Development of new cowpea (Vigna unguiculata) mutant 1
genotypes, analysis of their agromorphological variation, genetic 2
diversity and Population structure 3
Made DIOUF1, Sara DIALLO1, François Abaye BADIANE1,2, Oumar DIACK1 and Diaga 4
DIOUF1* 5
1Laboratoire Campus de Biotechnologies Végétales, Département de Biologie Végétale, 6
Faculté des Sciences et Techniques, Université Cheikh Anta Diop, Dakar-Fann, Code Postal 7
10700, Dakar, Sénégal 8
2Faculté des Sciences et Technologies de l’Education et de la Formation, Université Cheikh 9
Anta Diop, Dakar-Fann, Code Postal 10700, Dakar, Sénégal 10
* Correspondence: 11
13
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Keywords: Cowpea; Vigna unguiculata; Gamma rays; Induced mutagenesis; Plant breeding; 14
Agromorphological characterization; ISSR 15
Abstract 16
Cowpea is one of the most important legume grain in the SubSaharian region of Africa used 17
for human consumption and animal feeding but its production is hampered by biotic and 18
abiotic constraints raising the need to broaden its genetic basis. For this purpose, the seeds of 19
two cowpea varieties Melakh and Yacine were respectively irradiated with 300 and 340 Gy. 20
The developed mutant populations were agromorphologically characterized from M5 to M7 21
while the genetic diversity of the last were evaluated using 13 ISSR markers. Based on 22
agromorphological characterization, variation of flower color, pod length, seed coat color and 23
seed weight with respectively 78.01, 68.29, 94.48 and 57.58% heritability were recorded in 24
the mutant lines. PCA analyses allowed to identify the elite mutants based on their 25
agromorphological traits while Pearson’s correlation results revealed a positive correlation 26
between yield component traits. Three subpopulations were identified through STRUCTURE 27
analyses but assignment of the individuals in each group was improved using DAPC. 28
Analysis of Molecular Variance revealed that the majority (85%) of the variance rather 29
existed within group than among (15%) group. Finally, our study allowed to select new 30
promising mutant genotypes which could be tested for multi local trials to evaluate their 31
agronomic performance. 32
Introduction 33
Cowpea [Vigna unguiculata (L.) Walp. 2n = 2x =22] is an important crop legume for tropical 34
and subtropical regions grown in Africa, Southern Europe, Latin America, Southeast Asia, 35
and southwestern regions of North America on 12 496 305 hectares 36
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
(http://www.fao.org/faostat/en/#data/QC/visualize). Its production is estimated to 7 233 408 37
tons, Nigeria, Niger, Burkina Faso, Ghana, United Republic of Tanzania, Myanmar, Mali, 38
Cameroon, Sudan (including Sudan and South Soudan) and Kenya, are the top producers in 39
the world. In the Sahelian part of Africa, the crop plays a major role in human nutrition. For 40
instance, the fresh seeds, grilled on wood fire, are consumed in Senegal while the dry seeds 41
are used in a wide range of meal compositions. The young leaves cooked as spinach are eaten 42
in Eastern and Southern Africa while the hay as well as the seed are used for livestock feeding 43
in several African countries (for review see [1]). Compared to others legumes, its seed 44
contains higher amount of proteins which is estimated to 25% and a substantial amount of 45
minerals and vitamins based on the recent work published by [2], raising cowpea as a 46
valuable crop to fight malnutrition for the low-income farmers. Based on the estimation 47
realized by [3], cowpea which establishes a symbiosis with Bradyrhizobium, is a good 48
nitrogen fixing crop (70 to 350 kg nitrogen per hectare) leading to soil fertility improvement. 49
Despite its importance, the production of the crop is hampered by a wide range of biotic 50
(virus, bacteria, fungi, parasitic weeds and nematodes) and abiotic (drought, heat, etc.) 51
constraints [1, 4]. This susceptibility to a wide range of biotic and abiotic stresses is attributed 52
to the narrow genetic basis of the crop due to single domestication event and its self-53
pollinating pattern of reproduction [5]. To overcome these constraints, the genetic diversity 54
hidden into the existing germplasms which contains relevant agronomic traits have been 55
exploited during the past decades to increase the production of crops through the development 56
of outstanding cultivars, based on methods such as pure line selection, mass selection, 57
pedigree breeding, single seed descent and back crossing [6]. Despite these efforts, the genetic 58
basis of the realized lines is still narrow which arise the need to develop novel outstanding 59
varieties particularly in the era of climate change. For this purpose, technique such as 60
mutagenesis is a valuable tool to induce genetic variation for cultivar improvements. 61
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Mutagenesis has widely been used for the past seventy years in order to improve many 62
economically important crops leading to the selection of 3,320 varieties which are superior to 63
the natural cultivar in term of productivity, resistance to biotic and/or abiotic constraints 64
(https://mvd.iaea.org/#!Search). According to this database, only 16 cowpea varieties were 65
bred using gamma ray mutation techniques and 5 of them were selected from Africa precisely 66
in Kenya, Zambia and Zimbabwe. 67
Gamma rays are ionizing radiation which deeply penetrate into the cells of the target tissues 68
where they interact with molecules to generate reactive oxygen species (ROS) causing base 69
substitutions, genome rearrangements such as insertions, deletions, inversions and 70
translocations [7]. The base substitution caused by ROS is due to the conversion of guanines 71
into 8-oxo-Gs which induces mispairings with adenine while genome rearrangements are 72
caused by error-prone non homologous end joining (NHEJ) rather than error-free homologous 73
recombination which result from double-strand breaks (DSBs). When DSBs and NHEJ occur 74
in several genomic regions, they create favorable conditions for copy number variations 75
(CNVs), presence/absence variations (PAVs) and translocations [8,7]. Presently, it is well 76
documented that these genetic modifications induce agro morphological variations affecting 77
plant height, growth habit, number of leaves per plant, leaves color, number of branches per 78
plant, days to first flower, flower color, flowering ability, maturity, number of pods per plant, 79
number of seed per plant, pod and seed coat color, seed eye color, weight of 100 seeds and 80
tolerance to Maruca vitrata pod borer which were observed among cowpea mutants [9,10,11] 81
[12,13,14]. In view of these variations, several research teams attempted to characterize 82
cowpea mutant population using morphological traits, yield components [15,16,10,12,13,17] 83
and recently seed storage proteins [10,12]. To overcome the limits of using these traits, 84
Random amplified polymorphic DNA (RAPD; [10]) and inter-simple sequence repeat (ISSR; 85
[10,12]) were recently used to understand the genetic organization of some cowpea mutant 86
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
populations. of these markers. The analysis of ISSR has generated more informative results, 87
since these sequences are abundant, widely distributed across the eukaryotic genome, highly 88
reproducible and use SSR as primers allowing the amplification of inter SSR region [18]. 89
ISSR are useful to study genetic diversity, genome mapping or evolutionary biology in many 90
crop species and they overcome the limitations of other markers, such as low reproducibility 91
of RAPD and high cost of AFLP (Amplified fragment length polymorphic) and the use a 92
single primer during the amplification processes [18,19]. ISSR combine the advantages of 93
SSR, AFLP and RAPD markers which do not require prior genome sequence information and 94
are efficient to detect genetic variation among cowpea varieties [20,21] or mutant lines 95
[10,12]. ISSR primers can be unanchored with 1 to 4 degenerate nucleotides at 3’ or 5’ end to 96
avoid their slippery within the repeat units and smears apparition during amplification but 97
previous studies showed that primers anchored at 3’ gave more clear bands [22,12]. 98
The aim of this study was to broaden the genetic basis of cowpea using gamma irradiation 99
technique specifically to develop mutant populations for which their genetic diversity and 100
allelic richness were assessed and to use the generated information to select new elite 101
genotypes. 102
Materials and Methods 103
Plant materials and gamma irradiation 104
Two inbred cowpea varieties Melakh and Yacine (Table 1) widely cultivated in Senegal were 105
selected from the national germplasm and used in this study [23]. They belong to the early 106
maturity group which reach physiological maturity 64 days after sowing (DAS) under well-107
watered conditions [24,25]. In total 216 dry and healthy seeds for each variety Melakh and 108
Yacine were exposed respectively to 300 and 340 Gy of gamma ray. The irradiation was 109
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
performed at the International Atomic Energy Agency (IAEA), Agriculture and 110
Biotechnology Laboratory, A-2444 Seibersdorf, Austria using a cobalt 60 source Gammacell 111
Model No. 220. The control seeds were not exposed to gamma irradiation. 112
113
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Table 1: Agromorphological characters and pedigree of 40 cowpea mutants and their parent used in this study. 114 115
1 : White ; 2 : White with purple border ; 3 : Light purple ; 4 : Dark purple 116 Genotypes used for DNA analysis are indicated in bold. For the mutant line codes, the first or second letter refers to the parent’s name, the number which 117 appeared before « M » indicates the selected individual during the M generation. 118
Genotype codes Pedigree Growth habit
Leaflet length (cm)
Flower color
Pod length (cm)
Pod weigth (g)
Seed length (mm)
Seed coat color
Eye color
Y1-M4-9M5-2M6-M7 Yacine Erect 11.5 3 11.70 2.49 8.67 Brown Brown Y1-M4-9M5-3M6-M7 Yacine Erect 10 2 8.30 1.03 9.18 White-brown Brown Y1-M4-8M5-1M6-M7 Yacine Erect 9 1 14.40 2.59 10.68 White Brown Y1-M4-3M5-2M6-M7 Yacine Erect 8.1 2 16.60 2.13 9.82 White Brown Y1-M4-11M5-1M6-M7 Yacine Erect 12.3 2 12.40 1.09 7.67 White-brown Brown Y7-M4-1M5-1M6-M7 Yacine Prostrate 12.3 2 8.50 0.74 9.50 White Brown Y7-M4-6M5-1M6-M7 Yacine Erect 11.2 2 10.70 1.27 9.41 White-brown Brown Y7-M4-4M5-3M6-M7 Yacine Erect 10.2 2 12.10 1.71 9.63 Brown Brown Y13-M4-4M5-1M6-M7 Yacine Erect 9.6 1 10.00 0.99 8.11 White Brown Y7-M4-12M5-3M6-M7 Yacine Prostrate 9 4 12.10 1.54 10.80 Brown Brown Y7-M4-10M5-1M6-M7 Yacine Erect 9 4 14.10 1.88 9.77 White Brown Y1-M4-1M5-2M6-M7 Yacine Erect 12 2 13,15 1,66 9,62 White Brown Y1-M4-5M5-1M6-M7 Yacine Erect 10 1 8,70 1,25 10,55 Brown Brown Y1-M4-10M5-2M6-M7 Yacine Prostrate 13,5 2 15,80 2,78 8,43 white Brown Y1-M4-11M5-3M6-M7 Yacine Prostrate 10,5 1 19,00 3,26 9,05 White Brown Y1-M4-12M5-1M6-M7 Yacine Erect 8,5 2 10,65 0,93 9,36 White-brown Brown Y1-M4-14M5-1M6-M7 Yacine Erect 10,5 1 13,55 2,31 11,13 Brown Brown Y1-M4-15M5-1M6-M7 Yacine Erect 10,5 1 14,30 2,52 10,59 Brown Brown Y7-M4-4M5-2M6-M7 Yacine Erect 9,2 4 13,60 2,43 8,60 White Brown Y7-M4-3M5-1M6-M7 Yacine Erect 12 2 18,2 2,88 10,90 White-brown Brown Y7-M4-1M5-3M6-M7 Yacine Semi erect 12,5 2 11,90 1,51 9,84 White Brown Y1-M4-16M5-2M6-M7 Yacine Prostrate 10,5 3 14,40 1,51 8,38 White Brown Y7-M4-7M5-3M6-M7 Yacine Erect 6,5 4 12,45 1,46 9,51 White Brown Y7-M4-7M5-1M6-M7 Yacine Erect 12,6 2 19,30 2,91 9,99 White Brown YD7-M4-4M5-3M6-M7 Yacine Erect 10,2 2 12,10 1,71 9,63 Brown Brown Y7-M4-9M5-2M6-M7 Yacine Erect 8,6 4 10,40 1,27 9,99 Brown Brown Yacine Ndiaga Aw x Melakh Erect 11 1 14.20 2.71 11.02 Brown Brown
was not certified by peer review
) is the author/funder. All rights reserved. N
o reuse allowed w
ithout permission.
