Report-DGDC Feb 2008 - V to print - Betuco Annual Report 2006-2007.pdf · livelihoods in central...

FEDERAL PUBLIC SERVICE

FOREIGN AFFAIRS,

FOREIGN TRADE AND

DEVELOPMENT CO-OPERATION

________________________

Directorate General

Development Co-operation

____________________

The Consortium for

Improving Agriculture-based

Livelihoods in Central Africa

(CIALCA)

Progress ReportNovember 2006 –

December 2007

The Consortium for Improving Agriculture-based Livelihoods in Central Africa (CIALCA)

Progress Report November 2006 – December 2007

EXECUTIVE SUMMARY 5

MAJOR FINDINGS 5

1. INTRODUCTION 9

2. BENCHMARK AREA CHARACTERIZATION AND ORGANIZATION 13

3. PROGRESS WITH CHARACTERIZATION ACTIVITIES 16

3.A. CHARACTERIZATION STRATEGY 16 3.B. BASELINE SURVEYS 17 3.C. FINAL CHARACTERIZATION ACTIVITIES 23 3.C.1. DETAILED CHARACTERISATION STUDY ON LEGUME PRODUCTION, MARKETING AND

CONSUMPTION, AND NUTRITIONAL STATUS OF RURAL HOUSEHOLDS 23 3.C.2. DETAILED CHARACTERISATION STUDY ON BANANA PRODUCTION, MARKETING AND

CONSUMPTION 29

4. PROGRESS WITH STRATEGIC SOIL FERTILITY-RELATED ACTIVITIES 32

4.A. RELATIONSHIP BETWEEN SOIL FERTILITY AND NUTRITIONAL QUALITY OF BIO-

FORTIFIED BEANS 32 4.B. ASSESSMENT OF NUTRIENT DEFICIENCIES IN SOILS ON THE WALUNGU AXIS IN

SUD-KIVU 36 4.C. ASSESSMENT OF BANANA – ARBUSCULAR MYCORRHIZAL FUNGI RELATIONSHIPS 39

5. PROGRESS WITH BANANA GERMPLASM-RELATED ACTIVITIES 40

5.A IN-SITU GERMPLASM EVALUATION 40 5.B STRATEGIC RESEARCH AT K.U.LEUVEN 42

6. PROGRESS WITH LEGUME GERMPLASM-RELATED ACTIVITIES 44

6.A. LEGUME GERMPLASM DEMONSTRATION AND EVALUATION 44 6.A.1. ON-STATION LEGUME GERMPLASM EVALUATION 44 6.A.2. LEGUME GERMPLASM EVALUATION WITH FARMER ASSOCIATIONS AT THE

ACTION SITES 47

3

6.B. LEGUME SEED MULTIPLICATION 53 6.B.1. ON-STATION LEGUME SEED MULTIPLICATION 53 6.B.2. FARMER ASSOCIATION-LED LEGUME SEED MULTIPLICATION 54

7. PROGRESS WITH NATURAL RESOURCE MANAGEMENT-RELATED

ACTIVITIES 57

7.A. NATURAL RESOURCE MANAGEMENT OPTIONS FOR LEGUME-BASED SYSTEMS 57 7.A.1. OVERVIEW OF OPTIONS CURRENTLY BEING TESTED 57 7.A.2. SOIL CONSERVATION TECHNOLOGIES TESTED IN SUD-KIVU (“ERO-1” AND “ERO-2) 59 7.A.3. IMPROVED AGRONOMY AND SOIL FERTILITY MANAGEMENT IN

CASSAVA-LEGUME SYSTEMS 62 7.A.4. OPTIONS FOR SOIL FERTILITY AMENDMENT ON THE WALUNGU AXIS IN SUD-KIVU 64 7.B. NATURAL RESOURCE MANAGEMENT OPTIONS FOR BANANA-BASED SYSTEMS 65 7.B.1. ON-FARM TRIALS 65 7.B.2. ON-STATION TRIALS 66 7.B.3. INITIALISATION OF BANANA DISEASE CONTROL STRATEGIES 67

8. PROGRESS WITH MARKET-RELATED ACTIVITIES 68

8.A. BANANA VALUE CHAIN ANALYSIS 68 8.B. LEGUME VALUE CHAIN ANALYSIS 69

9. PROGRESS WITH NUTRITION-RELATED ACTIVITIES 72

9.A. SOYBEAN PROCESSING AND UTILIZATION 72 9.B. INITIALISATION OF BANANA NUTRITION-HEALTH-RELATED ACTIVITIES 74

10. PROGRESS WITH MONITORING AND EVALUATION 75

11. PROGRESS WITH DEGREE-RELATED ACTIVITIES 76

12. ANNEXES 80

ANNEX 1. LOG-FRAME EVALUATIONS 80 ANNEX 1.A. LOG-FRAME OF THE TSBF-CIAT-LED PROJECT 80 ANNEX 1.B. LOG-FRAME OF THE BIOVERSITY-LED PROJECT 83 ANNEX 1.C. LOG-FRAME OF THE IITA-LED PROJECT 85 ANNEX 2: FINAL CHARACTERIZATION TOOL FOR THE LEGUME-BASED SYSTEMS 87 ANNEX 3: LEG-2: LEGUME GERMPLASM DEMONSTRATION TRIALS 104 ANNEX 4: QUESTIONNAIRE USED FOR LEGUME GERMPLASM EVALUATION BY

FARMER ASSOCIATIONS 110 ANNEX 5: ACTION PLAN FOR INFORMAL LEGUME SEED SYSTEMS, 2007B–2008B,

TSBF-CIALCA 115 ANNEX 6: TRAINING OF CIALCA STAFF ON LEGUME SEED SYSTEMS AND SEED

PRODUCTION 117 ANNEX 7: ATELIER DE FORMATION SUR LA MULTIPLICATION DES SEMENCES DES

LEGUMINEUSES 118 ANNEX 8: FICHE FOR DATA COLLECTION IN LEGUME MULTIPLICATION FIELDS 129 ANNEX 9: ERO-1: EROSION CONTROL IN SUD-KIVU 130

4

ANNEX 10: ERO-2: COMPARISON OF VARIOUS FORAGE SPECIES FOR

EROSION CONTROL 141 ANNEX 11: EVALUATION DES FOURRAGES EN MILIEU PAYSAN (ERO-2) 143 ANNEX 12: CAS-1: IMPROVED CASSAVA AGRONOMY 147 ANNEX 13: CAS-2: IMPROVED CASSAVA AGRONOMY 149 ANNEX 14: CAS-3: IMPROVED FERTILITY MANAGEMENT IN CASSAVA SYSTEMS 153 ANNEX 15: PREFERENCES DES ESSAIS D’AMELIORATION DES SYSTEMES

AGRICOLES BASES SUR LE MANIOC PAR LES AGRICULTEURS DE KABAMBA 157 ANNEX 16: FER-1: IDENTIFICATION OF INPUTS REQUIRED FOR SOIL FERTILITY

AMENDMENT 164 ANNEX 17: PROTOCOL FOR ON-STATION AND ON-FARM MULCH TRIALS 167 ANNEX 18: RELEVANT REPORTS, PRESENTATIONS, AND PUBLICATIONS 171

5

EEEXXXEEECCCUUUTTTIIIVVVEEE SSSUUUMMMMMMAAARRRYYY Following a call for proposals of the Directorate General for Development Cooperation (DGDC - Belgium) in April 2004, three proposals were approved:

• ‘Sustainable and Profitable Banana-based Systems for the African Great Lakes Region’, led by the International Institute of Tropical Agriculture (IITA), Kampala, Uganda.

• ‘Enhancing the resilience of agro-ecosystems in Central Africa: a strategy to revitalize agriculture through the integration of natural resource management coupled to resilient germplasm and marketing approaches’, led by the Tropical Soil Biology and Fertility Institute of the International Center for Tropical Agriculture (TSBF-CIAT), Nairobi, Kenya.

• ‘Building Impact Pathways for Improving Livelihoods in Musa-based Systems in Central Africa’, led by the International Network for the Improvement of Banana and Plantain of the International Plant Genetic Resources Institute (INIBAP-IPGRI), Kampala, Uganda.

As the above projects proposed to operate largely in the same parts of Rwanda, Burundi, and the Democratic Republic of Congo (DRC), with similar national partner institutes, and due to the complimentary nature of the activities proposed, above institutes agreed to operate as a Consortium to ensure cooperation and complimentarity and avoid technical and financial duplication at the national level. The Consortium for Improving Agriculture-based Livelihoods in Central Africa (CIALCA) is a Consortium of the International Agricultural Research Centers (IARCs) and their national research and development partners that aims at close technical and administrative collaboration and planning in areas of common interest, thereby enhancing returns to the investments made by DGDC and accelerating impact at the farm level.

This report gives technical details and evaluates progress made in the 3 projects during the period November 2006 – December 2007, and describes some of the strategies and planned activities for 2008. Formal logframe evaluations are included in Annex 1.

MMMAAAJJJOOORRR FFFIIINNNDDDIIINNNGGGSSS

• After completion of the PRA activities and baseline survey, action sites were selected in all mandate areas and currently, activities are on-going with about 40 farmer associations across the different action sites. Activities in satellite sites, in which partner NGOs take the lead, have been initiated with 90 farmer associations.

• Through the baseline survey, implemented with 2,800 households across the 10 mandate areas, detailed information on various livelihood dimensions of rural livelihoods has been obtained. This information covered the areas of household structure and economics, social capital, agriculture, market access and postharvest processing and handling, food and nutrition, food security, and health. Some interesting facts are that (i) most agricultural production is based on relatively small amounts of organic inputs or no inputs at all, (ii) food insecurity is a major problem for over 40% of all households although large differences between mandate areas were observed, (iii) generally, most households have poor access to large regional or urban markets, especially in the more isolated areas, and (iv) social capital is quite extensive in most areas with relatively large proportions of households belonging to farmer, credit and savings, women, self-help, or health insurance groups.

6

• Detailed characterization of banana systems was implemented with 30 farms per action site and covered the thematic areas of production, marketing, and consumption. Productivity varied widely, ranging from 21-63 ton per hectare per cycle. In terms of pest and diseases, Fusarium was widespread while banana Xanthomonas wilt and banana bunchy top virus affected specific mandate areas. Deficiencies of P and Mg were prominent on strongly weathered soils. Lastly, drought stress resulted in yield losses, estimated between 30 and 70%.

• A detailed characterisation study was conducted to complement the earlier baselines survey with quantitative information on aspects of legume cropping, soil fertility status, marketing of legume produce and nutritional status. Following preliminary conclusions were drawn: • Low soil fertility, drought, climatic variability and erosion are the dominant

constraints for crop production. • Legume grain price variability in time and space opens up market opportunities

through storage and transport. • Anthropometric measures identify notable levels of malnutrition in young

children, with 8 – 23% being at risk and 2 – 12% suffering moderate malnutrition. Malnutrition was most pronounced in Sud-Kivu and least in Rwanda.

• An analysis was conducted on variability in Fe and Zn contents in grains of bio-fortified bean varieties. While in some varieties, micronutrient contents vary with environmental conditions (and appears to be related to the organic matter content to the soil), a number of varieties contain high amounts of Fe across environments (e.g., ARA-4), and can therefore be recommended.

• The Walungu area is very unproductive due to low soil fertility constraints. A pot trial was conducted on a large number of soils to assess nutrient deficiencies. Preliminary results identified low P as the major constraint, but symptoms of other deficiencies were observed. Detailed measurements are at present pending.

• A survey was carried out in 188 fields in Rwanda to identify arbuscular mycchorizal fungi (AMF) infection and plant parasitic nematode infection on banana roots. Highly variable infection rates were observed, providing an insight in the role that AMF play in banana production systems and possible benefits for future use of AMF to improve plant health and vigour.

• Related to in-situ conservation of banana germplasm, various varieties obtained through the SMIP project, existing tissue culture labs in the region, and ITC in Leuven, were planted at about 20 sites in the various mandate areas. Measurements on productivity, pest and disease tolerance, profitability, and genotype x environment interactions will be taken. Local macro-propagation facilities have been installed near each germplasm trial for rapid and clean multiplication of the best varieties.

• Strategic banana research at KULeuven, has focused on assessment of droughts stress in Musa cell cultures, cryopreservation, promoter tagging as a basis for developing cisgenic bananas, and understanding arbuscular-mycorrhizal fungal bio-control and its impact on banana productivity and survival. Various activities also were implemented around broadening the banana ITC, managed by KULeuven.

• On-station legume evaluation activities focused on identifying varieties tolerant to low-P conditions, and promiscuous varieties producing high amounts of biomass. Promising varieties were identified but differed between mandate areas. Currently, a selection of varieties is being fully characterised and multiplied to enable homologation and official release in the countries.

7

• Legume germplasm demonstrations and evaluations were conducted in 4 mandate areas, involving a total of 44 farmer associations. Performance differed between sites and mandate areas, and farmers used differing criteria to select preferred varieties. Researcher-defined criteria were explained and taken into account for final selection. In each site, at least 3 bush beans, 2 climbing beans and 2 soybean varieties were retained.

• Informal seed multiplication of farmer-preferred legume varieties was initiated and involves at present both associations in action and satellite sites. About half of the associations have produced sufficient seed quantities to satisfy their own needs and activities have been initiated to promote, disseminate and commercialize the production of improved germplasm seed. Associations are being supported through training, facilitation by NGO partners and involvement of the national seed service.

• A number of technologies targeting the main constraints for legume cropping (principally improved agronomy and nutrient input management) were demonstrated, involving a total number of 56 farmer associations in the mandate areas. Technologies for soil erosion control in Sud-Kivu and rain water harvesting in Rwanda to counteract seasonal drought spells are currently being tested on-station: • Hedgerow planting and reduced tillage are a valid alternative to terrace

construction, and do not negatively affect yields while being relatively effective in soil erosion control in the short term.

• Improved agronomic practices, using modified spacing and high-biomass yielding legumes can significantly improve legume yields in cassava intercropping systems.

• Fertilizer application, preferably applied in combination with organic resources, considerably increases legume yields in different cropping systems and sites.

• Demonstration activities have attracted large interest of farming communities and will proceed into an adaptation phase in 2008.

• Mulch application, manure application, soil and water conservation measures, and general plantation sanitation and husbandry were the themes that were retained by farmers in the context of improved management of banana plantations. About 15 on-farm trials per action site will be installed covering above themes. Parallel to the on-farm trials, on-station trials have been established at 8 sties, focusing on mulch, tillage, and bean intercropping. Specific trails looking at competition between mats for light, water, and nutrients as affected by banana planting density have been installed in 3 sites in Rwanda.

