Upload
others
View
6
Download
0
Embed Size (px)
Citation preview
GENOMIC AND BREEDING RESOURCES TO PRODUCE SEEDED AND HIGH BIOMASS INTERSPECIFIC HYBRIDS OF NAPIERGRASS AND PEARL MILLET
By
DEV RAJ PAUDEL
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2018
© 2018 Dev Raj Paudel
To my late Mom
4
ACKNOWLEDGMENTS
I wish to express my appreciation to the members of my advisory committee: Dr.
Fredy Altpeter, Dr. Jianping Wang, Dr. Patricio Munoz, Dr. Calvin Odero, and Dr.
Salvador Gezan, who have guided, supported, and encouraged me throughout the
course of my research project. Sincere thanks to my research advisor Dr. Fredy
Altpeter for allowing me to join his group and pursue the work described here. Thank
you for all the support, guidance, and mentorship that you have provided during my
graduate studies. I am truly inspired by your professionalism and leadership role.
I am particularly indebted to my co-advisor Dr. Jianping Wang who provided me
with space in her lab to do my experiments and provided me with opportunities to
develop skills in molecular and computational biology. Your mentorship has been an
invaluable gift over the past couple of years. One day, I hope to inspire others as you've
inspired me.
I would like to gratefully acknowledge the University of Florida Graduate School
Fellowship for funding the first four years of my PhD. I am thankful to the Graduate
School Doctoral Dissertation Award for funding my final semester. I am grateful to the
Florida Plant Breeders Working Group for providing funds for this research.
I was grateful to be surrounded by many wonderful people during my journey.
Many thanks to talented undergraduates, Fan Wen, Erik Hanson, and Stephanie Maya,
who helped not only with the experiments, but also helped me learn and hone my
mentorship skills.
In addition, I would like to express my gratitude to the following people:
• Dr. Baskaran Kannan without whose help, support, and wisdom, this work
wouldn’t have been completed.
5
• Staff at Plant Science Research & Education Unit, Citra, FL for coordinating and
helping in planting, managing, and harvesting of research plots.
• Dr. Calvin Odero, Venancio Fernandez, Raphael Mereb Negrisoli, and Nikol
Havranek at the Everglades Research and Education Center for providing
support to do fieldwork in Belle Glade, FL.
• Dr. Tina Strauss, Dr. Eshan Gurung, Er. Prasan Gurung, and Dr. Laxman
Adhikari for your continued support, valuable insights in my research, and
friendship.
• Members of the Altpeter Lab: Derek Hurley, Dr. Saroj Parajuli, Dr. Ratna Karan,
Dr. Simon Gere, and Bryant Brown, for helping during field work and research
activities.
• Members of the Wang Lab: Dr. Liping Wang, Dr. Xiping Yang, Dr. Ze Peng, Dr.
Zhou Hai, Aleksey Kurashev, Dr. Yu-Chien Tseng, and Dr. Song Jian for helping
in fieldwork, lab experiments, as well as discussions related to research that
were instrumental in providing insights into some of the work described in this
document.
• Marco Sinche for kindly providing yield and flowering data of mapping
experiments.
• Dr. Rajeev Varshney for providing the pearl millet reference genome sequence.
• Dr. Karen-Harris Schultz for providing napiergrass sequences.
• Cynthia Hight for playing a phenomenal role in supporting academic endeavors
and ensuring that I was on track every semester.
6
• Members of the Plant Science Council for guidance, support, friendship, and
humor that has helped me survive graduate school.
• Nepalese community in Gainesville, FL for always being there for me.
Obtaining this advanced degree was only possible due to the sheer sacrifice,
patience, unconditional love and understanding, long-term support, and encouragement
in all aspects of my career and life from the love of my life, my dear wife, Ashmita
Guragain, my father Bala Bhadra Paudel, mother late Dhan Kumari Paudel, brother
Dipendra Paudel, sister-in-law Kalpana Sapkota, and niece Deleena Paudel. I am
thankful to my parents-in-law and family, and other relatives for persistent
encouragement and support.
Finally, this dissertation is dedicated to my true champion and idol, my mother,
late Dhan Kumari Paudel, whom I lost during the final semester. She was my confidant
and mentor. She was the epitome of love, sacrifice, simplicity, wisdom, and strength, to
whom I owe everything. Life isn't the same without you.
I hope I have made you proud.
7
TABLE OF CONTENTS page
ACKNOWLEDGMENTS .................................................................................................. 4
LIST OF TABLES .......................................................................................................... 10
LIST OF FIGURES ........................................................................................................ 12
LIST OF ABBREVIATIONS ........................................................................................... 17
ABSTRACT ................................................................................................................... 18
CHAPTER
1 BACKGROUND ...................................................................................................... 20
Introduction ............................................................................................................. 20
Botany .............................................................................................................. 21 Napiergrass Ancestry ....................................................................................... 22 Napiergrass in the United States ...................................................................... 23
Pearl Millet ........................................................................................................ 24 Interspecific Hybrids of Napiergrass and Pearl Millet (PMN Hybrids) ............... 25
Cytoplasmic Male Sterility (cms) ...................................................................... 25 Molecular Tools Applied in Plant Breeding ....................................................... 27
Genotyping by Sequencing (GBS) ................................................................... 29 Target Enrichment Sequencing ........................................................................ 30
Quantitative Trait Loci Analysis ........................................................................ 30 Biomass yield ................................................................................................... 31 Flowering Time ................................................................................................. 31
Genes Related to Flowering ............................................................................. 33 Objectives ............................................................................................................... 34
2 SURVEYING THE GENOME AND CONSTRUCTING A HIGH-DENSITY GENETIC MAP OF NAPIERGRASS (CENCHRUS PURPUREUS SCHUMACH.) ......................................................................................................... 37
Introduction ............................................................................................................. 38
Methods .................................................................................................................. 41 Napiergrass Genome Survey ........................................................................... 41 SSR Identification and Marker Development .................................................... 42
Plant Materials and DNA Extraction ................................................................. 43 Genotyping-by-sequencing ............................................................................... 43 Comparative Genomics .................................................................................... 43 Sequence Analysis and SNP Calling ................................................................ 44 Linkage Map Construction ................................................................................ 45 Comparison Between Napiergrass and Pearl Millet Genome........................... 46
8
Results .................................................................................................................... 46 Napiergrass Genome Survey ........................................................................... 46
Genotyping-by-sequencing ............................................................................... 47 SNP Calling by Various SNP Callers ................................................................ 49 Genetic Linkage Map Construction .................................................................. 49 Comparison Between Genomes of Napiergrass and Pearl Millet ..................... 51
Discussion .............................................................................................................. 52
3 MAPPING QTLS CONTROLLING FLOWER NUMBER AND FLOWERING TIME IN NAPIERGRASS ........................................................................................ 84
Introduction ............................................................................................................. 84 Materials and Methods............................................................................................ 87
Development of a Mapping Population ............................................................. 87 Phenotyping the Mapping Population ............................................................... 88
Genetic Map ..................................................................................................... 88 QTL Analysis .................................................................................................... 89 Candidate Gene Identification .......................................................................... 90
Results .................................................................................................................... 91 Phenotypes ...................................................................................................... 91
Number of flowers ...................................................................................... 91 Flowering Time .......................................................................................... 91
QTL Analysis .................................................................................................... 91
Number of flowers ...................................................................................... 91 Flowering time ............................................................................................ 92
Candidate Genes ............................................................................................. 92
Discussion .............................................................................................................. 93
Conclusion .............................................................................................................. 97
4 EVALUATE THE GENETIC BACKGROUND OF FLOWERING TIME IN A NAPIERGRASS GERMPLASM COLLECTION .................................................... 113
Introduction ........................................................................................................... 113 Materials and Methods.......................................................................................... 117
Plant Materials and Phenotyping .................................................................... 117
DNA Extraction ............................................................................................... 117 Targeted Candidate Genes ............................................................................ 118 Probe Design .................................................................................................. 118
Probe Synthesis, Selection, and Sequencing ................................................. 119 Sequence Read Trimming and Mapping ........................................................ 120 SNP Calling .................................................................................................... 120 Population Structure ....................................................................................... 120
Results .................................................................................................................. 121 Candidate Genes Related to Flowering .......................................................... 121 Probe Design .................................................................................................. 121 Sequence processing ..................................................................................... 123 Phylogenetic analysis ..................................................................................... 124
9
Genome wide association analysis ................................................................ 124 Discussion ............................................................................................................ 125
Conclusion ............................................................................................................ 129
5 GENERATION OF INTERSPECIFIC HYBRIDS BETWEEN PEARL MILLET AND NAPIERGRASS AND EVALUATION OF THEIR PERFORMANCE. ............ 142
Introduction ........................................................................................................... 142 Materials and Methods.......................................................................................... 147
Description of Male Sterile Line of Pearl Millet ............................................... 147 Production of cms Lines of Pearl Millet .......................................................... 147 Production of PMN Hybrids ............................................................................ 148 Experimental Design ...................................................................................... 149
Traits Evaluated ............................................................................................. 149 Data Analysis ................................................................................................. 150
Results .................................................................................................................. 150 Plant Height .................................................................................................... 150 Tiller Number .................................................................................................. 150
Stem Diameter ............................................................................................... 151 Leaf Length .................................................................................................... 151
Leaf Width ...................................................................................................... 152 Plant Biomass ................................................................................................ 152 Dry Biomass ................................................................................................... 153
Coefficient of Variation ................................................................................... 153 Correlation ...................................................................................................... 154
Discussion ............................................................................................................ 154
6 CONCLUDING REMARKS ................................................................................... 179
Summary .............................................................................................................. 179 Future work ........................................................................................................... 181
LIST OF REFERENCES ............................................................................................. 183
BIOGRAPHICAL SKETCH .......................................................................................... 208
10
LIST OF TABLES
Table page 2-1 Parameters used for SNP calling for each software ........................................... 58
2-2 Repetitive elements present in the napiergrass genome .................................... 61
2-3 The sequence alignment of ten napiergrass sequence contigs to the pearl millet genome ..................................................................................................... 62
2-4 Frequency of classified repeat types (considering sequence complementary) in napiergrass ..................................................................................................... 63
2-5 Primer pairs developed for napiergrass SSR markers ........................................ 67
2-6 Summary of the alignment of non-redundant tags of napiergrass (Cenchrus purpureus) to the available genomes of different species .................................. 68
2-7 Alignment of individual napiergrass reads using Bowtie2 ................................... 69
2-8 Summary of the combined linkage map of napiergrass and the percentage of gaps less than 5 cM in male and female parent linkage maps ........................... 70
2-9 Summary of napiergrass single nucleotide polymorphism (SNP) markers mapped on the combined linkage map using 9 different software pipelines ....... 71
3-1 Descriptive statistics of flowering date and number of flowers for 185 F1
hybrids of a cross (N190 N122) at Citra, FL., in 2012, 2013 (Sinche 2013) and 2016 .......................................................................................................... 109
3-2 Detailed information about the QTLs for number of flowers. LG = Napiergrass linkage groups, Peak markers represent the marker at the highest peak of QTL, LSI represents the LOD-1 support interval in cM. .................................... 110
3-3 Detailed information about the QTLs for flowering time. LG = Napiergrass linkage groups, Peak markers represent the marker at the highest peak of QTL, LSI represents the LOD-1 support interval in cM. .................................... 111
3-4 List of putative flowering related genes from the genome of pearl millet (Varshney et al. 2017). ..................................................................................... 112
4-1 Genes and publications related to flowering used in this research ................... 130
5-1 Details of the cross types used in the experiment. ........................................... 175
11
5-2 Descriptive statistics for different types of crosses. P-value represents Pr>F for Levene’s test for homogeneity of variance (center = median), CV = coefficient of variation. ...................................................................................... 176
5-3 Correlation coefficients and p-values for biomass weight and biomass-related traits for PMN hybrids evaluated in Citra, FL. ................................................... 177
12
LIST OF FIGURES
Figure page 1-1 Distribution of napiergrass. Black dots represent local napiergrass and red
dots represent napiergrass listed as invasive species ........................................ 36
2-1 Sequence variation for SNPs called in various regions of the pearl millet genome. ............................................................................................................. 72
2-2 Micro-collinearity between contigs from napiergrass to the pearl millet genome. ............................................................................................................. 73
2-3 Inversion duplication between napiergrass and pearl millet (shown in bottom figure). 74
2-4 Estimated coverage of PstI restriction sites in the pearl millet genome. ............. 75
2-5 Histogram of uniquely mapped reads to the pearl millet genome. ...................... 76
2-6 Venn diagram showing concordant napiergrass SNPs called by five reference-based SNP callers, SAMtools, GBS-SNP-CROP, GATK, FreeBayes, and TASSEL. Numbers in parenthesis after the program name shows the total number of SNPs called by each program. ................................. 77
2-7 Genetic linkage map of the napiergrass female parent N190. ............................ 78
2-8 Genetic linkage map of the napiergrass male parent N122. ............................... 79
2-9 Genotyping by sequencing single nucleotide polymorphism (GBS-SNP) marker distribution for the 14 linkage groups of napiergrass. A black bar means a GBS-SNP marker. A blue bar represents segregation distortion region. The left scale plate represents genetic distance (centiMorgan as unit). . 80
2-10 Consensus genetic linkage map of napiergrass. ................................................ 81
2-11 Circos plot of the mapped TASSEL de-novo UNEAK napiergrass markers with pearl millet reference genome. Pearl millet pseudomolecules start with “PM” and are color coded for each pseudomolecule. Napiergrass linkage groups start with “LG” and are in green color. Each line corresponds to tags that showed significant BLAST hits to the pearl millet genome (> 80% identity and > 50 bp length)............................................................................................. 82
2-12 Syntenic regions between napiergrass linkage groups and the pearl millet genome. PM01 to PM07 are pearl millet pseudomolecules, LG01 to LG14 are napiergrass linkage groups. The small dots represent significant BLAST hits of mapped UNEAK tags to the pearl millet genome (>80% identity and >50 bp length). ................................................................................................... 83
13
3-1 Histogram of number of flowers in the mapping population in 2012 in Citra, FL. X-axis represents the number of flowers and y-axis represents the count of plants. Average number of flowers of two parental lines, N122 and N190 are plotted. Data is adapted from Sinche, 2013. ................................................. 98
3-2 Histogram of number of flowers in mapping population for 2013 in Citra, FL. X-axis represents the number of flowers and y-axis represents the count of plants. Average number of flowers of two parental lines, N122 and N190 are plotted. Data is adapted from Sinche, 2013. ....................................................... 99
3-3 Scatterplot of number of flowers between 2012 and 2013 in Citra, FL. X-axis represents the number of flowers in 2012 and y-axis represents the number of flowers in 2013. Data is adapted from Sinche, 2013. ................................... 100
3-4 Histogram of number of days to first flowering in 2012 in Citra, FL. X-axis represents the number of days to flowering (FT) and y-axis represents the number of accessions. Days to first flowering of two parental lines, N122 and N190 are plotted. Data is adapted from Sinche, 2013. ..................................... 101
3-5 Histogram of number of days to first flowering in 2013 in Citra, FL. X-axis represents the number of days to flowering (FT) and y-axis represents the number of accessions. Days to first flowering of two parental lines, N122 and N190 are plotted. Data is adapted from Sinche, 2013. ..................................... 102
3-6 Histogram of first flowering date in the mapping population in 2016 in EREC, Belle Glade, FL. X-axis represents the number of days to flowering (FT) and y-axis represents the number of accessions. Days to first flowering of two parental lines, N122 and N190 are plotted. ...................................................... 103
3-7 Scatterplot of first date of flowering between different years and locations. X-axis and Y-axis represent the number of days to flowering (FT) in different years. Data for 2012 and 2013 are from Citra, FL and adapted from Sinche, 2013. Data for 2016 are from Belle Glade, FL. ................................................. 104
3-8 Scatterplot between number of flowers and days to flowering in 2012 in Citra, FL. X-axis represents the number of days to flowering (FT) in 2012 and y-axis represents the number of flowers per line in 2012. Data is adapted from Sinche, 2013..................................................................................................... 105
3-9 Scatterplot between days to flowering and number of flowers in 2013 in Citra, FL. X-axis represents the number of days to flowering (FT) in 2013 and y-axis represents the number of flowers per line in 2013. Data is adapted from Sinche, 2013..................................................................................................... 106
3-10 Linkage map of male parent N122 showing potential and stable QTLs. QTLs for number of flower (NF) are color coded in blue and that for flowering time (FT) are colored in green. A black bar means a GBS-SNP marker. Markers
14
are labeled on the left and genetic distance (centimorgan as unit) are labeled on right. LG = Linkage group. ........................................................................... 107
3-11 Linkage map of female parent N190 showing potential QTLs. QTLs for number of flower (NF) are color coded in blue and that for flowering time (FT) are colored in green. A black bar means a GBS-SNP marker. Markers are labeled on the left and genetic distance (centimorgan as unit) are labeled on right. LG = Linkage group. ................................................................................ 108
4-1 Histogram of length of flowering related genes................................................. 131
4-2 Number of probes designed per gene. ............................................................. 132
4-3 Number of probes designed as a factor of the size of the gene. ...................... 133
4-4 Distribution of the targeted flowering genes in the genome of pearl millet. Each blue bar represents a flowering gene mapped on the pearl millet genome. ........................................................................................................... 134
4-5 Number of paired-end reads per sample. ......................................................... 135
4-6 Bayesian Information Criteria (BIC) vs. number of clusters in k-means clustering suggests K=3 in the germplasm collection. ...................................... 136
4-7 Projection of the napiergrass germplasm collection using first two linear discriminants (LDs) from discriminant analysis of principal components (DAPC). The shape of the points represent grouping by DAPC (circle = group 1; triangle = group 2; square = group 3) and the colors represent the origin continent: black = Americas; blue = Asia; and cyan = Africa. ........................... 137
4-8 Evolutionary relationships of taxa. G stands for group assigned by DAPC and are in different shapes (circle = group 1; triangle = group 2; square = group 3) and the fill colors represent the origin: black = Americas, blue = Asia, cyan = Africa, and white=unknown). ......................................................................... 138
4-9 Histogram for days to flowering trait in napiergrass germplasm collection. ...... 139
4-10 QQ plots for different models using GWASpoly using 78,129 SNP markers. DTF = Days to flowering. .................................................................................. 140
4-11 Manhattan plots for different models using GWASpoly. P values adjusted with FDR at 0.05. DTF = Days to flowering. ..................................................... 141
5-1 Seeds of pearl millet dwarf cms line (A), cms high biomass pearl millet line (B), PMN hybrid (C), and napiergrass (D), respectively from left to right. ......... 158
5-2 Boxplot of plant height (cm) for four different types of crosses studied. Small colored dots represent individual plant height of the cross. Different letters in
15
blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ...................................................................... 159
5-3 Histogram of plant height (cm) for four different crosses. X-axis represent plant height in cm and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ........... 160
5-4 Boxplot of number of tillers for four different types of crosses studied. Small colored dots represent number of tillers for individual plants of the cross. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ............................................ 161
5-5 Histogram of number of tillers for four different crosses. X-axis represent the number of tillers and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ........... 162
5-6 Boxplot of stem diameter (mm) for four different types of crosses studied. Small colored dots represent the stem diameter (mm) for individual plants of the cross. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17................................ 163
5-7 Histogram of stem diameter (mm) for four different crosses. X-axis represent the stem diameter (mm) and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ... 164
5-8 Boxplot of leaf length (cm) for four different types of crosses studied. Small colored dots represent the leaf length (cm) for individual plants of the cross. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ............................................ 165
5-9 Histogram of leaf length (cm) for four different crosses. X-axis represent the leaf length (cm) and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ........... 166
5-10 Boxplot of leaf width (mm) for four different types of crosses studied. Small colored dots represent the leaf width (mm) for individual plants of the cross. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ............................................ 167
5-11 Histogram of leaf width (mm) for four different crosses. X-axis represent the leaf width (mm) and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ........... 168
16
5-12 Boxplot of fresh biomass per plant (kg) for four different types of crosses studied. Small colored dots represent the fresh biomass per plant (kg) for individual plants of the cross. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ............................................................................................................. 169
5-13 Histogram of fresh biomass per plant (kg) for four different crosses. X-axis represent the fresh biomass per plant (kg) and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ................................................................................................ 170
5-14 Boxplot of projected dry biomass (tons per ha) for four different types of crosses evaluated at seven months of growth. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ............................................................................................ 171
5-15 Histogram of dry biomass (tons per ha.) for four different crosses evaluated at seven months of growth. X-axis represent the dry biomass (tons per ha.) and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ....................................... 172
5-16 Coefficient of variation (%) for the different traits evaluated. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17. ........... 173
5-17 Picture of four different crosses during harvest in Citra, FL. Cross A = Tift 85 × 25-17, Cross B = 787-S5 × 25-17, Cross C = MS 787 BC4 × 25-17, Cross D = P787 × 25-17. ............................................................................................ 174
17
LIST OF ABBREVIATIONS
cms Cytoplasmic male sterility
GBS Genotyping by sequencing
LG Linkage Group
NGS Next-generation sequencing
PMN Pearl millet napiergrass
SNP Single nucleotide polymorphism
SSR Simple Sequence Repeat
TES Targeted exome sequencing
18
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
GENOMIC AND BREEDING RESOURCES TO PRODUCE SEEDED AND HIGH BIOMASS INTERSPECIFIC HYBRIDS OF NAPIERGRASS AND PEARL MILLET
By
Dev Raj Paudel
December 2018
Chair: Fredy Altpeter Cochair: Jianping Wang Major: Agronomy
Napiergrass (Cenchrus purpureus Schumach) is a promising candidate for
forage and lignocellulosic biofuel production due to its high biomass potential. However,
napiergrass is listed as an invasive species in Florida due to wind dispersed seeds.
Seed production in napiergrass is compromised by low temperatures. Therefore, late
flowering genotypes, which are not able to flower before low temperatures come in
Florida, could be utilized to improve biosafety of napiergrass. However, genetic and
genomic resources for napiergrass are limited that preclude exploiting marker assisted
selection (MAS) for crop improvement. Genetic linkage map is an important tool for
MAS. In this research, we constructed the first high-density genetic map of napiergrass
by genotyping-by-sequencing a bi-parental mapping population of 185 F1 hybrids. As a
result, we mapped 1,913 single nucleotide polymorphism (SNP) markers into 14 linkage
groups of napiergrass, spanning a length of 1,410 cM with a density of one marker per
0.73 cM.
This genetic map enabled us to identify three stable and three potential
quantitative trait loci (QTLs) controlling number of flowers and flowering time,
19
respectively in the mapping population. We also identified five candidate genes related
to flowering in close proximity to the QTLs detected.
Full characterization of germplasm collections is very critical to efficiently utilize
them in breeding programs. We used targeted enrichment sequencing to characterize
the napiergrass germplasm collection and identified 78k SNPs in the collection. We
inferred the structure of the germplasm collection and constructed its phylogeny.
Genome wide association studies revealed one significant SNP for flowering time.
Napiergrass is mostly vegetatively propagated, which makes the planting
process much complicated and labor intensive. To ameliorate this, we introgressed
cytoplasmic male sterility (cms) into elite lines of pearl millet and hybridized them with
napiergrass to produce seed derived, sterile pearl millet napiergrass (PMN) hybrids. We
evaluated the biomass yield and uniformity of PMN hybrids generated by using different
parental backgrounds. There was a tremendous variation in different biomass related
traits among the crosses. These seeded-yet-sterile PMN hybrids could have a major
impact in the forage and biofuel industry if large scale production of high-quality seeds
that give rise to high yielding progenies can be developed.
20
CHAPTER 1 BACKGROUND
Introduction
Napiergrass (Cenchrus purpureus Schumach, syn. Pennisetum purpureum
Schumach) is a tropical perennial grass that originated from Africa (Singh, Singh, and
Obeng 2013). It is commonly called merker grass, elephant grass, or Uganda grass.
Napiergrass is a monocotyledonous flowering plant that belongs to the grass family
(Poaceae) and genus Cenchrus. Recently, species of Pennisetum and Odontelytrum
were merged and designated to the unified genus Cenchrus (Chemisquy et al. 2010).
Pennisetum genus consists of a heterogeneous group of over 140 species (Brunken
1977) and it includes species with basic chromosome number of 5, 7, 8 or 9 and ploidy
ranging from diploid to octoploid (Martel et al. 1997).
Napiergrass is an important fodder crop as well as an important cellulosic energy
crop due to its high dry biomass yield compared to sorghum, maize, sugarcane,
switchgrass, johnsongrass, and Erianthus (Ra et al. 2012). The first published note on
this grass was by Mynhardt, a Hungarian missionary in Barume (then north west
Rhodesia), who sent material to Zurich Botanical Gardens in 1905. However, the name
napiergrass recognizes the contribution of Colonel Napier of Bulawayo who first wrote
to the Rhodesian Agricultural Department to make them aware of the value of this plant
(Boonman 1988). Napiergrass is a short-day plant and flowering in tropical climates
occurs from autumn through winter (Singh, Singh, and Obeng 2013). It is an open-
pollinated species. The cultivars were mainly developed by vegetative propagation of
superior clones derived from natural out-crossing material (Bhandari, Sukanya, and
Ramesh 2006; Augustin and Tcacenco 1993). Germplasm collections of napiergrass
21
are maintained in several countries in Africa, Brazil, Puerto Rico, United States,
Australia, China, Pakistan, and India (Azevedo et al. 2012; Bhandari, Sukanya, and
Ramesh 2006).
Botany
Napiergrass is a perennial, rhizomatous grass, propagated vegetatively,
commonly from joints or cuttings of the canes (Thompson 1919). It tillers vigorously,
and mature tillers have 20 or more internodes with plant height ranging from 2 to 6 m.
Its leaves are long and narrow. Reproductive part of this grass is panicles,
characterized by tawny or purplish color, sessile fascicles, and sparsely plumose
bristles (Singh, Singh, and Obeng 2013). Napiergrass is protogynous where the stigma
exerts 3-4 days, starting at the top of the inflorescence, before anthesis, followed by
anthers shedding pollen for 3-4 days. The seeds are tiny in size averaging 3.8 million
seeds/kg and seed set is poor and prone to shattering with low germination rate.
Napiergrass requires a day length of 11 h or less to flower (Hanna 2004). When it is
grown in sub-temperate regions, it produces no or very few flowers because by the time
days shorten for flowering, the temperature drops too low to support flower growth.
Deep roots and rhizomes of napiergrass bind up the soil and prevent soil erosion (Juma
2014). The plant can be harvested in the first year itself unlike other grasses like
Miscanthus, that require the first year for establishment.
Napiergrass adapts to a wide range of soil types and pH range, while maximum
growth is attained on well-drained loamy soils that have high organic matter. It is
susceptible to waterlogging and traps cereal stem borer insect pest. Cereal farmers in
western Kenya control pests like stem borer using push-pull technology by establishing
a hedge of napiergrass around their cereal plots (Khan et al. 2001).
22
Napiergrass is highly heterozygous and gives rise to a very heterogeneous
population of seedlings, that are not morphologically uniform (Juma 2014). High
morphological variability of napiergrass acts as a valuable source of genetic variation.
Mass selection and hybridization has been used to develop several varieties of
napiergrass such as Uganda hairless, Cameroon, Gold Coast, and Clone 13 (Juma
2014).
Napiergrass Ancestry
Napiergrass is an allotetraploid (2n=4x=28, A’A’BB) (Jauhar 1981) with two sub-
genomes, A’ and B. The chromosomes in the A’ sub-genome of napiergrass is
homologous to the A genome of pearl millet (Cenchrus americanus (L.) Morrone, syn.
Pennisetum glaucum, 2n=2x=14) (Jauhar 1981). The two species, napiergrass and
pearl millet, can naturally hybridize to produce hybrids that are triploids with AA’ B
genomes, thus are sterile (Singh, Singh, and Obeng 2013). Based on chromosome
pairing in triploid hybrids (2n=3x=21; AA’B), it was inferred that the two species basically
share a genome (A and A’ being very similar). Genomic in situ hybridization revealed a
high level of homeology between genomes A and A’ (Reis et al. 2014). Approximately
29% of napiergrass genomes (A’B) were hybridized by the genome of pearl millet. It
was inferred that napiergrass and pearl millet had concomitantly diverged from the
common ancestor. The origin of napiergrass occurred at the interspecific hybridization
event, combining genome A from the common ancestor with genome B whose source is
still unknown (Reis et al. 2014; Jauhar and Hanna 1998). Genome A is 24% larger than
genome A’ of napiergrass. This difference in genome size could be related to genic
duplication in pearl millet and to genomic rearrangements in napiergrass. Pearl millet
has genome DNA content of 4.72 pg while the genome DNA content of napiergrass is
23
4.60 pg (Martel et al. 1997). Both species have approximately similar DNA content (pg)
but are different with regards to the basic chromosome number. Napiergrass is a
tetraploid and has about half the DNA content (1.15 pg) of the pearl millet monoploid
genome (2.36pg). Napiergrass also has smaller chromosomes than that of pearl millet
(Martel et al. 1997; Reis et al. 2014).
Napiergrass in the United States
Napiergrass was first introduced to Florida by forage breeders based on its high-
yielding forage attributes and now occurs on the banks of canals, waterways and
roadsides. Napiergrass cultivars that flower early produce a large amount of wind
dispersed seeds that enhance its invasive potential. Napiergrass can spread to natural
vegetation (D’Antonio and Vitousek 1992; Loope, Hamann, and Stone 1988; Schofield
1989) (Figure 1-1) and is listed as an invasive species by The Florida Exotic Plant Pest
Council (FLEPPC 2011). Therefore, environmental biosafety has become one of the
major concerns in napiergrass cultivation. It is also a major concern for sugarcane
cultivation as both have similar growth habits and reproduction modes (Singh, Singh,
and Obeng 2013) making it difficult to manage in the crop. In order to enhance the
environmental biosafety of napiergrass, measures to prevent invasiveness need to be
implemented. Napiergrass usually flowers as day lengths decreases and late flowering
has been observed in a number of accessions. Seed formation in napiergrass is
compromised due to low temperatures at the end of the season. To improve the
biosafety of napiergrass, it is important to select plants for late flowering and
hybridization. Therefore, the development of cultivars that are sterile or are late
flowering is important for napiergrass management in the field to mitigate invasiveness.
24
Pearl Millet
Pearl millet is a highly cross-pollinated monocot (Rajaram et al. 2013) belonging
to Paniceae tribe within the Panicoideae subfamily of the Poaceae family (Devos and
Gale 2000; Devos et al. 2000). It is a coarse grass grown primarily for grain in Africa
and Asia and as a feed and forage in America (Gupta and Mhere 1997). Pearl millet
accounts for almost half of the global millet production and is important for food security
in some of the world’s driest and hottest areas. It is grown mostly for its ability to
produce grain under hot, dry conditions on infertile soils with low water-holding capacity
(ICRISAT and FAO 1996; Rajaram et al. 2013). Pearl millet has much larger seeds
compared to napiergrass and is easy to establish in the field via seeds. Availability of a
reference draft genome of pearl millet (Varshney et al. 2017) provides an excellent
resource that can be used extensively for genomic improvement of pearl millet and
related species.
Cytogenetic studies on napiergrass and pearl millet have classified pearl millet as
primary and napiergrass as the secondary gene pool of the genus Pennisetum (Harlan
and Wet 1971; Martel et al. 1997). Primary gene pool of the genus Pennisetum includes
three species (one cultivated and two wild species) with 2n=2x=14 chromosomes.
Secondary gene pool is represented by one allotetraploid species, Pennisetum
purpureum (2n=4x=28) with A’A’BB genomes (Martel, Ricroch, and Sarr 1996). The
primary gene pool corresponds to the traditional concept of biological species and
includes spontaneous races (wild and/or weedy) as well as cultivated races. Crossing
and gene transfer is generally easy among species in the primary gene pool to form
generally fertile hybrids. Species in the secondary gene pool can hybridize with species
in the primary gene pool, but the hybrids tend to be sterile (Harlan and Wet 1971).
25
Interspecific Hybrids of Napiergrass and Pearl Millet (PMN Hybrids)
The hybrids of pearl millet and napiergrass, commonly called PMN hybrids,
combine the superior forage quality of pearl millet and the high yielding ability of
napiergrass (Gupta and Mhere 1997; Osgood, Hanna, and Tew 1997). These hybrids
are male and female sterile due to triploidy (2n=3x=21) (Gupta and Mhere 1997) and
have reduced persistence and ratooning ability as compared to napiergrass (Cuomo,
Blouin, and Beatty 1996). The resulting hybrid with AA’B genome is morphologically
more similar to napiergrass due to larger genetic contribution (66.7% chromosomes)
and dominance of napiergrass grass B genome over the pearl millet A genome for
genetic characters such as earliness, inflorescence and leaf characteristics, and seed
size (Obok, Ova, and Iwo 2012; Gonzalez and Hanna 1984). The dominance of B
genome over the A’ genome masks genetic variability (consequently phenotypic
variability) on the A’ genome. Most of the characteristics like resistance to pests,
vigorous growth, and outstanding forage yield potential are contributed by the B
genome (Hanna 1987). PMN hybrids do not set seed and thus do not pose a threat of
uncontrolled establishment through dissemination of seeds (Hanna and Monson 1980)
and are not considered invasive (Jessup 2013). Vegetative propagation of PMN hybrids
or napiergrass is a major cost factor. Development of varieties that allow large scale
production of high quality seeds giving rise to sterile plants represents a significant step
in biomass grass breeding because field establishment using seeds will allow for
automation resulting in significant cost reduction (Osgood, Hanna, and Tew 1997).
Cytoplasmic Male Sterility (cms)
Cytoplasmic male sterility (cms) results from the interaction between organellar
and nuclear genomes that renders functional pollen production obsolete (Ram, Ram,
26
and Yadava 2007). cms plants do not produce viable pollen while being completely
female fertile. cms is important in hybrid seed production, because it helps to control
pollination for commercial production of F1 hybrid seeds (Smith and Chowdhury 1983).
cms has been extensively and cost effectively exploited in a number of agronomic and
horticultural crops including pearl millet (Havey 2004). cms in pearl millet was
discovered in the 1955 (Burton 1958) and was first released as male-sterile inbred
Tifton 23A (A1 or milo cytoplasm). Other cms sources were also later identified (Burton
and Athwal 1967) and the A4 cytoplasm was found useful for forage hybrids that do not
require male-fertility restoration (Havey 2004). The availability of the cms pearl millet
‘Tift 23A’ paved the way to commercially produce seed-propagated interspecific hybrids
that can facilitate seed harvest and allows the production of hybrids from easily drilled
seeds (J. B. Powell and Burton 1966). cms can be transferred to elite varieties to make
them sterile. To facilitate this, the method of backcrossing is suitable, where established
cultivars that are deficient in one or two specific traits can be improved through
crossing. In a backcross method, F1 hybrid is repeatedly crossed with a desirable parent
to get the desired traits (Schlegel 2009; Acquaah 2012). This has been commonly used
to transfer entire sets of chromosomes from foreign cytoplasm in order to create
cytoplasmic male-sterile genotypes (Acquaah 2012). The availability of cms lines of
pearl millet will help to develop homogenous lines of pearl millet that are male sterile.
These lines can then be used to cross with near-inbred napiergrass to produce uniform
progenies of male and female sterile PMN hybrids. While producing seeds at a
commercial scale, it should be noted that napiergrass usually flowers in November –
27
December, so seed production should be established in areas that stay frost-free until
December.
Molecular Tools Applied in Plant Breeding
As a group of important forage and biofuel species, napiergrass has received
increasing attention in recent years. Molecular studies in napiergrass have concentrated
on assessing genetic diversity, cultivar identification, origin, and relatedness using
limited genetic marker technologies (López et al. 2014; Harris-Shultz, Anderson, and
Malik 2010; Kandel et al. 2016; Dowling, Burson, and Jessup 2014; Dowling et al.
2013). Despite some advances in genetic research of napiergrass, the genetic and
molecular mechanisms of agronomic traits that must be improved for commercialization
are still poorly understood. A good genetic understanding of biomass related traits such
as emergence date, flowering time control, nutrient uptake, abiotic and biotic stress
tolerance is needed to aid genetic improvement of crops which can be facilitated by a
genetic linkage map (Ma et al. 2012) that helps to understand genome structure and to
identify trait-specific molecular markers (Cai et al. 2015). In order to accelerate
molecular tool development high-throughput genotyping is required (F. Lu et al. 2013).
Molecular markers have recently gained popularity because they are not subject to
environmental influence, are available in vast numbers, and are more objective as
compared to phenotypic markers (Kim et al. 2015). Today, plant breeders have access
to several types of DNA markers and molecular breeding tools to choose from.
A DNA marker is a specific DNA sequence on a chromosome that shows
polymorphism between individuals (Andersen and Lübberstedt 2003; Agarwal,
Shrivastava, and Padh 2008; Kumar 1999). Genomic variation leading to DNA
polymorphisms can be directly linked to differences in phenotype, used to find
28
relationships between individuals in a population, and can be used as genetic markers
(Deschamps, Llaca, and May 2012; Rafalski 2002). Most crop agronomic traits are
controlled by several loci that contribute to minor phenotypic effects and are known as
quantitative trait loci (QTL)s (Falconer, Mackay, and Frankham 1996). Since, molecular
markers are stable and detectable in all tissues and are not confounded by the
environment and other effects (Agarwal, Shrivastava, and Padh 2008), it is possible to
assign chromosomal positions to individual QTLs and to determine the effect of
individual QTLs (Kumar 1999). In order to utilize these markers in a breeding program,
these markers should be tightly linked to the desired QTL. One of the major objectives
of molecular breeding in plant species is to connect phenotype to the genotype and then
use this knowledge in MAS by making phenotypic predictions based on genotypes
(Poland and Rife 2012). Various marker technologies are currently available like
restriction fragment length polymorphism (RFLP), random amplified polymorphic DNA
(RAPD), amplified fragment length polymorphism (AFLP), simple sequence repeat
(SSR), and single nucleotide polymorphism (SNP) (Agarwal, Shrivastava, and Padh
2008; Kumar 1999).
SSRs or microsatellites are short (2-6bp) repetitive DNA sequence distribute
randomly in genomes. Since SSR markers are hyper-variable, multi-allelic, often co-
dominant, highly reproducible, and readily multiplexed, they are one of the best choices
for foreground selection in marker-assisted programs (Rajaram et al. 2013). They have
been commonly used in genotyping populations, constructing linkage maps, and
identifying QTLs for desirable traits. Linkage maps create a framework for trait mapping
and QTL analysis (Rajaram et al. 2013). With the advent of next generation sequencing
29
(NGS) platforms, genotyping a large number of progenies of mapping populations
become much efficient, which further speedup the QTL identification.
In recent years, SNP markers have gained much consideration in the breeding
community as they occur in abundance and every SNP in a single copy DNA is a
potentially useful marker (Ganal, Altmann, and Röder 2009). SNPs can be categorized
according to nucleotide substitutions either as transitions (C/T or G/A) or transversions
(C/G, A/T, C/A or T/G) (Jiang 2013). SNPs are typically identified from sequences, such
as expressed sequence tag (EST) sequence data, array analysis, amplicon sequencing,
and next generation sequencing technologies (Ganal, Altmann, and Röder 2009). For
SNP genotyping, the widely used platforms include BeadXpressTM, GoldenGateTM,
Infinitum®, GeneChipTM, GenFlex TM, SNAPshotTM, TaqMan TM, SNPstreamTM,
SNPWaveTM, iPLEX GoldTM, ARRAYTM, and KASPTM among others. (Semagn et al.
2014).
Genotyping by Sequencing (GBS)
Sequence-based genotyping methods like GBS have enabled simultaneous
marker discovery and genotyping (Poland and Rife 2012). GBS uses direct genome
sequencing to produce genotyping information utilizing high throughput and multiplexing
capacities of NGS (Paux et al. 2012; Deschamps, Llaca, and May 2012). Restriction
enzyme is utilized for complexity reduction followed by multiplex sequencing, which
requires less DNA as well as avoids random shearing and size selection (Poland,
Brown, et al. 2012). The resulting sequences allow direct SNP detection, which avoids
the marker assay development stage (Deschamps, Llaca, and May 2012) and thus
make GBS an efficient genotyping method (Elshire et al. 2011). GBS has proven to be
effective for marker discovery and trait mapping in several species like wheat (Poland,
30
Endelman, et al. 2012), switchgrass (F. Lu et al. 2013), and potato (Uitdewilligen et al.
2013). Since, GBS performs SNP discovery and genotyping simultaneously, it is
particularly of high value to understudied crops lacking reference genomes (Kim et al.
2015).
Target Enrichment Sequencing
Target Enrichment Sequencing (TES) is another NGS-enabled approach, which
focuses on genes or genomic regions of interest for sequencing. In this approach,
probes are designed to enrich capturing of the target genomic regions based on
sequence homology, which are then sequenced. TES has been used in several crops
such as barley (Russell et al. 2016), switchgrass (Grabowski et al. 2016), and peanuts
(Peng et al. 2017), to identify sequence variations but hasn't been previously applied in
napiergrass.
