University of Ghana http://ugspace.ug.edu.gh
BREEDING SORGHUM [Sorghum bicolor (L.) MOENCH] FOR HIGH QUALITY 
STOVER FOR NIGER 
 
 
 
By 
 
 
OUSMANE SEYNI DIAKITE 
(10512763) 
 
 
THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON, IN PARTIAL 
FULFILMENT OF THE REQUIREMENTS FOR THE AWARD OF DOCTOR OF 
PHILOSOPHY DEGREE IN PLANT BREEDING 
 
 
 
WEST AFRICA CENTRE FOR CROP IMPROVEMENT 
COLLEGE OF BASIC AND APPLIED SCIENCES 
UNIVERSITY OF GHANA 
LEGON 
 
 
DECEMBER, 2018 
 
 
University of Ghana http://ugspace.ug.edu.gh
DECLARATION 
I hereby declare that except for references to works of other researchers, which have been duly 
cited, this work is my original research and that neither part nor whole has been presented 
elsewhere for the award of a degree. 
 
.................................................. 
Ousmane Seyni Diakité 
            Student 
                                                                           .................................................. 
                                                                                 Prof. Pangirayi Tongoona 
  Supervisor 
 
                                                                           .................................................. 
                                                                                 Prof. Eric Y. Danquah 
  Supervisor 
                                                                           .................................................. 
                                                                                    Prof. Niaba Teme 
                                                                                     Supervisor 
 
                                                                           .................................................. 
                                                                              Dr. Daniel K. Dzidzienyo 
  Supervisor 
 
i 
 
University of Ghana http://ugspace.ug.edu.gh
ABSTRACT 
Pasture shortage during March to July is the main constraint in cattle productivity in Niger. To 
alleviate this pasture shortage, farmers use dual purpose sorghum varieties where the grain is used 
as food and the nutritionally low quality stover as feed for their cattle during dry season. 
Consequently, cattle productivity is reduced. One sustainable means is to pave the way for better 
stover quality availability through incorporation of the brown midrib trait, which is known to 
improve quality of stover, into sorghum varieties. The main objective of this study was to develop 
dual purpose sorghum lines with high quality stover for sustained livestock productivity. The 
specific objectives were to (1) identify farmers’ and stover traders’ preferences on sorghum stover 
varieties; (2) introgress brown midrib6 and brown midrib12 genes in two Nigerien elite sorghum 
lines; (3) determine the agronomic and nutritive potential of new brown midrib (bmr) derived lines 
and (4) identify Single Nucleotide Polymorphism (SNP) mutations in Ethly methanesulfonate 
(EMS) sorghum mutants used in the national sorghum improvement programme. A Participatory 
Rural Appraisal (PRA) consisting of focus group discussions followed by semi-structured 
interviews, was conducted in three agro-ecological areas in Niger. Two bmr genes (bmr6 and 
bmr12) from three donor parents were introgressed into two Nigerien recurrent parents using hand 
emasculation in Sotuba, Mali. Ninety-four (94) derived BC1F3 families plus 6 checks were 
phenotyped using alpha lattice design in two sites (Tillabery and Konni) in Niger in 2017 rainy 
season. Selected derived families’ chemical compositions were analyzed at Sotuba Animal 
Nutrition Laboratory (LNA) in 2018 in Mali. Ten (10) EMS mutated accessions out of 554 
phenotyped in Niger, were sequenced for candidate SNP discovery in Purdue University in 2018. 
PRA results revealed that farmers cultivated sorghum for dual purpose (grain and stover) with high 
preference for sorghum stover compared to pearl millet for cattle feeding. Farmers complained of 
ii 
 
University of Ghana http://ugspace.ug.edu.gh
feed shortage during dry season coupled with poor quality of their millet and sorghum stovers. 
Stover trading was a growing business despite the traditional poor management practices. Main 
criteria for forage quality traits were higher biomass, juiciness, stay-green for both farmers and 
traders. There were significant to highly significant differences among bmr derived families for 
grain and dry matter yield potential within and across sites. Grain yield ranged from 3271.2 kg/ha 
to 1205.4 kg/ha while dry matter yield varied from 6909.8 kg/ha to 2867.0 kg/ha. Significant 
variations were observed for the dry matter, Neutral Detergent Fiber (NDF), Acid Detergent Fiber 
(ADF) and Acid Detergent Lignin (ADL) contents between parental and derived lines. Results 
from 10 EMS lines revealed bmr2, bmr6, bmr12 and stacked bmr2/bmr6 genotypes. No candidate 
SNP was found for 2 EMS mutants for the known bmr genes suggesting that these could represent 
new mutations. Further investigations may identify the nature of their candidate mutations. Deeper 
phenotyping on field and nutritional values are essential to detect higher yielding lines to confirm 
the already interesting data collected during the course of this study for improved cattle production 
in Niger. 
 
 
 
 
 
 
 
 
 
iii 
 
University of Ghana http://ugspace.ug.edu.gh
DEDICATION 
This work is dedicated to my wife Ramatou Adamou Djibo, our adorable daughters Aïchatou and 
Samira, my mother Aissata Timbo and my brothers and sisters. I love you and say thank you for 
your exceptional and uncommon characters. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
iv 
 
University of Ghana http://ugspace.ug.edu.gh
ACKNOWLEDGEMENTS 
I thank God the Almighty for the opportunities I was offered. 
Sorghum and Millet Innovation Laboratory (SMIL) for the financial support.  
The WACCI administration and the University of Ghana for making my stay and learning easy.  
Special thanks to Professor Niaba Teme, my In country supervisor for his wonderful mentoring.  
I am grateful to my supervisory committee: Professor Pangirayi Tongoona, Professor Eric Y. 
Danquah, Professor Niaba Teme and Doctor Daniel K. Dzidzienyo. 
I would like to thank Professor Mitch Tuinstra (Purdue University) for proving brown midrib 
germplasm sources and assisting me for the laboratory work. 
I would like to thank Doctor Issoufou Kapran for his guidance throughout my progression.  
I am also grateful to Dr. Vaksmann Michel, Dr. Mamadou Aissata, Mr. Aliou Sissoko and Mr. 
Sekouba Sanogo for their sincere implication in this work. 
Your sacrifices and understanding are deeply appreciated. 
 
 
 
 
 
 
 
 
v 
 
University of Ghana http://ugspace.ug.edu.gh
TABLE OF CONTENTS 
DECLARATION ................................................................................................................................................ i 
ABSTRACT ...................................................................................................................................................... ii 
DEDICATION ................................................................................................................................................. iv 
ACKNOWLEDGEMENTS ................................................................................................................................. v 
TABLE OF CONTENTS .................................................................................................................................... vi 
LIST OF TABLES .............................................................................................................................................. x 
LIST OF FIGURES ........................................................................................................................................... xi 
LIST OF ABBREVIATIONS .............................................................................................................................xiii 
CHAPTER ONE ............................................................................................................................................... 1 
1.0. GENERAL INTRODUCTION ...................................................................................................................... 1 
CHAPTER TWO .............................................................................................................................................. 6 
2.0. LITERATURE REVIEW .............................................................................................................................. 6 
2.1. Nature of the Brown midrib character .............................................................................................. 7 
2.2. Types of genes controlling brown midrib trait in sorghum ............................................................... 8 
2.2.1. Sorghum bmr6 ............................................................................................................................ 9 
2.2.2. Sorghum bmr12 ........................................................................................................................ 10 
2.3. Lignin biosynthesis and composition ............................................................................................... 11 
2.4. Lignin mutation degradation ........................................................................................................... 15 
2.5. Feeding cattle with bmr mutants ..................................................................................................... 16 
2.6. Sorghum crossings techniques......................................................................................................... 17 
2.6.1. Hand emasculation ................................................................................................................... 18 
2.7. Sorghum improvement: The backcross breeding method .............................................................. 20 
2.7.1. Genetic issues of backcross method ......................................................................................... 21 
2.8. Bmr genes introgression .................................................................................................................. 23 
2.9. SNP markers for bmr genes ............................................................................................................. 24 
2.10. Dual purpose cereals: Farmers and traders’ preferences in SSA ................................................... 25 
CHAPTER THREE .......................................................................................................................................... 26 
3.0. SURVEY ON CROP RESIDUES FOR CATTLES FEEDING IN NIGER ........................................................... 26 
3.1. Introduction ......................................................................................................................................... 26 
3.2. Material and methods ..................................................................................................................... 27 
3.2.1. Farmers ..................................................................................................................................... 27 
vi 
 
University of Ghana http://ugspace.ug.edu.gh
3.2.2. Stover traders ........................................................................................................................... 30 
3.3. Study areas ....................................................................................................................................... 30 
3.3.1. Description ................................................................................................................................ 30 
3.4. Results .............................................................................................................................................. 31 
3.4.1. FGD Results ............................................................................................................................... 31 
3.4.2. Results Semi Structured Interview (SSI) .................................................................................... 33 
3.4.2.1. Farmers .............................................................................................................................. 33 
3.4.2.2. Stover traders .................................................................................................................... 37 
3.5. Discussion ......................................................................................................................................... 39 
3.6. Conclusion ........................................................................................................................................ 47 
CHAPTER FOUR ........................................................................................................................................... 49 
4.0. INTROGRESSION OF BROWN MIDRIB GENES INTO TWO NIGERIEN SORGHUM VARIETIES ................ 49 
4.1. Introduction ......................................................................................................................................... 49 
4.2. Material and methods ..................................................................................................................... 51 
4.2.1. Plant materials .......................................................................................................................... 51 
4.2.2. Development of bmr population .............................................................................................. 52 
4.3. Data analysis .................................................................................................................................... 56 
4.4. Results .............................................................................................................................................. 56 
4.5. Discussion ......................................................................................................................................... 59 
4.6. Conclusion ........................................................................................................................................ 61 
CHAPTER FIVE ............................................................................................................................................. 62 
5.0. AGRONOMIC POTENTIAL OF TWO NIGERIEN SORGHUM VARIETIES WITH BMR 6 AND 12 GENES IN 
TILLABERY AND KONNI ................................................................................................................................ 62 
5.1. Introduction ......................................................................................................................................... 62 
5.2. Material and methods ..................................................................................................................... 63 
5.2.1. Sites ........................................................................................................................................... 63 
5.2.2. Plant Materials .......................................................................................................................... 64 
5.2.3. Methodology ............................................................................................................................. 65 
5.2.3.1. Experimental Design .......................................................................................................... 65 
5.2.3.2. Cultural practices ............................................................................................................... 65 
5.3. Phenotypic data collection ............................................................................................................... 66 
5.3.1. Description of observations ...................................................................................................... 66 
5.4. Data analysis .................................................................................................................................... 68 
vii 
 
University of Ghana http://ugspace.ug.edu.gh
5.5. Results .............................................................................................................................................. 72 
5.5.1. Traits analysis of variances........................................................................................................ 72 
5.1.1. Time of flowering (50% FL) ....................................................................................................... 72 
5.1.2. Plant height (PH) ....................................................................................................................... 74 
5.1.3. Grain yield (GY) ......................................................................................................................... 75 
5.1.4. Fresh Stover Yield (FS) ............................................................................................................... 77 
5.1.5. Dry matter Yield (DM) ............................................................................................................... 80 
5.6. Combined ANOVA results ................................................................................................................ 82 
5.7. Interaction genotypes by environment-GxE .................................................................................... 85 
5.8. Genetic and Agronomic parameters ................................................................................................ 87 
5.9. Qualitative data analysis: Lodging and foliar diseases ..................................................................... 88 
5.9.1. Lodging ...................................................................................................................................... 88 
5.9.2. Foliar diseases ........................................................................................................................... 89 
5.10. Discussion....................................................................................................................................... 92 
5.11. Conclusion ...................................................................................................................................... 98 
CHAPTER SIX ................................................................................................................................................ 99 
6.0. DETERMINATION OF NUTRIENT CONTENT OF BROWN MIDRIB DERIVED POPULATION .................... 99 
6.1. Introduction ......................................................................................................................................... 99 
6.2. Material and methods ................................................................................................................... 100 
6.2.1. Plant material .......................................................................................................................... 100 
6.2.2. Methodology ........................................................................................................................... 102 
6.2.3. Stover nutrient content data analysis ..................................................................................... 102 
6.2.3.1. Data collection ................................................................................................................. 102 
6.3. Results ............................................................................................................................................ 105 
6.4. Discussion ....................................................................................................................................... 117 
6.5. Conclusion ...................................................................................................................................... 122 
CHAPTER SEVEN ........................................................................................................................................ 123 
7.0. EMS SORGHUM SNP MUTATIONS IDENTIFICATION .......................................................................... 123 
7.1. Introduction ....................................................................................................................................... 123 
7.2. Material and methods ................................................................................................................... 124 
7.2.1. Plant materials ........................................................................................................................ 124 
7.2.2. Methodology ........................................................................................................................... 125 
7.2.2.1. Leaf tissue sampling ......................................................................................................... 125 
viii 
 
University of Ghana http://ugspace.ug.edu.gh
7.2.2.2.DNA extraction .................................................................................................................. 126 
7.2.2.3. DNA quality and quantity measurements........................................................................ 126 
7.2.2.4. MutMap analysis .............................................................................................................. 127 
7.2.2.5. SNPs identification ........................................................................................................... 129 
7.3. Results ............................................................................................................................................ 129 
7.3.1. DNA quantity and quality ........................................................................................................ 129 
7.3.2. SbEMS phenotypes identification ........................................................................................... 131 
7.3.2.1. Family 1: SbEMS 1453 mut-1 - Niger 13 x Tx623 ............................................................. 131 
7.3.2.2. Family 2: SbEMS 2354 mut-1 - Niger 18 x Tx623 ............................................................. 131 
7.3.2.3. Family 3: SbEMS 3779 mut-1 - Niger 27 x Tx623 ............................................................. 132 
7.3.2.4. Family 4: SbEMS 4465 mut-1 - Niger 29 x Tx623 ............................................................. 132 
7.3.2.5. Family 5: SbEMS 4791 mut-2 - Niger 31 x Tx623 ............................................................. 133 
7.3.2.6. Family 6: SbEMS 4848 mut (#1)-1 - Niger 32 x Tx623 ...................................................... 134 
7.3.2.7. Family 7: SbEMS 5095 (#1)-1 - Niger 34 x Tx623 ............................................................. 134 
7.3.2.8. Family 8: SbEMS 5199 mut (#1)-1 - Niger 38 x Tx623 ...................................................... 135 
7.3.2.9. Family 9: SbEMS 5398 mut-1 - Niger 39 x Tx623 ............................................................. 135 
7.3.2.10. Family 10: SbEMS 932-1-1 - Niger 553 x Tx623 ............................................................. 135 
7.4. Discussion ....................................................................................................................................... 138 
7.5. Conclusion ...................................................................................................................................... 141 
CHAPTER EIGHT ......................................................................................................................................... 142 
8.0. GENERAL CONCLUSION AND RECOMMANDATIONS ......................................................................... 142 
8.1. GENERAL CONCLUSION .................................................................................................................. 142 
8.2. RECOMMENDATIONS..................................................................................................................... 143 
REFERENCES .............................................................................................................................................. 144 
APPENDICES .............................................................................................................................................. 171 
 
 
 
 
 
 
  
ix 
 
University of Ghana http://ugspace.ug.edu.gh
LIST OF TABLES 
Table 2.1 Advantages and disadvantages of sorghum different crossing technics ....................... 19 
Table 2. 2. Advantages and disadvantages of the backcross breeding method. ........................... 23 
Table 3. 1 FGD: Topics and evaluation methods. ........................................................................ 29 
Table 3. 2. Summary results of the FGDs ..................................................................................... 32 
Table 4. 1.  Summary of agronomic characteristics of the different parental lines. ..................... 51 
Table 4. 2. Quantities of F1 seeds and seedlings generated per cross. ......................................... 57 
Table 4. 3. Quantities of BC1F1 seeds produced per population. ................................................ 57 
Table 4. 4. Segregation in the F2 families. ................................................................................... 58 
Table 4. 5. Segregation in the BC1F2 families. ............................................................................ 58 
Table 5. 2. Formula for phenotypic, genetic and heritability components estimation.................. 71 
Table 5. 3. Mean squares of bmr derived families. ....................................................................... 72 
Table 5. 4. Summary of the single ANOVA and homogeneity of variances................................ 82 
Table 5. 5. Results Combined ANOVA........................................................................................ 84 
Table 5. 6. Trait means dispersion per population and site of experiment. .................................. 86 
Table 5. 7. Top 20 stable entries across Tillabery and Konni in Niger in 2017. .......................... 87 
Table 5. 8. Agronomic and genetic parameters of bmr derived families at Tillabery. ................. 88 
Table 5. 9. Agronomic and genetic parameters of bmr derived families at Konni. ...................... 88 
Table 5. 10. z-values distribution of bmr derived families and checks. ....................................... 91 
Table 6. 1. List of selected families and parental lines. .............................................................. 101 
Table 6. 2. Agronomic characteristics and BSD values of the entries. ....................................... 106 
Table 6. 3. Results t-test population one. .................................................................................... 116 
Table 6. 4. Results t-test population two..................................................................................... 116 
Table 6. 5 Results t-test population three.................................................................................... 117 
Table 7. 1. Pedigree and size of the different EMS families. ..................................................... 125 
Table 7. 2. Concentration, absorbance of bulk DNA of SbEMS families. ................................. 130 
Table 7. 3. Summary SNPs results.............................................................................................. 137 
 
 
 
x 
 
University of Ghana http://ugspace.ug.edu.gh
 
LIST OF FIGURES 
Figure 2. 1. The monolignol biosynthetic parthway in sorghum. Sattler et al. (2014) ................. 14 
Figure 3. 1. Locations of PRA villages. ........................................................................................ 31 
Figure 3. 2. Land tenure in the three villages. .............................................................................. 33 
Figure 3. 3. Farmers’ sorghum seed supply sources at three locations surveyed ......................... 34 
Figure 3. 6. Farmers’ perception on stover quality trait at Lossa ................................................. 35 
Figure 3. 7. Farmers’ perception on stover quality trait at Dolli. ................................................. 35 
Figure 3. 8. Farmers’ perception on stover quality trait at Soumarana......................................... 36 
Figure 3. 9. The major constraints of milk production in Lossa, Dolli and Soumarana. .............. 37 
Figure 3. 10. Purchased stover quality requirement. .................................................................... 38 
Figure 4. 1. The different types of sorghum midribs involved in this study................................. 52 
Figure 4. 3. The general sheme for bmr genes introgression. (RP=Recurrent parent; DP=Donor 
parent). .......................................................................................................................................... 55 
Figure 5 1. Ten years average annual rainfalls in Tillabery and Konni. ....................................... 64 
Figure 5 2. Grain yield data distribution ....................................................................................... 83 
Figure 5 3. Fresh stover yield data distribution ............................................................................ 83 
Figure 5 4. Dry matter yield data distribution............................................................................... 84 
Figure 6. 1. Dry matter content of the selected families and the checks. ................................... 107 
Figure 6. 2. Net energy content of the selected families and the checks. ................................... 108 
Figure 6. 3. Crude proteins content of the selected families and the checks. ............................. 109 
Figure 6. 4. Calcium and Phosphorus content of the selected families and the checks. ............. 111 
Figure 6. 5. NDF content of the selected families and checks. ................................................... 113 
Figure 6. 6. ADF content of the selected families and the checks. ............................................. 114 
Figure 6. 7. ADL content of the selected families and the checks. ............................................. 115 
Figure 7. 1. Simplified scheme for application of MutMap to rice. ........................................... 128 
Figure 7. 2. DNA bands on the agarose gel. ............................................................................... 130 
Figure 7. 3. The T:C mutation contrasting the EMS1453 mutant to the reference genome. ...... 131 
Figure 7. 4. The T:C mutation contrasting the EMS2354 mutant to the reference genome. ...... 132 
Figure 7. 5. The G:A mutation contrasting the EMS3779 mutant to the reference genome. ..... 132 
xi 
 
University of Ghana http://ugspace.ug.edu.gh
Figure 7. 6. The C:T mutation contrasting the EMS4465 mutant to the reference genome. ...... 133 
Figure 7. 7. The T:C mutation contrasting the EMS 4791 mutant to the reference genome. ..... 133 
Figure 7. 8. The C:T mutation contrasting the EMS 4791 mutant to the reference genome. ..... 134 
Figure 7. 9. The A:G mutation contrasting the EMS4848 mutant to the reference genome. ..... 134 
Figure 7. 10. The different mutations contrasting the EMS 5095 to the reference genome. ...... 135 
Figure 7. 11. The A:G mutation contrasting the EMS932-1-1 mutant to the reference genome.136 
 
 
 
  
xii 
 
University of Ghana http://ugspace.ug.edu.gh
LIST OF ABBREVIATIONS 
FAO: Food and Agriculture Organization of the United Nations. 
SSA: Sub-Saharan Africa. 
INS : Institut National de la Statistique. 
NGO: Non-Governmental Organization. 
Bmr: Brown midrib. 
CGIAR: Consultative Group on International Agricultural Research. 
EMS: Ethyl MethaneSulfonate. 
BC: Before Christ. 
KG: Kilogram. 
DNA: Deoxyribonucleic acid. 
INTSORMIL: International Sorghum and Millet. 
PRA: Participatory Rural Appraisal 
GPS: Global Positioning System 
INRAN: Institut National de la Recherche Agronomique du Niger 
ICRISAT: International Crops Research Institute for the Semi-Arid Tropics 
UNDP: United Nations Development Program 
ECOWAS: Economic Community of West African States 
xiii 
 
University of Ghana http://ugspace.ug.edu.gh
IRAT: Institut des Recherches Agronomiques Tropicales et des Cultures Vivrières 
SNP: Single Nucleotide Polymorphism 
CTAB: Cetyl TrimethylAmmonium Bromide 
 
 
 
 
 
 
 
 
 
 
 
 
xiv 
 
University of Ghana http://ugspace.ug.edu.gh
CHAPTER ONE 
1.0. GENERAL INTRODUCTION 
Sorghum bicolor (L.) Moench is an important food, feed and industrial crop world-wide. Sorghum 
ranks fifth in global production among cereals and is the staple food crop for 500 million people 
in Sub-Saharan Africa (SSA) with an annual production of 60 million tons (FAO, 2014). More 
than 35% of sorghum is grown for human consumption, the remaining is used for animal feed, 
alcohol production and industrial products (Awika and Rooney, 2004).  
In Niger, sorghum is the second most important cereal crop after pearl millet and the country is 
ranked fifth in Africa in terms of production (FAO, 2018).  All five of the cultivated races of 
sorghum according to Harlan's classification (bicolor, caudatum, durra, guinea and kaffir) are 
found in Niger but, the caudatum and durra races are more common over much of Niger (Clark 
and Kitch, 1982). Total area under sorghum cultivation was 3,111,100 ha with a total production 
of 1,375,700 tons leading to an averages yield of 0.442 t ha-1 (INS, 2013). Sorghum’s low yield is 
attributed to biotic and abiotic constraints that negatively affect the agricultural systems and food 
supply.  Sorghum is indeed grown in the marginal areas of Sahel / Sudan zones with an average 
annual rainfall of 350 mm, and up to 600 - 800 mm in Sudan savannah zone and more than 800 
mm in the small area covered by the Guinea savannah zone in Niger (JAICAF, 2009). 
Agricultural activities in Niger face serious climatic risks with soils providing low nutrients to 
sorghum plant which is grown from June to September. During this cropping season, rains are 
erratic and not well distributed. This situation is worsened by a low use of chemical fertilizers. 
Nevertheless, sorghum shows a good adaptation under these difficult growing conditions 
1 
 
University of Ghana http://ugspace.ug.edu.gh
conferring its strategic role in the search for food and feed security. Sorghum stover is used to feed 
livestock (cattle and small ruminants) and its stems as building materials by farmers.  
In rural areas of Niger, sorghum is a staple food where a wide range of foods and beverages are 
prepared from the grain (thin = fura in local language) and thick (tuwo = in local language) 
porridge or couscous. Furthermore, sorghum residues constitute an important source of forage for 
the important Niger livestock. In fact, Niger ranks as the eighth largest livestock holder in Africa 
(Yameogo et al., 2014). In the Sahel region, Niger with 10.5 million tropical livestock units (TLUs) 
has the largest herd population (World Bank, 2013). There is thus an important potential for beef 
production and exportation. Spore (2015) claimed that pastures and crop residues represent about 
90% of feed for large ruminant animals in Niger. However, crop residues as dry roughages from 
cultivated grain are rich in fiber and low in nitrogen, minerals and vitamins, furthermore with 
forage maturity, there is decreased nitrogen, digestibility and increased fiber and lignin contents 
(Sultan et al., 2011). 
Consequently, a large part of sorghum stover is thus not efficiently used by animals. Suad et al. 
(2015) highlighted that in view of the pressing demand for fodder coupled with the importance of 
the grain of sorghum for people in SSA, it is imperative to reconsider the present mono-commodity 
breeding strategy of sorghum. Furthermore, Spore (2015) forecasted that between 2005 and 2030, 
meat and milk consumption in Africa will grow by 2.8 and 2.3% per year respectively and 
according to Aschalew (2014) many scientists can agree with rapid population growth, 
urbanization and rise in per capita income are opportunities for livestock revolution that is 
expected to be happening in the 2020s. However, Africa’s livestock producers are currently not 
able to meet the rising demand. In particular, providing Africa’s livestock with sufficient quantity 
2 
 
University of Ghana http://ugspace.ug.edu.gh
and quality feed which is a critical issue that researchers, NGOs (Non-Governmental 
Organizations) and policymakers, amongst others have to address (Spore, 2015).  
The use of sorghum as a forage crop is in fact growing in many regions of the world due to its high 
biomass productivity and ability to utilize efficiently water even under drought conditions 
(Sanchez et al., 2002). Forage sorghums constitute a particular group that was classified by Reed 
et al. (1988) into (i) greenchop-haylage-pasture grazing types that generally include sudangrass 
(subrace sudanense), sudangrass x sudangrass hybrids, or sorghum (bicolor grain type) x 
sudangrass hybrids, (ii) dual-purpose sorghums, usually bicolor grain types, tall 3 dwarf or normal 
2 dwarf in height, that are characterized by high grain:stover ratios, (iii) sweet sorghum cultivars 
or hybrids involving grain x sweet hybrids fed as greenchop, silage, or hay, (iv) silage types, 
usually 0, 1, or 2 dwarf types, that are ensiled, or (v) stubble residue after grain harvest that is 
supplemented with high-protein legume forages and/or grain concentrates as a good temporary 
roughage source. 
The genus Sorghum covers various species with high forage potential. Among them, the brown 
midrib types are well known for forage quality improvement. Incorporating the bmr trait into these 
crops could result in increased feed efficiency due to the reduction in lignin content with no 
repercussions on plant physiology. Bmr genotypes utilization in practical plant breeding 
programmes would not be difficult. Indeed, bmr genes inheritance is simple recessive and the bmr 
genes exhibit a useful morphological marker because it produces an easily identifiable phenotype 
(Bittinger et al., 1981). Another advantage of bmr sorghum lines is that they tend to be less 
susceptible to insects and pest damage compared to their wild types (Dowd et al., 2016), tolerant 
to fungal diseases at equal or possibly greater level than their wild-type (Funnell and Pedersen, 
2006) and the resistance to lodging compared to their wild types (Oliver et al., 2005). In addition, 
3 
 
University of Ghana http://ugspace.ug.edu.gh
Hisano et al. (2009) revealed that plants with low lignin had high total carbohydrate levels in 
biomass reflecting a compensation for the reduction in lignin level on a mass balance basis. The 
compensation mechanism suggest the existence of a cell wall sensing system which relays 
information from the cell wall to the interior of the cell, so it can react with regard to the synthesis 
of new cell wall components (Hannes et al., 2015).  
Several studies have been reported on the digestibility of stover. Chemical and genetic approaches 
have been used to improve forage fiber digestibility by reducing the amount of lignin or the extent 
of lignin cross linking with cell wall carbohydrates. Gerhardt et al. (1994) and Oliver et al. (2004) 
reported that forage sorghum digestibility can be significantly improved with the use of bmr genes.  
In Niger, as well as in the other countries of West Africa, technology for treating cereals straw 
with urea and supplementation with multi-nutritional blocks have been popularized but poorly 
adopted by Farmers (Abdou, 2010). On the other hand, enhancing the production quality with 
farmers and stovers traders’ implication of this abundant and available stover through genetic 
improvement may have a significant impact on livestock productivity. Indeed, Kelley et al. (1996) 
concluded that farmers’ new variety adoption is highly linked to its stover productivity. In this 
view, Tunde and Augustine, (2014) reported that farmers empirically estimate the value of a given 
feed from observations of animal responses. Crop residues play a strategic role for livestock 
feeding at small scale farmers level because they are harvested from crops that are primarily 
cultivated for grain (CGIAR, 2011). 
To date no studies have been reported on the nutritive value and the impact of sorghum containing 
the bmr trait on livestock feeding in Niger. Bmr6 and bmr12 genes are currently use to develop 
new dual purpose sorghum lines with high stover quality by introgressing bmr traits in local elite 
4 
 
University of Ghana http://ugspace.ug.edu.gh
varieties. In addition to bmr6 and bmr12, the sorghum breeding programme hypothesized that the 
EMS mutations induction in BTx623 (Sorghum bicolor reference genome) already cultivated in 
Niger will lead to new bmr variants with better nutritive quality. Indeed, according to Yusuff et al. 
(2016) mutagenesis is now a pillar of modern plant breeding. In fact, in plants, spontaneous 
mutations are very rare and random in terms of time of occurrence and that makes them more 
difficult to use in plant breeding programmes (Lönnig, 2005). 
The main objective of this thesis was to develop dual purpose sorghum lines with high quality 
stover to sustain livestock productivity in Niger. 
The specific objectives were to: 
i. identify farmers’ and stover traders’ preferences on sorghum stover varieties,  
ii. introgress bmr6 and bmr12 genes in two Niger elite sorghum lines, 
iii. determine the agronomic and nutritive potential of new bmr lines, and 
iv. identify SNP mutations in EMS-induced sorghum mutants used in the national sorghum 
improvement programme. 
 
 
 
 
 
 
 
 
5 
 
University of Ghana http://ugspace.ug.edu.gh
CHAPTER TWO 
2.0. LITERATURE REVIEW 
Sorghum bicolor (L.) Moench is a tropical grass. It is considered to be predominantly a self-
pollinated diploid species (2n = 20) and belongs to the Poaceae family (House, 1985). It is treated 
as an annual in temperate or subtropical climes where cold temperature limits growth and 
development (Duncan, 1996). Sorghum has a great genetic diversity composed mainly of bicolor, 
guinea, caudatum, durra, and kafir races with ten intermediate races recognizable by spikelet or 
the panicle morphology (Harlan and de Wet, 1972). It is also unique in its ability to thrive under a 
wide array of harsh environmental conditions and in some areas like Sub-Saharan Africa (SSA) 
and India, it is arguably the most important cereal crop and it is used as source of food and feed 
(Suad et al., 2015). Xin et al. (2009) reported the high water efficiency use of sorghum and its 
tolerance to low soil fertility in comparison with the major cereal crops. These traits make sorghum 
particularly advantageous as feedstock crop. Getachew et al. (2016) highlighted the potential of 
sorghum as forage for livestock in areas with limited water availability.  
History on forage crops can be traced back to about 1300 BC when alfalfa was cultivated in areas 
of modern Turkey (Iqbal et al., 2015). Forage was defined as a succulent green plant form and as 
crops harvested and cured as hay or silage to feed animals during lean periods when there was 
scarcity of green forage. During feed scarcity, livestock rely mostly on crop residues for feeding 
which can be up to 45% of their total annual feed intake, and up to 80% during critical periods 
(Sandford, 1989). In countries like Niger natural pasture and crop residues represent about 90% of 
feed for large ruminant animals (Spore, 2015). In the case of Sudan, where the second largest 
animal wealth in the African continent exists, sorghum straw has a great contribution in 
maintaining the national herd (Maarouf et al., 2014). In this view, Buso et al. (2011) emphasized 
6 
 
University of Ghana http://ugspace.ug.edu.gh
that the advantage of using sorghum is the possibility of producing greater biomass volume, 
associated with better quality silage, with digestible fibers for high animal consumption and 
performance in areas with cultivation restrictions for maize and other forages. Another advantage 
of sorghum is its great genetic resource variability including brown midrib types. 
2.1. Nature of the Brown midrib character 
 
The brown midrid (bmr) is a phenotypic marker. Sattler et al. (2010) reported that this phenotype 
has been observed in maize (Zea mays), sorghum (Sorghum bicolor) and pearl millet (Pennisetum 
glaucum) and was associated with reduced lignin levels and altered lignin composition compared 
to wild-type. In sorghum, Ayyangar and Ponnaiya (1941) reported that the abbreviation bmr was 
adopted to distinguish it from bm, already in use for the sorghum bloomless mutants (lack of waxy 
coating, which gives the typical colour to the stem and leaves). The bmr plants display a useful 
brown morphological marker on leaves and stems (Rao et al., 2012). The mutation responsible of 
the bmr phenotype occurs spontaneously or is induced by chemical or physical mutagenesis agents. 
Bmr mutants can also be created by genetic engineering.  
Bout and Vermerris (2003), stated that the bmr phenotype is controlled by a single recessive gene 
and it is the result of a recessive mutation in the lignin biosynthesis pathway. Hanna (1982) 
reporting Porter et al. (1978) treated seeds with 0.1 and 0.2% diethyl sulphate (DES) to induce 
bmr mutants with reduced lignin which were identified in segregating M3 head rows. The same 
conclusion was drown by Nagaraja et al. (2008). Furthermore, Casler et al. (2003) claimed that 
when the mutation is present in the homozygous recessive state, it results in reduction of 
lignification, cell-wall concentration, increases digestibility and voluntary intake of feed by 
ruminants. In addition, Batista de Aguilar et al. (2014) reported likewise that this mutation 
7 
 
University of Ghana http://ugspace.ug.edu.gh
increases forage digestibility, dry matter intake and productivity per animal. Consequently, plants 
with the bmr genes contain less lignin and are highly digestible. Cherney et al. (1991), by 
illustration asserted that lignin concentration of bmr lines has been reduced by 5 to 50%; a 10 g 
kg-1 decrease in lignin generally resulted in a 40 g kg-1 increase in digestibility thus animal 
performance may increase by up to 30%.  
In terms of physiological properties, Jung and Fahey (1983) proposed that the lignin content in 
bmr plants is poorly polymerized and contain fewer phenolic monomers that can affect digestion. 
It is well known that modifying lignin content or composition is important to enhance stover 
digestibility. In fact, plant lignin, after cellulose, is the second most important polymer in nature 
and while being beneficial to plants, the complex linkages between lignin and structural 
carbohydrates are detrimental to the digestibility of plant residues by livestock (Humphreys and 
Chapple, 2002). Moreover, according to Aydin et al. (1999) lignin is the primary indigestible 
component of plant cell walls, it inhibits digestion of cell wall carbohydrates in the rumen. 
2.2. Types of genes controlling brown midrib trait in sorghum 
 
Bmr mutants were first discovered in maize in 1924 in a self-pollinated line of a northwestern dent 
maize, resulting finally in the description of a total of four genes with each gene segregating as a 
simple Mendelian recessive character (Barriere et al., 2004). Following the finding of these genes, 
a series of nineteen bmr mutants were isolated from diethyl sulfate (DES) mutagenized populations 
at Purdue University in the 1970s (Porter et al., 1978). Those bmr mutants were designated bmr-1 
through bmr-19. However, six mutants were deleted due to sterility and poor phenotypic 
expression and three were selected as most agronomically acceptable (Fritz et al., 1988). Later 
additional spontaneous bmr mutants of sorghum were identified in breeding populations (Vogler 
8 
 
University of Ghana http://ugspace.ug.edu.gh
et al., 1994). Porter et al. (1978) suggested that in mature bmr sorghum plant the lignin content is 
half lesser than the wild type, whereas its leaf lignin content was just one quarter. These results 
were also confirmed by Grant et al. (1995). Indeed, those authors showed that the bmr mutants of 
sorghum have significantly lower levels of lignin content; they estimated around 51% less in their 
stems and 25% less in their leaves and furthermore bmr sorghum silage had 17% less lignin than 
regular sorghum silage making it more digestible. Allelism tests on the sorghum bmr mutants 
showed that several of the mutations are allelic (Bittinger et al., 1981). Saballos et al. (2008) 
proved that up to date four allelic classes of sorghum bmr mutants have been identified which are 
bmr2, bmr6, bmr12 and bmr19, with the first three showing the greatest potential for increasing 
biomass conversion. Sattler et al. (2009) confirmed that the bmr6 and bmr12 genes have been 
cloned; later Saballos et al. (2012) stated that mutations in 4-coumarate coenzyme a ligase (4CL), 
cinnamyl alcohol dehydrogenase (CAD2) and caffeic O-methyltransferase (COMT) genes 
contribute to the phenotype of bmr2, bmr6 and bmr12 groups respectively.  
 
2.2.1. Sorghum bmr6 
 
In sorghum, bmr6 encodes the major cinnamyl alcohol dehydrogenase (CAD) protein involved in 
lignin synthesis (Sattler et al., 2009). Reduced CAD activity altered cell wall architecture, reduced 
lignin level, and the incorporation of phenolic aldehydes into lignin in sorghum and maize (Palmer 
et al., 2008). Bmr6 plays an important role in sorghum lignin biosynthesis. Sattler et al. (2009) 
affirmed that bmr6 affects phenylpropanoid metabolism, resulting in reduced lignin concentrations 
and altered lignin composition in sorghum. Further, Sattler et al. (2010) confirmed that bmr6 plants 
were shown to have limited cinnamyl alcohol dehydrogenase (CAD which is a precursor for 
polymerization to lignin) activity, the enzyme that catalyzes the conversion of hydroxycinnamoyl 
9 
 
University of Ghana http://ugspace.ug.edu.gh
aldehydes (monolignals) to monolignols. Humphreys and Chapple (2002) demonstrated that the 
last step in monolignol biosynthesis is the reduction of cinnamyl aldehyde precursors, catalyzed 
by the enzyme cinnamyl alcohol dehydrogenase (CAD). Reduced CAD activity results in 
increased digestibility on dry weight basis, altered cell wall architecture, reduced lignin level, and 
the incorporation of phenolic aldehydes into lignin in sorghum and maize (Pillonel et al., 1991). 
Sattler et al. (2009) affirmed that genomic and cDNA sequences of Sb04g005950, the gene whose 
predicted amino acid sequences shared significant similarity to the Arabidopsis proteins AtCAD4 
(67.7% and 83.4%) indicated that a C-to-T transition mutation is present in the bmr6 gene and the 
mutation changed amino acid 132 of the protein from Glutamine (CAG) to a stop codon. 
2.2.2. Sorghum bmr12 
Bout and Vermerris (2003) claimed that among a number of the sorghum bmr mutants having a 
significantly lower lignin content in stems and leaves compared to their wild-type counterparts, 
the most severe reduction in lignin was found in the bmr12 and bmr18 mutants. Allelism tests by 
Bittinger et al. (1981) indicated that bmr12 and bmr18 are allelic and different from bmr6. 
However, Bout and Vermerris (2003) demonstrated that regarding their flowering time, a 
difference exists with bmr12 mutant flowering later than its wild-type whereas no significant 
difference was observed between the bmr18 and its wild-type. In addition, the cell wall 
composition of leaf blade tissue from bmr12 mutant contained less p-coumaric acid which is 
primarily esterified to syringyl residues (Ralph et al., 1994), therefore a reduction in amounts of 
syringyl residues is observed. 
Sattler et al. (2010) indicated that bmr12 encodes a caffeic O-methyltransferase (COMT) which 
catalyzes the penultimate step in monolignol biosynthesis, the transfer of a methyl group from S-
adenosyl-methionine (SAM) to the 5-hydroxyl group of 5-hydroxy-coniferyl substrates to form 
10 
 
University of Ghana http://ugspace.ug.edu.gh
sinapyl products in conformity with Li et al. (2000) whose study showed that caffeic acid O-
methyltansferase (COMT) controls the synthesis of syringyl lignin units.  
 
