i NODULATION AND NITROGEN FIXATION POTENTIAL OF SESBANIA SPP. RHIZOBIA ON SESBANIA SESBAN (L.) MERR. AND ROSE COCO (PHASEOLUS VULGARIS L.) MAKATIANI EMMANUEL TENDWA (B.Sc.) REG. NO. I56/22389/2012 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE AWARD OF THE DEGREE OF MASTER OF SCIENCE (MICROBIOLOGY) IN THE SCHOOL OF PURE AND APPLIED SCIENCES OF KENYATTA UNIVERSITY. MARCH, 2018 ii iii DEDICATION This work is dedicated to my family Rose Nasimiyu (Mama Goddy), Linda D. Indakwa, Godfrey Musa, Alex Imbahala, Linda I. Makatiani, Diana Ihavi and Larisa Makatiani. iv ACKNOWLEDGMENT First I thank The Almighty God for health and life. I express my deepest gratitude to my supervisors Dr. John M. Maingi, Dr. Omwoyo Ombori and Dr. David W. Odee for accepting me as their M.Sc. student and for providing all necessary support to finish my research work. I owe my deepest gratitude to all of them for their critical advice in research, scientific guidance, confidence and enormous support during my study period. I am deeply indebted to KEFRI administration for financial support towards my field and laboratory work as well as research facility (molecular, microbiology laboratories and glasshouse) at Muguga complex. I am grateful for the technical advice, support and encouragement provided by the KEFRI Biotechnology staff and in particular Joseph Machua, John Gicheru, John Ochieng, Stephen Omondi and Milton Esitubi. Special thanks go to Charles Oduor for his technical assistance during my work under glasshouse conditions. I appreciate Mr. Morris Muthini for his technical advice in molecular protocols. I am deeply grateful to my wife, Rose Nasimiyu for her limitless inspiration, understanding, sacrifice and taking care of our children during the time of my study. I am grateful to AFORNET project and Dr. Herta Kolberg for nodule collection in East Africa and Namibia respectively. I can never thank you enough and also it is hard to include all persons by name in this page. May God bless you all in abundance. v TABLE OF CONTENTS TITLE. .......................................................................................................................... i DECLARATION ............................................................. Error! Bookmark not defined. DEDICATION ............................................................................................................ ii ACKNOWLEDGMENT ............................................................................................ iv TABLE OF CONTENTS ............................................................................................ v LIST OF FIGURES .................................................................................................. xii LIST OF PLATES ................................................................................................... xiv ACRONYMS AND ABBREVIATIONS .................................................................. xv ABSTRACT............................................................................................................ xviii CHAPTER ONE ......................................................................................................... 1 INTRODUCTION ....................................................................................................... 1 1.1 Background information.......................................................................................... 1 1.2 Problem statement ................................................................................................... 4 1.3 Justification ............................................................................................................. 4 1.4 Research hypotheses ............................................................................................... 5 1.5 Objectives ............................................................................................................... 6 1.5.1 General objective ................................................................................................. 6 1.5.2 Specific objectives ............................................................................................... 6 1.6 Significance of the study ......................................................................................... 6 CHAPTER TWO ........................................................................................................ 8 LITERATURE REVIEW ........................................................................................... 8 2.1 Nitrogen reserves in the atmosphere and soil ........................................................... 8 2.2 Mechanisms of nitrogen fixation ............................................................................. 9 2.2.1 Mechanisms of biological nitrogen fixation ....................................................... .10 2.3 Legume–rhizobia symbiotic promiscuity ............................................................. ..13 2.4 The degree of specificity and nitrogen fixation effectiveness between legumes and rhizobia ......................................................................................................... 14 2.5 Sesbania sesban and legume nodulating bacteria ................................................... 16 2.6 Common beans ..................................................................................................... 17 2.6.1 Origin and distribution of common beans ........................................................... 17 2.6.2 Uses of common beans ....................................................................................... 18 2.6.3 Common beans production and its limitations .................................................... 18 2.6.4 Biology of common beans .................................................................................. 19 vi 2.6.5 Symbiosis of common beans with rhizobia ........................................................ .20 2.7 Characterization and identification of rhizobia ..................................................... .20 2.7.1 Methods for rhizobial identification................................................................... .21 2.7.1.1 Phenotypic identification......................................................................................22 2.7.1.2 Molecular techniques used in rhizobial identification..........................................22 2.8 Characteristics and current classification of rhizobia ............................................. 26 CHAPTER THREE .................................................................................................. 28 MATERIALS AND METHODS .............................................................................. 28 3.1 Study site .............................................................................................................. 28 3.2 Source of nodules .................................................................................................. 28 3.3 Media for culturing of rhizobia .............................................................................. 30 3.4 Isolation, purification and preservation of root nodule bacterial isolates ................ 30 3.5 Phenotypic characteristics of sesbania rhizobia ..................................................... 31 3.5.1 Determination of rhizobial growth rate and colony characteristics ...................... 32 3.5.2 YEMA‒BTB medium colour change by sesbania rhizobial isolates .................... 32 3.5.3 Gram staining ..................................................................................................... 32 3.6 Determination of genotype composition of sesbania rhizobia using PCR‒RFLP .... 33 3.6.1 Intrinsic antibiotic resistance assay ..................................................................... 33 3.6.1.1 Preparation of YEMA‒antibiotics media.............................................................35 3.6.1.2 Rhizobial culturing for intrinsic antibiotic resistance assay.................................35 3.6.1.3 Inoculation of rhizobial cultures on YEMA‒ antibiotics media..........................36 3.6.2 Screening of rhizobial isolates for salt tolerance levels ....................................... 36 3.6.3 Selection of sesbania rhizobia for PCR‒RFLP assays ......................................... 36 3.6.4 Rhizobial DNA extraction .................................................................................. 37 3.6.5 PCR mastermix preparation ................................................................................ 39 3.6.6 16S rRNA‒PCR amplification conditions ........................................................... 40 3.6.7 Gel electrophoresis of rhizobial 16S rRNA PCR products .................................. 40 3.6.8 Restriction of 16S rRNA gene PCR amplicons ................................................... 40 3.6.9 Gel electrophoresis of restriction fragments of rhizobial 16S rRNA .................... 41 3.7 Isolates effectiveness test with Sesbania sesban and common bean plants using Leonard jar assembly ........................................................................................... 41 3.7.1 Plant materials .................................................................................................... 41 3.7.2 Experimental design ........................................................................................... 42 3.7.3 Seed pretreatment and pre‒germination .............................................................. 42 3.7.4 Preparation of Leonard jar assembly ................................................................... 42 vii 3.7.5 Preparation of nitrogen‒free plant nutrient solution ............................................ 43 3.7.6 Transfer of pre‒germinated seeds to vermiculite in Leonard jars ........................ 43 3.7.7 Inoculant preparation and inoculation of the host plant ....................................... 44 3.7.8 Harvesting of Sesbania sesban and common beans ............................................. 45 3.7.9 Determination of shoot total nitrogen ................................................................. 45 3.7.10 Determination of symbiotic efficiencies (%) ..................................................... 46 3.8 Data analysis ......................................................................................................... 47 CHAPTER FOUR..................................................................................................... 48 RESULTS .................................................................................................................. 48 4.1 Morphological characterization ............................................................................. 48 4.2 Characteristics of sesbania rhizobia on YEMA‒BTB media .................................. 57 4.3 Gram staining........................................................................................................ 59 4.4.1 Intrinsic antibiotic resistance of sesbania isolates ............................................... 60 4.4.2 Rhizobia tolerance to sodium chloride ................................................................ 70 4.4.3 Selection of rhizobial isolates for PCR‒RFLP .................................................... 71 4.5.1 Estimation of rhizobial 16S rRNA amplicons using gel electrophoresis .............. 81 4.5.2 Separation of 16S rRNA restriction bands using gel electrophoresis ................... 83 4.5.3 Phylogenetic clusters of sesbania rhizobia .......................................................... 85 4.6 Symbiotic efficiency test of sesbania rhizobia on Sesbania sesban plants using Leonard jars in the greenhouse ............................................................................. 96 4.6.1Correlation among shoot dry weight, nodule number, nodule dry weight, nitrogen concentration and SE of S. sesban ……………………………………...110 4.7 Symbiotic effectiveness test of sesbania rhizobial isolates on Rose coco bean plants using Leonard jars .................................................................................... 113 4.7.1 Correlation among shootdry weight, nodule number, nodule dry weight, nitrogen concentration and SE of common bean plants…………………………..124 4.8 Multiple rhizobial occupancy in root nodules of Sesban sesban ........................... 129 4.9 Sesbania sesban and common bean cross‒inoculating isolates ............................ 131 CHAPTER FIVE..................................................................................................... 133 DISCUSSION, CONCLUSIONS AND RECOMMENDATIONS ........................ 133 5.1 Discussion........................................................................................................... 133 5.1.1 Morpho–cultural characterization of sesbania rhizobia ..................................... 133 5.1.2 Intrinsic antibiotic resistance and salt tolerance ................................................ 135 5.1.3 Molecular characterization of sesbania rhizobia...................................................138 5.1.4 Nodulation and nitrogen fixation potential of sesbania rhizobia on S. sesban .... 138 viii 5.1.5 Nodulation and nitrogen fixation potential of sesbania rhizobia on Rose coco bean variety………………………………………………………………………..141 5.2 Conclusions ........................................................................................................ 143 5.3 Recommendations ............................................................................................... 143 REFERENCES ........................................................................................................ 145 APPENDICES ......................................................................................................... 164 ix LIST OF TABLES Table 3.1: Origin of sesbania nodules used in this study ………………... 29 Table 3.2: Reference strains used in PCR‒RFLP assays, their host plants and sources …………………………………..………….......... 37 Recipe for Hexadecyltrimethylammonium bromide extraction buffer ……………………………............................................. 38 Characteristics of sesbania rhizobial isolates on YEMA‒CR media …………………………………………………………. 49 Percent composition of sesbania rhizobia in each of the nine morphotypes …………………………………………………. 51 Distribution of S. sesban isolates from four Kenyan sites in nine morphotypes …………………………………………...... 54 Distribution of S. sesban isolates from three Ugandan sites in nine morphotypes….……………………………….................. 55 Distribution of S. sesban isolates from two Tanzanian sites in nine morphotypes…..……………………………………........ 56 Distribution of sesbania isolates from eight Namibian sites in nine morphotypes……………………………………………... 57 pH reaction of sesbania rhizobial isolates on YEMA‒BTB media……………………………………………….................. 58 Percent intrinsic antibiotic resistance of S. sesban rhizobia from Kenya................................................................................. 64 Percent intrinsic antibiotic resistance of S. sesban rhizobial isolates from Uganda………………………………………..... 65 Percent intrinsic antibiotic resistannce of S. sesban rhizobial isolates from Tanzania............................................................... 67 Percent antibiotic resistance of Sesbania spp. rhizobial isolates from Namibia………………………………………… 69 List of sesbania rhizobial isolates and reference strains used for PCR‒RFLP assay ………………………………………… 80 Table 3.3: Table 4.1: Table 4.2: Table 4.3a: Table 4.3b: Table 4.3c: Table 4.3d: Table 4.4: Table 4.5a: Table 4.5b: Table 4.5c: Table 4.5d: Table 4.6: Table 4.7a: Sesbania sesban rhizobia from Kenya clustered using UPGMA based on combined patterns of 16S rRNA PCR‒RFLP (MspI, HinfI and HaeIII) compared to reference strains ………………………………………………………… 88 x Table 4.7b: Sesbania sesban rhizobia from Uganda clustered using UPGMA based on combined patterns of 16S rRNA PCR‒RFLP (MspI, HinfI and HaeIII) compared to reference strains ……………………..…………………………………. 90 Table 4.7c: Sesbania sesban rhizobia from Tanzania clustered using UPGMA based on combined patterns of 16S rRNA PCR‒RFLP (MspI, HinfI and HaeIII) compared to reference strains ………………………………………………………… 93 Table 4.7d: Sesbania sesban rhizobia from Namibia clustered using UPGMA based on combined patterns of 16S rRNA PCR‒RFLP (MspI, HinfI and HaeIII) compared to reference strains ………………………………………………………… 96 Table 4.8: Effect of sesbania rhizobial isolates from Kenya, Uganda, Tanzania and Namibia on shoot dry weight, nodule dry weight and nodule number of S. sesban...…………………….............. 99 Nodulation and nitrogen fixation phenotypes of Sesbania sesban rhizobial isolates from Kenya on S. sesban…………... 103 Table 4.9a: Table 4.9b: Nodulation and nitrogen fixation phenotypes of Sesbania sesban rhizobial isolates from Uganda on S. sesban…………. 104 Table 4.9c: Nodulation and nitrogen fixation phenotypes of Sesbania sesban rhizobial isolates from Tanzania on S. sesban………... 105 Table 4.9d: Nodulation and nitrogen fixation phenotypes of sesbania rhizobial isolates from Namibia on S. sesban………………… 106 Table 4.9e: Nodulation and nitrogen fixation phenotypes of reference strains on S. sesban…………………………………………… 107 Table 4.10: Effect of some sesbania rhizobial isolates on percent nitrogen, N content, N fixation ratio and the symbiotic effectiveness of S. sesban……………………………………………………… 108 Table 4.11: Correlation between shoot dry weight, nodule number, nodule dry weight, nitrogen concentration and symbiotic effectiveness of rhizobia in S. sesban………………………… 112 Table 4.12: Effect of sesbania rhizobial isolates on shoot dry weight, nodule dry weight and nodule number of common beans……. 116 Table 4.13: Effect of sesbania rhizobial isolates on percent nitrogen, N content, N fixation ratio and the symbiotic effectiveness of common beans ……………………………………………….. 118 Table 4.14a: Nodulation and nitrogen fixation phenotypes of S. sesban rhizobial isolates from Kenya on common beans ……………. 120 xi Table 4.14b: Nodulation and nitrogen fixation phenotypes of S. sesban rhizobial isolates from Uganda on common beans..………….. 121 Table 4.14c: Nodulation and nitrogen fixation phenotypes of S. sesban rhizobial isolates from Tanzania on common beans………….. 122 Table 4.14d: Nodulation and nitrogen fixation phenotypes of sesbania rhizobial isolates from Namibia on common beans…………... 123 Table 4.14e: Nodulation and nitrogen fixation phenotypes of reference rhizobial strains from Kenya on common beans…………….... 124 Table 4.15: Correlation between shoot dry weight, nodule number, nodule dry weight, nitrogen concentration and SE on common beans........................................................................................... 126 Table 4.16: Sesbania rhizobia with infective traits on common beans...….. Table 4.17: Growth response of S. sesban on inoculation using multiple nodules co-occupant rhizobia…………………………………. 130 128 xii LIST OF FIGURES Fig. 4.1: Neighbour joining dendrogram constructed based on colony morphological traits using UPGMA showing relatedness of sesbania rhizobial isolates …………………………………....... 52 Fig. 4.2: Colour change of YEMA-BTB by sesbania rhizobial isolates..... 59 Fig. 4.3 Schematic diagram showing orientation of isolates on Plate 4.5................................................................................................. 61 Fig. 4.4: Resistance of sesbania rhizobial isolates to antibiotics................ Fig. 4.5: Tolerance of sesbania rhizobia to different sodium chloride concentrations……………………………………....................... 71 Fig. 4.6a: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Bumala …………………………………………………………. 72 Fig. 4.6b: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Kuinet ………………….............................................................. 72 Fig. 4.6c: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Gituamba ………………………………………………………. 73 Fig. 4.6d: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Kavutiri …………………........................................................... 74 Fig. 4.6e: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from 75 Tororo ………………………………………………………… Fig. 4.6f: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from 75 Mbale …………………………………………………………. Fig. 4.6g: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from 76 Kabale ………………………………………………………… 62 xiii Fig. 4.6h: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Lushoto ………………………………………………………… 77 Fig. 4.6i: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from SUA …………………................................................................. 78 Fig. 4.6j: Unrooted neighbour joining dendrogram constructed based on combined IAR and salt tolerance using UPGMA method showing relatedness clusters I‒V of sesbania rhizobial isolates 79 from Namibia ………………………………………………… Fig. 4.7a: Unrooted UPGMA dendrogram showing clusters generated by combined restriction patterns of amplified 16S rRNA (HinfI+HinfI+HaeIII) of S. sesban rhizobia from Kenya and 86 reference strains.……………………………………………… Fig. 4.7b: Unrooted UPGMA dendrogram showing clusters generated by combined restriction patterns of amplified 16S rRNA (HinfI+HinfI+HaeIII) of S. sesban rhizobia from Uganda and reference strains ……………………………………………….. 89 Fig. 4.7c: Unrooted UPGMA dendrogram showing clusters generated by combined restriction patterns of amplified 16S rRNA (HinfI+HinfI+HaeIII) of S. sesban rhizobia from Tanzania and reference strains ………………………………........................... 91 Unrooted UPGMA dendrogram showing clusters generated by combined restriction patterns of amplified 16S rRNA (HinfI+HinfI+HaeIII) of sesbania rhizobia from Namibia and reference strains…….................................................................... 94 Fig. 4.7d: Fig. 4.8: Effectiveness of some selected sesbania rhizobial isolates on S. sesban plants.…………............................................................ 110 Fig. 4.9: Effectiveness of sesbania rhizobial isolates on common beans... 119 Fig. 4.10a: Response of inoculation on shoot dry weight (g) of S. sesban..... 131 Fig. 4.10b: Response of inoculation on shoot dry weight (g) of common bean …………………………………………………………...... 132 xiv LIST OF PLATES Growth of sesbania rhizobial on YEMA‒CR media showing colony morphotypes I‒IX …………………………………...... 50 Rhizobial colonies on YEMA media showing H, hollow centre.......................................................................................... 51 pH reaction characteristics of sesbania rhizobia on YEMA‒BTB media …………………………………………... 58 Plate 4.4: Gram’s negative rod‒shaped rhizobia……………………….... 60 Plate 4.5: Sensitivity of sesbania isolates to antibiotics ………................ 61 Plate 4.6: Sensitivity of sebania rhizobia to high salt levels (NaCl w/v) compared to controls ……………………………………….... 70 Plate 4.7: Agarose gel electrophoresis showing amplicon bands of 16S rRNA region …………………………………………………. 82 Plate 4.8a: Agarose electrophoresis patterns of 16S rRNA gene region of sesbania rhizobial isolates generated as a result of restriction digestion using endonuclease MspI............................................ 83 Plate 4.1: Plate 4.2: Plate 4.3: Plate 4.8b: Agarose electrophoresis patterns of 16S rRNA gene region of sesbania rhizobial isolates generated as a result of restriction digestion using endonuclease HinfI…………………………... 84 Plate 4.8c: Agarose electrophoresis patterns of 16S rRNA gene region of sesbania rhizobial isolates generated as a result of restriction digestion using endonuclease HaeIII………………………..... 84 Plate 4.9: Eight week-old S. sesban plants inoculated with sesbania rhizobial isolates.…………………………………………........ 97 Plate 4.10: Transverse section of nodules formed by sesbania isolates on S. sesban……………………………………………………..... 98 Plate 4.11: Four week‒old common bean plants inoculated with sesbani rhizobial isolates.…………………………………………….... 113 Plate 4.12: Nodules on roots of four‒week old common bean inoculated using S. sesban rhizobial strain MASS133………………….... 114 Plate 4.13: Transverse section of nodules formed by sesbania isolates on common bean…………………………………………………. 115 xv ACRONYMS AND ABBREVIATIONS ACIAR Australian Centre for International Agricultural Research AFORNET African Forest Research Network ALFP Amplified Fragment Length Polymorphisms ANOVA Analysis of Variance ARISA Automated Ribosomal Intergenic Spacer Analysis atm atmospheres ATP Adenosine Triphosphate BA Beatrice Anyango BMC BioMed Central BNF Biological Nitrogen Fixation bp base pair BSA Bovine Serum Albumin BTB Bromothymol Blue CA California CFN Centro de Investigacion sobre Fijacion de Nitrogeno CIAT Centro Internacional d'Agricultura Tropical CIRA Centre International de Recherches sur l'Anarchisme CIRAD Center for International Cooperation in Agronomic Research for Development CR Congo Red CTAB Cetyl trimethyl ammonium bromide DGGE Denaturing Gradient Gel Electrophoresis DNA Deoxyribonucleic acids DWO David Warambo Odee EDTA Ethylenediaminetetra‒acetic acid EPS Exopolysaccharide FAO Food and Agriculture Organization FEMS Federation of European Microbiological Societies g gram GenAlEx Genetic Analysis in Excel GENSTAT General Statistics HSD Honest Significant Difference xvi IAR Intrinsic Antibiotic Resistance IGS Intergenic Spacer IVS Intervening Sequences KARI Kenya Agricultural Research Institute kb kilobase KEFRI Kenya Forestry Research Institute KFR Kenya Forestry Research kJ Kilo Juoles LNB Legume Nodulating Bacteria masl metres above sea level MASS Makatiani Sesbania sesban MEGA Molecular Evolutionary Genetics Analysis MIRCEN Microbiological Resources Centers mL Millilitre µL Microlitre MLSA Multilocus Sequence Analysis MN Makatiani Namibia MoALF Ministry of Agriculture, Livestock and Fisheries (Kenya) MPa Mega Pascals NDWt. Nodule Dry Weight NGS Next-Generation Sequencing NiFTAL Nitrogen Fixation by Tropical Agricultural Legumes NJ Neighbour Joining NNo. Nodule Number PABRA Pan-African Bean Research Alliance PAST PAleontological STatistics PCR Polymerase Chain Reaction PGP Plant Growth Promoting pH potential Hydrogen PLOS Public Library of Science Psi Pascal per square inch PVP Polyvinylpyrrolidone RAPD Random Amplified Polymorphic DNA RFLP Restriction Fragment Length Polymorphism xvii RISA Ribosomal Intergenic Spacer Analysis rRNA ribosomal RNA SDWt. Shoot Dry Weight SE Symbiotic effectiveness SNWt. Specific Nodule Weight SUA Sokoine University of Agriculture TE Tris‒EDTA TGGE Temperature Gradient Gel Electrophoresis TSBF Tropical Soil Biology and Fertility VE Very Efficient UN United Nations UPGMA Unweighted Pair Group Method with Average USDA United States Department of Agriculture UVP Ultraviolet Products YEMB Yeast Extract Mannitol broth YEMA Yeast Extract Mannitol Agar xviii ABSTRACT Global crop and energy production are fast dwindling inversely to population growth. Common bean (Phaseolus vulgaris L.) yield has reduced worldwide due to soil infertility which can be reversed through application of chemical fertilizers. The chemical fertilizers used to ameliorate nitrogen, phoshorus and potassium are expensive and cause both deleterious physico‒chemical modification of soil and water mass eutrophication. The common bean has ability to fix atmospheric nitrogen symbiotically with rhizobia but only a few effective strains have been recovered from African soils, most of them ineffective under field conditions. Prospecting from local pool of strains trapped by wild native species like sesbania can increase the number of elite inoculant production strains for both species in agroforestry systems practiced under diverse soil and eco-climatic conditions. The specific objectives of this study were: to assess the phenotypic and genotypic characteristics of rhizobia from root nodules of East Africa and Namibia Sesbania spp. (here after referred to as sesbania) using morpho-cultural characteristics and PCR‒RFLP methods; to assess the nitrogen fixation potential of sesbania isolates on S. sesban using growth parameters; and to determine the infectiveness and symbiotic effectiveness of sesbania rhizobia on common beans. Experiments were carried out at the Kenya Forestry Research Institute, Muguga‒Nairobi, Kenya. Morpho-cultural techniques were used to characterize and cluster 128 presumptive sesbania rhizobia collected from Kenya, Uganda, Tanzania and Namibia. The diverse growth characteristics of rhizobia on YEMA media, intrinsic antibiotic resistance and salt tolerance were used to select for 79 sesbania isolates that were later subjected to fingerprinting assays using PCR‒RFLP of the 16S rDNA in comparison with 17 reference strains. The presumptive sesbania rhizobial isolates were used to inoculate S. sesban and common bean under glasshouse controlled conditions to test for their infectiveness and symbiotic effectiveness. Reference inoculants strains KFR647 and BA37 for S. sesban and bean respectively, were included in the test. Uninoculated positive control (70 ppm N as KNO3) and a negative control (0 ppm N) were included for strain effectiveness comparison and to check for contamination. The sesbania rhizobia were grouped into nine morphotypes and various ribotypes per site. The rhizobia varied in their infectiveness and symbiotic effectiveness on S. sesban and the common bean resulting in three categories viz: (1) highly effective (2) effective and (3) ineffective. The mean shoot dry weight, nodule number and nodule dry weight were all significantly different (p < 0.001). The shoot N content range was 0.16‒5.66 mg plant-1 and 0.34‒3.08 mg plant-1 for S. sesban and common beans at 8 and 4 weeks of growth respectively. Based on shoot dry weight due to inoculation, rhizobial isolate KFR402 was preferred as a common inoculant production strain for both common beans and S. sesban. However, data in the present study shows that the highest shoot dry weight was obtained with strain MASS133 (S. sesban) inoculated on Rose coco bean variety (0.87 g plant-1) and MASS172 (S. sesban) on S. sesban (1.06 g plant-1). Rhizobia recovered from sesbania grown in East Africa and Namibia are phenotypically and genetically diverse. The isolates exhibit great variations in effectiveness and nitrogen fixation efficiency on S. sesban and common beans (variety: Rose coco). Prospecting for elite rhizobia inoculant strains should be prioritized and tested for effectiveness on both S. sesban and common bean grown in diverse edaphic and agro-ecological conditions under agroforestry systems. 1 CHAPTER ONE INTRODUCTION 1.1 Background information Food production in heavily populated regions has remained steady, but at levels too low to mitigate widespread nutritional deficiencies. This has led to famines worldwide, which has resulted in the clearing of swathes of forest lands to pave way for crop production thereby reducing forest acreage to less than the recommended 10 % cover (FAO, 2012). Although at the onset, converted forest land is rich in soil nutrients, leaching and fixation of nutrients together with over‒cultivation and crop harvesting methods like 'cut and carry' eventually diminishes the soil fertility (Yeshaneh, 2015). Apart from inadequate supply of water to crops, tropical soils have nitrogen as the number one deficient plant nutrient followed by phosphorus. The only environmental friendly and feasible way to alleviate nitrogen deficiency in soils is the inclusion of nodule forming herbaceous crops and tree legumes on farms, which constitutes agro‒forestry. Common beans, peas, groundnuts are among legume crops grown in East and Southern Africa. The common bean appears on many agricultural farmlands worldwide (Morosan et al., 2017). The pulse is an annual herb that matures between one and three months (Koskey et al., 2017), hence used to alleviate hunger during food scarcity. Although common bean also provides high protein content that replaces meat which has become very expensive for the poor (Romero-Arenas et al., 2013), its current production is far less than the expected yields. For instance, there was a common bean deficit of 26 kilo-tonnes in 2014 in the Kenyan market (MoALF, 2015). This is more often than not due to low soil fertility experienced in 2 East and Southern Africa. Therefore, a common bean crop must be supplemented with nitrogen‒containing chemical fertilizers especially when growing in intensively farmed lands or naturally infertile soils. However, common bean is a legume that is expected to fix atmospheric nitrogen for their protein precursors if they successfully enter into symbiosis with compatible micro‒symbionts (Ribeiro et al., 2013). But being a naturalised legume in African soils, it often fails to fix atmospheric nitrogen due to the presence of ineffective rhizobia or occurrence of inadequate effective strains or total absence of compatible rhizobia (Catroux et al., 2001). Such scenarios call for a mandatory inoculation of seeds before they are sown and hence the need to select for rhizobia of high nitrogen fixation effectiveness. One of the methods used for selection of superior rhizobia is the determination of cross‒inoculation groups that may consist of crops and wild legumes (Gaur and Sen, 1979) whose members have ability to share the symbionts. Few trials have been made to cross‒inoculate the common beans with rhizobia baited using wild legumes including multipurpose trees grown on‒farm (Bala and Giller, 2001). A search for elite strains begins with collection of nodules from roots of legumes which act as baits for rhizobia from a consortium of microbes found in the soil followed by cross‒inoculation assays (Somasegaran and Hoben, 1994). Inoculation of hosts using a prospective inoculant strain is the only option to test for the symbiosis because identification and taxonomy does not globally reflect the symbiotic features of rhizobia, particularly their legume host range (Laguerre et al., 2001). Leguminous trees are also grown on agricultural lands in East and Southern Africa but their choice is dependent on the benefits, suitability and interaction with crops. 