The copyright holder for this preprint (w
hichthis version posted M
ay 25, 2020. .
https://doi.org/10.1101/2020.05.22.109041doi:
bioRxiv preprint
Table 1: Agro-morphological characters and pedigree of 40 cowpea mutants and their parent used in this study (continued). 119 120
1 : White ; 2 : White with purple border ; 3 : Light purple ; 4 : Dark purple 121 Genotypes used for DNA analysis are indicated in bold. For the mutant line codes, the first or second letter refers to the parent’s name, the number which 122 appeared before « M » indicates the selected individual during the M generation. 123
Genotype codes Pedigree Growth habit
Leaflet length (cm)
Flower color
Pod length (cm)
Pod weigth (g)
Seed length (mm)
Seed color Eye color
Me51M4-10M5-1M6-M7 Melakh Prostrate 11.5 2 13.4 1.806 9.065 White Brown Me51M4-11M5-2M6-M7 Melakh Prostrate 12.2 1 14.4 1.49 10.2 White Brown Me51M4-14M5-2M6-M7 Melakh Erect 12.9 1 16.4 2.586 12.155 White White Me51M4-25M5-3M6-M7 Melakh Semi erect 10 1 13.1 2.38 7.46 White White Me51M4-29M5-1M6-M7 Melakh Prostrate 11 1 16.2 1.912 9.315 White White Me51M4-39M5-1M6-M7 Melakh Prostrate 10 1 15.58 2.47 10 White White Me51M4-77M5-M7 Melakh Erect 13 1 6.16 1.25 16.55 White White Me51M4-8M5-3M6-M7 Melakh Erect 10,5 1 16,6 1,73 8,685 White Brown Me51M4-9M5-1M6-M7 Melakh Prostrate 10 2 15,8 1,76 9,035 White Brown Me51M4-9M5-3M6-M7 Melakh Semi erect 12 1 14 134 7,99 White Brown Me51M4-20M5-1M6-M7 Melakh Prostrate 11,3 2 16,74 2,63 9,235 White White Me51M4-24M5-1M6-M7 Melakh Prostrate 10 2 19,67 2,804 11,475 White Wite Me51M4-36M5-1M6-M7 Melakh Prostrate 13,2 1 12,4 1,25 8,455 White White Me51M4-102M5-M7 Melakh Erect 9,33 1 11 0,85 12,33 White White Melakh IS86-292 x IT83s-742-13 Semi erect 12.1 1 13.4 2.423 10.25 White White
was not certified by peer review
) is the author/funder. All rights reserved. N
o reuse allowed w
ithout permission.
The copyright holder for this preprint (w
hichthis version posted M
ay 25, 2020. .
https://doi.org/10.1101/2020.05.22.109041doi:
bioRxiv preprint
Experimental design and development of mutagenized 124
populations 125
The development of the mutagenized population was performed in different experimental 126
fields located in different sites in the western part of Senegal. In each site, the seeds of each 127
generation were sown using intra row spacing of 50 cm and 75 cm of inter raw spacing. The 128
irradiated seeds (M1) for each variety (Melakh and Yacine) were sown in a separate field in 129
August 2013 around Bambey during the raining season for the development of M1 130
population. Based on their yield, 12 and 7 mutant plants of Melakh and Yacine respectively 131
were selected and harvested for the development of the M2 population. For these purposes, 132
103 and 87 seeds respectively from M1 plant of Melakh and Yacine were sown during the dry 133
season in April 2014 in the “Centre National de Recherches Agronomiques (CNRA)” at 134
Bambey (Senegal) to develop M2 population. At maturity, the most productive mutant plants 135
12 for Melakh and 7 for Yacine were selected, harvested and their seeds sown in the same site 136
(CNRA) in September 2015 to develop the M3 population. From the M3 mutant plants, the 137
most productive were harvested and 88 and 81 seeds for the Mutant Melakh and Yacine were 138
respectively sown in September 2016 in the experimental field located in Ngolgane in the 139
vicinity of Niakhar (Senegal) in accordance with the same experimental design previously 140
described for M2, for the development of M4 population. At maturity stage, the plants were 141
harvested and a single descent method was used to develop M5 population. Thirty-nine (39) 142
and thirty-six (36) seeds of Yacine and Melakh mutants were respectively sown in December 143
2017 in a pot filled with sand from Sanghalkam (Senegal) and watered with tap water 3 times 144
a week. The mutant plants were grown in the Shadehouse of the “Département de Biologie 145
Végétale” at University Cheikh Anta Diop. The M6 and the M7 population were sown 146
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
respectively in May 2017 and August 2017 in the field at the Teaching and Research farm of 147
the “Département de Biologie Végétale” at University Cheikh Anta Diop. 148
Agro-morphological characterization 149
Based on the previous studies it is well documented that irradiation promotes the expression 150
of recessive characters in the advanced generations [26,27]. Therefore, qualitative and 151
quantitative parameters were analyzed from M5 to M7 populations. For instance, in our 152
studies, seed color and pod length variation were noticed in the 5th generation (M5). The 153
scored qualitative parameters encompassed: rate of germination, leaflet abnormalities, leaflet 154
shape, growth habit, flower color, day to first flowering and seed coat color. The quantitative 155
parameters included, percentage of germination, plant height, mean of pod length, number of 156
pod per plant, number of seeds per pod, width and length mean of the seed and weight of 100 157
seeds. The plant height was measured using a tape measure (Cow head brand) from the 158
cotyledonary node to the top of the plant at the appearance of the first flower [28]. The mean 159
length of 5 pods was measured with a tape measure. Width and length mean of the seed was 160
measured using a vernier caliper (Mutshito) and weighted using a balance (sartoruis®). The 161
data of the quantitative traits recorded throughout M5 to M7 were used for statistical analyses. 162
Statistical Analysis of Data 163
Analysis of variance (ANOVA) and correlation of the quantitative traits were carried 164
out using R software, (R Development Core Team, 2011, version 3.6.1). In order to determine 165
the association between quantitative traits such as the seed mean length, seed mean width, 166
seed mean weight, pod mean length, pod mean width, mean number of seed per pod, plant 167
height and qualitative parameters (like leaflet shape, flower color, growth habit and 168
abnormalities, seed color and flowering date), a standardized Principal Component Analysis 169
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
(sPCA) was performed with R software using the adjusted means of the measured traits to 170
assess the contribution of each of them on the genetic variability. The phenotypic coefficient 171
of variance, the genotypic coefficient of variation (GCV), the genetic advance (GA) and the 172
broad sense heritability (h2) was calculated using R software (R Development Core Team, 173
2011, version 3.6.1). 174
The genotypic variance (σ2g) was calculated using the following formula: 175
σ2g= (MSG- MSE)/ r, 176
Where MSG is the mean square of genotypes, MSE is the mean square of error, and r is the 177
number of advanced generation. 178
The Phenotypic Variance (σ2p) was assessed as follow: 179
σ2p= σ2g+ σ2e, 180
Where σ2g is the genotypic variance and σ2e is the mean squares of error. Error variance 181
σ2e=MSE. 182
According to [29], the estimation of phenotypic and genotypic coe�cient of variation were 183
calculated as follows: 184
��� �����
�X100 185
��� �√���
�X100, X is the mean. 186
GCV and PCV values were considered as low (0-10%), moderate (10-20%), and high (�20%) 187
following [30]. 188
The Heritability Estimate (broad sense) (h2%), h2%�σ2g
σ2p100 189
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
The heritability percentage was considered as low (0-30%), moderate (30-60%), and high 190
(�60%) [31]. 191
The Expected and Estimated Genetic Advance (GA) 192
GA= k.σp. h2, and The Genetic Advance as Percentage of Mean (GA %) is carried using 193
��% ��
� 100 194
GA was calculated using the method of [32] and selection intensity (k) was assumed to be 195
5%; where k = 2.06, a constant and σp is the phenotypic standard deviation. 196
Genetic advance as percentage of mean was categorized as low (0-10%), moderate (10-20%), 197
and high (>20%) [33]. 198
The Pearson’s correlation coefficient (r) for trait linkage evaluation were performed 199
using R software to determine the association between quantitative characters. The genetic 200
distance between mutants and their parents based on quantitative traits was tested using 201
multivariate analysis. To generate the dendrogram, similar matrices were used based on the 202
Ward methods cluster analysis [34]. 203
DNA genotyping 204
DNA extraction 205
Five hundred (500) mg of young leaflet collected from each individual plant of 1 month old 206
randomly selected were grounded in mortar in accordance with the protocol developed by 207
Fulton et al. (1995). RNA was removed by adding 50 µg/ml of RNAse A (CalBiochem) in 208
each tube which was incubated at room temperature for 1 h and then DNA was purified 209
according the protocol described by [23]. After precipitation, DNA was dried for 20 min with 210
a Speed Vac ® Plus Sc110 (Savant) and dissolved in 100 µl of TE x 0.1. The quantity and the 211
quality of the DNA extract were performed using a NanoDrop™ One/OneC Microvolume 212
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
UV-Vis Spectrophotometer (Thermo Scientific) at A260, A280, A260/A280 and A260/A230, 213
then the samples were stored at -20°C or used for amplification. 214
215
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Amplification of DNA and electrophoresis 216
The amplification reaction was performed in 0.2 ml tube puReTaq Ready-To-GoTM PCR 217
beads (GE Healthcare) containing 2.5 U of lyophilized PuReTaq, 200 µM dNPT and 1.5 mM 218
MgCl2, 0.5 µM of each ISSR primer (TSINGKE, China) and 25 ng of DNA in a final volume 219
of 25 µl. The tubes were loaded in a Prime thermocycler (TECHNE, UK) programmed for 220
pre-denaturating step of 5 min at 95°C followed by 40 cycles of 30 s 95°C, 1 min at 38 to 221
52°C (depending on primer, Table 2), 1 min at 72°C and a final extension of 8 min at 72°C. 222
After amplification, the PCR products were separated on 2% agarose gel (Sigma) for 2 h at 223
70V. The gel was finally stained during 30 min with GelRed X10,000 (Biotium) according to 224
the manufacturer's instructions and photographed under UV light using Gel Doc system (High 225
