INDIGENOUS KNOWLEDGE AND EFFECTS OF . , INTEGRATED SOIL FERTILITY MANAGEMENT ON GROWTH, GRAIN YIELD AND RHIZOBIA GENETICS OF SELECTED COWPEA VARIETIES I( JACINT A MALIAIKIMITI Reg. No. 184/7687/00 A Thesis Submitted in Partial Fulfillment of the Requirement for the Award of the Degree of Doctor of Philosophy in the School of Pure and Applied Sciences of Kenyatta University Department of Plant and Microbial Sciences December 2008 ©Tl'lr.intl'l Malia Kimiti Kamiti, Jacinta Malia Indigenous know/edge and effects of 111111111111111 2009/336643 u E LI HARY1' DECLARATION This thesis is my original work and has not been presented for a degree in any other university or any other award JACINTA MALIA KIMITI Reg No. 184/7687/000 oR d--\ ~ Signature DECLARATIO Date BY SUPERVISORS We confirm that the work reported in this thesis was carried out by the candidate under our supervision Dr. Gitonga N. M. Department of Plant and Microbial Sciences, Kenyatta University P.O Box 43844, Nairobi 11 6 J 0 i Signature D~te JwcJI I Dr. Odee D. W. KEFRI, P.O. Box 20412-00200 Nairobi. , Signature Date Dr. Vanlauwe B. TSBF·CIAT, Clo ICRAF P.O Box 30677, Nairobi. ~~ Signature ~/~/~1 Date 11 DEDICATION This thesis is dedicated to Theresia Mumbua III ACKNOWLEDGEME TS I am grateful for financial support from my employer, Kenya Forestry Research Institute (KEFRI) without which this study would have been impossible. I also thank KEFRI for granting me a study leave so as to undertake this study. I greatly appreciate financial support from The Directorate of Personnel Management (DPM) that paid for my tuition fees at Kenyatta University. I am indebted to AfNet (TSBF/CIAT) through Desert Margins Project (DMP) for the financial support that helped me to start off my PhD. research activities. I am very grateful to Dr. Andre' Bationo who ensured that I got some seed money from AfNet to start off my research work. I am grateful for the assistance from DMP Secretariat of KARI during data collection. Further, am grateful to the management of the KEFRI Muguga Regional Research Centre for their tireless support with transport to travel to the research sites. Biotechnology laboratory staff greatly assisted me and it is worthwhile mentioning the support accorded me both in the laboratory and in the field by Dr. David W. Odee, James Otieno, John Ochieng and John Gicheru. I am grateful to the KEFRI taxonomist, Mr. Francis Gachathi who gave botanical names to plant samples that I collected during my research activities. I am grateful to my supervisory committee for providing the guidance needed to make this study a reality. In addition, I am indeed grateful to Kenyatta University for registering me as their PhD. Student. Finally, I am grateful to my family for prayers and support throughout my study period. To crown it all I thank the Almighty God for doing everything well for me, may my God take all the glory in Jesus' great name. Amen. IV TABLE OF CO TE TS DECLARATION i DEDICATION ii ACKNOWLEDGEMENTS iii TABLE OF CO TE TS iv ABSTRACT vii LIST OF FIGURES viii LIST OF TABLES .ix LIST OF APPENDICES xi LIST OF ABBREVIATIONS xii CHAPTER ONE 1 1. 1 GENERAL INTRODUCTION 1.1 Research problem 3 1.2 Justification 4 1.3 Research questions 4 1.4 Study hypotheses 5 1.5 Research objectives 5 CHAPTER TWO 8 2. 8 LITERATURE REVIEW 2.1 Participatory rural appraisal (PRA) and questionnaire techniques 8 2.2 Integrated soil fertility management (ISFM) 9 2.3 Cowpea (Vigna unguiculata) 10 2.4 Rhizobia 11 CHAPTER THREE 18 3. 18 GENERAL MATERIALS A D METHODS v 3.1 Reconnaissance 3.2 Cowpea screening and selection 18 3.3 School demonstration 19 3.4 On-farm 3.5 surveys, community meetings and on-farm urveys 18 plots trials 19 itrogen fixation 19 3.6 Rhizobia populations 3.7 Determination in soil samples 19 of rhizobia diversity 20 CHAPTER FOUR 4. FARMER 21 AWARE AVAILABILITY ESS 0 SOIL FERTILITY STATUS A D FOOD IN MAKUENI DISTRICT 21 4.1 Introduction 21 4.2 Materials 23 4.3 Results 24 4.4 Discussion 31 and Methods CHAPTER FIVE AREA 5. 35 UNDER GRAIN LEGUMES AND PROBLEMS FACED BY FARMERS IN LEGUME PRODUCTIO 35 5.1 Introduction 35 5.2 Materials 5.3 Results 41 5.4 Discussion 46 and Methods CHAPTER SIX SCREE 6. 6.1 ING .40 50 EW COWPEA VARIETIES FOR DR YLA DS OF EASTER KE YA 50 Introduction 50 VI 6.2 Materials and Methods 52 6.3 Results and Discussion 54 CHAPTER SEVEN 7. 68 EFFECTS OF ISFM ON NODULATION, GROWTH AND GRAIN YIELD OF SELECTED COWPEA VARIETIES 68 7.1 Introduction 68 7.2 Materials and Methods 71 7.3 Results 73 7.4 Discussion 81 CHAPTER EIGHT 8. NITROGEN 86 FIXATION, POPULATION AND DIVERSITY RHIZOBIA UNDER ISFM 86 8.1 Introduction 86 8.2 Materials and Methods 89 8.3 Results 94 8.4 Discussion 101 CHAPTER NINE 9. OF COWPEA CONCLUSIONS 104 AND RECOMMENDATIONS 104 9.1 Conclusions 104 9.2 Recommendations 105 LITERATURE CITED 106 APPENDICES 131 Vll ABSTRACT The main objectives of this study were; (1) To find out whether farmers in Makueni District were aware of soil fertility status in their farms and annual food availability, (2) To find out the proportions of cultivated areas under grain legume production and the problems faced by farmers in grain legume production, (3) To screen and select high yielding cowpea varieties for dryland Makueni District, (4) To determine the effects of integrated soil fertility management (ISFM) on nodulation, growth and grain yield of selected cowpea varieties, and (5) To determine the effects of ISFM on nitrogen fixation, indigenous soil rhizobia populations and rhizobia diversity. Farmer participatory meetings were used to establish whether farmers recognized soil fertility as a problem in legume production. Results obtained revealed that farmers in the selected sites recognized soil fertility as a problem and included it in the list of general problems affecting them. Participating farmers indicated that only 2% of the cultivated farms in the study sites had fertile soils. To document grain yields, area under legume cultivation and problems faced by farmers in grain legume production, a structured questionnaire was used to collect information from farmers. Results obtained showed that grain yields ranged from 30 kg/ha to 416 kg/ha and area under legume cultivation from 48% to 92%. Problems faced by farmers in legume production included low soil fertility, inadequate farm inputs, weeds, pests and diseases. To select pioneer cowpea varieties, 34 cowpea varieties were selected and screened for two seasons at Kiboko Dryland Research Station. Some of the cowpea parameters assessed included pod length, plant biomass, grain yield and weights of 100 seeds. From the screening studies, nine cowpea varieties were selected for on-farm trials. To determine the effects of integrated soil fertility management (ISFM) on nodulation growth and grain yield of selected cowpea varieties, on-farm trials were established at two sites. The nine cowpea varieties that had been selected during the screening studies were planted in the trials. Treatments applied included a control, farmyard manure at 2.5 t/ha, phosphorus as triple superphosphate (TSP) (P20S, 0:46:0) at 15 kg/ha and a combination of both manure and TSP at the singly applied rates. Data collected included nodule and shoot biomass, and grain yields. Results obtained revealed that treatment application enhanced nodule and shoot biomass, and grain yields. Nitrogen fixation was estimated using J sN natural abundance method whi le rhizobia populations were determined using most probable number (MPN) experiment. Rhizobia diversity was determined using culture characterization and direct PCRRFLP of the 16S-23S rRNA intergenic spacer region (IOR) of rhizobia genome. Results of nitrogen fixation showed that 46-53% nitrogen (N) was fixed at a wetter site, while no N fixation took place at a drier site. Results from rhizobia population assessment revealed population counts of 4.89x I02 to 2.0x 104 cells/gram of soil with lower rhizobia counts at planting relative to the harvesting time while high rhizobia counts were recorded in amended soils relative to the controls. Further, restriction of eighteen rhizobia isolates from cowpea nodules with MspI restriction endonuclease revealed four rhizobia IOS groups. Vlll LIST OF FIGURES Figure 1.1 Map of Makueni District showing sub-locations with study sites Figure 4.1 Food availability calendars of Mbitini and Nguu Divisions 30 Figure 8.1 Amount of nitrogen fixed (%) at Kavuthu during the long rains 94 Figure 8.2 15N Figure 8.3 15N values for site by variety during the long rains 95 Figure 8.4 Nodule biomass of nodules recovered from MPN experiment.. 97 Figure 8.5 Shoot biomass of cowpea plants harvested from MP 98 Figure 8.6 Four PeR-amplified in sites 1 and 2 during the long rains 16S-23S rRNA 6 95 IGS patterns experiment.. obtained restriction of rhizobia strains with MpsI restriction endonuclease after 99 IX LIST OF TABLES Table 4.1 Population densities and number of households at Kavuthu, Matiku, Yikivumbu and Ndunguni Sub-locations in 1999 and 2003 Table 4.2 General problems faced by farmers in the selected sites of Makueni District. Table 4.3 22 , 25 Soil characterization and fertility status as described by farmers in selected sites of Makueni District.. 26 Table 4.4 Soil fertility management across farmer wealth classes 28 Table 4.5 Crop preference in Nguu and Mbitini Divisions of Makueni District 29 Table 4.6 Household sources of income and expenditure in the study sites 30 Table 5.1 Area under grain legume production in selected sub-locations .41 Table 5.2 Average legume yields (kg/ha) in selected sub-locations .42 Table 5.3 Soil fertility status of farms in the selected sites of Makueni District. 42 Table 5.4 Farmers (%) using inputs to enhance soil fertility Table 5.5 Sources of animal manure used by farmers 43 Table 5.6 Farmers (%) that reported common weeds in their farms 44 Table 5.7 Farmers (%) that reported pests and diseases in their farms 44 Table 5.8 Level of formal education (%) of household heads 45 Table 5.9 Food availability and source (%) during long dry spells 45 Table 6.1 Pod characters assessed during the short rains 57 Table 6.2 Pod characters assessed during the long rains 58 Table 6.3 Variety characters assessed during the short rains 60 Table 6.4 Variety characteristics assessed during the long rains 61 .43 x Table 6.5 Character ranking of varieties selected for on-farm trials Table 7.1 Chemical characteristics of soils collected at on-farm trial sites and animal manure applied to the on-farm trial.. Table 7.2 72 Nodule biomass (mg/plant) at 50% flowering during the long rains Table 7.3 67 74 Nodule biomass (mg/plant) at 50% flowering during the short rains 75 Table 7.4 Shoot dry weight (kg/ha) at crop maturity during the long rains 77 Table 7.5 Shoot dry weight (Kg/ha) at crop maturity during the short rains 78 Table 7.6 Grain yield at Kavuthu during the long rains 79 Table 7.7 Grain yield (Kg/ha) during the short rains 80 Table 8.1 Rhizobia estimates (rhizobia cells/gram of soil) in soils collected at the beginning and at the end oflong rain season 97 Table 8.2 Rhizobia isolate characterization using culture 99 Table 8.3 Fragment sizes in base pairs (bp) obtained after rhizobia restriction Table 8.4 100 Strain group, rRNA IGS pattern, cowpea varieties and treatments used in the PCR-RFLP analysis of 16S-23S rDNA IGS 100 Xl LIST OF APPENDICES Appendix 1 Maps of Yikivumbu, Ndunguni, Kavuthu and Matiku Sub-locations drawn by farmers during farmer meetings 131 Appendix 2 Questionnaire 135 Appendix 3 Experimental design (ISFM trial) 139 Appendix 4 Rhizobia cultures isolated from cowpea plant nodules 140 Xll LIST OF ABBREVIATIONS AfNet The African Network for Soil Biology and Fertility ASALs Arid and semi arid lands CIAT International Centre for Tropical Agriculture cm Centimeters Eds. Editors g Grammes ISFM Integrated soil fertility management kg/ha Kilogrammes per hectare Kgha-l yr' Kilogrammes per hectare per year Km Kilometer Km2 Square kilometer Litres M.a.s.l Metres above sea level mg Milligrammes mg/ml Microgrammes per millilitre mg/plant Milligrams per plant mgl' Milligrammes per litre ml Millilitres mM Micromolar mm Millimetres s Seconds TSBF Tropical Soil Biology and Fertility USA United States of America CHAPTER 1. ONE GENERAL INTRODUCTIO Soil nutrient depletion of arable lands is a major constraint to crop production in most parts of Africa. It results from continuous cropping with little or no inputs to replenish soil fertility, removal of crop residues to feed animals, overgrazing between the cropping seasons and soil erosion (Sanchez et al., 1997; Smaling et al., 1997; Stoorvogel et al., 1993). Soil nutrient depletion lowers the returns to agricultural investment, reduces food security through low crop yields, increases food prices and reduces government revenue through reduced taxes collected on agricultural goods (Sanchez et al., 1997). Nutrient depletion is not uniform and varies with soil characteristics, and is usually greater in sandy soils, although total nutrient loss is greater in clay soils (Swift et al., 1994). In addition, management of an individual farm directly affects nutrient depletion rates of that farm (Sanchez et al., 1997) and therefore nutrient depletion is not uniform across cultivated fields. Over the history of farming, farmers have used inputs to replenish soil fertility and animal manure has been used over a long time in tropical Africa to replenish soil fertility and enhance crop production (Dennison, 1961; Giller et al., 1997; Hartley, 1937; McCown et al., 1992). For example, in the arid and semi-arid lands (ASALs) of eastern Kenya, smallholder farmers use animal manure to improve soil fertility (Mathuva et al., 1996; Probert et al., 1995). However, animal manure, especially cattle manure, which is commonly used by farmers, is of low nutrient content and the KE VAT A IJNIVFR.~frY I H1RllAV} 2 amounts available are not usually adequate (Giller et al., 1997; Kimani and Lekasi, 2004; Probert et al., 1995). In addition, chemical composition of manure is highly varied depending on the diet fed to the livestock, collection and storage methods (Gill er et al., 1997; Powell, 1986). For example in Kenya nitrogen (N) content of cattle manure ranges from 0.2%-2.2%, while phosphorus (P) content ranges from 0.08% to 0.95% (Kimani and Lekasi, 2004). Inorganic fertilizers can supplement low nutrient animal manures but their prices especially to smallholder farmers on staple food crops are uneconomical (lama et al., 1997). For example in the arid and semi arid lands (ASALs) of eastern Kenya, adoption of inorganic fertilizers is constrained by high costs, low farm returns, and lack of right fertilizers to the resource poor farmers (Ikombo, 1984). Organic inputs in form of plant residues, especially legumes offer a cheap alternative source of N to expensive fertilizers for smallholder estimated that legume-Rhizobium farmers. For example, it is symbioses in tropical soils can fix up to almost 600 kg ha-1 yr' N (Gibson et al., 1982). However, unlike N, which may be fixed biologically, P can only be supplied to crops through addition of manures, compost, rock phosphates and inorganic fertilizers (Ayaga and Brookes, 2005). In addition, organic materials are low in P (Ayaga and Brookes, 2005; Giller et al., 1997; Palm 1995; Sanchez et al., 1997) and they should be supplemented with inorganic fertilizers (lama et al., 2000; Vanlauwe et al., 2001). For example, Janssen (1994) and Smaling and Braun (1996), found that application of combined inorganic P with farmyard manure significantly increased yields compared to when either of the inputs were applied as single dressing. 3 Soil nutrient depletion of arable lands is a major constrain to crop production in the drylands of Eastern Kenya (Siderius and Muchena, 1977). Need for increased food production has led to intensive agriculture, associated with ecological disturbance and soil fertility decline (Sande et al., 2004), which has resulted to reduced crop, yields. Soil productivity can be improved by use of inorganic fertilizers but the fertilizers are expensive and have negative effects on environment. Organic amendments, animal manure, crop residues and compost are used but are poor in nutrients and not adequate (Prober et al., 1995). Combinations of organic residues and inorganic fertilizers have been found to increase crop yields (Nziguheba et al., 2004, Ojiem et al., 2004 and Okalebo et al., 2004). combination Therefore the key to soil fertility improvement is the between organic amendments and mineral fertilizers. The combination of organic and inorganic fertilizers in a management system is termed as integrated soil fertility management (ISFM). 1.1 Research problem Soils in eastern Kenya drylands are low in nutrients especially phosphorus nitrogen (N) but the magnitude Makueni District knowledge. of the soil infertility has not been well documented using local farmers' In addition, there exists limited information cow pea varieties in comparison distribution on selection (P) and especially in indigenous of improved to the locally grown ones in the drylands. Further studies on ISFM effects on N fixation and rhizobia diversity drylands of eastern Kenya have not been reported. on cowpea in the 4 1.2 Justification There is need to document the magnitude of soil fertility status using indigenous knowledge from the local farmers so as to understand the actual soil infertility status at the farm level. Cowpea is the most important legume in the eastern Kenya drylands and there is therefore need to select high yielding varieties for the drylands so as to boost cowpea production in the region. In addition, it will be necessary to carry out ISFM studies on cowpea so as to understand the best nutrient management options for the crop under dryland conditions. There is also need to carry out studies on nitrogen fixation and rhizobia diversity of cowpea so as to document the effects of ISFM on these parameters. 1.3 Research questions 1) Are the farmers at Makueni District aware of soil fertility status in their farms and annual food availability? 2) What proportions of the cultivated areas do grain legumes occupy and what are the problems faced by farmers in grain legume production? 3) Are there better legume varieties, especially of cowpea, which can be selected for field trials at dryland Makueni District? 4) What effects would integrated soil fertility management (lSFM) have on nodule formation, growth and yield of selected cowpea varieties? 5) How would ISFM affect nitrogen fixation, indigenous rhizobia populations and diversity? 5 1.4 Study hypotheses 1) Farmers in Makueni District are not aware of soil fertility status in their farms and annual food availability. 2) Grain legumes dot not occupy large proportions of cultivated areas in Makueni District and farmers do not face any problems in grain legume production in the district. 3) Locally grown cowpea varieties cannot perform as well as the improved cowpea varieties in dryland Makueni District. 4) Integrated soil fertility management (ISFM) has no effect on nodulation, growth and grain yield of selected cowpea varieties. 5) Integrated soil fertility management has no effect on nitrogen fixation and, indigenous rhizobia population and diversity in the drylands ofMakueni District. 1.5 Research objectives 1) To find out whether farmers in Makueni District were aware of soil fertility status in their farms and annual food availability. 2) To find out the proportions of cultivated areas under grain legumes production and the problems faced by farmers in grain legume production. 3) To screen and select high yielding cow pea varieties for dryland Makueni District. 4) To determine the effects of integrated soil fertility management (ISFM) on nodulation, growth and grain yield of selected cowpea varieties. 5) To determine the effects of ISFM on nitrogen rhizobia populations and rhizobia diversity. fixation, indigenous soil 6 This study was implemented (Figure 1.1). Objectives namely, Yikivumbu Mbitini Division. 111 the ASALs of eastern Kenya in Makueni one and two were and Ndunguni Objective 111 implemented in Nguu Division three was implemented District in four sub-locations, and Kavuthu at Kiboko Station (not shown in the map) while the rest of the objectives and Matiku in Dryland Research were implemented using field trials that were established at Kavuthu and Ndunguni Sub-locations. I~ I I :;., Legend • Study sites • Market centres RO<1dclasses ~~ 'b o ~ivl slon bound.ry Location boundary a sue-recanons sub-tce atton with study situ II boundary ~ _. Figure 1.1 Map ofMakueni The study is represented conclusions .__ ...._.~_ ~_~_~_4~ District showing sub-locations 8 Kilomete: J with study sites in chapters four, five, six, seven and eight. A summary of and recommendations of the study is also included as chapter nine. In chapter four, results from farmer participatory meetings, that were held in four sub- locations in Makueni District are documented. Chapter five documents results from 7 farm surveys using a structured questionnaire. The surveys were carried out in the four sub-locations where farmer participatory meetings had been previously held. Chapter six documents uguiculata Agricultural commonly cultivated results obtained from on-station screening of cowpea (Vigna (L.) Walp) varieties Research grown Institute legume at Kiboko, a dryland (KARI). after common In Kenya, research station of Kenya cow pea is the second most beans (Phaseolus vulgaris for food (leaves, young pods and grain), while secondary fodder and soil fertility improvement. (L.) and is uses include It has high protein content with mean crude protein of leaves, grain and crop residues estimated at 32-34%, 23-35% and 11-12%, respectively (Imungi and Porter, 1983). Cowpea is mostly grown in the ASALs and about 85% is cultivated in eastern province (Muli and Saha, 2000). Chapter seven documents" the effects of integrated formation, biomass production soil fertility management and grain yields of nine selected cowpea varieties under on-farm conditions in contrasting rainfall conditions chapter eight documents (ISFM) on nodule over two seasons. Lastly, the effects of ISFM on nitrogen fixation and, soil rhizobia populations and diversity in the ISFM trial plots. 8 CHAPTER 2. 2.1 Participatory 2.1.1 PRA Participatory LITERATURE rural appraisal rural appraisal REVIEW (PRA) and questionnaire was developed including Rapid Rural Appraisal developed TWO from several (RRA) and Agroecosystem in Britain by the Institute for Development Institute for Environment and development Participatory and Development institutions (Conway techniques participatory Analysis, which were studies and the International (IIED) and several agricultural and McCracken, National Environment Secretariat and Clark University. approach has developed and expanded to other countries (Bronson, aim of PRA is to change the attitude of development view participation own problems about rural efforts of the Since then the 1995). The main partners and change agents to as an end rather than a means. Participatory attempts to make communities in 1988). The term PRA was coined in Kenya in early 1986 through collaborative Kenya 1983). techniques that it shortens the process of data collecting and analysing information (Conway and McCracken, research 1988; Chambers, rural appraisal differs from the previous data collecting communities methods Rural Appraisal directly involved in and responsible for assessing their and arriving at a consensus of action that needs to be taken. The assumption of PRA is that unless people feel a project is theirs, there will be no commitment and the outside experts must consult on the community needs since they do not know enough on the community participation discussions, is ensured by gathering visual aids, and observations in question. Under PRA approach, data in group discussions, (K WAP, 1993). farmer using informal 9 Farmer participation has been used in many development projects in the sub-Saharan Africa. For example, in Southern Malawi it was used to determine farmer preferences for bean varieties by combining ranking and scoring by farmers (Abeyasekera 2004), while in Kenya farmer participation was used in crop prioritization et al., (Onyango et al., (2000) and Okoko and Makwaro (2000) to establish the most preferred crop by selected communities. 2.1.2 Questionnaire In this technique structured questions targeting particular subjects are developed. Data is collected through visits to targeted communities or institutions. The disadvantage of this technique is that it is expensive and time consuming (PRA Programme University, 1995) because of the time taken to administer questionnaire Egerton and to analyse the information collected. 2.2 Continuous Integrated soil fertility management (ISFM) cropping, removal of crop residues to feed animals and overgrazing between cropping seasons with little or no external inputs have reduced the productive capacity of arable lands and thus threaten the sustainability systems in Sub-Saharan of food production Africa (Sanchez et al., 1997; Stoorvogel Kenya, decline in crop yields has been a major problem-facing et al., 1993). In smallholder farmers (Mathuva et aI., 1996). This is attributed to the high costs of inputs that make the use of inorganic fertilizers on staple food crops uneconomical for most smallholder farmers (Jama et al., 1997). Use of organic inputs as an external source of soil nutrients is a logical cheap alternative to expensive fertilizers to smallholder farmers. However, organic inputs are low in nutrient concentration compared to inorganic 10 fertilizers (Sanchez et al., 1997). Despite the fact that organic inputs are low in nutrients (Gill er et al., 1997), cattle manure is an integral component of soil fertility management, and manure application is one of the most commonly used and effective way of soil fertility improvement for crop production in Africa (Dennison, 1961; Hartley, 1937). For example, in the semi-arid areas of eastern Kenya where nitrogen and phosphorus limit crop production, manure has been used to enhance soil fertility and crop production (Gibberd, 1995; Ikombo, 1984; Kihanda et al., 2004). Recent research has shown that combination of organic and inorganic inputs enhances crop production and reduces cost of inorganic fertilizers ((Nziguheba et al., 2004; Ojiem et al., 2004; Okalebo et al., 2004). The combination of organic and inorganic inputs is termed integrated soil fertility management (ISFM). Studies on integrated nutrient management are common. For example, in Tanzania, Tithonia diversifolia superphosphate (Hems!.) improved A. Gray and Mijingu phosphorus phosphate rock or triple content and increased maize yields (Ikera, 2006). Further in the coastal lowlands of Kenya, a combination of inorganic fertilizers and manure showed that half of the recommended rates of organic manure and half of the inorganic fertilizers in combination had same effect on maize yields as the full recommended inputs when not combined (Saha and Muli, 2000). 