3 Some of the common tree legumes include Sesbania spp., Acacia spp., Leucaena spp., Calliandra spp. and Gliricidia spp. Like the common beans, sesbania are members of sub‒family Papilionaceae, family Leguminosae (Fabaceae) (Heuzé et al., 2015). Sesbania sesban is common in crop lands in East and Southern Africa intercropped with common beans and maize. Sesbania sesban, as well as other species of the genus have multiple uses that include fuelwood, soil nutrient enrichment fodder for both domestic animals and wild browsers and as medicinal plants (Gomase et al., 2012). The common bean has been associated with a broad range of Rhizobium spp., all of which have been described in different areas where common beans have been introduced globally (Anyango et al., 1995). Although the common bean has its origin in the Mesoamerican it has been reported to nodulate with rhizobia native to Africa, some of them ineffective in N2-fixation (Michiels et al., 1998). Such ineffectiveness in N2-fixation has partly been attributed to the ability of P. vulgaris to perceive nodulation signals from diverse rhizobial strains some of which are ineffective (Dall’Agnol et al., 2014). However, rhizobia similar to Rhizobium etli, R. phaseoli and R. tropici have been described in African soils (Anyango et al., 1995; Aserse et al., 2012) and shown to nodulate common beans. It is routine to use common beans as trap hosts during rhizobial population studies and bioprospecting for elite strains in new bean growing fields. The native rhizobial populations may have superior traits of resistance to adverse abiotic conditions found in most soils of the tropics (Zahran, 1999). Hence, the key role of this research was to determine the diversity of sesbania rhizobia and identify rhizobial strain(s) to be used as common inoculant for both S. sesban and the common beans in agroforestry systems. 4 1.2 Problem statement There is a great need to increase crop production and energy sources worldwide due to population increase (UN, 2013), which comes along with reduced purchasing power, especially for food. Global common bean production is way far below the expected yield mainly due to the ever declining soil fertility which can be reversed through application of chemical fertilizers (Fischer et al., 2014). The demand for common bean is on the increase as the world population and poverty increases (Porch et al., 2013). However, the use of chemical fertilizers is not only expensive, but has for long been implicated for both deleterious physico‒chemical modification of soil and water mass eutrophication (Savci, 2012). Common beans can establish symbiotic associations with several rhizobia species (Anyango et al., 1995); but the effectiveness of most strains under field conditions has been reported as very low (Ribeiro et al., 2013). So far, the pulse has been symbiotically associated with members of the genera Rhizobium and Sinorhizobium (Laguerre et al., 2001). The primary nitrogen fixation microsymbiont of the common beans is Rhizobium tropici which is found in native bean growing Mesoamerican and the Andean centre. Rhizobium tropici has been discovered associating with other legumes growing where the common bean is not native (Grange et al., 2007). 1.3 Justification Unlike the common bean, sesbania is native to Africa and associate effectively with a number of rhizobia genera including Rhizobium spp., Mesorhizobium spp. and Ensifer (Odee et al., 2002; Helene et al., 2015) among others. Exploration of new habitats and investigating nodulation of sesbania and other wild legumes cannot only 5 help to discover unidentified rhizobia, but also supports efforts of selecting effective combinations of legume‒Rhizobium genotype to exploit the huge potential of increased nitrogen fixation. Both common bean and S. sesban rhizobia are affiliated to members of the genus Rhizobium and therefore have a likelihood of associating with similar strains in the absence of their specific types. There are reports of frequent occurrence of common bean rhizobial isolates in East and Southern African soils that are more effective than CIAT899 (Anyango et al., 1995). However, little work has been carried out to determine cross-inoculating group containing P. vulgaris and other native legumes. Currently there is paucity of information on cross‒inoculation regarding the common bean and sesbania. Discovery and identification of elite rhizobial strains through cross‒inoculation experiments serve to make superior and multiple host inoculants hence reducing the cost of its production and purchase. Inoculation of legumes is not only environmental friendly way to enhance crop production but also a cheap way of reducing environmental pollution. 1.4 Research hypotheses i. H0: There are phenotypic and genotypic differences among the rhizobia isolated from various sesbania grown in East Africa and Namibia. ii. H0: Rhizobia from root nodules of sesbania are effective and efficiently fix nitrogen with S. sesban. iii. H0: Sesbania rhizobial isolates are effective and efficiently fix nitrogen with Rose coco bean variety. 6 1.5 Objectives 1.5.1 General objective To screen for common and highly effective rhizobial isolates for S. sesban and common beans from sesbania grown in diverse conditions of East Africa and Namibia for inoculant production. 1.5.2 Specific objectives i. To assess the phenotypic and genotypic characteristics of rhizobia isolates from root nodules of sesbania growing in East Africa and Namibia. ii. To assess effectiveness and nitrogen fixation efficiency of sesbania rhizobial isolates on S. sesban. iii. To determine the infectivity and nitrogen fixation potential of sesbania rhizobial isolates on Rose coco bean variety. 1.6 Significance of the study Rhizobia differ in the ability to compete, infect and fix nitrogen due to host compatibility and a myriad of abiotic conditions (Zahran, 2001). Identified rhizobia isolates from nodules of sesbania growing in East Africa and Namibia from this study will increase the knowledge of the diversity of rhizobia capable of forming nodules on S. sesban. Common beans‒S. sesban cross‒inoculating elite strains identified in the present study will be used for production of a common S. sesbanbeans inoculant instead of separate inoculants, thus reduce inoculant production cost and their purchase price. 7 A profusely nodulated perennial S. sesban crop following successful inoculation using a Rose coco–S. sesban elite rhizobial strain will perpetuate soil population of the strain when beans are off season hence, less need for inoculation of subsequent bean crop. Successful inoculation of these legumes using the selected elite strain(s) will lead to less use of nitrogenous fertilizers hence; mitigate eutrophication and hypoxia of water masses. The discovered rhizobial elite strain(s) for common beans from the present study will enhance production and hence alleviate the existing malnutrition. 8 CHAPTER TWO LITERATURE REVIEW 2.1 Nitrogen reserves in the atmosphere and soil Nitrogen (N) has globally received enormous attention probably much more than all other essential elements. It remains a requisite element for all living organisms for the reason that it determines the synthesis of nucleic acids which are essential for cellular functioning. Nitrogen represents about 2 % of the total plant dry matter that enters the food chain (Santi et al., 2013) while Bothe et al. (2007) estimated that all living organisms have nitrogen content of about 6.25 % dry weight. Naturally, approximately 1015 tonnes of dinitrogen gas (approximately 78 % N) is present in the atmosphere (Jordan, 1984) but unavailable to all living organisms. The available nitrogen reserves in the soil has remained small due to the comparatively large amounts withdrawn by plants, loss due to erosion, runoffs and leaching of nitrates (O'Leary et al., 1989). Nitrogen exists in soils as soluble inorganic ammonium and nitrate compounds; organic nitrogen (associated with soil humus) and ammonium nitrogen fixed by clay minerals (Schulten and Schnitzer, 1998). Soluble inorganic ammonium and nitrate compounds are the only forms readily available to higher plants, but rarely exceed 1‒2 % of the total N present in the soil. Most of the N in the soil is always metabolically unavailable to the higher plants leading to nitrogen deficiency-caused symptoms that may include yellowing and dropping of leaves, stunted growth, delayed flowering and fruiting (Dashora, 2011). The use of chemical nitrogen fertilizers in nitrogen deficient soils leads to a significant increase in crop yields. World fertilizer consumption trend has been on a steady increase since the 1960s. 9 For instance, between 2001 and 2009, the use of chemical nitrogenous fertilizers grew by 13 % (Stewart and Roberts, 2012). Food and Agricultural Organization (2012) estimated an annual increase of global chemical demand at 2 % between 2012 and 2016 with N fertilizer accounting for up to 60 % of the total costs of nitrogen, phosphorus and potassium (N+P+K) containing fertilizers. However, because of the inherent economical and environmental negative implications due to excessive use of nitrogen‒containing fertilizers (Kanimozhi and Panneerselvam, 2010), efforts have consistently been made to maximize on biological nitrogen fixation (BNF) (Baddeley et al., 2013). Soil nitrogen is mainly lost through volatilization and drainage of solutions into water masses. According to Saikia and Jain (2007) and Erisman et al. (2008), some of the adverse environmental effects of overreliance on nitrogenous fertilizers include: (i) depletion of large amounts of fossil fuels, a non‒renewable energy resources, (ii) eutrophication due to N in surface water, (ii) plant toxicity due to high levels of NO2 and ammonia (NH4) in soils and (iii) excessive plant growth due to excess nitrogen availability, depletion of stratospheric ozone due to nitric oxide (NO) and nitrous oxide (N2O). Consequently, the overreliance on chemical nitrogen fertilizers is a serious problem for sustainability and can be avoided by augmented use of BNF. 2.2 Mechanisms of nitrogen fixation The atmospheric diatomic nitrogen is inert due to its strong triple bond (N≡N) of energy 930 kJ mol‒1 (Wagner, 2012) and consequently entirely unavailable to all living organisms. However, the high energy barrier can be broken through three nitrogen fixation processes. First, the atmospheric fixation which is a result of lightning that breaks the triple bond and allows oxidation of nitrogen to form 10 nitrogen oxides which in turn dissolve in rain and falls back to earth. Second, industrial fixation by the Haber‒Bosch, a process in which nitrogen and hydrogen gases are heated to temperatures of 400‒450 °C and 20 MPa (200 atm) pressure (Hopkins and Dungait, 2010) to form liquid ammonia used in the manufacture of commercial chemical fertilizer. Finally, BNF that reduce N2 to NH3 at atmospheric pressure and environmental temperature through an ATP‒dependent process mediated by a multimeric enzyme complex, the nitrogenase found only in prokaryotes (Dighe et al., 2010). The Haber-Bosch process combines nitrogen from the air with hydrogen derived mainly from natural gas (methane) into ammonia. The reaction is exothermic, reversible and can be represented by the following equation: 2.2.1 Mechanisms of biological nitrogen fixation Mobilization of nutrients has been studied for decades, focusing mainly on biological nitrogen fixation, in particular, on the symbiosis between microbes and higher plants. Whilst symbiotic nitrogen fixation is currently well known, it is a still mystery to how long ago the first legumes (family Fabaceae) engaged members of protobacteria. A biochemical interplay between legumes and rhizobia leads to infection and organogenesis of nodules on roots or stems of legumes as well as some non‒legumes (Santi et al., 2013). The nodule forming microbes symbiotically fix atmospheric dinitrogen to the benefit of the plants and in return the plant supplies organic acids in particularly malate and succinate, which are produced from sucrose via sucrose synthase and glycolytic enzymes to rhizobia (Noisangiam et al., 2012). Rhizobia form nodules on legumes while actinomycetes known as Frankia (genus) cause root nodules and fix nitrogen on actinorhizal plants (non‒legume) such as Alnus spp., 11 Comptonia spp., Cassuarina spp. and Elaeagnus spp. (Kumar and Rao, 2012). Infection by rhizobia and Frankia normally lead to the formation of root nodules however, a few species for instance, Sesbania rostrata can form stem nodules especially when grown in submerged environments (Goormachtig et al., 2004). 2.2.1.1 Nodule formation on legumes Prokaryotes (bacteria or archaea) can fix nitrogen with or without direct association with eukaryotes (Figueiredo et al., 2013). The nitrogen fixation process is spontaneously switched on in presence of adequate carbon but limited nitrogen, resulting to estimated 1.44 x 108 tonnes N year-1 globally (Kumar et al., 2007). Divergent ideas and theories have attempted to explain the process of nodule formation and the subsequent nitrogen fixation in legumes. However, all agree to presence of a number of genes in bacteria that control different aspects of nodulation process and subsequent BNF. The complex process is orchestrated by a multitude of bacterial and plant signals (Ferguson et al., 2014) that starts with plant roots secreting specific cocktail of phenolic molecules, predominantly flavonoids and isoflavonoids into its rhizosphere. These molecules bind the transcriptional regulator NodD of compatible rhizobia and concomitantly stimulate a set of bacterial nodulation genes that lead to the synthesis of a highly specific signal molecule called Nod factors (lipochitooligosaccharide) which are perceived by legumes via LysM domain receptor kinases [lysin‒motif (LysM)] present on the root (Mandal et al., 2010; Black et al., 2012). For instance, host determinants of symbiosis specificity in soybean are GmNFR1 and GmNFR5 as main Nod factor receptors (Indrasumunar et al., 2011). Although only a small percent of legumes has been surveyed for 12 nodulation to date, Odee and Sprent (1992) described Acacia brevispica as a nod˗ factor transducer mutant found among a prominently nodulating legume genus. Nod factor perception triggers a subsequent signaling cascade that is necessary for appropriate nodule establishment and maintenance. In addition to the nod factors, many other bacterial lipopolysaccharides, compounds K‒antigen including exopolysaccharides polysaccharides, cyclic (EPS), β‒glucan, high‒molecular‒weight neutral polysaccharides (glucomannan) and gel‒forming polysaccharides affect different stages of the interaction (Lira et al., 2015). The attachment on the root hair of nodule forming bacteria possessing the capacity to synthesize nod factors, initiates root tip curling which culminates in bacteria entrapment and formation of an infection thread followed by initiation of cortical cell division (Shtark et al., 2011). Legumes root nodules are either indeterminate or determinate depending on whether or not the meristem remains active for the entire life of the nodule respectively (Subramanian, 2013). The indeterminate nodules, as those formed on roots of pea plants have their cell division first observed in the inner cortex. Determinate nodules whose cell divisions are first observed in sub‒epidermal cell layer are found on roots of common beans and sesbania among others legumes (Ferguson et al., 2010). In contrast to roots, both types of legume nodules have a peripheral vasculature. 2.2.1.2 The biochemical process of nitrogen fixation The biochemical process of nitrogen fixation is catalyzed by the complex enzyme called nitrogenase which is composed of two component proteins (dinitrogenase and 13 dinitrogenase reductase) encoded by the family nif genes and whose presence is characterized by pink coloration of the nodule tissues. The universal reaction N2+8H++8e-+16ATP 2NH3+H2+16ADP+16Pi), requires hydrogen as well as energy in form of ATP to progress. Leghemoglobin maintains a high oxygen flux, but at a concentration of approximately 104 to 105 times lower than in aerobic cultures to avoid inactivation of oxygen‒labile nitrogenase (Hunt and Layzell, 1993; Boyd and Peters, 2013). Excesses of oxygen hinder the transcription of the gene nifHDK or activity of the functional protein in aerobes but never a problem of anaerobic bacteria. 2.3 Legume–rhizobia symbiotic promiscuity Legumes and rhizobia have the ability to select their symbionts alike. This is achieved through the complex chemical signalling pathways that determine symbiosis specificity between rhizobia and the host plants (Angus and Hirsch, 2010). The ability to form nodules on roots or stems of more than one host rely on the promiscuity either of the legumes, or their rhizobia, or both (Peix et al., 2015). Some rhizobial strains have ability to nodulate individual or an extremely limited number of host species while promiscuous types can form nodules with a wide range of hosts across subfamilies. In the same way, some legume species are extremely restrictive and tolerate only a very narrow range of rhizobia whereas others are promiscuous and are nodulated by a wide range of rhizobia. Perret et al. (2000) and Lira et al. (2015) suggested that the rhizobia of herbaceous host species can nodulate more host plant species than those of woody legumes and tropical legumes are typically more promiscuous than temperate legumes respectively. 14 2.4 The degree of specificity and nitrogen fixation effectiveness between legumes and rhizobia Rhizobia represent great diversity between the genera with some closely related to non‒nodulating bacteria such as Agrobaterium or Brucella compared to each other (Sprent, 2001; Maróti and Kondorosi, 2014). The common feature is their ability to form nodules on legumes but not on all. The disparity in the ability to form nodules and fix nitrogen among rhizobia is due to the difference in their genome (Okazaki et al., 2010) and it is termed as host specificity (Györgypal et al., 1988). According to Wang et al. (2012), host specificity can occur during bacterial infection and nodule development as well as at the nitrogen fixation stage. Rhizobia whose nod genes are not transcribed (nod-) initiate no nodules on their legume hosts. Similarly, legume mutants with no nod factor transducing mechanisms fail to form nodules even in presence of rich rhizobial diversity. High degree of specificity between legumes and rhizobia has been described. For example, although the Nod factors produced by Rhizobium etli and Rhizobium loti are indistinguishable, the two species have distinct host ranges. Phaseolus spp. is restricted to Rhizobium etli while Lotus spp. forms symbiosis exclusively with Rhizobium loti (Cárdenas et al., 1995). Additionally, two Rhizobium strains that form nodules on one plant species may secrete different Nod factors. Rhizobium tropici and R. etli produce sulfated and acetylfucosylated Nod factors, respectively, yet both effectively form nodules on Proteus vulgaris (Mus et al., 2016). On the other hand, Bradyrhizobium japonicum, B. elkanii, Sinorhizobium fredii (strain USDA257) and Rhizobium sp. (strain NGR234) have a number of common hosts, but their Nod factors vary considerably (Perret et al., 2000). 15 Nitrogen fixation efficiency always varies with different host‒rhizobial combinations (Schumpp and Deakin, 2010). Rhizobia that offer insignificant benefits to their hosts are common in nature (El‒Maksoud and Keyser, 2010). In such scenarios the microsymbionts have been referred to as parasites (Denison and Kiers, 2004) or cheats (Kiers et al., 2003). Several symbiotic phenotypes exist and can be manifested by mutants as: the nodule deficient mutants impaired in the first steps of infection (nod˗); the bacterial mutants which induce nodules that present an early nodule senescence phenotype with blocked process of bacteroid differentiation (nod+ fix˗); fully differentiated but unable to reduce nitrogen to ammonium (nod+fix˗); differentiated into bacteroids less efficient in nitrogen fixation compared to the wild‒type strain (nod+ fix+/˗) (Maunoury et al., 2010; Liu et al., 2011; Saeki, 2011; Melino et al., 2012). Whilst, certain rhizobial strains can infect a host genotype and remain ineffective within the symbiosomes (nod+fix˗), the same strains can establish an effective interaction (nod+fix+) with other alternative host genotypes (Simsek et al., 2007). A perfect match between host and the microsymbiont result in effective nodules that have a red pigmentation on cross‒sections, signifying presence of leghemoglobin. Conversely, many isolates are known to produce ineffective nodules that border more on parasitism than symbiosis. Such ineffective rhizobia produce nodules that are greyish green or whitish on cross‒sections (Pommeresche and Hansen, 2017). Although symbiosis between rhizobia and legumes is a precise match, in some cases a certain level of mismatching is tolerated. A majority of rhizobia that form nodules on roots of wild legumes growing in most soils are diverse, most of which are promiscuous and a few with selective and specificity characteristics (Zahran, 2001). 16 Matching systems for many important legumes have been studied and categorized into cross‒inoculation groups, each of which consists of all the species that develop nodules when inoculated with rhizobia obtained from any one of them (Gaur and Sen, 1979). Although the cross‒inoculation matching system is used mostly as a guide for farmers, it can also be used for selecting the suitable rhizobial inoculant strain for the desired legume host(s). For instance, Syed et al. (2010) attempted to examine the host range of rhizobia isolated from nodules of Tephrosia purpurea, Crotolaria medicaginea, Leucaena leucocephala and Sesbania sesban in other legumes grown in Ajmer and Bikaner regions of Rajasthan. In this work, it was concluded that S. sesban was more specific with its rhizobia being compatible with few species outside the genus Sesbania. The only known way of testing for nitrogen fixation ability of a rhizobial isolate is when the host plant partners successfully forms nodules on the roots (infective) due to the inoculants (biofertilizers) with an ultimate alleviation of nitrogen deficiency symptoms (effective) of the host (Somasegaran and Hoben, 1994). 2.5 Sesbania sesban and legume nodulating bacteria Sesbania sesban is a fast growing tree species (Evans and Macklin, 1990) that has the ability to associate and fix atmospheric nitrogen in presence of its compatible rhizobia and nodulates most effectively with homologous strains (Turk and Keyser, 1992; Bala et al., 2001). However, many soils are devoid of rhizobia that are compatible with S. sesban. For example, S. sesban growing in some soil of South African failed to nodulate due to the absence of effective native rhizobia (Bala et al., 2002). Makatiani and Odee (2007) emphasized the need to determine the symbiotic status of sesbania whenever they are cultivated for the first time in order to make 17 appropriate decisions on whether to inoculate or not, which requires proper selection of actively fixing strains for inoculants production. Whereas common bean forms nodules with only members of Rhizobium spp., the genus Sesbania can form root or stem nodules with Azorhizobium (Goncalves and Moreira, 2004; Lee et al., 2008), Ensifer spp. syn. Sinorhizobium (Sharma et al., 2005), Mesorhizobium (McInroy et al., 1999; Odee et al., 2002) and Rhizobium spp. (Bala et al., 2002; Vinuesa et al., 2005) but their nitrogen fixation effectiveness remain largely unknown. Other reports from tropical Africa, Asia, Puerto Rico, Central and South America indicate that sesbania can be nodulated effectively by Bradyrhizobium sp. (Doignon‒Bourcier et al., 2000), Rhizobium huautlense (Wang et al., 1998), R. gallicum and R. tropici (Zurdo-Piñeiro et al., 2004), Sinorhizobium saheli and S. terangae (Lorquin et al., 1997), Azorhizobium caulinodans (Dreyfus et al., 1988), Mesorhizobium sp. and M. plurifarium (Wang and Martinez‒Romero, 2000; Wang et al., 2003). Bala et al. (2002) and Odee et al. (2002) reported frequent recovery of strains closely related to the genus Agrobacterium from sesbania nodules. 2.6 Common beans 2.6.1 Origin and distribution of common beans The common beans (Phaseolus vulgaris) has the Mesoamerican and the Andean as the two main centres of origin but are now widely distributed on many continents such as Asia, Europe and Africa (Becerra et al., 2011). It is the best known species of the genus Phaseolus in the family Fabaceae of about fifty plant species, all native to America (Romero-Arenas et al., 2013). Although the pulse was introduced to Africa approximately five centuries ago (Evans, 1976), common beans development to 18 increase on its yield through selection of genotypes suitable for diverse environmental conditions and their microsymbionts, is still under way. 2.6.2 Uses of common beans The common bean is a source of high concentration protein complex (cheap substitute for meat globally), carbohydrates, fiber, prebiotic, vitamin B, and chemically diverse micronutrient composition recommended for human consumption (Câmara et al., 2013). Besides providing income for rural smallholder families in Kenya, common bean is also a popular food to both the urban and rural population. In Africa, bean products are consumed at different stages of plant development hence, offering a prolonged and spread out food supply in the form of leaves, green pods, fresh grain, as well as dry grains. The common bean containing food is distributed in many forms including unprocessed seeds, canned products and as part of mixes. The pulse matures rapidly (Koskey et al., 2017) and serves as a key component in intensifying production through intercropping systems. Its ability to form symbiosis and fix nitrogen with rhizobia provides a long term solution for soil fertility improvement. 2.6.3 Common beans production and its limitations The world largest producer of common bean is Brazil with an estimated 3.3 million tons produced in 2016 (FAO, 2017). Although the common bean production in Kenya remained stable in 2014, 2015 and 2016 at 622.8, 714.4 and 616.0 kilo-tonnes respectively (FAO, 2017), its production is still way far below the expected yields due to the ever increasing population. The major limitation to common bean production in many smallholder farms is the declining soil fertility as a result of 19 continuous cropping with minimal inputs or rotation to replenish soil nutrients (Mungai and Karubiu, 2011). Other factors that influence bean production include seed density, chemical toxicities, pests and diseases, weeds, extreme environmental conditions among others (Beebe et al., 2013; Buruchara et al., 2011; Porch et al., 2013). For instance, in 2011 there was a net bean deficit in Kenya of 51.7 kilo-tonnes (FAO, 2013). In Kenya bean crop is mainly cultivated in rotation with maize and grows in diverse soil condition including low nitrogen and phosphorus, low pH, high salinity and low moisture (Maingi et al., 2001; Massawe et al., 2016). 2.6.4 Biology of common beans Common beans grow twining or sub‒erect and have numerous varieties which can be described in many shapes, sizes and colours. For example, kidney beans, lima beans, pinto beans, navy beans, green beans, wax beans and butter beans are just but a few of the many types found in the Americas (Gepts, 2014). Likewise, there also exists a number of Phaseolus spp. in Kenya each with a number of cultivars. Phaseolus vulgaris is the most dominant of all with over 100 cultivars grown in the country and its most common varieties include the Nyayo, Amini, Rose coco, Nyayo short, Kakunzu, Mwezi moja, Mwitemania, Wairimu, Kitui, Kitui small, Kayellow, Ikoso and Kamwithiokya which grow in the different eco‒climatic zones of Kenya (Ramaekers et al., 2013). Apart from seed colour and size, the varieties are differentiated by the growth habit (determinate and indeterminate), seed shape and days to maturity. 20 2.6.5 Symbiosis of common beans with rhizobia The common bean is known to form nodules with several Rhizobium species but the effectiveness of these strains under most field conditions is low (Dall’Agnol et al., 2013; Ribeiro et al., 2013). According to Hardarson et al. (1993), different varieties of common beans vary in their capacity to fix atmospheric nitrogen. Although common bean is generally considered more responsive than other legumes to nitrogen fertilization, it has been reported as a less nitrogen fixer than other crop legumes (Graham, 1981). Common bean is considered a promiscuous host that form nodules with Bradyrhizobium sp. (Han et al., 2005), Sinorhizobium meliloti (ZurdoPiñeiro et al., 2009), S. americanum (Mnasri et al., 2012), Rhizobium mesoamericanum (Lopez‒Lopez et al., 2012), R. freirei (Dall'Agnol et al., 2013), R. etli, R. tropici, R. leguminosarum bv. phaseoli, R. gallicum, R. giardinii, R. lusitanum, R. phaseoli, R. azibense (Mnasri et al., 2014) all in alpha‒proteobacteria and Burkholderia phymatum (Gyaneshwar et al., 2011) in class beta‒proteobacteria found in diverse soil ecosystems of the world. Although, R. tropici was originally isolated from host in native bean growing areas, it has been reported to associate with other legumes growing where the common bean is not native (Grange et al., 2007). 2.7 Characterization and identification of rhizobia Preliminary identification of cultivable microbes often involves morphological and physiological characteristics (Somasegaran and Hoben, 1994), which is attainable through microbial isolation, culturing and later visualized by observation using microscopes. Typing of microbes into their respective biotypes, serotypes, bacteriocin and phage groups remain also crucial during identification of microbes residing in animals, plants and soil. Data is derived from well‒established and 21 observable growth parameters, physiological and biochemical profiles. The physiological tests used for identification of microbes include fermentation of various carbohydrates, growth on carbon and nitrogen sources, determination of vitamin requirements, growth at various temperatures, growth on media containing various levels of sugar and sodium chloride, ability to hydrolyze urea and resistance to antibiotics (Fakruddin and Mannan, 2013). Accuracy and speed of microbial characterization and identification have increased with the advent of molecular biology techniques. The genotypic techniques involved in identification of microbes include restriction digestion and PCR amplification or hybridization, all of which employ DNA bands visualization (Auch et al., 2010), multilocus sequence analysis (MLSA) of different protein‒coding housekeeping genes and whole‒genome sequence analysis (Berrada and Fikri‒Benbrahim, 2014). These molecular tools are now used for investigating the diversity and structure of microbial communities including the uncultivable (Hill et al., 2000). 2.7.1 Methods for rhizobial identification 2.7.1.1 Phenotypic identification Phenotypic identification methods comprise the study of the morphology and biochemical reactions by bacteria whose properties can be observed after incubation of the cultures grown on solid media for a definite time. Morphological characteristics include colony characteristics such as size, shape, pigmentation, elevation, opacity, margin, surface texture and consistency. The biochemical tests use specific growth media, nutrients, chemicals or growth conditions to elicit visible characteristics. Some simple tests like the Gram's reaction, acid-fast reaction, 22 motility, arrangement of flagella, presence of spores, capsules and inclusion bodies aid in identification of the organism. However, phenotypic characteristics like isoenzyme profile, antibiotic susceptibility profile and chromatographic analysis of cellular fatty acids are sensitive enough for strain characterization. Although vital in the preliminary classification of microbes, Petti et al. (2005) noted that phenotypic traits are always not static hence, can change with stress or evolution. 2.7.1.2 Molecular techniques used in rhizobial identification Development of molecular techniques which complement the analysis of morphocultural traits has enhanced the ability to rapidly detect, identify and classify microbes. Moreover, these techniques have also been used to establish the taxonomic relationship among closely related genera and species (Mishra et al., 2016). Identification using molecular methods relies on the comparison of the nucleic acid (RNA and DNA) sequences or protein profiles of a microorganism with documented data on known organisms. The molecular methods are considered sensitive enough to allow detection of low concentrations of viable or non‒viable microbes in both pure cultures and complex samples as in soils and water. These include methods such as DNA re-association, DNA–DNA and mRNA-DNA hybridization, DNA cloning and sequencing and other PCR-based methods such as 16S rRNA, comparisons of restriction fragment length polymorphisms (RFLP), amplified fragment length polymorphisms (AFLP) or G+C % content in the genomic DNA with corresponding data on known species or strains (Vinay et al., 2013). Other molecular currently used to delineate in situ microbes include denaturing gradient gel electrophoresis (DGGE) temperature gradient gel electrophoresis (TGGE) (Smalla et al., 2007), ribosomal 23 intergenic spacer analysis (RISA) and automated ribosomal intergenic spacer analysis (ARISA) (Fakruddin and Mannan 2013; Namkeleja et al., 2016). After elucidation of double helical structure of DNA by Watson and Crick, the quest to detect its sequences from many organisms including human and microbial genomes pioneered the discovery of molecular methods such as Maxam and Gilbert’s technique (Maxam and Gilbert, 1977) and Sanger sequencing (Sanger and Coulson, 1975). As with many technologies, genomics has evolved at a remarkable pace leading to creation of dozens of next-generation sequencing companies, technologies and the corresponding field of bioinformatics. For instance, in the midto-late 1990s, microarrays were developed as highly parallel assays to measure RNA and DNA offering the first genome-scale parallel analysis of the nucleic acids (Heather and Chain, 2016). Second- and third-generation sequencing techniques emerged and have enabled unbiased means to scrutinize trillions of templates of DNA and RNA in a single instrument run. Coupled to current, high-performance computing and a host of bioinformatics tools that have been developed to analyze the data, whole genome sequences from an individual organism are generated (JimenezLopez et al., 2013). Collectively these technologies are referred to as next-generation sequencing (NGS) and share several characteristics including massive parallelization of chemical sequencing reactions, micro- to nano-scale reaction volumes. Additionally, there is immense amount of computational power to capture the raw data and process to formats interpretable by analysis software running on external computers (Levy and Myers, 2016). The NGS has opened new frontiers of genomics research which include clinical applications (Rizzo and Buck, 2012) and identification of soil and plant microbial communities (Lakshmanan et al., 2014). 