2.3. Lignin biosynthesis and composition  
Lignin is a complex, amorphous biopolymer matrix; it serves critical functions such as the 
structural integrity of cell wall tissues for vascular plants. Indeed, lignin is required for vascular 
elements to transport water under negative pressure, an important function for adaptability (Boyce 
et al., 2004). Zheng-Hua et al. (2000) added that lignin offers rigidity to the walls of conducting 
tracheary elements (TEs) for enduring the negative pressure generated from transpiration, it also 
renders the walls of maturing TEs undigestible by hydrolytic enzymes released during autolysis of 
xylogenesis. In addition, Zheng-Hua et al. (2000) affirmed that the deposition of lignin in the walls 
of sclerenchyma cells adds physical toughness and chemical durability to the walls but may deter 
feeding by herbivores. Lignin action may also block the liberation of sugars from the cell wall 
polysaccharide moieties, release compounds that can inhibit microbes used for fermenting sugars 
to fuels, and adhere to hydrolytic enzymes (Sattlet et al., 2009). However, lignin can also offer a 
physical barrier for protection of adjacent tissues from damage after deposition at the sites of 
wounding or pathogen invasion. Therefore, lignin content reduction in plant cells has become an 
area of interest for bioenergy feedstock improvement (Chen and Dixon, 2007). Lignin content of 
crop plants has been reduced by traditional plant breeding, natural and induced mutations, and 
insertion of transgenes (Pedersen et al., 2005). Lignin, the third cell wall component, is an 
ubiquitous polymer whose abundance in the world is exceeded only by cellulose (Jung and Ni, 
1998). It accounts for approximately 30% of organic carbon in the biosphere (Boerjan et al., 2003). 
Conversely, severe reduction of lignin content may affect plant fitness (Sattler et al., 2014). Indeed, 
11 
 
University of Ghana http://ugspace.ug.edu.gh
many studies such Jones et al. (2001); Piquemal et al. (1998); Ruel et al. (2009) reported the 
collapse of vascular elements in mutants extremely impaired in lignin synthesis. In the same idea, 
Mir Derikvand et al. (2008) showed that at very low levels of lignin, plant growth and development 
are affected.  
The biogenesis of the lignified plant cell wall is a complex metabolic process in which the 
formation and modification of several cell wall polymers need to be coordinated (Saballos et al., 
2008). In the same way, lignin biosynthesis is a complex metabolic process and Dauwe et al. 
(2007) define lignin as an aromatic heteropolymer and an important component of secondarily 
thickened cell walls contributing to make them rigid. They have attracted significant research 
attention because they represent a major obstacle in chemical pulping, forage digestibility and 
processing of plant biomass to biofuels (Vanholme et al., 2008). In addition, understanding lignin 
biosynthesis and properties were important targets for feed improvement (Mackay et al., 1997). 
The deposition of lignin in plant cell walls was part of the mechanisms which allowed the 
development of upright plants of large size adapted to a terrestrial habitat (Alain, 2000). However, 
the exact mechanism of lignin biosynthesis is still unclear, and many researchers such as Saballos 
et al. (2008) affirmed that lignin is synthesized via the oxidative coupling of several phenolic 
monomers and the traditional view holds that lignin is derived from the monolignols p-coumaryl, 
coniferyl and sinapyl alcohol, which give rise to p-hydroxyphenyl (H), guaiacyl (G), and syringyl 
(S) residues respectively. Bonawitz and Chapple (2010) affirmed that the deposition of each lignin 
type in plant tissues is spatially and temporally controlled: first being deposited H units, followed 
by G units, finally S units. It is important to notice that Chloe et al. (2002) affirmed that in 
angiosperms, lignin is composed of guaiacyl (G) and syringyl (S) units, the G unit is singly 
methylated on the 3-hydroxyl group, whereas the S subunit is methylated on both the 3- and 5-
12 
 
University of Ghana http://ugspace.ug.edu.gh
hydroxyl moieties. The ratio of S-to-G subunits dictates the degree of lignin condensation by 
allowing for different types of polymeric linkages, in fact increased G content leads to more highly 
condensed lignin composed of a greater proportion of biphenyl and other carbon-carbon linkages, 
whereas S subunits are commonly linked through more labile ether bonds at the 4-hydroxyl 
position (Guo et al., 2001). For Pond et al. (1987), in forage crops harvested at maturity, lignin 
content increases, particularly the S subunits which may negatively affect the digestibility and 
ultimately the nutrient availability of alfalfa and other forage grasses. Rabinovitch et al. (2001) 
reported that the conventional nomenclature for the monomers are called monolignols. According 
to Sattler et al. (2012), lignin subunits are derived from the amino acid phenylalanine, hence both 
phenylalanine metabolism and lignin biosynthesis occupy central roles in plant metabolism. 
Boerjan et al. (2003) detailed that Phenylalanine, the primary substrate in the lignin biosynthetic 
pathway, is derived from the shikimate pathway and then modified by the enzyme phenylalanine 
ammonia lyase (PAL) to form cinnamic acid. Subsequently, a series of enzymes such as 
hydroxylases [p-coumarate-3-hydroxylase (C3H), cinnamate-4-hydroxylase (C4H), ferulate-5-
hydroxylase (F5H)], methyl transferases [caffeic acid-O-methyltransferase (COMT), caffeoyl-
CoA-O-methyltransferase (CCoAOMT)], reductases [cinnamoyl-CoA-reductase (CCR)], and 
dehydrogenases [cinnamyl alcohol dehydrogenase (CAD), sinapyl alcohol dehydrogenase (SAD)], 
produce lignin biosynthetic pathway components such as alcohols, acids and aldehydes. 
Alain (2000), added that lignin synthesis requires significant energy and the final product is highly 
heterogeneous in nature but always exhibits a higher C/H and C/O ratios than the other polymers 
of the cell wall resulting in a higher calorific value. Recently, Konovalov et al. (2015) asserted that 
the biosynthesis of lignin takes place on the external side of the cell, on a polysaccharide matrix at 
13 
 
University of Ghana http://ugspace.ug.edu.gh
certain positions which contain hydroxycinnamic (n-coumaric and ferulic) acids bound to the cell 
wall; their phenolic oxy-groups serve as “anchors” for the process of lignification. 
The composition of lignin contains aromatic metabolites: aldehydes and acids (Dalimova et al., 
1994). Aldehydes (monolignans) are included in the lignin in the same manner as monolignols, 
via peroxidase and laccase (Ros Barceló et al., 2001). Konovalov et al. (2015) believe that the 
lignin, having aldehydes in its composition, becomes colored red or reddish-brown.  
 
 
Figure 2. 1. The monolignol biosynthetic parthway in sorghum. Sattler et al. (2014) 
The enzymatic steps (gray) are as follows: phenylalanine ammonia lyase (PAL); cinnamate 4-
hydroxycinnamol CoA:shikimate transferase (HCT); p-coumarate 3-hydroxylase (C3H); caffeoyl 
shikimate esterase (CSE); caffeoyl CoA O-methyltransferase (CCoAOMT); cinnamyl CoA 
reductase (CRR); ferulate 5-hydroxylase (F5H); caffeic acid O-methyltransferase (COMT); and 
cinnamyl alcohol dehydrogenase (CAD). The bmr2, bmr6 and bmr12 mutants are impaired in 4CL, 
CAD and COMT enzymatic activities, respectively. Source: Sattler et al. (2014) 
 
14 
 
University of Ghana http://ugspace.ug.edu.gh
2.4. Lignin mutation degradation   
Plants depend on their cell walls for rigidity and fitness (Santiago et al., 2013). The cell wall of 
plant biomass is composed of cellulose, hemicellulose, lignin, and minor components (Cassie et 
al., 2015). In many regions, grasses are generally used in livestock feeding however for an efficient 
nutrition, the degradation of fiber in animal stomach is an important factor limiting forages 
utilization. Carpita and Gibeaut (1993) reported that cell walls of Gramineae differ from the cell 
walls of many other species in terms of both their carbohydrate and lignin composition while Jung 
and Fahey, (1983); and Grisebach (1981) confirmed that lignin is a factor limiting the extent of 
digestibility of cell wall polysaccharides by animals. In addition, Jung and Allen (1995) asserted 
that it may highly influence rumen microbes’ action on plants cellulose and hemicellulose 
therefore in breeding for improving the feeding value of a crop, cell wall digestibility is the target. 
Barrière et al. (2003) reported that this is obvious when the harvested plant has no grain. Most 
research efforts to control lignin have focused on biosynthesis of the monolignols, most of the 
enzymes involved in the monolignol synthesis have been cloned and characterized in Arabidopsis 
thaliana (At) and others dicot species (Anterola and Lewis, 2002).  
The most rapid method of improving forage sorghum quality is to improve In Vitro Dry Matter 
Digestibility (IVDMD). Pedersen et al. (1982) and Boerjan et al. (2003) reported that the lignin 
polymer cross-links the polysaccharides, rigidifying and reinforcing the secondary cell wall 
structure. Sattler et al. (2012) reported that this resulted in a barrier that is chemically and 
microbially resistant. Furthermore, Sattler et al. (2014) showed that this resulting matrix is 
refractory to both chemical degradation and biological digestion, which impairs hydrolysis of the 
polysaccharides into their monomeric sugars in ruminant livestock or cellulosic bioenergy 
systems.  
15 
 
University of Ghana http://ugspace.ug.edu.gh
Chemical mutagenesis by soaking sorghum seeds in diethyl sulfate or ethyl methane sulphonate 
(EMS) induces mainly point mutations, and are thus ideal for producing missense and nonsense 
mutations that are known to produce phenotypic changes which could be desirable in crop 
improvement. Among the chemical agents, the alkylating agent (EMS) is the most commonly used 
in plants because of its potency and ease to use. Another advantage of EMS is the low level of 
chromosomal breaks and lethal effects (Greene et al., 2003). EMS alkylates guanine bases and 
leads to mispairing-alkylated G pairs with T instead of C, resulting in primarily G/C- to-A/T 
transition mutations (Ali et al., 2012). Sorghum bmr genotypes have been shown to contain less 
lignin, therefore they are more digestible than normal forage sorghum. Oliver et al. (2005) found 
that digestibility is due to two different modifications in biochemical pathways by which lignin is 
produced in the plant. The bmr phenotype is characterized by the presence of a brown coloured 
midrib in sorghum, maize and pearl millet (Rao et al., 2012). Halpin et al. (1998) stated that the 
brownish pigments are strongly associated with the lignin as they persist in the cell wall after the 
removal of hemicellulose and cellulose. The bmr phenotype has been widely used mutants’ 
identification (Sattler et al., 2014). For the phenotype evaluation, it is important to notice that, the 
bmr phenotype is visible at seedling stage but tends to fade at physiological maturity. Although 
the intensity cannot be taken as a measure of reduction in lignin, it is a clear indicator that the bmr 
genes are present.  
2.5. Feeding cattle with bmr mutants 
Lima et al. (2005) considered that sorghum has the potential to be used as ruminants feed, 
especially in semi-arid regions where it is a major source of fodder and for being resistant to 
drought and high temperatures, high yield, and high nutritional value. In fact sorghum has been 
shown to be an excellent silage in areas characterized by unfavourable growing conditions for 
16 
 
University of Ghana http://ugspace.ug.edu.gh
maize. In many milk producing regions sorghum constitutes an important forage crop for grazing 
dairy cows. Some studies have demonstrated that the bmr silage with or without protein 
supplements increases significantly the milk production of lactating cows (Cherney et al., 1991; 
Oba and Allen, 1999) and animal performance. These properties make the use of bmr sorghum for 
cattle feeding very attractive in many countries. According to a paper from INTSORMIL 
(International Sorghum and Millet), University of Nebraska-Lincoln (2013) a trial conducted in 
some countries in central America (Costa Rica, El Salvador, Nicaragua, Honduras, Guatemala, 
Panama and Haiti) using bmr sorghum for feeding dairy cows, dairy farmers had an average 
increase of 20% in milk production and weight gain of the cattle was 900 g per animal per day 
which represents an increase of 10% over conventional sorghum. The bmr sorghum lines were 
very rapidly spread throughout Central America and had a significant impact on increasing rural 
income and promoting food security. 
2.6. Sorghum crossings techniques 
Sorghum inflorescence is a panicle. Indeed, Brown et al. (2006) reported that the inflorescences 
show primary, secondary and tertiary branching. Morphologically, the panicle (or head) varies 
from compact to open. Sorghum is thus a perfect-flowered plant and is a normally self-pollinated 
crop with however a certain level of outcrossing. This level of outcrossing is contingent on the 
specific genotype being grown and the environmental conditions encountered prior to and during 
anthesis (Schertz and Dalton 1980). House (1985) estimated the outcrossing from 1 to 10% and 
can also reach to 30-60% depending to the panicle compactness. In general, the knowledge of plant 
reproduction mode is important for a successful improvement programme. For sorghum 
improvement, the discovery of genetic inheritance and the development of plant breeding 
principles in the early 20th century were the levers rapidly adopted by sorghum breeders and used 
17 
 
University of Ghana http://ugspace.ug.edu.gh
to address critical issues (Rooney, 2004). Sorghum flowering starts a few days after panicle 
emergence from the top toward the bottom of the panicle. Based on the flowering habits, sorghum 
breeders have developed several methods of hybridization and outcrossing elimination (genetic 
male sterility; cytoplasmic male sterility; hot-water emasculation; anther dehiscence control by 
use of humidity; chemical control of anther dehiscence). Indeed, fertile line panicles’ early selfing 
ensure pure line maintenance whereas several crossing methods were established for segregating 
populations’ development. Below, the principle crossing technics are described and the table 2.1. 
summarizes the advantages versus disadvantages of each technic. 
2.6.1. Hand emasculation 
The chosen panicle for emasculation is prepared 1-2 days before. Flowers are emasculated the day 
before anthesis. Such florets occur below and within about 3 cm of opened florets in a sorghum 
panicle. All open spikelets are removed with scissors. In addition, all florets except those that are 
to be emasculated are removed, leaving only the florets that are expected to open the next day. The 
three anthers are coaxed out of the enclosing lemma and palea by inserting a sharpened pencil or 
similar pointed instrument. Care must be taken not to break the anthers, and if the anther is 
breached, that flower should be removed and instruments rinsed to avoid contaminating the next 
floret. Every anther must be removed before the set of florets is completely emasculated. (Rooney, 
2004). 
  
18 
 
University of Ghana http://ugspace.ug.edu.gh
Table 2.1 Advantages and disadvantages of sorghum different crossing technics  
Method Advantages Disadvantages 
Hand emasculation Require simple tools (House, 1985) Varieties differ in ease of emasculation 
(House, 1985) 
 
Produce small quantities of seeds (Schertz 
and Dalton 1980) 
Takes special skill and time consuming 
(Schertz and Dalton 1980) 
Genetic male sterility Simple and breeders can make a Can not be use for commercial hybrids 
large number of crosses in a short seeds production 
 
amount of time (Rooney, 2004) 
Cytoplasmic male sterility Simple and breeders can make a Makes possible the commercial hybrid 
large number of crosses in a short seeds production (House, 1985) 
amount of time (Rooney, 2004) 
Hot-water emasculation Simple to use in green house High level of seedlings from self- 
(Rooney, 2004) fertilization in the nurseries. 
 
Varieties respond differently to this technic 
(House, 1985)  
Anther dehiscence control Simple and breeders can make a A certain level of self-pollination 
with humidity large number of crosses in a short can occur in seed from a poured cross. 
amount of time (Rooney, 2004) (Rooney, 2004) 
 
In most cases, the proportion of progeny 
that are F1 hybrids will vary based on the 
specific genotype used as a female parent, 
the fecundity of the pollen parent, and the 
environmental conditions during the 
process. (Rooney, 2004) 
Chemical control of anther  Require a good control of the technic 
dehiscence 
Source: House (1985); Rooney (2004); Schertz and Dalton (1980) 
 
19 
 
University of Ghana http://ugspace.ug.edu.gh
2.7. Sorghum improvement: The backcross breeding method 
According to Reddy et al. (2012), sorghum, improvement research has come a long way from 
using simple methods like mass selection to advanced level of selection using molecular markers 
for trait improvement. The backcross breeding method was first proposed by Harlan and Pope 
(1922). The method is widely used by sorghum breeders to transfer a gene from one source (donor 
parent) to another line (recurrent parent). Backcrossing is useful for several breeding objectives 
such us insects, diseases, drought tolerance, high lysine or proteins content (House, 1985). This 
author added that the relationship between the donor and recurrent parents is an important aspect 
in a backcross breeding method. In other words, if a desired character is found in a kafir and a 
durra type, and it is desired to transfer this character to a kafir, then it would be better to use the 
kafir rather than the durra source as the donor parent (House, 1985). In plant breeding, the 
backcrossing method is used to improve a well-adapted line that lacks a particular gene. In the 
crossing program, the well adapted desirable parent is called recurrent parent and the source of the 
desirable gene lacked in the recurrent parent background is called donor parent. Even though the 
chief role of the donor parent is to provide the missing gene, it should not be significantly deficient 
in other desirable traits. In fact, the rationale of backcross breeding is to replace a specific 
undesirable gene with a desirable alternative while preserving all other qualities (adaptation, 
productivity, etc.) of an adapted line (Acquaah, 2007). Therefore, this breeding method enables 
breeders to transfer desirable traits in a specific genetic background. Instead of inbreeding the F1 
as normally done, the F1 is crossed with the desirable parent. In successive generations, progenies 
are selected for the characteristic of interest and then backcrossed to the recurrent parent in order 
to recover the maximum genome of the RP carrying the trait of interest (Hospital, 2005). Acquaah 
(2007) mentioned that a selection pressure is used to identify and eliminate undesirable 
20 
 
University of Ghana http://ugspace.ug.edu.gh
individuals. For recessive trait, a progeny test is necessary to determine the genotype of a backcross 
progeny before continuing with the next cross. This authors emphasized that backcross method of 
breeding is best suited to improving established cultivars that are later found to be deficient in one 
or two specific traits. However, Fehr (1987) reported that the success of backcrossing is influenced 
by the number of genes controlling a character and the role of the environment on the phenotypic 
expression of a gene(s). For Acquaah (2007) the backcross method is most effective and easy to 
conduct when the missing trait is qualitatively (simply) inherited, dominant, and produces a 
phenotype that is readily observed in a hybrid plant. If the trait is simply inherited, the RP can be 
effectively covered with the trait incorporated (House, 1985). The procedure for transferring a 
recessive trait is similar to that for dominant traits, but entails an additional step. 
2.7.1. Genetic issues of backcross method 
In the breeding process to improve the recurrent parent, with each backcross step the progeny 
becomes more like the recurrent parent. Hospital (2005) affirmed that if selection is applied for 
the desired characteristic only, then the proportion of donor genome is expected to be reduced by 
one-half (50%) at each generation, except on the chromosome holding the characteristic. On the 
chromosome carrying the trait of interest, the rate of decrease is slower due to linkage drug (Stam 
and Zeven 1981). On the overall genome of the RP; House (1985) reported that with each 
generation of backcrossing, the genetic composition of the donor parent is reduced by a factor 
(1/2)n where ‘n’ is the number of generation of backcrossing. It is important to consider the status 
of gene to be transfer. Indeed, the process of dominant gene transfer is easy and fast whereas for a 
recessive gene it is slightly different because it is not expressed in the heterozygote (House, 1985). 
In this case, self-pollination is required at each generation in the segregation populations in order 
to identify the genotypes carrying the gene of interest. Moreover, Acquaah (2007) reported that an 
21 
 
University of Ghana http://ugspace.ug.edu.gh
additional step is needed after each backcross to produce an F2 generation in order to identify the 
recessive trait. If many genes control the character to be transfer, it is probably not possible to 
backcross for more than one or two generations before a new line must be selected by a pedigree 
breeding method (House, 1985). Moreover, continuous backcrossing becomes not possible because 
the expression of the trait becomes less or not measurable.  
The genetic advance in backcross breeding depends on several factors. The table 2.2 below 
describe the advantages and disadvantages of the backcross breeding method. 
 
 
 
 
 
 
 
  
22 
 
University of Ghana http://ugspace.ug.edu.gh
Table 2. 2. Advantages and disadvantages of the backcross breeding method. 
Advantages Disadvantages 
Does not have to be conducted in the environment in which the Not suitable for transferring quantitative traits. 
RP is adapted because all that is needed is to be able to identify The trait should be highly heritable and readily 
and select the target trait identifiable in each generation 
Does not require extensive testing: the new and the adapted Presence of undesirable linkages may prevent 
cultivars are similar except for the newly incorporated trait.  
the cultivar being improved from attaining the 
performance of the original recurrent parent 
It is possible to transfer two or more genes by simultaneous Recessive traits are more time consuming to 
selection among the progeny. This undertaking requires a larger transfer. 
population that would be necessary if two genes are transferred 
 
independently 
It is a conservative method, not permitting new recombination  
to occur 
Backcross breeding is repeatable. If the same parents are used,  
the same backcrossed cultivar can be recovered. 
It is useful for introgressing specific genes from wide crosses.  
 
It is applicable to both self-pollinated and cross pollinated  
species 
Source: Acquaah (2007) 
2.8. Bmr genes introgression 
Rooney (2000) reported that there are three basic colours pigmentation that occur throughout the 
leaves and stems of sorghum plants: purple, red and tan. These colours are developed on the leaves 
or culm where the plant has been damaged by insects, diseases or mechanical means. The colours 
are conditioned by two genes P and Q. Tan plants are genetically ppqq and ppQQ, purple plants 
are genetically PPqq and PPQQ (Ayyangar et al., 1933).  
23 
 
University of Ghana http://ugspace.ug.edu.gh
Morphologically inherited traits such as the bmr phenotype observed in maize, sorghum and pearl 
millet are of a great importance in plant breeding. Roy et al. (2004) affirmed that such traits are 
important from the point of identification of a cultivar. In maize, Barriere and Argillier (1993) 
reported that the bmr genes segregate in a simple Mendelian recessive character (3:1). Gupta 
(1995) also concluded that the bmr gene is controlled by a recessive allele and the segregation in 
fits 3:1 ratio in pearl millet. In sorghum, the same conclusions were drowning by Porter et al. 
(1978). 
2.9. SNP markers for bmr genes 
In sorghum, Porter et al. (1978) informed that there are at least four independent bmr loci 
exhibiting the bmr phenotype (bmr2, bmr6, bmr12 and bmr19). For the bmr6 phenotype, Saballos 
et al. (2009) revealed that the SNP mutation C-to-T transition mutation at position 2800 of the 
SbCAD2 genomic sequence. This particular nucleotide change creates a premature stop codon; the 
resulting protein contains 131 amino acids and lacks the nucleotide binding site (Saballos et al., 
2009). For the bmr12 phenotype, Sattler et al. (2012) identified from an EMS-mutagenized 
TILLING population, four missense mutations in the coding region, which changed evolutionarily 
conserved amino acids Ala71Val, Pro150Leu, Gly225Asp, and Gly325Ser while for the previously 
characterized bmr12 mutants all contain premature stop codons. The missense mutations found 
had all impact on enzyme activity, Klason lignin content, lignin subunit composition and 
saccharification yield (Sattler et al., 2012). For the bmr2 phenotype, Saballos et al. (2012) reported 
that a missense mutation (G to A within the genomic coding sequence at position +468) leads to 
an amino acid substitutions Gly111Asp and proved that the bmr2 gene encoded a 4-coumarate 
coenzyme A ligase (4CL), which catalyzes an early step in monolignol biosynthesis. A single allele 
represents the bmr19 locus at this time despite the large numbers of brown midrib mutants 
24 
 
University of Ghana http://ugspace.ug.edu.gh
originally isolated from sorghum (Saballos et al., 2008). For Sattler et al. (2010) this may indicate 
that either genetic saturation has not been achieved or alternatively, bmr19 is not a simple loss of 
function mutation. Furthermore, Sattler et al. (2014) reported that the bmr19 mutant is not publicly 
available (effectively reducing the available sorghum brown midrib mutants to a set of three 
independent loci: bmr2, bmr6, and bmr12). Moreover, for Saballos et al. (2008), the bmr19 appears 
to be of limited value for forage and bioenergy applications, because it did not significantly reduce 
lignin concentration and did not markedly alter lignin subunit composition.  
2.10. Dual purpose cereals: Farmers and traders’ preferences in SSA 
According to Jacob et al. (1997) many countries in semi-arid regions need dual purpose crops to 
help alleviate feed and food shortage. In Niger, farmers struggle to feed the animals during the 
long dry season with the poorly nutritional stored crop residues (Abdou, 2010). According to 
Reddy et al. (2004), dual purpose sorghum development requires effort on traits that affect yield 
and quality of forage and grain. In fact, several studies reported on the importance of residues yield 
in new varieties adoption for small scale farmers (Magnan et al., 2012; Kelly and 
Parthasarathy,1994). Indeed, for the majority of smallholders, crop residues from dual-purpose 
crops constitute 40-60% of total dry matter intake (Rao et Hall, 2003). In semi-arid regions, crop 
residues from cereal and leguminous crops, are by far the most important feed source available to 
farmers (Rao et Hall, 2003). For Miller and Stroup (2004), the potential of sorghum as dual purpose 
crop is enormous while Tunde and Augustine (2014) reported that farmers in two villages of Niger 
appreciate stovers qualities based on physical aspects. Chikuta et al., (2014) reported that farmers 
for dual purpose preferred a variety with high grain yield potential combined with high biomass 
and high sugar content in Zambia. 
  
25 
 
University of Ghana http://ugspace.ug.edu.gh
CHAPTER THREE 
3.0. SURVEY ON CROP RESIDUES FOR CATTLES FEEDING IN NIGER 
3.1. Introduction 
In Niger agro-pastoralism is the major food and feed production system. In this system, crop 
residues represent an important resource used for soil fertility management and livestock feeding. 
Valbuena et al. (2014) reported that in areas with low cereal intensity in SSA, crop residues 
constitute a vital dry season feed resource for livestock particularly when other resources are 
scarce. Witcombe (1996), reported that in marginal areas, most farmers have not adopted new 
cultivars because of limited access to improved seeds and perhaps recommended varieties did not 
have attributes such as high stover yield needed or did not appear as productive as expected. 
Furthermore, the CGIAR (2011) concluded that crop residues, especially stover and straw are 
increasingly important commodities that increase the overall value of dryland cereals. In Niger, 
the important role of cereal stovers in livestock feeding during dry season for small scale farmers 
is obvious particularly in the context of limited availability of natural pasture. However, the correct 
feeding of livestock for more productivity in terms of fattening, milk or field work is a challenge 
that most farmers face. This situation combined to the recurrent feed deficit generated another 
recent aspect in crop residues utilization. Indeed, Valbuena et al. (2012) informed that the 
increasing use of crop residues for purposes other than mulching implies that these have become 
a private good with an explicit economic value. Currently in Niger’s large cities, appearance of 
small shops specialized in forage trading is obvious everywhere on the edges of the roads and in 
the markets in an informal system. The main objective of this chapter was to appreciate well this 
growing market. 
26 
 
University of Ghana http://ugspace.ug.edu.gh
The specific objectives were to: 
- determine livestock feeding constraints, stover production and utilization in Niger during 
the dry season, 
- identify farmers’ and stover traders’ preferences for sorghum stover varieties. 
3.2. Material and methods 
The current survey was conducted on sorghum farmers and stover traders based on focus group 
discussion (FGD) and semi structured interviews (SSI). 
3.2.1. Farmers  
In collaboration with livestock extension service, three villages well known for their livestock 
rearing and sorghum cultivation in the country were randomly sampled. Extension service agents 
assisted in the sampling and gathering sorghum farmers with intensive experience beneficial to 
PRA survey. A FGD with an average of a group of twelve farmers was firstly conducted to have 
an overview on sorghum production, cattle rearing and feeding constraints. During the FGD, males 
and female groups were met separately to avoid men's negative influence on women’s free 
expression of their opinions. The different topics discussed during the FGD were evaluated by 
ranking, rating, listing or sorting by the group of farmers. After the FGD, a list of sorghum farmers 
rearing cattle in each village was recorded and 30 farmers per village were randomly sampled. The 
selected farmers were interviewed separately in a semi-structured interview based on a 
questionnaire structured on four major’ points: farmer identification, main activities, sorghum 
production and livestock rearing (constraints, opportunities, contribution in income, stover quality 
traits. 
 
27 
 
University of Ghana http://ugspace.ug.edu.gh
The following material was used during FGD: 
- Discussion guide line to address keys topics  
- Posters on sorghum life cycle for a better understanding  
- bmr sorghum pictures for illustrations 
- Forage sorghum pictures for illustrations 
- Exotic improved cow pictures  
- Tape recorder to collect oral information 
- GPS to reference geographical position of each village 
- Camera for pictures 
- Excel software for data analyses 
FGDs team was composed of: 1 facilitator and moderator: local extension agent; 1 Livestock 
scientist; 1 PhD candidate for FGD topic introduction and data collection.  
 
 
 
 
 
 
 
 
 
 
28 
 
University of Ghana http://ugspace.ug.edu.gh
Table 3. 1 FGD: Topics and evaluation methods. 
Topics Evaluation methods 
Importance of sorghum crop in  Sorghum 
the area compared to millet and                   Ranking Millet 
maize   Maize 
Sorghum production system  Ranking: association vs pure 
 Diseases 
                  Rating Drought 
Main sorghum production 
 Fertilizer constraints (grain/stover)  
 Soil 
  Others 
 Commercialization  
Domestic sorghum stem/forage                   Sorting Animal feeding 
utilization  Domestic utilization 
  Other 
 Sorghum 
Importance in animals feeding 
                 Rating Millet 
compared to millet and maize 
  Maize 
 
Livestock feeding constraints 
                  Listing 
Farmers’ empirical preferences  
for specific traits and attributes                   Listing 
of forage sorghum   
 Milking cow 
Which cow is fed during the dry                   Sorting Pregnant cow 
season?  Draught cow 
  Bullock 
 Stem chopped and put 
                   Rating container 
How is cow fed during the dry Chopped stems put on 
season?  ground 
Without chopping put 
  on ground 
 
 
29 
 
University of Ghana http://ugspace.ug.edu.gh
3.2.2. Stover traders 
A census of all the stover salesmen in each village and its surrounding was conducted. This census 
facilitated the collection of personal information (name, and cellphone number) and the place of 
every stover owner shop. A questionnaire was then conceived based on the identification, supply, 
methods of storage and the profitability of the activity to collect information using a semi 
structured interviews. 
3.3. Study areas 
3.3.1. Description 
- Soumarana: is a village in the South-East Niger next to Maradi located at the geographical 
position of 13°26’38’’ North and 6°54’11’’ East. The average rainfall is 400-500 mm per year. 
- Dolli: is a village located in the central Niger next to Tahoua region at the geographical 
position of 13°47’30’’ North and 5°15’00’’ East. The average rainfall is 400-500 mm per year. 
- Lossa: is a village in located in the southwestern Niger next to Tillabery region at the 
geographical position: 13°56’2’’ North and 1°34’29’’ East. The average rainfall is 500-600 mm 
per year. 
30 
 
University of Ghana http://ugspace.ug.edu.gh
 
Figure 3. 1. Locations of PRA villages. 
 
3.4. Results 
3.4.1. FGD Results 
Sorghum was the second cereal crop after pearl millet in all three villages. Livestock rearing was 
well practiced and integrated in crop farming. Farmers were well aware of livestock feeding 
constraints during a particular period (March-July) of the year. Indeed, they revealed feed shortage, 
stover poor quality, pasture unavailability which impact seriously animal and milk productivity, 
transport and farm management particularly for draughts (land preparation, sowing, and weed 
management) animals. Drought was reported as the main constraint for sorghum production in all 
three localities while, sorghum midge was cited as source of low grain production in Dolli and 
Soumarana. However, farmers were not aware of the real cause of sorghum midge causing low 
grain yield. One major information collected was that sorghum stover was important in livestock 
feeding. Indeed, at Dolli and Soumarana, it was the main feed whereas at Lossa millet stover came 
first. For stover, biomass was the most important attribute of preference for farmers.  
31 
 
University of Ghana http://ugspace.ug.edu.gh
Farmers had never experienced growing bmr sorghum and its attributes. This was confirmed 
through the exposition of brown midrib varieties’ posters. However, they noticed the existence of 
white and yellow midribs. 
Table 3. 2. Summary results of the FGDs 
Topics Results 
Importance of sorghum crop in the area compared to millet 2nd cereal crop after pearl millet in all the 
and maize localities 
Sorghum production system  Mostly sorghum was cultivated in pure 
Drought was the 1st at Lossa however in 
Dolli and Soumarana drought and disease 
Main sorghum production constraints (grain/stover)  
(sorghum midge) were the major 
constraints 
Livestock feeding was the main utilization 
Domestic sorghum stem/forage utilization 
of the stover 
Sorghum was 2nd behind millet stover at 
Importance in animals feeding compared to millet and 
Lossa, however sorghum stover was at the 
maize 
1st position at Dolli and Soumarana 
-Shortage of forage 
-Poor quality of forage 
Livestock feeding constraints 
-Cereal bran price 
-Pasture unavailability 
1st Biomass 
Farmers’ empirical preferences for specific traits and 
2nd palatability 
attributes of forage sorghum 
3th stay-green 
1st Milking cow 
2nd Bullock 
Which cow is fed during the dry season? 
 
  
Whole stem without chopping put on 
How is cow fed during the dry season? 
ground 
 
 
F 
32 
 
University of Ghana http://ugspace.ug.edu.gh
3.4.2. Results Semi Structured Interview (SSI) 
3.4.2.1. Farmers 
➢ Identification 
Ninety-one farmers were interviewed of which 93% were males and 7% were females. The 
traditional land tenure system (land belongs to extended family) was the most wide spread across 
the three villages (72%). (figure 3.2) 
There were 4 females at Soumarana and 2 females at Dolli. Their farm land tenure was dominated 
by the rent followed by the legally owned (obtained from a legal structure). No female farmer was 
met at Lossa in the western part of the country.  
On average each famer had 4.59 ha of land. 
 
Land tenure
80.0% 72.5%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
12.1% 12.1%
10.0% 2.2% 1.1%
0.0%
Traditinal Legally Lease Borrow Bought
Types
 
Figure 3. 2. Land tenure in the three villages. 
Traditional = land belongs to extended family; Legally owned (Obtained from a legal structure 
(administrative); Lease /Rent; Borrow = Land owned by a parent. 
 
 
33 
 
Frequencies
University of Ghana http://ugspace.ug.edu.gh
➢ Main activities 
Sixty-eight per cent of the investigated farmers were involved in other activities such as trading, 
or fattening livestock throughout the year to improve their income and 32% were just farmers. 
Trading was more developed at Dolli (80% of farmers) than the other two villages, Soumarana 
(71%) and Lossa (53%). 
➢ Sorghum production: 
All investigated farmers produced sorghum and reared livestock. In fact, Sorghum production and 
rearing livestock was a criterion of their selection. However, the survey revealed that they all 
produced sorghum for both grain and stover. Eighty-six per cent of interviewed farmers were 
favorable for adopting improved varieties while 46% bought improved seeds frequently. 
 
Figure 3. 3. Farmers’ sorghum seed supply sources at three locations surveyed 
 
 
34 
 
University of Ghana http://ugspace.ug.edu.gh
➢ Livestock rearing: A profitable livestock rearing in semi-arid regions involves the control 
of key strategies to overcome numerous constraints. 
 Stover quality trait: Three types of answers were given by farmers on stover quality: 
biomass (59.3%) was ranked first followed by juiciness (20.9%) and the stay-green (19.8%) comes 
which were ranked about the same. 
Lossa: Stover quality traits
50
45 43
40 37
35
30
25
20
20
15
10
5
0
Biomass Juiciness Stay-green
Traits
 
Figure 3. 4. Farmers’ perception on stover quality trait at Lossa 
Dolli: Stover quality traits
50 47
45
40
40
35
30
25
20
15 13
10
5
0
Biomass Juiciness Stay-green
Traits
 
Figure 3. 5. Farmers’ perception on stover quality trait at Dolli. 
35 
 
Frequencies (%) Frequencies (%)
University of Ghana http://ugspace.ug.edu.gh
Soumarana: Stover quality traits
100
87
90
80
70
60
50
40
30
20 13
10
0
0
Biomass Juiciness Stay-green
Traits
 
Figure 3. 6. Farmers’ perception on stover quality trait at Soumarana. 
 Milk production: Majority of farmers (59.3%) mentioned stover shortage as the main 
constraint for milk production, followed by malnourished cows (15.4%) and expensive cereal bran 
price (12.1%). However, seven per cent (7.7%) of farmers did not have constraints in milk 
production. Additional information showed that these farmers (7.7%) kept small size of herd (2 to 
5 cattle) which were kept in the village throughout the year and fed mostly industrial feed as 
complement in addition to the cereal bran for their diet. 
 
36 
 
Frequencies (%)
University of Ghana http://ugspace.ug.edu.gh
Milk production
70
59.3
60
50
40
30
20 15.412.1
7.7
10 4.4
1.1
0
Forage shortage Cereal ban Malnutrition of Shortage of No constraint Unavailability
price cattles forage and of concatrated
cereal ban feed
Constraints
 
Figure 3. 7. The major constraints of milk production in Lossa, Dolli and Soumarana. 
 Fattening: Sixty-nine per cent of the investigated farmers carried out this activity and 68% 
responded that it was an important source of income. Thirty-one per cent did not have a particular 
view and for 1% fattening is a thrift.  It is important to notice that those 31% with no view 
correspond to those who are not doing this activity. 
3.4.2.2. Stover traders 
 Twenty-four (24) stover salesmen were identified. Among them, 21 were from Konni the principal 
city close to Dolli (14 miles separate the 2 localities) and 3 were from Soumarana. No stover 
salesman was found at Lossa and its surroundings. All traders ran their business individually and 
have never received a particular training. Main types of stover marketed forage were from sorghum 
and pearl millet crops, cowpea hay all in bundles, groundnut haulms, cowpea pod shells and cereals 
brans in bags. 
 Supply: Farmers are the main suppliers for stover traders. Indeed 87.5% of traders 
purchased from farmers principally at Konni while at Soumanarana stover salesmen are producers. 
37 
 
Frequencies (%)
University of Ghana http://ugspace.ug.edu.gh
A bundle of 18 kg to 20 kg each purchased price was comprised between 250F-300FCFA ($0.40-
0.48) at the harvest time (sorghum) which was their main period of acquisition. Standard qualities 
requested when purchasing were good biomass (42%), stay-green+biomass (21%), stay-green 
(17%) and stover with enough leaves (21%). 
 