2.3 Cow pea (Vigna unguiculatay Cowpea (Vigna unguiculata L. Walp.) is an annual or bi-annual grain cereal legume commonly referred to as cowpeas, southern pea, black- eyed pea, Crowder pea, lubia, niebe, coupe or frijole. It is native to Africa, Asia and the Middle East. It is grown in the Savanna regions of the tropics and subtropics. It is widely grown in Africa, Latin 11 America, cowpea Southeast is grown commonly Asia and Southern United States (Davis et al., 1991). Most in west and central African countries. In Kenya, cowpea is grown in the arid and semi-arid areas (ASALs) of mostly eastern Kenya. Cowpea value lies in its high protein content, its ability to tolerate drought and fix atmospheric N, which allows it to grow in and improve poor soils (Bressani, 1985). Cowpea is used for food, fodder and as a source of income. Leaves, young pods and grain are the parts of the plant used for food. The same plant parts are sold to generate cash for farmers. Cowpea has an added advantage to soil fertility improvement it is a nitrogen-fixing through residual in that legume and can grow in very poor soils and replenish them effect on a subsequent crop. Studies on cowpea are common in Africa. For example, in Nigeria, Saidou et aI., (2006) evaluated cowpea tolerance and response to external application of phosphorus both in the field and in the greenhouse. In Kenya, Muli and Saha (2000) investigated adaptation and yield performance of cowpea cultivars along the coast of Kenya. 2.4 Rhizobia Rhizobia are aerobic gram-negative legumes have Bradyrhizobium. been grown. There bacteria that live freely in soil, especially where are two types The term Rhizobium of rhizobia, refers to the fast-growing Rhizobium and types whereas Bradyrhizobium is used for the slow-growing types (Date and Halliday, 1987). They form specialised structures on roots of leguminous the sites of symbiotic energy requiring plants called nodules, which are nitrogen fixation (Lara et al., 1988). Nitrogen fixation is an process and utilises phosphorus 1990). To fix one molecule of nitrogen in form of ATP (Bockman et al., 16 ATPs are required. Thus, legume 12 association aspect with rhizobia of symbiosis number, nodule nitrogen fixation. 2.4.1 contributes is enhanced weight, to high phosphorus by phosphorus. plant dry matter This production, uptake. includes Virtually traits plant nitrogen every such a nodule concentration and Rhizobia diversity Population diversity stability, is an important population be investigated Ribosomal aspect must be as diverse using differences of population as possible. genetics. The microsymbiont in 16S ribosomal ribonucleic RNA genes are the best targets for studying phylogenetic of an "evolutionary clock". all bacteria, and more domains highly conserved variable diversity acid (rRNA) rRNA one can find all the properties have For population can genes. In relationships. The rRNA present in et al., (Vandamme 1996). The 5S rRNA (Vandamme was the et al., 1996), sequence comparison sequences can (Ochman investigated used first marker to a lesser molecule but cannot it is only provide and Wilson, molecules extent and to be sequenced 120 nucleotides information only partial sequences et al., 1997). The 16S sequences and above (Young, 1996), whereas sequence change topology of the 16S and 23S rRNA in the 23S rRNA is faster and it can therefore of rRNA gene dendrograms has become bacteria therefore as and most the the larger extensively The later has so far been of some are reliabie molecule, rhizobia have been at the level of genus the average for close relationships. is however, obligatory and accurately useful be valuable numerous long are the 16S and 23S rRNAs. (Tesfaye Sequencing as 1987). The most determined 1995). for similar (Ludwig and together with rate of The et al., DNA- 13 DNA reassociation they form the key methods that have to be used when new rhizobial species or genes are to be described. Several methods have been used in the past to assess composition, genetic (2) diversity. Isozyme Restriction Fragment PCR-based fingerprinting on repetitive Some of the methods electrophoresis, length polymorphism, elements, (5) Pulsed-field hybridization, gel electrophoresis, (8) Amplification fragment length polymorphism, (4) (6) based and (9) of amplified rRNA gene fragments. DNA-base composition The unique taxonomic cytosine content organisms DNA-DNA (1) DNA base with random or arbitrary primers, (7) Amplification Restriction fragment length polymorphism 2.4.1.1 (3) include: feature used in this method was mole percent guanine plus (Mol% G+C). However, the disadvantage of this method is that having the same Mol% G+C values are not necessarily since Mol% G+C does not take into account the linear arrangement in the DNA (Johnson, 1984). However, the technique closely related of the nucleotides is still used as part of the standard description of bacterial taxa (Vandamme et al., 1996) 2.4.1.2 DNA-DNA hybridization The property of DNA to dissociate and reassociate is used in hybridization The complementary strands are separated and then let to associate another single-stranded DNA (Johnson, used are: the hydroxyapatite nuclease method (Vandamme reassociation analysis. to itself or to 199 I). Some of the reassociation methods method, the optical renaturation rates method and the SI et al., 1996). The hydroxyapatite method was the first method used in rhizobia taxonomy (Wedlock and Jarvis, 1986; Crow et al., 1981; Hollis et al., 1981). In this method, the DNA of the reference organism is 14 labeled with 32p (in vivo) and let to associate with an excess of unlabelled DNA from the test organism. Then, the double stranded DNA is separated DNA with hydroxyapatite to which the double-stranded into single-stranded DNA binds. The single- stranded DNA can then be eluted from the solution. In the optical renaturation method, the DNA spectrophotometer is not labeled and the reassociation at 260 nm (Johnson, difference in the temperature between the homologous 1991). However, is measured the determination rates with a of the corresponding to half the increase in relative absorbance and heterozygous hybrids formed under standard conditions (Vandamme et al., 1996). In the S I nuclease method, the DNA is labeled with tritium eH). The reassociation mixture is treated with single-strand specific nuclease SI, which degrades single strand regions including the loops. Double stranded DNA is collected by precipitation the association or by binding to nitrocellulose is calculated counter (Johnson, from the radioactivity membrane. The percent of values obtained by a scintillation 1991). Currently, the SI nuclease method (Nour et 01., 1995; Nour et al., 1994; Laguerre et 01., 1993) has replaced the hydroxyapatite used high levels of radioactivity. The disadvantage of DNA-DNA method, which hybridization method is that, it is laborious and when rhizobial species are many, only those strains, which are phenologically closely related, can be analysed (de Lajudie et al., 1994, Nour et al., 1994). 2.4.1.3 Restriction fragment length polymorphism (RFLP) This method is usually used together with southern blot hybridization, DNA fragments are used as probes (Mason and Williams, used with plasmids but the disadvantage where, specific 1985). RFLP can also be is that different strains of bacteria do not always contain or keep their plasmids (Vandamme et al., 1996). RFLP analysis was 15 good typing method before the routine application rhizobial taxonomy (Kaijalainen of polymerase it has been used to study chromosomally and Lindstrorn, 1989), especially hybridization chain reaction. In encoded polymorphism patterns of nod and nif genes which are often plasmid coded (Eardly et al., 1992; Laguerre et al., 1993). The RFLP analysis using symbiotic genes is still very useful (UI1Z and Elkan, 1996; Paffetti et al., 1996). 2.4.1.4 Pulse-field gel electrophoresis (PFGE) In PFGE the D A molecules are forced to change their direction of migration with modification of current and their speed of reorientation varies as a function of their size. Resolution of DNA molecules of up to 9 Mb can be obtained. Using this method, rhizobia can be differentiated at species, biovar and strain level. The PFGE method has mainly been used to study the genome rearrangements (Sobral et al., 1991) and in estimating the genome size (Huber and Selenska-Pobell, 1994; Sobral et al., 1991). However, PFGE has not been the method of choice in taxonomy, the methods based on amplification by polymerase chain probably because reaction are more convenient. 2.4.1.5 PeR-based fingerprinting with random or arbitrary This method is based on the theory that the distribution oligonucleotide primers of any randomly chosen sequence will vary within the genome of different individuals. Thus when a single randomly chosen sequence is used as a primer, the amplification product of individual organisms will be different (Harrison et aI., 1992). This method has mainly been used to study genetic diversity within a single rhizobial species (Sikora et al., 1997; Paffetti et al., 1996; Dye et al., 1995; Selenska-Pobell et al., 16 1995). The main disadvantage of this method is the low sensitivity to the reaction conditions (Vos et al., 1995). 2.4.1.6 Amplification based 011 repetitive elements This method is based on the principle that prokaryotic repetitive sequences (rep-Pf.R) genomes contain interspersed (Stern et al., 1984; Higgins et al., 1982; Sharples and L1oyd, 1990; Hulton et al., 1991; Martin et al., 1992; Versalovic et al., 1994). PCR-primers are designed from these elements and they enable the simultaneous amplification of many DNA fragments of different sizes, originating from the sequences lying between the repetitive within elements. a single This method has only been used to study the differences rhizobial species using primers of only REP or ERIC element (Laquerre et al., 1996). 2.4.1.7 Amplification Amplification fragment length polymorphism the fingerprinting fragment length polymorphism (AFLP) (AFLP) is the newest technique among methods. In AFLP, the genomic DNA is digested with a rare cutting and a frequent cutting restriction enzyme. Adapters are then ligated to the ends of the fragments to generate template DNA for amplification. In this technique, primers are used. They consist of three parts: a core sequence, an enzyme-specific sequence and a selective to the adapter extension of three nucleotides, and is complementary sequence and to the adjacent restriction site. The selective extension makes sure that only a subset of the fragments fingerprint is amplified and a specific and highly reproducible is obtained. The advantage of this method is that no prior knowledge of nucleotide sequence is needed. From 50 to 100 amplified fragments are separated on polyacrylamide gel and visualised autoradiographically as one of the primers is 17 labeled radioactively visualization (Vos et al., 1995). If a non-radioactive can be made possible bradyrhizobial and mesorhizobial by silver staining. AFLP method is used, fingerprinting of strains have been found to be highly reproducible and the results in agreement with rep-PCR and with RFLP analyses of the 16S and 23S rDNA amplified by PCR. 2.4.1.8 Restriction fragment length polymorphism gene fragments of amplified rRNA There are three rRNA molecules of different sizes, 5S, 16S and 23S rRNA. However, the RFLP analysis of PCR-amplified rRNA fragment is based on the two large molecules, the 16S rDNA and 23S rDNA. The 16S molecule has also been amplified with the 16S-23S rDNA intergenic spacer region (IGS). Analysis of PCR-RFLP, also called amplified identification ribosomal method DNA-restriction for bacteria. analysis, has been found to be a rapid The procedure involves amplification of PCR product with universal primers located in the conserved regions of the rRNA genes. The amplified enzymes, fragment and reproducible the is then digested with several, selected, profiles and composed thus obtained are mainly of 2 to 10 fragments similarity of the bacteria can be visualised frequently species-specific, (Vandamme by a dendrogram cutting highly et al., 1996). The constructed from the patterns. The PCR-RFLP analysis of the 1.5 kb 16S rDNA has been the most popular DNA fingerprinting identification method. Studies have shown that, this method is not only a rapid method but also a simple tool in taxonomy and gives an estimation of genetic relationships at species and higher levels (Laguerre et al., 1994). 18 CHAPTER THREE 3. 3.1 GENERAL MATERIALS AND METHODS Reconnaissance surveys, community meetings and on-farm surveys A reconnaissance survey was done in Makueni District with the help of agricultural extension officers. Two divisions were selected and in each division two sub-locations were identified administrators as study sites. After site identification, visits were made to the (chiefs and assistant chiefs) to brief them on the aim of the project. During the same visits dates for farmer participatory meetings were set. During farmer participatory meetings, farmers formed discussion groups so as to allow all the farmers participating farmers to contribute to the discussions. All discussions were led by one of the community from farmer discussions farm surveys, members using a checklist provided to each group. Results were documented on flip charts in the local language. For on- a questionnaire was formulated so as to seek specific answers to selected questions from farmers in the selected sites. Individual interviews were done with household heads on random visits to the homesteads. 3.2 Cowpea screening and selection Thirty-four cow pea varieties, with 30 varieties from International Institute of Tropical Agriculture (lIT A) and 4 varieties from local sources were multiplied and screened at Kiboko, a Kenya Agricultural Research Institute (KARI) centre. The centre is a dry land research station located about 30 km from the research sites. The cowpea screening activities rains of2005. were implemented during the short rains of 2004 and the long After screening, seven cow pea varieties were selected for on-farm trials after assessment of some yield parameters such as plant biomass production, yield, pod length and pod numbers. grain 19 3.3 School demonstration School demonstration plots trials were established during the short rains of 2005 in two primary schools, Kavuthu and Ndunguni. All the 34 cowpea varieties were planted with four treatments (control, manure at 2.5 t/ha, phosphorus as TSP (P20S: 0:46:0) at 15 kg/ha and munure+ TSP at the singly applied rates) and replicated From the demonstration plots, two cowpea varieties in 3 blocks. that were recommended by farmers were used in on-farm trials (Chapter 7). 3.4 On-farm trials Cow pea varieties selected on-station (Kiboko) and those selected by farmers were tested using integrated soil fertility management (ISFM) in on-farm trials established at Kavuthu and Ndunguni. Treatments that were used in the trials were the same as those used in the school demonstration 3.5 Nitrogen Nitrogen fixation plots. fixation was estimated using ISN natural abundance method. Selected cowpea varieties were sent to USA for ISN analysis. 3.6 Rhizobia populations in soil samples Soils for this study were sampled from the integrated soil fertility management trials. The soils were collected at the beginning of the trials and during harvesting time in the first season of the trials. To determine rhizobia populations in the sampled soils, most probable number (MPN) method was used. Two cowpea varieties selected from among the nine varieties used in the ISFM trials were used as test crops. A fivefold dilution series replicated four times at each dilution level was used. The dilution series 20 ranged from s' to y6. A 1 ml volume from each dilution level was applied directly to the cowpea root systems using sterile 1.0 ml pipettes. Leonard jars were used and infection counts were assessed after 28 days. Data sheets were prepared to allow for the entry of each experimental unit displaying either positive or negative nodule formation. 3.7 Determination of rhizobia diversity 3.7.1 Using rhizobia cultures Rhizobia cultures were obtained from root nodules of two selected cowpea varieties. Using culture characteristics such as colony size, shape, colour and mucus production, nine rhizobia strain groups were identified. The rhizobia strains were stored in a 15% glycerol yeast manitol at -70°C for future studies. 3.7.2 Direct polymerase Dried beads chain reaction (Pf.R) amplification (one for each amplification process) Ready-to-go of 16s rRNA PCR beads (GE Healthcare illustra™) containing nucleotides and buffers, and combined with primers, were used to amplify rhizobia genome from fully-grown amplification, a programmable rhizobia culture cells. For Thermal Controller (PTC-100TM MJ Research Inc., Watertown, MA) was used as described here: Initial denaturation at 93°C for 2 min; 35 cycles of denaturation at 72 0c) (45 s at 93°C), annealing (45 s at 62°C), extension (2 min and final extension at 72 -c for 5 min. Amplified rRNA products were subjected to restriction enzyme and separated by horizontal gel electrophoresis agarose gel, stained with ethidium bromide and photographed under UV. in 1% 21 CHAPTER FOUR 4. FARMER AWARENESS ON SOIL FERTILITY STATUS AND FOOD AVAILABILITY IN MAKUENI DISTRICT 4.1 Introduction Makueni district is one of the districts in Eastern Province of Kenya and occupies a total area of7,966 km", this being 5.16% of Eastern Province which has a total area of 154,354 km", and 2.69% of the total land area of Kenya (581,677 krn '). The district was curved from Machakos district and borders Kitui district to the east, Machakos district to the north and west and Kajiado district and Taita Taveta districts to the south. Makueni District rises between 1000-1500 m above sea level and has vegetation generally described as bush and thicket. It has a bimodal rainfall that has an annual average of 200-400 November/December. Makueni is described mm, varying with altitude, Average annual temperature as recent sedimentary and falls in March/May and of the district is 28°C. Geology of rocks and major soils are generally described as dark-red sandy loams (The MacmiJlan Atlas, 1997). The 1999 population census carried out in Kenya indicated that the total population of the country was 28, 686,607, with 6,371,370 households and a population density of 49 persons population per square kilometre. At the same time Makueni of 77 I ,545 with 144,320 households district had a total and a population density of 97 persons per square kilometre (Republic of Kenya, 2001). Some of the sub-locations Makueni District are Kavuthu and Matiku in Mbitini Division and, Yikivumbu le in and 22 Ndunguni in Nguu Division. According to Republic of Kenya (2001), some of these sub-locations had very high population densities in 1999 (Table 4.1). Table 4.1 Population densities and number of households at Kavuthu, Matiku, Yikivumbu Sublocation Population Kavuthu Matiku Ndunguni Yikivumbu 1,060 2,273 957 1,321 and Ndunguni Sub-locations Area of the sub-location (km2) 7.8 8.7 8.5 21.4 in 1999 and 2003 Population densities (persons/krrr') in 1999 136 261 113 62 Number of households in 1999 171 372 165 205 Smaling et al., (1997) reported that loss of soil nutrients leading to soil depletion is a worldwide problem affecting 135 million hectares, most of them in South America and African, and that the problem of soil infertility in sub-Saharan continuous cropping without sufficient replenishment depletion. Sanchez of nutrients leading to nutrient et al., (1997) further reported that, soil fertility smallholder farms is the fundamental food production Africa is due to in sub-Saharan depletion in biophysical root cause for declining per capita Africa. Continuous cropping, removal of crop residues to feed animals and overgrazing between cropping seasons with little or no external inputs reduce the productive capacity of arable lands and thus threaten the sustainability of food production systems in sub-Saharan Africa (Sanchez et aI., 1997; Stoorvogel et al., 1993). In Kenya, decline in crop yields is a major problem facing smallholder farmers (Mathuva et al., 1996), which has been attributed to high costs of inputs that make the use of inorganic fertilizers on staple food crops uneconomical most smallholder farmers (lama et al., 1997). for 23 4.2 A reconnaissance and Methods survey was carried out in Makueni district to select study sites. Division agricultural site identification. existence Materials extension officers from Mbitini and Nguu Divisions assisted in The criteria for site selection included accessibility of on-going donor funded projects. Chiefs to the sites and and assistant chiefs of the selected sites advised on sites where no other projects were going on to avoid conflict of interest. Two divisions, Mbitini and Nguu, were identified and in each division, two study sites were selected. The sites selected included Kavuthu and Matiku sublocations in Mbitini division, and Yikivumbu and Ndunguni division. After site selection, help of the extension location. The assistant officers mobilized announcements farmer participatory sub-locations meetings were arranged with the officers and the local chiefs and their assistants, chiefs of each sub-location the communities in public meetings, in Nguu in each sub- and the agricultural to attend the meetings through extension posters and locally called 'barazas'. Meetings were held in central points, in local primary schools, since the schools had closed for August holidays. Participating of between representation farmers democratically divided themselves eight to thirty on attendance, and with an equal of gender. Each group was assigned a topic to discuss and issued with a flip chart where to document general problems economics. people depending into groups of sizes affecting group discussion outcomes. Topics discussed included farmers, soil fertility, crop production and household 24 4.3 4.3.1 General Participating problems affecting farmers in each study site farmers in all selected sites identified and discussed general problems facing them. Problems identified products, water accessibility, and communication, decreasing Results were poor soils, seed scarcity, livestock feed scarcity, decreasing crop pests and diseases, unreliable marketing farm land, poor transport rainfall, education costs, crop yields, distant hospitals and poor services, social problems, lack of post primary education, low livestock prices, soil erosion, poor planting methods, lack of farm implements, problems inadequate labour and noxious weeds. In each site, all identified were ranked University, using pairwise ranking matrix (PRA Programme Egerton 1995). However for easy comparison of the site problem prioritization, the results from pairwise ranking matrixes from each site were summarized (Table 4.2). Problems identified and priorities varied with the study sites. However, from farmer discussions, priority problems were accessibility to water, unreliable soils, seed scarcity due to famines, decreasing land, education rainfall, poor costs, poor planting methods, pests and diseases and distant hospitals. 4.3.2 Soil types, distribution Soil identification maps (Appendix and women and fertility status and characterization was made possible by drawing of sub-location 1). In all the study sites, sub-location maps were drawn by both men representatives and used them to identify and quantity various soil types. This became possible when the farmers location into villages, distribution the extent of sub-divided each sub- which helped them to discuss the extent of soil type and in detail. Soil types identified were red soils, sandy soils, black cotton soils, sandy gravel and stony soils. Soil fertility status were described as fertile, 25 average, low or very low (Table 4.3). Low and very low fertility study sites and covered and Yikivumbu, were described Table 4.2 40%, respectively. 100%, 75% and 98% of Kavuthu, soils dominated Ndunguni, Only 2% of the soils, and at Yikivumbu by the participating General problems Matiku Sub-location, farmers as fertile. faced by farmers in the selected sites of Makueni District Problem Poor soils Crop pests and diseases Unreliable rainfall Seed scarcity due to famine Marketing farm products Water accessibility Lack of livestock feeds and diseases Education costs Decreasing land Decreasing crop yields Poor transport and comm unication Distant hospitals and poor services Social problems Lack of post primary institutions Low livestock prices Soil erosion Poor planting methods Lack of farm implements Inadequate labour Noxious weeds Site and Rank Kavuthu 3 4 N/A 7 9 I 8 Matiku 7 N/A N/A 5 10 1 Ndunguni 3 8 2 11 10 1 Yikivumbu 7 8 1 4 6 2 9 3 N/A N/A 2 11 2 3 5 5 5 N/A N/A N/A 12 8 7 12 N/A the 6 4 6 11 N/A N/A N/A N/A 10 NIA N/A N/A N/A NI A NI A NI A 12 9 6 N/A NI A NI A 12 NI A N/A 4 N/A NI A NI A NI A NI A N/A N/A 9 11 10 NI A means the item was not discussed at the selected site, while 1 and 12 means most important and 12 least important, respectively. 26 Table 4.3 Soil characterization and fertility status as described by farmers in selected sites of Makueni District Sublocation Kavuthu Ndunguni Matiku Yikivumbu Soil type Important soil characteristics 'I1ivi (Black cotton soils) 'Kitune' (Red Soils) 'Nthangathi' (Sandy Soils) Black 111 colour, high water holding capacity Red, productive, good water holding capacity Whitish, low water holding capacity, poor In nutrients, dries quickly a) Cracking type: black, cracks b) Non-cracking cotton soil, high water holding capacity a) Hard red soils ('Kitune Kyumu'). Gives low yields b) Soft red soils (' lturn bekethe') High water holding capacity and good for crop production Whitish or greyish in colour, poor moisture retention Black, high water holding capacity Red, good water retention capacity Whitish III colour, easily drained Reddish white in colour, have gravel, low water retention capacity Black 111 colour, high water holding capacity 'Yumba' or 'Ilivi' (Black cotton soils) 'Kitune' (Red Soils) 'Nthangathi' (Sandy soils) 'I1ivi' (Black cotton soil) 'Kitune' (Red Soils) 'Nthangathi' (Sandy soil) 'Kivuthi' (Sandy) gravel 'Yumba' or 'ilivi' (Black cotton soil) 'Kitune' (Red soils) 'Nthangathi' Sandy soil 'Mavia', (Stony soil) Red 111 colour, Low water holding capacity Greyish in colour, poor water retention Rocky, stony and shallow, low water holding capacity Fertility Status Average Cover (%) 60 Low 25 Very low 15 Low 30 Low 60 Low 10 Low 15 Average 25 Low 45 Very low 15 Fertile 2 Low 15 Low 75 Very low 8 27 4.3.3 Wealth in relation to soil fertility management Farm size was an important conservation indicator of riches and soil fertility and food production. farmer classes, Farmers in the four study sites identified various farm sizes they owned and soil fertility employed in each farmer-class improvement, management strategies (Table 4.4). The richest farmers, class 1 farmers, had largest farm sizes and commonly bought animal manure from farmer class 3 and 4. However, farm sizes owned by various farmer-classes varied with site. For example farmer class 1 at Kavuthu owned 10 to 15 acres of land while the same farmer class at Yikivumbu owned more than 20 acres. 