performance UV Transilluminator UVP). 226
Genetic variation analysis 227
Amplifications were repeated three times for each single ISSR primer in order to retain the 228
clear and reproducible bands. On this basis, the total number of amplified bands was 229
calculated: their size estimated and the percentage of polymorphic bands evaluated. The 230
polymorphic bands were scored using a binary code as presence (1) or absence (0) to 231
construct a data matrix. 232
The Shannon diversity index, heterozygosity (Nei index) and the private alleles were 233
calculated using GenAlex 6.5 software [35]. To evaluate the discriminatory power of each 234
marker, the Polymorphic Information Content (PIC) was calculated using the PowerMarker 235
software 3.25 [36], while assessing the genetic variation among and within groups an analysis 236
of molecular variance (AMOVA) was performed using GenAlex 6.5 software [35]. 237
238
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Population structure analysis 239
The population structure was analyzed using the Bayesian clustering approach implemented 240
in the software STRUCTURE 2.3.3 [37] while the number of subpopulations was tested from 241
1 to 10 independent runs. Using the admixture model [38] each simulation set to 100 000 burn 242
in periods and 10 runs of 200 000 iterations of Markov chain Monte Carlo (MCMC) were 243
performed. These results were uploaded to the online software, Structure Harvester [39] to 244
determine the most likely number of subpopulations using the Evanno Δk method [40]. To 245
assign the individuals into clusters, a membership coefficient q ≥ 0.7 was used. The genotypes 246
within a cluster with membership coefficients q < 0.7 were considered as genetically 247
admixed. 248
Discriminant Analysis of Principal Components (DAPC) was performed on the basis of the 249
binary matrix data, in order to confirm or invalidate the pattern of the genetic structure 250
obtained with STRUCTURE and to identify the loci responsible for possible differentiation 251
between genetic groups. This analysis was performed using the package « adegenet » [41] of 252
R software [42]. 253
A Neighbor Joining [43] dendrogram was constructed using the inter-individual distance 254
matrix, calculated on the basis of the [44, 45] index. This analysis was performed using 255
Darwin 6.0 software [46]. 256
Results 257
Agromorphological characterization of the munants 258
Variation of qualitative traits among the mutants 259
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
The germination rate of the irradiated seeds of Melakh and Yacine was respectively 98.15% 260
and 99.08% in the M1 generation. This germination rate was respectively for Melakh and 261
Yacine 92.23% and 96.5 in the M2, 93.3% and 92.4% in the M3, 63.64 and 81.5% in M4, 262
90% and 92% in M5, 88% and 93% in M6 and 87.5% and 93.82% in M7 while it was 100% 263
for the control. These results suggested that gamma irradiation at 300 or 340 Gy affected 264
negatively the germinative power of the seeds. Growth habit variability appeared in the M5 265
for Melakh mutants where 94% were prostrate, 3% erected and 3% semi erected like the 266
control. During the M6, 3% were prostrate, 97% erected and semi erected phenotype was not 267
observed among the mutants. The M7 included 62% prostrate, 7% semi erected and 31% 268
erected. In the Yacine mutants the prostrate phenotype appeared in the M5 with 38% and 61% 269
were erected like the control but in the M6 the percentage of the individuals with these traits 270
was respectively 4% and 96%. The semi erected phenotype appeared for the first time in the 271
M7 with 4% while 18% and 78% were respectively prostrate and erected (Table 1). The 272
leaflet shape within the M7 mutants of Melakh revealed the existence of 4 leaflet forms which 273
were globular (6%), subglobular (76%), hastate (6%) and subhastate (12%) while the leaflet 274
form of the control was subhastate. The leaflet of Yacine was subglobular but 4% of its 275
mutants had subhastate leaflet indicating that gamma irradiation induced leaflet shape 276
variability (Fig 1A-D). Foliar number abnormalities were 11.11%, 11.5% and 0% respectively 277
in M5, M6 and M7 (S1 Fig). These foliar number abnormalities, such as three primary leaves 278
around the node in some mutants instead of two opposite leaves and tetraleaflet, revealed that 279
the gamma irradiation affected the genes controlling leaflet shape in these mutants. 280
Phenotypic variability in flower color was first observed in the M5 of the mutants of Yacine 281
and Melakh. Among the mutants of Melakh 22%, 50% and 36% had white flowers with 282
purple border respectively in M5, M6 and M7 unlike to the control which had white flowers 283
(Fig 1E-H). In contrast, three patterns of flower coloration were observed among the mutants 284
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
of Yacine. In the M5, 42.5% of the mutant showed white flowers like the control but 35 and 285
22.5% had white flowers with purple border and purple flowers respectively. The proportion 286
of the mutants with white flowers with purple border was 48% in M6 and 43% in M7 whereas 287
the mutants with purple flowers represented 13 and 26% in the M6 and M7 respectively. 288
These data suggested that gamma irradiation affected the genes controlling flower color of the 289
cowpea samples. Sterility characterized by flower abortion was observed only among the 290
mutants of Melakh in M5, M6 and M7 with 11%, 5% and 2% respectively. Among our 291
population, 50% of flowering was reached 45 days after sowing (DAS) in M6 and M7 for 292
Melakh mutants while this value was respectively 46 and 50 DAS for the mutants of Yacine 293
(S2 Fig). The color of the seed coat was unchanged from M1 to M7 for the mutants of Melakh 294
but some of them showed brown or beige eye. In the M4 of Yacine mutants, the seed coat was 295
brown like the control except one mutant where brown seeds, white seeds and pigmented 296
brown and white were harvest (Fig 1J). In the M5, 43% had white seeds, 38% of the mutants 297
had brown seeds as the control and 19% had light brown seeds (Fig 1K-M). 298
Fig 1. Variation of qualitative traits observed in the populations. 299
A-D: Leaflet shape, Bar = 6.25 cm; E-H: Flower color, Bar = 0.84 cm; I-M: Seed coat color, 300
Bar = 1 cm 301
A: Globular (Yacine); B: Hastate; C: Subglobular; D: Suhastate (Melakh); E: White flowers 302
(Yacine and Melakh), F: White flower with purple border; G: Dark purple; H: Light purple; I: 303
Brown seed coat (Yacine); J: Brown, white, white pickleed brown seed coat (from on single 304
plant in M4 of the mutant of Yacine); K: light brown speckled in dark brown seed coat; L: 305
White seed coat; M: Light brown seed coat 306
Variation of quantitative traits and yield components among the mutants 307
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
To advance the mutant populations from M1 to M4, pedigree method was used and the 308
selection criteria was based on the plant yield and no shattering pods for the mutants of 309
Yacine only. In contrast, from M5 to M7, single seed descent method was used. In M5, the 310
mutants of Melakh were selected based on 100 seed weights but in M6 and M7 one more 311
yield component (pod length) was including in the selection criteria. The average pod length 312
of Melakh control measured across generations was 19 cm while the value obtained in the 313
mutants ranged from 12.5 to 25 cm cross generation M5, M6 and M7 (Fig 2A). At the same 314
time, the average pod length of Yacine control was estimated to 14.65 cm but variability of 315
the pod length was observed among the mutants of Yacine ranging from 8.70 to 19 cm cross 316
generations M5, M6 and M7. 317
The number of pods per plant ranged from 1 to 6, 1 to 20 and 2 to 15 in M5, M6 and M7 318
respectively for the mutants of Yacine, for the control it ranged from 3 to 6. These values 319
ranged from 1 to 7, 2 to 35 and 1 to 43 for the mutants of Melakh and from 3 to 10 for the 320
control. The mean seed length varied from 8.75 to 10.72, 8.1 to 11.66, 7.46 to 12.16 mm 321
respectively for M5, M6 and M7 for the mutants of Melakh and 9.2 mm for the control (Fig 322
2B). For the mutants of Yacine the mean seed length varied from 6.12 to 10.04, 9.12 to 11.20 323
and 8.11 to 11.13 mm respectively for M5, M6 and M7 and 10 mm for the control. The 324
number of seeds per pod varied from 9 to 15, 4 to 12 and 3 to 12 for the mutants of Melakh 325
respectively for M5, M6 and M7 and 10 for the control. For the mutants of Yacine, the 326
number of seeds per pod ranged from 7 to 13, 5 to 12 and 4 to 16 respectively in M5, M6 and 327
M7 and 9 for the control. The 100 seed-weight ranged from 16.67 to 28.52, 20.07 to 32.28 328
and 13.55 to 38 g respectively for M5, M6 and M7 for the mutants of Melakh but the values 329
recorded for the control ranged from 18.35 to 32.12 g. At the same time, 100 seed-weight 330
ranged from 10.78 to 24.5, 17.27 to 33.16, 13.83 to 30.03 g respectively in M5, M6 and M7 331
population of the mutants of Yacine and from 16.4 to 30.40 g for the control. Two mutant 332
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
lines of Melakh produced more seeds (44 to 184) per plant regardless of the generation. 333
Similar results were observed among the mutants of Yacine (17 to 150 seeds). 334
Fig 2. Variation of the pod length (A) and seed size (B) in M5 among the mutants of Melakh. 335
On the top: Pods and seeds harvested from the mutant Me51M4-39M5. Bar = 1.54 cm 336
On the bottom: Pods and seeds harvested from the parent control Melakh. Bar = 1 cm 337
Genotypes clustering based on Principal Components Analysis and 338
correlation between traits 339
The projection of agromorphological parameters collected from M7 in the PCA Biplot 340
showed that the axis 1 explained 28% of the variation. This axis encompassed the elite 341
mutants in term of seed length, pod weight and seed weight (Y7-M4-3M5-1M6-M7 and 342
Me51M4-14M5-1M6-M7) and the early maturing mutant lines (Me51M4-36M5-1M6-M7, 343
Y1-M4-16M5-2M6-M7 and Y7-M4-1M5-1M6-M7) compared to their control parents Yacine 344