• Initial steps have been taken to start Xanthomonas wilt activities in Rwanda, focusing on screening of banana germplasm, Xanthomonas wilt control options and replanting time. Additional work on systemicity of the bacteria will be conducted in Uganda.

• Market surveys with 400 traders and 150 transporters aiming at detailing banana value chains revealed that most actors deal mainly with beer banana. Various costs along the value chain were quantified and critical market constraints were also identified. Cross-border studies between Rwanda, DR Congo, Burundi, and Uganda were also completed.

• Legume (soybean, bean, and groundnut) value chain analysis has been completed in the Bas-Congo and Sud-Kivu mandate areas. Most farmers did not sell legumes and for those that sell a proportion of their produce, revenues were mostly used to meet urgent needs and not to re-invest in agriculture. Prices were strongly

8

influenced by periods of abundance and shortage. Most traders in the local markets were women.

• A strategy to train farmers and trainers in soybean processing and utilization has been completed and training materials that accompany those training events have been produced. At least 10 25-are demonstration gardens for soybean production have been installed near health centers at the action sites. It is expected that in first instance 200 trainers and 400 farmers will be trained. A strategy has also been developed for the trainers to continue training individual farmers. Planning for banana nutrition-related activities has also been completed.

• A detailed M&E framework has been put in place to monitor and evaluate progress with project implementation and project interventions. The baseline survey will serve as a basis for evaluation of initial project-related impact to be assessed towards the end of the first phase of CIALCA.

• In terms of degree-related training, CIALCA is currently engaging 8 PhD students, 13 MSc students, and 13 undergraduates. Four post-doctoral fellows are leading specific project activities. Various training sessions with farmer associations, for instance, on seed multiplication and participatory evaluation of technologies, have also been conducted. On-the-job training of national system scientists has resulted in substantial improvements in the technical capacity of those partners.

9

111... IIINNNTTTRRROOODDDUUUCCCTTTIIIOOONNN Following a call for proposals of the Directorate General for Development Cooperation (DGDC - Belgium) in April 2004, focusing on Central Africa, three proposals were approved:

• ‘Sustainable and Profitable Banana-based Systems for the African Great Lakes Region’, led by the International Institute of Tropical Agriculture (IITA), Kampala, Uganda.

• ‘Enhancing the resilience of agro-ecosystems in Central Africa: a strategy to revitalize agriculture through the integration of natural resource management coupled to resilient germplasm and marketing approaches’, led by the Tropical Soil Biology and Fertility Institute of the International Center for Tropical Agriculture (TSBF-CIAT), Nairobi, Kenya.

• ‘Building Impact Pathways for Improving Livelihoods in Musa-based Systems in Central Africa’, led by Bioversity International, Kampala, Uganda.

The purpose of the project led by IITA is to develop and disseminate in partnerships with all stakeholders technologies that improve the sustainability and profitability of banana-based cropping systems. Emphasis is put on identifying and exploring markets as a driving force for changing banana-based farming systems. Technologies promoted include amongst others locally adapted natural resource management options (including integration of legumes), integrated pest management options, the introduction of new banana hybrids, and improved post-harvest technologies. The project emphasizes strong partnerships and capacity building with NARS, Universities, non-governmental organizations (NGO), community-based organizations (CBO), and the private sector. The project will also put emphasis on strategic research on sustainable use of the natural resource base through collaboration with UCL.

The purpose of the Bioversity-led project is to strengthen national and regional mechanisms to plan and orient investments, projects and research for development synergies by increasing the contribution of Musa to rural well-being. The project will also strengthen national frameworks for conserving local Musa germplasm, introducing and evaluating new cultivars and multiplying and disseminating clean planting material of superior cultivars. An important part of the project will be to support the global collection of Musa germplasm and research on stress responses on banana, both occurring in Belgium. The project further aims to identify, with scientists, extension agencies, NGOs, and farmers, market opportunities for bananas and banana products, to validate options for integrated pest and soil fertility management, and to develop improved Musa production systems. The project emphasizes strong partnerships and capacity building with NARS, universities, NGOs, CBOs, and the private sector. In the INIBAP-IPGRI-led project, strategic research backing is given by K.U.Leuven.

The purpose of the project led by TSBF-CIAT is to develop and disseminate in partnerships with all stakeholders resilient agro-ecosystems through integration of stress-tolerant and bio-fortified germplasm, inclusion of locally adapted natural resource management (NRM) options, market-led diversification and intensification, and revitalisation of research for development capacity of all stakeholders. The main entry points are multi-purpose legumes that will address issues related to declining soil fertility, low income, and food insecurity and malnutrition which are major constraints to improved rural livelihoods in the target areas. The project aims at integrating strategic, applied, and adaptive research for development with strong involvement of various partners with

10

expertise in all the above sectors. In the TSBF-CIAT-led project, strategic research backing is given by K.U.Leuven.

As the above projects proposed to operate largely in the same parts of Rwanda, Burundi, and the Democratic Republic of Congo (DRC), with similar national partner institutes, and due to the complimentary nature of the activities proposed, above institutes agreed to operate as a Consortium to ensure cooperation and complimentarity and avoid technical and financial duplication at the national level. The Consortium for Improving Agriculture-based Livelihoods in Central Africa (CIALCA) is a Consortium of the International Agricultural Research Centers (IARCs) and their national research and development partners that aims at close technical and administrative collaboration and planning in areas of common interest, thereby enhancing returns to the investments made by DGDC and accelerating impact at the farm level. The overall goal of CIALCA is to facilitate all above interactions in order to obtain integration between the activities of each project to the extent that this is desirable. It is not to create a ‘super-project’ directing all activities within each of the three projects.

In order to operationalize CIALCA, various initiatives were taken both at the administrative and at the technical level. At the administrative level, a Memorandum of Understanding between IITA, Bioversity, and CIAT has been approved; a CIALCA Consultative Committee (CCC) has been constituted; CIALCA offices or representations have been installed and CIALCA facilitators or representatives engaged; and communication and reporting channels have been formalized. At the technical level, various annual planning meetings have been organized (Photograph 1) and a CIALCA website was set up (WWW.CIALCA.ORG) (Photograph 2). The CIALCA website is a knowledge platform for CIALCA reports, newsletters, publications, student theses, but also contains info on partner organisations and has links to other interesting websites. It is envisaged to upload most of the available regional agricultural information on the CIALCA website.

The coordinates of the CIALCA offices are:

• Burundi: Mr. Sylvestre Hakizimana, IRAZ, PO Box 91, Gitega, Burundi, Tel: (+257) 403020/21, Mobile: (+257) 903315, email: [email protected] or [email protected].

• Rwanda: Mrs. Kantengwa Speciose, c/o CIAT Rwanda, Kacyiru, Boulevard the l'Umuganda, Concorde building, 1st floor, Kigali, Tel:(+250) 55 104708 or 08518471, email: [email protected].

Photograph 1: Participants at an annual CIALCA planning meeting.

Photograph 2: Homepage of the CIALCA website.

11

• DRC – Sud-Kivu: Mr Dieudonné Katunga Musale, Coordinator, 6 Av. Kasongo, Commune d'Ibanda, Bukavu, Eastern D.R.Congo, Tel:(+243) 98 669793, email: [email protected].

• DRC – Bas-Congo: Mr Jean-Paul Lodi Lama, c/o INERA office, 13 Avenue des Cliniques, Kinshasa-Gombe, B.P.2037 Kinshasa 1, Tél: (+243) 815136746. email: [email protected] or [email protected].

Various benefits have been observed while operating as CIALCA in the target areas. At the administrative level, (i) although no funds were originally budgeted for setting up offices in various locations, due to the combined effort of the three projects, sufficient funds were identified to have such offices installed, which has proven to be an essential component in implementing project activities; (ii) managing CIALCA has been very efficient because of a clear distribution of tasks between the three projects; and (iii) participatory rural appraisals and baseline surveys were implemented jointly by the three projects which has resulted in a very cost-effective way to obtain the data and in tools that are richer in terms of covering a wider range of topics important for rural livelihoods in the mandate areas.

In terms of partnerships, (i) CIALCA partners have been actively using facilities of the CIALCA offices (e.g., meeting venue, internet access) and (ii) visibility of CIALCA in the region has been relatively high since the various partners openly identify themselves with the Consortium. Partners involved in CIALCA activities are summarized as follows:

• Belgian institutes: Katholieke Universiteit Leuven (K U Leuven), Université Catholique de Louvain-la-Neuve (UCL), Faculté Universitaire des Sciences Agronomiques de Gembloux (FUSAGx)

• Non-governmental organizations (NGOs) & extension: DRC: Diobass; Bureau Diocésaine de Développement (BDD), Association pour la Promotion de la Démocratie et du Développement de la République Démocratique du Congo (APRODEC); Burundi: Catholic Relief Services (CRS); Rwanda: Rwanda Rural Rehabilitation Initiative (RWARRI), Rwanda Development Organisation (RDO)

• National agricultural research institutes (NARS): DRC: Institut National des Etudes et de la Recherche Agricole (INERA), Centre de Recherche des Sciences Naturels (CRSN); Rwanda: Institut des Sciences Agronomiques de Rwanda (ISAR): Burundi: Institut des Sciences Agronomiques du Burundi (ISABU), Institut de Recherche Agronomiques et Zootechnique (IRAZ).

• Regional networks: BARNESA, AFNET, FOODNET, ECABREN

• National universities: DRC: Université Catholique de Bukavu (UCB), Université Catholique de Graben (UCG), Université de Kinshasa (UNIKIN); Rwanda: Université Nationale du Rwanda (UNR), Burundi: Université du Burundi (UB), other universities in East and Southern Africa.

• Private sector: DRC: Gourmet Gardens; Burundi: Agro-biotech, Phytolab.

• Health partners: Rwanda: Rwandese Health Environment Project Initiative (RHEPI); DRC: Centre Olame; DRC, Burundi: Healthnet-TPO; DRC, Rwanda, Burundi: Various health and nutrition centers, Ministries of Health.

• Farmer groups (FGs): DRC, Rwanda, Burundi: Various community-based organizations (CBOs).

At the technical level, (i) the different projects have been leading activities that are relevant for all three projects, depending on their in-house capacity; (ii) due to the wide thematic coverage of the three projects as a whole, CIALCA activities cover all most important realms of rural livelihoods and all major components of the farming systems in the

12

mandate areas; (iii) various MSc or PhD-related activities are co-supervised by colleagues who have research links to several of the CIALCA projects; and (iv) within the banana research group, there are strong complementarities in skills, which is optimally exploited by designing collaborative research activities. Last but not least, some extra proposals have been accepted to strengthen CIALCA activities. Examples are (i) the project on ‘Mobilizing Innovation Platforms for Bringing More Quality Benefits to More People in Post-Conflict Central African Great Lakes Region’, supported by the CSO-CGIAR Competitive Grants Program, (ii) the project on ‘Amélioration de la productivité agricole en incitant l’utilisation efficace et rendable des intrants inorganiques dans le cadre de la gestion intégrée de la fertilité de sol dans la province de Sud-Kivu au République Démocratique du Congo’, supported by VLIR, Belgium, and (iii) a project on ‘Banana macro-propagation technology for rapid multiplication of improved banana cultivars’, supported by BTC through ISAR (i.e. technology introduced by IITA-CIALCA, project written by IITA-CIALCA scientist, training provided by IITA-CIALCA staff).

This report describes progress with CIALCA activities between November 2006 and December 2007, largely covering the second year of the 3-year project, and is a follow up on the first progress report, covering the period September 2005 – October 2006. As for the former project, based on advice from DGDC, the current project covers the various activities implemented by the three projects, constituting CIALCA.

13

222... BBBEEENNNCCCHHHMMMAAARRRKKK AAARRREEEAAA CCCHHHAAARRRAAACCCTTTEEERRRIIIZZZAAATTTIIIOOONNN

AAANNNDDD OOORRRGGGAAANNNIIIZZZAAATTTIIIOOONNN Based on the project proposals, several mandate areas have been identified for focussing activities of the projects (Figure 1). Mandate areas are the geographic boundaries within which the project will operate and criteria for choosing these were: (i) high levels of poverty and environmental degradation, associated with low food and nutrition security, (ii) high potential for productivity increases, (iii) good potential access to local and regional markets, and (iv) existence of active agricultural development networks. Furthermore, in the mandate areas, (v) the major cropping systems are based on Musa or plantain (Musa), cassava, and legumes, which are mandate crops for the 3 institutes constituting CIALCA-II, (vi) mid-season drought is occurring more frequently, induced by climate change, (vii) social and biophysical heterogeneity is substantial, and (viii) recent civil strife has reduced research-for-development capacity and service and market infrastructure. Mandate areas are defined as areas with similar agro-ecological conditions and poverty profiles that have nonetheless relatively good access to large urban markets and where the target cropping systems are important components of smallholder farmers. Mandate areas represent large areas (i.e. hundreds to thousands of square kilometers), often corresponding to provinces, e.g. Kibungo (Rwanda), Gitega (Burundi), South Kivu highlands (DRC) and the number of people living in each mandate area can vary between 100,000 and 1,000,000. All the mandate areas have been characterized in terms of population density, altitude, agro-ecological potential, and access to markets (www.cialca.org).

Within each mandate area, action sites and satellite sites are identified based on the relative access to markets (Table 1). Action sites are geographical areas encompassing a community or a limited cluster of communities in each mandate area in which the field activities related to technology identification, evaluation, and adaptation will take place in partnership with development agencies. Action sites are selected to reflect contrasts in

Figure 1: CIALCA mandate areas in DRC, Rwanda, and Burundi (www.cialca.org).

14

specific key variables, presumed to substantially influence the nature of best-bet technologies and their mode of dissemination. The number of people living in each action site can vary between 500 and 5000. Satellite sites are similar in terms of geographical area, population, and other general characteristics as action sites but more numerous. These sites will be used to evaluate best-bet options, developed in the action sites, under leadership of associate partners. The selection of satellite sites is an on-going process and . NGO partners will especially be implicated to select suitable satellite sites within each mandate area.

Most NRM development, testing, and adaptation work will take place in the action sites with the satellite sites serving as a means to scale out project products. Within each action site, baseline information has been collected on farmer typologies, within-farm soil fertility gradients, farming systems, post harvest value addition activities and potentials, markets, social structures, nutrition and health status, indigenous coping mechanisms, the current contributions of legumes and bananas to human and livestock nutrition and cropping system productivity, etc. These data will provide a framework for impact assessment later in the project. Another important aspect of these activities has been the identification of active farmer groups that will lead all evaluation activities, as the current project works mainly with communities rather than individual farmers.

15

Table 1: Summary of the active Action and Satellite sites, and associations engaged in the different Mandate Areas.