4.3.4 Crop preference Main cereals grown in the study sites were maize (Zea mays L.), pearl millet tPennisetum glaucum (L.) R.Br.), sorghum (Sorghum bicolor (L.) Moench) and finger millet tEleusine corcana (L.) Gaertn), while grain legumes included common bean (Phaseolus vulgaris L.), cowpea (Vigna unguiculata (L.) Walp), pigeonpea (Cajanus cajan (L.) Millsp), green gram (Vigna radiata (L.) R. Wilcz) and dolichos (Lablab purpures L.). Using pairwise ranking crop preference was prioritized the results summarized the most preferred important in all sites and (Table 4.5). Crop preference ranking showed that maize was crop followed legume preferred by common bean. Pigeon pea was the second by farmers while finger millet was the second most preferred cereal. Pearl millet and Dolichos were only discussed at Yikivumbu location. sub- 28 Table 4.4 Site 2 3 4 Soil fertility management across farmer wealth classes Wealth classes and Soil management strategy Class 4 Class 1 Class 2 Class 3 N/A Own less than 7 Own 10-15 acres -Own 7-9 acres of land, all farms of land, most acres of land, farms terraced, rarely terrace are well terraced, buy their farms, sells manures all may farms, buys manure from their manure to manure from class 3 class 1 and 2 class 3 N/A Own 20-50 acres Own 5-9 acres Own less than 5 of land, heavily of land, acres of land, limited manures farms manures farms, have may purchase manure use, a few and purchases manure from manure from terrace their class 3, high class 3, farms, farms are is not fenced and degree of terracing terracing, well common, most are commonly fenced farms, farms fenced, overgrazed by farms have fertile most farms roaming animals, have fertile Soil infertility is soils. soils very common. N/A Own more than Own 10-15 Own less than 10 15 acres, have acres, most acres of land, well- fenced terraced farms, provide labour for farms, well- many fence class 2 and 1 and terraced farms, their farms, their farms are high harvest due some have good not manured nor to fertile soils harvest due to terraced and fertile soils commonly no harvest. Own more than Own 10-19 Own 5-9 acres, Own less than 5 20 acres, all farms acres, all their have scattered acres of land, are terraced, farms are braches and sisal farms not fenced, manures farms terraced, fencing, have few do not terrace and buys manure manures and terraces, does their farms, from class 3 and buys manure limited manuring, limited manuring, 4, from class 3 and limited harvest little if any high harvest 4, high harvest due to poor soils harvest because of fertile due to fertile soils soils Site (sub-location); 1- Kavuthu, 2- Matiku, 3- Ndunguni and 4 Yikivumbu. Class 1 consisted of the richest farmers while class 4 was the poorest farmers. NI A means farmer class was not described in the site. 29 Table 4.5 Crop preference in Nguu and Mbitini Divisions ofMakueni District Site and crop priority position Crop Yikivumbu Ndunguni Kavuthu Matiku Maize ] 1 1 I Common bean 2 2 2 4 Green gram 4 6 5 7 Pigeonpea 5 3 3 2 Cowpea 3 5 6 3 Finger millet 6 4 4 5 Sorghum 9 7 7 6 Dolichos 8 N/A N/A N/A Pearl millet 7 N/A N/A N/A N/A means the item was not discussed at the selected site, whilel means most important and 9 means least important. 4.3.5 Participating Annual food availability farmers discussed food security in all the sub-locations. Results obtained from the study showed that food availability peaked in February and slowly dropped over the year in all sites except Matiku where it rose between June and August (Figure 4.1). At Yikivumbu, from September through Ndunguni and Kavuthu, food availability dropped to zero November. However, from November food availability started rising again at Kavuthu due to availability of green vegetables from cowpea that provided food from November, a month following the on-set of short rains. 4.3.6 Household income and expenditure Results obtained from the farmer participatory meetings revealed that crop production and livestock keeping were the main source of income in the study sites (Table 4.6). Income from crops was 55%, 10%, 40% and 60% at Yikivumbu, and Matiku, respectively Ndunguni, Kavuthu compared to income from livestock that was 25%, 70%, 30% and 15% at Yikivumbu, Ndunguni, Kavuthu and Matiku, respectively. However, 30 the percentage expenditure on food purchases was 40%, 60%, 40% and 50% at Yikivumbu, Ndunguni, Kavuthu and Matiku, respectively. 120 -----,--- ---- .--.-.----.----._------------ .........-Matiku -Kavuthu 100 C _Ndunguni ____ Yikivumbu 80 "0 0 <2 0) 60 > 40 :0 -'.<;;" '" < 20 0 Jan Feb Mar Apr Ma JWl Jul Aug Sept Oct Nov Dec Months of the year Figure 4.1 Food availability calendars ofMbitini Table 4.6 Household sources of income and expenditure in the study sites Sub-location Yikivumbu Ndunguni Kavuthu Matiku Income Crops Livestock Others Contribution (%) 55 25 20 Crops Livestock Others 10 70 20 Crops Livestock Monthly salary Casual labour Crops Livestock Employment Others 40 30 10 20 60 15 15 10 and Nguu Divisions Expenditure Food Farm inputs Health Others Food Farm inputs School fees Others Education Food Others Education Food Health Others Contribution (%) 40 IS 20 25 60 II 5 24 30 40 30 IS 50 18 17 n ' 31 4.4 Accessibility Discussion to water was the most important problem in three out of the four study sites, where it took position number I. According to the farmers in all the sites, there was no clean water for domestic and livestock use because all water in the sublocations came from shallow wells sunk in seasonal rivers. The farmers said that the wells occasionally sometimes dried up in dry weather far away from their homesteads forcing them to sink deeper wells that caused them to spend most time fetching water for domestic use. However, rainfall unreliability was very important at Yikivumbu and Ndunguni sub-locations where it ranks position 1 and 2, respectively. It was noted that the two sites had experienced Related to the rainfall unreliability priority position commonly planted 4 at Yikivumbu. drought between 1998 and 2003. was the problem of seed scarcity that was given It was noted that farmers in the study sites local crop varieties most of which had been wiped out by the frequent droughts. The problem of decreasing land was more important at Kavuthu and Matiku, where it took positions 2 and 3, respectively. It was observed that most farm units were smaller than in the other sites. The farmers said that in addition to population increases there were influxes of farmers from drier sites who purchased because the sub-locations land in the sub-locations were wetter than most parts of the district. In addition, soil fertility problem was more important at Kavuthu and Ndunguni sub-locations than other sites, where the problem took position 3. It was noted that the farmers in the two sub-locations practiced irrigation of vegetables and cereals along nearby seasonal rivers and had noted that the soils in these sub-locations had low fertility status. However, crop pests and diseases were important at Kavuthu where it took position 4 32 out of the listed 11 problems. Farmers in this site cited pest problems in cowpea and pigeonpea, which they said, lowered crop yields. However at Ndunguni poor planting methods was listed at position number 4 where farmers cited lack of equipment especially ploughs that are commonly used for cultivation in the sub-location. Distant hospital and poor services at the hospitals took position 4 out of 12 problems at Matiku. It was noted that the nearby shopping centre, Matiku market, lacked health care facilities, clinics or health centres and farmers took their sick to hospital that were over 20 km away for treatment, where the services where poorly offered. It was also noted that the sub-location was hilly that made transportation The commonest difficulty. soil types were black cotton, red and sandy soils with means of 26.8%, 31.3% and 36.3% respectively across the sites. In addition, dominant soil type at Kavuthu was black cotton soil (60%), Ndunguni was red soil (60%), Matiku and Yikivumbu was sandy soils with estimated cover of 45% and 75%, respectively. According to participating farmers, some of the characteristics of infertile soils are powdery nature of a soil, presence of bright red colour, soil hardness to till. The soils in all the study sites were described as infertile with low to very low fertile soils covering 40%, 100%, 75% and 98% of Kavuthu, Ndunguni, Matiku and Yikivumbu, respectively. These results can be compared to the findings of Mach aria (2003) where farmers classified classification, soils at Kasikeu division of Makueni Unlike scientific which considers whole soi I profile, farmer classification based on physical characteristics hardness. district. of soils was of the soil, such as surface colour, drainage and 33 Soil fertility status in the sites could be tied to soil fertility management practices and farmer beliefs. From the results in soil fertility status, it appeared that most farmers in the study sites were in farmer-class 3 or 4 of farmers, who rarely terraced or used animal manure to enhance soil fertility in their farms. In addition, farmers, animal manure was the main input used by farmers to enhance soil fertility. A similar finding was also documented by Probert et a/., (1995), who found out that in the drylands of eastern Kenya animal manure was the main farm input and was applied without measurement. A related finding was documented in West Pokot where farmers used farmyard manure for crop production without measuring the amount of manure they used (Wanjekeche et a/., 2000). Maize was their staple food crop and of highest priority 111 all sites followed by common bean that took position 2 in all sites except at Matiku where bean took position 4, because farmers experience was that pigeonpea and cowpea yields had j i i,j been better than those of bean in the site. Pigeon pea took position 3 at Kavuthu and Ndunguni and was therefore the second most important gram was the third most important legume at Yikivumbu legume in the sites. Green after common bean and cowpea because farmers planted the legume largely for income generation. It was also found out that in all the study sites legumes were sold to generate income because they fetched better cash in the markets than cereals. Finger millet was the second most important cereal after maize in all sites because it was used for domestic consumption and for income generation. Farmer crop prioritization was also reported by Onyango et al., (2000); Okoko and Makwaro (2000) who found out that maize was the most preferred crop in other parts of Kenya. 34 Food availability to households was highest between February and April because this was the harvesting season for the short rains crop. The farmers also revealed that this was the most reliable rain season compared to the long rains. After the harvesting season, household food availability dropped in all sites, with sharp drops at Ndunguni and Yikivumbu. According to the agricultural extension officers, the two sites had high incidences of crop failures due to rainfall unreliability. the farmers from Matiku Sub-location, the sub-location However, according to occasionally got a crop in the short rains that matured in August and boosted household food availability in the sublocation. The participating sites by estimating farmers measured household food availability in the study proportions of farmers buying food from the food stores in the market centres at the study sites. Some farmers started purchasing food as from June and their numbers increased as the year progressed and peaked at September, October and November at Yikivumbu, Ndunguni and Kavuthu, respectively. Farmers obtained their livelihood from crop growing and livestock keeping. However, contribution to household income from crops on average (41 %) was higher than that from livestock (35%). This largely indicated that most farmers grew crops compared to keeping livestock. The farmers also spend most of their income to purchase food, thus confirming the annual food deficit in the study sites. 35 CHAPTER FIVE 5. AREA UNDER GRAIN LEGUMES AND PROBLEMS FACED BY FARMERS IN LEGUME PRODUCTION 5.1 Introduction Grain legumes are valued for their multiple uses as food, green manure, fodder and cover for crops. Besides these direct benefits, legumes generate cash and grain legumes fetch more cash when sold compared to cereals such as maize. In Kenya, many types of grain legumes are grown in various parts of the country. However common legumes cultivated in the country include bean (Phaseolus pigeon pea (Cajanus cajan (L.) Millsp), cowpea (Vigna unguiculata vulgaris L.), (L.) Walp) and green grams (Vigna radiata Wilczek). 5.1.1 Bean (Common bean) (Phaseolus Bean is the commonest legume incorporated vulgaris L.) in different cropping systems in Kenya (Chui and Nadar, 1984). The main bean production areas in the country are Eastern, Central, Western and Nyanza Provinces at altitudes varying from 1,500 to 2,500 m above sea level. In addition, common beans are grown in Eastern Province, there are two rainy seasons (Masumba, with a mean of between 1984; Stoetzer and Waite, dietary protein and contributes 500 to 800 mm annually 1984). Like other legumes, to the maintenance where bean provides of soil fertility (Chui and Nadar, 1984). Both pests and diseases attack beans either in the field or during storage. Common insect pests of economic of importance to bean include: 36 • Bean fly or bean stem maggot (Opyomyia phaseoli), usually common in the seedling stage where it feeds on the stem base from inside the bean plant. • Black bean aphid (Aphis fabae Scop.), found usually during cool dry periods and sacks sap from pods, leaves, leaf stalks and stems. • Bean leafhopper (Empoasca solana Delong), common during vegetative stages where it feeds on leaves and flowers. • Bean weevil (Acanthoscelides obtectus Say.), common in seed storage, where it feeds on the grain. Dimethoate is used to control bean fly, black aphid and leafhopper, and planting early is recommended to avoid insect infestation. However, bean weevil is controlled by dusting bean seeds with super atelic at 50 g per bag before storage. Common diseases of bean in the production Synonym: Uromyces appendiculatus vignae BarcI.), that cause defoliation stem blight (Macrophomina monocultures areas are rust (Uromyces phaseolina) Pars.; Frios and yield losses; Ashy and Southern blight (Scleoatinia roflsii). In of bean, most diseases are severer than when bean are intercropped with maize. These common diseases are controlled using Benomyl and copper oxychloride, as well as rotating beans with cereals. (Van Rheenen et al., 1981; Audi et al., 1996). Some of the improved bean varieties from Kenya Agricultural Research Institute (KARI) are Katumani bean 1, Katumani bean 2, Katumani bean 9 and Kat-x-56. With an exception of Kat-x-56, that matures within 62-68 days other varieties mature within 60-65 days. The potential yield of Katumani bean I, Katumani bean 9 and Katx-56 is 1440-1980 kg/ha while that of Katumani bean 2 is 1350-1530 kg/ha (Audi et al., 1996). 37 5.1.2 Pigeon pea (Cajanus cajan (L.) MiIIsp) Kenya ranks second in pigeon pea production after India in the world (Remanandan et al., 1982). Eastern, Coast and Central provinces are the major pigeon pea growing areas of Kenya. Eastern Province leads in hectarage and about 90% of the total area under pigeon pea is in the province. In Kenya, both green and dry grain of pigeon pea is used as food while the plants are used for animal feed, green manure and as a cover crop (Khan, (Sheldrake 1973). Pigeon and Narayana, pea has a residual nitrogen (N) of about 40 kg/ha 1979). Both diseases and pests attack Pigeon pea. For example wilt that is caused by Fusarium undum Butler is the commonest disease of pigeon pea in Kenya, while serious pigeon pea pests in the country include thrips (Megaluro-thrips rjostedti Tryhom), Heliothis tHeliothis armigera) and pod-sucking bug (Clavigralla gibbosa) (Shakoor et al., 1984a). Pigeon pea has several improved varieties that include Kat 60/8, Kat 81/3/3, Kat 777 and ICPL 89091 that mature within 136-150, 170-185, 160-180 and 120 days respectively. Their yield potentials are 1200-1500 kg/ha, 1400-2500 kg/ha, 1400-2200 kg/ha and 1000 kg/ha respectively (Audi et al., 1996). 5.1.3 Green gram (Vigna radiata Wilczek) Green grams (mung bean) are an important crop in the warm dry parts of eastern Kenya where it is grown for both subsistence and as a cash crop (Shakoor et al., 1984b). Dry whole grain is used for food, although the Asian community, the largest consumer of the crop, cooks it as split grains (Dhai). Grain protein content of green gram varies from 21-29% depending on the variety and environment where the crop has been grown. Mung bean is generally free from flatulence inducing factors that are 38 common in many grain legumes. Sulphur amino acids, methionine and cystine content are low as in other legumes and lysine is high (8gll OOg protein dry weight) (Shakoor et 01., 1984b) Mung bean is pan-tropical and is able to grow in adverse conditions. It can escape drought through its early maturing ability (Rowe, 1980) and some varieties are more resistant to drought than cow pea (Waite et al., 1984). Pests of green gram are thrips, aphids, pod-sucking bugs, apion beetle and bruchids while common diseases include powdery mildew and yellow mosaic virus (Audi et al., 1996). In addition, improved varieties of green gram include KVR 22 and KVR 26 with potential yields of 1000 to l300 kg/ha (Audi et al., 1996). 5.1.4 Cowpea (Vigna unguiculata (L.) Walp) Cowpea is an annual or bi-annual grain legume commonly referred to as cowpeas. In Kenya, it is the third most important grain legume after beans and pigeon pea and covers about 18000 ha, excluding the cowpea grown in home gardens (Muthamia and Kanampiu, 1996). About 85% of the total area under cowpea is in arid and semi-arid lands (ASALs) of Eastern Province and 15% in the Coast, Western and Central Provinces (Muruli et al., 1980; Muthamia and Kanampiu, both indeterminate predominant 1996). In Eastern Province, and semi-erect types are grown, while indeterminate in Western and Central Provinces. Chaturverdi that indeterminate types give better yields under drought land races are et aI., (1980) indicated conditions than the determinate types. Cowpea is commonly grown in mixtures with maize, sorghum or pigeon pea. In Eastern Province cowpea is grown for both grain and leaves while in Western and Central Provinces it is mainly grown for its leaves (Shakoor et al., 39 1984a). Cowpea is used for food, fodder and as a source of income. Leaves, young pods and grain are the parts of the plant used for food. The same plant parts are sold to generate cash for farmers. Cowpea value lies in its high protein content of 10% 35% (Imungi and Porter 1983), its ability to tolerate drought and fix atmospheric N, which allows it to grow and improve poor soils. It has a well-developed deep root system and grows well under drought conditions (Shakoor et al., 1984a; Muruli et al., 1980). Major insect leafhoppers pests of (Empoasca cow pea are thrips (Megalurothrips sp), legume pod borers (Maruca armigera), aphids (Aphis craccivora Koch), pod-sucking and apion beetle (Apion solenortum). rjostedt testulalis Tryhom), and Heliothis bugs (Clavigralla gibbosa) Bruchid is a pest of dry cowpea grain (Shakoor et al., 1984a; Audi et al., 1996). Some of the improved cowpea varieties from KARI include Machakos 66, Katumani 80, KVU-419 and KVU HB 48EI0. These varieties matures within 85-95, 75-85, 65-72 and 85-95 days, respectively, and have yield potentials of 1200-1800 kg/ha, 1500-1800 kg/ha, 1170 kg/ha and 1200-1500 kg/ha, respectively (Audi et al., 1996). Legume production in the dryland is very low partly because soils are commonly nutrient deficient, especially of nitrogen (N) and phosphorus (P). This is iargely due to continuous cropping without external inputs (McCown et al., 1992). However many farmers use manure because they are aware of its benefits but the quanties avilable are insufficient and of poor quality (Probert et al., 1995). As a result, crop yields are low and yields of grain legumes rarely exceed 500 kg/ha (Mathuva et al., 1996). 40 5.2 Materials and Methods On-farm surveys using a structured questionnaire four sub-locations covering Kenya. The sub-locations Kavuthu and Matiku reconnaissance survey Agricultural two divisions (Appendix of dryland Makueni included Yikivumbu and Ndunguni In Mbitini division. The 2) were carried out in sites had on the basis of their accessibility district in eastern in Nguu division, and been during selected rainy seasons. extension officers of each division assisted in site identification the reconnaissance surveys. To implement questionnaires, random visits to homesteads. The number of household in a during the process involved single During the visits, household heads were interviewed. heads interviewed were 32, 23, 21 and 43 at Yikivumbu, Ndunguni, Kavuthu and Matiku, respectively, and were determined with the help of a biometrician Ndunguni, participatory from the total households Kavuthu meetings and Matiku, (Chapter (350, 234, 216 and 443 at Yikivumbu, respectively) documented during farmer 4). Samples of weeds, pests and disease-affected crops were taken from Ndunguni and Kavuthu during the on-farm trials and brought to Kenya Forestry Research Institute (KEFRI) laboratories where a plant taxonomist, an entomologist and a pathologist identified relevant samples. Data collected was analysed using SPSS for windows Release 10.0 of 1999. The analysis was done by sub-location to allow for comparison between sites. 41 5.3 5.3.1 Results Grain legume production and area under legumes Yikivumbu had largest farm sizes followed by Ndunguni, Kavuthu and finally Matiku and the total cultivated operated generally farm sizes were in the same order. The number of farms increased with decreasing total farm size. In addition, the area occupied by cultivated legumes increased with decreasing farm size and ranged from 48% to 92% (Table 5.1). Table 5.1 Site Yikivumbu Ndunguni Kavuthu Matiku SED Area under grain legume production in selected sub-locations Area Under legumes (ha) 1.2 l.1 Proportions ofthe Cultivated area under grain legumes (%) 48 55 l.5 83 1.2 l.l 92 0.69 0.57 Average Total Cultivated number of farm size farm size (ha) fields (ha) 1.1 4.9 2.5 1.4 4.4 2.0 l.8 2.1 2.3 2.0 0.25 1.5 1.28 Main legumes grown in the study sites included common bean, cowpea, green gram and pigeon pea; with grain yield ranges of 30 to 416 kg/ha. Yikivumbu yields of cowpea had highest and green gram of 239 kg/ha and 416 kg/ha, respectively. addition highest amounts In of common bean and pigeon pea of 250 kg/ha and 189 kg/ha, respectively, were recorded at Ndunguni (Table 5.2). 42 Table 5.2 Average legume yields (kg/ha) in selected sub-locations Sub-location and legume yield Crop Kavuthu Matiku Ndunguni Bean 42 91 250 Pigeon pea 81 178 189 Cowpea 107 130 102 Green grams 178 30 60 Totals 260 459 719 SED means standard errors of differences of means 5.3.2 Soil fertility distribution, According to farmers' indigenous Yikivum bu 63 187 239 416 905 Grain yield totals 446 635 578 684 SED 29 35 42 88 farm inputs and their sources knowledge, soil fertility status in the study sites could be described as fertile, moderate or poor (Table 5.3). Most of the farms in the study sites had moderate fertility status. With exception of Kavuthu, fertile soils in other sites covered less than 10% of the cultivated areas. In addition, Kavuthu and Matiku had highest percentages of farms with poor soils that covered about 24% and 23%, respectively. Table 5.3 Soil fertility status of farms in the selected sites of Makueni District Sub-location Yikivumbu Ndunguni Kavuthu Matiku Means Fertile 0 21.7 5 7 8.4 Soil Fertility Status (%) Moderate 100 69.6 71 70 77.7 Poor 0 8.7 24 23 13.9 Farmers in the study sites used farmyard manure (animal manure), compost and crop residues and inorganic fertilizers to enhance soil fertility (Table 5.4). Animal manure was the main farm input and between 84% and 97% of the farmers interviewed said KENYATTA UNiVERSITY LIBRARY} ! 43 they used it, while farmers using crop residues ranged from 6% to 39%, and those using compost ranged from 0 to 70%. Use of inorganic fertilizers was only recorded at Ndunguni. In addition, between 84% and 100% of the farmers interviewed used animal manure from their own animal kraals (Table 5.5). Table 5.4 Farmers (%) using inputs to enhance soil fertility Sub-location Yikivumbu Ndunguni Kavuthu Matiku Means Table 5.5 Fertilizer 6 3 o 39 70 9 o o o Crop residues 87 91 10 14 17 16 22 84 90 2 Sources of animal manure used by farmers Sub-location Yikivumbu Ndunguni Kavuthu Matiku Means 5.3.3 Compost Manure 97 Manure sources (%) Bought Owned 100 3 3 35 91 100 84 94 Borrowed 17 5 o 2 11 9 7 Problem weeds, diseases and pests There were two main types of problem weeds recorded, grass and non-grass types. Non-grass weeds were 'Mukuutu' 'Mung'oi' (Acanthospernul11 (Trichodesma hispidum zeylanicum (DC», 'Munzee' (Burm.f.) R. Br.), black jack (Bidens pilosa (L.» and 'Uthunga' (Launaea cornuta (Oliv. & Hiern) O. Jeffrey). The grass weeds included 'Kithangai' couch grass tDigitaria scalarum (Schweinf.) Chiov.), 'Ikoka' star grass (Cynodon dactylon (L) Pers.) and 'Mbiu' (Cyperus rotundus (1.». Presence of weeds in farmers' fields varied across sites. For example all farmers interviewed (100%) at Yikivumbu said that 'Mung'oi' and 'uthunga' were present in their farms 44 while at Kavuthu 91 % of the farmers interviewed said that 'Mukuutu', 'Mung'oi' 'Munzee' and were present in their farms (Table 5.6). Pests in this study were divided into two, field and storage pests. Field pests included animals such as squirrels and monkeys, birds and insects such as aphids and white flies (Bemisia tabaci Gennadius), while weevils were the storage pests. Documented diseases were rust and powdery mildew (Erysiphe polygoni 5.7). All sites experienced DC ex St.-Am) (Table pest and disease infestation at varying levels. For example monkeys were commonly repotted at Ndunguni (39%) while squirrels and birds were commonly reported at Yikivumbu and Ndunguni. Table 5.6 Farmers (%) that reported common weeds in their farms Site 1 2 3 4 5 6 7 Yikivumbu 9 28 97 41 78 100 100 Ndunguni 52 74 96 83 96 96 96 Kavuthu 62 62 91 91 67 81 91 Matiku 40 49 81 72 61 61 81 Means 41 53 91 72 76 85 92 1 - Digitaria scalarum, 2- Cynodon dacylon, 3 - Trichodesma zeylanicum, 4 Bidens pilosa, 5- Cyperus rotundus, 6- Launaea cornuta and 7- Acanthospernum hispidum Table 5.7 Farmers (%) that reported pests and diseases in their farms Storage Field pest Site Monkeys Squirrels Birds Pests Aphi ds White flies Weevil s 6 lOO lOO 97 97 2 lOO 39 lOO 74 74 3 6 95 86 86 95 4 72 84 9 97 61 Mean 15 98 90 85 82 1) - Yikivumbu, 2)- Ndunguni, 3- Kavuthu and 4)- Matiku I Diseases Powdery mildew Rust lOO 78 88 74 86 88 88 91 81 84 100 lOO 74 lOO 94 45 5.3.4 Household head education and food sources and availability Majority of the household heads (between 43% and 70%) had primary education only (Table 5.8). In addition, about 90% of households experienced food shortage during the long dry spells and on average over 95 % of the household heads purchased food during the long dry spells experienced in the drylands (Table 5.9). Table 5.8 Level of formal education (%) of household heads Sub-location None Yikivumbu 15.6 Ndunguni 4 Kavuthu 19 Matiku 12 Means 13 N/A means not available. Table 5.9 Site Yikivumbu Ndunguni Kavuthu Matiku Means Primary 65.6 70 43 58 59 Secondary 15.6 26 38 30 28 University 3.1 N/A N/A N/A 0.75 Food availability and source (%) during long dry spells Food shortage was experienced during the long dry spells Yes No 100 91.3 8.7 95.2 4.8 95.3. 