24 2.7.2 Morphological and molecular characterization of rhizobia Distribution and diversity of rhizobia within tropical soils vary as their hosts and also with the ever changing diverse eco‒climatic conditions of their natural habitats (Odee et al., 1997; Hungria and Vargas, 2000). The diversity among nodulating bacteria may be due to the dynamic structure of bacterial genomes. For example, the organization of symbiotic genes within easily transmissible symbiosis islands or plasmids can permit the conversion of non‒symbiotic bacteria into nitrogen‒fixing plant endosymbionts in a single step (Marchetti et al., 2010). The host compatibility spectrum, cultural methods (media, morphology, antibiotic and other biochemical tests), serological methods, bacteriophage typing, molecular biology methods (nucleic acid hybridization and PCR-based techniques) and genomic sequencing have been used as criteria to consider during the study of rhizobia (Jia et al., 2015). The morphological properties include: size, shape, cell wall characteristics (Gram staining), surface characteristics and pigmentation, sporulation characteristics, mechanisms of motility, and other cellular inclusions and ultra‒structural characteristics (Somasegaran and Hoben, 1994). The recent advancement in molecular techniques have have been used to reorganized some genera and describe many other new genera and species of bacteria associated with tropical legumes (Odee et al., 2002; Menna et al., 2006; Toolarood et al., 2012; Hassen et al., 2014). These molecular techniques are highly versatile and have better resolution over the traditional cultural methods which they complement (Pierre and Didier, 2002; Capote et al., 2012; Marinkovic et al., 2013). They have led to phylogeny inferences that include 16S rRNA PCR‒RFLP analysis (Oger et al., 1998; Romdhane et al., 2005), 16S rRNA sequencing, RAPD (El‒Fiki, 2006), DNA‒DNA 25 hybridization (Murray et al., 2001; Auch et al., 2010), multilocus sequence analysis of different protein‒coding housekeeping genes and whole‒genome sequence analysis (Berrada and Fikri‒Benbrahim, 2014). 2.7.2.1 The bacterial 16S rRNA gene The rRNA genes play a great role in the protein synthesis and are therefore essential for the survival of all organisms. Among the rRNA genes is the 16S (small subunit) rRNA gene which is 1500 bp long with ten well conserved and ten divergent regions (Clarridge, 2004). However, cases of abnormally large 16S rRNA gene (larger than 1500 bp) have been documented and attributed to foreign DNA sequences, usually called intervening sequences (IVS) with about 140 bp long (Linton et al., 1994). For instance, Laguerre et al. (1994) found two strains, CFN299 and C‒05‒35 (both Rhizobium tropici type II) that produced a single band of 1600 bp when 16S rRNA was amplified which was attributed to insertion of 72 bp nucleotides. Constant mutation of the divergent regions of the 16S rRNA gene occurs at a rate of about 1 % every 50 years. Insertion or deletion of bases result in polymorphism of rRNA gene sequences and has been used in phylogenetic detection for more than two decades now (Moreira et al., 1998). The taxonomy of Rhizobiaceae like many other bacteria has undergone significant changes with more importance put on genotypic rather than phenotypic methods for the identification of strains to their species level (Babalola, 2003). This has resulted in a change in the number of recognized species of rhizobia and the reorganization of the family. These genotypic methods rely on the conserved nature of rRNA such that isolates from the same species maintain the same sequence, whereas the more 26 phylogenetically diverse the species is, the greater the divergence in the sequences of their rRNA. More over, although the 16S rRNA is sufficiently conserved and contains conserved regions, it also contains variable and hypervariable sequences (Spratt, 2004). 2.7.2.2 PCR‒RLFP of bacterial 16S rRNA Bacterial strains identification has widely been carried out using PCR‒RLFP of 16S rRNA, a technique which is currently regarded as quick and versatile (Wolde‒Meskel et al., 2005). The PCR‒RLFP procedures involve isolation of DNA, amplification of desired DNA sequence using primers that specifically target and amplify a region of desired genes of bacteria, digestion of the DNA with restriction endonucleases and size fractionation of the resultant DNA fragments by electrophoresis (Vinay et al., 2013). This method has extensively been used to assign bacterial collection to their likelihood phylogenetic groups. For instance, Manceau and Horvais (1997) used RFLP of rRNA operons to assess phylogenetic diversity among strains of Pseudomonas syringae pv. Tomato successfully established the close relationships existing between P. syringae and P. viridiflava. The PCR‒RFLP of 16S rRNA gene has been used broadly in assessment of phylogenetic diversity among strains of rhizobia (Mwenda et al., 2011; Mnasri et al., 2012). 2.8 Characteristics and current classification of rhizobia Rhizobia are non‒spore forming, aerobic, heterotrophic and motile bacteria (Sheu et al., 2016). They are Gram negative microscopic rods and defined as nitrogen fixing bacteria (diazotroph) capable of forming root or stem nodules on legumes (Mia and Shamsuddin, 2010). The family Rhizobiaceae of order Rhizobiales in classes alpha‒ 27 and beta‒ protobacteria have generally been called rhizobia (from rhiza‒bios; which live in a root). These Legume Nodulating Bacteria (LNB), to avoid confusion between the general term of Rhizobium and the genus name, grow as soil free‒living organisms, but can also induce and colonize root nodules in legume plants resulting in symbiotic relationships that benefits both organisms (Zakhia et al., 2004). Ever since Beijerinck isolated the first Rhizobium culture which he named Bacillus radicicola, later changed to Rhizobium leguminosarum by Frank in 1889 and assigned to the genus Rhizobium (Young et al., 2001; Willems, 2006), more than 113 species belonging to 11 genera within the Proteobacteria classes had been described by 2013 using phenotypic features and molecular tools. The genera and species under α‒Proteobacteria include: Rhizobium (43), Bradyrhizobium (15) Azorhizobium (3), Mesorhizobium (29) Ensifer/Sinorhizobium (13) Neorhizobium, Devosia (1) Methylobacterium (1) Ochrobacterium (2), Phyllobacterium (1) Shinella (1). Nine other species have been classified under β‒Proteobacteria and they include Burkholderia (6), Herbaspirillum (1) (http://edzna.ccg.unam.mx/rhizobialtaxonomy/node/4, and Cupriavidus (2) http://www.bacterio.cict.fr/ and http://www.rhizobia.co.nz/taxonomy/rhizobia). Recently, the γ‒Proteobacteria was also suggested and has been assigned the genera Xanthomonas and Pseudomonas (Berrada and Fikri‒Benbrahim, 2014). Unlike other members of Rhizobiaceae, the genus Agrobacterium is mainly soil‒inhabiting. However, some Agrobacterium strains possessing symbiotic plasmid as a result of horizontal or vertical transfer have ability to cause nodules on roots of legume plants (Cummings et al., 2009; Zhao et al., 2014). 28 CHAPTER THREE MATERIALS AND METHODS 3.1 Study site Controlled experiments were carried out in the laboratory and glasshouses located at the KEFRI, Muguga, Nairobi, Kenya which is globally positioned at 1°15'34.29'' S 36°37'36.05'' E and 1267 metres above sea level. 3.2 Source of nodules Nodules of varying sizes were randomely collected from roots of sesbania that were grown in diverse conditions of Namibia, Kenya, Uganda and Tanzania (Table 3.1). All the sites of nodule collection had no known previous sesbania inoculation using rhizobia. Nodules of Sesbania sesban, S. pachycarpa, S. sphaerosperma, S. microphylla, S. rostrata, S. macowaniana and S. bispinosa were collected by Dr. Herta Kolberg in Namibia, desiccated over unhydrous silica gel contained in air‒tight capped plastic bottles and transported to KEFRI. Nodules of S. sesban grown in East Africa were collected by the author of this thesis. The nodules were desiccated over silica gel contained in air‒tight capped plastic bottles before they were transported to KEFRI. 29 Table 3.1: Origin of sesbania nodules used in this study Country Origin Latitude / Longitude Sesbania Isolates Kenya Bumala (Busia) 00° 18' 9.1'' N 34° 12' 23.1'' E S. sesban MASS120‒128; 156‒178. Kuinet (Eldoret) 00° 36' 7.8'' N 35° 18' 28.0'' E " MASS169‒175. Kavutiri (Embu) 00° 25' 0.04'' S 37° 30' 06.2'' E " MASS129‒155. Gituamba (Murang'a) 00° 52' 45.7'' S 36° 54' 21.5'' E " MASS110‒117. SUA (Morogoro) 06° 56' 8.3'' S 37° 06' 3.6'' E " MASS41‒46. Lushoto 04° 07' 09'' S 38° 26' 56.6'' E " MASS29‒40. Tororo 00° 63' 6.8'' S 37° 19' 9.36'' E " MASS47‒53. Mbale 01° 06' 6.6'' S 34° 17' 7.9'' E " MASS54‒61. Kabale 01° 25' 12'' S 29° 53' 6.2'' E " MASS62‒69. Okahandja 21° 39' 35.5'' S 16° 52' 22'' E S. macowaniana MN1, MN15. Khorixas‒Outijo 20° 30' 43'' S 14° 57' 24.7''E " MN5, MN24, MN25, MN26, MN27, MN39. Swakop 22° 38' 23'' S 14° 44' 35'' E S. pachycarpa MN9, MN44, MN49, MN51. Rio Tinto Gorge 22° 27' 12.4'' S 15° 07' 24.8'' E " MN16, MN36, MN43, MN45, MN46, MN47, MN48, MN50, MN4, MN53, MN54, MN55, MN56, MN58, MN59, MN60, MN61, MN62. Epupa falls 16° 59' 39'' S 13° 17' 28'' E S. sesban MN7, MN8, MN13, MN18, MN19, MN21, MN22, MN23, MN30, MN35, MN38, MN57. Otjinungua 17° 49' 54'' S 12° 31' 20'' E " MN67, MN68. Suclabo 18° 07' 32.4'' S 21° 35' 51'' E S. cinerascens MN20, MN28, MN31, MN32, MN37, MN40. Omuramba 18° 05' 42'' S 20° 23' 36'' E S. bispinosa MN11, MN17, MN33, MN34. Bunya 17° 51' 29'' S 19° 21' 49'' E S. microphylla MN69, MN70. Rooidrom 17° 49' 53'' S 12° 31' 22'' E S. sphaerosperma MN2, MN10. Sesfontein 19° 02' 37'' S 13° 45' 09'' E " MN12, MN41, MN71. Korokoko 17° 59' 14'' S 20° 57' 06'' E S. rostrata MN50. Tanzania Uganda Namibia 30 3.3 Media for culturing of rhizobia Yeast Extract Mannitol broth (YEMB) or Yeast Extract Mannitol agar (YEMA) was used as an artificial media for culturing of the rhizobia recovered from root nodules. The recipe for YEMB included 10.0 g mannitol; 0.50 g di‒Potassium orthophosphate (K2HPO4); 0.20 g magnesium sulphate (MgSO4.7H2O); 0.10 g sodium chloride (NaCl); 0.50 g yeast extract and distilled water (1.0 litre) while YEMA contained YEMB and agar (16.0 g) as described in Somasegaran and Hoben (1994). The pH of YEMB was adjusted to 6.8 and sterilized or before addition of agar when YEMA was desired. The media was sterilized at a temperature of 121 °C and a pressure of approximately 15 pounds per square inch for 15 minutes using an autoclave. YEMA containing 0.00125 mg kg‒1 Congo red (Diphenyldiazo‒α‒naphthylaminesulfonate) dye (YEMA‒CR) was used as differential media while acid reaction of isolates was determined on YEMA containing 0.00125 mg kg‒1 bromothymol blue (BTB) as indicator. 3.4 Isolation, purification and preservation of root nodule bacterial isolates The dry nodules preserved over silica gel were immersed in tap water contained in a petri dish for two hours to imbibe. Thereafter, isolation procedure was carried out according to Somasegaran and Hoben (1994) with minor modifications at the nodule sterilization step where acid was omitted and nodules were immersed for 60 seconds in 3.5 % sodium hypochlorite instead. Each of the nodules in which the rhizobia was isolated was then immersed in 95 % (v/v) ethanol for 10 seconds to break the surface tension and was followed by an immersion in 2 % (v/v) sodium hypochlorite (NaOCl) for two minutes to decontaminate the surface. The nodules were successively rinsed in five changes of sterile distilled water before being transferred 31 into a drop (approximately 100 µL) of sterile distilled water contained in a petri dish using a pair of sterile forceps. The nodules were squashed using a sterile blunt glass rod to release bacteroids and or bacteria. A sterile tungsten inoculating loop was used to lift a loopful of the nodule suspension and was streaked on YEMA media. The streaking was performed in a dilution manner to yield isolated colonies on the final streak‒line. Inoculated plates were incubated in an inverted position in the dark at a temperature of 28±1 °C and monitored daily to observe colony emergence. Whenever more than one type of colony appeared growing on media from a single nodule, they were re‒streaked on fresh YEMA media and considered as separate rhizobia isolates. A culture from each nodule was assigned MASS or MN and a numeral for East African or Namibian nodules respectively. However, in the case of dual or multiple nodule occupants these labels were followed by a consecutive lower case letters respectively (for example, MASS133a and MASS133b). Cultures from single nodules but stored as mixed forms were asigned a similar labels followed by respective lower case letters (for example, MASS133ab). Pure isolates were preserved in 16 % (v/v) glycerol‒YEMB contained in well labeled autoclavable plastic vials at a temperature of ˗70 °C. 3.5 Phenotypic characteristics of sesbania rhizobia During culturing, rhizobia isolates were characterized by their morphological traits as cell (Gram stain), colony (absorption of Congo red, form, elevation, margin, appearance, optical property, pigmentation, texture), mucous (extracellular polysaccharides) production and pH reaction on YEMA as described by Mpepereki et al. (1997) and Odee et al. (1997). All these traits were recorded and used to group rhizobia into their respective morphotypes. 32 3.5.1 Determination of rhizobial growth rate and colony characteristics Yeast extract Mannitol Agar containing 25 mg L‒1 (w/v) Congo red dye (YEMA‒CR) was used as differential media for identification of typical rhizobia from other soil inhabiting bacteria (Somasegaran and Hoben, 1994). Morphological characteristics of the colonies were recorded as diameter, inability to absorb Congo red dye, form, elevation, margin, appearance, optical properties, pigmentation, texture and mucous (extracellular polysaccharides) production on YEMA. 3.5.2 YEMA‒BTB medium colour change by sesbania rhizobial isolates Rhizobial isolates were inoculated on YEMA media supplemented with 0.00125 mg kg‒1 bromothymol blue (BTB) as a pH reaction indicator. Three- to seven-day old cultures were observed for their ability to change colour of YEMA‒BTB medium at pH 6.8 (green) to yellow, blue or retain the green colour (Somasegaran and Hoben, 1994). The observed colour of YEMA‒BTB media after three -seven days was recorded. 3.5.3 Gram staining Gram staining was performed according to the method by Somasegaran and Hoben (1994). A loopful of actively growing rhizobia in liquid culture was transferred to the surface of a clean glass slide and spread over a small area. The culture film was air dried then fixed by passing the slide five times over a bunsen burner flame without exposing the dried film directly to the flame. The slide was flooded with crystal violet solution for one minute then washed off for 5 seconds using gently running tap water. The water was drained by by holding the slide in a vertical position before flooding the slide with Gram's iodine solution and allowed to act (as a mordant) for 33 one minute. The Gram's iodine solution was washed off using tap water. Excess water was drained from the slide by by holding the slide in a vertical position and blotted dry using a blotting paper. The slide was flooded with 95 % (v/v) alcohol for 10 seconds and washed off using tap water. The water was drained by holding the slide in a vertical position before the slide was flooded with safranin solution and allowed to counter stain for 30 seconds. The slide was washed in running tap water, drained and allowed to air‒dry before bacterial examination using a total magnification of x600 on Olympus microscope (model: 1X2‒ILL100 T5SN). The documentation system interphased to the microscope was then used to photograph the images. 3.6 Determination of genotype composition of sesbania rhizobia using PCR‒RFLP In this study, the number of sesbania isolates was reduced from 129 to 79 through selection of representative isolate(s). The selection was aided by dendrogram clusters obtained using UPGMA method based on combined similarity matrix of data from the diverse growth characteristics of rhizobia on YEMA media, intrinsic antibiotic resistance (IAR) and salt tolerance (NaCl). The 79 sesbania isolates were later subjected to fingerprinting assays using PCR‒RFLP of the 16S rDNA in comparison with 17 reference strains. 3.6.1 Intrinsic antibiotic resistance assay Resistance to antibiotics was evaluated by inoculating each rhizobial isolate on YEMA media supplemented with antibiotics. The isolates were tested for their 34 ability to resist two concentration levels [50 mg L‒1 and 100 mg L‒1 (w/v)] of each of the 12 different antibiotics. 3.6.1.1 Preparation of YEMA‒antibiotics media Yeast extract mannitol agar media supplemented with 12 antibiotics were prepared as described by Odee et al. (1997). Stock solutions of the following antibiotics were prepared separately: spectinomycin dihydrochloride (C14H24N2O7.2HCl), tetracycline (C22H24N2O8), penicillin‒G sodium salt (C16H17N2O4SNa), novobiocin sodium salt (C31H35N2O11Na), kasugamycin hydrochloride (C14H25N3O9.HCl), neomycin sulphate (C23H46N6O13.3H2SO4), streptomycin sulphate (C21H39N7O12)2.3H2SO4), erythromycin (C37H67NO13), rifampicin (C43H58N4O12), kanamycin mono sulphate (C18H36N4O11.H2SO4), ampicillin (C16H19N3O4S) and carbanicillin. All antibiotic stock solutions were prepared by dissolving 125 g of each antibiotic in 25 mL distilled water except erythromycin which was dissolved in 25 mL of 99 % ethanol due to its low solubility in water as compared to ethanol according to the method of Manna et al. (2004). Each of the stock solutions was then filter‒sterilized using 0.45 µM millipore (Whatman cellulose nitrate membrane filter). Meanwhile, YEMA was prepared in aliquots of 500 mL, sterilized and allowed to cool to approximately 50 °C before addition of the sterile antibiotics to make either 50 mg L‒1 or 100 mg L‒1 of each of the antibiotics. The media was swirled clockwise and anticlockwise to uniformly disperse the antibiotics then poured in sterile plastic petri dishes under aseptic conditions and allowed to set. 35 3.6.1.2 Rhizobial culturing for intrinsic antibiotics resistance assay A loopful of rhizobial culture growing on YEMA media was inoculated in a 10 mL YEMB and vigorously shaken to disperse the cells introduced in the media. The inoculated broth was placed on a horizontal orbital incubator shaker (Model: Tour‒120‒2) and aerated at 100 rpm, at a temperature of 28±1 °C for three and seven days for fast and slow growing types respectively. Zero point two mililitres (0.2 mL) of the broth culture was transferred to 10 mL of YEMB and allowed to grow for three and seven days for fast and slow growing rhizobia respectively. 3.6.1.3 Inoculation of rhizobial cultures on YEMA‒antibiotics media Rhizobia grown to turbid were aseptically transferred onto sterile wells. The well containing 0.5 mL of broth and a plate with YEMA containing antibiotics were put in the respective positions on a Denly multipoint inoculator. Multipoint inoculator pins were then lowered to pick the cultures and transfer to YEMA media supplemented with twelve antibiotics at two concentration levels of 50 mg L‒1 and 100 mg L‒1. Three replicates per level of each antibiotics were allowed for each tested isolate. Inoculated media was incubated in the dark at a temperature of 28 ±1 °C. After 4 days of incubation growth of rhizobia was recorded in an excel spreadsheet as growth present (1) or no growth (0). During scoring, growth of rhizobial cultures on YEMA‒antibiotics media was compared to those grown on control media (YEMA with no antibiotics). 36 3.6.2 Screening of rhizobial isolates for salt tolerance levels Rhizobial isolates were grown in YEMB media to turbid and inoculated as described in sections 3.6.1.2 and 3.6.1.3 on YEMA media containing 0.1 %, 1 %, 3 %, 5 % 6 %, 7 %, 8 % and 10 % NaCl (w/v) using a multi‒point inoculator then incubated at 28 ±1 °C. Four days‒old rhizobia inoculated plates were recorded as growth present (1) or no growth (0) in an excel spreadsheet. During scoring, growth of rhizobial cultures on YEMA + salt levels was compared to those grown on control media (YEMA + 0.1 % NaCl). 3.6.3 Selection of sesbania rhizobia for PCR‒RFLP assays Respective intrinsic antibiotic resistance and salt tolerance data for isolates from each site in Excel format were merged and exported to statistical software PAST (Hammer et al., 2005) for clustering of the isolates into their similarity matrix using the Unweighted Pair Group Method with Average (UPGMA). Dendrograms were constructed for isolates per site and used to select for rhizobia samples used in PCR‒RFLP fingerprinting practical. The clusters were then used to select characteristically unique rhizobial isolates per site. A maximum of three isolates were picked from each cluster of closely related isolates per site. Isolates from Namibia were merged to generate a single dendrogram. In cases of nodule co‒occupancy, all the isolates were picked for PCR‒RFLP assays even if they exhibited 100 % similarity in IAR and salt tolerance. Reference strains (Table 3.2) obtained from various culture collection centres were also included in the study. 37 Table 3.2: Reference strains used in PCR‒RFLP assays, their host plants and sources Strain Host plant Source KFR8 (Bradyrhizobium sp.) Acacia nubica KEFRI KFR84 (Mesorhizobium sp. type II) A. tortilis KEFRI KFR459 (Agrobacterium sp. type I) A. polyacantha KEFRI USDA9030 (Rhizobium leguminosarum) Phaseolus vulgaris USDA BA37 (R. leguminosarum) P. vulgaris MIRCEN DWO253 (R. leguminosarum) P. vulgaris KEFRI DWO461 (R. tropici type IIB) Sesbania sesban KEFRI KFR647 (M. huakuii ) S. sesban KEFRI KFR402 (Mesorhizobium sp. type I) S. sesban KEFRI Azorhizobium sp. S. rostrata Senegal USDA1002 (Ensifer meliloti) Medicago sativa USDA ORS177 (Bradyrhizobium sp.) Faidherbia albida Senegal KFR552 (B. elkanii) F. albida KEFRI USDA76 (B. elkanii) Glycine max USDA USDA110 (B. japonicum) G. max USDA DWO100 (Rhizobium sp.) Prosopis juliflora KEFRI USDA2370 (R. leguminosarum bv. Viciae) Pisum sativum USDA KFR, Kenya Forestry Research ‒ Rhizobial Culture Collection; MIRCEN, Microbiological Resources Centres; USDA, United states Department of Agriculture, National Rhizobium Culture Collection, Beltsville Agricultural Research Center, USDA, Beltsville, USA. 3.6.4 Rhizobial DNA extraction Bacterial DNA was extracted using the method by Terefework et al. (2001) with minor alteration. The fast and slow growing rhizobial isolates were grown in 10 mL YEMB contained in McCartney bottles for three and seven days respectively. One hundred microlitres (100 µL) rhizobial cultures grown in YEMB to 109 cells mL‒1 were aseptically seeded in fresh 10 mL YEMB. Two millilitres of 48 hours‒old broth culture was aseptically transferred into an eppendorf tube and centrifuged at 10 000 x g (25 °C) for 10 minutes to pellet the rhizobial cells. The supernatant containing YEMB was carefully discarded by decanting leaving behind a pellet of bacteria at the 38 bottom of eppendorf tube. This process was repeated two more times for the slow growing rhizobia. The pellet was washed two times using sterile distilled water by centrifuging at 10 000 x g (25 °C) for 5 minutes. Cell pellets were re‒suspended in 100 μL of TE buffer (10 mM Tris/HCl, 1 mM EDTA, pH 8.0) and vortexed for 5 seconds to mix. The suspension was subjected to boiling at 96 °C in a water bath for 10 minutes to lyse the bacteria and cooled for 30 minutes at temperature of 25 °C. Four hundred millilitres (400 mL) of 65 °C‒preheated CTAB (Hexadecyltrimethylammonium bromide) extraction buffer (Table 3.3) was added to the lysed bacteria, mixed gently and incubated in a water bath at 65 °C for 20 minutes before cooling for 30 minutes at 25 °C. Four hundred microlitres (400 µL) of chloroform: isoamyl alcohol (24:1, v/v) was added and mixed gently then incubated at a temperature of ˗20 °C for 20 minutes. The samples were then centrifuged at 14 000 x g at a temperature of 4 °C for 10 minutes and the supernatant was transferred into a fresh eppendorf tube. Table 3.3: Recipe for Hexadecyltrimethylammonium bromide extraction buffer Reagent mL CTAB 1.28 PVP 2.56 5 M NaCl 3.60 0.5 M EDTA 0.51 1 M Tris‒HCl, pH 7.6 1.28 Mercaptoethanol 0.064 H2O (distilled sterile) 3.6 Total 12.894 CTAB, Hexadecyltrimethylammonium bromide; PVP, polyvinylpyrrolidone. 39 Four hundred microlitres of ice‒cold 99.5 % ethanol was added and mixed well by inverting the tube 10 times. The DNA was precipitated by centrifugation at 14 000 x g at a temperature of 10 °C for 30 minutes. The supernatant was carefully discarded by aspirating the ethanol. The DNA pellet was washed in 400 µL of 70 % ice‒cold ethanol by inverting the tube gently and then centrifuged at 14 000 x g (10 °C) for 10 minutes. The supernatant was carefully discarded by aspirating and the eppendorf tube was inverted over a styrofoam and placed on the bench at room temperature to dry the DNA pellet for 14 hours. The DNA pellet contained in the eppendorf tube was re‒suspended in 100 µL of DNase/RNase free water and mixed by flicking the tube with a finger until DNA dissolved. The quality and quantity of the extracted DNA was measured using a spectrophotometer (Model: Biospec‒nano for life science, Shimadzu Biotech‒Japan) before each was diluted to 10 ng µL‒1 using DNase and RNase free water. The bacterial DNA was stored under ˗20 °C before being used for 16S rRNA gene amplification. 3.6.5 PCR mastermix preparation Bacterial DNA extracted using the method of Terefework et al. (2001) was used as a template to amplify the nearly full‒length 16S rRNA gene using fD1 (5'‒AGAGTTTGATCCTGGCTCAG‒3') and rD1 5'‒AAGGAGGTGATCCAGCC‒3') primers (Weisburg et al., 1991). A 24 µL reaction PCR mastermix enough for 100 samples was prepared each to include 12.88 µL DNase and RNase free water; 2.5 µL of 10x PCR buffer; 2.92 µL of 2.5 mM MgCl2; 1.0 µL of 10 mM dNTPs; 1.2 µL of 10 µM each of the universal oligonucleotide primers fD1 and rD1 which corresponds to Escherichia coli 16S rRNA gene position 8‒27 and 1524‒1540 respectively which amplified near 40 full‒length rhizobia 16S rDNA; 2.0 µL Q‒solution; 0.3 µL of 5 units µL‒1 Taq polymerase and mixed using a vortex for 5 seconds. Twenty four microlitres (24 µL) of mastermix aliquots were transferred to DNase‒ and RNase‒free thin‒walled PCR 96 plate before 1 µL of 10 ng µL‒1 rhizobial DNA was added using a Veriti® 96‒well thermal cycler (Applied biosystems). 3.6.6 16S rRNA‒PCR amplification conditions PCR amplification of the 16S rRNA gene region of rhizobia from 79 sesbania, P. vulgaris and 16 reference strains was carried out as per the protocol described by Tan et al. (1997) using the following reaction conditions: initial denaturation at 94 °C for 5 minutes, 30 cycles (denaturation at 94 °C for 30 seconds, primer annealing at 53 °C for 40 seconds, extension at 72 °C for 90 seconds) and a final extension at 72 °C for 7 minutes. 3.6.7 Gel electrophoresis of rhizobial 16S rRNA PCR products The size of the PCR amplicons were determined by staining the PCR products using SYBR® green dye then separated on 0.8 % agarose gel for 45 minutes at 90 volts using 0.5 x TBE buffer (0.484 g L‒1 Tris, 0.037 g L‒1 EDTA, pH 8). The PCR products were then viewed and documented using a gel documentation bioimaging system (Model: Gel Doc‒it 300, UVP Bioimaging Systems, Upland, CA). The size of 16S rRNA band of each sample was recorded. 3.6.8 Restriction of 16S rRNA gene PCR amplicons Restriction of the16S rRNA gene PCR amplicons was carried out in a mastermix that comprised of 2.4 µL DNase‒free and RNase‒free PCR water, 1.0 µL Cutsmart 41 buffer, 0.1 µL (100 ng 50 µL‒1) Bovine serum albumin (BSA) and 0.5 µL of each of the endonucleases HaeIII (5′‒GG/CC‒3′), HinfI (5′‒G/ANTC‒3 and MspI (5′‒C/CGG‒3′) separately, as per the manufacturer′s instructions (Invitrogen™ Life Technologies). Six microlitres of the 16S rRNA gene PCR products were added to each of the 4 µL mastermix aliqouts in DNase‒ and RNase‒free thin‒walled un‒skirted low profile 96 plates and incubated at 37 °C for 2 hours. 3.6.9 Gel electrophoresis of restriction fragments of rhizobial 16S rRNA Separation of restriction fragments by electrophoresis was performed as described by Laguerre et al. (1994). Restriction fragments were stained using SYBR® green dye and separated by horizontal electrophoresis in a 2.5 % agarose gel for 2.5 hours at 90 V using 0.5 x TBE buffer. Restriction fragment profiles were then viewed and documented using a gel documentation bioimaging system (Model: Gel Doc‒it 300, UVP Bioimaging Systems, Upland, CA). The restriction fragment patterns were used to generate binary data (1) or (0) and recorded in an Excel spreadsheet which was later separated per the four countries of origin and used to draw likelyhood dendrograms. 3.7 Isolates effectiveness test with Sesbania sesban and common bean plants using Leonard jar assembly 3.7.1 Plant materials Quality S. sesban seeds were obtained from a raised seed stand in Malava forest, Kakamega, Kenya. A healthy seeded S. sesban tree was selected for seed harvesting. Seeds from a single mother tree were preferred to minimize genetic variations as opposed to the general collections. Sesbania sesban was chosen for this work since it 42 was not possible to acquire seeds for other local and exotic Sesbania species. Common bean seeds (variety Rose coco LOT Number 14‒13‒7561; REF Number 13/14/03/2116) were obtained from Simlaw Seeds Company Limited (Kenya). 3.7.2 Experimental design The experiment was made in a randomized complete block design with four replicates each containing two plants. There were 115 treatments that consisted of one hundred and seven sesbania rhizobial isolates, six reference strain (DWO253 Phaseolus vulgaris, KFR269 Siratro, KFR209 Faidherbia albida, BA37 P. vulgaris, KFR647 S. sesban and KFR402 S. sesban), uninoculated plants supplemented with 0 ppm N and 70 ppm N (controls). 3.7.3 Seed pretreatment and pre‒germination Healthy S. sesban and common bean seeds with uniform seed size were selected before they were pre‒treated. Common bean‒seeds were cleaned in several changes of tap water to remove seed preservatives. The seeds were surface sterilized using 3 % sodium hypochlorite for ten minutes, rinsed in five changes of sterile distilled water, soaked in sterile distilled water for 12 hours to imbibe and break the dormancy before they were aseptically pre‒germinated on 0.5 % (w/v) water agar contained in petri dishes for three days at a temperature of 28 °C. 3.7.4 Preparation of Leonard jar assembly The assembling of the high density (autoclavable) Leonard jars was accomplished as described by Somasegaran and Hoben (1994). Leonard jar assembly was made up of a reservoir (bottom) containing plant nutrient solution and a substrate holder (top). 43 Medium size vermiculite, an inert substrate, was soaked for 48 hours and cleaned in several changes of tap water before its pH was adjusted to 6.8 using either 1 N sodium hydroxide or 1 N hydrochloric acid. The top part of the assembly was parked with vermiculite then placed over the reservoir which contained 500 mL nitrogen‒free plant nutrient solution. The entire assembly was inserted in a khaki paper bag No. 4 and covered using an aluminum foil. Assembled jars were sterilized using an autoclave at a temperature of 121 °C and pressure of 1.2 bars for 60 minutes. Sterile jars were allowed to cool overnight. 3.7.5 Preparation of nitrogen‒free plant nutrient solution Five millilitres of N‒free plant nutrient stock solutions (Broughton and Wilworth, 1971) containing 294.1 g L‒1 CaCl2.2H2O; 136.1 g L‒1 KH2PO4; 6.7 g L‒1 Fe‒citrate; 123.3 g L‒1 MgSO4.7H2O; 87.0 g L‒1 K2SO4; 0.338 g L‒1 MnSO4.H2O; 0.247 g L‒1 H3BO3; 0.288 g L‒1 ZnSO4.7H2O; 0.100 g L‒1 CuSO4.5H2O; 0.056 g L‒1 CoSO4.7H2O and 0.048 g L‒1 Na2MoO2.2H2O was each added to 5 litres of distilled water, stirred using a glass rod to mix then diluted to ten litres. The pH of the nutrient solution was adjusted to 6.8 using 1 N NaOH or 1 N HCl before it was added to Leonard jar reservoir. 3.7.6 Transfer of pre‒germinated seeds to vermiculite in Leonard jars The aluminum foil on the sterile Leonard jar was removed before three holes deep enough for inserting 2 cm‒long radicles were made into the substrate of each Leonard jar using a sterile pair of forceps. Three pre‒germinated seeds were then inserted gently into holes made in each jar and covered using vermiculite of the same 44 Leonard jar. The aluminum foils were placed back to cover the jars until emergence of the seedlings. 