Mandate area Action sites Satellite sites Area Number of associations Area Number of associations

Lemfu 3 (ADERKI, APEKI, ADPN) Lemfu 30 (managed by BDD-Kisantu, in ‘groupements’ Kisantu, Kiyanika and Ngufu)

Kanga-Kipeti 2 (ADEKO, ACKI) Muala-Nzundu 2 (managed by APRODEC) Mbanza-Nzundu 2 (ACDPP, APDKI) Kinkewa 1 (managed by APRODEC) Zenga 3 (CALDZ, ADN, AFEPA) Kiwembo 1 (managed by APRODEC) Bovin 1 (managed by APRODEC) Zenga 4 (managed by CLD, in ‘groupements’ Nkolo,

Nkolo-Tava and Makuta) Vunda 4 (managed by CLD) Sadi 12 (managed by CLD)

Bas-Congo

Tumba 5 (managed by BDD-Matadi) Kabamba 3 (Tuungane, Maendeleo, ADEPB) Kavumu 1 (managed by DIOBASS) Luhihi 3 (ATM, Rhusimane, Rhubehaguma) Lurhala 2 (APACOV, CINAMULA)

Sud-Kivu

Mwegerera 4 (ALEMALU, Abagwasinye, Rhuchihane, Bololoke) Murambi 2 (Dufashanye, Iriba) Nyakigando 30 individual households, involved through ISAR Rugarama 3 (Twisungane, Imbaraga, Giribakwe) Kabarore 2 (Isoko y’ubumwe, Abahujumugambi)

Umutara

Nyakigando 2 (Ingandurarugo, Dufatanye) Kabare 2 (Duterimbere A, Duterimbere B) Gatore 2 (Benyshiaka, Dutabarante) Musenyi 2 (Turwanyinzara, Turwanyubwigunge)

Kigali-Kibungo

Mayange 2 (Tubanenabose, Abiwguruye)

16

333... PPPRRROOOGGGRRREEESSSSSS WWWIIITTTHHH CCCHHHAAARRRAAACCCTTTEEERRRIIIZZZAAATTTIIIOOONNN

AAACCCTTTIIIVVVIIITTTIIIEEESSS

333...AAA... CCCHHHAAARRRAAACCCTTTEEERRRIIIZZZAAATTTIIIOOONNN SSSTTTRRRAAATTTEEEGGGYYY Three characterization-related activities were originally planned: (i) participatory rural appraisals (PRA), (ii) a baseline survey, and (iii) specific follow-up or detailed characterization studies (Figure 2). As explained above, the PRAs and baseline survey were implemented as CIALCA, through common planning and cost-sharing between the three projects. After having delineated the mandate areas (see above), PRAs were held in about 8 ‘villages’ per Mandate area, thereby ensuring that villages with relatively good and relatively poor access to markets were included. ‘Villages’ corresponded with political units containing about 500 households and were named differently in different region (the ‘village’ equivalent is underlined):

• Sud-Kivu: province - territoire - chefferie/collectivité - groupement - localité (chef, 500-1000 HH) - village

• Bas-Congo: province - district - territoire - secteur - groupement - village (chef, 500-1000 HH)

• Rwanda: province - district - secteur (chef, 500 HH) - cellule - nyumba kumi

• Burundi: province - commune - zone - colline (chef, 500 HH) - secteur

Based on the information obtained through the PRAs, Action Sites were identified following a set of specific criteria, including the presence of active farmer groups, accessibility, etc. A baseline survey was then implemented with about 2800 households across all Action Sites identified, focusing on aspects of livelihoods, markets, nutrition, bananas, and legumes. The baseline survey also aimed at collecting information to enable constructing farmer typologies, based on the presence of specific production units or access to resources (e.g., land, labour, capital, knowledge). As a last step in the characterization work, detailed characterization studies have been implemented to get all required information related to the specific themes of the individual projects. The Bioversity and IITA projects

Baseline study PRA/Specific studies

X Baseline

implementation

XDraft tools

(incl Sampling)

X Training/testing

X Final tools

X Team identification

XData entry

X PRA checklist

X PRA’s/confirm sites

Selected

Villages

XCommunity

selection

X Specific studies(markets, health,soil, legumes,bananas)

X Tools for

spec studies

X Training/testingTypologies

Field activities

XData analysis

X Baseline report X PM&E with

communities

X Training/testing

X Teams

X Teams

Figure 2: Strategy for the selection and characterization of the Action Sites.

17

have focused on banana production systems and access to markets while the TSBF-CIAT project have focused on legume production and current contribution to rural livelihoods, market access, and nutritional status of the rural population. More details related to the baseline survey, and the specific studies are given in the sections below.

333...BBB... BBBAAASSSEEELLLIIINNNEEE SSSUUURRRVVVEEEYYYSSS The baseline surveys are a follow-up and in-depth study of the above described PRA exercise. The baseline surveys provide counterfactual data for ex-post impact assessment and information for priority setting and technology assessment. The data cover the following fields: household systems and socio-economic structures, farming system agronomics and economics, access to markets and marketing patterns for the focus crops, post harvest handling and processing of the focus crops, social structure of households and households’ embedding in social structures within the sites, status and determinants of food security, and health and nutritional status. The baseline survey was implemented in 30 villages across all mandate areas and involved about 2800 households (Table 2).

Table 2: Baseline sites and sample size.

Country Action sites Villages Households/village Total

DR Congo Sud-Kivu montagneux 5 100 500 Nord-Kivu montagneux 4 100 400 Bas-Congo 4 100 400 Rwanda Umutara 4 50 200 Kigali-Kibungo 4 100 400 Gitarama 2 100 200 Kibuye-Gisenyi 2 100 200 Burundi Gitega 2 100 200 Kirundo 2 100 200 Ruzizi plains 1 100 100 TOTAL 30 2,800

The baseline data has been entered into the SPSS software package. Copies of the database containing all data have been given to all national partners. The major findings will be published in a book in the course of 2008. A draft report summarizing the major findings is already available, and some highlights and examples of these findings are provided below for the major themes covered by the baseline survey: • Household systems and social and economic structures, including assets and financial

resources

• Farming systems agronomics and economics

• Access to markets and marketing patterns for the focus crops (banana, legumes, cassava)

• Post harvest handling and processing of the focus crops

• Social structure of households and households’ embeddedness in social structures

• Status and determinants of food security, including food consumption patterns and nutrition

• Health (including access to and utilization of health services) Household structure and economics Some key household characteristics are shown in Table 3. A relatively high percentage of households have attained at least primary level of education in all the mandate areas. However Sud-Kivu showed only 28.4% of the households having attained primary education. Literacy level was also substantially lower in Sud Kivu (67.2) compared to Rwanda East and Ouest (>94.5). The high percentage of respondents (i.e. mostly HH

18

head or spouse) having no formal education in areas such as Cibitoke (53.2%) and Sud Kivu (47.4%) is likely to present some bottlenecks in technology uptake in these areas.

Table 3: Household characteristics

Burundi DR Congo Rwanda

Cibitoke

Gitega

Kirundo

Bas-Congo

Nord-Kivu

Sud-Kivu

Est

Ouest

Sud

Total

HH head gender (%) Male 78.2 80.6 92.1 82.7 85.2 84.8 80.5 80.9 82.0 83 Female 21.8 19.4 7.9 17.3 14.8 15.2 19.5 19.1 18.0 17 Level of education (%) No formal education 53.2 36.0 39.8 13.8 39.9 47.4 28.6 20.2 14.9 32 Adult alphabetization 2.1 10.8 7.6 1.0 0.3 5.5 5.8 4.0 5.9 4.3 Primary education 36.2 45.2 40.7 32.2 45.7 28.4 53.6 64.6 73.3 44 Secondary education 7.4 2.7 5.9 40.6 13.8 17.8 10.8 10.1 3.0 16 Other 1.1 5.4 5.9 12.3 0.3 0.9 1.2 1.0 3.0 3 Literacy of respondent (%) Literate 82.6 82.8 83.7 81.3 78.0 67.2 94.5 97.3 85.7 60 Non Literate 17.4 17.2 16.3 18.7 22.0 32.8 5.5 2.7 14.3 40 Age of respondent 45.0 44.5 40.2 42.8 41.1 45.2 42.5 43.0 39.9 43 Household size 5.5 6.6 6.2 5.6 6.4 7.0 5.8 6.4 6.3 6

The average number of people with in each household is about 6 persons however, Sud-Kivu shows a higher household size of about 7 persons while Cibitoke showed the lowest of about 5 persons. Overall, the average age of the household head is about 43 years which indicates that most of the household heads are adults that can take informed household decisions. However, household members under the age of 18 take up the largest composition (45.1-50.3%) of household members at all sites. The dependency ratio which represents the economic burden for resources imposed on the working population was computed as follows: Dependency ratio equals the number of individuals aged below 18 or above 59 years, divided by number of individuals aged 18 – 59 years. The results indicate that there is a high dependency ratio in all CIALCA sites; there are more people who are not of working age and that need to be looked after. This was especially the case in the provinces of Kirundo, and Nord & Sud-Kivu which have a dependency ratio of 1.17, 1.27 and 1.34, respectively. Social capital Membership to a professional organization is of great importance in that households benefit from these associations through collective marketing, credit access, sharing of new ideas and experiences, and a high level of output and return on goods sold through better bargaining power. Figure 3 highlights farmer membership to an association in Burundi, Congo and Rwanda. The results reported indicate higher percentages (about 70%) reported in Rwanda provinces for respondents who belonged to a health insurance group.

19

Other countries however showed minimal or no involvement in a health insurance group. The East province in Rwanda indicated that about 41% of respondents belonged to a self help group. Higher proportions of 42% and 45% of respondents in Ouest and Sud provinces respectively belonged to a credit and savings group. Fewer proportions of respondents belonged to a women’s group.

Agriculture Table 4 shows the average household land holding and land holding by land type and province. Results show that households in Bas-Congo have a larger farm size of 2.12 hectares followed by Cibitoke at 1.97 hectares. Households in Gitega have the smallest farm size. Households in Cibitoke had the largest farm size for field on hill (0.96ha) and homestead area of (0.45ha). Households in Bas-Congo had the largest farm size for field under marsh (0.40ha) and forested area (0.44ha) accounting for 18.9% and 20.8% of total farm size respectively. Area under grazing fields was reported to be largest in Nord-kivu (0.60ha). Cropped area that is, homestead area, field under marsh and field on hill account for the biggest proportion. These figures show that land pressure is very high, especially in Gitega, Kirundo, Sud Kivu and Ouest Rwanda.

Table 4: Household land

Land type Homestead

area Field under marsh

Field on hill

Forested area

Grazing field

average farm size (ha)

Burundi Cibitoke 0.45 0.06 0.96 0.36 0.14 1.97 Gitega 0.10 0.04 0.09 0.05 0.22 0.50 Kirundo 0.20 0.04 0.24 0.06 0.37 0.91 Congo Bas-Congo 0.36 0.40 0.53 0.44 0.39 2.12 Nord-Kivu 0.19 0.34 0.50 0.19 0.60 1.82 Sud-Kivu 0.3 0.20 0.34 0.09 0.20 1.13 Rwanda Est 0.25 0.17 0.62 0.12 0.26 1.47 Ouest 0.27 0.12 0.44 0.15 0.09 1.07 Sud 0.33 0.16 0.95 0.10 0.40 1.94

Figure 3: Household membership of social groups

20

Table 5: Types of organic inputs applied in homestead plots

% of HH Burundi DR Congo Rwanda Cibi-

toke Gite-ga

Kirun-do

Bas-Congo

Nord-Kivu

Sud- Kivu

Est Ouest Sud Total

Nothing 97.3 58.3 90.2 97.0 76.0 44.0 77.1 64.3 54.9 67.0 Manure 0 1.4 4.9 0 0 0 0 0 0 0.3 Compost 2.7 38.9 4.9 3.0 4.1 42.7 21.3 31.8 41.2 26.1 Biomass 0 1.4 0 0 15.7 9.2 1.4 3.9 3.9 5.1 Other 0 0 0 0 4.1 4.1 0.3 0 0 1.6

Homestead plots in areas with very high land pressure, such as Gitega, South Kivu, and Rwanda Ouest, generally received more often (>35% of farmers) organic nutrient inputs than plots in areas with less high land pressure such as Bas Congo and Cibitoke (<3%).

Manure is seldom applied in a pure form, but is often mixed into compost or left unused near the kraal (Table 5). The availability of manure is closely linked to the number of cattle owned. Livstock numbers are generally low at all sites, with some 15% of farmers owning cattle on average. However, cattle ownership was particularly low in Gitega, Cibitoke and DRC, where 10% or less of the farmers owned cattle (Figure 4). These findings highlight the need for CIALCA to look for biomass-related methods to improve nutrient recycling and availability to improve crop production. Table 6: Value of output by crop grown per season (USD)

Burundi DR Congo Rwanda Cibi-

toke Gite-ga

Kirun-do

Bas- Congo

Nord- Kivu

Sud- Kivu

Est Ouest Sud

Banana beer 40.5 13.2 17.6 0 9.4 14.0 13.8 41.4 12.1 Dessert banana 0 1.9 11.0 9.0 2.7 4.3 8.7 18.4 24.8 Cooking banana 57.2 13.2 17.6 4.9 12.6 5.2 23.0 70.0 23.0 Plantain banana 0 0 0 17.9 10.7 4.5 0 0 0 Bitter cassava 18.6 5.2 11.6 43.7 14.3 22.6 1.6 1.1 2.8 Sweet cassava 0 29.0 0 35.8 1.4 35.8 0.7 1.5 1.0 Cassava leaves 0 1.9 8.6 5.4 1.1 0 0.2 0 0 Fresh beans 0 0 0 0.4 0 5.4 0.6 19.3 0 Grain beans 17.6 1.3 4.2 26.2 10.7 10.0 0.6 0.8 1.0 Bean leaves 0 0 0 0 0 17.9 0.6 0 0 Green beans 0 0 0 0 0 53.7 2.0 2.8 1.3 Ground nuts 0 8.4 3.2 15.2 10.7 15.2 2.4 0 22.1 Soybean 0 0 0 8.6 3.6 9.7 2.3 0.5 9.4 Cowpea 0 6.6 0 1.6 0 0 0 0 0 Cowpea leaves 0 0 0 0 0 0 1.8 0 0

Figure 4: Cattle ownership

21

In terms of cropping systems and the value of the crops produced, the baseline confirmed that banana and legumes are amongst the dominant crops at all sites (Table 6). Bananas are the primary crop in terms of output value in Cibitoke, Kirundo, Nord Kivu, and all Rwandan sites. Legumes are very important in all sites, and cassava is particularly important in Gitega, Bas Congo, and South Kivu. Market access, agricultural marketing, post harvest processing and handling Markets outlets that are dominantly used by farmers to sell their products are the farm gate and the local markets (Table 7). Urban and big regional markets are still important for a substantial portion of the farmers (>20%) in areas that are relatively close to big cities, such as Cibitoke (Bujumbura), Bas-Congo (Kinshasa), South Kivu (Bukavu), and Rwanda Ouest (Kigali). However, in general most households have poor access to the big regional or urban markets, and this was particularly true for the more isolated areas such as Kirundo, which have no big city or big regional market nearby.

