sci. 44:283 288.

PAGE 174

174 Fulton, T.M., J. Chunwongse, and S.D. Tanksley. 1995. Microprep protocol for extraction of D NA from Tomato and other he rbaceous plants. Plant Mol. Biol. Rep 13(3):207 209. Futuyma, D.J. 1979. Evolutionary Biology. Sinauer Associates, Inc., Sunderland, MA. Gilman, E.F., J.G. Norcini and T. Howe. 2007. Coreopsis leavenworthii FPS142. Glenn, T.C. and N.A. Schable. 2005. Isolating micro satellites DNA loci. Method Enzymol 395:202 222. Goldstein, D.B. and D.D. Pollock. 1997. Launching microsatellites: A review of mutation processes and methods of ph ylogenetic inference. J. Hered 88:335 342. Grant, V 1981. Plant Speciation. Columbia University Press, New York. Greenleaf, S.S., N.M. Williams, R. Winfree, and C. Kremen. 2007. Bee foraging ranges and their relationship to body size. Oecologia. 153:589 596. Hamrick, J.L. 1983. The distribution of genetic variation within and among natural plant populations, p. 335 348. In: S.M. Chambers and C.M. Schonewald (eds.). Genetics and wild population management. Addison Wesley, New York. Harper, J.L., J.N. Clatworthy, I.H. McNaughton and G.R. Sagar. 1961. The ev olution an d ecology of closely related species living in the same area. Evolution. 15(2):209 227. Harris, H. 1966. Enzyme polymorphism in man. Proc of the Royal Soc of London 169: 298 310. Hartl, D.L. 2000. A primer of population genetics. 3 rd ed. Sinaue r Associates, Inc., Sunderland, MA. Hausbeck, M.K., W.R. Quackenbush and S.D. Linderman. 2002. Evaluation of cultivars of African daisy for resistance to powdery mildew. < http://plantpathology.msu.edu/labs/hausbeck/Publications/2003%20B&C%2018 ~O004.pdf > H avens, K. 1998. The genetics of plant restoration: An overview and a surprise. Restor M g t Notes. 16:68 72. Heywood, V.H. 1993. Flower plants of the world. Oxford University Press, New York. Hoekstra, F.A. and J. Bruinsma. 1975. Viability of Compositae po llen: Germination in vitro and influences of climatic conditions during dehiscence. Z. Pflanzenphysiol. Bd. 76:36 43.

PAGE 175

175 Hufford, K.M. and S.J. Mazer. 2003. Plant Ecotypes: Genetic differentiation in the age of ecological restoration. TREE. 18(3):147 155. Kab at, S.M., J.G. Norcini, and B. Deghan. 2007. Temperature and light affects coreopsis ( Coreopsis leavenworthii ). Native Plants J. 8(3):236 247. Knapp, E. E. and K. J. Rice. 1994. Starting from seed: genetic issues in using native grasses for restorat ion. Restor M g t Notes 12:40 45. Kumar, S., K. Tamara, and M. Nei. 2004. MEGA 3: Integrated software for molecular evolutionary genetics analysis and s equence alignment. Briefings in Bioinform atics 5: 150 163. Langridge, P. and K. Chalmers. 2004. The principle: Identification and application of molecula r markers, p. 3 54. In: H. Lorz and G. Wenzel (eds.). Mol Marker Systems. Springer Verlag, Berlin. Levin, D.A. 1981. Dispersal versus gene flow in plants. Ann Missouri Bot. Gard 68(2):233 253. Levin, D.A. and H.W. Kerster. 1969. The dependence of bee mediated pollen and gene dispersal upon plant density. Evolution. 23(4):560 571. Li, Y.C., A.B. Korol, T. Fahima, A. Beiles, and E. Nevo. 2002. Mic rosatellites: Genomic distribution, putative functions and mutational mechanisms : A review. Mol. Ecol 11: 2453 2465. Meyer, S.E. and S. B. Monsen. 1993. Genetic considerations in propagating native shrubs, forbs, and grasses from seed. p.47 54. In: T. D. L andis (technical coordinator). Proceedings, Western Forest Nursery Association Annual Meeting, Fallen Leaf Lake CA, September 14 18, 1992. USDA Forest Service General Technical Report RM 221. Morgante, M. and A.M. Olivieri. 1993. PCR amplified microsatelli tes as markers in plant genetics. Plant J 3(1):175 182. Mueller, U.G. and L.L. Wo lfenbarger. 1999. AFLP genotypi ng an d fingerprinting. TREE. 14(10): 389 394. Muller, A. and M. Kuhlmann. 2008. Pollen hosts of w estern palaearctic bees of the g enus Colletes ( Hymenoptera: Colletidae): The Ast eraceae paradox. Biol. J. Linn. Soc 95:719 733. Nei, M. 1978. Estimation of average heterozygosity and genetic distance form a small number of individuals. Genetics. 89:583 590. Nei, M. 1987. Molecular Evolutionary Genetic s. Columbia University Press, NY.