Stover standards qualities
45 42
40
35
30
25 21 20
20 17
15
10
5
0
Good biomass Stay-green+biomass Stay-green Enough leaves
Stover traits
 
Figure 3. 8. Purchased stover quality requirement. 
 Storage: Except two salesmen at Soumarana who stored on shade their stover, the 
remaining piled up their stover outside at their business place stover bundles. The storage duration 
was not determined since selling can happen any time, but in case of slow selling it can reach 12 
months. In such situation, sellers complained about rodents, mites, thieves, fire or rains. 
Unfortunately, 50% of the salesmen do not have any solution for those constraints management 
whereas 29% utilized chemicals to control mites and rodents’ attacks. The remaining 21% affirmed 
that vigilant and regular inspection of stock constitute their solution. 
38 
 
Frequencies (%)
University of Ghana http://ugspace.ug.edu.gh
 Sales: In majority, salesmen affirmed peak selling time runs from April to July. During 
this particular period, the stover bundle price was comprised between 750-1000F CFA ($1.20-
1.61) and is mostly established according to its demand and availability. The standard qualities 
wished were the presence of the leaves, good biomass and the general quality of the stover. 
For 75% of the salesmen, stover business is a source of income and 25% were conducting to ensure 
their family daily expenditure. In fact, 75% save money for a continuous acquisition of stover. 
3.5. Discussion 
Farmer identification provided a useful stating point and served also to locate farmer establishment 
(experience in sorghum cultivation, farm equipment …). The results on land tenure in the three 
villages in Niger showed that land tenure was mainly traditional (72.5%); in the other hand, land 
belonged to extended family, distributed between members or not under the responsibility of the 
head of the family. Land tenure revealed therefore a capital importance as it determines access to 
resources such as crop residues (FAO, 2014). In fact, Yazon and Peter (2003) reported that out of 
1.267.000 km2 in Niger, arable land occupies only 3.94% and agriculture is mainly practiced in 
the southern part of the country where rainfall rarely exceeds 800 mm. Further, concerning land 
distribution and exploitation, Ngaido (1996) reported that in a village of the country the following 
types of lands: individual lands, family lands and village common lands also known as chieftaincy 
lands. The results of this survey are consistent with Terraciano (1998) who explained that the land 
inheritance is through a patrilineal system. However, in the particular case of women land tenure, 
it was observed that 83% of their farm land was owned by a parent and only 17% belongs to them. 
In addition, women farm size was between 2 to 3 ha with an acceptable fertility. The overall results 
indicated that the average farm size was 4.59 ha. Waha et al. (2016) revealed that defining the 
farm size in Africa is country-specific and depends on the self-assessment of the households and 
39 
 
University of Ghana http://ugspace.ug.edu.gh
affirmed that the average farmed land area of small farms in Niger is 5.5 ha or less. With regards 
to land fertility, 85% of farmers consider their farms to be acceptable to good for sorghum 
production. This can be explained by the fact that sorghum is generally cultivated on clay soil 
opposed to millet which is cultivation on sandy soil.  
The majority of interviewed farmers (69.2%) were practicing fattening to raise their income and 
for 75.8% of farmers, livestock contribution into their income was important, which is in harmony 
with Kamuanga et al. (2008) and Tunde et al. (2015). In addition, Tunde and Augustine (2014) 
confirmed that livestock are kept by most farmers to complement crop activities, because 
ruminants provide fertilizer for crop production and are also valued as assets that can be readily 
liquidated to meet household and farm financial obligations. According to Kamuanga et al. (2008), 
livestock rearing is one of the main economic activities on which the poorest population depends 
for food and income. The profitability of fattening depends however on a good control of 
investments in animal feeding. For this purpose, in addition to crop residues, cereal brans and other 
industrial concentrated feed played an important role. Findings reports that only 41.8% of farmers 
utilized cereals brans and concentrate as supplement due to their high cost which were also 
mentioned as constraints for milk production. Tunde et al. (2015) found that many rural 
households frequently used crop residues from their farm and brans from processed grains to 
reduce feed cost. In this view, the Food and Agriculture Organization (2014) highlighted the 
excessive price of wheat bran in ECOWAS countries and Niger seems to be the biggest for maize 
bran imports. At a single village level, the study revealed that fattening was more practiced at 
Soumarana (84%) than at Lossa (63%) and Dolli (60%), the proximity of Nigeria republic can 
explain this situation. Indeed, Soumarana is closer to Nigeria border which is one of the main 
destinations for livestock exportation from Niger. Guilbert et al. (2009) estimated that from the 
40 
 
University of Ghana http://ugspace.ug.edu.gh
Eastern network, 20% of livestock at international market in Maiduguri is thought to come from 
Niger and from the Western network livestock are collected from Mali, Niger, Burkina Faso and 
Benin to the international market in Kano although it is difficult to assess import flows of live 
animals from Niger, Chad or Mali due to the porosity of borders. 
In general, the principal sources of household income in Niger are still the sale of crop, other 
income generating activities from migrants’ remittances, livestock and byproducts sales (Rep. of 
Niger, 2002). Farming system is oriented on subsistence based characterized by low inputs 
utilization and not very productive material (equipment) leading thus to low productivity. 
Livestock holding, well integrated in farmer activities, contributes to income improvement and 
diversification. Among the production constraints, farmers explained that drought was the major 
limitation followed by sorghum midge at Soumarana and Dolli. The major biotic stresses in Niger 
include sorghum midge (Stenodiplosis sorghicola), Striga and various diseases and drought for 
abiotic stress. Hamidou et al. (2018) reported that sorghum midge is one of the most limiting 
factors of sorghum production and most farmers in major sorghum growing areas of Niger still 
rely on the low yielding, midge-susceptible local sorghum varieties. In the PRA study area, farmers 
considered a night wind coming from East as the cause of sorghum midge. On this subject, Kadi 
Kadi (1993) reported that farmers in Maradi and Konni thought a night wind caused empty glumes 
and reduced sorghum production. The World Bank (2013) also reported that drought was the 
principal risk for farmers in Niger and the country has experienced seven droughts between 1980-
2010 with adverse impact on national agriculture production in agreement with investigated 
farmers’ views included in this study. 
Concerning improved seeds, 46% of investigated farmers affirmed frequent utilization, which is 
in disagreement with Olembo et al. (2010) who reported in their previous improved sorghum 
41 
 
University of Ghana http://ugspace.ug.edu.gh
varieties utilization in Niger was about 17.10%. This high level of improved seed utilization can 
be explain by the proximity of research farms, NGOs activities in the areas of the study but also 
the released varieties are well adapted with farmer preference traits. Indeed, the INRAN research 
farms are about at: less than three km for Lossa, fourteen km for Dolli and less than twenty km for 
Soumarana. 
Investigated farmers produced sorghum for dual purpose. Dual purpose lines are very important 
for the farmers in Niger because they care as much about stover yield as feed as they are for the 
grain for food. Current PRA findings are in harmony with those of Thornton et al. (2003) who 
reported that mixed crop-livestock systems is well practiced and continues to be a substantial area 
of interest because farmers can produce both for human and livestock consumption. Singh et al. 
(2004) reported that in West Africa, farmers do not grow food and fodder separately and therefore 
crop residues are the major source of feed for livestock. Furthermore, Kelly and Parthasarathy 
(1994) reported that yield and stover quality of sorghum and millet are important choice standards 
in new variety adoption by farmers in semi-arid India. Despite harsh environmental conditions in 
Niger, sorghum shows a good adaptation and plays a strategic role in the search for food and feed 
security. According to JAICAF (2009), cultivars of sorghum in Niger are mostly traditional 
varieties and the production quantity accounts for one quarter of that for principal cereals. In 
certain areas however it surpasses pearl millet in importance as staple food. 
During the study, farmers used biomass, the juiciness and stay-green as main parameters to 
evaluate the quality of cereal crop residues. Tunde and Augustine (2014) reported that in a previous 
study in two villages (Milli and Gourdjia) of Maradi region that farmers described 5 different 
physical parameters (coloration, texture, odor, age at harvest and animal behavior) to determine 
feed quality. In general, concerning the quality of the purchased stover, Blummel et al. (2013) 
42 
 
University of Ghana http://ugspace.ug.edu.gh
considered that crop residues, fodders from different crops are not considered the same by farmers 
and traders. Further, surveys of crop residues, fodder trading in SSA and India showed that fodder 
traders and customers were well aware of differing fodder quality from crop residues of different 
crops. Findings from this survey support the conclusions by the above authors. 
During the survey in the 3 localities, the average of milk production per day was: 48 liters at 
Soumarana, 138 liters at Dolli and 216.5 liters at Lossa. The low level of milk production at 
Soumarana can be explain by the severe feed shortage (94% of famers’ complained about feed 
shortage). Abdou (2010) reported feed deficiency and crop residues cannot satisfy the animal 
maintenance in Niger between April and June. At farmer exploitation level, the overall average of 
milk production was 4.42 liters, the maximum was 60 liters and the minimum was 0.5 liter. It is 
important to notice that only milk cow production was considered and 56% of the investigated 
farmers did not sell milk unlike the remaining 44%. The main reason was that the quantity of milk 
produced was not enough (42.9%) and the milk production was on seasonal basis. Indeed, milk is 
more produced when cows calving correspond to the rainy season thus the availability of feed. The 
major constraints for milk production was forage shortage (59.3%) and malnutrition of cattle 
(15.4%) according to farmers, in addition most of their milking cows were not improved breeds. 
This is consistent with Singh et al. (2004); Diester and Hassane (2006) who reported in a previous 
study that Nigerien improved cattle such as the Azawak or the Kouri rank among the best milk 
cows in West Africa; however the actual production is limited by malnutrition. Another reason for 
low milk production was the poor animal health maintenance. Health and economic constraints 
were mentioned by Diester and Hassane (2006) as part of low performance in term of milk 
production in Niger. Indeed, at Lossa and Dolli animal veterinary service was lacking. However, 
in case of emergency, farmers have to contact the closest livestock service and pay for the cost of 
43 
 
University of Ghana http://ugspace.ug.edu.gh
drugs and transport. The same report was observed by Tunde and Augustine (2014) in Gourdjia 
(Maradi region). At Dolli, the average milk production per day was 4.6 liters and 90% of farmers 
mentioned stover shortage as a constraint whereas the average milk production per day was 7.21 
liters with lack of forage for 73% of farmers at Lossa. Tunde and Augustine (2014) found a high 
feed scarcity from April to June in two villages in Maradi region, in general. UNDP (2006) 
reported that cattle breeding activity is seriously affected by the decrease of fodder production and 
the reduction of pastoral areas as a result of climatic events. In addition, the availability of rice 
straw at Lossa may favor farmers to feed their livestock. Beside the fact that livestock products is 
bringing important nutritional benefits to large segments of population from developing countries 
(Muehlhoff et al., 2013), milk production offers an important source of cash income to many of 
the over 200 million poor livestock keepers estimated in developing regions (Pica-Ciamarra et al., 
2011). However, Zezza et al. (2014) affirmed that at the household level data and studies on the 
role of milk production for human nutrition and livelihoods are severely hampered by the difficulty 
of producing reliable estimates of milk production in small-scale livestock production system.  
In addition to this, farmers mentioned several uses of draught animal in their farming activities 
(farm work: ploughing, fertilizer, transport…) as advantage gains from their livestock, this result 
provide confirmatory evidence from previous Mohamed-Saleem and Von Kaufmann (1988) 
survey on farmers in Nigeria.  
The results of the survey revealed also that farmers are stover market suppliers. Berazneva and 
Dyson (2013) supposed that in the context of African smallholder agriculture, crop residues have 
the most valuable application among on-farm resources. Socio economic factors such as the short 
term gain might influence farmers’ decisions concerning this resource. In a previous study on crop 
residues trade in SSA and South Asia, Valbuena et al. (2011) reported that the need to cover their 
44 
 
University of Ghana http://ugspace.ug.edu.gh
own food requirements and household expenses pushes smallholder farmers to favor practices with 
positive returns in the short term. In the context of growing demand of crop residues in Niger, 
cereal stover represents a source of emergency revenue and can play an important role for short-
term benefits for farmers. According to CGIAR (2011), the increasing value of stover has been a 
prominent trend in Asia and stover markets are emerging in drier, more densely populated areas 
of West Africa. Further, as the demand for livestock and livestock products increases, so too will 
the importance of fodder and feed, therefore dryland cereals will thus focus on increasing the 
quantity and quality of stover and straw as well as grain. 
 During the dry season, crop residues were important sources of diets used to feed livestock in all 
three villages. However, in terms of ranking, sorghum residues constituted the first choice in Dolli 
and Soumarana and second at Lossa. This is in harmony with Tesfaye (1998) who reported that 
sorghum stover was the major source of dry fodder for urban and peri-urban dairy production in 
Hyderabad (India). For 87.5% of traders, farmers are the suppliers for stover particularly at the 
harvest time, confirming Mohammad (2008) findings. Indeed, he affirmed that un-chopped stover 
purchased by traders from small cities are brought to Hyderabad and sold in informal fodder 
market. Improving storage methods can maintain the quality of the stover from harvest to feeding 
time during dry season. Unfortunately, 92% of the local traders piled up their stover outside with 
no respect to conservation and storage techniques. Mohammad (2008) confirmed that for different 
feed resources, trade and marketing, conservation, storage and processing technologies to mitigate 
spatial and temporal imbalances in feed supply and demand is not well understood in East Africa 
(Ethiopia and Tanzania) and South Asia (India and Bangladesh). The same constraint is reported 
from the current survey. 
45 
 
University of Ghana http://ugspace.ug.edu.gh
Stover trading is a source of livelihood for many farmers and traders in Konni and Soumarana. 
This is in agreement with Tesfaye (1998) who affirmed that stover selling, transporting, trading 
and its use in dairy production support the livelihood of many people from producers (farmers) to 
the end users in Hyderabad (India). Grunert (2005), reported that in Hyderabad buyers judge 
quality based on experience then use that knowledge in purchase decisions and traders infer quality 
differences based on intensity of demand for different types of stover. For Mohammad (2008) 
roughage feeds are traded in a variety of forms and generally there is no formally defined quality 
or standard though local buyers and sellers use informal indicators to differentiate quality to some 
extent. Further in Bangladesh, traders make varying sizes of small bundles of rice straw for sale 
but are not weighed to determine actual weight and quality is mainly differentiated based on visual 
observations.  
 
 
 
 
 
 
 
 
 
 
 
  
46 
 
University of Ghana http://ugspace.ug.edu.gh
3.6. Conclusion 
Small scale farmers with mixed crop-livestock dominate the production system in all the 3 
localities. Throughout the survey farmers’ preference for dual purpose sorghum variety was 
evident. They highlighted the significant position of sorghum stover in their livestock feeding. 
Farmers also explained the importance of livestock contribution in their household income and 
commodities. In this context, the long dry season (November to May) constitutes an opportunity 
for practicing additional activities such as fattening and trading. However, those activities were 
seriously impacted by the shortage and mere quality of feed from their sorghum landraces stover. 
Drought is one of the major cause of low stover production that affect animal productivity like 
milk production. Indeed, in all the localities the low milk production reduces the opportunity of 
milk marketing. This restriction was therefore a limit of a sustainable long term revenue gain for 
farmers. 
Farmers and stover traders appreciate essentially sorghum stover quality at harvest time based on 
physical aspect. High biomass and stay-green were important choice standards. The integration of 
those key information in breeding program might favor the rate of adoption of new sorghum dual 
purpose based on the bmr attributes. 
Farmers rely on crop residues to feed their livestock during the dry season. During this particular 
moment, shortage of feed constitutes one of the main constraints for livestock rearing. As a result, 
livestock face harsh feeding conditions. This situation leads to a growing demand for stover. 
However, this demand is an opportunity that increases the value to crop residues particularly 
sorghum stover due to its high preference for livestock feeding (cattle and draught feeding) in 
Niger. Indeed, this survey showed that crop residues trading is a growing and profitable business 
47 
 
University of Ghana http://ugspace.ug.edu.gh
in the Southern part of the country. Un-chopped fodder was the main handling method throughout 
the cereal stover chain from the farm to the end user.  Moreover, with the locally growing stover 
market, the economic value of crop residues as feed and/or source of income might have 
implications on farmers’ variety adoption as they are the main suppliers of stover to traders. 
Therefore, the improvement could lead to a substantial increase of farmers’ income since cereal 
residues are the major feed for their livestock and also farmers are the main sources of supply to 
traders in Niger.  
 
 
 
 
 
 
 
 
 
 
 
 
  
48 
 
University of Ghana http://ugspace.ug.edu.gh
CHAPTER FOUR 
4.0. INTROGRESSION OF BROWN MIDRIB GENES INTO TWO NIGERIEN 
SORGHUM VARIETIES 
4.1. Introduction 
The first sorghum breeding efforts in Niger began with the creation of IRAT (Institut des 
Recherches Agronomiques Tropicales et des Cultures Vivrières) in 1964. The varieties Bagoba 
(Durra), Jan Jare (Caudatum), Gourma (Guinea caudatum), El-Mota (Guinea-caudatum) and 
Babbadiah (Durra) were improved by mass selection and released to farmers (Clark and Kitch, 
1982). INRAN (the National Agricultural Research Institute for Niger) was created in 1975 with 
the mandate of research on crops and animals in Niger. According to Adamou et al. (1985), the 
objectives of sorghum improvement is the development of varieties with high and stable yield, 
good grain quality, and resistance to drought, insects and diseases. In fact, in rural areas of Niger 
sorghum is a staple food. In addition, sorghum residues also constitute an important source of 
forage for livestock feeding. However, sorghum residues like other cereal residues are known to 
have poor feeding value due to their high lignin content. Indeed, Wilson and Kennedy (1996) 
revealed that the presence of lignin reduces the forage quality by hindering access to digestible 
energy and nutrients (i.e. proteins and minerals). Lignin is cross-linked to the cell wall matrix 
whereas an important aspect of a forage crop is its feeding value which involves a number of 
criteria affecting both plants and animals. However, Akinola et al. (2015) affirmed that the cereal 
crop residues’ potential as livestock feed is enormous if particularly if appropriate methods of 
improving their nutritive value are used.  Jung and Fahey (1983) reported that bmr plants have 
lignin which is less polymerized and contain less phenolic monomers that can affect digestion. 
49 
 
University of Ghana http://ugspace.ug.edu.gh
Ayyangar et al. (1936) explained that in sorghum, the D gene which affect the leaf midribs also 
controls stem juiciness, with pithy (D) being dominant to juicy (d) stems. According to Kumari et 
al. (2017) green color of leaf midrib in sorghum is an indicator of sweetness whereas brown colour 
indicates reduced level of enzyme resistant polymer 'lignin' in plants. According to Rooney (2000) 
a white (and dry) midrib is conditioned by a dominant allele at the D locus, while recessive 
genotypes have green and juicy midribs. 
Porter et al. (1978) produced 19 bmr sorghum mutants by chemically treating seed from two grain 
sorghum lines. These mutants were numbered bmr 1 to bmr 19. Porter et al. (1978) suggested that 
the bmr6, bmr12, and bmr18 should be selected for further evaluation. However, according to 
McCollum et al. (2005) the bmr6 and bmr12 genotypes are more prevalent than the bmr18. Adding 
bmr trait to local cultivars and newly developed lines will be more sustainable and will bring a 
significant impact on livestock productivity in countries like Niger. However, exotics bmr6 and 
bmr12 sources were less adapted into the Nigerien typical sahelian environment (Diakité et al., 
2018). Two well adapted sorghum varieties (Sepon82, and El mota) were selected as suitable 
candidates for their stover quality improvement. In the Nigerien national plants catalogue Sepon82 
is an elite variety for grain yield. In addition, it has a good stover potential and aptitude to stay 
green at the physiological maturity. El mota is a local variety widely cropped by farmers for its 
earliness, adaptation to erratic rainfall, poor soil fertility and its high yield potential (Kapran et al., 
2007). 
The main objective of this chapter was develop bmr breeding population with Sepon82, and El 
mota as recurrent parents using the backcross breeding method. 
 
 
50 
 
University of Ghana http://ugspace.ug.edu.gh
The specific objectives were to: 
i. introgress bmr6 and bmr12 genes to local sorghum varieties, 
ii. identify and advance superior genotypes with bmr trait in the breeding populations. 
4.2. Material and methods 
4.2.1. Plant materials 
The plant materials were composed of Sepon82, and El mota as recurrent parents (RP). The bmr 
donor parents were Redlanbmr6, Wheatlandbmr12 and Tx630bmr12 were used as donor parents 
(DP). Those donor parents were from Purdue University Indianapolis (USA). Table 4.1 
summarizes the agronomic characteristics of the different plant materials.  
Table 4. 1.  Summary of agronomic characteristics of the different parental lines. 
  Plant type Grain Yield (t/ha) Biomass Yield Lodging Midrib color 
Recurrent Parents      
Sepon82 Tan 2-2.5 Good Resistant Dull-green 
El mota Anthocyanin 1-1.5 Acceptable Less resistant White 
Donor Parents      
Redlanbmr6 Anthocyanin 1.1-1.4 Low Resistant Brown 
Wheatlandbmr12 Anthocyanin 1-1.5 Low Resistant Brown 
Tx630bmr12 Tan 1-1.5 Low Resistant Brown 
 
Source: République du Niger (2012) 
51 
 
University of Ghana http://ugspace.ug.edu.gh
 
Figure 4. 1. The different types of sorghum midribs involved in this study 
White (El mota: left), brown (Redlanbmr6: centre) and dull green (Sepon82: right). 
4.2.2. Development of bmr population 
bmr genes introgression for the breeding population development activities took place in Mali 
from January 2016 until May 2017 at Sotuba Research Station. The geographic coordinates of 
Sotuba: 12° 39’ N; 07° 56’ O at 320 m altitude. 
The crossing blocks nurseries were prepared and managed for a good growth of seedlings. The 
Standard agronomic practices were implemented for the growth of the plants. 
Crossing blocks were established in early January 2016 off season under drip irrigation system. 
As sorghum is a short-day plant, its cultivation in January, when photoperiod is short, causes early 
flowering thus allowing crossing synchronization between the two parents. Nevertheless, two 
planting dates (27/01/2016 and 9/02/2016) for the donor parents (DP) and one date of planting for 
the recurrent parents (RP) (27/01/2016) were used to ensure flowering synchronization required 
52 
 
University of Ghana http://ugspace.ug.edu.gh
for crossings. Furthermore, too early parents were pruned during vegetative period to allow new 
shoots emergence for flowering synchronization between parents. There were 10 hills planted for 
expected 10 plants per DP or RP line. 
At flowering stage (March 2016: period of the high temperature at Sotuba), RPs were hand 
emasculated using tweezers/forceps. The hand emasculation technique using tweezers was chosen 
to minimize the amount of seedlings from self-fertilization in F1 plants.  Fifty to sixty flowers of 
each RP per cross were chosen for each crossing. Flowers from the top of each panicle were 
emasculated. Anthers were removed in each female and male flower. Two to three days after 
emasculation the recurrent parents were dusted with donor parent pollen. Flowers from the bottom 
of the same female panicle were bagged in order to produce identical recurrent parents’ seeds. 
Other flowers on the same female plant were completely discarded to avoid seed contamination. 
Likewise, each donor parent was crossed to each recurrent parent to produce F1 seed for each 
population. Mature crosses were harvested and kept separately. Donor parent plant seed was also 
self-pollinated. RP and F1 seeds were threshed separately to avoid any seed mixtures and stored 
at 4°C.  
F1s seeds from all successful crosses and their respective parental line seeds were treated with a 
Caïman Rouge P (Permethrine 25 g/kg + Thirame 250 g/kg) an insecticide fungicide, pre-
germinated in petri-dishes, and then were transferred in plastic pots containing 1 kg of compost 
for BC1F1 seed production. After 20 days period, F1 were transplanted in one row of 3 m. Parental 
seedlings were planted next to F1 for conformity control. Heterosis and presence of anthocyanin 
were used to differentiate F1 from their parental phenotypes. Hybrid vigour from crosses is easy 
to distinguish from parental lines. Anthocyanin (purple plant) is dominant over tan plant. If male 
parent is purple and the female parent tan, then the F1 derived from these two parents must be 
53 
 
University of Ghana http://ugspace.ug.edu.gh
purple. Conversely, if F1 is tan plants, this is considered a self, therefore discarded. These two 
techniques allowed the identification of few F1 plants. From F1 deriving from two purple or tan 
plants, segregation of bmr plants in F2 or BC1F2 was a sure mean of verification of true F1 cross. 
The identified F1s were then used as pollen donor on the emasculated top of RP panicles while the 
bottom of the RP panicles were self-pollinated. This method allowed BC1F1s seeds production 
simultaneously with pure RP seeds while by self-pollinating the F1s’ panicles F2 seeds were 
produced. 
BC1F1 seeds obtained during 2016 cropping season crosses between F1s and recurrent parents 
were planted for segregation study and generation advancement to obtain BC1F2.  F2 seeds were 
also planted in other nurseries for bmr trait segregation study and F3 seeds production. Each 
BC1F1 and F2 plant seeds were germinated in petri dishes (21-31 October), transplanted in plastic 
pots containing 1 kg of compost when primary root appeared, then transplanted in nursery field 
for BC1F2 plant identification. As the RP (El mota) population heights were heterogeneous, each 
crossing within this population crosses was kept separate for population identity. Harvest was done 
in early January.  
At harvest, each bmr BC1F2 panicle and F3 panicles were threshed separately. Seeds were pre-
germinated in petri-dish first, then transferred in pots after the appearance of their primary root 
and finally transplanted in the field in late January to early February. Photoperiod sensitive 
sorghum cultivation in long days (starting March) delays flowering until September avoiding 
harvest in May or early June. To speed up panicle initiation, seedlings after 21 days of 
transplantation in open field, were covered with cages every day from 5 pm to 8 am, during 30 
days. This artificial shortening of day length induces reproductive phase and seed for day length 
sensitive plants. The cages comprised of black plastic layer under each white cotton tissue for 
54 
 
University of Ghana http://ugspace.ug.edu.gh
complete shade to cover each sorghum plot. All panicles were self-pollinated using appropriate 
paper bags during all generation advancement. Figure 4.3. below shows the general breeding 
scheme for the bmr genes introgression. 
 
 
 
Figure 4. 2. The general sheme for bmr genes introgression. (RP=Recurrent parent; DP=Donor 
parent).  
 
All BC1F1 were planted and non bmr segregating BC1F2 populations were discarded. 
 
 
 
 
 
55 
 
University of Ghana http://ugspace.ug.edu.gh
4.3. Data analysis 
Data for bmr segregating phenotypes were collected in the field at BC1F2 and F2 generation for 
Chi-square (X2) test on every breeding population for the 3:1 segregation ratio in order to select 
the genotypes with the bmr genes incorporated. 
The following formula was used for the Chi-square test. 
 
𝑑2
𝑋2 = ∑( ) 
𝑒
Where X2= Chi-square value; ∑ for sum; d=observed value-expected value and e=expected value. 
4.4. Results 
A total of thirty-four (34) seeds were obtained as possible F1s (Table 4.2) in April 2016. All 
progenies were examined as follows and out of twenty-six seedlings, eight (8) F1s were identified. 
 Sepon82 (tan plant) x Redlan bmr6 (purple, anthocyanin): among the nine (9) F1s plants, 
only 1 plant (F1-1-3) was anthocyanin, therefore it was considered true F1 plant. The remaining 
seedlings were tan plants and were discarded from the breeding populations’ development. 
 Sepon82 (tan plant) x Wheatland bmr12 (purple, anthocyanin plant):  two (2) F1 plants 
were anthocyanin, thus were considered as F1s. 
 El mota (purple, anthocyanin) x Tx630 bmr12 (tan plant): twelve (12) F1 seedlings were 
generated and among them 3 were selected as F1s because of the heterosis observed. 
 El mota (purple anthocyanin) x Redlan bmr6 (purple, anthocyanin): both parental lines 
were anthocyanin. 3 F1 seedlings were produced but 2 were selected as F1s based on their visible 
heterosis effect. 
 
56 
 
University of Ghana http://ugspace.ug.edu.gh
Table 4. 2. Quantities of F1 seeds and seedlings generated per cross. 
Crosses F1 seeds harvested Seedlings generated True F1 
El mota x Tx630bmr12 14 12 3 
El mota x Wheatlandbmr12 0 0 0 
El mota x Redlanbmr6 4 3 2 
Sepon82 x Wheatlandbmr12 2 2 2 
Sepon82 x Redlanbmr6 14 9 1 
Sepon82 x Tx630bmr12 0 0 0 
Total 34 26 8 
 
 BC1F1 populations development 
Table 4.3. summarizes the quantity of seeds harvested in the crosses between the selected F1 (bmr 
donor) and the recurrent parents. 
Table 4. 3. Quantities of BC1F1 seeds produced per population. 
Population BC1F1 Seeds 
El mota//El mota/Tx630bmr12 1-1 15 
El mota//El mota/Tx630bmr12 1-2 2 
El mota//El mota/Tx630bmr12 1-3 31 
El mota//El mota/Tx630bmr12 1-4 2 
Sepon82//Sepon82/Wheatlandbmr12 2-1 36 
Sepon82//Sepon82/Wheatlandbmr12 2-2 12 
Sepon82//Sepon82/Redlanbmr6 3-1 3 
Sepon82//Sepon82/Redlanbmr6 3-2 15 
Total 116 
 
In the F2 populations, the bmr genes segregation gives the following results summarized in Table 
4.4.  
 
 
 
 
57 
 
University of Ghana http://ugspace.ug.edu.gh
Table 4. 4. Segregation in the F2 families. 
Population Phenotype Observed Expected Cal. X2 X2 Critical value 
 Dull-green 13 14 3.84 
Sepon82//Sepon82/Redlanbmr6 Brown 6 5 0.271  
Total   19 19     
 Dull-green 14 15 3.84 
Sepon82//Sepon82/Wheatlandbmr12 Brown 6 5 0.266  
Total   20 20     
 White 15 15  
El mota//El mota/Tx630bmr12 Brown 5 5 0 3.84 
Total   20 20     
For a test with 1 degree of freedom, the "critical" value of the chi-square statistic is 3.84: when X2 > to the 
critical value, then the data did not fit the X2 model or deviated from the expected value. Observed: number 
of bmr plants counted in the field out of a total of 20 plants. Expected: number of bmr plants derived from 
the single dominant gene segregation. 
 Bmr trait segregation study in BC1F2 populations 
In the BC1F2 populations, the bmr genes segregation gave the following results summarized in 
Table 4.5.  
Table 4. 5. Segregation in the BC1F2 families. 
Population Phenotype Observed Expected Cal. X2 X2 Critical value 
 Dull-green 42 45 3.84 
Sepon82//Sepon82/Redlanbmr6 Brown 18 15 0.8  
Total   60 60     
 Dull-green 47 45 3.84 
Sepon82//Sepon82/Wheatlandbmr12 Brown 13 15 0.35  
Total   60 60     
 White 43 45  
El mota//El mota/Tx630bmr12 Brown 17 15 0.35 3.84 
Total   60 60     
For a test with 1 degree of freedom, the "critical" value of the chi-square statistic is 3.84. 
 
  
58 
 
University of Ghana http://ugspace.ug.edu.gh
 4.5. Discussion 
Tan and anthocyanin sorghum plants types were used for the genetic transmission of the bmr trait 
in order to produce F1 plants. The morphological traits associated with distinct phenotype are of 
great importance in plant selection. In this view, Rooney (2001) stated that F1 hybrid plants must 
be identified by the breeder on some specific phenotypic or genetic basis. Moreover, the author 
concluded that this is typically accomplished by using a simply inherited phenotypic trait or 
heterosis that occurs between parents of different origin. In the F1 progenies derived from crosses 
involving tan and anthocyanin plants in this study anthocyanin marker was observed. This result 
showed the dominance of anthocyanin over tan character. This finding is in agreement with 
Nagaraja et al. (2008). In addition, anthocyanin marker is of great importance for sorghum biomass 
utilization in livestock feeding. To this end, Porter et al. (1978) stated that the anthocyanin 
pigments of sorghum leaf sheath have been found to be associated with economically important 
traits such as increased fodder quality. After crossing the parental lines (recurrent and donor 
parents), thirty-four (34) F1 seeds were harvested from six (6) crosses. This low level of success 
of the crosses can be due to the high temperatures occurring at this period of the year (almost 
45°C). In this view, House (1985) demonstrated that crossing at a temperature of 40°C (with low 
humidity) usually results in failure of seed set. Regarding the low success of crosses, only eight 
(8) true F1s over twenty-six (26) seedlings, breakage of anthers and pollen contamination on 
tweezers tips during the emasculation process can be responsible. This hypothesis was supported 
by House (1985) who described prevention of anther rupture and pollen sticking to the 
emasculation tool. In addition, Schertz and Dalton (1980) reported that the success of hybridization 
varies with personal skill, and the amount of injuries sustained by the floret part during the 
emasculation.  
59 
 
University of Ghana http://ugspace.ug.edu.gh
The IBPGR and ICRISAT (1993) produced a useful descriptor of many morphological traits for 
sorghum. For the leaf midrib colour; the white, dull green, yellow, brown and purple were 
considered. Further Acquaah (2007) showed that sorghum midrib colour was controlled by a single 
dominant (D) gene. For the particular case of the bmr type, Bout and Vermerriss (2003) revealed 
that sorghum brown midrib colour is controlled by single recessive gene. Following this, Nagaraja 
et al. (2008) reported the dominance of white midrib over brown midrib colour in F1 population. 
Further those authors showed in the following F2 segregating population, a good fit of the ratio 
3:1 revealing bmr trait as single recessive gene. The same report was observed in nurseries at 
BC1F1 and F1 levels in this study. Indeed, all the individual plants in every breeding population 
exhibited the parental midrib colour. Those results are also consistent with the basic 3:1 Mendelian 
dominant and recessive phenotypic traits inheritance ratio for a single gene. House (1985) too 
stated that a recessive trait is not expressed in F1s. Later with the advancement to BC1F2 and F2 
breeding populations, segregation for the bmr trait was observed. Results on segregation patterns 
in this study confirmed the recessive inheritance of the bmr genes as concluded by earlier authors 
(Akira et al., 2012) on recessive mutation in a single locus. Based on this information, Chi square 
test was performed to verify the segregation ratios of bmr trait in all the breeding populations. The 
results were in harmony with the following authors Acquaah (2007); Bout and Vermerriss (2003); 
Nagaraja et al. (2008). Therefore, in breeding population advancement each identified bmr plant 
in every breeding population was self-pollinated.  
 
 
 
60 
 
University of Ghana http://ugspace.ug.edu.gh
4.6. Conclusion 
The results from Chi-square test confirmed that bmr6 and bmr12 genes were transmitted 
successfully in the local varieties El mota and Sepon82 genetic background. Three breeding 
populations were generated out of the six (6) breeding populations expected. In each population, 
every panicle was kept separately as a particular family. The genetic transfer of the bmr6 and 
bmr12 genes laid out the dual purpose sorghum varieties with high stover quality programme. 
Indeed, in most cases, in the breeding programmes sorghum new lines were bred for high grain 
yield potential. However, with the increased demand of livestock feed, breeding new dual-purpose 
varieties need to be developed. A total of 94 bmr BC1F3 families were selected for phenotyping in 
the Nigerien environment for their potential in terms of grain, fresh stover and dry matter yield. 
 
  
61 
 
University of Ghana http://ugspace.ug.edu.gh
CHAPTER FIVE 
5.0. AGRONOMIC POTENTIAL OF TWO NIGERIEN SORGHUM VARIETIES WITH 
BMR 6 AND 12 GENES IN TILLABERY AND KONNI 
5.1. Introduction 
In Niger like other South Saharan Africa countries, the mixed crop-livestock production system 
dominates in most small scale farmer holders. Cereal crops are mostly cultivated for dual purpose 
targeting human consumption and livestock feeding. Indeed, as the grain is used for human 
consumption, the stover part is used as feed. In this situation, farmers therefore care equally for 
grain and stover yields. Cereals residues are significant sources for feeding livestock. In this 
context, Youngquist et al. (1990) reported that in low rainfall environments, livestock depend 
mainly on crop residues for feed during dry season and drought years. Stover yield is particularly 
important after the years of low rainfall because its unavailability can affect animal survival. The 
World Bank (2013) reported that drought is the principal risk in Niger where the country 
experiences high frequency of drought with adverse impact on the national agricultural production. 
However, little research has been done to increase the quantity and the nutritive value of dryland 
crop residues (CGIAR, 2011). In this situation, the development of dual purpose cereal varieties 
constitutes a pressing need and an important goal for the SSA countries. Dual purpose sorghum 
lines with high stover quality can constitute a sustainable solution if agronomic potentials are 
acceptable. Indeed, Oliver et al. (2005) affirmed that it may be possible to add value to crop and 
animal systems by enhancing the digestibility of the stover residues by the use of bmr genes if 
grain yields can be maintained. This will make genetic improvement of dryland cereals more 
efficient and effective in addressing the needs of producers and consumers of grain, and crop 
62 
 
University of Ghana http://ugspace.ug.edu.gh
residues from these crops (CGIAR, 2011). To ascertain the latter two ideas (CGIAR, 2011; Oliver 
et al., 2005), conventional breeding method was used to introgress bmr6 and 12 genes in two 
adapted Nigerien sorghum varieties for grain yield, stover yield and quality improvement. The 
main objective was to enhance stover and grain yield potentials of bmr6 and bmr12 derived 
populations. The specific objectives were to: 
- determine the agronomic performance of 94 BC1F3 bmr 6 and 12 derived families and, 
- identify superior bmr 6 and 12 families for grain and stover yields. 
5.2. Material and methods 
5.2.1. Sites 
Phenotyping trials were conducted in two Niger INRAN research farms (Tillabery and Konni). 
These localities are the closest research farms to PRA villages, thus in the same agro ecological 
environment. Tillabery is in the western part of the country (1°26'98’’ North; 14°13'59’’ East) and 
its annual average rainfall is 400 mm. Konni (13°47’23’’ North and 5°14’ 57’’ East) and has 500 
mm rainfall annually. The Figure 5.1 shows the rainfall patterns during the last ten years in the 
two localities. 
63 
 
University of Ghana http://ugspace.ug.edu.gh
 
Figure 5 1. Ten years average annual rainfalls in Tillabery and Konni. 
5.2.2. Plant Materials 
The genetic material was composed of hundred (100) entries, of which 94 BC1F3 were bmr6 and 
12 derived families from 3 breeding populations and six checks. The bmr derived families were 
composed of three populations. Population1 (Pop1) called BCS1 resulted from El mota//El mota 
/Tx630bmr12 had 60 families. Population2 (Pop2) named BCS2, derived from 
Sepon82//Sepon82/Redlanbmr6, had 27 families. Population3 (Pop3) called BCS3 was from 
Sepon82//Sepon82/Wheatlandbmr12 covered 7 families. There were 6 checks added to the bmr 
derived families. Checks were composed of three bmr donor parent (Readlanbmr6, 
Wheatlandbmr12 and Tx630bmr12), two recurrent parental checks (El mota, Sepon82) and a local 
check called Tillabery landrace. 
 