Table 7: Most dominant market outlet used to sell farm produce.


toke Gitega Kirun

-do Bas-Congo

Nord-Kivu

Sud- Kivu

Est Ouest Sud

Total

On farm 7.6 17.0 18.9 31.1 34.8 22.4 21.5 22.2 13.5 24.5 Local market 37.5 47.9 63.3 31.4 21.4 30.3 44.4 25.1 53.4 35.0 Neighbor market 27.8 22.9 15.6 14.0 25.6 23.3 20.1 26.7 23.6 21.5 Urban market 16.7 12.2 2.2 19.4 12.5 5.7 6.9 17.3 4.1 11.3 Regional market 10.4 0 0 4.2 5.7 18.2 7.1 8.6 5.4 7.7

Food and nutrition, food security Food insecurity is a major problem for over 40% of all households in the CIALCA sites (Table 8). However, large differences exist, and areas with particularly high levels of food insecurity (>60% food insecure) are Gitega and Kirundo in Burundi and South Kivu in DR Congo. Female headed households are in general more food insecure than male headed households. These are also the areas with very high land pressure and relatively little land available per household.

Table 8: Percentage of male and female headed households that are food-insecure for at least part of the year.


toke Gitega Kirun-

do Bas-Congo

Nord-Kivu

Sud- Kivu

Est Ouest Sud

Total

Male headed 34.0 61.2 62.7 47.5 8.8 56.4 33.3 28.2 47.4 40.0 Female headed 46.5 67.0 65.1 58.6 11.6 67.8 36.1 48.3 40.9 46.2

Health We recorded the illnesses affecting children under 5 years (Table 9) and adults over 18 years of age. Illnesses can be a major set back to the household’s progress towards low mortality rates, food security and poverty reduction. Among the illnesses that were mostly afflicting the children and adults were malaria/headache. Higher proportions of above 80% were reported for adults over 18 years faced with malaria in the areas of Cibitoke, Gitega, Kirundo, Est, Ouest and Sud compared to 45% in Bas-Congo which was the lowest. The highest proportion among children less than 5 years faced with malaria was reported in Rwanda Est and Ouest (78%) and the lowest (38%) was recorded in Nord-Kivu. Diarrhea for children under 5 was particularly important (>12% of HH) in Burundi and Nord Kivu. The data suggests that health, food security, and the available land for

22

each household are strongly related; e.g., areas such as Gitega, Kirundo, and South Kivu are amongst the highest in terms of diseases recorded, while also having the highest levels of food insecurity and the smallest average land holdings.

Table 9: Percentage of households that recorded illnesses for children under 5 years


toke Gitega

Kirun-do

Bas-Congo

Nord-Kivu

Sud- Kivu

Est Ouest Sud

Total

fever/flu 3.9 19.0 10.5 29.9 3.5 6.8 3.2 0.7 5.3 9.8 vomiting 0 0 0 0 0 3.1 1.4 0.7 0 0.9 worms 2.0 0 3.5 1.9 0 8.6 8.8 12.8 13.3 5.7 malaria/headaches 68.6 43.1 68.6 47.5 37.8 48.8 77.9 78.0 69.3 58.0 cough/bronchitis/tbc 0 0 0 10.2 1.8 18.5 3.7 3.5 1.3 6.5 stomach 0 0 0 1.3 0.4 2.2 0 0 0 0.7 diarrhea 13.7 26.7 12.8 1.9 12.0 4.0 1.6 2.1 2.7 6.3 others 11.8 10.3 3.5 2.9 44.5 7.1 3.0 2.1 8.0 11.0 measles 0 0 0 4.5 0 0 0 0 0 0.8

The way forward with the baseline study Although a draft report with results of the baseline survey is available, CIALCA would like to publish a book with the major findings and descriptives for this study. This book should be printed and released in the course of 2008. In addition, we will continue to further explore the data with in-depth studies on the following topics:

• Production systems of the focal crops

• Marketing systems of the focal crops, in particular bananas

• Potential for processing and value adding, in particular for bananas

• Sociology, nutrition and health studies

23

333...CCC... FFFIIINNNAAALLL CCCHHHAAARRRAAACCCTTTEEERRRIIIZZZAAATTTIIIOOONNN AAACCCTTTIIIVVVIIITTTIIIEEESSS

33..CC..11.. DDEETTAAIILLEEDD CCHHAARRAACCTTEERRIISSAATTIIOONN SSTTUUDDYY OONN LLEEGGUUMMEE

PPRROODDUUCCTTIIOONN,, MMAARRKKEETTIINNGG AANNDD CCOONNSSUUMMPPTTIIOONN,, AANNDD

NNUUTTRRIITTIIOONNAALL SSTTAATTUUSS OOFF RRUURRAALL HHOOUUSSEEHHOOLLDDSS A detailed characterisation study was conducted during June and July 2007 in the mandate areas of the TSBF-CIAT project. Prior to the implementation of the study, a two-day training and discussion session was organized at the CIALCA office in Kigali on 24-25 May 2007, involving agronomists, socio-economists and nutritionists from the various regions. Local training sessions were then organized with poll-taker teams of the various disciplines, followed by trial runs with a number of households in the field (Photograph 3).

This study complemented the earlier conducted baseline survey with quantitative information on aspects of legume cropping, soil fertility status, marketing of legume products and nutritional status of rural livelihoods. Households were randomly selected within the set of households interviewed during the baseline survey. Between 15 and 20 households in each of the 4 action sites in all mandate areas were fully characterized; nutritional status was evaluated in twice as many households. The characterisation study entailed detailed questionnaires with farmers, soil and plant sampling, agronomic measurements in legume-grown fields, collection of essential socio-economic data for market chain analysis, anthropometric measurements in children between 2-5 years old, and an assessment of dietary intake and diversity. The questionnaire used is presented in Annex 2. Data entry has currently been concluded and some preliminary analyses have been conducted (presented below). In depth examination will involve factor and multivariate analysis. Soil and plant sample analysis are at this time pending. Legume production In each household, a map of the farm was drawn by the household head, indicating the location of the manure/compost storage system, livestock facilities and the different fields, relative to the homestead. The farmer was asked to specify the crops grown in the past two seasons. All fields cultivated by legumes were than highlighted and visited by an agronomist and a household member (see example in Figure 5). Detailed measurements were taken, including soil sampling, and measurement of crop and weed densities. Farmers gave details on crop management and inputs applied, appraised the soil fertility according to the local classification system and indicated the major constraints for crop production in the fields. Finally, compost and manure facilities were sampled for determination of organic matter quality and nutrient contents, and legume grain samples were taken for analysis of nutritional quality. Analysis results are pending.

Photograph 3: Field testing of the survey with chief enumerators.

24

80m

2007 A

beans

2007 B

sorghum

cocoyam

beans

P1 (civu)

owned

P7

P2 (civu)

owned

120m

P3 (civu)

owned

160m

P5

SK/02/54

10km

P6 (civu)

owned

400m

320m

P4 (civu)

owned

livestock

2 cows, grazing, with manure collection

7 goats, tethered, daily manure collection

2007 A

beans

2007 B

beans

maize

cassava

2007 A

soybean

sugarcane

2007 B

beans

sorghum

sugarcane

2007 A

banana

soybean

beans

2007 B

banana

beans

2007 A

cassava

2007 B

cassava

2007 A

banana

2007 B

banana

beans

organic inputs

compost (FYM, crop residues, green

manure, HH waste)

2007 A

banana

2007 B

banana

80m

2007 A

beans

2007 A

beans

2007 B

sorghum

cocoyam

beans

2007 B

sorghum

cocoyam

beans

P1 (civu)

owned

P7

P2 (civu)

owned

120m

P3 (civu)

owned

160m

P5

SK/02/54SK/02/54

10km

P6 (civu)

owned

400m

320m

P4 (civu)

owned

livestock



livestock



2007 A

beans

2007 A

beans

2007 B

beans

maize

cassava

2007 B

beans

maize

cassava

2007 A

soybean

sugarcane

2007 A

soybean

sugarcane

2007 B

beans

sorghum

sugarcane

2007 B

beans

sorghum

sugarcane

2007 A

banana

soybean

beans

2007 A

banana

soybean

beans

2007 B

banana

beans

2007 B

banana

beans

2007 A

cassava

2007 A

cassava

2007 B

cassava

2007 B

cassava

2007 A

banana

2007 A

banana

2007 B

banana

beans

2007 B

banana

beans

organic inputs


manure, HH waste)

organic inputs


manure, HH waste)

2007 A

banana

2007 A

banana

2007 B

banana

2007 B

banana

Figure 5: An example of a farm map of relatively wealthy household in Luhihi (Sud-Kivu), with livestock and manure facilities, and 7 fields (all owned by the household). Five fields are cultivated by legumes (yellow fields are of medium fertility and green fields of high fertility, according to the farmer’s appraisal).

Common legume systems differ between the mandate areas (Figure 6). Legumes are commonly associated with cassava in Bas-Congo, with cassava and/or sweet potato in Sud-Kivu, and with cereals in Umutara. In Kibungo, legumes are frequently grown in association with both cereals and banana. Mixed cropping systems, with 3 or more crop types grown in the same field, are also common (except in Bas-Congo); these include mostly associations of root and tuber crops with cereals and legumes, and to a lesser extent banana with cereals and legumes (except in Kibungo). Legume mono-cropping is uncommon and almost never practiced during two consecutive seasons; farmers are aware of the disease accumulations, particularly for beans. Pure rotation systems are likewise rare, and some legumes are usually planted in association during the cereal season. Planting in line is very rarely practiced for legumes or cereals; seeds are usually simply broadcast.

25

Kibungo and Bugesera

Umutara

Bas-Congo

Sud-Kivu

cereal-legume association

cereal-legume rotation

root/tuber-legume association

banana-legume association

legume mono-cropping

mixed cropping

Figure 6: Relative importance of common legume production systems in the 4 mandate areas of the TSBF-CIAT project.

In Sud-Kivu and Rwanda, more than 70% of the legume-grown fields are positioned on slopes. While in Rwanda, conservation structures are common and well-maintained, in Sud-Kivu these are almost entirely absent. Almost 90% of the legume-grown fields on slopes are unprotected (only some physical embankments without hedgerows were observed), and two thirds of the fields show visible signs of erosion. Farmers however consider low soil fertility, drought and climatic variability as the major constraints for legume production. Legume commercialisation Legume varieties were characterized and farmers specified the minimal and maximal prices at which they sold their legume grains during the year on the local market. A preliminary analysis was conducted in Sud-Kivu (Figure 7). Groundnuts are primarily produced in Kabamba, and are sold at a much higher price (min. price = 1.6 $ kg-1) than beans and soybean (min price = 1.0 $ kg-1 and 0.6 $ kg-1, respectively). Maximal groundnut purchase prices in Kabamba are 1.9 $ kg-1 during periods of scarcity. Soybean prices differ between sites. Prices are lowest in Kabamba and Luhihi (on average 0.65 $ kg-1), where soils are relatively more fertile and soybean is more commonly produced than in the Walungu area (Lurhala and Mwegerera). Minimal soybean purchase prices in the Walungu area are almost twice as high as on the northern axis (Kabamba and Luhihi). For beans, purchasing prices fluctuate around 1 $ kg-1, and are slightly lower in Lurhala. Maximal price increases during periods of scarcity occur in Kabamba (up to 0.5 $ increase kg-1). Price differences in space and time open up a number of marketing opportunities through transportation and storage, allowing producers to sell where and when the price is highest. Moreover, prices depend on grain traits such as size and colour. Prices for grains of the preferred colour (red and white) are higher than for less-preferred colours (black), and usually increase for larger grain sizes.

26

Bbc

Cc

A

Aab

Bbcc

ab

aa

0.0

0.5

1.0

1.5

2.0

groundnut beans soybean

minimum purchase price (USD kg-1)

Aa

BaAb

Bb

ca

b a

0.0

0.5

1.0

1.5

2.0

beans soybean

maximal price increase during year (USD kg-1)

Kabamba

Luhihi

Lurhala

Mwegerera

Figure 7: Minimal legume grain purchasing prices (left) and price increases during periods of scarcity (right) as reported by farmer-producers in the 4 action sites in the Sud-Kivu mandate area. Letter labels indicate a significant (P<0.05) difference in price between species (capital letters) or sites (lower-case letters).

Nutritional status of rural households In Sud-Kivu, malnutrition is very prevalent in younger children; more than 30% of 2- to 3-year-old children show at least mild symptoms of marasmus or suffer from kwashiorkor (Figure 8). Lack of muscular tissue, swollen abdomen, wrinkled or flaky skin, and scanty, pale hair were the most commonly observed symptoms. Relatively less symptoms of malnutrition were observed in Rwanda (across both mandate areas). In Bas-Congo, symptoms of kwashiorkor or marasmus were very rarely observed. In Rwanda and in Bas-Congo, malnutrition was more pronounced in girls than in boys, while in Sud-Kivu, symptoms of marasmus or kwashiorkor were more often observed in boys than in girls.

0

10

20

30

40

50

24-35

months

36-47

months

48-59

months

60-71

months

proportion of age group showing

symptoms of malnutrition (%) girls

boys

Rwanda

0

10

20

30

40

50

24-35

months

36-47

months

48-59

months

60-71

months



boys

Sud-Kivu

0

10

20

30

40

50

24-35

months

36-47

months

48-59

months

60-71

months



boys

Bas-Congo

Figure 8: Prevalence of malnutrition symptoms (marasmus/kwashiorkor) in 2- to 5-year-old children of rural households in Rwanda, Sud-Kivy and Bas-Congo.

27

Anthropometric measures were taken in 2- to 5-year-old children, and included the weight, height and mid-upper-arm circumference (MUAC). In Sud-Kivu, depending on the measures and cut-off points used, between 2 and 12% of the children suffer moderate acute malnutrition, and between 14 and 25% are at risk of malnutrition (Figure 9). In Rwanda, anthropometric measures identified very few children with moderate malnutrition, and up to about 10% were at risk. In Bas-Congo, somewhat conflicting results for malnutrition prevalence were observed for the both measures. While MUAC measurements suggested mild and moderate malnutrition in over 20% and almost 10 % of 2- to 5-year-old children, respectively, malnutrition was almost entirely absent according to weight-for-height (WFH) measurements. Estimates of malnutrition rates tend to be larger using MUAC measurements than WFH measurements (Bairagi and Ahsan, 1998).