Quantitative Trait Loci Analysis
Quantitative trait loci (QTL) refers to a specific region on chromosomes that
harbors gene(s) controlling traits. QTL analysis is performed by estimating correlation
between phenotype data with genotype (markers) data of segregating populations
(Miles and Wayne 2008). Primary types of segregating population for QTL mapping
include F2, recombinant inbred lines (RILs), BC1 (backcross 1), double haploid lines
(DHL), near-isogenic lines (NILs) and full-sib F1 (pseudo-testcross) (Schneider 2005).
Quantitative traits, such as yield, height, flowering time, pest and disease resistance, in
plants have been mapped in genomes of several species. QTLs are valuable
information to develop markers linked to traits of interests for MAS in molecular
breeding programs. Genetic variability for flowering time exists in napiergrass (Sinche
2013). Therefore, identifying flowering time related QTLs will help in identifying markers
31
linked to late flowering trait, an important trait to reduce invasiveness. Sexual
hybridization of genetically distant parents and selection of late flowering, high yielding
accessions would increase the biofuel yield and enhance the biosafety of napiergrass.
Biomass yield
In a comparison of major energy crops for ethanol production, napiergrass
showed the highest dry biomass yield than sorghum, maize, sugarcane, switchgrass,
johnsongrass, and Erianthus (Ra et al. 2012). Napiergrass yields up to 84.8 Mg ha-1 yr-1
have been obtained in Puerto Rico (Vicente-Chandler, Silva, and Figarella 1959), while
in Florida, yields ranged between 35-45 Mg ha-1 yr-1 (Woodard and Sollenberger 2012;
Erickson et al. 2012). Biomass yield is a complex trait and studies have been conducted
to identify QTLs for various components affecting yield. In sorghum, QTLs for yield
related traits such as plant height, tiller number, leaf length, leaf width, stem diameter,
and flowering time have been identified (Hart et al. 2001; Murray et al. 2008; Xiao-ping
et al. 2011). Similarly, QTLs for leaf yield, stem yield, plant height and flowering time
have also been identified in Miscanthus sinensis (Gifford et al. 2015; Atienza et al.
2003). In Miscanthus, QTLs for yield co-segregated with other traits like number of
tillers, leaf area, leaf length, and leaf width (Gifford et al. 2015). In napiergrass, biomass
yield showed high correlations with number of tillers, plant height, and stem diameter
(Sinche 2013).
Flowering Time
Floral transition is the switch from vegetative growth to reproductive growth in
plants and it primarily determines flowering time. Flowering time is a key factor in plant
adaptation and is linked to various attributes like plant height, yield, number of leaves,
etc. (Durand et al. 2012). In many species, flowering is induced in response to the
32
length of periods of lightness and darkness associated with day and night length. These
species are categorized as short-day, long-day, intermediate-day, or day-neutral based
on their day length requirement (Schlegel 2009; Bastow and Dean 2002). Plants in
which flowering is favored by day lengths shorter than the critical and corresponding
long nights are called short-day plants (eg. Glycine max, Oryza sativa, and Zea mays).
The plants in which flowering is initiated when the day length is longer than the critical
are called long-day plants (eg. Hordeum vulgare, Triticum aestivum, and Solanum
tuberosum) (Garner 1933; Schlegel 2009; Bastow and Dean 2002). Napiergrass
belongs to the short-day plants (Osgood, Hanna, and Tew 1997; Singh, Singh, and
Obeng 2013).
Many studies have detected QTLs related to flowering time or earliness in
various crops. The genetic control of flowering time is in general quantitative in nature
(H. Lu et al. 2014). For example, in rice, 15 QTLs were associated with days to
flowering (Maheswaran et al. 2000), and in tomato, three QTLs related to earliness were
identified that were associated with flowering time, fruit set time, and ripening time
(Lindhout et al. 1994). However, a previous study in maize showed that flowering time
was a complex trait and no major QTL should be expected (Buckler et al. 2009). QTLs
for flowering related traits like 50% anthesis and heading date colocalized with other
QTLs for number of tillers, tiller diameter, leaf width, and leaf area in Miscanthus
(Gifford et al. 2015). In orchardgrass (Dactylis glomerata L)., 11 QTLs for heading date
and flowering time were found to be distributed on three linkage groups where
candidate genes such as hd1 and VRN1 were annotated (Zhao et al. 2016).
33
Flowering time in napiergrass is of prime importance as it is related to biosafety
and biomass quality. Late flowering napiergrass lines serve as potential bio-safe
biofuels as their flowering is compromised by low temperatures which usually occurs in
early December in Florida. Late flowering cultivars, therefore have less potential for
invasiveness than early flowering genotypes and may produce higher yields due to a
longer period of vegetative growth. Extensive phenotypic variation in flowering time is
an indication that flowering time is quantitative in nature and several genes might be
controlling the trait. Identifying QTLs for flowering time in napiergrass will help to
shorten the breeding cycle by the successful utilization of MAS. A biparental mapping
population from a cross between early-flowering and late-flowering lines will generate
progenies segregating for flowering time and this population can be genotyped in order
to identify QTLs for flowering time.
Genes Related to Flowering
Flowering time has been studied in some grass species and genomic studies
have identified a number of genes involved in flowering. Potential candidates can be
FRI, LEAFY, CO, DNF, MADS-box, and RID1 that have some roles in flowering
regulation in other species like Arabidopsis, wheat, maize, rice (Lee, Bleecker, and
Amasino 1993; Wuxing Li et al. 2013; Suárez-López et al. 2001; Morris et al. 2010; C.
Wu et al. 2008). For example, high levels of proteins encoded by FLOWERING LOCUS
T (FT) are correlated with early flowering and the lack of these causes late flowering
(Samach 2012). Major genes involved in photoperiod of flowering are highly conserved
between rice and Arabidopsis (C. Wu et al. 2008). For example, Hd1 QTL in rice that
promotes heading under short-day conditions corresponds to a gene homolog of
CONSTANS in Arabidopsis (Yano et al. 2000); Hd17 corresponds to a homolog of
34
Arabidopsis ELF3 (EARLY FLOWERING 3) (Matsubara et al. 2012; Matsubara et al.
2008); and Hd3a encodes a protein that is closely related to Arabidopsis FT (Kojima et
al. 2002). Mining of genome sequences that are available for several grass species for
flowering related genes and their characterization can identify candidate genes in
napiergrass. These candidate genes can be used for screening germplasm collections
for identification of haplotypes that confer late flowering. In addition to this, integrating
genomics with conventional breeding will help to shorten the breeding cycle for
selection.
Objectives
The overall objectives of this research are to identify genetic components (QTL
and candidate genes) underlying flowering time in napiergrass and to develop male and
female sterile PMN hybrids that are yielding high amounts of biomass and display
uniform seed progenies.
The substantial variation in flowering time in napiergrass can be utilized for
breeding late flowering varieties. This helps to make napiergrass a bio safe genotype for
biomass production in northern Florida where a freeze event typically occurs before
flowering of these late lines. The identification of QTLs controlling flowering time will
make the pre-selection for late flowering lines in breeding materials highly efficient.
Identifying polymorphisms within flowering genes in napiergrass germplasm collection
helps to reveal allelic variation in the germplasm. Development of a male sterile and
homozygous line of pearl millet will prevent its self-pollination, thus facilitating easy
crossing with napiergrass under field conditions. These high-biomass yielding male
sterile lines can then be used for hybridization with napiergrass.
35
The specific objectives of this research were to 1) construct genetic maps of
napiergrass, 2) identify QTLs for flowering time in napiergrass, 3) evaluate sequence
variations in a napiergrass germplasm collection, and 4) develop and evaluate male and
female sterile PMN hybrids.
36
Figure 1-1. Distribution of napiergrass. Black dots represent local napiergrass and red dots represent napiergrass listed as invasive species
-50
0
50
-100 0 100 200
Longitude
La
titu
de
37
CHAPTER 2 SURVEYING THE GENOME AND CONSTRUCTING A HIGH-DENSITY GENETIC
MAP OF NAPIERGRASS (CENCHRUS PURPUREUS SCHUMACH.)
Napiergrass (Cenchrus purpureus Schumach.) is a tropical forage grass and a
promising lignocellulosic biofuel feedstock due to its high biomass yield, persistence,
and nutritive value. However, its utilization for breeding has lagged behind other crops
due to limited genetic and genomic resources. In this study, next-generation sequencing
was first used to survey the genome of napiergrass. Napiergrass sequences displayed
high synteny to the pearl millet genome and showed expansions in the pearl millet
genome along with genomic rearrangements between the two genomes. An average
repeat content of 27.5% was observed in napiergrass including 5,339 simple sequence
repeats (SSRs). Furthermore, to construct a high-density genetic map of napiergrass,
genotyping-by-sequencing (GBS) was employed in a bi-parental population of 185 F1
hybrids. A total of 512 million high quality reads were generated and 287,093 SNPs
were called by using multiple de-novo and reference-based SNP callers. Single dose
SNPs were used to construct the first high-density linkage map that resulted in 1,913
SNPs mapped to 14 linkage groups, spanning a length of 1,410 cM and a density of 1
marker per 0.73 cM. This map can be used for many further genetic and genomic
studies in napiergrass and related species.
This chapter was published in Scientific Reports and is licensed under a Creative Commons Attribution 4.0 International License.
Paudel, D., Kannan, B., Yang, X., Harris-Shultz, K., Thudi, M., Varshney, R.K., Altpeter, F. and Wang, J., 2018. Surveying the genome and constructing a high-density genetic map of napiergrass (Cenchrus purpureus Schumach). Scientific reports, 8(1), p.14419.
38
Introduction
Napiergrass (Cenchrus purpureus Schumach., syn. Pennisetum purpureum
Schumach.), also known as elephant grass, is a tropical perennial grass native to
eastern and central Africa. It is cultivated primarily for forage and widely used by
smallholder dairy farmers due to its high growth rate, leaf nutritive value, perennial
nature, persistence, ease of propagation, and broad adaptation (Bhandari, Sukanya,
and Ramesh 2006; Farrell, Simons, and Hillocks 2002; Singh, Singh, and Obeng 2013;
Chemisquy et al. 2010). As a C4 grass species, napiergrass is a promising candidate
feedstock for biofuel production due to its superior yield of biomass (Ra et al. 2012;
Anderson, Casler, and Baldwin 2008; Somerville et al. 2010). Napiergrass cultivars are
typically developed from natural out-crossings (Bhandari, Sukanya, and Ramesh 2006;
Augustin and Tcacenco 1993). It is an allotetraploid (2n = 4x = 28, A’A’BB) (Jauhar
1981) with an average amount of DNA per G1 nucleus of 5.78 pg (M. G. Taylor and
Vasil 1987). The chromosomes in the A’ genome of napiergrass are believed to be
homologous to the A genome of pearl millet (Pennisetum glaucum, 2n = 2x = 14, AA)
(Jauhar 1981). Pearl millet and napiergrass form a monophyletic group (Martel et al.
1997) and were initially classified as primary and secondary gene pool of the genus
Pennisetum, respectively (Harlan and Wet 1971; Martel et al. 1997). Recently, species
of Pennisetum and Odontelytrum were transferred to the unified genus Cenchrus
(Chemisquy et al. 2010). Pearl millet and napiergrass can hybridize to produce hybrids
called kinggrass (Dowling, Burson, and Jessup 2014) or Pearl Millet-Napiergrass (PMN)
hybrids (Hanna and Monson 1980; Burton 1944; Muldoon and Pearson 1979). These
hybrids are sterile due to triploidy (2n = 3x = 21, AA’B genome) (Gupta and Mhere
1997), thus preventing the unintended spreading into natural areas or other cropping
39
systems by wind dispersed seeds. Some PMN hybrids show high heterosis for biomass
yield and forage quality while the perennial, persistent nature is often reduced
compared to napiergrass (Singh, Singh, and Obeng 2013).
The targeted improvement of napiergrass includes identification of agronomically
superior genotypes and studies assessing genetic diversity and relatedness using
random amplification of polymorphic DNA (RAPD), amplified fragment length
polymorphism (AFLP), isozymes, and simple sequence repeats (SSRs) developed for
other species like pearl millet and buffelgrass (Pennisetum ciliare) (Lowe et al. 2003;
Bhandari, Sukanya, and Ramesh 2006; Harris-Shultz, Anderson, and Malik 2010;
Kandel et al. 2016; Dowling et al. 2013; Dowling, Burson, and Jessup 2014; López et al.
2014; Smith et al. 1993). Other than these, genetic information on napiergrass is very
meager (Smith et al. 1993). A genetic map is lacking and molecular tools are not yet
deployed in napiergrass breeding programs (Dowling et al. 2013; Negawo et al. 2017).
Development of molecular markers for detection and utilization of DNA polymorphisms
will help to understand the molecular basis of various agronomic traits (Song et al.
2015). Molecular breeding for yield components, flowering date, nutrient uptake, abiotic
and biotic stress tolerance will accelerate genetic improvement of napiergrass. This can
be greatly facilitated by having access to marker resources like SSR, single nucleotide
polymorphisms (SNPs), and genetic linkage maps. SSRs as molecular markers are very
advantageous because they are locus specific, multi-allelic, co-dominant, and easy to
detect by polymerase chain reaction (PCR) (W. Powell, Machray, and Provan 1996;
Kannan et al. 2014). SNP markers have gained increasing consideration in molecular
breeding and linkage map construction as they occur in a large number and high
40
density (Ganal, Altmann, and Röder 2009). Access to these resources will support
marker-assisted selection (MAS) by making phenotypic predictions based on the
genotype (Poland and Rife 2012).
Recently, next generation sequencing (NGS) technology has simplified linkage
map construction by using high throughput genotyping-by-sequencing (GBS), which
allows simultaneous SNP discovery and genotyping across the whole genome of the
population of interest (Elshire et al. 2011; Poland and Rife 2012; Deschamps, Llaca,
and May 2012). GBS has been effective for marker discovery, genetic mapping,
quantitative trait locus (QTL) analysis, population genetics, and comparative genomics
studies in several diploid species, and has recently gained popularity in polyploid
species such as wheat (Triticum aestivum) (Poland, Endelman, et al. 2012), switchgrass
(Panicum virgatum) (F. Lu et al. 2013), potato (Solanum tuberosum) (Uitdewilligen et al.
2013), and sugarcane (Saccharum spp.) (Yang, Sood, et al. 2017) among others.
However, the presence of highly similar homeologous copies of two genomes in
allopolyploid species complicates SNP detection which relies on delineating true allelic
SNPs from homeologous SNPs because sequences from homeologous loci mimic
allelic SNPs and can introduce false-positives. Distinguishing allelic SNPs from
homeologous SNPs relies on the use of high-stringency sequence read alignment,
specifically uniquely aligned reads (Clevenger et al. 2015). Despite challenges of using
GBS for genotyping of polyploid species, genetic mapping without a reference genome
has been carried out in switchgrass by defining linkage groups with the modulated
modularity clustering (MMC) method (Stone and Ayroles 2009) referring to the genome
of foxtail millet (Setaria italica) (F. Lu et al. 2013). A genetic map of wheat was
41
constructed by using the bin-mapping procedure with homozygous genotypes of a
double-haploid population (Poland, Brown, et al. 2012). Each program for calling
variants utilizes different models or algorithms to identify potential polymorphisms,
therefore, multiple software programs need to be evaluated in order to identify the best
SNP caller for polyploids (Clevenger et al. 2015).
Linkage maps are important tools for map-based cloning, marker-assisted
breeding, QTL identification, genome organization, and comparative genomics of
important species. A number of linkage maps have been constructed for several
grasses including pearl millet (Punnuri et al. 2016). However, so far, napiergrass SSR
markers, genetic linkage map, or reference genome assembly are lacking. The purpose
of this study was to survey the napiergrass genome and to construct a high-density
genetic linkage map. Here, for the first time, we have surveyed whole genome
sequences in napiergrass, developed SSR markers, and constructed high-density
genetic map of napiergrass to investigate its genomic and genetic architecture.
Methods
Napiergrass Genome Survey
The genomic DNA of napiergrass cultivars Merkeron and UF1 was sequenced
using Illumina Genome Analyzer and 454 GS-FLX. For Illumina reads, reads that
contained more than 50% low-quality bases (Q20) were removed and adaptor
sequences were trimmed. Quality and adapter trimming of 454 reads was done using
default Newbler v2.8 (454 Life Sciences, Roche, Branford, CT) settings. Illumina reads
were assembled using ABySS/1.3.4 (Simpson et al. 2009) with kmer size ranging from
25 to 60 at intervals of 5. The 454 reads were assembled using Newbler v2.8 (454 Life
Sciences, Roche, Branford, CT) with default parameters. The assemblies were
42
completed using CAP3 (Xiaoqiu Huang and Madan 1999). The largest 10 contigs of the
assembly were selected for further analysis. Repeats on these contigs were masked
using a comprehensive public repeat database compiled from TIGR plant repeats
(http://plantrepeats.plantbiology.msu.edu/), Plant miniature inverted-repeat transposable
elements (P-MITE) database (http://pmite.hzau.edu.cn/django/mite/), MIPS Repeat
Element Database (http://mips.helmholtz-muenchen.de/plant/recat/), and Repbase from
RepeatMasker software (http://www.repeatmasker.org/). Unique repeats were extracted
from this database by removing redundant repeats with 98% identity using CD-HIT/4.6
(Garrison and Marth 2012). SNPs and indels were called using FreeBayes/0.9.15
(Garrison and Marth 2012) excluding alleles with depth less than 20. The annotation of
SNPs was performed using SnpEff/4.0 (http://snpeff.sourceforge.net/) (Cingolani et al.
2012). In order to identify sequence similarity among the two genomes, clean reads
from Illumina and 454 were aligned to the pearl millet genome v1 (Varshney et al. 2017)
using bowtie2/2.2.5.
SSR Identification and Marker Development
The napiergrass assembly was used to identify SSR markers that
contained repeat motifs ranging in length from 1 to 6 bp. The minimum number of
repeats was 10 for Mono-, 6 for Di-and 5 for Tri-, Tetra-, Penta- and Hexa-. SSRs were
analyzed based on their types, number of repeats, and percentage frequency of
occurrences of each SSR motif. SSRs in napiergrass were detected using
MIcroSAtellite identification tool (MISA) (Thiel et al. 2003) and primers were developed
using primer3 software (Rozen and Skaletsky 2000). SSR search results were input into
scripts p3_in.pl and p3_out.pl in order to identify SSR primer pairs for napiergrass.
43
Plant Materials and DNA Extraction
A mapping population of 185 F1 hybrid progenies were developed from a cross
between two napiergrass accessions (N122 and N190) described previously (Sinche
2013; Sinche et al. 2018). The 185 F1 hybrids were planted in the field at the Plant
Science Research and Education Unit (PSREU), Citra, Florida, along with the parental
accessions.
Young and healthy leaf tissues were harvested from each individual of the
mapping population. DNA extraction was done following the protocol described
previously (Dellaporta, Wood, and Hicks 1983). The extracted DNA samples were run
on a 2% agarose gel to check the quality and quantified with PicoGreen to meet the
requirements of GBS. 185 F1 plants that were confirmed to be true hybrids using SSR
markers (Sinche et al. 2018) were selected for GBS.
Genotyping-by-sequencing
GBS data was generated at the Institute of Biotechnology, Cornell University as
described previously (Elshire et al. 2011). Briefly, DNA samples were digested with the
restriction enzyme PstI followed by ligation of adapters, that consisted of Illumina
sequencing primers and barcode adapters, to the DNA fragment ends. After ligation, 95
samples were combined into a pool and PCR amplification was performed to create a
GBS library and sequenced on Illumina HiSeq 2000.
Comparative Genomics
Unique tags of napiergrass from TASSEL de-novo UNEAK/3.0 (Glaubitz et al.
2014) pipeline were used for comparative genomic analysis. CD-HIT/4.6.4 (Weizhong Li
and Godzik 2006) was used to cluster the tags. Genomes of rice (Osativa_323_v7.0),
Brachypodium (Bdistachyon_314_v3.0), maize (Zmays_284_AGPv3), sorghum
44
(Sbicolor_313_v3.0), foxtail millet (Sitalica_312_v2), switchgrass (Pvirgatum_273_v1.0),
wheat (Taestivum_296_v2), Arabidopsis (Athaliana_167_TAIR9) were downloaded from
Phytozome v11 (https://phytozome.jgi.doe.gov/pz/portal.html). The barley genome
(ASM32608v1) was downloaded from Ensembl (http://www. ensembl.org). We used
BLASTN (BLAST v2.5.0) with the default settings and an e-value cutoff of 1×10-8 to
blast the uniquely clustered tags of napiergrass against different genomes in order to
find the percentage similarity of napiergrass reads among the various grass species.
Tags of 64 bp with > 80% identity and alignment length > 50 bp to the respective
genomes were counted as a hit.
Sequence Analysis and SNP Calling
Raw data processing and SNP identification was performed using both de novo
and reference-based approaches. Common software capable of calling SNPs de novo
used in this research were TASSEL/3.0 UNEAK (Glaubitz et al. 2014), Stacks/1.24
(Catchen et al. 2013), and GBS-SNP-CROP 1.1 (Melo, Bartaula, and Hale 2016). For
the reference based approach, pearl millet reference genome v1 (Varshney et al. 2017)
was used. The reference genome consists of seven pseudomolecules. Six different
reference based pipelines were evaluated to call SNPs viz., TASSEL 4.3 (Glaubitz et al.
2014), Stacks 1.24 (Catchen et al. 2011), GBS-SNP-CROP 1.1 (Melo, Bartaula, and
Hale 2016), SAMtools 1.2 mpileup (H. Li et al. 2009), FreeBayes 0.9.21 (Garrison and
Marth 2012), and GATK 3.3 (McKenna et al. 2010). Parameters used in each software
are provided in Table 2-1. Sequence variants called were filtered with a minimum depth
of 48 per sample.
45
Linkage Map Construction
QC-filtered SNPs were further filtered by the following standards for map
construction: 1) markers must be genotyped in at least 180 individuals; 2) Individuals
with over 10% missing data were discarded; and 3) Redundant markers were removed
by standard of similarity = 1. For each parental map construction, only single dose
markers were used. Markers segregating at a distorted Mendelian ratio (expected ratio
for ‘lmxll’ type and ‘nnxnp’ type is 1: 1, χ2 test, 0.001 < P < 0.05) were marked. The
single dose markers from the maternal and paternal parent were analyzed separately
using JoinMap 4.1 (van Ooijen 2006) and outcross pollinated family (CP) was selected
as the population type. Markers that were heterozygous in N122 and homozygous in
N190 (‘lmxll’ type) were selected to build N122 linkage groups. Markers that were
heterozygous in N190 and homozygous in N122 (‘nnxnp’ type) were selected to build
N190 linkage groups. The linkage groups were built using regression mapping
algorithm, with a minimum logarithm of odds (LOD) value at 20, and a maximum
recombination frequency at 0.40. Marker positioning calculation was performed with a
goodness-of-fit jump at 5, followed by a “ripple” procedure (value =1). Map distances
were estimated using the Kosambi mapping function. Genetic distance between SDR
markers were corrected using DistortedMap (S. Q. Xie, Feng, and Zhang 2014).
Linkage maps were drawn with MapChart (Voorrips 2002). For integrated map
construction, markers that were heterozygous in both parents (‘hkxhk’ type) were
selected to build combined linkage groups. Markers segregating at distorted Mendelian
ratio (expected ratio for ‘hkxhk’ type is 1:2:1, χ2 test, 0.001< P < 0.05) were marked. The
retained markers were then added with the markers from male and female parents to
construct a combined map. The linkage groups were built using regression mapping
46
algorithm, with a minimum logarithm of odds (LOD) value at 20, and a maximum
recombination frequency of 0.40. Other parameters were the same with linkage map
construction above. Regions showing segregation distortion (0.001< P <0.05) with more
than three adjacent loci were marked as SDR regions (Paillard et al. 2003; Z. Zhang et
al. 2016).
Comparison Between Napiergrass and Pearl Millet Genome
Consensus sequence of mapped markers from TASSEL de-novo UNEAK were
used to compare with the reference genome of pearl millet with same parameters
(BLASTN defaults with an e-value cutoff of 1x10-8). Markers that showed significant hits
to the genome sequence and /or gene models of pearl millet with > 80% identity and
alignment length > 50 bp were extracted and used for comparative genomics study. A
circos plot was drawn using circos/0.69-2 (Krzywinski et al. 2009).
Results
Napiergrass Genome Survey
Approximately 211 million raw reads from Illumina and 97 thousand raw reads
from 454 sequencing were subjected to a sequence quality check. After filtration and
trimming, 161 million clean Illumina reads and 96,000 clean 454 reads were aligned to
the pearl millet genome v1 (Varshney et al. 2017). A total of 62.5 million (38.8%) reads
were able to align with the pearl millet genome. Polymorphisms were detected between
the napiergrass and pearl millet aligned reads, of which 619,708 SNPs and 24,135
indels were identified. Most of the sequence variations (58.7% SNP) were in intergenic
regions (Figure 2-1). The clean reads were assembled into 113,789 contigs with a total
size of 44.5 Mbp and a N50 of 435 bp and a GC content 43.45%. The largest 10 contigs
of the sequence assembly, which ranged from 8,506 to 25,329 bp in length, were
47
selected as representative napiergrass genome fragments. The repeat content of the 10
longest contigs ranged from 5% to 90% with an average of 27.5% and a total of 164
repetitive elements (Table 2-2). Two contigs had no hits in the pearl millet genome due
to a high repeat content (over 80%). The rest of the contigs had one or more large hits
(>500 bp) to the pearl millet genome. The micro-synteny showed mostly collinear
relationship between the genomes of the two species (Figure 2-2). However,
chromosome inverted duplications were also observed in the pearl millet genome
(Figure 2-3), indicating that the chromosome rearrangement occurred after the
speciation of napiergrass and pearl millet. The length of stringently (>500 bp and >80%
sequence similarity) aligned regions accounted for 36.3% of the examined contig
sequences of napiergrass (Table 2-3). The total length of the alignment was 25.1%
higher in pearl millet than in napiergrass aligned regions.
From the assembled napiergrass survey sequences, 5,339 SSRs were identified.
Mono- type repeats were most common in napiergrass, followed by Tri-, Di- and Tetra-
type repeats (Table 2-4). From these identified SSRs, 1,926 were successfully used for
primer design (Table 2-5). All of the primer sequences aligned to the assembly of
napiergrass and 89% of the primers were uniquely aligned. On the other hand, the
overall alignment rate of the primers with pearl millet genome v1 (Varshney et al. 2017)
was 31% with 15% uniquely aligned. These SSR primers will undoubtedly serve as an
abundant resource for molecular markers in napiergrass.
Genotyping-by-sequencing
To construct the linkage map for napiergrass, an F1 bi-parental population was
developed, which consisted of 185 true hybrid individuals (Sinche et al. 2018). These
hybrids were subjected to GBS. A total of 549 million raw reads were generated. After
48
trimming and filtering, 512 million high quality reads were retained. The average number
of reads per sample was 2.6 million and ranged from 44 thousand to 5.4 million reads
per sample. In silico digestion of the pearl millet genome v1 (Varshney et al. 2017) with
PstI yielded DNA fragments in the range of 170-350 bp, which suggest that an
estimated average depth for the mapping population was 36.5 reads per locus per
sample (Figure 2-4), which should allow us to call the SNPs confidently at most of the
loci.
A total of 695,602 unique tags were identified from the clean reads generated
from the mapping population by using the TASSEL de-novo UNEAK pipeline. These
tags were further clustered into 182,934 non-redundant tags by CD-HIT. To examine
the sequence similarity between napiergrass and other grass species, we aligned the
non-redundant tags of napiergrass against several grass species with complete genome
sequences including rice (Oryza sativa) (Osativa_323_v7.0), Brachypodium
(Bdistachyon_314_v3.0), maize (Zmays_284_AGPv3), sorghum (Sbicolor_313_v3.0),
foxtail millet (Sitalica_312_v2), switchgrass (Pvirgatum_273_v1.0), wheat
(Taestivum_296_v2), pearl millet v1(Varshney et al. 2017) , and barley (ASM32608v1),
with Arabidopsis (Athaliana_167_TAIR9) as an outgroup control. The results showed
that the percentage of napiergrass sequence tags aligned to these grass species
ranged from 2.6% to 37.9% for barley and pearl millet genome, respectively (Table 2-6),
indicating a relatively close relationship between napiergrass and pearl millet.
49
SNP Calling by Various SNP Callers
Three de-novo SNP calling pipelines, TASSEL-UNEAK, Stacks, and GBS-SNP-
CROP identified 10,799, 6,871, and 4,521 SNPs, respectively. Reference based
pipelines were also applied by using pearl millet v1 (Varshney et al. 2017) as the
reference genome. However, the alignment rate was relatively low due to the
differences between the napiergrass and pearl millet genomes. The percentage of clean
reads aligned to the pearl millet genome using Bowtie2 ranged from 5.60% to 44.62%
with an average of 39.68%. Two samples had a small number of sequences (< 10% of
the average number of sequences per sample) and also the lowest percentage of
uniquely mapped reads (Table 2-7, Figure 2-5). Therefore, these samples were
removed from linkage map construction. Six different reference-based pipelines were
employed to call SNPs viz., TASSEL 4.3 (Glaubitz et al. 2014), Stacks 1.24 (Catchen et
al. 2011), GBS-SNP-CROP (Melo, Bartaula, and Hale 2016), SAMtools 1.2 mpileup (H.
Li et al. 2009), FreeBayes 0.9.21 (Garrison and Marth 2012), and GATK 3.3 (McKenna
et al. 2010). TASSEL 4.3, Stacks, and SAMtools identified 7,326, 4,920, 27,082 SNPs,
respectively in the mapping population, whereas FreeBayes, GBS-SNP-CROP, and
GATK that can handle ploidy identified 25,193, 2,906 and 197,475 SNPs, respectively.
The six reference-based SNP callers concordantly called only 11 SNPs (Figure 2-6,
only five programs are shown in figure due to Venn-diagram display limitations) and
207,391 non-redundant SNPs.
Genetic Linkage Map Construction
From a total of 549,944 SNPs called by both reference based and de-novo
pipelines, 287,093 SNPs were filtered for further analysis. Out of these, a total of 18,286
single-dose SNPs were genotyped in more than 180 progenies. Three individuals with
50
more than 10% missing sites were removed from further analysis. For linkage map
construction of each parental line, only the SNPs showing heterozygous in one parent
but homozygous in the other parent were selected. A total of 3,276 loci were
heterozygous in female parent but homozygous in male parent and segregated with an
expected ratio of 1:1 in the population, thus can be used for female parent linkage map
construction. Similarly, 3,417 loci were heterozygous in male parent but homozygous in
female parent and segregated with an expected ratio of 1:1 in the population, thus can
be used for male parent linkage map construction. For the female parental line, a total
of 1,606 SNPs were grouped and 899 loci were mapped on 14 linkage groups with a
total length of 1,555.17 cM averaging 1 marker every 1.72 cM (Figure 2-7). Inclusion of
segregation distorted (SD) markers increased the genetic distance of the female map by
28.13%. For the male parent, a total of 1,509 markers were grouped into 14 linkage
groups and 1,073 markers were mapped onto these 14 linkage groups with a total
length of 1,939.19 cM averaging 1 marker every 1.80 cM (Figure 2-8). Inclusion of SD
markers increased the total genetic distance of the male map by 38.41%.
A combined linkage map containing markers that segregated from both female
and male parents was constructed, which can facilitate future QTL mapping of the
population. To construct a combined linkage map, the markers showing heterozygous
on both parents in addition to male-parent heterozygous and female-parent
heterozygous markers were used. Therefore, a parent-averaged combined map was
constructed by using 378 heterozygous markers for both parents that segregated in a
1:2:1 ratio in the population, in combination with 3,417 male-parent heterozygous and
3,276 female-parent heterozygous markers. In total, 4,058 markers were grouped into
51
14 linkage groups out of which 1,913 markers were mapped. The final composite
linkage map spanned a length of 1,410.10 cM with an average of 0.73 cM between
markers. The largest linkage group was Linkage group 2 (LG 2), which spanned 142.40
cM and contained 170 markers (Table 2-8). Length of each linkage group ranged from
70.18 cM to 142.40 cM and density ranged from 0.88 to 1.77 markers per cM (Figure 2-
9, Table 2-8, Figure 2-10). Results of the χ2 test indicated that 114 (6.06%) of the 1,879
markers showed significant segregation distortion (0.001<P<0.05) on the combined
map. These distorted markers showed clustered distribution on three segregation
distortion regions (SDRs) in linkage groups LG07 and LG08 (Fig 2-9).
Among the different reference-based SNP callers, GATK called the highest
number of SNPs (197,475) followed by SAMtools and FreeBayes (Table 2-9). After
accounting for segregation ratio and missing data, SAMtools retained the largest
number of SNPs followed by TASSEL de-novo UNEAK. However, when considering the
total number of markers mapped on the combined linkage groups, TASSEL de-novo
UNEAK showed the highest percentage of SNPs mapped followed by Stacks (Table 2-
9).
Comparison Between Genomes of Napiergrass and Pearl Millet
Sequence tags of the markers that mapped on napiergrass linkage groups were
extracted and compared to the pearl millet genome. Among the 1,156 TASSEL de-novo
UNEAK tags positioned on the combined map, 663 were found to have significant
sequence similarities to the genome sequence of pearl millet. Considerable collinearity
was observed between the napiergrass and pearl millet genomes (Figure 2-11). For
each pearl millet pseudomolecule, two corresponding regions in the linkage groups
(LGs) of napiergrass genome were identified (Figure 2-11, Figure 2-12). However, some
52
pearl millet genomic regions had more than two corresponding regions on napiergrass
genome. For example, pseudomolecule 3 of pearl millet had regions corresponding to
three linkage groups LG03, LG12, and LG14 of napiergrass indicating possible
chromosomal rearrangement between the two species after speciation (Figure 2-11,
Figure 2-12).
Discussion
Despite its importance as a forage grass and its enormous potential as a biofuel
crop, molecular, genetic, and genomic studies have been severely limited in
napiergrass. Currently, there was no equivalent genome sequence in the public domain
to be used as a reference for napiergrass. In this study, an initial comparison between
the napiergrass survey sequences to 10 available grass genomes revealed that
napiergrass genomic sequences had the highest similarity with the pearl millet genome,
which could be explained by the presence of the A’A’ genome of napiergrass that is
homologous to the AA genome of pearl millet. Consequently, in this study we have
utilized pearl millet genome v1 (Varshney et al. 2017) as a reference for SNP calling
and also performed de-novo SNP calling without a reference genome. A total of 38.8%
of the napiergrass reads aligned to the pearl millet genome using Bowtie 2, which
performed better over BWA, another popular aligner (Langmead and Salzberg 2012;
Yang, Song, et al. 2017). The large portion of unaligned reads might be from the B
genome or the divergent chromosome regions of A genome between the two species.
From the genome survey comparison, the total length of all the alignments of
napiergrass reads was 25.1% longer in pearl millet indicating genic duplication or
expansion in pearl millet and genomic rearrangements between the two species during
evolution from their ancestral genome. This is consistent with a previously reported
53
genomic in situ hybridization, which verified that the pearl millet genome A was 24%
larger compared to the chromosomes of genome A’ of napiergrass (Reis et al. 2014).
For the 10 longest contigs in our assembly, average repeat content (27%) was lower
than reported from other grasses including sorghum (61%) (Paterson et al. 2009),
maize (85%) (Schnable et al. 2009), foxtail millet (46%) (G. Zhang et al. 2012), rice
(43.3%) (J. Yu et al. 2002), and pearl millet (77%) (Varshney et al. 2017). Low repeat
content in napiergrass could be attributed to the loss of genomic sequences after
hybridization. Rearrangements and loss of genomic sequences are common events
after hybridization (Kellis, Birren, and Lander 2004). Similar to other plant genomes,
long-terminal repeat (LTR) retrotransposons comprised the most abundant class (62.19
%) of repeats (Table 2-2). Significant relationships between napiergrass, pearl millet,
and P. squamulatum suggested their common origin and it was inferred that
napiergrass and pearl millet had concomitantly diverged from a common ancestor (Reis
et al. 2014; Martel et al. 2004; Martel et al. 1997) and the origin of napiergrass occurred
at the interspecific hybridization event, by combining genome A of the ancestor with
genome B of a still unknown second ancestor (Reis et al. 2014). Our study showed that
the napiergrass genome had considerable microcolinearity with the pearl millet genome
showing evidence of their relatedness and shared ancestry. Chromosome inverted
duplications on pseudomolecule 3 of pearl millet showed possible rearrangement after
speciation of napiergrass and pearl millet. Two corresponding regions on the
napiergrass linkage groups for each pearl millet chromosome corroborate the
hypothesis that these two genomes evolved from a common ancestor.
54
We developed a limited genomic assembly of napiergrass based on Illumina and
454 sequences. Nearly two thousand SSR markers were developed, which could be
immediately useful for applications in napiergrass breeding and genetics. With the
advancement of NGS, high throughput NGS-enabled genotyping technologies are
becoming readily accessible. Yet, SSR markers remain as a popular tool for genetic
studies, variety identification, monitoring of seed purity, and hybrid quality. They are
particularly important in laboratories which have limited resources and lack access to
NGS facilities or bioinformatic expertise. To our knowledge, this is the first study in
napiergrass where SSR markers were developed based on napiergrass genome
survey.
A genetic linkage map is an important tool to reveal the genome structure and to
identify marker-trait associations (Cai et al. 2015) which ultimately help in MAS (F. Lu et
al. 2013) to improve precision of selection. In this study, we used the GBS approach to
construct a combined high-density linkage map that spanned 1,473.9 cM with 1,917
markers on 14 linkage groups, which is a very critical tool for further genetic and
genomics studies of napiergrass. GBS has been extensively used for genotyping many
diploid organisms, however, SNP calling from the NGS data in allotetraploids like
napiergrass is particularly challenging due to existence of highly similar homeologous
copies, one corresponding to A genome and the other to B genome (Nagy et al. 2013).
Therefore, different strategies have been devised to construct linkage map in
allopolyploids. For example only uniquely aligned reads (single copy) were considered
for SNP calling and subsequent map construction (Trick et al. 2009; X. Zhou et al. 2014)
to avoid the collapsed alignment of homoeologous reads due to low divergence, recent
55
polyploidization event, and severe domestic bottlenecks (Pandey et al. 2012). SNP
calling in allotetraploid Brassica napus L. (rapeseed; 2n = 4x =38; AACC) was done by
utilizing only uniquely mapped reads (single copy) and a read depth minimum of three
to four reads at each potential SNP (Trick et al. 2009). Linkage map construction in
zoysiagrass (Zoysia matrella) was performed by utilizing single-dose markers after
calling SNPs using the maximum likelihood method in Stacks (Xiaoen Huang et al.
2016). Similarly, single dose markers from TASSEL de-novo UNEAK were used to
construct linkage maps in prairie cordgrass (Spartina pectinate)(Crawford et al. 2016).
In this study, we applied multiple SNP callers and strategies to maximize SNP
calling for linkage map construction for napiergrass. In the final combined genetic map,
the number of markers identified by different software varied dramatically. GATK called
the highest number of SNPs followed by SAMtools and FreeBayes initially. Both GATK
and SAMtools apply Bayesian method to compute the posterior probability for each
possible genotype and then choose the genotype with the highest probability as the
consensus genotype (X. Yu and Sun 2013). GBS-SNP-CROP and TASSEL showed a
low matching percentage, which is similar to results from previous research (Melo,
Bartaula, and Hale 2016). The number of useful markers for linkage group construction
was the highest in SAMtools (47.75%) followed by TASSEL de-novo UNEAK (35.68%).
However, the TASSEL de-novo UNEAK pipeline had the highest number of markers
mapped on the linkage groups (60.43%) followed by Stacks (13.43%). This indicated
that the network-based SNP discovery in TASSEL de-novo UNEAK and UStacks
pipeline (Kim et al. 2015) could be efficiently utilized for constructing linkage maps in
non-model species. Even though TASSEL was primarily designed for diploids, it is
56
powerful enough to give a large number of mapped markers compared to other
programs that handle polyploidy like FreeBayes, GATK, or GBS-SNP-CROP.
The SNP markers were relatively evenly distributed among the linkage groups
with more than 97.45% of marker interval being less than 5 cM. To our knowledge, this
linkage map with an average inter-marker distance of 0.7 cM is the first genetic linkage
map of napiergrass to date. A study based on an interspecific population of a cross
between pearl millet and napiergrass has been previously reported to link RAPD
markers with biomass related traits in Pennisetum (Smith et al. 1993). The large number
of markers and their even distribution in our study facilitate full-scale map coverage.
Few regions where the interval space was > 5 cM might be due to stretches of large
repeats or due to low coverage sequencing of GBS (Poland and Rife 2012; Mathew et
al. 2014). Segregation distortion is regarded as a potential evolutionary force and
including these markers for linkage map construction could increase genome coverage
as well as benefit QTL mapping (S. Xu 2008; D. R. Taylor and Ingvarsson 2003).