3.7.7 Inoculant preparation and inoculation of the host plant One hundred and seven representative sesbania rhizobial isolates of various morphological groups and origin of sesbania isolates and reference strains (DWO253, Phaseolus vulgaris, KFR269 Siratro, KFR209 Faidherbia albida, BA37 P. vulgaris, KFR647 S. sesban and KFR402 S. sesban, used for inoculants preparation for their respectively hosts at KEFRI) were grown in YEMB for 3‒7 days on a rotary shaker with a rotation of 100 rpm depending on their rates of growth. Slow growing rhizobia were inoculated in broth three days before the fast growers. Turbid cultures at late exponential phase, with approximately 10 9 rhizobial cells mL‒1 broth were used to inoculate the pre‒germinated S. sesban and common bean seedlings. One millilitre of the rhizobial culture was introduced at the root collar of each one of the three‒day old seedling growing in Leonard jar using a pipetman with a sterile pipette tip. Four replicates per treatment of sesbania rhizobial cultures and a similar set of uninoculated Minus N treatments (controls) replenished with N‒free plant nutrient and uninoculated Plus N treatments (controls) replenished with 70 ppm N applied as KNO3 solution were allowed. Vermiculite in all jars was covered with a layer of sterile ballast as a barrier against external contamination. Ten‒day old seedlings were thinned to two per jar using a sterile scalpel blade. The four replicates of each treatment were arranged into a completely randomized design on a clean bench in a glasshouse with controlled conditions (70 % shaded, 80 % humidity and at mean temperatures of 30/18 °C day/night). Plants were replenished 45 with 500 mL of respective sterilized nutrient solution every two weeks after pouring off the previous contents of the Leonard jar reservour until harvest. 3.7.8 Harvesting of Sesbania sesban and common beans Sesbania sesban and common beans were destructively harvested after 8 and 4 weeks of growth respectively. The substrate holding plant roots were pulled out of the Leonard jar then immersed in tap water contained in a spacious trough to wash off the vermiculite. Plants with washed roots were transferred to the laboratory where nodules were detached from the roots and roots separated from the shoots. Nodules were counted and recorded. Shoots, roots and nodules were then dried separately at a temperature of 70 °C for 72 hours. Shoot dry weight, nodule dry weight per plant and specific nodule dry weight were determined and recorded separately. 3.7.9 Determination of shoot total nitrogen Shoots of all nodulated common beans and selected S. sesban treatments were ground to fine powder using a hammer grinder (Model; Retsch Gmbh MM400, Germany) and their nitrogen content determined using Kjeldhal method as described by Anderson and Ingram (1993) and Okalebo et al. (2002). A 0.3 g plant sample was added to a digestion tube containing 4.4 mL of dissolved 0.42 g of selenium (Se) powder and 14 g of lithium sulphate in 350 mL of 30 % hydrogen peroxide and 420 mL. of concentrated sulphuric acid and digested in a block digester (Skalar Block Digester System, Model SA 5640). The digestion tubes were heated at a temperature of 360 ºC for 2 hours. Twenty five millitre (25 mL) of water was added to the cooled sample and topped to the 50 mL mark using distilled water. The sample was allowed to settle and a 10 mL clear solution was transferred to the distillation tubes, 10 mL of 46 40 % NaOH added to the sample then distilled. The extract was steamed immediately into 5 mL of 1 % boric acid and 4 drops of mixed indicator. Distillation was continued for 2 more minutes after the indicator turned green then followed by titration using 0.1 M HCl until the colour changed from green through grey to a definite pink. Nitrogen content in shoot sample was calculated using the following formula by Okalebo et al. (2002): N (%) = Titre x HCl Molarity x Extract Volume x 0.014 / aliquot volume x plant weight (where 0.014 is milliequivalent weight of nitrogen). 3.7.10 Determination of symbiotic efficiencies (%) Symbiotic efficiencies percent (SE %) of the sesbania rhizobial isolates on S. sesban and common beans was determined as described by Yaman and Cinsoy (1996). Nitrogen content was calculated as: (shoot dry weight x N concentration) per plant and the N fixed (plant N content in inoculated treatments ˗ plant N content in un‒inoculated treatments). Symbiotic effectiveness was estimated by comparing each inoculated plant with the plus 70 ppm N treatments (i.e. plant N content in inoculated plant/plant N content in 70 ppm N plant) x 100 (Beck et al., 1993). The sesbania isolates forming symbiosis with S. sesban were then arbitrarily rated as: very efficient (VE) when the shoot dry weight (SDWt.) of the nodulated host in the Leonard jar was higher than the total mean of all inoculation treatments added to the standard deviation (0.867 + 0.179); efficient (E) when its yield was between that of the mean + standard deviation (between 0.867 + 0.179) and mean ˗ standard deviation (0.867 ˗ 0.179) and inefficient (I) when its yield was smaller than the mean ˗ standard deviation (0.867 ˗ 0.179), as described by Lalande et al. (1990). This 47 rating was repeated using the total mean of shoot dry weight of nodulated common beans and the standard deviation (0.789 + 0.109 and 0.789 ˗ 0.109). 3.8 Data analysis The combined binary data from antibiotic resistance and salt tolerance recorded in an Excel spreadsheet was exported to statistical computer software (Hammer et al., 2005) for dendrogram construction using the UPGMA. The binary data obtained from PCR‒RFLP fragment profiles and recorded in an Excel spreadsheet was exported to the statistical programs GenAlex 6.2 (Peakall and Smouse, 2006; 2012) and statistical program MEGA version 4 (Tamura et al., 2007) for phylogenetic diversity analysis and dendrogram construction using the UPGMA. Isolates that had identical RFLP patterns were designated as one ribotype. Data for shoot dry weight, root dry weight, nodule number and nodule dry weight per plant were assessed for normality of distribution before they were subjected to a one-way ANOVA using GenStat computer software 16th Edition (VSN International, 2012). Pearson correlation coefficient was used to determine the relationship between plant growth parameters. Tukey’s HSD pairwise comparison at p ≤ 0.05 was used to separate the means. 48 CHAPTER FOUR RESULTS 4.1 Morphological characterization One hundred and twenty eight nodule forming bacteria (rhizobia) were isolated from root nodules of sesbania that were grown in diverse conditions in Kenya, Uganda, Tanzania and Namibia (Appendices I–IV). The isolates were either flat, raised or dome shaped with entire margins. They were grouped into nine unique morphotypes based on their growth characteristics on YEMA‒CR (Table 4.1). Morphotypes I‒V were fast growers (2‒3 days), morphotypes VI‒VII (4‒6 days) were moderate growers while morphotypes VIII and IX exhibited slow growth (7‒9 days) on growth media incubated at 28±1 °C. Morphotypes II, V and VI produced copious exopolysaccharides (EPS), morphotypes I, III, IV and VII secreted moderate EPS with gummy properties while morphotypes VIII and IX were non‒EPS producers (Table 4.1). 49 Table 4.1: Characteristics of sesbania rhizobial isolates on YEMA‒CR media Characteristics Morphotype 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides I 5 mm diameter, milky, translucent, shiny, dome and copious friable exopolysaccharides II 4 mm diameter, red, opaque, shiny, raised and moderate gummy exopolysaccharides III 2 mm diameter, brown centre, clear margin, raised, shiny, moderate gummy purple exopolysaccharides IV 5 mm diameter, pink suspensions, opaque, raised, dull and copious watery exoplysaccharides V 4 mm diameter, transparent, shiny, dome and copious viscous exopolysaccharides VI 2 mm diameter, milky opaque, raised, shinny and gummy moderate exopolysaccharides VII 1 mm diameter, milky opaque, dome, exopolysaccharides VIII shiny and no < 1.0 mm diameter, pink, translucent, flat, dull, dry and no exopolysaccharides IX Colony size of the rhizobia ranged from < 1.0 mm to 5.0 mm in diameter. Morphotype VI colonies were transparent. Morphotypes I, II and IX were translucent, morphotypes III, V, VII and VIII were opaque while morphotype IV had a brown centre with clear margins. With the exception of rhizobia clustered in morphotype III, all others were nonchromogenic during growth on YEMA media. All the isolates were shiny except those in morphotypes V and IX which were dull (Plate 4.1). 50 I II III IV V VI VII VIII IX 5mm Plate 4.1: Growth of sesbania rhizobial on YEMA‒CR media showing colony morphotypes I ‒ IX. During culturing of the sesbania rhizobia on YEMA media, some cultures developed hollow centres with uneven margins irrespective of the number of sub‒culturing onto fresh media (Plate 4.2). 51 H Plate 4.2: Rhizobial colonies on YEMA media showing H, hollow centre. Sesbania rhizobia assigned to the nine morphotypes had the following percent composition: morphotype I (23.7 %), II (5.3 %), III (5.3 %), IV (2.6 %), V (18.4 %), VI (14.5 %) VII (14.5 %), VIII (6.6 %) and IX (9.2 %) (Table 4.2). Table 4.2: Percent composition of sesbania rhizobial isolates in each of the nine morphotypes Morpho‒ Phenotype groups of isolates type I Percent isolates MASS37a, MASS51a, MASS59, MASS137a, MASS140, MASS177a, MASS170, MN2, MN10, MN18, MN26, MN27, MN35, MN39, MN50, MN51, MN56, MN62. 23.7 II MASS171, MASS57, MASS138, MN4. 5.3 III MASS31a, MASS130, MASS147, MN22. 5.3 IV MASS40a, MN44. 2.6 V MASS31b, MASS40b, MASS62, MASS114, MASS117a, MASS126, MASS127, MASS159, MASS174, MASS175, MN19, MN31, MN34, MN38. 18.4 VI MASS117b, MASS141, MASS172, MASS177b, MASS37b, MASS41, MASS49, MASS53, MN9, MN36, MN71. 14.5 VII MASS42, MASS43, MASS51b, MASS60, MASS65, MASS112, MASS132, MASS133, MASS137b, MASS176, MN12. 14.5 VIII MASS30, MASS33, MASS50, MASS117c, MN13. 6.6 IX MASS29, MASS36, MASS136, MASS160, MN11, MN17, MN26. 9.2 52 Sesbania rhizobial isolates were grouped into nine distinct morphotypes (I–IX) as per the Neighbour joining (NJ) dendrogram constructed based on colony morphological traits using UPGM A (Figure 4. 1). I II III IV V VI VII VIII IX Figure 4.1: Neighbour joining dendrogram constructed based on colony morphological traits using UPGMA showing relatedness of sesbania rhizobial isolates. 53 Rhizobial isolates recovered from Kenyan sites were distributed in all morphotypes except IV (Table 4.3a). Morphotype VIII contained a single isolate (MASS117c). Morphotype VI was universal in the four Kenyan sites. No rhizobia isolated from Kavutiri were grouped into morphotypes IV, V and VIII. The three Kuinet rhizobial isolates were distributed in morphotypes I, II and VI. Two isolates from Bumala site were grouped into morphotype V (MASS177b) and morphotypes VI (MASS176) while three isolates were grouped into morphotype IX. No isolates from Bumala conformed to characteristics of morphotypes II, III, IV and VIII. Gituamba isolates were grouped into morphotypes V, VI, VII and VIII with two isolates in morphotype V and one each in morphotypes VI, VII and VIII. No rhizobial isolates from Gituamba were grouped into morphotypes I, II, III, IV and IX (Table 4.3a). 54 Table 4.3a: Distribution of S. sesban rhizobial isolates from four Kenyan sites in nine morphotypes Morphotype group Kenya Kavutiri Kuinet Bumala Gituamba I MASS137a MASS140 MASS170 MASS177a ‒ II MASS138 MASS171 ‒ ‒ III MASS130 MASS147 ‒ ‒ ‒ IV ‒ ‒ ‒ ‒ V ‒ ‒ MASS126 MASS127 MASS159 MASS174 MASS175 MASS114 MASS117a VI MASS141 MASS172 MASS177b MASS117b VII MASS132 ‒ MASS176 MASS112 MASS133 MASS137b VIII ‒ ‒ ‒ MASS117c IX MASS136 ‒ MASS168 MASS178 ‒ MASS160 The only two rhizobial isolates from Kabale were grouped each in morphotypes V and VII (Table 4.3b). Tororo isolates were distributed in groups I, VI, VII and VIII. Mbale nodules had isolates that conformed to characteristics similar to those of morphotypes I, II and VII only. 55 Table 4.3b: Distribution of S. sesban rhizobial isolates from three Ugandan sites in nine morphotypes Morphotype group Uganda Kabale Tororo Mbale I ‒ MASS51a MASS59 II ‒ ‒ MASS57 III ‒ ‒ ‒ IV ‒ ‒ ‒ V MASS62 ‒ ‒ VI ‒ MASS49 ‒ MASS53 VII MASS65 MASS47 MASS60 MASS51b VIII ‒ MASS50 ‒ IX ‒ ‒ ‒ The three isolates from SUA in Tanzania were grouped into morphotypes VI (one isolate) and VIII (two isolates) (Table 4.3c). Lushoto isolates were spread in all morphotypes except in morphotypes II and VII. 56 Table 4.3c: Distribution of S. sesban rhizobial isolates from two Tanzanian sites in nine morphotypes Morphotype group Tanzania SUA Lushoto I ‒ MASS37a II ‒ ‒ III ‒ MASS31a IV ‒ MASS40a V ‒ MASS31b MASS40b VI MASS41 MASS37b VII MASS42 ‒ MASS43 VIII ‒ MASS30 MASS33 IX ‒ MASS29 MASS36 All the rhizobial isolates recovered from Khorixas Outijo and Rooidrom were grouped into morphotype I (Table 4.3d). Isolates from Swakop were distributed equally in morphotypes I, IV and VI. Rio Tinto Gorge isolates were clustered into morphotypes I (3 isolates) and one isolate in each of morphotypes II and VI. Isolates from Epupa falls were grouped into morphotypes I, III, V and VIII. Suclabo and Omuramba isolates were grouped into V and IX while those from Sesfontein were grouped in morphotypes VI and VII. 57 Table 4.3d: Distribution of sesbania rhizobial isolates from eight Namibian sites in nine morphotypes MN27 MN56 MN62 MN39 ‒ Sesfontein ‒ Rooidrom MN18 MN35 Omuramba Suclabo MN50 falls MN51 Epupa Rio Tinto Gorge MN26 Swakop Khorixas I Outijo Morphotype Namibia MN2 MN10 ‒ II ‒ ‒ MN4 ‒ ‒ ‒ ‒ ‒ III ‒ ‒ ‒ MN22 ‒ ‒ ‒ ‒ IV ‒ MN44 ‒ ‒ ‒ ‒ ‒ ‒ V ‒ ‒ ‒ MN19 MN31 MN34 ‒ ‒ MN38 VI ‒ MN9 MN36 ‒ ‒ ‒ ‒ MN71 VII ‒ ‒ ‒ ‒ ‒ ‒ ‒ MN12 VIII ‒ ‒ ‒ MN13 ‒ ‒ ‒ ‒ IX ‒ ‒ ‒ ‒ MN28 MN11 ‒ ‒ MN17 4.2 Characteristics of sesbania rhizobia on YEMA‒BTB media Actively growing sesbania rhizobia characteristically changed the green colour of YEMA‒BTB at pH 6.8 to blue or yellow colour. Some of the isolates did not change the green colour to either blue or yellow (Plate 4.3). 58 A B C D Plate 4.3: pH reaction characteristics of sesbania rhizobia on YEMA‒BTB media. A, control plate (with no rhizobia); B, yellow colour; C, green colour; D, Blue colour. Seventy eight sesbania rhizobial isolates were acid producers, 25 were neutral and 21 were alkaline producers (Table 4.4). Table 4.4: pH reaction of sesbania rhizobial isolates on YEMA‒BTB media pH reaction Isolate Acid MASS30, MASS31a, MASS32, MASS33, MASS34, MASS35, MASS37, MASS38, MASS39, MASS40, MASS41, MASS42, MASS43, MASS44, MASS45, MASS46, MASS47, MASS48, MASS49, MASS50, MASS51, MASS53, MASS54, MASS55, MASS57, MASS60, MASS62, MASS63, MASS65, MASS66, MASS67, MASS68, MASS69, MASS110, MASS111, MASS112, MASS113, MASS114, MASS115, MASS161, MASS162, MASS167, MASS168, MASS170, MASS172, MASS173, MASS176, MASS177, MN1, MN4, MN8, MN13, MN16, MN18, MN19, MN20, MN21, MN22, MN24, MN25, MN26, MN27, MN28, MN31, MN34, MN37, MN38, MN40, MN41, MN42, MN43, MN44, MN45, MN51, MN58, MN59, MN9, MN60. Neutral MASS116, MASS131, MASS132, MASS133, MASS134b, MASS135, MASS136, MASS137b, MASS156, MASS158, MASS159, MASS169, MASS171, MASS174, MASS181, MN2, MN10, MN12, MN15, MN39, MN49, MN56, MN57, MN62, MN68. Alkaline MASS29, MASS31b, MASS36, MASS61, MASS64, MASS117, MASS130, MASS157, MASS163, MASS164, MASS165, MASS166, MASS178, MASS180, MN11, MN17, MN35, MN50, MN69, MN70, MN71. 59 Out of the 124 isolates tested on YEMA-BTB media, the colour change from green to yellow accounted for 64.3 % followed by the green colour (19.4 %) and blue at 16.3 % (Figure 4.2). Figure 4.2: Colour change of YEMA-BTB by sesbania rhizobial isolates. 4.3 Gram staining All the sesbania isolates retained the pink color on counter staining using safranin, typical of legume root nodule forming bacteria. The stained isolates were rod‒shaped on observation under a microscope (Plate 4.4). 60 Plate 4.4: Gram negative rod‒shaped rhizobia (Magnification x 600). 4.4 Selection of rhizobia for PCR‒RFLP using intrinsic antibiotic resistance (IAR) and sodium chloride (NaCl) tolerance assays 4.4.1 Intrinsic antibiotic resistance of sesbania isolates A total of 128 rhizobia from root nodules of sesbania that were tested for their intrinsic antibiotic resistance on YEMA supplemented with 12 different synthetic antibiotics each at two concentration levels of 50 mg L‒1 and 100 mg L‒1 were variously sensitive to the type and concentration of antibiotics used. For instance, isolate MASS140 was totally suppressed by 100 mg L‒1 of kanamycin but produced copious EPS on 100 mg L‒1 penicillin compared to control plate. Also, MASS50a produced copious EPS on plates with 100 mg L‒1 penicillin compared to the controls (Plate 4.5). 61 A C B Plate 4.5: Sensitivity of sesbania isolates to antibiotics. A, kanamycin; 100 mg L‒1 ; B, penicillin 100 mg L‒1; C, control. MASS 125 MASS MASS MASS 136 141 MASS 137a MASS 51c 48 MASS MASS 51a 51b MASS 62 OP. MASS MASS 178 MASS 140 59 MASS MASS MASS 111 147 40 MASS MASS MASS 110 50a MASS MASS 65 57 60 Figure 4.3: Schematic diagram showing orientation of isolates on Plate 4.5. OP = orientation point). 62 Sixty one point three percent (61.3 %) of the sesbania rhizobia tested were resistant to all the antibiotics except spectinomycin and streptomycin at a concentration of 50 mg L‒1 (Figure 4.4). Spectinomycin and streptomycin at a concentration of 50 mg L‒1 allowed 26.8 % and 49.3 % respectively of sesbania isolates to grow. The isolates were most sensitive to concentrations of 100 mg L‒1 of rifampicin, spectinomycin, streptomycin, neomycin and kanamycin with resistance of 14.8 %, 12 %, 12 %, 10.6 % and 8.5 % respectively. Figure 4.4: Resistance of sesbania rhizobial isolates to antibiotics. Kas, kasugamycin; Car, carbanicillin; Neo, neomycin; Ery, erythromycin; Pen, penicillin‒G; Tet, tetracycline; Rif, rifampicin; Amp, ampicillin; Nov, novobiocin; Kan, kanamycin; Spe, spectinomycin; Str, streptomycin. Rhizobia from Bumala, Gituamba, Kuinet and Kavutiri (Kenya) had varying resistance to the 12 antibiotics at the two concentration levels 50 mg L‒1 and 100 mg L‒1 (Table 4.5a). The mean percent resistance ranged from 64.6 % to 81.1 % and 26.7 % to 44.2 % for antibiotics concentration level of 50 mg L‒1 and 100 mg L‒1 63 respectively. Bumala site isolates had a mean percent resistance of 79.5 % and 44.2 % to antibiotics at a concentration of 50 mg L‒1 and 100 mg L‒1 respectively. All isolates from Bumala were resistant to 50 mg L‒1 rifampicicn, kasugamycin, neomycin and carbanicillin. Rhizobia from Bumala had the least resistance (7.7 %) to spectinomycin and neomycin. Gituamba site isolates had a mean percent resistance of 64.6 % and 29.2 % on antibiotics at a concentration of 50 mg L‒1 and 100 mg L‒1 respectively. There was a total growth inhibition of isolates from Gituamba site by streptomycin and spectinomycin at a concentration of 50 mg L‒1 and kanamycin, streptomycin, spectinomycin, novobiocin and neomycin at a concentration of 100 mg L‒1. An equal number of isolates from Gituamba had similar resistance to ampicillin and penicillin (37.5 %) and erythromycin (87.5 %). Kuinet site isolates had a mean percent resistance of 66.7 % and 26.7 % to antibiotics at a concentration of 50 mg L‒1 and 100 mg L‒1 respectively. All the isolates from Kuinet grew on media supplemented with kanamycin, penicillin and novobiocin at a concentration of 50 mg L‒1. Kanamycin, streptomycin, spectinomycin and neomycin at a concentration of 100 mg L‒1 totally inhibited growth of Kuinet site isolates. Kavutiri site isolates had a mean percent resistance of 81.1 % and 39.4 % to antibiotics at a concentration of 50 mg L‒1 and 100 mg L‒1 respectively. All the isolates from Kavutiri were resistant to penicillin, kasugamycin, erythromycin, neomycin and carbanicillin at a concentration of 50 mg L‒1. Kanamycin at a concentration of 100 mg L‒1 had a total growth inhibition to Kavutiri isolates (Table 4.5a). 64 Table 4.5a: Percent intrinsic antibiotic resistance of S. sesban rhizobial isolates from Kenya Bumala 50 Gituamba Kuinet Kavutiri 100 50 100 50 100 50 100 Antib mg iotics L‒1 mg mg mg mg mg mg mg ‒1 L ‒1 L ‒1 L ‒1 L ‒1 L ‒1 L L‒1 Amp 84.6 61.5 37.5 37.5 40.0 40.0 80.0 53.3 Kan 84.6 30.8 87.5 0 100.0 0 66.7 0 Rif 100.0 53.8 100.0 37.5 80.0 20.0 66.7 20.0 Str 23.1 15.4 0 0 20.0 0 40.0 26.7 Pen 92.3 69.2 37.5 37.5 100.0 40.0 100.0 66.7 Tet 84.6 61.5 37.5 25 40.0 20.0 80.0 26.7 Kas 100.0 46.2 100.0 62.5 80.0 20.0 100.0 40.0 Spe 15.4 7.7 0 0 20.0 0 53.3 33.3 Nov 76.9 23.1 100.0 0 100.0 40.0 86.7 13.3 Ery 92.4 76.9 87.5 87.5 80.0 80.0 100.0 93.3 Neo 100.0 7.7 87.5 0 60.0 0 100.0 6.7 Car 100.0 76.9 100.0 62.5 80.0 60.0 100.0 93.3 Mean % 79.5 44.2 64.6 29.2 66.7 26.7 81.1 39.4 Rhizobial isolates from Tororo, Mbale and Kabale (Uganda) exhibited variation in their resistance to the 12 antibiotics used in this study (Table 4.5b). The mean percent resistance ranged from 66.7 % to 91.7 % and 26.7 % % to 45.8 % for antibiotics concentration level of 50 mg L‒1 and 100 mg L‒1 respectively. Isolates from Tororo had a mean percent resistance of 66.7 % at 50 mg L‒1 and 29.2 % at 100 mg L‒1. inhibition to Tororo isolates exhibited the lowest resistance of 20 % and 0 % to streptomycin at concentrations 50 mg L‒1 and 100 mg L‒1 had the highest growth respectively. All isolates from Tororo were resistant to carbanicillin at a concentration of 50 mg L‒1 while 90 % of the isolates were resistant at 100 mg L‒1. Mbale isolates had a mean percent resistance of 75 % and 26.7 % at 50 mg L‒1 and 65 100 mg L‒1 respectively. All isolates from Mbale were susceptible to kanamycin, rifampicin, streptomycin, spectinomycin, novobiocin and neomycin at a concentration of 100 mg L‒1. Mbale isolates were all resistant to 50 mg L‒1 and 100 mg L‒1 carbanicillin. Rhizobial isolates from Kabale had a mean percent resistance of 91.7 % and 45.8 % at 50 mg L‒1 and 100 mg L‒1 respectively. All isolates from Kabale site were susceptible to kanamycin, rifampicin, streptomycin, spectinomycin and neomycin at a concentration of 100 mg L‒1. All Kabale isolates were resistant to ampicillin, penicillin, erythromycin and carbanicillin at concentrations 50 mg L‒1 and 100 mg L‒1. Table 4.5b: Percent intrinsic antibiotic resistance of S. sesban rhizobial isolates from Uganda Tororo 50 Antibiotics Mbale 100 ‒1 mg L 50 ‒1 Kabale 100 ‒1 50 ‒1 100 ‒1 mg L mg L mg L mg L mg L‒1 Ampicillin 70.0 60.0 60.0 60.0 100.0 100.0 Kanamycin 60.0 0.0 40.0 0 50.0 0 Rifampicin 80.0 10.0 100.0 0 100.0 0 Streptomycin 20.0 0 40.0 0 50.0 0 Penicillin 80.0 70.0 100.0 60.0 100.0 100.0 Tetracycline 40.0 10.0 60.0 20.0 100.0 50.0 Kasugamycin 70.0 20.0 100.0 40.0 100.0 50.0 Spectinomycin 20.0 10.0 40.0 0 100.0 0 Novobiocin 80.0 10.0 60.0 0 100.0 50.0 Erythromycin 90.0 60.0 100.0 40.0 100.0 100.0 Neomycin 90.0 10.0 100.0 0 100.0 0 Carbanicillin 100.0 90.0 100.0 100.0 100.0 100.0 Mean % 66.7 29.2 75.0 26.7 91.7 45.8 66 Rhizobial isolates from Lushoto and SUA sites (Tanzania) varied in their intrinsic resistance to the 12 antibiotics (Table 4.5c). The mean percent resistance ranged from 73.6 % to 77.8 % and 30.6 % to 34.0 % for antibiotics concentration level of 50 mg L‒1 and 100 mg L‒1 respectively. Lushoto site isolates had a mean percent resistance of 73.6 % and 34 % to antibiotics at concentrations of 50 mg L‒1 and 100 mg L‒1 respectively. At a concentration of 100 mg L‒1 rifampicin inhibited the growth of all the isolates from Lushoto. The highest IAR was recorded with Lushoto isolates at 100 mg L‒1 erythromycin (91.7 %). A 58.3 % resistance was recorded for rhizobial isolates from Lushoto on YEMA media containing 50 mg L‒1 and 100 mg L‒1 concentrations of ampicillin. Similarly, an equal resistance of 91.7 % was recorded for rhizobial isolates recovered from Lushoto on YEMA media containing erythromycin at concentrations of 50 mg L‒1 and 100 mg L‒1. Rhizobia from SUA site had a mean percent resistance of 77.8 % and 30.6 % at concentration of 50 mg L‒1 and 100 mg L‒1 respectively. Kanamycin and streptomycin conferred inhibition to all SUA isolates at both concentrations. SUA isolates recorded 100 % resistance to all antibiotics at 50 mg L‒1 except ampicillin, kanamycin, streptomycin and kasugamycin. Kanamycin, rifampicin, streptomycin, tetracycline spectinomycin and neomycin at a concentration 100 mg L‒1 had total inhibition of SUA isolates. 67 Table 4.5c: Percent intrinsic antibiotic resistance of S. sesban rhizobial isolates from Tanzania Antibiotics Lushoto ‒1 50 mg L SUA ‒1 100 mg L ‒1 50 mg L 100 mg L‒1 Ampicillin 58.3 58.3 66.7 66.7 Kanamycin 83.3 8.3 0 0 Rifampicin 75.0 0 100.0 0 Streptomycin 33.3 16.7 0 0 Penicillin 83.3 66.7 100.0 66.7 Tetracycline 58.3 33.3 100.0 0 Kasugamycin 100.0 25.0 66.7 33.3 Spectinomycin 50.0 16.7 100.0 0 Novobiocin 75.0 16.7 100.0 66.7 Erythromycin 91.7 91.7 100.0 66.7 Neomycin 75.0 16.7 100.0 0 Carbanicillin 100.0 58.3 100.0 66.7 Mean % 73.6 34.0 77.8 30.6 The mean percent resistance ranged from 25 % to 92 % and 8 % to 75 % for antibiotics concentration level of 50 mg L‒1 and 100 mg L‒1 respectively (Table 4.5d). Fifty milligrams per litre and one hundred milligrams per litre (50 mg L‒1 and 100 mg L‒1) kanamycin and streptomycin totally inhibited the growth of sesbania isolates from Okahandja while 50 mg L‒1 and 100 mg L‒1 kanamycin inhibited growth of isolates from Khorixas‒Outijo. Fifty milligrams per litre and one hundred milligrams per litre (50 mg L‒1 and 100 mg L‒1) ampicillin, streptomycin, tetracycline, spectinomycin and novobiocin totally inhibited the growth of sesbania isolates from Rio Tinto Gorge. Likewise, 50 mg L‒1 and 100 mg L‒1 ampicillin, streptomycin, tetracycline and novobiocin inhibited the growth of isolates from Suclabo. Fifty milligrams per litre and one hundred milligrams per litre (50 mg L‒1 and 100 mg L‒1) kanamycin and streptomycin inhibited the growth of isolates from 68 Bunya while 50 mg L‒1 and 100 mg L‒1 novobiocin totally inhibited the growth of isolates from Rooidrom. Also 50 mg L‒1 and 100 mg L‒1 neomycin totally inhibited the growth of isolates from Sesfontein. Apart from the complete resistance to tetracycline, kasugamycin and spectinomycin at a concentration of 50 mg L‒1 and to kasugamycin at a concentration of 100 mg L‒1, all isolates from Korokoko were susceptible to the antibiotics used in the present study. Penicillin, novobiocin, erythromycin and carbanicillin at concentrations of 50 mg L‒1 and 100 mg L‒1 did not affect sesbania isolates from Okahandja (Table 4.5d). Erythromycin and carbanicillin did not affect growth of sesbania isolates from Rio Tinto Gorge. Sesbania isolates from Epupa falls were all resistant to kasugamycin while ampicillin, streptomycin and carbanicillin at concentrations 50 mg L‒1 and 100 mg L‒1did not affect growth isolates from Otjinungua. Growth of all the isolates from Rooidrom was not affected by concentrations 50 mg L‒1 and 100 mg L‒1of ampicillin, streptomycin, penicillin, tetracycline, kasugamycin, spectinomycin, erythromycin, neomycin and carbanicillin. Growth of all the sesbania isolates from Sesfontein and Korokoko were not affected by concentrations 50 mg L‒1 and 100 mg L‒1of kasugamycin. Isolates from Rooidrom had the highest mean percent survival at antibiotics concentration 100 mg L‒1 (75.0 %) while those from Korokoko had the lowest mean percent survival (8 %). 69 Table 4.5d: Percent antibiotic resistance of Sesbania spp. rhizobial isolates from Namibia Okahandja Khorixas Outijo Antib -iotics 50 100 50 mg L‒1 mg L‒1 Amp 50 Kan Swakop 50 mg L‒1 100 mg L‒1 0 25 0 0 Rif 100 Str Rio Tinto Gorge Epupa falls Otjinungua 50 100 50 100 50 mg L‒1 100 mg L‒1 mg L‒1 mg L‒1 mg L‒1 mg L‒1 25 57 6 0 0 50 0 0 43 0 100 0 50 25 0 57 14 100 0 0 50 25 29 14 Pen 100 100 25 25 71 Tet 100 0 25 25 Kas 100 50 100 Spe 100 0 Nov 100 Ery Suclabo Omura- Bunya Rooidrom mba 50 mg L‒1 100 mg L‒1 50 100 50 100 50 mg L‒1 100 mg L‒1 mg L‒1 mg L‒1 mg L‒1 mg L‒1 7 100 100 0 0 50 20 100 17 0 50 0 33 0 30 0 0 50 0 50 0 33 0 40 0 0 50 17 50 0 0 0 71 100 0 67 67 100 100 33 86 0 0 0 100 17 50 0 75 100 43 100 0 100 100 100 75 25 100 14 0 0 100 0 100 50 25 57 43 0 0 33 100 100 25 25 71 57 100 100 Neo 100 0 25 25 57 0 100 Car 100 100 25 25 71 71 Mean % 79 42 38 25 31 11 Sesfontein 50 mg L‒1 100 mg L‒1 50 100 0 0 0 100 20 0 0 70 0 0 0 100 50 0 17 50 50 50 0 50 100 100 55 25 Korokoko 50 100 mg L‒1 100 mg L‒1 mg L‒1 mg L‒1 100 7 3 0 0 100 0 67 33 0 0 0 100 0 67 0 0 0 0 0 100 100 33 0 0 0 40 50 50 100 100 67 67 0 0 60 30 100 0 100 100 100 0 100 0 67 80 60 100 50 100 100 100 100 100 100 67 0 70 10 50 0 100 100 100 0 100 0 0 0 0 40 20 100 0 0 0 67 33 0 0 100 50 33 33 70 60 100 50 100 100 67 0 0 0 33 50 0 33 0 50 10 100 0 100 100 0 0 0 0 50 17 100 100 33 33 80 50 100 50 100 100 67 0 0 0 67 30 58 17 60 28 71 29 75 21 92 75 72 33 25 8 70 4.4.2 Rhizobia tolerance to sodium chloride Sesbania rhizobial isolates growth on YEMA containing salt levels (NaCl w/v) of 1 %, 3 % and 5 % were similar to those on 0.1 % NaCl (control). Higher salt levels (6 % to 10 %) increasingly inhibited growth of the rhizobial isolates (Plate 4.6). A H B C D G F E Plate 4.6: Sensitivity of sebania rhizobial isolates to salt levels (NaCl w/v) compared to controls. A,1 %; B, 3 %; C, 5 %; D, 6 %; E, 7 %; F, 8 %; G, 10 % (NaCl levels); H, 0.1 % (control). Rhizobia obtained from all the study sites were tolerant to 0.1‒3 % NaCl (Figure 4.4), there after an increase in salt concentration became inversely proportional to the percent of tolerant sesbania rhizobia (3 % NaCl = 99.2 %, 5 % NaCl = 98.4 %, 6 % NaCl = 93 %, 7 % NaCl = 80.1 %, 8 % NaCl = 64.3 % and 10 % NaCl = 54.3 % of the isolates (Figure 4.5). 71 Figure 4.5: Tolerance of sesbania rhizobial isolates to different sodium chloride concentrations. Error bars with 5 %. 4.4.3 Selection of rhizobial isolates for PCR‒RFLP Combined binary patterns generated by growth (1) or no growth (0) of rhizobial isolates on YEMA media supplemented with various levels of antibiotics or NaCl and analysed using UPGMA method resulted in dendrograms per site with distinctive clusters (Figures 4.6a‒j). Thirteen Bumala isolates were split into clusters I and II. Cluster I had one isolate MASS176. Cluster II comprised of split into two groups of nine isolates (MASS165, MASS157, MASS126, MASS127, MASS167, MASS177a, MASS178, MASS125 and MASS168) in one group and three MASS160, MASS123 and MASS177b in the other group at node with 26 % similarity index. Two isolates MASS165 and MASS157 were split within cluster Cluster II at nodes with 6 % and 9 % respectively. Isolates MASS126, MASS127, MASS167, MASS177a and MASS178 had a similarity index of 89 %. Similarly, isolates MASS125 and MASS168 had a similarity index of 89 % (Figure 4.6a). 72 I II Figure 4.6a: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Bumala. Numbers on branches are bootstrap % from 1000 replicates. Rhizobial isolates from Kuinet formed two similarity clusters. Cluster I consisted of two perfectly similar isolates (MASS117a and MASS117c) at similarity index 100 %) but were 72 % similar to MASS117b. Cluster II contained MASS112, MASS114 and MASS115 (100 % similar) and MASS111 and MASS110 (100 % similar) but the two groups had a nodal similarity index of 97 % (Figure 4.6b). I II Figure 4.6b: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Kuinet. Numbers on branches are bootstrap % from 1000 replicates. 73 Gituamba isolates MASS174 and MASS170 shared the same similarity node at 76 % in cluster I. Isolates MASS172 and MASS173 were similar (100 %) but also shared a cluster with MASS171 at a node with similarity index of 63 % in cluster II (Figure 4.6c). I II Figure 4.6c: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Gituamba. Numbers on branches are bootstrap % from 1000 replicates. Rhizobial isolates from Kavutiri were grouped into cluster I and II (Figure 4.6d). Cluster I contained isolates split into two groups at similarity index 34 %. Rhizobial isolates MASS134, MASS130 and MASS129 were similar (100 %) and had a similarity index of 82 % with MASS141. The similarity index of isolate MASS140 to isolates MASS138, MASS147, MASS149 and MASS131 was at 42 %, isolate MASS131 to isolates MASS138, MASS147 and MASS149 was at 44 % and isolate MASS138 to isolates MASS147and MASS149 was 58 %. Rhizobial isolates in cluster II had a node at 61 % and they included MASS132, MASS133 (100 % similarity index) and MASS137a and MASS 137b (100 % similarity index). Isolate MASS136 was similar to isolates MASS132 and MASS133 by 52 % while 74 MASS142 was similar to isolates MASS137a and MASS 137b by 42 % (Figure 4.6d). I II Figure 4.6d: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Kavutiri. Numbers on branches are bootstrap % from 1000 replicates. Rhizobial isolates of nodules collected from Tororo segregated into two clusters (I and II). Cluster I had a node at 38 % relatedness which split the isolates into two groups. Isolates MASS50b and MASS47 with relatedness of 21 % formed one group while a node with 7 % relatedness formed another group of six isolates. Isolates MASS48 and MASS49 were 100 % similar but differed with MASS51b and MASS53 (29 % similarity) at a node with 14 % similarity. Isolates MASS50c and MASS51a were similar by 44 %. Custer II contained isolates MASS50a and MASS52 which had a relatedness index of 57 % (Figure 4.6e). 75 I II Figure 4.6e: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Tororo. Numbers on branches are bootstrap % from 1000 replicates. Five S. sesban rhizobial isolates from Mbale were segregated into two clusters (I and II). Cluster I contained isolates MASS60 and MASS54 which had a similarity of 100 %. Cluster II had a node with similarity of 58 % and split isolate MASS59 from MASS57 and MASS61. Rhizobial isolates MASS57 and MASS61 had a similarity of 100 % (Figure 4.6f). I II Figure 4.6f: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Mbale. Numbers on branches are bootstrap % from 1000 replicates. 