and Melakh. The axis 2 explaining 19.7% of the variation, was constituted by the mutant lines 345
(Y1-M4-11M5-3M6-M7, Y7-M4-1M5-3M6-M7, Y1-M411M5-3M6-M7, Me51M4-39M5-346
1M6-M7, Me51M4-29M5-1M6-M7, Me51M4-20M5-1M6-M7, Me51M4-10M5-1M6-M7, 347
Me51M4-9M5-1M6-M7, Me51M4-11M5-1M6-M7) which acquired a new growth habit i.e 348
prostrate compared to their parents (Fig 3). The evaluation of the Pearson’s coefficient 349
between agromorphological characters cross generation (from M5 to M7) showed that steam 350
pigmentation (Pg) was significantly and negatively correlated to seed color (SC) (r = -0.5, p = 351
0.01 in M5; r = -0.59, p = 0.001 in M6 and r = -0.64, p = 0.001 in M7), the number of seeds 352
per plant (NSP) was significantly and positively correlated to the pod length (PdL) (r = 0.42 p 353
= 0.05 in M5, r = 0.62 p = 0.001 in M6 and r = 0.82 p = 0.001 in M7), Seed weight (SWg) and 354
Seed Length (SL) (r = 0.86 p = 0.001 in M5, r = 0.60 p = 0.001 in M6 and r = 0.74 p = 0.001 355
in M7) (S1 Table, S2 Table and Table 2). In addition to these, in the M7, which is supposed to 356
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
be more stable, the yield component parameters such as the pod number (PdN) and the growth 357
habit (GH) are significantly correlated (r = 0.37 p = 0.05) and pod weight (PWg) and PdL (r = 358
0.78 p = 0.001). The pod width is significantly and positively correlated to the SWg (r = 0.54 359
p<0.001), PWg (r = 0.40.01 to SL (r = 0.46 p = 0.01) as do the pod weight (PdWg) and SWg 360
(r = 0.56 p = 0.001). In contrast, Seed width (SW) and PdW are negatively and significantly 361
correlated (r= -0.47 p = 0.01, Table 2). 362
Fig 3. Principal Component Analysis of 40 mutant M7 lines and their parent based on agro-363
morphological data. 364
365
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Table 2. Estimation of Pearson’s correlation between the agromorphological characters in the M7 of the mutant lines. 366 367 PH Pig GH DF FC PdL PdW SL SW SWg NSP PdN PdWg SC PH 1 Pig 0,31* 1 GH 0,72*** 0,38* 1 DF 0,10 0,05 0,11 1 FC -0,12 0,13 -0,09 -0,07 1 PdL 0,18 0,17 0,29 0,14 -0,18 1 PdW 0,01 -0,44** -0,14 -0,04 0,01 0,23 1 SL -0,09 -0,26 -0,19 -0,27 -0,10 0,13 0,46** 1 SW 0,11 0,15 0,22 -0,19 -0,17 -0,04 -0,47** 0,02 1 SWg -0,02 -0,29 -0,13 -0,31* -0,18 0,33* 0,54*** 0,74*** 0,27 1 NSP 0,31* 0,18 0,36* 0,11 -0,20 0,82*** 0,26 0,10 0,04 0,31* 1 PdN 0,33* 0,20 0,37* -0,19 0,00 0,19 0,15 0,17 0,14 0,19 0,30 1 PdWg 0,10 -0,06 0,13 0,10 -0,13 0,78*** 0,40** 0,20 0,06 0,55*** 0,67*** 0,06 1 SC -0,35* -0,64*** -0,40** -0,04 0,20 -0,40** 0,33* 0,25 0,03 0,29 -0,34* -0,18 -0,08 1
* significant at 5% level of probability; ** significant at 1% level of probability; *** significant at 0.1% level of probability 368
was not certified by peer review
) is the author/funder. All rights reserved. N
o reuse allowed w
ithout permission.
The copyright holder for this preprint (w
hichthis version posted M
ay 25, 2020. .
https://doi.org/10.1101/2020.05.22.109041doi:
bioRxiv preprint
Relationship among mutants based on agromorphological parameters 369
Based on the agro-morphological characters the M7 population was divided into 3 groups. 370
The first group which encompassed 52% of the individuals are divided into 3 subgroups. The 371
subgroup 1 included only the mutants (Me51M4-102M5-M7, Me51M4-36M5-1M6-M7, and 372
Me51M4-77M5-M7) which are sister of Y13M4-4M5-1M6-M7, Y1M4-11M5-1M6-M7 and 373
Me51M4-25M5-3M6-M7. The subgroup 2 only comprised mutants of Yacine except 374
Me51M4-8M5-3M6-M7 which clustered with genotypes sharing the same characters such as 375
erected port, white brown-eyed seed. The subgroup 3 included only mutants of Yacine. The 376
second group included 14% of the genotypes which had all erected port and brown-eyed seed 377
and contained only mutants of Yacine and their parent. The third group was the second largest 378
one with 33% of the genotypes was divisible into 2 main subgroups. The first subgroup 379
encompassed Melakh its mutants and 2 mutants of Yacine. The second included mutants of 380
Melakh and 2 mutants of Yacine which shared several agromorphological characters with 381
Melakh among these long pod and white seed (Fig 4). 382
Fig 4. Hierarchical classification of 40 mutant lines and their parent based on agro-383
morphological characters using Ward’s method in R software. 384
Heritability of agromorphological characters in the mutants 385
Using R software, it was possible to analyze the phenotypic (PCV) and genotypic (GCV) 386
coefficient of variance and the heritability values for the different measured traits in the 387
mutant populations. The PCV varied from 100.59 for pod number (PdN) to 9.12 for Seed 388
length (SL). Plant height (PH) recorded high value (75.1%) of PCV in contrast, day of 389
flowering (DF), pod width (PdW), eye color (HC) and leaflet length (LF) showed respectively 390
9.37, 10.14, 10.32 and 15.89%. The recorded GCV values ranged from 67.12% for pod 391
number (PdN) to 5.97% for date of flowering (DF). All the studied traits showed a genetic 392
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
coefficient of variance (GCV) below 50% except pod number and seed color (SC). The 393
genetic advance (GA) as a percentage of the mean recorded in the mutants ranged from 0.45% 394
for petiole length to 92.27% for pod number. High values of heritability were recorded for 395
seed color (94.48%), steam pigmentation (92.83%), flower color (78.01%), eye color 396
(68.98%), pod length (68.29%), pod width (59.06%), seed weight (57.58%), number of seed 397
per plant (52.11%), seed length (51.19%), the remaining characters showed values below 50% 398
(Table 3). 399
400
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Table 3. Estimation of mean values of phenotypic coefficient of variance (PCV), genotypic coefficient of variation (GCV %), broad sense 401
heritability (h2bs%) and genetic advance as% of the mean (GA %) of eighteen traits (quantitative and qualitative) in the M7 of 40 402
cowpea mutant lines. 403
PH LF LW PeL Pig GH DF FC PdL PdW SL SW SWg NSP PdN PdWg SC EC Mean 51.8 10.4 5.8 7.04 - - 49.6 - 15.4 8.8 9.5 6.6 0.2 10.6 6.4 1.9 - - PCV% 75.1 15.89 24.34 27.45 20.85 46.9 9.37 46.67 21.73 10.14 9.12 14.65 25.69 20.1 100.59 25.76 54.1 10.32 GCV% 23.67 6.7 13.31 9.28 20.09 17.59 5.97 41.22 17.96 7.79 6.53 9.48 19.49 14.51 67.12 17.97 52.59 8.57 GA % @ 5%
20.18 5.82 14.98 0.45 - - 7.84 - 30.56 12.33 9.61 12.62 30.47 21.57 92.27 25.82 - -
h2(%) 17.14 17.78 29.89 11.43 92.83 14.06 40.62 78.01 68.29 59.06 51.19 41.85 57.58 52.11 44.53 48.66 94.48 68.98
404
PH: Plant Height, LF: Leaflet Length; PeL: Petiole; Pig: Pigmentation; GH: Growth Habit; DF: Date of Flowering; FC: Flower Color; PdL: Pod Length; 405 PdW: Pod Weight; SL: Seed Length; SW: Seed Weight; NSP: Seed Number per Plant; PdN: Pod Number; PdWg: Pod Weight; SC: Seed Color; EC: Eye 406 Color 407
was not certified by peer review
) is the author/funder. All rights reserved. N
o reuse allowed w
ithout permission.
The copyright holder for this preprint (w
hichthis version posted M
ay 25, 2020. .
https://doi.org/10.1101/2020.05.22.109041doi:
bioRxiv preprint
Genetic characterization of the mutants 408
Genetic diversity induced by mutagenesis revealed by ISSR markers 409
In total, the 13 ISSR markers used in this study gave polymorphism amplifying 129 bands 410
(loci) in the 18 mutant lines and their 2 parents. The size of the amplified bands ranged from 411
150 to 2000 bp. Of these, 111 (86%) bands were polymorphics with numbers ranging from 5 412
(UBC827) to 12 (UBC825 and UBC844) for each primer (Table 4). The percentage of the 413
polymorphic bands per primer ranged from 60% (UBC809) to 100% (UBC825, UBC844, 414
17899A, 17899B and, HB10). The band frequencies ranged from 0.318 (UBC841) to 0.808 415
(HB12) with a mean of 0.473±0.145. The genetic diversity ranged from 0.142 (UBC841 and 416
UBC823) to 0.293 (UBC844 and 17899B) with a mean of 0.220±0.049. The Polymorphic 417
Information Content (PIC) values for each primer varied between 0.167 (UBC841) and 0.307 418
(HB09) with a mean of 0.238±0.044 (Table 5). Based on these data, our analyses showed that 419
the average diversity level for all the mutants and their parents was equal to 0.222 (Nei 420
index). The level of the genetic diversity observed in group 1 (h = 0.308) was higher than the 421
one in group 2 (h = 0.146; p = 0.0001) and in group 3 (h = 0.212; p = 0.0001). The level of 422
genetic diversity in group 3 was higher than in group 2 (p = 0.016). The genotypes which 423
belong to group 1 also recorded a greater number of private alleles (17 bands vs. 4 for group 2 424
and 7 for group 3, Table 6). The Shannon’s information index was higher in group 1 (0.463) 425
than in group 2 and 3 with respectively 0.211 and 0.319 but for the entire population this 426
value was 0.331. To investigate the genetic variance within and among genetic pools, 427
AMOVA was carried out in this study. Our results revealed that the majority of the variance 428
rather existed within group (85%) than among group (15%) (Table 7). 429
430
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Table 4. List of the Inter-simple sequence repeat (ISSR) primers with their annealing temperatures (Tm) used to genotype the mutants 431
M7 and their parent, total number of bands, number of polymorphic bands, percentage of polymorphic bands and the band size range. 432
ISSR primer codes
Sequences (5’-3’)
Annealing temperature (°C)
Total number of bands
Number. of polymorphic bands
Percentage of polymorphic bands
(%)
Band size range (pb)
UBC809 (AG)8G 52.2 10 6 60 1400-200 UBC825 (AC)8T 49.8 12 12 100 1500-200 UBC841 (GA)8YC 52.6 14 11 78.57 1800-150 UBC844 (CT)8AC 52.6 12 12 100 1300-300 17899A (CA)7AG 49.2 11 11 100 2000-250 17899B (CA)7GG 51.7 11 11 100 1500-300 HB8 (GA)6GG 44 8 6 75 1300-250 HB9 (GT)6GG 44 10 8 80 1200-250 HB10 (GA)6CC 44 8 8 100 1000-350 HB12 (CAC)3GC 38 7 6 85.71 1500-300 UBC827 (AC)8G 52.2 7 5 71.42 1000-400 UBC823 (TC)8C 52.2 10 8 80 1500-400 UBC807 (AG)8T 49.8 9 7 77.77 1500-200
was not certified by peer review
) is the author/funder. All rights reserved. N
o reuse allowed w
ithout permission.