Including SDR markers and correcting for bias led to an increase in genetic distance
between distorted markers (S. Q. Xie, Feng, and Zhang 2014). The deviation from
expected Mendelian ratio shows disturbances in the transmission of genetic information
from one generation to the next and can be caused by chromosome loss or
rearrangements, genetic load, gametic selection, zygotic selection, or both (Faris,
Laddomada, and Gill 1998; Karkkainen, Koski, and Savolainen 1996; Bodénès et al.
2016). Napiergrass generally outcross through wind pollination that could result in high
levels of gene flow leading to genetic load. The assignment of napiergrass linkage
57
groups according to the pearl millet genome allows for future fine mapping and QTL
analysis.
In summary, this study reports for the first time a high-density genetic linkage
map using NGS-derived SNP markers, as well as the development of SSRs from
napiergrass genomic sequences. The napiergrass genome showed considerable
collinearity with the pearl millet genome and the genetic map contains 14 linkage groups
with low inter-marker interval. The results will be useful for future molecular breeding
programs such as identification of QTLs for important traits as well as MAS for the
genetic improvement of napiergrass and comparative genomics. These resources will
play a critical role in future whole genome sequencing projects and leveraging
molecular breeding of napiergrass.
58
Table 2-1. Parameters used for SNP calling for each software.
Reference based Parameters Remarks. [defaults]
TASSEL 4.3 -c 5 Min. number of times a tag must be present to be output <5> [1]
-mnMAF 0.01 Min. minor allele frequency <0.01>[0.01]
-mnMAC 100000 Min. minor allele count <100000>[10] (SNPs that pass either -mnMAF or -mnMAC will be output)
-misMat 2 Threshold genotypic mismatch rate above which the duplicate SNPs won’t be merged <2>[0.05]
-callHets When two genotypes at a replicate SNP disagree for a taxon, call it a heterozygote
Stacks -A CP CP type for genetic map
-m 3 Min. number of identical, raw reads required to create a stack <3>[3]
GBS-SNP-CROP -l 30
-sl 4:30
-tr 30
-m 32
Trimmomatic LEADING parameter
Trimmomatic SLIDINGWINDOW parameter
Trimmomatic TRAILING parameter
Trimmomatic MINLEN parameter
-rl 100
-pl 32
-p 0.01
-id 0.93
Raw GBS read length
Min. length required after merging to retain read
p-value for PEAR
Nucleotide identity value required for USEARCH read clustering
-Q 30
-q 0
-f 0
-F 2308
Phred score base call quality
Alignment quality
SAMtools flags
SAMtools flags
-mnHoDepth0 11 Min. depth required for calling a homozygote when the alternative allele depth = 0
-mnHoDepth1 48 Min. depth required for each allele when calling a heterozygote
-mnHetDepth 3 Min. depth required for each allele when calling a heterozygote
59
Table 2-1. Continued
Reference based Parameters Remarks. [defaults]
-altStrength 0.9 Across the pop. For a given putative bi-allelic SNP, this alternate allele strength is the minimum proportion of non-primary allele reads that are the secondary allele
-mnAlleleRatio 0.1 Min. required ratio of less frequent allele depth to more frequent allele depth
-mnCal 0.75 Min. acceptable proportion of genotyped individuals to retain a SNP
-mnAvgDepth 4 Min. avg. depth of an acceptable SNP
-mxAvgDepth 200 Max avg. depth of an acceptable SNP
SAMtools mpileup -uf Default
FreeBayes -C 2
--min-alternate-count Require at least <2> observations supporting an alternate allele within a single individual in order to evaluate the position [1]
-p 4 --ploidy <4> [2]
--use-best-n-alleles 4
Evaluate only the best N SNP alleles ranked by sum of supporting quality scores [all]
--min-coverage 5 Require at least <5> coverage to process a site [0]
GATK -T UnifiedGenotyper
Call SNPs and indels on a per-locus basis
-stand_call_conf 30
The min. phred-scaled confidence thresholds at which variants should be called <30> [30]
-stand_emit_conf 10
The minimum phred-scale confidence threshold at which variants should be emitted (and filtered with LowQual if less than the calling threshold) <10> [30]
-ploidy 4 Ploidy <4> [2]
-mbq 20 Minimum base quality required to consider a base for calling
-glm BOTH Genotype likelihoods calculation model <BOTH> includes SNPs and INDELs [SNP]
60
Table 2-1. Continued
de-novo based Parameters Remarks. Defaults []
UNEAK -e PstI Restriction enzyme used
-c 5 Min. count of a tag must be present to be output [5]
-e 0.03 Error tolerance rate in the network filter [0.03]
-mnMAF 0.05 Min. minor allele frequency [0.05]
-mxMAF 0.5 Max. minor allele frequency [0.5]
-mnC 0 Min. call rate (proportion that how many taxa are covered by at least one tag)
-mxC 1 Max. call rate [1]
Stacks -m 3 Min. number of identical, raw reads to create a stack
-M 2 No. of mismatches allowed between loci when processing a single individual [2]
-n 1 No. of mismatches allowed between loci when building the catalog [1]
-t Remove, or break up, highly repetitive RAD-Tags in the ustacks program
GBS-SNP-CROP Same as reference based except, script 3 settings not required
61
Table 2-2. Repetitive elements present in the napiergrass genome.
Transposable Element Count
DNA transposon
Tc1/Mariner 8
hAT 20
PIF/Harbinger 16
EnSpm 1
CACTA 4
Polinton 1
LTR Retrotransposon
LTR 37
LTR/Copia 5
LTR/Gypsy 5
Copia 4
Retrotransposon 3
Retroelement 1
Gypsy 2
Non-LTR Retrotransposons
L1 2
Pseudogene
tRNA 1
rRNA 8
rDNA-like 3
Others
Mutator 15
Telomeric 4
MobileElement 2
Micro-like sequence 1
Low-complexity 3
Simple Repeats 12
Unspecified 6
Total 164
62
Table 2-3. The sequence alignment of ten napiergrass sequence contigs to the pearl millet genome.
No.
Napiergrass contig
Size of contig (bp)
GC content (%)
Repeat content (%)
Corresponding pearl millet pseudomolecule
Number of hits above 500bp
Number of predicted genes
Alignment length in napiergrass
Alignment length in pearl millet
Sum of expanded length in pearl millet
1 Contig1 8,506 55.3 80.3 Pg5 0 0 / / /
2 Contig3434 9,653 42.5 9.8 Pg3 5 0 5,407 5,384 -23
3 Contig5516 11,920 44.9 17.5 Pg3 11 1 6,776 13,509 6733
4 Contig5578 8,558 42.8 5.9 Pg2 3 0 5,948 6,487 539
5 Contig5588 8,595 44.6 14 C26927002 1 1 591 591 0
6 Contig5729 9,088 46 20.4 Pg5 3 0 2,250 2,246 -4
7 Contig5798 8,651 43.1 5.3 C27370090 1 0 1,385 1,385 0
8 Contig5878 25,329 43.4 12.8 Pg6 4 0 5,692 5,676 -16
9 Contig5902 13,801 46.1 16.4 Pg6 8 0 6,209 7,596 1387
10 Contig6139 14,330 42.2 93.1 Pg7 0 0 / / /
Total 10 118,431 45.09 27.5 8 36 2 34,258 42,874 8,616
63
Table 2-4. Frequency of classified repeat types (considering sequence complementary) in napiergrass.
Repeats 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 ≥20 Total
A/T - - - - - 1045 338 173 76 49 34 17 9 10 3 12 1766
C/G - - - - - 116 71 36 26 11 7 7 6 2 1 283
AC/GT - 118 43 27 18 14 6 2 1 1 1 2 1 9 243
AG/CT - 250 137 82 45 31 19 14 13 9 6 3 5 3 2 6 625
AT/AT - 180 81 55 40 17 4 5 4 8 5 1 4 12 416
CG/CG - 30 10 1 41
AAC/GTT 45 15 14 6 1 2 1 84
AAG/CTT 260 85 46 27 13 13 8 7 12 3 2 476
AAT/ATT 62 31 6 4 5 2 1 2 1 1 2 117
ACC/GGT 86 35 15 7 3 1 2 1 1 151
ACG/CGT 45 3 7 1 56
ACT/AGT 35 12 4 2 1 54
AGC/CTG 131 53 30 15 4 3 236
AGG/CCT 129 34 10 5 1 2 1 182
ATC/ATG 86 38 13 13 4 3 3 3 1 1 1 166
CCG/CGG 157 39 18 3 1 1 219
AAAC/GTTT 2 2
AAAG/CTTT 14 5 3 2 2 26
AAAT/ATTT 19 1 2 1 1 2 26
AACC/GGTT 3 1 4
AAGC/CTTG 1 1
AAGG/CCTT 1 1
AATC/ATTG 1 1
AATG/ATTC 2 1 1 4
AATT/AATT 1 1
ACAG/CTGT 5 1 1 7
ACAT/ATGT 6 2 1 1 1 5 1 1 4 22
ACCC/GGGT 1 1
ACCG/CGGT 1 1
64
Table 2-4. Continued
Repeats 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 ≥20 Total
ACCT/AGGT 1 1
ACGC/CGTG 2 2
ACGG/CCGT 1 1
ACTC/AGTG 2 2
ACTG/AGTC 1 1
AGAT/ATCT 7 2 2 1 1 13
AGCC/CTGG 1 1 2
AGCG/CGCT 1 1
AGCT/AGCT 1 1
AGGC/CCTG 1 1 2
AGGG/CCCT 2 2
ATCC/ATGG 6 1 1 1 9
ATGC/ATGC 3 3
CCGG/CCGG 2 2
AAAAC/GTTTT 2 2
AAAAG/CTTTT 1 1 2
AAAAT/ATTTT 1 1
AAACC/GGTTT 1 1
AAAGG/CCTTT 1 1 2
AAAGT/ACTTT 1 1
AAATC/ATTTG 1 1
AACAG/CTGTT 1 1
AACAT/ATGTT 1 1
AACCT/AGGTT 1 1
AACTG/AGTTC 1 1
AAGAG/CTCTT 1 1
AAGAT/ATCTT 1 1
AAGCT/AGCTT 1 1
AAGGG/CCCTT 1 1
65
Table 2-4. Continued
Repeats 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 ≥20 Total
AATAC/ATTGT 1 1
AATAG/ATTCT 2 1 1 4
AATCC/ATTGG 1 1
AATCT/AGATT 1 1 2
AATGG/ATTCC 1 1
AATGT/ACATT 1 1
ACACC/GGTGT 1 1 2
ACACG/CGTGT 2 2
ACCAG/CTGGT 1 1
ACCGC/CGGTG 1 1
ACGGC/CCGTG 1 1
ACGTC/ACGTG 1 1
ACTAT/AGTAT 1 1 2
ACTCC/AGTGG 2 2
AGAGG/CCTCT 2 2
AGATG/ATCTC 1 1 2
AGCCG/CGGCT 2 2
AGCCT/AGGCT 1 1
AGGCC/CCTGG 1 1
ATCCC/ATGGG 1 1
AAAAAG/CTTTTT 1 1
AAATCG/ATTTCG 1 1
AACCCG/CGGGTT 1 1
AACTCC/AGTTGG 1 1
AAGAGC/CTCTTG 1 1
AAGAGG/CCTCTT 2 2
AAGATG/ATCTTC 2 1 2 5
AAGCAC/CTTGTG 1 1
AAGGAG/CCTTCT 3 3
66
Table 2-4. Continued
Repeats 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 ≥20 Total
AAGTAG/ACTTCT 1 1
AATAGG/ATTCCT 1 1
AATAGT/ACTATT 1 1
AATCCC/ATTGGG 1 1
AATCGG/ATTCCG 1 1
ACCAGC/CTGGTG 1 1
ACCATC/ATGGTG 1 1 2
ACCCCG/CGGGGT 1 1
ACCGTG/ACGGTC 1 1
ACGGGG/CCCCGT 1 1
ACTGCT/AGCAGT 1 1 2
AGATAT/ATATCT 1 1
AGATGG/ATCTCC 1 1
AGCAGG/CCTGCT 1 1
AGCATC/ATGCTG 1 1
AGCATG/ATGCTC 1 1
AGCCGG/CCGGCT 1 1
AGCCTG/AGGCTC 1 1
Total 1172 953 451 252 142 1263 459 243 136 84 56 31 26 16 6 49 5339
67
Table 2-5. Primer pairs developed for napiergrass SSR markers. Available at https://www.nature.com/articles/s41598-018-32674-x#Sec18
68
Table 2-6. Summary of the alignment of non-redundant tags of napiergrass (Cenchrus purpureus) to the available genomes of different species.
Genome used (species name) Number of tags
with blast hits
Percentage of tags
with blast hits (%)
Arabidopsis (Arabidopsis thaliana) 120 0.07
Purple false brome (Brachypodium
distachyon) 6,029 3.30
Barley (Hordeum vulgare) 4,751 2.60
Rice (Oryza sativa) 6,879 3.76
Pearl millet (Pennisetum glaucum) 69,385 37.93
Switchgrass (Panicum virgatum) 24,163 13.21
Sorghum (Sorghum bicolor) 14,654 8.01
Foxtail millet (Setaria italica) 40,849 22.33
Wheat (Triticum aestivum) 6,459 3.53
Maize (Zea mays) 11,972 6.54
69
Table 2-7. Alignment of individual napiergrass reads using Bowtie2. Number of raw reads Clean reads (retained
by Stacks) % of reads retained
Total reads mapped to pearl millet
Uniquely mapped % uniquely mapped
Min. 44,858 20,350 45.37% 16,658 1,140 6.00%
Max. 5,046,114 4,739,605 97.53% 3,022,185 1,902,058 45.00%
Avg. 2,893,836.07 2,696,528.11 92.71% 1,727,194.04 1,077,802.96 39.68%
Total 549,828,854 512,340,341 93.18% 328,166,867 204,782,562 39.97%
70
Table 2-8. Summary of the combined linkage map of napiergrass and the percentage of gaps less than 5 cM in male and female parent linkage maps.
Napier grass linkage group
Pearl millet syntenic pseudomolecule
Number of grouped markers
Mapped markers
Unmapped markers
Length (cM)
Density (markers per cM)
Combined map Gaps <=5 cM (%)
Female map Gaps <= 5cM (%)
Male map Gaps <=5 cM (%)
LG01 PM01 411 163 248 109.33 1.49 98.15 93.98 93.75
LG02 PM06 378 170 208 142.40 1.19 97.63 89.19 92.45
LG03 PM03 360 156 204 89.81 1.74 98.06 96.97 85.71
LG04 PM05 339 127 212 74.83 1.70 99.21 95.51 85.53
LG05 PM02 324 182 142 105.50 1.73 98.34 96.55 96.36
LG06 PM04 300 129 171 112.53 1.15 96.09 96.23 96.83
LG07 PM07 279 99 180 96.84 1.02 96.94 89.8 84.72
LG08 PM06 279 120 159 98.45 1.22 97.48 84 89.86
LG09 PM02 278 96 182 108.70 0.88 92.63 94.64 95.24
LG10 PM01 254 181 73 102.49 1.77 98.89 92.65 75.76
LG11 PM07 237 141 96 97.35 1.45 98.57 91.53 88.06
LG12 PM03 230 146 84 105.72 1.38 97.24 89.29 90.11
LG13 PM05 224 101 123 70.18 1.44 98.00 83.67 94.12
LG14 PM03 165 102 63 96.00 1.06 97.03 89.09 90.91
Total (average) 4,058 1,913 2,145 1,410.10 (1.37) (97.45) (91.65) (89.96)
71
Table 2-9. Summary of napiergrass single nucleotide polymorphism (SNP) markers mapped on the combined linkage map using 9 different software pipelines.
Software Number of
SNPs called
Total SNPs
used for map
construction
No. of SNPs on
map
Percentage of
SNPs on the
map (%)
FreeBayes 25,193 6 0 0.00
GATK 197,475 52 5 0.26
SAMtools 27,082 3,377 151 7.89
GBS-SNP-CROP 2,906 115 52 2.72
TASSEL 7,326 116 56 2.93
Stacks 4,920 447 257 13.43
GBS-SNP-CROP
de-novo
4,521 96 51 2.67
Stacks de-novo 6,871 339 185 9.67
TASSEL de-novo
UNEAK
10,799 2,523 1,156 60.43
Total 287,093 7,071 1,913
72
Figure 2-1. Sequence variation for SNPs called in various regions of the pearl millet genome.
0%
10%
20%
30%
40%
50%
60%
70%Intergenic
Up
5' UTR
Exon
Donor
Intron
Acceptor
Exon
3' UTR
Down
Intergenic
73
Figure 2-2. Micro-collinearity between contigs from napiergrass to the pearl millet genome.
74
Figure 2-3. Inversion duplication between napiergrass and pearl millet (shown in bottom figure).
75
Figure 2-4. Estimated coverage of PstI restriction sites in the pearl millet genome.
0
5
10
15
20
25
30
35
40
0 5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
10
0
Mo
re
Num
ber
of
sam
ple
s
Estimated coverage (X)
76
Figure 2-5. Histogram of uniquely mapped reads to the pearl millet genome.
0
10
20
30
40
50
60
70
80
90
100
<6 // 34 36 38 40 42 44 46 48
Num
ber
of
sam
ple
s
Percentage of uniquely mapped reads (%)
77
Figure 2-6. Venn diagram showing concordant napiergrass SNPs called by five reference-based SNP callers, SAMtools, GBS-SNP-CROP, GATK, FreeBayes, and TASSEL. Numbers in parenthesis after the program name shows the total number of SNPs called by each program.
78
Figure 2-7. Genetic linkage map of the napiergrass female parent N190.
Napiergrass: Female Parent Linkage Map
dT20387dT41575dT1702 dT3027dT909C1_10467974dT38795S3_64306230dT3328dT49252dT30456dT16886dT53891dT16764dT35147dT49010dT2473dT1951dT16527C1_220011116dT50264dT29470dT32431dT9029dT50826dT47548dT45924 dT7985dT24786 dT732C1_102656250dT13338dT19473dT18485dT46948dT51201dT23838dT23364dT23237dT28498dT51779dT43458dT6993dT29597dT5196dT14180dT48800dT3431dT48497S1_17257463dT9167dT28049dT48243S1_17257559S1_17257551dT18424dT973dT8459dT33654dT48112dT53672dT12272dC_359278C1_133709193dC_339765S1_171369178dT40041dT18064dT971dT22725dT10843 dT19448dT47057C1_234266014TS1_273230575dT30844dT7314dT766dT36979S1_217117736dT30297dT3813S1_241274562
1
S6_80432248
S3_200493878
dT46838
dT13195
dT35970
dT7946dT15344dT3773dT22636dT9283dT40427
dT34184dT17674dT14387dT1832
dT11200
S6_199487878
S6_197918666
dT40470dT31571dT45101dT50977dT53887dT50932dT25232dT32392dT44224dT42117dC_193114dT18155dT16658dT51346dT35381dT10670dT15893 dC_23232dT47482 dT35323dT16487dT16409dT4135dT35318dT16616dT16875dT17908dT35582dT15707dT49861dT50319dT22787dT4190 C7_82320373dT49133dT32996dT15879dT8813S6_398797
dT29058
dT46613
dT7678
dT42236
dT1466
dT40376
dT42316
dT53493
dT3930
dT159dT590dT6548dT41013dT38488dT6810
dT50779
dT17665
2
dT28623dT35743
dT30402
dT53129C6_237238320dT7257
dT42723
dT27350
dT165S3_58228915dT17556dT24472dT11506dT38048dT7396
dT22681dT44178dT17379dT19812
S3_234961343
dT12951dT19939
S3_65392670
dT35908
dT25817
dT6688dC_42042dT2147S3_68670889dC_28051 dT5016dT22363S3_68670908dT33771dT51194dT13022
dT40851
dT11544dT42647dT42768dT50758dT6927S3_92611362dT37158dT12246dT31485dT18863dT18489dT35462dT19043dT11234dT26079dT16041dT13169dT25285dT18211dT49009dT28196dT6590dT50718dT22708dT27517S3_35602619dT3175dT35742dT43871dT2820dT492dT1427dT12717dT32542dT42678dT18556dT14307C3_245369357dT43387dT43702TS3_205733261dT24042dT52226dT48403dT15989dT23856dT42771dT207dT21277dT38486C3_202051529dT8152dT15649dT15821dT22318dT31754dT3519dT25849dT35984dT23371dT25577S3_41221675
3
S1_6210777
S5_128073888
S1_6210761
dT24515
dT17924TS5_114098156dT16773dT40995dT15817dT14651dT18480TS5_128053793 dT13827C5_156998414 C5_112498054dC_24267dT27345 dT30788dT9315S5_40839446dT4737dT31217dT32027dT46427dT32695dT43636dT31735
dT50145
dT43048C5_132196063dT20788dT32207dT24792
dT46652
dT27717dT15391dT1524dT14375dT31410dT31722
dT18219dT36949
dT24276
dT7856
dT5489
dT43892
dT48559
dT42534
dT49410dT52149
dT32206
dT6664
dT8114dT53440dT40421dT14806dT22647S5_31904873dT39799dT3053S5_26740824dT4924dT37762
dT26721dT26387dT6733dT17794S5_16898738dT37016dT47995dT34310dT51527TS5_13509703dT54119dT53202dT32125dT27794dT2435dT46426dT20384C5_3132406dT15117dT29259dT780dT15952dT26168dT36121dT32758dT30470
4
S2_16737601S2_16737644
dT5178S2_174624948dT43720dT28282dT38918S2_24418753dT12948S2_24418742dT42842dT19311dT38665dT47411dT3578dT50366dT33944dT1873dT14669dT8327dT31449dT49212 dT13764C2_21121892S2_110636956dT16449dT36036dT29465dT49950S2_76988086TS2_76988086S2_76982335dT5796dT50180dT16824dT17391dT20397dT50336dT27602dT19647dT23469dT10549dT4979dT53077dT15063dT348dT34541dT45884S2_192440020dC_212370dT51287dT10378dT46228dT11302dT31033dT45970dT2290dT3272dT10652dT44368dT9233dC_167176dT3180dT44853dT25838dC_64162dT51942dT30560dT27325dC_88689dT48120S2_246378092dT32534dT6193dT20949dT12622dT50361dT31503dT31430dT43605dT13489dT33550dT10926dT37626dT37138dT37696dT2179
5
S4_43915384S4_43915389
dT20158
dT12265
dT21313dT16746
S4_12461988
S4_88888650
S4_17912300
dT49047
dT52847
dT12456dT18515dT9337dT35551dT37277
dT1848dT23823dT49039
dT17535 dT23182S4_49830629dC_23097 S4_53822034C4_54042612 dC_15219dT20953
S4_14538257
dT33655dT18867dC_45605dT19245dT3906dT4427 dC_60915dC_107792dT51291 TS4_14538257dC_10879C4_99397060 dC_90916dC_31636dT51677dT2317TS4_171269213dT30722dT29654dT12209TS4_171269280C4_194142272 C4_194142277dT8377dT9213
6
S7_182881700
dT30583S7_112269G7_212623362dT46230dT910dT26620dT43326dT27451dT50763dC_388382dT50419dT24186dT24408TS7_180945596dT46209
S7_133861267dT8557dT14658
S7_100645462
dT37501
S7_90744125dT28918S7_201830035dT10131dT429S7_32127059dT33050S7_117332098dT36791dT7088dT526dT30134dT35232dT45831 dT53184TS7_167047574dT52369S7_102687604 C7_89265590C4_41131304 dC_221873dT40952dT3713dT22267dT12440dT32512S8_44730456
S7_89265546
7
dT50316
dT37150
dC_93403 dC_63791
dT15691
dT13837dT18102
dT3402dT46843dT23515
dT11877dT28906
dT44856C6_223591048dT1593dT26926
S6_209490431dT43186C6_200342230dT35845
dT9008
dT47714
S6_9770308
dT44581
dT4343
dT1085
TS6_16445661
dT47034dT34102dT6207TS6_395354dT26411dT42227C6_73259826C6_59409932 dT24125C6_59409930dT771dT1888 dT40500dT47247dT12621dT26512dT49424dT29318dT29815dT38935dT31316dT358dT45995
8
dT14190
dT44709
dT12313
dT27092
C2_22518425dT48390
dT741
dT38026
dT21645dT33261dT15678dT3554dC_183703dT47400dT21265dT41244dT11408S2_78894878S2_78894944dT36645dT35239dT8425dT30235dT52869S2_78894908dT22956dT22796dT36870C2_206875269dT40868dT28517dT48482dT45174dT7119 dT5487dT1149 TS1_248242945dT37165dT3390 dT44398dT22957dT8998dT3928dT24570dT1352dT2536dT7910
dT17835dT44636dT32308dT17639
dT35127dT6817dT15991dT48401
dT48669
9
dT29619
dT39834
dT42418
dT47966dT47917dT5844dT3588dT29992dT7436dT20060dC_23137dC_15179C4_182430864dT30922 dT50307dT12488dT32245dT5081dT45821dT48530dT28837dT38881dT51420dT13498dT4836dT21114dT42109dT40903dC_1926dC_9578 dC_4103dC_4097dC_9584C4_136757390dT47650dT34779dT10689dT28312dT14780dT11768dT46085dT10507dT50903dT48992S1_135580638
dT47961
dT47132dT50618dT51615S2_236302464dT45124dT24369dT33254dT10533dT48984dT40084dT23737dT52173S1_268164908dT37270C8_45958102dT27051dT31375dT36311dT33782dT46112dT40810dT37914
10
dT49930S4_131830044 G4_131830067S4_131830067S4_131830066S4_131830063dT22876dT12269C4_131830077dC_61912 dC_55022dT22567dT26116dT38883dT36340 dT10040dT26884dT10739 dT621dT17325 S7_89250999dT17578dT25711dT5186dT8749dC_10270dT649dT53247dT53022 dT12194S7_39646339S7_39646309S7_39646292dT33677dT3510dT19298dT1762dT41196dT30018dT40278S7_83881022dT7491dT38871dT29364
dT8910
dC_223673
dT43632dT34571dT3675S7_213184560dT22858dT32036dT29409
dT19025dT7711
dT39297
dT42995
dT47137
dT37923
11
dT25904
dT43814
dT8647
dT51744
dT37862dT17058
dT9730dT24050dT38555dT38857dT19291 dT3341C7_118131103dC_15863dT8512C3_216897716dT37132dT28896dT47840dT5917S3_296004281dT50258dT16793dT34879dT50794
dT5877
dT28463
dT14945
12
dT6944
dT6724
dT21137
dT10917dT15762
dT51352dT12719dT5926dT8126dT36120dT18762dT3360dT50697dT26006dT15569dT37911
dT12924dT54030 dC_79210dC_79234dT1355dT43106dT46299dT32011
dT31151dT23798dT15747dT15626dT5625
dT14825
dT22667dT12193dT8566TS5_23967052S5_34147564dT51949
dT46811
dT19739
dT6729
TS5_12313000
dT10696
dT40635
S5_7812618
dT46712
dT30275
dT49612
dT22104dT50943dT46312
13
S7_177490027
dT19814
dT24731dT51987dT21395
dT34497
S3_237947843
dT11865
dT22841
dT10571
S3_255053440S3_255053465S3_257143847dT35964dT17340dC_399018S3_253954916
dT29965dT45243dT48835
dC_1015698dT1549S3_244121865dT29691dT51116dT12496dT41167dT33642S3_276917001dT12709S7_182689743S7_182689737S7_182689721dT19423dT51331dT23909dT30367dT23052dT2763dT40596dT42183dT37207dT7307 dT31619dT21724 dT39129dT8609dT6698dT30790dT34207dT32848dT40659dT18945
dT28003
S4_86603238
14
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120
125
130
135
140
145
150
155
160
165
170
175
79
Figure 2-8. Genetic linkage map of the napiergrass male parent N122.
Napiergrass: Male Parent Linkage Map
S3_66915283S3_66915258dT6838dT44555dT27134S3_66915278S1_7068544dT3437dT46011dT38476dT35347dT33533dT28618dT39822dT28250C5_27253496TS7_74798545dT31263dT6267dT30050C1_119093525 C1_119093531dT37316dT12054S1_71014695dT17414dT29057S3_284635389dT2475dT18370 dT12390dT41113dT30582dC_618193dT44577dT34797dT51180dT28611dT35867dT23729dT34784dT37028S1_193475897dT36388dT3762dT38724TS1_74139971 dT11106dT6132dT7984dT13772dT53762dT8946dT18605dT29128dT52531dT41949dT47752
dT27185
S4_21013305
dT8342dT52469dT37356dT35128dT53136dT39139dT36338dT19693dT44366dT28974S1_249594415dT5535dT8792dT42596dT48305dT32706dT49414C1_251517594dC_259973S1_251483789dT41404dT42403dT9863S1_260924412dT35453dT45297dT28492dT47753dT32523dT21677dC_2450640dT48962dT50248dT27136dT32605S1_267892930
1
dT11902dT17348dT46533dT53963dT38576C3_124558897dT10674TS6_10754586dT10631dT27467dT2555dT45550dT11037S5_31045825S6_35473070dT35938C6_39611566dT6835dT16233dT38911dT38805dT6794dT14931dT16992dT46041dT16346dT47787dT42989dT50566dT9987
dT29726
dT23975dT31396dT43235
dT2439
dT39167
dT35142
dT5958dT8821
dT25121dT39882dT30721 dT19708dT36233dT18641dT28669dT15498 dC_1941dC_5486367 dC_428936dT50064 dT41460dT5222dT43295dT31359dT44933dT33834dT46138
dT987dT17187
dT3299dT29166
dT2229
dT19151dT22571dT36501dT11509dT4155dT1086dT20499
S6_118469241dT14954dT39428dT19912dT53574dT43321dT34936dT2297dT8340dC_234325dT20274dT22921dT30440dT14807dT30703dT6973dC_196800dT18761dT36412dT24998dT25533dT29143S6_55827481
dT43184C6_200342258dT9665dT18516dT18539dT13751dT50350
dT22426
dT17747
dT45130
dT29238
dT13151
dT31763
2
S3_260011343
S3_43511354
S3_248431947
dT14717
dT29788dT42777
TS3_102980429
S3_277598650dT3273G3_243475604 S3_243475604S3_243475658dT36593dT50908 C3_273261021S3_243475784dC_75778C2_168637894dT41484dT6670dT17358dT40996dT7695dC_242221dC_150274dT17019dT43835dT17380dT52592dT24195dT5635S3_267455176C3_119371432 C2_56055562S3_12254914dT2704dT24290dT35854dT34136dT29456dT1963dT6453dT53771dT52360
dT9432TS3_132668864dT15069dT40139dT41347dT723
dT38859
dT33605dT50383dT4281dT45959
dT21342dT12737
dT48862C3_67469347dT26209dT15881
S3_215627633dT53635dT36969dC_45074S3_58384211S3_58384176dT25438dT18781dT26851dT35166TS3_52460829dT44698
S3_61665185
dT20803
dT13705
dT24779
3
S5_3204810S5_3204852 S5_3204827S5_3204849S5_3204807
dT8760dT16974dT23430C5_4168325dT21831dT36067dT50761dT941dT52160dT22371dT32024dT3345dT43dT34717
S5_15050945
S5_32287078
dT35978
dT24372
C5_64002238dT8498dT19279dT5137dT24075dT19864S5_77555777dT4800S5_52320341
S5_59674699
dT22005dT28525
dT2126
dT32675dT48427dT51146dT23969
dT7381
dT12898
dT47043
dT22927dT21823dT38799 dT9509dT27949C5_65406593dT49533dT20066dC_282880
S5_142555945dT41084dT47758dT9747dT20176dT7627dT40123C5_154569850dT32999dT717dT49728dT51649dT10471dT17979dT31973dT26005dT53506dT36750dT13412dT36706dT15816dT5428S5_116583893S5_116583902
4
S2_14223558dT30464S2_16747159dT8117dT28731dT41513dT12415dT15685C2_174576603dT1897dT53530dT29388dT7310dT48623dT19877dT23648S2_30824637dT45539dT22002dT51889dT42630dT1791dT32449dT28199dT41204dT32273dT51314dT6913dT13912dT15477dT11178 dT3792C2_69640554 dT777dT8400dT34457dT3108dT13121dT4303dT1297dT6238dT43507dT34243dT53270dT9893dT27178dT24088dT3606dC_37024dT13934dT47510dT50561dT743S2_151241666C2_151270539dT52704dT39328dT45195dT2081dT44451dT31232dT34738dT52854dT11711dT35398dT33063dT35972dT27463dT24964dT20846S2_191695057S2_191695048dT49783dT49964dT32474dT53079dT52542S2_116787266dT51978dT51060S2_201762543dT33021dT1490dT51442dT26122dT35797dT35508S2_214316973dT10485dT9291dT45475S5_46066941dC_247492dT2450S2_100034039dT14670dT38863dT38020S2_226559654S2_216176645dT9060dT25565dT44700S2_176915862dT17024dT5816dT13896dT21007dT24044dT35466
5
S4_192354592
dT43261
dT3062
dT20572
dT9434
dT50501
dT13383
dT25619dT15831dT29846dT20879 dT3030dT27514dT40535 dT18538dT48489C4_200698121dT4430dT23119dT43805dT3891dC_13461S4_116088663dT13255dT32416C4_90405411dT8320S4_29656682dT42579dT4620dT1631
dT43233dT18870dT8454dT7751
dT29230dT4496dT6466dT16151TS4_49707946dT11273dT43759S4_109511769TS4_79794887dT36845dT33420dT25340dC_124707dT31777S4_17921468dT39406dT45232dT4301 C4_13275000dT24376dT36267dT26600dT52456dT10193dT34107dT22232dT9256S4_17796331
6
S2_174467255
S1_203215691C2_174467255 dC_58121dC_58157dT1325dT24339dT33835dT10459dT21208dT41386dT16455dT38126dT20154dT24498
dT50610
dT44347dT6436dT4353
dT5027dT23432dT39900dT861dT23905dT42142S7_227510
S7_161179045 S7_161179072
dT21588
dT14775
dT34305
dT5751
dT17613
dT7877
dT7992dT22362S7_168784407S7_168784355
dT26372
S7_146917990
S7_100737594
dT46088
dT36827dT36253
C7_38012075 S7_52134705
S7_52134761
S7_52134764
dT25902dT43823
dT19286dT43692S5_1337832dT23441dT28159dT9688
dT11814dT16180dT26922S7_119895303dT8614dT29758dT10112
dT6553
dT37830S7_66630588dT39337dT44668dT25362dT28445dT9471dT3618
7
S1_256402076
dT7731dT4991dT45191dT14502dT40880dT5478dT32461
dT40798
dT4388dT8297dT15729dT33449dT28736
dT45351
dT36410
S6_161712969
dT5999
dT47337
dT20275
S5_15660406
dT21669
dT45910
S6_69712882S6_69712890dT41069dT31386dT9518dT51317dT20193dT3770dT51103dT26827dT29908dT34795dT48051dT53110dT16879C5_27253540 dT9307C6_186956079 dT47900dT37205dT12426dT18627dT32069dT46240dT43200dT47089dT38416S6_99869597S6_99869584dT1280dT3322dT8391dT52684dT30730dT33180dT35795dT41539dT33330dT1374dT29657TS2_146426838dT31286dT17087dT43816dT10476dT53934
8
dT29596
dT6813dT8719dT33885dT31744dT26674dT44870dT30975dT13445dT8898dT24568dT46387dT36424dT33344dT4180dT39395dT25394dT47535dT1174dT2510dT3073dT19937dT37466dT38885dT48964dT22920dT1264dT11516dT52321dT12612dT9304dT42843dT1869dT18940dT13681dT2286 dT30154dT16070dT30575dC_161538dT25258dT38621dT27267S2_187091485dT33297dT1716dT53659dC_66514 dT30639dT40485dT53712dT32613dT51740C2_186899428dT23775dT9800dT32303S2_15995145 S2_15995173S2_15995166dT25206dT31778dT10587dC_198183dT34859dT30984dT7580dT25812dT11183dT35258dT10669dT21093dC_299275dT51487dT334dT50663dT10724dT18564dT36456S2_225137713S1_49814862S2_225245713dT32617S2_223315521
9
dT25323
dT41707dT45868
dT46296
dT19710
dT38592
dT53592
dT36198
S8_31103785
TS4_61117193dT16762
dT39279dT4379dT2661C4_54525188dT27692
dT43430dT11454
S1_211915643
dT5896
dT8867
dT41621
dT49164
dT46261
dT44455
dT30147
dT14352dT29968
dT32403dT29535
dT27669
dT35529
dT11525
10
dT47338dT35994
TS7_92021240dT13178
dT8606
dT11256dT42952C4_131830009
dT52290
dT20991
dT36533dT30022dT52180
dT31904
dT11891TS4_175268085
dT13560
dT29868
dT45689
dT52876dT15502dT18206
dT8487
dT26938
dT31173
dT10167dT2668dT8511dT34945dT17028dT16149dT28164dT838dT34453dT47823dT5924dT49429S7_12814263
dT45454
dT32197dT46410dT30920dT31921S7_146917952dT15629dT41891dT20567dT27310
dT11287
dT20514
dT24863
dT13782
dT46507
dT40245
dT24540dT17489
dT29640dC_75137dC_162440dT35101dT35136dT36597dT50685dT10621dT9181
S7_216787106
dT32609
11
dT52537
dT16456dT26692dT43606C2_14084192 C8_42747377dT14701
dT31069
dT9771
dT2397dT41528dT11071
dT20355dT47937dT47375dT23509dT13802C2_38440300dT47931dT9367dT35753dT16483dT29644dT36437dT5577dT29458dT22493dT5545dT304dT24621dT30035dT18050dT51168dT14389 dT1961dT12204dT991dT12112dT44755dT26322dT23286dT38411dT45471dT38599dT12000G3_47707669S3_47707650G2_236219679dT27423dT35339dT40541S3_50037874dT25091dT50771dT29828dT14016dT5445C3_198785997dT574dT53719dT14813dT11551dT40850dT6981dT32941dT27447dT10535dT2057dT2770dT221dT39089dT36711dT52029dT48485dT29482dT16591dT38040C3_64621327dT11606dT332 dT23060dC_45050dT24014dT41922
dT14847
dT5281
dT20166
S3_45974916dT10443
dT20483
dT23539
12
dT36565dC_4861C5_3132373dT37861S5_2055822dT40955dT25964dT39601C5_1635981dT51406dT51571dT27547dT43154dT34088dT41728dT25168dT44314dT9903dT9904dT5307dT1844dT24760dT48033dT13687dT43027dT6561dT24969dT43643dT1444dC_121093 C5_13708967dT53121dT1436dT6043S3_123721011dT49798dT18120C5_23963287dT52425dT45026dT34144
dT31045
dT14412dT16dT48116dT46519dT10955dT46588dT5755dT46787dT49723dT33950dT48885dT40660dT7562dT22374dT24215dT29452dT40958dC_26149dT22682dT10481dT49967dT13083dT43713dT39674dT1564dT21673dT48272dT1781dT16697dT28451dT13262dT11880dT48311dT25766dT34492dT14203S5_136028687dT47563S5_132636460 S5_132636448S5_132636508dT30544dT28689
13
S1_78149353S1_78149366dT4830dT25275dT18503dT10276dT32396dT39911dT52605dC_94047dT12412dT32705
dT25225
dT31852dT48139dT14455S6_32125800dT9078dT32166dT11671dT53836
dT7886
dT22293dT6248
dT35828
dT11704
dT17003dT6792dT24471
TS3_259950719dT47792TS3_71103541dT34857dT48904dT48523dT10105dT47317dT38317dT40493dT43569
S3_258304119S3_258304047 S3_258304050S3_258304063
14
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120
125
130
135
140
145
150
155
160
165
170
175
180
185
190
195
200
205
210
80
Figure 2-9. Genotyping by sequencing single nucleotide polymorphism (GBS-SNP) marker distribution for the 14 linkage groups of napiergrass. A black bar means a GBS-SNP marker. A blue bar represents segregation distortion region. The left scale plate represents genetic distance (centiMorgan as unit).
81
Figure 2-10. Consensus genetic linkage map of napiergrass.