76 Four S. sesban rhizobial isolates from Kabale site were segregated into clusters I and II. Cluster I contained isolates MASS69 and MASS62 with 100 % similarity. Cluster II consited of MASS65 and MASS68 with 100 % similarity (Figure 4.6g). I II Figure 4.6g: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Kabale. Numbers on branches are bootstrap % from 1000 replicates. Isolates from Lushoto were separated into two clusters I and II. Cluster I contained five isolates which were split into two groups of isolate MASS37a and four isolates MASS38, MASS29, MASS30 and MASS31a respectively at node with 75 % similarity. The group with four isolates was again split into two isolates each (MASS38; MASS29) and (MASS30; MASS31a) at node 38 % similarity with similarity of 83 and 100 % repectively. Cluster II was again split into two subclusters with a single isolate (MASS36) on one side and six isolates on ther other side at node 35 %. The six were split into four and two isolates respectively at node with similarity of 22 % (MASS33, MASS40b, MASS40a and MASS37b) and (MASS31b and MASS39). The clusters containing isolates MASS33 and MASS40b and isolates MASS40a and MASS37b were split at node with similarity of 22 %. Isolates MASS33 and MASS40b; MASS40a and MASS37b; MASS31b and MASS39 had similarities of 46 %, 33 % and 46 % respectively (Figure 4.6h). 77 I II Figure 4.6h: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from Lushoto. Numbers on branches are bootstrap % from 1000 replicates. The three S. sesban rhizobial isolates from SUA (Sokoine University of Agriculture) were grouped into clusters I and II (Figure 4.6i). Cluster I contained isolates MASS42 and MASS43 (100 % similar) while cluster II had one isolate (MASS41). 78 I II Figure 4.6i: Unrooted neighbour joining dendrogram constructed based on IAR and salt tolerance using UPGMA method showing relatedness clusters I and II of sesbania rhizobial isolates from SUA. Numbers on branches are bootstrap % from 1000 replicates. Sesbania rhizobial isolates from Namibia clustered into I‒V (Figure 4.6j). Cluster I included isolates MN2, MN10, MN31, MN12, MN18, MN39, MN45, MN11, MN62, MN57, MN13, MN15, MN19, MN40, MN70, MN71, MN1, MN4, MN8, MN28 and MN58. Isolates MN28 and MN58 with similarity index 45 % were separated from other isolates in cluster I by a node with 16 % similarity. Two other groups were separated by a node with 9 % similarity. A subcluster comprising isolates MN2, MN10, MN31, MN12, MN18, MN39, MN45, MN11, MN62 and MN57 had similarities varying between 4 and 100 %. The subcluster containing isolates MN13, MN15, MN19, MN40, MN70, MN71, MN1, MN4 and MN8 had similarities between 18 and 98 %. Cluster II contained isolates MN9, MN16, MN51, MN25, MN26, MN24, MN20, MN32, MN60, MN38, MN59, MN43, MN27, MN37, MN17, MN21 and MN36. The isolates in this cluster had similarities that varied from 2 to 100 % at different nodes. Isolates MN17 and MN21 were 100 % similar. Cluster III contained a single isolate MN69. Cluster IV comprised of three isolates separated by a node with 90 % similarity into MN44 and MN68 (62 % similarity) and MN47. Cluster V comprised of MN50 and MN56 with a node with 26 % similarity (Figure 4.6j). 79 I II III IV V Figure 4.6j: Unrooted neighbour joining dendrogram constructed based on combined IAR and salt tolerance using UPGMA method showing relatedness clusters I‒V of sesbania rhizobial isolates from Namibia. Numbers on branches are bootstrap % from 1000 replicates. A total of 79 test rhizobia from East Africa and Namibia, were selected for PCR‒RFLP assays. In addition, 17 reference strains were included in the assays, two of them a S. sesban and common bean inoculant strain each (Table 4.6). 80 Table 4.6: List of sesbania rhizobial isolates and reference strains used for PCR‒RFLP assay No. Isolate No. Isolate No. Isolate No. Isolate 1. MASS29 25. MASS117b 49. MASS172 73. MN39 2. MASS30 26. MASS117c 50. MASS174 74. MN44 3. MASS31a 27. MASS126 51. MASS114 75. MN50 4. MASS31b 28. MASS127 52. BA37 76. MN51 5. MASS36 29. MASS159 53. MN2 77. MN56 6. MASS37a 30. MASS160 54. MN4 78. MN62 7. MASS37b 31. MASS168 55. MASS53 79. MN71 8. MASS40a 32. MASS175 56. MN9 80. MASS33 9. MASS40b 33. MASS176 57. MN10 81. KFR459 10. MASS41 34. MASS177a 58. MN12 82. USDA9030 11. MASS42 35. MASS177b 59. MN13 83. KFR84 12. MASS43 36. MASS178 60. MN11 84. DWO253 13. MASS47 37. MASS130 61. MN17 85. DWO461 14. MASS49 38. MASS132 62. MN18 86. USDA1002 15. MASS51a 39. MASS133 63. MN19 87. ORS177 16. MASS51b 40. MASS136 64. MN34 88. KFR647 17. MASS50 41. MASS137a 65. MN22 89. KFR402 18. MASS57 42. MASS137b 66. MN26 90. KFR8 19. MASS59 43. MASS138 67. MN27 91. Azorh. sp. 20. MASS60 44. MASS140 68. MN28 92. KFR552 21. MASS62 45. MASS141 69. MN31 93. DWO100 22. MASS65 46. MASS147 70. MN35 94. USDA337 23. MASS112 47. MASS170 71. MN36 95. USDA76 24. MASS117a 48. MASS171 72. MN38 96. USDA110 Nos. 1‒51 and 53‒80, sesbania rhizobial isolates; No. 52, common bean inoculant strain; Nos. 81‒96, reference strains; Azorh. sp., Azorhizobium sp. 81 4.5 Genotype composition of sesbania rhizobia 4.5.1 Estimation of rhizobial 16S rRNA amplicons using gel electrophoresis The DNA samples extracts from the 79 rhizobial isolates and 17 reference strains had OD260/280 of 1.78‒2.04. The PCR amplicons of the 16S rRNA gene amplified by universal primers fD1 and rD1 yielded similar bands of approximately 1500 bp when quantified on 0.8 % agarose gel (Plate 4.7). 82 M 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 M 1500 bp 1000 bp M 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 M 1500 bp 1000 bp M 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 M 1500 bp 1000 bp M 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 M 1500 bp 1000 bp Plate 4.7: Agarose gel electrophoresis showing amplicon bands of 16S rRNA region. Lanes 1‒51 sesbania rhizobial isolates, Lane 1, MASS29; 2, MASS30; 3, MASS31a; 4, MASS31b; 5, MASS36; 6, MASS37a; 7, MASS37b; 8, MASS40a; 9, MASS40b; 10, MASS41; 11, MASS42; 12, MASS43; 13, MASS47; 14, MASS49; 15, MASS51a; 16, MASS51b; 17, MASS50; 18, MASS57; 19, MASS59; 20, MASS60; 21, MASS62; 22, MASS65; 23, MASS112; 24, MASS117a; 25, MASS117b; 26, MASS117c; 27, MASS126; 28, MASS127; 29, MASS159; 30, MASS160; 31, MASS168; 32, MASS175; 33, MASS176; 34, MASS177a; 35, MASS177b; 36, MASS178; 37, MASS130; 38, MASS132; 39, MASS133; 40, MASS136; 41, MASS137a; 42, MASS137b; 43, MASS138; 44, MASS140; 45, MASS141; 46, MASS147; 47, MASS170; 48, MASS171; 49, MASS172; 50, MASS174; 51, MASS114; Lane 52, common bean inoculant strain, BA37; Lanes 53‒80 sesbania rhizobial isolates; 53, MN2; 54, MN4; 55, MASS53; 56, MN9; 57, M10; 58, MN12; 59, MN13; 60, MN11; 61, MN17; 62, MN18; 63, MN19; 64, MN34; 65, MN22; 66, MN26; 67, MN27; 68, MN28; 69, MN31; 70, MN35; 71, MN36; 72, MN38; 73, MN39; 74, MN44; 75, MN50; 76, MN51; 77, MN56; 78, MN62 79, MN71; 80, MASS33; Lanes 81‒96, reference strains 81, KFR459; 82, USDA9030; 83, KFR84; 84, DWO253; 85, DWO461; 86, USDA1002; 87, ORS177; 88, KFR647; 89, KFR402; 90, KFR8; 91, Azorhizobium sp.; 92, KFR552; 93, DWO100; 94,USDA337; 95, USDA76; 96, USDA110; M, 100 bp (Invitrogen). 83 4.5.2 Separation of 16S rRNA restriction bands using gel electrophoresis Digestion of the 16S rRNA amplicons using endonucleases MspI, HinfI and HaeIII, resulted in specific and unique band patterns for DNA extracts from the 79 test isolates and 17 reference strains. Restriction using each of the three endonucleases yielded one to five polymorphic bands (Plates 4.8a, 4.8b and 4.8c). The 50 bp bands were not included in analysis since most were not clear. 1000 bp 500 bp 250 bp 100 bp 1000 bp 500 bp 250 bp 100 bp 1000 bp 500 bp 250 bp 100 bp 1000 bp 500 bp 250 bp 100 bp Plate 4.8a: Agarose electrophoresis patterns of 16S rRNA gene region of sesbania rhizobial isolates generated as a result of restriction digestion using endonuclease MspI. Samples in lanes 1‒96 are as in Plate 4.7. M, 50 bp marker (Invitrogen). 84 1000 bp 500 bp 250 bp 100 bp 1000 bp 500 bp 250 bp 100 bp 1000 bp 500 bp 250 bp 100 bp 1000 bp 500 bp 250 bp 100 bp Plate 4.8b: Agarose electrophoresis patterns of 16S rRNA gene region of sesbania rhizobial isolates generated as a result of restriction digestion using endonuclease HinfI. Samples in lanes 1‒96 are as in Plate 4.7. M, 50 bp marker (Invitrogen). 1000 bp 500 bp 250 bp 100 bp 1000 bp 500 bp 250 bp 100 bp 1000 bp 500 bp 250 bp 100 bp 1000 bp 500 bp 250 bp 100 bp Plate 4.8c: Agarose electrophoresis patterns of 16S rRNA gene region of sesbania rhizobial isolates generated as a result of restriction digestion using endonuclease HaeIII. Samples in Lanes 1‒96 are as in Plate 4.7. M, 50 bp marker (Invitrogen). 85 4.5.2 Phylogenetic clusters of sesbania rhizobia The combined data for presence (1) or absence (0) of fragments obtained from restriction of 96 PCR products of the 16S rRNA using endonucleases (MspI, HinfI and HaeIII) separated into country of origin and subjected to cluster analysis using UPGMA by GenAlex statistical program and MEGA version 4 statistical programs resulted in phylogenetic diversity dendrograms. The mean cophenetic correlation coefficient was 62.1 % after 1000 bootstrap replications were applied to the analysis to determine branch support in the consensus trees. Rhizobia isolated from nodules of S. sesban grown in Kenya were grouped into nine clusters which comprised of ribotypes I‒IX (Figure 4.7a). Ribotype I contained five isolates closely affiliated to Rhizobium spp. Type A. Isolate MASS132 had a similarity of 40 % with R. leguminosarum DWO253 but shared a branch support at 3 % similarity with other Rhizobium reference strains. A cluster of four isolates sharing the same node at 16 % split into MASS126 and MASS159 with similarity of 84 % and MASS117a and MASS174 with similarity of 34 %. No S. sesban rhizobial isolates from Kenya clustered with Bradyrhizobium spp. that formed ribotype II. Ribotype III was split into two sub-clusters. The first sub-cluster contained isolates that were unmatched to any of the reference strains and contained two pairs of isolates with similarity indeces of 44 % (MASS141 and MASS172) and 53 % (MASS177a and MASS178) repectively but separated by a common branch support of 23 % similarity. The second sub-cluster of ribotype III was similar to the first subcluster ribotype by a branch support with only 1 % similarity. Two isolates were clustered in the second sub-cluster and were similar to Sinorhizobium meliloti USDA1002 by 24 % (MASS170) and 6 % (MASS177b) (Figure 4.7a). 86 I II III IV V VI VII VIII IX Figure 4.7a: Unrooted UPGMA dendrogram showing clusters generated by combined restriction patterns of amplified 16S rRNA (HinfI+ HinfI + HaeIII) of S. sesban rhizobia from Kenya and reference strains. Ribotypes are indicated by Roman numbers I‒IX. Numbers on branches are bootstrap % from 1000 replicates. 87 Agrobacterium sp. shared a branch support at 6 % similarity index with the three isolates in the second sub-cluster. Two sub-clusters of isolates with a branch support of 1 % formed ribotype IV. The first sub-cluster contained isolates MASS117b, MASS168, MASS137b, MASS133 and MASS140 with varying similarities of between 1 % and 48 %. The second sub-cluster comprised of MASS138 which was similar to Azorhizobium sp. by 24 %. The unmatched ribotype V contained isolate MASS112 by a bootstrap of 44 % to MASS175 and similar to MASS136 by 12 %. Isolates MASS171, MASS147, MASS130 and MASS137a were all grouped in ribotype VI with varying similarities between 5 % and 56 % and were clustered together with Mesorhizobium spp. Ribotype VII contained MASS127 which was similar to Rhizobium spp. Type B with a similarity of 4 %. Ribotype VIII consisted of three isolates that were not matched to any of the reference strains. Isolate MASS117c was similar to MASS160 by a bootstrap of 81 % while MASS176 had a bootstrap of 31 %. Ribotype IX had a single isolate (MASS114) that was unmatched to any of the reference strains used in the present study (Figure 4.7a). The highest percentage of isolates (20.70 %) was in ribotype III and IV (Table 4.7a). No rhizobial isolate from the Kenyan sites clustered with Bradyrhizobium spp. Bumala isolates were clustered with all ribotypes except VI. Isolates MASS117a, MASS117b and MASS117c from a single nodule collected from Gituamba site were grouped into ribotypes I, IV and VIII respectively. Isolate MASS137a and MASS137b (Kavutiri) clustered into ribotypes VI and IV respectively. Dual isolates MASS177a and MASS177b were both grouped with III (Table 4.7a). 88 Table 4.7a: Sesbania sesban rhizobial isolates from Kenya clustered using UPGMA based on combined patterns of 16S rRNA PCR‒RFLP (MspI, HinfI and HaeIII) compared to reference strains Ribotype Reference Sesbania sesban rhizobial isolate % clusters ribotype I Rhizobium spp. Type A MASS126 (BK), MASS159 (BK), MASS117a (GK), MASS174 (BK), 17.20 MASS132 (KvK). II Bradyrhizobium spp. III Agrobacterium spp./ Sinorhizobium spp. MASS141 (KvK), MASS172 (KuK), MASS177a (BK), MASS178 (BK), 20.70 MASS170 (KuK), MASS177b (BK). IV Azorhizobium spp. MASS117b (GK), MASS168 (BK), MASS137b (KvK), MASS133 (KvK), 20.70 MASS140 (KvK). V Unmatched MASS112 (GK), MASS175 (BK), 10.30 MASS136 (KvK). VI Mesorhizobium spp. MASS130 (KvK), MASS137a (KvK), MASS147 (KvK), MASS171 (KuK). - 0.00 VII Rhizobium spp. Type B MASS127 (BK). VIII Unmatched MASS117c (GK), MASS160 (BK), MASS176 (BK). IX Unmatched MASS114 (GK). 13.80 3.50 10.30 3.50 Sites of nodule collection, KuK, Kuinet; KvK, Kavutiri; GK, Gituamba; BK, Bumala; GK, Gituamba. (All in Kenya). Rhizobial isolates from Uganda formed two clusters (I and II) (Figure 4.7b). Cluster I comprised of Bradyrhizobium spp., Rhizobium spp., Sinorhizobium spp., Azorhizobium sp. and isolates MASS60, MASS51a and MASS53. Isolate MASS60 was similar to R. leguminosarum DWO253 by 33 % similarity index and with S. meliloti USDA1002 by 6 %. Isolate MASS51a and MASS53 had a similarity of 39 % and shared the same branch support with R. tropici USDA9030 and R. tropici IIB DWO461 at a similarity of 6 %. 89 I II Figure 4.7b: Unrooted UPGMA dendrogram showing clusters generated by combined restriction patterns of amplified 16S rRNA (HinfI+ HinfI + HaeIII) of S. sesban rhizobial isolates from Uganda and reference strains. Ribotypes are indicated by Roman numbers I‒II. Numbers on branches are bootstrap % from 1000 replicates. 90 The isolates MASS47, MASS51b, MASS57, MASS59, MASS62 and MASS65 with similarity between 5 % and 99 % shared a branch support at 1 % with Mesorhizobium spp. Isolate MASS49 was similar by 35 % to Agrobacterium sp. Type I KFR459 which also shared a branch support with Mesorhizobium spp. at 2 % similarity index (Figure 4.7b). Ribotype I contained isolates MASS51a (Tororo), MASS53 (Tororo) and MASS60 (Mbale) which were closer to Rhizobium spp. Type A than Bradyrhizobium spp. (Table 4.7b). Isolates MASS47, MASS50, MASS51b and MASS49 from Tororo (Uganda), MASS57 and MASS59 from Mbale (Uganda), MASS62 and MASS65 from Kabale (Uganda) were in one cluster with Mesorhizobium spp. and Agrobacterium sp. and accounted for 72.70 % of all isolates. Table 4.7b: Sesbania sesban rhizobial isolates from Uganda clustered using UPGMA based on combined patterns of 16S rRNA PCR‒RFLP (MspI, HinfI and HaeIII) compared to reference strains Ribotype Reference strains rhizobial Sesbania sesban rhizobial isolate % clusters ribo‒ type I Bradyrhizobium spp./ Rhizobium spp. Type A MASS51a (TU), MASS53 (TU), 27.30 MASS60 (MU). II Mesorhizobium spp./ Agrobacterium spp. MASS47 (TU), MASS50 (TU), 72.70 MASS51b (TU), MASS57 (MU), MASS59 (MU), MASS62 (KU), MASS65 (KU), MASS49 (TU). Sites of nodule collection in parenthesis, TU, Tororo; KU, Kabale; MU, Mbale. (All in Uganda). Rhizobial isolates from nodules of S. sesban grown in Tanzania were grouped into four clusters of ribotypes I‒IV (Figure 4.7c). Cluster I comprised of Bradyrhizobium 91 spp. and was supported by a branch with similarity of 5 % with MASS42, Sinorhizobium meliloti USDA1002 and Agrobacterium Type I KFR459. I II III IV Figure 4.7c: Unrooted UPGMA dendrogram showing clusters generated by combined restriction patterns of amplified 16S rRNA (HinfI+ HinfI + HaeIII) of S. sesban rhizobial isolates from Tanzania and reference strains. Ribotypes are indicated by Roman numbers I‒IV. Numbers on branches are bootstrap % from 1000 replicates. 92 Isolate MASS42 was closely clustered with S. meliloti USDA1002 with a similarity of 19 % than with Agrobacterium Type I KFR459 with 5 % similarity. Cluster II was composed of MASS31b and MASS40b with a similarity of 49 % and shared a branch support of 4 % with Rhizobium spp. Isolates MASS30 and MASS31a (100 % similar) and were grouped in cluster III together with Mesorhizobium spp. Isolate MASS33 shared one node with Mesorhizobium spp. with a similarity of 45 %. Rhizobium spp. shared a branch support with Mesorhizobium spp. at a similarity of 10 % while isolates MASS37a and MASS37b had a similarity of 33 % and shared the same branch support at similarity of 14 % with MASS29 in cluster III. Mesorhizobium spp. and Rhizobium spp. shared a branch support with isolates MASS37a, MASS37b and MASS29 at 2 % similarity. Cluster IV comprised of isolates MASS36, MASS40a, MASS41 and MASS43 which shared a branch support at 5 % with Azorhizobium sp. Isolates MASS41 and MASS43 had a similarity of 93 % while MASS36 and MASS40a had a similarity of 65 %. The two set of isolates of cluster IV shared a similarity branch support at 32 % (Figures 4.6c). A S. sesban rhizobial isolate MASS42 collected from SUA (Tanzania) clustered in ribotype I while isolates MASS41 and MASS43 from the same site clustered in ribotype IV (Table 4.7c). Rhizobial isolates from Lushoto clustered in ribotype II and IV. Two pairs of nodule co‒occupants recovered from Lushoto segregated separately into different ribotypes i.e. MASS31b (II) and MASS31a (III), MASS40b (II) and MASS40a (IV). Nodule co‒occupants MASS37a and MASS37b were both clustered in ribotype III. Ribotype III had the highest percentage (46.10 %) of rhizobial isolates and comprised of isolates from Lushoto only. No isolate clustered with ribotype I. 93 Table 4.7c: Sesbania sesban rhizobial isolates from Tanzania clustered using UPGMA based on combined patterns of 16S rRNA PCR‒RFLP (MspI, HinfI and HaeIII) compared to reference strains Reference rhizobial strains Sesbania sesban rhizobial isolate % ribo‒ clusters type Bradyrhizobium spp. - 0.00 I Sinorhizobium spp./ Agrobacterium spp. MASS42 (SUAT). 7.70 II Rhizobium spp. Type A MASS31b (LT), MASS40b (LT). III Mesorhizobium spp./ Rhizobium spp. Type B MASS29 (LT), MASS30 (LT), 46.10 MASS31a (LT), MASS33 (LT), MASS37a (LT), MASS37b (LT). IV Azorhizobium spp. MASS36 (LT), MASS40a (LT), 30.80 MASS41 (SUAT), MASS43 (SUAT). Ribotype 15.40 Sites of nodule collection in parenthesis, LT, Lushoto; SUAT, Sokoine University of Agriculture. (All in Tanzania). Rhizobia from Namibia were clustered into ten groups (I-X) (Figure 4.7d). Cluster I comprised of MN17, MN9, MN28, MN12 and Bradyrhizobium spp. Isolate MN17 had 83 % similarity with Bradyrhizobium sp. ORS177 and shared a branch support of 26 % with MN9 and MN28 which had a similarity of 41 %. Isolate MN12 shared a 7 % branch support with the Bradyrhizobium spp. Isolate MN34 was similar to the Rhizobium spp. by a similarity of 38 % in cluster II. Cluster III with three isolates MN51, MN50 and MN71 were similar to Mesorhizobium sp. Type II KFR84 by 13 %, 27 % and 37 % respectively. Cluster IV comprised of isolate MN56 which was similar by 46 % to S. meliloti USDA1002. Cluster V contained isolates MN18, MN19, MN35, MN36 and MN39 which did not match with the reference strains and had similarity indices of between 1 % and 31 %. Cluster VI had one isolate MN38 which had a similarity of 71 % with R. tropici IIB DWO461. 94 I II III IV V VI VII VIII IX X Figure 4.7d: Unrooted UPGMA dendrogram showing clusters generated by combined restriction patterns of amplified 16S rRNA (HinfI+ HinfI + HaeIII) of sesbania rhizobial isolates from Namibia and reference strains. Ribotypes are indicated by Roman numbers I‒X. Numbers on branches are bootstrap % from 1000 replicates. 95 Isolates MN10, MN22 and MN2 in cluster VII had a similarity of 49 % and shared a branch support with Mesorhizobium spp. at a similarity of 12 % (Figure 4.6d). The isolates in cluster VIII comprised MN4, MN31, MN11 and MN13 which did not match with the reference strains and had similarity of between 1 % and 31 %. The two isolates in cluster IX were both similar to each other by 30 % and to Agrobacterium sp. Type I KFR459 by 30 %. Cluster X comprised of isolates MN26, MN27 and Azorhizobium sp. MN26 was similar to MN27 by 36 % and were both similar to Azorhizobium sp. by 6 %. Sesbania rhizobial isolates recovered from Namibia were not ribotype specific apart from isolates from Khorixas Outijo (MN26, MN27) which entirely formed ribotype X and MN39 in ribotype V (Table 4.7d). Isolates MN9, MN51 and MN44 from Swakop were distributed in ribotypes I, III and IX respectively. Five isolates MN50, MN56, MN36, MN4 and MN62 from Rio Tinto Gorge were distributed in ribotypes III, IV, V, VIII and IX respectively. Isolates from Epupa falls were distributed in ribotypes V (MN18, MN19 and MN35), VI (MN38), VII (MN22) and VIII (MN13). Isolates MN28 and MN31 from Suclabo were clustered in ribotypes I and VIII respectively. Isolates MN17 in ribotype I, MN34 in ribotype II and MN11 in ribotype VIII were all isolated from sesbania root nodules collected from Omuramba. Both isolates MN2 and MN10 from Rooidrom were grouped into ribotype VII. Isolates MN12 and MN71 from Sesfontein were distributed in ribotypes I and III respectively. Ribotype V had the highest percent (19.30 %) of isolates while ribotypes II, IV and VI had the lowest percent (3.80 %) of isolates (Table 4.7d). 96 Table 4.7d: Sesbania spp. rhizobial isolates from Namibia clustered using UPGMA based on combined patterns of 16S rRNA PCR‒RFLP (MspI, HinfI and HaeIII) compared to reference strains Ribo- Reference strains Sesbania isolate clusters type % ribo‒ type I Bradyrhizobium spp. MN9 (Nb), MN12 (Nh), MN17 (Nf), 15.40 MN28 (Ne). II Rhizobium spp. MN34 (Nf). 3.80 Type B III Mesorhizobium spp. Type A MN50 (Nc), MN51 (Nb), MN71 (Nh). 11.60 IV Sinorhizobium spp. MN56 (Nc). V Unmatched MN18 (Nd), MN19 (Nd), MN35 (Nd), 19.30 MN36 (Nc), MN39 (Na). VI Rhizobium spp. MN38 (Nd). 3.80 3.80 Type A VII Mesorhizobium spp. Type B MN2 (Ng), MN10 (Ng), MN22 (Nd). 11.60 VIII Unmatched MN4 (Nc), MN11 (Nf), MN13 (Nd), 15.40 MN31 (Ne). IX Agrobacterium spp. MN44 (Nb), MN62 (Nc). 7.70 X Azorhizobium spp. MN26 (Na), MN27 (Na). 7.70 Sites of nodule collection in parenthesis, Na, Khorixas Outijo; Nb, Swakop; Nc, Rio Tinto Gorge; Nd, Epupa falls; Ne, Suclabo; Nf, Omuramba; Ng, Rooidrom; Nh, Sesfontein. (All in Namibia). 4.6 Symbiotic efficiency test of sesbania rhizobia on Sesbania sesban plants using Leonard jars in the greenhouse Eighty three percent (83 %) of the 79 sesbania rhizobial isolates formed nodules on roots of S. sesban. Leaf colour of the eight-week old S. sesban plants varied from dark green to yellow. The most effective isolates caused dark green colour akin to the 97 Plus N treatment plants while the non-effective isolates caused yellow plants as those of Minus N treatment (Plate 4.9). A B C D Plate 4.9: Eight week-old S. sesban plants inoculated with sesbania rhizobial isolates (A, MN8; B, KFR647); C, Plus N treatment (70 ppm N) (control) and D, Minus N treatment (control). The nodules formed by all the isolates were morphologically similar but differed in symbiotic effectiveness. Cross-sections of nodules formed by effective isolates on roots of S. sesban were pink (Leghemoglobin present). Nodules formed by ineffective isolates were pale to white on their cross-sections (Leghemoglobin absent) (Plate 4.10). Nodules on roots of S. sesban were either oval or spherical with smooth surfaces. 98 A B 1 cm Plate 4.10: Transverse section of nodules formed by sesbania isolates on S. sesban. A, Leghemoglobin present; B, Leghemoglobin absent. Shoot dry weight of S. sesban caused by inoculation using sesbania rhizobia ranged between 0.06 g (isolate MN17) and 1.06 g (isolate MASS172) (Table 4.8). Apart from MN19 (S. macowaniana) which caused a mean shoot dry weight of 0.79 g, fifteen S. sesban isolates recorded higher shoot dry weight compared to 0.36 g Minus N treatments (Control) of between 0.66 g (MASS177b) and 1.06 g (MASS172). Nodule dry weight ranged between 1.0 mg (MASS177a) and 94.0 mg (MASS125) while nodule number per plant ranged between 1 (MASS50) and 36 (MASS31a) (Table 4.8). Rhizobia included in this experiment from legumes other than sesbania [strains BA37 and DWO253 (Phaseolus vulgaris), KFR269 (Siratro) and KFR209 (Faidherbia albida)] also elicited nodules on roots of S. sesban. Plus N (uninoculated) treatments had a mean dry weight of 0.98 g while non-inoculated Minus N (control) treatments had 0.36 g. Both control treatments did not nodulate. 99 Table 4.8: Effect of sesbania rhizobial isolates from Kenya, Uganda, Tanzania and Namibia on shoot dry weight, nodule dry weight and nodule number of S. sesban Isolate‒Origin SDWt. (g) NDWt. (mg) NNo. MASS170 Sesbania sesban 0.36±0.13 34.00±15.00 11.00±4.00 MASS171 S. sesban 0.75±0.06 64.00±14.00 24.00±4.00 MASS172 S. sesban 1.06±0.10 91.00±12.00 23.00±4.00 MASS173 S. sesban 0.44±0.05 53.00±10.00 21.00±4.00 MASS174 S. sesban 0.20±0.04 24.00±11.00 3.00±2.00 MASS129 S. sesban 0.89±0.17 84.00±20.00 23.00±4.00 MASS130 S. sesban 0.72±0.13 58.00±26.00 14.00±3.00 MASS133 S. sesban 0.21±0.06 31.00±19.00 3.00±1.00 MASS134 S. sesban 0.79±0.18 72.00±14.00 20.00±3.00 MASS136 S. sesban 0.09±0.03 16.00±7.00 20.00±7.00 0.12±0.04 14.00±4.00 6.00±2.00 MASS138 S. sesban 0.20±0.03 21.00±3.00 18.00±3.00 MASS141 S. sesban 0.61±0.10 62.00±15.00 15.00±2.00 MASS147S. sesban 0.82±0.22 87.00±11.00 29.00±5.00 MASS149 S. sesban 0.25±0.03 7.00±3.00 2.00±1.00 MASS110 S. sesban 0.57±0.18 78.00±12.00 21.00±5.00 MASS111 S. sesban 0.72±0.08 73.00±13.00 29.00±4.00 MASS114 S. sesban 0.37±0.07 33.00±8.00 11.00±2.00 MASS115 S. sesban 0.38±0.06 39.00±8.00 14.00±2.00 MASS117a S. sesban 0.16±0.08 19.00±1.00 5.00±4.00 MASS117b S. sesban 0.16±0.05 3.00±1.00 1.00±0.00 MASS117c S. sesban 0.40±0.08 51.00±13.00 18.00±3.00 MASS117abc S. sesban 0.34±0.11 26.00±15.50 11.00±6.00 MASS123 S. sesban 0.33±0.06 44.00±9.00 21.00±3.00 MASS125 S. sesban 0.60±0.14 94.00±16.0 19.00±5.00 MASS126 S. sesban 0.14±0.02 9.00±3.00 2.00±1.00 MASS160 S. sesban 0.31±0.04 35.00±8.00 11.00±3.00 MASS168 S. sesban 0.15±0.06 23.00±16.00 10.00±5.00 MASS177a S. sesban 0.10±0.02 1.00±0.20 5.00±2.00 Kuinet (Kenya) Kavutiri (Kenya) MASS137b S. sesban . Gituamba (Kenya) Bumala (Kenya) 100 MASS177b S. sesban 0.66±0.12 67.00±22.00 21.00±5.00 MASS178 S. sesban 0.24±0.03 20.00±16.00 6.00±3.00 MASS29 S. sesban 0.11±0.02 14.00±3.00 10.00±2.00 MASS30 S. sesban 0.32±0.05 27.00±6.00 11.00±3.00 MASS31a S. sesban 0.47±0.08 59.00±13.00 36.00±5.00 MASS31b S. sesban 0.09±0.02 1.00±0.10 1.00±1.00 MASS33 S. sesban 0.09±0.03 2.00±0.70 1.00±1.00 MASS36 S. sesban 0.10±0.02 12.00±3.00 6.00±1.00 MASS37a S. sesban 0.07±0.02 8.00±2.00 8.00±2.00 MASS37b S. sesban 0.30±0.08 30.00±11.00 15.00±3.00 MASS38 S. sesban 0.22±0.05 17.00±5.00 12.00±6.00 MASS39 S. sesban 0.68±0.03 64.00±8.00 24.00±5.00 MASS40b S. sesban 0.35±0.10 50.00±15.00 13.00±3.00 MASS42 S. sesban 0.13±0.02 6.00±3.00 2.00±1.00 MASS41 S. sesban 0.38±0.05 36.00±5.00 13.00±1.00 MASS43 S. sesban 0.28±0.05 29.00±5.00 15.00±2.00 MASS47 S. sesban 0.33±0.05 39.00±10.00 17.00±4.00 MASS48 S. sesban 0.16±0.03 4.00±2.00 3.00±1.00 MASS49 S. sesban 0.22±0.04 19.00±9.00 4.00±2.00 MASS50ab S. sesban 0.08±0.03 1.00±0.30 1.00±1.00 MASS50b S. sesban 0.45±0.07 47.00±11.00 13.00±4.00 MASS51a S. sesban 0.15±0.03 10.00±5.00 4.00±2.00 MASS51abc S. sesban 0.14±0.02 4.00±1.00 4.00±1.00 MASS51c S. sesban 0.18±0.02 9.00±6.00 2.00±1.00 MASS53 S. sesban 0.33±0.06 42.00±8.00 19.00±4.00 MASS62 S. sesban 0.42±0.07 63.00±12.00 20.00±3.00 MASS65 S. sesban 0.17±0.03 12.00±6.00 3.00±2.00 MASS69 S. sesban 0.46±0.18 50.00±7.00 19.00±3.00 MASS54 S. sesban 0.35±0.05 33.00±5.00 15.00±4.00 MASS57 S. sesban 0.24±0.06 43.00±21.00 6.00±4.00 MASS59 S. sesban 0.83±0.18 71.00±17.00 27.00±5.00 Lushoto (Tanzania) SUA (Tanzania) Tororo (Uganda) Kabale (Uganda) Mbale (Uganda) 101 MASS60 S. sesban 0.16±0.04 15.00±5.00 3.00±2.00 MASS61 S. sesban 0.74±0.12 61.00±8.00 19.00±3.00 MN1Sesbania macowaniana 0.14±0.04 20.00±8.00 8.00±2.00 MN2 S. sphaerosperma 0.13±0.04 12.00±6.00 4.00±2.00 MN8 S. sesban 0.78±0.12 73.00±9.00 30.00±4.00 MN10 S. sphaerosperma 0.22±0.05 48.00±35.00 5.00±3.00 MN11 S. sphaerosperma 0.21±0.06 25.00±10.0 26.00±3.00 MN12 S. sphaerosperma 0.27±0.04 18.00±7.00 10.00±4.00 MN13 S. sesban 0.06±0.02 16.00±0.20 1.00±1.00 MN17 S. bispinosa 0.06±0.01 6.00±1.00 10.00±1.00 MN18 S. sesban 0.84±0.19 75.00±18.00 25.00±5.00 MN19 S. macowaniana 0.79±0.12 65.00±7.00 21.00±2.00 MN21 S. sesban 0.16±0.04 18.00±9.00 4.00±3.00 MN22 S. macowaniana 0.18±0.03 12.00±9.00 6.00±3.00 MN28 S. cinerascens 0.06±0.01 8.00±1.00 14.00±2.00 MN31 S. cinerascens 0.17±0.02 24.00±5.00 23.00±6.00 MN39 S. macowaniana 0.14±0.03 2.00±1.00 2.00±1.00 MN40 S. cinerascens 0.37±0.10 34.00±10.00 9.00±3.00 MN44 S. pachycarpa 0.32±0.06 43.00±19.00 7.00±3.00 MN45 S. pachycarpa 0.15±0.02 14.00±4.00 7.00±2.00 MN49 S. pachycarpa 0.28±0.06 29.00±5.00 11.00±3.00 MN50 S. pachycarpa 0.18±0.02 18.00±3.00 20.00±4.00 MN51 S. pachycarpa 0.23±0.01 28.00±2.00 24.00±3.00 MN57 S. sesban 0.59±0.13 43.00±11.00 14.00±3.00 MN58 S. pachycarpa 0.32±0.41 20.00±15.00 4.00±1.00 MN62 S. pachycarpa 0.13±0.02 9.00±8.00 2.00±2.00 MN68 S. sesban 0.93±0.14 54.00±10.00 32.00±6.00 MN69 S. microphylla 0.19±0.03 22.00±5.00 19.00±3.00 MN70 S. microphylla 0.13±0.01 19.00±1.00 31.00±3.00 MN71 S. microphylla 0.15±0.04 10.00±6.00 3.00±1.00 DWO253 Phaseolus vulgaris 0.21±0.02 18.00±4.00 11.00±5.00 KFR269 Siratro 0.59±0.03 72.00±17.00 25.00±5.00 Namibia Reference strains Kibwezi (Kenya) Loruk (Kenya) 102 KFR209 Faidherbia albida 0.60±0.17 52.00±12.00 16.00±3.00 BA37 P. vulgaris 0.16±0.07 15.00±3.00 28.00±3.00 KFR647 S. sesban 0.60±0.19 60.00±11.00 22.00±3.00 KFR402 S. sesban 0.70±0.04 85.00±12.00 20.00±3.00 0 ppm N 0.36±0.17 - 0 70 ppm N 0.98±0.21 - 0 Nyamonye (Kenya) Controls SDWt., shoot dry weight; NDWt., nodule dry weight; NNo., nodule number; Sokoine University of Agriculture Tanzania. Values represent the mean ± SD of eight replicates. Ten of the 13 isolates from Kavutiri elicited nodules on roots of S. sesban but only five had effective nodules (nod+nif+ phenotypes) (Table 4.9a). All isolates from Kuinet nodulated S. sesban, but only two exhibited nod+nif+ phenotypes. Eight of the 12 isolates from Bumala had nod+ but only two had nod+nif+ phenotypes. All isolates from Gituamba nodulated S. sesban, two of the isolates (MASS110 and MASS111) were effective while one of the isolates (MASS117c) was partially effective. A pair of nodule co‒occupants were isolated each from S. sesban growing in Kavutiri and Bumala while triple isolates co‒occupied a nodule collected from Gituamba. The co‒occupant isolates included MASS137a and MASS137b (Kavutiri), MASS177a and MASS177b (Bumala) and MASS117a, MASS117b and MASS117c (Gituamba). Co‒occupant isolates MASS137a and MASS137b of nodules collected from Kavutiri site were non‒infective (nod˗) and infective/ineffective (nod+nif˗) phenotypes respectively. The Bumala isolate MASS177a was nod+nif˗ while its nodule co‒occupant MASS177b was both infective and effective (nod+nif+) phenotype. The three Gituamba nodule co‒occupant isolates were of phenotypes nod+nif˗ (MASS117a and MASS117b) and partially effective (nod+nif+/˗) phenotype (MASS117c). 103 Table 4.9a: Nodulation and nitrogen fixation phenotypes of Sesbania sesban rhizobial isolates from Kenya on S. sesban Site nod˗ nod+ nif˗ nod+nif+/˗ nod+ nif+ Kavutiri MASS131 MASS137a MASS133 MASS136 MASS137b MASS138 ‒ MASS129 MASS130 MASS134 MASS141 MASS137ab MASS140 Kuinet Bumala Gituamba ‒ MASS127 MASS157 MASS167 MASS176 MASS177ab ‒ MASS149 MASS147 MASS170 MASS173 MASS174 ‒ MASS171 MASS172 MASS123 MASS126 MASS160 ‒ MASS125 MASS177b MASS117c MASS110 MASS111 MASS168 MASS177a MASS178 MASS114 MASS115 MASS117a MASS117b MASS117abc nod˗, non‒infective; nod+, infective; nif˗, ineffective; nif+/˗, partially effective; nif+, effective. Two isolates from Kabale (Uganda) were partially effective on S. sesban; one was ineffective while one failed to nodulate S. sesban (Table 4.9b). Similarly, isolates MASS50b and MASS53 from Tororo were partially effective; five isolates (MASS47, MASS48, MASS49, MASS51a and MASS51c) were ineffective while three isolates (MASS50a, MASS51b and MASS52) were non‒infective on S. sesban. Mixed isolates from single nodules (MASS50ab and MASS51abc) were non‒effective on their host of origin. Two isolates (MASS59 and MASS61) recovered from nodules from Mbale were effective on S. sesban while MASS54 isolate was partially effective. 104 Table 4.9b: Nodulation and nitrogen fixation phenotypes of Sesbania sesban rhizobial isolates from Uganda on S. sesban Site nod˗ nod+nif˗ nod+nif+/˗ nod+nif+ Kabale MASS68 MASS65 MASS62 ‒ MASS69 Tororo MASS50a MASS51b MASS47 MASS52 MASS49 MASS48 MASS50b MASS53 ‒ MASS54 MASS59 MASS50ab MASS51a MASS51c MASS51abc Mbale ‒ MASS57 MASS60 MASS61 nod˗, non‒infective; nod+, infective; nif˗, ineffective; nif+/˗, partially effective; nif+, effective. Three isolates from SUA formed nodules on the host of origin but only MASS41 was partially effective (Table 4.9c). Eleven of the twelve isolates from Lushoto formed nodules on S. sesban. Isolate MASS39 was effective while MASS31a and MASS40b were partially effective. Mixed isolates from single nodules (MASS37ab and MASS40ab) were ineffective on S. sesban. 