  
64 
 
University of Ghana http://ugspace.ug.edu.gh
5.2.3. Methodology 
5.2.3.1. Experimental Design 
The experimental design was a square lattice: (number of treatment t =100; number of block per 
replicate b =10; number of plot per block p =10) with 3 replications in every location. Entries were 
randomized using GenStat 12th Edition. Each entry was sown on 3 rows. The row length was 3 m 
and the sowing density was 0.80 m between rows and 0.50 m between hills on the row (0.80 m x 
0.50 m). All phenotypic data were collected on plants in the middle row. After thinning, in every 
planting hill, 3 plants were retained.  
5.2.3.2. Cultural practices 
The following standard cultural practices were performed during the phenotyping: 
 Land preparation: The experimental land of each location was cleaned of debris. Seed bed 
was plowed and ridged.  
 Sowing: Seeds were treated with Caïman Rouge P (Permethrine 25 g/kg + Thirame 250 
g/kg) an insecticide fungicide.  Sowing was done after the first good rainfall.  
 Fertilizer applications: Diammonium phosphate (DAP) was applied before planting. 
Ridge sides were ‘opened’ and DAP (18N-46P2O5-0K2O) fertilizer was applied at the rate of 100 
kg/ha. Urea with 46%N was applied (100 kg/ha) in two sets. The first set (50 kg/ha) was applied 
at tillering and the second one (50 kg/ha) at booting stage.  
 Thinning: thinning was done two weeks after the emergence of seedlings and 3 plants per 
hill were maintained. The final density per hectare was 75000 plants. 
65 
 
University of Ghana http://ugspace.ug.edu.gh
 Field management: weed control was performed three times during plant development in 
every locality. The first one was done at tillering stage, the second at booting and the third at 
heading stage.  
 The Carbodan 3% (a systemic insecticide containing cabofuran 30 g/kg) was applied at 
seedling stage to avoid stem borer attack. 
5.3. Phenotypic data collection 
Two types of data were collected: quantitative and qualitative. Phenotypic data were collected on 
7 variables across the two environments for every family and recorded in Excel spread sheet.  
The following data were collected: 
- 50% flowering days (50%FL) 
- Plant height (PH) 
- Grain yield (GY) 
- Fresh stover yield (FSY) 
- Dry matter yield (DMY) 
- Lodging (LGG) 
- Foliar diseases (FDS) 
 
5.3.1. Description of observations 
 50% flowering days: This parameter was noted daily by using a visual observation since 
the appearance of the first panicle in each experimental plot. The number of days to 50% flowering 
66 
 
University of Ghana http://ugspace.ug.edu.gh
of a family was recorded as the number of days to 50% of the plants in the middle row of the 
experimental unit reached the anthesis at least halfway down the panicle.  
 Plant height: this measurement was done at each experimental plot physiological maturity. 
In the middle row of every family, 5 plants were randomly taken and measured from the base of 
the plant to the top of the panicle. Then the average height was calculated and recorded as the 
family height in the replication in cm. 
 Grain yield: the middle rows were harvested. Then the panicles were dried and threshed. 
For every family, the quantity of grain was weighted and finally the grain yield was calculated in 
kg/ha based on the sowing density in the experimental unit surface per replication (experimental 
unit = (0.80 m*0.50 m) * 6 (hills) =2.4 m2. 
 Fresh stover yield: this parameter was computed just after harvesting panicles. The fresh 
stover of every family per experimental unit was harvested and directly weighted in kg. The fresh 
stover yield per hectare was then estimated in kg/ha per net plot. 
 Dry matter yield: For the dry matter estimation, a sample of 1 kg of fresh stover was taken 
per experimental unit after harvesting sorghum panicles. Then each sample was dried in an oven 
at 105°C. The samples were daily weighted until dry weight stability of each sample. This final 
dry weight was obtained after 5 days in general. The obtained value per sample was used to 
calculate the total moisture of every sample and finally the dry matter per family per plot. 
The following formula was used to obtain the total humidity: 
- Total Humidity (H):  
𝐹𝑆𝑤 − 𝐷𝑆𝑤
𝐻 = ∗ 100 
𝐷𝑆𝑤
- FSw: Fresh Sample weight 
- DSw: Dried Sample weight 
67 
 
University of Ghana http://ugspace.ug.edu.gh
The dry matter was calculated using the following formula: 
𝐻
𝑆𝐷𝑀𝑡𝑜𝑡 = 𝐹𝑆𝑆𝑡𝑜𝑡 − ∗  𝐹𝑆𝑆 100 𝑡𝑜𝑡
 
Where SDMtot is the Sample Dry Matter total; FSStot is the Fresh Sample Stover and H is the 
moisture. 
 Foliar disease: Sorghum foliar diseases can be due to the harmful attacks from various 
pathogens. Consequently, in the normal growing conditions a wide range of foliar diseases are 
observed on sorghum plants. Biomass free of disease is important for the quality of the stover. 
Therefore, throughout the experiment regular inspection was done to identify and record all the 
plants in every family with visible infected leaves (abnormal colour or shape of leaves) or insect 
damages. At the physiological maturity, data for the tolerance disease damage were collected based 
on (House, 1985) scale: 1=experimental unit free of disease; 2=1 to 10% plant with disease; 3=25% 
plant with disease; 4=40% plant with disease and 5=more than 40% plant with disease. From this 
scale, 1=absence of foliar disease; 2 and 3=presence of foliar diseases and finally 4 and 5=severe 
presence of foliar disease. 
 Lodging: this variable was measured at each family physiological maturity. In practice, 
the number of stems lodged in the middle row of each family and replication was counted based 
on House (1985) scale: 1=0-10% plant lodged; 2=10-25% plant lodged; 3=25-50% plant lodged; 
4=50-75% and 5=75-100% plant lodged. 
5.4. Data analysis 
Qualitative data: Descriptive analysis using Excel followed by the Kruskal-Wallis non-parametric 
test using Minitab Eds.14 software were performed. 
68 
 
University of Ghana http://ugspace.ug.edu.gh
Quantitative data: different analyses were carried out: 
- A descriptive analysis with Excel.  
- Normality of data distribution using the Anderson-Darling test. 
- Homogeneity of the variances using the Hartley’s Fmax test. 
- Analysis of variance (ANOVA):  two types of ANOVA were performed. Firstly, a single 
separate ANOVA was performed for every variable in each site. These ANOVA analyses were 
done with R 3.4.1 environment for statistical computing using PBIB.test function (package 
‘agricolae’) for the alpha design analysis of variance using the linear model as follow:  
Yijl = µ + ti + rj + bi(j) +eijl 
Where; Yijl = value of the observed trait; ti = treatment effect and (i=1…100); rj = replication effect 
and (j=1..3); b = block within replicate effect and (j=1…10); e = random error. 
Secondly, a combined ANOVA was also performed for the variables showing homogeneity of 
variances. Indeed, data for each site were checked for homogeneity of variance using the Hartley’s 
Fmax method. The homogeneity of variance is an assumption of ANOVA. The Hartley’s Fmax 
method consist of a ratio of the larger variance/smaller variance and a result less than 3 indicate 
similarity of the variances. Therefore, the combined ANOVA was performed for the following 
variables: grain yield, fresh stover and dry matter yields. The combined ANOVA was performed 
using R 3.4.1 software of ‘agricolae’ package. The sites, replicates and blocks were considered as 
random while the entries were fixed. Statistical model was: site + replicates (sites) + site.replicates 
(blocks) + site*varieties. The mean separation was done with the Tukey Honestly Significant 
Difference (HSD).  
69 
 
University of Ghana http://ugspace.ug.edu.gh
- Genotypes by Environment (GxE) analysis: In order to check the response of the different 
entries across the locations, a GxE study was performed. The following statistical general linear 
model: site + replicates (site) + entries + entries*site was used with the Minitab Eds.14 software. 
The sites were considered as random factor while the entries were fixed. The different genotypes 
interactions and performances in the two environments (Tillabery and Konni) were performed 
using the Breedingview software. The entries sensitivity and stability were checked using the GGE 
biplot and ranking. A twenty per cent (20%) selection pressure was applied to detect the best and 
stable lines across the two sites. For this purpose, the Wricke's ecovalence test was used.  
- Genotypic variance (σ2g)  and Phenotypic variance (σ2p) Phenotypic coefficient of 
variation (PCV) and Genotypic coefficient of variation (GCV) Broad sense heritability (H2), 
Genetic Advance (GA), Genetic Advance of the Mean (GAM) were computed. 
 The genetic and phenotypic variances were computed based on the mean square genotype, 
the residual and the number of the replication of the trial in each location according to Wricke and 
Weber (1986) and Prasad et al. (1981) respectively. 
 The Phenotypic and genotypic coefficients of variation were calculated using the 
phenotypic, genotypic variances and the grand mean of the trait under consideration according to 
Burton (1952) and Johnson et al. (1955). The results were then classified from low (0-10%), 
moderate (10-20%) to high (>20%) according to Sivasubramanian and Menon (1973). 
 The broad sense heritability was also calculated and hierarchized as follow 0-30%=low, 
30-60%=moderate and >60%=high as described by Allard et al. (1960) for every trait. 
 The genetic advance was calculated using the data from the broad sense heritability, the 
selection intensity using K value (k is a constant; at 20% selection pressure k=1.40) and the 
70 
 
University of Ghana http://ugspace.ug.edu.gh
phenotypic variance then it was classified based on the following scale: 0-10%=low, 10-
20%=moderate, >20=high with Fehr et al. (1987) method.  
 The genetic advance of the mean was calculated in percentage using the genetic advance 
and the mean value of the trait according to Johnson et al. (1955) and categorized as: low=0-10%; 
moderate=10-20% and high=>20%. 
Table 5. 1. Formula for phenotypic, genetic and heritability components estimation. 
Parameters Formulas Significance 
Genetic variance GV = (MSG-MSE)/r MSG = mean square genotype, MSE = 
mean square error, r = number of 
replications 
Phenotypic variance PV = GV+MSE GV=genetic variance, MSE = mean 
square error 
Phenotypic coefficient of PCV = √𝑃𝑉 ∗ (100/̅̅𝑋̅) PV=phenotypic variance, x̄ = the 
variation grand mean of the trait 
Genotypic coefficient of GCV = √𝐺𝑉 ∗ (100/̅̅𝑋̅) GV = Genotypic variance, x̄ the grand 
variation mean of the trait 
Broad sense Heritability H2B= (GV/PV) GV = Genotypic variance, PV = 
Phenotypic variance 
Genetic advance 2GA = H2B*K*√𝑃𝑉 H B = Broad sense heritability, K is a 
constant (1.40 at 20% selection 
pressure), PV = Phenotypic variance 
Genetic advance of mean 𝐺𝐴GA (%) = ( ) ∗ 100 GA=Genetic advance, x̄ = the grand 
?̅?
mean of the trait 
 
71 
 
University of Ghana http://ugspace.ug.edu.gh
5.5. Results 
5.5.1. Traits analysis of variances 
The single ANOVA (Table 5.3) showed in both locations very highly significant statistical 
differences between the entries for maturity, plant height, grain yield, stover fresh and dry matter 
yields. Replications were not homogenous in Tillabery while; they were in Konni. The blocks were 
homogenous within replication in both sites. The table 5.3 below describes the statistical 
expression of the different traits in the two locations. 
Table 5. 2. Mean squares of bmr derived families. 
Source of 
Site Df 50%FL PH GY FSY DM 
variance 
Replication 2 147.343*** 5670.9*** 6168268*** 431619342*** 31231918*** 
Treatment 99 249.869*** 5918.1*** 1161950*** 44170996*** 4623682*** 
Tillabery 
block/replication 27 10.903ns 385.6ns 276610ns 15021836ns 1959056ns 
Residual 171 11.07 425.1 286471 14385282 1909355 
Replication 2 22.773ns 499.5ns 1447635** 5299815ns 466812ns 
Treatment 99 124.567*** 3765.1*** 503894*** 24382567*** 1753026*** 
Konni 
block/replication 27 31.886ns 1541.2ns 174572ns 8034003ns 676789ns 
Residual 171 34.5 1658.9 242280 13015877 932540 
0.05=*; **=0.01; ***=0.001: significant at probability level; ns=not significant; Df =Degree of 
freedom; 50%FL=Days to 50% flowering; PH=Plant height; GY=Grain yield; FSY=Fresh stover 
yield; DM=Dry matter yield. 
 
5.1.1. Time of flowering (50% FL) 
 Tillabery, 
there were highly significant differences observed among families. BCS1-33 with 51 days was the 
earliest family and BCS2-150 with 88 days was the latest. Families from population1 (BCS1) were 
earlier while their two parents were different in flowering by 21 days (52 days for El mota and 73 
for Tx630bmr12).  
72 
 
University of Ghana http://ugspace.ug.edu.gh
In population1, all families flowered between 51 to 67 days: BCS1-33 was the earliest followed 
by BCS1-9 while BCS1-66 the latest. 
In population2, days to 50% flowering varied from 60 to 88 days whereas their RP (Sepon82) and 
DP (Redlanbmr6) flowered at 76 and 68 days, respectively. Families BCS2-165 (60 days), BCS2-
149 (64 days), BCS2-155 (65 days), BCS2-140 (65 days), BCS2-167 (67 days), BCS2-144 (67 
days) were the earliest while BCS2-150 (88 days) was the latest.  
In population3, days to 50% flowering ranged from 70 days (BCS3-192, earliest) to 80 days 
(BCS3-179, latest). The RP (Sepon82) and the DP (Wheatlandbmr12) flowered at 76 and 65 days 
respectively.  
Among local checks, El mota (52 days) flowered earlier than Sepon82 (76 days) and Tillabery 
landrace was the latest at 77 days.  
 Konni 
Highly significant differences were observed among families. Days to 50% flowering distribution 
ranged from 49 (BCS2-160) to 85 (BCS2-203). Families from population1 (BCS1) flowered 
earlier than the families from the other two populations. 
In population1, days to 50% flowering ranged from 52 days (BCS1-13) to 73 days (BCS1-106). 
The earliest families were BCS1-13, BCS1-40, BCS1-15 (53 days) while the latest were BCS1-20 
(70 days), BCS1-106 (73 days). The RP (El mota) and DP (Tx630 bmr12) flowered in 62 and 65 
days respectively.  
73 
 
University of Ghana http://ugspace.ug.edu.gh
In population2, days to 50% flowering ranged from 49 days (BCS2-160) the earliest to BCS2-203 
(85 days) the latest in flowering. Their RP (Sepon82) and DP (Redlanbmr6) flowered at 74 and 66 
days, respectively.  
In population3, days to 50% flowering ranged from 55 days (BCS3-183, the earlier) to 69 days 
(BCS3-210 the latest). Their RP (Sepon82) and DP (Wheatlandbmr12) flowered at 74 and 62 days, 
respectively.  
El mota flowered (62 days) earlier than Sepon82 (74 days) and the Tillabery (76 days) later. 
5.1.2. Plant height (PH) 
 Tillabery 
Very highly significant differences were observed among the entries. Plant height varied from 3.18 
m to 0.88 m. The Tillabery landrace was the tallest (3.18 m) followed by BCS1-102 (2.86 m) and 
El mota (2.54 m). Sepon82 had 2.01 m.  
In population1, BCS1-102 (2.86 m) was the tallest followed by BCS1-82 (2.47 m), BCS1-95 (2.45 
m) while BCS1-45 (1.53 m) was the shortest. RP (El mota) reached 2.54 m and the donor parent 
Tx630 bmr12 had 1.09 m. 
In population2, plant height varied from 2.22 m (BCS2-142), followed by BCS2-138 (1.73 m) to 
1.10 m (BCS2-157). The RP (Sepon82) had 2.01 m and the donor parent (Redlanbmr6) had 1.32m. 
In population3, BCS3-181 with 2.37 m was the tallest followed by BCS3-210 (2.32 m) and BCS3-
179 (1.31 m) was the shortest. The RP (Sepon82) and DP (Wheatlandbmr12) reached 2.01 m and 
0.87 m, respectively. 
 
74 
 
University of Ghana http://ugspace.ug.edu.gh
 Konni 
Very highly significant differences were observed among families. Plant height varied from 3.01 
m to 0.82 m. The local Tillabery landrace was the tallest (3.01m) followed by BCS1-102 (2.57m). 
In population1, BCS1-102 (2.57 m) was the tallest followed by BCS1-92 (2.47 m) and the shortest 
families were BCS1-20 (1.15 m) and BCS1-44 (1.12 m). Parental lines reached 2.43 m (El mota) 
and 1.14 m (Tx630bmr12) respectively. 
In population2, BCS2-160 (2.17 m), BCS2-148 (2.13 m), BCS2-133 (2.09 m), BCS2-166 (2.07 
m) and BCS2-158 (2.01m) were the tallest while BCS2-203 (1.15 m), BCS2-144 (1.22 m), BCS2-
159 (1.36 m) were the shortest. The RP (Sepon82) had 1.82 m whereas the DP (Redlanbmr6) had 
1.19 m. 
In population3, BCS3-183 (2.28 m), BCS3-179 (2.02 m), BCS3-197 (1.87 m), BCS3-192 (1.83m), 
and BCS3-210 (1.83 m) were the tallest, while BCS3-181 (1.43 m) and BCS3-188 (1.71 m) were 
the shortest families. RP (Sepon82) had 1.82 m and the DP (Wheatlandbmr12) had 0.82 m 
5.1.3. Grain yield (GY) 
 Tillabery 
Very highly significant differences were observed among families. GY varied from 3504.2 kg/ha 
(BCS1-88) to 700.2 kg/ha (BCS3-181). Among local lines, Sepon82 (2631.1 kg/ha; 33.6% harvest 
index) produced better, followed by El mota (2415.0 kg/ha; 27.7% harvest index) and the Tillabery 
landrace (2402.06 kg/ha; 21.5% harvest index). The bmr donors produced less than the recurrent 
parents.  
75 
 
University of Ghana http://ugspace.ug.edu.gh
In population1, several families yielded better than their two parents (El mota and Tx630bmr12). 
BCS1-88 (3504.2 kg/ha and 46.5% harvest index) produced the highest grain yield followed by 
BCS1-102 (3428.5 kg/ha; 35.2% harvest index) and BCS1-17 (1083.6 kg/ha) had the lowest grain 
yield. The parental lines of this population yielded respectively 2415.0 kg/ha (El mota) and 1512.5 
kg/ha (Tx630bmr12).     
In population2, five families yielded better than their recurrent parents (Sepon82 and Redlanbmr6). 
BCS2-203 (2851.1 kg/ha; 48.8% harvest index) was the best followed by BCS2-155 (2731.2 
kg/ha; 34.5% harvest index) and BCS2-148 (1155.8 kg/ha) was the least yielding family. Their 
parental lines yielded respectively 2631.1 kg/ha (Sepon82) and 1131.0 kg/ha (Redlanbmr6). 
In population3, no derived lines yielded more than the recurrent parent Sepon82 (2631.1 kg/ha).  
BCS3-192 (2240.8 kg/ha; 27.5% harvest index) was the highest yielding family whereas BCS3-
181 was the least yielding family (700.2 kg/ha). The DP yielded 1363.2 kg/ha less than the best 
family. 
 Konni 
There were very highly significant differences among entries. Families from population1 were 
better yielding. Sepon82 (2123.55 kg/ha; 38.7% harvest index) yielded better than El mota 
(1974.10 kg/ha; 31.7% harvest index) and the Tillabery landrace (2014.01kg/ha). The bmr donor 
lines produced lesser than the local sorghum lines. 
In population1, grain yields varied from 3226.6 kg/ha (BCS1-60) to BCS1-17 (1285.8 kg/ha). The 
most productive families were: BCS1-60 (3226.6 kg/ha; 37.6% harvest index); BCS1-62 (3104.44 
kg/ha; 48.5% harvest index); BCS1-95 (2945.93 kg/ha; 40.3% harvest index); BCS1-102 (2919.35 
kg/ha; 35.3% harvest index); BCS1-127 (2828.23 kg/ha; 44.6% harvest index); BCS1-71 (2751.45 
76 
 
University of Ghana http://ugspace.ug.edu.gh
kg/ha; 35.4% harvest index). Their parental lines yielding respectively 1974.10 kg/ha (El mota) 
and 1472.6 (Tx630bmr12). 
In population2, BCS2-203 with 2706.5 kg/ha yielded more while BCS2-157 yielded less. The 
highest yielding families were: BCS2-203 (2706.5 kg/ha; 35.3% harvest index); BCS2-138 
(2355.69 kg/ha; 29.8% harvest index); BCS2-155 (2346.49 kg/ha; 38.8% harvest index). The 
parental lines produced 2123.55 kg/ha and 1414.0 kg/ha respectively for Sepon82 and 
Redlanbmr6. 
In population3, BCS3-181 (2250.4 kg/ha; 41.3% harvest index) and BCS3-192 (2239.5 kg/ha; 
28.7% harvest index) were highest grain yielding. BCS3-188 with 1859.8 kg/ha was the lowest. 
The recurrent parent (Sepon82) yielded 2123.55 kg/ha and Wheatlandbmr12 had 1119.8 kg/ha. 
Grain yield performance was more variable in Tillabery (StDev=781.1) than in Konni (StDev= 
575.2).  
5.1.4. Fresh Stover Yield (FS) 
 Tillabery 
There were very highly significant differences observed among families. The Tillabery landrace 
yielded 26053.54 kg/ha followed by BCS2-138 and BCS2-201 with respectively 23372.33 kg/ha 
19843.02 kg/ha. Wheatlandbmr12 had the lowest yield (3206.22kg/ha). Recurrent parent El mota 
and Sepon82 yielded on average 16658.29 kg/ha and 15776.26 kg/ha, respectively.  
Population1: Fresh stover yields varied from 18945.42 kg/ha (BCS1-102) to 5568.40 kg/ha 
(BCS1-49). The RP (El mota) yielded 16658.29 kg/ha while the donor parent (Tx630bmr12) had 
8161.53 kg/ha. In this population, superior families were BCS1-102 (18945.42 kg/ha), BCS1-66 
77 
 
University of Ghana http://ugspace.ug.edu.gh
(18288.44 kg/ha), BCS1-78 (17680.11 kg/ha), BCS1-71 (17141.93 kg/ha), BCS1-20 (17131.79 
kg/ha), BCS1-127 (16972.37 kg/ha). 
Population2: fresh stover yields varied from 23372.33 kg/ha (BCS2-138) to 8073.09 kg/ha (BCS2-
150). The recurrent parent had 15776.26 kg/ha and the donor parent yielded 6841.42 kg/ha. The 
families with superior fresh stover yield were respectively: BCS2-138 (23372.33 kg/ha); BCS2-
201 (19843.02 kg/ha); BCS2-199 (18090.62 kg/ha) and BCS2-145 (16969.51 kg/ha). 
Population3: BCS3-201 with 16965.95 kg/ha was the highest yielding followed by BCS3-197 
(16470.96 kg/ha) and BCS3-192 (16039.83 kg/ha). Parental lines yielded respectively 15776.26 
kg/ha (Sepon82) and 3206.22 kg/ha (Wheatlandbmr12). 
 Konni 
Very highly statistical significant differences were observed among families. BCS1-102 
(20623.18kg/ha), BCS2-138 (20537.39kg/ha) and BCS2-166 (20481.78kg/ha) had arithmetically 
better yield than all families. In contrast, Wheatlandbmr12 (2783.81 kg/ha), Tx630bmr12 (6780.08 
kg/ha) and Redlanbmr6 (9488.77 kg/ha) were respectively less yielding compared to their derived 
families. The Tillabery farmer landrace (18839.01kg/ha) yielded more fresh stover than El mota 
(14761.07 kg/ha) and Sepon82 (11869.29 kg/ha).  
Population1: fresh stover yield ranged from 20623.18 kg/ha (BCS1-102) to 10499.142 kg/ha 
(BCS1-44). The RP and DP yielded respectively 14761.07 kg/ha and 6780.08 kg/ha. Better 
yielding families in this population were BCS1-102 (20623.18 kg/ha), BCS1-27 (20356.56 kg/ha), 
BCS1-60 (18150.42 kg/ha), BCS1-71 (18100.66 kg/ha). 
78 
 
University of Ghana http://ugspace.ug.edu.gh
Population2: BCS2-138 and BCS2-166 with respectively 20537.39 kg/ha and 20481.78 kg/ha 
were the elite families followed by BCS2-201 (20082.99 kg/ha) and BCS2-203 (18486.30 kg/ha). 
Their parental lines produced 11869.29 kg/ha (RP) and 2783.81 kg/ha (Wheatlandbmr12). 
Population3: BCS3-192 with 20272.22 kg/ha yielded more followed by BCS3-183 (19471.22 
kg/ha). The RP yielded 11869.29 kg/ha while the DP produced 2783.81 kg/ha. 
  
79 
 
University of Ghana http://ugspace.ug.edu.gh
5.1.5. Dry matter Yield (DM) 
 Tillabery 
Families exhibited very highly significant differences among them. The farmer landrace was 
higher yielding (8802.63 kg/ha) followed by BCS2-138 (8221.073 kg/ha), BCS1-102 (6336.27 
kg/ha) and BCS3-210 (6197.56 kg/ha). Among the two RP, El mota (6297.64 kg/ha) yielded better 
than Sepon82 (5195.44 kg/ha).   
Population1: In this population the dry matter yield ranged from 6336.27 kg/ha (BCS1-102) to 
2116.65 kg/ha (BCS1-49). The parental lines gave respectively 6297.64 kg/ha (RP) and 2710.27 
kg/ha (DP). The highest yielding families were respectively: BCS1-102 (6336.27 kg/ha), BCS1-
66 (5849.70 kg/ha), BCS1-20 (5792.33 kg/ha), BCS1-78 (5640.15 kg/ha), BCS1-68 (5598.24 
kg/ha). 
Population2: In this population, the dry matter yields varied from 8221.07 kg/ha (BCS2-138) to 
2596.5 kg/ha (BCS2-150). Sepon82 (RP) yielded 5195.44 kg/ha whereas Redlanbmr6 (DP) 
yielded 2729.92 kg/ha. Among the different families of this population, BCS2-138 (8221.07 
kg/ha), BCS2-167 (6070.52 kg/ha), BCS2-199 (5990.39 kg/ha), BCS2-201 (5623.36 kg/ha), 
BCS2-145 (5592.97 kg/ha) were higher yielding. 
Population3: In population3, BCS3-210 (6197.56 kg/ha) had better dry matter yield followed by 
BCS3-192 (5909.12 kg/ha), BCS3-197 (5202.07 kg/ha) while BCS3-179 (3880.63 kg/ha) had the 
lowest dry matter yield. The RP (Sepon82) yielded 5195.44 kg/ha and the DP (Wheatlandbmr12) 
had 1261.84 kg/ha.  
  
80 
 
University of Ghana http://ugspace.ug.edu.gh
 Konni 
Very highly significant differences were observed among families. The Tillabery landrace yielded 
higher dry matter yield (5639.55kg/ha). This landrace was followed respectively by BCS3-192 
(5566.28 kg/ha), BCS2-138 (5536.52kg/ha), BCS1-27 (5510.09kg/ha) and BCS2-166 
(5507.19kg/ha). El mota (4258.90 kg/ha) yielded better than Sepon82 (3367.06 kg/ha).  
Population1: BCS1-27 with 5510.09 kg/ha had the highest dry matter yield while BCS1-103 
(2766.28 kg/ha) had the lowest dry matter yield. The RP (El mota) yielded 4258.90 kg/ha and the 
DP (Tx630bmr12) had 2091.69 kg/ha. Superior families of this population were: BCS1-27 
(5510.09 kg/ha), BCS1-102 (5359.95 kg/ha), BCS1-60 (5345.31 kg/ha), BCS1-118 (5052.28 
kg/ha), BCS1-71 (5016.62 kg/ha). 
Population2: dry matter yields varied from 5536.52 kg/ha (BCS2-138) to 3095.56 kg/ha (BCS2-
140). Parental line performances were respectively, Sepon82 (3367.06 kg/ha) and DP, Redlan 
bmr6 (2387.87 kg/ha), while BCS2-138 (5536.52 kg/ha), BCS2-166 (5507.19 kg/ha), BCS2-201 
(5385.87 kg/ha) and BCS2-203 (4958.55 kg/ha) were superior families. 
Population3: dry matter yields ranged from 5566.28 kg/ha (BCS3-192) to 3205.00 kg/ha (BCS3-
181). The RP (Sepon82) yielded 3367.06 kg/ha while the DP (Wheatlandbmr12) yielded 1074.56 
kg/ha. Superior families were BCS3-192 (5566.28kg/ha), BCS3-183 (5034.00 kg/ha), BCS3-179 
(4646.96 kg/ha). 
 
 
 
81 
 
University of Ghana http://ugspace.ug.edu.gh
Table 5. 3. Summary of the single ANOVA and homogeneity of variances. 
Traits Site Mean Sq Df F value Fmax 
TL 249.869*** 99 22.5722 6.251 
50% Flowering-(50%FD) 
KN 124.567*** 99 3.6106  
TL 5918.1*** 99 13.922 6.133 
Plant Height (PH-cm) 
KN 3765.1*** 99 2.2697  
TL 1161950*** 99 4.0561 1.95 
Grain Yield (GY-kg/ha) 
KN 503894*** 99 2.0798  
TL 44170996*** 99 3.0706 1.639 
Fresh Stover Yield (kg/ha) 
KN 24382567*** 99 1.8733  
TL 4623682*** 99 2.4216 1.288 
Dry matter Yield (kg/ha) 
KN 1753026*** 99 1.8798   
0.05=*; **=0.01; ***=0.001: significant at probability level; Df =Degree of freedom;  
TL=Tillabery and KN=Konni. 
 
5.6. Combined ANOVA results 
The Hartley’s Fmax results indicate that variances are homogeneous for Grain yield, fresh stover 
yield and dry matter yield because the ratio of maxi variance divided by mini variance is less than 
3. A combined ANOVA was therefore performed on those variables to selected bmr families 
exhibiting stable and acceptable performance across the two sites (table 5.5). Highly statistical 
differences were observed between the sites. Entrees performed diversely with high interaction 
between sites. 
The data normality test was checked (Figure 5.2.). The normality test showed that, phenotypic data 
follow normal distributions for traits under study. However, a tendency of shift was observed at 
the tails for the fresh stover and dry matter yields.  
 
82 
 
University of Ghana http://ugspace.ug.edu.gh
TL KN
Normal 
99.9
Mean 2131
StDev 462.3
99
N 100
AD 0.576
95 P-Value 0.131
90
80
70
60
50
40
30
20
10
5
1
0.1
1000 1500 2000 2500 3000 3500
GY
 
Figure 5 2. Grain yield data distribution 
TL KN
Normal 
99.9
Mean 13728
StDev 2738
99
N 100
AD 1.015
95 P-Value 0.011
90
80
70
60
50
40
30
20
10
5
1
0.1
5000 10000 15000 20000 25000
FS
 
Figure 5 3. Fresh stover yield data distribution 
83 
 
Percent Percent
University of Ghana http://ugspace.ug.edu.gh
TL KN
Normal 
99.9
Mean 4207
StDev 843.9
99
N 100
AD 0.942
95 P-Value 0.016
90
80
70
60
50
40
30
20
10
5
1
0.1
1000 2000 3000 4000 5000 6000 7000 8000
DM
 
Figure 5 4. Dry matter yield data distribution 
Table 5. 4. Results Combined ANOVA. 
Variables Source of variation Df Mean Sq 
VAR 99 1281462*** 
Site 1 7286547*** 
Site*Replicates 4 3807953*** 
Grain yield 
VAR*Site 99 384383** 
Site*Replicates*Blocks 54 236568ns 
Residuals 342 262642 
   Source of variation Df Mean Sq 
 VAR 99 44580000.00*** 
 Site 1 1288000000.00*** 
Fresh stover yield Site*Replicates 4 218500000.00*** 
 VAR*Site 99 23970000.00*** 
 Site*Replicates*Blocks 54 11310000.00ns 
  Residuals 342 13740000 
   Source of variation Df Mean Sq 
 VAR 99 4251992*** 
Site 1 5169888ns 
Site*Replicates 4 15848919*** 
Dry matter yield VAR*Site 99 2124832** 
Site*Replicates*Blocks 54 1309670ns 
Residuals 342 1422247 
0.05=*; **=0.01; ***=0.001: significant at probability level; ns = not significant 
84 
 
Percent
University of Ghana http://ugspace.ug.edu.gh
5.7. Interaction genotypes by environment-GxE 
Results for the grain yield, fresh stover and dry matter yields indicate variation between the sites. 
Varieties performed diversely across the two sites (table 5.6).  Indeed, the entries stability across 
the two environments revealed diverse adaptation levels for the grain, fresh stover and dry matter 
yield. The Wricke's ecovalence for genotypes consistency response results indicated that families 
from population1 and population2 were more consistent across the two locations (table 5.7).  
 
 
85 
 
University of Ghana http://ugspace.ug.edu.gh
Table 5. 5. Trait means dispersion per population and site of experiment. 
  
Populations means dispersion 
    
Traits Entries StDev Mean Maxi Mini 
        
   TL KN TL KN TL KN TL KN 
Whole Pop 9.541 8.000 63.003 61.553 99 89 47 46 
  Pop1 4.138 6.902 56.81 59.711 70 81 47 46 
  
50%FL Pop2 7.576 9.45 72.42 63.89 99 89 58 48 
  Pop3 4.392 7.27 75.095 63.14 86 80 65 51 
Whole Pop 47.70 48.35 184.83 184.07 356 350 76.33 72.3 
  
Pop1 31.05 45.49 205.81 191.14 294.66 350 125.33 87.7 
  
PH Pop2 32.23 42.11 136.56 170.19 356 263.7 90 107.3 
  Pop3 38.18 49.3 191.82 185.5 259.33 310 125.33 119 
Whole Pop 784.1 575.2 2241.7 2021.3 4971.7 3753.9 450 862.9 
  Pop1 754.2 599.5 2524.2 2104.1 4971.7 3753.9 844.6 889.8 
  
GY Pop2 620.3 547.0 1872.7 1895.9 3525.4 3145 1026.7 862.9 
  Pop3 541 338.8 1528 2058.7 2398 2576.3 450 1412.5 
Whole Pop 5205 4035 12263 15193 34083 29583 1667 1750 
  Pop1 4564 3657 11625 15300 24583 29583 4167 8333 
  
FS Pop2 5363 3806 13301 15628 26250 22917 3333 8790 
  Pop3 5004 4405 13260 16230 23750 26250 6250 8750 
Whole Pop 1735 1085.4  4300 4114.6 11119 7980 543 642 
  
Pop1 1509 989.7  4119 4128.6 8626 7980 1355 1808 
  
DM Pop2 1832 1035 4496 4231 11119 6483 1541 2455 
  Pop3 1643 1138 4923 4361 8047 6509 2021 2359 
50%FL=days to 50% flowering; PH=plant height; GY=grain yield; FS=fresh stover yield; 
DM=dry matter yield; StDev=Standard deviation; Maxi=Maximum; Mini=Minimum; 
TL=Tillabey; KN=Konni. 
 
  
86 
 
University of Ghana http://ugspace.ug.edu.gh
Table 5. 6. Top 20 stable entries across Tillabery and Konni in Niger in 2017. 
   GY FSY DMY 
Rank  Genotypes Value  Genotypes Value  Genotypes Value 
1 BCS2-201 3 BCS3-188 53 BCS1-59 314 
2 BCS2-165 3 Wheatlandbmr12 126 BCS1-41 365 
3 BCS1-26 126 BCS1-44 362 BCS2-149 755 
4 BCS2-142 396 BCS1-59 787 BCS2-141 919 
5 BCS1-86 618 BCS1-108 1831 BCS1-73 1279 
6 BCS1-90 679 BCS1-33 1997 BCS1-128 1886 
7 BCS1-65 999 BCS2-133 5409 BCS2-156 1903 
8 BCS2-172 1042 BCS1-72 9381 BCS1-121 2086 
9 BCS1-10 1113 BCS1-88 14120 BCS1-60 2215 
10 BCS2-159 1147 BCS3-192 15361 Tx630bmr12 3024 
11 BCS2-203 1403 BCS2-141 30221 BCS1-44 3199 
12 BCS2-199 1484 BCS2-143 34001 BCS1-88 3827 
13 BCS1-80 2174 BCS1-26 34720 BCS1-80 4159 
14 BCS1-9 2220 BCS1-60 86041 Redlanbmr6 4539 
15 BCS1-105 2957 BCS1-80 88379 BCS2-133 4945 
16 BCS2-138 3460 BCS2-156 116123 BCS2-159 5485 
17 BCS1-112 3648 BCS3-181 117842 BCS2-201 7801 
18 BCS3-192 3748 BCS1-73 121070 BCS3-188 8167 
19 BCS1-72 3853 BCS1-42 174930 BCS1-46 8725 
20 Local Tillabery 3985 BCS1-82 239011 BCS1-33 9170 
GY=grain yield; FSY=fresh stover yield; DM=dry matter yield; Value=coefficient of consistency. 
5.8. Genetic and Agronomic parameters 
The PCV and GCV were moderate for days to 50% flowering but were high for other variables at 
Tillabery (table 5.8). At Konni, the PCV was moderate for days to 50% flowering but high for 
other variables whereas the GVC values were low for 50% flowering and moderate for the 
remaining variables (table 5.9).  
At Tillabery, the broad sense heritability (H2B) was high for days to 50% flowering and plant 
height but moderate for the remaining variables. In Konni, H2B was moderate for days to 50% 
flowering but was low for the all other variables.  
87 
 
University of Ghana http://ugspace.ug.edu.gh
The genetic advance (GA) was high for the plant height and the fresh stover yield, moderate for 
days to 50% flowering and finally low for the dry matter and grain yield in Tillabery. In Konni, 
the GA was moderate for the fresh stover yield but low for all the traits. 
The GAM was moderate for days to 50% flowering and the dry matter but high for the other traits 
in Tillabery while it was moderate for the plant height and grain yield but low for all the traits at 
Konni.  
Table 5. 7. Agronomic and genetic parameters of bmr derived families at Tillabery. 
Traits Site VG VP PCV GCV H2B GA GAM (%) 
50%FL(days) TL 79.599 90.669 15.113 14.16 0.877 0.117 18.575 
PH(cm) TL 1831 2256.1 25.698 23.15 0.811 0.539 29.198 
DM(kg/ha) TL 904775.7 2814131 39.01 22.119 0.321 7.55 17.559 
FSY(kg/ha) TL 9928571 24313853 40.21 25.695 0.408 28.189 22.987 
GY(kg/ha) TL 291826.3 578297.3 33.923 24.098 0.504 5.372 23.966 
TL=Tillabery 
Table 5. 8. Agronomic and genetic parameters of bmr derived families at Konni. 
Traits SITE VG VP PCV GCV H2B GA GAM(%) 
50%FL(days) KN 30 64.5 13.05 8.902 0.465 0.052 8.5 
PH(cm) KN 702 2360.9 26.397 14.394 0.297 0.202 10.9 
DM(kg/ha) KN 273495.3 1206035.3 26.69 12.71 0.226 3.486 8.4 
FSY(kg/ha) KN 3788896.6 16804774 26.981 12.811 0.225 12.939 8.5 
GY(kg/ha) KN 87204.6 329484.6 28.397 14.609 0.264 2.126 10.5 
KN = Konni 
5.9. Qualitative data analysis: Lodging and foliar diseases 
5.9.1. Lodging 
Recurrent (RP) and bmr donor (DP) parents showed a good resistance to lodging and tolerance to 
foliar diseases. Indeed, among parental lines, El mota was less sensitive to lodging. El mota 
showed 5.55% lodging at Tillabery (3 plants over 54 in the 3 replications) and 3.70% at Konni (2 
88 
 
University of Ghana http://ugspace.ug.edu.gh
plants over 54 at Konni in the 3 replications). One plant lodged in Sepon82 and no plant lodged in 
the Tillabery farmer landrace and the DP. At both Tillabery and Konni, 81.4% and 82.3% of the 
families showed resistance to lodging effect respectively.  
At Tillabery resistance to lodging varied very highly among entries (H0 = 223.23, P = 0.000). In 
population1, BCS1-45 and BCS1-86 with -1.86 and 2.89 respectively had the lowest and the 
highest z-value (z-value corresponding statistically to the standard deviation). In population2 
BCS2-150 had the lowest z-value (-2.91), while a group of families (BCS2-152, BCS2-153, BCS2-
155, BCS2-156, BCS2-157, BCS2-158, BCS2-159, BCS2-160, BCS2-165, BCS2-166, BCS2-
167, BCS2-172, BCS2-199, BCS2-201 and BCS2-203) with -1.34 had the highest z-value. In 
population3, z-value varied from -1.34 (BCS3-179, BCS3-181, BCS3-183, BCS3-188, BCS3-192) 
to 1.03 (BCS3-210).  
At Konni (H0 = 218.47 and P = 0.000). The z-values for resistance to lodging varied from -1.46 
to 2.88 (BCS1-86) in population1. In population2, z-values were -1.46 for all the families except 
BCS2-203 (-0.77) whereas in population3 z-values were -1.46 for all the families.  
5.9.2. Foliar diseases 
Few families (6%) showed leaf disease symptoms. Indeed, at Tillabery only 2.7% of the entries 
showed severe leaf disease symptoms (more than 40% of the plant with infected leaves). El mota 
was free of diseases while Sepon82 showed 3% presence of infected leaves. Local Tillabery 
exhibited 7% foliar disease presence. Among bmr donor parents, only Tx630bmr12 was 
susceptible (25% of plant with foliar disease). In population1, BCS1-41 and BCS1-68 with 8% 
each and BCS1-43 (3%) were susceptible families. In population2, BCS2-143 (8%) was the most 
susceptible family whereas in population3 BSC3-179 (17%) was the only susceptible family.  
89 
 