Napiergrass: Integrated linkage map
S1_267892930S1_241274562dT36338dT39139dT53136dT35128dT3813dT37356dS_22638dT52469dS_58585St_7983dT8342St_8341TS1_57392744dT30392dT36979dT766dT36115dT7314TS1_242079107dT30844TS1_273230575C1_234266014dT47057dT19448dT10843dT22725dT971dT18064dT37631dT40041S1_171369178dC_339765C1_133709193dC_359278dT13670TS1_149192348dT45509dT12272dT53672dT48112dT33654dT8459S1_144792807dT973dT16562dT18424St_2847dT50103dT28049dS_5480S1_17257551St_1775St_1908St_3838S1_17257559dS_38214dS_33009dT48243S1_17257463dT48497dT3431dT48800dT14180dT42594C1_212950043TS1_134040623dT46801dT29597dT6993C1_117276997dT36356dT38584St_4264St_4265dT28498dT13814St_4395dT23237dT13772dT23364dT53762dT23838dT51201dT45924dT7984dT46948TS1_74139971dT11106St_11047dS_83781dT19473St_119dS_25631dT732dT13338C1_102656250dT24786dT7985St_11044dS_9398dT39765dT18485dT47548dT9029dT50826dT32431dT29470dT50264St_5341St_203dT16527St_202dT33481dS_62356dS_8179dT46699dT52420dT1951dT35867dT49010dT35147dS_20460St_8592St_26736dS_64991St_26737dT10713dT3328S3_64306230C1_10467974dT34797dT1702dT3027dC_42488dT38795dT31457dT9151dT51180dT2680dT44577dT36538St_4579dS_37868dT18370St_857St_84477St_795dT30050St_5465G8_2178054dS_38908dS_53100St_1128St_8964St_10547dS_896dT3437St_44026St_43733dT14768dT7777
LG01
S6_80432248dT46838dT35970dT13151dT29238S6_197918666dT17747dT18539dT50350dS_64625St_65272dS_36562dT22426dT11200dT29974St_63794St_63444dT2241dT47054dT14807dS_24618dT30703dC_196800dT24998dT30440St_68017dS_24309dS_50465dT41470dT8340dS_78648dT2297dT41082dS_20475dT42484S6_242118514dT40470dT31260dT42554dT45101dT50977dT25232dT15707St_52312dT13790dC_193114dT42117dT2283dT11665St_67496dT16875dS_65719St_70712dT4135dS_32733St_69789dT18264St_12035dT35582St_864dT35318dT16616St_58109dS_35557dT16409dS_66127TS8_83771289dT35381dC_23232dT47482dT10670dT16487dT16658dT22201dT21000S7_154422455dT7220dT15893dT661dT41813dT7888dT38698dT42580dT5181dT50319St_45778St_30560dT5239dT39882dT30721dC_100348C2_129356854dT4190C7_82320373dT22787St_38861dT28192dT36233dT31759St_68287dC_428936dT28669dT5222dT37647dT43295dC_193851dT19708dT42197dT49133St_66871dS_16944St_68737dS_66207dS_19764dT46586dS_50880dS_50638St_63927St_63925dT46613dT38873dS_65237dS_10139dT43417dS_27918dT37638dT9987dT1466dT50566dT47787dT16346dS_68159dT3930dT6794dT14931dT53493dT46041dS_33008dS_58337dT16992dT38805dT38911dT16233dS_64711dT6835dT17665C6_39611566dT35938dT6810S6_35473070dT38488S5_31045825dT11037dT6548dT41013dT590dT45550dT2555dT159dT27467dT50779dT10674dT10631TS6_10754586C3_124558897dT38576dT53963dT46533dT17348dT11902
LG02
St_36922St_27049St_27050dT24472S3_85079300dT11506dT38048St_36360St_36361dT44178dT7396S3_234961343S3_65392670St_35372dT35908St_31668TS3_40995143dT26209dT2147dC_42042S3_68670889St_36963S3_68670908dT5016dC_28051dT22363dT33771dT51194St_27334St_27337dT13022dT40851dT42647dT42768St_26265TS3_109270808dT50758dT6927St_27574St_26191St_27576dT24195TS3_34404935S3_92611362dT52592dT12246dT24290dT2704St_27619St_27618S3_12254914dT31485St_27764dS_63784St_28126dT20803dT22019dT18489dT18863dT37926S3_99632078dT13705dS_55508dT26079dT31367dT16041dT42473dS_37174dT13169dC_150274St_33851S3_37058907St_30077St_35238St_33736TS3_274563801TS3_274563794TS3_274563751St_31749dT24779TS3_123337609St_12442dT40996TS3_219893756dT50718dT7695dT28196dS_67157St_38576St_27151dT50908dS_40283dT22708dT5249St_32104dS_2204dT6670dT41484S3_35602619S3_243475658dT1934S3_243475784C2_168637894dC_75778C3_273261021St_34784dT43871dT36593dT2820G3_243475604S3_243475604dT492dS_51063dT3273dT35742dT4631S3_277598650dT12717dS_24557St_32307dS_60100dT40657dT19568dS_10652TS3_102980429dS_68998dT32542dT20435St_32659St_30275dT43387St_32493dT24042St_50788dT25239dS_20958dT15989dT42777dT29788dS_53679dT21277dT15721C3_202051529dT24470dT8152dT15649dT38486dT14717dS_69156dT18085S3_258397449S1_55738760S3_16736775S3_41221675dT33250S3_260011343
LG03
S5_3204810S5_3204852S5_3204827S5_3204849S5_3204807St_55016dS_69164St_83036St_55233S5_34645373St_56474dS_12036St_56231dT8760St_57688St_50301dS_3673dS_33240dS_55772St_56111dT22629C5_4168325St_52503dT36067dT21831dT50761dT941dT50341dT23430St_52805St_54858dS_40368St_54860dT41977dT22371St_53749dS_24959St_53600dT36706St_53555dT31058dT13412dT36750dT53506dS_64191dT26005dT17979dT717dT32999St_54089C5_154569850dT20176dT15952dT7627dT780dT9747dT47758S5_92980877C5_3132406St_57085dT15117dT46426dT20384St_55733St_55732dS_40895dS_35332St_54510St_57106dC_282880dT10717St_53813dT20066dT27794C5_65406593dT9509dT21823dT32419S5_65361310St_56639dT2435dT50835dT15164dT48427dT5772dT23323dT30226dT8428TS5_114098156S5_40839446dT41901dT40995dT34274dT28525dT48365dT18480dT9315St_57290S5_63556385dS_51396TS5_128053793dT13827dC_24267C5_112498054C5_156998414St_51734S5_63556383dT27345dT30788dT14651dT15817St_50448dS_62152St_49453dT4737St_50451dT22005dT16773dT47857dT27775dT24515S5_59674699dT51425dT51S5_52320341dT24372S1_6210777
LG04
S2_30030493dT13183dT41204dT51314dT6913dT11178dT8400dT3792C2_69640554dT777dT34457dT13912dT15477dT32273dT3108dT28199S2_16737601dT1791S2_16737644dT42630dT51889St_22811dS_12677dT22002St_23394S2_174624948dT5178dT43720dT28282dT40697dT38918S2_24418753dT12948S2_24418742dT42842dT19311St_16460dT38665dT33063dT47411St_7438St_33827St_33826dT3578dT50366dT33944dT1873dT14669dT8327St_17032dT31449dT13764dT49212S2_110636956C2_21121892dS_26130St_24033dS_49336dT16449St_41249dT36036St_12768dT29465dT49950St_18051St_13306St_75341S2_76988086TS2_76988086S2_76982335dT5796S2_14223558dS_3168dT50180dT16824St_14790dT1183dS_68166dS_22487dT1114dT17391dT20397dT50336dT27602dT19647dS_66177dT23469dT10549dT4979S2_214301707dT53077dT15063dT348dS_7312St_14914dT34541St_15285St_15284dT45884S2_216176645S2_192440020St_16949St_18971dC_212370dS_17901dT51287dS_50616dT10378dT46228dT11302dT31033dT9291dT45970St_19647dT2290St_15938dT3272dT10652dT44368dT9233dS_69530dC_167176S2_226559654S2_230011940dT3180St_20919dT38020TS2_230011891TS2_230011940dS_292dT44853dT10355St_20865St_20885dC_64162dT25838dS_43657dC_88689dT51942dT30560dT27325dT48120dT25304St_22003St_13438dT19734TS2_210211525dT32534dT6193dT20949TS2_102571886dT12622dT50361dT5902dT44728dS_68599St_58349C2_79892636dC_299059dT31503dT2426dT31430St_17805dT12632dT43605dT44700dT25565dT13489dT33550St_22721dT47765dT10926S7_61864734dT32441dT640dT13896dT21007dT37138dT37626dT24945dT37696dT2179
LG05
dT42156dT36401dT53206dT12209dT49934TS4_171269213dT22615dT2317dT23827dS_49354dT27674St_48501dT51677dC_31636C4_99397060dC_90916S4_43915384dT27183S4_43915389TS4_199335191dT21313dT30293dT47539St_40435TS4_17374522dT52456dT23773S4_17921468St_40403dT45232dS_43008dT39406dS_62092dS_59686St_44404St_40987S4_56719430dT12456dS_16097dT49039dT9886dT1848St_39084dT38946TS4_56235778St_47561St_47563dT35551TS4_79794887dT28999St_47706TS4_79798648TS4_79798642St_47414dT33420dC_124707C3_186584463dT37277St_47413dT30894dT9337dT36845dT18515dT52450TS4_114218088St_32476S4_55762918dT10273S4_71863256dT52847dT5808St_46189dT30598S4_109511769St_46150dT11273dT16151dT6466S4_143679490St_54552dT8454dT18870dT43233dT4620dT42579dS_55849S4_29656682dT3285dT21909dT8320C4_90405411St_41693dT32416dT13255dS_39004St_42645St_13970dT29340dS_38178St_48484St_31330St_31333St_42445St_56407dS_59418St_42444dC_13461dT32924St_47777dS_57611St_47778dS_25947dT43805dS_29894dT23119C4_200698121dT27514dT3030dT40535dT20879dT18538dT48489dT29846dT15831dT25619dT3891dT22674dT12425S4_199859831
LG06
S7_219034180
dT1325dC_58157dT24339dT33835dT10459dT41386dT21208dT38126dT20154S7_182881700dT16455dT24498dT16860dT6436dT4353dS_40842dT39900dT23432dT861dT23905S7_227510dT30583dT42142S7_161179045S7_161179072S7_112269G7_212623362dT46230dT34305dT910dT14775dT26620dT43326dT27451dT50763dC_388382dT11554dT24186dT50419dT24408S7_162635983dT46209S7_100737594S7_133861267dT8557S7_52134764dT36827dT46088dT14658C7_146983287dT36253S7_100645462C7_38012075dT25902dT43823dT37501dT43692dT19286S5_1337832dT14025S7_90744125dT28159dT9688dT47365dT23441dT11814dT26922dT16180dT10131S7_32127059dT36791S7_119895303dT33050dT8614dT7088dT526dT30134dT35232dT45831dT53184dT29758TS7_167047574dT52369S7_102687604C7_89265590C4_41131304dC_221873dT40952dT10112dT3713dT22267dT12440S7_66705818dT32512dT47732S8_44730456S7_89265546S7_136957144
LG07
St_62484St_63298S6_15199694S6_37610339dT45995dT2684dT31673dS_29023dS_7808dT7834dT52932dT19776dT29318dT1374dT41539dC_9265dT33330TS2_146426838dT32668dS_68537St_67003dT25459St_67005St_69926St_8239dT771St_71627dT35795dT40500St_60317St_65388dT33180St_65387dS_43802dT52684dT47247C6_59409930dT24125C6_59409932C6_73259826dT47034dS_59431dT1281TS6_101968558dS_15188dT454dT17688dT22268TS6_16445661C6_105814832dT29814dT52515dT31219dT13258dT52774dT9518dT32504dT38233St_45444St_60942dT4343S6_59035298S6_199704613S6_241915017St_62949S6_241915045St_59677dT45221dT23999dT7731dT4991St_19657St_67385dT45191dT14502dT40880dS_49183dT4808St_64847dT47341dT5478dS_30976dT39816dT32461St_67353TS6_200342427TS6_200143111dT46758dT47729St_63770St_65266St_7565dT51264dS_53085dT330S6_209490431dT15729dS_31075dS_7074dT33449St_63375St_70658dT26926dT8297dT1593C6_223591048dT44856dT28906dT11877dT23515dT46843dT18102dT13837dT15691dT47337dT5999dC_93403dC_63791dT37150dT50316
LG08
dT27092
dT14190
dT44709TS8_54685526
dT12313
dT18940S5_74991349dT30575dT47221dT38621dT9800dT45548dT36456dT2536dT9428dT1352dT18564dT24570dT1670dT7119dT5487dT50663dT44398dT3390dT48482dT37165dT10724dT1149TS1_248242945dT8998dT22957dT3928dT45174dT26839dT21093dC_299275dT28517dT40868St_13976C2_206875269dT36870dT334S2_78894908St_58417dS_34607dS_44017dT52869dS_13200C2_143807176dS_24485dT36645dT191dS_13845dT50464dS_35262dS_13192dT51911S2_225137713S1_49814862dT40583St_16188dS_64130S2_225245713dT17835St_33671dS_57503St_21261dT44636dT32617dT32308dS_23912St_24952S2_144268311dS_56758S2_144268276dS_14880dS_57817dS_41244St_25765dT40973dT29502dS_30742dT14696dT20326dT35127dT52863dS_47193St_17971St_17970dS_68589St_20466St_20650dT48611dS_60240St_18499St_80242
LG09
dT37914dT46112dT41558dT40810dT33782dT36311dT31375dT18082dT32644dS_15838dT27051C8_45958102dT35420dS_28564St_43790dT40690dT37270dT4159S1_268164908S1_268948076dT9656dT27080St_8337dS_20051dS_5062TS1_270056181dT23820dT52173dT23737dT40084dT48984dT10533dT33254dT26907St_7976St_7494dT24369dS_35489S2_236302464dT45124dT20774dT47132dT51615dT50618dT43193S1_256031621dT47532TS1_55816576dT144dT40320dT6229dC_220532dT50746dT45573dS_30445dT7085dT20685dT17967dT6028dT36552dT49336dT23044dS_59001dT47746dT29658dC_105219dT2822dT39692dT40261C1_149814711dT31390dT43343S1_235015034S1_235015054dT42355dT12855dT34010dT5814dT41014dT30197dT5876St_3875dT12762St_9029St_9025dT49435C1_231986176dT40418dS_54187dT20183dT48992dS_8637dT50903dT10507dT46085dT11768dT14780dT28312St_9993St_8983dT47682dT10689St_45245dT34779dT47650C4_136757390dC_9584dC_4097dC_4103dC_9578dC_1926dT40903dS_6690dT1078TS4_5133343dT42109dT21114dT4836dT13498dS_2186dT51420dT38881dT28837dT48530dT45821dT5081dT32245dT50307dT30922dT12488dS_16469C4_182430864dT41621dC_15179dC_23137dT20060St_11615St_46107dT16799dT7436dT5896St_36519dT8867St_39105dT29992dS_55275dS_63047S1_211915643dT3588St_46582dT33996dT5844dT47966dT43430dT47917St_70710C4_54525188St_46842St_46843dS_68339dT42418dT44125dT47502St_41094dS_42852dT53273dT5708dT53333dT39834dT30409dT51639dT11173dT29622dT13526dT25986dT49885dT19710dT17365dT53592dT36198dT38592
LG10
dT34863dT43981dT32288dT34945dT16149dT8511dT2668dT49930dT26938St_26088dS_48139dT31173dT30424dT18438S4_131830044dT44002dT15502dT52876St_72633dT45689G4_131830067dT29868S4_131830067S4_131830066TS4_175268085S4_131830063dT13560dT11891dT22876dT31904dT12269dC_55022dT22567dT38883dT28123dT26116dT10040dS_61343dT36340dT26152dT10739dT26884dT621dT17578St_80208S7_89250999dT17325dT25711dS_21720St_16173dS_353dS_12928dT5186dT8749dC_10270dT15950TS8_38834569dT649dT53247dT53022dT12194S7_39646339S7_39646309S7_39646292dT33677dS_17159dT3510dS_36568dS_18627dT19298St_35836dT1762dT41196St_76782dT30018dT1053dT40278S7_83881022dS_4342dT7491St_81113S4_175268085dS_51373S4_175268116dT32609dT9181dT10621St_83873dT29364dT20560dT35101dT37141dT35136dT49802dT29640dT42825dC_162440dT43565dC_75137dS_53783St_116030dS_89120St_116029dT50685dT43419dT53109dS_217dT10058dT31974dT53392dT42256dT24540dT34174dT47569St_75947dT952dT54084TS7_195845045TS7_195845122dT21991dT44682dT46307dT8910dT24512dT12987dC_223673St_77447St_77432dT43632dT34571dT3675S7_213184560dT22858dT32036dT29409dT19025dT7711dT39297dT42995dT47137dT37923
LG11
dT52537dT16456dT26692dT43606C2_14084192C8_42747377dT14701dT31069St_83499dT9771dT2397dT41528dT11071dS_70286dT20355dT47937dT47375dT29957St_23648dT23509dT13802C2_38440300dT47931dT9367dT35753dT16483dT29644dT36437dT5577dT54067dT36404dT29458dT37862dT5545dT304dT24621St_18279dT17058dT30035dT18050dT51168St_19470dT14389dT1961dT12204dT27243dT991dT12112dT29731dT19317dT44755dT37721dT26322dT39135dT23286dT38411dT45471dT38599dT12000dT4552dS_3981G3_47707669dS_27715dS_68831St_28245St_28242dT26266S3_47707650dT9730dS_58321dT13363dT27423dS_2736dT38555dT24050St_26083dT35339dC_15863dT3341C7_118131103dT40541S3_146735713S3_50037874dT25091dT5917dT28896dT37132dT8512dT50771dT29828dT47840dT14016C3_216897716dT38857dT5445dT19291C3_198785997dT574dT53719dT14813dT11551dT24476S3_58732195dT4006dT40850dT34879dT6981dT32941dT27447dT50794dT10535dT2057dT9107dT2770dT6635dT39970St_37822dT221dT39089dT36711dT52029dT48485dT29482dT16591dT38040C3_64621327dT11606dT332dT23060dS_63173dS_2101dC_45050dT2954dT24014dT5877dT41922dT14847dT5281dT20166S3_45974916dT20483dT23539dT13850dT28463dT4710dT14945
LG12
dT19752St_57083dT6944St_51502St_52583dS_53273dS_30098dT47563dT31151dT15747dT14203dT34137S5_136028687dT8126dS_23166St_51337dT36120dT51352dT18762dT3360St_50213dT12719dT34794dT25766dT34959dT15762dT19372dT34492dT43443dT41039dT45156dS_63438dT44382St_57235dT20917dS_5311dT7338dT48311dT33358dT11880dS_66625St_68482dS_4586St_52257dS_46016dT28451dT15626dT16697dS_12408dT16760dS_26459dS_49110dT14825dT13865dT22667dT12193dS_63968dT8566St_54452TS5_23967052dT36565S5_34147564dT51949S5_875176dC_4861C5_3132373TS5_12313000S5_2055822dT10696dT51406dT51571dT34088dS_26201dT41728dT25168S5_7812618dS_29656dT48033dT40635dT9904dT9903dT44314dT3855dS_21070St_77252dT46712TS5_6700729St_57050St_56490dS_15609St_59001St_77253dT30275dT53121dC_121093C5_13708967dT1436dT49612dT50943dT22104dT46312
LG13
S7_177490027
dT19814dT24731dT51987dT21395S3_258304063S3_258304050S3_258304047dT34497S3_258304119S3_237947843dT11865S3_259562492TS3_202192213dT22841dT43569dT40493dT27298dT27432dT47317TS3_52779446TS3_52779443dT48523dT10571dT10105dT48904TS3_71103541dT43079dT47792S3_255053440dT4657S3_255053465S3_257143847dS_45674dT35964dT17340dT6792dT17003dC_399018St_32898S3_253954916dS_3349dS_1370dT29965dS_33295dT11704dT45243dT48835dT35828dT6248dT22293dS_335dC_1015698dT1549S3_244121865dT29691dT51116dT12496dT41167dT33642S3_276917001dT12709S7_182689721dT317dT9078dT19423dS_45055TS6_161742923dT51331dT23909dT36064dT30367dS_40112St_31374TS3_228287887dT7231dS_37949dS_36977dS_27204dT2763dT40596dT37207dT8609dT7307dT31619dT21724dT39129dT6698dT30790dT42183dT12412dT34207dT32705dT40659dT52605dT18945St_30603dT27593dT39911S4_86603238dS_31623St_32380
LG14
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120
125
130
135
140
82
Figure 2-11. Circos plot of the mapped TASSEL de-novo UNEAK napiergrass markers with pearl millet reference genome. Pearl millet pseudomolecules start with “PM” and are color coded for each pseudomolecule. Napiergrass linkage groups start with “LG” and are in green color. Each line corresponds to tags that showed significant BLAST hits to the pearl millet genome (> 80% identity and > 50 bp length).
83
Figure 2-12. Syntenic regions between napiergrass linkage groups and the pearl millet genome. PM01 to PM07 are pearl millet pseudomolecules, LG01 to LG14 are napiergrass linkage groups. The small dots represent significant BLAST hits of mapped UNEAK tags to the pearl millet genome (>80% identity and >50 bp length).
84
CHAPTER 3 MAPPING QTLS CONTROLLING FLOWER NUMBER AND FLOWERING TIME IN
NAPIERGRASS
Introduction
Napiergrass (Cenchrus purpureus Schumach) is a tropical perennial grass that
originated from Africa (Singh, Singh, and Obeng 2013) and is an important fodder crop
widely used as feed for dairy cows (Farrell, Simons, and Hillocks 2002). In addition, due
to its high biomass potential, napiergrass is considered as a promising crop for
cellulosic biofuel with higher dry biomass yield compared to sorghum (Sorghum bicolor),
maize (Zea mays), sugarcane (Saccharum sp.), switchgrass (Panicum virgatum),
johnsongrass (Sorghum halepense), and Erianthus (Ra et al. 2012). Napiergrass was
introduced to the United States in 1913 (Burton 1990). Being non-edible, napiergrass
escapes the 'food versus fuel' debate as the biofuel feedstock and has competitive
advantage over tree species because it can be harvested for biomass in the first year
after planting. Furthermore, the lignin content, which is considered a hindrance to the
fermentation of biomass to ethanol, is much lower in napiergrass than woody biomass,
10% in napiergrass compared to 20-30% in woods (Tong, Smith, and Mccarty 1990;
Mckendry 2002). In addition, napiergrass tolerates multiple harvests after which it
shows better ratooning ability than energycane (Cuomo, Blouin, and Beatty 1996). This
supports a constant feedstock supply and minimizes transportation costs of biomass.
Napiergrass is a short-day plant and flowering in tropical climates occurs from
autumn through winter (Singh, Singh, and Obeng 2013). Early flowering cultivars
produce abundant wind dispersed seeds, which contribute to the high potential of
invasiveness or weediness (D’Antonio and Vitousek 1992; Loope, Hamann, and Stone
85
1988; Schofield 1989). Therefore, The Florida Exotic Plant Pest Council has listed
napiergrass as an invasive species (FLEPPC 2011). Controlling flowering or modifying
flowering time of napiergrass can minimize invasiveness and boost its potential as
biofuel feedstock. Late flowering will reduce the total seed production and low
temperatures during late season may even compromise development of viable seeds
(Grabowski et al. 2016), thus reducing invasiveness. Significant genotypic variation for
agronomic traits have been documented in napiergrass (Sinche et al. 2018). Flowering
time showed a considerable variation within an F1 population of napiergrass, thus there
is a good reason to believe that improvements of flowering time in napiergrass can be
achieved using genetic approaches. Genome editing tools could be utilized to regulate
flowering time (Jung et al. 2018; Jung and Müller 2009) and to avoid unintended
spreading of napiergrass. However, routine transformation and genome editing
protocols still need to be developed for napiergrass (S. Zhou et al. 2018; J. Wang et al.
2017). So far there are no quantitative trait loci (QTL) mapping reports for agronomic
traits in napiergrass. Even in this genomic era, there is no noticeable genomic data
publicly available for napiergrass. This presents a challenge to improve agronomic traits
in napiergrass breeding utilizing marker assisted selection (MAS). Therefore, more
genomic resources and a better understanding of the genetic basis of flowering time is
necessary in napiergrass.
Most studies on napiergrass were limited to assessing genetic diversity and
relatedness using random amplification of polymorphic DNA (RAPD), amplified
fragment length polymorphism (AFLP), isozymes, and simple sequence repeats (SSRs)
developed for other species like pearl millet (Cenchrus glaucum) and buffelgrass
86
(Pennisetum ciliare) (Lowe et al. 2003; Bhandari, Sukanya, and Ramesh 2006; Harris-
Shultz, Anderson, and Malik 2010; Kandel et al. 2016; Dowling et al. 2013; Dowling,
Burson, and Jessup 2014; López et al. 2014; Smith et al. 1993). Environmental
biosafety of napiergrass can be increased by utilizing molecular markers that are linked
to specific traits of interest such as late flowering. This MAS allows effective selection of
breeding materials. Identification of QTLs related to flowering will support MAS
programs for breeding napiergrass with delayed or less flowering characteristics,
limiting invasive potential of napiergrass. This may enhance the potential for utilizing
napiergrass as a forage and biofuel crop.
QTL mapping tries to identify stretches of DNA that are closely linked to genes
underlying the trait of interest by performing statistical analysis of genome-wide
molecular markers and traits measured in progeny of controlled crosses (Stinchcombe
and Hoekstra 2008). The advancement of next generation sequencing (NGS) has
hugely facilitated QTL identification and mapping. High density genetic maps developed
by genotyping-by-sequencing (GBS) have been used successfully to identify genes
related to flag leaf traits in wheat (Triticum aestivum) (Hussain et al. 2017), winter
hardiness and fall dormancy in Medicago sativa (Adhikari et al. 2018), bunch fruit weight
and height in palm (Elaeis guineensis) (Pootakham et al. 2015), and bloom date in
peach (Prunus persica) (Bielenberg et al. 2015). QTLs related to flowering time have
also been identified in other species as well. In pearl millet, QTLs for flowering time co-
mapped with QTLs for stover yield, grain yield, and biomass yield in LG4 and LG6
(Yadav et al. 2003). Similarly, three QTLs on LG2, LG3, and LG4 were identified for
grain yield across variable post-flowering moisture environments in pearl millet (Bidinger
87
et al. 2007). In wheat and barley, vernalization (Vrn) and photoperiod (Ppd) genes were
involved in flowering time variations (Cockram et al. 2007). Transcription factors such as
AP2 and agamous-like MADS-box were detected in Adzuki bean for QTLs related to
flowering time, maturity, and seed coat color (Y. Li et al. 2017). In rice, several different
QTLs for flowering time were identified and cloned in early and late flowering cultivars
that may have been involved in adaptation to cold regions (Izawa 2007; Xue et al.
2008). However, no QTL analyses in napiergrass have been reported so far. Whether
the orthologs in other species underlying flowering time also control the flowering time in
napiergrass is unknown.
The variation of flowering time in napiergrass accessions can be exploited to
understand the genetic basis of flowering in napiergrass. QTL analyses in napiergrass
is enabled by the generation of the first genetic map of napiergrass, which contains
1,913 SNP markers called from GBS of a pseudo-F2 mapping population, derived from
a cross between an early flowering line and a late flowering line of napiergrass (Paudel
et al. 2018). This map facilitates the identification of QTLs for various traits including
flowering related traits that can be utilized in the future for MAS. The main objective of
this research was to use this first genetic map of napiergrass to identify markers linked
to genes controlling flowering time and flower number through QTL analyses.
Materials and Methods
Development of a Mapping Population
A mapping population of 185 F1 hybrids was developed by crossing an early
flowering accession N122 and a late flowering accession N190 of napiergrass (Sinche
88
2013). Accession N190 produced a higher biomass and had a reduced number of thick
tillers compared to N122 (Sinche 2013).
The 185 F1 hybrids along with the two parents, N122 and N190, and an
established cultivar Merkeron were phenotyped as replicated single row plots at PSREU
in 2011 and 2012. Clones of the whole population were also planted at the Everglades
Research and Education Center (EREC) in Belle Glade, FL in 2015 and flowering date
was recorded in the following year after planting.
Phenotyping the Mapping Population
The field experimental design followed a randomized complete block design
(RCBD) with three replicates of 187 lines (185 F1 hybrids and 2 parents). One block
contained a single plant as a replicate of each line. Within each block the lines were
randomly assigned to estimate block effects. The first flowering date and flower
numbers of each line were recorded in October ~ December of 2012, 2013 on the plants
established at PSREU. The first flowering date was also recorded on the plants
established in EREC in October ~ December of 2016. The flowering date, defined as
the date when the first flower was visible, was documented weekly from the first week of
October to the first week of December. Flowering traits in 2012 and 2013 at PSREU
were obtained from a previous study (Sinche 2013). Flowering time (FT) was calculated
as the number of days between the first appearance of the flower and vernal equinox
(March 20) for the specific year (Lambert, Miller-Rushing, and Inouye 2010).
Genetic Map
The genetic map was generated based on the SNPs from GBS of the 185 lines
of the population (Paudel et al. 2018). Briefly, GBS was used to genotype the 185 F1
89
individuals and the two parents. SNPs were identified using various software tools to
construct a linkage map for maternal and paternal parent by employing a pseudo test-
cross strategy (Paudel et al. 2018). For the female parental line, a total of 899 SNP loci
mapped on 14 linkage groups with a total length of 1,555.17 cM were used. Similarly, for
the male parental line, a total of 1,073 markers that were grouped into 14 linkage
groups spanning a length of 1,939.19 cM were used (Paudel et al. 2018).
QTL Analysis
For QTL detection we chose the composite interval mapping method on
WinQTLcart 2.5 (S. Wang et al. 2005). Mean values of each trait across three replicates
from different years were used for the QTL analysis. The walking speed chosen for all
traits was 2 cM. A forward and backward stepwise regression method with a probability
of 0.1 and a window size of 10 cM were utilized to determine cofactors. LOD thresholds
used to determine the significance of identified QTLs was identified by using the
thousand-permutation test to each data set (p ≤ 0.05) (Churchill and Doerge 1994).
Adjacent QTLs on the same chromosome for the same trait were considered different
when the support intervals did not overlap (Haggard, Johnson, and St. Clair 2015). The
95% confidence interval was calculated for each QTL considering a 2-LOD support
interval (van Ooijen 1992). The QTL span was delimited using LOD-1 support interval
(LSI). The contribution rate (R2) was calculated as the percentage of phenotypic
variance explained by each QTL in proportion to the total phenotypic variance. QTLs
were named according to McCouch et al. (McCouch et al. 1997). Specifically, the QTLs
detected for number of flowers on the linkage map constructed for early flowering parent
N122 were designated “qNFE” (qtl Number of Flowers Early) followed by a linkage
90
group number and QTL number on the same linkage group for the same trait, separated
by a dash “-“. Similarly, the QTLs detected for flowering number based on the linkage
map constructed for late flowering parent N190 were named “qNFL”. The QTL for
flowering time on linkage map for early flowering parent N122 were named as “qFTE”
(qtl Flowering Time Early) and the QTLs for flowering time on map for late flowering
parent N190 were given “qFTL” (qtl Flowering Time Late). QTLs with a positive or
negative additive effect for a trait imply that the increased or decreased phenotypic
value is contributed by the QTL, respectively.
We categorized the identified QTLs as either stable and potential QTLs. QTLs
detected based on the phenotypic data from more than one year were considered as
stable, while those detected based on only one year’s phenotypic data were considered
as potential QTLs. The positions of the QTLs identified on each linkage map were
indicated by using MapChart 2.3 (Voorrips 2002).
Candidate Gene Identification
Sequences were extracted from the sequence tags, which generated the SNP
markers flanking QTL region. The extracted sequences were then BLASTed against the
pearl millet genome v1.1 (Varshney et al. 2017) to identify nucleotide matches of these
sequences to identify the QTL sequence intervals. Gene models and KEGG annotation
(Varshney et al. 2017) of pearl millet genes within the identified QTL sequence interval
were extracted. The function of each gene model was checked for its role in flowering
and a list of potential candidates was curated.
91
Results
Phenotypes
Number of flowers
The number of flowers of the 185 F1 individuals, and their parental lines, early
flowering line (N122) and a late flowering line (N190) ranged from 0 to 46 with an
average of 6 in 2012 (Figure 3-1, Table 3-1) and from 5 to 256 with an average of 61 in
2013 (Figure 3-2, Table 3-2). Many F1 individuals actually didn’t flower before the
harvest date (Dec. 6th, immediately before the first predicted frost) in 2012.The
correlation between number of flowers for 2012 and 2013 was 0.62 (Fig 3-3). The
heritability estimates for number of flowers ranged from 79% to 85% (Table 3-1).
Flowering Time
The days to flowering of the whole F1 population including the two parental lines,
ranged from 219 to 270 days with an average of 240 days in 2012 (Fig 3-4) and from
220 to 258 days with an average of 238 days in 2013 (Table 3-1, Fig 3-5). The average
number of days to flowering in 2016 was 233 days (Fig 3-6, Table 3-1).
The correlation between days to flowering for 2012 and 2013 was 0.34, while that
between 2012 and 2016 was 0.22. Correlation of days to flowering between 2013 and
2016 was 0.26 (Fig 3-7). The heritability for the days to flowering were estimated
ranging from 60% to 87% (Table 3-1). Days to flowering was negatively correlated with
number of flowers (r = - 0.42 for 2012 and r = -0.64 for 2013, Fig 3-8 and Fig 3-9).
QTL Analysis
Number of flowers
Three stable QTLs (qNFE-1-1, qNFE-1-2, and qNFE-1-3) were identified on LG1
of the genetic map for early parent N122 for number of flowers (Table 3-2, Figure 3-10).
92
Three potential QTLs (qNFE-2-1, qNFE-6-1, and qNFE-6-2) were identified on LGs, 2
and 6 (Table 3-2). The most important number of flowers QTL (qNFE-1-2) for N122
parent (R2=0.20, LOD=10.76) was found in LG1and was located at 123.7 ~ 126.8 cM.
The same linkage group harbored the two additional stable QTLs (qNFE-1-1 and qNFE-
1-3) that explained 15-17% of variance with LOD score ranging from 8.13 to 9.10 at
peak intervals from 133.5 cM - 134.6 cM and 116.9 cM - 118.2 cM. This suggests that
LG1 plays an important role for number of flowers in napiergrass.
Five potential QTLs for number of flowers were identified on LG 1, 4, and 5 of
the N190 map (Fig 3-11, Table 3-2). These potential QTLs had a PVE of 6-10% with a
LOD value ranging from 3.18 - 4.78. Four out of those five QTLs had additive effects in
favor of trait value. One potential QTL (qNFL-11-2) had a negative effect (Table 3-2).
Flowering time
A total of six potential QTLs were detected for FT. Two potential QTLs out of the
six were detected on the data of year 2013 on LG 1 of N190 and two potential QTLs
were detected on the same LG of N122 with R2 ranging from 0.11- 0.14 and LOD score
of 5.40-6.77. Two additional potential QTLs were identified on LG 7 of map N122. The
R2 for these potential QTLs ranged from 0.09 - 0.12 with LOD scores ranging from 3.91-
5.57. Both QTLs showed negative effect of small value (R2 = 0.07) on the trait.
Candidate Genes
Candidate genes were only identified in the genome regions corresponding to the
three stable flower number QTLs: qNFE-1-1, qNFE-1-2, and qNFE-1-3. Based on
BLAST analysis of sequences flanking the QTL marker against the pearl millet genome
(Varshney et al. 2017), three genome intervals spanning a genome size of 4 Mb, 12.37
93
Mb and 1.4 Mb, respectively, were identified on pseudomolecule 1 of the pearl millet
genome, corresponding to the three QTLs. A total of 624 gene models were identified
within the three QTL intervals of the genome (160 gene models on qNFE-1-1, 397 on
qNFE-1-2, and 67 on qNFE-1-3). 295 of these models had KEGG annotation results (71
on qNFE-1-1, 191 on qNFE-1-2, and 33 on qNFE-1-3) (Varshney et al. 2017). This
region harbored potential candidate flowering genes such as AGAMOUS, DELLA, Floral
homeotic protein DEFICIENS, PPM1 (MIKCC MADS-domain protein), WRKY, and
SERK1 (Table 3-4).
Discussion
Napiergrass is an important forage crop with high potential as a cellulosic biofuel
feedstock. Breeding of non- or late flowering varieties that have an extended vegetative
growth is not only important for obtaining a high yield but also for reducing invasiveness.
So far, genes involved in flowering regulation have not been identified in napiergrass,
which has severely hindered the improvement of flowering time traits in napiergrass
breeding. Over the last few years, mapping of QTLs for economically important traits
and genome assembly has largely facilitated the breeding programs by the
development of MAS (Y. Wang et al. 2011). Recently, a high-density genetic map of
napiergrass has been developed (Paudel et al. 2018) that will ease identifying QTLs.
Exploring QTLs is important because many studies have identified candidate loci near
QTLs. For example in cereals, a QTL controlling heading date variation was close to loci
controlling photoperiod and vernalization (Laurie et al. 1994; Bezant et al. 1996). In rice,
four genes controlling flowering time (Hd1, Hd6, Hd3a, SE5) were identified from 14
QTLs (Yano et al. 2001). In maize, it was shown that flowering time variation was the
94
result of cumulative effect of several small QTLs and by using 5,000 RILs, a total of 36
QTLs were identified for days to anthesis (Dell’Acqua et al. 2015). Flowering time QTLs
in brassicas mapped to similar regions in homologous chromosomes within and
between species (Lagercrantz et al. 1996; Osborn et al. 1997). Flowering related genes
in Brassica showed a close sequence similarity to one end of chromosome 5 of
Arabidopsis, which contained many flowering related genes (Bohuon et al. 1998). This
type of syntenic results showed that the same QTLs may exist in other related species.
The availability of the reference genome of pearl millet (Varshney et al. 2017) largely
facilitates the comparative genomic approaches for napiergrass as these two species
share one genome (Jauhar and Hanna 1998).
In this study, we phenotyped a biparental F1 or pseudo-F2 population that
segregated for flowering time to identify QTLs related to flowering. The traits studied
showed a continuous variation indicating that the traits are quantitative in nature. We
observed a negative phenotypic correlation between the number of flowers and
flowering time in the F1 population. Days to flowering and number of flowers were
consistent in different years and locations for the population. The heritability for
flowering date ranged from 0.60-0.87 for napiergrass. It was higher than the heritability
reported for heading date in rice which ranged from 0.37 to 0.55 (L. Zhou et al. 2016)
and lower than that reported for maize (0.82 to 0.93) (C. Wang et al. 2010). For linkage
analysis, we employed a pseudo-testcross strategy. Linkage analysis of quantitative
traits in outcrossing polyploid species by using single dose markers (1:1) is a common
practice due to the limitation of the linkage analysis software (K. K. Wu et al. 1992). In
this approach, a pseudo-testcross strategy using heterozygous markers for one parent
95
and recessively homozygous markers in other parent is employed for linkage map
construction (S. Wu et al. 2010; W. Zhou et al. 2015). This strategy has been
successfully applied in tree plants such as Pinus elliottii and P. caribaea (Shepherd et
al. 2003), legumes such as alfalfa (Adhikari et al. 2018), as well as grasses such as
orchard grass (W. Xie et al. 2011), and sugarcane (Yang, Islam, et al. 2018).
In this study, three stable QTLs for number of flowers were identified on the
linkage map constructed for N122. An initial study done in pearl millet identified QTLs
related to grain number and panicle number in LG1 of pearl millet (Bidinger et al. 2007;
Yadav et al. 2003). However, the lack of shared markers did not allow a definitive
conclusion if these two QTLs represent the same locus.
Several QTL regions related to flowering time and number of flowers were
located at the end of the LG 1 of napiergrass. Across all years, the early flowering QTLs
with the greatest effect was found on LG 1 that correspond to pseudomolecule 1 of
pearl millet (Varshney et al. 2017). This highlighted the importance of this chromosome
for flowering and it might harbor genes that control flowering in napiergrass. QTL
identification has enabled us to link variations at the trait level to variations at the
sequence level. Since a QTL can harbor tens to hundreds of genes (Gelli et al. 2016),
the identification of genes responsible for phenotypic variation poses a major challenge.
Accurate identification of the underlying genes responsible for trait variation in QTL
regions is difficult in napiergrass due to the lack of genomic and transcriptomic
resources in napiergrass. Nevertheless, comparative genomic approaches utilizing
reference genome of pearl millet has helped us to identify putative candidate genes
related to flowering in napiergrass. Several candidate genes affecting flowering time in
96
model plants have been discussed previously (Kuittinen, Sillanpa, and Savolainen 1997;
Bouché et al. 2016). In rice, QTL analysis of flowering time identified genes such as
Hd1 (CO orthologue in Arabidopsis), Hd6 (CK2 orthologue in Arabidopsis), Hd3a (FT
orthologues in Arabidopsis), and SE5 (HY1 orthologue in Arabidopsis) (Yano et al.
2001).
In our study, we located important flowering time related genes including
AGAMOUS, DELLA, DEFICIENS, PPM1 (MIKCC MADS-domain protein), WRKY, and
SERK1 (Table 3-4) in the QTL area of napiergrass that showed sequence similarity to
the pearl millet genome. The AGAMOUS gene encodes a transcription factor that
regulates genes to determine stamen and carpel development (Yanofsky et al. 1990).
DELLA protein plays a key role in negative regulation of gibberellin biosynthesis that
regulates many cellular and developmental events include flowering, pollen maturation,
and the transition from vegetative growth to flowering (Yoshida et al. 2014). DEFICIENS
is an ortholog of APETALA3 in Arabidopsis that functions in petal and stamen organ
identity (Zahn et al. 2005). PPM1 is ubiquitously expressed throughout vegetative and
reproduction tissues and may have diverse functions (Singer, Krogan, and Ashton
2007). WRKY has been related to several abiotic responses and accelerates flowering
by regulating FLOWERING LOCUS T and LEAFY (Phukan, Jeena, and Shukla 2016).