0

20

40

60

80

100

<11cm 11-12.5cm 12.5-13.5cm >13.5cm

Malnutrition prevalence in children

according to M

UAC (%) girls

boys

(a) Rwanda

0

20

40

60

80

100

z-score<-3 -3<z-score<-2 -2<z-score<-1 -1<z-score


according to W

FH (%)

girls

boys

(b) Rwanda

0

20

40

60

80

100

<11cm 11-12.5cm 12.5-13.5cm >13.5cm


according to M

UAC (%) girls

boys

(c) Sud-Kivu

0

20

40

60

80

100



according to W

FH (%)

girls

boys

(d) Sud-Kivu

0

20

40

60

80

100

<11cm 11-12.5cm 12.5-13.5cm >13.5cm


according to M

UAC (%) girls

boys

(e) Bas-Congo

0

20

40

60

80

100



according to W

FH (%)

girls

boys

(f) Bas-Congo

Figure 9: Malnutrition prevalence according to (a,c,e) mid-upper-arm circumference (MUAC) and (b,d,f) weight-for-height (WFH) in 2- to 5-year-old children of rural households in Rwanda, Sud-Kivu and Bas-Congo. Categories signify (from left to right): severe, moderate, mild and absent acute malnutrition (Lancet and Morley, 1974); z-scores are relative to the median of the respective populations.

28

Legume consumption in Sud-Kivu Diets of 2- to 5-year-old children were determined by asking mothers to recall the foods given to their child during the past week and month. Presented below are the data for the action sites in the Sud-Kivu mandate area. Legumes constitute the principal source of protein for children in Sud-Kivu. In Lurhala and Mwegerera, more than 50% of the children consume meat- or fish-derived protein less than once a week, and usually only once or twice a month (Figure 10). About 30% of the households feed their children with meat or fish once or twice a week. In Kabamba and Luhihi, which are close to the lakeside, consumption of small fish (‘frétins’) is very common, but meat is less frequently eaten. The consumption of eggs is very rare in all sites. Among the four legumes grown in the area, soybean and bean are the most important legumes in the diets of young children. Cowpea is rather uncommon and groundnut is principally grown as a cash crop. In Luhihi, a productive soybean area, about three quarters of the children are fed with soybean more than five times a week (Figure 11). Soybean is either prepared as soymilk, tea, or porridge (commonly mixed with maize and/or sorghum). In the other sites, soybean remains an important constituent of young children and is consumed on average 3-4 times per week. Beans are also important, and fed at least once or twice a week in 70% of the households.

Kabamba

0

20

40

60

80

100

0 1-2 3-5 >5

frequency of consumption (week-1)

proportion of children (%)

Luhihi

0

20

40

60

80

100

0 1-2 3-5 >5



Lurhala

0

20

40

60

80

100

0 1-2 3-5 >5



Mwegerera

0

20

40

60

80

100

0 1-2 3-5 >5


proportion of children (%) fish

meat

eggs

Figure 10: Frequency of consumption of meat, fish and eggs by 2- to 5-year-old children in the 4 action sites in the Sud-Kivu mandate area.

29

Kabamba

0

20

40

60

80

100

0 1-2 3-5 >5



Luhihi

0

20

40

60

80

100

0 1-2 3-5 >5



Lurhala

0

20

40

60

80

100

0 1-2 3-5 >5



Mwegerera

0

20

40

60

80

100

0 1-2 3-5 >5


proportion of children (%) beans

soybean

Figure 11 Frequency of consumption of beans and soybean by 2- to 5-year-old children in the 4 action sites in the Sud-Kivu mandate area.

33..CC..22.. DDEETTAAIILLEEDD CCHHAARRAACCTTEERRIISSAATTIIOONN SSTTUUDDYY OONN BBAANNAANNAA

PPRROODDUUCCTTIIOONN,, MMAARRKKEETTIINNGG AANNDD CCOONNSSUUMMPPTTIIOONN The objective of the farm-level diagnostic studies was to identify and quantify the biophysical and economic parameters driving the banana cropping systems. During the CIALCA planning meetings, the objectives, tools and sampling strategies for this activity were further developed and the final tool was completed and field-tested in October 2006. Subsequently, staff and students executing the diagnostic surveys in North Kivu, Rwanda, South Kivu, and eventually Burundi all underwent a minimum 2 day on-site training by IITA-Uganda research staff. At each action site, 30 farms representing a subsample of the baseline survey were visited. The 30 farms per site represented 10 wealthy, 10 medium, and 10 poor farmers, following local wealth classifications as set during the PRA’s and as identified by key local resource persons. This sampling framework resulted in a total sample size of over 540 farms in 18 sites. Banana production and constraints and farmer perceptions. Banana productivity averaged by site ranged from 21-43, 25-53 and 35-63 Mg ha-1 cycle-1 for Burundi, Rwanda, and South Kivu, respectively. With the cycle duration (i.e. period between two harvests from a single mat) averaging about 1 year, these production levels are much higher than those reported in general by FAO for the region (6-15 Mg ha-1 yr-1). Banana pest (i.e. nematodes and weevils) and disease (i.e. Black Sigatoka) damage parameters varied distinctly between regions, but were strongly negatively related to altitude. Fusarium wilt was widespread and limited productivity of exotic cultivars (i.e. Pisang Awak, Apple banana, Gros Michel). Banana Xanthomonas wilt (BXW) severely affected production in certain sites in Central Uganda, Western Rwanda and Eastern DRC, while banana bunchy top virus (BBTV) was observed in the Rusizi valley. However, BXW

30

and BBTV currently affect a limited geographic area and still have relatively small impact on national banana production. Highest productivity was observed near the Albertine rift, where soils are relatively young, rainfall is high (> 1400mm yr-1), and plant densities are high (1800-3300 mats ha-1), compared to much of Eastern and Central Rwanda and Burundi, which have strongly weathered soils, low rainfall (<1100mm yr-1), and low plant densities (1000-1700 mats ha-1). These findings suggest that much of the current research focus is based on flawed data and assumptions. Water stress, which has previously been overlooked, is an important production constraint, while several ‘traditional’ banana pests and diseases may not be as important as is often assumed.

Lab analysis of foliar samples revealed that nutrient deficiencies such as P and Mg are indeed predominant on strongly weathered soils, such as the Walungu axis south of Bukavu (Figure 12). Potassium deficiency generally predominates on soils that have a low stock of weatherable nutrients (quartzite and granite) such as in Kibuye, Ruhango (both Rwanda) and Gitega (Burundi). Areas that know high levels of crop management and where external nutrient inputs (e.g. external mulch) is frequently applied (e.g. Kibungo) tend to have less nutrient deficiency problems.

Ruhango

Nzenga

Munoli

Lurhala

Luhihi

Kirundo-

Kibuye

Kibungo

Kaliva

Kabam

ba

Gitega-2

Gitega-1

Cyanangu

Cibotoke

Burhale

Bingo

K (% of dry m

atter)

5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.5

Ruhango

Nzenga

Munoli

Lurhala

Luhihi

Kirundo-

Kibuye

Kibungo

Kaliva

Kabam

ba

Gitega-2

Gitega-1

Cyanangu

Cibotoke

Burhale

Bingo

N (% dry m

atter)

5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

Ruhango

Nzenga

Munoli

Lurhala

Luhihi

Kirundo-

Kibuye

Kibungo

Kaliva

Kabam

ba

Gitega-2

Gitega-1

Cyanangu

Cibotoke

Burhale

Bingo

P (% dry m

atter)

.5

.4

.3

.2

.1

Ruhango

Nzenga

Munoli

Lurhala

Luhihi

Kirundo-

Kibuye

Kibungo

Kaliva

Kabam

ba

Gitega-2

Gitega-1

Cyanangu

Cibotoke

Burhale

Bingo

MG (% dry m

atter)

1.2

1.0

.8

.6

.4

.2

0.0

Figure 12: Foliar nutrient concentrations (in % dry matter weight) for the diagnostic survey sites in Burundi (Gitega, Cibitoke, Kirundo), South Kivu (Burhale, Kabamba, Luhihi, Lurhala), North Kivu (Bingo, Kaliva, Munoli, Nzenga), and Rwanda (Cyangugu, Kibungo, Cyangugu, Ruhango)

We also compared farmers’ perceptions of constraints and coping strategies related to banana production, and verify these with field assessments. Average farmer bunch weight estimates were variable, ranging from large (76%) over estimations in Bugesera (Rwanda) to large under estimations (63%) in Karongi. In most Rwandan and Burundian sites,

31

farmers reported drought stress (85%) and poor soil fertility (74%) as major constraints to banana productivity, whereas a minority (26%) of farmers mentioned pest and diseases as major constraints. Farmer perceptions are in line with our assessments. Drought stress yield losses based on average rainfall were estimated at between 30 and 70% for sites in central and eastern Rwanda and Burundi. Banana weevil (Cosmopolites sordidus) damage was low at all sites (<3.5%) and root necrosis moderate (13 – 23%), with the exception of Cibitoke where root necrosis was only 8%, due to the dominance of the nematode resistant Yangambi km5 cultivar. Fusarium wilt strongly affected exotic banana production (i.e., AB and ABB beer and dessert bananas) in all sites and 10– 67% of plantations. Farmers have therefore resorted to replacing Kayinja (ABB) with AAA-EA beer bananas. Farmers’ crop management was varied. Banana yield was positively correlated (r2=0.13, P<0.001) with amount of mulch applied. However, few farmers (22%) applied external mulch. While most farmers interviewed (86%) owned some cattle or small ruminants, only a minority of them (39%) reported application of manure/compost. None of the farmers applied mineral fertilizers. This study shows that farmers correctly perceive abiotic stress factors as the most yield limiting. However, only a minority of farmers seem to adopt technologies (i.e., application of mulch, manure, and compost) to overcome these yield-limiting factors.

0

10

20

30

40

50

60

70

80

90

0 100 200 300 400 500 600

Penetrometer resistance

Yield (t/ha/cycle)

Figure 13 Penetrometer measurements from Rwanda, Burundi and South Kivu suggest that maximum attainable yields (t/ha/cycle) decrease with increasing penetrometer resistance.

32

444... PPPRRROOOGGGRRREEESSSSSS WWWIIITTTHHH SSSTTTRRRAAATTTEEEGGGIIICCC SSSOOOIIILLL

FFFEEERRRTTTIIILLLIIITTTYYY---RRREEELLLAAATTTEEEDDD AAACCCTTTIIIVVVIIITTTIIIEEESSS

444...AAA... RRREEELLLAAATTTIIIOOONNNSSSHHHIIIPPP BBBEEETTTWWWEEEEEENNN SSSOOOIIILLL FFFEEERRRTTTIIILLLIIITTTYYY AAANNNDDD

NNNUUUTTTRRRIIITTTIIIOOONNNAAALLL QQQUUUAAALLLIIITTTYYY OOOFFF BBBIIIOOO---FFFOOORRRTTTIIIFFFIIIEEEDDD BBBEEEAAANNNSSS A G by E analysis of Fe contents in grains of beans grown in Sud-Kivu and Umutara In the legume evaluation trials, 27 bush bean and 9 climbing bean varieties, along with the local variety, were tested by 2 farmer associations in 4 sites in Sud-Kivu, DRC and in 4 sites in Umutara, Rwanda (16 associations in total). In each association, separate blocks were set up with a control and a treatment with goat manure application at 5 t dry matter (DM) ha-1. After harvest, a representative sample of grains was taken, oven-dried and manually ground using an agate mortar and pestle. A subset of the grain samples was analysed for Fe contents using radial ARL ICP-AES by ARI laboratories, Adelaide.

A preliminary analysis was conducted by fitting following general linear model:

Fe_Ca,t,g = µ + αa + βt + γg + θa,t + θ’g,t + εa,t,g with Fe_Ca,t,g = grain Fe content for genotype ‘g’ in association ‘a’ with treatment ‘t’ [mg Fe kg-1], µ = grand mean, αa = environment (association) mean deviations, βt = treatment mean deviations, γg = genotype mean deviations, θa,t = association × treatment interaction residuals, θ’g,t = genotype × treatment interaction residuals and εa,t,g = the error term; the association CINAMULA (Lurhala) was excluded from the model as no observations in the control treatment were available.

Analysis of variance demonstrates that grain Fe content is largely determined by environment (association) and genotype and unaffected by FYM application (Table 10). Genotypic and environmental effects alone can explain 17 and 44% of the total variation, respectively. Interaction effects between genotype and environment are part of the error term (35.1% of the total variation).

Table 10: ANOVA for significance of genotype, environment (association), treatment and environment × treatment and genotype × treatment interaction effects on Fe contents in bean grains

source of variation df SS MS F-value P-value % of total SS

total 214 27684.3 model 53 17971.4 339.1 5.62 <0.0001 64.9 environment (assoc.) 10 12304.1 1230.4 20.40 <0.0001 44.4 treatment 1 0.4 0.4 0.01 0.9335 0.0 genotype 16 4682.9 292.7 4.85 <0.0001 16.9 environment × treat. 10 534.3 53.4 0.89 0.5480 1.9 genotype × treat. 16 449.6 28.1 0.47 0.9599 1.6 error 161 9712.9 60.3 35.1

Bean varieties Marungi and ARA4 generally contained highest grain Fe contents while BRB194, CIM9314-36 and Kiangara contained lowest grain Fe contents (Table 11). Highest Fe contents were observed in the ALEMALU and APACOV associations (which are rather infertile sites), while lowest Fe contents were observed in the MAENDELEO, RUSINAME and IRIBA associations (which are rather fertile sites).