University of Ghana http://ugspace.ug.edu.gh
At Konni, entries showed a good level of resistance to foliar diseases. Indeed, only 3.3% showed 
severe leaf disease symptoms. Sepon82 was free of foliar diseases while El mota and the Tillabery 
farmer landrace each 6.66%. Between the bmr donor parents, Tx630bmr12 with 20% was the most 
susceptible. Readlanbmr6 showed 16.66% and Wheatlandbmr12 3.33%. The repartition between 
populations showed that in population1, BCS1-72 (16.33%) and BCS1-127 (21.66%) were the 
most susceptible families. In population2 and population3 respectively BCS2-143 (6%) and BCS3-
179 (4.66%) were also the most susceptible families like in Tillabery. 
Results from Tillabery on foliar diseases showed highly significant differences among entries (H0 
= 135.96 and P = 0.008). The z-values were -0.28 for all the families excepted BCS1-68 and BCS1-
41 which had z value of 0.74 in population1 while in population2, the values varied from -0.28 
(BCS2-138; BCS2-140; BCS2-141; BCS2-142; BCS2-145; BCS2-146; BCS2-149; BCS2-158; 
BCS2-159; BCS2-165; BCS2-167 and BCS2-201) to 2.81 (BCS2-203). In population3, only 
BCS3-179 with 0.79 showed different z-values from the -0.28 of the other families (BCS3-181, 
BCS3-183, BCS3-188, BCS3-192, BCS3-197, BCS3-210). At Konni, (H0 = 172.69 and P = 
0.000). In population1, the z-values were -0.75 however, BCS1-61 with 2.98 showed the highest 
value. In population2, the z values varied from -0.75 to 2.71 (BCS2-160) whereas in population3 
only BCS3-179 with 2.26 exhibited a different value from the other families (-0.75 in general). 
The z-values were used to compare the average rank of each population to the overall rank of all 
families (entries). In both Tillabery and Konni, lodging was more spread in population1 (TL=-1.86 
to 2.89; KN=-1.46 to 2.88) compared to population2 (TL=-2.91 to -1.34; KN=-1.46 to -0.77) and 
population3 (TL=-1.34 to 1; KN=-1.46). The recurrent parents score for lodging were negative 
thus lesser than the mean. Sepon82 and the local Tillabery exhibited also lesser z-values compared 
to the overall mean. However, El mota with 0.11 and -0.05 at respectively Tillabery and Konni 
90 
 
University of Ghana http://ugspace.ug.edu.gh
was closer to the overall entries mean for the susceptibility to lodging. This may indicate a 
tendency for susceptibility to lodging of the recurrent parent El mota. 
With regards to foliar diseases, population2 had more varied tolerance at Tillabery (TL=-0.28 to 
2.81) while in Konni the range was most widely spread in the population1 (KN=-0.75 to 2.98). 
The z-value scores showed that at Konni, the parental lines, were poorly resistant to foliar diseases 
except Sepon82. However, at Tillabery, Tx630bmr12 was the most susceptible line. 
Table 5. 9. z-values distribution of bmr derived families and checks. 
  Konni Tillabery 
Population Lodging Foliar Disease Lodging Foliar Disease 
Population1  -1.46 to 2.88  -0.75 to 2.98  -1.86 to 2.89  -0.28 to 0.74 
Population2  -1.46 to -0.77  -0.75 to 2.71  -2.91 to -1.34  -0.28 to 2.81 
Population3  -1.46  -0.75 to 2.26  -1.34 to 1.03  -0.28 to 0.79 
Checks 
Redlanbmr6 -1.46 2.86 -1.34 -0.28 
Wheatlandbmr12 -1.46 2.2 -1.34 -0.28 
Tx630bmr12 -1.46 1.6 -1.34 1.81 
El mota 0.11 2.41 -0.05 -0.28 
Local Tillabery -1.46 2.41 -1.34 1.63 
Sepon82 -0.85 -0.75 -0.79 0.67 
 
 
 
 
 
 
 
 
91 
 
University of Ghana http://ugspace.ug.edu.gh
5.10. Discussion 
The results of this study showed that the families derived from population1 exhibited earliness 
compared to populations2 and population3 and the Tillabery landrace. In addition to their earliness, 
families from population1 showed high grain yield in both locations. Shapiro et al. (1993) reported 
that earliness gives drought escape and thus have the potential to increase yields in low-rainfall 
years. The results of this study are thus in agreement with those authors. Earliness combined with 
acceptable grain and healthy dry matter stover yields are essential breeding traits for new sorghum 
lines for local small-scale farmers facing shortage, erratic and variability in rainfall. Camara et al. 
(2006) revealed that characteristics of the variety such as early maturity, less diseases and drought 
resistance incidences and productivity influenced sorghum and millet adoption by farmers. Current 
study findings pave the road for sustainable livestock feeding in Niger with rain fed cropping 
environments. 
The Tillabery landrace had the highest fresh stover and dry matter yield. El mota, the farmer variety 
also ranked among the best for dry matter yield. High stover yield of local landraces indicates 
farmers’ varietal preference for varieties with high stover potential and shows the high importance 
of stover from cereal crops in the country. Magnan et al. (2012) reported that the importance of 
crop residue as feed has an implication for farmers’ cereal variety choice. On the other hand, 
superior varieties with low biomass yield might not be selected by farmers in areas where stover 
value is high. The results of the current study are in agreement with those authors. There were 
families performing better than their parental lines in fresh and dry matter yields. Current results 
are in harmony with Kotasthane et al. (2015) who evaluated bmr sorghum parental lines and their 
derivatives for fresh and dry biomass and found some derived lines superior to their parental lines 
checks.  
92 
 
University of Ghana http://ugspace.ug.edu.gh
The results of this study indicate that several elites bmr derived families produced better grain 
yield than their recurrent parents in both locations. Oliver et al. (2005) reported that bmr genes 
were found to have negative agronomic impact on grain and stover yields in grain sorghum lines 
but the yield-drag can be overcome with heterosis. Furthermore, the improvement of the stover 
quality due to bmr genes may raise the potential of bmr derived families and compensate grain 
yield-drag for other families. Indeed, with an equal to greater harvest index, their dry matter yields 
were higher than the recurrent parental lines. In this view, the CGIAR (2011) reported that the 
classical target of research has been on increasing the output of grain, while the value of the stover 
was of secondary interest. 
The current study results revealed that across locations and for all the traits, phenotypic variances 
were superior to genotypic variances. The results are thus in harmony with Sawadogo et al. (2014) 
who found similar results in a genetic evaluation of sweet sorghum accessions from Burkina Faso. 
This shows variability in plant material used (BC1F3) and selection opportunities to identify 
superior families. According to Bhagasara et al. (2017) variability availability is a pre requisite for 
any improvement program. The phenotypic and genotypic coefficients of variation were moderate 
for the 50% flowering time in both two locations. The current results are in agreement with those 
of Godbharle et al. (2010). This indicates that different families seemed to mature within a short 
interval. For the plant height, grain, fresh stover and dry matter yields, the phenotypic coefficients 
of variation were higher than the genotypic coefficient of variations. This indicates that those traits 
were highly influenced by the environmental conditions, corroborating with findings from 
Bhagasara et al. (2017).  
In Tillabery, high heritability was observed for days to 50% flowering and plant height, moderate 
heritability was observed for the fresh stover, dry matter and grain yields while the genetic advance 
93 
 
University of Ghana http://ugspace.ug.edu.gh
was high for plant height and fresh stover yield. Bhagasara et al. (2017) also reported high 
heritability for days to 50% flowering and plant height. The heritability of a trait is important in 
plant breeding because it is required in determining its response to selection. The traits exhibiting 
high heritability (days to 50% flowering and plant height) can be improved through direct 
selection. The traits with moderate and low heritability can be improved through pedigree 
selection. In contrast, low heritability was observed at Konni for all traits excepted days to 50% 
flowering. This reveals high environmental conditions on families in this location.  
The genetic advance was low for grain and dry matter yields at both locations. This indicates that 
those two traits are under the control of several genes and suggest that phenotypic selection may 
not be effective. Govindarasu et al. (1990) revealed that high heritability of a trait does not signify 
a high genetic advance. The results are thus in agreement with these authors. 
Families performed diversely across and within location on lodging and foliar diseases during 
agronomic performance trials. In breeding for high stover quality, lines resistant to lodging and 
free of foliar diseases are important assets to consider. Lodging and foliar diseases can 
considerably decrease productivity (grain and stover yields) and also their quality. Therefore, those 
two qualitative variables were visually and daily investigated from vegetative until maturity time 
for each family across and within location. A good resistance to lodging was observed in both 
locations. The score1 (very good resistance) and score2 (good resistance) accumulated 81.4% at 
Tillabery and 82.3% at Konni. Indeed, 59.7 families were scored 1 and 21.7 families had score 2 
at Tillabery. At Konni, 62 families obtained score 1 while 20.3 families scored 2. It is important 
to know that only 2.3% at Tillabery and 3.67% at Konni of families were very susceptible to 
lodging (score5). This good level of bmr derived families’ resistance to lodging was supported by 
Bean et al. (2013), who observed no significant differences in lodging between bmr and 
94 
 
University of Ghana http://ugspace.ug.edu.gh
conventional forage sorghum. Sattler et al. (2010) reported that increases in lodging attributable 
to bmr genes were not detected and the genetic background of sorghum affects lodging more than 
brown midrib. Li et al. (2015) also concluded that genetic characteristics are likely to play a 
significant role in the lodging resistance of bmr12 plants, and lodging resistance associated with 
bmr12 plants can be significantly improved by heterosis. Getachew et al. (2016) informed that in 
bmr hybrid variety, lodging was more influenced by the variety or delayed harvest period than the 
bmr genotype. At population levels, Population1 was more susceptible to lodging than population2 
and population3 across locations. Bmr12 gene was successfully introgressed in both population1 
and population3. Thus, lodging observed in population1 can be due to the genetic background of 
its recurrent parents or a combination between the two parents. Indeed, the recurrent parent El 
mota reached 254.99 cm at Tillabery and 196.33 cm at Konni. Sepon82 was 201.88 cm at Tillabery 
and 124.33 cm at Konni. Saballos (2008) affirmed that tall plants are more prone to lodging. 
Findings from this current study are in agreement with those reported by (Cherney et al., 1991 and 
Oliver et al., 2005) who asserted that the effect of bmr mutations on forage quality varies 
depending on the genetic background of the recurrent parent. To prevent the effect of such 
phenomenon, Vogler et al. (2009) suggested the importance of identifying a suitable genetic 
background that allows for optimal impact of the mutation. Kruskal-Wallis non-parametric test for 
families’ resistance to lodging showed a high range of z-value in population1 compared to 
population2 and population3 across locations.  This result shows a great diversity among families 
in population1. It is important to mention that population1 was derived from a farmer variety with 
heterogeneous plant height. The results are in agreement with Sawadogo et al. (2014) who reported 
the high diversity in farmers’ varieties due to their seed production and management systems. 
Indeed, farmers produce their own seeds in areas where several varieties coexist. In contrast, the 
95 
 
University of Ghana http://ugspace.ug.edu.gh
low range of z-values in population2 and population3 explains the character of improved varieties 
of both RP and DP.  
Families showed resistance to foliar diseases across both locations. This good resistance to foliar 
diseases observed is in agreement with Dowd et al. (2016) who observed in laboratory no 
consistent increase in insects or pathogen damage for the low lignin sorghum lines. Later under 
field condition, the same authors concluded that sorghum lines bmr6 and bmr12 generally do not 
have increased susceptibility to insects and disease as compared to wild type.  For the leaf diseases, 
the identified pathogens in Niger include: Ramulispora sorghicola and R. sorghi, Cercospora 
sorghi, Colletotrichum graminicola, Ascochyta sorghina, Helminthosporium turcicum and 
Gloeocercospora sorghi (Sidibe, 1978). In general, a low level of sorghum susceptibility to foliar 
diseases was reported in previous study (Bandyopadhyay et al., 1987) who claimed that the 
Acremonium strictum disease has been observed in some African sorghum growing countries like 
Lesotho, Zimbabwe, Malawi, Tanzania and Mali but does not appear to be a serious disease on 
varieties currently grown by farmers. The comparison between different populations showed that 
in population1, five different families BCS1-41, BCS1-68, BCS1-43, BCS1-72 and BCS1-127 
were susceptible to foliar diseases in the two locations. This result can be due to the susceptibility 
or the poor adaptability of Tx630bmr12, the donor parent. Indeed, this line was the most 
susceptible line among the bmr donor parents and exhibited a high level of 25% susceptibility at 
Tillabery and 20% at Konni. In contrast, in population2 and population3, families BCS2-143 and 
BSC3-179 were susceptible in both locations at different degree respectively. Kruskal-Wallis non-
parametric test tolerance to foliar diseases showed that families derived from population1 
exhibited the lowest z-value in Tillabery, but the highest z-value in Konni. Those results show the 
96 
 
University of Ghana http://ugspace.ug.edu.gh
good adaptability of the RP in the Tillabery compared to Konni in addition to the high level of 
susceptibility of the bmr DP in Tillabery and Konni. 
The following bmr derived families BCS1-60, BCS1-102, BCS2-138, BCS2-201, BCS3-192, 
BCS3-210 exhibited promising potential for dual purpose target. However, susceptibility to 
lodging in BCS1-60, BCS1-102 is observed and additional backcrossing may contribute for more 
genetic stability.  
The GxE analysis was performed for three quantitative traits (grain, fresh stover and dry matter 
yields) in order to identify the most stable entries across the two locations. The different lines 
performed very diversely with various ranking stability levels. Kang (2002) found that for 
quantitatively inherited traits, the genotype values and their relative rankings can change from one 
environment to another. The results are in agreement with this author. Furthermore, Kimbeng et 
al. (2009) asserted that the rank changes confound the determination of the overall true genetic 
value of the prospective varieties. 
  
97 
 
University of Ghana http://ugspace.ug.edu.gh
5.11. Conclusion 
Different bmr derived families performed diversely across locations for all seven variables studied. 
Recurrent parents had good grain and dry matter yields. The bmr donor parents’ agronomic 
performances were poor to moderate in grain and dry matter productivity in the local growing 
conditions. However, the bmr breeding populations productivity was acceptable for both grain and 
dry matter yields. The agronomic performance of some bmr derived families from population1 
and population2 for both grain and dry matter yields were higher than those of their recurrent 
parent alones. However, in population3, bmr derived families had grain yield reduction tendency 
while dry matter production was superior in all populations compared to recurrent parents’ 
performance. Families from population1 exhibited early maturity pattern while families from 
population3 were later maturing.  
Stover quality traits of bmr6 and bmr12 genes introgression did not significantly decrease the 
resistance to lodging and foliar disease in the breeding populations. 
The recurrent parent El mota exhibited a little sensitivity to lodging.  
Within each population, the following bmr derived families BCS1-102; BCS1-60; BCS2-138 and 
BCS3-192 were identified respectively in Population1, Population2 and in Population3 as superior 
families across locations for grain and stover yields. These families are potential candidates for 
stover and grain yield improvement for sustainable cattle productivity (milk and beef) and will 
open for the next future cattle intensive production. Opportunity to use bmr6 and bmr12 genes 
enriched sorghum lines as an alternative for sorghum stover quality improvement at small scale 
farmers’ level is opened. 
 
98 
 
University of Ghana http://ugspace.ug.edu.gh
CHAPTER SIX 
6.0. DETERMINATION OF NUTRIENT CONTENT OF BROWN MIDRIB DERIVED 
POPULATION 
6.1. Introduction 
Stover quality can be defined in various ways. Practically it can be considered as the quantity 
intake to which it has the potential to produce a desired animal response. According to Ball et al. 
(2001), factors that influence forage quality include the following: intake, palatability, digestibility 
and nutrient content. The nutrient content of a given line is therefore of great importance in the 
extensive livestock husbandry. In the developing countries, the FAO declared that the cereals 
residues are major forages and sorghum residues was qualified as a valued feed, especially if cut 
and dried immediately after the heads have been harvested for grain (www.fao.org). In small-scale 
farming systems, stover is usually harvested and dried after grain harvest. The unchopped stover 
is further stored and used as livestock feed. In maintaining sustainable production systems, crop 
residues can play a strategic role if high quality of stover is used for livestock feeding. Indeed, 
Elbasha et al. (1999) foresee a shift to greater importance of crop residues to increase livestock 
production. FAO (2012) reported that in tropical countries, dairy animals are primarily fed on crop 
residue based diets with very little green fodder/hay/silage, which if available may only be for a 
limited time.  
A participatory rural appraisal conducted in 2016 in three localities in Niger revealed that during 
the dry season, which coincides with shortage of feed, milking cows, bullocks and then draught 
animals were the main beneficiaries of sorghum stovers (Diakité et al., 2017). However, those 
stovers were qualified of low nutritive values by farmers and stover traders. In this view, Suharti 
99 
 
University of Ghana http://ugspace.ug.edu.gh
et al. (2011) asserted that low cattle production may be caused by inadequate nutrient supply in a 
high-forage based ration in developing countries.  
In countries like Niger, there is a pressing demand for efficient stover improvement for both 
quantity and quality to sustain livestock productivity. To date no studies have been reported on 
sorghum stover nutritive value. Furthermore, the value of local sorghum varieties containing the 
bmr trait need to be assessed. Ninety-four (94) BC1F3 derived bmr families were phenotyped for 
agronomic performances specially for the grain and dry matter yields. Among them, 20 elite 
families were selected based on their superior phenotypic performance in one location (Tillabery) 
to determine their nutritional contents. Families' parental lines were also analysed along with their 
progenies. The main objective of this study was to assess nutrient content of selected bmr derived 
families and respective parental lines. The specific objectives were to: 
- determine the nutritional composition of the bmr derived families, and 
- identify superior families with high nutritional values.  
6.2. Material and methods 
6.2.1. Plant material 
The plant material was composed of 20 entries selected for their superior agronomic performance 
plus six checks composed of the recurrent and bmr donor parents and one farmer landrace (Table 
6.1.).  
The Baker’s Standard Deviation for multi traits selection method (BSD) was used to select the 
elite bmr derived families. Early maturing (days to 50% flowering time), grain and dry matter 
yields were considered.  
100 
 
University of Ghana http://ugspace.ug.edu.gh
BSD = Σ?̅?𝐼/𝜎𝑝𝑖 : 
Where Σ?̅?𝐼 is the sum of mean of the ith trait and  
𝜎𝑝𝑖 is the phenotypic standard deviation of the ith mean. 
The selected families were set out as follow: Population1 (El mota//El mota/Tx630bmr12) was 
represented by 12 families. Population2 (Sepon82//Sepon82/Redlanbmr6) contains 6 progenies 
families and finally population3 (Sepon82//Sepon82/Wheatlandbmr12) includes 2 families.  
Table 6. 1. List of selected families and parental lines. 
 
  Populations 
  Number Code   Number Code 
 1 BCS1-18 13 BCS2-138 
 2 BCS1-60 14 BCS2-145 
 3 BCS1-66 15 BCS2-172  
Population2 
 4 BCS1-68 16 BCS2-199 
 5 BCS1-71 17 BCS2-201 
Population1 6 BCS1-82 18 BCS2-203 
 7 BCS1-88 19 BCS3-192 
Population3 
 8 BCS1-92 20 BCS3-210 
 9 BCS1-95 21 El mota 
 10 BCS1-102  22 Local Tillabery 
 11 BCS1-118 23  Redlanbmr6  
Checks 
 12 BCS1-127 24 Sepon82 
   25  Tx630bmr12 
      26 Wheatlandbmr12 
 
 
 
 
 
 
101 
 
University of Ghana http://ugspace.ug.edu.gh
6.2.2. Methodology 
Stover samples (leaves + stems) of every family was collected at after panicle harvest at maturity. 
Sorghum plants were cut at ground level. The collected stover samples were dried in shade on 
shelves during two months.  
Stover chemical composition determination was conducted at the Animal Nutrition Laboratory 
(LNA) of IER/Sotuba in Mali in 2018. All samples were dried at 50°C for 48 hours before the 
chemical analysis. Drying sample was to eliminate all remaining humidity in all samples in order 
to work with complete dried samples.  
6.2.3. Stover nutrient content data analysis 
A descriptive analysis was firstly done for the mean distribution of every variable (Appendix 1). 
Then a one sample t-test for differences of the means was performed for all the variables. In each 
population, the recurrent parents’ mean was tested against its derivative progenies’ mean. The t-
value were used to measure the size of the difference while the p-value were used for statistical 
differences observation. 
 6.2.3.1. Data collection 
The stover quality analysis were based on the following elements: Dry matter, Net Energy, Crude 
Protein content, mineral content (calcium and phosphorus), Cellulose content, Neutral Detergent 
Fiber, Acid Detergent Fiber, and the Acid Detergent Lignin using the sequential filter bags method 
of the ankom technology for the last three elements (Ankom, 2010). For the other elements the 
analysis were conducted according to the description below: 
102 
 
University of Ghana http://ugspace.ug.edu.gh
 Dry matter (DM): The samples dry matter consisted as their final mass when they were 
completely dried except their water content. Its estimation was done by sampling 2 grams (P1) in 
each sample. After that, those 2 grams were placed in the oven at 105°C for 4 hours. Then all the 
samples were weighed to get their dry weight (P2). The difference P1-P2 gives the dry matter 
content of each sample. 
 Net Energy (NE): The energy is a quality associated with the nutrient content of feedstuffs 
and mixed diets and is therefore, the most accurate and unbiased way to date of characterizing the 
energy content of feed (Moehn et al., 2005). In each sample 0.22 g was weighted. Then the 0.22 g 
was put into pellets to prevent its spraying throughout the calorimeter during the influx of oxygen. 
Pure oxygen was used to combust the sample in the calorimeter using a bomb calorimeter.  
 Crude Proteins content (CP): CP is widely used to determine the protein requirements in 
a feed. It is thus a significant indicator of feed quality. CP determination was conducted using the 
Kjeldahl method. In practice, this method consists of 3 main steps (digestion, distillation and the 
titration) leading to the determination of the amount of nitrogen (N) in each sample. In general, 
the term CP refers to all the nitrogenous compounds in a feed therefore the total amount of CP was 
finally obtained as follow: CP = N*6.25. 
 Minerals: The major mineral requirement for cows feed include calcium and phosphorus. 
The calcium content in every sample was estimated using a photometer while the phosphorus 
content was determined using the spectrolic21. 
 Cellulose content: The Wend method was used to estimate the cellulose content in every 
sample. In practice, 1 g was taken from every dried sample and put in a flask. Then sulfuric acid 
was added and the flask was heated in an oven at 150°C for 2 hours. After that the sample was 
103 
 
University of Ghana http://ugspace.ug.edu.gh
weighted to get P1. At this point, the sample was burned at 700°C for 2 hours and weighted again 
to get P2. Finally, the percentage of cellulose of the sample was obtained by doing (P1-P2) *100. 
 Neutral Detergent Fiber (NDF): The NDF is basically a mixture of hemicellulose + 
cellulose + lignin. With a very highly sensitive balance, 0.5 g portion was taken from every sample 
and put in a centrifugal mill with a 2 mm screen. A heat sealer was utilized to completely seal the 
upper edge of the filter bag within 4 mm of the top. Two blank bags were added to the experiment 
to get the blank bag correction.  The experiment was conducted in conformity to the ankom 
protocol (Ankom, 2010) and finally the percentage of NDF of every sample was determined using 
the following formula: 
%NDF = (W3-(W1*C1)) *100/W2 
Where: W1= bag tared weight; W2= Sample weight; W3= Dry weight of bag with fiber after 
extraction process; C1= Blank bag correction (running average of final oven-dried weight divided 
by the original blank bag weight). 
 Acid Detergent Fiber (ADF): The ADF is a mixture of cellulose + lignin. It was 
determined after the NDF. The same bags containing the same samples from the previous protocol 
were used in conformity to the ankom protocol. The percentage of ADF was calculated as follow: 
%ADF= (W3- (W1 * C1)) * 100/W2 
Where: W1=Bag tared weight; W2 = Sample weight; W3 = Dry weight of bag with fiber after 
extraction process; C1 = Blank bag correction factor (running average of final oven-dry weight 
divided by original weight). 
 Acid Detergent Lignin (ADL): The ADL correspond to the lignin content in a sample. It 
was determined on the residue from the ADF procedure. Practically, an acidified detergent solution 
was used to dissolve cell soluble, hemicellulose leaving a residue of cellulose, lignin, and heat 
104 
 
University of Ghana http://ugspace.ug.edu.gh
damaged protein, a portion of cell wall protein plus the minerals. ADL was determined 
gravimetrically as the residue remaining upon ignition after 72% sulfuric acid (H2SO4) treatment. 
% Lignin = (Crucible Weight after Acid Soak - Crucible Weight after Ignition) / (Sample Weight x DM) 
6.3. Results 
The Baker’s Standard Deviation (BSD) for multi traits selection method allowed a clear 
identification of the top 20% superior families in every population. Table 6.2. shows the agronomic 
characteristics and BSD score of every selected BC1F3 family. The mean distribution gave the 
spread tendency and the variability in every population for all the variables (Appendix 1). The one 
sample t-test allowed the comparison between the recurrent parental lines and their derivative bmr 
progenies for the nutrient content in each population. Table 6.3. shows the t-test results. 
 
 
 
 
 
 
 
 
105 
 
University of Ghana http://ugspace.ug.edu.gh
Table 6. 2. Agronomic characteristics and BSD values of the entries. 
50%FL GY DMY BSD BSD BSD BSD 
Families 
(Day) kg/ha kg/ha  50%FL GY DM TOTAL 
BCS1-66 67 2732.64 5849 19 5.19 5.48 29.89 
BCS1-95 63 3334.58 5014.47 18 6.33 4.69 29.29 
BCS1-102 58 3429.58 6322.07 17 6.51 5.92 29.25 
BCS1-127 62 3125.42 5219.5 18 5.93 4.89 28.6 
BCS1-82 58 3163.33 5614.68 17 6.01 5.26 27.99 
BCS1-71 57 3113.19 5516.24 17 5.91 5.16 27.61 
BCS1-118 56 3294.31 5490.05 16 6.26 5.14 27.54 
BCS1-18 57 3035.69 5505.19 16 5.76 5.15 27.35 
BCS1-88 59 3505 4037.18 17 6.66 3.78 27.35 
BCS1-92 55 3373.06 5318.14 16 6.41 4.98 27.34 
BCS1-60 55 3330.97 5324.22 16 6.33 4.98 27.17 
BCS1-68 58 2665.42 5600.68 17 5.06 5.24 27.13 
BCS2-138 71 2536.53 8199.72 11 5.5 6.98 23.22 
BCS2-172 87 2100 4780.64 13 4.55 4.07 21.67 
BCS2-203 86 2854.16 2991.09 13 6.19 2.54 21.64 
BCS2-199 74 2434.37 6004.29 11 5.28 5.11 21.48 
BCS2-145 69 2561.81 5599.27 10 5.55 4.76 20.76 
BCS2-201 76 1881.11 5628.04 11 4.08 4.79 20.36 
BCS3-210 74 2114.91 6169.08 24 4.01 7.15 35.14 
BCS3-192 70 2238.61 5908.43 23 4.24 6.84 33.99 
Local Tillabery 77 2405 8805.6 22 4.57 8.24 34.92 
Sepon82 76 2634.72 5200.26 22 5 4.87 31.88 
Tx630bmr12 73 1514.3 2713.1 21 2.88 2.54 26.37 
Redlanbmr6 68 1131.66 2732.45 20 2.15 2.56 24.22 
Wheatlandbmr1
66 1356.7 1287.07 19 2.58 1.21 22.72 
2 
EL mota 52 2413.05 6296.52 15 4.58 5.89 25.57 
50%FL=50% Flowering time; GY=Grain yield; DMY= Dry matter yield; BSD= Baker’s Standard Deviation 
 
 Dry matter (DM): In population1 the dry matter varied from 94.05% (BCS1-71) to 
96.67% (BCS1-60).  The dry matter content of their two parental lines was 95.32% and 95.25% 
respectively for the bmr donor and the recurrent parents. The t-test analysis revealed a t-value of 
2.71 while the p-value was 0.019. Those results revealed the difference between the RP mean and 
the progenies mean. 
106 
 
University of Ghana http://ugspace.ug.edu.gh
In population2, the dry matter varied from 94.42% (BCS2-203) to 95.98% (BCS2-138).  The dry 
matter content of their two parental lines was 95.00% and 95.32% respectively for the bmr donor 
and the recurrent parents. The t-value was 0.23 while the p-value was 0.826. Therefore, no 
statistical significant difference was observed between the RP mean and the progeny mean. 
In population3 the dry matter varied from 94.48 % (BCS3-210) to 95.15% (BCS3-192).  The dry 
matter content of their two parental lines was 95.02% and 95.32% respectively for the bmr donor 
and the recurrent parent. There were no significant differences between the means of the recurrent 
parental line and the progenies means (t-value=-1.31 while p-value=0.320). 
Among the local sorghum lines, the Tillabery farmer landrace with 95.35% exhibited the highest 
dry matter content followed by Sepon82 and finally El mota. The figure 6.1. shows the 
performance of every family.  
Dry Matter
97.00
96.50
96.00
95.50
95.00
94.50
94.00
93.50
93.00
92.50
Derived bmr families and checks
 
Figure 6. 1. Dry matter content of the selected families and the checks. 
 Net Energy (NE): In population1 the net energy varied from 2893.41 kcal/kg DM (BCS1-
118) to 3755.02 kcal/kg DM (BCS1-71). Their two parents produced respectively (3392.76 and 
107 
 
% DM
BCS1 - 18
BCS1 - 60
BCS1 - 66
BCS1 - 68
BCS1 - 71
BCS1 - 82
BCS1 - 88
BCS1 - 92
BCS1 - 95
BCS1 - 102
BCS1 - 118
BCS1 - 127
BCS2 - 138
BCS2 - 145
BCS2 - 172
BCS2 - 199
BCS2- 201
BCS2 - 203
BCS3 - 192
BCS3 - 210
EL mota
Local Tillabery
Redlanbmr6
Sepon82
Tx630bmr12
Wheatlandbmr12
University of Ghana http://ugspace.ug.edu.gh
3611.37 kcal/kg DM) for the bmr donor and the recurrent parent. The t-value (2.44) and p-value 
(0.031) revealed the statistical difference between the RP and progenies means. 
In population2, the net energy varied from 3506.83 kcal/kg DM (BCS2-172) to 3732.31 kcal/kg 
DM (BCS2-201) while their parental lines produced respectively 3664.32 for the bmr donor and 
3525.24 for the recurrent parent. The t-test revealed a t-value of 2.17 while the p-value was 0.073. 
Those results indicate that there were no statistical differences between the RP mean and its’ 
progenies mean. 
In population3, the net energy varied from 3673.57 kcal/kg DM (BCS3-192) to 3692.49 kcal/kg 
DM (BCS3-210). The t-value was 1.99 and the p-value was 0.185. Therefore, the DP mean and 
the derivative progenies’ mean were statistically equal. 
Among the local lines, the Tillabery farmer landrace with 3714.26 kcal/kg DM and had the highest 
net energy followed by El mota and finally sepon82. 
 
N.E kcal/kg DM
4000.00
3500.00
3000.00
2500.00
2000.00
1500.00
1000.00
500.00
0.00
Derived bmr families and checks
 
Figure 6. 2. Net energy content of the selected families and the checks. 
 Crude Proteins (CP): In population1 the CP content varied from 2.75% (BCS1-60) to 
5.63% (BCS1-82) while their two parents had respectively 3.81% for the bmr donor and 3.75% 
108 
 
B.E kcal/kg DM
BCS1 - 18
BCS1 - 60
BCS1 - 66
BCS1 - 68
BCS1 - 71
BCS1 - 82
BCS1 - 88
BCS1 - 92
BCS1 - 95
BCS1 - 102
BCS1 - 118
BCS1 - 127
BCS2 - 138
BCS2 - 145
BCS2 - 172
BCS2 - 199
BCS2- 201
BCS2 - 203
BCS3 - 192
BCS3 - 210
EL mota
Local Tillabery
Redlanbmr6
Sepon82
Tx630bmr12
Wheatlandbm…
University of Ghana http://ugspace.ug.edu.gh
for El mota. The t-test revealed a t-value and p-value of 1.51; 0.156 respectively. The results 
indicate no statistical difference between the RP mean and its’ derivative progenies’ means. 
In population2, the CP content varied from 2.63% (BCS2-201) to 7.00% (BCS2-172). Results of 
their two parental lines were 5.31% for the bmr donor and 3.94% for Sepon82. The results from 
the t-test indicate a statistical equality between the RP mean and the derivative progenies means: 
(t and p-values were respectively: 1.27 and 0.251). 
In population3, the CP content varied from 4.06% (BCS3-210) and 4.13% (BCS3-192). Parental 
lines produced respectively 5.44% for the bmr donor and 3.94% for Sepon82. The t-test revealed 
that the RP mean was not greater than the derivative progenies mean (t-value=1.86 and p-
value=0.204). 
The Tillabery farmer landrace with 4.25% had more CP than sepon82 which was followed by El 
mota. Among all the entries, the families from population2 (BCS2-172 and BCS2-138) exhibited 
the highest CP content (figure 6.3.) 
Crude Proteins content
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
Derived bmr families and checks
 
Figure 6. 3. Crude proteins content of the selected families and the checks. 
109 
 
Quantity (%)
BCS1 - 18
BCS1 - 60
BCS1 - 66
BCS1 - 68
BCS1 - 71
BCS1 - 82
BCS1 - 88
BCS1 - 92
BCS1 - 95
BCS1 - 102
BCS1 - 118
BCS1 - 127
BCS2 - 138
BCS2 - 145
BCS2 - 172
BCS2 - 199
BCS2- 201
BCS2 - 203
BCS3 - 192
BCS3 - 210
EL mota
Local Tillabery
Redlanbmr6
Sepon82
Tx630bmr12
Wheatlandbmr12
University of Ghana http://ugspace.ug.edu.gh
 Minerals:  Minerals, particularly the macro minerals play an important role in cattle 
correct nutrition. Generally, the calcium/phosphorus ratio is reported as important because an 
imbalance can cause infertility. 
In population1 the calcium content varied from 0.10% (BCS1-127) to 0.39% (BCS1-68). 
Tx630bmr12 the bmr donor line produced 0.05% while El mota the recurrent parent had 0.10%. 
The t-test results revealed a t-value of 2.99 while the p-value was 0.011. 
In the population2 the calcium content varied from 0.05% (BCS2-199 and BCS2-172) to 0.15% 
(BCS2-203 and BCS2-138). Redlanbmr6 the bmr donor parent had 0.13% while Sepon82 the 
recurrent parent had 0.15%. There was no statistical difference between the RP mean and the 
progenies mean. Indeed, t-value=1.27 while the p-value=0.251. 
In the population3, the calcium content varied from 0.04% (BCS3-210) to 0.15% (BCS3-192). 
Their parental lines had respectively 0.21% for Wheatlandbmr12 the bmr donor and 0.15% for 
Sepon82 the recurrent parent. The t-test revealed a t-value=1.86 and p-value=0.204. The calcium 
content was therefore homogenous in the population3. 
The Tillabery farmer landrace with 0.17% had the highest calcium content among the local lines 
and was followed by Sepon82 and finally by El mota. 
In population1 the phosphorus content fluctuated from 0.10% (BCS1-118 and BCS1-92) to 0.32% 
(BCS1-127). Tx630bmr12 the bmr donor line contained 0.05% while El mota the recurrent parent 
held 0.21%. The t-test revealed no statistical difference between the RP and the derivative 
progenies. Indeed, t-value=-1.01 while the p-value=0.333. 
In population2, the phosphorus content ranged from 0.06% (BCS2-203) to 0.27% (BCS2-199). 
Redlanbmr6 the bmr donor parent contained 0.13% while Sepon82 the recurrent parent held 
110 
 
University of Ghana http://ugspace.ug.edu.gh
0.17%. No statistical difference was observed between the RP and its’ progenies: t-value=0.1 while 
p-value=0.926. 
In population3, the phosphorus content varied from 0.02% (BCS3-210) to 0.23% (BCS3-192). 
Wheatlandbmr12 the bmr donor contained 0.12% while Sepon82 the recurrent parent had 0.17%. 
No statistical difference was observed between the RP and its’ progenies: t-value=-0.48 while p-
value=0.678. 
Among the local lines, the Tillabery farmer landrace with 0.25% had the highest values and was 
followed by Sepon82 and finally El mota. 
Calcium and Phosphorus content
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
Derived bmr families and checks
Calcium (%) Phosphorus (%)
 
Figure 6. 4. Calcium and Phosphorus content of the selected families and the checks. 
 Fibers content: Fiber can be described as the carbohydrate fraction of the feed that cannot 
be digested. Indeed, fibers are resistant to digestive enzymes. Fibers are mostly composed of 
111 
 
Quantity (%)
BCS1 - 18
BCS1 - 60
BCS1 - 66
BCS1 - 68
BCS1 - 71
BCS1 - 82
BCS1 - 88
BCS1 - 92
BCS1 - 95
BCS1 - 102
BCS1 - 118
BCS1 - 127
BCS2 - 138
BCS2 - 145
BCS2 - 172
BCS2 - 199
BCS2- 201
BCS2 - 203
BCS3 - 192
BCS3 - 210
EL mota
Local Tillabery
Redlanbmr6
Sepon82
Tx630bmr12
Wheatlandbmr12
University of Ghana http://ugspace.ug.edu.gh
cellulose, neutral detergent fiber, acid detergent fiber and acid detergent lignin. Each of the cited 
element was measured and the following results were obtained. 
- Cellulose content: Cellulose is an important structural component of the plants cell walls. 
The cellulose content obtained fluctuated from 28.56% (BCS1-92) to 32.56% (BCS1-68) in the 
population1. Tx630bmr12 the bmr donor parent had 30.53% while El mota the recurrent parent 
held 34.70%. The t-test revealed a very highly statistical difference between the RP and its’ 
derivative progenies. Indeed, the t-value=-7.95 while p-value=0.000. 
In the population2, the cellulose content ranged from 24.2% (BCS2-203) to 31.75% (BCS2-172). 
Redlanbmr6 the bmr donor parent contained 31.58% while Sepon82 held 33.31%. The t-value was 
-3.5 while the p-value was 0.013. A statistical difference exists in population2 for the cellulose 
content. 
In population3, the cellulose content varied from 26.20% (BCS3-192) to 28.31% (BCS3-210). 
Wheatlandbmr12 the bmr donor contained 33.83% while Sepon82 had 33.31%. The results from 
the t-test showed an uniformity of cellulose content between the RP and the progenies: t-value=-
1.91 while the p-value=0.196. 
Among the local lines, El mota had the highest cellulose content followed Sepon82 and finally the 
Tillabery farmer landrace. Several families from population1 exhibited high cellulose content. 
- Neutral Detergent Fiber (NDF): In population1 the NDF content fluctuated from 55.66% 
(BCS1-71) to 65.17% (BCS1-60) while for the parental lines the NDF values were 62.77% for 
Tx630bmr12 the bmr donor and 67.15% for El mota the recurrent parent. A highly statistical 
difference exists between the RP mean and the progenies: t-value=-6.34 and p-value=0.000. 
In population2, BCS2-145 with 53.99% had the lowest NDF and BCS2-172 the highest value 
(62.15%). Redlanbmr6 the bmr donor held 62.58% while Sepon82 contained 62.45%. The t-test 
112 
 
University of Ghana http://ugspace.ug.edu.gh
indicate a statistical difference between the RP mean and the progenies: t-value=-3.34 and p-
value=0.016. 
In population3 were 59.29% for BCS3-192 and 63.32% for BCS3-210. Wheatlandbmr12 the bmr 
donor contained 64.82%. No statistical difference was observed between the RP and the progenies 
in the population3: t-value=-0.62 while the p-value=0.597. 
Among the local lines, El mota had the highest NDF value followed respectively by the Tillabery 
farmers’ landrace (64.90%) and finally Sepon82. 
NDF %
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
Derived bmr families and checks
 
Figure 6. 5. NDF content of the selected families and checks. 
- Acid Detergent Fiber (ADF): In the population1 the ADF values fluctuated from 31.43% 
(BCS1-102) to 37.60% (BCS1-60). Tx630bmr12 the bmr donor contained 33.95% while El mota 
had 38.23%. A very highly statistical difference was observed between the RP and its’ derivatives 
progenies: t-value=-6.62 and p-value=0.000.  
In the population2, the ADF values ranged from 28.08% (BCS2-138) to 33.39% (BCS2-199) while 
the results of their two parental lines were respectively 33.38% for Redlanbmr6 and 33.56% for 
113 
 
%NDF
BCS1 - 18
BCS1 - 60
BCS1 - 66
BCS1 - 68
BCS1 - 71
BCS1 - 82
BCS1 - 88
BCS1 - 92
BCS1 - 95
BCS1 - 102
BCS1 - 118
BCS1 - 127
BCS2 - 138
BCS2 - 145
BCS2 - 172
BCS2 - 199
BCS2- 201
BCS2 - 203
BCS3 - 192
BCS3 - 210
EL mota
Local Tillabery
Redlanbmr6
Sepon82
Tx630bmr12
Wheatlandbmr12
University of Ghana http://ugspace.ug.edu.gh
Sepon82. The t-test indicate a statistical difference between the RP mean and the progenies: t-
value=-2.75 and p-value=0.033. 
In population3, the ADF values were 31.85% (BCS3-192) and 34.04% (BCS3-210). 
Wheatlandbmr12 the bmr donor contained 37.61%. No statistical difference was observed between 
the RP and the progenies in the population3: t-value=-0.62 and p-value=0.600 
Among the local lines, El mota had the highest ADF value followed by the Tillabery farmers 
landrace (34.85%) and finally Sepon82. 
ADF %
45.00
40.00
35.00
30.00
25.00
20.00
15.00
10.00
5.00
0.00
Derived bmr families and checks
 
Figure 6. 6. ADF content of the selected families and the checks. 
- Acid Detergent Lignin (ADL): In the population1, the ADL values fluctuated from 1.70% 
(BCS1-118) to 4.45% (BCS1-127). Tx630bmr12 the bmr donor contained 2.56% while El mota 
the recurrent parent had 4.67%. A very highly statistical difference was observed between the RP 
and its’ derivatives progenies: t-value=-7.34 and p-value=0.000. 
In the population2, the ADL values ranged from 1.93% (BCS2-145) to 3.93% (BCS2-138) while 
their parental lines had respectively 2.83% for Redlanbmr6 and 4.06% for Sepon82. The t-test 
114 
 
%ADF
BCS1 - 18
BCS1 - 60
BCS1 - 66
BCS1 - 68
BCS1 - 71
BCS1 - 82
BCS1 - 88
BCS1 - 92
BCS1 - 95
BCS1 - 102
BCS1 - 118
BCS1 - 127
BCS2 - 138
BCS2 - 145
BCS2 - 172
BCS2 - 199
BCS2- 201
BCS2 - 203
BCS3 - 192
BCS3 - 210
EL mota
Local Tillabery
Redlanbmr6
Sepon82
Tx630bmr12
Wheatlandb…
University of Ghana http://ugspace.ug.edu.gh
indicate a statistical difference between the RP mean and the progenies: t-value=-3.45 and p-
value=0.014 
In population3, the ADL values varied from 2.65% (BCS3-192) to 3.45% (BCS3-210). No 
statistical difference was observed between the RP and the progenies: t-value=-1.65 and p-
value=0.241. 
Between the local lines, the Tillabery farmers’ landrace had the highest ADL content with 5.17% 
followed by El mota 4.67% and finally Sepon82 4.06%. 
ADL %
6.00
5.00
4.00
3.00
2.00
1.00
0.00
Derived bmr families and checks
 
Figure 6. 7. ADL content of the selected families and the checks. 
 