SERK1 (SOMATIC EMBRYOGENESIS RECEPTOR-LIKE KINASE1) acts as a
negative regulator of abscission metabolism and is required for anther development
(Lewis et al. 2010). Further investigations of these genes involved in flowering in wild
accessions of napiergrass are necessary to elucidate the basis for intraspecific variation
which is of great relevance to breeders. Future fine-mapping analysis by identifying
97
additional genetic variants in the QTL regions or by targeted resequencing of the QTL
intervals in a subset of individuals can lead towards accurate identification of targets for
improving flowering time in napiergrass and related crops. To verify the napiergrass
QTLs detected in our analysis, phenotyping should be repeated in multiple locations.
Breeders are more interested in QTLs that have large effect (Kearsey and Farquhar
1998) and the QTLs identified in this study can be used in the future for MAS in order to
breed late flowering napiergrass.
Conclusion
In summary, this study reports for the first time QTL analysis in napiergrass.
Three stable QTLs controlling number of flowers in napiergrass were identified, which
can explain 15%-20% of the phenotypic variation. We also identified three potential
QTLs controlling flowering time in napiergrass, which can explain 11%-14% of the
phenotypic variation. Gene models in pearl millet that mapped to two stable QTLs
harbored MADS-box transcription factors such as AGAMOUS and DEFICIENS, along
with other proteins such as DELLA, WRKY, and SERK1 that are involved in flowering
time regulation in plants. The QTLs detected in this study will be valuable information for
napiergrass breeding programs and help to understand the genetic basis of flowering.
This study confirms that flowering time and flower number are highly heritable traits in
napiergrass. Therefore, the late flowering napiergrass genotypes developed by Sinche
(2013) constitute a valuable germplasm resource to develop late flowering napiergrass
cultivars. In addition, validation of the putative candidate genes identified in this study
should lead to targets for genome editing to manipulate flowering time in napiergrass.
98
Figure 3-1. Histogram of number of flowers in the mapping population in 2012 in Citra, FL. X-axis represents the number of flowers and y-axis represents the count of plants. Average number of flowers of two parental lines, N122 and N190 are plotted. Data is adapted from Sinche, 2013.
99
Figure 3-2. Histogram of number of flowers in mapping population for 2013 in Citra, FL. X-axis represents the number of flowers and y-axis represents the count of plants. Average number of flowers of two parental lines, N122 and N190 are plotted. Data is adapted from Sinche, 2013.
100
Figure 3-3. Scatterplot of number of flowers between 2012 and 2013 in Citra, FL. X-axis represents the number of flowers in 2012 and y-axis represents the number of flowers in 2013. Data is adapted from Sinche, 2013.
0
100
200
0 10 20 30 40
Number of Flowers 2012
Nu
mb
er
of F
low
ers
20
13
101
Figure 3-4. Histogram of number of days to first flowering in 2012 in Citra, FL. X-axis represents the number of days to flowering (FT) and y-axis represents the number of accessions. Days to first flowering of two parental lines, N122 and N190 are plotted. Data is adapted from Sinche, 2013.
102
Figure 3-5. Histogram of number of days to first flowering in 2013 in Citra, FL. X-axis represents the number of days to flowering (FT) and y-axis represents the number of accessions. Days to first flowering of two parental lines, N122 and N190 are plotted. Data is adapted from Sinche, 2013.
103
Figure 3-6. Histogram of first flowering date in the mapping population in 2016 in EREC, Belle Glade, FL. X-axis represents the number of days to flowering (FT) and y-axis represents the number of accessions. Days to first flowering of two parental lines, N122 and N190 are plotted.
104
Figure 3-7. Scatterplot of first date of flowering between different years and locations.
X-axis and Y-axis represent the number of days to flowering (FT) in different years. Data for 2012 and 2013 are from Citra, FL and adapted from Sinche, 2013. Data for 2016 are from Belle Glade, FL.
R2
0.11
220
230
240
250
220 230 240 250 260 270
Days to Flowering 2012
Da
ys to
Flo
we
rin
g 2
01
3
R2
0.04
220
230
240
250
220 230 240 250 260 270
Days to Flowering 2012
Da
ys to
Flo
we
rin
g 2
01
6
R2
0.06
220
230
240
250
220 230 240 250
Days to Flowering 2013
Da
ys to
Flo
we
rin
g 2
01
6
105
Figure 3-8. Scatterplot between number of flowers and days to flowering in 2012 in Citra, FL. X-axis represents the number of days to flowering (FT) in 2012 and y-axis represents the number of flowers per line in 2012. Data is adapted from Sinche, 2013.
0
10
20
30
40
220 230 240 250 260 270
Days to Flowering 2012
Nu
mb
er
of F
low
ers
20
12
106
Figure 3-9. Scatterplot between days to flowering and number of flowers in 2013 in Citra, FL. X-axis represents the number of days to flowering (FT) in 2013 and y-axis represents the number of flowers per line in 2013. Data is adapted from Sinche, 2013.
R2
0.41
0
100
200
220 230 240 250
Days to Flowering 2013
Nu
mb
er
of F
low
ers
20
13
107
Figure 3-10. Linkage map of male parent N122 showing potential and stable QTLs. QTLs for number of flower (NF) are color coded in blue and that for flowering time (FT) are colored in green. A black bar means a GBS-SNP marker. Markers are labeled on the left and genetic distance (centimorgan as unit) are labeled on right. LG = Linkage group.
Linkage map of male parent N122
S3_66915283 0.0S3_66915258 0.6
dT6838 5.5dT44555 6.5dT27134 6.9
S3_66915278 7.5S1_7068544 8.9
dT3437 10.2dT46011 10.4dT38476 13.3dT35347 13.8dT33533 14.1dT28618 14.7dT39822 15.3dT28250 15.4
C5_27253496 16.0TS7_74798545 16.5
dT31263 19.7dT6267 21.1
dT30050 21.8C1_119093525 C1_119093531 22.0
dT37316 22.3dT12054 22.6
S1_71014695 24.0dT17414 24.9dT29057 25.8
S3_284635389 26.2dT2475 27.9
dT18370 dT12390 28.0dT41113 28.1dT30582 28.3
dC_618193 30.4dT44577 31.3dT34797 32.9dT51180 33.5dT28611 36.3dT35867 37.1dT23729 40.2dT34784 41.0dT37028 41.5
S1_193475897 41.8dT36388 42.3dT3762 42.5
dT38724 48.8TS1_74139971 dT11106 49.2
dT6132 49.9dT7984 50.2
dT13772 50.6dT53762 51.2dT8946 51.8
dT18605 52.3dT29128 53.6dT52531 56.0dT41949 60.2dT47752 68.8
dT27185 84.6
S4_21013305 91.5
dT8342 114.3dT52469 116.0dT37356 117.2dT35128 118.2dT53136 118.6dT39139 120.7dT36338 122.1dT19693 124.8dT44366 125.5dT28974 126.0
S1_249594415 126.5dT5535 127.0dT8792 127.2
dT42596 127.4dT48305 128.0dT32706 129.2dT49414 129.9
C1_251517594 130.2dC_259973 130.3
S1_251483789 130.4dT41404 130.8dT42403 131.3dT9863 131.6
S1_260924412 133.5dT35453 134.2dT45297 134.7dT28492 135.8dT47753 136.7dT32523 137.3dT21677 137.7
dC_2450640 138.4dT48962 139.3dT50248 139.6dT27136 140.5dT32605 141.2
S1_267892930 149.1
qN
FE
-1-1
qN
FE
-1-2
qN
FE
-1-3
qF
TE
-1-1
qF
TE
-1-2
LG1
dT11902 0.0dT17348 0.5dT46533 1.3dT53963 2.5dT38576 3.2
C3_124558897 3.3dT10674 4.2
TS6_10754586 4.7dT10631 4.8dT27467 6.0dT2555 8.6
dT45550 9.1dT11037 10.5
S5_31045825 10.8S6_35473070 12.0
dT35938 14.3C6_39611566 14.8
dT6835 15.9dT16233 20.7dT38911 21.0dT38805 23.3dT6794 24.4
dT14931 25.2dT16992 25.7dT46041 26.6dT16346 29.2dT47787 29.6dT42989 32.8dT50566 34.0dT9987 36.3
dT29726 41.8
dT23975 44.8dT31396 46.2dT43235 47.3
dT2439 49.2
dT39167 52.1
dT35142 56.8
dT5958 60.6dT8821 61.9
dT25121 69.1dT39882 69.2
dT30721 dT19708 69.3dT36233 69.5dT18641 71.2dT28669 71.7
dT15498 dC_1941dC_5486367 dC_428936
dT50064 dT4146074.3
dT5222 74.5dT43295 74.8dT31359 75.0dT44933 77.3dT33834 81.5dT46138 82.2
dT987 91.3dT17187 92.4
dT3299 95.9dT29166 96.5
dT2229 98.5
dT19151 108.9dT22571 110.8dT36501 111.1dT11509 111.6dT4155 111.9dT1086 112.6
dT20499 112.8
S6_118469241 119.5dT14954 121.1dT39428 123.1dT19912 124.3dT53574 124.7dT43321 127.6dT34936 127.9dT2297 129.9dT8340 131.0
dC_234325 132.0dT20274 134.9dT22921 135.8dT30440 139.1dT14807 140.0dT30703 141.4dT6973 142.1
dC_196800 142.9dT18761 143.2dT36412 143.8dT24998 144.5dT25533 145.0dT29143 145.7
S6_55827481 146.2
dT43184 153.6C6_200342258 154.5
dT9665 155.1dT18516 156.7dT18539 157.3dT13751 158.5dT50350 160.4
dT22426 163.1
dT17747 168.9
dT45130 171.6
dT29238 173.2
dT13151 177.6
dT31763 184.3
qN
FE
-2-1
LG2
S4_192354592 0.0
dT43261 3.2
dT3062 5.4
dT20572 7.7
dT9434 9.3
dT50501 14.2
dT13383 16.9
dT25619 20.8
dT15831 23.7dT29846 26.5
dT20879 dT3030dT27514
27.1
dT40535 dT18538 27.3dT48489 27.7
C4_200698121 28.2dT4430 28.9
dT23119 29.7dT43805 30.0dT3891 30.9
dC_13461 31.7S4_116088663 33.3
dT13255 37.0dT32416 38.2
C4_90405411 39.4dT8320 39.9
S4_29656682 41.2dT42579 44.4dT4620 44.5dT1631 47.5
dT43233 52.7dT18870 53.3dT8454 53.8dT7751 55.4
dT29230 66.0dT4496 66.5dT6466 67.8
dT16151 68.9TS4_49707946 69.2
dT11273 70.1dT43759 70.7
S4_109511769 71.8TS4_79794887 76.3
dT36845 76.6dT33420 77.6dT25340 77.7
dC_124707 79.3dT31777 84.2
S4_17921468 85.4dT39406 85.9dT45232 86.6
dT4301 C4_13275000 87.4dT24376 89.5dT36267 89.7dT26600 90.2dT52456 91.1dT10193 91.7dT34107 92.0dT22232 92.6dT9256 94.2
S4_17796331 95.7
qN
FE
-6-1
qN
FE
-6-2
LG6
S2_174467255 0.0
S1_203215691 14.4
C2_174467255 dC_58121dC_58157
18.9
dT1325 20.5dT24339 21.7dT33835 22.9dT10459 23.8dT21208 24.3dT41386 24.5dT16455 26.4dT38126 26.9dT20154 27.2dT24498 28.0
dT50610 37.5
dT44347 41.9dT6436 42.3dT4353 43.7
dT5027 48.8dT23432 49.4dT39900 51.7
dT861 52.5dT23905 53.4dT42142 55.0
S7_227510 55.3
S7_161179045 S7_161179072 57.7
dT21588 60.1
dT14775 64.4
dT34305 66.6
dT5751 69.4
dT17613 77.8
dT7877 83.4
dT7992 88.3dT22362 88.7
S7_168784407 90.3S7_168784355 91.7
dT26372 98.5
S7_146917990 109.0
S7_100737594 114.2
dT46088 118.4
dT36827 120.7dT36253 121.3
C7_38012075 S7_52134705 124.9
S7_52134761 126.6
S7_52134764 128.5
dT25902 136.4dT43823 137.3
dT19286 145.8dT43692 146.0
S5_1337832 147.8dT23441 149.3dT28159 150.8dT9688 151.5
dT11814 156.0dT16180 157.3dT26922 157.7
S7_119895303 159.3dT8614 160.3
dT29758 161.1dT10112 162.1
dT6553 173.8
dT37830 178.0S7_66630588 179.5
dT39337 180.7
dT44668 183.3dT25362 183.6dT28445 183.8dT9471 185.2dT3618 186.5
qF
TE
-7-1
qF
TE
-7-2
LG7
108
Figure 3-11. Linkage map of female parent N190 showing potential QTLs. QTLs for number of flower (NF) are color coded in blue and that for flowering time (FT) are colored in green. A black bar means a GBS-SNP marker. Markers are labeled on the left and genetic distance (centimorgan as unit) are labeled on right. LG = Linkage group.
Linkage map of male parent N122
dT20387 0.0dT41575 3.7
dT1702 dT3027 4.6dT909 4.8
C1_10467974 5.2dT38795 5.4
S3_64306230 5.8dT3328 6.1
dT49252 6.4dT30456 6.5dT16886 6.7dT53891 7.0dT16764 7.2dT35147 8.3dT49010 8.6dT2473 8.9dT1951 9.6
dT16527 9.9C1_220011116 12.0
dT50264 20.4dT29470 21.6dT32431 22.3dT9029 22.5
dT50826 22.7dT47548 23.0
dT45924 dT7985dT24786 dT732
23.5
C1_102656250 23.6dT13338 23.7dT19473 23.8dT18485 23.9dT46948 24.3dT51201 24.5dT23838 24.8dT23364 25.2dT23237 30.6
dT28498 38.0dT51779 38.5dT43458 39.3dT6993 41.3
dT29597 42.0dT5196 43.9
dT14180 44.3dT48800 44.6dT3431 45.2
dT48497 45.3S1_17257463 45.8
dT9167 46.1dT28049 46.3dT48243 46.6
S1_17257559 46.9S1_17257551 47.2
dT18424 47.8dT973 48.2
dT8459 50.7dT33654 50.8dT48112 51.1dT53672 52.0dT12272 53.7
dC_359278 58.8C1_133709193 58.9
dC_339765 59.0S1_171369178 59.2
dT40041 59.7dT18064 61.2
dT971 61.9dT22725 62.2
dT10843 dT19448 62.3dT47057 63.4
C1_234266014 65.4TS1_273230575 66.0
dT30844 66.4dT7314 67.2dT766 68.5
dT36979 69.2S1_217117736 73.5
dT30297 75.1
dT3813 78.0
S1_241274562 87.0
qN
FL-1
-1
qF
TL-1
-1qF
TL-1
-2
LG1
S1_6210777 0.0
S5_128073888 4.6
S1_6210761 8.3dT24515 11.1dT17924 15.6
TS5_114098156 16.0dT16773 16.2dT40995 16.5dT15817 16.8dT14651 18.3dT18480 18.8
TS5_128053793 dT13827C5_156998414 C5_112498054
dC_2426719.1
dT27345 dT30788 19.2dT9315 19.4
S5_40839446 19.6dT4737 20.2
dT31217 22.1dT32027 22.2dT46427 22.4dT32695 23.0
dT43636 29.3
dT31735 33.3
dT50145 38.1
dT43048 41.0C5_132196063 41.2
dT20788 41.7dT32207 41.9dT24792 43.5
dT46652 46.4
dT27717 49.6dT15391 50.1dT1524 51.2
dT14375 53.4dT31410 54.0dT31722 54.3
dT18219 59.0dT36949 59.7
dT24276 63.6
dT7856 72.9
dT5489 76.2
dT43892 77.6
dT48559 82.4
dT42534 91.8
dT49410 93.8
dT52149 94.9
dT32206 99.6
dT6664 101.1
dT8114 104.1
dT53440 105.2
dT40421 108.2dT14806 108.5dT22647 108.9
S5_31904873 109.7dT39799 110.2dT3053 110.6
S5_26740824 111.4dT4924 111.7
dT37762 113.4
dT26721 118.4dT26387 119.1dT6733 119.8
dT17794 120.1S5_16898738 120.9
dT37016 121.4dT47995 123.0dT34310 123.7dT51527 124.4
TS5_13509703 124.6dT54119 125.1dT53202 125.4dT32125 126.4dT27794 126.6dT2435 127.9
dT46426 129.8dT20384 130.1
C5_3132406 131.6dT15117 131.9dT29259 132.8
dT780 133.3dT15952 134.3dT26168 135.2dT36121 137.1dT32758 139.0
dT30470 142.4
qN
FL-4
-1qN
FL-4
-2
LG4
dT49930 0.0
S4_131830044 G4_131830067 6.6S4_131830067 6.7S4_131830066 7.2S4_131830063 7.3
dT22876 8.3dT12269 9.4
C4_131830077 9.5dC_61912 dC_55022 9.7
dT22567 10.3dT26116 10.7dT38883 11.0
dT36340 dT10040 11.3dT26884 11.6
dT10739 dT621 11.7dT17325 S7_89250999 11.9
dT17578 12.0dT25711 12.1dT5186 12.3dT8749 12.6
dC_10270 13.2dT649 13.7
dT53247 13.9dT53022 dT12194 14.0
S7_39646339 15.6S7_39646309 15.9S7_39646292 16.2
dT33677 16.7dT3510 17.3
dT19298 17.9dT1762 21.7
dT41196 22.3dT30018 22.7dT40278 23.3
S7_83881022 23.7dT7491 25.5
dT38871 28.1dT29364 35.0
dT8910 50.9
dC_223673 59.4
dT43632 63.3dT34571 64.0dT3675 64.6
S7_213184560 64.9dT22858 65.4dT32036 66.2dT29409 67.8
dT19025 71.6dT7711 72.5
dT39297 74.1
dT42995 78.6
dT47137 80.2
dT37923 90.6
qN
FL-1
1-1
qN
FL-1
1-2
LG11
109
Table 3-1. Descriptive statistics of flowering date and number of flowers for 185 F1
hybrids of a cross (N190 N122) at Citra, FL., in 2012, 2013 (Sinche 2013) and 2016.
Statistic Flowering datea Number of flowers (plot-1) b
2012 2013 2016 2012 2013
Min 10/25 10/26 10/23 0 5
Max 12/06c 12/03 12/02 46 256
Mean 11/15 (240) d 11/13 (238) d 11/07 (232) d 6 61
SE 1 d 0.5 d 0.05 d 0.76 2.86
H2 0.60 0.87 0.64 0.85 0.79
a Flowering date: mm/dd
b 1.8 m long plots.
c The harvest concluded before the first predicted frost on 12/06/2012 and several genotypes did not flower until that day.
d Value in parenthesis indicates number of days to first flowering counted from March 20 of that year.
SE Standard error H2 Broad sense-heritability estimated on entry mean basis
110
Table 3-2. Detailed information about the QTLs for number of flowers. LG = Napiergrass linkage groups, Peak markers represent the marker at the highest peak of QTL, LSI represents the LOD-1 support interval in cM.
Parent QTL code LG Year Peak markers Peak LOD PVE
(R2)
Allele
dir.
LSI (cM) Flanking Markers
N122 qNFE-1-1 1 2012, 2013 dT37356 8.13 0.15 +
116.9 cM - 118.2
cM dT52469, dT35128
N122 qNFE-1-2 1 2012, 2013 S1_249594415 10.76 0.20 +
123.7 cM - 126.8
cM dT36338, dT5535
N122 qNFE-1-3 1 2012, 2013 dT35453 9.10 0.17 -
133.5 cM - 134.6
cM
S1_260924412,
dT45297
N122 qNFE-2-1 2 2013 TS6_10754586 3.48 0.08 + 3.1 cM - 4.8 cM dT38576, dT10631
N122 qNFE-6-1 6 2012 S4_17921468 3.64 0.06 + 84.3 cM - 86 cM dT31777, dT45232
N122 qNFE-6-2 6 2012 dT10193 3.45 0.06 + 91.1 cM - 91.9 cM dT52456, dT34107
N190 qNFL-1-1 1 2013 dT3328 4.04 0.10 + 5.6 cM - 6.7 cM dT38795, dT53891
N190 qNFL-4-1 4 2012 S1_6210761 3.66 0.07 + 7.6 cM - 8.8 cM
S5_128073888,
dT24515
N190 qNFL-4-2 4 2012 dT40995 4.78 0.10 + 16.2 cM - 16.7 cM dT16773, dT15817
N190 qNFL-11-1 11 2013 dT8910 4.22 0.08 + 47.5 cM - 52.9 cM dT29364, dC_223673
N190 qNFL-11-2 11 2013 dT43632 3.18 0.06 - 57.9 cM - 64.2 cM dT8910, dT3675
111
Table 3-3. Detailed information about the QTLs for flowering time. LG = Napiergrass linkage groups, Peak markers represent the marker at the highest peak of QTL, LSI represents the LOD-1 support interval in cM.
Parent QTL
code
Linkage
Group
Year Peak
markers
Peak
LOD
PVE (R2) Allele dir. LSI (cM) Flanking Markers
N122 qFTE-1-1 1 2013 dT53136 6.77 0.14 +
118.3 cM - 120.8
cM dT35128, dT39139
N122 qFTE-1-2 1 2013 dT48305 5.40 0.11 -
126.5 cM - 128.3
cM
S1_249594415,
dT32706
N122 qFTE-7-1 7 2012 dT44347 3.91 0.09 - 31 cM - 42.1 cM dT24498, dT6436
N122 qFTE-7-2 7 2016 dT39337 5.57 0.12 -
178.3 cM - 182.1
cM dT37830, dT44668
N190 qFTL-1-1 1 2013 dT41575 3.60 0.07 - 3.3 cM - 4.6 cM dT20387, dT3027
N190 qFTL-1-2 1 2013 dT1951 3.33 0.07 - 8.9 cM - 11.2 cM
dT2473,
C1_220011116
112
Table 3-4. List of putative flowering related genes from the genome of pearl millet (Varshney et al. 2017). Gene name Description Gene ID Chr Start position End position Exon
count QTL
AGAMOUS AGAMOUS-like protein; K09264 MADS-box transcription factor
Pgl_GLEAN_10032931 chr1 268933236 268935744 7 qNFE-1-3
DELLA DELLA domain GRAS family transcription factor rga-like protein; K14494 DELLA protein
Pgl_GLEAN_10026530 chr1 248097573 248099492 1 qNFE-1-2
DEFICIENS Floral homeotic protein DEFICIENS, putative; K09264 MADS-box transcription factor
Pgl_GLEAN_10000416 chr1 250352485 250350575 6 qNFE-1-2
PPM1 PPM1; MIKCC MADS-domain protein PPM1; K09264 MADS-box transcription factor
Pgl_GLEAN_10029410 chr1 253208367 253209680 1 qNFE-1-2
WRKY K13424 WRKY transcription factor 33
Pgl_GLEAN_10027907 chr1 242557234 242558164 3 qNFE-1-2
SERK1 SOMATIC EMBRYOGENESIS RECEPTOR-LIKE KINASE 1; kinase/ transmembrane receptor protein serine/threonine kinase
Pgl_GLEAN_10000999 chr1 250195809 250194274 2 qNFE-1-2
113
CHAPTER 4 EVALUATE THE GENETIC BACKGROUND OF FLOWERING TIME IN A
NAPIERGRASS GERMPLASM COLLECTION
Introduction
Napiergrass (Cenchrus purpureus) is an important forage and biofuel candidate
(Singh, Singh, and Obeng 2013). However, there has been relatively little effort in
improving napiergrass as compared to cereal crops for increased yield and agronomic
characteristics (Wanjala et al. 2013). Napiergrass cultivation in the US is limited by its
potential for invasiveness (Sollenberger et al. 2014). Typically, napiergrass is very
persistent and only few diseases like napiergrass stunt disease and napiergrass head
smut disease have been reported with significant impact on biomass yield or crop
persistence (Farrell, Simons, and Hillocks 2002).
Existing genetic diversity of napiergrass needs to be assessed in order to
strengthen breeding programs for development of high yielding and late flowering
cultivars that can be considered as non-invasive (Wanjala et al. 2013; Sinche 2013).To
initiate this, there is a need to have accurate information on the available germplasm to
identify clones that form the base population for breeding late flowering lines. The first
step of breeding for flowering time is to create variability within the germplasm so as to
widen the base population of the candidate clones for selection (Juma 2014). Variability
can be due to intrinsic reasons or due to random genetic drift, natural selection,
mutation, gene flow and transfer (Y. Xu et al. 2017). This variability can be determined
by assessing the level of polymorphism between accessions and within each accession.
This information can be used for clone development. To determine the variability of the
available germplasm, stable and informative molecular markers are critical to study the
114
genetic diversity. Molecular characterization of germplasm will increase breeding
efficiency by providing important genetic information of the breeding materials and will
ultimately help in identification and introgression of desirable agronomic traits into high
yielding clones through marker assisted selection (MAS). This will lead to increased
napiergrass yield with decreased invasiveness.
Napiergrass cultivar discrimination has been mostly done based on
morphological and agronomic characteristics that lead to inconsistency in identification
of specific accessions (Struwig 2007). Consequently, many cultivars of napiergrass are
described with more than one name/identifier. Marker informed elimination of
redundancy in collections of napiergrass accessions will facilitate the maintenance of
the collection and its use in crop improvement programs. In this regards, molecular
markers have proven to be very effective to distinguish between morphologically related
individuals of the same species (Wanjala et al. 2013).
Hence, genetic assessment of napiergrass accessions with molecular markers
was initiated several decades ago. Characterization of different napiergrass
germplasms has been carried out using Restricted Fragment Length Polymorphism
(RFLP) (Smith et al. 1993), Random Amplification of Polymorphic DNA (RAPD) (Lowe
et al. 2003), Amplified Fragment Length Polymorphism (AFLP) (Harris-Shultz,
Anderson, and Malik 2010; Wanjala et al. 2013), inter-simple sequence repeats (ISSRs)
(de Lima et al. 2011), sequence-related amplified polymorphism (SRAP) (X.-M. Xie et
al. 2009), and simple sequence repeat (SSR) markers (Azevedo et al. 2012; López et
al. 2018; López et al. 2014). Currently, single nucleotide polymorphism (SNP) markers
are gaining popularity as they are ubiquitous in the genome, occur in a large number
115
with high density, relatively inexpensive, codominant, and can act as functional markers
if located in coding regions (Rafalski 2002; Ganal, Altmann, and Röder 2009). Further,
genome wide SNP markers can be used as diagnostic tools for fingerprinting
germplasm, to correct discrepancies that are inherent in taxonomic methods based on
morphological characteristics and to estimate genome composition of clones (Yang,
Song, et al. 2018).
Flowering time is very important to plant breeders because agronomic traits like
biomass yield and quality depend on the timing and intensity of flowering (Veeckman et
al. 2016). Flowering time is a key factor in plant adaptation and is linked to various
attributes like plant height, yield, and number of leaves (Durand et al. 2012). In many
species, flowering is induced in response to day length. Different flowering responses
are categorized as short-day, long-day, intermediate-day, or day-neutral based on the
day length requirement (Schlegel 2009; Bastow and Dean 2002). Napiergrass belongs
to the short-day plants (Osgood, Hanna, and Tew 1997; Singh, Singh, and Obeng
2013), in which flowering is favored by day lengths shorter than the corresponding
nights.
Early flowering in perennials limits vegetative growth, and delayed flowering will
hinder seed development to progress completely before the onset of adverse weather
(Grabowski et al. 2016). Delaying flowering could potentially lead to increases in
biomass yields as seen in other grasses like switchgrass where a 10 days delayed
flowering increased biomass by > 25% (Price and Casler 2014). Flowering time is an
important target trait in napiergrass breeding programs because different flowering time
helps to adapt cultivars in different geographic locations or seasons. It is also important
116
because flowering has been associated with invasiveness of napiergrass in Florida
(FLEPPC 2011). High number of wind-dispersed seeds has contributed to categorizing
napiergrass as an invasive species in Florida, thus limiting its utilization as an industrial
crop for forage or biofuel (López et al. 2014). For this reason, development of cultivars
that are late flowering is critical for napiergrass management in the field because late
flowering lines can be harvested before flowering, as development of viable seeds is
impaired by low temperatures in Florida during Nov-Feb. Studying flowering time
variation in napiergrass can help to identify potential genes that promote or hinder seed
formation in napiergrass. Mining genome sequences that are available for several grass
species for flowering related genes and their characterization can also identify
candidate genes. These candidate genes can be investigated in different populations for
late flowering. In addition to this, integrating genomics with conventional breeding will
shorten the breeding cycle for selection.
The substantial variations in flowering time within the population of napiergrass
can be exploited to not only select lines with desirable traits but also to identify
sequence variations that are associated with it or could be causal. By characterizing the
relationships between genetic variations and flowering time we might be able to identify
loci affecting flowering time variation in napiergrass. Flowering time can be controlled by
a large number of quantitative trait loci (QTLs) with small effects (Buckler et al. 2009).
Genetic and genomic approaches that have been used in other species to characterize
flowering time genes might not be feasible in napiergrass due to its large genome size
and high polyploid level. This challenge can be mitigated by using genome reduction
approaches like exome-capture sequencing methods (Evans et al., 2014). Exome
117
sequencing targets only genic regions of the genome with high depth and can produce
gene-level resolution of genome-wide patterns of sequence variation (Grabowski et al.
2016).
To acquire a global perspective of sequence variations and to obtain an accurate
population structure of napiergrass germplasm, we deeply sequenced the coding
regions of 94 accessions from the Tifton nursery. Then we used genome-wide SNPs
identified to evaluate the genetic relationships of the germplasm as well as performed
genome wide association studies targeting flowering time in napiergrass.
Materials and Methods
Plant Materials and Phenotyping
A total of 94 napiergrass accessions from the germplasm collection available at
the USDA-ARS, Tifton nursery were used in this study. Nodes from those accessions of
napiergrass were cut and planted in 6” pots on August 19, 2015 in Everglades
Research and Education Center (EREC), Belle Glade, FL. Transplanting of the pots
was done on October 12, 2015. Plots were established in a randomized complete block
design with three replications and spacing of 4 ft x 4 ft plant to plant as well as row to
row distance were maintained. Emergence of the first flower was noted every week
during October – December 2016. Days to flowering was calculated as the number of
days between the first appearance of the flower and vernal equinox (March 20)
(Lambert, Miller-Rushing, and Inouye 2010).
DNA Extraction
Young and healthy leaf tissues were harvested from each individual of the
germplasm collection. DNA extraction was done following the protocol described
118
previously (Dellaporta, Wood, and Hicks 1983). The extracted DNA samples were run
on a 2% agarose gel to check the quality and quantified with PicoGreen to meet the
requirements of exome sequencing.
Targeted Candidate Genes
Genes related to flowering in different monocots and other species were mined
using various publications (Table 5-1). In addition, Refseq database was searched for
“flowering” and filtered for “Green plants” to get potential flowering related genes (Pruitt,
Tatusova, and Maglott 2007). All gene models present in the pearl millet genome
(Varshney et al. 2017) were used for probe design to achieve a genome-wide coverage
of all coding sequences (CDS).
Probe Design
The protein sequences for flowering genes were clustered using CD-HIT
(Weizhong Li and Godzik 2006). Probes of 120 bp long with an overlap of 60 bp were
developed using Emboss 6.5.7 (Rice, Longden, and Bleasby 2000). These probes were
mapped back to the pearl millet genome (Varshney et al. 2017) using BLAT (Kent
2002). A hit was defined under cutoff: e-value ≤ 1e-05; alignment identity = alignment
length * percentage of identity ≥96 (120 bp * 80%=96 bp). Probes that hit the genome
less than three times were selected and probes containing repeats were removed. 2kb
promoter region and 500 bp downstream region of the flowering genes were also
included for probe design with the same criteria.
A relatively relaxed criteria was chosen to target missing flowering time related
genes. For this, 120 bp long probes were designed with 90 bp overlap and good probes
were selected with the criteria (Alignment * Identity) / 120 >= 80. The probes that hit the
119
genome less than 4 times were selected and probes containing repeated regions were
removed.
Probes were also designed to target the coding regions of the whole genome.
For this, sequences of 38,579 genes of pearl millet were used to retrieve 120 bp probe
sequence with maximum allowable overlap of 30 bp. Good probes were selected under
the criteria alignment * identity / 120 >= 80. Probes with multiple hits on the genome or
containing repeat regions were removed. A maximum of 2 probes were kept for each
CDS.
A relaxed criterion was implemented for genes, where no probes were designed
according to the standards above. For those genes, probes were designed with 90 bp
overlap and good probes were selected as alignment * identity /120 >= 80. Probes with
up to 2 hits on the genome were kept and probes containing repetitive regions were
discarded.
Probe Synthesis, Selection, and Sequencing
The designed probes were submitted to RapidGenomics LLC (FL, USA) for
quality check. Probes that passed quality criteria of RapidGenomics were then selected
using the following criteria: for flowering related genes, a total of 18 probes were
targeted per gene while for genome wide CDS, 1 probe per gene was targeted. Four
probes were targeted per gene for 2 kb promoter region and 1 probe per gene was
selected for 500 bp downstream region of the gene. Finally, a total of 37,000 probes
were submitted for synthesis. The synthesized probes were used to capture the DNA
fragments of the germplasm collection. The captured DNA fragments were sequenced
using the Illumina HiSeq 3000 platform (150 bp paired-end reads). The probe synthesis,
120
library preparation, target enrichment, and sequencing were performed by
RapidGenomics LLC (FL, USA).
Sequence Read Trimming and Mapping
Raw reads were trimmed using Trimmomatic (Bolger, Lohse, and Usadel 2014).
Reads that contained more than 50% low-quality bases (Q20) were removed and
adaptor sequences were trimmed. Trimmed reads were aligned to the pearl millet
genome v1.1 (Varshney et al. 2017) using BWA-mem (H. Li and Durbin 2010). For SNP
calling, we retained only uniquely mapped reads with a combination of mapping quality
‘0’ and ‘XA:Z’ tag (both indicates multiple alignment).
SNP Calling
SNPs were called using GATK 3.2.2 (McKenna et al. 2010) with settings (-T
UnifiedGenotyper -glm BOTH -R -ploidy 4 -mbq 20). For each sample, homozygous
SNPs required a depth of ≥11; heterozygous SNPs required a depth of ≥31; and at least
2 reads for minor allele were required to make a call. At the population level a mapping
quality of ≥30; minor allele frequency ≥0.05; and call rate of ≥95% was employed.
Population Structure
In order to infer population structure of the germplasm accessions, we used
DAPC available in the adegent package for R (Jombart 2008). Groups were clustered
using k-means clustering, where the best k minimized the Bayesian Information
Criterion (BIC). GWASpoly (Rosyara et al. 2016) was used with six different models for
GWAS using sequence variants that had a call rate ≥ 95% and MAF ≥ 0.05. Molecular
Evolutionary Genetics Analysis version 6.0 (Tamura et al. 2013) was used to infer
genetic relationships among the germplasm collection and to perform phylogenetic
121
analysis. Less than 5% alignment gaps, missing data, and ambiguous bases were
allowed at any position. Neighbor-Joining method was used to infer the evolutionary
history. The evolutionary distances were computed using the p-distance method and
are in the units of the number of base differences per site (Tamura et al. 2013).
Results
Candidate Genes Related to Flowering
A total of 3,542 flowering time related genes were curated from different
organisms including Arabidopsis, Brachypodium, maize, rice, Setaria, sorghum and
switchgrass (Table 4-1). This list included 1,285 flowering related genes retrieved from
the Refseq database (Pruitt, Tatusova, and Maglott 2007). CD-HIT clustered the
flowering genes into 1,279 non-redundant genes that were further compared to peptides
from the pearl millet genomes using BLAT. In pearl millet, 506 genes had BLAT hits to
the non-redundant flowering amino acid sequences. Thus the 506 genes were targeted
for the probe design.
Probe Design
The CDS sequences of 506 ortholog genes in pearl millet related to flowering
were subjected to probe design. A total of 8,314 probes were designed and 4,908
probes targeting 473 pearl millet flowering time related genes remained after filtering.
The number of probes designed per gene ranged from 1-91 (Figure 4-2) and the
number of probes designed per gene was proportional to the size of the gene (Figure 4-
3). To further cover the genes without probe designed, the selection criteria were
relaxed (see Materials and Methods) and an additional 9,099 probes were designed
targeting 422 additional genes.
122
We extracted 2kb region upstream of the flowering genes to design probes for
the promoter regions. A total of 3,215 probes were designed targeting promoter regions
of 243 pearl millet genes. We further extracted 500bp downstream of the flowering
genes and designed probes from the terminator regions. A total of 1,025 probes were
designed from the terminator regions targeting 239 flowering related genes.
Probes were also designed to target the coding regions of the whole genome.
Thus the sequences of 38,579 gene models annotated in the pearl millet genome v1.1
(Varshney et al. 2017) were included to design genome wide probes. The length of the
genes ranged from 100 bp to 5600 bp (Figure 4-1), with most of them in the range of
500 bp -1600 bp. A total of 53,525 probes targeting 29,015 out of 38,579 genes were
designed. For the genes that couldn’t be targeted, we relaxed selection criteria and
further design 7,306 probes that targeted additional 4,318 genes (Materials and
Methods).
Thus, a total of 78,988 probes were totally designed and were submitted for
quality check, after which 63,967 probes passed the quality test provided by Rapid
Genomics. From these probes, a total of 37,000 unique probes were further selected for
synthesis and sequencing. This final probe set targeted a total of 818 flowering related
genes with a maximum of 18 probes per gene. These genes were distributed on all
chromosomes of the pearl millet genome (Figure 4-4). The probe set also targeted a
total of 28,519 genome wide CDS with 1 probe per CDS. Promoter region of 240 genes
were targeted with a maximum of 4 probes per gene and terminator region of 231 genes
were targeted with a maximum of 1 probe per gene.
123
Sequence processing
A total of 375,135,615 paired-end reads (2×150 bp) were generated by Illumina
sequencing. The number of paired-end reads per sample ranged from 2.3 M to 5.6 M,
with an average of 3,990,804 read pairs per sample (Figure 4-5). A total of 354,016,510
(94.35%) read pairs passed the quality control. From the good quality reads, a total of
20,901,032 raw SNPs were called by GATK (McKenna et al. 2010). After filtering for
quality, depth, minor allele frequency (MAF), and call rate, a total of 78,129 high quality
SNPs remained that were subjected to genome wide association analysis using
GWASpoly (Rosyara et al. 2016).
A total of 2,893 SNPs (one SNP per 10 Kbp) were selected to perform DAPC.
Based on the Bayesian Information Criteria (BIC), the germplasm could be clustered
into 3 clusters (k=3) (Figure 4-6). Using these SNPs, the germplasm could be projected
into three linear discriminants (LD)s (Figure 4-7) that corresponded to their geographic
origin. Group 1 had accessions originating from Africa, group 2 had accessions that
originated from America (central and south America), and group 3 had a mix of lines
from Asia and America. However, three accessions whose origin were in Taiwan (Asia)
were grouped into the same group as accessions originating from Africa (group 1).
Seven accessions whose origin was listed as America were also grouped into the
African group 1. Group 2 exclusively contained accessions originating in America.
Group 3 contained accessions from the Americas and Asia, with one exception that
came from Africa.
124
Phylogenetic analysis
The phylogenetic analysis involved 94 samples in MEGA6 by using 2,872 SNPs
(one SNP every 10 Kbp). Positions that had less than 95% site coverage were
removed. Evolutionary analyses showed that most of the lines could be assigned based
on the geographic locations, however, there were some mixtures (Figure 4-8). Group 1
mostly contained lines that originated in Africa, while group 2 contained lines with origin
from America. Many accessions in group 2 had unknown origin. Group 3 included a mix
of African, Asian, and American lines. Asian lines were grouped in group 1 and group 3.
Genome wide association analysis
Average days to flowering in the germplasm collection ranged from 232 days to
279 days (Figure 4-9). The heritability for days to flowering was 0.58. All the high-quality
SNPs after filtering (78,129) were used for genome wide association analysis by using
GWASpoly with six different models (Figure 4-10, Figure 4-11). For every model, the
majority of p-values lie close to the 1:1 dashed line (Figure 4-10). For days to flowering
(DTF), the strongest association signal was observed with the 1-dom-alt model (Fig 4-
11) for a marker on Chromosome 5 at location 129,365,754 with an estimated marker
effect of -16.79. Other models didn’t give significant associations (Fig 4-11).
Subsequently, we explored the region 100kb upwards and 100kb downwards of this
marker to identify potential candidate genes. Ten gene models were located within this
region and we identified potential flowering candidate genes including calcium-binding
protein CML and enoyl-CoA hydratase in this region.
125
Discussion
Dissecting genetic loci controlling flowering time is critical for napiergrass
breeding as flowering is related to invasiveness and biomass accumulation. In this
research, we applied TES for high throughput genotyping of tetraploid napiergrass. We
were able to investigate sequence variants using TES which reduced genome
complexity by selectively sequencing only the coding regions. After applying quality
filter, we identified 78,129 high quality SNPs in the germplasm collection that serve as
valuable and novel genomic resource for napiergrass breeding programs. Moreover, we
performed GWAS in the napiergrass germplasm collection using these SNPs and
identified one SNP as well as two candidate genes that was associated with flowering
time.