4.7 95.5 4.6 o Purchasing was the main source of food during long dry spells Yes 100 87 100 No 95.3 95.6 4.7 4.4 o 13 o 46 5.4 Discussion Grain legume cover ranged from 48% to 90% increasing with decrease in cultivated farm size. The intercropping increase in land under of grain legumes legume with cereals, cover was commonly due maize. legume and cereal crops is a common practice of smallholder to increased Intercropping of farmers throughout the tropics (Sakala et al., 2000) and in East Africa maize is commonly intercropped or rotated with grain legumes (Pilbeam et al., 1995). Sites with relatively larger cultivated areas, Yikivurnbu and Ndunguni higher grain yields probably important had relatively indicating that farm sizes in the study sites were more than the area under grain legume cultivation. Commonly grown grain legumes had mean grain yields recorded as 250 kg/ha, 189 kg/ha, 239 kg/ha, and 4 I6 kg/ha for bean, pigeon pea, cowpea and green grams, respectively, which were below the potential grain legume yields documented by KARI, where potential grain yield of common bean lies between 1350- 1980 kg/ha, that of pigeon pea between 1200-2500 kglha, green gram between 1000- 1500 kg/ha and that of cowpea between 1200-1800 kg/ha (Audi et al., 1996). Most farms in the study sites had moderate soil fertility status followed by poor soils and finally fertile soils with overall means of 77.7%, 13.9% and 8.4%, respectively. This implied that fertile soils covered less than 10% of the cultivated farms in the selected sites. These participatory results agreed with the findings from farmer meetings in the same study sites (Chapter 4), which revealed that, fertile soils covered less than 10% of cultivated farms. According to the farmers, soil erosion was the main cause of soil infertility and 81 %, 22%, 62% and 71 % of the farmers interviewed at 47 Yikivumbu, Ndunguni, they had problems Kavuthu and Matiku Sub-locations, of soil erosion reported in the sub-locations respectively, said that in their farms. The high erosion observations were largely due to free-range mode of livestock keeping in the study sites that was noted at 97%, 87%, 81 % and 84% at Yikivumbu, Ndunguni, Kavuthu animal keeping and Matiku often causes overgrazing However, higher respectively. between seasons Free-range (Sanchez mode of et al., 1997; et al., 1993) that may cause soil erosion and soil Smaling et al., 1997; Stoorvogel degradation. Sub-locations, proportions of poor soils at Kavuthu (24%) and Ndunguni (23%) were most probably due to continuous cropping without farm inputs, removal of nutrients in harvested crops, runoff and erosion, and leaching (Sanchez et al., 1997; Smaling et al., 1997) Farmers improved their soils mainly by the use of farmyard manure (manure). The manure was readily available because most households kept livestock. This was confirmed by the incidence of livestock ownership of 100%, 96%, 100% and 98% at Yikivumbu, Ndunguni, Kavuthu and Matiku, respectively. and 86% of farmers at Yikivumbu, respectively, manure Ndunguni, Kavuthu However, 98%, 91 %,67% and Matiku sub-locations, said that animal manure was not adequate. from their own animal Farmers commonly used kraals and only a few farmers, especially from Ndunguni, bought or borrowed manure from their neighbours. Other inputs were crop residues, compost and fertilizers, that were used at small amounts and commonly at Ndunguni. Ndunguni sub-location borders Muoni River and the residents practiced irrigation of vegetables that required a lot of farm inputs, which explained high farm inputs and higher proportion of fertile soils in the sub-location. suggested that low soil fertility and low use of organic Bekunda et al., (1997) and inorganic mineral 48 fertilizers are productivity the greatest biophysical constraints to increasing agricultural in the farming systems in the semiarid regions of Africa. There were seven problem weeds recorded in the study sites of both grass and nongrass types. According to the farmers these weeds were very competitive controlled could significantly and if not lower legume yields. They said some of the weeds were most difficult to control especially 'Kithangai' and 'Mbiu' rhizomes. Two of the non-grass weeds 'Mukuutu' which have underground and 'Mung'oi' became aggressive after first weeding and farmers were forced to do about three weedings to ensure good legume yields. The 'Munzee' and according and 'Uthunga' to the farmers weeding germinate immediately after planting had to be done two weeks after planting. Majority of the farmers used hand hoes to control weeding after legume germination. However, where the legumes were grown in rows, the farmers used ox-ploughs remove weeds between the crop rows. Farmers 'Mukuutu' Farmers and 'Mung'oi' at Ndunguni also said they occasionally to burnt during the dry seasons, commonly before planting seasons. said that they culturally controlled 'Kithangai' by planting pumpkin (Cucurbita maxima Duchesne) plants that chocks the weed. Pest infestation varied across sites because of variation in location and vegetation cover. For example monkeys were commonest in Ndunguni because the sub-location bordered called Nguu where an unsettled and forested Squirrels and birds were commonest scheme in Yikivumbu monkeys lived. and Ndunguni because the sub- locations had large uncultivated and bushy areas, which provided hiding places for the birds and the squirrels. Monkeys, birds and squirrels were controlled by scaring them away and sometimes by use of locally made traps. According to farmers, pests and 49 diseases were controlled by use of chemicals although cultural control was also used. Farmers at Yikivumbu mildew (Erysiphe sub-location po lygoni De. said they used wood ash to control powdery Es St.-Am.) and white flies (Bemisia tabaci Gennadius). However, farmers in all the sites said that they control rust by use of crop rotation. Most household Ndunguni heads had attained 70%, Kavuthu primary school education (Yikivumbu 43% and Matiku 58%) and commonly 66%, relied on crop growing and livestock keeping. It was noted that 88%, 87%, 95% and 70% of the household heads at Yikivumbu, Ndunguni, Kavuthu and Matiku, respectively, were engaged in livestock keeping and crop growing. This most probably resulted to the food shortage documented in the study sites and consequent buying food. Poshiwa et aI., (2006) found out that additional years of schooling have significant positive effect on farm output and gross values of farm production. It was interesting to note that not all farmers who experienced food shortage at Ndunguni and Matiku bought food most probably because Ndunguni borders Muoni River and most farmers plant crops, especially maize, that provide food to households during the long dry spells. It was observed that (Chapter 4) farmers at Matiku got a crop during the long rains that boosted household food availability during the long dry spells. 50 CHAPTER SIX 6. SCREENING NEW COWPEA VARIETIES FOR DRYLANDS OF EASTERN KENYA 6.1 Introduction Cowpea (Vigna unguiculata [L.] walp) belongs to the tribe Phaseoleae Van der Maesen, 1985). It was domesticated Africa (Ng and Marechal, in Sub-Saharan 1995) and is a major component (Polhil and Africa, probably West of traditional cropping systems in the drier parts of the tropics. Cowpea is important because of its multiple uses, which suppression include improving soil fertility through biological nitrogen fixation, of weeds, the green pods and dry grains are eaten as well as leaves and green pods. Green leaves may be consumed or sold to generate household income (Muli and Saha, 2000). It is also grown for forage and for use as a green manure (Tarawali et al., 1997). Cowpea is a source of protein with mean crude protein of leaves, grains and crop residues ranging from 32-34%, respectively, and contains 62% soluble carbohydrates. carbohydrate content cowpea is used in nutritional 23-35% and 11-25%, Due to its high protein and products (Imungi and Porter, 1983). In Africa, West Africa is the key cowpea producing zone mainly in the dry Savannah and semi-arid agro-ecological zones. Nigeria is the largest cowpea producer and consumer in the world while Niger is the largest cowpea exporter in the world with an estimated 215, 000 metric tones (MT) exported annually mainly to Nigeria. World cowpea production was estimated at 3,319,375 metric tones with 75% of that production being from Africa, 21 % from South America, 1% from Europe, 2% from Asia, and 1% from North America (FAOST AT, 2000). West and Central Africa 51 account for over 64% of the estimated wide followed 12.5 million ha cultivated by Central and South America to cowpea world (19%), Asia (10%) and East and Southern Africa (6%) (Singh et al., 1997). Cowpea is commonly grown in the arid and semi-arid second to pigeon pea in productivity areas of Kenya and rates (Waite et al., 1984). The area under cowpea in Kenya is 1800 ha excluding cowpea in home garden, with about 85% of the area under cowpea production and Kanampiu, as sorghum being in the arid and semi-arid lands (ASALs) (Muthamia 1996). Cowpea is commonly grown in mixtures with other crops such (Sorghum bieolor (L.) Moench), (Manihot eseulenta Crantz) (Mortimore advantages of intercropping maize (Zea mays (L.), and cassava et al 1997; and Van Ek et al., 1997). The are that it provides crops that enhance balanced diet and reduces labour demand; it minimizes crop failure risk and adverse effects of pests, maximizes returns per unit area of land and reduces soil erosion (Nadar, 1984). However, cowpea grown in mixtures has low yields than those in pure stands. For example Muleba and Ezumah (1995) found that grain yields of cowpea in field intercrops ranged between 0-133 kg/ha compared to its potential yield of 1500 kg/ha to 3000 kg/ha. Factors limiting cowpea yields are low plant population density, low yield potential of local cultivars, insect pests and diseases, shading by the cereals, drought stress and low soil fertility. For example cow pea growth is retarded especially low phosphorus and micronutrients by poor soil fertility (Bationo et al., 1991). 52 6.2 Materials and Methods A total of 34 improved cowpea varieties were used in this study. Of the 34 varieties, 30 varieties were obtained from (lJTA) and consisted maturing varieties; of 15 early and 15 medium two varieties from Maseno, TSBF -CIA T trials, and 2 varieties were obtained from a local farmer. Early maturing varieties obtained from lIT A with given codes were: IT97K-568-18 IT97K-461-4 (E4), IT99K-1060 IT97K-568-18 (E5), IT97K-570-18 IT99K-II22 (E8), IT98K-429-2 (EI2), IT96K-6IO (EI3), IT98K-428-3 Medium maturing IT98K-I31-2 (El), (E9), IT98K-506-1 (E2), IT97K-494-3 (E3), (E6), IT97K-499-38 (E7), (EIO), IT97K-356-1 (El I), (EI4) and IT98K-1399 (EI5). varieties with given codes included IT99K-491-7 (M1), IT95K- 193-12 (M2), IT98K-503-1 (M3), IT95K-207-22 (M4), IT99K-205-9 (M5), IT98K- 128-4 (M6), IT97K-499-35 (M7), IT95K-1073-57 (M8), IT97K-818-35 (M9), IT97K- 556-4 (MIO), IT98K-463-7 (MI I), IT97K-I021-24 (MI2), IT97K-I075-7 (M13), IT95K-52-34 (M14) and IT97K-564-1 (MI5). Varieties from Maseno included IT98K 247-2 (CPI4) and IT97K -1068-7 (CP21), while local varieties used were Kathoka and Kang'au. The experiment Institute (KARI) dryland was established at a Kenya Agricultural research centre, Kiboko, Research 960m above sea level, 2015'S; 37°45'E and ran for two seasons, during the short and the long rain seasons. The experiment was set up as a completely randomized design (CRD) with each variety replicated 3 times, in rows (Muli and Saha, 2000). Each row had 48 plants spaced at 20 x 60 cm (Audi et al., 1996), covering an area of 5.76 m2. The experiment was both rain-fed and irrigated. Overhead sprinklers that were set to run for 3 hours at each irrigation cycle, discharging an equivalent of about 32 mm rainfall, were used. In the first season, planting was done on IOIl 11 2004 and irrigation was done when rainfall was limiting. In the second season, planting was done on 6/4/2005 and irrigation was 53 done as in short rains. In both seasons and at planting time, phosphorus triple superphosphate (TSP) (P20S, was not limiting. Weeding in the form of 0:46:0) was applied at 20kg/ha to ensure that P was done 3 times each season. At flower onset, early podding and at full podding in both seasons, the crop was sprayed with Karate® at a rate of20 ml per 20 litres of water to control white flies and aphids. To eliminate edge effect, only 40 plants were harvested leaving 4 plants on either side of the row. Data collected included pod numbers, pod length, seeded pod weights, number of seeds in each pod, seed weight per pod, number of branches, plant biomass production, grain yields and, shell biomass, total above ground and weights of 100 seeds. To determine grain dry weights, shell and shoot biomass, sub-samples and dried in a Sanyo Connection Oven (MOV-2l2F, were taken to the laboratory Sanyo Electric Company, Japan) at 60°C for 72 hours. However, to select pioneer species for further studies, means of the assessed characters were selected were totaled. Cowpea varieties that had highest mean totals for on-farm trials. The data obtained was analysed Discovery Edition 1, GenStat Procedure Library Release PL12.2. using GenStat 54 6.3 Results and Discussion Pod weights, seeds per pod and seed weight per pod were generally higher during the short rains than during the long rains (Table 6.1 and 6.2). This was probably because of an infestation (Nephtys of the cowpea spp). In addition, length, seeded pod weights, plants during early growth stages by cat worms medium maturing varieties had generally higher pod seeds per pod and seed weight per pod than the early maturing varieties in both the short and long rain seasons (Table 6.1 and 6.2). Pod lengths ranged from 11.2 cm to 17.2 cm and from 10.6 cm to 17.3 cm, respectively, during the short and the long rains. These pod lengths were close to the range of 13.0 cm to 19.8 cm obtained by Muli and Saha (2000) in a cowpea screening trial at Msabaha, in the Kenyan Coast. The pod lengths obtained in this study were also close to the range of 7.3 cm to 15 cm recorded by Amanullah and Hatam (2000) in Pakistan. During the short and the long rains, mean pod length for all 34 varieties were 13.4 cm and 13.5 cm, respectively, and varieties that had above average pod lengths in both seasons were El, E2, E6, E7, E9, EIO, M2, M4, M6, M7, M8, M9, MIO, MII, M13, MI4, MI5, Kathoka and CP21. Of these varieties 58% were medium maturing and only 32% were early maturing varieties. Longest pod lengths were recorded in E6 (17.2 cm) and Kathoka (17.3 cm) during the short and long rain seasons, respectively. This implied that the local variety, Kathoka, had potential for long pod formation that was comparable to that of the improved varieties. Pod weight ranged from 1.1g to 2.2 g during the short rains and 0.9 g to 2.5 g during the long rains with overall means of 1.7 g and 1.5 g, respectively (Table 6.1 and 6.2). During the two rain seasons, varieties that had above average pod weights were El, E2, E7, E8, EIO, M2, M4, M7, M8, M9, MIO, MII, MI2, M13, MI4, MI5, Kathoka, 55 CPI4 and CP21 with 58% of the varieties being medium maturing and 26% early maturing varieties. However, varieties with heaviest pods of 2.2g were E8, M9 and CP21 during the short rains and M 13 (2.5 g) during the long rains. Kathoka had a pod weight of2.3 g during the long rain, which was higher than the highest pod weight in the short rains implying that the local variety had a high potential to produce heavy pods. The correlations respectively, between pod length and pod weight were 54% and 53%, during the short and the long rains, implying that pod length influenced pod weight in both seasons. During both the short and the long rains, seed numbers ranged from 4 to 12. However, the average seeds per pod (seed numbers) for the 34 varieties studied were 8 and 7, respectively, during the short and long rain (Table 6.1 and 6.2). These seed numbers were comparable to cowpea seeds/pod recorded by Amanullah and Hatam (2000) that ranged from 7 to 15 with a mean of 9 seeds/pod. Varieties that had above average seed numbers during both the short and the long rains seasons were E2, E7, E8, M2, M4, M7, M8, M9, Mll, M13, M14, Kathoka and CP21. Of these varieties 62% were medium maturing while 23 % were early maturing varieties. Highest seed numbers (12 seeds per pod) were recorded in Kathoka and E8 during the short rains, and in Kathoka (12 seeds per pod) during the long rains. This meant that the local variety, Kathoka, was a heavy seeder and was comparable to the improved variety E8 with respect to the number of seeds produced. Pod length influenced the number of seeds as reflected by significant (p<0.05) correlations between pod length and seed numbers of 61 % and 69%, respectively, during the short and the long rains. Seed numbers and pod weights correlations had significant of 72% (p<O.OI) and 61 % (p<0.05), respectively, during the short and the long rains. Significant correlations between seed 56 numbers and pod length were also reported by Nakawuka and Adipala (1999) 111 Uganda. Seed weight per individual pod ranged from 0.9 to 1.7 g during the short rains and from 0.6 to 1.8 g during the long rains. Average seed weights per pod for 34 varieties were 1.3 g and 1.0 g, respectively, during the short and the long rains (table 6.1 and 6.2). During the two seasons varieties that had above average seed weights were El, E2, E5, E7, E8, ElO, E13, M2, M4, M7, M8, M9, MIO, Mll, M12, M13, M14, Kathoka, CP14 and CP21. Of these varieties medium maturing varieties contributed 50% while early maturing ones had 35%. This was the highest contribution by the early maturing varieties in the parameters so far measured (pod length, pod weight, seed numbers and seed weight per pod). Highest seed weight per pod was recorded in M8 and M9 (1.7) and M13 (1.8 g) during the short and long rains, respectively. variety, Kathoka had seed weight per pod of 1.6 g and 1.5 g, respectively, Local during the short and the long rains, implying that the variety had high potential for heavy seed production. There were significant (p<0.05) correlations between seed weight and pod lengths of 58% and 65%, respectively, weights were also significantly of 71 and 62 %, respectively, significant during the short and the long rains. Seed (p<O.Ol) correlated to pod weights, with correlations during the short and the long rains. The strongest and (p<O.01) correlations of 81% and 83%, respectively, during the short and the long rains were between seed number and seed weight. This implied that seed weight was strongly influenced by seed numbers in both seasons. 57 Table 6.1 Pod characters assessed during the short rains Variety El E2 E3 E4 E5 E6 E7 E8 E9 EI0 Ell E12 E13 E14 E15 Ml M2 M3 M4 M5 M6 M7 M8 M9 M10 Ml1 M12 M13 M14 MI5 CP14 CP21 Kathoka Kang'au Pod length (cm) 12.5 12.5 13.3 11.2 11.6 17.2 11.9 12.8 13.8 14.2 11.3 13.1 12.9 12.7 12.6 11.8 15.1 12.9 14.2 11.6 13.8 13.6 11.8 15.0 15.4 13.6 13.3 12.3 Mean of 34 varieties 15.3 13.5 13.3 16.3 15.2 12.7 13.4 SED 0.42 Seeded pod weight (g) 1.3 1.4 1.5 1.1 1.3 1.6 1.7 2.2 1.4 1.9 1.4 1.3 1.7 1.4 1.7 1.3 2.0 1.6 1.6 1.4 1.5 1.8 2.0 2.2 2.0 2.0 2.0 1.7 1.9 1.9 1.7 2.2 2.1 1.6 1.7 Seeds numbers Seed weight per pod (g) 7 7 7 4 0.9 1.1 1.2 6 8 8 12 7 8 6 7 7 6 8 8 9 7 6 7 8 8 8 9 7 9 7 10 9 7 8 9 12 8 8 0.44 0.9 0.9 1.1 1.3 1.6 1.1 1.6 1.1 1.0 1.4 1.0 1.4 1.0 1.6 1.2 1.2 1.2 1.2 1.5 1.7 1.7 1.5 1.5 1.4 1.3 1.4 1.3 1.3 1.5 1.6 1.3 1.3 0.12 0.09 Significance ** ** ** ** ** Means p<O.OOI. E and M means early and medium maturing cowpea varieties, respectively. Data values are means of 3 replicates. 58 Table 6.2 Pod characters assessed during the long rains Variety El E2 E3 E4 E5 E6 E7 E8 E9 EI0 Ell E12 E13 E14 E15 Ml M2 M3 M4 MS M6 M7 Pod length (cm) 14 15.5 12.9 10.8 12.3 15.7 14.9 10.6 14.2 13.9 12.2 11.7 12.6 12.9 11.9 11.6 14.3 11.7 13.8 11.4 16.0 15.0 13.8 14.1 15.7 13.7 11.9 M8 M9 M10 M11 M12 M13 M14 MI5 CP14 CP21 Kathoka Kang'au 15.9 17.3 11.7 Mean of 34 varieties 13.5 SED 0.75 15.3 15.3 12.7 13.2 Seeded pod weight (g) 1.7 2.0 1.2 0.9 1.2 1.1 1.6 1.3 1.4 1.4 1.4 1.2 1.4 1.0 1.3 1.2 1.7 1.0 1.9 1.0 1.3 1.9 2.1 2.3 1.6 1.4 1.0 2.5 1.5 1.2 1.7 1.8 Seeds per pod 7 9 6 4 6 6 8 7 6 7 5 6 6 5 6 6 9 6 8 6 7 8 10 6 7 10 4 Seed weight per pod (g) 1.1 1.5 0.9 0.7 1.5 0.7 1.1 1.0 1.0 1.1 0.7 0.7 0.8 0.8 1.0 0.9 1.3 0.8 1.2 0.7 0.9 1.4 1.6 0.9 1.1 1.0 0.6 1.8 1.0 1 1.5 11 6 6 7 9 12 7 7 0.7 1.0 0.32 0.93 0.18 2.3 0.8 1.2 1.2 1.5 Significance ** ** ** ** Means p<0.001. E and M means early and medium maturing cow pea varieties, respectively. Data values are means of 3 replicates. ** 59 Branch numbers, pods per plant, plant biomass, grain yield, shell weight, total above ground and weight of 100 seeds were generally higher during the short rains than during the long rains. In addition, medium maturing varieties had generally higher values for all parameters measured during both the short and the long rain seasons compared to the early maturing varieties (Table 6.3 and 6.4). Cowpea branching is an adaptation for high and efficient translocation development branches, by of a substantial root system (Olufajo and Singh, 2000). The numbers of nodes intercropping system that is usually accompanied and internodes length (Nelson and Robichaux, are plant traits that are important in 1997). Thus cultivars with a bushy type habit are high yielding under sole cropping, whereas the cultivars with a spreading habit are higher yielding under intercropping (Nelson and Robichaux, 1997). Branch numbers varied from 2 to 6 during the short rains and from 0 to 5 during the long rains (Table 6.3 and 6.4). These branch numbers were lower than number of branches of between 6 and 14 documented by Amanullah and Hatam (2000) at Pakistan, the differences being probably due to differences in cowpea genotypes used in the two studies. Out of the 34 varieties, mean branch numbers were 3 in both seasons. Varieties that had branches above average were El, E4, E7, Ell, E14, Ml, M2, M9, MIO, M12, Ml3, MI4, MI5, Kathoka and Kang'au. varieties 53% and 33%, respectively. contributed Of these varieties medium and early maturing Highest branch numbers were recorded in the local variety, Kang'au (6 branches) during the short rains, and in El (5 branches) during the long rains. The later observation probably implied that Kang'au had a high potential of branch formation than most of the improved varieties. 60 Table 6.3 Variety charactersassessedduring the shortrains Variety El E2 E3 E4 E5 E6 E7 E8 E9 EI0 Ell E12 E13 E14 EIS Ml M2 M3 M4 MS M6 M7 M8 M9 MI0 Mll M12 M13 M14 MI5 CP14 CP21 Kathoka Kang'au Mean of 34 varieties SED Significance BN " 2 " 4 3 3 4 2 " 3 3 3 3 4 3 " 3 3 3 " 3 3 3 4 4 2 4 4 4 4 3 3 4 6 3 0.5 ** .) .) .) .) .) PN 19 15 20 16 15 12 24 15 20 16 15 16 14 24 12 22 10 19 12 15 13 17 12 8 17 17 13 15 18 21 17 17 15 46 17 3 ** PB 1005 676 984 647 436 638 2881 803 1306 1773 986 687 784 1654 623 1226 559 1121 678 729 690 1539 1324 1178 2382 2232 791 2855 2280 1630 761 1386 1961 3933 1328 201 ** GY 1563 1359 1995 1190 1310 1094 2466 2023 1875 2213 1265 1378 1600 2052 1398 1867 1197 1845 1236 1424 1279 1955 1722 1079 1978 1917 1569 1567 2132 2318 1901 2203 1970 3435 1746 359 ** SWt 421 347 529 311 372 460 661 576 490 513 372 384 352 658 324 517 262 495 345 336 357 461 336 270 658 496 404 553 795 850 513 765 636 1110 489 lIS ** TAG 2989 2383 3508 2148 2117 2192 6007 3401 3671 4500 2623 2449 2736 4363 2345 3610 2018 3461 2259 2488 2327 3955 3382 2526 5018 4644 2763 4975 5206 4798 3175 4354 4569 8479 3572 597 ** WI00 17 16 16 22 18 16 16 14 18 18 19 17 18 16 17 14 16 17 20 16 17 19 21 19 19 16 19 14 19 18 15 18 14 18 17 2 ** ** Means p<O.OOl. E and M means early and medium maturing cowpea varieties, respectively. BNbranch numbers; PN- Pod numbers; PB- plant biomass kg/ha; GY - Grain yield kg/ha; SWt- shell biomass kg/ha; TAG- Total aboveground kg/ha and WIOO- weight of 100 seeds. 61 Table 6.4 Variety El E2 E3 E4 E5 E6 E7 E8 E9 E10 Ell E12 E13 E14 E15 M1 M2 M3 M4 M5 M6 M7 M8 M9 MI0 Ml1 M12 M13 M14 MI5 CP14 CP21 Kathoka Kang'au Mean of34 SED Significance Variety characteristics assessedduring the long rains BN 5 3 2 4 0 3 3 3 3 2 4 3 2 4 3 4 4 3 2 2 3 3 3 2 4 3 2 3 3 3 3 3 3 4 3 0.7 ** PN 30 18 14 15 4 14 16 13 12 13 13 10 10 16 13 6 10 12 8 10 13 12 18 6 19 8 7 15 12 14 10 13 8 10 12 3 ** PB 1762 1077 354 614 98 494 758 535 825 737 548 352 362 929 358 329 508 527 531 441 681 693 1320 737 1329 367 171 1479 691 462 423 746 1332 469 678 201 ** GY 2686 2245 )066 854 304 816 1379 1070 835 1214 756 585 631 460 1050 431 1035 841 693 581 1026 1445 2484 412 1722 718 367 2220 1264 985 956 1277 1031 643 1061 359 ** SWt 1120 754 352 289 106 456 646 371 371 337 312 221 206 375 267 148 298 258 254 221 427 439 579 202 704 252 133 873 569 389 299 572 508 229 398 115 >!