 
 
 
 
115 
 
%ADL
BCS1 - 18
BCS1 - 60
BCS1 - 66
BCS1 - 68
BCS1 - 71
BCS1 - 82
BCS1 - 88
BCS1 - 92
BCS1 - 95
BCS1 - 102
BCS1 - 118
BCS1 - 127
BCS2 - 138
BCS2 - 145
BCS2 - 172
BCS2 - 199
BCS2- 201
BCS2 - 203
BCS3 - 192
BCS3 - 210
EL mota
Local Tillabery
Redlanbmr6
Sepon82
Tx630bmr12
Wheatlandbm…
University of Ghana http://ugspace.ug.edu.gh
Table 6. 3. Results t-test population one. 
Variable N Mean 95% C.I SEM T P 
DM 13 95.7592 (95.3494, 96.1690) 0.1881 2.71 0.019 
Ash 13 7.33846 (6.45940, 8.21752) 0.40346 5.82 0 
Cellulose 13 31.3415 (30.4209, 32.2621) 0.4225 -7.95 0 
CP 13 4.14615 (3.57610, 4.71621) 0.26164 1.51 0.156 
Lipids 13 1.34385 (1.03038, 1.65731) 0.14387 5.31 0 
Ca 13 0.180769 (0.121898, 0.239641) 0.02702 2.99 0.011 
P 13 0.193077 (0.156536, 0.229618) 0.016771 -1.01 0.333 
NDF 13 61.1585 (59.1007, 63.2162) 0.9444 -6.34 0 
ADF 13 34.3892 (33.1250, 35.6534) 0.5802 -6.62 0 
ADL 13 2.86923 (2.33474, 3.40372) 0.24531 -7.34 0 
E.B 13 3420.2 (3249.76, 3590.65) 78.23 -2.44 0.031 
N=Sample size; C. I=Confidence interval; SEM=Standard error of the mean; p<0.05 
 
Table 6. 4. Results t-test population two 
Variable N Mean 95% C.I SEM T P 
DM 7 95.365 (94.8799, 95.8516) 0.1986 0.23 0.826 
Ash 7 9.03143 (6.52659, 11.53627) 1.02367 -1.9 0.106 
Cellulose 7 29.0443 (26.0631, 32.0255) 1.2183 -3.5 0.013 
CP 7 4.64429 (3.28831, 6.00026) 0.55416 1.27 0.251 
Lipids 7 1.64857 (1.32231, 1.97483) 0.13334 -0.24 0.821 
Ca 7 0.11 (0.065646, 0.154354) 0.018127 -2.21 0.069 
P 7 0.172857 (0.100287, 0.245427) 0.029658 0.1 0.926 
NDF 7 57.9786 (54.7000, 61.2572) 1.3399 -3.34 0.016 
ADF 7 31.1314 (28.9728, 33.2900) 0.8822 -2.75 0.033 
ADL 7 2.97857 (2.21261, 3.74453) 0.31303 -3.45 0.014 
E.B 7 3596.04 (3516.29, 3675.79) 32.59 2.17 0.073 
N=Sample size; C. I=Confidence interval; SEM=Standard error of the mean; p<0.05. 
 
 
 
 
 
 
116 
 
University of Ghana http://ugspace.ug.edu.gh
Table 6. 5 Results t-test population three. 
Variable N Mean 95% C.I SEM T P 
DM 3 94.9833 (93.8801, 96.0866) 0.2564 -1.31 0.32 
Ash 3 8.17333 (1.86165, 14.48502) 1.46693 -1.91 0.196 
Cellulose 3 29.2733 (20.2023, 38.3444) 2.1082 -1.91 0.196 
CP 3 4.04333 (3.80463, 4.28203) 0.05548 1.86 0.204 
Lipids 3 1.35 (0.63735, 2.06265) 0.16563 -1.99 0.185 
Ca 3 0.113333 (-0.044431, 0.271097) 0.036667 -1 0.423 
P 3 0.14 (-0.128701, 0.408701) 0.06245 -0.48 0.678 
NDF 3 61.6867 (56.4186, 66.9547) 1.2244 -0.62 0.597 
ADF 3 33.15 (30.2904, 36.0096) 0.6646 -0.62 0.6 
ADL 3 3.38667 (1.63006, 5.14328) 0.40826 -1.65 0.241 
E.B 3 3630.43 (3402.91, 3857.96) 52.88 1.99 0.185 
N=number of entries; C.I=confidence interval; SEM=Standard Error of the Mean; p<0.05. 
 
6.4. Discussion 
The results of the current study revealed high values of dry matter content of the selected families 
(94.05% to 96.67%). The results of this study are higher than those of Sultan et al. (2011). Indeed, 
those authors found 88.38% dry matter in sorghum stover dry roughage from common crops 
residues fed to livestock in India. In cereal crops, the dry matter content reveals an important aspect 
as it can be considered as the overall indicator of the total amount of nutrient available for the 
animal. Furthermore, in Niger farmers and stover traders preferred varieties with high biomass 
potential (Diakité et al., 2017). With regards to the differences between the parental lines and the 
derived progenies, a tendency of increased dry matter yield was observed across all three 
populations. These findings are in contrast to those of Beck et al. (2014) who showed that dry 
matter at harvest was greater for the non bmr Sweet Sunny Sue (SSS) than NutriPlus (NP-BMR), 
and Dry Stalk bmr (DS-BMR). However, the authors affirmed that the superiority of the nonbmr 
SSS versus the two bmr types was less than 3%. In this study the elite families were BCS1-60, 
117 
 
University of Ghana http://ugspace.ug.edu.gh
BCS1-68, BCS1-18, BCS1-127, BCS1-95, BSC2-138, BCS1-88, BCS2-172. Those families’ 
performances were 1.45%, 1.22%, 1.17%, 0.8%, 0.73%, 0.7%, 0.65%, 0.54% respectively superior 
to the parental lines’ performance. The best families belong mostly to population1. The 
phenotyping results showed that population1 contains the tallest families. Costa et al. (2016) 
reported that the dry matter production of forage sorghum is generally correlated with plant height 
and that is, the dry matter production potential increases with height. The low progress observed 
can be explained by the complexity of the trait. Indeed, it is generally reported that if a character 
is governed by non-additive gene action, it may give high heritability but low genetic advance as 
confirmed by the moderate heritability and low genetic advance observed for this trait during its’ 
phenotyping. The net energy values for this study varied from 2893.41 kcal/kg DM to 3755.02 
kcal/kg DM. For sorghum cultivars stovers, Singh et al. (2018) found 0.95-1.33 kcal/g DM less 
than current study findings. The energy content of cereal stovers remains of great importance in 
feed diet determination. According to Moehn et al. (2005) the net energy is the most precise and 
unbiased way of a feed's energy value. However, FAO (2012) asserted that crop residues are also 
low in energy content, which therefore has to be supplied from concentrate feeds. Families superior 
to their parental lines were identified: BCS1-71 (7.22%), BCS1-68 (6.33%), BCS1-18 (5.29%), 
BCS1-66 (4.62%); BCS2-201 (3.82%), BCS2-145 (2.24%), BCS3-210 (8.47%) and BCS3-192 
(7.92%).  
The results of this study showed that the recurrent parental lines El mota with 3.75% and Sepon82 
with 3.94% exhibited low levels of Crude Proteins (CP) content. However, the results from 
different derived breeding populations varied from (2.63% to 7%) indicating a low to acceptable 
levels of CP in comparison to the acceptable standards and significant level of improvement. The 
results of this study for the CP content are in harmony with Sultan et al. (2011); Singh et al. (2018). 
118 
 
University of Ghana http://ugspace.ug.edu.gh
Indeed, these authors found that dry roughages mainly from the cereal straw and stover particularly 
from tropical countries are usually low in CP and high in cell-wall constituents. CP content of their 
tested dry roughages except one (LS), were below the ruminant maintenance requirement. They 
found CP content of 3.7 to 6.7% in sorghum stovers. Kubkomawa et al. (2015) and asserted that 
cereals straw or residues are inherently low in CP. They also found that cows generally require 
crude protein in the range of 7 to 14% of daily dry matter intake with lactating cows requiring 
especially more. Beck et al. (2014) found that concentrations of CP decreased with increasing 
phenological growth stage. Moreover, Elliot and Topps (1964) reported that maintenance 
requirements of Digestible CP (DCP) of cattle can be reduced by supplementing low quality 
roughages with concentrate, and the practice is normally undertaken during the long dry season. 
In this study, some families from population2 had higher CP content (5.81% and 7%) compared 
to the other two populations.  Badve et al. (1993) also found varietal differences for sorghum stover 
quality for CP content and cell wall concentrations. The superior families among the three bmr 
breeding populations with percentage of improvement compared to the average level of their 
parental lines were respectively: BCS2-172 (51.51%), BCS1-82 (48.94%), BCS1-95 (42.32%), 
BCS1-92 (40.47%), BCS1-71 (37.30%), BCS2-138 (25.75%), BCS1-88 (9.25%), BCS2-203 
(8.22%) and BCS1-66 (5.82%).  
The findings from this study showed differences between bmr breeding populations with low 
levels of calcium and phosphorus corroborating with those of Kubkomawa et al. (2015). These 
authors asserted that cereal hays and silages and such crop residues are relatively low in calcium. 
For a given diet, Kubkomawa et al. (2015) suggested that the total ration should provide a 
calcium/phosphorus ratio of 1.2 to 2:1, with cows at minimum of 1.2:1. The current results of this 
study are insufficient for such requirement. However, phosphorus has been described as the most 
119 
 
University of Ghana http://ugspace.ug.edu.gh
prevalent mineral deficiency for grazing cattle worldwide (Onyeonagu et al., 2013). Khan et al. 
(2005) reported that levels of minerals in plants is a function of interaction between several factors 
which include soil type, plant species, stage of maturity, dry matter yield, grazing management 
and climate.  
A diverse distribution pattern was observed for the fibers content in the current study. For the 
cellulose content, the bmr breeding populations ranged between 24.20% to 32.56% while El mota 
and Sepon82 had respectively 34.70% and 33.31%. This results are lower than Singh et al. (2018) 
who found 27.9% to 33.8%. Harper and McNeill (2015) reported that among the fiber elements, 
the NDF plays a considerable role in forage intake and digestibility. Schroeder (2004) concluded 
that NDF increases with the advancing maturity of forages and as the NDF in forages increases, 
animals will be able to consume less forage. The NDF content in the bmr breeding populations 
ranged from 53.99% (BCS2-145) to 65.17% (BCS1-60) whereas El mota and Sepon82 had 67.15% 
and 62.45%. Singh et al. (2018) found 55.0% to 68.2% in stover of sorghum cultivars. The NDF 
content in the bmr breeding populations are lower than 59.9% to 79.3% found by Hamed et al. 
(2015) in sorghum stover. The low NDF content can be attributable to the improvement made with 
the introgression of the bmr genes in the breeding populations. In this view, Beck et al. (2014) also 
found and asserted that at all phenological growth stages, bmr hybrids were predicted to contain 
less NDF than nonbmr. Furthermore, in a study of comparison of the effect of the sorghum bmr6, 
bmr12 genes in six grain sorghum lines for their grain yield, stover yield and stover quality, Oliver 
et al. (2005) found on average more NDF content in the wild-type compared to its derivative 
counterpart. Fritz et al. (1981) published similar results. The current results are therefore in 
harmony with those authors. The ADF content of the bmr breeding populations varied from 
28.08% (BCS2-138) to 37.60% (BCS1-60) while for the recurrent parents it was 38.23% (El mota) 
120 
 
University of Ghana http://ugspace.ug.edu.gh
and 33.56% (Sepon82). Singh et al. (2018) found 35.3% to 43.1% in stover of sorghum cultivars. 
All bmr breeding families contain less ADF than their respective recurrent parents. For the same 
element Oliver et al. (2005) found similar results. Casler et al. (2003) also reported that bmr 
Sudangrass hybrids contained 3% less ADF than normal hybrids when harvested at heading stage. 
For the sorghum x Sudangrass hybrids, Fritz et al. (1988) found that presence of the bmr gene 
decreased ADF concentration by 8 to 10%. The results of the current study are thus in harmony 
with those authors. However, McCollum et al. (2005) reported no difference in ADF concentration 
between nonbmr and bmr forage sorghum varieties. The ADL content in the bmr breeding 
populations ranged from 1.70% (BCS1-118) to 4.45% (BCS1-127) while the two recurrent parents 
had 4.67% and 4.06% respectively for El mota and Sepon82. Singh et al. (2018) found 4.33% to 
5.79% lignin in stover of sorghum cultivars. All the bmr breeding families contain less ADL than 
their respective recurrent parents. For the same element Oliver et al. (2005) found similar results.   
The cellulose, NDF, ADF and ADL content were very highly significant between the recurrent 
parent El mota and its’ derivatives bmr progenies (population1) and significant in population2. On 
the contrary, in the population3, the recurrent parent Sepon82 and its’ progenies were statistically 
equal. Current results are in harmony with (Badve et al., 1993; Hamed et al., 2015).  
To sum up, the NDF, ADF and ADL content of this study ranged from 53.99% to 65.17%; 28.26% 
to 37.60% and 1.70% to 4.45% respectively. For the same elements, Singh et al. (2018) found 
55.0% to 68.2%; 35.3% to 43.1%; and 4.33% to 5.79% respectively. The current findings are 
therefore inferior to those authors’ results. Moreover, it is obvious that the lignin content in the 
bmr breeding populations is lower than their findings. 
  
121 
 
University of Ghana http://ugspace.ug.edu.gh
6.5. Conclusion 
 
Several bmr derived families exhibited an acceptable to good levels of dry matter, net energy, and 
crude protein. A good level of ADL reduction was observed in the bmr derived families. However, 
their minerals content (calcium and phosphorus) remains low and may imply supplementation 
efforts. For their fiber content, a great variability was found across the different families. The 
relatively lower cellulose content in the bmr breeding populations compared to their parental lines 
may suggest better digestibility quality. Some families (BCS1-102, BCS1-66, BCS2-138, BCS2-
145, BCS2-203, BCS2-201, BCS3-192, BCS3-210) combined well the considerable lignin 
reduction without impacting their fitness to an acceptable level of NDF and ADF.  
 
 
 
 
 
 
 
 
 
 
 
 
  
122 
 
University of Ghana http://ugspace.ug.edu.gh
CHAPTER SEVEN 
 
7.0. EMS SORGHUM SNP MUTATIONS IDENTIFICATION 
7.1. Introduction 
The molecular identification of allelic variation responsible for induced mutant phenotypes is one 
of the most effective and convincing methods to demonstrate gene function (Addo-Quaye et al., 
2016a). According to Vermerris, (2008) mutant phenotypes are of great help in the identification 
of genes that are involved in a product of interest. These make bmr mutations induction a 
sustainable alternative breeding method for dual purpose sorghum production and stover quality 
improvement. Among the chemical agents, the alkylating agent (EMS) is the most commonly used 
in plants because of its potency and ease to use. Another advantage of EMS is the low level of 
chromosomal breaks and lethal effect (Greene et al., 2003). Practically, the ethyl group of EMS 
reacts with guanine in DNA, forming the abnormal base O6-ethylguanine. During DNA 
replication, DNA polymerases that catalyse the process frequently place thymine, instead of 
cytosine, opposite O6-ethylguanine. Following subsequent rounds of replication, the original G:C 
base pair can become an A:T pair (a transition mutation); this changes the genetic information 
(Wikipedia.org). For Per Sikora et al. (2011), chemical mutagens tend to cause single base-pair 
(bp) changes or single-nucleotide polymorphism (SNPs) and provide a very high mutation 
frequency. In plant improvement and genetics, mapping the genes responsible for traits remains 
one of the most basic goals (Brauer et al., 2006). A large number of EMS mutants (~12,000) 
derived from Tx623 were generated at Purdue University (Addo-Quaye et al., 2016b). Selection 
for agronomic potential was further conducted and finally 554 EMS mutant lines were evaluated 
123 
 
University of Ghana http://ugspace.ug.edu.gh
in the Niger sahelian environment. Ten (10) EMS mutants were finally selected based on their 
clear brown midrib colour exhibition during the rainy season 2015.  
In Niger, proper animal feeding is one of the major challenge for livestock productivity. The main 
objective of this chapter was to identify the EMS candidate SNP mutations responsible of the bmr 
trait from those ten (10) EMS mutant lines to speed up sorghum stover quality improvement. 
The specific objectives were to: 
-i identify the nucleotide change and the type of mutation observed, 
-ii identify the amino acid change, 
-iii classify the different EMS mutants’ phenotype. 
7.2. Material and methods 
7.2.1. Plant materials 
The plant materials were composed of twelve (12) sorghum entries set up as follow: ten (10) 
Sorghum bicolor EMS (SbEMS) families and two (2) bmr6 and bmr12 sources. The ten (10) 
families were crossed to Tx623 for F1 production. The obtained F1 were self-pollinated to produce 
F2 seeds. In the F2 segregating nurseries, selection based on the midrib colour was conducted, thus 
only the genotypes exhibiting the bmr character were self-pollinated. The two (2) bmr sources 
(Tx631 bmr6 x N223 and Tx631 bmr12 x N223) were developed in the same breeding conditions 
and used as checks. The progenies of every cross were considered as a family therefore the entire 
population was constituted of twelve (12) families. The Table 7.1 shows the entries pedigrees and 
family sizes 
 
124 
 
University of Ghana http://ugspace.ug.edu.gh
Table 7. 1. Pedigree and size of the different EMS families. 
Number Entries Source Family Size 
1 SbEMS 1453 mut-1 - Niger 13 x Tx623 WL16-2700-1 26 
2 SbEMS 2354 mut-1 - Niger 18 x Tx623 WL16-2704-1 13 
3 SbEMS 3779 mut-1 - Niger 27 x Tx623 WL16-2708-1 27 
4 SbEMS 4465 mut-1 - Niger 29 x Tx623 WL16-2712-1 21 
5 SbEMS 4791 mut-2 - Niger 31 x Tx623 WL16-2720-1 9 
6 SbEMS 4848 mut (#1)-1 - Niger 32 x Tx623 WL16-2724-1 21 
7 SbEMS 5095 (#1)-1 - Niger 34 x Tx623 WL16-2752-1 32 
8 SbEMS 5199 mut (#1)-1 - Niger 38 x Tx623 WL16-2728-1 16 
9 SbEMS 5398 mut-1 - Niger 39 x Tx623 WL16-2732-1 24 
10 SbEMS 932-1-1 - Niger 553 x Tx623 WL16-2736-1 18 
11 Tx631 bmr-6 x N223 WL16-2762-1 60 
12 Tx631 bmr-12 x N223 WL16-2795-1 64 
 
7.2.2. Methodology 
The seedlings were cultivated in a green house. Each family was composed of many plots (9 to 
64: Table 7.1.). In each plot twelve (12) seeds were sown on a slightly alkaline soil (pH=7.8) for 
germination. The green house was kept at 27°C and seedlings were irrigated daily with tap water. 
Labels were used to distinguish each family. Two weeks after germination, miracle Growth 
fertilizer containing both macro and micro nutrients was applied to maintain seedlings in good 
growing condition.  
7.2.2.1. Leaf tissue sampling 
DNA extraction was conducted on three weeks old seedlings. To avoid any contamination among 
families, sampling was done by wearing laboratory gloves. In order to form a bulk of leaves of 
each family, one plant per plot was randomly selected considering that in each plot all the plants 
carry the same mutation. Subsequently, this bulk made up of equal size of pieces of leaves was 
125 
 
University of Ghana http://ugspace.ug.edu.gh
designed to represent every family. Twelve (12) envelopes containing each family were kept in 
liquid nitrogen from the green house to the storage at -80°C in the laboratory. 
7.2.2.2.DNA extraction  
For the DNA extraction a Cetyltrimethylammonium bromide (CTAB) protocol method was used. 
In practice, every family’s bulk of leaves was ground separately in a cold mortar and the entire 
powder was used for DNA extraction. Thus DNA obtained was a pooled DNA and was 
representative of each family. This allows the capture of all the mutations in a family during a 
single run sequencing. Kyle et al. (2011) asserted that by analysing DNA from a mixture of 
phenotypically mutant progeny from the cross, it is possible to define the genetic region containing 
the mutation and even to identify the mutation itself. DNA quantification and quality 
measurements were also performed; those measurements are important to get an optimal 
performance during the sequencing step.  
7.2.2.3. DNA quality and quantity measurements 
The DNA quantities and qualities were checked twice in order to satisfy the Illumina sequencing 
for the whole genome sequence requirements. Firstly, with the electrophoresis technique using 
1µg DNA per family on agarose gel. This method was used to determine the presence or absence 
of the DNA in all the samples. The agarose gel was at 0.8% concentration and the electrophoresis 
duration was 1 hour at 100 Volts; 400 mA. Secondly the DNA amount was estimated with a 
spectrophotometer (NanoDrop 1000) using 1µl of no diluted DNA sample from every family. The 
spectrophotometer was set to the default that is DNA-50. 
The DNA bands from the agarose gel electrophoresis were appreciated by visualization under 
ultraviolet light. For the spectrophometer, the ratios of absorbance at 260 nm/280 nm was used to 
126 
 
University of Ghana http://ugspace.ug.edu.gh
assess the DNA purity and 260/230 to measure the nucleic acid purity. The Spectrophotometer 
gives also the DNA concentration based on the absorbance at 260 nm. Indeed, DNA has a maximal 
absorbance near 260 nm, higher levels of absorbance are indicative of greater concentrations of 
DNA present within the samples (www.lgcgroup.com/genomics). For the first ratio (260/280), a 
value of ~1.8 is generally accepted as ‘pure’ DNA while nucleic acid purity was accepted in the 
range 1.8 to 2.2 looking at the ratio 260/230. (Thermo Scientific 2008). All the DNA samples were 
checked to Electropherogram for the sequencing requirements and were further sequenced with a 
>8.33x genome coverage. In order to identify the causative nucleotide change candidate of the bmr 
phenotype observed in the derived EMS sorghum families the Whole Genome Sequencing results 
were analyzed with the MutMap method. 
7.2.2.4. MutMap analysis 
The following bioinformatics softwares were used during the sequencing data processing: 
MobaXterm v10-9 for the Unix command utilization and sequencing data transfer in the WinSCP.   
Akira et al. (2012) proposed and described the MutMap as an elegant and rapid method for the 
identification of nucleotides changes in plant genomes. The method involves F2 segregating 
populations derived from a cross between a mutant line and the wild type (the reference genome) 
used to generate the mutant. This MutMap is however based on the whole genome resequencing 
of pooled DNA from segregating population of plant that show a useful marker (Akira et al., 2012).  
In practice, the EMS mutants’ sequencing data processing started firstly with the download of the 
BTx623 (Rooneys’ data) parent from European BioInformatic. This BTx623 was used as the 
Sorghum bicolor reference genome and each EMS family was aligned to BTx623 for the MutMap 
configuration. A high quality read was selected for the genome sequencing data using the ml bwa 
127 
 
University of Ghana http://ugspace.ug.edu.gh
and ml samtools/0.1.8. The BWA (Burrows-Wheeler Aligner) is a software package for mapping 
low-divergent sequences against a large reference genome (Li and Durbin, 2009). The 
samtools/0.1.8. are genetic format for storing large nucleotide sequence alignments (Kwangsik, 
2014). Then to develop a reference sequence for one of the parental lines used for crossing, the 
module load gcc/5.2.0 r/3.3.1 was used. Consensus map was made using QTL-
seq_test/2.make_consensus/90.align_to_this_fasta with module load gcc/5.2.0 r/3.3.1. Figure 7.1. 
describes the MutMap principles.  
 
Figure 7. 1. Simplified scheme for application of MutMap to rice. 
A rice cultivar with a reference genome sequence is mutagenized by ethyl methanesulfonate (EMS). The mutant 
generated, in this case a semi-dwarf phenotype, is crossed to the wild-type plant of the same cultivar used for the 
mutagenesis. The resulting F1 is self-pollinated to obtain F2 progeny segregating for the mutant and wild-type 
phenotypes. Crossing of the mutant to the wild-type parental line ensures detection of phenotypic differences at the 
F2 generation between the mutant and wild type. DNA of F2 displaying the mutant phenotype are bulked and subjected 
to whole-genome sequencing followed by alignment to the reference sequence. SNPs with sequence reads composed 
only of mutant sequences (SNP index of 1) are closely linked to the causal SNP for the mutant phenotype. Source: 
Akira et al. (2012) 
128 
 
University of Ghana http://ugspace.ug.edu.gh
7.2.2.5. SNPs identification 
The Integrative Genomics Viewer (IGV) and the MultAlin (Multiple alignment) softwares were 
used for SNPs identification and amino acid changes across the mutants’ genome. In practice the 
sequencing BAM file of every EMS mutant was compared to the new sorghum reference genome 
(Sorghum bicolor v3.1.1).  Akira et al. (2011) reported that in F2 progeny, the majority of SNPs 
segregate in a 1:1 mutant/wild type ratio except the SNP responsible of the phenotype change 
which is homozygous. After alignment to the reference genome, the SNP-alleles frequencies were 
determined as the percentage of the reference genomes’ allele versus the percentage of the 
alternative allele. The SNPs alleles frequencies follow a 50% mutant genotype and 50% wild type 
genotype for all the majority of the positions across the sequences excepted at causative mutations’ 
position. In total, our number of reads depth varied from 1 to 16. The SNP index were calculated 
in percentage for all the EMS families as defined by Akira et al. (2011): SNP INDEX = Number 
of read of the mutant SNP/Total reads corresponding to the SNP*100.  
All the mutants were firstly screened for the known bmr2 (Sobic.004G062500), bmr6 
(Sobic.004G071000.2) and bmr12 (Sobic007G047300) genes positions. 
7.3. Results 
7.3.1. DNA quantity and quality 
For the DNA quantity and quality requirements, the following results were obtained for every 
sample: Table 7.2 shows results obtained from the spectrophotometer readings of the different 
families while the Figure 7.2 shows the results from the agarose gel. The results obtained revealed 
the presence of DNA that all the DNA extracted from all the twelve (12) families were acceptable 
for sequencing. 
129 
 
University of Ghana http://ugspace.ug.edu.gh
Table 7. 2. Concentration, absorbance of bulk DNA of SbEMS families. 
Concentration Absorbance Absorbance
Number Entries
(ng/µl) 260/280 260/230
1 SbEMS 1453 mut-1 - Niger 13 x Tx623 662.84 1.85 1.62
2 SbEMS 2354 mut-1 - Niger 18 x Tx623 590.62 1.88 2.06
3 SbEMS 3779 mut-1 - Niger 27 x Tx623 513.64 1.85 1.99
4 SbEMS 4465 mut-1 - Niger 29 x Tx623 554.14 1.82 1.92
5 SbEMS 4791 mut-2 - Niger 31 x Tx623 510.52 1.92 2.14
6 SbEMS 4848 mut (#1)-1 - Niger 32 x Tx623 441.94 1.91 1.84
7 SbEMS 5095 (#1)-1 - Niger 34 x Tx623 405.82 1.89 2.1
8 SbEMS 5199 mut (#1)-1 - Niger 38 x Tx623 532.08 1.86 2.1
9 SbEMS 5398 mut-1 - Niger 39 x Tx623 968.72 1.89 2.12
10 SbEMS 932-1-1 - Niger 553 x Tx623 1190.01 1.91 2.34
11 Tx631 bmr-6 x N223 623.15 1.93 2
12 Tx631 bmr-12 x N223 598.89 1.93 2.15  
 
 
Figure 7. 2. DNA bands on the agarose gel. 
Samples loaded in the order above at the indicated amounts. 6ul DNA ladder (1kb) was loaded 
onto the gel on the first and last wells. 
 
130 
 
University of Ghana http://ugspace.ug.edu.gh
7.3.2. SbEMS phenotypes identification 
The following results were obtained for every SbEMS mutant: EMS 2354; 932-1-1 as bmr2, EMS 
1456; 4848; 3779; 4465 as bmr6 and EMS 5095 as bmr12. The EMS 4791 was classified as 
possible double mutant bmr2/bmr6 while EMS 5199 and 5398 were suggested as new bmr 
variants. Below the detailed SNP information on every EMS mutant is described. 
7.3.2.1. Family 1: SbEMS 1453 mut-1 - Niger 13 x Tx623 
The sequence alignment of the SbEMS 1453 mut-1 - Niger 13 with the Sorghum bicolor reference 
genome v3.1.1 revealed a Cytosine:Thymine (C:T) nucleotide change in the Sobic.004G071000.2 
gene on the chromosome 4 at the genetic position 5731439 base pair. Indeed, a Cytosine nucleotide 
was found at the reference genome while for the same position with four read depth a 100% 
Thymine nucleotide was present for the mutant. This transition mutation in the protein coding 
sequence (CDS) resulted in a change of the reading frame TGC/TGT indicating a candidate 
mutation for the bmr6 phenotype. In fact, an amino acid change Alanine to Cysteine was observed 
in the amino acid chain.  
 
Figure 7. 3. The T:C mutation contrasting the EMS1453 mutant to the reference genome. 
7.3.2.2. Family 2: SbEMS 2354 mut-1 - Niger 18 x Tx623 
The sequence alignment of the SbEMS 2354 mut-1 - Niger 18 with the Sorghum bicolor v3.1.1 
reference genome showed a Cytosine:Thymine (C:T) nucleotide change in the Sobic.004G062500 
gene on the chromosome 4 at the genetic position 5042955 base pair. The reference genome carried 
a Cytosine nucleotide while a 100% Thymine nucleotide with three read depth was observed on 
131 
 
University of Ghana http://ugspace.ug.edu.gh
the mutant. The following transition mutation CCC/CTC in the protein coding sequence resulted 
in a modification from Proline to Leucine amino acid. This point mutation constitutes a candidate 
mutation for the bmr2 phenotype.  
 
Figure 7. 4. The T:C mutation contrasting the EMS2354 mutant to the reference genome. 
7.3.2.3. Family 3: SbEMS 3779 mut-1 - Niger 27 x Tx623 
The sequence alignment of the SbEMS 3779 mut-1 - Niger 27 with the sorghum bicolor v3.1.1 
reference genome showed a Guanine:Adenine (G:A) nucleotide change in the 
Sobic.004G071000.2 on the chromosome 4 at the genetic position 5730127 base pair. The 
reference genome contains Guanine while Adenine was present on the mutant’s sequence at this 
position for one read depth. The initial TGG from the reference genome was converted to TAG*, 
therefore the initial amino acid (Proline) became a stop codon. This nonsense mutation in the 
protein coding sequence places the mutant as a candidate for the bmr6 phenotype.  
 
Figure 7. 5. The G:A mutation contrasting the EMS3779 mutant to the reference genome. 
7.3.2.4. Family 4: SbEMS 4465 mut-1 - Niger 29 x Tx623 
The sequence alignment of the SbEMS 4465 mut-1 - Niger 29 with the Sorghum bicolor v3.1.1 
reference genome showed a C:T mutation in Sobic.004G071000.2 sorghum bmr gene. This 
mutation at the genetic position 5732336 base pair with six read depth on the chromosome 4. 
Indeed, a TCC/TCT change was observed in an intron close to the protein coding sequence. This 
132 
 
University of Ghana http://ugspace.ug.edu.gh
mutation suggested a possible splicing site change and indicates the mutant as a candidate bmr6 
phenotype. 
 
Figure 7. 6. The C:T mutation contrasting the EMS4465 mutant to the reference genome. 
7.3.2.5. Family 5: SbEMS 4791 mut-2 - Niger 31 x Tx623 
The sequence alignment of the SbEMS 4791 mut-2 - Niger 31 with the Sorghum bicolor v3.1.1 
reference genome showed two candidate mutations in two different bmr genes.  
 In the Sobic.004G062500 on the chromosome 4 at the genetic position 5042074 base pair, 
a C:T mutation was found. Indeed, while the reference genome exhibited a Cytosine; a 100% 
Thymine was observed on the mutants’ sequence at the corresponding genetic position with 4 read 
depth. As a result of this nonsense mutation in the protein coding sequence, the initial triplet CAG 
coding for the amino acid glutamine was changed into a stop codon (TAG*). This mutation 
resulted to a bmr2 candidate phenotype for this mutant.  
 
Figure 7. 7. The T:C mutation contrasting the EMS 4791 mutant to the reference genome. 
 In the Sobic.004G071000.2 on the chromosome 4 at the genetic position 5730873 base pair 
a C:T mutation was found. Indeed, an alternative T nucleotide was observed instead of C from the 
reference genome. This possible splicing site change with four read depth may also be a candidate 
mutation responsible of a bmr6 phenotype. 
133 
 
University of Ghana http://ugspace.ug.edu.gh
 
Figure 7. 8. The C:T mutation contrasting the EMS 4791 mutant to the reference genome. 
7.3.2.6. Family 6: SbEMS 4848 mut (#1)-1 - Niger 32 x Tx623 
The sequence alignment of the SbEMS 4848 mut (#1)-1 - Niger 32 with the Sorghum bicolor 
v3.1.1 reference genome revealed an Guanine:Adenine (G:A) nucleotide change on the 
chromosome 4 at the genetic position 5731386 base pair for the Sobic.004G071000.2 gene. For 
the four read depth, a 100% Adenine was observed whereas the reference genome presented 
Guanine. This missense mutation in the protein coding sequence resulted a change GCG to ACG. 
Consequently, the expected arginine amino acid was changed to threonine and the mutant was 
classified as a candidate bmr6 phenotype. 
 
Figure 7. 9. The A:G mutation contrasting the EMS4848 mutant to the reference genome. 
7.3.2.7. Family 7: SbEMS 5095 (#1)-1 - Niger 34 x Tx623 
The sequence alignment of the SbEMS 5095 (#1)-1 - Niger 34 with the Sorghum bicolor v3.1.1 
reference genome revealed three different mutations on the chromosome 7 for the 
Sobic007G047300 gene. The mutations were found in the introns. This suggest a possible splicing 
site change. Their read depth was respectively of 7, 9 and 12 at the genetic position 4722728 base 
pair, 4722588 base pair and 4722669 base pair with the following nucleotide change T:C, T:A and 
C:T. The results of those mutations in the Sobic007G047300 gene suggested this mutant as a 
candidate for the bmr12 phenotype. 
134 
 
University of Ghana http://ugspace.ug.edu.gh
 
 
 
Figure 7. 10. The different mutations contrasting the EMS 5095 to the reference genome. 
7.3.2.8. Family 8: SbEMS 5199 mut (#1)-1 - Niger 38 x Tx623 
The sequence alignment of the SbEMS 5199 mut (#1)-1 - Niger 38 x Tx623 with the Sorghum 
bicolor v3.1.1 reference genome did not reveal any homozygous SNP for all the known sorghum 
bmr genes.  
7.3.2.9. Family 9: SbEMS 5398 mut-1 - Niger 39 x Tx623 
The sequence alignment of the SbEMS 5398 mut-1 - Niger 39 x Tx623 with the Sorghum bicolor 
v3.1.1 reference genome did not reveal any homozygous SNP for all the known sorghum bmr 
genes. 
7.3.2.10. Family 10: SbEMS 932-1-1 - Niger 553 x Tx623 
The sequence alignment of the SbEMS 932-1-1 Niger 553 x Tx623 with the Sorghum bicolor 
v3.1.1 reference genome revealed a G:A nucleotide change on the chromosome 4 in the 
Sobic.004G062500 at the genetic position 5042945 base pair. Two read depth was observed and 
135 
 
University of Ghana http://ugspace.ug.edu.gh
GAG/AAG change in the protein coding sequence resulted in glutamic acid to lysine change. This 
mutation places the mutant as a candidate bmr2 phenotype. 
 