105 Table 4.9c: Nodulation and nitrogen fixation phenotypes of Sesbania sesban rhizobial isolates from Tanzania on S. sesban Site nod˗ nod+nif˗ nod+nif+/˗ nod+nif+ SUA ‒ MASS42 MASS41 ‒ MASS39 MASS43 Lushoto MASS37ab MASS29 MASS31a MASS40a MASS30 MASS40b MASS40ab MASS31b MASS33 MASS36 MASS37a MASS37b MASS38 nod˗, non‒infective; nod+, infective; nif˗, ineffective; nif+/˗, partially effective; nif+, effective. Five isolates from nodule collection sites in Namibia did not form nodules on S. sesban (Table 4.9d). Twenty one isolates formed ineffective nodules. Two isolates (MN44 and MN49) from Swakop were partially effective on S. sesban. Four isolates (MN8, MN18, MN19 and MN57) from Epupa falls and a single isolate (MN68) from Otjinungua were effective on S. sesban. 106 Table 4.9d: Nodulation and nitrogen fixation phenotypes of sesbania rhizobial isolates from Namibia on S. sesban Site Khorixas Outijo Swakop nod˗ ‒ MN9 nod+nif˗ nod+nif+/˗ nod+nif+ MN39 ‒ ‒ MN51 MN44 ‒ MN49 Rio Tinto Gorge MN54 MN45 MN56 MN58 ‒ ‒ MN62 Epupa falls ‒ MN13 ‒ MN8 MN21 MN18 MN22 MN19 MN57 Suclabo MN37 MN28 ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ MN31 MN40 Omuramba ‒ MN11 MN17 Rooidrom ‒ MN2 MN10 Sesfontein ‒ MN12 MN71 Okahandja Otjinungua Bunya MN15 MN1 ‒ ‒ ‒ ‒ MN69 ‒ ‒ ‒ ‒ ‒ MN68 MN70 Korokoko ‒ MN50 nod˗, non‒infective; nod+, infective; nif˗, ineffective; nif+/˗, partially effective; nif+, effective. All the reference strains (KFR209, DWO253, KFR269, KFR647 and KFR402) used in S. sesban nodulation test were effective except the common bean inoculant strain, BA37 which was infective but ineffective (Table 4.9e). 107 Table 4.9e: Nodulation and nitrogen fixation phenotypes of reference strains on S. sesban Site nod˗ nod+nif˗ nod+nif+/˗ nod+nif+ Gituamba ‒ BA37 ‒ ‒ (P. vulgaris) Loruk ‒ ‒ ‒ KFR209 (F. albida) Kibwezi ‒ ‒ ‒ DWO253 (P. vulgaris), KFR269 (Siratro) ‒ ‒ ‒ KFR647 (S. sesban) Nyamonye ‒ ‒ ‒ KFR402 (S. sesban) Yala swamp nod˗, non‒infective; nod+, infective; nif˗, ineffective; nif+/˗, partially effective; nif+, effective. There were significant differences (p < 0.001) in mean percent nitrogen and nitrogen content per shoot of S. sesban inoculated using selected rhizobia from various legume hosts (Table 4.10). The lowest percent nitrogen was 1.35 (isolate MASS134) while the highest was 7.9 (isolate MN19). Common bean inoculant strain (BA37) caused the lowest nitrogen content per shoot (0.16 mg shoot-1) while isolate MASS129 caused the highest (5.66 mg shoot-1). Nitrogen fixation ratio varied from ˗0.28 (BA37) to 5.23 (isolate MASS129). The symbiotic effectiveness of selected rhizobia on S. sesban also varied with the rhizobial isolate used. Three isolates were ineffective, twenty effective and three highly effective. 108 Table 4.10: Effect of some sesbania rhizobial isolates on percent nitrogen, N content, N fixation ratio and the symbiotic effectiveness of S. sesban Isolate/Origin % N shoot ‒1 N content (mg) shoot‒1 N fixation SE ratio Kuinet (Kenya) MASS171 S. sesban 4.08±0.18fg 4.15±0.49de 3.71 E MASS172 S. sesban 1.87±0.23ab 1.99±0.31b-e 1.56 HE MASS129 S. sesban 5.65±0.67h 5.66±0.64e 5.23 HE MASS130 S. sesban 5.56±0.11h 2.78±0.29a-e 4.54 E MASS134 S. sesban 1.35±0.10a 1.03±0.56abc 0.14 E MASS141 S. sesban 4.09±0.13fg 2.43±0.32abc 1.99 E MASS147S. sesban 4.00±0.42fg 3.10±0.49a-e 2.66 E MASS110 S. sesban 2.12±0.20abc 1.11±0.24a-d 0.68 E MASS111 S. sesban 2.12±0.15abc 1.52±0.17ab 0.64 E MASS125 S. sesban 5.41±0.23h 2.00±0.41b-e 1.57 E MASS177b S. sesban 3.22±0.12c‒f 2.03±0.07a-e 1.59 E 1.98±0.18ab 1.35±0.10ab 0.46 E MASS62 S. sesban 3.39±0.03ef 1.13±0.26ab 0.70 I MASS69 S. sesban 2.36±0.09a‒e 1.52±0.19abc 1.08 I MASS59 S. sesban 2.25±0.13a‒d 1.86±0.36abc 1.62 E MASS61 S. sesban 1.37±0.08a 1.01±0.01ab 0.13 E MN8 S. sesban 5.46±0.17h 4.01±0.40cde 3.58 E MN18 S. sesban 2.96±0.18b‒f 2.47±0.32a-e 1.59 E MN19 S. macowaniana 7.90±1.02i 4.14±0.02a-d 3.71 E MN57 S. sesban 3.88±0.20fg 3.80±0.41b-e 3.36 E MN68 S. sesban 3.35±0.28def 2.71±0.12a-e 2.28 HE 1.66±0.14a 0.97±0.14ab 0.09 E Kavutiri (Kenya) Gituamba (Kenya) Bumala (Kenya) Lushoto (Tanzania) MASS39 S. sesban Kabale (Uganda) Mbale (Uganda) Namibia Reference strains Kibwezi (Kenya) KFR269 Siratro 109 Loruk (Kenya) KFR209 Faidherbia albida 4.54±1.15gh 1.74±0.26a-e 1.31 E BA37 P. vulgaris 2.04±0.19ab 0.16±0.03a ˗0.28 I KFR647 S. sesban 3.23±0.20c‒f 2.19±0.16a-d 1.76 E KFR402 S. sesban 2.20±0.20abc 1.20±0.21abc 0.77 E 0 ppm N 2.45±0.10a‒e 0.43±0.05a 0.00 ‒ 70 ppm N 3.93±0.15fg 3.70±0.39b-e 0.00 ‒ Nyamonye (Kenya) Controls Means within a column followed by the same letter(s) are not significant different according to Tukey HSD at 5 %. Values are means of four replicates ± standard errors. SE (%), symbiotic effectiveness; I, ineffective; E, effective; HE, highly effective. Rhizobial isolates MN57, MN8, MN19, MASS171, MASS130, MASS129 and MASS172, recovered from sesbania had a higher SE on S. sesban compared to plus N control (Figure 4.8). Isolate MASS172 had the highest SE compared to all other isolates used in this study. 110 Figure 4.8: Effectiveness of some selected sesbania rhizobial isolates on S. sesban plants. Error bars with 5 %. 4.6.1 Correlation among shoot dry weight, nodule number, nodule dry weight, nitrogen concentration and SE of S. sesban plants There was no correlation (r2 = 0.191, p =0.097) between the number of nodules per S. sesban plant and the respective shoot dry weight (Table 4.11). There was a very strong positive correlation (r2 = 0.836, p <0.001) between shoot dry weight of S. sesban inoculated using sesbania rhizobia and the respective nodule dry weight. There was also a strong positive correlation (r 2 = 0.818, p < 0.001) between nodule number per plant and dry weight of nodules following inoculation using sesbania rhizobia. There was no correlation (r2 = 0.171, p = 0.395) between percent N per shoot of S. sesban and shoot dry weight. No correlation (correlation (r2 = ‒ 0.142, p = 0.473) was recorded between percent N per shoot of S. sesban and number of 111 nodules per plant (Table 4.11). Similarly, there was no correlation (r2 = 0.257, p = 0.089) between percent N per per shoot and nodule dry weight. Higher symbiotic effectiveness significantly correlated positively (r 2 = 0.499, p < 0.05) with shoot dry weight. There was no significant correlation (r 2 = 0.083, p =0.224) between SE and the number of nodules per S. sesban plant. A significant correlation (r 2 = 0.479, p < 0.05) was recorded between SE and nodule dry weight per plant. Likewise, there was a significant correlation (r2 = 0.933, p < 0.001) between % N per plant and SE in S. sesban (Table 4.11). 112 Table 4.11: Correlation between shoot dry weights, nodule number, nodule dry weight, nitrogen concentration and symbiotic effectiveness of rhizobia in S. sesban SDWt. SDWt. NNo. NDWt. % N shoot‒1 SE NNo. NDWt. % N shoot‒1 SE 1 Pearson Correlation 0.191 1 Sig. (2‒tailed) 0.097 Pearson Correlation 0.836** 0.818** Sig. (2‒tailed) <0.001 < 0.001 Pearson Correlation 0.171 ‒0.142 0.257 Sig. (2‒tailed) 0.395 0.473 0.089 Pearson Correlation 0.499* 0.083 0.479* 0.933** Sig. (2‒tailed) <0.05 0.224 < 0.05 < 0.001 1 1 1 *,** correlation is significant at 0.05 and 0.001 level (2‒tailed) respectively. SDWt., shoot dry weight; NDWt., nodule dry weight; NNo., nodule number; SE (%), symbiotic effectiveness. 113 4.7 Symbiotic effectiveness test of sesbania rhizobial isolates on Rose coco bean plants using Leonard jars The deliberate introduction of rhizobial isolates from sesbania on roots of common bean seedlings allowed determination of microsymbionts' infectivity and effectiveness. The leaf colour of common beans inoculated using sesbania rhizobial isolates were yellowish green while those of Minus N treatment (controls) were yellow at harvesting time (Plate 4.11). A C B D Plate 4.11: Four week‒old common bean plants inoculated with sesbania rhizobial isolates (A, MASS133); B, BA37 (common bean inoculant strain); C, Plus N treatment (70 ppm N) (control) and D, Minus N treatment (control). 114 Nodules recovered from the roots of common bean were globose, determinate and measured between 1‒5 mm in diameter. Some of the nodules had a smooth surface as those found on roots of sesbania while others had a rough surface typical of bean nodules. Nodules were observed on entire fibrous roots of the rhizobia infected common beans (Plate 4.12). Nodules Plate 4.12. Nodules on roots of four‒week old common bean inoculated with S. sesban rhizobial strain MASS133. Cross section of common bean nodules formed by effective isolates were large and pink (Leghemoglobin present) while those formed by ineffective isolates were small and green (Leghemoglobin absent) (Plate 4.13). Nodules borne on roots of common beans were spherical with either rough or smooth surfaces. 115 A B 1 cm Plate 4.13: Transverse section of nodules formed by sesbania isolates on common bean. A, Leghemoglobin present; B, Leghemoglobin absent. Fifteen out of the 79 rhizobia recovered from root nodules of various sesbania elicited nodules on the roots of common bean. There was no nodulation on roots of Plus N and Minus N treatments (control). The number of nodules on roots inoculated bean treatments ranged between 3 and 63 and was significantly different (p < 0.001). The nodule dry weight varied significantly (p < 0.001) and ranged between 0.5 mg (MASS31a) and 153 mg (MASS149). Seventy three percent of the infective isolates originated from S. sesban root nodules. The mean shoot dry weight accumulated by nodulated common bean was not significantly different (p =0.244) and ranged between 0.45 g (MASS31a) and 0.87 g (MASS133) per plant (Table 4.12). 116 Table 4.12: Effect of sesbania rhizobial isolates on shoot dry weight, nodule dry weight and nodule number of common beans Isolate‒Origin SDWt. (g) NDWt. (mg) NNo. MASS133 Sesbania sesban 0.87±0.03abc 145.6±44.20bc 63.00±15.00b MASS137a S. sesban 0.69±0.09ab 61.8±20.10a‒c 36.00±9.00ab MASS149 S. sesban 0.55±0.05a 153.0 ±62.40c 33.00 ±16.00ab MASS30 S. sesban 0.51±0.11a 5.0±1.90a 5.00±2.00a MASS31a S. sesban 0.45±0.06a 0.5±0.30a 3.00±2.00a MASS38 S. sesban 0.81±0.13abc 7.0± 4.90a 3.00±2.00a MASS40ab S. sesban 0.79±0.08abc 5.8±2.80a 5.00±2.00a MASS42 S. sesban 0.55±0.11a 5.4±3.30a 13.00±9.00a 0.67±0.07ab 20.4±11.70a 8.00±3.00a 0.76±0.06bc 16.7±4.80a 26.00±7.00a MN2 S. sphaerosperma 0.69±0.10abc 12.9±9.30a 5.00±3.00a MN28 S. cinerascens 0.74±0.04ab 10.6±4.40a 6.00±3.00a MN39 S. macowaniana 0.74±0.07ab 16.8±9.40a 10.00±4.00a MN69 S. microphylla 0.75±0.07ab 47.1±26.70ab 25.00±10.00a DWO253 Phaseolus vulgaris 0.70±0.10a 30.9± 15.30a 18.00±9.00a KFR269 Siratro 0.68±0.10ab 16.2±4.80a 29.00±9.00ab 0.60±0.17 52.00±12.00 16.00±3.00 0.71±0.10a 34.1±16.20a 13.00±6.00a 0.60±0.10ab 36.0±10.50a 20.00±6.00a 0 ppm N 0.36±0.17 - - 70 ppm N 0.98±0.21 - - Kavutiri (Kenya) Lushoto (Tanzania) Tororo (Uganda) MASS51c S. sesban Mbale (Uganda) MASS57 S. sesban Namibia Reference strains Kibwezi (Kenya) Loruk (Kenya) KFR209 Faidherbia albida Nyamonye (Kenya) KFR402 S. sesban Gituamba (Kenya) BA37 P. vulgaris Controls SDWt., shoot dry weight; NDWt., nodule dry weight; NNo., nodule number; Means within a column followed by the same letter(s) are not significant different according to Tukey HSD at 5 %. Values represent the mean ± SD of eight replicates. 117 There was a significant difference (p < 0.001) of shoot % N due to inoculation of common bean plants using rhizobial isolates from sesbania and some reference strains from various hosts (Table 4.13). Isolate MASS133 caused the highest shoot N content (3.14 %). Nitrogen content per plant ranged between 0.34 mg (MASS31a) and 3.08 mg (MASS57) and had a significant difference (p < 0.001). Similarly, the nitrogen fixation ratio of nodulated common beans plants ranged between ˗0.4‒2.16. Fifteen isolates were symbiotically effective, four of them described as highly effective. Four isolates MASS149, MASS30, MASS31a and MASS42 were ineffective on the common bean plants. The computed symbiotic effectiveness (SE) of four elite strains (MASS38, MASS57, MASS133 and MASS40ab) was rated as very effective (Table 4.13). 118 Table 4.13: Effect of sesbania rhizobial isolates on percent nitrogen, N content, N fixation ratio and the symbiotic effectiveness of common beans % N shoot ‒1 Isolate N content (mg) shoot ‒1 N fixation ratio SE Kavutiri (Kenya) MASS133 3.14±0.28g 2.75±0.22e 2.01 HE MASS137a 1.92±0.16c‒f 1.32±0.27a-e 0.58 E MASS149 2.03±0.13c‒f 1.11±0.07a-e 0.37 I MASS30 2.05±0.31c‒f 1.04±0.19abc 0.30 I MASS31a 0.75±0.04a 0.34±0.08a ˗0.40 I MASS38 2.05±0.36c‒f 1.65±0.17a-e 0.91 HE MASS40ab 1.90±0.33b‒f 1.50±0.37a-e 0.76 HE MASS42 1.36±0.06a‒d 0.75±0.19abc 0.01 I 1.24±0.26abc 0.83±0.28abc 0.09 E 2.69±0.09f‒g 3.08±0.23abc 2.16 HE MN2 1.89±0.21b‒f 1.86±0.32c-e 1.12 E MN28 2.51±0.08e‒g 3.03±0.19b-e 1.13 E MN39 2.85±0.12f‒g 2.12±0.20d-e 1.38 E MN69 2.65±0.10e‒g 1.98±0.22b-e 1.24 E DWO253 Phaseolus vulgaris 1.65±0.20a‒e 1.16±0.30a-d 0.42 E KFR269 Siratro 1.11±0.11abc 0.76±0.12abc 0.02 E Faidherbia 2.41±0.25d‒g 1.59±0.34b-e 0.85 E 2.31±0.19d‒g 1.64±0.40a-e 0.90 E BA37 P. vulgaris 2.10±0.14c‒f 1.25 ±0.27a-e 0.51 E 0 ppm N 1.10±0.04abc 0.90±0.10ab 0.00 ‒ 70 ppm N 1.48±0.12a‒d 2.33±0.10 a-e 0.00 ‒ Lushoto (Tanzania) Tororo (Uganda) MASS51c Mbale (Uganda) MASS57 Namibia Reference strains Kibwezi (Kenya) Loruk (Kenya) KFR209 albida Nyamonye (Kenya) KFR402 S. sesban Gituamba (Kenya) 119 Mean % N: f (20, 83) = 12.37, p‒value < 0.001, l.s.d = 0.4402. Mean N content (mg) plant -1: f (20, 83) = 5.21, p‒value < 0.001, l.s.d = 0.8348. Means within a column followed by the same letter(s) are not significant different according to Tukey HSD at 5 %. Values are means of four replicates ± standard errors. SE (%), symbiotic effectiveness; I, ineffective; E, effective; HE, highly effective. Symbiotic effectiveness of rhizobia that nodulated common bean varied significantly (p < 0.05) and ranged between 28.4 % for MASS31a and 229.6 % for MASS133 when compared to uninoculated Plus N treatments (control) rated as 100 %. The common bean inoculant (BA37) caused a symbiotic effectiveness of 104.4 % (Figure 4.9). Figure 4.9: Effectiveness of sesbania rhizobial isolates on common beans. Error bars with 5 % value. Apart from MASS133 and MASS137a sesbania rhizobial isolates that were effective nitrogen fixers (nod+nif+ phenotypes) in common beans, all other isolates recovered from the Kenyan sites were non-infective (nod˗) on the common bean (Table 4.14a). 120 Table 4.14a: Nodulation and nitrogen fixation phenotypes of S. sesban rhizobial isolates from Kenya on common beans Site Kavutiri nod˗ nod+nif˗ MASS129, MASS130, MASS149 nod+ nif+/˗ ‒ MASS131, MASS134, nod+ nif+ MASS133, MASS137a, MASS136, MASS137b, MASS137ab, MASS138, MASS140, MASS141, MASS147. Kuinet MASS160, MASS168, ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ MASS170, MASS171, MASS172, MASS174. Bumala MASS173, MASS123, MASS125, MASS126, MASS127, MASS157, MASS167, MASS176, MASS177a, MASS177b, MASS177ab, MASS178. Gituamba MASS110, MASS111, MASS114, MASS115, MASS117a, MASS117b, MASS117c, MASS117abc. All isolates affiliated to S. sesban from Ugandan soils did not form nodules on common beans (nod˗ phenotypes) except MASS51c and MASS57 which were infective and effective (nod+nif+ phenotypes) on the common beans. None of the S. sesban rhizobial isolates obtained from Kabale, Tororo and Mbale sites in Uganda caused ineffective or partially effective nodules (nod+nif˗ or nod+nif+/˗ phenotypes) on common beans (Table 4.14b). 121 Table 4.14b: Nodulation and nitrogen fixation phenotypes of S. sesban rhizobial isolates from Uganda on common beans Site nod˗ nod+nif˗ nod+nif +/˗ nod+nif+ Kabale MASS62 MASS65 ‒ ‒ ‒ ‒ ‒ MASS51c ‒ ‒ MASS57 MASS68 MASS69 Tororo MASS47 MASS48 MASS49 MASS50a MASS50b MASS50ab MASS51a MASS51b MASS51abc MASS52 MASS53 Mbale MASS54 MASS59 MASS60 MASS61 Sesbania sesban isolates MASS38 and MASS40a + MASS40b (as a mix inoculant of a nodule co‒occupants) from Lushoto formed effective nodules (nod+nif+ phenotypes) on common beans while MASS42 (SUA), MASS30 and MASS31a (Lushoto) formed ineffective nodules on common beans (nod+nif˗ phenotypes) (Table 4.14c). There were no partially effective nodules formed on the common beans by S. sesban rhizobial isolates from the two Tanzanian sites. All other isolates from the two sites did not form nodules (nod˗ phenotypes) on common bean plants. 122 Table 4.14c: Nodulation and nitrogen fixation phenotypes of S. sesban rhizobial isolates from Tanzania on common beans Site nod˗ nod+nif˗ nod+nif+/˗ nod+nif+ SUA MASS41 MASS42 ‒ ‒ MASS30 ‒ MASS38 MASS40ab MASS43 Lushoto MASS29 MASS31b MASS31a MASS33 MASS36 MASS37a MASS37b MASS37ab MASS39 MASS40a MASS40b Four isolates (MN2, MN28, MN39 and MN69) from various Namibian sites were effective (nod+nif+ phenotype) on common beans (Table 4.14d). All other sesbania rhizobia from Namibia were non‒infective (nod‒ phenotypes) on common bean plants. None of the sesbania rhizobial isolates from Namibia caused ineffective or partially effective nodules. 123 Table 4.14d: Nodulation and nitrogen fixation phenotypes of sesbania rhizobial isolates from Namibia on common beans Site Khorixas Outijo Swakop nod˗ nod+nif˗ nod+nif +/˗ nod+nif+ ‒ ‒ ‒ MN39 ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ MN28 ‒ ‒ ‒ ‒ ‒ MN2 ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ MN69 ‒ ‒ ‒ MN9, MN44, MN49, MN51. Rio Tinto MN45, MN50, Gorge MN54, MN56, MN58, MN62. Epupa falls MN8, MN13, MN18, MN19, MN21, MN22, MN57. Suclabo MN31, MN40. MN37, Omuramba MN11, MN17 Rooidrom MN10. Sesfontein MN12, MN71. Okahandja MN1, MN15. Otjinungua MN68 Bunya MN70 Korokoko MN50 Five reference isolates viz: BA37 (P. vulgaris), KFR209 (F. albida), DWO253 (P. vulgaris), KFR269 (Siratro) and KFR402 (S. sesban) exhibited nod+nif+ phenotype on common bean plants (Table 4.14e). A S. sesban inoculant production strain (KFR647) failed to induce nodules on the common bean plants. None of the reference rhizobial strains from Kenya caused ineffective or partially effective nodules on common beans. 124 Table 4.14e: Nodulation and nitrogen fixation phenotypes of reference rhizobial strains from Kenya on common beans Site nod˗ nod+nif˗ nod+nif+/˗ nod+nif+ Gituamba ‒ ‒ ‒ BA37 (P. vulgaris) Loruk ‒ ‒ ‒ KFR209 (F. albida) Kibwezi ‒ ‒ ‒ DWO253 (P. vulgaris), KFR269 (Siratro) Yala swamp KFR647 ‒ ‒ ‒ ‒ ‒ KFR402 (S. sesban) (S. sesban) Nyamonye ‒ 4.7.1 Correlation among shoot dry weight, nodule number, nodule dry weight, nitrogen concentration and SE of common bean plants There was a significant positive correlation (r 2 =0.323, p < 0.05) between number of nodules per plant and the shoot dry weight of the plants (Table 4.15). There was no significant correlation (r2 = 0.025, p = 0.0529) between nodule dry weight and shoot dry weight following rhizobial inoculation of common bean using sesbania rhizobia. There was also a significant positive correlation (r2 = 0.461, p < 0.05) between % N per shoot and shoot dry weight. A strong positive significant correlation (r2 = 0.6943, p < 0.001) between symbiotic effectiveness (SE) and shoot dry weight. There was a strong positive correlation (r2 = 0.676, p < 0.001) between number of nodules and nodule dry weight. No significant correlation (r2 = 0.204, p = 0.089) was recorded between % N per shoot and number of nodules per plant. There was a weak correlation (r2 = 0.302, p < 0.05) between SE and number of nodules per plant. There was a significant positive correlation (r2 =0.386, p < 0.05); (r2 =0.388, p < 0.05) between % N per shoot and nodule dry weight; SE and nodule dry weight 125 respectively (Table 4.15). A strong positive correlation (r2 = 0.946, p < 0.001) was recorded between nitrogen concentration (% N) per shoot and symbiotic effectiveness on common bean treatments due to inoculation with sesbania isolates (Table 4.15). 126 Table 4.15: Correlation between shoot dry weight, nodule number, nodule dry weight, nitrogen concentration and SE on common beans SDWt. (g) NNo. NDWt. (mg) % N shoot‒1 SDWt. Pearson Correlation (g) Sig. (2‒tailed) NNo. Pearson Correlation 0.323* Sig. (2‒tailed) < 0.05 NDWt. Pearson Correlation 0.025 0.676** (mg) Sig. (2‒tailed) 0.529 < 0.001 Pearson Correlation 0.461* 0.204 0.386* Sig. (2‒tailed) < 0.05 0.089 < 0.05 Pearson Correlation 0.694** 0.302* 0.388* 0.946** Sig. (2‒tailed) < 0.001 < 0.05 < 0.05 < 0.001 %N shoot SE ‒1 SE 1 1 1 1 1 *, ** correlation is significant at 0.05 and 0.001 level (2‒tailed) respectively. SDWt., shoot dry weight; NDWt., nodule dry weight; NNo., nodule number; SE (%), symbiotic effectiveness. 127 Fourteen (14) sesbania rhizobial isolates from S. sesban, S. sphaerosperma, S. macowaniana, S. cinerascens and S. microphylla were infective on common bean while 97 isolates were non‒infective (Table 4.16). Isolate KFR402 (S. sesban) caused nodules on roots of common beans. Rhizobial isolates from S. pachycarpa, S. bispinosa and S. rostrata did not form association with common beans. Other infective isolates with origin from non‒sesbania included BA37, DWO253 (common bean), KFR209 (Faidherbia albida) and KFR269 (Siratro). Isolates from Acacia tortilis and A. xanthophloea were non‒infective on common beans. 128 Table 4.16: Sesbania rhizobia with infective traits on common beans Host of origin Sesbania sesban Infective rhizobia Non‒infective rhizobia MASS30 MASS31a MASS38 MASS29, MASS31b, MASS33, MASS36, MASS37a, MASS37b, MASS37ab, MASS39, MASS40a, MASS40b, MASS41, MASS43, MASS47, MASS48, MASS49, MASS50a, MASS50b, MASS50ab, MASS51a, MASS51b, MASS51abc, MASS52, MASS53, MASS54, MASS59, MASS60, MASS61, MASS62, MASS65, MASS68, MASS69, MASS110, MASS111, MASS114, MASS115, MASS117a, MASS117b, MASS117c, MASS117abc, MASS123, MASS125, MASS126, MASS127, MASS129, MASS130, MASS131, MASS134, MASS136, MASS137b, MASS137ab, MASS138, MASS140, MASS141, MASS147, MASS157, MASS160, MASS167, MASS168, MASS170, MASS171, MASS172, MASS173, MASS174, MASS176, MASS177a, MASS177b, MASS177ab, MASS178, KFR647, MN8, MN13, MN18, MN21, MN57, MN68 MASS40ab MASS42 MASS51c MASS57 MASS133 MASS137a MASS149 KFR402 MN9, MN44, MN45, MN49, MN51, MN56, MN58, MN62 S. pachycarpa S. sphaerosperma MN2 MN10, MN12 S. macowaniana MN39 MN1, MN22, MN15, MN19, MN71 S. cinerascens MN28 MN31, MN37, MN40 S. microphylla MN69 MN70 S. bispinosa MN11, MN17 S. rostrata MN50 P. vulgaris BA37 DWO253 KFR84 Acacia tortilis Faidherbia albida KFR209 Siratro A. xanthophloea KFR269 DWO75 129 4.8 Multiple rhizobial occupancy in root nodules of Sesban sesban Several rhizobial isolates from S. sesban grown in East African sites were found to co‒occupy the nodules. Test for ability of the co‒occupants to infect S. sesban resulted in non‒infective or sub‒obtimal nitrogen fixation. Nodulated S. sesban inoculated using either of the nodule rhizobial companions or in their mixed forms resulted in significantly (p < 0.001) lower shoot dry weight compared to Plus N (control) and the S. sesban inoculant production strain (KFR647) treatments. Isolate MASS177b had the highest nodule dry weight (0.67 mg) and specific nodule weight of 3.14 mg. A mixed inoculant MASS177a + MASS177b (MASS177ab) was non‒infective on S. sesban while in their separate forms isolates MASS177a and MASS177b initiated 4.6 and 21.4 nodules respectively. Similarly, a mixture of isolate MASS31a + MASS31b (MASS31ab) was non‒infective on S. sesban. Isolate MASS31a caused 36 nodules per plant with shoot dry weight of 0.47 g while MASS31b elicited only one nodule per plant with shoot dry weight of 0.09 g. A combined rhizobial inoculation using companions MASS117a + MASS117b + MASS117c (MASS117abc) of a single nodule recorded insignificant difference of shoot dry weight 0.34 g compared to a single isolate MASS117c (0.40 g). The shoot dry weight 0.16 g caused by isolates MASS117a and MASS117b was not significantly different (p > 0.05) from shoot dry weight of Minus N treatment (Table 4.17). 130 Table 4.17: Growth response of S. sesban on inoculation using multiple nodules co-occupant rhizobial isolates from roots of S. sesban Isolate SDWt.(g) NNo. NDWt.(mg) SNWt. (mg) MASS177a 0.10±0.02ad 4.60±2.40ab 0.01±0.01a 0.25±0.12a MASS177b 0.48±0.12def 21.40±4.70cde 0.67±0.02d 3.14±0.85c MASS177ab 0.11±0.01ad Bumala (Kenya) - - - Gituamba (Kenya) MASS117a 0.16±0.08ad 4.60±3.90ab 0.09±0.01ab 0.57±0.28ab MASS117b 0.16±0.05ad 0.60±0.40a 0.01±0.01a 0.85±0.63ab MASS117c 0.40±0.08ae 17.60±3.00bcd 0.51±0.01bcd 2.69±0.30bc MASS117abc 0.34±0.11ae 11.00±6.00ad 0.26±0.02ad 1.42±0.40abc MASS31a 0.47±0.08 cf 36.00±4.90e 0.59±0.11cd 1.74±0.26abc MASS31b 0.09±0.02abc 0.90±0.30a 0.01±0.01a 0.49±0.16ab MASS31ab 0.14±0.03ad - - - MASS37a 0.07±0.02a 6.80±2.00abc 0.06±0.02ab 1.24±0.70abc MASS37b 0.30±0.08ae 15.00±3.00ad 0.30±0.01ad 1.99±0.40abc MASS37ab 0.12±0.03ad - - - MASS50a 0.15±0.03ad - - - MASS50b 0.45±0.07bf 13.00±4.00abcd 0.04±0.01ad 1.96±0.44abc MASS50ab 0.08±0.03ab 1.00±0.50a 0.01±0.01a 0.25±0.14a MASS51a 0.15±0.03ad 3.90±2.30ab 0.05±0.03ab 0.73±0.32ab MASS51b 0.10±0.03ad - - - MASS51c 0.18±0.02ad 2.30±1.00a 0.06±0.04ab 1.13±0.50abc MASS51abc 0.14±0.02ad 4.00±1.00ab 0.04±0.01ab 1.23±0.49abc 0.58±0.09ef 22.00±3.30de 0.60±0.01d 2.69±0.24bc Lushoto (Tanzania) Tororo (Uganda) Reference strains KFR647 Controls 0 ppm N 0.15±0.03ad - - - 70 ppm N 0.98±0.14f - - - SDWt. (g) f (22, 159) = 7.01, p‒value < 0.001, l.s.d = 0.2141, NNo. f (22, 159) = 10.47, p‒value < 0.001, l.s.d = 8.185, NDWt. (mg), f (22,159) =6.39, p‒value < 0.001, l.s.d = 0.026 and SNWt. (mg), f (22,159) =4.72, p‒value < 0.001, l.s.d = 1.213. Means within a column followed by the same letter(s) are not significant different according to Tukey HSD at 5 %. Values are means of four replicates ± standard errors. SDWt., shoot dry weight; NNo., nodule number; NDWt., nodule dry weight; SNWt., specific nodule weight. 131 4.9 Sesbania sesban and common bean cross‒inoculating isolates All the isolates with cross‒inoculating ability on S. sesban and the common beans caused low shoot dry weight on both species compared to the respective Plus N treatments (control). There was no significant difference (p ≤ 0.05) in shoot dry weight of S. sesban inoculated using strains KFR269, KFR209 and KFR647 (S. sesban inoculant strain). Similarly, there was no significant difference (p ≤ 0.05) in relative effectiveness of isolate MASS133, DWO253 and BA37 (inoculant strain for common beans) on S. sesban (Figure 4.9a). Isolate KFR402 showed the highest relative effectiveness (0.35) on S. sesban. b a bc a Figure 4.10a: Response of inoculation on shoot dry weight (g) of S. sesban. Standard error (SE) bars with the same letters are not significantly different (p > 0.05) according to Tukey’s HSD test. The isolate MASS133 caused a significant difference (p ≤ 0.05) of a higher relative effectiveness of 0.2 on common bean compared to all other isolates. The common 132 bean inoculant production strain BA37 caused the least relative effectiveness (-0.07) compred to Plus N treatment (control). Common bean shoot dry weight caused by isolates KFR402, KFR269, DWO253 and BA37 were not significantly different (p > 0.05) from Minus N treatment (control) (Figure 4.9b). a e b b b b c bd d Figure 4.10b: Response of inoculation on shoot dry weight (g) of common bean. Bars with the same letters are not significantly different (p > 0.05) according to Tukey’s HSD test. 133 CHAPTER FIVE DISCUSSION, CONCLUSIONS AND RECOMMENDATIONS 5.1 Discussion 5.1.1 Morpho–cultural characterization of sesbania rhizobia The morpho-cultural characteristics exhibited by sesbania isolates including colonies with either flat, raised or dome shaped (all with entire margins), fast, moderate or slow growth, production of copious exopolysaccharides (EPS), moderate EPS with gummy properties or non‒EPS producers, acid, alkaline or neutral on YEMA–BTB media were similar to those of root nodules isolated from legumes grown worldwide (Somasegaran and Hoben, 1994; Legesse and Assefa, 2014; Bhargava et al., 2016; Nahar et al., 2017). Majority of fast growing rhizobia were acid producers while the slow growers were alkaline producers on YEMA medium. However, the pH reaction for fast and slow growers was not universal, findings that agree with those by Okaron et al. (2017). The time taken for sesbania rhizobial colonies to attain maximum growth (i.e. 2‒3 days, 4‒6 days and 7‒9 days for fast, moderately fast and slow growers respectively) with a colony size range from < 1.0 mm to 5.0 mm in diameter on YEMA medium incubated in the dark at 28±1 °C were in close agreement with Dereje et al. (2015). Bacteria recovered from the nodules of sesbania were Gram negative, a characteristic shared by all members of α- and β-protobacteria (Berrada and Fikri-Benbrahim, 2014). This characteristic aided in the presumptive identification of rhizobia from the study sites. However, the recovery of fast and slow growing rhizobia from root nodules of sesbania contrasted in the slow growth with the observations by Atangana et al. (2014) that sesbania form nodules with fast-growing strains of Rhizobium. 134 Most of the sesbania rhizobial isolates recovered were clustered in morphotypes I, II, VII, VIII and IX (Figure 4.1), which are probably the most compatible morphotypes possessing high affinity for sesbania as opposed to morphotypes III, IV, V and VI. The high diversity of morphotypes observed with sesbania isolates could be as a result of horizontal gene transfer which confers survival mechanisms over time to bypass adverse conditions found in some tropical soils. For instance, the presence of rhizobial isolates grouped in morphotypes II, IV and V produced copious exopolysaccharides (EPS) a property also characterized by Hewedy et al. (2014) and which is essential for infection process and nodule formation; protect rhizobia in the soil against deleterious biotic and abiotic stress factors and which restrict oxygen diffusion through the nodular cells to protect the oxygen sensitive nitrogenase in the nodules (Bhargava et al., 2016). Rhizobia falling in morphotype III presented a unique characteristic of chromogenic colonies which is contrary to growth of most rhizobia. The results also point out to non-specificity of the rhizobial morphotypes to sesbania used in the study and that the infection process is not stringently controlled by a single variant of flavanoids or isoflavanoids even within the same host (Liu and Murray, 2016). The recovery of rhizobia with heterogeneous colony characteristics in the present study indicates that sesbania symbiotically associate with diverse rhizobial strains. Hence the acceptance of hypothesis that there are phenotypic differences among the rhizobia isolated from various sesbania grown in East Africa and Namibia. Lack of site specificity for morphotypes implies that rhizobial strains compatible to sesbania are readily available in many soils of East Africa and Namibia. However, this finding contradicts that of Bala et al. (2002) in which they found no rhizobia 135 compatible with S. sesban in 39 out of 53 soils sampling sites of Southern Africa. The similarities of colony traits of Mesorhizobium and Sinorhizobium (morphotype II) and Bradyrhizobium and Azorhizobium (morphotype IX) together with the variation of traits for the genus Rhizobium (morphotypes V and VII) suggest presence of colony dimorphism within rhizobia nodulating sesbania which may lead to identification inconsistencies when only morphological traits are relied upon. An intriguing observation in the present study is the evidence of hollow centres (Plate 4.2) in some rhizobial colonies isolated from Namibian sesbania nodules. The hollow centres could be a symptom of rhizobial infection by rhizobiophages. Rhizobiophages are important vectors for rhizobial transformation, a potential means of genetic diversity among rhizobia through conferring of traits like resistance to antibiotics, tolerance to salts and infectiveness (Santamaría et al., 2013). However, rhizobiophages could as well have a negative effect on elite strains in the soil through elimination of susceptible rhizobial populations or transform infective and effective strains into non-infective forms (Msimbira et al., 2016). 5.1.2 Intrinsic antibiotic resistance and salt tolerance The intrinsic antibiotic resistance (IAR) and salt tolerance are important techniques used for discrimination and identification of rhizobial strains. Sesbania rhizobial isolates showed different degrees of susceptibility of between 26.8 % and 90.8 % (Figure 4.3) to the twelve antibiotics which can be attributed to genetic variation in target genes and resistance genes acquired through horizontal gene transfer (Bhargava et al., 2016). Notable is the less resistance to kanamycin and which support findings by The isolates also showed a generally low IAR for kanamycin, which is a very common characteristic 136 for tropical rhizobia. This phenotypic trait is highly desired for inoculant strains persistence amidst antibiotics producing microbes present in many agricultural soils (Shetta et al., 2011; Adegboye and Babalola, 2013). Sesbania grow in diverse conditions that include water-logged and dry soils with varying salinity levels in different parts of the world. Salinity is one of the adverse conditions that affect rhizobia‒legume interactions leading to poor nitrogen fixing through inhibition of the initial steps of the symbiosis. It was intriguing to note that some of the sesbania isolates were highly halotolerant at a concentration of 10 % NaCl (w/v) which is higher than the 5 % recorded by Messaoud et al. (2014) but agrees with the findings of Bouzeraa-Bessila et al. (2015). In their findings, the most salt-tolerant strains MSC2 and MSC9 isolated from Chatt soil and MSF21, MSF20 and MSF19 isolated from Fetzara soil were tolerant at 10 % of NaCl. This might be as a result of the soil chemical characteristics of the site where sesbania root nodules were recovered. Results of the present study suggest that IAR and salt tolerance dendrogram patterns (Figures 4.5a–4.5j) were sufficiently used in grouping of rhizobial isolates from sesbania grown in diverse habitats of East Africa and Namibia, which supports the use of some phenotypic traits for strain identification (Chanway and Holl, 1986). 