<* TAG 5569 4075 1772 1749 508 1766 2882 1976 2030 2289 1616 1156 1200 1764 1674 908 1841 1626 1478 1243 2130 2578 4384 1352 3755 1337 671 4571 2524 1834 1679 2595 2872 1341 2140 597 ** WI00 17 17 16 20 17 17 16 14 20 17 17 14 17 15 17 14 19 19 21 18 18 18 20 14 19 14 17 21 19 16 17 18 11 18 17 9 ** ** Means p<O.OOI. E and M means early and medium maturing cowpea varieties, respectively. BNbranch numbers; PN- Pod numbers; PB- plant biomass kg/ha; GY - Grain yield kg/ha; SWl- shell biomass kg/ha; TAG- Total aboveground kg/ha and WIOO- weight of lOO seeds. Data values are means of3 replicates. I • 62 Pod numbers ranged from 8 to 46 during the short rains but dropped to a range of 4 to 30 during long rains. The ranges were higher than those obtained by Muli and Saha (2000) of 8 to 18 at Mtwapa and 7 to 12 at Msabaha, at the Kenyan Coast. However, pod numbers obtained were lower than those obtained by Owolade et al., (2006), at Ibadan, Nigeria, that ranged from 50 to 75 pods per plant. Of the varieties tested, varieties that had above average pod numbers were El, E2, E3, E4, E6, E7, E8, E9, EIO, Ell, E13, E14, E15, MI, M3, M6, M8, MIO, M13, M14, MI5, Kang'au and CP21. Of these varieties, 57% were early maturing while 35% were medium maturing varieties. This observation probably indicated that early maturing varieties produce high number of pods compared to medium maturing ones. In addition, highest pod numbers of 46 pods per plant during the short rains was recorded in the local variety, Kang'au while during the long rains El had the highest pods per plant (of 30). This observation probably meant that the local variety was superior to the tested improved varieties in pod production production. during the short rains and had high potential Branch and pod numbers were significantly (p<0.05) correlated for pod (55%) during the short rains but the correlation dropped to 46% during the short rains. A positive correlation between branch and pod numbers of 46% was also documented by Nakawuka and Adipala (1999) in Uganda. Plant biomass of the tested varieties ranged from 436 to 3933 kg/ha during the short rains and from 98 to 1762 kg/ha during the long rains. This biomass was comparable to the biomass production of between 960 to 3590 kg/ha obtained by Olufajo and Singh (2000) in Nigeria. Average biomass of the 34 varieties was 1372 kg/ha during the short rains and 678 kg/ha during the long rains. Varieties that had above average biomass in both seasons were El, E2, E7, E9, EIO, E14, M6, M7, M8, M9, MIO, 63 M11, M13, contributed produced Kang'au M14, MI5, Kathoka, Kang'au 33% while medium-maturing and CP21. varieties Early maturing varieties had 50% of the varieties that biomass above average. Highest biomass of 3933 kg/ha was recorded in during the short rains but E I produced highest biomass (1762 kg/ha) during the long rains. This observation implied that Kang'au biomass to the tested production significant compared (p<0.05) correlation had a very high potential for improved varieties. between biomass production There was a and branch numbers of 62% during the sh0I1 rains that dropped to 45% during the short rains. However, the correlations between biomass production and pod numbers were significant 71 % (p<O.Ol) and 66 % (p<0.05), respectively, during the short and the long rains. Grain yield ranged from 1079 to 3435 kg/ha during the short rains and from 304 to 2686 kg/ha during the long rains. These ranges were close to the seed yields obtained by Muli at Saha (2000) of 1600 to 2800 kg/ha at Mtwapa and 1300 to 1980 kg/ha at Msabaha along the Kenyan Coast. The yields were also close to grain yields of 1499 to 2739 kg/ha obtained by Olufajo and Singh (2000) at Nigeria. The grain yields were also close to the yields obtained by Owolade et al., (2006) of between 983 to 1498 kg/ha. Average grain yield for the 34 varieties was 1746 kg/ha during the short rains but dropped to 1061 kg/ha during the long rains. Varieties that had above average grain yields in both rain seasons were El, E2, E3, E7, E8, E9, EIO, E14, MI, M3, M7, M8, MIO, M11, M13, M14, Kathoka, Kang'au, CP14 and CP21. Of these varieties early maturing varieties contributed 40% compared to 40% contribution from the medium yielding varieties. During the short rains, highest grain yield (3435 kg/ha) was recorded in the local variety, Kang'au while in the long rains highest grain yield (2686 kg/ha) was recorded in variety El. This observation indicated a very high grain 64 yield potential of the local variety, Kang'au, varieties. observation The significantly during the long compared rains diversification correlation during Kang'au improved grain yields of the local varieties to of local varieties accompanied of the locally grown varieties. Grain yield had a significant by (p<0.05) (59%) with branch numbers during the short rains that dropped to 32% the long rains. Pod numbers correlations where dropped could have indicated the susceptibility pests and therefore need for efficient management to the tested and grain yields of 89% and 73%, respectively, had significant (p<O.Ol) during the short and the long rains. In addition, grain yield had significant (p<O.O1) correlations with biomass production of 80% and correlations 73%, respectively, between during and the grain yield and branch numbers, yields of 86% and 89% respectively (1999) in Uganda. the short were documented long rains. Significant and pod numbers and grain by Nakawuka and Adipala From the data obtained in the study, it appeared that cat worm attack not only affected yield parameters but also their correlation coefficients. Shells (pod shells) remain after seeds are removed from pods. Unlike leaves, most of which disintegrate during threshing, shells make a good proportion of legume biomass that is fed to livestock after grain is removed. Shells are therefore considering legume remains after grains are removed. important when In this study, shell biomass (biomass) ranged from 262 to 1110 kg/ha during the short rains and from 106 to 1120 kg/ha during the long rains. However, average shell biomass for the 34 varieties was 489 kg/ha during the short rains and 398 kg/ha during the long rains. Varieties that had shell biomass above average in both seasons were El, E2, E3, E6, E7, E8, E9, EIO, E14, M1, M3, M6, M7, M8, M10, Mll, M13, M14, MI5, Kathoka, Kang'au, CP14 and CP21. Of these varieties 39% were early maturing varieties and 43% 65 medium yielding varieties. Highest shell biornass (1110 kg/ha) was recorded Kang'au during the short rains while El had highest shell biomass in (1120 kg/ha) during the long rains. These are the same varieties that had highest grain yield during the short and the long rains, respectively. potential for shell biornass production. Again the local variety showed high Shell weight had positive correlations with branch num bers of 57% (p<0.05) and 39%, respectively, during the long and the short rains. However, significant respectively. the relationship between shell weight and biornass produced was (p<O.OI) being 74% and 80% during A higher and significant the short and the long rains, (p<O.OI) relationship shell biornass and grain yield of 89% and 85%, respectively, long rains. Further, there was a significant (p<0.05) biomass and pod numbers of 72 and 66%, respectively was obtained between during the short and the relationship between shell during the short and the long rams. Total aboveground was the sum of plant biornass, seed yield and shell biornass. It ranged from 2018 to 8479 kg/ha during the short rains and from 508 to 5569 kg/ha. Mean of the 34 varieties was 3572 kg/ha during the short rains and 2140 kg/ha during the long rains. Varieties with above average total aboveground (TAG) were El, E2, E7, E9, EI0, E14, MI, M7, M8, MIO, MII, M13, M14, MI5, Kathoka, Kang'au and CP21. Medium maturing varieties made up 47% while early yielding varieties made 35% of the varieties with TAG in both seasons. Highest TAG (8479 kg/ha) during the short rains was recorded in Kang'au while highest TAG (5561 kg/ha) during the long rains was obtained in El. Kang'au again showed superiority in TAG. over improved varieties 66 Weight of 100 seeds (WIOO) ranged from 14 to 22 g during the short rains and 11 to 21 g during the long rains. This seed weight range was close to the seed range of I I to 20 g documented by Ogbonnaya et al., (2003) in Nigeria. Mean W I00 of the 34 varieties was 17 in both seasons. Varieties that had above average W 100 were E4, E5, E9, EIO, Ell, Kang'au EI3, M2, M3, M4, M5, M6, M7, M8, M9, M 10, Ml2 M14, MI5, and CP2I. Of these varieties 30% were early maturing and 60% medium maturing. During the short rains highest (22 g) W I00 was recorded in E4 and in the long rains M4 and M13 had highest (21 g) WIOO. In addition, local variety, Kang'au was among the above average varieties, again placing the local variety among the superior cowpea varieties. To determine best performing best cowpea varieties, all characters varieties that were assessed that were identified in order were summed up. Nine included Kang'au, M13, E7, MIO, El, M8, MI4, Kathoka and CP21 (Table 6.5). However, M13 was found to be very susceptible to drought and insect attack and was therefore not included in the list of the nine cowpea varieties selected for on-farm trials. Kathoka was also omitted to give way to CP2] that was obtained from Maseno and closely resembled Kathoka, showed resistance to drought in a rain-fed school demonstration recommended and had close character totals. Varieties M7 and E6 trial and farmers the two cowpea varieties for further testing. Therefore, nine cow pea varieties that included Kang'au, lT97K-568-18 (E 1), IT97K-570-18 (E6), IT97K-499- 38 (E7), IT97K-499-35 (M8), IT97K-556-4 (MlO), IT95K-52- (M7), IT95K-I073-57 34 (MI4) and IT97K-1068-7 (CP21), were selected for on-farm trials. 67 Table 6.5 Character ranking of varieties selected for on-farm trials Varieties selected Kang'au E7 MIO El M8 MI4 CP21 M7 E6 Variety Kang'au M13 E7 MIO El M8 M14 Kathoka CP21 EI0 MI5 M7 E2 E14 Ml1 E9 E8 E3 M3 CP14 Ml M6 Ell E15 E6 E13 E4 M9 M2 M4 M5 E12 M12 E5 Position Character total I 2 19785 19220.9 17807.5 17679.3 17251.5 3 4 5 6 7 8 15659 15587.4 14998 14026.9 13694.1 13386.4 13188.2 9 10 I1 12 13 14 15 13037 12374.8 12075.2 11524.9 10864.5 10675 10289.2 9819.4 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 9140.8 9033.7 8588.1 8147.9 8032.4 7978.8 7916.6 7860.2 7834 7587.9 7567.3 7317 .6972.2 33 34 5348.8 IlJ:NVATTll IIMI\/::P.~lTV I 1r:lIlADVJ ! 68 CHAPTER SEVEN EFFECTS OF ISFM ON NODULATION, GROWTH AND GRAIN YIELD OF SELECTED COWPEA VARIETIES 7.1 Cowpea (Vigna unguiculata Introduction (L) Walp.) is one of the earliest plants cultivated by man and the third most important grain legume in Kenya after common beans and pigeon pea. It is grown in the semi-arid areas, with most of it in eastern province. Cowpea is used for fodder, food and for soil fertility improvement (Muthamia and Kanampiu, 1996; Tarawali et aI., 1997). Cowpea growth is retarded by poor soil fertility especially micronutrients (Bationo et al., 199 I). However, low phosphorus its growth (P) and can be improved addition of inorganic P and nitrogen (N) (Audi et al., 1996). Phosphorus by can be added at a rate of 13-25 Kg/ha as diammonium phosphate or as triple superphosphate and N at 10-15 kg/ha (Audi et al., 1996; Shetty et al., 1995; Subarao et al., 1999). Phosphorus is an essential component of all living systems and in higher plants it functions as a constituent of nucleic acids and proteins important in cell division and induces root growth, promotes increases disease resistance Nelson, 1975). Phosphorus anthocyanin (Marschner, 1996; Owolade seed formation and et al., 2006; Tisdale and of purple pigment and stagnant growth. The condition is more pronounced during seed Deficiency 1996). It is deficiency in young plants leads to accumulation the periods of adverse weather when growth processes during (Marschner, formation, the leaves turn yellow are slow. If P is deficient starting with the older leaves. of P is also shown by formation of small, dark green leaves that stand 69 more upright than do the normal leaves (Marschner, 1996). Crops growing In P deficient soils also exhibit delayed crop maturity, and reduced quality and quantity of crop yields (Fairhurst et al., 1999). The primary source of P is mineral apatite found in primary rocks and apatite is the most commonly occurring phosphate mineral in rocks. Other sources of P are organic matter and secondary and complex compounds in the soil. The amount of organic P varies greatly among soils and within soil profile and processes that lead to increased soil organic matter also increase P content of the soil (Wild, 1988). Phosphorus absorbed into plants in ionic form, mainly as H2P04- availability phosphates is greatest precipitate between (Marschner, 1986) and its pH 6.0 to 7.2. Iron (Fe) and aluminium at low pH «5.5) and calcium (Ca) phosphates is (AI) at high pH (> 8). However, the available phosphorus supply in the soils depends on the amounts and forms ofP present in the soil (Kamprath, 1991). Phosphorus deficiency is widespread covering an area estimated at over 2 billion ha (lama et al., 1997; Smaling, 1993) and it may result from 10w-P status of the parent material, weathering, between nutrient (Fairhurst long-term anthropogenic mismanagement inputs and exports, and P loss by erosion et al., 1999). However, soil P deficiency primarily through imbalance and surface run-off resuits from either inherent low P levels or depletion of soils (Buresh et al., 1997) but soil P can be replenished by addition of inorganic fertilizers, organic matter in form of plant and animal residues or phosphate (Chien and Menon, 1995). rocks such as Busumbu and Mijingu phosphate rocks 70 Continuous cropping, removal of crop residues to feed animals and overgrazmg between cropping seasons with little or no external inputs have reduced the productive capacity of arable lands and thus threatened systems in sub-Saharan Kenya, decline the sustainability of food production Africa (Sanchez et al., 1997; Stoorvogel in crop yields is a major problem facing et al., 1993). In smallholder farmers (Mathuva et al., 1996). This is attributed to the high costs of inputs that make the use of inorganic fertilizers on staple food crops uneconomical for most smallholder farmers (Jama et al., 1997). Use of organic alternative inputs as an external to expensive source of soil nutrients fertilizers to smallholder are low in nutrient concentration is a logical cheap farmers. However, organic inputs compared with inorganic fertilizers (Sanchez et al., 1997) although cattle manure has been one of the most commonly used ways of soil fertility improvement for crop production 1997). For example, in the semi-arid phosphorus 1961; Giller et al., areas of eastern Kenya where nitrogen and limit crop production, farmyard manure is commonly used to enhance soil fertility and crop production (Gibberd, Recent in Africa (Dennison, research 1995; Ikombo, has shown that a combination 1984; Kihanda et al., 2004). of organic and inorganic inputs enhances crop production and reduces cost of inorganic fertilizers (Ojiem et al., 2004; Okalebo et al., 2004). The combination of organic and inorganic inputs is termed integrated soil fertility management (ISFM). Therefore the objective of this study was to determine the effect of integrated soil fertility management and grain yield of selected amounts in Makueni District. improved cowpea varieties on nodulation, growth, under contrasting rainfall 71 7.2 Materials and Methods Two farms were selected in two sites with contrasting were located at Kavuthu Sub-location 7"S, 037°25'23.6"E, in Mbitini division, and at Ndunguni Sub-location 02°04'44. 2"S, 037°34'44.3"E. rainfall amounts. The farms 1223 m.a.s.l, 0 I°59' 52. in Nguu division, 1082 m.a.s.l, The farms were selected based on the willingness of the farmers to relinquish their farms for trials, farm uniformity and availability of the required farm sizes. First planting was done in the long rains of 2006 and a second one in the short rains in the same year. Ox-ploughs were used to prepared land for planting during the long rains but hand hoes were used to prepare land during the short rains to avoid mixing up of the treatments applied during the long rains. Before planting, soils were collected in the upper 20 cm at 10 locations in each plot, mixed and sub-samples taken for analyses (Mathuva et al, 1996). The soils along with animal manure used in on-farm trials were analysed as described by Okalebo et al., (2002) (Table 7.1). Each trial site had 3 blocks and 36 sub-plots in each block. Each sub-plot measured 3 m2 (100 cm x 300 cm). A distance of 60 cm and 80 cm separated the sub-plots and blocks, respectively. treatments within the blocks GenStat programme and varieties within was used to randomize the treatments (Appendix 3). Treatments applied consisted of; I) A control, no inputs were applied (T!), 2) Animal manure at 2.5 t/ha (T2), 3) P as TSP (P20S, 0:46:0) at 15 kg/ha (T3) and 4) Animal manure + TSP (T4) at the above rates. Treatment application application was done using spot method to ensure that the crop was in contact with the nutrients. Nine cowpea varieties, eight of which were improved varieties (Chapter 6); IT97K-568- 18, IT97K-570-18, IT95K-52-34 IT97K-499-38, Kang'au IT97K-499-35, and IT97K-I068-7 Kang'au and CP21, respectively, IT97K-I073-57, IT95K-556-4, coded El, E6, E7, M7, M8, MIO, M14, were planted in each trial site. The experiment was 72 set up as a split plot design with nine varieties and four treatments that were replicated in three blocks. Cowpea was planted at a spacing of 20 cm x 60 cm within sub-plots. During planting, 3-4 seeds were planted per hole and thinned to one plant per hole after 2 weeks. Weeding was done as required to ensure the plots remained clear of weeds over the cropping seasons. In both seasons, spraying was done at 50% crop flowering to control insect pests using Bestox® 100EC at a rate of 10 ml per 20 litre water/ha. Plant samples were taken at 50% flowering and at crop maturity. At 50% flowering, plant samples were carefully dug out, cut at the root color, the root system wrapped in a paper, stored in a cooling box and brought to the laboratory. laboratory, nodules were dried and nodule biomass was recorded. weight determination, Connection the samples In the For nodule dry were dried at 60°C for 72 hours in a Sanyo oven (Model MOV-212 F, Sanyo Electric Company, Japan). At crop maturity, mature pods were separated from the plants and the plants were cut at the at root color. In the laboratory, pods were shelled and the seeds and shoot system were dried and weighed as above. Data obtained was analysed using GenStat Discovery Edition 1, GenStat Procedure Library Release PL12.2, and treatment means were separated using standard errors of differences of means (SED). Table 7.1 Chemical characteristics of soils collected at on-farm trial sites and animal manure applied to the on-farm trials Source Manure On-farm Ndunguni On-farm Kavuthu PHin water 8.8 Conductivity Nitrogen (mmhos ern") (%) 1.4 4.2 I Carbon (%) 18 Organic matter (%) 30.9 Phosphorus (ppm) (Olsen P) 5 7.6 6.0 0.09 12 19.9 4 7.7 2.9 0.05 6.5 11.1 3 73 7.3 7.3.1 Results Nodule biomass at 50% flowering During the long rains, nodule biomass ranged from 3 to 334 mg plant" at Kavuthu while highest nodule biomass of 28 mg plant" was recorded at Ndunguni (Table 7.2). Generally treatment significant (p<0.05) addition enhanced nodule formation increases especially at Kavuthu. in most varieties However, with the highest nodule biomass of 334 mg plant" at Kavuthu and 28 mg plant" at Ndunguni were recorded in M 14 in a manure +TSP treatment. During the short rains, highest nodule biomass of 300 rng plant" was recorded at Kavuthu, whi le the highest nodule biomass at Ndunguni was 833 mg plant" (Table 7.3). The highest nodule biomass of 300 mg plant" recorded at Kavuthu was recorded in Kang'au in a manure+TSP treatment, while at Ndunguni highest biomass of 833 mg plant" was again recorded in Kang'au but in a P (TSP) treatment. In both sites, treatment addition enhanced most varieties. nodule biomass with significant (p<0.05) increases in 74 Table 7.2 Nodule Site Variety/Treatment El E6 biomass Kavuthu Control 23a 100a c (mg/plant) at 50% flowering Manure 14a 49b SC E7 3 SSb 22C M7 c SOb 12 M8 98c 74C MIO SOd 182b M14 Kang'au 97a IS0a 34c CP21 74 47 Treatment means 0.069 P-Value Variety P-value Treatment 0.377 P-Value treatment*Variety O.SIS 44 SED Variety 27 SED Treatment SED variety*treatment 82 Same letter on the same row by site indicates replicates. 8r TSP 39a 55b 93a 67b 109a 212a 113c 104a 67b during the long rains Manure+TSP 13a 32b 9S Ndunguni Control 3a Manure Oa a SIb 3 87a l2c 14Sb 334a 24b 2Sc 80 Ib Ib 7a lC TSP Manure+TSP 2a 2a Oa Oa 4a la la Oa 2a 1b 6a Sa 1b Sa 3b 2b c 28a 16 a 7b 3b S 63 Ob Ob b 20 8b 1be 4b Oc 3 3 9 6 0.034 0.388 0.318 4 2 7 that treatments are not significantly different at (p<O.OS). Data values are means of 3 75 Table 7.3 Nodule biomass (mg/plant) at 50% flowering during the short rains Site Kavuthu Ndunguni Variety/Treatment Manure TSP Manure Control TSP Manure+TSP Control Manure+TSP 33c 133a 6r5 675 El 100a 67a6 335 33c a a a a b a Ob OC 133 100 E6 133 100 33 100 a b a b Ob Ob Ob Ob 133 67a 67 E7 33 a a Ob Oa Oa Ob Ob M7 67 33 a a b a b Oc OC Ob M8 100 133 33a 100 33 iooM10 200a 100b 200a 267a 100b 67d 133b 100b roe= M14 133ab 33c 133a 33c 167a 33c Kang'au 133c 233b 833a 133c 233b 300a 267b 33d c b c b c bc 100a 3673 33 67 CP21 167 67 33 100 Treatment means 41 78 178 105 118 104 70 85 P-Value Variety 0.021 0.054 P-value Treatment 0.185 0.620 P-Value 0.416 treatment*Variety 0.894 SED Variety 42 92 SED Treatment 28 73 SED variety*treatment 85 212 Same letter on the same row by site indicates that treatments are not significantly different at (p<0.05). Data values are means of 3 replicates. 6r 76 7.3.2 Shoot biomass production at crop maturity At Kavuthu biomass production ranged from 1219 to 3301 kg/ha compared to 131 to 1608 kg/ha at Ndunguni (Table 7.4). In addition, the highest shoot biomass of 3301 kg/ha at Kavuthu was recorded in Kang'au in a manure+TSP treatment, while at Ndunguni highest biomass of 1608 kg/ha was recorded for M7 in a TSP treatment. Generally, addition of treatments enhanced biomass accumulation the long rains and treatment addition resulted in both sites during to significant (p<0.05) biomass increases in most varieties During the short rams, a generally compared to that recorded higher biomass was recorded during the long rains. At Kavuthu in the trial sites biomass produced ranged from 3090 to 7283 kg/ha compared to a range of 4834 to 12725 kg/ha at Ndunguni (Table 7.5). However, recorded in Kang'au kg/ha at Ndunguni highest biomass of 7283 kg/ha at Kavuthu in a manure+TSP treatment while highest biomass of 12725 was recorded in M8 in a control treatment. short rains, treatment application was enhanced biomass production Further, during the in both Kavuthu and Ndunguni with significant (p<0.05) increases in most varieties. 7.3.3 Grain yield During the long rains, grain was only harvested at Kavuthu (Table 7.6) and ranged between 586 to 2235 kg/ha and highest yield of 2235 kg/ha was recorded in variety El in a manure treatment. As in biomass production, grain yield was generally enhanced by treatment application with significant (p<0.05) increases in most cow pea varieties. 77 Table 7.4 Shoot dry weight (kg/ha) at crop maturity during the long rains Site Variety/ Treatment El E6 E7 M7 M8 MIO M14 Kang'au CP2l Treatment means P-Value Variety Kavuthu P-Value Treatment P-Value Variety* Treatment 0.291 0.987 0.677 0.45 Control 1663cb 2294b 1569b 1424bc 1794ab 1630c 1369c 2036c 1846a 1736 0.386 Ndunguni Manure 2630a 1835c 1863a 1513b 1821 ab 2141 b 1944b 1341 d 2505a 1954 TSP 1869b 2766a 1933a 1219c 1869a 1760c 2435a 2485b 1249c 1953 M+TSP 1444c 2530ab 1388b 2396a 1566b 2977a 2024b 3301 a 2521a 2238 Control 722a 325c 344c 367c 1544a 711c 472c 1147a 614b 694 0.358 Manure 4835 800b 269c 702b 525b 827b 1211 a c 555 925a 700 TSP 180c 772b 689b 1608a 131 c 1130a 455c a 1180 c 197 705 Manure + TSP 392b 1127a 986a 664b 161 c 811 b 752b b 962 950a 756 SED variety 209 336 SED treatment 191 257.6 SED variety * treatment 749 538.4 Same letter on the same row by site indicates that treatments are not significantly different at (p<0.05). Data values are means of 3 replicates. 78 Table 7.5 Site Shoot dry weight (Kglha) at crop maturity during the short rains Kavuthu Ndunguni Variety/Treatment Manure Control TSP M+TSP C c a 3671 El 3829 5989 55596 E6 4615c 4373c 5042b 5542a c a a 3224 E7 5478 5148 4215b b a a M7 3468 4854 4737 4576a c a 4851 b 5426a M8 4296 5817 a MI0 4332b 4951 4743ab 5026a ab c bc M14 5253 4762 5115 6150a c 5501 b Kang'au 5528b 3090 7283a CP21 4296b 4454b 3290c 5642a Treatment means 4891 4398 4667 5491 P-value variety 0.375 P-value treatment 0.042 P-value variety*Treatment 0.582 SED variety 479 SED Treatment 414 SED Variety*Treatment 1176 Same letter on the same row by site indicates that treatments are not means of 3 replicates. 11- Control Manure TSP 5764b 7719a 61756 b b 5559 6064 7447a a b 6650 7139a 5723 ab c 7914 6370 7336b a b 12725 8685 8530b 6695c 1231 7072c a a 8677 8985 5323b a b 10748 6253 5017c 7119b 10835a 7108b 7880 8208 6794 0.343 0.438 r Manure TSP 5631b 7369a 7455a 8533a 4834c 8144b 9088a 11048a 6242b 7593 + 0.297 1116 893 2574 significantly different at (p<0.05). Data values are 79 During the short rains, grain yield ranged from 586 to 2524 kg/ha at Kavuthu and from 344 to 2368 kg/ha at Ndunguni (Table 7.7). In addition, at Kavuthu, highest grain yield of 2524 kg/ha was recorded in E I in manure+ TSP treatment whi le at Ndunguni same variety had highest grain yield of 2368 kg/ha in a control treatment. In addition, treatment addition generally enhanced grain yield in most varieties with significant (p<0.05) increases in both study sites. Table 7.6 Grain yield at Kavuthu during the long rains Manure Variety Control Manure TSP TSP c a b c El 716 2235 1352 586 a c b I747b E6 2060 1013 1674 b a 841 c E7 1075b 1094 1552 a a 891 a M7 844 702a 750 c b a M8 855 1I02 1397 747c 1427a 1272ab MIO 1322a 1105b c a 741 c M14 594 1910 1416b Kang'au 1344b 2127a 922c 1280b CP21 1408b 164r 1455ab 1183c Treatment 1213 means 1152 1186 1380 P-value variety 0.545 P-value treatment 0.671 P-value Variety*treatme nt 0.392 SED variety 363.7 SED treatment 198.5 SED variety*treatme nt 631 Same letter on the same row indicates that treatments are not significantly different at (p<0.05). Data values are means of 3 replicates. + 80 Table 7.7 Grain yield (Kg/ha) during the short rains Site VarietylTreatment El E6 E7 M7 M8 MI0 M14 Kang'au CP21 Kavuthu Control 914d 1100d 586c 1258d 1069c 1433a 1491 c 1558c 1619b Ndunguni Manure 2169b 1202c 1438b 1760c 1508b 1041 b 822d 1813b 1358c TSP 1685c 2010a 1483b 1958b 2163a 1055b 1758b 1816b 1122d M+TSP 2524a 1735b 1724a 2380a 1477b 1150b 2235a 2452a 1849a Control 2368a 1413b 922c 1266b 1083c 1113b 127r 344b 1452a Manure 969d 783c 1488b 1572a 1599a 1647a 866b 536b 1508a TSP 1827b 1261b 1763a 1449a 1369b 1472a 991 b 397b 1597a Manure+ TSP 1619c 1624a 977c 1444ab 722d 1166a 964b 875a 1438a 1249 1497 Treatment means 1203 1225 1457 1672 1347 1219 0.114 P-value variety 0.119 P-value treatment 0.852 0.006 P-value 0.73 variety*Treatment 0.767 329.3 SED variety 249.8 178.4 SED Treatment 202 SED Variety*Treatment 581.2 568.6 Same letter on the same row per site indicates that treatments are not significantly different at (p<0.05). Data values are means of 3 repl icates. - - ---- --- - .. -- •...... ..... .-.-,-. --~ 81 7.4 7.4.1 Discussion Nodule biomass Lower nodule biomass recorded at Ndunguni compared to that recorded at Kavuthu was most likely due to low total rainfall received at Ndunguni (194 mm) compared to a higher rainfall (233 mm) received at Kavuthu during the long rains. In addition, it was noted that most of the rainfall recorded at Ndunguni fell during the last two weeks of April, which were also the first two weeks after cowpea was planted and about three weeks before cowpea reached 50% flowering. Therefore samples were being taken for nodule assessment when cowpea at 50% flowering, the soil was already dry (had limited moisture) and therefore very few nodules were recovered and in some plants no nodules were recovered from cowpea plants at Ndunguni. The few nodules recovered at Ndunguni under limited soil moisture could have been caused , • by water stress on the cowpea plants that most probably reduced nodule formation. Sprent (1971) suggested that water stress affect formation and longevity of leguminous root nodules. In addition to rainfall differences, treatment differences in nodule biomass were noted in both study sites although most treatment differences were noted at Kavuthu than at Ndunguni treatment were probably indicating effectiveness. recorded that high rainfall At Kavuthu, in TSP manure+ TSP (80 mg/plant) treatments probably amounts at Kavuthu enhanced highest mean nodule biomass (95 mg/plant) compared to manure (74 mg/plant) and indicating that the cowpea plants used TSP more efficiently for nodule formation than they did use other nutrient forms. Miao et al., (2007) studying in the effect of P supply on nodule formation in soybean I 82 observed that nodule mass was increased by P supply through increased nodule development and functioning. Results of nodule biomass application did not always during the long rams further indicated that treatment increase nodule biomass and that cow pea response to treatment application varied with site and variety. For example, during the long rains, nodule biomass of M14 at Kavuthu was increased by all treatments while at Ndunguni nodule biomass of the same variety was increased by only two treatments, manure and manure+TSP treatments. In the same season, nodule biomass of Kang'au was not affected by treatment application at Kavuthu but was increased by addition of TSP at Ndunguni. This observation agreed with Audi et al., (1996) who suggested that cowpea does not always respond to nutrient inputs. During the short rains, a relatively higher nodule biomass was recorded at Ndunguni than at Kavuthu although more total seasonal Ndunguni rainfall was still recorded at Kavuthu (397 mm) compared to (386 mm). The differences observed in nodule biomass in the two sites during that season could have been caused by site differences that probably included soil chemical characteristics (Table 7.1). During the short rains, the highest mean nodule biomass (178 mg/plant) was recorded in TSP treatment at Ndunguni nutrient for nodule formation. further indicating that TSP was the most effective However, as opposed to the observations made on treatment effect on nodule biomass during the long rains, where treatment effect were common at Kavuthu, probably indicating nodule biomass in both sites indicated clear treatment the importance of rainfall in treatment effectiveness. effect Also as 83 observed in the long rains, not all treatment additions enhanced nodule biomass and that variety response to treatment addition varied with site. For example, during the short rains, there was no effect of treatment addition in nodule biomass of M 14 at Kavuthu while at Ndunguni nodule biomass of the same variety was increased by two treatments, manure and manure+ TSP, as in the first season. Besides site variation, treatment effect on nodule response to treatments. biornass also indicated Observations seasonal variation of variety made on nodule biomass could be compared to the studies done by Jemo et al., (2006), who, when studying the effect of P in low phosphorus soils of southern treatments, Cameroon, recorded highest nodule biomass in TSP and obtained nodule biomass that ranged from 234 to 676 mg/plant that was within the nodule biomass range obtained in this study. 