Figure 7. 11. The A:G mutation contrasting the EMS932-1-1 mutant to the reference genome. 
136 
 
University of Ghana http://ugspace.ug.edu.gh
Table 7. 3. Summary SNPs results. 
SNP 
Mutation 
Mutant ID Gene Read Chr Position Candidate Phenotype Ref. Allele Alt. Allele Sequence Amino Acid
type
(bp)
SbEMS1453 Sobic.004G071000.2  4 4 5731439 W Bmr6 W C T tgC/tgT A(Alanine)/Cysteine Missense
SbEMS2354 Sobic.004G062500 3 4 5042955 Bmr2 W W C T cCc/cTc P(Proline)/Leucine Missense
SbEMS3779 Sobic.004G071000.2 1 4 5730127 W Bmr6 W G A tGg/tAg* P(Proline)/Stop Codon Nonsense
SbEMS4465 Sobic.004G071000.2 6 4 5732336 W Bmr6 W C T tcC/tcT SSC
SbEMS4791 Sobic.004G071000.2 4 4 5730873 W Bmr6 W T C gTa/gCa SSC
SbEMS4791 Sobic.004G062500 4 4 5042074 Bmr2 W W C T Cag/Tag* Q(Glutamine)/Stop Codon Nonsense
SbEMS4848 Sobic.004G071000.2  4 4 5731386 W Bmr6 W G A Gcg/Acg R(Arginine)/Threonine Missense
SbEMS5095 Sobic007G047300 9 7 4722588 W W Bmr12 T A aTt/aAt SSC
SbEMS5095 Sobic007G047300 7 7 4722728 W W Bmr12 T A aTt/aAt SSC
SbEMS5095 Sobic007G047300 12 7 4722669 W W Bmr12 C T tCg/tTg SSC
SbEMS932-1-1 Sobic.004G062500 2 4 5042945 Bmr2 W W G A Gag/Aag E(Glutamic acid)/Lysine Missense
SbEMS5199 No homozygous SNP data available
SbEMS5398 No homozygous SNP data available  
W=Wild type;  Chr=Chromosome; Ref.=Reference; Alt=Alternative; SSC=Splice Site Change
137 
 
University of Ghana http://ugspace.ug.edu.gh
 
7.4. Discussion 
According to Kharkwal (2012) mutant forms showing both large and small effects on the 
phenotype arise for all kinds of traits. In Purdue University, Addo-Quaye et al. (2016) induced 
EMS mutations in Tx623 and developed a large population for SNPs discovery and gene function 
study. Among the 554 EMS families tested in Niger in 2015, ten (10) families exhibited clearly 
the bmr phenotype. This level of bmr phenotype exhibition (1.8%) is in agreement with Sattler et 
al. (2014). Those authors isolated 46 bmr mutants from M2 generation of EMS mutagenized 
TILLING population of BTx623 out of approximately 3,000 M1 lines. In this view, Sarada (1989) 
reported that the effectiveness of the mutagen agent depends on several conditions such as the 
properties of the tissues treated, the treatment conditions. Furthermore, Konzak et al. (1965) 
observed that pre-soaking results in higher efficiency compared to dry seeds treatment. 
DNA from the 10 EMS bmr sorghum families were sequenced at 8.33x genome coverage and a 
wide variety of mutations were identified across their genome. In the majority the SNPs mutations 
observed were C:T and A:G. In this view, Greene et al. (2003) affirmed that chemical mutagenesis 
by soaking sorghum seeds for example in diethyl sulfate or ethyl methane sulphonate (EMS: 
C3H8SO3) can induce point mutations in plant genome leading to phenotypic changes, which could 
be desirable traits in crop improvement. Ali et al. (2012) indicated that EMS alkylates guanine 
bases and leads to mispairing-alkylated G pairs with T instead of C, resulting in primarily G:C to 
A:T transitions. Bradley et al. (2007) concluded that 70-99% of the changes are G:C to A:T base 
pairs transitions. The results of the current study showed mainly point mutations with transition 
mutations C:T (64%); G:A (27%) and transversion mutation T:A (9%).  Lee et al. (2011) reported 
that transition mutations occurred more frequently than transversion in mutational events. The 
138 
 
University of Ghana http://ugspace.ug.edu.gh
SNP mutations observed resulted in: amino acid changes (36%); stop codons (18%) and possible 
splice site change (45%). The current study found for the bmr6 phenotype amino acid change 
(Alanine to Cysteine, Arginine to Threonine), a possible splice site change and a change of Proline 
to a premature stop codon. Saballos et al. (2009) reported an amino acid change to a premature 
stop codon for the bmr6 phenotype. For the bmr12 phenotype, the current study identified a 
possible splice site change as a causative SNP while Sattler et al. (2012) reported missense 
mutations in the coding region responsible of amino acid change contrasting with the premature 
stop codon previously reported. The results of the SNP identification of the current study revealed 
amino acid change (Proline to Leucine, Glutamic acid to Lysine) and Glutamic acid to a premature 
stop codon for the bmr2 phenotype. Saballos et al. (2012) also reported an amino acid substitution 
for the bmr2 phenotype. 
The results of this study indicate the following EMS 1453, EMS 3779, EMS 4465, EMS 4848 as 
good candidate for the following bmr6 phenotype; the EMS 2354, EMS 932-1-1 for the bmr2 
phenotype and the EMS 5095 for the bmr12 phenotype. The EMS 4791 was found as candidate 
for the stacked bmr2/bmr6 phenotype. A complementation test conducted at Purdue University 
(laboratory 2304 unpublished data) also classified the EMS mutants 1453, EMS 4848, EMS 5398 
for the bmr6 phenotype; EMS 3779, EMS 4791, EMS 5095, EMS 932-1-1 for the bmr2 phenotype 
and the EMS 4465 for the bmr12 phenotype. The results of the SNP identification are therefore in 
agreement with the complementation test for the mutant EMS 1453, EMS 4848, EMS 932-1-1. 
For the EMS 4791 the sequencing data revealed a SNP mutation in the Sobic.004G062500 
resulting in a premature stop codon and a mutation in the Sobic.004G071000.2 (splice site change). 
According to Per Sikora et al. (2011) that mutations in coding regions can be silent, missense or 
nonsense whereas in noncoding regions, mutations can change promoter sequences or other 
139 
 
University of Ghana http://ugspace.ug.edu.gh
regulatory regions resulting in up-or downregulation of gene transcription; in addition, aberrant 
splicing of mRNA, altered mRNA stability and changes in protein translation may also occur as a 
result of mutagenesis. In this view, Black (2003) stated that alternative pre-mRNA splicing is a 
central mode of genetic regulation in higher eukaryotes and variability in splicing patterns is a 
major source of protein diversity from the genome. The SNP identification results for the EMS 
4791 suggest a stacked bmr2/bmr6 phenotype. For forage quality trait, Pedersen et al. (2008) 
suggested that double bmr mutant may contain lesser lignin and may be more digestible than the 
single bmr mutant. However, they revealed a reduction in the total biomass yield of the double 
mutant compared with the single mutants and closest wild-type used. For the EMS 5398, no 
homozygous SNP was found for the reported bmr genes, however the complementation test 
classified the mutant as a bmr6 phenotype. The result from the current study disagrees with the 
complementation test for this mutant.  
For the EMS 5199, no homozygous SNP was found for the reported bmr genes. The 
complementation test also did not classify this mutant for the known bmr phenotypes. The SNP 
identification result is in agreement with the complementation test for this mutant. The findings 
are in agreement with Sattler et al. (2014).  Indeed, for those authors, among a set of 20 EMS 
mutants, six bmr mutants were not allelic to bmr2, bmr6 or bmr12 and a further complementation 
testing showed these mutants represent four novel bmr loci. 
 
 
 
 
140 
 
University of Ghana http://ugspace.ug.edu.gh
7.5. Conclusion 
The bmr phenotype is obviously visible from seedlings to mature plants, thus forward genetic was 
used to identify the candidate mutations. The comparison of the EMS mutants’ DNA to the 
Sorghum bicolor reference genome showed some candidates SNPs mutations. The missense 
mutations resulted in an amino acid changes while the nonsense mutations resulted to premature 
stop codon. The transition mutations observed resulted in general into amino acid change whereas 
the transversion mutation led to a possible splice site change. 
Two of the mutants were found to be different from the reported bmr genes. Further investigation 
(allelism test) may be useful for novel bmr loci identification. 
The EMS bmr mutants will undergo in a backcrossing program before their utilization in the 
sorghum stover improvement program. However, the candidate SNPs discovered needs to be 
confirm.  
 
 
 
 
 
 
 
 
 
  
141 
 
University of Ghana http://ugspace.ug.edu.gh
CHAPTER EIGHT 
 
8.0. GENERAL CONCLUSION AND RECOMMANDATIONS 
8.1. GENERAL CONCLUSION 
Participatory rural appraisal revealed mixed crop-livestock was the main production system for all 
farmers involved in the study zones. Farmers were well aware of the impact of the feed shortage 
on animal performance. Feed shortage during the hot and dry season (March-July) is recurrent. 
Cattle (milking cows and bullocks) were the first to be fed with sorghum residues. Stover with 
greener and enough biomass was preferred by traders and buyers for feeding cattle.    
Three populations composed of 94 families (BC1F3) from three different bmr6 and bmr12 donor 
parents were developed using hand emasculation technique.  
Bmr derived populations (BC1F3) were phenotyped at Tillabery and Konni in Niger for earliness, 
grain yield, dry matter yield, resistance to lodging and foliar diseases. Several bmr BC1F3 families 
exhibited a great potential in grain and dry matter yields better than their recurrent and donor 
parental lines. High yielding dual purpose lines were identified showing stability across locations 
for dry matter and grain yields.  Bmr6 and bmr12 genes in locally adapted sorghum varieties with 
higher stover yield were identified with improved harvest index.  
Bmr derived elite families showed significant improvement in dry matter, NDF, ADF and ADL 
contents.  
EMS bmr lines were tested in the Nigerien environment and their genetic analysis was performed 
at Purdue University.  The genotyping revealed bmr2, bmr6, bmr12 and stacked bmr2/bmr6 
142 
 
University of Ghana http://ugspace.ug.edu.gh
phenotypes as results of the EMS mutagenesis. Bmr6 and bmr12 are already used for sorghum 
stover improvement. The stacked bmr2/bmr6 may result in a useful bmr donor line. For two other 
bmr EMS families, no homozygous SNP were found. This indicate a new mutation in a non-
reported bmr gene due to the EMS mutagenesis. Further investigations may identify the nature of 
their candidate mutations.  
8.2. RECOMMENDATIONS 
Breeding effort should continue to identify elite lines from promising families (BCS1-102, BCS1-
66, BCS1-60, BCS2-138, BCS2-201, BCS2-203, BCS2-145, BCS3-192, BCS3-210). 
More field phenotyping of derived elite lines should continue to study GxE effect. 
Additional backcrossing is needed to decrease recurrent parent background for better digestibility. 
Digestibility study should be undertaken to bring in the market nutritionally valuable lines. 
Develop dual purpose sorghum varieties well adapted with high stover quality.  
Establish stover business in the country by assisting traders with appropriate technics of stover 
conservation and bunching. 
Feeding milking cow, draught animal, small ruminants, etc. for improved productivity should be 
undertaken. 
Establishing pilot study for intensive milk cows farming by feeding them with bmr stover is 
required for efficient use of bmr lines. 
 
 
 
143 
 
University of Ghana http://ugspace.ug.edu.gh
REFERENCES 
Abdou, N. (2010). The effect of supplementing millet stover with groundnut haulms and cereal 
  brans  on Feed intake and growth performance of sheep. University of Kwazulu-Natal. 
  101p. 
Acquaah, G. (2007). Principles of Plant Genetics and Breeding. Second Edition. John Wiley & 
   Sons, Ltd. 584p 
Adamou, M., Axtell, J. D., Clark, W. J. (1985). Niger Sorghum and Millet. Workshop Report of 
  Research Collaboration in Niger. 3-10p. 
Addo-Quaye, C., Tuinstra, M., Carraro, N., Weil, C. and Dilkes, B. P. (2016). Whole genome 
  sequence accuracy is improved by replication in a population of mutagenized sorghum. 
 Purdue University. 82p 
Addo-Quaye, C., Buescher, E., Norman, B., Chaikam, V., Baxter, I. and Dilkes, B. P. (2016). 
 Forward genetics by sequencing EMS variation induced inbred lines. Purdue University. 
 51p 
Akinola, A. A., Ayedun, B., Abubacar, M., Sheu, M. and Abdoulaye, T. 2015. Crop residue usage
  and its determinants in Kano State, Nigeria. Journal of Development and Agricultural 
 Economics. 7(4) 162-173 
Akira, A., Shunichi, K., Kentaro, Y., Satoshi, N., Hiroki, T., Hiroyuki, K., Hideo, M., Kakoto, Y.,
  Chikako, M., Muluneh, T., Hideki, I., Liliana, C., Sophien, K. and Ryohei, T. (2012). 
  Genome sequencing reveals agronomically important loci in rice using MutMap. Nature 
  biotechnology. 30 (2). 174-178 
Alain, M. B. (2000). Lignins and lignification: Selected issues. Plant Physiology Biochemistry 
 38 (1/2), 81-96 
144 
 
University of Ghana http://ugspace.ug.edu.gh
Ali, B. T., Amin, B. T., and Behzad, S. (2012). Ethyl methane sulphonate (EMS) Induced 
 Mutagenesis in Malaysian Rice (cv. MR219) for Lethal Dose Determination. American 
 Journal of Plant Sciences. 3, 1616-1665. 
Allard, R. W. (1960). Principle of plant breeding. John Wiley and Sons, inc., New York.   
Ankom Fiber Analyzer. (2010). Operator’s manual. 28p  
Anterola, A. M., Lewis N., G. (2002). Trends in lignin modification: a comprehensive analysis of 
 the effects of genetic manipulations/mutations on lignification and vascular integrity. 
 Phytochemistry 61: 221-294 
Aschalew, A. (2014). Feed constraint to livestock revolution in Ethiopia: A review. Academic 
  Journal of Nutrition. 3 (2): 15-18 
Awika, J. M., Rooney L., W.  (2004). Sorghum phytochemicals and their potential aspects on 
 human health. Phytochemistry. 65:1199-1221. 
Aydin, G., Grant, R. J., and O’rear, J. (1999). Brown Midrib Sorghum in Diets for Lactating Dairy 
 Cows. Journal of Dairy Science. Vol. 82 (10): 2127-2135 
Ayyangar, G. N. R., and Ponnaiya, B., W. X. (1941). The occurrence and inheritance of a 
 bloomless sorghum. Current Science. 10: 408-409. In Sorghum: Origin, History, 
 Technology and Production. Wayne C., S. and Frederiksen R., A. 2000. 824p. 
Ayyangar G., N., R., Vijiaraghavan C., Pillai V., G. and Ayyar M., A., S. 1933. Inheritance of 
 characters in sorghum-the great millet. II. Purple pigmentation on leaf-sheath and glume. 
 Indian Journal of  Agricultural Science. 3:589-604 
Ayyangar, G. N. R., Ayyar, M. A. S. Rao, V. P. and Nambiar, A. K. (1936). Mendelian  
  segregations for juiciness and sweetness in sorghum stalks. Madras Agricultural Journal. 
 24:247-248 
145 
 
University of Ghana http://ugspace.ug.edu.gh
Badve, V. C., Nisal, P. R., Joshi, A. L. and Rangnekar, D. V. (1993). Variation in quality of 
 sorghum stover. In: Singh K; Schiere JB, eds. Feeding of ruminants on fibrous crop 
 residues. ICAR, New Delhi and Wageningen Agricultural University, Wageningen, The 
 Netherlands. 370-377. 
Ball, D. M., Collins, M., Lacefield, G. D., Martin, N. P., Mertens, D. A., Olson, K. E., Putnam, D. 
 H., Undersander, D. J., and Wolf, M. W. (2001). Understanding forage quality. American 
 Farm Bureau Federation. Publication 1-01, Park Ridge, IL. 20p. 
Bandyopadhyay, R., Mughogho, L. K., and Satyanarayana, M. V. (1987). Systemic infection of 
 sorghum by Acremonium strictum and its transmission through seed. Plant Disease. 
  71:647-650 
Barriere, Y. and Argillier, O. (1993). Brown midrib genes of maize. A review. Agronomie. 13. 
  865-876. 
Barriere, Y., Ralph, J., Mechin, V., Guillaumie, S. (2004). Genetic and molecular basis of grass 
 cell wall biosynthesis and degradability. II. Lessons from brown-midrib mutants. Comptes 
 Rendus Biologies. 327. 847-860 
Barriere, Y., Carine, G., Deborah, G. and Magalie, P. (2003). Genetic variation and  
  breeding strategies for improved cell wall digestibility in annual forage crops. A review. 
  Animal Research, EDP Sciences, 52 (3), pp.193-228. <10.1051/animres:2003018>. 
 <hal00889970> 
Batista de Aguilar, P., Ananias de Assis Pires, D., Barros Frota, B. C., Santo Rodrigues, J. A., 
 Tavares dos Reis, S. and Rocha Junior, V. R. (2014). Nutritional characteristics of bmr 
 mutant and normal sorghum genotypes used for cutting and grazing. Acta Scientiarum. 
 Animal Sciences. 36, N°3. 259-264. 
146 
 
University of Ghana http://ugspace.ug.edu.gh
Bean, B. W., Baumhardt, R. L., McCollum, F., T. and McCuistion K. C. (2013). Comparison of 
  sorghum classes for grain and forage yield and forage nutritive value. Field Crops 
 Research. 142: 20-26. 
Beck, P. A., Hutchison, S., Gunter, S. A., Losi, T. C., Stewart, C. B., Capps, P. K. and Phillips, J. 
 M. (2014). Chemical composition and in situ dry matter and fiber disappearance of 
 sorghum x Sudangrass hybrids. Journal of Animal Science. 85:545-555 
Berazneva, J. and Dyson, C. H. (2013). Economic Value of crop residues in African smallholder 
  agriculture. Cornell University. 25.  
Bhagasara, V. K., Ranwah, B. R., Meena, B. L. and Rumana, K. (2017). Estimation of GCV, 
  PCV, Heritability and Genetic Gain for Yield and its Related Components in Sorghum 
  [Sorghum bicolor (l.) Moench]. International Journal of Current Microbiology and 
 Applied Sciences.  6 (5): 1015-1024.  
Bittinger, T. S., Cantrel, R. P. and Axtell, J., D. (1981). Allelism tests of the brown midrib mutants 
 of sorghum. Journal of Heredity 72:147-148.   
Black, D. L. (2003). Mechanisms of Alternative Pre-Messenger RNA Splicing. Annual Review 
 of Biochemistry. 72:291-336 
Blümmel, M., Homann-Kee Tui, S., Valbuena, D. Duncan, AJ. and Herrero, M. (2013). Biomass 
 in crop-livestock systems in the context of the livestock revolution. Secheresse. 24, 
 (4): 330-339.  
Boerjan, W., Ralph, J. and Baucher, M. (2003). Lignin biosynthesis. Annual Review of Plant 
  Biology. Vol. 54:519-546 
Bonawitz, N., D. and Chapple, C. (2010). The genetics of lignin biosynthesis: Connecting 
 genotype to phenotype. Annual Review of Genetics. 44: 337-363 
147 
 
University of Ghana http://ugspace.ug.edu.gh
Bout, S. and Vermerris, W. (2003). A candidate gene approach to clone the sorghum Brown midrib 
 gene encoding caffeic acid O-methyl transferase. Molecular Genetics and Genomics. 
 269: 205-214 
Boyce, C. K., Zwieniecki, M. A., Cody, G. D., Jacobsen, C., Wirick, S., Knoll, A. H. and Holbrook 
 N., M. (2004). Evolution of xylem lignification and hydrogel transport regulation. 
 Proceedings of the National Academy of Sciences of the USA 101(50): 17555-17558. 
Bradley, J. T., Jennifer, C., Thomas, H. T., Peter, C., Elizabeth, A. G., Steven, H. and Luca, C. 
 (2007). Discovery of chemically induced mutations in Rice by TILLING. BMC Plant 
 Biology. 7:19.  
Brauer, M. J., Christianson, C. M., Pai, D. A, and Dunham, M. J. (2006). Mapping Novel Traits 
 by Array-Assisted Bulk Segregant Analysis in Saccharomyces cerevisiae. Genetics. 173: 
  1813-1816 
Brown, P. J., Klein, P. E., Bortiri, E., Acharya, C. B., Rooney, W. L., and Kresovich, S. (2006). 
 Inheritance of inflorescence architecture in sorghum. Theoretical and Applied Genetics. 
 113(5) 931-942 
Burton, G. W. (1952). Quantitative inheritance in grasses. In: Proceedings. 6th Inter. Grassland 
 Congress 1. 277-283. 
Buso, W.H.D., Morgado, H.S., Silva, L.B., França, A.F.S. (2011). Utilização do sorgo forrageiro 
 na alimentação animal. Publicações em Medicina Veterinária e Zootecnia, Londrina, V.5, 
 n.23, p.160-189 
Camara, Y., Bantilan, M. C. S. and Ndjeunga, J. (2006). Impacts of Sorghum and Millet research 
  in West and Central Africa (WCA): A Synthesis and Lessons Learnt p.16. 
148 
 
University of Ghana http://ugspace.ug.edu.gh
Carpita, N. C Gibeaut, D. M. (1993). Structural models of the primary cell walls in flowering 
 plants:  consistency of molecular structure with the physical properties of the wall during 
 growth. The Plant Journal.  3:1-30 
Casler, M. D., Pederson, J. F. and Undersander, D. J. (2003). Forage yield and economic losses 
  associated with the brown-midrib trait in sudangrass. Crop Science. 43:782-789. 
Cassie, M. W., Vimal, K. B., Carloalberto, P., Krishan, M. R., Seth, D. and Venugopal, M. (2015). 
 Engineering Plant Biomass Lignin Content and Composition for Biofuel and Bioproducts.
  Energies. 8, 7654-7676 
CGIAR (Consultative Group on International Agricultural Research). (2011). Dryland cereals. A 
 global alliance for improving food security, nutrition and economic growth for the world’s 
 most vulnerable poor. 237p. 
Chen, F., Dixon, R. A. (2007). Lignin modification improves fermentable sugar yields for biofuel 
 production. Nature Biotechnology. 25: 759-761 
Cherney, J. H., Cherney, J. R., Akin, D. E., Axtell, J., D. (1991). Potential of brown midrib, low 
 lignin mutants for improving forage quality. Advances in Agronomy. 46. 157-198. 
Chikuta, S., Odong, T., Kabi, F., Mwala, M. and Rubaihayo, P. (2014). Farmers perceptions on 
  dual purpose sorghum and it’s potential in Zambia. International Journal of Agricultural 
 Research Innovation and Technology. 4 (2): 76-81 
Chloe, Z., Parvathi, K., Jean-Luc, F., Richard, A. D., and Joseph, P. N. (2002). Structural Basis 
 for the Modulation of Lignin Monomer Methylation by Caffeic Acid/5 Hydroxyferulic 
 Acid 3/5- O –Methyltransferase. The Plant Cell. 14, 1265-1277 
149 
 
University of Ghana http://ugspace.ug.edu.gh
Clark, J. and Kitch, L. (1982). The potential role of improved agricultural technologies in Niger: 
 plant breeding. Characteristics of Sorghums in Niger. In Niger cereals research project. 
 Final technical report. 85p. 
Costa, R. F., de Assis Pires, D. A, Moura, M. M. A., de Sales, E. C. J., Rodrigues, J. A. S. and 
 Rigueira, J. P. S. (2016). Agronomic characteristics of sorghum genotypes and nutritional 
 values of silage. Acta Scientiarum Animal Sciences Maringá. 38, n. 2. 127-133. 
Dalimova, G. N. and Abduazimov, Kh. A. (1994). Lignins in the herbs, Khim. Prirod. Soedinenii, 
 Nº. 2. 160-177. 
Dauwe, R., Morreel, K., Goeminne, G., Gielen, B., Rohde, A., Van Beeumen, J., Ralph, J., Boudet, 
 A-M., Kopka, J., Rochange, S. F., Halpin, C., Messens, E., Boerjan, W. (2007). Molecular
  phenotyping of lignin-modified tobacco reveals associated changes in cell wall 
 metabolism, primary metabolism, stress metabolism and photorespiration. The Plant 
 Journal 52, 263-285. 
Diakité, O. S., Teme, N., Aissata, M., Mahamane, A. K., Seidou, G., Pangirayi, T. (2017). Survey
  on Crop Residues for Cattle Feeding in Niger. Journal of Agriculture and Food  
  Technology. 7(8) 7-15. 
Diester, G. and Hassane, D. (2006). Niger: Country pasture/Forage resource profiles. FAO. 24p  
Dowd, P. F., Funnell-Harris, D. L. and Sattler, S. E. (2016). Field damage of sorghum (Sorghum 
 bicolor) with reduced lignin levels by naturally occurring insect pests and pathogens. 
  Journal of Pest Science. 89:885-895 
Duncan, R. R. (1996). Breeding and improvement of forage s o r g h u m  for the 
 tropics. Agronomy. 57. p27 
150 
 
University of Ghana http://ugspace.ug.edu.gh
Elbasha, E., Thornton, P. K. and Tarawali, G. (1999). An ex post economic assessment of planted 
 forages in West Africa. ILRI Impact Assessment Series. no. 2. 61p.  
Elliot, R. C. and Tropps, J. H. (1964). Studies of protein requirements of ruminants. Nitrogen 
 balance trials of Bald head Persian sheep given diets of different energy and protein 
 contents. British Journal of Nutrition. 18:245-252. 
Food and Agriculture Organization. (2012). Balanced feeding for improving livestock 
 productivity – Increase in milk production and nutrient use efficiency and decrease in 
 methane emission, by M.R. Garg. FAO Animal Production and Health Paper No. 173. 46p. 
 Rome, Italy. 
Food and Agriculture Organization. (2014). Crop residues and agro-industrial by-products 
 in West Africa. Situation and way forward for livestock production. 64p 
Food and Agriculture Organization. (2018). Global Information and Early Warning System 
 on Food and Agriculture.  Country Brief. The Niger. 2p. 
Fehr, W. R., Fehr, E., L. and Jessen, H. J. (1987). Principles of cultivar development: theory and 
 technique. Vol. 1. Macmillan. New York. UAS. 23-27. 
Fritz, J. O., Cantrell, R. P., Lechtenberg, V. L., Axtell, J. D. and Hertel, J. M. (1981). Brown Midrib
  Mutants in Sudangrass and Grain Sorghum. Crop science. 21. 707-709 
Fritz, J. O., Moore, K. J. and Jaster, E. H. (1988). In situ digestion kinetics and ruminal turnover 
 rates of normal and brown midrib mutant sorghum x sudangrass hays fed to non-lactating 
 Holstein cows. Journal of Dairy Science. 71 (12):3345-3351. 
Fritz, J. O., Moore, K. J. and Jaster, E. H. (1988). Digestion kinetics and cell wall composition of 
 brown midrib sorghum x sudangrass morphological components. Crop Science. 30:213-
 219. 
151 
 
University of Ghana http://ugspace.ug.edu.gh
Funnell, D. L. and Pedersen, J. F. (2006). Reaction of Sorghum Lines Genetically Modified for 
 Reduced Lignin Content to Infection by Fusarium and Alternaria spp. Plant Disease. 90 
 (3):331-338 
Gerhardt, R. L., Fritz, J. O., Moore, K. J. and Jaster, E. H. (1994). Digestion kinetics and 
 composition  of normal and brown midrib sorghum morphological components. Crop 
  Science. 34:1353-1361. 
Getachew, G. Putnam, D. H., De Ben, C. M. and De Peters, E. J. (2016). Potential of Sorghum as 
  an Alternative to Corn Forage. American Journal of Plant Sciences. 7. 1106-1121. 
Godbharle, A. R., More, A. W. and Ambekar, S. S. (2010). Genetic Variability and Correlation 
 Studies in elite ‘B’ and ‘R’ lines in Kharif Sorghum. Electronic Journal of Plant Breeding, 
 1(4): 989-993 
Govindarasu, R., Rathenam, M., Sivasubramanian, P. (1990). Genetic variability in sesame 
 (Sesamum  indicum L) Madras Agricultural Journal. 77: 450-452. 
Grant, R. J., Haddad, S. G., Moore, K. J. and Pedersen, J. F. (1995). Brown mid-rib sorghum silage
  for mid-lactation dairy cows. Journal of Dairy Science 78(9):1970-80. 
Greene, E. A., Codomo, C. A., Taylor, N. E., Henikoff, J. G., Till, B. J., Reynolds, S. H., Enns, L.
  C., Burtner, C., Johnson, J. E., Odden, A. R., Comai, L. and Henikoff, S. (2003). Spectrum
  of chemically induced mutations from a large-scale reverse genetic screen in 
 Arabidopsis. Genetics. 164. N°2. 731-740. 
Grisebach, H. (1981). Lignins. In "The Biochemistry of Plants" (P. K. Stumpf and E. E. Conn, 
 Eds.), Vol. 7. "A Comprehensive Treatise," pp. 457-478. Academic Press, New York. 
Grunert, K. G. (2005). Food quality and safety: consumer perception and demand. European 
  Review of Agricultural Economics 32 (3):369-391. 
152 
 
University of Ghana http://ugspace.ug.edu.gh
Guilbert, B., Bnazhaf, M., Soule, B. G., Balami, D. H., Ide, G. (2009). Etude regionale sur les 
 contexts de la commercialization du betail/acces aux marches et defies d’amelioration des 
 conditions de vie des communautes pastorals. 119p 
Guo, D., Chen, F., Inoue, K., Blount, J., and Dixon, R., A. (2001). Downregulation of caffeic acid
  3-O-methyltransferase and caffeoyl CoA 3-O-methyltransferase in transgenic alfalfa: 
  Impacts on lignin structure and implications for the biosynthesis of G and S lignin. Plant 
 Cell 13, 73-88. 
Gupta, S. C. (1995). Inheritance and Allelic Study of Brown Midrib Trait in Pearl Millet. The 
  Journal of Heredity. 86(4). 301-303 
Halpin, C., Holt, K., Chojecki, J., Oliver, D., Chabbert, B., Monties, B., Edwards, K., Barakate, 
  A., Foxon, G. A. (1998). Brown- midrib maize (/bm1/): a mutation affecting the cinnamyl 
 alcohol dehydrogenase gene. The Plant Journal. 14 (5) 545-553. 
Hamed, A. H. M., Abbas, S. O., Ali, K. A. and Elimam, M., E. (2015). Stover Yield and Chemical 
 Composition In Some Sorghum Varieties In Gadarif State, Sudan. Animal Review. 2(3): 
 68-75 
Hamidou, M., Souley, A. M., Soulemane, O., Danquah, E. Y., Ba, M., Ofori, K., Gracen, V. and 
 Kapran, I. (2018). Farmers’ awareness of sorghum production constraint and their 
 preferences for sorghum traits in Niger. Agricultural Science Research Journal. (8) 2.  
 31-19. 
Hanna, W. W. (1982). Mutation breeding of pearl millet and sorghum. International atomic energy 
 agency. INIS-XA-387. 15p 
Hannes, V., Dimitrios, F., Bradley, J. N. and Ueli, G. (2015). Measuring the Mechanical Properties 
 of Plant Cell Walls. Plants (4) 167-182 
153 
 
University of Ghana http://ugspace.ug.edu.gh
Harlan, H. V. and Pope, M. N. (1922). The use and value of back-crosses in small-grain breeding. 
 Journal of Heredity. 13. 319-322. 
Harlan, J. R. and de Wet, J. M. J. (1972). A simplified Classification of sorghum. Crop Science 
 12:172-176. 
Harper, K. J. and McNeill, D. M. (2015). The Role iNDF in the Regulation of Feed Intake and the 
 Importance of Its Assessment in Subtropical Ruminant Systems (the Role of iNDF in the 
 Regulation of Forage Intake). Agriculture. 5. 778-790 
Hisano, H., Nandakumar, R. and Wang, Z-Y. (2009). Genetic modification of lignin biosynthesis 
  for improved biofuel production. In Vitro Cellular and Developmental Biology-Plant 
 45(3):306-313 
Hospital, F. (2005). Selection in backcross programmes. Philosophical Transaction of the Royal 
 Society B: Biological Sciences. 360 (1459). 1503-1511. 
House, L. R. (1985). A Guide to Sorghum Breeding. International Crops Research Institute for the 
Semi-Arid Tropics. Patancheru, India. 212p. 
Humphreys, J. M. and Chapple, C. (2002). Rewriting the lignin road map. Current. Opinion in 
 Plant Biology. 5: 224-229. 
IBPGR, and ICRISAT. (1993). Descriptors for sorghum [Sorghum bicolor (L.) Moench].  
  International Board for Plant Genetic Resources, Rome, Italy; International Crops 
  Research Institute for the Semi-Arid Tropics. Patancheru, India. 45p   
INTSORMIL, University of Nebraska-Lincoln. 2013. Bmr forage sorghum project. Work plan and 
 progress report 2010-2012. 16p. 
154 
 
University of Ghana http://ugspace.ug.edu.gh
Iqbal, M. A., Bilal, A., Shah, M. H., and Ali, K. 2(015). A Study on Forage Sorghum (Sorghum 
  bicolor L.) Production in Perspectives of White Revolution in Punjab, Pakistan: Issues and 
 Future Options. American-Eurasian Journal of Agricultural and Environmental Sciences. 
 15 (4): 640-647 
Jacob, J. P., Mitaru, B. N., Mbugua, P. N. and Blair, R. (1997). The nutritive value of Kenyan 
 sorghum for poultry. Tropical Science. 37: 43-48. 
JAICAF (Japan Association for International Collaboration of Agriculture and Forestry). (2009). 
 Minor cereal in Niger: Focusing on pearl millet and sorghum. 117 pages. Institut national 
  de la statistique, 2013, Annuaire statistique, Edition 2013, 238 pages. 
Johnson, H. W., Robinson, H. F. and Comstock, R. E. (1955). Estimates of genetic and 
 environmental variability in soybean. Agronomy Journal. 47(7). 314-318. 
Jones, L., Ennos, A. R.  and Turner, S. R.  (2001). Cloning and characterization of irregular xylem4 
 (irx4) (irx4): A severely lignin-deficient mutant of Arabidopsis. The Plant Journal 26: 205-
 216. 
Jung, H. G., Allen, M. S. (1995). Characteristics of plant cell walls affecting intake and 
 digestibility of forages by ruminants. Journal of Animal Science. 73.  2774-2790 
Jung, H. G. and Fahey, G. C. Jr. (1983). Nutritional implications of phenolic monomers and 
  lignin: A review. Journal of Animal Science. 57. 206-219. 
Jung, H. J. G. and Ni, W. (1998). Lignification of plant cell walls: impact of genetic manipulation.
  Proceedings of the National Academy of Sciences of the USA. 95(22):12742-12743 
Kadi Kadi, A. H. (1993). Rapport de campagne (1993) Entomologie du sorgho. INRAN CERRA, 
 Maradi, Niger. 
155 
 
University of Ghana http://ugspace.ug.edu.gh
Kamuanga, M. J. B., Somda, J., Sanon, Y. and Kagone, H. (2008). Livestock and regional market
  in the Sahel and West Africa: Potentials and challenges. 170p 
Kang, M. S. (2002). Genotype-environment interaction: Progress and prospects. In: MS Kang (Ed)
  Quantitative Genetics, Genomics and Plant Breeding. CAB International, New York, 
 USA. 400p. 
Kapran, I., Grenier, C. and Ejecta, G. (2007). Introgression of genes for striga resistance into 
  African landraces of sorghum. In Integrating New Technologies for Striga Control: 
 towards ending the witch-hunt. ISBN-10: 981-270-708-5. 345p. 
Kelley, T. G., Parthasarathy, R. P., Weltzien, R., Purohit, M. L. (1996). Adoption of improved 
 cultivars of pearl millet in arid environment: straw yield and quality considerations in 
 western Rajasthan. Experimental Agriculture. 32: 161-72. 
Kelley, T. G. and Parthasarathy, R. P. (1994). Yield and quality characteristics of improved and 
 traditional sorghum cultivars: farmers‟ perceptions and preferences. In: Joshi, A.L., Doyle, 
 P.T. and Oosting, S.J. (eds), Variation in the Quantity and 31 Quality of Fibrous Crop 
 Residues. Proceedings of a National Seminar held at the BAIF Development Research 
 Foundation, Pune, pp. 133-145. 
Khan, Z. I., Ashraf, M., Hussain, A. and McDowell, L. R. (2005). Seasonal Variation of Trace 
  Elements in a Semiarid Veld Pasture. Communications in Soil Science and Plant Analysis, 
 37:9-10, 1471-1483. 
Kharkwal, M. C. (2012). A brief history of plant mutagenesis. In: Shu Q., Y, Forster B., P, 
 Nakagawa H. Plant mutation breeding and biotechnology. ISBN 9781780640853. 616p. 
156 
 
University of Ghana http://ugspace.ug.edu.gh
Kimbeng, C. A., Zhou, M. M. and da Silva, J. A. (2009). Genotype x environment interactions and 
 resource allocation in sugarcane yield trials in the Rio Grande valley region of Texas. 
  Journal of the American Society of Sugar Cane Technologists. 29: 11-24. 
Konovalov, A. A., Shundrina, I. K. and Karpova, E. V. (2015). Polymorphism of Lignification 
 Enzymes in Plants: Functional Importance and Applied Aspects. Uspekhi Sovremennoi 
 Biologii,. 135. (5). 496-513. 
Konzak, C. F., Nilan, R. A., Wagner, J. and Foster, R. J. (1965). Efficient chemical mutagenesis. 
 The use of induced Mutations in Plant Breeding. 49-70. In Manual on Mutation Breeding. 
  2nd edition. FAO/AIEA. ISBN 92-0-115077-6. Vienna 1977. 298p.  
 Kotasthane, T. V., Umakanth, A. V., Madhu, P., Bhat, V. and Patil, J. V. (2015). Evaluation of 
 Brown-Midrib Genotypes for Fresh and Dry Biomass Suitable for Bio fuel Production. 
 International Journal of Mechanical and Industrial Technology. (2) Issue 2, pp: 104-
 108 
Kubkomawa, H., Olawuye, H. U., Krumah, L. J., Etuk, E. B. and Okoli, I. C. (2015). Nutrient 
 requirements and feed resource availability for pastoral cattle in the tropical Africa: A 
 review. Journal of Agricultural and Crop Research. 3(7). 100-116  
Kumari, P., Pahuja, S. K., Panchta, R., Arya, S., Satpal, Tokas, J. and Aruna, C. (2017). Evaluation 
 of forage sorghum brown midrib lines for quality biomass production. Global Journal of 
 Bio-Science and Biotechnology. 6 (2) 234-239 
Kwangsik, Nho. (2014). Practical Guideline for Whole Genome Sequencing. Indiana University 
 School of Medicine Indiana University Department of Radiology and Imaging Sciences 
 Center for Computational Biology and Bioinformatics. 48p. 
157 
 