We applied DAPC to analyze the population structure of the napiergrass
germplasm collection. DAPC is a multivariate method to analyze genetic structure and
does not rely on Hardy-Weinberg equilibrium assumptions or linkage disequilibrium
assumptions (Jombart 2008). We selected 2k high quality and evenly distributed SNPs
for the analysis and inferred that three groups existed in the germplasm collection. This
is in contrast with results from (Kandel et al. 2016) who inferred 5 groups in a subset of
this population. Those results were based on only 29 SSR markers in contrast with our
study that used 2k uniformly distributed SNPs. In this study, we had 24 lines that were
not used in the previous study (Kandel et al. 2016). Out of these, 14 lines were of
unknown origin. Our study grouped accessions N43 and N51 into the same group
similar to previous study that grouped biofuel type and naturalized napiergrass into
126
different groups (López et al. 2014). However, our results grouped N13 in a separate
group compared to N43 and N51.
We identified a total of 20 million raw SNPs and compiled a comprehensive
sequence variation database for napiergrass that included 78k high quality SNPs. This
dataset provides accurate genotypes for the germplasm collection and serves as the
foundation for population genomics analyses in napiergrass, like gene mapping and
GWAS. We were able to identify potential candidate variations using GWASpoly
(Rosyara et al. 2016), which is primarily designed for autopolyploid species, however, it
has also been successfully used in diploids, as well as allopolyploids like wheat (Phan
et al. 2018).
The regulation of flowering time is important for plant adaptation and biomass
production. In this study, we used genotypes derived from exome-capture sequencing
and flowering time information from a germplasm collection of napiergrass to
characterize the genetic architecture underlying flowering time regulation in napiergrass.
This is the first GWAS study performed in napiergrass. Recently, QTL mapping was
conducted in a bi-parental mapping population of 185 progenies with multiple QTLs
detected for flowering time and number of flowers (Chapter 3). We detected one
significant QTL in this study on chromosome 5 of the pearl millet genome, which
corresponded to linkage group 4 of napiergrass, where two QTLs (qNFL-4-1 and qNFL-
4-2) for number of flowers were identified in the previous study (Chapter 3). Other QTLs
identified in Chapter 3 were not identified by GWAS. This could be explained by the
type of population studied (germplasm collection vs. bi-parental population), statistical
method employed (GWAS vs. linkage mapping) and strict threshold to declare
127
significance (p<0.05 after Bonferroni correction vs. logarithm of odd <3) (Yang, Islam, et
al. 2018). In the 200kb region surrounding the most significant marker, we identified two
genes related to flowering. Calcium-binding protein CML cause alterations in flowering
time and can act as a switch in response to day length perception and the gain of CML
function results in early flowering (Tsai et al. 2007). Similarly, enoyl-CoA hydratase is
important for seed germination (Richmond and Bleecker 1999). Other potential
flowering candidates like phytochrome B and AP2-like factors were located at 150 kb
and 165 kb respectively away from the significant marker. The markers and genes
associated with flowering time could be fully utilized in napiergrass breeding programs
after they have been validated in a much larger population.
Identifying genes that are responsible for trait variation is a major challenge (Gelli
et al. 2016). Due to the unavailability of transcriptome sequences for napiergrass, it will
be difficult to pinpoint the underlying genes responsible for trait variation. Future
evaluations could lead us towards potential targets for improving flowering time in
napiergrass and related crops. Currently there is no report that has utilized TES in
napiergrass and we have made this first attempt to use exome sequencing for calling
variations in napiergrass for flowering related genes as well as performed GWAS.
GWAS in polyploids is challenging because most studies are based on diploid genetics
and software that are modeled using diploid genetics (Bourke et al. 2018). While there
is no software available specifically to perform GWAS in allotetraploids, recently, two
software were released that accept polyploid data, namely GWASpoly (Rosyara et al.
2016) and SHEsisPlus (Shen et al. 2016). In this study, we found 1 significant SNP
using GWASpoly. Modeling GWAS studies based on polyploidy as compared to diploidy
128
can result in identification of more associated SNPs (Ferrão et al. 2018). Therefore, the
selection of a proper model is critical in GWAS. In the future a software that is
particularly modeled for allopolyploidy might become available and may help to identify
more SNPs associated with the trait. Moreover, we were also limited by the size of the
population. Our population of 94 accessions might not have enough power to detect the
variations in flowering time, which is mostly controlled by minor effect QTLs (Dell’Acqua
et al. 2015). In addition, phenotyping may lack accuracy since it was only completed for
one year and at one location. These factors might have played a role in the limited
number of SNPs identified for the flowering time. Phenotyping the germplasm collection
at multiple locations and for multiple years may allow to identify significant associations
to the traits of interest. Further, genotyping more accessions of napiergrass germplasm
could also help in identifying other important variations that were not detected in this
study. In addition, this germplasm collection can also be phenotyped for other traits like
those related to biomass accumulation to associate these traits with sequence
variations and potential candidate genes.
Probe design is critical for improving the performance of TES for deep
sequencing. Probes can be designed from transcribed sequences (expressed sequence
tags or RNA-seq reads) or genomic sequences (whole genome sequencing, reduced
representation library, genotyping by sequencing) of the species of interest or related
species. We used the genome of a closely related species of napiergrass, pearl millet
(Varshney et al. 2017), to design probes for this study. Using 37k probes, we targeted
818 flowering genes, and 28,519 overall genes according to pearl millet genome.
129
Since napiergrass is allotetraploid with two similar sub-genomes, most of the
probes designed for this study might have multiple hits. With the unavailability of
reference genome of napiergrass, we are unable to separate the probes that could have
come from each sub-genome. While TES provides high coverage data, some non-
primary targets could also be captured, which is unavoidable for organisms that do not
have a reference genome (Peng et al. 2017).
Conclusion
Our study showed the feasibility to applying targeted enrichment sequencing with
probes designed from CDS of pearl millet genome to target the genomic regions in
napiergrass. We inferred the structure of the germplasm collection as well as
constructed phylogeny for the germplasm. TES allowed us to perform a GWAS on this
population, and we identified one significant SNP and two candidate genes for flowering
time variation. The success of this approach could be improved by increasing the
number of accessions studied, phenotyping for multiple years at multiple locations, as
well as using the reference genome of napiergrass once it becomes available. With the
availability of the genome reference, we can assess uniqueness of probes that will
increase unique mapping rate as well as minimize the proportion of off-target capture.
130
Table 4-1. Genes and publications related to flowering used in this research.
Organism Genes Reference
Arabidopsis 306 (Bouché, Lobet, Tocquin, & Périlleux, 2016)
Brachypodium 23 (Higgins, Bailey, & Laurie, 2010)
Maize 692 (Li et al., 2016)
Maize 148 (Xu et al., 2012)
Rice 201 Wiki pathways
Rice 39 (Lee & An, 2015)
Setaria 53 (Mauro-Herrera et al., 2013)
Sorghum 141 (Mace, Hunt, & Jordan, 2013)
Switchgrass 654 (Grabowski et al., 2016)
mRNA 1,285 Refseq db: flowering + greenplants
Grand Total 3,542
131
Figure 4-1. Histogram of length of flowering related genes.
0
10
20
30
40
50
60
70
80
100
-199
300
-399
500
-599
700
-799
900
-999
110
0-1
19
9
130
0-1
39
9
150
0-1
59
9
170
0-1
79
9
190
0-1
99
9
210
0-2
19
9
230
0-2
39
9
250
0-2
59
9
270
0-2
79
9
290
0-2
99
9
310
0-3
19
9
330
0-3
39
9
350
0-3
59
9
370
0-3
79
9
390
0-3
99
9
410
0-4
19
9
430
0-4
39
9
450
0-4
59
9
470
0-4
79
9
490
0-4
99
9
510
0-5
19
9
530
0-5
39
9
Num
be
r o
f ge
ne
s
Length of gene (bp), n=1213
132
Figure 4-2. Number of probes designed per gene.
0
10
20
30
40
50
60
0 20 40 60 80 100
Nu
mb
er
of
ge
ne
s
Number of probes designed per gene
133
Figure 4-3. Number of probes designed as a factor of the size of the gene.
134
Figure 4-4. Distribution of the targeted flowering genes in the genome of pearl millet. Each blue bar represents a flowering gene mapped on the pearl millet genome.
135
Figure 4-5. Number of paired-end reads per sample.
0
1,000,000
2,000,000
3,000,000
4,000,000
5,000,000
6,000,000
Num
be
r o
f re
ad
pa
irs
Sample
136
Figure 4-6. Bayesian Information Criteria (BIC) vs. number of clusters in k-means clustering suggests K=3 in the germplasm collection.
5 10 15 20
42
04
30
44
04
50
Value of BIC
versus number of clusters
Number of clusters
BIC
137
Figure 4-7. Projection of the napiergrass germplasm collection using first two linear discriminants (LDs) from discriminant analysis of principal components (DAPC). The shape of the points represent grouping by DAPC (circle = group 1; triangle = group 2; square = group 3) and the colors represent the origin continent: black = Americas; blue = Asia; and cyan = Africa.
-4
-2
0
2
-2.5 0.0 2.5 5.0
LD1
LD
2
138
Figure 4-8. Evolutionary relationships of taxa. G stands for group assigned by DAPC and are in different shapes (circle = group 1; triangle = group 2; square = group 3) and the fill colors represent the origin: black = Americas, blue = Asia, cyan = Africa, and white=unknown).
N43 G
2N
228 G
2
N51 G
2N
199 G
2
N210 G
2N
222 G
2N
203 G
2N
-Mer G
2N
212 G
2N
198
G2
N8
G2
N11
4 G
2N22
3 G
2
N21
4 G
2
N215 G2
N116 G2
N204 G2
N127 G2
N211 G2
N225 G2
N-6X G2
N205 G2
N128 G2
N129 G2
N12 G2
N132 G2
N74 G3N130 G3N9 G3
N14 G3N37 G3N239 G3N122 G3N168 G3N16 G3N20 G
3N172 G
3
NILR
I-16791 G1
N40 G
3
N35 G
3
N36 G
3
N34 G
3
N42 G
3
N39 G
3
N41 G
3
N32 G
3
N23 G
1
N24 G
1
N66 G
1
N56 G
1
N67 G
1
N68 G
1
N69 G
1N243 G
1
N244 G
1
N242 G
1
N240 G
1
N147 G
1
N17
9 G
1
N13
1 G
1
N15
2 G
1
N16
3 G
1
N17
1 G
1N164 G
1N181 G
1N183 G
1N170 G1N186 G1N182 G1N185 G1
N178 G1N137 G1N138 G1
N150 G1
N166 G1
N151 G1
N180 G1
N161 G1
N71 G1
N72 G1
N155 G1
N188 G1
N226 G1
N238 G1
N19 G1
N22 G1
N7 G
1
N70 G
1N
13 G1
N109 G
1N
ILRI-16786 G
1
N157 G
1N
190 G
1
NIL
RI-1
4984 G
1
139
Figure 4-9. Histogram for days to flowering trait in napiergrass germplasm collection.
0
5
10
15
230 240 250 260 270 280
Days to flowering
140
Figure 4-10. QQ plots for different models using GWASpoly using 78,129 SNP markers. DTF = Days to flowering.
141
Figure 4-11. Manhattan plots for different models using GWASpoly. P values adjusted with FDR at 0.05. DTF = Days to flowering.
142
CHAPTER 5 GENERATION OF INTERSPECIFIC HYBRIDS BETWEEN PEARL MILLET AND
NAPIERGRASS AND EVALUATION OF THEIR PERFORMANCE
Introduction
Napiergrass or elephantgrass (Cenchrus purpureus Schumach) is an important
forage crop that is widely used in Africa for dairy cows because of its high yields and
nutrient value (Singh, Singh, and Obeng 2013). Mature napiergrass reach plant heights
of 5-6 m and up to 20 nodes per stem (Boonman 1988) and out-yield other grasses by a
significant margin (Ra et al. 2012). Napiergrass is generally propagated via stem
cuttings. Clonal propagation increases propagation costs as cultivation is typically done
manually. Therefore, even cultivars of napiergrass with high biomass quality like Mott,
which produced average cattle gains of 0.97 kgd-1 compared to industry standard of
0.39 kgd-1 of ‘Pensacola’ bahiagrass (Paspalum notatum Flugge) (Sollenberger et al.
1989), have not been used commercially in the US because of the limitations
associated with the vegetative propagation (Diz and Schank 1993). Production of large
seeds that can be harvested and planted mechanically will propel the commercial
success of napiergrass.
Napiergrass is the fastest growing plant in the world (Karlsson and Vasil 1986)
with reported dry biomass yield of 45 dry t ha−1 year−1 in Florida (Woodard and
Sollenberger 2012) and as high as 80 dry t ha−1 in tropical countries (Vicente-Chandler,
Silva, and Figarella 1959). Therefore, napiergrass has a great potential as
lignocellulosic feedstock for biofuel production. Being non-edible and able to grow on
marginal land, napiergrass escapes the 'food versus fuel' debate and has competitive
advantage over tree species as it can be harvested for biomass in the first year after
143
planting. Additionally, lignin content, which is considered a hindrance to the
fermentation of biomass to ethanol, is much lower in napiergrass than woody biomass,
10% in napiergrass compared to 20-30% in woods (Tong, Smith, and Mccarty 1990;
Mckendry 2002). Most perennial biomass crops like switchgrass (Panicum virgatum L.)
have establishment issues due to small seed size, slow growth rate, dormancy, and
negative response to high planting densities (Noble Research Institute 2007). Leading
bioenergy candidate grasses like switchgrass, Miscanthus, and energycane are not
capable of both direct seeding and high production of biomass in the establishment year
and in contrast to napiergrass do not tolerate multiple harvests per year (Singh, Singh,
and Obeng 2013).
Napiergrass is a short-day plant and flowering in tropical climates occurs from
autumn through winter (Singh, Singh, and Obeng 2013). Early flowering cultivars
produce an abundant amount of small and wind dispersed seeds making seed
collection difficult and increasing its invasive potential (D’Antonio and Vitousek 1992;
Loope, Hamann, and Stone 1988; Schofield 1989). Therefore, napiergrass is listed as
an invasive species by the Florida Exotic Plant Pest Council (FLEPPC 2011).
Invasiveness in napiergrass can be effectively controlled by developing interspecific
triploid hybrids between napiergrass and pearl millet (Cenchrus americanus, 2n=2x=14)
(Hanna 1981). The chromosomes in the A’ genome of napiergrass are homologous to
the A genome of pearl millet (Jauhar 1981) and these two species hybridize naturally to
produce pearl millet napiergrass (PMN) hybrids that are triploids (AA’ B genome) and
thus are sterile (Singh, Singh, and Obeng 2013). PMN hybrids do not set seed and so,
do not pose a threat of uncontrolled establishment through dissemination of seeds
144
(Hanna and Monson 1980) and are not considered invasive (Jessup 2013). PMN
hybrids combine superior forage quality of pearl millet and high yielding ability of
napiergrass (Gupta and Mhere 1997; Osgood, Hanna, and Tew 1997). Some of these
hybrids produced higher biomass (18.9 Mg ha-1) than napiergrass (17.5 Mg ha-1) and
pearl millet (13.2 Mg ha-1) in Louisiana (Cuomo, Blouin, and Beatty 1996). Pearl millet
seeds are larger in comparison to napiergrass and no seed shattering occurs on the
pearl millet panicle (Fig 5-1). The shape and size of PMN seeds are similar to the
female pearl millet parent. These seeds can be planted using seed drills. Development
of seeded varieties that are sterile will represent a significant step in napiergrass
breeding because establishment of fields by seeds will allow automation of planting,
thus a significant cost reduction (Osgood, Hanna, and Tew 1997). Normally, the
resulting hybrid (2n=21) with AA’B genome has greater similarity to the napiergrass type
due to larger genetic contribution (66.7% chromosomes) and dominance of the
napiergrass B genome over the pearl millet A genome for genetic characters such as
earliness, inflorescence and leaf characteristics (Obok, Ova, and Iwo 2012; Gonzalez
and Hanna 1984). Most of the characteristics like resistance to pests, vigorous growth,
and high forage yield potential are on the B genome (Hanna 1987).
Both napiergrass and pearl millet are protogynous in nature, a phenomenon
where stigmas are exerted prior to anther exertion, therefore, they are predominately
cross-pollinated that results in high heterozygosity (Dowling, Burson, and Jessup 2014).
The heterozygous out-crossing nature of napiergrass and pearl millet leads to
significant segregation and lack of uniformity in progenies. Because of the
heterozygosity in napiergrass and pearl millet, PMN hybrids exhibit a high level of
145
heterosis (Dowling, Burson, and Jessup 2014). For successful commercial application, it
is important to get a high level of heterosis in the hybrids while maintaining uniformity in
biomass and persistence of the progenies. Uniformity, the ability of a cultivar to
produce a specific phenotype instead of a varying phenotype, increases the amount of
predictability on the total biomass yield (Makumburage and Stapleton 2011). Using
PMN hybrids for commercial cultivation requires seeds that produce plants with a high
biomass yield and a certain level of uniformity. Selection for stand uniformity was found
to be associated with increased tolerance to environmental stress in maize (Tollenaar
and Wu 1999). PMN hybrid combinations can produce non-germinating seeds
(Aken’ova and Chheda 1973), or produce hybrids with varying yield potential (Hanna
and Monson 1980). Phenotypic variation can be due to genetic variation caused by
alleles segregating within a population, epistasis and mutations, or environment, caused
by fluctuating external condition (Fraser and Schadt 2010).
Commercial production of PMN hybrids can be facilitated by utilizing cytoplasmic
male sterile (cms) lines of pearl millet. ‘Tift 23A’, a cms line of pearl millet paved the way
to produce seed-propagated PMN hybrids that can facilitate seed harvest (J. B. Powell
and Burton 1966). However, the cms lines of pearl millet are dwarf type forages. It is
critical to choose the right parental combination in order to maximize yield of
interspecific PMN hybrids. For this, the cms trait in the dwarf forage type pearl millet
needs to be introgressed into high biomass pearl millet lines and homogenized so that
progenies from these will be uniform. The backcross method has been commonly used
to transfer entire sets of chromosome from foreign cytoplasm in order to create
cytoplasmic male-sterile genotypes (Acquaah 2012). These lines can then be used to
146
cross with inbred napiergrass to produce progenies of male and female sterile PMN
hybrids.
Since PMN hybrids can produce forage until the season’s first frost, they have
the potential to address the fall forage deficit in the southeast USA (Cuomo, Blouin, and
Beatty 1996). PMN hybrids can be easily established using seeds and may produce
high yields in the first year itself. PMN hybrids are more resistant to most pests and
diseases than pearl millet (Hanna and Monson 1980). Expressed sequence tags –
simple short repeat (EST-SSR) markers can be used in order to confirm triploid hybrids
in a breeding program (Dowling et al. 2013). Therefore, developing seeded PMN
hybrids with high biomass yield and a certain level of uniformity should have a major
impact in the forage and biofuel industry. As such, the impact of different levels of
homozygosity/selfing of pearl millet and napiergrass parents on biomass yield and
uniformity will inform us about the best strategy to manage field breeding of PMN
hybrids. Therefore, it is necessary to identify the level of heterozygosity that is present
in the progenies from different crosses by evaluating biomass yield and uniformity of the
hybrids under field conditions. There have been no previous studies done on the impact
of different levels of selfing of pearl millet and napiergrass parents on the agronomic
performance of the PMN hybrids.
Introgression of cms into high biomass pearl millet should facilitate field
production of PMN seeds by open pollination of cms pearl millet with napiergrass. In this
study we report the biomass yield and uniformity of progenies produced from crosses
between four different parental types of female pearl millet (cms forage, high biomass,
147
cms high biomass, and homozygous high biomass) and a napiergrass male parent (25-
17).
Materials and Methods
Description of Male Sterile Line of Pearl Millet
In this research we used a forage type pearl millet, Tift 85D2A4 (Tift 85) which is a
cytoplasmic male sterile line that flowers in about 75 days. The cms line of pearl millet
Tift 85 was received from Dr. Wayne Hanna, University of Georgia. The A4 cytoplasm
was transferred from a wild subspecies of pearl millet (Wayne Hanna, personal
communication). Tift 85D2B1 is used to maintain sterility of Tift 85D2A4 and it is self-
fertile with good seed set. It was developed by selecting a rust and leaf spot resistant
plant from a selfed population of a BC5 plant developed by backcrossing Tift23D2B1 to a
wild grassy introduction [Pennisetum glaucum (L.) R. Br. Subsp. Monodii (Maire)
Brunken] (Hanna, Wells, and Burton 1987).
Production of cms Lines of Pearl Millet
Cytoplasmic male sterility present in Tift 85 was introgressed into three high
biomass yielding pearl millet lines based on vigor: PI 288787 01 SD (787), PI 215603 01
SD (603), and DLSBF. PI 288787 01 SD is a late flowering accession collected in India
with an average of 19 nodes, 0.880 gm seed weight and 370 cm plant height (USDA,
ARS, and NGRP 1963). PI 215603 01 SD is a late flowering accession collected in India
with an average of 370 cm plant height and 0.720 gm seed weight (USDA, ARS, and
NGRP 1954). DLSBF is an African introduction from Burkina Faso (Wayne Hanna,
personal communication). Pollen from the pearl millet elite lines was dusted onto stigma
of dwarf type cms Tift 85. Thus produced F1 progeny was back crossed to a selfed high
148
biomass line of pearl millet. Male sterility on individual plants were tested by visual
observation prior to crossing in each generation. Selfing of parental line and back
crossing with the most advanced line of the pearl millet continued. Finally, we were able
to introgress cms into one of the three elite pearl millet lines (line 787). We continued to
back cross the cms line 787 four times in order to get a cms version of the elite line of
pearl millet.
Production of PMN Hybrids
In order to obtain near inbred lines of napiergrass, two lines of napiergrass were
selfed, namely, Schank and 25-17. These two lines were selfed two times and flowering
was induced by controlling photoperiod in a growth chamber. However, selfed second
generation of napiergrass didn’t flower due to problems with photoperiod control in the
greenhouse. Therefore, crosses of parental napiergrass (25-17) with different types of
female pearl millet lines were performed (Table 5-1). The interspecific triploid (3x=21)
hybrids were produced in the growth chamber by pollinating pearl millet female plants
with napiergrass pollen.
These four crosses represent the cross of male napiergrass parent (25-17) with
four different types of female pearl-millet lines (cms parent – Tift85 [A], selfed 5th
generation -787S5 [B], BC4 generation – MS 787 BC4 [C], and original parent – P787
[D]).
Seedlings from these four crosses were grown and evaluated for triploidy based
on flowering. Nodes from the confirmed triploids were cut and replanted as replicated
clones into 6” pots. Five clones (nodes) for each plant were grown in the greenhouse in
UF/IFAS Plant Science Research and Education Unit (PSREU), Citra, FL.
149
The pots were filled with sandy soil from the field in PSREU to mimic the soil
conditions of the field. The nodes were planted on Jan 30, 2018 and were irrigated twice
daily. Fertilizer at the rate of 20-20-20 was supplied through irrigation using Dosatron.
After the nodes germinated, the ‘N’ fertilizer rate was increased to 46-0-0 on Feb 21,
2018 in order to promote vegetative growth of the plants. The plants were transplanted
to the field on March 28, 2018. The plants were irrigated with 12mm weekly in the
absence of rainfall after planting. On April 20, 2018 the plants were fertilized with 33.62
kg ha-1 N; 11.21 kg ha-1 and 33.62 kg ha-1 K plus micronutrient package followed by
irrigation. The plants were further fertilized on May 10, 2018 with 84.06 kg ha-1 N; 20.17
kg ha-1 P and 84.06 kg ha-1 K plus micronutrient followed by irrigation. On June 18,
2018 the plants were re-fertilized with 67.25 kg ha-1 N and 89.66 kg ha-1 K using the 6-
0-8 liquid fertilizer plus micronutrient package followed by irrigation.
Experimental Design
Completely randomized block design (RCBD) with 5 replications was used for the
field experiment. Individuals of the same cross type were grouped together and
randomized. Row to row distance was 1.22 m (4 ft) and plant to plant distance was
maintained at 0.91 m (3ft).
Traits Evaluated
We measured yield related attributes of the PMN hybrids. Plant height, number of
tillers, stem diameter, leaf length, leaf width, fresh biomass, and dry biomass were
evaluated. For fresh biomass, all the tillers arising from individual plants were cut using
a brush cutter and fresh weight measured with a hanging scale after seven months of
planting. Samples for dry biomass were taken on August 20, 2018 by cutting 1-2 tillers
150
per plant for 3 replications of each line and weighed. They were then dried in a plant
sample dryer at 48.8°C for 1 month after which the dry biomass weight of the samples
was measured. The dry biomass for replicates was averaged and its derived function
was used to extrapolate dry biomass from each line’s fresh biomass.
Data Analysis
Levene's test was done to compare the variances using 'car' package in R (Fox
et al. 2012). Analysis of variance (ANOVA) and Tukey's HSD test were carried out using
'agricolae' package in R (De Mendiburu 2014). Linear model was fit using the trait value
as dependent variable and cross type and replication as independent variable.
Results
Plant Height
Plant height of the four different crosses were significantly different to each other
(HSD, p<0.05). Variance due to block was significant which was accounted for in the
model. The interaction between plant height and block was not significant. Cross D had
significantly highest height followed by cross B, cross C, and cross A (Figure 5-2). The
coefficient of variation of height was highest in cross type A (23.98%) and the lowest CV
was on cross D (11.31%) (Table 5-2). Maximum height among all the groups was in
cross D, with a height of 426.72 cm. Lowest range of plant height was in cross B
(228.60 cm – 396.24 cm) and the highest in cross D (86.36 cm – 426.72 cm) (Figure 5-
3).
Tiller Number
Tiller number varied significantly between the four types of crosses (p<0.05).
Block effect was not significant for tiller number. Tiller number was significantly higher in
cross A followed by cross D (HSD, p<0.05) (Figure 5-4). Tiller number of cross B was at
151
par with cross C. Lowest number of tillers (1) was found in cross C and the highest
number of tillers (43) was found in cross A. The CV of number of tillers was highest in
cross A and lowest in cross D (Table 5-2). Similarly, the range of number of tillers was
much greater in cross A (2 - 43) as compared to other crosses (3 – 22 in cross B, 1 – 27
in cross C, and 5 – 15 in cross D) (Figure 5-5).
Stem Diameter
Stem diameter between the groups were significantly different (p<0.05) and the
block effect was not significant. Significantly highest (17.25 mm) stem diameter was
found in cross B followed by cross C which was at par with cross D (Figure 5-6). Cross
A showed the highest CV (24.81%) for stem diameter while cross D showed the lowest
CV (14.16%) (Table 5-2). The range of stem diameter was wider in cross A (5.92 mm –
21.95 mm) compared to other crosses (10.04 – 25.71 mm for cross B, 9.63 mm – 22.65
mm for cross C, and 10.65 mm – 22.06 mm for cross D) (Figure 5-7).
Leaf Length
Leaf length was significantly different among the four crosses (p<0.05). Block
effect was not significant. Leaf length of cross D was at par with that of cross B, while it
was significantly higher than cross C or cross A (Figure 5-8). The minimum leaf length
was found in cross A (82.25 cm) and the maximum leaf length was found in cross D
(106.84 cm). The CV for leaf length was highest in cross A (28.32%) and the lowest CV
of leaf length was on cross D (10.15%) (Table 5-2). Range of leaf length was lower in
cross D (72.39 cm – 132.08 cm) as compared to other crosses (38.10 cm – 130.81 cm
for cross A, 26.67 cm – 129.54 cm for cross B, and 29.21 cm – 146.05 cm for cross D)
(Figure 5-9).
152
Leaf Width
Leaf width was significantly different among the four crosses (p<0.05). Block
effect was not significant. Leaf width of cross D was at par with that of cross B, while it
was significantly higher than cross C or cross A (Figure 5-10). On the other hand, leaf
length of cross B was at par with cross D and cross C and significantly higher than
cross A. The minimum leaf width was found in cross C (25 mm) and the maximum leaf
width was found in cross B (69 mm). The CV for leaf width was highest in cross A
(18.80%) and the lowest CV of leaf width was on cross D (13.83%) (Table 5-2). Range
of leaf width was lower in cross D (27 mm – 62 mm) as compared to other crosses (23
mm – 59 mm for cross A, 30 mm – 69 mm for cross B, and 25 mm – 64 mm for cross C)
(Figure 5-11).
Plant Biomass
Fresh biomass per plant (kg) was significantly different (p < 0.05) among the four
crosses and the block effect was also significant. The mean response was different in
block 4 and block 5 as compared to the other blocks and the block effect was accounted
for in the model. The interaction between treatment and block was not significant.
Significantly highest biomass was found in cross D (8.54 kg/plant) followed by cross A
(6.88 kg/plant). Biomass for Cross C (4.97 kg/plant) and cross B (5.67 kg/plant) were at
par with each other (Fig 5-12). The lowest biomass was found in cross B and the
highest biomass was found in cross C (20.59 kg). The CV of biomass was lowest in
cross D (33.63%) and highest in cross A (67.40%). Similarly, the spread of biomass
values was wider in cross A (0.18 kg/plant – 19.16 kg/plant) as compared to other
153
crosses (1.56 kg/plant – 14.28 kg/plant in cross B, 0.12 kg/plant – 20.59 kg/plant in
cross C and 3.06 kg/plant – 19.70 kg/plant in cross D) (Figure 5-13).
Dry Biomass
Dry biomass data from single plant was extrapolated to a larger land area in
order to get dry biomass in tons ha-1. Projected dry biomass (tons ha-1) was significantly
different (p < 0.05) among the four crosses. Significantly highest biomass was found in
cross D (24.83 tons ha-1). Dry biomass for the remaining three cross types were at par
with each other (18.39 tons ha-1, 18.34 tons ha-1, and 16.30 tons ha-1 for cross A, cross
B, and cross C, respectively) (Fig 5-14). Highest dry biomass was found in cross A
(54.87 tons ha-1) and the lowest was found in cross C (0.27 tons ha-1). The CV of dry
biomass was lowest in cross D (33.63%) and highest in cross A (67.40%). Similarly, the
range of biomass values among different progeny plants was much wider in cross A
(0.36 tons ha-1 - 54.51 tons ha-1) as compared to other crosses (3.84 tons ha-1 – 47.56
tons ha-1 in cross B, 0.27 tons ha-1 – 53.72 tons ha-1 in cross C, and 8.73 tons ha-1 –
48.59 tons ha-1 in cross D) (Figure 5-15).
Coefficient of Variation
The coefficient of variation (CV) for all measured traits is shown in Figure 5-16.
For all the traits evaluated, it was observed that the CV was least on cross D and
highest on cross A. CV was the highest for dry biomass (tons/ha.) (30.95% - 73.42%)
(Table 5-2) and least for leaf width (13.83-18.80%) among all types of crosses. A
relatively lower CV was observed for stem diameter (14.16-24.81%), plant height
(10.16-23.98%), and leaf length (10.15-28.32%). On the other hand, higher CV was
154
observed for number of tillers (31.93-48.47%). A picture of four different crosses during
harvest is shown in Figure 5-17.
Correlation
All of the evaluated traits had significant, positive correlations with biomass
weight. Correlation coefficient was the highest for number of tillers (0.60) and lowest for
stem diameter (0.18). Correlation coefficients and p-values between biomass weight
and each trait are presented in Table 5-3.
Discussion
Both pearl millet and napiergrass are protogynous and are predominantly cross-
pollinated. This increases their heterozygosity which leads to a high level of heterosis in
PMN hybrids (Dowling, Burson, and Jessup 2014; J. B. Powell and Burton 1966).
Development of a cms pearl millet line optimized for biomass production will be the
most efficient system for commercial scale production of large seeded PMN hybrids
according to the method described by Powell and Burton (1966) (Dowling 2011). For
commercial success of PMN hybrids, phenotypic stability and robustness of the
genotype to consistently produce a specific phenotype is critical. To achieve this, we
introgressed cms into high biomass pearl millet lines and assessed the phenotypic
variation in biomass yield and related traits of PMN hybrids depending on different pearl
millet sources. It should be noted that our attempt to produce a male napiergrass parent
with higher level of homozygosity was not accomplished in time and therefore a
heterozygous napiergrass parent was used in all these crosses. Inherent heterozygosity
in this parent probably contributed to a high level of phenotypic variability in the
progenies.
155
The biomass produced after seven months of growth was the highest in the cross
involving the original pearl millet parent (cross D) followed by the cms parent, Tift 85
(cross A). The CV for biomass was also lowest in cross D (30.95%). However, CV for
biomass of cross A was the highest (73.42%) as compared to cross B (39.66%) and
cross C (56.55%). For the other traits evaluated, lowest CV was mostly found in cross D
followed by cross B. As expected the high biomass pearl millet parent P787 used in
cross D had a higher combining ability than the forage-type pearl millet parent Tift 85
resulting in significant differences in biomass yield. The backcross of Tift 85 with P787
was not able to restore the superior combining ability of P787. Interestingly, progenies
from cross C using 787-S5 as parent showed a significant lower biomass yield than
progenies from crosses involving the heterozygous P787. Testing the biomass
accumulation of selfed progenies from P787 at every generation and selecting superior
individuals may help to generate male sterile 787 with superior combining ability. The
difference among CV of cross D and cross B was low. For plant height, the CV of cross
B was actually the lowest among the different crosses. CV for the various traits involving
cross C (BC4 generation of Tift85 and 25-17) was lower than cross A but higher than
cross B or cross D. We saw that the CV decreased in BC4 but hasn't reached the level
of P787 or 787-S5. In theory, BC4 genotype will be 93.75% identical to the recurrent
parent (Acquaah 2012). Therefore, in the line MS 787 BC4, we are in theory able to
introgress approximately 94% of the genome of P787 parent. Continuing the back cross
to more than BC7 generation will give 99% similarity to the P787 parent and may result
in cms progenies that are more uniform.
156
Even though the average yield of PMN hybrids using cms parent Tift85 in cross A
was high, the high variability in yield would make these lines less acceptable for
commercial production because phenotypic uniformity contributes to a predictable yield.
Many factors such as variable genetic yield potential, uneven germination, variable
planting depth, soil clods, insect damage, and moisture might be responsible for non-
uniformity of plant stands etc. (Martin et al. 2005) which can be controlled by proper
agronomic practices, experimental setup, and statistical analysis. Also, the cultivar
resulting from back-crossing could differ from the initial cultivar beyond the transferred
gene(s) because of linkage drag from the association of undesirable traits with the
genes from the donor(Acquaah 2012).
Biomass yield of grasses is affected by different morphological traits. In sorghum,
biomass is correlated with plant height, number of tillers, leaf length, leaf width, stem
diameter, and flowering time (Hart et al. 2001; Murray et al. 2008; Xiao-ping et al. 2011).
Similarly, in Miscanthus, plant height, stem diameter, late flowering, and growth rate
showed the highest positive correlation with yield (Zub, Arnoult, and Brancourt-Hulmel
2011). In Trichloris crinite, foliage height and basal diameter were strongly correlated
with biomass yield (Cavagnaro et al. 2006). In napiergrass, plant height, number of
tillers, and stem diameter were significantly correlated with plant biomass (Sinche
2013). Our results show that in PMN hybrids, traits like plant height, leaf length, leaf
width, stem diameter, and number of tillers are significantly correlated with biomass
yield. This indicates that tall PMN hybrids with numerous thick tillers and wide and long
leaves produced more biomass than those with the opposite characteristics.
157
In this study, we have developed a cms version of elite pearl millet line P787.
The development of this non-dwarf biomass-type cms line of pearl millet provides new
resources for pearl millet and napiergrass breeding that can be exploited for commercial
production of uniform PMN hybrids.
158
Figure 5-1. Seeds of pearl millet dwarf cms line (A), cms high biomass pearl millet line (B), PMN hybrid (C), and napiergrass (D), respectively from left to right.
159
Figure 5-2. Boxplot of plant height (cm) for four different types of crosses studied. Small colored dots represent individual plant height of the cross. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
d b c a100
200
300
400
A B C D
Cross Type
Pla
nt h
eig
ht (c
m)
160
Figure 5-3. Histogram of plant height (cm) for four different crosses. X-axis represent plant height in cm and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
C D
A B
100 200 300 400 100 200 300 400
0
10
20
30
0
10
20
30
Plant height (cm)
161
Figure 5-4. Boxplot of number of tillers for four different types of crosses studied. Small colored dots represent number of tillers for individual plants of the cross. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
a c c b0
10
20
30
40
A B C D
Cross Type
Nu
mb
er
of tille
rs
162
Figure 5-5. Histogram of number of tillers for four different crosses. X-axis represent the number of tillers and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
C D
A B
0 10 20 30 40 0 10 20 30 40
0
20
40
60
0
20
40
60
Number of tillers
163
Figure 5-6. Boxplot of stem diameter (mm) for four different types of crosses studied. Small colored dots represent the stem diameter (mm) for individual plants of the cross. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
c a b b5
10
15
20
25
A B C D
Cross Type
Ste
m d
iam
ete
r (m
m)
164
Figure 5-7. Histogram of stem diameter (mm) for four different crosses. X-axis represent the stem diameter (mm) and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
C D
A B
5 10 15 20 25 5 10 15 20 25
0
10
20
30
40
0
10
20
30
40
Stem diameter (mm)
165
Figure 5-8. Boxplot of leaf length (cm) for four different types of crosses studied. Small colored dots represent the leaf length (cm) for individual plants of the cross. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
c a b a
50
100
150
A B C D
Cross Type
Le
af le
ng
th (
cm
)
166
Figure 5-9. Histogram of leaf length (cm) for four different crosses. X-axis represent the leaf length (cm) and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
C D
A B
50 100 150 50 100 150
0
10
20
30
40
0
10
20
30
40
Leaf length (cm)
167
Figure 5-10. Boxplot of leaf width (mm) for four different types of crosses studied. Small colored dots represent the leaf width (mm) for individual plants of the cross. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
c ab b a
30
40
50
60
70
A B C D
Cross Type
Le
af w
idth
(m
m)
168
Figure 5-11. Histogram of leaf width (mm) for four different crosses. X-axis represent the leaf width (mm) and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
C D
A B
30 40 50 60 70 30 40 50 60 70
0
10
20
30
0
10
20
30
Leaf width (mm)
169
Figure 5-12. Boxplot of fresh biomass per plant (kg) for four different types of crosses studied. Small colored dots represent the fresh biomass per plant (kg) for individual plants of the cross. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
b c c a0
5
10
15
20
A B C D
Cross Type
Fre
sh
Bio
ma
ss (
kg
/pla
nt)
170
Figure 5-13. Histogram of fresh biomass per plant (kg) for four different crosses. X-axis represent the fresh biomass per plant (kg) and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
C D
A B
0 5 10 15 20 0 5 10 15 20
0
20
40
0
20
40
Fresh Biomass (kg/plant)
171
Figure 5-14. Boxplot of projected dry biomass (tons per ha) for four different types of crosses evaluated at seven months of growth. Different letters in blue color indicate significant differences among crosses based on Tukey’s HSD test at p ≤ 0.05. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
b b b a0
20
40
A B C D
Cross Type
Dry
Bio
ma
ss (
ton
s/h
a.)
172
Figure 5-15. Histogram of dry biomass (tons per ha.) for four different crosses evaluated at seven months of growth. X-axis represent the dry biomass (tons per ha.) and y-axis represents the count of plants. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
C D
A B
0 20 40 0 20 40
0
5
10
15
20
0
5
10
15
20
Dry Biomass (tons/ha)
173
Figure 5-16. Coefficient of variation (%) for the different traits evaluated. A = Tift 85 × 25-17, B = 787-S5 × 25-17, C = MS 787 BC4 × 25-17, D = P787 × 25-17.
LeafWidth StemDiam TillerNum
BioMass height LeafLength
A B C D A B C D A B C D
0
20
40
60
0
20
40
60
Cross type
Co
effic
ien
t o
f va
ria
tio
n (
%)
174
Figure 5-17. Picture of four different crosses during harvest in Citra, FL. Cross A = Tift 85 × 25-17, Cross B = 787-S5 × 25-17, Cross C = MS 787 BC4 × 25-17, Cross D = P787 × 25-17.
175
Table 5-1. Details of the cross types used in the experiment.
Cross type Female parent ♀ (Pearl millet)
Male parent ♂ (Napiergrass)
Cross details Number of plants
A Tift85 × 25-17 cms parent 40 B 787-S5 × 25-17 Selfed 5th gen. 50 C MS 787 BC4 × 25-17 BC4 generation 27 D P787 × 25-17 Original parent 22
176
Table 5-2. Descriptive statistics for different types of crosses. P-value represents Pr>F for Levene’s test for homogeneity of variance (center = median), CV = coefficient of variation.