33

Table 11: Adjusted grain Fe content means and standard errors for the different bean varieties (across associations) and for the different associations (across varieties)

association action site mean SE species variety mean SE

ALEMALU Lurhala 82.55 3.11 BB Marungi* 74.06 1.91 APACOV Burhale 77.58 1.64 BB ARA4* 72.91 1.85 DUFATANYE Nyakigando 73.24 1.39 BB ZAA5/2 72.82 2.11 TWISUNGANE Rugarama 71.53 1.90 BB ECAPAN021 72.44 2.04 RHUBEHAGUMA Luhihi 71.00 2.01 CB VCB81013* 72.36 2.80 ABAGWASINYE Burhale 70.26 4.19 BB HM21-7* 71.90 2.03 TUUNGANE Kabamba 64.84 2.00 BB ZKA93-10m* 69.77 1.80 ISOKO Y’UB’MW. Kabarore 62.59 1.44 CB MLV06* 69.63 2.99 MAENDELEO Kabamba 59.52 1.74 BB AFR708 68.47 2.10 RUSINAME Luhihi 59.36 1.68 BB CODMLB003* 68.28 2.19 IRIBA Murambi 58.93 1.77 CB AND10* 68.08 2.39

CB VCB81012* 67.78 3.01 CB local variety CB 67.16 3.00

BB local variety BB 65.50 2.19

BB BRB194* 61.95 2.19 BB CIM9314-36 59.64 2.04

CB Kiangara* 58.51 2.68

The AMMI model analyses G×E interactions by combining ANOVA (with additive parameters) and PCA (with multiplicative parameters). As it requires a fully balanced dataset (i.e. each genotype in each environment), only a sub-selection of the full dataset was submitted. This sub-selection included 10 bush bean varieties (AFR708, ARA4, BRB194, CIM9314-36, CODMLB003, ECAPAN021, HM21-7, Marungi, ZAA5/2 and ZKA93-10m/95) in 7 associations (APACOV, RHUBEHAGUMA, RUSINAME, TUUNGANE, MAENDELEO, ISOKO Y’UBUMWE and DUFATANYE. This sub-selection includes both some of varieties with the highest and lowest grain Fe contents but does not include the associations with highest (ALEMALU) or lowest (IRIBA) grain Fe contents. Applying the simple regression model to the data sub-selection shows that purely genotypic effects remain significant (explaining 33% of the total variation) but purely environmental effects become insignificant (P<0.37). The AMMI analysis can be used to study G×E interactions in the data sub-selection; however, it does not take full account of the environmental variability in the entire dataset. As manure application did not significantly affect grain Fe contents, treatments were considered as replications in the AMMI analysis. Missing values (18) were predicted using the model described in the preliminary analysis (i.e. assuming no G×E interaction) to balance the dataset. The AMMI model is described as:

Fe_Ca,g = µ + αa + γg +∑−

ηζλN

1n

n,an,gn +ρa,g + εa,g

with Fe_Ca,g = grain Fe content for genotype ‘g’ in association ‘a’ [mg Fe kg-1], µ = grand mean, αa = environment (association) mean deviations, γg = genotype mean deviations, N = the number of singular value decomposition (SVD) axes retained in the model, λn = singular value for SVD axis n, ζg,n = genotype singular vector value for SBD axis n, ηa,n = association singular vector value for SVD axis n, ρa,g = AMMI residuals, and εa,g = the error term.

The AMMI analysis shows that grain Fe contents are significantly affected by environment (association) and genotype, which explained 5 and 46% of the model variation, respectively (Table 12). G×E interaction accounted for 49% of the total model variation. Two IPCA factors could significantly explain 74% of the G×E interaction variation.

The IPCA factors were tested for correlation with soil characteristics (pH, org. C, total N, extractable P and exchangeable bases). IPCA 1 was found to be negatively correlated with

34

soil organic C and total N contents (r = -0.71 and -0.77 with P-values 0.07 and 0.04, respectively). Environments with higher IPCA 1 scores would thus be less fertile environments with lower organic C and total N contents. Sites with higher soil organic C and total N contents were RUSINAME, RHUBEHAGUMA, MAENDELEO and APACOV. Sites with lowest soil organic C and total N contents were ISOKO Y’UBUMWE and DUFATANYE.

Table 12: AMMI analysis of variance for significance of genotype, environment (association) and genotype × environment (association) interaction effects on grain Fe contents, and the partitioning of interaction effects into AMMI axes

source of variation df SS MS F-value P-value % of G×E SS

total 139 17636 126.9 model 69 12656 183.4 2.41 0.0003 genotype 9 5795 643.9 8.45 <0.0001 environment (assoc.) 6 615 102.5 3.96 0.0020 block 7 181 25.9 0.34 0.9326 genotype × environ. 54 6246 115.7 1.52 0.0553 IPCA1 14 2496 178.3 2.34 0.0113 40.0 IPCA2 12 2126 177.2 2.33 0.0156 34.0 residuals 28 1623 58.0 0.76 0.7846 26.0 error 63 4799 76.2

The AMMI biplot (Figure 14) shows 71% of the model variation with 46%, 5% and 20% due to genotype, environment and G×E interaction (IPCA 1 only), respectively. For any G-E combination in the biplot, the AMMI-calculated grain Fe content can be estimated by adding the G and E means minus the grand mean (69.9 mg Fe kg-1) to the product of the G and E IPCA 1 scores. For variety BRB194 grown in ISOKO Y’UBUMWE for example, this becomes:

Fe_CAMMI = 76.0 (G mean) + 68.7 (E mean) – 69.9 (grand mean) + 2.99 (G IPCA 1 score) × 4.37 (E IPCA 1 score) = 87.9 mg Fe kg-1

This fits the observed value Fe_Cobserved = 90.3 mg Fe kg-1. Actual calculated values by the AMMI model (in this case: 88.2 mg Fe kg-1) may differ slightly from these simple calculated values as the AMMI model also considers the G×E interaction effect accounted for by ICPA 2 (17% of the model variation).

This example illustrates the positive interaction between this genotype and environment. Inversely, growing AFR708 in the same association would lead to a negative interaction. Genotypes can be selected based on high Fe contents and/or on stability (i.e. less variable grain Fe contents across environments). When selecting genotypes for high Fe contents, genotypes need to be chosen with high means and positive interaction with a given environment. For ISOKO Y’UBUMWE, TUUNGANE and DUFATANYE, these genotypes would thus be BRB194 and ZKA93-10m, while for MAENDELEO, RHUBEHAGUMA and APACOV, these genotypes would be ARA4 and AFR708. Varieties with stable grain Fe contents across environments are ECAPAN021 and ZAA5/2. However, these varieties both have low Fe contents (below the grand mean). The most suitable variety across all environments would thus be ARA4, with the highest mean Fe content (80.0 mg Fe kg-1) and a relatively small ICPA 1 score (-1.56 mg Fe kg-0.5), indicating relatively high stability.

35

IPCA 1 (mg Fe kg

-1)0.5

MAENDELEO

RHUBEHAGUMAAPACOV

RUSINAME

DUFATANYE

TUUNGANE

ISOKO Y'UBUMWE

ARA4

AFR708

CODMLB003

HM21-7

ECAPAN021ZAA5/2

Marungi

CIM9314-36

ZKA93-10m

BRB194

-3

-2

-1

0

1

2

3

4

5

55 60 65 70 75 80 85

environment (association)

genotype

grain Fe content (mg Fe kg-1)

Figure 14: AMMI biplot showing the main and IPCA 1 effects of both genotypes and environments (associations) on grain Fe content; an estimate of the G×E interaction effect for a specific genotype – environment (association) combination equals the product of their corresponding IPCA1 scores.

Identification of varieties with highest Fe contents for specific environments can also be done through plotting the AMMI-calculated grain Fe contents for each genotype in function of the environments’ IPCA 1 scores (Figure 15). Genotypes thus become ranked for grain Fe content in each environment. Varieties ARA4, AFR708, ZKA93-10m and BRB194 are amongst the 4 varieties with highest Fe contents in most associations (Table 13). Except for AFR708, these are known biofortified varieties with elevated Fe and/or Zn contents in the grains. These varieties also show opposite trends in grain Fe content versus IPCA 1 score relationship: while BRB194, and ZKA93-10m show increasing grain Fe contents with increasing IPCA 1 score (decreasing soil C and N content), ARA4, AFR708 and CODMLB003 show decreasing grain Fe contents with increasing IPCA 1 scores (increasing soil C and N content).

Table 13: Varieties ranked following grain Fe contents based on AMMI-calculated values in each environment; varieties marked with an asterisk (*) are known biofortified varieties with elevated grain Fe and/or Zn contents.

association action site 1st variety 2nd variety 3rd variety

APACOV Burhale ARA4* AFR708 CIM9314-36* RHUBEHAGUMA Luhihi ARA4* AFR708 ZKA93-10m* RUSINAME Luhihi ARA4* ZKA93-10m* BRB194* TUUNGANE Kabamba ZKA93-10m* BRB194* ARA4* MAENDELEO Kabamba ARA4* AFR708 CODMLB003* ISOKO Y’UB’MW. Kabarore BRB194* ZKA93-10m* CIM9314-36* DUFATANYE Nyakigando CIM9314-36* ARA4* AFR708

36

AMMI-calculated grain Fe content (mg Fe kg

-1)

ZKA93-10m

CODMLB003

AFR708

ARA4

BRB194

CIM9314-36

ECAPAN021

HM21-7

Marungi

ZAA5/2

MAENDELEO

RHUBEHAGUMA

APACOV

RUSINAME

DUFATANYE

TUUNGANE

ISOKO

Y'UBUMWE

40

45

50

55

60

65

70

75

80

85

90

-3 -2 -1 0 1 2 3 4 5

←higher soil C+N content IPCA 1 (mg Fe kg-1)0.5 lower soil C+N content→

Figure 15: Calculated grain Fe contents of 10 bean varieties based on the AMMI model equation across environment IPCA 1 scores. Full lines are regressions for varieties BRB194, ZKA93-10m and CIM9314-36; dotted lines are regressions for varieties ARA4, AFR708 and CODMLB003.

Conclusion A simple linear regression model applied to the entire dataset showed that 44% of the total variation in grain Fe contents could be attributed to purely environmental effects, while 17% is related to purely genotypic effects. An AMMI analysis on selected varieties and environments (associations) was conducted to study G×E interaction effects. Although the sub-selection did not represent the environmental variability of the entire dataset, it revealed significant G×E interaction effects accounting for 35% of the total variation in the data sub-selection. This demonstrates the necessity of taking G×E interactions into account when selecting bean varieties for high grain Fe contents.

The variety ARA4 (Photograph 4) can be recommended as a genotype with high Fe contents (73 mg Fe kg-1) across environments. However, other varieties such as BRB194 and ZKA93-10m have higher grain Fe contents in specific environments, likely characterized by less fertile soils with lower organic C and total soil N contents.

444...BBB... AAASSSSSSEEESSSSSSMMMEEENNNTTT OOOFFF NNNUUUTTTRRRIIIEEENNNTTT DDDEEEFFFIIICCCIIIEEENNNCCCIIIEEESSS IIINNN SSSOOOIIILLLSSS

OOONNN TTTHHHEEE WWWAAALLLUUUNNNGGGUUU AAAXXXIIISSS IIINNN SSSUUUDDD---KKKIIIVVVUUU The Walungu area in Sud-Kivu is very unproductive due to low soil fertility constraints (see also section 7.A). The exact nature of the soil constraints remains unknown. Results from a set of field trials established in February 2007 (“FER-1”) instigated a series of greenhouse pot trials, set up between August and December 2007 at the research center of

Photograph 4: ARA4, a biofortified bean variety with stable and high Fe contents.

37

INERA (Institut National pour l’Etude et la Recherche Agronomique) in Mulungu, Sud-Kivu (Photograph 5). Preliminary results from these trials are highlighted below.

Pot trial I At each of the 8 locations of the field trial, 60kg of soil was sampled from the 0-20 soil layer, sun-dried and sieved to pass 4 mm. These soils were then used to set up a pot trial with 8 different treatments: a reference treatment with all nutrients applied (540 mg N kg-1 soil, 54 mg P kg-1, 450 mg K kg-1, 225 mg Ca kg-1, 99 mg Mg kg-1, 36 mg S kg-1, 3.2 mg Fe kg-1, 3.2 mg Mn kg-1, 2.0 mg Zn kg-1, 2.0 mg Cu kg-1, 0.6 mg B kg-1 and 0.6 mg Mo kg-1), six treatments with K, P, Mg, S, Zn and B omitted, respectively, and a treatment with application of N, P and K only (at the same rates as in the reference treatment). The above rates were assumed to eliminate deficiencies for the respective plant nutrients. Following salts were used for the nutrient additions: KH2PO4, NH4H2PO4, Mg(NO3)2.6H2O, NH4NO3, KNO3, Ca(NO3)2.4H2O, MgSO4.7H2O, (NH4)2S04, ZnCl2, CuCl2.2H2O, FeCl3, MnCl2.4H2O, (NH4)6Mo7O24.4H2O, and Na2B4O7.10H2O. The pots were filled with 2.5 kg of soil, after which the soil of each pot was mixed with nutrient solutions according to the different treatments. The pots were then placed on tables in the greenhouse following a randomized complete block design with 3 replicates, and rotated daily. In each pot, 3 maize (Zea mais L., cv. Katumani) seeds were sown, and thinned to one plant per pot after one week (1 WAP). Moisture conditions were kept optimal in the course of the trial. The average minimum and maximum temperatures during the growth period equalled 15.6°C and 43.0 ºC, respectively. Regularly, the height of the plants (i.e. the distance from the plant basis to the highest tip of the three youngest fully developed leafs) was measured. At 5 WAP, plants were harvested and the biomass was sun-dried in closed paper bags. Subsequently, plants were oven-dried (65ºC) and weighed.

All plants showed severe visual signs of P deficiency (Photograph 6), and no significant differences in height or final dry weight were observed between soils or treatments. At 5 WAP, all plants had a similar height as low as 45 cm. As such, the assumed rate of 54 mg P kg-1, which corresponds to a (broadcast) field rate of about 120 kg P ha-1, is inadequate to lift available P levels sufficiently to eliminate P deficiency for maize. In conclusion, P deficiency was the principal constraint for crop growth in the soils tested.

Pot trial II A second pot trial was established to investigate the extent of P deficiency in the two action sites. To this end, a wider range of soils was collected from other the sites, including the fields chosen for legume evaluation, legume seed multiplication fields and randomly selected soils sampled during the final characterisation study. While the former two sets of

Photograph 5: Missing nutrient pot trial established at a greenhouse at INERA-Mulungu, Sud-Kivu, DRC.

Photograph 6: Visual symptoms of P deficiency in young maize plants.

38

soils were presented by farmer associations for project activities and of variable fertility according to the association’s ability to meet the expenses of providing land, the latter soils are typically cultivated by legumes following the local practices. The selected soil collection is considered a representative set of soils used for legume cultivation in the Walungu area.

The pot trial was set up following procedures for planting, watering and measuring plants during plant growth equal to the first trial. Six different treatments were imposed in a single replicate: a control without nutrient additions, a reference treatment with additions of all nutrients (396 mg N kg-1, 360 mg P kg-1, 360 mg K kg-1, 210 mg Ca kg-1, 92 mg Mg kg-1, 42 mg S kg-1, 2.9 mg Mn kg-1, 1.9 mg Zn kg-1, 1.9 mg Cu kg-1, 0.6 mg B kg-1 and 0.6 mg Mo kg-1), 3 treatments with N, P and K omitted respectively, and a treatment with application of N, P and K only (at the same rates as in the reference treatment). The P rate was increased 10-fold as compared to the first trial to ensure P deficiency was entirely eliminated. The average minimum and maximum temperatures during the growth period equalled 15,6°C and 45,6°C, respectively. Plants were cut at 38 days after planting (DAP). At harvest the youngest fully developed leaf was cut and dried separately for leaf nutrient analysis.