University of Ghana http://ugspace.ug.edu.gh
Kyle, R. P., Kristina M. S. and Michael F. (2011). Bulk Segregant Analysis Followed by High-
 Throughput Sequencing Reveals the Neurospora Cell Cycle Gene, ndc-1, To Be Allelic 
  with the Gene for Ornithine Decarboxylase, spe-1. Eukaryotic Cell. 10 (6). 724-733 
Lee, S., Costanzo, S.  and Jia, Y. (2011). The Structure and Regulation of Genes and Consequences
  of Genetic Mutations. In Plant Mutation Breeding and Biotechnology. ISBN 978-92-5-
 105000-0. 595p. 
Li, H. and Durbin, R. (2009). Fast and accurate short read alignment with Burrows-Wheeler 
 Transform. Bioinformatics, 25:1754-60. [PMID: 19451168]  
Li, L., Popko, J. L., Umezawa, T., Chiang, V. L. (2000). 5-Hydroxyconiferyl aldehyde modulates 
 enzymatic methylation for syringyl monolignol formation, a new view of monolignol 
 biosynthesis in angiosperms. Journal of Biological Chemistry. 275. 6537-6545 
Li, Y., Liu, G., Li, J., You, Y., Zhao, H., Liang, H. and Mao, P. (2015). Acid detergent lignin, 
 lodging resistance index, and expression of the caffeic acid O-methyltransferase gene in 
  brown midrib-12 sudangrass. Breeding Science. 65:291-297. 
Lima, C. B., Carneiro, J. C., Novaes, L. P., Lopes, F. C. F. and Rodrigues, J. A. S. (2005). 
 Potencial forrageiro e avaliação bromatológica de híbridos de sorgo com capim-sudão. Juiz 
 de Fora: Embrapa Gado de Leite.  (Boletim de Pesquisa e Desenvolvimento, 18). 
Lönnig, W. E. (2005). Mutation breeding, evolution, and the law of recurrent variation. Recent 
 Res. Dev. Genetics Breeding. 2: 45-70 
Maarouf, I. M. and Zeinab, A. Z. (2014). Quality Attributes in Sudanese Sorghums Improved for 
 Forage Yield. American Journal of Experimental Agriculture. 4 (2): 171-182 
Mackay, J. J., O’Malley, D. M. Presnell, T. Booker F. L., Campbell, M. M., Whetten, R. W., 
 Sederoff, R. R. (1997). Inheritance, gene expression, and lignin characterization in a 
158 
 
University of Ghana http://ugspace.ug.edu.gh
 mutant pine deficient in cinnamyl alcohol dehydrogenase. Proceedings of the National 
 Academy of Sciences of the USA. 94: 8255-8260 
Magnan, N., Larson, D. M. and Taylor, J. E. (2012). Stuck on stubble? the non-market value of 
 agricultural byproducts for diversified farmers in morocco. American Journal of 
 Agriculture Economics. 94(5): 1055-1069 
McCollum, III F., McCuistion, T. K. and Bean, B. (2005). Brown mid-rib and photoperiod-
 sensitive forage sorghums. Proc. Plains Nutrition Council. 5-20p. 
Mekbib, F. (2006). Farmer and formal breeding of sorghum (Sorghum bicolor (L.) Moench) and 
  the implications for integrated plant breeding. Euphytica 152:163-176. 
Miller, F. R. and Stroup, J. A. (2004). Growth and management of Sorghums for forage production. 
 In: Proceedings, National Alfalfa symposium 3-5 December, 2004, San Diego, CA, UC 
 Cooperative Extension University of California, Davis 95616. 
Mir Derikvand, M., Sierra, J. B., Ruel, K., Pollet, B., Do, C-T., Thevenin, J., Buffard, D., Jouanin, 
 L., Lapierre, C. (2008). Redirection of the phenylpropanoid pathway to feruloyl malate in 
 Arabidopsis mutants deficient for cinnamoyl-CoA reductase 1. Planta, 227: 943-956. 
Moehn, S., Atakora, J. and Ball, R. O. (2005). Using Net Energy for Diet Formulation: Potential 
 for the Canadian Pig Industry. Advances in Pork Production. 16. 119p. 
Mohamed-Saleem, M. A. and Von Kaufmann, R. R. (1988). A rapid survey of feeding regimes for 
 draught cattle in Niger State, Nigeria. International Livestock Center for Africa. 8p 
Mohammad, A. J. (2008). Feed and fodder markets in South Asia and East Africa: A synthesis of 
 four PRA case studies. International Livestock Research Institute. 34p 
Muehlhoff, E., Bennett, A. and McMahon, D. (2013). Milk and dairy products in human nutrition. 
 In Milk and dairy products in human nutrition. FAO Publications. 404p 
159 
 
University of Ghana http://ugspace.ug.edu.gh
Nagaraja, Reddy R., Murali Mohan, S., Madhusudhana, R., Umakanth, A.V., Satish, K. and 
  Srinivas, G. (2008). Inheritance of morphological characters in sorghum. Journal of SAT 
  Agricultural Research. (6). 3p. 
Ngaido, T. (1996). “Legislation and institutions in the Sahel: the case of Niger.” (In: Agriculture 
 + Rural Development, 3:2. 18-21. 
Oba, M., Allen, M. S. (1999). Effects of brown midrib-3 mutation in corn silage on organic matter 
 intake and productivity of high yielding dairy cows. Journal of Dairy Science. 82, 135-
 142. 
Olembo, K. N., M’mboyi, F., Kiplagat, S., Sitieney J., K. and Oyugi, F. K. (2010). Sorghum 
 breeding in Sub-Saharan Africa: The success stories. The African Biotechnology  
  Stakeholders Forum (ABSF). 40p 
Oliver, A. L., Pedersen, J. F., Grant, R. J., Klopfenstein, T. J. (2005). Comparative effects of the 
  sorghum bmr-6 and bmr-12 genes: I Forage sorghum yield and quality. Crop Science. 
 45:2234-2239. 
Oliver, A. L., Grant, R. J., Pedersen, J. F., O'Rear, J. (2004). Comparison of brown midrib-6 and -
 18 forage sorghum with conventional sorghum and corn silage in diets of lactating dairy 
 cows. Journal of Dairy Science. 87:637-644. 
Onyeonagu, C. C., Obute, P. N., Eze, S. M. (2013). Seasonal variation in the anti-nutrient and 
  mineral components of some forage legumes and grasses. African Journal of  
  Biotechnology. 12(2):142-149. 
Palmer, N. A., Sattler, S. E., Saathoff, A. J., Funnell D., Pedersen, J. F., Sarath, G. (2008). Genetic 
 background impacts soluble and cell wall-bound aromatics in brown midrib mutants of 
 sorghum. Planta 229: 115-127 
160 
 
University of Ghana http://ugspace.ug.edu.gh
Pedersen, J. F., Gorz, H. J., Haskins, F. A. and Ross, W. M. (1982). Variability for quality and 
 agronomic traits in forage sorghums hybrids. Crop Science. 22, 853-856  
Pedersen, J. F., Vogel, K. P., and Funnell D., L. (2005). Impact of Reduced Lignin on Plant Fitness. 
 USDA-ARS /UNL Faculty. 163. 9p. 
Pedersen, J. F., Toy, J. J., Funnell D., L., Sattler, S. E., Oliver, A. L. and Grant, R., A. (2008). 
 Registration of BN611, AN612, BN612, and RN613 Sorghum Genetic Stocks with Stacked 
 bmr-6 and bmr-12 Genes. Journal of Plant Registrations 2:258-262 
Per Sikora, Aakash, C., Mikael, L., Johanna, O. and Olof, O. (2011). Mutagenesis as a Tool in 
 Plant Genetics, Functional Genomics, and Breeding. International Journal of Plant 
 Genomics. 13p. 
Pica-Ciamarra, U., Tasciotti, L., Otte, J. and Zezza, A. (2011). Livestock Assets, Livestock  
Income and Rural Households Cross-Country Evidence from Household Surveys. 18p. 
Pillonel, C., Mulder, M. M., Boon, J. J., Forster, B., Binder, A. (1991). Involvement of cinnamyl-
 alchol dehydrogenase in the control of lignin formation in Sorghum bicolor (L.) Moench. 
 Planta, 185: 538-544 
Piquemal, J., Lapierre, C., Myton, K., O’Connell, A., Schuch, W., Grima-Pettenati J. and Boudet, 
 A-M. (1998). Down-regulation of Cinnamoyl-CoA Reductase induces significant changes
  of lignin profiles in transgenic tobacco plants. The Plant Journal. 13(1). 71-83 
Pond, K. R., Ellis, W. C., Lascano, C. E., and Akin, D. E. (1987). Fragmentation and flow of 
 grazed  coastal Bermudagrass through the digestive tract of cattle. Journal of Animal 
  Science. 65, 609-618. 
161 
 
University of Ghana http://ugspace.ug.edu.gh
Porter, K. S., Axtell, J. D., Lechtenberg, V. L. and Colenbrander, V. F. (1978). Phenotype, fiber 
 composition, and in vitro dry matter disappearance of chemically induced brown midrib 
 (bmr) mutants of sorghum. Crop Science. 18: 205-208 
Prasad, S. R., Prakash, R., Sharma, C. M. and Haque, M., F. (1981). Genotypic and phenotypic 
  variability in quantitative characters in oat. Indian Journal of Agricultural Science. 54: 
 480-482. 
Rabinovich, M. L., Bolobova, A. V. and Kondrashchenko, V., I. (2001). Teoreticheskie osnovy 
 biotekhnologii drevesnykh kompozitov. Kniga 1. Drevesina i razrushayushchie ee griby 
  (Theoretical Principles of Biotechnology of Wood Components, Book 1: Wood and 
 Wood_Destroying Fungi), Moscow: Nauka. 264p. 
Ralph, J., Hatfield, R. D., Quideau, S., Helm, R. F., Grabber, J. H., Jung, H. J. G. (1994). Pathway 
 of p-coumaric acid incorporation into maize lignin as revealed by NMR. Journal of the 
  American Chemical Society. 116: 9448-9456 
Rao, P. S., Deshpande, S., Blummel, M., Reddy, V. S. and Hash, T. (2012). Characterization of 
 brown  midrib mutants of sorghum (sorghum bicolor (L.) Moench). The European Journal 
 of Plant Science and Biotechnology. 6 (Special Issue 1), 71-75.  
Rao, P. P and Hall, A. J. (2003). Importance of crop residues in crop-livestock systems in India 
 and farmers’ perceptions of fodder quality in coarse cereals. Field Crops Research. 84: 
 189-198 
Reddy, B. V. S., Ramesh, S. and Reddy, P. S. (2004). Sorghum breeding research at ICRISAT -
 goals, strategies, methods and accomplishments. International Sorghum and Millets 
 Newsletter. 45: 5-12. 
162 
 
University of Ghana http://ugspace.ug.edu.gh
Reddy, B.V.S., Ashok Kumar, A., Sharma, H. C., Srinivasa, R. P., Blümmel, M., Ravinder Reddy, 
 Ch., Sharma, R., Deshpande, S. P., Datta, M. S., and Dinakaran E. (2012). Sorghum 
 improvement (1980-2010): Status and way forward. Journal of SAT Agricultural Research. 
 10. 1-14. 
Reed, J. D., Kebede, Y. and Fussell, L. K. (1988). Factors affecting the nutritive value of sorghum 
 and millet crop residues. III   "Plant Breeding and The Nutritive Value of Crop Residues, 
 Proc. Workshop" (1.  D.  Reed, B.  S.  Capper, and P. 1. H. Neate, Eds.) pp.  233-251.   
République du Niger. (2002). Poverty Reduction Strategy. 187p   
République du Niger. (2012). Catalogue National des Espèces et Variétés Végétales. 276 pages. 
Rooney, W. L.  (2000). Genetics and Cytogenetics. In Sorghum: Origin, History, Technology, and 
 Production. John Wiley & Sons, Inc. ISBN 0-471-24237-3. 824p.  
Rooney, W. L. (2001). Sorghum genetics and cytogenetics. In “Sorghum: Evolution, History, 
  Production and Technology” (C. W. Smith and R. A. Frederiksen, Eds.). 261-307 pp. 
Rooney, W. L. (2004). Sorghum Improvement-Integrating Traditional And New Technology To 
  Produce Improved Genotypes. Advances in Agronomy, Volume 83. 73p. 
Ros Barcelo, A. and Pomar, F. (2001). Oxidation of cinnamyl alcohols and aldehydes by a basic 
  peroxidase from lignifying Zinnia elegans hypocotyls. Phytochemistry. 57. (7). 1105-
 1113. 
Roy, J. K., Lakshmikumaran, M. S., Balyan, H. S. and Gupta, P. K. (2004). AFLP-based genetic 
 diversity and its comparison with diversity based on SSR, SAMPL, and phenotypic traits 
 in bread wheat. Biochemical Genetics. 42(1/2):43-59. 
163 
 
University of Ghana http://ugspace.ug.edu.gh
Ruel, K. Berrio-Sierra, J. Mir Derikvand, M. Pollet, B., Thévenin, J., Lapierre, C., Jouanin, L. and 
 Joseleau, J.-P. (2009). Impact of CCR1 silencing on the assembly of lignified secondary 
 walls in Arabidopsis thaliana. New Phytologist. 184(1):99-113 
Saballos, A. Ejeta, G. Sanchez, E. Kang, C. and Wilfred Vermerris. (2009). A Genomewide 
 Analysis of the Cinnamyl Alcohol Dehydrogenase Family in Sorghum [Sorghum bicolor 
  (L.) Moench] Identifies SbCAD2 as the Brown midrib6 Gene. The Genetics Society of 
  America. 181: 783-795 
Saballos, A. (2008). Development and Utilization of Sorghum as a Bioenergy Crop IN Genetic 
  Improvement of Bioenergy Crops. (ed.), ISBN: 978-0-387-70804-1. 467p. 
Saballos, A., Sattler, S. E., Sanchez, E., Foster, T. P., Xin, Z. (2012). Brown midrib2 (Bmr2) 
  encodes the major 4-coumarate: Coenzyme A ligase involved in lignin biosynthesis in 
 sorghum [Sorghum bicolor (L.) Moench]. The Plant Journal. 70: 818-830. 
Saballos, A., Vermerris, W., Rivera, L. and Ejeta, G. (2008). Allelic association, chemical 
 characterization and saccharification properties of brown midrib mutants of sorghum 
 (Sorghum bicolor (L.) Moench). BioEnergy Research. 1:193-204 
Sanchez, A. C., Subudhi, P. K., Rosenow, D. T., Nguyen, H. T.  (2002). Mapping QTLs associated 
 with drought resistance in sorghum (Sorghum bicolor (L.) Moench). Plant Molecular 
 Biology. 48. 713-726. 
Sandford, S. (1989). Crop residues/livestock relationships. In: Renard, C., Vandenbeldt, R.C. and 
 Parr, J.F. (eds), Soil, Crop and Water Management Systems for Rainfed Agriculture in the 
 Sudano-Sahelian Zone. Proceedings of an International Workshop, 11-16 January 1987, 
 ICRISAT. 169-182 
164 
 
University of Ghana http://ugspace.ug.edu.gh
Santiago, R., Barros-Rios, J. and Malvar, R. A. (2013). Impact of Cell Wall Composition on Maize 
 Resistance to Pests and Diseases. International Journal of Molecular Science. 14.  6960-
 6980 
Sarada, N. Mani. (1989). EMS-Induced mutagenesis in Sorghum bicolor (L.) Moench.  
  Proceedings  of the Indian National Science Academy. 477-482. 
Sattler, S. E., Funnell-Harris, D. L. and Pedersen, J. F. (2010). Efficacy of singular and stacked 
  brown midrib 6 and 12 in the modification of lignocellulose and grain chemistry. 
 Journal of Agriculture and Food Chemistry. 58 (6). 3611-3616 
Sattler, S. E., Funnell-Harris, D. L., Pedersen, J. F. (2010). Brown midrib mutations and their 
  importance to the utilization of maize sorghum and pearl millet lignocellulosic tissues. 
  Plant Science 178, 229-238 
Sattler, S. E., Palmer, N., Saballos, A., Greene, A. M., Xin, Z., Gautam, S., Vermerris, W., 
 Pedersen, J. F. (2012). Identification and characterization of four missense mutations in 
 brown midrib  12 (bmr12), the caffeic O-methyltransferas (COMT) of sorghum. 
 BioEnergy Research. 5: 855-865 
Sattler, S. E., Saathoff, A. J., Haas, E. J., Palmer, N. A., Funnell-Harris, D. L., Sarath, G.  and 
 Pedersen, J. F. (2009). A Nonsense Mutation in a Cinnamyl Alcohol Dehydrogenase Gene 
 Is Responsible for the Sorghum brown midrib6 Phenotype. Plant Physiology.150. 584-595  
Sattler, S. E., Saballos, A., Zhanguo, X., Deanna, L., Funnell-Harris, Vermerris, W. and Pedersen, 
 J. F. (2014). Characterization of novel sorghum brown midrib mutants from an EMS-
 mutagenized population. The Genetics society of America. 34p 
Sattler, S. E., Palmer, N. A., Saballos, A., Greene, A. M., Zhanguo, X., Gautam, S., Vermerris, W. 
 and Pedersen, J. F. (2012). "Identification and Characterization of Four Missense 
165 
 
University of Ghana http://ugspace.ug.edu.gh
 Mutations in Brown midrib12 (Bmr12), the Caffeic O Methyltranferase (COMT) of 
 Sorghum". Agronomy & Horticulture. Faculty Publications. Paper 690. 17p 
Sawadogo, N., Nanema, R. K., Kando, P. B., Traore, R. E., Nebie, B., Tiama, D., Sawadogo, M. 
 and Zongo, J-D. (2014). Évaluation de la diversité génétique des sorghos à grains sucrés 
  (Sorghum bicolor (L.) Moench) du Nord du Burkina Faso. Journal of Applied Biosciences
  84:7654-7664. 
Schertz, K. F. and Dalton, L. G. (1980). Sorghum. In “Hybridization of Crop Plants” Fehr, W. R. 
 and Hadley, H. H. American Society of Agronomy. 577-588 
Schroeder, J. W. (2004). Forage Nutrition for Ruminants. North Dakota State University  
  Agriculture and University Extension. 22p. 
Shapiro, B., Sanders, J. H., Reddy, K. C., Baker, T. G. (1993). Evaluating and adapting new 
 technologies in a high risk agricultural system-Niger. Agricultural Systems 42, 153-171. 
Sidibe, O. (1978). Principal Sorghum Diseases in Niger. In Proceedings of the International 
 Workshop on Sorghum Diseases. ICRISAT. 478p. 
Singh B., B., Larbi A., Tabo R and Dixon A., G., O. (2004). Trends in development of crop 
  varieties for improved crop-livestock systems in West Africa. 18p 
Singh, S., Venktesh Bhat, B., Shukla, G. P., Singh, K. K. and Gehrana, D. (2018). Variation in 
 carbohydrate and protein fractions, energy, digestibility and mineral concentrations in 
 stover of sorghum cultivars. Tropical Grasslands-Forrajes Tropicales. 6(1):42-52 
Sivasubramanian, S. and Menon, M. (1973). Heterosis and inbreeding depression in rice. Madras 
 Agricultural Journal. 60: 11-39 
Spore, February-March (2015). Feeding Africa’s Livestock. Fodder and forage solutions   
166 
 
University of Ghana http://ugspace.ug.edu.gh
Stam, P. and Zeven, A. C. (1981). The theoretical proportion of the donor genome in near-isogenic
  lines of self-fertilizers bred by backcrossing. Euphytica 30, 227-238.  
Suad, A. H., Maarouf, I. M., Samia, O., Y. (2015). Breeding for dual purpose attributes in sorghum: 
 Effects of harvest option and genotype on fodder and grain yields. Journal of plant 
 breeding and crop science. 7(4) 101-106. 
Suharti, S., Astuti, D. A., Wina, E.  and Toharmat, T. (2011). Rumen microbial population in the 
  in vitro fermentation of different ratios of forage and concentrate in the presence of whole 
 lerak (Sapindus rarak) fruit extract. Asian-Australasian Journal of Animal Sciences.  
 24(8):1086-1091. 
Sultan, S., Kushwaha, B. P., Nagl, S. K., Mishra1, A. K., Bhattacharya, S., Gupta, P. K. and Singh, 
 A. (2011). In Vitro Methane Emission from Indian Dry Roughages in Relation to Chemical 
 Composition. Current Science. 101, 57-65. 
Teferedegne, B. (2000). New perspectives on the use of tropical plants to improve ruminant 
  nutrition. The Proceedings of the Nutrition Society, 59 (2): 209-214 
Terraciano, A. M. (1998). “Contesting land, contesting laws: tenure reform and ethnic conflict in 
 Niger.” (In: Columbia Human Rights Law Review, 29:3 723-754p. 
Tesfaye, A. (1998). Economics of milk production in and around Hyderabad of Andhra Pradesh. 
 M.Sc. thesis, Acharya NG Ranga Agricultural University, Hyderabad 500 030. 
Thermo Scientific. (2008). NanoDrop 1000 Spectrophotometer. V3.7 User’s Manual. 105p  
Thornton, P. K., Kristjanson, P. M. and Thorne, P. J. (2003). Measuring the potential impacts of 
 improved food-feed crops: methods for ex ante assessment. Field Crops Research. 84: 199-
 212. 
167 
 
University of Ghana http://ugspace.ug.edu.gh
Tunde, A. A., and Augustine, A. (2014). Assessment of existing and potential feed resources to 
 improve livestock productivity in dryland areas of Niger. ILRI project report. 29p 
Tunde, A. A., Augustine, A. and Dangoma A. (2015). Least-cost rations for sheep fattening: A 
 manual for livestock farmers and extension workers in the West African Sahel. Version I. 
 15p 
UN Secretary General for Food Security and Nutrition and UN System Inflenza Coordinator. 
 (2014). African Livestock Futures. Policy Brief. 12p 
UNDP. (2006). National Adaptation Program of Action. 83p  
Valbuena, D., Gerard, B., Duncan, A., Teufel, N., Homann, S. and Bhatia, M. (2011). Trade-offs 
  of crop residue use in smallholder mixed farming systems in Sub-Saharan Africa and 
 South Asia. CGIAR Systemwide Livestock Project (SLP). 5p. 
Valbuena, D., Erenstein, O., Tui, S., H-K., Abdoulaye, T., Claessens, L., Duncan, A. J., Gerard, 
 B., Rufino, M. C., Teufel, N., Van Rooyen. A and Van Wijk Mark, T. (2012). Conservation 
 Agriculture in mixed crop-livestock systems: Scoping crop residue trade-offs in Sub-
 Saharan Africa and South Asia. Field Crops Research 132. 175-184 
Valbuena, D., Tui, S., H-K., Erenstein, O., Teufel, N., Duncan, A., Abdoulaye, T., Swain, B., 
 Mekonnen, K., Germaine, I. and Gerard, B. (2014). Identifying determinants, pressures and 
 trade-offs of crop residue use in mixed smallholder farms in Sub-Saharan Africa and South 
 Asia. Agricultural Systems 134. 107-118. 
Vanholme, R., Morreel, K., Ralph, J. and Boerjan, W. (2008). Lignin engineering. Current 
  Opinion in Plant Biology. 11(3):278-285 
Vermerris, W. (2008). Genetic Improvement of Bioenergy Crops. ISBN: 978-0-387-70804-1. 
 467p  
168 
 
University of Ghana http://ugspace.ug.edu.gh
Vogler, R., Ejeta, G., Johnson, K., Axtell, J. (1994). Characterization of a new brown midrib 
  sorghum line. Agronomy Abstracts. 124. 
Vogler, R. K., Tesso, T. T., Johnson, K. D. and Ejeta, G. (2009). The effect of allelic variation on 
 forage quality of brown midrib sorghum mutants with reduced caffeic acid O methyl 
 transferase activity. African Journal of Biochemistry Research. 3(3), 070-076. 
Waha, K., Zipf, B., Kurukulasuriya, P. and Hassan, R. M. (2016). An agricultural survey for more 
 than 9,500 African households. Scientific Data. 8p 
Wilson, J. and Kennedy, P. 1996. Plant and animal constraints to voluntary feed intake associated 
 with fiber characteristics and particle breakdown and passage in ruminants. Australian 
 Journal of Agricultural Research. 47: 199-225 
 
Witcombe, J. R. (1996). Participatory Approaches to Plant Breeding and Selection. Biotechnology 
 and Development Monitor. 29. 26. 
World Bank. (2013). Agricultural sector risk assessment in Niger. Moving from crisis 
 response to long-term risk management. Report number 74322-NE. 90p.  
Wricke, G. and Weber, W. E. (1986). Quantitative genetics and selection in plant breeding. Walter 
 de Gruyter & Co. Berlin, Germany. 
www.lgcgroup.com/genomics. A comparison of DNA quantification values obtained by UV 
 spectrophotometry (including NanoDrop) and PicoGreen analysis. Technical note. 3p 
Xin, Z., Wang, M. L., Burow, G., Burke, J. (2009). An induced sorghum mutant population 
  suitable for bioenergy research. BioEnergy Research. 2, 10-16 
Yameogo, N. D., Nabassaga, T., Ncube, M. (2014). Diversification and sophistication of livestock 
 products: The case of African countries. African Development Bank Group: working paper 
 series N°212. 29p. 
169 
 
University of Ghana http://ugspace.ug.edu.gh
Yazon, G. and Peter, C. B. (2003). Niger country brief: Property rights and land markets. 27p  
Ye, Z.-H., Kneusel, R. E., Matern, U. and Varner, J. E. (1994). An alternative methylation pathway
  in lignin biosynthesis in Zinnia. Plant Cell 6, 1427-1439 
Youngquist, J. B., Carter, D. C. and Clegg, M. D. (1990). Grain and forage yield and stover quality 
 of sorghum and millet in low rainfall environments. Experimental Agriculture. 26 (3). 
 279-286 
Yusuff, O., Mohd, Y. R., Norhani, A., Ghazali, H., Asfaliza, R., Harun, A. R., Gous, M. and 
 Magaji, U. (2016). Principle and application of plant mutagenesis in crop improvement: a 
  review. Biotechnology & Biotechnological Equipment. 30:1.1-16 
Zezza, A., Federighi, G., Adamou, K. and Hiernaux, P. (2014). Milking the Data: Measuring 
  Income from Milk Production in Extensive Livestock Systems. Experimental Evidence 
  from Niger. World Bank Group. Policicy Research Working Paper 7114. 34p 
Zheng-Hua Ye, Ruiqin Z., Morrisson, H. W., Himmelsbach, D. S. (2000). Caffeoyl coenzyme A 
  O-methyltransferase and lignin biosynthesis. Phytochemistry 57. 1177-1185 
 
 
 
 
 
 
 
 
 
 
 
170 
 
University of Ghana http://ugspace.ug.edu.gh
APPENDICES 
Appendix 1.  Descriptive Statistic of 20 best bmr derived families and checks conducted at 
Tillabery in 2017 for chemical contents.  
DM  Mean SEM SD Mini Maxi 
Pop1 (Elmota//El mota/Tx630bmr12)) 95.80 0.20 0.69 94.05 96.67 
Pop2 (Sepon82//Sepon82/Redlanbmr6)) 95.37 0.24 0.58 94.42 95.98 
Pop3 (Sepon82//Sepon82/Wheatlandbmr12)) 94.82 0.34 0.47 94.48 95.15 
T1 (Redlanbmr6) 95.25     
T2 (Wheatlandbmr12) 95.35     
T3 (Tx630bmr12) 95.00     
T4 (El mota) 95.32     
T5 (Local Tillabery) 95.32     
T6 (Sepon82) 95.02     
Ash % Mean SEM SD Mini Maxi 
Pop1 (Elmota//El mota/Tx630bmr12)) 7.53 0.38 1.33 5.05 9.92 
Pop2 (Sepon82//Sepon82/Redlanbmr6)) 8.71 1.15 2.81 5.56 13.30 
Pop3 (Sepon82//Sepon82/Wheatlandbmr12)) 6.77 0.74 1.05 6.03 7.51 
T1 (Redlanbmr6) 4.99     
T2 (Wheatlandbmr12) 8.36     
T3 (Tx630bmr12) 6.86     
T4 (El mota) 10.98     
T5 (Local Tillabery) 11.09     
T6 (Sepon82) 11.10     
Cellulose % Mean SEM SD Mini Maxi 
Pop1 (Elmota//El mota/Tx630bmr12)) 31.06 0.34 1.19 28.56 32.56 
Pop2 (Sepon82//Sepon82/Redlanbmr6)) 28.33 1.17 2.87 24.20 31.75 
Pop3 (Sepon82//Sepon82/Wheatlandbmr12)) 27.25 1.05 1.49 26.20 28.31 
T1 (Redlanbmr6) 34.70     
T2 (Wheatlandbmr12) 30.05     
T3 (Tx630bmr12) 31.58     
T4 (El mota) 33.31     
T5 (Local Tillabery) 30.53     
T6 (Sepon82) 33.83     
Crude Proteins % Mean SEM SD Mini Maxi 
Pop1 (Elmota//El mota/Tx630bmr12)) 4.18 0.28 0.98 2.75 5.63 
Pop2 (Sepon82//Sepon82/Redlanbmr6)) 4.76 0.64 1.57 2.63 7.00 
Pop3 (Sepon82//Sepon82/Wheatlandbmr12)) 4.10 0.04 0.05 4.06 4.13 
T1 (Redlanbmr6) 3.75     
171 
 
University of Ghana http://ugspace.ug.edu.gh
T2 (Wheatlandbmr12) 4.25     
T3 (Tx630bmr12) 5.31     
T4 (El mota) 3.94     
T5 (Local Tillabery) 3.81     
T6 (Sepon82) 5.44     
Lipids %  Mean SEM SD Mini Maxi 
Pop1 (Elmota//El mota/Tx630bmr12)) 1.41 0.14 0.49 0.89 2.82 
Pop2 (Sepon82//Sepon82/Redlanbmr6)) 1.64 0.16 0.39 1.25 2.29 
Pop3 (Sepon82//Sepon82/Wheatlandbmr12)) 1.19 0.03 0.04 1.16 1.21 
T1 (Redlanbmr6) 0.58     
T2 (Wheatlandbmr12) 0.47     
T3 (Tx630bmr12) 1.37     
T4 (El mota) 1.68     
T5 (Local Tillabery) 2.20     
T6 (Sepon82) 2.32     
Ca % Mean SEM SD Mini Maxi 
Pop1 (Elmota//El mota/Tx630bmr12)) 0.19 0.03 0.10 0.10 0.39 
Pop2 (Sepon82//Sepon82/Redlanbmr6)) 0.10 0.02 0.05 0.05 0.15 
Pop3 (Sepon82//Sepon82/Wheatlandbmr12)) 0.10 0.06 0.08 0.04 0.15 
T1 (Redlanbmr6) 0.10     
T2 (Wheatlandbmr12) 0.17     
T3 (Tx630bmr12) 0.13     
T4 (El mota) 0.15     
T5 (Local Tillabery) 0.05     
T6 (Sepon82) 0.21     
P %   Mean SEM SD Mini Maxi 
Pop1 (Elmota//El mota/Tx630bmr12)) 0.19 0.02 0.06 0.10 0.32 
Pop2 (Sepon82//Sepon82/Redlanbmr6)) 0.17 0.04 0.09 0.06 0.27 
Pop3 (Sepon82//Sepon82/Wheatlandbmr12)) 0.13 0.11 0.15 0.02 0.23 
T1 (Redlanbmr6) 0.21     
T2 (Wheatlandbmr12) 0.25     
T3 (Tx630bmr12) 0.13     
T4 (El mota) 0.17     
T5 (Local Tillabery) 0.05     
T6 (Sepon82) 0.12     
NDF %   Mean SEM SD Mini Maxi 
Pop1 (Elmota//El mota/Tx630bmr12)) 60.66 0.87 3.02 55.66 65.17 
Pop2 (Sepon82//Sepon82/Redlanbmr6)) 57.24 1.32 3.23 53.99 62.15 
Pop3 (Sepon82//Sepon82/Wheatlandbmr12)) 61.31 2.02 2.85 59.29 63.32 
T1 (Redlanbmr6) 67.15     
T2 (Wheatlandbmr12) 64.90     
T3 (Tx630bmr12) 62.58     
172 
 
University of Ghana http://ugspace.ug.edu.gh
T4 (El mota) 62.45     
T5 (Local Tillabery) 62.77     
T6 (Sepon82) 64.82     
ADF %   Mean SEM SD Mini Maxi 
Pop1 (Elmota//El mota/Tx630bmr12)) 34.07 0.53 1.82 31.43 37.60 
Pop2 (Sepon82//Sepon82/Redlanbmr6)) 30.73 0.93 2.27 28.08 33.40 
Pop3 (Sepon82//Sepon82/Wheatlandbmr12)) 32.94 1.09 1.55 31.85 34.04 
T1 (Redlanbmr6) 38.24     
T2 (Wheatlandbmr12) 34.85     
T3 (Tx630bmr12) 33.38     
T4 (El mota) 33.57     
T5 (Local Tillabery) 33.95     
T6 (Sepon82) 37.61     
ADL %    Mean SEM SD Mini Maxi 
Pop1 (Elmota//El mota/Tx630bmr12)) 2.72 0.21 0.73 1.70 4.45 
Pop2 (Sepon82//Sepon82/Redlanbmr6)) 2.80 0.30 0.74 1.93 3.93 
Pop3 (Sepon82//Sepon82/Wheatlandbmr12)) 3.05 0.40 0.56 2.65 3.45 
T1 (Redlanbmr6) 4.67   4.67 4.67 
T2 (Wheatlandbmr12) 2.41     
T3 (Tx630bmr12) 2.83     
T4 (El mota) 4.06     
T5 (Local Tillabery) 2.56     
T6 (Sepon82) 5.41     
Energy kcal/kg DM Mean SEM SD Mini Maxi 
Pop1 (Elmota//El mota/Tx630bmr12)) 3404.30 83.30 288.40 2893.40 3755.00 
Pop2 (Sepon82//Sepon82/Redlanbmr6)) 3607.80 35.90 88.00 3506.80 3732.30 
Pop3 (Sepon82//Sepon82/Wheatlandbmr12)) 3683.00 9.46 13.40 3673.60 3692.50 
T1 (Redlanbmr6) 3611.40     
T2 (Wheatlandbmr12) 3714.30     
T3 (Tx630bmr12) 3664.30     
T4 (El mota) 3525.20     
T5 (Local Tillabery) 3392.80     
T6 (Sepon82) 3282.60     
 
 
 
 
 
 
173 
 
University of Ghana http://ugspace.ug.edu.gh
 
Appendix 2: Questionnaire SSI sorghum farmers 
Title: Improved forage quality of sorghum stover 
General objective: Generate information on the production and use of sorghum biomass in 
animal feed in the village of Soumarana in Niger 
1. Identification 
Full name of the manager: _________________________________ 
Age of manager: ______       Sex      Male Female 
Village ___________________________ Commune of: ___________________ 
Interviewer: _______________________________________ 
Date: ______________ 
Are you a member of an organisation?  YES        NO 
If YES which one? ____________________________________ 
Have you received any training? _______________________________ 
In what field? _________________________________________ 
Farm equipment: _______________________________ 
2. Activities 
What are your activities? ___________________________________________________ 
_________________________________________________________________________ 
__________________________________________________________________________ 
Among these which ones are the most important? (NB: from the most important to the least 
important) 
___________________________________________________________________________ 
__________________________________________________________________________ 
How do you manage these activities? _____________________________________________ 
174 
 
University of Ghana http://ugspace.ug.edu.gh
_________________________________________________________________________ 
How many hectares do you have for your farming activities? _____________________ 
 
3. Land tenure  
Use /soil type  Crops  Land Tenure: ha 
(TLBMCP)1 
-Irrigated Plot  Sorghum   
 
Millet   
-Non irrigated Plot  
Cowpea    
Others    
1T = traditional (belongs to extended family, distributed between members or not under the 
responsibility of the head of the family); 
L = legally owned (Obtained from a legal structure (administrative, town hall, etc.) ; 
B = lease /Rent;  
M = Share-cropping (with part of the crop each year);  
C = Land owned by a parent;  
P = Lent land (no reward: borrow). 
What is the soil condition of your plots in terms of fertility: ___________________________? 
___________________________________________________________________________ 
 Do you use organic fertilizers?     YES NO 
If YES for which crop? _________________________ 
Where is the manure from? ______________________________________________________ 
___________________________________________________________________________ 
What quantity did you use last year? ___________  
In case of purchase specify the price: ____________ 
 
175 
 
University of Ghana http://ugspace.ug.edu.gh
3.1.Sorghum production  
Do you grow sorghum each year?                 YES                         NO  
Are you interested in new varieties developed by research? YES              NO  
Do you buy improved varieties each year?  OUI                   NO     
What are your production objectives? 
1. Grain yield only………………………….Why? 
 
 
2. Biomass only ………………..                            Why? 
o If sorghum forage crop; specify harvest period ……………….. 
 
3. Double usage                                              Why? 
 
Uses  Quantities (kg) 
Grain  
Biomass  
Marketing   
Others   
 
4. Animal breeding and milk production  
 
Do you any animals?                               YES                              NO 
If yes how many: 
-Cattle _______ 
-Sheep_______ 
-Goats _______ 
-Camels_______ 
-Donkeys_______ 
Do you produce any milk?               YES                            NO      
How much do you produce per day? ______ 
Do you sell milk?           YES                                                   NO  
Why? _______________ 
 
176 
 
University of Ghana http://ugspace.ug.edu.gh
Where is the herd? In the dry season: _______________ 
         In the rainy season: ______________ 
How many cows do you milk on average per day? __________ 
How many dairy cows do you have on average per year in the herd? ________ 
 
4.1. Animal Feed  
Do you have forage constraints in feeding of your animals? 
List them: 
- 
- 
- 
Quantity of animal feed purchased per season? ________ 
Purchase price per unit of measure (bag, bundle, etc.):  
Type of animal feed  Quantity Price  
Concentrated animal feed    
Cereal bran    
Cultivated fodder    
                                                     Crop Residues 
Sorghum   
Millet   
Cowpea    
Other   
 
 
According to you what criterion influences forage quality? 
- 
- 
177 
 
University of Ghana http://ugspace.ug.edu.gh
- 
Food quantity produced on the farm: 
Type of food  Quantity Price  
Crop residues    
Cultivated fodder    
Others    
 
4.2. Animal Health  
Costs of veterinary treatment _________________________________________ 
Number of veterinarian visits per year? ______________________________________ 
4.3. Reproduction  
Did you do any artificial insemination?  YES                    NO 
If YES Number of inseminated cows? ___________ How much? ______________ 
Success rate? _______________ 
 
What are the difficulties in milk production? _______________________________ 
___________________________________________________________________________ 
___________________________________________________________________________ 
What are the solutions according to you? ____________________________________________ 
___________________________________________________________________________ 
___________________________________________________________________________ 
 
 
Are you doing fattening?            YES                                      NO  
Why? __________________ 
                            
178 
 
University of Ghana http://ugspace.ug.edu.gh
 
What benefits do you derive from breeding your animals? 
- 
- 
- 
 What is the share of the income obtained with the marketing of the animals? 
-Important         
-Average           
- Weak             
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
179 
 
University of Ghana http://ugspace.ug.edu.gh
Appendix 3: Questionnaire SSI stover traders 
Title:. Improved marketing of sorghum fodder 
General objective:. Generate information on the marketing of sorghum straw and other fodder 
in Lossa, Dolli, and Soumarana 
 
4. Identification 
Full name of the manager: _________________________________ 
Age of manager: ______       Sex      Male Female 
Village ___________________________ Commune of: ___________________ 
Interviewer: _______________________________________ 
Date: ______________ 
Are you a member of an organisation?  YES        NO 
If YES, which one? ____________________________________ 
Have you received any training? _______________________________ 
In what field? _________________________________________ 
 
 
5. Types of marketed fodder  
 
Crops Quantity (kg) 
Sorghum   
 
Millet  
 
Cowpea   
 
Peanut   
 
Others   
 
180 
 
University of Ghana http://ugspace.ug.edu.gh
 
 
6. Supply  
 
- What is your period of supply? _________________ 
- What is the purchase price? _________________ 
- Who are your main suppliers? _________________ 
- What quality standards are required to purchase? _________________ 
 
7. Storage 
 
 
- What are the storage conditions? _________________ 
- How long is storage? _________________ 
- List the main constraints encountered during storage: _________________ 
- How do you manage these constraints? _________________ 
 
8. Sale  
 
- What is the sales period of the straw?  
181 
 
University of Ghana http://ugspace.ug.edu.gh
- What is the peak sales period? _________________ 
- What is the selling price of the bundle of straw throughout the year? _________________ 
- What are the quality standards required by agro-pastoralists for sale? 
9. Profitability  
- Do you have a rolling capital? YES                           NO 
- What benefit do you derive from this commercial activity? __________________ 
- What are your prospects in this business? __________________     
 
 
 
 
182