Trait Cross type P-value Mean Standard Deviation CV (%) Min Max
Plant height 0.0000
Plant height A 0.0778 223.84 53.67 23.98 86.36 340.36
Plant height B 0.9584 313.41 31.84 10.16 228.60 396.24
Plant height C 0.8109 285.09 50.58 17.74 111.76 375.92
Plant height D 0.6924 328.32 37.13 11.31 86.36 426.72
Number of tillers 0.0000
Number of tillers A 0.0758 17.88 8.67 48.47 2.00 43.00
Number of tillers B 0.6613 9.46 3.69 39.03 3.00 22.00
Number of tillers C 0.1015 8.98 4.29 47.73 1.00 27.00
Number of tillers D 0.6265 12.82 4.09 31.93 5.00 15.00
Stem diameter 0.0000
Stem diameter A 0.3174 13.31 3.30 24.81 5.92 21.95
Stem diameter B 0.9829 17.26 2.73 15.80 10.04 25.71
Stem diameter C 0.4913 16.42 2.61 15.90 9.63 22.65
Stem diameter D 0.8523 16.41 2.32 14.16 10.65 22.06
Leaf length 0.0000
Leaf length A 0.2382 82.26 23.29 28.32 38.10 130.81
Leaf length B 0.5612 106.01 11.75 11.08 26.67 129.54
Leaf length C 0.6267 101.17 15.05 14.88 29.21 146.05
Leaf length D 0.1045 106.84 10.84 10.15 72.39 132.08
Leaf width 0.0529
Leaf width A 0.4596 40.75 7.66 18.80 23.00 59.00
Leaf width B 0.8497 47.61 6.88 14.46 30.00 69.00
Leaf width C 0.7394 46.04 7.92 17.20 25.00 64.00
177
Table 5-2. Continued Trait Cross type P-
value
Mean Standard
Deviation
CV (%) Min Max
Leaf width D 0.1875 48.89 6.76 13.83 27.00 62.00
Fresh biomass 0.0000
Fresh biomass A 0.0214* 6.88 4.64 67.40 0.18 19.16
Fresh biomass B 0.6081 6.23 2.29 36.75 1.56 14.28
Fresh biomass C 0.0794 5.47 3.10 56.63 0.12 20.59
Fresh biomass D 0.7515 8.54 2.87 33.63 3.06 19.70
Dry biomass 0.0000
Dry biomass A 0.0207* 18.39 13.50 73.42 0.36 54.87
Dry biomass B 0.6462 18.34 7.27 39.66 3.84 47.56
Dry biomass C 0.1246 16.30 9.22 56.55 0.27 53.72
Dry biomass D 0.9255 24.83 7.68 30.95 8.73 48.59
178
Table 5-3. Correlation coefficients and p-values for biomass weight and biomass-related traits for PMN hybrids evaluated in Citra, FL.
Plant height Leaf length Leaf width Stem diameter Number of tillers
R 0.43 0.36 0.25 0.18 0.6
p-value <0.001 <0.001 <0.001 <0.001 <0.001
179
CHAPTER 6 CONCLUDING REMARKS
Summary
Napiergrass is an important forage and biofuel crop. However, its commercial
utilization has been lagging behind other crops due to limited genetic and genomic
resources. Before this study, there were no SSR markers derived from napiergrass
sequences publicly available, no genetic linkage map was available, and no QTL
studies for any traits were published. The lack of these genetic resources limits the
advances that can be achieved via breeding. In this project, we have developed genetic
and genomic resources that will be important in napiergrass breeding. In Chapter 2, we
constructed the first high-density genetic linkage map of napiergrass using NGS-derived
SNP markers. We also identified 5,339 SSRs using napiergrass genomic sequences
and successfully designed primers for 1,926 SSRs. These results will be useful for
future molecular breeding programs such as identification of QTLs for important traits
as well as MAS for the genetic improvement of napiergrass and comparative
genomics.
Early flowering cultivars of napiergrass produce abundant wind dispersed seeds,
which contribute to high potential of invasiveness. Controlling flowering or modifying
flowering time of napiergrass can help in reducing its invasiveness and boost its
potential as biofuel feedstock. Therefore, a better understanding of the genetic basis of
flowering time in napiergrass is necessary. To facilitate this, in Chapter 3, we
demonstrated that flowering time and number of flowers are highly heritable traits.
Therefore, we conducted the first QTL analysis in napiergrass and identified three
stable QTLs controlling number of flowers. We also identified three potential QTLs that
180
control flowering time in napiergrass. We were able to identify potential candidate genes
such as AGAMOUS, DEFICIENS, DELLA, WRKY, and SERK1 that were harbored by
the QTL regions. The QTLs detected in this study will be valuable tools for napiergrass
breeding and marker assisted selection. Similarly, the candidate genes identified could
be potential targets for genome editing to modify flowering time in napiergrass.
To improve our understanding of the genetic basis of flowering time in
napiergrass and to evaluate the genetic diversity of napiergrass, we used exome-
capture sequencing to study the germplasm collection of napiergrass, described in
Chapter 4. We identified 78,129 high quality SNPs in the germplasm collection that
serve as novel genomic resources for napiergrass breeding programs. We also
identified potential candidate genes like Calcium-binding protein CML, enoyl-CoA
hydratase, phytochrome B, and AP2-like factors in the germplasm collection. This study
showed the feasibility to apply targeted exome sequencing with probes designed from
CDS of pearl millet to target the genomic regions in napiergrass.
We also used traditional breeding approaches to increase biosafety of
napiergrass. For this, in Chapter 5, we introgressed cms available in forage type dwarf
pearl millet lines to high biomass yielding pearl millet lines. We then hybridized these
elite cms lines of pearl millet with napiergrass to generate PMN hybrids that are male
and female sterile and will not contribute to invasiveness with wind dispersed seeds. We
studied the uniformity of different types of parental combinations. Substantial variation
within each cross was found for biomass accumulation and yield related traits.
Uniformity of the progenies could be enhanced by using homozygous parents to make
the crosses. If efficient seed production can be accomplished under field conditions and
181
if uniformity can be improved without compromising biomass yield, PMN hybrids may
outperform alternative forage and biofuel crops in the near future.
Future work
The construction of the first genetic map of napiergrass has opened up new
avenues in napiergrass breeding. With this map, we were able to identify QTLs related
to flowering time and number of flowers in napiergrass. There are other traits like stem
diameter, tiller number, and yield, whose genetic basis needs to be further explored. It is
our hope that the genetic map developed in Chapter 2 can be utilized to explore other
traits of agronomic importance to identify QTLs and candidate genes for these traits.
Further, the sequence variations identified in the germplasm collection can be utilized in
the future to study the genetic architecture of other traits and to perform GWAS for
these traits. Moreover, introgression of cms into high biomass lines like P787 will
continue to obtain a homozygous cms line with superior combining ability with
napiergrass. Similarly, generation of homozygous or near-homozygous napiergrass
lines will have to be continued by repeated self-fertilization and selection of the best
performing lines. PMN hybrids from these homozygous or near homozygous parents
should be evaluated not only for increased uniformity but also for biomass yield which
may mainly depend on the different level of heterosis in the A and B genomes of the
hybrids and the combining ability of the parents. For commercial production of PMN
hybrid seeds under field condition different accessions need to be evaluated to
synchronize flowering time and identify parents that contribute to the highest seed yield.
Alternatively, highest yielding napiergrass lines can be evaluated in countries where
182
napiergrass is a native species, the preferred forage crop and invasiveness is not a
concern.
183
LIST OF REFERENCES
Acquaah, George. 2012. “Breeding Self Pollinated Species.” In Principles of Plant Genetics and Breeding, Second Edition, 303–36. Wiley Online Library.
Adhikari, Laxman, Orville M Lindstrom, Jonathan Markham, and Ali M Missaoui. 2018. “Dissecting Key Adaptation Traits in the Polyploid Perennial Medicago Sativa Using GBS-SNP Mapping.” Frontiers in Microbiology 9: 1–19. doi:10.3389/fpls.2018.00934.
Agarwal, Milee, Neeta Shrivastava, and Harish Padh. 2008. “Advances in Molecular Marker Techniques and Their Applications in Plant Sciences.” Plant Cell Reports 27 (4): 617–31. doi:10.1007/s00299-008-0507-z.
Aken’ova, M. E., and H. R. Chheda. 1973. “Interspecific Hybrids Pennisetum Typhoides S & H × P. Purpureum Schum. for Forage in the Hot Humid Lowland Tropics of West Africa.” The Nigerian Agricultural Journal 10 (10): 82–90.
Andersen, Jeppe R, and Thomas Lübberstedt. 2003. “Functional Markers in Plants.” Trends in Plant Science 8 (11): 554–60. doi:10.1016/j.tplants.2003.09.010.
Anderson, William F., Michael Casler, and Brian Baldwin. 2008. “Improvement of Perennial Forage Species as Feedstock for Bioenergy.” In Genetic Improvement of Bioenergy Crops, edited by Wilfred Vermerris, 309–45. Springer Science+Business Media, LLC.
Atienza, S. G., Z. Satovic, K. K. Petersen, O. Dolstra, and A. Martín. 2003. “Identification of QTLs Influencing Combustion Quality in Miscanthus Sinensis Anderss. II. Chlorine and Potassium Content.” Theoretical and Applied Genetics, no. 107: 857–63. doi:10.1007/s00122-003-1218-z.
Augustin, Eiiane, and Fernando Adami Tcacenco. 1993. “Isoenzymatic Characterization of Elephantgrass (Pennisetum Purpureum Schum.) Germplasm.” Rev. Brasil. Genet. 16 (3): 685–96.
Azevedo, Ana Luisa Sousa, Pricila Palla Costa, Juarez Campolina Machado, Marco Antonio Machado, Antônio Vander Pereira, and Francisco José da Silva Lédo. 2012. “Cross Species Amplification of Pennisetum Glaucum Microsatellite Markers in Pennisetum Purpureum and Genetic Diversity of Napier Grass Accessions.” Crop Science 52: 1776–85. doi:10.2135/cropsci2011.09.0480.
Bastow, Ruth, and Caroline Dean. 2002. “The Molecular Basis of Photoperiodism.” Biological Rhythm Research 37 (4): 353–80. doi:10.1016/S1534-5807(02)00296-4.
184
Bezant, Jeremy, David A. Laurie, Nicky Pratchett, Jan A N Chojecki, and Mike J. Kearsey. 1996. “Marker Regression Mapping of QTL Controlling Flowering Time and Plant Height in a Spring Barley (Hordeum Vulgare L .) Cross.” Heredity 77: 64–73.
Bhandari, Anitha P, D H Sukanya, and C R Ramesh. 2006. “Application of Isozyme Data in Fingerprinting Napier Grass (Pennisetum Purpureum Schum.) for Germplasm Management.” Genetic Resources and Crop Evolution 53. Springer: 253–64.
Bidinger, F. R., T. Nepolean, C. T. Hash, R. S. Yadav, and C. J. Howarth. 2007. “Quantitative Trait Loci for Grain Yield in Pearl Millet under Variable Postflowering Moisture Conditions.” Crop Science 47 (3): 969–80. doi:10.2135/cropsci2006.07.0465.
Bielenberg, Douglas Gary, Bradley Rauh, Shenghua Fan, Ksenija Gasic, Albert Glenn Abbott, Gregory Lynn Reighard, William R. Okie, and Christina Elizabeth Wells. 2015. “Genotyping by Sequencing for SNP-Based Linkage Map Construction and QTL Analysis of Chilling Requirement and Bloom Date in Peach [Prunus Persica (L.) Batsch].” PLoS ONE, 1–14. doi:10.1371/journal.pone.0139406.
Bodénès, Catherine, Emilie Chancerel, François Ehrenmann, Antoine Kremer, and Christophe Plomion. 2016. “High-Density Linkage Mapping and Distribution of Segregation Distortion Regions in the Oak Genome.” DNA Research 23 (2): 115–24. doi:10.1093/dnares/dsw001.
Bohuon, E J R, L D Ramsay, J A Craft, A E Arthur, D F Marshall, D J Lydiate, and Mike J. Kearsey. 1998. “The Association of Flowering Time Quantitative Trait Loci with Duplicated Regions and Candidate Loci in Brassica Oleracea.” Genetics 150: 393–401.
Bolger, Anthony M., Marc Lohse, and Bjoern Usadel. 2014. “Trimmomatic: A Flexible Trimmer for Illumina Sequence Data.” Bioinformatics 30 (15): 2114–20. doi:10.1093/bioinformatics/btu170.
Boonman, Joseph G. 1988. East Africa’s Grasses and Fodders: Their Ecology and Husbandry. Edited by Helmut Lieth and Harold A. Mooney. Springer-Science+Business Media, B.V.
Bouché, Frédéric, Guillaume Lobet, Pierre Tocquin, and Claire Périlleux. 2016. “FLOR-ID: An Interactive Database of Flowering-Time Gene Networks in Arabidopsis Thaliana.” Nucleic Acids Research 44: D1167–71. doi:10.1093/nar/gkv1054.
Bourke, Peter M., Roeland E. Voorrips, Richard G. F. Visser, and Chris Maliepaard. 2018. “Tools for Genetic Studies in Experimental Populations of Polyploids.” Frontiers in Plant Science 9: 1–17. doi:10.3389/fpls.2018.00513.
185
Brunken, Jere N. 1977. “A Systematic Study of Pennisetum Sect . Pennisetum (Gramineae).” American Journal of Botany 64 (2): 161–76.
Buckler, Edward S, James B Holland, Peter J Bradbury, Charlotte B Acharya, Patrick J Brown, Chris Browne, Elhan Ersoz, Sherry Flint-Garcia, Arturo Garcia, and Jeffrey C. Glaubitz. 2009. “The Genetic Architecture of Maize Flowering Time.” Science 325. American Association for the Advancement of Science: 714–18.
Burton, Glenn W. 1944. “Hybrids between Napier Grass and Cattail Millet.” The Journal of Heredity, 227–32.
———. 1958. “Cytoplasmic Male-Sterility in Pearl Millet (<i>Pennisetum glaucum)(L.) R. Br.” Agronomy Journal 50 (4). American Society of Agronomy: 230.
———. 1990. Grasses : New and Improved. Edited by J. Janick and J.E. Simon. Advances in New Crops. Portland, Oregon: Timber Press.
Burton, Glenn W., and D S Athwal. 1967. “Two Additional Sources of Cytoplasmic Male-Sterility in Pearl Millet and Their Relationship to Tift 23A.” Crop Science 7 (3). Crop Science Society of America: 209–11.
Cai, C, F Y Cheng, J Wu, Y Zhong, and G Liu. 2015. “The First High-Density Genetic Map Construction in Tree Peony (Paeonia Sect. Moutan) Using Genotyping by Specific-Locus Amplified Fragment Sequencing.” PLoS One, 1–16. doi:10.1371/journal.pone.0128584.
Catchen, Julian, A. Amores, P. Hohenlohe, W. Cresko, J. H. Postlethwait, and D.-J. De Koning. 2011. “Stacks: Building and Genotyping Loci de Novo from Short-Read Sequences.” G3.Genes|Genomes|Genetics 1: 171–82. doi:10.1534/g3.111.000240.
Catchen, Julian, Paul A. Hohenlohe, Susan Bassham, Angel Amores, and William A. Cresko. 2013. “Stacks: An Analysis Tool Set for Population Genomics.” Molecular Ecology 22: 3124–40. doi:10.1111/mec.12354.
Cavagnaro, Pablo F, Juan B Cavagnaro, José L Lemes, Ricardo W Masuelli, and Carlos B Passera. 2006. “Genetic Diversity among Varieties of the Native Forage Grass Trichloris Crinita Based on AFLP Markers, Morphological Characters, and Quantitative Agronomic Traits.” Genome 49: 906–18. doi:10.1139/g06-060.
Chemisquy, M. Amelia, Liliana M. Giussani, María A. Scataglini, Elizabeth A. Kellogg, and Osvaldo Morrone. 2010. “Phylogenetic Studies Favour the Unification of Pennisetum, Cenchrus and Odontelytrum (Poaceae): A Combined Nuclear, Plastid and Morphological Analysis, and Nomenclatural Combinations in Cenchrus.” Annals of Botany 106: 107–30. doi:10.1093/aob/mcq090.
186
Churchill, Gary A., and R. W. Doerge. 1994. “Empirical Threshold Values for Quantitative Trait Mapping.” Genetics 138: 963–71. doi:10.1534/genetics.107.080101.
Cingolani, Pablo, Adrian Platts, Le Lily Luan Wang, Melissa Coon, Tung Nguyen, Le Lily Luan Wang, Susan J. Land, Xiangyi Lu, and Douglas M. Ruden. 2012. “A Program for Annotating and Predicting the Effects of Single Nucleotide Polymorphisms, SnpEff.” Fly 6 (2): 80–92.
Clevenger, Josh, Carolina Chavarro, Stephanie A. Pearl, Peggy Ozias-Akins, and Scott A. Jackson. 2015. “Single Nucleotide Polymorphism Identification in Polyploids: A Review, Example, and Recommendations.” Molecular Plant 8 (6). Elsevier Ltd: 831–46. doi:10.1016/j.molp.2015.02.002.
Crawford, Joseph, Patrick J. Brown, Thomas Voigt, and D. K. Lee. 2016. “Linkage Mapping in Prairie Cordgrass (Spartina Pectinata Link) Using Genotyping-by-Sequencing.” Molecular Breeding 36 (5). Springer Netherlands: 1–12. doi:10.1007/s11032-016-0484-9.
Cuomo, Greg J, David C Blouin, and James E Beatty. 1996. “Forage Potential of Dwarf Napiergrass and a Pearl Millet X Napiergrass Hybrid.” Agronomy Journal 88: 434–38.
D’Antonio, Carla M, and Peter M Vitousek. 1992. “Biological Invasions by Exotic Grasses, the Grass/fire Cycle, and Global Change.” Annual Review of Ecology and Systematics 23. JSTOR: 63–87.
de Lima, R S N, R F Daher, L S a Gonçalves, D a Rossi, a T do Amaral Júnior, M G Pereira, and F J S Lédo. 2011. “RAPD and ISSR Markers in the Evaluation of Genetic Divergence among Accessions of Elephant Grass.” Genetics and Molecular Research : GMR 10 (3): 1304–13. doi:10.4238/vol10-3gmr1107.
De Mendiburu, Felipe. 2014. “Agricolae: Statistical Procedures for Agricultural Research.” R Package Version.
Dell’Acqua, Matteo, Daniel M. Gatti, Giorgio Pea, Federica Cattonaro, Frederik Coppens, Gabriele Magris, Aye L. Hlaing, et al. 2015. “Genetic Properties of the MAGIC Maize Population: A New Platform for High Definition QTL Mapping in Zea Mays.” Genome Biology 16 (167). Genome Biology: 1–23. doi:10.1186/s13059-015-0716-z.
Dellaporta, Stephen L, Jonathan Wood, and James B Hicks. 1983. “A Plant DNA Minipreparation: Version II.” Plant Molecular Biology Reporter 1 (4). Springer: 19–21.
187
Deschamps, Stéphane, Victor Llaca, and Gregory D. May. 2012. “Genotyping-by-Sequencing in Plants.” Biology 1. Molecular Diversity Preservation International: 460–83. doi:10.3390/biology1030460.
Devos, Katrien M, and Mike D Gale. 2000. “Genome Relationships: The Grass Model in Current Research.” The Plant Cell 12. Am Soc Plant Biol: 637–46.
Devos, Katrien M, T S Pittaway, A Reynolds, and M D Gale. 2000. “Comparative Mapping Reveals a Complex Relationship between the Pearl Millet Genome and Those of Foxtail Millet and Rice.” Theoretical and Applied Genetics 100. Springer: 190–98. doi:10.1007/s001220050026.
Diz, D. A., and S. C. Schank. 1993. “Characterization of Seed Producing Pearl Millet X Elephantgrass Hexaploid Hybrids.” Euphytica 67: 143–49. doi:10.1007/BF00022737.
Dowling, Charlie D. 2011. “Marker-Assisted Verification of Hybrids in Pearl Millet-Napiergrass (Pennisetum Glaucum [L.] R. BR. X Pennisetum Purpureum Schumach.).” Texas A&M University.
Dowling, Charlie D., Byron L Burson, Jamie L Foster, Lee Tarpley, and Russell W Jessup. 2013. “Confirmation of Pearl Millet-Napiergrass Hybrids Using EST-Derived Simple Sequence Repeat (SSR) Markers.” American Journal of Plant Sciences 4. Scientific Research Publishing: 1004–12.
Dowling, Charlie D., Byron L Burson, and Russell W Jessup. 2014. “Marker-Assisted Verification of Kinggrass (Pennisetum Purpureum Schumach . X Pennisetum Glaucum [ L .] R . Br .).” Plant Omics Journal 7 (2): 72–79.
Durand, Eléonore, Sophie Bouchet, Pascal Bertin, Adrienne Ressayre, Philippe Jamin, Alain Charcosset, Christine Dillmann, and Maud I Tenaillon. 2012. “Flowering Time in Maize: Linkage and Epistasis at a Major Effect Locus.” Genetics 190. Genetics Soc America: 1547–62.
Elshire, Robert J., Jeffrey C. Glaubitz, Qi Sun, Jesse A. Poland, Ken Kawamoto, Edward S. Buckler, and Sharon E. Mitchell. 2011. “A Robust, Simple Genotyping-by-Sequencing (GBS) Approach for High Diversity Species.” PLoS ONE 6 (5). doi:10.1371/journal.pone.0019379.
Erickson, John E., Arkorn Soikaew, Lynn E. Sollenberger, and Jerry M. Bennett. 2012. “Water Use and Water-Use Efficiency of Three Perennial Bioenergy Grass Crops in Florida.” Agriculture 2: 325–38. doi:10.3390/agriculture2040325.
Falconer, Douglas S, Trudy F C Mackay, and Richard Frankham. 1996. “Introduction to Quantitative Genetics.” In Trends in Genetics, 4th ed. [Amsterdam, The Netherlands: Elsevier Science Publishers (Biomedical Division)], c1985-.
188
Faris, J. D., B. Laddomada, and B. S. Gill. 1998. “Molecular Mapping of Segregation Distortion Loci in Aegilops Tauschii.” Genetics 149 (1): 319–27.
Farrell, G., S. A. Simons, and R. J. Hillocks. 2002. “Pests, Diseases and Weeds of Napier Grass, Pennisetum Purpureum: A Review.” International Journal of Pest Management 48 (1). Taylor & Francis: 39–48. doi:10.1080/09670870110065578.
Ferrão, Luís Felipe V., Juliana Benevenuto, Ivone de Bem Oliveira, Catherine Cellon, James Olmstead, Matias Kirst, Marcio F. R. Resende, and Patricio Munoz. 2018. “Insights into the Genetic Basis of Blueberry Fruit-Related Traits Using Diploid and Polyploid Models in a GWAS Context.” Frontiers in Ecology and Evolution 6. doi:10.3389/fevo.2018.00107.
FLEPPC. 2011. “List of Invasive Plant Species.” Florida Exotic Pest Plant Council. http://www.fleppc.org/list/2011plantlist.pdf.
Fox, John, Sanford Weisberg, Daniel Adler, Douglas Bates, Gabriel Baud-Bovy, Steve Ellison, David Firth, Michael Friendly, Gregor Gorjanc, and Spencer Graves. 2012. “Package ‘car.’” Vienna: R Foundation for Statistical Computing.
Fraser, Hunter B., and Eric E. Schadt. 2010. “The Quantitative Genetics of Phenotypic Robustness.” PLoS ONE 5 (1). doi:10.1371/journal.pone.0008635.
Ganal, Martin W, Thomas Altmann, and Marion S Röder. 2009. “SNP Identification in Crop Plants.” Current Opinion in Plant Biology 12: 211–17. doi:10.1016/j.pbi.2008.12.009.
Garner, Wightman Wells. 1933. “Comparative Responses of Long-Day and Short-Day Plants to Relative Length of Day and Night.” Plant Physiology 8 (3). American Society of Plant Biologists: 347–56.
Garrison, Erik, and Gabor Marth. 2012. “Haplotype-Based Variant Detection from Short-Read Sequencing.” arXiv:1207.3907.
Gelli, Malleswari, Sharon E. Mitchell, Kan Liu, Thomas E. Clemente, Donald P. Weeks, Chi Zhang, David R. Holding, and Ismail M. Dweikat. 2016. “Mapping QTLs and Association of Differentially Expressed Gene Transcripts for Multiple Agronomic Traits under Different Nitrogen Levels in Sorghum.” BMC Plant Biology 16 (16). BMC Plant Biology: 1–18. doi:10.1186/s12870-015-0696-x.
Gifford, Justin M., Won Byong Chae, Kankshita Swaminathan, Stephen P. Moose, and John A. Juvik. 2015. “Mapping the Genome of Miscanthus Sinensis for QTL Associated with Biomass Productivity.” Global Change Biology Bioenergy 7: 797–810. doi:10.1111/gcbb.12201.
189
Glaubitz, Jeffrey C., Terry M. Casstevens, Fei Lu, James Harriman, Robert J. Elshire, Qi Sun, and Edward S. Buckler. 2014. “TASSEL-GBS: A High Capacity Genotyping by Sequencing Analysis Pipeline.” PLoS ONE 9 (2). doi:10.1371/journal.pone.0090346.
Gonzalez, Boldomero, and Wayne W. Hanna. 1984. “Morphological and Fertility Responses in Isogenic Triploid and Hexaploid Pearl Millet X Napiergrass Hybrids.” The Journal of Heredity 75: 317–18.
Grabowski, Paul P., Joseph Evans, Chris Daum, Shweta Deshpande, Kerrie W. Barry, Megan Kennedy, Guillaume Ramstein, et al. 2016. “Genome-Wide Associations with Flowering Time in Switchgrass Using Exome-Capture Sequencing Data.” New Phytologist. doi:10.1111/nph.14101.
Gupta, S C, and O Mhere. 1997. “Identification of Superior Pearl Millet by Napier Hybrids and Napiers in Zimbabwe.” African Crop Science Journal 5 (3). African Crop Science Society: 229–37.
Haggard, J. Erron, Emily B. Johnson, and Dina A. St. Clair. 2015. “Multiple QTL for Horticultural Traits and Quantitative Resistance to Phytophthora Infestans Linked on Solanum Habrochaites Chromosome 11.” G3.Genes|Genomes|Genetics 5: 219–33. doi:10.1534/g3.114.014654.
Hanna, Wayne W. 1981. “Method of Reproduction in Napiergrass and in the 3X and 6X Alloploid Hybrids with Pearl Millet.” Crop Science. Crop Science Society of America, 123–26.
———. 1987. “Utilization of Wild Relatives of Pearl Millet.” In Proceedings of the International Pearl Millet Workshop, 33–42. Patancheru, India: ICRISAT.
Hanna, Wayne W., and Warren G Monson. 1980. “Yield, Quality, and Breeding
Behavior of Pearl Millet ✕ Napiergrass Interspecific Hybrids.” Agronomy Journal
72 (2). American Society of Agronomy: 358–60.
Hanna, Wayne W., H D Wells, and Glenn W. Burton. 1987. “Registration of Pearl Millet Inbred Parental Lines, Tift 85D2A1 and Tift 85D2B1.” Crop Science 27: 1324–25.
Harlan, J R, and J M J de Wet. 1971. “Toward a Rational Classification of Cultivated Plants.” Taxon 20 (4). International Association for Plant Taxonomy (IAPT): 509–17. doi:10.2307/1218252.
Harris-Shultz, Karen R., William F. Anderson, and Ravindra Malik. 2010. “Genetic Relationships among Napiergrass (Pennisetum Purpureum Schum.) Nursery Accessions Using AFLP Markers.” Plant Genetic Resources: Characterization and Utilization 8 (1). Cambridge Univ Press: 63–70.
190
Hart, G. E., K. F. Schertz, Y. Peng, and N. H. Syed. 2001. “Genetic Mapping of Sorghum Bicolor (L.) Moench QTLs That Control Variation in Tillering and Other Morphological Characters.” Theoretical and Applied Genetics 103: 1232–42. doi:10.1007/s001220100582.
Havey, Michael J. 2004. “The Use of Cytoplasmic Male Sterility for Hybrid Seed Production.” In Molecular Biology and Biotechnology of Plant Organelles, edited by H. Daniell and C. D. Chase, 623–34. doi:10.1007/978-1-4020-3166-3_23.
Huang, Xiaoen, Fangfang Wang, Ratnesh Singh, James A Reinert, M C Engelke, Anthony D Genovesi, Ambika Chandra, and Qingyi Yu. 2016. “Construction of High-Resolution Genetic Maps of Zoysia Matrella ( L .) Merrill and Applications to Comparative Genomic Analysis and QTL Mapping of Resistance to Fall Armyworm.” BMC Genomics 17 (562): 1–16. doi:10.1186/s12864-016-2969-7.
Huang, Xiaoqiu, and Anup Madan. 1999. “CAP3: A DNA Sequence Assembly Program.” Genome Research 9: 868–77. doi:10.1101/gr.9.9.868.
Hussain, Waseem, P. Stephen Baenziger, Vikas Belamkar, Mary J. Guttieri, Jorge P. Venegas, Amanda Easterly, Ahmed Sallam, and Jesse A. Poland. 2017. “Genotyping-by-Sequencing Derived High-Density Linkage Map and Its Application to QTL Mapping of Flag Leaf Traits in Bread Wheat.” Scientific Reports 7. Springer US: 1–15. doi:10.1038/s41598-017-16006-z.
ICRISAT, and FAO. 1996. The World Sorghum and Millet Economies: Facts, Trends and Outlook. Patancheru, India: International Crops Research Institute for the Semi-Arid Tropics, and Rome, Italy: Food and Agricultural Organization of the United Nations.
Jauhar, Prem P. 1981. Cytogenetics and Breeding of Pearl Millet and Related Species. Alan R. Liss, Inc. Progress and Topics in Cytogenetics, Volume 1.
Jauhar, Prem P, and Wayne W. Hanna. 1998. Cytogenetics and Genetics of Pearl Millet. Edited by Donald L. Sparks. Advances in Agronomy. Vol. 64. New York: Academic Press, 1949-. doi:10.1017/CBO9781107415324.004.
Jessup, Russell W. 2013. “‘Seeded-yet-Sterile’ Perennial Biofuel Feedstocks.” Advances in Crop Science and Technology 1 (2): 1–2.
Jiang, Guo-liang. 2013. “Molecular Markers and Marker-Assisted Breeding in Plants.” In Plant Breeding from Laboratories to Fields, 45–83. InTech.
Jombart, Thibaut. 2008. “Adegenet: A R Package for the Multivariate Analysis of Genetic Markers.” Bioinformatics 24 (11): 1403–5. doi:10.1093/bioinformatics/btn129.
191
Juma, Awalla Borniface. 2014. “Genetic Diversity and Population Structure of Napier Grass (Pennisetum Purpureum Schumach) in Western Kenya.” University of Eldoret, Kenya.
Jung, Christian, Gina Capistrano-Gossmann, Janina Braatz, Niharika Sashidhar, and Siegbert Melzer. 2018. “Recent Developments in Genome Editing and Applications in Plant Breeding.” Plant Breeding 137 (1): 1–9. doi:10.1111/pbr.12526.
Jung, Christian, and Andreas E Müller. 2009. “Flowering Time Control and Applications in Plant Breeding.” Trends in Plant Science 14 (10): 563–73. doi:10.1016/j.tplants.2009.07.005.
Kandel, Raju, Hari P. Singh, Bharat P. Singh, Karen R. Harris-Shultz, and William F. Anderson. 2016. “Assessment of Genetic Diversity in Napier Grass (Pennisetum Purpureum Schum.) Using Microsatellite, Single-Nucleotide Polymorphism and Insertion-Deletion Markers from Pearl Millet (Pennisetum Glaucum [L.] R. Br.).” Plant Molecular Biology Reporter 34: 265–72. doi:10.1007/s11105-015-0918-2.
Kannan, Baskaran, Senthilvel Senapathy, Arcot Gajaraj Bhasker Raj, Subhash Chandra, Arunachalam Muthiah, Arun Prabhu Dhanapal, and C Thomas Hash. 2014. “Association Analysis of SSR Markers with Phenology, Grain, and Stover-Yield Related Traits in Pearl Millet (Pennisetum Glaucum (L.) R. Br.).” The ScientificWorld Journal 2014: 1–14. doi:10.1155/2014/562327.
Karkkainen, Katri, Veikko Koski, and Outi Savolainen. 1996. “Geographical Variation in the Inbreeding Depression of Scots Pine.” Evolution 50 (1): 111. doi:10.2307/2410785.
Karlsson, Sylvia B., and Indra K. Vasil. 1986. “Growth, Cytology and Flow Cytometry of Embryogenic Cell Suspension Cultures of Panicum Maximum Jacq. and Pennisetum Purpureum Schum.” Journal of Plant Physiology 123. Gustav Fischer Verlag, Stuttgart: 211–27. doi:10.1016/S0176-1617(86)80072-4.
Kearsey, Mike J., and A. G. L. Farquhar. 1998. “QTL Analysis in Plants ; Where Are We Now?” Heredity 80: 137–42.
Kellis, Manolis, Bruce W Birren, and Eric S Lander. 2004. “Proof and Evolutionary Analysis of Ancient Genome Duplication in the Yeast Saccharomyces Cerevisiae.” Nature 428 (6983): 617–24. doi:10.1038/nature02424.
Kent, W James. 2002. “BLAT — The BLAST -Like Alignment Tool.” Genome Research 12: 656–64. doi:10.1101/gr.229202.
192
Khan, Z. R., J. A. Pickett, L. Wadhams, and F. Muyekho. 2001. “Habitat Management Strategies for the Control of Cereal Stemborers and Striga in Maize in Kenya.” Insect Science and Its Application 21 (4): 375–80. doi:10.1017/S1742758400008481.
Kim, Changsoo, Hui Guo, Wenqian Kong, Rahul Chandnani, Lan-Shuan Shuang, and Andrew H. Paterson. 2015. “Application of Genotyping by Sequencing Technology to a Variety of Crop Breeding Programs.” Plant Science 242. Elsevier Ireland Ltd: 14–22. doi:10.1016/j.plantsci.2015.04.016.
Kojima, Shoko, Yuji Takahashi, Yasushi Kobayashi, Lisa Monna, Takuji Sasaki, Takashi Araki, and Masahiro Yano. 2002. “Hd3a, a Rice Ortholog of the Arabidopsis FT Gene, Promotes Transition to Flowering Downstream of Hd1 under Short-Day Conditions.” Plant & Cell Physiology 43 (10): 1096–1105. doi:10.1093/pcp/pcf156.
Krzywinski, Martin, Jacqueline Schein, Inanc Birol, Joseph Connors, Randy Gascoyne, Doug Horsman, Steven J Jones, and Marco A Marra. 2009. “Circos: An Information Aesthetic for Comparative Genomics.” Genome Research 19 (9). Cold Spring Harbor Lab: 1639–45.
Kuittinen, H, M J Sillanpa, and O Savolainen. 1997. “Genetic Basis of Adaptation : Flowering Time in Arabidopsis Thaliana.” Theoretical and Applied Genetics 95: 573–83.
Kumar, Lalitha Sunil. 1999. “DNA Markers in Plant Improvement: An Overview.” Biotechnology Advances 17 (September): 143–82.
Lagercrantz, Ulf, Joanna Putterill, George Coupland, and Derek Lydiate. 1996. “Comparative Mapping in Arabidopsis and Brassica, Fine Scale Genome Collinearity and Congruence of Genes Controlling Flowering Time.” The Plant Journal 9 (1): 13–20.
Lambert, Allison M., Abraham J. Miller-Rushing, and David W. Inouye. 2010. “Changes in Snowmelt Date and Summer Precipitation Affect the Flowering Phenology of Erythronium Grandiflorum (Glacier Lily; Liliaceae).” American Journal of Botany 97 (9): 1431–37. doi:10.3732/ajb.1000095.
Langmead, Ben, and Steven L Salzberg. 2012. “Fast Gapped-Read Alignment with Bowtie 2.” Nature Methods 9 (4): 357–59. doi:10.1038/nmeth.1923.
Laurie, David A., Nicky Pratchett, Jeremy H Bezant, and John W Snapet. 1994. “Genetic Analysis of a Photoperiod Response Gene on the Short Arm of Chromosome 2(2H) of Hordeum Vulgare (Barley).” Heredity 72: 619–27.
193
Lee, Ilha, Anthony Bleecker, and Richard Amasino. 1993. “Analysis of Naturally Occurring Late Flowering in Arabidopsis Thaliana.” Molecular and General Genetics MGG 237. Springer: 171–76.
Lewis, Michael W., Michelle E. Leslie, Emilee H. Fulcher, Lalitree Darnielle, Patrick Healy, Ji-Young Youn, and Sarah J. Liljegren. 2010. “The SERK1 Receptor-like Kinase Regulates Organ Separation in Arabidopsis Flowers.” Plant Journal 62 (5): 817–28. doi:10.1111/j.1365-313X.2010.04194.x.
Li, Heng, and Richard Durbin. 2010. “Fast and Accurate Long-Read Alignment with Burrows-Wheeler Transform.” Bioinformatics (Oxford, England) 26 (5). Wellcome Trust Sanger Institute, Wellcome Genome Campus, Cambridge, CB10 1SA, UK. DOI - 10.1093/bioinformatics/btp698 SRC - Pubmed ID2 - 20080505 FG - 0: 589–95. doi:10.1093/bioinformatics/btp698.
Li, Heng, Bob Handsaker, Alec Wysoker, Tim Fennell, Jue Ruan, Nils Homer, Gabor Marth, Goncalo Abecasis, Richard Durbin, and 1000 Genome Project Data Processing Subgroup. 2009. “The Sequence Alignment/map Format and SAMtools.” Bioinformatics 25 (16). Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Cambridge, CB10 1SA, UK, Broad Institute of MIT and Harvard, Cambridge, MA 02141, USA. DOI - 10.1093/bioinformatics/btp352 SRC - Pubmed ID2 - 19505943 FG - 0: 2078–79. doi:10.1093/bioinformatics/btp352.
Li, Weizhong, and Adam Godzik. 2006. “Cd-Hit: A Fast Program for Clustering and Comparing Large Sets of Protein or Nucleotide Sequences.” Bioinformatics 22 (13): 1658–59. doi:10.1093/bioinformatics/btl158.
Li, Wuxing, Yun Zhou, Xing Liu, Peng Yu, Jerry D Cohen, and Elliot M Meyerowitz. 2013. “LEAFY Controls Auxin Response Pathways in Floral Primordium Formation.” Science Signaling 6 (270): 1–6. doi:10.1126/scisignal.2003937.
Li, Yuan, Kai Yang, Wei Yang, Liwei Chu, Chunhai Chen, Bo Zhao, Yisong Li, et al. 2017. “Identification of QTL and Qualitative Trait Loci for Agronomic Traits Using SNP Markers in the Adzuki Bean.” Frontiers in Plant Science 8: 1–14. doi:10.3389/fpls.2017.00840.
Lindhout, Pim, Sjaak Van Heusden, Gerard Pet, Johan W. van Ooijen, Hans Sandbrink, Ruud Verkerk, Ria Vrielink, and Pim Zabel. 1994. “Perspectives of Molecular Marker Assisted Breeding for Earliness in Tomato.” Euphytica 79 (3). Springer: 279–86.
Loope, Lloyd L, Ole Hamann, and Charles P Stone. 1988. “Comparative Conservation Biology of Oceanic Archipelagoes: Hawaii and the Galapagos.” BioScience 38 (4). JSTOR: 272–82.
194
López, Yolanda, Aleksey Kurashev, Christine Chase, Maria Gallo, Lynn E. Sollenberger, Fredy Altpeter, and Jianping Wang. 2018. “Developing and Validating Microsatellite Markers in Elephant Grass (Pennisetum Purpureum S.).” Euphytica 214 (185): 1–`3. doi:10.1007/s10681-018-2256-6.
López, Yolanda, Jeffery Seib, Kenneth R. Woodard, Karen Chamusco, Lynn E. Sollenberger, Maria Gallo, S. Luke Flory, and Christine Chase. 2014. “Genetic Diversity of Biofuel and Naturalized Napiergrass (Pennisetum Purpureum).” Invasive Plant Science and Management 7 (2). Weed Science Society of America 810 East 10th Street, Lawrence, KS 66044-8897: 229–36. doi:10.1614/IPSM-D-13-00085.1.
Lowe, Andrew J, W Thorpe, A Teale, and J Hanson. 2003. “Characterisation of Germplasm Accessions of Napier Grass (Pennisetum Purpureum and P. Purpureum× P. Glaucum Hybrids) and Comparison with Farm Clones Using RAPD.” Genetic Resources and Crop Evolution 50 (2). Springer: 121–32.
Lu, Fei, Alexander E. Lipka, Jeffrey C. Glaubitz, Robert J. Elshire, Jerome H. Cherney, Michael D. Casler, Edward S. Buckler, and Denise E. Costich. 2013. “Switchgrass Genomic Diversity, Ploidy, and Evolution: Novel Insights from a Network-Based SNP Discovery Protocol.” PLoS Genetics 9 (1): 1–14. doi:10.1371/journal.pgen.1003215.