PAGE 176

176 Norcini, J. G. 20 02. Seed production of Leavenworth C oreopsis ENH868. Norcini, J.G. and J.H. Aldrich. 2004. Establishment of native wildflower plantings by seed. Florida Cooperative Extension Service Publication ENH 968 Norcini, J.G. and A.H. Aldrich. 2007. Storage effects on dormancy and germination of native tickseed species. HortTechnology. 17(4):505 512. Norcini, J.G. and J.H. Aldrich. 2008. Assessing viability of prevariety germplasm of native Coreopsis species. Ho rtScience. 43(6):1870 1881. Norcini, J.G., J.H. Aldrich, and F.G. Martin. 2006. Harvest year and fertilizer effects on Coreopsis J. Environ. Hort. 24(2):63 67. Norcini, J.G., A.L. Frances, and C. Rei nhart Adams. 2009. Establishment of Coreopsis leavenworthii ) on roadside right of ways. ENH1104. Norcini, J.G., M. Thetford, K.A. Klock Moore, M.L. Bell, B.K. Harbaugh, and J.H. Aldrich. 2 001. Black eyed susan success: Y our seed sou rce is important. Greenhouse Prod News 11(2):58 60. Osborne, J.L., A.P. Martin, N.L. Carreck, J.L. Swain, M.E. Knight, D. Goulson, R.J. Hale, and R.A. Sanderson. 2008. Bumblebee flight distances in relation to the forage landscape. J. Anim. Ecol 77:406 415. Parker, H.M. 1973. A biosystematic study of section Calliopsis of Coreopsis (Compositae). Univ. of Ark. Littlerock, PhD Diss Potts, B.M., W.C. Potts, a nd B. Cauvin. 1987. Inbreeding and interspecific hybridization in Eucalyptus gunnii Silvae Geneti ca. 36:194 199. Price, P.W. 1997. Insect Ecology 3 rd Ed. John Wiley and Sons, Inc., New York. Reed, D. H. and R. Frankham 2003. Correlation between Fitness and Genetic Diversity. Conservation Biol 17: 230 237. Rincker, C.M., R.G. May, H.H. Rampton, C.S. G arrison, and J.G. Dean. 1982. Genetic stability of three Italian ryegrass cultivars during seed multiplication in diverse environments. Crop Sci 22: 377 380. Rogers D.L. and A.M. Montalvo. 2004. Genetically appropriate choices for plant materials to maint ain biological diversity. University of California Report to the USDA Forest Service, Rocky Mountain Region, Lakewood, CO. Rohlf, F.J. 2005. NTSYSpc numerical taxonomy and multivariate analysis system version 2.2. Exeter Software, Setauket New York.

PAGE 177

177 Rukin i, D. 2008. Dormancy in pre variety germplasm of native Coreopsis species. University of Florida Gainesville, PhD Diss Ryding, O. and K. Bremer 1992. Phylogeny, distribution, and classification of the Coreopsid eae (Asteraceae). Syst. Bot. 17: 649 659. SA S Institute, Inc. 1997. SAS Users Guide. SAS Inst., Inc. Raleigh, NC. Schmitt, J. 1980. Pollinator foraging behavior and gene dispersal in Senecio (Composita e). Evolution. 34(5):934 943. Sharma, J.N. 1971. The genetics, nature and occurrence of self incomp atibility and cross incompatib i lity in four annual species of Coreopsis L. Montana State University Bozeman, PhD Diss Sherff, E.E. 1955. Core opsis and Bidens p. 4 40. In: E.E Sherff and E.J. Alexander, C ompositae Heliantheae Coreopsid inae. N. Amer. Fl. Series II. Part 2, New York Botanical Gardens, New York. Simon, U. and A. Kastenbauer. 1979. Gro w th type and yield comparisons of forage species after seed multiplication in Germany and in the United St ates. II. Meadow Fescue, Timoth y, and Perennial Ryegra ss. Crop Sci 19:209 213. Sinclair, E.A. and R.J. Hobbs. 2009. Sample size effects on estimates of population genetic structure: Implications for ecological restoration. Restor. Ecol 17(6):837 844. Slatkin, M. 1987. Gene flow and the geographic structure of natural populations. Science. 236:787 792. Smith, E.B. 1975. The chromosome numbers of North American Coreopsis with phyletic interpretations. Bot. Gaz. 136(1):78 86. Smith, E.B. 1976. A biosystematic survey of Coreopsis in eastern United States and Can ada. SIDA. 6(3):123 215. Smith, E.B. 1983. Phyletic tr ends in sections Eublepharis an d Calliopsis of the genus Coreopsis (Compositae). Amer. J. Bot. 70(4):549 554. Smith, E.B. and H.M. Parker. 1971. A biosystematic study of Coreopsis tinctoria an d C. c arda minefolia (Compositae). Brittonia. 23:161 170. Song, Z.P., B.R. Lu, B. Wang, and J.K. Chen. 2004. Fitness estimation t hrough performance comparison o f F 1 hybrids with their parental species Oryza rufipogon and O. sativa Ann. Bot London 93:311 316. Stebbi ns, G.L. 1959. The role of hybridization in evolution. P. Am. Philos. Soc 103(2):231 251.