7.4.2 Shoot biomass As observed significantly biomass However, with nodule biomass during the long rains, shoot biomass was higher at Kavuthu than at Ndunguni and as with the nodule biomass, low at Ndunguni was attributed to low total rainfall received in the site. as opposed to highest overall mean nodule biomass at Kavuthu that was observed in TSP treatment during the long rains, highest shoot biomass was recorded in manure+T P treatment in both sites. This observation was in agreement with Ojiem et al., (2004), who suggested that combining organic and inorganic P results in synergistic effects, particularly in drier moisture stressed growing seasons. In addition, treatment effect for same varieties on shoot growth did not always match treatment effect on nodule biomass indicating that treatments that enhanced nodule 84 biomass did not necessarily enhance shoot biomass and probably that response of the shoots to nutrients was different from that of nodules. As observed in nodule biomass, during the short rains, significantly higher shoot biomass was recorded at Ndunguni compared to Kavuthu and this observation attributed to site characteristics as opposed production to the observations Ndunguni other than rainfall differences in the sites. In addition, made during the long rains where shoot biomass in both sites was in overall enhanced by addition of manure+TSP, only at Kavuthu was where the treatment in overall enhanced it was shoot biomass while at highest shoot biomass in overall was recorded in manure treatment. The later observation could have been caused by mineralization (Campo et al., 1998; Cornejo et al., 1994) of the high carbon content in manure and soil (Table 7.1) under high rainfall at Ndunguni. These observations confirmed the need to combine organic and inorganic inputs during drier seasons in both sites and in all seasons at Kavuthu. It also indicated inorganic that during high rainfall seasons there might be no need to use inputs at Ndunguni observed that majority application to enhance cowpea biomass production. of the varieties at Ndunguni It was also did not respond to treatment probably indicating that there was a release of nutrients to the soil from carbon mineralization at the site. Loomis and Connor (1992) suggested that organic matter could increase soil microbial biomass that can lead to an increase nutrient availability to a growing crop. 85 7.4.3 Grain yield As observed in other parameters (nodule and shoot biomass) treatment application generally enhanced grain yield at Kavuthu during the long rains. However, in overall highest mean grain yield was recorded in the TSP treatment most efficient nutrient in grain filling. In addition, increase grain yield of about 50% of the varieties responded to treatment application. indicating that P was treatment indicating application did not that not all varieties Most grain yields recorded during the long rain season was within cowpea yield ranges of 1653 to 1827 kg/ha recorded by Kihanda et al., (2004) in the semi-arid eastern Kenya. As observed during the long rains, highest grain yield overall was recorded in TSP treatment in both sites probably indicating that P applied as TSP was the most efficient nutrient in grain filling in both seasons. This observation while inorganic inputs did not enhance biomass accumulation also indicated that at Ndunguni, inorganic inputs were needed for grain filling. Varieties responded differently to nutrient inputs depending on the site indicating treatments. Most grain yield values recorded during the short rains ranged 791 to by Jemo et al., (2006) in low phosphorus 1596 kg/ha obtained Cameroon. In addition, yields of improved 1996). the effect of site differences on response to soils of southern most cow pea varieties had grain yields within the potential cow pea varieties (1200 to 1800 kg/ha) in Kenya (Audi et al., 86 CHAPTER EIGHT 8 NITROGEN FIXATION, POPULATION AND DIVERSITY OF COWPEA RHIZOBIA UNDER ISFM 8.1 Introduction There is a trend towards soil fertility loss in many soils in the semi-arid zones of Africa and planting legumes that can be used for both soil fertility and food may arrest this trend. Grain legumes have been important crops and have been used for centuries as food (van Kessel and Hartley, 2000). They supply proteins for animals and mankind and cover about 11% of the world's arable lands (FAO, 1993). Grain legumes are also unique in that they acquire atmospheric symbioses with bacteria collectively (N) through called rhizobia in a process commonly termed biological nitrogen fixation (Stevenson and van Kessel, fixation (BNF) is an important aspect of sustainable food production crop productivity and long-term nitrogen 1997). Biological nitrogen and environmentally friendly (van Kessel and Hartley, 2000). However, grain legumes contribute N to the soil only when the total quantity of N fixed symbiotically is larger than the amount of N removed at harvest in seed or in crop residue (Evans et al., 1989; Giller et al., 1994; Stevenson and van Kessel, 1997). Several methods have been used to estimate BNF and the most commonly used methods are: (1) Nitrogen difference method also called N balance method. This method is based on difference in total nitrogen between a legume and a non-fixing reference plant. It is based on the assumption that both legume and the non-fixing reference plant must absorb the same amount of soi I 87 (2) 15N-isotope dilution either with enriched 15N-fertilizers (Fried et al., 1983) or through changes at the natural 15N abundance level (Shearer and Kohl, 1986). Both methods are used to estimate nitrogen fixed over a season. In both methods, it is assumed that the rates of mineral N uptake are same for both the N-fixing legume and the non-fixing reference plant. However, differences in seasonal N accumulation patterns of legume and reference crop under field conditions, the concurred decline in %15N of available soil N pool and differences in root distribution can lead to erroneous estimates of N fixation. In addition, use of different non-fixing reference plants can also lead to variable estimates of N fixation (Danso et al., 1993; Witty, 1983). (3) Xylem-solute method, where N is estimated by measuring changes in ureid content in the xylem sap (Herridge et al., 1990). This method is limited to ureid producing legumes such as cowpea (Vigna unguiculata L.) and can only be used to estimate N fixed over a short period of time. (4) Acetylene observation Reduction Assay (A RA) (Hardy et al., 1968) is based on the that N2-fixing enzyme, nitrogenase, catalyses the reduction of acetylene (C2H2) to ethylene (C4H4) ARA method (HC=CH + 2H+2e- ~ is that it only provides activities under prevailing H2C=CH2). The disadvantage an instantaneous measure of of nitrogenase assay conditions and lacks linearity in the rate of C2H2 reduction during the assay. Rhizobia by, definition, are bacteria that establish symbioses with legumes forming root or stem nodules on the host and fixing atmospheric nitrogen (N2) (Bala and Giller, 2001). Exploitation of the low-input tropical of BNF to improve productivity 88 cropping systems populations. demands in part characterization of indigenous rhizobial Minimum criteria for describing new rhizobial species include cultural, morphological, physiological al., 1991). Antibiotic symbiotic traits and phylogenetic characters (Graham et resistance (Mueller et al., 1988), colony morphology (Zhang et al., 1991), stress tolerance to external temperatures, salinity and acidity (Keyser at al., 1979) and nucleic acids (Johnson, 1984) have been used as important criteria to distinguish rhizobia. Cowpea rhizobia were first classified rhizobia nodulating promiscuous "Cowpea cross-inoculation in heterogeneous tropical and sub-tropical group of slow-growing legume species called group" (Allen and Alien, 1981) but were later transferred to the genus Bradyrhizobia that consisted of three species, B. japonicum (groups 1 and la), B. elkanii and B. liaoningese (Hollis et al., 1981; Jordan, 1982; Kuykendall and Saxena, 1991; Xu et al., 1995). However, Willems et al., (2001), using DNADNA hybridization Brydyrhizobium and consists 16S-23S rDNA of four highly IGS related sequence analysis genospecies showed (B. japonicum, that B. liaoningese), and at least three other genospecies one of which is B. elkanii). "Cowpea miscellany" well characterized promiscuous rhizobia indigenous to African soil environments (Mpepereki nodulating have not been et al., 1996). These bacteria are often described as a wide range of legumes but with poor effectiveness (Singleton et al., 1992). In addition, bacteria from cow pea nodules are said to contain both slow- and fast-growing rhizobia strains (Giller, 2001). 89 8.2 8.2.1 Estimation Materials and Methods of biological nitrogen fixation (BNF) Plant samples were taken from two on-farm trial sites, Kavuthu and Ndunguni in the semi-arid Makueni District where nine cowpea varieties had been planted in the long rains. Treatments in the trails included a control (no amendment used) (Tl), manure at 2.5 tlha (T2), 15 kg/ha P applied as TSP (P20S) (T3) and a combination manure+TSP of (T4) at the singly applied rates. Cowpea varieties tested included a very good nodulator in the trial sites (MI4), a local variety (Kang'au), a medium nodulator (E6) and a poor nodulator (M8) (full variety descriptions have been given in chapter 6). Plants used for BNF estimates were sampled at 50% flowering along with nine common non-N-fixing plants consisted reference plants, consisting of shrubs and herbs. The reference of Cordia sinensis, Trichodesma zeylanicum, Zea mays, Solanum incanum, Melia volkensii, Ipomea sp. (4 species). The amount of nitrogen fixed was assessed using the ISN natural abundance method. For ISN analysis, dry and ground plant samples were sent to the UC Davis Stable Isotopes Facility, California, USA, where they were analysed using a continuous flow Isotope Ratio Mass Spectrometer (IRMS). An average olsN value for the 9 non-fixing plants (5.63) was used as a reference value and %N fixed was calculated as follows: 100[(Non-fixing-Legume sample)/(Non-fixing-l00% fixation value)] (SyiJa et al., 2002) A olsN value of 0 was assumed for plants fixing 100% of their nitrogen. Data were analysed using two-way ANOV A. 90 8.2.2 Rhizobia enumeration Soils were collected and assessment of nodule and shoot biomass at the top 20 cm from Kavuthu on-farm trial during the long rains, at planting and during harvesting. In the trial, nine cow pea varieties had been planted but only two of the test varieties (M14 and Kangau) were used in this study because they were the best nodulators among the nine varieties used in the on-farm trials. During the planting time, soils were collected as described in chapter 7 but during harvesting time soi Is were collected by treatment and variety. Treatments used at planting were a control, animal manure at 2.5 t/ha, P as TSP at 15kg/ha and a combination of animal manure and TSP at the singly applied rates (Coded T I, T2, T3 and T4, respectively). The collected soil samples were stored at 4°C immediately after collection. Rhizobia enumeration was done using the Most Probable Number Method (MPN), as described M14 and Kang'au, by (Vincent, 1970). Clean seeds of the test cow pea varieties, were placed in a flask and shaken for 3 minutes in 95% ethanol. After ethanol was drained, the seeds were further shaken in 3% sodium hypochlorite (NaOCI) for 3 minutes and rinsed in 10 changes of sterile water. The seeds were then soaked for 2 hours in sterile warm water and then transferred to a sterile beaker and incubated at 28°C in an oven for 48 hours. Pregerminated transferred into sterilized Leonard jar assembly (CaCh2H20, KH2P04, ZnS04.7H20, CuS04.7H20, 1985). An uninoculated Iron-citrate, containing MgS04.7H20, CoS04.7H20, seedlings were aseptically N-free nutrient K2S04, MnS04.H20, NaMo02.2HO) (Somasegaran and a +N nitrogen (0.75g/L KN03) media H3B03, and Hoben, controls were included. After 4 days, the root systems of the plants were inoculated with I ml of a 1:9 (Soil: water) soil suspension diluted serially to give dilutions of s', y2, s', y4, y5 and y6. All jars were labeled to include variety, soil treatment, dilution, replication and date 91 of planting. To reduce evaporation, the top surfaces of the jars were covered with sterile gravel (ballast). The jars were placed in a green house whose floor and walls had been cleaned using tap water. The plants were destructively weeks. Data collected at harvesting harvested after 4 included number of positive jars (nodulated) in each dilution as well as nodule and shoot biomass of the harvested cow pea plants. Rhizobia estimates were done using MPNES Programme (Bennet et af., 1990). To determine shoot and nodule biomass cow pea plants were destructively harvested after 4 weeks and nodules carefully plucked from the root systems while the shoots were cut at the shoot base. Nodule and shoot biomass were determined as described in chapter 7. 8.2.3 Determination 8.2.3.1 Characterization of rhizobia diversity of rhizobia using cultures Fresh nodules obtained from the MPN experiment (above) were surface sterilized by immersion in 95% ethanol and finally in 3% sodium hypochlorite (NaOCI) for 3-4 minutes. The nodules were rinsed in six changes of sterile water and then each nodule crushed in a drop of sterile water. A suspension from each crushed nodule was streaked on Yeast extract mannitol agar (YMA), (lOg mannitol, 0.5 g K2HP04, MgS04, 0.1 g NaCI, CuS04.7H20, 0.5 g, and K2S04, MnS04.H20, CoS04.7H20, NaMo02.2H20 H3B03, and iron citrate) 0.2 g ZnS04.7H20, yeast extracts and topped to 1000 ml with distilled water, adjusted to pH 6.8 using IM NaOH and IM HCI, 15 g agar was added and the media was sterilized by autoclaving at 121°C and 103 kPa for (non-rhizobia contaminants 15 minutes) containing 25 mgl' Congo red dye absorb Congo red). The plates were incubated at 28°C until visible 92 bacterial growth could be seen on the streak line. A total of 376 rhizobia isolates were cultured. The isolates purification. on a fresh Congo However only 151 cultures were used to determine in the cultures production were then sub-cultured using colony shape and plasticity. characteristics such red media for strain types present as colony colours, mucous To determine growth rate, 70 cultures out of the purified 151 cultures were selected and streaked on YMA without Congo red. The rate emergence and growth of the rhizobia colonies to maximum line helped to characterize the isolates into fast or slow-growing size on the streak types. To determine whether isolates were neutral, acidic or alkaline, the pure isolates were streaked on YMA containing bromo-thymol blue (25 ppm), a pH indicator, and colour changes from green (neutral) to blue (alkaline) or yellow (acid) helped to categorize the 70 isolates (Appendix 4) into neutral, alkaline or acidic culture types. 8.2.3.2 i) Determination of rhizobia diversity by direct polymerase chain reaction (peR) amplification of 16s rRNA gene from rhizobia culture cells rRNA amplification Eighteen strains of rhizobia were analysed using PCR-Restriction polymorphism (PCR-RFLP) of 16S-23S rRNA intragenic spacer region (IGS). For RFLP analysis, primers used were, a forward primer Yl (5'-TGG AAC GCT GGC GGC-3') corresponding (5'-TAC CTT GTT ACG ACT TCA 1482-1507 in the Escherichia fragment length CTC AGA ACG to positions 20-43 and a reverse primer Y3 ccc CAGTC-3') corresponding to positions coli 16s rRNA gene sequence (Odee et al., 2002; Cruz et al., 2001). 25 1-11 of PCR reaction product contained 50 ng pure DNA from fully- KENYATTA UNiVERSITY LIBRARY 93 grown cultures, one dried bead (Ready-to-go PCR beads (GE Healthcare illustra TM), 50 mM KCI and 1.5 mM MgCI2), 1.5 III Y I, 1.5 III Y2 and 22 III PCR grade water. 10 III of the PCR reaction programmable product were used for amplification. Thermal a Controller (PTC-IOO™ MJ Research Inc., Watertown, MA) was used as described denaturation For amplification, here: Initial denaturation at 93 QC for 2 min; 35 cycles of (45 sat 93 QC) annealing (45 sat 62 QC), extension (2 minutes at 72 QC) and final extension at 72 QC for 5 minutes. Amplified DNA products were separated by horizontal gel electrophoresis of 2 f.ll aliquots of PCR product (Odee et al., 2002) in 1% agarose (Invitrogen ™ Life Technologies) (107.8 g tris-base, in TBE (Tris-borate EDTA) buffer 7.44 g EDTA and 55 g boric acid) for 2'/2 hours at 80 V and stained with (1 ug/ml) ethidium bromide for 30 minutes. Stained DNA profiles were photographed under UV with Kodak Molecular tOEastman Kodak Company, ii) Imaging Software Version 4.0, 1998-2005. rRNA restriction Aliquots of 6 III from the PCR reaction endonucleses product MspI in 0.3 III enzyme (in a restriction PCR reaction product, Serum Albumin were digested with restriction product that consisted of 6 III I III enzyme buffer, 0.3 III restriction enzyme, 0.1 III Bovine and 2.6 III water giving a total volume of 10 Ill) for 2 hours. The restricted rRNA was analysed by horizontal electrophoresis in 2.5% (w/v) agarose gel in TBE buffer, stained and photographed as described above. To determine the size of base pairs of the restricted rRNA, the pattern bands obtained after restriction were compared bromide. to the base pairs of the standard, 1.5% agarose stained with ethodium 94 8.3 8.3.1 Results Biological nitrogen fixation Large and significant differences (p<0.05) were found between amounts of nitrogen fixed at the two sites with plants at Kavuthu fixing 46-53% of their nitrogen while only one treatment (Kang'au treated with TSP, 815N=4.96) at site 2. The local variety (Kang'au) of its nitrogen (815N=2.28), (815N=I.58). nitrogen showed fixation of 12% N supplied with Triple super phosphate fixed 60% while Ml4 treated with manure+TSP However, differences between varieties fixed 72% of its and differences between treatments were not significant (Figures 8.1-8.3). 56 54 .-----------------------------~===, DTI DT2 ~ 52 '2f. '--' 50 "2x 48 t;:: 46 Z 44 bJT3 DT4 42 40 Treatment Figure 8.1 Amount of nitrogen fixed (%) at Kavuthu during the long rains Treatments (TI-T4); I) Control, 2) Manure at 2.5 tlha, 3) P at 15kglha and 4) manure+ P at the singly applied rates. 95 12 10 8 ~ 6 4 ,[====;------ - - DS1T1 DS1T2 GS1T3 DS1T4 DS2T1 DS2T2 DS2T3 DS2T4 ~'------' 2 Site and treatment Figure 8.2 N in sites 1 and 2 during the long rains J5 S1 and S2 represents site 1 (Kavuthu) and 2 (Ndunguni), respectively. Treatments, T1-T4 means; 1) Control, 2) Manure at 2.5 t/ha, 3) P at 15kg/ha and 4) manure + P at the singly applied rates. la 8 6 4 OLS1 OM14S1 OSlM8 OS2L J S2M14 DS2E6 .1 -, T ;.l: .. . T .L T 2 I T ..I.. .,... ~ I ,., -- ........... .. .~ zI,' o Variety by site Figure 8.3 15N values for site by variety during the long rains SI and S2 represents site 1 and 2, respectively, while L, M14 and E6 represents cowpea varieties Kang'au, M4 and E6, respectively. 96 8.3.2 Rhizobia populations in the soil, and nodule and shoot biomass Rhizobia counts ranged between 4.89x 102 and 2.0x 104 cells/gram of soil (Table 8.1). With an exception of rhizobia counts in manure treatment in M 14 and in manure +TSP treatment in local (L) variety, Kang'au, generally rhizobia counts were lower at planting than during harvesting. Further, the highest rhizobia count in variety M 14 was recorded in TSP and manure+TSP treatments, while in Kang'au highest rhizobia counts were recorded in manure and TSP treatments. Nodule biomass ranged from 0.9 to 3.4 mg with highest nodule biomass of 3.4 mg being recorded in a control treatment ofKang'au (Figure 8.4). Generally a low nodule biomass was recorded in cowpea plants grown in soils collected at the start of the season and treatment application overall a higher and significant did not improve nodule biomass. In addition, (p<0.05) nodule biomass was recorded in in Kang'au compared to M14. Shoot biomass ranged from 29 to 52 mg/plant, with highest biomass production being recorded in M 14 in a manure treatment (Figure 8.5). There was no difference in shoot biomass between cowpea plants grown in the soils collected at the start of the season and the control treatments. In addition, treatment application enhanced shoot growth in manure and manure+ TSP treatments. significantly (p<O.O1) 97 Table 8.1 Rhizobia estimates (rhizobia cells/gram of soil) in soils collected at the beginning and at the end of long rain season Variety Time of soil collection Treatment Cells/gram soil M14 Planting 0 1.074x 103 M14 Harvest 1 1.625xl04 MI4 Harvest 2 4.89x102 MI4 Harvest 3 2.0x104 M14 Harvest 4 2.0x104 Kang'au Planting 0 6.98x102 Kang' au Harvest 1 1.625x 104 Kang'au Harvest 2 2.0x104 Kang' au Harvest 3 2.0x 104 Kang'au Harvest 4 8.75x103 TO- Soils collected before ISFM experiment was started; Tl- Soils collected from the control treatment; T2- Soils collected from 2.5 t/ha manure amended soils; T3- Soils collected from 15 kg/ha P amended soils and T4- Soils collected from manure+P amendments. M14 and Kang'au are cowpea test varieties. 4 3.5 C '"P- 3 M S 2.5 VI VI '"E .~ oD ~ 2 I.5 =' -0 0 Z 0.5 0 TO TI T2 T3 T4 Treatment Figure 8.4 Nodule biomass of nodules recovered from MPN experiment TO- Soils collected before ISFM experiment control treatment; T2- Soils collected from collected from 15 kg/ha P amended soils amendments. MI4 and Kang'au are cowpea was started; Tl- Soils collected from the 2.5 t/ha manure amended soils; T3- Soils and T4- Soils collected from manure+P test varieties. 98 60 --~ El Kang'au -z: 50 c OMl4 C<:! 0.. 00 40 E '--' Vl ~ 30 E o ..D 20 o o ..c if) 10 TO T2 Tl T3 T4 Treatment Figure 8.5 Shoot biomass of cow pea plants harvested from MPN experiment TO- Soils collected before ISFM experiment was started; Tl- Soils collected from the control treatment; T2- Soils collected from 2.5 t/ha manure amended soils; T3- Soils collected from 15 kg/ha P amended soils and T4- Soils collected from manure+P amendments. M14 and Kang'au are cow pea test varieties. 8.3.3 Rhizobia characterization Using colony characteristics of rhizobia isolates, 9 groups of rhizobia were identified (Table 8.2). Most of the strains fell under strain group 3, while only two strains fell under group 9. In addition, assessment of 70 isolates for growth rate revealed that 97% (68) of the isolates were fast growing while only 3% isolates were slow growing. Slow growing isolates had dry colonies while most of the fast growing ones had wet colonies. With an exception of strain group 9, all other strain groups were fast growing. Further, analysis of 18 using PCR-RFLP of 16S rRNA intragenic spacer region revealed four rhizobia groups represented by four patterns (Figure 8.6). The patterns had base pairs (bp) that ranged from 271 to 827 (Table 8.3). Variety and treatment did not appear to affect restriction patterns (Table 8.4). 99 Table 8.2 Rhizobia isolate characterization Isolate Colony characteristic group using 376 cultures Round, small, flat, clear with brown centres Tiny, sticky and 2 mucoid Large, round, translucent, white, 3 gummy Tiny, dry, flat, sticky 4 brown 5 Large, pink, dome Small, round, dry, 6 flat, orange Large, brown, 7 mucoid, spreading Large, watery, spreading, clear with 8 suspensions Flat, orange, shiny, 9 dry using cultures Colour pH Number of cultures tested for growth rate and acid Growt characteristi h rate cs Yellow Acid Fast 6 33 Green Neutral Fast 11 11 Yellow Acid Fast 18 59 Yellow Yellow Acid Acid Fast Fast 8 8 8 21 Yellow Acid Fast 3 3 Yellow Acid Fast 9 9 Yellow Acid Fast 5 5 Blue Alkaline Slow 2 2 M Number of cultures selected for detai led studies bp 1,353 1,078 872 Figure 8.6 Four PCR-amplified 16S-23S rRNA IGS patterns obtained restriction of rhizobia strains with MpsI restriction endonuclease M is 1.5 agarose gel stained with ethidium bromide and bp is base pairs. after 100 Table 8.3 Fragment sizes in base pairs (bp) obtained after rhizobia restriction rRNA IGS patterns Fragment size (bp) of amplified rRNA IGS digested with MspI 271 383 603 684 II 271 684711 III 271 408657 IV Table 8.4 271 408603 827 Strain group, rRNA IGS pattern, cowpea varieties and treatments used in the PCR-RFLP analysis of 16S-23S rDNA IGS Pattern number 1 2 3 4 5 6 7 8 9 la 11 12 13 14 I5 16 17 18 Strain group 8 8 8 8 9 9 7 7 7 7 7 2 2 1 2 2 2 2 rDNA group I I II II III III III III III I I I II II I I IV IV TO- Soils collected before ISFM experiment control treatment; T2- Soils collected from collected from 15 kg/ha P amended soils amendments. M14 and Kang'au are cowpea IGS Variety Treatment M14 Kang'au Kang'a Kang'au M14 Kang'au Kang'au Kang'au MI4 Kang'au Kang'au M14 Kang'au Kang'au Kang'au MI4 MI4 MI4 TO T3 T3 Tl T4 T2 Tl T2 TO T2 T2 T4 T2 T3 T3 TI T4 T4 was started; Tl- Soils collected from the 2.5 t/ha manure amended soils; T3- Soils and T4- Soils collected from manure+P test varieties. 