5.1.3 Molecular characterization of sesbania rhizobia According to the 16S rRNA PCR‒RFLP unrooted tree topologies (Figure 4.6a–4.6d), rhizobial isolates from root nodules of sesbania were tentatively grouped into the genera Rhizobium, Sinorhizobium, Bradyrhizobium, Mesorhizobium, Azorhizobium and Agrobacterium with an overall mean cophenetic correlation coefficient of 62.1 % (bootstrap value) compared to reference strains. Except for Bradyrhizobium, these results support those of Young and Haukka (1996) that S. sesban rhizobia are 137 genetically diverse, with isolates represented in the genera Rhizobium, Mesorhizobium, Sinorhizobium and Azorhizobium. The presence of unmatched isolates (dissimilar to all the reference strains used in this study) from Kenya and Namibia was an indication that there exists rhizobial genera other than Rhizobium, Sinorhizobium, Bradyrhizobium, Mesorhizobium and Agrobacterium that also form symbiosis with sesbania. These data demonstrated a high diversity of rhizobia associated with sesbania and therefore the hypothesis that there are genotypic differences among the rhizobia isolated from various sesbania grown in East Africa and Namibia is accepted. These results agree with those of Nahar et al. (2017) where rhizobia affiliated with S. bispinosa showed genotypic diversity using 16S rRNA sequences. Results from the present study have also shown that the PCR-RFLP markers were not able to resolutely split Agrobacterium spp. and Sinorhizobium spp. isolated from sites in Kenya and Tanzania; Bradyrhizobium spp. and Rhizobium spp. Type A, Mesorhizobium spp. and Agrobacterium spp. isolated from sites in Uganda; Sinorhizobium spp. and Agrobacterium spp., Mesorhizobium spp. and Rhizobium spp. Type B from sites in Tanzania. This may possibly indicate that some of the strains in the sites of nodule collection are actively exchanging their genome or acquiring genes from the surroundings to suit both host and abiotic conditions for survival. The phylogenetic disimilarities range of 2 % ‒ 100 % is an indication of a rich diversity of rhizobia affiliated symbiotically to sesbania growing in East Africa and Namibia. Findings of the present study show that the bulk of sesbania isolates from various sites were of the genera Rhizobium, Sinorhizobium, Bradyrhizobium and Mesorhizobium. With the exception of Bradyrhizobium spp., these results are in line 138 with those of Bala and Giller (2001) who reported that S. sesban are effectively infected by rhizobia that belong to the genera Rhizobium, Mesorhizobium and Sinorhizobium. However, these contrast with those of Singh et al. (2013) where only fast growers (Rhizobium spp.) were recovered from S. sesban which were grown in Mumbai and its suburban areas. In the present study isolates with Agrobacteriumlike 16S rRNA PCR-RFLP patterns were recovered from sesbania of Namibia which is an indication that some Agrobacterium spp. acquired the ability to nodulate legumes through transformation, a finding that supports that of Kondorosi et al. (1982) that Agrobacterium mutants supplied with symbiotic plasmids of rhizobia are conferred the ability to form nodules and fix nitrogen. However, the 16S rRNA PCRRFLP clustering together of Agrobacterium-like rhizobia from Kenya and Tanzania with Sinorhizobium spp. and those from Uganda clustered with Mesorhizobium spp. might be a pointer to the source of symbiotic genes that transforms Agrobacterium spp. to legume nodulating bacteria. 5.1.4 Nodulation and nitrogen fixation potential of sesbania rhizobia on S. sesban A high number of sesbania rhizobia (83 %) caused nodulation on S. sesban an indication that members of this genus Sesbania easily cross-inoculate with similar rhizobia. Sesbania rhizobial infection and the associated nodule formation on roots of S. sesban were not host or site specific which can plausibly be attributed to wide distribution of sesbania and their compatible rhizobia in the tropics following widespread seed dispersal to new sites mainly accompanied by compatible microsymbionts as suggested by Grange et al. (2007). 139 In the present study S. sesban responded to inoculation using sesbania rhizobia with the four possible nodulation and nitrogen fixation phenotypes nod˗, nod+fix‒, nod+fix+/˗ and nod+fix+ as was earlier described by Maunoury et al. (2010), Liu et al. (2011), Saeki, (2011) and Melino et al. (2012). Symbiotic effectiveness test of sesbania rhizobia on S. sesban revealed a huge disparity in the ability to fix nitrogen as indicated by shoot dry weight range of 0.06 g ‒ 1.06 g per plant. This variation in effectiveness supports the hypothesis that not all rhizobia from root nodules of sesbania are effective in fixing nitrogen with S. sesban. The unexpected failure of isolates recovered from S. sesban to nodulate the host of isolation resulting in negative Koch's postulates may be attributed to premature abortion of bacteroid differentiation process, failed bacterial endocytosis from infection threads into cortical cells or a reduced pool of functional bacteroids after they underwent premature senescence (Melino et al., 2012). The highly positive significant correlation (r2 = 0.836, p < 0.001) between shoot dry weight of S. sesban and the respective nodule dry weight imply that high nodule biomass is responsible for fast growth of S. sesban in the absence of external nitrogen sources. The high nodule biomass observed in treatments with efficient nitrogen fixation could be as a consequence of rhizobial affinity for root cortex cells and the subsequent entry into the cells resulting in higher hyperplasic and greater hypertrophic effects on the symbiosomes. Results in the present study also show no significant correlation (r2 = 0.191, p = 0.097) between nodule number and shoot dry weight of S. sesban which is inconsistent with that of Chemining’wa et al. (2011) where they reported a positive and significant correlation between the two parameters. In the present study, the high weight of nodules can have compensated for high number of nodules and the incidences of low plant biomass of profusely nodulated legume plants 140 could imply failure of host to sanction ineffective nod+fix˗ or nod+fix+/˗ phenotypes rhizobial strain(s) (Sachs et al., 2010). Six isolates recovered from S. sesban [MASS59 (Mbale‒Uganda), MN18 and MN68 (Nambia), MASS147 and MASS129 (Kavutiri‒Kenya) and MASS172 (Kuinet‒Kenya)] caused high shoot dry weight compared to the uninoculated plus N (control) treatments. Most of the isolates had significant (p < 0.001) low shoot dry weights compared to the uninoculated minus N (control) treatments (0.36 g per plant) denoting parasitism nature of some highly infective microsymbionts which concurs with the findings by Sachs et al. (2010) that legumes have no ability to detect effectiveness of rhizobia at the point of infection. The significant differences (p < 0.001) in mean percent nitrogen and nitrogen content per shoot of S. sesban inoculated using selected rhizobia from various legume hosts indicate polymorphism in genes involved in nitrogen fixation present in different rhizobial isolates. Nodulation of S. sesban by rhizobia from legumes other than sesbania [i.e. strains BA37 and DWO253 (P. vulgaris), KFR269 (Siratro) and KFR209 (F. albida), all included in the present study] indicates that S. sesban and the common beans fall in a common cross-inoculation group with a wide host range inluding other herbaceous legumes and tree species. These results match and support those of Wolde-Meskel et al. (2016) where most of the rhizobia strains identified as effective symbionts of sesbania were originally isolated from trees and leguminous food crops other than sesbania suggesting an availability of broad range isolates in soils where other legumes are grown. 141 5.1.5 Nodulation and nitrogen fixation potential of sesbania rhizobia on Rose coco bean variety Sesbania isolates formed root nodules on Rose coco bean plants (Table 4.16) which plausibly indicate that common beans together with sesbania fall within the same cross-inoculation group. These findings support the hypothesis that sesbania rhizobial isolates infect common beans. The cross nodulation ability of the two legumes was further confirmed by the reciprocal nodulation of S. sesban by strains of common bean origin (Table 4.9e). These results support the idea that tropical legumes nodulate with rhizobial strains found outside their centre of diversity as a result of promiscuity in host range (Giller, 2001). A limited number of 14 out of the total 128 test sesbania rhizobial isolates formed nodules on roots of Rose coco bean cultivar. The most probable explanation for these results is the lack of preference by sesbania rhizobia for the Rose coco common bean cultivar used in the present study (Gicharu et al., 2013). The symptomatic nitrogen deficiency frequently manifests in plants growing in nitrogen poor soils, non-nodulated legumes or legumes associated with ineffective rhizobial strains. Unexpectedly Rose coco bean variety nodulated by sesbania rhizobial nod+nif+ phenotypes with pink nodules (Plate 4.13) and which rated as highly effective (Table 4.13) had yellowish green leaves compared to the dark green colour of the non‒inoculated Plus N treatments (control) (Plate 4.11). However, the yellow leaf colour of the bean plants with effective nodules not only points to the high nitrogen demand by bean plants especially during the onset of flowering and podding but also to the low nitrogen fixation ability of the pulse compared to other crop legumes (Graham, 1981; Hardarson et al., 1993). Inspite of the leaf yellow 142 colour of nodulated common bean, these findings support the null hypothesis that sesbania rhizobial isolates infect and effectively fix nitrogen with Rose coco beans variety. These results contrast with those of Degefu et al. (2011) where none of the rhizobial strains recovered from the tree species S. sesban, Acacia abyssinica, A. tortilis and A. senegal were able to induce nodules in crop legume species. Rose coco bean variety was also nodulated by non-sesbania isolates KFR209 (F. albida) and KFR269 (siratro), findings which support observations by Michiels et al. (1998) and Perret et al. (2000) that common bean has the ability to recognize signals that trigger the nodulation process from many rhizobia although the resultant interactions are often not effective. The negative relative effectiveness of MASS133, DWO253 and BA37 on S. sesban and KFR209 and BA37 on Rose coco beans can be attributed to failure of host-entry restriction mechanisms and symbiotic selection centres against 'parasitic' rhizobial strains (Perret et al., 2000) hence the rejection of the hypothesis that sesbania rhizobial isolates efficiently fix nitrogen with Rose coco bean variety. 5.1.6 Multiple nodule occupancy Two or three rhizobial isolates were found co‒occupying nodules collected from roots of East African S. sesban, results that support earlier findings by Nguyeni et al. (2010) where 10‒50 % of nodules were simultaneously occupied by two or three genotypes. When and how the diverse genotypes of nodule forming rhizobia enter and co‒occupy a single nodule has remained unclear. The significant difference (p < 0.001) between shoot dry weight caused by two nodule companions and the absence of additive effect in mixed inoculants could be attributed to detection and subsequent 143 elimination of the 'cheater' strains as earlier described by Jones et al. (2015) and Checcucci et al. (2016). The frequent low shoot dry weight of S. sesban inoculated using mixed co-occupants compared to one of the co-occupants support (Friesen and Mathias, 2010) that legume hosts discriminate between strains within a nodule resulting in emergence of polymorphism that cause precipitous decline in host benefit. 5.2 Conclusions i. Rhizobia nodulating sesbania grown in East Africa and Namibia are phenotypically and genetically diverse but not site specific. ii. Most sesbania isolates are infective on S. sesban. However, they exhibit great variations in nodulation and effectiveness. iii. Rose coco bean variety fall in the same cross-inoculation group as sesbania but only 14 out of the total 128 sesbania rhizobial isolates caused nodulation. 5.3 Recommendations i. Sesbania are nodulated by diverse rhizobia which calls for keen selection before they are used as inoculant production strains for legumes within its cross-inoculating group. ii. Prospecting for sesbania-common bean cross-inoculating rhizobia in tropical soils where sesbania are endemic should be prioritized in order to recover even more superior strains. iii. Superior genetic tools for example DNA and gene sequencing should be employed to evaluate the genetic diversity of rhizobia nodulating S. sesban and common bean to species level. 144 iv. Isolates MASS172 and MASS133 from root nodules of S. sesban grown in Kuinet and Kavutiri (Kenya) respectively and KFR402 (reference strain from S. sesban) have the potential as inoculant production strains for both species. However, the isolates should be evaluated for effectiveness in soils from different agro-ecological zones where common beans and S. sesban are grown. v. The isolates MASS172, MASS133 and KFR402 should be evaluated for compatibility and nitrogen fixing effectiveness on many other common bean varieties given that there are frequent reports of variations in nodulation and biomass accumulation among bean varieties under controlled greenhouse conditions. 145 REFERENCES Adegboye M. F. and Babalola O. O. (2013). Phylogenetic characterization of culturable antibiotic producing Streptomyces from rhizospheric soils. Molecular Biology, 13:1‒7. Aggani S. L. (2013). Development of biofertilizers and its future perspective. Scholars Academic Journal of Pharmacy, 2:327–332. Anderson J. M. and Ingram J. S. I. (1993). Tropical soil biology and fertility: A handbook of methods 2nd edition. CAB International, Wallingford. Angus A. A. and Hirsch A. M. (2010). Insights into the history of the legume-beta proteobacterial symbiosis. Molecular Ecology, 19:28–30. Anyango B., Wilson K. J., Beynon J. L. and Giller K. E. (1995). Diversity of rhizobia nodulating Phaseolus vulgaris L. in two Kenyan soils with contrasting pHs. Applied and Environmental Microbiology, 61(11):4016‒4021. Aserse A. A., Räsänen L. A., Assefa F., Hailemariam A. and Lindström K. (2012). Phylogeny and genetic diversity of native rhizobia nodulating common bean (Phaseolus vulgaris L.) in Ethiopia. Systematic and Applied Microbiology, 35:120‒131. Atangana A., Khasa D., Chang S. and Degrande A. (Eds.) (2014). Tropical agroforestry. Springer Publishers, p 186 Auch A. F., Jan M., Klenk H. P. and Göker M. (2010). Digital DNA‒DNA hybridization for microbial species delineation by means of genome‒to‒genome sequence comparison. Standards in Genomic Sciences, 2:117‒134. Babalola O. O. (2003). Molecular techniques: An overview of methods for the detection of bacteria. African Journal of Biotechnology, 2(12):710‒713 Baddeley J. A., Jones S., Topp C. F. E., Watson C. A., Helming J. and Stoddard F. L. (2013). Biological nitrogen fixation (BNF) by legume crops in Europe. Legume Futures Report 1.5. Bala A. and Giller K. (2001). Symbiotic specificity of tropical tree rhizobia for host legumes. New Phytologist, 149:495‒507. Bala A., Murphy P. and Giller K. E. (2001). Genetic diversity of rhizobia from natural populations varies with the soil dilution sampled. Soil Biology and Biochemistry, 33:841‒843. Bala A., Murphy P. and Giller K. E. (2002). Occurrence and genetic diversity of rhizobia nodulating Sesbania sesban in African soils. Soil Biology and Biochemistry, 34:1759‒1768. 146 Becerra V. V., Paredes C. M. and Debouck D. (2011). Genetic relationships of common bean (Phaseolus vulgaris L.) race chile with wild Andean and Mesoamerican germplasm. Chilean Journal of Agricultural Research, 71(1):3‒15. Beck D. P., Materon L. A. and Afandi F. (1993). Practical Rhizobium‒legume technology manua; Manual No. 9. International Centre for Agricultural Research in the Dry Areas, Aleppo, Syrian Arab Republic, pp 1‒245. Beebe S. E., Rao M. I., Blair W. M. and Acosta-Gallegos A. J. (2013). Phenotyping common beans for adaptation to drought. Africa Crop Science, 4(35):1‒20. Berrada H. and Fikri‒Benbrahim K. (2014). Taxonomy of the rhizobia: Current perspectives. British Microbiology Research Journal, 4(6):616‒639. Bhargava Y., Murphy J. S. R., Kumar T. V. R. and Rao M. N. (2016). Phenotypic, stress tolerance and plant growth promoting characteristics of rhizobial isolates from selected wild legumes of semiarid region, Tirupati, India. Advances in Microbiology, 6:1‒12. Black M., Moolhuijzen P., Chapman B., Barrero R., Howieson J., Hungria M. and Bellgard M. (2012). The genetics of symbiotic nitrogen fixation: Comparative genomics of 14 rhizobia strains by resolution of protein clusters. Genes, 3:138‒166. Bothe H., Ferguson S. J. and Newton W. E. (Eds.) (2007). Biology of the nitrogen cycle. Elsevier B.V., Amsterdam. Bouzeraa-Bessila M., Beddiar A. and Bouzeraa H. (2015). Effect of abiotic factors on the Sinorhizobium sp. growth isolated from Medicago sativa cultivated on Northeastern Algeria soils and the selection of the highly tolerant strains. Advances Environemental Biology, 9(8):103–115. Boyd E. S. and Peters J. W. (2013). New insights into the evolutionary history of biological nitrogen fixation. Frontiers in Micribiology, 4(201):1–12. Broughton W. J. and Dilworth M. J. (1971). Control of leghaemoglobin synthesis in snake beans. Biochemistry Journal, 125:1075‒1080. Buruchara R., Chirwa R., Sperling L., Mukankusi C., Rubyogo C. J., Muthoni R. and Abang M. M. (2011). Development and delivery of bean varieties in Africa: The Pan-African Bean Research Alliance (PABRA) model. African Crop Science, 29(4):227‒245. Câmara C. R. S., Urrea C. A. and Schlegel V. (2013). Pinto beans (Phaseolus vulgaris L.) as a functional food: Implications on human health. Agriculture, 3:90– 111. 147 Capote N., Pastrana A. M., Aguado A. and Sanchez‒Torres P. (2012). Molecular tools for detection of plant pathogenic fungi and fungicide resistance. In: C. J. Cumagun (Ed.). Plant Pathology, InTech Publishers-Croatia, p 7151. Cárdenas L., Domínguez J., Quinto C., López‒Lara I. M., Lugtenberg B. J., Spaink H. P., Rademaker G. J., Haverkamp J. and Thomas‒Oates J. E. (1995). Isolation, chemical structures and biological activity of the lipo‒chitin oligosaccharide nodulation signals from Rhizobium etli. Plant Molecular Biology, 29:453‒464. Chanway C. P. and Holl F. B. (1986). Suitability of intrinsic antibiotic resistance as a method of strain identification in Rhizobium trifolii. Plant and Soil, 93:287‒291. Checcucci A., Azzarello E., Bazzicalupo M., Galardini M., Lagomarsino A., Mancuso S., Marti L., Marzano M. C., Mocali S., Squartini A., Zanardo M. and Mengoni A. (2016). Mixed nodule infection in Sinorhizobium meliloti–Medicago sativa symbiosis suggests the presence of cheating behavior. Frontiers in Plant Sciences, 7(835):1‒9. Chemining’wa G. N., Theuri S. M. and Muthomi J. W. (2011). The abundance of indigenous rhizobia nodulating cowpea and common bean in Central Kenyan soils. African Journal of Horticultural Science, 5:92‒97. Clarridge J. E. (2004). Impact of 16S rRNA gene sequence analysis for identification of bacteria on clinical microbiology and infectious diseases. Clinical Microbiology Reviews, 17(4):840–862. Cummings S. P., Gyaneshwar P., Vinuesa P., Farruggia F. T., Andrews M., Humphry D., Elliott G. N., Nelson A., Orr C., Pettitt D., Shah G. S., Santos S. R., Krishnan H. B., Odee D., Moreira F. M. S., Sprent J. I., Young J. P. W. and James E. K. (2009). Nodulation of Sesbania species by Rhizobium (Agrobacterium) strain IRBG74 and other rhizobia. Environmental Microbiology, 11:2510‒2525. all’Agnol R. ., Ribeiro R. A, elamuta . R. M, Orme o-Orrillo E., Rogel M. A., Andrade D. S., Mart ne -Romero E. and Hungria M. (2014). Rhizobium paranaense sp. nov., an effective N2-fixing symbiont of common bean (Phaseolus vulgaris L.) with broad geographical distribution in Brazil. International Journal of Systematic and Evolutionary Microbiology, 64:3222–3229. all’Agnol R. ., Ribeiro R. A., Ormenõ‒Orrillo E., Rogel M. A., Marcon . R., elamuta, Sou a A. ., Mart ´ne ‒Romero E. and Hungria M. (2013). Rhizobium freirei sp. nov., a symbiont of Phaseolus vulgaris that is very effective at fixing nitrogen. International Journal of Systematic and Evolutionary Microbiology, 63:4167‒4173. Dashora K. (2011). Nitrogen yielding plants: The pioneers of agriculture with a multipurpose. American‒Eurasian Journal of Agronomy, 4(2):34‒37. 148 Degefu T., Wolde-meskel E. and Frostegard A. (2011). Multilocus sequence analyses reveal several unnamed Mesorhizobium genospecies nodulating Acacia species and Sesbania sesban trees in Southern regions of Ethiopia. Systematic and Applied Microbiology, 34:216–226. Denison R. F. and Kiers E. T. (2004). Life histories of symbiotic rhizobia and mycorrhizal fungi. Current Biology, 21(18):R775‒R785. Dereje T. D., Assefa F., Gebrekidan H. and Keneni G. (2015). Nutritional, ecophysiological and symbiotic characteristics of rhizobia nodulating faba bean (Vicia faba L.) collected from acidic soils of Ethiopia. African Journal of Environmental Science and Technology, 9(7):646–654. Dighe N. S., Shukla D., Kakotwar R. S., Laware R. B., Bhawar S. B. and Gaikwad R. W. (2010). Nitrogenase enzyme: A review. Pelagia Research Library‒Der Pharmacia Sinica, 1(2):77‒84. oignon‒Bourcier ., Willems A., Coopman R., Laguerre G., Gillis M. and de Lajudie P. (2000). Genotypic characterization of Bradyrhizobium strains nodulating small Senegalese legumes by 16S‒23S rRNA intergenic gene spacers and amplified fragment length polymorphism fingerprint analyses. Applied Environmental Microbiology, 66:3987‒3997. Dreyfus B., Garcia J. L. and Gillis M. (1988). Characterization of Azorhizobium caulinodans gen. nov., sp. nov., a stem nodulating nitrogen‒fixing bacterium isolated from Sesbania rostrata. International Journal of Systematic Bacteriology, 38:89‒98. El‒ iki A. A. (2006). Genetic diversity in rhizobia determined by Random Amplified Polymorphic DNA analysis. Journal of Agriculture and Social Sciences, 2(1):1‒4. El‒Maksoud A. H. K. and Keyser H. H. (2010). Restriction specificity of some soybean genotypes to Bradyrhizobium japonicum serogroups. International Journal of Biological, Biomolecular, Agricultural, Food and Biotechnological Engineering, 4(11):824‒827. Erisman J. W., Galloway J. N., Sutton M. A., Klimont Z. and Winiwater W. (2008). How a century of ammonia synthesis changed the world. Nature Geoscience, (1):636‒639. Evans A. M. (1976). Beans Phaseolus spp. (Leguminosae‒Papilionatae). In: N. W. Simmonds (ed.), Evolution of crop plants. Longmans, London, pp 168‒172. Evans D. O. and Macklin B. (1990). Perennial Sesbania species in agroforestry systems. Proceedings of a workshop, Nitrogen Tree Fixing Association, Paia, Hawaii, Nairobi, Kenya. March 27‒31, 1989. 149 Fakruddin M. and Mannan K. S. (2013). Methods for analyzing diversity of microbial communities in natural environments. Ceylon Journal of Science (Biological Sciences), 42(1):19‒33. FAO (2012). State of the world’s forests, p 60. FAO (2013). Statistical database of the Food and Agriculture Organization of the United Nations, Rome, Italy. FAO (2017). Food and agricultural commodities production. Ferguson B. J., Indrasumunar A., Hayashi S., Lin M‒H, Lin Y‒H, Reid D. E. and Gresshoff P. M. (2010). Molecular analysis of legume nodule development and auto-regulation. Journal of Integrative Plant Biology, 52 (1):61–76. Ferguson B. J., Li D., Hastwell A. H., Reid D. E., Li Y., Jackson S. A. and Gresshoff P. M. (2014). The soybean (Glycine max) nodulation‒suppressive CLE peptide, GmRIC1, functions interspecifically in common white bean (Phaseolus vulgaris), but not in a supernodulating line mutated in the receptor PvNARK. Plant Biotechnology Journal, 12:1085‒1097. Figueiredo M. B., Mergulhão A. C. S., Sobral J. K., Lira Junior M. A. and Ferreira de Araújo A. S. (2013). Biological nitrogen fixation: Importance, associated diversity, and estimates. In: N. K. Arora (Ed.). Plant Microbe Symbiosis, Fundamentals and Advances, 10:268‒289. Fischer R. A., Byerlee D. and Edmeades G. O. (2014). Crop yields and global food security: will yield increase continue to feed the world? ACIAR Monograph No. 158. Australian Centre for International Agricultural Research: Canberra. Friesen M. L. and Mathias A. (2010). Mixed infections may promote diversification of mutualistic symbionts: why are there ineffective rhizobia? Journal of Evolutionary Biology, 23:323‒334. Gaur Y. D. and Sen A. N. (1979). Cross inoculation group specificity in Cicer‒Rhizobium symbiosis. New Phytologist, 83:745‒754. Gepts P. (2014). Domestications of Phaseolus beans and their importance for conservation and genetic improvement. In: D. Rubiales (Ed.), A legume perspective, Phaseolus: A new world gift to mankind‒Why common beans are so common? The Journal of the International Legume Society, 2:8‒11. Gicharu G. K., Gitonga N. M., Boga H., Cheruiyot R. C. C. and Maingi J. M. (2013). Variation in nitrogen fixation among three bush bean cultivars grown in Kenya when inoculated with three rhizobia strains. Greener Journal of Agricultural Sciences, 3(11):748‒754. Giller K. E. (2001). Nitrogen fixation in tropical cropping systems (2nd ed.). New York, NY: CABI Publishing, p 264. 150 Gomase P., Anjum S., Shakil S. and Shahnavaj K. M. (2012). Sesbania sesban Linn: A review on its ethnobotany, phytochemical and pharmacological. Profile Asian Journal of Biomedical and Pharmaceutical Sciences, 2(12):11‒14. Gonçalves M. and Moreira F. M. S. (2004). Specificity of the legume Sesbania virgata (Caz.) pers. and its nodule isolates Azorhizobium johannae with other legume hosts and rhizobia. International Symbiosis, 36:57‒68. Goormachtig S., Capoen W. and Holsters M. (2004). Rhizobium infection: Lessons from the versatile nodulation behaviour of water‒tolerant legumes. Trends in Plant Science, 9:518‒522. Graham P. H. (1981). Some problems of nodulation and symbiotic nitrogen fixation in Phaseolus vulgaris L.: A review. Field Crops Research, 4:93–112. Grange L., Hungria M., Graham P. H. and Martínez‒Romero E. (2007). New insights into the origins and evolution of rhizobia that nodulate common bean (Phaseolus vulgaris) in Brazil. Soil Biology and Biochemistry, 39(4):867‒876. Gyaneshwar P., Hirsch A. M., Moulin L., Chen W. M., Elliott G. N., Bontemps C., Santos P. E., Gross E., Reis F. B. Jr., Sprent J. I., Young J. P. W. and Euan K. J. (2011). Legume‒nodulating betaproteobacteria: Diversity, host range, and future prospects. Molecular Plant‒Microbe Interactions, 24(11):1276‒1288. Györgypal Z., Iyer N. and Kondorosi A. (1988). Three regulatory nodD alleles of diverged flavonoid‒specificity are involved in host‒dependent nodulation by Rhizobium meliloti. Molecular and General Genetics, 212:85‒92. Hammer Ø., Harper D. A. T. and Ryan P. D. (2005). PAST‒Palaeontological statistics, Computer software version 1.34. Han S. Z., Wang E. T. and Chen W. X. (2005). Diverse bacteria isolated from root nodules of Phaseolus vulgaris and species within the genera Campylotropis and Cassia grown in China. Systematic and Applied Microbiology, 28(3):265‒276. Hardarson G., Bliss F. A., Cigales-Rivero M. R., Henson R. A., Kipe-Nolt J. A., Longeri L., Manrique A., Peňa-Cabriales J. J., Pereira P. A., Sanabria C. A. and Tsai S. M. (1993). Genotypic variation in biological nitrogen fixation by common bean. Plant and Soil, 152:59–70. Hassen A. I., Bopape F. L. and Trytsman M. (2014). Nodulation study and characterization of rhizobial microsymbionts of forage and pasture legumes in South Africa. World Journal of Agricultural Research, 2(3):93‒100. Heather J. M. and Chain B. (2016). The sequence of sequencers: The history of sequencing DNA. Genomics, 107:1‒8. 151 Helene L. C. F., Delamuta J. R. M., Ribeiro R. A., Ormeño-Orrillo E., Rogel M. A. Martínez-Romero E. and Hungria M. (2015). Bradyrhizobium viridifuturi sp. nov., encompassing nitrogen-fixing symbionts of legumes used for green manure and environmental services. International Journal of Systematics and Evolution of Microbiolology, 65:4441–4448. Heuzé V., Tran G., Bastianelli D., Lebas F. (2015). Sesban (Sesbania sesban). Feedipedia, a programme by INRA, CIRAD, AFZ and FAO. Hewedy O. A., Eissa R. A., Elzanaty A. M., Nagaty H. H. and Abd Elbary M. I. (2014). Phenotypic and genotypic diversity of rhizobia nodulating faba bean from various Egyptian locations. Journal of Bioprospecting and Biotechniques, 4(5):1‒8. Hill G. T., Mitkowski N. A., Aldrich-Wolfe L., Emele L. R., Jurkonie D. D., Ficke A., Maldonado-Ramirez S., Lynch S. T. and Nelson E. B. (2000). Methods for assessing the composition and diversity of soil microbial communities. Applied Soil Ecology, 15:25–36. Hopkins D. W. and Dungait J. A. J. (2010). Soil microbiology and nutrient cycling. In: G. R. Dixon and E. L. Tilston (Eds.), Soil microbiology and sustainable crop production. Springer Science and Business Media B. V., pp 71‒80. Hungria M. and Vargas M. A. (2000). Environmental factors affecting N2 fixation in grain legumes in the tropics, with an emphasis on Brazil. Field Crops Research, 65:151‒164. Hunt S. and. Layzell D. B. (1993). Gas exchange of legume nodules and the regulation of nitrogenase activity. Annual Review of Plant Physiology and Plant Molecular Biology, 44:483‒511. Indrasumunar A., Searle I., Lin M. H., Kereszt A., Men A., Carroll B. J. and Gresshoff P. M. (2011). Nodulation factor receptor kinase 1a controls nodule organ number in soybean (Glycine max L. Merr.). The Plant Journal, 65:39‒50. Jia R. Z., Zhang R. J., Wei Q., Chen W. F., Cho I. K., Chen W. X. and Li Q. X. (2015). Identification and classification of rhizobia by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Journal of Proteomics and Bioinformatics, 8:98–107. Jimenez-Lopez J. C., Gachomo E. W., Sharma S. and Kotchoni S. O. (2013). Genome sequencing and next-generation sequence data analysis: A comprehensive compilation of bioinformatics tools and databases. American Journal of Molecular Biology, 3:115–130. ones E. I., Afkham M. E., Akḉay E., Bronstein . L., Bshary R., rederickson M. E. Heath K. D., Hoeksema J. D., Ness J. H., Pankey M. S., Porter S. S., Sachs J. L., Scharnagl K. and Friesen M. L. (2015). Cheaters must prosper: reconciling theoretical and empirical perspectives on cheating in mutualism. Ecological Letters, 18:1270‒1284. 152 Jordan D. C. (1984). Family III. Rhizobiaceae Conn 1938. In: Krieg N. R., Holt J. C. (Eds.), Bergey’s manual of systematic bacteriology. The Williams and Wilkins Co., Baltimore, pp 234‒236. Kanimozhi K. and Panneerselvam A. (2010). Studies on isolation and nitrogen fixation ability of Azospirillum spp. isolated from Thanjavur district. Der Chemica Sinica, 1(3):138‒145. Kiers E. T., Rousseau R. A., West S. A. and Denison R. F. (2003). Host sanctions and the legume–Rhizobium mutualism. Nature, 425:78‒81. Kondorosi A., Kondorosi E., Pankhurst C. E., Broughton W. J. and Banfalvi Z. (1982). Mobilization of a Rhizobium meliloti megaplasmid carrying nodulation and nitrogen fixation genes into other rhizobia and Agrobacterium. Molecular and General Genetics, 188:433–439. Koskey G., Mburu S. W., Njeru E. M., Kimiti J. M., Ombori O. and Maingi J. M. (2017). Potential of native rhizobia in enhancing nitrogen fixation and yields of climbing beans (Phaseolus vulgaris L.) in contrasting environments of Eastern Kenya. Frontiers in Plant Science, 8(443):1-12. Kumar G. R., Kumar N. V., Mohan M. K., Samal D. and Ghosh P. (2007). Extraction of bulk DNA from Thar desert soils for optimization of PCR‒DGGE base microbial community analysis. Electronic Journal of Biotechnology, 10:400‒408. Kumar S. and Rao B. (2012). Biological nitrogen fixation: A review. International Journal of Advanced Life Sciences, 1:1‒9. Laguerre G., Allard M. R., Revoy F. and Amarger N. (1994). Rapid identification of rhizobia by Restriction Fragment Length Polymorphism analysis of PCR‒Amplified 16S rRNA genes. Applied and Environmental Microbiology, 60(1):56‒63. Laguerre G., Nour S., Macheret V., Sanjuan J., Drouin P. D. and Amarger N. (2001). Classification of rhizobia based on nodC and nifH gene analysis reveals a close phylogenetic relationship among Phaseolus vulgaris symbionts. Microbiology, 147:981‒993. Lakshmanan V., Gopinath Selvaraj G. and Bais H. P. (2014). Functional soil microbiome: Belowground solutions to an aboveground problem. Plant Physiology, 166:689–700. Lalande R., Bigwaneza P. C. and Antoun H. (1990). Symbiotic effectively of strains of Rhizobium leguminosarum biovar phaseoli isolated from soils of Rwanda. Plant Soil, 121:41‒46. Laranjo M., Alexandre A. and Oliveira S. (2014). Legume growth‒promoting rhizobia: An overview on the Mesorhizobium genus. Microbiological Research, 169:2‒17. 153 Lee K., Backer P., Aono T., Liu C., Suzuki S., Suzuki T., Kaneko T., Yamada M., Tabata S., Kupfer D. M., Najar F. Z., Wiley G. B., Roe B., Binnewies T. T., Ussery D. W., D'Haeze W., Herder J. D., Gevers D., Vereecke D., Holsters M. and Oyaizu H. (2008). The genome of the versatile nitrogen fixer Azorhizobium caulinodans ORS571. BMC Genomics, 9:271‒284. Legesse S. and Assefa F. (2014). Symbiotic and phenotypic characteristics of rhizobia nodulating faba bean (Vicia faba) from Tahtay Koraro, Northwestern zone of Tigray regional state, Ethiopia. International Journal of Technology Enhancements and Emerging Engineering Research, 2(11):15–23. Levy S. E. and Myers R. M. (2016). Advancements in Next-Generation Sequencing. Annual Review of Genomics and Human Genetics, 17:95-115. Linton D., Dewhirst F. E., Clewley J. P., Owen R. J., Burnens A. P. and Stanley J. (1994). Two types of 16S rRNA gene are found in Campylobacter helveticus: analysis, applications and characterization of the intervening sequence found in some strains. Microbiobgy, 140:847‒855. Lira M. A., Nascimento L. R. S. and Fracetto G. G. M. (2015). Legume‒rhizobia signal exchange: Promiscuity and environmental effects. Frontiers in Microbiology, 6(945):1‒9. Liu C.