The copyright holder for this preprint (w
hichthis version posted M
ay 25, 2020. .
https://doi.org/10.1101/2020.05.22.109041doi:
bioRxiv preprint
Table 5: Band frequency, genetic diversity and polymorphism information content (PIC) of 433
each ISSR locus. 434
Marker Band Frequency Gene diversity PIC
ISSR 825 0.338 0.203 0.219
ISSR841 0.318 0.142 0.167
ISSR823 0.381 0.142 0.172
ISSRHB-10 0.369 0.232 0.228
ISSRHB-8 0.575 0.231 0.236
ISSR17899A 0.468 0.212 0.224
ISSRUBC-807 0.55 0.228 0.242
ISSR827 0.46 0.236 0.293
ISSR844 0.417 0.293 0.288
ISSRHB-09 0.413 0.274 0.307
ISSR17899B 0.368 0.293 0.274
ISSR HB-12 0.808 0.164 0.196
ISSR809 0.683 0.211 0.258
Mean 0.473±0.145 0.220±0.049 0.238±0.044
435
Table 6. Statistical analyses of genetic diversity level. 436
Population Number of gentoypes
% of polymorphic
loci Nei diversity
Shannon's Information
Index Group1 7 85.59 0.308 0.463
Group2 4 34.23 0.146 0.211
Group3 9 61.26 0.212 0.319
All genotypes 20 60.36 0.222 0.331
437
438
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Table 7. Genetic variance within and among groups based on Analysis of Molecular 439
Variance (AMOVA). 440
Source of variance Degree of freedom
Sum of Square
Mean Square
Estmated Variance
Percentage of total variance
Among group 2 65.008 32.504 2.729 15 Within group 17 257.992 15.176 15.176 85 Total 19 323.000 17.905 100
441
Population structure and genetic relationship among genotypes 442
Using STRUCTURE software [37], the evaluation of the delta k according to the Evanno 443
method showed the highest peak at k = 3 (Fig 5A) and the mean value of the logarithm of 444
likelihood (LnP (D) for k = 1 was lower than that of k = 3 which is the higher peak (Fig 5B). 445
The representation of the ancestry at k = 3, revealed three (3) genetic pools (Fig 5C). Group I 446
included 30% of the individuals while group II and group III encompassed respectively 50 447
and 20% of the genotypes. The genetic relationship revealed by the dendrogram was in 448
agreement with STRUCTURE analysis which clearly distinguished three groups. The first 449
group contained 2 mutants of Melakh (Me51M4-14M5-2M6-M7 and Me51M4-39M5-1M6-450
M7) and 4 other mutants of Yacine (Y1-M4-8M5-1M6-M7, Y1-M4-3M5-2M6-M7, Y1-M4-451
11M5-1M6-M7 and Y7-M4-4M5-3M6-M7). This group encompassed the highest number of 452
admixed individuals. The second genetic pool contained exclusively the mutants of Yacine. 453
Of these, Y1-M4-9M5-2M6-M7 and Y1-M4-9M5-3M6-M7clustering with a high bootstrap 454
value (96%). The admixed Y7-M4-1M5-1M6-M7 clustered with Y7-M4-6M5-1M6-M7 with 455
80% bootstrap value. The third genetic pool contained Yacine, 3 mutants of Yacine (Y13-M4-456
4M5-1M6-M7, Y7-M4-12M5-3M6-M7 and Y7-M4-10M5-1M6-M7) and 5 other mutants of 457
Melakh (Me51M4-10M5-1M6-M7, Me51M4-11M5-2M6-M7, Me51M4-25M5-3M6-M7 and 458
Me51M4-29M5-1M6-M7). The variety Melakh was clustering with its mutant, Me51M4-459
77M5-M7 with a high bootstrap value (92%) (Fig 6A). 460
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
The clustering of the genotypes resulting from the DAPC analysis identified 3 groups which 461
showed any individual genetically admixed (Fig 6B). This results suggested that DAPC 462
analysis was appropriate to assess mutant populations structure by achieving better separation 463
among group as it was also showed by the number of clusters identified (Fig 7A). Group 1 464
and Group3 differed from each other based on the first axis of the DAPC. This classification 465
was based on the loci 84, 37, 57, 105, 35, 10 and 106 which were the most discriminating, in 466
decreasing order (Fig 7B). Group 2 differed from the others on the second axis. These 467
findings were based on the loci 97, 78, 87, 21, 105, 7, 82, 64, 47 and 12 which were the most 468
discriminating, in decreasing order (Fig 7C). 469
Fig 5. Structure of the mutant population. 470
A: Probability of subdivision into genetic groups by the method of Evanno et al. 2005; B: 471
Probability of subdivision into genetic groups by the log mean likelihood method and C: 472
Ancestry for K = 3. 473
Fig 6. Neighbor Joining Dendrogram based on ISSR data representing the grouping of 18 474
cowpea mutant lines and their parent. The dendrogram was constructed using the distance 475
matrix between individuals, calculated using the Jaccard, (1902, 1912) Index. 476
A and B: Dendrograms based on STRUCTURE and DAPC analyses respectively. 477
Fig 7. Discriminant Analysis of Principal Components (DAPC) for 18 cowpea mutant lines 478
and their parent based on ISSR data. 479
A: graphical representation of the groups; B: contribution of the alleles based on the first axis; 480
C: contribution of the alleles based on the second axis 481
482
483
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Discussion 484
Wide genetic variability is a prerequisite for a successful breeding program particularly in the 485
era of climate change with its adverse effects leading to the erosion of the plant genetic 486
resources. Thus, to broaden the crop’s genetic basis, a wide range of techniques have been 487
used in the last decade but among these, induced mutagenesis has been proved to be best for 488
creating novel variation in crop genome and was used in this study to expand variability in 489
cowpea which experienced a single domestication event during the course of evolution. 490
Agromorphological variability analysis of the mutants 491
In this study the percentage of germination of the irradiated seeds decreased compared to the 492
control (Melakh and Yacine) regardless of the dose and the generation used. These results 493
were in agreement with the findings of Melki and Marouani [47], Horn et al. [13] and 494
Olasupo et al. [17] who recorded similar observations during their studies. In contrast, Horn et 495
al. [13] recorded zero germination of cowpea irradiated seeds at 300 Gy, in this study 98.15% 496
of field establishment were observed for Melakh M1 white-seeded at this dose, this value 497
reached 99.08% for Yacine M1 brown-seeded at 340 Gy. These results were in agreements 498
with the findings of Olasupo et al. [17] who suggested that radio-sensitivity is genotype 499
dependent as it was associated with seed characteristics (seed coat color, water content, 500
thickness and weight). Presently, it is well documented that ionizing radiation is injurious of 501
enzymes and growth hormone leading to biochemical and physiological modifications, cell 502
death, abnormal cell division, tissue and organ failure and growth disturbance [48, 49, 50]). 503
We can assume that these type of changes occurred in our irradiated seed materials as 11.11% 504
and 11.5% of the Melakh mutant lines in M5 and M6 respectively showed abnormities of 505
leave numbers, corroborating the discoveries of Girija and Dhanavel [51] and Nair and Mehti 506
[11] performed in cowpea mutants that were bifoliated, tetrafoliated, pentafoliated and 507
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
hexafoliated (S1 Fig). These disturbances would explain the change observed in leaves shape 508
of our mutants with 3 new forms e.i. globular, subglobular and hastate (Fig 1). In contrast, the 509
investigations performed by Gnanamurthy et al. [16] led to the discovery of only a globular 510
form within their mutant populations. Similar modifications induced early flowering in some 511
Melakh and Yacine mutant lines which is an important agronomic trait for farmers and 512
breeders particularly in the Sahel zone where the rainy season has become shorter and shorter. 513
In addition, mutagenesis treatment affected flower coloration which changed from white to 514
white with purple border in Melakh and in Yacine mutant population it changed from white to 515
white with purple border, purple or dark purple with a high heritable value (h2 = 78.01%) (Fig 516
1, Table 3). These findings were in accordance to the observations of Horn et al. [13] and 517
Girija et al. [52] who also noticed flower color variation in their cowpea mutant populations. 518
In this study, similar seed coat color variation was observed in the mutants of Yacine, first 519
time during the M4 generation where one single plant produced white seeds, brown seeds and 520
white-browned seed (Fig 1J) but from M5 brown seed and white were only recorded as it was 521
previously reported that seed coat color is an important trait for consumer preference 522
depending on the regions [53]. On the other hand, the seed coat color of Melakh mutants 523
remained unchanged e.g white as the control regardless of the generation. These results 524
explained the high heritability value (h2 = 94.48%; Table 3) recorded for seed coat color in 525
our populations. In accordance with this study, mutants seed coat color variation was also 526
recorded by Gaafar et al. [12] and Horn et al. [13] suggesting that mutation affected the 527
candidate genes involved in the control of late stages in flavonoid biosynthesis pathway 528
namely the basic helix–loop–helix gene for the C locus, the WD-repeat gene for the W locus 529
and the E3 ubiquitin ligase gene for the H locus [54]. 530
During this study, yield component (pod length, pod width, number of seeds per plant, seed 531
length and seeds weight) variation with a high heritability values was noticed among the 532
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
mutant genotypes compared to the control. Similar results were recorded by Gnanamurthy et 533
al. [16] Goyal and Khan [55] and Horn et al. [13] which suggested that mutagenesis can be 534
used to improve crop yield, one of the most important agronomic characteristics for breeders. 535
High heritability, genetic advance and genetic coefficient of variation values of pod length 536