101 8.4 8.4.1 Discussion Biological nitrogen fixation Biological nitrogen fixation was recorded in the wetter site (Kavuthu), where it varied from 46 to 53%. Nitrogen fixation at the drier site (Ndunguni) was most probably limited by low and unevenly distributed rainfall (chapter 7) that could have caused water stress to the cowpea plants to the extent that nitrogen fixation was inhibited. Sprent (1971) suggested leguminous root that water stress reduces the formation nodules thereby reducing nitrogen fixation and longevity of in nodules. This suggestion agrees with the observation that very low nodule biomass was recorded at the drier site (Ndunguni) during the long rains (Chapter 7) compared to the wetter site (Kavuthu) and hence low or no nitrogen fixation was recorded in the cowpea plants recovered from Ndunguni. At Kavuthu, nitrogen fixation was higher in manure and TSP than in manure+ TSP treatment. manure and TSP in the manure+TSP This was probably because presence of both treatment could have resulted to high levels ofP in the soil that could have decreased N fixation. Miao et al., (2007) reported a decrease in N fixation with increasing P supply. Nitrogen fixation values obtained in this study were comparable to ranges of 46-59%, 36-54%, 35-50% and 36-51 % obtained by Gathumbi et al., (2002) in western Kenya in Sesbania sesban, Calliandra callothyrsus, Macroptilium (groundnut), respectively, shrubs to provide atropurpureta hypogaea (Siratro) and Arachis hypogaea indicating that cowpea can be used to replace leguminous fixed N for crops. Cowpea will have added advantage over leguminous shrubs because of its multiple uses. In addition to edible grain production, and its leaves that can be used locally as a green vegetable, generate income for households. it can also be sold to 102 8.4.2 Rhizobia Rhizobia counts harvesting time. February/March populations in the soil were generally This lower at the beginning observation was most probably of the season than at due to a dry spell of that could have lowered rhizobia populations to the levels observed at the beginning of the long rains, in the soils collected at cowpea planting time. This observation was in agreement with the observation made by Mapfumo et al., (2000), where low rhizobia counts were recorded in soil samples collected following a dry spell. Seasonal documented increase by Mulongoy in rhizobia populations and Ayanaba in the cowpea (1985) while studying trials was also seasonal rhizobia populations in three locations in West Africa. In the study, they noted that rhizobia populations were higher in wetter than in the drier sites. Rhizobia populations observed in this study were generally lower than the rhizobia populations observed by Maingi et al., (2006) at Kiboko research station about 40 km east of Kavuthu trial using cowpea, cultivar Ken Kunde 1, which ranged from 2.59x104 to 1.89xl05. differences in rhizobia counts could have been due to moisture differences The between the two sites given that the Kavuthu trial was rain fed while all crops at Kiboko Research Station are usually cultivated using irrigation. In addition, the differences could have also been due to the differences in cowpea cultivars used in both studies. High rhizobia populations growth hormones multiplication are important in cultivated soils because rhizobia produce which can stimulate plant growth and compounds of pathogenic bacteria (Williams and Bartholomew, which inhibit 2008). Soils collected at harvesting had relatively higher nodule biomass compared to those collected at planting confirming the observation that more rhizobia populations were 103 present in the soil at harvesting compared to the start of the season. This indicated that the high rhizobia populations present in the soil at harvesting enhanced nodule formation when soils collected at harvesting time were inoculated to cow pea plants. Shoot rhizobia biomass was enhanced population sizes by treatment indicating application that shoot that had also enhanced growth responded to treatment application than did nodule formation in this study. 8.4.3 Rhizobia characterization Most fast growing cultures formed wet colonies while slow growing colonies formed dry colony types. These observations were most probably the characteristic colony types of both slow and fast-growing colonies of cowpea. For example, Mpepereki et al., (1997) reported similar observations in cowpea rhizobia at Zimbabwe. However, they observed similar proportions of both fast-slow growing rhizobia (49% and 51 %, respectively) as opposed to this study where 97% of the isolates were fast growing. In this study indigenous studied after trapping rhizobia populations rhizobia obtained from an ISFM trial were strains using two cowpea varieties. PCR-RFLP analysis of 16S-23S IGS region groups genetically related strains (Jensen et al., 1993; Vinuesa et al., 1998). Therefore the 18 strains represented four genotypes. Wade et al., (2003), studying on diversity of indigenous rhizobia associated with three cowpea cultivars, also observed four genotypes in the rhizobia isolated from the cultivars. 104 CHAPTER NINE 9. CONCLUSIONS In this chapter conclusions AND RECOMMENDATIONS drawn from this study and recommendations for further research are summarized. 9.1 1) Conclusions Farmers considered in Makueni District were aware of declining it as a problem to crop production. soil fertility and This was shown by use of animal manure by fanners to enhance soil fertility in their farms. 2) Most farmers in Makueni District experienced food deficit for about eight months in a year and short rains were most reliable for crop production in the study sites. 3) Grain legumes occupied substantial proportions of cultivated land with the cultivated area occupied by the grain legumes increasing with decrease in cultivated farm size. 4) Local cowpea variety, Kang'au, had very high potential for nodule formation, biomass production and grain yield and performed better than all improved cowpea varieties, to the belief that local varieties are low yielding. contrary, Low yields observed by farmers may be probably due to poor management of the local varieties. 5) Cowpea nodulation, biomass production and grain yields were enhanced by addition of inputs in form of manure, TSP or manure+ TSP and that manure+ TSP treatment was commonly effective during the drier season, as well as during the wetter season at Kavuthu. This observation sites to enhance cowpea production. implied need to add inputs in the study 105 6) Grain yield was enhanced by addition of phosphorus implying that phosphorus was limiting cowpea grain production in the study sites. 7) Treatment addition enhanced nitrogen fixation in cowpea and improved indigenous soil rhizobia populations. 9.2 1) Recommendations There is need to train farmers on the importance of soil fertility improvement in the study sites, especially in the management and use of locally available farm inputs such as animal manure, compost and crop residues. Farmers should also be sensitized on the use of inorganic inputs, especially phosphorus (P), which can significantly be improved in the soils by addition ofP fertilizers. 2) Improved cowpea varieties, especially those used for on-farm trials in chapter seven should be included in the farming systems of Makueni District and cultivated along with the locally available varieties to enhance cowpea diversity and boost cowpea yields in the district. 3) Soil fertility improvement management nitrogen (ISFM) fixation, should growth interventions such as use of integrated soil fertility be considered to maximize and in grain legume cowpea production Techniques such as spot application of inputs and microdozing to enhance effective use of inputs in the study sites. performance in the study in sites. should be introduced 106 LITERA TURE CITED Abeyasekera, S., J.M. Ritchie and Lawson-McDowall. scores to determine Malawi. Experimental Alien, O.N. and E.K. characteristics, farmers' preferences 2004. Combining for bean varieties ranks and in Southern Agriculture. 38:97-107. Alien. 1981. The uses and nodulation. leguminoseae: University a source of Madison, book of Wisconsin, U.S.A. Amanullah, P. and M. Hatam. 2000. Yield potential of cow pea germplasm. Pakistan Journal of Biological Sciences. 3(5): 858-861. Audi, P.O., e.K. research, Ngugi, F.P. Wandera and R. Ndengwa. extension Recommendations KARI-Katumani, and development 1996. Dryland farming project. Dryland for smallholder farmers. Proceedings farming: of a workshop held in Machakos, Kenya. Pp. 47. Ayaga, G. and P.e. Brookes. 2005. Increasing fertilizer phosphorus use efficiency through organic matter cycling for enhanced maize production in Kenya. Pp. 345-362. In: Mugendi, P.N. Macharia, D.N., G. Kironchi, M. Mburu, J.G. Mureithi P.T. Gicheru, and F. Maina e.K.K. Gachene, (eds.). Capacity building for land resource management to meet the challenges of food security in Africa. Proceedings of the 21 st annual conference. Soil Science Society of East Africa, Nairobi, Kenya. Bala, A. and K.E. Giller. 2001. Symbiotic specificity of tropical tree rhizobia for host legumes. New Phytologist. Bationo, A., BJ. Ndunguru, 149: 195-207. B.R. N'tare, B.e. Christianson 1991. Fertilizer management strategies for legume-based and A.V. Mokwunye. cropping systems in 107 West Africa semiarid tropics. Pp 213-225. In: Johansen, C, K.K. Lee and K.L. Sahrawat (eds.). Phosphorus tropics. Proceedings nutrition of grain legumes of the international' workshop in the semi-arid held at International Crop Research Institute for the Semi-Arid Tropics (lCRISAT), Patancheru, India. Bekunda, M.A., A. Bationo, and H. Ssali. 1997. Soil fertility management A review of selected research trials. Replenishing in Africa: soil fertility in Africa. SSSA special publication. 51 :63-79. Bennett, J.E., P.L. Woomer and R.S. Yost. 1990. Users manual for MPNES mostprobable-number enumeration system. Version 1.0. NiITAL Project and 1990. Agriculture and University of Hawaii, Paia. Bockman, c., O. Kaarstad, O.H. Lie and 1. Richards. fertilizers. Agriculture group, Olso. Pp. 245 . Bressani, R. 1985. Nutritive value of cowpea. Cowpea research, production and utilization. In: S.R. Singh and K.O Rachie (eds.). John Wiley and sons Ltd. Bronson, J. 1995. Conducting resource management Buresh, R.J., P.c. capital Publication. in natural from Vanuatu. Clark University, Worcester, U.S.A. Smithson in Africa. PRA in The South Pacific: Experiences and D.T. Hellums. Replenishing 1997. Building Soil Fertility in Africa. soil phosphorus SSSA Special 51:111-149. Campo, J., V.J. Jaramillo, J.M. Maas. 1998. Pulses of soil phosphorus availability in a Mexican Oecologia. tropical dry forest: Effects of seasonality and levels of wetting. 115: 167-172. Chambers, R. 1983. Rural development: Putting the last first, Longman scientific and technical. Essex, U.K. 108 Chaturverdi, G.S., P.K. Aggarwal and S.K. Simba. determinate and indeterminate cowpeas Agricultural Sciences. 94: 137-144. 1980. Growth in dryland and yield of agriculture. Journal of Chien, S.H. and R.G. Menon. 1995. Factors affecting the agronomic effectiveness of phosphate rock for direct application. Fertilizer Research. 41 :227-234. Chui, J.N and M. Nadar. 1984. Evaluation of effects of Rhizobium phaseoli strains on nodulation, dry matter and grain yield of two bean (Phaseolus vulgaris) varieties. East African Agricultural and Forestry Journal. 44: 109-112. Conway, G. and J. McCracken. development: Development, 1988. Training notes for agroecosystem Ethiopia. International Institute for analysis for Environment and London. Cornejo, F.H., A. Varela and J. Wright. 1994. Tropical forest litter decomposition under seasonal drought: nutrient release, fungi and bacteria. Oikos. 70: 183190. Crow, V.L., B.D.W. Jarvis and R.M. Greenwood. homologies among acid-producing of Systematic Bacteriology. 1981. Deoxyribonucleic strains of Rhizobium. International 16S ribosomal isolated from banana DNA Merril). Applied Environmental problems characterization in the use of Microbiology. and F. Zapata. ISN (Ananas bacteria comosus (L.) 67(5):2375-2379. 1993. Misconceptions soil enrichment fixation. Plant and Soil. 152:25-52. and F.O. Pedrosa. of nitrogen-fixing (Musa spp.) and pineapple Danso, S.K.A., G. Hardarson Journal 31: I 52- 172. Cruz, L.M., E.M. Souza, O.B. Weber, J.1. Baldan, J. Dobereiner 200l. acid techniques and practical for estimating N2 109 Date, R.A. and J. Halliday. of rhizobia. 1987. Collection, isolation, cultivation and maintenance Pp 1-29. In: Elkan, G.!. (ed.). Symbiotic nitrogen fixation technology. Mercel Dekker, Inc. New York. Pp. 440. Davis, D.W., .E.A. Oelke, E.S. Oplinger, J.D. Dol, C.V. Hanson and D.H. Putman. 1991. Cowpea. Alternative field crops Manual. Pp. 1-9. de Lajudie, P., F. Laurent-Fulele, M. Neyra, A. Willems, B. Pot, D. Dewettinck, G. Maestrojuan, M.D. Coli ins, B. Dreyfus, K. Kersters and M. Gillis. 1994. Polyphasic taxonomy of rhizobia: emendation of the genus Sinorhizobium description of Sinorhizobium Sinrhizobium teranga Bacteriology. 44:715-733. Dennison, E.B. sp. melitoti comb, nov., S. saheli sp. Nov., and novo International 1961. The value of farmyard manure northern Nigeria. Empire Journal of Experimental Journal of in maintaining Systematic fertility extracts using leguminosarum randomly Journal of Microbiology. amplified biovar trifolii polymorphic 111 Agriculture. 29:330-336. Dye, M., L.Sk0t, L.R. Mytton, S.P. Harrison, J.J. Dooley and A. Cresswell. study of Rhizobium and populations DNA profiles. 1995. A from soil Canadian 41:336-344. Eardy, B.D., l.P.W. Young and R.K. Selander. 1992. Phylogenetic Rhizobium sp. strain Or 191, a symbiont Phaseoulus vulgaris, based on partial sequences genes. Applied Environmental Microbiology. of both Medicago position saliva of and of the 16S rRNA and nijH 58: 1809-1815. Evans, J., G.E. Oconnor, G.L. Turner, D.R. Coventry, N. Fettell, J. Mahoney, E.L. Armstrong and D. Walsgott. 1989. Nitrogen fixation and its value to soil N 110 Increases in Lupin, field pea and other legumes in south-eastern Australia. Australian Journal of Agricultural Research. 40:791-805. Fairhurst, T., R. Lefroy, E. Mutert and N. Batjes. 1999. The importance, and causes of phosphorus tropics. Agroforestry FAOSTA T. 2000. Site Organization distribution deficiency as a constraint to crop production in the Forum. 9(4): 2-8. Internethttp/www.fao.org/statists. Food and of United Nations (FAO). 1993. Year production Agriculture book. Volume 46, FAO, Rome. Food and Agriculture Organization of United Nations (FAO). 1993. Year production book. Volume 46, FAO, Rome. Fried, M., S.K.A. Danso and F. Zapata. 1983. The methodology fixation by non-legumes as inferred from field experiments Canadian Journal of Microbiology. ofN with legumes. 29:1053-1062. Gathumbi, S.M., G. Cadisch and K.E. Giller. 2002. for assessing N2-fixation of measurement N Natural abundance 15 as a tool of herbaceous, shrub and tree legumes in improved fallows. Soil Biology and Biochemistry. 34:1059-1071. Gibberd, V. 1995. Yield response offield crops to animal manure in semi-arid Kenya. Tropical Science. 35 :418-426. Gibson, A.H., B.L. Dreyfus and Y.R. Dormmergues. tropics. P. 328. In: Dormmergues, of tropical soils and plant 1982. Nitrogen fixation in the Y.R. and H.G. Diem (eds.). Microbiology productivity. Martinus NijhofflDr. W. Junk Publishers, The Hague. Giller, K.E. 2001. Nitrogen fixation in tropical cropping systems, 2nd Edition. CABl publishing, Wellingford, U.K. Pp. 423. 111 Giller, K.E., J.F. McDonagh and G. Cadisch. 1994. Can biological nitrogen fixation sustain Rimmer agriculture in the tropics? Pp. 173-191. (eds). Soil and sustainable International, In: Syers, J.K. and D.L. land management in the tropics. CAB Wellingford, U.K. Giller, K.E., G. Cadisch, C Halitosis, and E. Adams. 1997. Building soil nitrogen in Africa. Replenishing soil fertility in Africa. SSSA Special Publican. 51: 151- 192. Graham, P.H., MJ. Sadowsky, H.H. Keyser, Y.M. Barnet, R.S. Bradley, J.E. Cooper, DJ. Deley, B.D.W. Jarvis, E.B. Roslycky, B.W. Stijdom and J.P.W. young. 1991. Proposed minimal standards for description of root- and stem-nodulating Bacteriology. of new genera and species bacteria. International Journal of Systematic 41:582-587. Hardy, R.W.F., R.D. Holsten, E.K. Jackson and R.e. Burns. 1968. The acetyleneethylene assay for N2-fixation: laboratory and field evaluations. Plant Cresswell. 1992. Physiology. 43:1185-1204. Harrison, S.P., L.R. Characterization polymorphism Mytton, of L. Sket, Rhizobium using random M. Dye, isolates primers. by Canadian and C. amplification Journal of DNA of Microbiology. 38: 1009-1 0 15. Hartley, K.T. 1937. An explanation Nigeria. Experimental Herridge, of the effect of farmyard manure in Northern Agriculture. 19:224 -263. D.F., F.J. Bergerson and M.B. Peoples. 1990. Measurement of nitrogen fixation by soybean in the field using the ureid and natural 15N abundance methods. Plant Physiology. 90:708-716. 112 Higgins, C.F., G. Ferro-Luzzi 1982. A novel Ames, W.M. Barnes, J.M. Clement and M. Hofnung. intercistronic regulatory element of prokaryotic operons. hybridization studies of Nature. 289:760-762. Hollis., A.B. W.E. Kloos and G.H. Elkan. 1981. DNA:DNA Rhizobium Microbiology. Huber, 1. and japonicuni and Rhizobiaceae. Journal of General 123 :215-222. S. fingerprinting, related Selenska-Pobell. genome 1994. Pulsed-field size estimation gel electrophoresis- and rrn loci number of Rhizobium galegae. Journal of Applied Bacteriology. 77:528-533. Hulton, C.S.J., D.F. Higgins and P.M. Sharp. 1991. ERIC sequences: of repetitive typhimurium elements and other enterobacteria. Ikera, S.T. 2006. Combining improvement in the genomes of Escherichia a novel family coli, Salmonella Molecular Microbiology. Tithonia diversifolia 5:825-834. and Mijingu phosphate rock for of P availability and maize grain yields on a Chromic Acrisol in Morogoro Tanzania. Nutrient Cycling in Agroecosystems. 76(2-3): 249-260. Ikombo, B. 1984. Effects of farmyard manure and fertilizers on maize in semi-arid areas of eastern Kenya. East. African Agricultural and Forestry Journal. 44:266-274. Imungi, J.KJ. and N.N. Porter. 1983. Nutrient content of raw and cooked cowpea leaves. Food science. 48: 1252-1254. Jama, B., C.A. Palm, RJ. Buresh, A. Niang, C. Gachengo, Amadalo. 2000. Tithonia improvement diversifoloa G. Nzinguheba as a green manure in western Kenya: A review. Agroforestry and B. for soil fertility Systems. 49:201-221. 113 Jama, B., R.A. Swinkles and R.J. Buresh. 1997. Agronomic and economic evaluation of organic and inorganic sources of phosphorus in Western Kenya. Agronomy Journal. 89:597-604. Janssen, B.H. 1994. Integrated nutrient management: the use of organic and mineral fertilizers. Pp. 89-105. In: Van Reuler, H. and W.H. Prins (eds.). The role of plant nutrients for sustainable Wageningen, Jemo, M., R.e. food crop production in Sub-Saharan The Netherlands. Abaidoo, e. Nolte and W.J. Horst. 2006. Genotypic phosphorus Africa. variation uptake and dinitrogen fixation in cowpea on low-phosphorus for soils of southern Cameroon. Journal of Plant Nutrition and Soil Science. 169:816825. Jensen, M.A., J.A. Webster and N. Straus. 1993. Rapid identification the basis of Polymerase polymorphism. Chain Reaction-Amplified Applied Environmental and J.J. Holt (eds). Bergeys's rhibosomal Microbiology. Johnson, J.L. 1984. Nucleic acid in bacterial classification. of bacteria on DNA spacer 59:945-954. Pp. 8-11. In: Kreig, N.R. manual of systematic bacteriology. Volume 1. Williams and Wilkins, Baltimore, USA. Johnson, J.L. 1991. DNA reassociation and M. Goodfellow experiments. Pp. 21-44. In: Stackebrandt, (eds.). Nucleic acid techniques E. in bacterial systematics. John Willey & Sons Ltd, England. Jordan, D.e. 1982. Bradyrhizobium from leguminous 32:136-139. Transfer of Rhizobium japonicum genus novo a genus of slow-growing plants. International Buchanan 1080 to root nodule bacteria Journal of Systematic Bacteriology. 114 Kajalainen, S. and K. Lindstrorn. polymorphismanalysis 1989. Restriction fragment length of Rhizobium ga/egae strains. Journal of Bacteriology. 171 :5561-5566. Kamprath, EJ. 1991. Appropriate measurements of phosphorus tropics. (eds.). Phosphorus International Crop (ICRISAT), Patancheru, in the c., K.K Lee and K.L. nutrition of grain legumes in the semi-arid soils of the semi-arid tropics. Pp. 23-32. In: Johansen, Sahrawat availability Research Institute for Semi-arid Tropics India. Keyser, H.H., D.N. Munns and J.S. Hohenberg. 1979. Acid tolerance of rhizobia in culture and symbiosis with cowpea. Soil Science Society of American Journal. 43:719-722. Khan, T.N. 1973. A new approach to the breeding of pigeon peas (Cajanus cajan Millsp): Formation of composites. Euphytiea. 22: 373-377. Kihanda, F.M., G.P.Warren and S.S. Atwal. 2004. The influence of goat manure application on crop yield and soil nitrate variation in semi-arid eastern Kenya. Pp.163-186. In: Bationo, A. (ed.). Managing fertility in sub-Saharan nutrient cycles to sustain soil Africa. Tropical Soil Biology and Fertility Institute of CIA T, Nairobi, Kenya. Kimani, S.K. and J.K. Lekasi. 2004. Managing manures through their production cycles to enhance their usefulness: A review. Pp. 187-197. In: Bationo, A. (ed.). Managing nutrient cycles to sustain soil fertility in sub-Saharan Africa. Tropical Soil Biology and Fertility Institute of Cl AT, Nairobi, Kenya. 115 Kuykendall, L.D. and sinnorhizobium B. Saxena. 1991. Genetic diversity in Bradyrhizobium Jordan 1982 and a proposal for Bradyrhizobium L.M. elkanii sp. novo Canadian Journal of Microbiology. 38:501-505. KWAP Project. 1993. Extension methods and techniques: A manual for training of field extension workers. ETC-Kenya. Nairobi, Kenya. Pp. 80. Laguerre, G., M.P. Fernandez, heterogeneity V. Edel, P. among French vulgaris L. International ormand and N. Amarger. Rhizobium strains Journal of Systemic Bacteriology. Laguerre, G., M.R. AlIard, F. Revoy and N. Amarger. rhizobia by restriction isolated fragment 1993. Genomic from Phaseolus 43:761-767. 1994. Rapid identification length polymorphism amplified 16S rRNA genes. Applied Environmental analysis Microbiology. of of PCR60:56-63. Laguerre, G., P. Mavingui, M.R. Allard, M.P. Charnay, P. Louvrier, S.l. Mazurier, L. Rigottier-Gois and N. Amarger. fingerprinting and PCR-restriction chromosomal leguminosarum Microbiology. and symbiotic and its 1996. Typing by PCR DNA fragment length polymorphism gene different regions: Application biovars. Applied to analysis of Rhizobium Environmental 62:2029-2036. Lara, M., J.L. Ortega, B.E. Olgluin and F. Sanchez. nitrogen of rhizobia metabolism in Phaseolus 1988. Nodule expression vulgaris root nodules. Bothe, H., F.J. de Bruij and W.E. Newton (eds.). Nitrogen Pp. 617-622. and In. fixation: hundred years after. Gustav Fischer, Stuttgart. Loomis, R.S. and P.J. Connor. 1992. Crop Ecology. Productivity and management agricultural systems. Cambridge University Press, Cambridge. in 116 Ludwig, W., R. RosseIl6-Mora, R. Aznar, S. Klugbauer, Beimfohr, E. Brockmann, Klugbauer, N. Springer, D. Lane, R. Nietupsky, Schleifer. 1995. proteobacteria G. Kirchihof, Comparative sequence S. Dorn, M. Bachleitner, M. Weizenegger analysis system. Appl ied Microbiology. of 23S and K.H. rR A from 18: 164-188. Macharia, P.N. 2003. Integrating fanners' and scientists' soil mapping and management: S. Spring, K. Reetz, C. knowledge in participatory A case study from semi-arid eastern Kenya. East African Agricultural and Forestry Journal. 69:39-47. Maingi, J.M., N.M. Gitonga, C.A. Shisanya, B. Hornetz Population levels of indigenous Bradyrhizobia and G.M. Muluvi. 2006. nodulating soybean in two soils of the semi-arid and semi-humid Journal of Agriculture and Rural Development promiscuous agroecological zones. in the Tropics and Subtropics. 107(2):149-159. Mapfumo, P., S. Mpepereki and P. Mafongoya. 2000. Pigeon pea rhizobia prevalence and crop response to inoculation in Zimbabwean Experimental Marschner, smallholder-managed soils. Agriculture. 36:423-434. H. 1986. Mineral Nutrition of Higher Plants. Academic Press, London. Pp. 674. Martin, B., O. Humbert, M. Camara, E. Guenzi, J. Walker, T. Mitchell, P. Andrew, M. Prudhomme, G. Alloing, R. Hakenbeck, D.A. Marrison, G.J. Boulnois and J.P. Claverys. 1992. A highly conserved repeated DNA element located in the chromosome of streptococcus. Nucleic Acids Research. 20:3479-3483. 117 Mason, PJ. and J.G. Williams. 1985. Hybridization in the analysis of recombinant DNA. Pp. 113-137. In: Hames, B.D. and SJ. Higgins (eds.). Nucleic acid hybridization: a practical approach. ILR Press Ltd, Oxford, England. Pp. 245. Masumba, B.A.N. 1984. Physiological measurement for ranking bean cultivars with respect to relative drought and heat resistance. East African Agricultural and Forestry Journal. 44:290-296. Mathuva, M.N., M.R. Rao, P.e. Smithson and R. Coe. 1996. Improving mays) yields in semiarid highlands of Kenya: agroforestry maize (Zea or inorganic fertilizers? Field Crops Research. 55:57-72. McCown, R.I., Keating sustainable B.A., Probert M.E. and Jones R.K. crop production in semiarid Africa. Qiao, and M. An. 2007. 1992. Strategies Outlines of of Agriculture. 21(1):21-23. Miao, S.J., Y.F. development X.Z. Han formation in soybeans (Glycine max L.) in response to phosphorus in solution culture. Podosphere. Mortimore, Nodule and supply 17(1 ):36-43. M.J., B.B. Singh, F. Harris and S.F. Blade. 1997. Cowpea in traditional cropping systems. Pp. 99-113. In: Singh, B.B., D.R. Mohan, K.E. Dashiell and L.E.N. Jackai (eds.). Advances International in cowpea research. Institute of Tropical Agriculture Co-publication of (UTA) and Japan International Research Centre for Agricultural Sciences (JIRICAS). IITA, Ibadan, Nigeria. Mpepereki, S., A.G. Wollum specificity and F. Makonese. 1996. Diversity of cowpea rhizobia indigenous to Zimbabwean 186:167-171. in symbiotic soils. Plant Soil. 118 Mpepereki, S., F. Makonese of indigenous and A.G. Wollum. 1997. Physiological rhizobia nodulating characterization Vigna unguiculata in Zimbabwean soils. Symbiosis. 22:275-292. Mueller, J.G., H.D. Skipper, E.R. Shipe and L.W. Grimes. 1988. Intrinsic antibiotic resistance in Bradyrhizobium japonicum. Soil Biology and Biochemistry 20(66):879-882. Muleba, N. and. Ezumah H.C. 1995. Optimizing cultural practices for cowpea in Africa. Pp 289-295. In Singh, S.R. and K.O. Rochie (eds.). Cow pea research, production and utilization. John Wiley and Sons Ltd, Chichester, Uk. Muli, M.B. and H.M. Saha. 2000. Participatory evaluation of cowpea cultivars for adaptation and yield performance in coastal Kenya. Pp. 267-272. In: Mureithi. J.G., G.K.K. Gachene, F.N. Muyekho, M. Onyango, L. Mose and O. Magenya (eds.). Participatory technology development for soil management by small holders in Kenya. Kenya Agricultural Research Institute, Nairobi, Kenya. Mulongoy, K. and A. Ayanaba. 1985. Dynamics of the population and soybean rhizobia Microbiology and Biotechnology. sizes of cowpea at three locations in West Africa. World Journal of 2(2):301-308. Muruli, B.I., R.S. Pathak, D.M. Mukunya, A.K. Karel, S.O. Keya and H. Ssali. 1980. Cowpea research in Kenya. Tropical grain legume Bulletin. 19: 13-16. Muthamia, J.G.N. and F.K. Kanampiu. 1996. On-farm cowpea evaluation In the marginal areas of eastern Kenya. Pp. 677-685. In: Fungoh, P.O. and G.C.O. Mbadi (eds.). Focus on Agricultural Research for Sustainable Development Changing Economic Environment. Conference, Nairobi, Kenya. Proceedings in of the 5th KARI Scientific 119 Nadar, H.M. 1984. Intercropping rainfall conditions and intercrop component interaction under varying in eastern Kenya. I: Maize/bean intercrop. East African Agricultural and Forestry Journal. 44:166-175. Nakawuka, C.K. and E. Adipala. component interactions 1999. A path coefficient analysis of some yield in cowpea. African Crop Science Journal. 7(4):327- 331. Nelson, D.e. and R.H. Robichaux. 