Lu, Hongfeng, Tao Lin, Joël Klein, Shenhao Wang, Jianjian Qi, Qian Zhou, Jinjing Sun, Zhonghua Zhang, Yiqun Weng, and Sanwen Huang. 2014. “QTL-Seq Identifies an Early Flowering QTL Located near Flowering Locus T in Cucumber.” Theoretical and Applied Genetics 127. Springer: 1–9. doi:10.1007/s00122-014-2313-z.
Ma, Xue Feng, Elaine Jensen, Nickolai Alexandrov, Maxim Troukhan, Liping Zhang, Sian Thomas-Jones, Kerrie Farrar, et al. 2012. “High Resolution Genetic Mapping by Genome Sequencing Reveals Genome Duplication and Tetraploid Genetic Structure of the Diploid Miscanthus Sinensis.” PLoS ONE 7 (3): 1–11. doi:10.1371/journal.pone.0033821.
Maheswaran, M, Ning Huang, S R Sreerangasamy, and Susan R McCouch. 2000. “Mapping Quantitative Trait Loci Associated with Days to Flowering and Photoperiod Sensitivity in Rice (Oryza Sativa L.).” Molecular Breeding 6. Springer: 145–55.
Makumburage, G. Buddhika, and Ann E. Stapleton. 2011. “Phenotype Uniformity in Combined-Stress Environments Has a Different Genetic Architecture than in Single-Stress Treatments.” Frontiers in Plant Science 2: 1–8. doi:10.3389/fpls.2011.00012.
195
Martel, E., D De Nay, S Siljak-Yakoviev, S. Brown, A. Sarr, D. De Nay, S. Siljak-Yakovlev, et al. 1997. “Genome Size Variation and Basic Chromosome Number in Pearl Millet and Fourteen Related Pennisetum Species.” Journal of Heredity 88. Am Genetic Assoc: 139–43.
Martel, E., V. Poncet, F. Lamy, S. Siljak-Yakovlev, B. Lejeune, and A. Sarr. 2004. “Chromosome Evolution of Pennisetum Species (Poaceae): Implications of ITS Phylogeny.” Plant Systematics and Evolution 249: 139–49. doi:10.1007/s00606-004-0191-6.
Martel, E., A Ricroch, and A Sarr. 1996. “Assessment of Genome Organization among Diploid Species (2n = 2x = 14) Belonging to Primary and Tertiary Gene Pools of Pearl Millet Using Fluorescent in Situ Hybridization with rDNA Probes.” Genome 39. NRC Research Press: 680–87.
Martin, K. L., P. J. Hodgen, K. W. Freeman, Ricardo Melchiori, D. B. Arnall, R. K. Teal, R. W. Mullen, et al. 2005. “Plant-to-Plant Variability in Corn Production.” Agronomy Journal 97 (6): 1603–11. doi:10.2134/agronj2005.0129.
Mathew, Lisa S, Manuel Spannagl, Ameena Al-malki, Binu George, Maria F Torres, Eman K Al-dous, Eman K Al-azwani, et al. 2014. “A First Genetic Map of Date Palm (Phoenix Dactylifera) Reveals Long-Range Genome Structure Conservation in the Palms.” BMC Genomics 15 (285): 1–10.
Matsubara, Kazuki, I. Kono, K. Hori, Y. Nonoue, N. Ono, A. Shomura, T. Mizubayashi, et al. 2008. “Novel QTLs for Photoperiodic Flowering Revealed by Using Reciprocal Backcross Inbred Lines from Crosses between Japonica Rice Cultivars.” Theoretical and Applied Genetics 117: 935–45. doi:10.1007/s00122-008-0833-0.
Matsubara, Kazuki, Eri Ogiso-Tanaka, Kiyosumi Hori, Kaworu Ebana, Tsuyu Ando, and Masahiro Yano. 2012. “Natural Variation in Hd17, a Homolog of Arabidopsis ELF3 That Is Involved in Rice Photoperiodic Flowering.” Plant and Cell Physiology 53 (4): 709–16. doi:10.1093/pcp/pcs028.
McCouch, Susan R, Y. G. Cho, M. Yano, E. Paul, M. Blinstrub, H. Morishima, and T. Kinoshita. 1997. “Report on QTL Nomenclature.” Rice Genetics Newsletter 14: 11–13.
Mckendry, Peter. 2002. “Energy Production from Biomass (Part 1): Overview of Biomass.” Bioresource Technology 83: 37–46.
McKenna, Aaron, Matthew Hanna, Eric Banks, Andrey Sivachenko, Kristian Cibulskis, Andrew Kernytsky, Kiran Garimella, et al. 2010. “The Genome Analysis Toolkit: A MapReduce Framework for Analyzing next-Generation DNA Sequencing Data.” Genome Research 20 (9): 1297–1303. doi:10.1101/gr.107524.110.
196
Melo, Arthur T. O., Radhika Bartaula, and Iago Hale. 2016. “GBS-SNP-CROP: A Reference-Optional Pipeline for SNP Discovery and Plant Germplasm Characterization Using Variable Length, Paired-End Genotyping-by-Sequencing Data.” BMC Bioinformatics 17 (29): 1–15. doi:10.1186/s12859-016-0879-y.
Miles, Cecelia M., and Marta Wayne. 2008. “Quantitative Trait Locus ( QTL ) Analysis.” Nature Education 1 (208): 1–7.
Morris, Karl, Sarah Thornber, Lesley Codrai, Christine Richardson, Adam Craig, Ari Sadanandom, Brian Thomas, and Stephen Jackson. 2010. “DAY NEUTRAL FLOWERING Represses CONSTANS to Prevent Arabidopsis Flowering Early in Short Days.” The Plant Cell 22: 1118–28. doi:10.1105/tpc.109.066605.
Muldoon, D K, and C J Pearson. 1979. “The Hybrid between Pennisetum Americanum and Pennisetum Purpureum.” Herbage Abstracts 49 (5): 189–99.
Murray, Seth C., William L. Rooney, Sharon E. Mitchell, Arun Sharma, Patricia E. Klein, John E. Mullet, and Stephen Kresovich. 2008. “Genetic Improvement of Sorghum as a Biofuel Feedstock: II. QTL for Stem and Leaf Structural Carbohydrates.” Crop Science 48: 2180–93.
Nagy, Istvan, Susanne Barth, Jeanne Mehenni-Ciz, Michael T Abberton, and Dan Milbourne. 2013. “A Hybrid next Generation Transcript Sequencing-Based Approach to Identify Allelic and Homeolog-Specific Single Nucleotide Polymorphisms in Allotetraploid White Clover.” BMC Genomics 14 (100): 1–19. doi:10.1186/1471-2164-14-100.
Negawo, Alemayehu Teressa, Abel Teshome, Alok Kumar, Jean Hanson, and Chris S Jones. 2017. “Opportunities for Napier Grass (Pennisetum Purpureum) Improvement Using Molecular Genetics.” Agronomy 7 (28): 1–21.
Noble Research Institute. 2007. “Switchgrass Establishment Requires Patience.” Ag News & Views. https://www.noble.org/news/publications/ag-news-and-views/2007/july/switchgrass-establishment-requires-patience/.
Obok, Ekemini Edet, Micheal E Aken Ova, and Godfrey A Iwo. 2012. “Forage Potentials of Interspecific Hybrids between Elephant Grass Selections and Cultivated Pearl Millet Genotypes of Nigerian Origin.” Journal of Plant Breeding and Crop Science 4 (9): 136–43. doi:10.5897/JPBCS12.019.
Osborn, Thomas C, C. Kole, I. A. P. Parkin, A. G. Sharpe, M. Kuiper, D. J. Lydiate, and Martin Trick. 1997. “Comparison of Flowering Time Genes in Brassica Rapa, B. Napus and Arabidopsis Thaliana.” Genetics 146: 1123–29.
Osgood, RV V, Wayne W. Hanna, and TL L Tew. 1997. “Hybrid Seed Production of Pearl Millet × Napiergrass Triploid Hybrids.” Crop Science 37. Crop Science Society of America: 998–99.
197
Paillard, S., T. Schnurbusch, M. Winzeler, M. Messmer, P. Sourdille, O. Abderhalden, B. Keller, and G. Schachermayr. 2003. “An Integrative Genetic Linkage Map of Winter Wheat (Triticum Aestivum L.).” Theoretical and Applied Genetics 107: 1235–42. doi:10.1007/s00122-003-1361-6.
Pandey, Manish K., Emmanuel Monyo, Peggy Ozias-Akins, Xuanquiang Liang, Patricia Guimares, Shyam N. Nigam, Hari D. Upadhyaya, et al. 2012. “Advances in Arachis Genomics for Peanut Improvement.” Biotechnology Advances 30. Elsevier Inc.: 639–51. doi:10.1016/j.biotechadv.2011.11.001.
Paterson, Andrew H., John E Bowers, Rémy Bruggmann, Inna Dubchak, Jane Grimwood, Heidrun Gundlach, Georg Haberer, et al. 2009. “The Sorghum Bicolor Genome and the Diversification of Grasses.” Nature 457 (7229): 551–56. doi:10.1038/nature07723.
Paudel, Dev R., Baskaran Kannan, Xiping Yang, Karen R. Harris-Shultz, Mahendar Thudi, Rajeev K. Varshney, Fredy Altpeter, and Jianping Wang. 2018. “Surveying the Genome and Constructing a High-Density Genetic Map of Napiergrass (Cenchrus Purpureus Schumach).” Scientific Reports 8 (1). Springer US: 1–11. doi:10.1038/s41598-018-32674-x.
Paux, Etienne, Pierre Sourdille, Ian Mackay, and Catherine Feuillet. 2012. “Sequence-Based Marker Development in Wheat: Advances and Applications to Breeding.” Biotechnology Advances 30. Elsevier Inc.: 1071–88. doi:10.1016/j.biotechadv.2011.09.015.
Peng, Ze, Wen Fan, Liping Wang, Dev R. Paudel, Dante Leventini, Barry L. Tillman, and Jianping Wang. 2017. “Target Enrichment Sequencing in Cultivated Peanut (Arachis Hypogaea L.) Using Probes Designed from Transcript Sequences.” Molecular Genetics and Genomics. Springer Berlin Heidelberg, 1–11. doi:10.1007/s00438-017-1327-z.
Phan, Huyen T.T., Kasia Rybak, Stefania Bertazzoni, Eiko Furuki, Eric Dinglasan, Lee T. Hickey, Richard P. Oliver, and Kar Chun Tan. 2018. “Novel Sources of Resistance to Septoria Nodorum Blotch in the Vavilov Wheat Collection Identified by Genome-Wide Association Studies.” Theoretical and Applied Genetics 131. Springer Berlin Heidelberg: 1223–38. doi:10.1007/s00122-018-3073-y.
Phukan, Ujjal J., Gajendra S. Jeena, and Rakesh K. Shukla. 2016. “WRKY Transcription Factors: Molecular Regulation and Stress Responses in Plants.” Frontiers in Plant Science 7: 1–14. doi:10.3389/fpls.2016.00760.
Poland, Jesse A., Patrick J Brown, Mark E Sorrells, and Jean-Luc Jannink. 2012. “Development of High-Density Genetic Maps for Barley and Wheat Using a Novel Two-Enzyme Genotyping-by-Sequencing Approach.” PloS One 7 (2). Public Library of Science: 1–8.
198
Poland, Jesse A., J Endelman, J Dawson, J Rutkoski, S Y Wu, Y Manes, S Dreisigacker, et al. 2012. “Genomic Selection in Wheat Breeding Using Genotyping-by-Sequencing.” Plant Genome 5 (3): 103–13. doi:Doi 10.3835/Plantgenome2012.06.0006.
Poland, Jesse A., and Trevor W Rife. 2012. “Genotyping-by-Sequencing for Plant Breeding and Genetics.” The Plant Genome 5 (3). Crop Science Society of America: 92–102.
Pootakham, Wirulda, Nukoon Jomchai, Panthita Ruang-areerate, Jeremy R. Shearman, Chutima Sonthirod, Duangjai Sangsrakru, Somvong Tragoonrung, and Sithichoke Tangphatsornruang. 2015. “Genome-Wide SNP Discovery and Identification of QTL Associated with Agronomic Traits in Oil Palm Using Genotyping-by-Sequencing (GBS).” Genomics 105. Elsevier B.V.: 288–95. doi:10.1016/j.ygeno.2015.02.002.
Powell, Jerrel B, and Glenn W. Burton. 1966. “A Suggested Commercial Method of Producing an Interspecific Hybrid Forage in Pennisetum.” Crop Science 6. Crop Science Society of America: 378–79.
Powell, Wayne, Gordon C Machray, and Jim Provan. 1996. “Polymorphism Revealed by Simple Sequence Repeats.” Trends in Plant Science 1 (7). Elsevier: 215–22.
Price, David L., and Michael D. Casler. 2014. “Divergent Selection for Secondary Traits in Upland Tetraploid Switchgrass and Effects on Sward Biomass Yield.” Bioenergy Research 7: 329–37. doi:10.1007/s12155-013-9374-8.
Pruitt, Kim D., Tatiana Tatusova, and Donna R. Maglott. 2007. “NCBI Reference Sequences (RefSeq): A Curated Non-Redundant Sequence Database of Genomes, Transcripts and Proteins.” Nucleic Acids Research 33: 501–4. doi:10.1093/nar/gkl842.
Punnuri, Somashekhar M., Jason G. Wallace, Joseph E. Knoll, Katie E. Hyma, Sharon E. Mitchell, Edward S. Buckler, Rajeev K. Varshney, Bharat P. Singh, and T H E Plant Genome. 2016. “Development of a High-Density Linkage Map and Tagging Leaf Spot Resistance in Pearl Millet Using GBS Markers.” Plant Genome 9 (2): 1–13. doi:10.3835/plantgenome2015.10.0106.
Ra, Kaien, Fumitaka Shiotsu, Jun Abe, and Shigenori Morita. 2012. “Biomass Yield and Nitrogen Use Efficiency of Cellulosic Energy Crops for Ethanol Production.” Biomass and Bioenergy 37. Elsevier: 330–34.
Rafalski, Antoni. 2002. “Applications of Single Nucleotide Polymorphisms in Crop Genetics.” Current Opinion in Plant Biology 5. Elsevier: 94–100.
199
Rajaram, Vengaldas, Thirunavukkarasu Nepolean, Senapathy Senthilvel, Rajeev K. Varshney, Vincent Vadez, Rakesh K Srivastava, Trushar M Shah, Ambawat Supriya, Sushil Kumar, and Basava Ramana Kumari. 2013. “Pearl Millet [Pennisetum Glaucum (L.) R. Br.] Consensus Linkage Map Constructed Using Four RIL Mapping Populations and Newly Developed EST-SSRs.” BMC Genomics 14 (159). BioMed Central Ltd: 1–15.
Ram, Hari Har, Hari Har Ram, and Rakesh Yadava. 2007. Genetic Resources and Seed Enterprises: Management and Policies. New India Publishing.
Reis, Gabriela Barreto dos, Amanda Teixeira Mesquita, Giovana Augusta Torres, Larissa Fonseca Andrade-Vieira, Antônio Vander Pereira, and Lisete Chamma Davide. 2014. “Genomic Homeology between Pennisetum Purpureum and Pennisetum Glaucum (Poaceae).” Comparative Cytogenetics 8 (3): 199–209. doi:10.3897/CompCytogen.v8i3.7732.
Rice, Peter, Lan Longden, and Alan Bleasby. 2000. “EMBOSS: The European Molecular Biology Open Software Suite.” Trends in Genetics 16 (6): 276–77. doi:10.1016/S0168-9525(00)02024-2.
Richmond, Todd A., and Anthony B. Bleecker. 1999. “A Defect in Beta-Oxidation Causes Abnormal Inflorescence Development in Arabidopsis.” The Plant Cell 11: 1911–24. doi:10.1105/tpc.11.10.1911.
Rosyara, Umesh R., Walter S. De Jong, David S. Douches, and Jeffrey B. Endelman. 2016. “Software for Genome-Wide Association Studies in Autopolyploids and Its Application to Potato.” The Plant Genome 9 (2): 1–10. doi:10.3835/plantgenome2015.08.0073.
Rozen, S, and H Skaletsky. 2000. “Primer3 on the WWW for General Users and for Biologist Programmers.” Methods in Molecular Biology 132. Whitehead Institute for Biomedical Research, Cambridge, MA, USA. SRC - Pubmed ID2 - 10547847 FG - 0: 365–86.
Russell, Joanne, Martin Mascher, Ian K. Dawson, Stylianos Kyriakidis, Cristiane Calixto, Fabian Freund, Micha Bayer, et al. 2016. “Exome Sequencing of Geographically Diverse Barley Landraces and Wild Relatives Gives Insights into Environmental Adaptation.” Nature Genetics 48 (9). Nature Publishing Group: 1024–30. doi:10.1038/ng.3612.
Samach, Alon. 2012. “Control of Flowering.” In Plant Biotechnology and Agriculture:, First Edit, 387–404. Elsevier Inc. doi:10.1016/B978-0-12-381466-1.00025-0.
Schlegel, Rolf H J. 2009. Dictionary of Plant Breeding. CRC Press.
Schnable, Patrick S., Doreen Ware, RS Fulton, and JC Stein. 2009. “The B73 Maize Genome: Complexity, Diversity, and Dynamics.” Science 326 (5956): 1112–15.
200
Schneider, Katharina. 2005. “Mapping Populations and Principles of Genetic Mapping.” In The Handbook of Plant Genome Mapping: Genetic and Physical Mapping, edited by Khalid Meksem and Gunter Kahl, 1–21. WILEY-VCH Verlag GmbH & Co. KGaA. doi:10.1002/3527603514.ch1.
Schofield, Eileen K. 1989. “Effects of Introduced Plants and Animals on Island Vegetation: Examples from Galápagos Archipelago.” Conservation Biology 3 (3). Wiley Online Library: 227–39.
Semagn, Kassa, Raman Babu, Sarah Hearne, and Michael Olsen. 2014. “Single Nucleotide Polymorphism Genotyping Using Kompetitive Allele Specific PCR (KASP): Overview of the Technology and Its Application in Crop Improvement.” Molecular Breeding. doi:10.1007/s11032-013-9917-x.
Shen, Jiawei, Zhiqiang Li, Jianhua Chen, Zhijian Song, Zhaowei Zhou, and Yongyong Shi. 2016. “SHEsisPlus, a Toolset for Genetic Studies on Polyploid Species.” Scientific Reports 6. Nature Publishing Group: 1–10. doi:10.1038/srep24095.
Shepherd, M., M. Cross, M. J. Dieters, and R. Henry. 2003. “Genetic Maps for Pinus Elliottii Var. Elliottii and P. Caribaea Var. Hondurensis Using AFLP and Microsatellite Markers.” Theoretical and Applied Genetics 106: 1409–19. doi:10.1007/s00122-002-1185-9.
Simpson, Jared T, Kim Wong, Shaun D Jackman, Jacqueline E Schein, Steven J M Jones, Inanç Birol, Jared T Simpson, et al. 2009. “ABySS: A Parallel Assembler for Short Read Sequence Data.” Genome Research 19 (June): 1117–23. doi:10.1101/gr.089532.108.
Sinche, Marco. 2013. “Genetic Improvement of Elephantgrass (Pennisetum Purpureum Schum.) as a Bioenergy Feedstock.” University of Florida.
Sinche, Marco, Baskaran Kannan, Carlos Corsato, Dev R. Paudel, Jianping Wang, and Fredy Altpeter. 2018. “Development and Characterization of Napiergrass (Pennisetum Purpureum) Mapping Population for Flowering Time and Biomass Related Traits.” Submitted for Publication.
Singer, S. D., N. T. Krogan, and N. W. Ashton. 2007. “Clues about the Ancestral Roles of Plant MADS-Box Genes from a Functional Analysis of Moss Homologues.” Plant Cell Reports 26: 1155–69. doi:10.1007/s00299-007-0312-0.
Singh, Bharat P., Hari P Singh, and Eric Obeng. 2013. “Elephantgrass.” In Biofuel Crops: Production, Physiology and Genetics, edited by Bharat P Singh, 271–91. CAB International.
Smith, Rex L., and M K U Chowdhury. 1983. “Mitochondrial DNA Polymorphism in Male-Sterile and Fertile Cytoplasms of Pearl Millet.” Crop Science 29: 809–14.
201
Smith, Rex L., M. E. Schweder, M. K. U. Chowdhury, J. C. Seib, and S. C. Schank. 1993. “Development and Application of RFLP and RAPD DNA Markers in Genetic Improvement of Pennisetum for Biomass and Forage Production.” Biomass and Bioenergy 5 (1): 51–62.
Sollenberger, Lynn E., G M Prine, W R Ocumpaugh, Wayne W. Hanna, C S Jones, S C Schank, and R S Kalmbacher. 1989. “Registration of ‘Mott’dwarf Elephantgrass.” Crop Science 29. Crop Science Society of America: 827–28.
Sollenberger, Lynn E., Kenneth R. Woodard, João M.B. B. Vendramini, John E. Erickson, Kenneth A. Langeland, M. Kimberly Mullenix, Chaein Na, et al. 2014. “Invasive Populations of Elephantgrass Differ in Morphological and Growth Characteristics from Clones Selected for Biomass Production.” Bioenergy Research 7 (May): 1382–91. doi:10.1007/s12155-014-9478-9.
Somerville, Chris, Heather Youngs, Caroline Taylor, Sarah C Davis, and Stephen P Long. 2010. “Feedstocks for Lignocellulosic Biofuels.” Science 329: 790–93.
Song, Xiaoming, Tingting Ge, Ying Li, and Xilin Hou. 2015. “Genome-Wide Identification of SSR and SNP Markers from the Non-Heading Chinese Cabbage for Comparative Genomic Analyses.” BMC Genomics 16 (328): 1–18. doi:10.1186/s12864-015-1534-0.
Stinchcombe, JR, and HE Hoekstra. 2008. “Combining Population Genomics and Quantitative Genetics : Finding the Genes Underlying Ecologically Important Traits.” Heredity 100: 158–70. doi:10.1038/sj.hdy.6800937.
Stone, Eric A., and Julien F. Ayroles. 2009. “Modulated Modularity Clustering as an Exploratory Tool for Functional Genomic Inference.” PLoS Genetics 5 (5): 1–13. doi:10.1371/journal.pgen.1000479.
Struwig, M. 2007. “Fingerprinting Pennisetum Purpureum Schumach. Varieties and Cultivars Using AFLP Analyses.” North West University.
Suárez-López, P, K Wheatley, F Robson, H Onouchi, F Valverde, and G Coupland. 2001. “CONSTANS Mediates between the Circadian Clock and the Control of Flowering in Arabidopsis.” Nature 410: 1116–20. doi:10.1038/35074138.
Tamura, Koichiro, Glen Stecher, Daniel Peterson, Alan Filipski, and Sudhir Kumar. 2013. “MEGA6: Molecular Evolutionary Genetics Analysis Version 6.0.” Molecular Biology and Evolution, 2725–29. doi:10.1093/molbev/mst197.
Taylor, D R, and P K Ingvarsson. 2003. “Common Features of Segregation Distortion in Plants and Animals.” Genetica 117: 27–35. doi:10.1023/A:1022308414864.
202
Taylor, M G, and I K Vasil. 1987. “Analysis of DNA Size , Content and Cell Cycle in Leaves of Napier Grass (Pennisetum Purpureum Schum .).” Theoretical and Applied Genetics 74: 681–86.
Thiel, T., W. Michalek, Rajeev K. Varshney, and A. Graner. 2003. “Exploiting EST Databases for the Development and Characterization of Gene-Derived SSR-Markers in Barley (Hordeum Vulgare L.).” Theoretical and Applied Genetics 106. Institute of Plant Genetics and Crop Plant Research (IPK), Corrensstrasse 3, D-06466 Gatersleben, Germany. DOI - 10.1007/s00122-002-1031-0 SRC - Pubmed ID2 - 12589540 FG - 0: 411–22. doi:10.1007/s00122-002-1031-0.
Thompson, J. B. 1919. Napier and Merker Grass : Two New Forage Crops for Florida. Gainesville: University of Florida, Agricultural Experiment Station.
Tollenaar, M, and J Wu. 1999. “Yield Improvement in Temperate Maize Is Attributed to Greater Stress Tolerance.” Crop Science 39: 1597–1604. doi:10.2135/cropsci1999.3961597x.
Tong, Xinggang, Laurence H Smith, and Perry L Mccarty. 1990. “Methane Fermentation of Selected Lignocellulosic Materials.” Biomass 21: 239–55.
Trick, Martin, Yan Long, Jinling Meng, and Ian Bancroft. 2009. “Single Nucleotide Polymorphism (SNP) Discovery in the Polyploid Brassica Napus Using Solexa Transcriptome Sequencing.” Plant Biotechnology Journal 7: 334–46. doi:10.1111/j.1467-7652.2008.00396.x.
Tsai, Yu Chang, Nikkí A. Delk, Naweed I. Chowdhury, and Janet Braam. 2007. “Arabidopsis Potential Calcium Sensors Regulate Nitric Oxide Levels and the Transition to Flowering.” Plant Signaling and Behavior 2 (6): 446–54. doi:10.4161/psb.2.6.4695.
Uitdewilligen, Jan G A M L, Anne Marie A Wolters, Bjorn B. D’hoop, Theo J A Borm, Richard G F Visser, and Herman J. van Eck. 2013. “A next-Generation Sequencing Method for Genotyping-by-Sequencing of Highly Heterozygous Autotetraploid Potato.” PLoS ONE 8 (5): 1–14. doi:10.1371/journal.pone.0062355.
USDA, ARS, and NGRP. 1954. “PI 215603.” Germplasm Resources Information Network - (GRIN). [Online Database] National Germplasm Resources Laboratory, Beltsville, Maryland. http://www.ars-grin.gov/cgi-bin/npgs/acc/display.pl?1176422.
———. 1963. “PI 288787.” Germplasm Resources Information Network - (GRIN). [Online Database] National Germplasm Resources Laboratory, Beltsville, Maryland. http://www.ars-grin.gov/cgi-bin/npgs/acc/display.pl?1218439.
203
van Ooijen, Johan W. 1992. “Accuracy of Mapping Quantitative Trait Loci in Autogamous Species.” Theoretical and Applied Genetics 84: 803–11. doi:10.1007/BF00227388.
———. 2006. “JoinMap 4.0: Software for the Calculation of Genetic Linkage Maps in Experimental Populations.” Wageningen, Netherlands: Kyazma BV.
Varshney, Rajeev K., Chengcheng Shi, Mahendar Thudi, Cedric Mariac, Jason Wallace, Peng Qi, He Zhang, et al. 2017. “Pearl Millet Genome Sequence Provides a Resource to Improve Agronomic Traits in Arid Environments.” Nature Biotechnology 35 (10): 969–76.
Veeckman, E, K Vandepoele, T Asp, and T Ruttink. 2016. “Genomic Variation in the FT Gene Family of Perennial Ryegrass (Lolium Perenne).” In Breeding in a World of Scarcity: Proceedings of the 2015 Meeting of the Section “Forage Crops and Amenity Grasses” of Eucarpia, 121–26. doi:10.1007/978-3-319-28932-8.
Vicente-Chandler, Jose, Servando Silva, and Jacinto Figarella. 1959. “The Effect of Nitrogen Fertilization and Frequency of Cutting on the Yield and Composition of Three Tropical Grasses.” Agronomy Journal 51. American Society of Agronomy: 202–6.
Voorrips, Roeland E. 2002. “MapChart: Software for the Graphical Presentation of Linkage Maps and QTLs.” The Journal of Heredity, 77–78. doi:10.1093/jhered/93.1.77.
Wang, Cuiling, Yanhui Chen, Lixia Ku, Tiegu Wang, Zhaohui Sun, Fangfang Cheng, and Liancheng Wu. 2010. “Mapping QTL Associated with Photoperiod Sensitivity and Assessing the Importance of QTL×environment Interaction for Flowering Time in Maize.” PLoS ONE 5 (11): 8–10. doi:10.1371/journal.pone.0014068.
Wang, Jia, Zhitong Chen, Surong Jin, Zhongli Hu, Yibing Huang, and Ying Diao. 2017. “Development and Characterization of Simple Sequence Repeat (SSR) Markers Based on a Full-Length cDNA Library of Napier Grass (Pennisetum Purpureum Schum).” Genes and Genomics 39 (12). The Genetics Society of Korea: 1297–1305. doi:10.1007/s13258-017-0536-5.
Wang, Shengchu, CJ Basten, P Gaffney, and Z-B Zeng. 2005. “Windows QTL Cartographer 2.5. User Manual.” International Immunology. Department of Statistics, North Carolina State University, Raleigh, NC. doi:10.1093/intimm/dxs150.
Wang, Yan, Silong Sun, Bo Liu, Hui Wang, Jie Deng, Yongcui Liao, Qian Wang, Feng Cheng, Xiaowu Wang, and Jian Wu. 2011. “A Sequence-Based Genetic Linkage Map as a Reference for Brassica Rapa Pseudochromosome Assembly.” BMC Genomics 12 (1): 239. doi:10.1186/1471-2164-12-239.
204
Wanjala, Bramwel W., Meshack Obonyo, Francis N. Wachira, Alice Muchugi, Margaret Mulaa, Jagger Harvey, Robert A. Skilton, Janice Proud, and Jean Hanson. 2013. “Genetic Diversity in Napier Grass (Pennisetum Purpureum) Cultivars: Implications for Breeding and Conservation.” AoB PLANTS 5 (January): 1–10. doi:10.1093/aobpla/plt022.
Woodard, Kenneth R., and Lynn E. Sollenberger. 2012. “Production of Biofuel Crops in Florida: Elephantgrass.” SS-AGR-297.
Wu, Changyin, Changjun You, Caishun Li, Tuan Long, Guoxing Chen, Mary E Byrne, and Qifa Zhang. 2008. “RID1, Encoding a Cys2/His2-Type Zinc Finger Transcription Factor, Acts as a Master Switch from Vegetative to Floral Development in Rice.” Proceedings of the National Academy of Sciences of the United States of America 105 (35): 12915–20.
Wu, K. K., W. Burnquist, M. E. Sorrells, T. L. Tew, P. H. Moore, and S. D. Tanksley. 1992. “The Detection and Estimation of Linkage in Polyploids Using Single-Dose Restriction Fragments.” Theoretical and Applied Genetics 83: 294–300. doi:10.1007/BF00224274.
Wu, S, J Yang, Y Huang, Y Li, T Yin, S D Wullschleger, G a Tuskan, and R Wu. 2010. “An Improved Approach for Mapping Quantitative Trait Loci in a Pseudo-Testcross: Revisiting a Poplar Mapping Study.” Bioinform Biol Insights 4: 1–8. doi:10.4137/BBI.S13402.TYPE.
Xiao-ping, Lu, Yun Jin-feng, Gao Cui-ping, and Surya Acharya. 2011. “Quantitative Trait Loci Analysis of Economically Important Traits in Sorghum Bicolor × S. Sudanense Hybrid.” Canadian Journal of Plant Science 91: 81–90. doi:10.4141/cjps09112.
Xie, S. Q., J. Y. Feng, and Y. M. Zhang. 2014. “Linkage Group Correction Using Epistatic Distorted Markers in F2 and Backcross Populations.” Heredity 112: 479–88. doi:10.1038/hdy.2013.127.
Xie, Wengang, Xinquan Zhang, Hongwei Cai, Linkai Huang, Yan Peng, and Xiao Ma. 2011. “Genetic Maps of SSR and SRAP Markers in Diploid Orchardgrass (Dactylis Glomerata L.) Using the Pseudo-Testcross Strategy.” Genome 54: 212–21. doi:10.1139/G10-111.
Xie, Xin-Ming, Feng Zhou, Xiang-Qian Zhang, and Ju-Ming Zhang. 2009. “Genetic Variability and Relationship between MT-1 Elephant Grass and Closely Related Cultivars Assessed by SRAP Markers.” Journal of Genetics 88 (3). Springer: 281–90.
Xu, Shizhong. 2008. “Quantitative Trait Locus Mapping Can Benefit from Segregation Distortion.” Genetics 180: 2201–8. doi:10.1534/genetics.108.090688.
205
Xu, Yang, Pengcheng Li, Zefeng Yang, and Chenwu Xu. 2017. “Genetic Mapping of Quantitative Trait Loci in Crops.” Crop Journal 5. Elsevier B.V.: 175–84. doi:10.1016/j.cj.2016.06.003.
Yadav, R. S., F. R. Bidinger, C. T. Hash, Y. P. Yadav, O. P. Yadav, S. K. Bhatnagar, and C. J. Howarth. 2003. “Mapping and Characterisation of QTL X E Interactions for Traits Determining Grain and Stover Yield in Pearl Millet.” Theoretical and Applied Genetics 106: 512–20. doi:10.1007/s00122-002-1081-3.
Yang, Xiping, Md. S. Islam, Sushma Sood, Stephanie Maya, Erik A. Hanson, Jack Comstock, and Jianping Wang. 2018. “Identifying Quantitative Trait Loci (QTLs) and Developing Diagnostic Markers Linked to Orange Rust Resistance in Sugarcane (Saccharum Spp.).” Frontiers in Plant Science 9 (350): 1–10. doi:10.3389/fpls.2018.00350.
Yang, Xiping, Jian Song, James Todd, Ze Peng, Dev R. Paudel, Ziliang Luo, Xiaokai Ma, et al. 2018. “Target Enrichment Sequencing of 307 Germplasm Accessions Identified Ancestry of Ancient and Modern Hybrids and Signatures of Adaptation and Selection in Sugarcane ( Saccharum Spp.), a ‘sweet’ Crop with ‘bitter’ Genomes.” Plant Biotechnology Journal, 1–11. doi:10.1111/pbi.12992.
Yang, Xiping, Jian Song, Qian You, Dev R. Paudel, Jisen Zhang, and Jianping Wang. 2017. “Mining Sequence Variations in Representative Polyploid Sugarcane Germplasm Accessions.” BMC Genomics 18 (594). BMC Genomics: 1–16. doi:10.1186/s12864-017-3980-3.
Yang, Xiping, Sushma Sood, Neil Glynn, Md Sariful Islam, Jack Comstock, and Jianping Wang. 2017. “Constructing High-Density Genetic Maps for Polyploid Sugarcane (Saccharum Spp.) and Identifying Quantitative Trait Loci Controlling Brown Rust Resistance.” Molecular Breeding 37 (116). Molecular Breeding: 1–12.
Yano, Masahiro, Yuichi Katayose, Motoyuki Ashikari, Utako Yamanouchi, Lisa Monna, Takuichi Fuse, Tomoya Baba, Kimiko Yamamoto, Yosuke Umehara, and Yoshiaki Nagamura. 2000. “Hd1, a Major Photoperiod Sensitivity Quantitative Trait Locus in Rice, Is Closely Related to the Arabidopsis Flowering Time Gene CONSTANS.” The Plant Cell Online 12. Am Soc Plant Biol: 2473–83.
Yano, Masahiro, Shoko Kojima, Yuji Takahashi, Hongxuan Lin, and Takuji Sasaki. 2001. “Genetic Control of Flowering Time in Rice, a Short-Day Plant.” Plant Physiology 127. Am Soc Plant Biol: 1425–29. doi:10.1104/pp.010710.1.
Yanofsky, Martin F., Hong Ma, John L. Bowman, Gary N. Drews, Kenneth A. Feldmann, and Elliot M. Meyerowitz. 1990. “The Protein Encoded by the Arabidopsis Homeotic Gene Agamous Resembles Transcription Factors.” Nature 346: 35–39. doi:10.1038/346035a0.
206
Yoshida, H., K. Hirano, T. Sato, N. Mitsuda, M. Nomoto, K. Maeo, E. Koketsu, et al. 2014. “DELLA Protein Functions as a Transcriptional Activator through the DNA Binding of the INDETERMINATE DOMAIN Family Proteins.” Proceedings of the National Academy of Sciences 111 (21): 7861–66. doi:10.1073/pnas.1321669111.
Yu, Jun, Songnian Hu, Jing Juan Jian Jun Wang, Gane Ka-shu Wong, Songgang Li, Bin Liu, Yajun Deng, et al. 2002. “A Draft Sequence of the Rice Genome (Oryza Sativa L. Ssp. Indica).” Science (New York, N.Y.) 296 (5565): 79–92. doi:10.1126/science.1068037.
Yu, Xiaoqing, and Shuying Sun. 2013. “Comparing a Few SNP Calling Algorithms Using Low-Coverage Sequencing Data.” BMC Bioinformatics 14 (274): 1–15. doi:10.1186/1471-2105-14-274.
Zahn, Laura M., J. Leebens-Mack, C. W. DePamphilis, H. Ma, and G. Theissen. 2005. “To B or Not to B a Flower: The Role of DEFICIENS and GLOBOSA Orthologs in the Evolution of the Angiosperms.” Journal of Heredity 96 (3): 225–40. doi:10.1093/jhered/esi033.
Zhang, Gengyun, Xin Liu, Zhiwu Quan, Shifeng Cheng, Xun Xu, Shengkai Pan, Min Xie, et al. 2012. “Genome Sequence of Foxtail Millet (Setaria Italica) Provides Insights into Grass Evolution and Biofuel Potential.” Nature Biotechnology 30 (6): 549–54. doi:10.1038/nbt.2195.
Zhang, Zhen, Haihong Shang, Yuzhen Shi, Long Huang, Junwen Li, Qun Ge, Juwu Gong, et al. 2016. “Construction of a High-Density Genetic Map by Specific Locus Amplified Fragment Sequencing (SLAF-Seq) and Its Application to Quantitative Trait Loci (QTL) Analysis for Boll Weight in Upland Cotton (Gossypium Hirsutum).” BMC Plant Biology 16 (79). BMC Plant Biology: 1–18. doi:10.1186/s12870-016-0741-4.
Zhao, Xinxin, Linkai Huang, Xinquan Zhang, Jianping Wang, Defei Yan, Ji Li, Lu Tang, Xiaolong Li, and Tongwei Shi. 2016. “Construction of High-Density Genetic Linkage Map and Identification of Flowering-Time QTLs in Orchardgrass Using SSRs and SLAF-Seq.” Scientific Reports 6 (29345): 1–10. doi:10.1038/srep29345.
Zhou, Liyuan, Shouye Liu, Weixun Wu, Daibo Chen, Xiaodeng Zhan, Aike Zhu, Yingxin Zhang, et al. 2016. “Dissection of Genetic Architecture of Rice Plant Height and Heading Date by Multiple-Strategy-Based Association Studies.” Scientific Reports 6 (29718). Nature Publishing Group: 1–10. doi:10.1038/srep29718.
207
Zhou, Sifan, Chengran Wang, Taylor F. Frazier, Haidong Yan, Peilin Chen, Zhihong Chen, Linkai Huang, et al. 2018. “The First Illumina-Based de Novo Transcriptome Analysis and Molecular Marker Development in Napier Grass (Pennisetum Purpureum).” Molecular Breeding 38 (95). Molecular Breeding: 1–14. doi:10.1007/s11032-018-0852-8.
Zhou, Wencai, Zaixiang Tang, Jing Hou, Nan Hu, and Tongming Yin. 2015. “Genetic Map Construction and Detection of Genetic Loci Underlying Segregation Distortion in an Intraspecific Cross of Populus Deltoides.” PLoS ONE, 1–13. doi:10.1371/journal.pone.0126077.
Zhou, Xiaojing, Youlin Xia, Xiaoping Ren, Yulin Chen, Li Huang, Shunmou Huang, Boshou Liao, Yong Lei, Liyin Yan, and Huifang Jiang. 2014. “Construction of a SNP-Based Genetic Linkage Map in Cultivated Peanut Based on Large Scale Marker Development Using next-Generation Double-Digest Restriction-Site-Associated DNA Sequencing (ddRADseq).” BMC Genomics 15 (351): 1–14. doi:10.1186/1471-2164-15-351.
Zub, H. W., S. Arnoult, and M. Brancourt-Hulmel. 2011. “Key Traits for Biomass Production Identified in Different Miscanthus Species at Two Harvest Dates.” Biomass and Bioenergy 35. Elsevier Ltd: 637–51. doi:10.1016/j.biombioe.2010.10.020.
208
BIOGRAPHICAL SKETCH
Dev Paudel received his Ph.D. in Agronomy from the University of Florida (UF) in
the fall of 2018. His research focuses on utilizing traditional plant breeding approaches
and modern bioinformatic and molecular techniques in the genetic improvement of
napiergrass (elephant grass) and its interspecific hybrids with pearl millet. This will
eventually help in sustainability of forage and biofuel feedstock.
Dev received his MS degree in Crop Science from Texas Tech University,
Lubbock, Texas, USA where his research focused on evaluating the potential of new
testing methods for cotton breeding. Plant breeders, ginners, farmers, and spinning
mills can use the information obtained from his research to make informed decisions for
increased profitability in the premium yarn market. After his MS degree, he worked at
the Texas A&M AgriLife Research, Pecos, Texas where he optimized nutrient media for
algae production. He has a BS degree in Agriculture from Tribhuvan University, Nepal.