Treatment differences became apparent at about 3 WAP (Figure 16, Photograph 7). At 32 DAP, plant heights in the control and the treatments without P addition were similar and as low as 50 cm, while an average height of 85 cm was observed in the reference treatment. Visual signs of P deficiency were observed in the former two treatments, and were very pronounced in all 30 soils (without exception); final dry weights were more than 5 times lower, compared to the reference treatment. In the other treatments with N, K or micronutrients omitted, responses were differential, but no consistent differences with the reference treatment were observed. In some soils, however, maize plants clearly showed to suffer from nutrient deficiencies other than P. Leaf analyses (pending) are required to diagnose these.

SED treatment SED time

0

20

40

60

80

100

0 5 10 15 20 25 30 35

time (DAP)

plant height (cm)

control

PK + µnutrients

NK + µnutrients

NP + µnutrients

NPK

NPK + µnutrients

SED (a)

SED (b)

0

2

4

6

8

10

control PK +

µnutrients

NK +

µnutrients

NP +

µnutrients

NPK NPK +

µnutrients

DM yield (g)

legume evaluation soils (n=3)

multiplication soils (n=12)

final characterisation soils (n=15)

Figure 16: Maize plant height in function of time (top) and dry biomass weight after 38 days of growth (bottom), as affected by different nutrient application regimes, observed in a pot trial conducted at INERA-Mulungu, Sud-Kivu, DRC; error bars represent SED for comparison of treatment and time effects (left), and soil group (a) and treatment (b) effects (right).

39

No significant differences in maize response were observed between soils used for legume germplasm evaluation, legume multiplication and soils sampled during the final characterisation study. The similarity between the soil types was confirmed by soil analysis (Table 14). All soils were similarly characterized by a similar and low organic C content (0.6%), soil pH (5.2) and low cation exchange capacity (6 cmolc kg

-1).

Table 14: Soil properties for project soils (sampled in fields for legume germplasm evaluation and multiplication) and final characterisation soils (typically cultivated with legumes, sampled in farmers’ fields).

org. C (g kg-1)

pH (H20) CEC (cmolc kg-1)

Project soils 6.63 5.2 5.88 Final characterisation soils 5.36 5.3 6.71 SED 0.82 0.2 0.46

444...CCC... AAASSSSSSEEESSSSSSMMMEEENNNTTT OOOFFF BBBAAANNNAAANNNAAA ––– AAARRRBBBUUUSSSCCCUUULLLAAARRR

MMMYYYCCCOOORRRRRRHHHIIIZZZAAALLL FFFUUUNNNGGGIII RRREEELLLAAATTTIIIOOONNNSSSHHHIIIPPPSSS A survey was carried out in 188 fields in Rwanda to identify arbuscular mycorrhizal fungi (AMF) and plant parasitic nematode infection on banana roots in five regions: Ruhengeri (young volcanic soils), Gitarama–Butare (soils derived from granitic rocks), Kibungo (weathered soils from schistose materials), Gashonga (clay ferralsol on basalt) and Bugarama (volcanic alluvial soils). Data were recorded for the single cultivar Intuntu (AAA-EA). We recorded management practices, root health parameters, root colonization by arbuscular mycorrhizal fungi (AMF) and nematode infection in roots. Highly varying AMF colonization was observed in different soil types. Highest colonization was recorded in roots of banana plants grown on clay soils on basalt with frequency 62.6% and intensity 35.4%, while the lowest was on poor weathered soils on granite with 17.2% and 16.7% of frequency and intensity, respectively. Other soil types had intermediate infection. Higher AMF frequency was associated with slightly increased height (r=0.190, p<0.001) and girth of banana plants (r=0.144, p<0.05), significantly higher number of functional roots (r=0.185, p<0.05), higher root number per 20cm3 of soil volume (r=0.222, p<0.01) and much lower root necrosis (r= - 0.235, p<0.01). This study provides an insight on the role that AMF play in existing highland banana production systems and possible benefits of future use of AMF to improve plant health and vigor. Further studies should respond on the question how AMF may help to improve and sustain the yield of highland banana production systems.

Photograph 7: In most soils, treatments effects followed trends as observed in the example, from left to right: control, N omitted, P omitted, K omitted, NPK only, and reference treatment).

40

555... PPPRRROOOGGGRRREEESSSSSS WWWIIITTTHHH BBBAAANNNAAANNNAAA GGGEEERRRMMMPPPLLLAAASSSMMM---RRREEELLLAAATTTEEEDDD AAACCCTTTIIIVVVIIITTTIIIEEESSS

555...AAA IIINNN---SSSIIITTTUUU GGGEEERRRMMMPPPLLLAAASSSMMM EEEVVVAAALLLUUUAAATTTIIIOOONNN The objectives of these activities are to introduce, test and disseminate new germplasm to farmers in order to improve productivity and profitability under the given constraints (i.e. pests, soils, climate, and management practices) and given the actual and potential market and consumption requirements. This activity is executed in close collaboration with the INIBAP-IPGRI-led project. A protocol for the establishment, management, and data collection has been developed in collaboration with our partners. Germplasm trials have been installed in late 2006 / early 2007 across the region, representing the existing variation in soils and climate (i.e altitude) within the region (Photograph 8).

Sites: 1. South Kivu

a. Luhihi b. Burhale c. Lurhale d. Kabamba

2. North Kivu a. Maboya b. Munoli c. Mutwanga d. UCG

3. Rwanda a. Kibungo b. Ruhango c. Kibuye d. Cyangugu e. Kayonza f. Bugesera

4. Burundi a. Mugina-Cibitoke b. Kirundo-Kirundo c. Busoni-Kirundo d. Giheta-Gitega e. Mutaho-Gitega

Besides providing farmers with improved banana cultivars, the trials will provide much scientific insight into genotype × environment interactions, which will help breeding, IPM and agronomic research in the longer term. The germplasm trials will also serve as farmer demonstration sites for best-bet banana management practices, including management practices such as timely deleafing, weeding, mulching and desuckering, in combination with proper planting practices and cleaning techniques (boiling water treatment, paring) of sucker-derived planting material.

Photograph 8: CIALCA banana germplasm trial at the Mulungu station in Sud Kivu, DR Congo.

Photograph 9: Mg deficient plants in germplasm trial in Lurhala, Sud Kivu

41

Germplasm sources are: (i) three tissue-culture derived IITA highland banana hybrids developed under the DGDC-funded ‘Strategic Musa Improvement Project’ (NSH 20, 22, and 42), (ii) tissue-culture derived exotic cultivars already available in the region (FHIA 01, FHIA 03, FHIA 21, FHIA 25, Yangambi km5) in existing tissue culture labs (i.e. Agrobiotec, Burundi), (iii) tissue-culture derived exotic cultivars obtained from the banana International Transit Center in Leuven, (iv) 4 sucker derived local checks and best-bet highland varieties from within the Great lakes region. The three IITA hybrid varieties were added to the varieties that were mutplied in the course of 2006-2007 and were added to the field trials by late 2007, early 2008. Unfortunately, the banana varieties coming from ITC-Belgium and that were multiplied at Agrobiotech in Burundi got infected by a fungal pathogen in the lab and these plants have therefore not been established in the field yet. A second batch of ITC tissue culture plantlets has been sent to IRAZ for multiplication there. These plants should be added to the germplasm trials by September 2008.

Management practices are as much as possible uniform across sites. These will include: 1. Watering of the initial planting material if deemed necessary for the survival of the plants. 2. Planting pits of 60x60x60 cm. 3. Manure/Compost application in planting pit 4. Mulching of plants 2 x per year at the end of the wet season, preferably with grasses. 5. Desuckering of plants in order to have a 1-2-3 system (mother, daughter, granddaughter),

preferably with a circular movement of plants, so that original planting density are respected as much as possible.

6. Deleafing of leaves that have <50% of functional leaf surface area. 7. Bunding/ water traps in sloping areas. 8. All plant residues (leaves, stems, etc.) will remain within the plot around the mat of origin.

However, ALL bunches will be exported, even when not sold. 9. Methods for selling or distribution of harvested bunches will be negotiated with the partners

managing the trial. 10. No use of pesticide/insecticide – damage parameters will be taken at flowering (nematodes)

and harvest (weevil). 11. Herbicide treatment or superficial removal of weeds by hand or hoe, but without tilling the

soil = damaging of roots.

Data collection in these germplasm trials is ongoing, and a common data collection tool has been developed for all the sites.

In addition, presence of the BBTV vector, pentalonia nigronervosa is monitored in the 3 countries using “yellow traps” within the framework of the PhD project of Celestin Niyongere. Field trials on BBTV reinfection rates in newly planted fields (with both in vitro and sucker-derived plants) comprising 9 genotypes were established in December 2007.

Close to each germplasm trial, local macro-propagation facilities will be (and have been) developed for rapid, low-cost, and relatively clean multiplication of banana planting material at or near the sites of the germplasm trials. After the initial training at IITA-Uganda in 2006, Macro-propagation sites have been established at INERA-Mulungu in DRC, at ISAR-Butare in Rwanda, and at the RWARRI NGO in Kibungo, Rwanda. As we envisage expanding these activities, these first three sites will act as a site for the training of trainers on selection, cleaning and rapid multiplication of banana planting material.

Photograph 10: BBTV screening plot at the ISABU Mparambo station in Burundi.

42

555...BBB SSSTTTRRRAAATTTEEEGGGIIICCC RRREEESSSEEEAAARRRCCCHHH AAATTT KKK...UUU...LLLEEEUUUVVVEEENNN ITC During early 2007, 20 plantain accessions collected in the Congo-Basin were shipped from the Kisangani University in DRC to the ITC. In July, a selection of 36 accessions of East-African highland bananas, diploids and triploids, collected in the Morogoro and Mbeya region of Tanzania in 2005, were sent to the ITC for duplication in vitro. In the course of 2007, the active collection increased with 30 new accessions, and currently includes a total of 1,212 accessions. Currently, the collection comprises 808 accessions (66.6%) that are virus indexed negative and available for international distribution. Accessions in storage were indexed routinely for bacterial endophytes at annual subculturing. A total of 253 tests were performed on 58 accessions in storage. The majority concerned samples of accessions that were checked after virus therapy. Also 58 accessions that were newly introduced and were initiated in vitro for the first time were tested pre-storage. Rejuvenation of the collection was continued for 9 accessions and started for another 51. The plants were decapitated in the greenhouse and will be re-initiated in vitro early 2008. The lyophilized leaf collection presently counts 598 accessions. In 2007, 2,234 samples of 411 accessions were processed in the leaf bank; per accession a variable number of samples (1-37 replicates) are stored at -20°C. In 2007, 509 accessions, represented by 1654 individual tissue samples, were distributed from the ITC gene bank. Most materials were supplied as proliferating cultures (PT) (55%) followed by rooted plantlets (RP) (34%) and lyophilized leaf tissues (LT) (11%). A significant increase in the demand of lyophilized leaf tissue for molecular (DNA) study purposes was recorded. The International Transit Centre processed a total of 27 orders from 22 scientists in 18 countries. Samples of 408 accessions were supplied. The exchange of plant material from the ITC gene bank involved 265 different accessions keeping the utilization ratio of germplasm available for international distribution constant at about 30%. A small number of 84 accessions were distributed to underpin the gene bank activities including virus indexing and therapy, characterization and regional multiplication and distribution activities.

A feasibility study for the virus pre-indexing operation in collaboration with FUSAGx, was started. The applicability of the PhytoPASS technology for preliminary indexing of banana germplasm on different types of tissue (leaf tissue of field plants, corm tissue, and leaf tissue of in vitro plants), pre- and post- entry, and for the most common banana viruses, is currently being assayed on the new introductions from the Democratic Republic of Congo and Tanzania. Preliminary results of the material from DRC indicate that the pre-indexing allows the identification of material with an ‘initially better’ health status, and thus more suitable for introduction in the collection, reducing the need for post-entry therapy and repeated virus indexing. Preliminary research towards induction of drought stress in Musa cell cultures A better understanding of plant response to water deficit has become an increasingly important challenge. The effects of water deficit on plants can be investigated at different levels: (i) in the field (ii) in the green-house (iii) and in vitro. However, whole-plant factors such as differences in morphology, carbohydrate assimilation and partitioning render it very difficult to study the effects of induced drought in se. We explored the possibility to use cell cultures in an in vitro model mimicking drought stress in banana, induced by polyethylene glycol (PEG). The water content of ‘Cachaco (ABB)’, ‘Williams (AAA)’ and ‘Musa balbisiana (BB)’ cell samples were on average 89.6, 88.7 and 91.2% respectively. The biggest influence of the lowest PEG concentration (7.5%), was noted for ‘Musa balbisiana’

43

losing 5.4% of its water content in comparison to 3.5% measured for ‘Cachaco’ and 3% for ‘Williams’ cells. For all cell lines, a decrease in water content with 5-6% was obtained after increasing PEG concentration from 7.5% to 15% and also from 15% to 22.5%. As the water content decreases with increasing PEG concentration, PEG is indeed suitable to induce dehydration in banana cell cultures. Regeneration tests also showed that PEG negatively affected regrowth, and that this regrowth is variety-dependent and more affected by higher concentrations. In summary PEG can effectively be used to induce dehydration in Musa embryogenic cell cultures. The value of the in vitro model should be evaluated by research performed ex vitro, submitting plants from the same varieties to drought stress in the greenhouse and/or in the field. Cryopreservation In 2007, 108 accessions belonging to 19 genomic groups were successfully cryopreserved. For 12 accessions the genomic constitution is unknown. This brings the total to 527 accessions completely and definitively stored in liquid nitrogen. Also 192 accessions were shipped with a dry shipper to IRD and safely stored as “black box”. Molecular biology: Promoter tagging as a basis for developing cisgenic banana Based on a previously developed gene- and promoter tagging platform, a genome-wide screening strategy was performed for the identification and characterization of novel promoters in banana. Embryogenic Musa cell suspensions of the plantain were transformed with Agrobacterium tumefaciens containing a promoterless, codon-optimized luciferase (luc+). Tens of thousands of transgenic cell colonies were selected and screened. The frequency of low temperature-responsive cell colonies ranged between 0.17-1.69%. 94 colonies were regenerated to plantlets and screened throughout different regeneration stages. In total, twenty-four banana flanking DNA sequences were cloned in seven independently selected lines. So far we isolated and characterised a novel banana promoter. Understanding AMF biocontrol and its impact in banana-based cropping systems Fourteen intercrops with high mycorrhizal compatibility and low nematode susceptibility were selected. Effects of intercrops with different nematode susceptibility levels on nematode population build-up in an intercrop set-up were studied. Common bean looks promising as a banana intercrop for nematode control in a mixed cropping system. AMF and rhizobial colonisation both resulted in reduced nematode populations and improved growth of common bean. However, dual colonisation did not provide extra plant growth when compared to single AMF colonisation. Mechanisms responsible for the AMF bio-protection effect were unraveled: the split-root experiments indicate a combination of a locally and systemically induced effect. R. similis is less attracted to and penetrates less (+50%) in AMF colonized plant roots. Strong indications exist that root exudates play an important role in this process. However, based on the attraction bio-assays and exudates experiment it could not be determined whether the decreased penetration of the AMF colonised plants was partly due to a repellent effect of the root exudates. Once penetrated in the root, R. similis development and reproduction in the AMF-colonised roots goes slower than in non-colonised roots.