1997. Identifying associated with yield under intercropping plant implications architectural traits of genotype-cropping system interactions. Plant Breeding. 16: 163-170. Ng, N.Q. and R. Marechal. 1985. Cowpea taxonomy, origin and germplasm. Pp. 11- 21. In: Singh S.R. and K.O. Rachie. (eds.). Cowpea research, production and utilization. John Wiley and & Sons, New York. Nour, S., J.e. Cleyet-Marel, Genotyphic D. Beck, A Affosse and phenotyphic and M.P. diversity of Rhizobium S., J.C. Cleyet-Marel, heterogeneity description of strains of Rhizobium Systemic Bacteriology. Nziguheba P. Normand G, R Merckx, nodulating 40:345-354. and M.P. Fernandez. chickpea mediterraneum 1994. isolated from chickpea (Cicer arietinum L.). Canadian Journal of Microbiology. Nour, Fernandez. (Cicer 1995. Genomic arietinum sp. novo International L.) and Journal of 45:640-648. e.A. Palm and P.K. Mutuo. 2004. Combined use of Tithonia diversifolia and inorganic fertilizer for improved maize production in a phosphorus deficient soil in western Kenya. Pp. 329-346. In: Bationo, A. (ed.). Managing nutrient cycles to sustain soil fertility in sub-Saharan Tropical Soil Biology and Fertility Institute of Cl AT, Nairobi, Kenya. Africa. 120 Ochman, H. and A.c. substitution Wilson. 1987. Evolution in bacteria: evidence rate in cellular genomes. Journal of Molecular for universal Evolution. 26:74- 86. Odee, D.W., K Haukka, S.G. Mclnroy, J.I Sprent, J.M. Sutherland and LP.W. Young. 2002. Genetic and symbiotic characterization and herbaceous of rhizobia isolated from tree legumes grown in soils from ecologically diverse sites in Kenya. Soil Biology and Biochemistry. 34:801-811. Ogbonnaya, c.r, B. Selection Sarr, C. Brou, O. Diouf, N.N. Diop and H. Roy-Macauley. of cow pea genotypes in hydroponics, 2003. pots and field for drought tolerance. Crop Science. 43:1114-1120. Ojiem, J.O., C.A. combining Palm, and M.A. organic and inorganic phosphorus a humic-Nitosol Managing E.A. Okwuosa in western nutrient Kenya. Mudeheri. 2004. Effect of sources on maize grain yield in Pp. 347-358. cycles to sustain soil fertility In: Bationo, A. (ed.). in Sub-Saharan Africa. Tropical Soil Biology and Fertility Institute of CIAT, Nairobi, Kenya. Okalebo, J.R., C.A. Palm, J.K. Lekasi, S.M. Nandwa, C.O. Othieno, M. Waigwa and K.W. Ndung'u. 2004. Use of organic and inorganic resources to increase maize yields in some Kenyan infertile soils: A five-year experience. Pp. 359372. In: Bationo, A. (ed.). Managing nutrient cycles to sustain soil fertility in sub-Saharan Africa. Tropical Soil Biology and Fertility Institute of CIAT, Nairobi, Kenya. Okalebo, J .R., K.W. Gathua and P. Woomer. 2002. Laboratory plant analysis: A working Nairobi, Kenya. Pp. 128. manual. The second edition. methods of soil and SACRED Africa, 121 Okoko, E.N. and S. Makwaro. 2000. Evaluation of the effect of compost and inorganic fertilizer on maize yield in Nyamira District South West Kenya. Pp. 23-29. In: Mureithi, J.G., C.K.K. Gachene, F.N. Muyekho, M. Onyango, L. Mose and O. Magenya (eds.). Participatory management by smallholders conference technology in Kenya. Proceedings of the soil management development for soil of the 2nd scientific and legume research network projects. KARI, Nairobi, Kenya. Olufajo, 0.0. and B.B. Singh. 2000. Advances in cowpea cropping systems research. In: Fatokun, CA, S.A. Tarawali, Pp. 267-277. B.B. Singh, P.M. Kormawa and M. Tamo (eds.). Challenges and opportunities for enhancing sustainable cowpea production. lIT A, Ibadan Nigeria. Onyango, R.M.A., TJ. Mwangi, W.W. Kiinya, M.K. Kamidi and M.W. Wanyonyi. 2000. Evaluation production for smallholder maize in North Rift Kenya. Pp. 1-12. In: Mureithi, J.G., c.K.K. Gachene, F.N. Muyekho, technology of organic and inorganic fertilizers M. Onyango, L. Mose and O. Magenya (eds.). Participatory development Proceedings for soil management of the 2nd scientific conference by smallholders in Kenya. of the soil management and legume research network projects. KARI, Nairobi, Kenya. Owolade, O.F., M.O. Okande, B.S. Alabi and J.A. Adediran. 2006. Phosphorus level affects brown blotch disease development and yield of cowpea. World Journal of Agricultural Palm, C.A. Sciences. 2(1): 105-108. 1995. Contribution intercropped of agroforestry trees to nutrient plants. Agroforestry Systems. 30: 105-124. requirements 111 122 Peffetti, D., C. Scotti, S. Gnocchi, S. Fancelli and M. Bazzicalupo. 1996. Genetic diversity of an Italian Rhizobium meliloti population from different Medicago saliva varieties. Applied Environmental Microbiology. 62:2279-2285. Pilbeam, C.A., M. Wood and P.G. Mugane. 1995. Nitrogen use in maize-grain legume cropping systems in semi-arid Kenya. Biology of Fertilizers and Soils. 20:57-62. Polhill, R.M. and L.J.G. Van der Maesen. 1985. Taxonomy of grain legumes. Pp. 336. In: Summerfield, R.J. and E.H. Roberts (eds.). Grain Legume Crops. ColI ins, London. Poshiwa, X., R. Nyoka, N. Chikumba, T. Gatsi and 1. Chakoma. 2007. A baseline socio-economic evaluation of an European Union funded project on use of micro-symbionts in agroforestry systems in Zimbabwe. Journal of Sustainable Development in Africa. 8(1): 186-196. Powell, J.M. 1986. Manure for cropping: A case study from central Nigeria. Experimental Agriculture. 22: 15-24. PRA Programme Egerton University. rural appraisal practitioners. 1995. PRA field handbook for participatory 1995. The PRA Programme Egerton University, Njoro, Kenya. Pp. 84. Probert, M.E., J.R. Okalebo smallholders' and R.K. Jones. 1995. The use of manure fams in semiarid eastern Kenya. Experimmental on Agriculture. 31:371-381. Remanandan, P., A. Shakoor and E.C.K. Ngugi. 1982. Pigeon peas germplasm mission in Kenya. Genetic Resources Progress Report (ICRISAT). 48: 1-40. 123 Republic of Kenya. 2001. The 1999 population and housing census. In: Central Bureau of Statistics and Ministry of Finance and Planning. Volume 1. Pp. 415. Rowe, A.G. 1980. Alternative crops 3: Grain legumes. Zimbabwe Agricultural Journal. 79: 221-227 Saha, R.e. and M.B. Muli. 2000. Effect of combining green manure legumes, farmyard manure and inorganic nitrogen on yield of maize in coastal Kenya. Pp.l03-113. In: Mureithi, J.G., e.K.K. Gachene, F.N. Muyekho, M. Onyango, L. Mose and O. Magenya (eds.). Participatory soil management conference technology by smallholders in Kenya. Proceedings of the soil management development for of the 2nd scientific and legume research network projects. KARI, Nairobi, Kenya. Saidou, A.K., R.e. Variability Abaidoo, B.B. Singh, E.N.O. Iwuafor and N. Sanginga. of cowpea breeding lines to Iow phosphorus tolerance 2006. and response to external application of phosphorus. Pp. 413-422. In: Bationo A, B. Waswa, J Kihara and 1. Kimetu (eds.) Advances management in sub-Saharan Africa: Challenges in integrated soil fertility and opportunities. Springer, Dordrecht, The Netherlands. Sakala, WD., G. Cadisch, K.E. Giller. 2000. Interactions and pigeon pea and mineral nitrogen fertilizers nitrogen mineralization. between residues of maize during decomposition Soil Biology and Biochemistry. and 32:679-688. Sanchez, P.A., K.D. Shepherd, MJ. Soul, F.M. Place, RJ. Buresh and A-M. N. Izac. 1997. Soil fertility replenishment capital. Replenishing 46. in Africa: An investment in natural resource Soil Fertility in Africa. SSSA Special Publication. 51: 1- 124 Sande, S., c.K.P.O. Ogol and P.L. Woomer. 2004. Soil food webs as influenced by soil fertility management (ed.) Improving empowering practices. P. 55. In: Bationo, J Kimetu and J Kihara. human farmers welfare and to combat environmental soil fertility conservation degradation. by Tropical Soil Biology and Fertility Institute of Cl AT, Nairobi, Kenya. Selenska-Pobell, S., L. Gigova and N. Petrova. 1995. Strain-specific fingerprinting of Rhizobium galegae generated by PCR with arbitrary and repetitive primers. Journal of Applied Bacteriology. 79:425-431. Shakoor, A., E.C.K. Ngugi, M.S. Muthoka and J. W. Karnau. 1984a. Improvement cowpea varieties for dry areas in Kenya. East African Agricultural of and Forestry Journal. 44: 312-317. Shakoor, A., W.K. Rono and E.C.K. Ngugi. 1984b. Performance determinate varieties of green gram in semi-arid of early maturing areas. Eastern African Agricultural and Forestry Journal. 44:324-326. Sharma, R.C., P.M. Govindakrishnan, different needs green manures of potato-wheat R.P. Singh and H.c. Sharma. 1988. Effect of and fodder crops on crop yields and phosphorus systems in the north-western plains of India. Experimental Agriculture. 24:79-103. Sharples, G.J. and R.G. L1oyd. 1990. A novel repeated DNA sequence located in the intergenic regions of bacterial chromosomes. Nucleic Acids Research. 18:6503-6508. Shearer, G. and D.H. Kohl. 1986. N2-fixation natural in field settings: estimations N abundance. Australian Journal of Plant Physiology. J5 based on 13:699-756. 125 Sheldrake, A.R. and A. Narayana. 1979. Growth development pigeon peas tCajanus cajan). Journal of Agricultural Shetty, S.V.R., B.R. Ntare, A. Bationo and C. Renard. mixed cropping productivity and nutrient uptake in Sciences. 92: 513-526. 1995. Millet and cowpea in systems of the Sahel: A review of strategies and sustainability. Pp. 293-303. In: Powel, J.M., S. Fernandez- Rivera and T.O. Williams (eds.). in mixed farming for increased Livestock and sustainable systems of Sub-Saharan Africa. nutrient cycling International Livestock Centre for Africa (ILCA), Addis Ababa, Ethiopia. Siderius W. and F.M. Muchena. 1977. Soils and environmental agricultural research stations in Kenya. Miscellaneous conditions in paper No. M5. Ministry of Agricultural University. Sikora, S., S. Redzepovic, Bradyrhizobium I. Pejic and V. Kozumplik. japonicum field population revealed by RAPD fingerprinting. Journal of Applied Microbiology. Singh, B.B., O.L. Chamblis 1997. Genetic diversity of 82:527-531. and B. Sharma. 1997. Recent advances in cowpea breading. Pp. 30-49. In: Singh, B.B., D.R. Mohan, K.E. Dashiell. and L.E.N Jackai (eds.). Advances in cow pea research. IITAIJIRCAS, Ibadan, Nigeria. Ibadan, Nigeria. Singleton, P.W., B.B. Bohlool and P.L. Nakao. 1992. Legume rhizobia inoculation in the tropics: Myths and realities. Myths and science of soils in the tropics. ASA, SSSA. Special publication. 29: 135-155. Smaling, E.M.A. management 1993. An agroecological with special University Wageningen, reference framework to Kenya, The Netherlands. for PhD. integrating thesis. nutrient Agricultural 126 Smaling, E.M.A. and A.R. Braun. 1996. Soil fertility research in Sub-Saharan Africa: A new dimension, new challenges communication in soil science and plant. Analysis. 27:365-386. Smaling, E.M.A., S.M. Nandwa and B.H. Janssen. 1997. Soil fertility in Africa is at stake. Replenishing soil fertility in Africa. SSSA special publication. 51 :47- 61. Sobral, B.W.S., RJ. Honeycut and A.G. Atherly. 1991. The genomes of the family Rhizobiaceae: size, stability Journal of Bacteriology. Somasegaran, and rarely cutting restriction endonucleases. 173 :704-709. B. and HJ. Hoben. 1985. Methods in legume-Rhizobium technology. University of Hawaii NiIT AL * Project and MIRCEN*. Pp.367. Sprent, J.1. 1971. Effects of water stress on nitrogen fixation in root nodules. Plant and soil. 35(1):225-228. Stern, MJ., G. Ferro-Luzzi Ames, N.H. Smith, E.C. Robinson and C.F. Higgins. 1984. Repetitive extragenic palindromic sequences: A major component of the bacterial genome. Cell. 37:1015-1026. Stevenson, F.C. and C. Van Kessel. 1997. Nitrogen contribution of pea residue in hummocky terrain. Soil Science Society of American Journal. 61 :494-503. Stoetzer, H.M. and B.H. Waite. 1984. Evaluation of bean diseases in grain legume project trials, Katumani Research Station. East African Agricultural and Forestry Journal. 44:333-354. Stoorvogel, J.J., E.M.A. Smaling and B.H. Hansen. balances in Africa Research. 35:227-235. at different scales: 1993. Calculating soil nitrogen IJ supra-national scale. Fertility 127 Subarao, G.Y., C. Renard, A. Bationo, N. van-Duivenbooden Alternative technologies Pp. 121-132. and C. Bielders. 1999. for Sahelian crop production systems in West Africa. In: Faroda, advances in management A.S., N.L. Joshi and S. Kattju (eds.). Recent of arid ecosystems. Arid zone research Association of India, Jodhpur, India. Swift, MJ., P.D. Seward, P.G.H. Frost, J.N. Quresh and F.N. Muchena. term experiments 1994. Long- in Africa: Developing a database for suitable landscape use under global change. Pp. 229-251. In: Legh, R.A. and A.E. Johnson (eds.). Longterm experiments International in agricultural G. Gueye, A.T. Ba and B. Dreyfus. biological nitrogen fixation by pterocarpus park in Senegal Tarawali, ecological sciences. CAB Wellingford, England. Sylla, S.N., 1. Ndoye, Biotechnology. and using 15N 2002. Estimate of luvens in semi arid natural forest natural abundance method. African Journal of 1(2):50-56. S.A., B.B. Singh, M. Peters and S.F. Blade. 1997. Cowpea fodder. Pp 313-325. haulms as In: Singh, B.B., D.R. Mohan Raj, K.E. Dashiell L.E.N Jackai (eds.). Advances in cowpea research. lIT AIJIRCAS, and Ibadan, Nigeria. Tesfaye, M., DJ. Peterson and F.B. Brian Holl. 1997. Comparison of partial 23S rDNA sequences from Rhizobium species. Canadian Journal of Microbiology. 43:526-533. The MacmiJlan Atlas. 1997. The Macmillan primary school atlas. Macmillan (Publishers) Ltd. Nairobi, Kenya. Pp. 57. Kenya 128 Tisdale, S. and W. elson. Macmillan Publishing, 1975. Soil fertility and fertilizers. Third New York. Urtz, B.E. and G.H. Elkan. 1996. Genetic diversity among Bradyrhizobium that effectively Microbiology. edition. nodulate peanut (Arachis hypogaea). Canadian isolates Journal of 42:1121-1130. Van Ek, G.A., J. Henriet, S.F. Blade and B.B. Singh. 1997. Quantitative assessment of traditional cropping systems in the Sudan Savanna of Northern Nigeria. 11. Management and productivity Agricultural Research. of major cropping systems. Samaru Journal of 14:47-60. Van Kessel, C. and C. Hartley. 2000. Agricultural management of grain legumes: has it led to an increase in nitrogen fixation? Field Crops Research. 65: 165-181. Van Rheenen, H.A., O.E. Hasselbach and S.G.S. Muigai. 1981. The effect of growing bean together with maize on the incidence of beans diseases and pests. Netherlands Journal of plant pathology. 87:193-199. Vandamme, P., B. Pot, M. Gillis, P. de Vos, K. Kersters Polyphasic Microbiologyl Vanlauwe, taxonomy, a consensus approach to and J. Swings. bacterial 1996. systematics. Reviews. 60:407-438. B., K. Aihou, P. Houngnandan, J.Diels, N. Sanginga and R. Merckx. 2001. Nitrogen management in 'adequate' input maize-based agriculture in the derived savanna benchmark zone of Benin Republic. Plant and Soil. 228:61- 71. Versalovic, J., M. Schneider, fingerprinting F.J. de Bruijn and J.R. Lupski. of bacteria using repetitive sequences-based 1994. Genomic polymerase reaction. Methods in Molecular and Cellular Biology. 51 :25-40. chain 129 Vincent, J.M. 1970. International A manual Biological for practical Programme study Handbook of root-nodule bacteria. No. 15. Blackwell, Oxford. Pp. 164. Vinuesa, P., J.L.W. Rademaker, characterization FJ de Bruijin and D. Werner. of Bradyrhizobium strains nodulating endemic wood legumes of the canary islands by PCR-Restriction genes encoding 16S rRNA (l6S Spacers. Repetitive partial 16S Length Polymorphism rDNA) Extragenic Palindromic rDNA sequencing. 1998. Genotypic and 16S-23S rDNA analysis of Intergenic PCR genomic fingerprinting Applied Environmental and Microbiology. 64:2096-2104. Vos, P., R. Hogers, M. Beeker, M. Reijans, T. Van de Lee, M. Hornes, A. Frijters, J. Pot, J. Peleman, M. Kuiper and M. Zabeau. 1995. AFLP: a technique for DNA fingerprinting. Nucleic Acids Research. 23:4407-4414. Wade, T.K., 1. Ndoye, S. Braconnier, B. Sarr, P. de Lajudie and M. Neyra. 2003. Diversity of indigenous bradyrhizobia associated with three cowpea cultivars (Vigna unguiculata L.) grown under limited and favourable water conditions in Senegal (West Africa). African Journal of Biotechnology. Waite, B.H., A. Shakoor, W. Songa. and E.C. Ngugi. 2(10: 13-22. 1984. Principal diseases of pigeon peas, cowpeas, chickpeas and green gram in the semi-arid areas in Kenya. East African Agricultural and Forestry Journal. 44: 364-375. Wanjekeche, E., TJ. Mwangi, P. Powon and J. Khaemba. 2000. Farmer participation in the evaluation of grain legumes in north-western Kenya. Pp. 13-21. In: Mureithi, J.G., C.K.K. Gachene, F.N. Muyekho, M. Onyango, L. Mose and O. Magenya (eds.). Participatory technology development for soil management 130 by smallholders in Kenya. Proceedings of the 2nd scientific conference of the soil management and legume research network projects. KARI, Nairobi, Kenya. Wedlock, D.N. And B.D.W. Jarvis. 1986. DNA homologies fredii, rhizobia Bradyrhizobium that nodulate Galega sp., and between Rhizobium other Rhizobium species. International Journal of Systematic and Bacteriology. 35:550-558. Wild, A.1988. Russell's Soil Conditions and Plant Growth. Eleventh edition. John Wiley and Sons, New York. Willems, A., R. Coopman and G. Gillis. 2001. Comparison of sequence analysis of 16S-23S DNA spacer regions, AFLP analysis and DNA:DNA hybridization in Bradyrhizobium. International Journal of Systematic Evolutionary Microbiology. 51 :623-632. Williams, R.D. and P.W. Bartholomew. 2008. Effect of cowpea and pea inocula on cool season grass. Allelopathy Journal. 21(2):381-388. Witty, J.F. 1983. Estimating of N2-fixation in the field using 15N labeled fertilizer: some problems and solutions. Soil Biology and Biochemistry. Xu, L.M., C. Ge, Z. Cui, J. Li and H. Fan. 1995. Bradyrhizobium 15:631-640. Liaoningensis sp. novo Isolated from the root nodules of soybean. International Journal of Systematic Bacteriology. 45: 706-711. Young, lP.W. 1996. Phylogeny and taxonomy of rhizobia. Plant and Soil. 186:45-52. Zhang, X., R. Harpen, M. Karsists and K. Lindstrom. 1991. Diversity of Rhizobium bacteria isolated from the root nodules of leguminous Journal of Systematic Bacteriology. 41: 104-113. trees. International 131 APPENDICES Appendix 1 Maps ofYikivumbu, Ndunguni, Kavuthu and Matiku Sublocations drawn by farmers during farmer meetings VOLOLO SUB. LOCATION , \ I 1 I I \ \ \ \ , "- -" , , I I "" " , --- " , " \ ", \ r / I " " +6 I I I I I / :\p...~\ ~\)p... '-l\~ I I ~f>.0t-/\ • I +L> Assistant Chief's Office I I I I I ~ --------....----.-__ I /................... / NGOMANO VILLAGE / \\ @ \ '.•... ~ / tSsl. OD :::.::: > 1'--. z ::) W > I 0 I z (9 <C -.:~.•.. ,"'" ""'\ j ;; """:......,;::" \'.... \ -- .-"" KYENI \ VILL. SR I \ ~ \ @ LEGEND: -SChOOI SS -Seed store ~ -Market » -Shrin£>s CD - CaUl£> dip H -Hospital + - Figure 1 - - __ Village boundaries Boundaries __________ - River (Stream) (Dtspensary) Church Map ofYikivumbu Sub-location 132 ". ' S.Army Church ..'" K Muthyaamui I ,.SOWETO MARKET Dam ,! / .... MAKUTANO • MARKET ~~.::..- ~:luia • ':~'r~ J.- V': Nyumu ':\ .• Mullu \- •.• 1;? ~, //' <y V Killi .".\"'?/ ":'" -'..... -.\, Nzioka Z\,\" :\( ~ <\\',\~\"" ~ \'\- ~ N7..rmgi •..•• • Ale Kuvuva \~ \{ »>: .' ~.~::~_\~;::::::.:;"- +NI ••;-;- .•...::"',/ /" •• -" ,~ .y/'~<':, • .,Y/. \~\ .r,: .e> 7' / Ass, Chiefs Office :' : Donald \\ (/1--0 ./ ••. Ii 3 \ , go 'C-'-'MUlhusi Muteti \ \7. Mutuku -:>: :)1l ~ £,MilOlU \\~. S: ("' Wambua Ngule ""<::",. ",,- '-,:. .&. ~: '- ., River :::::=== Road ____ Figure 2 Dam • Outside boundary Map ofNdunguni Nyumu <:".............::.. -... ~,-_... Sub-location " 1\ \ ". "\ ~ Town/Market ! : i I, y :.:.\\ I !I 0 \\ v I •• \.. ..\-: •••••••••••....• ~ .• , •••• ----._~....-.~-•.•.-.,._... '-"-_ ... _ .._.-.---_ . Homestead ji ,I 1'- ~ \),- ~" •. A Ngule ...... .. .. " -.v ". , ~ ~. \'\. \~~. \;'~ •••. I ~\KinYili •••. CATHOLIC CHURCH Ndonye KEY: Muteli! Ngu,"~ ....~?~_--:;, ; : L..-;r';: ..--' 1\ .-<;:.-::y i 1 NDUNGUNI \\\±.. !I! 1:'< I' -.\ Nyamai", i I i !!l .:\ e \ ,: i If I .\ : \ r, l.- I!. : ~ \'-'\\ '7 ,: :\ \'\. ~ Mwau -. ~ ••.•·.NDUNGUNi\ '\ SCHOOL \,', .o.~· l' I: ..-.. • : \·r_.·;:-~OhfiKitenge ',-'P .' ;JUNCTION ,'~IKWAMBITI -:> .. I Mukula ••... :;:::~~:>/-.-".\ \\~~ KATlVANI ....·.TOWN i ~~' .~>-. '!:-;::~~ "V/ i~ ••••• 133 \ I ,I: lJ'~-~~!£J11 ~~rHu hI [ ------_ -, O( l _ MARKO- " (I, !If Kavuthu (j i 1+ Cdtholic ..J 0 j + Redeemtld /- J... J 1~ U) "11 t~ \ avu u _, PI!, School 'tAlC, \ Kavuthu I I +ABC I U\nhi ~/-------~ j-TK Church 51 GNCA K AV UT H U I ----- 0 KioSKS " , Salvation army + Church A~C\t Catholic Ngo~d Church Church I T U M BUN / • IN G 0 M A N 0 I V ILL I I I 9lnne, I ~ ~\\ll' AGE Liani Shrine MASUE SI LOCATION +AIC Tumbuni I IVILLAGEI VILLAGE -- + Catha/!: Church \ ~~ I I W E 5 MUIU V ILL A 'G E ,MUllSYA I\ VILLAGE I I Figure 3 Map ofKavuthu MULU \ Sub-location LEGEND ::== Roads - S/Location boundary •• - - Vil~ge boundary 134 IVISUNI VILLAGE N Church \j School Shrine - ~ Roads Dam ------- Riv~r • ® OD o Figure 4 LEG END + - SlLocation boundary - Market Map ofMatiku Sub-location Village boundary ~ 135 Appendix 2 Questionnaire Questionnaire No. D DD MM YY ITIJ Date of interview: Name of respondent District Code Division: codeD Location: 1 Sub Location: __________ Household code: code 1 ---'1 --' D code D --' __________ Village: 1 D code D D SECTION A: FAMILY INFORMATION Table A.I. Personal information for members of the household currently resident: Fill where appropriate ID Name Age (Group) 1-::;20yrs 2- 21-30 3- 31-40 4-41-50 5- 51-60 6-~61 yrs 1 2 3 Marital status Sex 1- Single 2- Married 3- Widowed 4- Divorced 5- Separated Male-I Female-2 4 5 Relationship to HIH head 1- Head 2- wife/Husband 3- Son/daughter 4- mother/father 5-other l-None 2- Primary 3-Secondary 4-University 5-0ther 7 6 I 2 3 Total resident HIH members Highest education Level . Main occupation I-Farming! Livestock keeping 2·Business 3-Employed 4- Other 5-None 8 136 I (A) What is the general condition of housing in the HH? Specify condition for up to five houses: Table A 2 Housing conditions within the Household Wall House No 1- Earth mud 2- Bricks/Stones 3- Iron sheets 4- Wood slabs 5- Maize stovers 6- Withies I 1. Main House 2 Floor Roof I Earth/mud 2- Cement 3- Withies I-Thatch 2-lron sheets 3-Tiles 4-0ther 3 4 2 3 4 5 6 SECTION B: FARM INFORMATION Farm owned and managed by household 1'--__ I(B) Which year did you settle in this farm? 2(B) How many fields do you operate? Field ID I Total area (Acres) 2 Cultivated area (Acres) Area under Legumes (Acres) 3 4 Tenure System I-surveyed 2-Lease 3-unsurveyed 4-Communal 5- Other Ownershi p I-Own 2-Rent Rent In Land cost: Rent out land income 5 6 7 8 Table B.l. Please provide fields information as specified in the table below for the current season 3(B) Besides the legumes you grow on your fields, would you like to try others? 1- Yes 2-N°D 4(B) Who makes decisions concerning types of crops grown in Iyour farm? 0 l-Household head 3- Household head and son 5- Wife of household head 0 0 2- Household head and wife 4- Son of household head 6- Other (specify) c=J 0 137 Table B.2. Information fields Field ID 1 on the crops grown in the fields and production Crops grown in field 1- Beans 2- Pigeon peas 3- Cow peas 4-Green grams Legume location where terraced 1- Inside 2- On top 3- On slopes 4 Away 5- Other 2 3 Table B.3: Please provide information Unit Of yield of the Total produced I 5 4 on soil fertility, Soil Erosion Problems and Control Field [D Soil fertility Status Fertile Moderately fertile Poor Farm inputs I-Manures 2-Fertilizers 3-Crop residues 4-Compost 5-None 6-0ther(Specify) Soil Erosion Problem Yes No Soil Conservatio n structures present Yes No Conservation structures types present Bench terrace Bench terrace with grass Bench terrace with trees Bench terrace with grass and trees Check dam Cut off drain Other (specify) I 2 3 4 5 6 5(B) In case you use fertilizer when required? I-Yes D 2-No In your farm, IS fertilizer available In local market D 6(B)Incase you have soil erosion problems and you do not have terraces, give reasons why you don't have them? 7(B) Where do you get your manure? I-own 2-buy 3-NeighbourD D D Other (specify) _ 138 D D 8(B)Would you like to try other fertility improvement methods? I-Yes 2-No 9(B) If you were trained on best fertilizer management practices, would you use fertilizer in your farm? l-yeSO 2-noO SECTION C: WEEDS, PESTS AND DISEASES T a bl e ID - W ee d s, pests an d diiseases Field Weeds in your farm ID l-Kithangai 2-Ikoka 3-Munyeeli (Mukuutu) 4-Munzee 5-Mbiu (kiviu) 6-Kikatu 7-Ukuku §.-Untunga 9-mbete (kavete) IO-Ikongo II-Song'e 12-Lamuyu I3-0ther (specify) 1 2 I-hoe/panga weeding 2-plough-weeding 3-use of herbicide 4-burning 5-carrying out offarm 6-other (specify) Pests problems I-Squirrels 2-birds 3-monkeys 4-scania "Osarna" 5-weevils 6-Wiu 7-Umuu 8-0ther (specify) 3 4 Weed control in your farm Disease problems I-mbaa 2-rust 3-other (specify) 5 SECTION D: LIVESTOCK KEEPING D D 1(0) Do you keep livestock? I-Yes 2-.No 2(0) Which of the following livestock do you ke~ l-Cattle 2-goat 3-sheep L-J 5-donkey 6-rabbit 7-other (specify) 3(0) How do you keep your livestock l-free-range 2-tethering 3-paddocking (Specify) _ c::::=::J c=::J c::::=::J C=::J C=::J 4-poultry _ C=::J 0 C=::J c:=J 4-other SECTION E: FOOD SECURITY 1(E) Do you have food shortage during the dry seasons? 1-Yes IL-__ 2(E) What is your ~ource of food during dry seasons l-reserve 4-neighbours (own) U 0 2-purchaseD 5-0ther (specify) --' 2-No 3-ReliefD _ 139 Appendix 3 Experimental design (ISFM trial) On-farm trial design with 4 treatments and 9 cow pea varieties in three blocks Block 1 Treatment/varieties 3 I 2 4 I 2 3 4 5 6 7 8 9 5 2 I 4 7 6 8 3 9 2 7 4 I 6 5 3 8 9 I 5 2 3 6 4 9 8 7 4 I 2 3 8 7 6 9 5 Block 2 1 2 3 4 5 6 7 8 9 ) 4 2 3 8 5 7 6 9 ) 2 ) 7 6 8 5 3 2 4 9 4 3 8 9 5 4 7 6 ) 2 3 1 9 2 3 6 4 8 7 5 Treatment/varieties Block 3 1 2 3 4 5 6 7 8 9 3 9 5 1 4 3 6 2 7 8 4 5 4 7 8 1 6 9 2 3 2 1 3 8 5 2 4 6 9 7 1 9 6 3 2 4 1 8 7 5 Treatment/varieties 140 Appendix a) 4 Rhizobia cultures isolated from cow pea plant nodules Isolates used to group rhizobia strains Strain group I Number of cultures .,., 11 2 3 11 59 4 5 8 21 6 7 8 3 9 5 2 b) Culture codes 3,6,1 1,12,15,16,19,20,3 1,32,69,70,2 I 1,212,2 I 5,2 I 6,2 I 9,240,247, 248,25 I ,252,261 ,284,286,299,303,304,343,345,346,374 7,8,152, 154,184,186, I 88, I 89,1 90,233,279,347 5,15,16,36,40,109,1 10,1 13,127, I 28, 129, 130,1 3 1,145, I 51,153,165 ,167,168,171, I 74, I 79, 181, I 82, 183,187,223,225,226,228,234,239, 262,263,27 I ,272,275,276,277,278,280,283,289,290,3 I 1,312,323, 324,333,334,335,336,342,343,348,354 39,111,112,133,166,207,208,227 22,24,25,26,27,29,30,38,107,108, 114,220,285,293,329,330,353,3 54,365,366,373 23,132,146 20 I ,202,204,231 ,232,245,246,34 I ,369 218,291,292,300,370 180,224 Isolates used to determine rhizobia acidity and growth rate Strain group I II III IV V VI VII VIII IX Number of cultures 6 11 18 Culture codes 3,6,212,211,247,346 7,8,152,154,186,188,189,190,233,279,347 5,l5,84, 110, 130, 131,151,153,174,179,181,182,228,27 2,280,290,333,348 39, 111,112,133,166,207,208,227 22,24, 25,38,107,285,330,354 23,132 201,202,204,231,232,245,246,341,369 218,291,292,300,370 180,224 8 8 3 9 5 2 Meaning of culture codes Numbers 1-50: M14 treatment one; Numbers 51-88: M14 treatment two; Numbers 145-192: M14 treatment four; Numbers 193-220: Kang'au treatment one; Numbers 221-264: Kang'au treatment two; Numbers 265-304: Kang'au treatment three; Numbers 305-344: Kang'au treatment four; Numbers 345-360: Kang'au April (soils collected before planting); Numbers 361-376: M14 April (soils collected before planting).