and number of seeds per plant recorded in this study suggest that these traits can be 537
considered as attributes for the improvement through selection of the mutants. In addition, 538
regardless of the generation (M5 to M7) and the environment, the pod length and the number 539
of seeds per plant were positively and significantly correlated as did seed length and seed 540
weight during this study. In contrast, the variation of the correlation values between traits 541
recorded might be due to the effect of interaction genotype x environment which could induce 542
epigenetic modifications impacting gene expression or gene segregation over generations. 543
The genetic relationship based on the agromorphological characters revealed that the mutant 544
populations were subdivided into 3 groups (Fig 4). The first group included genotypes which 545
notably deviated from their respective parents, thereby suggesting that the irradiation doses 546
used were efficient enough to induce heterogeneous population as any polytony was observed 547
in the dendrogram. In the remaining two others groups, most of the mutants were clustering 548
with their respective control revealing their high phenological diversity as reported by Laskar 549
and Khan [56]. Taking together, these findings meet the recommendation of Rohman et al. 550
[57] who suggested that the cluster contributing to the greatest divergence can help to choose 551
the parent in a breeding program. Based on these, PCA results (Fig 3) revealed in this study 552
that yield related traits such as pod length, number of seeds per plant and seed weight are 553
major contributor to the genetic divergence which was in accordance with the report of 554
Afuape et al. [58] who suggested that PCA is accurate to select the best genotypes for future 555
breeding progam. 556
557
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Genetic diversity of mutants based on molecular markers 558
In the present study, the gene diversity measures through the average heterozygosity and the 559
genetic distance among individuals in a population and the PIC values which is a good 560
indicator of the usefulness of a marker to determine its inheritance between the offspring and 561
the parent were estimated in our mutants. Our results showed that, in general, the gene 562
diversity and the PIC were close (Fig 5), suggesting the evenness of allele’s frequencies in the 563
mutant populations as reported by Shete et al. [59]. According to Botstein et al. [60], a marker 564
is considered highly informative if the PIC is ≥ 0.50, moderately informative with a PIC 565
values ranging from 0.25 to 0.5 and slightly informative at less than 0.25. Based on this, the 566
ISSR827, ISSR844, ISSRHB9, ISSR17899B and ISSR809 were moderately informative, the 567
remaining markers were slightly informative which were higher than the scores recorded in 568
the cowpea mutant population developed by Gaafar et al. [12]. These differences could be 569
attributed to the low irradiation dose (50 Gy) used which might induce less mutations and less 570
variability compared to the 300 Gy and 340 Gy employed in this study. In addition, Gaafar 571
and coworkers [12] used ethidium bromide for DNA staining which is less sensitive than 572
GelRed, the one we used in our study leading to more DNA bands scoring. In contrast, the 573
PIC scores reported in the natural populations of Vigna [61], mungbean, blackgram [62] and 574
in Vigna unguiculata [21] were a bit higher probably due to several mutations undergone by 575
these genomes during evolution. Our results show that ISSR are accurate markers to 576
discriminate cowpea mutants for new genotypes identification and selection. In this study, the 577
ISSR primers UBC825 (AC repeat motif), UBC844 (CT repeat motif), 17899A and 17899B 578
(CA repeat motif) and HB12 (GA repeat motif) gave 100% polymorphic bands which 579
suggested that these motifs are abundant in the genome of cowpea. Similar observations have 580
been reported in Arachis hypogea [63,64] and in Vigna mungo [65]. 581
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Population structure and genetic relationship of the mutants based on ISSR 582
markers 583
Analyzing population structure in mutants is relevant to understand the organization of the 584
genetic variation which is driven by the combined effect of recombination, mutation, 585
demographic history and natural selection. Based on this, and due to the informative nature of 586
the ISSR markers used, the generated data subjected to STRUCTURE analysis showed 3 587
subpopulations (optimal K = 3). These results were consistent with the organization of the 588
dendrogram (Fig 6A). The clustering of the Me51M4-39M5-1M6-M7 and Y1-M4-8M5-1M6-589
M7, two genotypes tolerant to the nematode Meloidogyne incognita, compared to their 590
parents, according to our preliminary studies, might suggest that this character is distributed in 591
this group (group 1). In addition, this group encompassed genotypes with long pod length 592
compared to their respective parents. These findings are relevant to select the best mutant 593
genotypes which can be proposed for variety registration and popularization but also for gene 594
discoveries and breeding programs. These results demonstrate the ability of gamma rays to 595
induce large genetic changes in DNA material as among the 18 mutants genotyped 7 derived 596
from 1 Melakh mutant and 11 from 1 Yacine mutant. The findings were in accordance with 597
previous studies recently performed on several crops such as cowpea [12,13,66] and chickpea 598
[67]. The dendrogram analysis revealed a group encompassing only Yacine mutants. In this 599
group 2, the Y7-M4-4M5-3M6-M7 genotype could be included as it shares several 600
morphological characters such as the erected shape, except Y7-M4-1M5-1M6-M7 which is 601
prostate, white flower with purple border except the mutant Y1-M4-9M5-2M6-M7 showing 602
light purple flower, brown coat seed, except Y7-M4-1M5-1M6-M7 with white coat seed, Y7-603
M4-6M5-1M6-M7 a white-brown seeded genotype as is Y1-M4-9M5-3M6-M7. Thus, ISSR 604
are accurate markers to discriminate these genetic lines. In contrast, the group 3 is more 605
heterogeneous including Yacine, its mutants and its progenitor Melakh and its mutants. Those 606
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
results suggested that the assignation of Melakh to the group could be an artefact. The 607
clustering of Yacine and its relative Melakh in the same group 1 demonstrate the 608
discriminating power of the DNA based molecular markers. 609
Genetic differentiation within and between groups 610
In the present study, STRUCTURE analysis and the dendrogram results showed 3 groups 611
which aroused our curiosity to understand the variability within and between clusters by 612
assessing the genetic parameters such as Nei’s genetic diversity and Shannon’s information 613
index which are important to measure the degree of genetic diversity among and within group 614
in a population. Our results showed that group 1 had a high Nei’s genetic diversity and 615
Shannon’s information index and a high number (17) of private alleles unlike group 2 which 616
had the lowest diversity, suggesting that the gamma irradiation doses were efficient to induce 617
new alleles in the mutants (Table 6). Identification of private alleles are important in plant 618
breeding and conservation as their presence in a single population might be linked to specific 619
agronomic traits usable for new genotypes selection in a mutant population. In accordance 620
with this, the AMOVA results revealed that a large majority of the total variation (85%) was 621
noticed within group variation suggesting a high level of differentiation while only 15% of the 622
variation were recorded between group (Table 7). According to Seyoum et al. [68], small 623
variation between groups might be an advantage due to its usefulness to study marker-traits 624
association. In contrast, a large variation between groups could reduce the possibility to detect 625
the effects of single genes in a genome wide association study [69]. Based on these and taking 626
into account the important agronomic traits recorded in our mutant populations a genome 627
wide association can be performed in order to detect the molecular markers associated with 628
these characters and usable in a molecular breeding program. 629
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
To analyze population structure, different methods such as STRUCTURE, Principal 630
Coordinate Analysis and DAPC can be used. The latter provides complementary information 631
leading to better assignments of individuals in the accurate group and its advantage is related 632
to the fact that the target population has not to be in Hardy-Weinberg equilibrium. In this 633
study, DAPC results divided the mutant populations in well-defined 3 groups with less 634
admixture compared to STRUCTURE results. Indeed, the ancestry value recorded in 635
STRUCURE analysis allowing the assignment Me51M4-14M5-2M6-M7 and Melakh to 636
group 1 was less than 70% in contrast, this abnormality was resolved using DAPC as it did 637
with Y7-M4-1M5-1M6-M7 and Yacine to group 2 and 3 respectively. These results were in 638
accordance with previous studies carried out in potato [70], in landraces and bred cultivars of 639
Prunus avium [71] and Ginseng germplasm [72] which revealed that DAPC gave a better 640
grouping resolution than STRUCTURE. In addition, DAPC analysis showed that the loci such 641
as 35 and 37 (amplified by ISSR primer HB10), 57 (amplified by ISSR primer 807) greatly 642
contributed for the discrimination of group 1 and group 3. In contrast loci 97, 78, 21 and 105 643