PAGE 178

178 Tadesse, M., D.J. Crawford, and E.B. Smith. 1995. Comparative capitular morphology and anatomy of Coreopsis L. and Bidens L. (Compositae), including a review of gene ric boundaries. Brittonia. 47(1):61 91. Tesfaye, T., B. Getachew and M. Worede. 1991. Morphological diversity in tetraploid wheat landrace populations from the central highlands of Ethiopia. Hereditas. 114:171 176. Thiele, J., T. Hansen, H.R. Siegismund, and T.P. Hauser. 2010. Genetic variation of inbreeding depression among floral and fitness traits in Silene nutans Heredity. 104:52 60. of insects. 7 th Ed. Thomson Broo ks/Cole, Belmont, CA. Turesson, G. 1922. The genotypical response of the plant species to the habitat. Heriditas. 3:211 350. USDA. 2011a. PLANTS profile Coreopsis leavenworthi i Torr. and A. Gray Leavenworth United States Department of Transport ation. 7 Feb. 2011. < http://www.plants.usda.gov/java/nameSearch?keywordquery=coreopsis+l eavenworthii&mode=sciname > USDA. 2011b. PLANTS profile Coreo psis tinctoria Nutt. Golden tickseed. United States Department of Transportation. 7 Feb. 2011. < http://www.plants.usda.gov/java/profile?symbol=COTI3 > van Treuren, R., R. Bulsma, N.J. Oubo rg, and W. van Delden. 1993. The significance of genetic erosion in the process of extinction. IV. Inbreeding depression an d heterosis effects caused by s e l fing and outcrossing in Scabiosa columbaria Evolution. 47(6):1669 1680. Vos, P., R. Hogers, M. Blee ker, M. Reijans, T. van de Lee, M. Hornes, A. Frijters, J. Pot, J. Peleman, M. Kuiper, and M. Zabeau. 1995. AFLP: A new technique for DNA finger printing. Nucl Acids Res 23(21):4407 4414. Weising, K., H. Nybom, K. Wolff and G. Kahl. 2005. DNA fingerprint ing in Plants: Principles, methods, and applications. 2 nd ed. CRC Press, Boca Raton, FL. Whitten, J.L. 2002. Plant fact sheet Calliopsis Coreopsis tinctoria Nutt. USDA. < plants.usda.gov/factsheet/doc/fs_coti3.doc > Wunderlin, R.P. 1998. Guide to the vascula r plants of Florida. Univ. Press of Fla., Gainesville, Fla. Wunderlin, R.P. and B.F. Hansen. 2004. Atlas of Florida vascular plants. Florida Center for Community Design and Research, and Institute for Systematic Botany, University of South Florida. 7 May 2 011. < http://www.plantsatlas.usf.edu >

PAGE 179

179 Yeh, F.C., R C. Yang, T.B.J. Boyle, Z H. Ye, and J.X. Mao. 1997. POPGENE, the user friendly shareware for population genetic analysis. Molecular Biology and Biotechnology Centre, University of Alberta, Canada. Zane, L., L. Bargelloni, and T. Patarnello. 2002. Strategies for microsatellit e isolation: A review. Mol. Ecol 11:1 16.

PAGE 180

180 BIOGRAPHICAL SKETCH Sarah Smith was born in Leechburg, PA. She attended The Pennsylvania State University in State College, PA, where she g raduated in May 2004 with a B.S. in h orticulture. She then joined the Horticultural Sciences department at the University of Florida in August 2004 and received her Master of Science degree in December 2006. She then continued at the University of Florid a to pursue her Doctor of Philosophy degree in the Horticultural Sciences Department which was awarded in August 2011