-W. and Murray J. D. (2016). The role of flavonoids in nodulation hostrange specificity: An update. Plants, 5(3):33. Liu W., Kohlen W., Lillo A., Op den Camp R., Ivanov S., Hartog M., Limpens E., Jamil M., Smaczniak C., Kaufmann K., Yangb W. C., Hooivelde G. J., Charnikhovac T., Bouwmeesterc H. J., Bisseling T. and Geurts R. (2011). Strigolactone biosynthesis in Medicago truncatula and rice requires the symbiotic GRAS‒type transcription factors NSP1 and NSP2. Plant Cell, 23:3853‒3865. Lopez‒Lopez A., Rogel‒Hernandez M. A., Barois I., Ceballos A. I. O., Martínez J., Orme~no‒Orrillo E. and Martínez‒Romero, E. (2012). Rhizobium grahamii sp. nov., from nodules of Dalea leporina, Leucaena leucocephala and Clitoria ternatea, and Rhizobium mesoamericanum sp. nov., from nodules of Phaseolus vulgaris, siratro, cowpea and Mimosa pudica. International Journal of Systematic and Evolutionary Microbiology, 62:2264‒2271. Lorquin J., Lortet G., Ferro M., Mear N., Dreyfus B., Prome J. C. and Boivin C. (1997). Nod factors from Sinorhizobium saheli and S. teranga bv. sesbaniae are both arabinosylated and fucosylated, a structural feature specific to Sesbania rostrata symbionts. Molecular Plant‒Microbe Interaction, 10:879‒890. Maingi J. M., Shisanya C. A., Gitonga N. M. and Hornetz B. (2001). Nitrogen fixation by common bean (Phaseolus vulgaris L.) in pure and mixed stands in semiarid south-east Kenya. European Journal of Agronomy, 14:1–12. 154 Makatiani E. T. and Odee D. W. (2007). Response of Sesbania sesban (L.) Merr. to rhizobial inoculation in an N‒deficient soil containing low numbers of effective indigenous rhizobia. Agroforest Systems, 70:211‒216. Manceau C. and Horvais A. (1997). Assessment of genetic diversity among strains of Pseudomonas syringae by PCR‒Restriction Fragment Length Polymorphism analysis of rRNA operons with special emphasis on P. syringae pv. tomato. Applied Environmental Microbiology, 63(2):498‒505. Mandal S. M., Chakraborty D. and Dey S. (2010). Phenolic acids act as signaling molecules in plant‒microbe symbioses. Plant Signal Behaviour, 5(4):359‒368. Manna P. K., Kumaran V., Mohanta G. P. and Manavalan R. (2004). Preparation and evaluation of a new erythromycin derivative–erythromycin taurate. Acta Pharmaceuticals, 54:231–242. Marchetti, M., Capela, D., Glew, M., Cruveiller, S., Chane‒Wong‒Ming, B., Gris, C., Timmers, T., Poinsot, V., Gilbert, L. B., Heeb, P., Medigue, C., Batut, J., and Masson‒Boivin, C. (2010). Experimental evolution of a plant pathogen into a legume symbiont. PLOS Biology, 8(1):e1000280. Marinkovic ., ordevic V., Baleševic‒Tubic S., Bjelic D., Vucelic‒Radovic B. and ošic . (2013). Osmotic stress tolerance, PGP traits and RAPD analysis of Bradyrhizobium japonicum strains. Genetika, 45(1):75‒86. Maróti G. and Kondorosi É. (2014). Nitrogen‒fixing Rhizobium‒legume symbiosis: are polyploidy and host peptide-governed symbiont differentiation general principles of endosymbionsis? Frontiers in Microbiology, 5(326):1‒6. Massawe P. I., Mtei K. M., Munishi L. K. and Ndakidemi P. A. (2016). Existing practices for soil fertility management through cereals-legume intercropping systems. World Research Journal of Agricultural Sciences, 3(2):080‒091. Maunoury N., Redondo‒Nieto M., Bourcy M., Van de Velde W., Alunni B., Laporte P., Durand P., Agier N., Marisa L., Vaubert D., Delacroix H., Duc G., Ratet P., Aggerbeck L., Kondorosi E. and Mergaert P. (2010). Differentiation of symbiotic cells and endosymbionts in Medicago truncatula nodulation are coupled to two transcriptome‒switches. PLOS One, 5(3):e9519. Maxam A. M. and Gilbert W. (1977). A new method for sequencing DNA. Proceedings of the National Academy of Sciences, U.S.A., 74 (2):560–564. McInroy S. G., Campbell C. D., Haukka K. E., Odee D. W., Sprent J. I., Wang W. J., Young J. P. W. and Sutherland J. M. (1999). Characterization of rhizobia from African acacias and other tropical woody legumes using BiologTM and partial 16S rRNA sequencing. FEMS Microbiology Letters, 170(1):111‒117. 155 Melino V. J., Drew E. A, Ballard R. A., Reeve W. G., Thomson G., White R. G. and O’Hara G. W. (2012). Identifying abnormalities in symbiotic development between Trifolium spp. and Rhizobium leguminosarum bv. trifolii leading to sub‒optimal and ineffective nodule phenotypes. Annals of Botany, pp 1‒14. Menna P., Hungria M., Barcellos F. G., Bangel E. V., Hess P. N. and Martínez‒Romero E. (2006). Molecular phylogeny based on the 16S rRNA gene of elite rhizobial strains used in Brazilian commercial inoculants. Systematic and Applied Microbiology, 29(4):315‒332. Messaoud B. B., Aboumerieme I., Nassiri L., El Fahime E. and Ibijbijen J. (2014). Phenotypic and genotypic characteristics of rhizobia isolated from Meknes‒Tafilalet soils and study of their ability to nodulate Bituminaria bituminosa. British Microbiology Research Journal, 4(4):405‒417. Mia M. A. B. and Shamsuddin Z. H. (2010). Rhizobium as a crop enhancer and biofertilizer for increased cereal production. African Journal of Biotechnology‒Academic Journals, 9(37):6001‒6009. Michiels J., Dombrecht B., Vermeiren N., Xi C., Luyten E. and Vanderleyden J. (1998). Phaseolus vulgaris is a non-selective host for nodulation. Microbiology Ecology, 26:193‒205. Ministry of Agriculture, Livestock and Fisheries (2015). Economic review of agriculture [ERA] 2015 in Kenya, ninth edition. Report, pp 1‒12. Mishra V., Kumar D., Srivastava V. K., Gautam A., Shukla S., Verma K. R. and Verma A. R. (2016). Identification of unknown bacteria by using biochemical tests and 16S rRNA sequencing from different soil samples. International Journal of Advance Research in Science and Engineering, 5(4):234–249. Mnasri B., Liu T. Y., Saidi S., Chen W. F., Chen W. X., Zhang X. X. and Mhamdi R. (2014). Rhizobium azibense sp. nov., a nitrogen fixing bacterium isolated from root nodules of Phaseolus vulgaris. International Journal of Systematic and Evolutionary Microbiology, 64:1501–1506. Mnasri B., Saïdi S., Chihaoui S. and Mhamdi R. (2012). Sinorhizobium americanum symbiovar mediterranense is a predominant symbiont that nodulates and fixes nitrogen with common bean (Phaseolus vulgaris L.) in a Northern Tunisian field. Systematic and Applied Microbiology journal, 35:263‒269. Moreira F. M. S., Haukka K. and Young J. P. W. (1998). Biodiversity of rhizobia isolated from a wide range of forest legumes in Brazil. Molecular Ecology, 7:889‒895. Morosan M., Hassan A. M., Naranjo M. A., López-Gresa M. P., Boscaiu M. Vicente O. (2017). Comparative analysis of drought responses in Phaseolus vulgaris (common bean) and P. coccineus (runner bean) cultivars. The EuroBiotech Journal, 1(3):1–6. 156 Mpepereki S., Makonese S. and Wollum A. G. (1997). Physiological characterization of indigenous rhizobia nodulating Vigna unguiculata in Zimbabwean soils. Symbiosis, 22:275–292. Msimbira L. A., Jaiswal S. K. and Dakora F. D. (2016). Identification and characterization of phages parasitic on Bradyrhizobia nodulating groundnut (Arachis hypogaea L.) in South Africa. Applied Soil Ecology, 108:334–340. Mungai N. W. and Karubiu N. M. (2011). Effectiveness of rhizobia isolates from Njoro soils (Kenya) and commercial inoculants in nodulation of common beans (Phaseolus vulgaris). Journal of Agriculture, Science and Technology, 12(1):47–59. Murray A. E., Lies D., Li G., Nealson K., Zhou J. and Tiedje J. M. (2001). DNA‒DNA hybridization to microarrays reveals gene‒specific differences between closely related microbial genomes. Proceedings of the National Academy of Sciences of the United States of America, 98(17):9853‒9858. Mus F., Crook M. B., Garcia K., Garcia Costas A., Geddes B. A., Kouri E. D., Paramasivan P., Ryu M. H., Oldroyd G. E. D., Poole P. S., Udvardi M. K., Voigt C. A., Ané J. M. and Peters J. W. (2016). Symbiotic nitrogen fixation and the challenges to its extension to non‒legumes. Applied Environmental Microbiology, 82(13):3698‒3710. Mwenda G. M., Karanja N. K., Boga H., Kahindi J. H. P., Muigai A. and Odee D. (2011). Abundance and diversity of legume nodulating rhizobia in soils of Embu district, Kenya. Tropical and Subtropical Agroecosystems, 13:1‒10. Nahar N., Rashid R. B., Begum A. and Akhter H. (2017). Isolation, identification and molecular characterization of Rhizobium species from Sesbania bispinosa cultivated in Bangladesh. African Journal of Agricultural Research, 12(22):18741880. Namkeleja Y., Mtei K. and Ndakidemi P. A. (2016). Isolation and molecular characterization of elite indigenous rhizobia nodulating Phaseolus bean (Phaseolus vulgaris L.). American Journal of Plant Sciences, 7:1905‒1920. Nguyeni M. T., Akiyoshi K., Nakatsukasa M., Siaeki Y. and Yokoyama K. (2010). Development of multiple occupancy nodules in correlation of the density dependent infection by soybean nodulating rhizobia. Soil Microorganisms 64(2):101‒106. Noisangiam R., Teamtisong K., Tittabutr P., Boonkerd N., Toshiki U., Minamisawa K. and Teaumroong N. (2012). Genetic diversity, symbiotic evolution, and proposed infection process of Bradyrhizobium strains isolated from root nodules of Aeschynomene americana L. in Thailand. Applied and Environmental Microbiology, 78(17):6236‒6250. 157 Odee D. W., Haukka K., McInroy S. G., Sprent J. I., Sutherland J. M. and Young J. P. W. (2002). Genetic and symbiotic characterization of rhizobia isolated from tree and herbaceous legumes grown in soils from ecologically diverse sites in Kenya. Soil Biology and Biochemistry, 34:801‒811. Odee D. W. and Sprent J. I. (1992). Acacia brevispica: A non‒nodulating mimosoid legume? Soil Biology and Biochemistry, 24:717‒719. Odee D. W., Sutherland J. M., Makatiani E. T., McInroy S. G. and Sprent J. I. (1997). Phenotypic characteristics and composition of rhizobia associated with woody legumes growing in diverse Kenyan conditions. Plant and Soil, 188:65‒75. Oger P., Dessaux Y., Petit A., Gardan L., Manceau C., Chomel C. and Nesme X. (1998). Validity, sensitivity and resolution limit of the PCR‒RFLP analysis of the rrs (16S rRNA gene) as a tool to identify soil‒borne and plant‒associated bacterial populations. Genetics Selection Evolution, 30(1):311‒321. Okalebo J. R., Gathua K. W. and Woomer P. L. (2002). Laboratory methods of soil and plant analysis: A working manual, KARI, SSSEA, TSBF, SACRED Africa, Moi University. Second Edition, Nairobi, Kenya, pp 128. Okaron V., Were B. A. and Nyongesa B. O. (2017). Phenotyping groundnut rhizobia native to phosphorus deficient soils in Western Kenya. American Journal of Microbiological Research, 5(1):15-24. Okazaki S., Okabe S., Higashi M., Shimoda Y., Sato S., Tabata S., Hashiguchi M., Akashi R., Göttfert M. and Saeki K. (2010). Identification and functional analysis of type III effector proteins in Mesorhizobium loti. Molecular Plant‒Microbe Interactions, 23:223‒234. O'Leary M., Rehm G. and Schmitt M. (1989). Understanding nitrogen in soils. Minnesota Extension service, AG‒FO‒3770, pp 1‒ 4. Peakall R. and Smouse P. E. (2006). GenAlEx 6: Genetic analysis in Excel. Population genetic software for teaching and research. Peakall R. and Smouse P. E. (2012). GenAlEx 6.5: Genetic analysis in Excel. Population genetic software for teaching and research‒an update. Bioinformatics, 28:2537‒2539. Peix A., Martha H. R., Velázquez E. and Bedmar E. J. (2015). Bacterial associations with legumes. Critical Reviews in Plant Sciences, 34:1–3, 17–42. Perret X., Staehelin C. and Broughton W. J. (2000). Molecular basis of symbiotic promiscuity. Microbiology and Molecular Biology Reviews, 64:180‒201. Petti C. A., Polage C. R. and Schreckenberger P. (2005). The role of 16S rRNA gene sequencing in identification of microorganisms misidentified by conventional methods. Journal of Clinical Microbiology, 43(12):6123‒6125. 158 Pierre H. and Didier R. (2002). Traditional and molecular technique for the study of emerging bacterial diseases: One laboratory’s perspective, 8(2):122‒131. Pommeresche R. and Hansen S. (2017). Examining root nodule activity on legumes. FertilCrop Technical Note, pp 1‒4. Porch T. G., Beaver S. J., Debouck G. D., Jackson A. S., Kelly D. J. and Dempewolf H. (2013). Use of wild relatives and closely related species to adapt common bean to climate change. Agronomy, 3:433‒462. Ramaekers L., Micheni A., Mbogo P., Vanderleyden J. and Maertens M. (2013). Adoption of climbing beans in the central highlands of Kenya: An empirical analysis of farmers’ adoption decisions. African Journal of Agricultural Research, 8(1):1‒19. Ribeiro R. A., Ormeno‒Orrillo E., Fuzinatto D. R., Graham P. H., Martinez‒Romero E., Hungria M. and Soja E. (2013). Novel Rhizobium lineages isolated from root nodules of the common bean (Phaseolus vulgaris L.) in Andean and Mesoamerican areas. Research in Microbiology, 164(7):740‒748. Rizzo J. M. and Buck M. J. (2012). Key principles and clinical applications of "Next-Generation" DNA sequencing. Cancer Prevention Research, 5(7):887–900. Romdhane S. B., Nasr H., Samba‒Mbaye R., Neyra M., Ghorbal M. H. and De Lajudie P. (2005). Genetic diversity of Acacia tortilis ssp. raddiana rhizobia in Tunisia assessed by 16S and 16S‒23S rDNA genes analysis. Journal of Applied Microbiology, 100:436‒445. Romero-Arenas O., Huato D. M. A., Tapia R. J. A., Báez S. B. A., Huerta L. M. and Huerta C. E. (2013). The nutritional value of beans (Phaseolus vulgaris L.) and its importance for feeding of rural communities in Puebla-Mexico. International Research Journal of Biological Sciences, 2(8):59‒65. Sachs J. L., Russell J. E., Lii Y. E., Black K. C., Lopez G. and Patil A. S. (2010). Host control over infection and proliferation of a cheater symbiont. Journal of Evolutionary Biology, 23:1919–1927. Saeki K. (2011). Rhizobial measures to evade host defense strategies and endogenous threats to persistent symbiotic nitrogen fixation: a focus on two legume‒Rhizobium model systems. Cell Molecular of Life Sciences, 68(8):1327‒1339. Saikia S. P. and Jain V. (2007). Biological nitrogen fixation with non‒legumes: An achievable target or a dogma? Review Article. Current Science, 92(3):317‒322. Sanger F. and Coulson A. R. (1975). A rapid method for determining sequences in DNA by primed synthesis with DNA polymerase. Journal of Molecular Biology, 94(3):441‒448. 159 Santamaría R. I., Bustos P., Sepúlveda-Robles O, Luis Lozano L., César Rodríguez C., Fernández J. L., Juárez S., Kameyama L., Guarneros G., Dávila G. and Víctor González V. (2013). Narrow-host-range bacteriophages that infect Rhizobium etli associate with distinct genomic types. Applied and Environmental Microbiology, 80:446–454. Santi C., Bogusz D. and Franche C. (2013). Biological nitrogen fixation in non‒ legume plants. Annals of Botany, 111:743–767. Savci S. (2012). An agricultural pollutant: Chemical fertilizer. International Journal of Environmental Science and Development, 3(1):77‒80. Schulten H. R. and Schnitzer M. (1998). The chemistry of soil organic nitrogen: a review. Biology and Fertility of Soils, 26:1–15. Schumpp O. and Deakin W. J. (2010). How inefficient rhizobia prolong their existence within nodules. Trends Plant in Sciences, 15:189‒195. Sharma R. S., Mohmmed A., Mishra V. and Babu C. R. (2005). Diversity in a promiscuous group of rhizobia from three Sesbania spp. colonizing ecologically distinct habitats of the semi‒arid Delhi region. Research in Microbiology, 156:57‒67. Shetta N. D., Al-Shaharani T. S. and Abdel-Aal M. (2011). Identification and characterization of Rhizobium associated with woody legume trees grown under Saudi Arabia condition. American-Eurasia Journal of Agriculture and Environmental Sciences, 10:410‒418. Sheu S. Y., Chen Z. H., Young C. C. and Chen W. M. (2016). Rhizobium ipomoeae sp. nov., isolated from a water convolvulus field. International Journal of Systematic and Evolutionary Microbiology, 66:1633–1640. Shtark O., Provorov N., Mikić A., Borisov A., Ćupina B. and Tikhonovich I. (2011). Legume root symbioses: Natural history and prospects for improvement. Field Vegetation and Crop Resesrch, 48:291‒304. Simsek S., Ojanen‒Reuhs T., Stephens S. B. and Reuhs B. L. (2007). Strain‒ecotype specificity in Sinorhizobium meliloti‒Medicago truncatula symbiosis is correlated to succinoglycan oligosaccharide structure. Journal of Bacteriology, 189(21):7733‒7740. Singh N., Luthra U. and Desai N. (2013). Phenotypic and genotypic characterization of Rhizobium species isolated from the root nodules of Sesbania sesban found in Mumbai and its suburban areas. Biotechnology, 3(7):60‒67. 160 Smalla K., Oros-Sichler M., Milling A., Heuer H., Baumgarte S., Becker R., Neuber G. Kropf S., Ulrich A. and Tebbe C. C. (2007). Bacterial diversity of soils assessed by DGGE, T-RFLP and SSCP fingerprints of PCR-amplified 16S rRNA gene fragments: Do the different methods provide similar results? Journal of Microbiological Methods, 69:470–479. Somasegaran P. and Hoben H. (1994). Methods in Legume‒Rhizobium technology. NIFTAL project. University of Hawaii, Paia, Honolulu. Spratt D. A. (2004). Significance of bacterial identification by molecular biology methods. Endodontic Topics, 9:5‒14. Sprent J. I. (2001). Nodulation in legumes. Royal Botanic Gardens: Kew, UK. Stewart W. M. and Roberts T. L. (2012). Food security and the role of fertilizer in supporting it. Procedia Engineering, 46:76–82. Subramanian S. (2013). Distinct hormone regulation of determinate and indeterminate nodule development in legumes. Journal of Plant Biochemistry and Physiology, 1(3):1–5. Swain H. and Abhijita S. (2013). Nitrogen fixation and its improvement through genetic engineering. Journal of Global Biosciences, 2(5):98‒112. Syed F. A., Rawat L. S., Meghvansi M. K. and Mahna S. K. (2010). Determination of the host range of tree and herb rhizobia for their alternative legumes. Journal of Phytology, 2(7):101‒106. Tamura K., Dudley J., Nei M. and Kumar S. (2007). MEGA 4: Molecular evolutionary genetics analysis (MEGA) software version 4.0. Molecular Biology and Evolution, 24:1596‒1599. Tan Z. Y., Xu X. ., Wang E. T., Gao . L., Mart ´ne -Romero, E. and Chen, W. X. (1997). Phylogenetic and genetic relationships of Mesorhizobium tianshanense and related rhizobia. International Journal of Systematic Bacteriology, 47:874–879. Terefework Z., Kaijalainen S. and Lindstro M. K. (2001). AFLP fingerprint as a tool to study the genetic diversity of Rhizobium galegae isolated from Galega orientalis and Galega officinalis. Journal Biotechnology, 91:169–180. Toolarood A. S., Alikhani H. A., Salehi G., Asadi‒Rahmani H., Khavazi K., Poorbabaee A. A. and Lindström K. (2012). Molecular diversity of rhizobia isolated from root nodules of alfalfa evaluated by analysis of IGS and 16S rRNA. Annals of Biological Research, 3(5):2058‒2063. Turk D. and Keyser H. H. (1992). Rhizobia that nodulate tree legume: specificity of the host for nodulation and effectiveness. Canadian Journal of Microbiology, 38:45‒460. 161 United Nations (2013). World population prospects social affairs economic and the 2012 revision. Department of economic and social affairs, population division. New York, p 118. Vinay O., Bhupendra P. and Kiran S. (2013). 16S rDNA-RFLP analysis of phylogenetic tree of Rhizobium bacteria. Indian Journal of Applied Research, 3(12):474‒476. Vinuesa P., Silva C., Lorite M. J., Izaguirre‒Mayoral M. L., Bedmar E. J. and Martínez‒Romero E. (2005). Molecular systematics of rhizobia based on maximum likelihood and Bayesian phylogenies inferred from rrs, atpD, recA and nifH sequences, and their use in the classification of Sesbania microsymbionts from Venezuelan wetlands. Systematics of Applied Microbiology, 28:702‒716. VSN International (2012). Genstat for Windows 14th Edition. VSN International, Hemel Hempstead, UK. Wagner S. C. (2012). Biological nitrogen fixation. Nature Education and Knowledge, 3(10):15. Wang D., Yang S., Tang F. and Zhu H. (2012). Symbiosis specificity in the legume–rhizobial mutualism. Cellular Microbiology, 14(3):334–342. Wang E. T., Kan F. L., Tan Z. Y., Toledo I., Chen W. X., Martinez‒Romero E. (2003). Diverse Mesorhizobium plurifarium populations native to Mexican soils. Archival Microbiology, 180:444‒454. Wang E. T. and Martinez‒Romero E. (2000). Sesbania herbacea‒Rhizobium huautlense nodulation in flooded soils and comparative characterization of S. herbacea‒nodulating rhizobia in different environments. Microbial Ecology, 40:25‒32. Wang E. T., van Berkum P., Beyene D., Sui X. H., Dorado O., Chen W. X. and Martinez‒Romero E. (1998). Rhizobium huautlense sp. nov., a symbiont of Sesbania herbacea that has a close phylogenetic relationship with Rhizobium galegae. International Journal of Systematic Bacteriology, 48:687‒699. Weisburg W. G., Barns S. M., Pelletier D. A. and Lane D. J. (1991). 16S Ribosomal DNA amplification for phylogenetic study. Journal of Bacteriology, 173:697‒703. Willems A. (2006). The taxonomy of rhizobia: An overview. Plant and Soil, 287:3‒14. Wolde-Meskel E., Dagne E. D. and Haile W. (2016). Response of sesbania (Sesbania sesban L. Merr.) to inoculation with indigenous isolates of Rhizobium strains. African Journal of Plant Science, 10(7):136–144. 162 Wolde‒Meskel E., Terefework Z., Frostegård A. and Lindström K. (2005). Genetic diversity and phylogeny of rhizobia isolated from agroforestry legume species in Southern Ethiopia. International Journal of Systematic and Evolutionary Microbiology, 55:143‒145. Yaman M. and Cinsoy A. S. (1996). Determination of the most effective Rhizobium strain (Rhizobium japonicum L.) in soybean. Journal Aegean Agricultural Research Institute, 6:84‒96. Yeshaneh G. T. (2015). Assessment of micronutrient status in different land use soils in Maybar lake watershed of Albuko District, South Wello Zone, North Ethiopia. Global Journal of Crop, Soil Science and Plant Breeding, 3(1):143‒150. Young . M., Kuykendall L. ., Mart ne -Romero E., Kerr A. and Sawada H. (2001). A revision of Rhizobium Frank 1889, with an emended description of the genus, and the inclusion of all species of Agrobacterium Conn 1942 and Allorhizobium undicola de Lajudie et al. 1998 as new combinations: Rhizobium radiobacter, R. rhizogenes, R. rubi, R. undicola and R. vitis. International Journal of Systematic and Evolutionary Microbiology, 51:89–103. Young J. P. W. and Haukka K. E. (1996). Diversity and phylogeny of rhizobia. New Phytologist, 133:87–94. Zahran H. H. (1999). Rhizobium‒Legume symbiosis and nitrogen fixation under severe conditions and in an arid climate. Microbiology and Molecular Biology Reviews, 63(4):968‒989. Zahran H. H. (2001). Rhizobia from wild legumes: Diversity, taxonomy, ecology, nitrogen fixation and biotechnology. Journal of Biotechnology, 91:143‒153. Zakhia F., Jeder H., Domergue O., Willems A., Cleyet‒Marel J. C., Gillis M., Dreyfus B., de Lajudie P. (2004). Characterization of wild legume nodulating bacteria (LNB) in the infra‒arid zone of Tunisia. Systematic of Applied Microbiology, 27:380‒395. Zhao L., Fan M., Zhang D., Yang R., Zhang F., Xu L., Wei X., Shen Y. and, Wei G. (2014). Distribution and diversity of rhizobia associated with wild soybean (Glycine soja Sieb. and Zucc.) in Northwest China. Systematic Applied Microbiology, 37:449–456. Zurdo-Piñeiro J. L., García‒ raile P., Rivas R., Peix, A., Le_on‒Barrios M., Willems A., Mateos P. ., Martíne ‒Molina E., Vel‒a que E. and Berkum P. (2009). Rhizobia from Lanzarote, the Canary Islands, that nodulate Phaseolus vulgaris have characteristics in common with Sinorhizobium meliloti isolates from mainland Spain. Applied and Environmental Microbiology, 75:2354‒2359. 163 Zurdo-Piñeiro J. L., Velazquez E., Lorite M. J., Brelles‒Marino G., Schroder E. C., Bedmar E. J., Mateos P. F. and Martinez‒Molina E. (2004). Identification of fast‒growing rhizobia nodulating tropical legumes from Puerto Rico as Rhizobium gallicum and Rhizobium tropici. Systematic Applied Microbiology, 27:469‒477. 164 APPENDICES Appendix I: Growth characteristics of Sesbania sesban rhizobial isolates from Kenya on YEMA media Isolate/Host Characteristics per study site BTB Gituamba 110 S. sesban 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 5 mm diameter, milky, gummy, dome shaped shinny A 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 2 mm, milky suspensions, shiny, raised and watery N <1 mm diameter, transparent, dome shaped, tiny B 2 mm diameter, milky suspensions, shiny, raised and watery B 130 S. sesban <1 mm diameter, transparent, dome shaped, tiny B 131 S. sesban 2 mm diameter, milky suspensions, shiny, raised and watery N 132 S. sesban <1 mm diameter, transparent, dome shaped, tiny N 133 S. sesban 2 mm diameter, milky suspensions, shiny, raised and watery N 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides N 134b S. sesban <1 mm diameter, transparent, dome shaped, tiny N 135 S. sesban <1 mm diameter, transparent, dome shaped, tiny N 136 S. sesban <1 mm diameter, transparent, dome shaped, tiny N 137a S. sesban <1 mm diameter, transparent, dome shaped, tiny N 111 S. sesban 112 S. sesban 113 S. sesban 114 S. sesban 115 S. sesban 116 S. sesban 117 S. sesban Kavutiri 129 S. sesban 134a S. sesban 165 137b S. sesban 2 mm diameter, milky suspensions, shiny, raised and watery N 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides N 157 S. sesban <1 mm diameter, transparent, dome shaped, tiny B 158 S. sesban 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides N 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides N 2 mm diameter, milky suspensions, shiny, raised and watery A 2 mm diameter, milky suspensions, shiny, raised and watery A 163 S. sesban <1 mm diameter, transparent, dome shaped, tiny B 164 S. sesban <1 mm diameter, transparent, dome shaped, tiny B 165 S. sesban 2 mm diameter, milky suspensions, shiny, raised and watery B 166 S. sesban <1 mm diameter, transparent, dome shaped, tiny B 167 S. sesban 2 mm diameter, milky suspensions, shiny, raised and watery A 168 S. sesban <1 mm diameter, transparent, dome shaped, tiny A 176 S. sesban 2 mm diameter, milky suspensions, shiny, raised and watery A 2 mm diameter, milky suspensions, shiny, raised and watery A <1 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides B 179 S. sesban <1 mm, transparent, dome shaped, tiny B 180 S. sesban <1 mm diameter, transparent, dome shaped, tiny B 181 S. sesban 2 mm diameter, milky suspensions, shiny, raised and watery N 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides N 3 mm diameter, pink, translucent, milky centre, A Bumala 156 S. sesban 159 S. sesban 161 S. sesban 162 S. sesban 177 S. sesban 178 S. sesban Kuinet 169 S. sesban 170 S. sesban 166 dome, shinny exoplysaccharides 171 S. sesban 172 S. sesban and moderate gummy 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides N 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 173 S. sesban 3 mm diameter, translucent, raised shinny, watery A 174 S. sesban 3 mm diameter, translucent, raised shinny, watery N BTB, Bromothymol blue; A, acid; B, alkaline; N, neutral. Appendix II: Growth characteristics of Sesbania sesban rhizobial isolates Uganda on YEMA media Isolate/Host Characteristics per study site BTB Tororo MASS47 S. sesban MASS48 S. sesban MASS49 S. sesban MASS50 S. sesban MASS51 S. sesban MASS52 S. sesban MASS53 S. sesban 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 4 mm, milky suspensions, shiny, raised and watery A 2 mm diameter, milky suspensions, shiny, raised and watery A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 1.0 mm diameter, milky opaque, dome, shiny, and no exopolysaccharides A 2 mm diameter, milky suspensions, shiny, raised and watery A 2.5 mm diameter, milky suspensions, shiny, raised and watery A 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A Mbale MASS54 S. sesban MASS55 S. sesban 167 MASS56 S. sesban MASS57 S. sesban MASS59 S. sesban MASS60 S. sesban MASS61 S. sesban 1.0 mm diameter., milky opaque, dome, shiny, and no exopolysaccharides A 2.5 mm diameter, milky suspensions, shiny, raised and watery A 1.0 mm diameter, milky opaque, dome, shiny, and no exopolysaccharides A 2 mm diameter, milky suspensions, shiny, raised and watery A 2 mm diameter, milky suspensions, shiny, raised and watery B 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A <1.0 mm diameter, transparent, dome shaped, tiny B 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A <1.0 mm diameter, transparent, dome shaped, tiny A <1.0 mm diameter, transparent, dome shaped, tiny A 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A Kabale 62 S. sesban 63 S. sesban 64 S. sesban 65 S. sesban 66 S. sesban 67 S. sesban 68 S. sesban 69 S. sesban BTB, Bromothymol blue; A, acid; B, alkaline. 168 Appendix III: Growth characteristics of Sesbania sesban rhizobial isolates from Tanzania on YEMA media Isolate/Host Characteristics per study site BTB SUA MASS41 S. sesban MASS42 S. sesban MASS43 S. sesban MASS44 S. sesban MASS45 S. sesban MASS46 S. sesban <1.0 mm diameter, transparent, dome shaped, tiny A <1.0 mm diameter, transparent, dome shaped, tiny A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 2 mm diameter, milky suspensions, shiny, raised and watery A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A <1.0 mm diameter, pink, translucent, flat, dull and dry, no exopolysaccharides B 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 2 mm diameter, milky suspensions, shiny, raised and watery A <1.0 mm diameter, pink, translucent, flat, dull and dry, no exopolysaccharides B 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 2 mm diameter, milky suspensions, shiny, raised and watery A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 1.0 mm diameter, milky opaque, dome, shiny, and no exopolysaccharides B Lushoto MASS29 S. sesban MASS30 S. sesban MASS31a S. sesban MASS31b S. sesban MASS32 S. sesban MASS33 S. sesban MASS34 S. sesban MASS35 S. sesban MASS36 S. sesban 169 MASS37 S. sesban MASS38 S. sesban MASS39 S. sesban MASS40 S. sesban 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 4 mm diameter, red, opaque, shiny, raised and moderate gummy exopolysaccharides A 2 mm diameter, milky suspensions, shiny, raised and watery A 4 mm diameter, red, opaque, shiny, raised and moderate gummy exopolysaccharides A BTB, Bromothymol blue; A, acid Appendix IV: Growth characteristics of sesbania rhizobial isolates from Namibia YEMA media Isolate/Host Characteristics per study site BTB Namibia 1 S. macowaniana 2 mm diameter, pink translucent, flat, dull A 2 S. sphaerosperma 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides N 4 mm diameter, transparent, margin brown centre, raised, gummy, mucoid, shinny A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 9 S. pachycarpa 0.5 mm diameter, red, raised dry, shinny A 10 S. sphaerosperma 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides N <1 mm diameter, pink, translucent, flat, dull and dry, no exopolysaccharides B 1 mm diameter., milky opaque, dome, shiny, and no exopolysaccharides N 4 mm diameter, transparent, raised, watery mucous, banana smell A <1 mm diameter, pink, translucent, flat, dull and dry, no exopolysaccharides N 16 S. pachycarpa <1 mm diameter, clear, raised, dry, shinny A 17 S. bispinosa 0.5 mm, pink, translucent, raised, dull, dry B 18 S. sesban 4 mm diameter, transparent, shiny, dome and copius viscous EPS A 3 mm diameter, pink translucent, milky, raised, mucoid, shinny A 4 S. pachycarpa 8 S. sesban 11 S. bispinosa 12 S. sphaerosperma 13 S. sesban 15 S. macowaniana 19 S. macowaniana 170 20 S. cinerascens 1 mm diameter., milky opaque, dome, shiny, and no exopolysaccharides A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A <1 mm diameter, pink, translucent, flat, dull and dry, no exopolysaccharides A 25 S. macowaniana 0.5 mm diameter, red, flat, dry, shinny A 26 S. macowaniana 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 27 S. macowaniana <1.0 mm diameter, red, flat, dry, shinny A 28 S. cinerascens 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A <1.0 mm diameter, pink, translucent, flat, dull and dry, no exopolysaccharides B 4 mm diameter, red, opaque, shiny, raised and moderate gummy exopolysaccharides B 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 38 S. sesban 2 mm diameter, pink translucent, flat, dull, dry A 39 S. macowaniana 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides N 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 42 S. macowaniana 3 mm diameter, milky, translucent, flat, dry A 43 S. macowaniana 2 mm diameter, milky, translucent, flat, dry A 44 S. pachycarpa 2 mm diameter, pink opaque, raised, dry, dull A 45 S. pachycarpa 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 4 mm diameter, red, opaque, shiny, raised and moderate gummy exopolysaccharides N 21 S. sesban 22 S. macowaniana 24 S. macowaniana 31 S. cinerascens 34 S. sphaerosperma 35 S. sesban 36 S. pachycarpa 37 S. cinerascens 40 S. cinerascens 41 S. sphaerosperma 49 S. pachycarpa 171 50 S. rostrata 51 S. pachycarpa 56 S. pachycarpa 57 S. sesban 58 S. pachycarpa 59 S. pachycarpa 60 S. pachycarpa 62 S. pachycarpa 68 S. sesban 69 S. microphylla 70 S. microphylla 71 S. macowaniana 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides B 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides N <1.0 mm diameter, pink, translucent, flat, dull and dry, no exopolysaccharides N 4 mm diameter, red, opaque, shiny, raised and moderate gummy exopolysaccharides A 4 mm diameter, red, opaque, shiny, raised and moderate gummy exopolysaccharides A 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides A 3 mm diameter, pink, translucent, milky centre, dome, shinny and moderate gummy exoplysaccharides N 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides N 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides B 4 mm diameter, transparent, shiny, dome and copious watery exopolysaccharides B 5 mm diameter, milky, translucent, shiny, dome and copious friable exopolysaccharides B BTB, Bromothymol blue; A, acid; B, alkaline; N, neutral.