amplified respectively by 17899B, ISSR844, ISSR 841 and ISSRHB12 was also involved in 644
the differentiation of group 2 from the others (Fig 7B). This latter ISSR primer was suspected 645
by Gaafar et al. [12] to be usable in a marker assisted selection for high yield genotypes. 646
These results suggested that DAPC is an appropriate method to analyze the organization of 647
the genetic variation within and between mutant population in order to identify the best 648
genotype as shown in our study, the individuals belonging to the same group shared several 649
agromorphological characters. For instance, group 1 (Fig 6B) included white-seeded 650
genotypes and long pod length unlike group 3 which showed short pod length. 651
Conclusion 652
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
In this study, it was obviously noticed that gamma irradiation is a powerful tool to induce 653
genetic variation in cowpea to broaden its genetic basis in order to overcome the bottleneck 654
effect undergone by this crop during its domestication. This assertion has been supported by 655
the various agromorphological variations affecting plant growth habit, flower color, pod 656
length and weight, seed color, size and weight noticed among our mutant populations. These 657
phenotypic variations were explained by the genetic variation revealed by 13 ISSR markers 658
and the use of the datasets generated by these molecular markers enhanced our understanding 659
on the mutant population structure. Overall, efficient exploitation of both agromorphological 660
and molecular data led to the identification of promising high yielding new mutant genotypes 661
which can be proposed for multi trial tests to assess their performance and stability in 662
different agroecological conditions in Senegal and abroad but also their nutritional value. 663
These mutant genotypes are valuable genetic resources usable for breeding programs but also 664
for gene discoveries such as the ones involved in growth habit, pod length or seed size in 665
cowpea. 666
Acknowledgments 667
We thank the IAEA for funding this project. We thank also Thuloane Bernard TSEHLO and 668
Dr Fatma SARSU respectively Programme Management Officer and Technical Officer for 669
this project at the IAEA for their advices and support. I would also like to thank the 670
anonymous reviewers for their valuable comments. 671
672
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
Author Contributions 673
MD, FAB and DD designed the study. MD and SD performed DNA extraction and 674
genotyping. MD and SD performed data analyses, OD run STRUCTURE and DAPC 675
analyses. MD, SD, FAB and DD drafted the manuscript. All authors contributed to the final 676
version. 677
References 678
1. Diouf D. Recent Advances in Cowpea [ Vigna Unguiculata (L.) Walp.] “Omics” 679
Research for Genetic Improvement. Afr. J. Biotechnol. 2011; 10 (15):2803–2810. 680
doi:10.5897/AJBx10.015. 681
2. Boukar O, Massawe F, Muranaka S, Franco J, Maziya-Dixon B, Singh B Fatokun C. 682
Evaluation of Cowpea Germplasm Lines for Protein and Mineral Concentrations in 683
Grains. Plant Genet. Resour. 2011; 9 (4):515–522. doi :10.1017/s1479262111000815 684
3. Quin FM. “Introduction,” in Advances in Cowpea Research, Eds B. B. Singh, D. R. 685
Mohan-Raj, K. E. Dashiell, and L. E. N. Jackai (Ibadan: Co-Publication of International 686
Institute of Tropical Agriculture (IITA),Japan International Research Center for 687
Agricultural Sciences (JIRCAS)). 1995, pp ix–xv. 688
4. Boukar O, Fatokun CA, Huynh BL, Roberts PA, Close TJ. Genomic Tools in Cowpea 689
Breeding Programs: Status and Perspectives. Front. Plant Sci. 2016; 7. 690
doi:10.3389/fpls.2016.00757. 691
5. Badiane FA, Diouf M, Diouf D. Cowpea. In Broadening the Genetic Base of grain 692
legumes. Springer. 2014; 95–114. doi :10.1007/978-81-322-2023-7_5 693
6. Li C-D, Fatokun CA, Ubi B, Singh BB, Scoles GJ. Determining Genetic Similarities 694
and Relationships among Cowpea Breeding Lines and Cultivars by Microsatellite 695
Markers. Crop Sci. 2001; 41 (1):189–197. doi. 10.2135/cropsci2001.411189x. 696
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
7. Morita R, Kusaba M, Iida S, Yamaguchi H, Nishio T, Nishimura M. Molecular 697
Characterization of Mutations Induced by Gamma Irradiation in Rice. Genes Genet. 698
Syst. 2009; 84 (5):361–370. doi :10.1266/ggs.84.361 699
8. Naito K, Kusaba M, Shikazono N, Takano T, Tanaka A, Tanisaka T, Nishimura M. 700
Transmissible and Nontransmissible Mutations Induced by Irradiating Arabidopsis 701
Thaliana Pollen with γ-Rays and Carbon Ions. Genetics. 2005;169 (2): 881–889. 702
doi :10.1534/genetics.104.033654 PMID: 15371348 703
9. Adekola OF,Oluleye F. Induced Tolerance of Cowpea Mutants to Maruca Vitrata 704
(Fabricius) (Lepidoptera: Pyralidae). Afr J Biotechnol. 2008, 7 (7), 878–883. 705
10. Badr A, Ahmed HIS, Hamouda M, Halawa M, Elhiti MA. Variation in Growth, Yield 706
and Molecular Genetic Diversity of M2 Plants of Cowpea Following Exposure to 707
Gamma Radiation. Life Sci. J. 2014; 11 (8):10–19. 708
11. Nair R, Mehta AK. Induced Mutagenesis in Cowpea [Vigna Unguiculata (L.) Walp] 709
Var. Arka Garima. Indian J. Agric. Res. 2014; 48 (4):247–257. doi :10.5958/0976-710
058x.2014.00658.1 711
12. Gaafar RM, Hamouda M, Badr A. Seed Coat Color, Weight and Eye Pattern 712
Inheritance in Gamma-Rays Induced Cowpea M2-Mutant Line. J. Genet. Eng. 713
Biotechnol. 2016; 14 (1):61–68. doi:10.1016/j.jgeb.2015.12.005. 714
13. Horn LN, Ghebrehiwot HM, Shimelis HA. Selection of Novel Cowpea Genotypes 715
Derived through Gamma Irradiation. Front. Plant Sci. 2016; 7. 716
doi:10.3389/fpls.2016.00262. 717
14. Olasupo FO, Ilori CO, Forster BP, Bado S. Selection for Novel Mutations Induced by 718
Gamma Irradiation in Cowpea (Vigna Unguiculata [L] Walp.). Int. J. Plant Breed. 719
Genet. 2018; 12:1–12. doi :10.3923/ijpbg.2018.1.12 720
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
15. Adekola OF, Oluleye F. Induction of Genetic Variation in Cowpea (Vigna Unguiculata 721
L. Walp.) by Gamma Irradiation. Asian J. Plant Sci. 2007; 6 (5):869–873. 722
doi :10.3923/ajps.2007.869.873 723
16. Gnanamurthy S, Mariyammal S, Dhanavel D, Bharathi T. Effect of Gamma Rays on 724
Yield and Yield Components Characters R3 Generation in Cowpea (Vigna Unguiculata 725
(L.) Walp). Int J Res Plant Sci 2012; 2:39–42. 726
17. Olasupo FO, Ilori CO, ForsterBP, Bado S. Mutagenic Effects of Gamma Radiation on 727
Eight Accessions of Cowpea (Vigna Unguiculata [L.] Walp.). Am. J. Plant Sci. 2016; 7 728
(2):339–351. doi :10.4236/ajps.2016.72034 729
18 Reddy MP, Sarla N, Siddiq EA. Inter Simple Sequence Repeat (ISSR) Polymorphism 730
and Its Application in Plant Breeding. Euphytica 2002; 128 (1):9–17. 731
doi:10.1023/a:1020691618797 732
19. Joshi P, Dhawan V. Assessment of Genetic Fidelity of Micropropagated Swertia 733
Chirayita Plantlets by ISSR Marker Assay. Biol. Plant. 2007; 51 (1):22–26. 734
doi:10.1007/s10535-007-0005-0 735
20 Nova MV, Leite NGA, Houllou LM, Medeiros LV, Neto AL, Hsie BS,Borges-Paluch 736
LR, Santos BS, Araujo CSF, Rocha AA. Genetic Variability and Resistance of 737
Cultivars of Cowpea [Vigna Unguiculata (L.) Walp] to Cowpea Weevil 738
(Callosobruchus Maculatus Fabr.). Genet. Mol. Res. 2014; 13 (1):2323–2332. 739
doi :10.4238/2014.march.31.12 740
21. Igwe DO, Afiukwa CA, Ubi BE, Ogbu KI, Ojuederie OB, Ude GN. Assessment of 741
Genetic Diversity in Vigna Unguiculata L.(Walp) Accessions Using Inter-Simple 742
Sequence Repeat (ISSR) and Start Codon Targeted (SCoT) Polymorphic Markers. 743
BMC Genet. 2017; 18 (1):98. doi: 10.1186/s12863-017-0567-6 PMID: 29149837 744
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
22. Blair MW, Panaud O, McCouch SR. Inter-Simple Sequence Repeat (ISSR) 745
Amplification for Analysis of Microsatellite Motif Frequency and Fingerprinting in 746
Rice (Oryza Sativa L.). Theor. Appl. Genet. 1999; 98 (5):780–792. 747
doi :10.1007/s001220051135 748
23. Badiane FA, Gowda BS, Cissé N, Diouf D, Sadio O; Timko MP. Genetic Relationship 749
of Cowpea (Vigna Unguiculata) Varieties from Senegal Based on SSR Markers. Genet. 750
Mol. Res. GMR. 2012; 11 (1):292–304. doi:10.4238/2012.February.8.4. 751
24. Cisse N, Ndiaye M, Thiaw S, Hall AE. Registration of’Melakh’cowpea. Crop Sci. 752
1997; 37 (6). doi :10.2135/cropsci1997.0011183x003700060054x 753
25. Cisse N, Ndiaye M, Sene A. Registration of’Yacine’cowpea. Crop Sci. 2005; 45 (1):13. 754
doi:10.2135/cropsci2005.0413a 755
26. Schum A. Mutation Breeding in Ornamentals: An Efficient Breeding Method? In XXI 756
International Eucarpia Symposium on Classical versus Molecular Breeding of 757
Ornamentals-Part I 612; 2003; 47–60. doi:10.17660/actahortic.2003.612.6 758
27. Shin JM, Kim BK, Seo SG, Jeon SB, Kim JS, Jun B, Kang SY, Lee JS, Chung MN, 759
Kim, SH. Mutation Breeding of Sweet Potato by Gamma-Ray Radiation. Afr. J. Agric. 760
Res. 2011; 6 (6):1447–1454. 761
28. IBPGR. Descriptors for Cowpea. 1983. 762
29. Singh RK, Chaudhary BD. Biometrical Methods in Quantitative Genetics Analysis. 763
Kalyani Publishers. 1985. doi :10.2307/2530404 764
30 Sivasubramanian SS, MadhavaMenon P. Genotypic and Phenotypic Variability in Rice. 765
Madras Agric. J. 1973; 1093–1096. 766
31. Robinson HF, Comstock RE, Harvey PH. Estimates of Heritability and the Degree of 767
Dominance in Corn1. Agron. J. 1949; 41 (8):353–359. 768
doi:10.2134/agronj1949.00021962004100080005x. 769
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
32. Assefa K, Ketema S, Tefera H, Nguyen HT, Blum A, Ayele M, Bai G, Simane B, 770
Kefyalew T. Diversity among Germplasm Lines of the Ethiopian Cereal Tef [Eragrostis 771
Tef (Zucc.) Trotter]. Euphytica 1999; 106 (1):87–97. 772
33. Johnson HW, Robinson HF, Comstock RE. Genotypic and Phenotypic Correlations in 773
Soybeans and Their Implications in Selection. Agron. J. 1955; 47 (10):477–483. 774
doi :10.2134/agronj1955.00021962004700100008x 775
34. Sneath PS, Soka L. Numerical Taxonomy. RR 1973 WH Freeman, San Francisco, CA. 776
35. Peakall R, Smouse PE. GenAlEx 6.5: Genetic Analysis in Excel. Population Genetic 777
Software for Teaching and Research—an Update. Bioinformatics. 2012; 28 (19):2537–778
2539. doi:10.1093/bioinformatics/bts460. 779
36. Liu K, Muse SV. PowerMarker: An Integrated Analysis Environment for Genetic 780
Marker Analysis. Bioinformatics 2005; 21 (9):2128–2129. 781
doi:10.1093/bioinformatics/bti282. 782
37. Pritchard JK, Stephens M, Donnelly P. Inference of Population Structure Using 783
Multilocus Genotype Data. Genetics. 2000; 155 (2):945–959. 784
38. Falush D, Stephens M, Pritchard JK. Inference of Population Structure Using 785
Multilocus Genotype Data: Linked Loci and Correlated Allele Frequencies. Genetics. 786
2003; 164 (4):1567–1587. 787
39. Earl D, Vonholdt B. Structure Harvester: A Website and Program for Visualizing 788
STRUCTURE Output and Implementing the Evanno Method. 2012. 789
doi :10.1007/s12686-011-9548-7 790
40. Evanno G, Regnaut S, Goudet J. Detecting the Number of Clusters of Individuals 791
Using the Software STRUCTURE: A Simulation Study. Mol. Ecol. 2005; 14 (8):2611–792
2620. doi :10.1111/j.1365-294x.2005.02553.x 793
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
41. Jombart T. Adegenet: A R Package for the Multivariate Analysis of Genetic Markers. 794
Bioinformatics. 2008; 24 (11):1403–1405. doi:10.1093/bioinformatics/btn129. 795
42. R Development Core Team. R: A Language and Environment for Statistical 796
Computing. 2008; R Foundation for Statistical Computing, Vienna, Austria. 797
43. Saitou N, Nei M. The Neighbor-Joining Method: A New Method for Reconstructing 798
Phylogenetic Trees. Mol. Biol. Evol. 1987; 4 (4):406–425. 799
doi:10.1093/oxfordjournals.molbev.a040454 800
44. Jaccard, P. Lois de Distribution Florale Dans La Zone Alpine. Bull Soc Vaudoise Sci 801
Nat. 1902; 38:69–130. 802
45. Jaccard P. The Distribution of the Flora in the Alpine Zone. New Phytol. 1912; 11 803
(2):37–50. doi :10.1111/j.1469-8137.1912.tb05611.x 804
46. Perrier X, Jacquemoud-Collet J. DARwin Software. 2006. Http. Darwin Cirad 805
Frdarwin 2015. 806
47 Melki M, Marouani A. Effects of Gamma Rays Irradiation on Seed Germination and 807
Growth of Hard Wheat. Environ. Chem. Lett. 2009; 8 (4):307–310. 808
doi:10.1007/s10311-009-0222-1. 809
48. Lagoda PJL. Effects of Radiation on Living Cells and Plants. Plant Mutat. Breed. 810
Biotechnol. 2012; 123–134. doi:10.1079/9781780640853.0123 811
49. Mudibu J, Nkongolo KKC, Kalonji-Mbuyi A, Kizungu RV. Effect of Gamma 812
Irradiation on Morpho-Agronomic Characteristics of Soybeans (Glycine Max L.). Am. 813
J. Plant Sci. 2012; 3 (3):720–726. doi:10.4236/ajps.2012.33039. 814
50 Ambavane AR, Sawardekar SV, Sawantdesai SA, Gokhale NB. Studies on Mutagenic 815
Effectiveness and Efficiency of Gamma Rays and Its Effect on Quantitative Traits in 816
Finger Millet (Eleusine Coracana L. Gaertn). J. Radiat. Res. Appl. Sci. 2015; 8 817
(1):120–125. doi:10.1016/j.jrras.2014.12.004. 818
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
51. Girija M, Dhanavel D. Mutagenic Effectiveness and Efficiency of Gamma Rays Ethyl 819
Methane Sulphonate and Their Combined Treatments in Cowpea (Vigna Unguiculata 820
L. Walp). Glob. J Mol Sci 2009; 4. 821
52. Girija M, Dhanavel D, Gnanamurthy S. Gamma Rays and EMS Induced Flower Color 822
and Seed Mutants in Cowpea (Vigna Unguiculata L. Walp). Adv Appl Sci Res 2013; 4 823
(2):134–139. 824
53. Jaeger SR, Antúnez L, Ares G, Swaney-Stueve M, Jin D, Harker FR. Quality 825
Perceptions Regarding External Appearance of Apples: Insights from Experts and 826
Consumers in Four Countries. Postharvest Biol. Technol. 2018; 146:99–107. 827
doi :10.1016/j.postharvbio.2018.08.014 828
54. Herniter IA, Lo R, Muñoz-Amatriaín M, Lo S, Guo YN, Huynh BL, Lucas M, Jia Z, 829
Roberts PA, Lonardi S, Close TJ. Seed Coat Pattern QTL and Development in Cowpea 830
(Vigna Unguiculata [L.] Walp.). Front. Plant Sci. 2019; 10. 831
doi:10.3389/fpls.2019.01346. 832
55. Goyal S, Khan S. Induced Mutagenesis in Urdbean (Vigna Mungo L. Hepper): A 833
Review. Int. J. Bot. 2010; 6 (3):194–206. doi:10.3923/ijb.2010.194.206 834
56. Laskar RA, Khan S. Assessment on Induced Genetic Variability and Divergence in the 835
Mutagenized Lentil Populations of Microsperma and Macrosperma Cultivars 836
Developed Using Physical and Chemical Mutagenesis. PloS One 2017; 12 837
(9):e0184598. doi:10.1371/journal.pone.0184598. 838
57. Rohman MM, Hakim MA, Sultana NA, Kabir ME, Hasanuzzan M, Ali M. Genetic 839
Divergence Analysis in Sorghum (Sorghum Bicolor L.). Asian J Plant Sci 2004; 3 840
(2):211–214. doi:10.3923/ajps.2004.211.214 841
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
58. Afuape SO, Okocha PI, Njoku D. Multivariate Assessment of the Agromorphological 842
Variability and Yield Components among Sweet Potato (Ipomoea Batatas (L.) Lam) 843
Landraces. Afr. J. Plant Sci. 2011; 5 (2):123–132. 844
59. Shete S, Tiwari H, Elston RC. On Estimating the Heterozygosity and Polymorphism 845
Information Content Value. Theor. Popul. Biol. 2000; 57 (3):265–271. 846
doi:10.1006/tpbi.2000.1452 847
60. Botstein D, White RL, Skolnick M, Davis RW. Construction of a Genetic Linkage Map 848
in Man Using Restriction Fragment Length Polymorphisms. Am. J. Hum. Genet. 1980; 849
32 (3):314. 850
61. Ajibade SR, Weeden NF, Chite SM. Inter Simple Sequence Repeat Analysis of Genetic 851
Relationships in the Genus Vigna. Euphytica 2000; 111 (1):47–55. 852
doi:10.1006/tpbi.2000.1452 853
62. Tantasawat P, Trongchuen J, Prajongjai T, Thongpae T, Petkhum C, Seehalak W, 854
Machikowa T. Variety Identification and Genetic Relationships of Mungbean and 855
Blackgram in Thailand Based on Morphological Characters and ISSR Analysis. Afr. J. 856
Biotechnol. 2010; 9 (27):4452–4464. doi:10.1016/j.scienta.2009.12.033 857
63. Ferguson ME, Burow MD, Schulze SR, Bramel PJ, Paterson AH, Kresovich S, 858
Mitchell S. Microsatellite Identification and Characterization in Peanut (A. Hypogaea 859
L.). Theor. Appl. Genet. 2004; 108 (6):1064–1070. doi:10.1007/s00122-003-1535-2 860
64. Mondal S, Sutar SR, Badigannavar AM. Assessment of Genetic Diversity in Cultivated 861
Groundnut (Arachis Hypogaea L.) with Differential Responses to Rust and Late Leaf 862
Spot Using ISSR Markers. Indian J. Genet. Plant Breeding 2009; 69 (3):219–224. 863
65. Souframanien J, Gopalakrishna T. A Comparative Analysis of Genetic Diversity in 864
Blackgram Genotypes Using RAPD and ISSR Markers. Theor. Appl. Genet. 2004; 109 865
(8):1687–1693. doi:10.1007/s00122-004-1797-3 866
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
66. Ezzat A, Adly M, El-Fiki A. Morphological, Agronomical and Molecular 867
Characterization in Irradiated Cowpea (Vignaunguiculata (L.) Walp.) and Detection by 868
Start Codon Target Markers. J. Radiat. Res. Appl. Sci. 2019; 12 (1):403–412. 869
doi:10.1080/16878507.2019.1686578 870
67. Amri-Tiliouine W, Laouar M, Abdelguerfi A, Jankowicz-Cieslak J, Jankuloski L, Till 871
BJ. Genetic Variability Induced by Gamma Rays and Preliminary Results of Low-Cost 872
TILLING on M2 Generation of Chickpea (Cicer Arietinum L.). Front. Plant Sci. 2018; 873
9. doi:10.3389/fpls.2018.01568 874
68. Seyoum M, Du XM, He SP, Jia YH, Pan Z, Sun JL. Analysis of Genetic Diversity and 875
Population Structure in Upland Cotton (Gossypium Hirsutum L.) Germplasm Using 876
Simple Sequence Repeats. J. Genet. 2018; 97 (2):513–522. doi:10.1007/s12041-018-877
0943-7. 878
69. Flint-Garcia SA, Thuillet AC, Yu J, Pressoir G, Romero SM, Mitchell SE, Doebley J, 879
Kresovich S, Goodman MM, Buckler ES. Maize Association Population: A High-880
Resolution Platform for Quantitative Trait Locus Dissection. Plant J. Cell Mol. Biol. 881
2005; 44 (6):1054–1064. doi:10.1111/j.1365-313X.2005.02591.x. 882
70. Deperi SI, Tagliotti ME, Bedogni MC, Manrique-Carpintero NC, Coombs J, Zhang R, 883
Douches D, Huarte MA. Discriminant Analysis of Principal Components and Pedigree 884
Assessment of Genetic Diversity and Population Structure in a Tetraploid Potato Panel 885
Using SNPs. PloS One 2018; 13 (3). doi:10.1371/journal.pone.0194398 886
71. Campoy JA, Lerigoleur-Balsemin E, Christmann H, Beauvieux R, Girollet N, Quero-887
García J, Dirlewanger E, Barreneche T. Genetic Diversity, Linkage Disequilibrium, 888
Population Structure and Construction of a Core Collection of Prunus Avium L. 889
Landraces and Bred Cultivars. BMC Plant Biol. 2016; 16 (1):49. doi:10.1186/s12870-890
016-0712-9 891
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
72. Lee KJ, Lee JR, Sebastin R, Cho GT, Hyun DY. Molecular Genetic Diversity and 892
Population Structure of Ginseng Germplasm in RDA-Genebank: Implications for 893
Breeding and Conservation. Agronomy 2020; 10 (1):68. 894
doi:10.3390/agronomy10010068 895
896
Supporting information 897
S1 Table. Estimation of Pearson’s correlation between the agromorphological 898
characters in the M6 of the mutant lines. 899
S2 Table. Estimation of Pearson’s correlation between the agromorphological 900
characters in the M5 of the mutant lines. 901
S1 Fig. First Leave abbnormalities observed among the Melakh mutants. 902
A: Two opposite leaves observed during the juvenile stage; B: Three leaves observed on the 903
mutant 904
C and D: Leaves observed during adult stage on the mutant (tetraleaflet) and the control 905
respectively 906
S2 Fig. Variation of the flowering date after sowing (DAS) in the populations. 907
A: Melakh and its mutants; B: Yacine and its mutants 908
909
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 25, 2020. . https://doi.org/10.1101/2020.05.22.109041doi: bioRxiv preprint