44

666... PPPRRROOOGGGRRREEESSSSSS WWWIIITTTHHH LLLEEEGGGUUUMMMEEE GGGEEERRRMMMPPPLLLAAASSSMMM---RRREEELLLAAATTTEEEDDD AAACCCTTTIIIVVVIIITTTIIIEEESSS

666...AAA... LLLEEEGGGUUUMMMEEE GGGEEERRRMMMPPPLLLAAASSSMMM DDDEEEMMMOOONNNSSSTTTRRRAAATTTIIIOOONNN AAANNNDDD

EEEVVVAAALLLUUUAAATTTIIIOOONNN

66..AA..11.. OONN--SSTTAATTIIOONN LLEEGGUUMMEE GGEERRMMPPLLAASSMM EEVVAALLUUAATTIIOONN On-station evaluation of bean, soybean, groundnut and pigeon pea germplasm started in March 2006 to identify potential varieties for testing at action site level with the various farmer associations. In these evaluation trials, specific measures were included to identify varieties tolerant to low-P conditions, and varieties with effective nodulation and fixation of atmospheric nitrogen. In September 2007, additional characterisation activities were started to fully characterize selected varieties, which is essential for homologation of the germplasm. Table 15 presents an overview of the various on-station germplasm evaluation and characterisation activities conducted at the ISAR and INERA stations.

Table 15: On-station legume germplasm evaluation since the project start; in March ’07, characterisation of germplasm was initiated to enable homologation of selected varieties.

station 2006 B (Mar’06-Jun’06)

2007 A (Oct’06-Jan’07)

2007 B (Mar’07-Jun’07)

2008 A (Oct’07-Jan’08)

Rubona- ISAR

CB, BB, SB - - -

Karama- ISAR

GN, CP, PP - - -

Nyagatare- ISAR

- - - BB, CB, SB (+characterisation)

Mulungu- INERA

CB, BB, SB, GN, CP, PP

CB, BB, SB - BB, CB, SB (+characterisation)

M’vuazi- INERA

BB, CB, GN, CP GN, CP BB, CB, SB (+characterisation)

GN, SB (+characterisation)

CB: climbing bean; BB: bush bean; SB: soybean; GN: groundnut; CP: cowpea; PP: pigeon pea. Biomass production, nodulation and atmospheric N fixation Specific observations on biomass production, nodulation and atmospheric N fixation were carried out in the trials at the various stations, for all of the species tested. Presented below is a selection of results obtained. Emphasis was given to soybean and groundnut germplasm because of their known capacity for soil fertility enhancement. Large differences between varieties were observed in nodulation, biological N fixation and biomass production. At all sites, a generally positive response to P fertilization was observed. At the ISAR-Rubona station, several soybean varieties showed higher

Photograph 11: Soybean germplasm evaluation trial at INERA-Mulungu (season 2008 A), Sud-Kivu, DRC.

45

nodulation and biomass production when supplied with TSP fertilizer (e.g., SB2, SB4, SB19, SB20); some varieties, however, did not respond (e.g., SB6, SB12) (Figure 17). An analysis using the 15N-isotopic dilution for soybean varieties grown at the INERA-Mulungu station showed that most of the newly introduced varieties are equally effective in fixing atmospheric N as the local variety (Figure 18). In most varieties, P fertilization increased the effectiveness of N fixation (proportion of N derived from the atmosphere).

Biomass production, nodulation and atmospheric N fixation are of interest, as they determine the potential of the varieties to improve soil fertility by supplying organic matter and atmospheric N that can benefit a subsequent crop. These characteristics therefore constitute a first researcher-defined criterion for variety selection. Several of the newly varieties are superior to the local variety in terms of biomass production, in particular soybean varieties SB19 and SB24 in Sud-Kivu and Rwanda, and SB20 in Bas-Congo.

Also within the 63 groundnut varieties tested (Spanish Bunch, Valencia and Virginia growth types), a wide variability in nodulation and biomass production was observed. Several varieties performed superiorly to the local variety (JL24) at INERA-M’Vuazi, Bas-Congo (Figure 19). Particularly varieties ICGV-SM93555 and ICGV-SM99551 were retained for further testing as they produced up to twice as much biomass as the local variety (both with and without additional P fertilisation. Grain yield and response to P application Grain and biomass yield were evaluated with and without P application, with the aim to identify varieties that are tolerant to low-P conditions and/or respond to P application. These observations were taken both on-station, and in germplasm evaluation trials in farmers’ fields. A multi-locational assessment is required to identify varieties that are

SB19

SB4

SB15SB20

SB14

SB17

SB2

SB6

SB12

SB17

SB15

SB2

SB20SB6

SB14

SB12

SB19

SB4

0

1000

2000

3000

4000

5000

0 100 200 300 400 500

nodulation [kg FM ha-1]

biomass yield [kg D

M ha-1

]

control

with TSP application

Figure 17: Left: Nodulation and biomass yield as affected by TSP application at 25 kg P ha-1 for selected soybean varieties grown at the ISAR-Rubona station, Rwanda; FM = fresh matter, DM = dry matter.

0

10

20

30

40

50

60

70

80

90

100

SB2 SB4 SB6 SB12 SB14 SB17 SB19 SB20 local

% N derived from atmosphere

with TSP application

control

Figure 18: Biological N fixation determined by 15N-isotopic dilution method as affected by TSP application at 25 kg P ha-1 for selected soybean varieties grown at the INERA-Mulungu station, Sud-Kivu, DRC; Imperial was used as the local variety.

46

consistently tolerant to low-P conditions and/or respond to P application. Data from the on-station trials will be merged with yields observed in farmers’ fields (with and without TSP fertilizer application) to identify promising varieties (currently on-going). Detailed studies will then be undertaken to unravel mechanisms of low-P tolerance in these selected varieties.

At the INERA-M’vuazi station, several soybean varieties showed to be tolerant to low-P conditions and at the same time responsive to P application (e.g., SB19, SB20, Bossier and SB44) (Figure 20). In addition, the variety SB44 produced high amounts of biomass without P application. Some varieties responded to P application but performed poorly under low-P conditions (e.g., SB24, 449/16/6), while others performed well under low-P conditions but did not respond to P application (e.g., SB46, SB51).

The tolerance to low-P conditions and responsiveness to P application constitute a second researcher-defined criterion for variety selection, as P deficiency is one of the major constraints for legume production in the areas (see also section 4.B).

JL24

0

1000

2000

3000

4000

0 5 10 15 20 25 30

nr of nodules (x 106 ha

-1)

biomass yield (kg D

M ha-1)

Spanish Bunch type(control)

Spanish Bunch type(with TSP application)

Valencia type (control)

Valencia type (with TSP application)

Figure 19: Nodulation and biomass yield as affected by TSP application at 25 kg P ha-1 for selected groundnut varieties (Spanish Bunch and Valencia type) grown at the INERA-M’vuazi station, Bas-Congo, DRC; JL24 is the local variety; DM = dry matter.

SB42

SB38

SB51

SB19

Duiker

SB44

Vuangi

SB9

SB24TGM1781

SB20

SB14

TGx814-26D

0

1000

2000

3000

4000

5000

6000

0 1000 2000 3000 4000 5000 6000

biomass yield in control [kg ha-1]

biomass yield with P application [kg ha-1]

non-responsive

susceptible to low-P

responsive


responsive

tolerant to low-P

non-responsive

tolerant to low-P

SB46

SB51

SB19

Duiker

SB44

Bossier

SB25

Vuangi

SB9

SB24449/16/6

TGM1781

SB20

SB15

SB2

SB14

Peka6

TGx814-26D

0

500

1000

1500

2000

0 500 1000 1500

grain yield in control [kg ha-1]

grain yield with P application [kg ha-1]

non-responsive


responsive


responsive

tolerant to low-P

non-responsive

tolerant to low-P

Figure 20: Biomass (top) and grain (bottom) yield for selected soybean varieties as affected by NPK application at 25 kg P ha-1 for selected soybean varieties grown at the INERA-M’vuazi station, Bas-Congo, DRC; local varieties tested include TGx 814-26D and Vuangi; quadrants classify varieties based to their tolerance to low-P conditions (along the X-axis) or their response to P application (along the Y-axis).

47

Legume variety characterisation In March 2007 in Bas-Congo, and in October 2007 in Rwanda, on-station trials were established to fully characterize a selection of soybean and bean varieties, retained after selection based on farmer and researcher criteria. Detailed measurements are being taken during 2 consecutive seasons on the physiology and the resistance to pests, diseases and environmental stresses. This information will be used to produce technical leaflets (“fiches techniques”) that are required for the homologation of new germplasm. These new varieties can then be introduced in national formal seed multiplication and dissemination systems.

66..AA..22.. LLEEGGUUMMEE GGEERRMMPPLLAASSMM EEVVAALLUUAATTIIOONN WWIITTHH FFAARRMMEERR

AASSSSOOCCIIAATTIIOONNSS AATT TTHHEE AACCTTIIOONN SSIITTEESS Bean and soybean germplasm evaluation trials were initiated in October 2006 with all associations in the action sites in the 4 mandate areas (Table 16). During the period 2007 A – 2008 A, a total of 44 farmer associations have been involved in germplasm evaluation activities across the 4 mandate areas. When variety selection was affected by abnormal drought or excessive rain, or when farmers insisted on evaluating the germplasm during multiple seasons, trials were repeated during one or two additional seasons. Groundnut and pigeon pea germplasm was included in season 2007 B. Cowpea germplasm poorly adapted to high altitude and was only tested with farmer associations in Bas-Congo. In season 2008A, germplasm evaluation trials were set up with a limited number of bean and soybean varieties through NGO partners with associations in satellite sites in Bas-Congo.

Table 16: Legume germplasm evaluation trials installed with farmer associations in between October 2006 and January 2008.

mandate area

2007 A (Oct’06-Jan’07)

2007 B (Mar’07-Jun’07)

2008 A (Oct’07-Jan’08)

BC BB, CB, SB: 8 associations SB: 2 associations

BB, CB: 6 associations GN: 8 associations CP: 8 associations

KK BB, CB, SB: 7 associations BB, CB, SB: 5 associations PP: 4 associations GN: 1 association

BB, CB, SB: 6 associations PP: 5 associations GN: 1 association

UM BB, CB, SB: 8 associations GN, PP: 2 associations

BB, CB: 1 association SB: 5 associations GN: 2 associations

SK

BB, CB, SB: 8 associations BB, CB: 4 associations GN: 8 associations

GN: 11 associations

Mandate areas: BC: Bas-Congo; KK: Kigali-Kibungo, UM: Umutara, SK: Sud-Kivu. Species: CB: climbing bean; BB: bush bean; SB: soybean; GN: groundnut; CP: cowpea; PP: pigeon pea.

Photograph 12: Soybean germplasm evaluation trial in Murambi, Umutara, Rwanda.

48

A number of measurements were taken in the germplasm evaluation trials (Annex 3). Firstly, grain yields were determined with and without P fertilizer application for soybean (to test for low-P tolerance), and with and without manure application (to test for resistance to low soil fertility) for the other species. Secondly, biomass samples were collected to assess atmospheric N fixation and potential for soil fertility improvement, and grain samples were collected to assess the genetic and environmental variability in micronutrient contents (particularly Fe and Zn) in bean grains (see section 4.A.). Furthermore, pest and disease occurrence was assessed and scored for bean varieties. Finally, participatory farmer evaluation events were organized to assess farmers’ selection criteria and preferences for the various legume species tested. Grain yield An analysis was conducted to assess the effects of site, input application and variety on grain yields of the various legume species evaluated. Generally, yields of the varieties were significantly site-dependent. In Sud-Kivu for example, some varieties performed well on the northern axis (e.g., SB4, SB6), but not on the southern axis (Figure 21). Other varieties, however, performed well across all sites (particularly Peka-6, SB19 and SB24). In all sites, new varieties performing equally well or better than the local variety could be identified. Soybean grain yields on the northern axis (Luhihi and Kabamba) were much larger than on the southern axis (Lurhala and Mwegerera). On the southern axis, soil fertility is very poor compared to the northern axis, and due to the high altitude (around 2000m) and short growing season, only early-maturing varieties were able to produce (e.g., SB19, Peka-6).

LurhalaSED

0

200

400

600

800

SB2

SB4

SB6

SB14

SB15

SB17

SB19

SB20

SB24

SB25

449/16/6

Bossier

Duiker

Ogden

Peka6

Soprosoy

TGM1781

local

grain yield [kg ha-1]

non-estimatable

non-estimatable

non-estimatable

non-estimatable

non-estimatable

non-estimatable

non-estimatable

MwegereraSED

0

200

400

600

800

SB2

SB4

SB6

SB14

SB15

SB17

SB19

SB20

SB24

SB25

449/16/6

Bossier

Duiker

Ogden

Peka6

Soprosoy

TGM1781

local


non-estimatable

non-estimatable

non-estimatable

non-estimatable

non-estimatable

non-estimatable

Kabamba SED

0

500

1000

1500

2000

2500

3000

3500

SB2

SB4

SB6

SB14

SB15

SB17

SB19

SB20

SB24

SB25

449/16/6

Bossier

Duiker

Ogden

Peka6

Soprosoy

TGM1781

local


non-estimatable

non-estimatable

LuhihiSED

0

500

1000

1500

2000

2500

3000

3500

SB2

SB4

SB6

SB14

SB15

SB17

SB19

SB20

SB24

SB25

449/16/6

Bossier

Duiker

Ogden

Peka6

Soprosoy

TGM1781

local


Figure 21: Soybean grain yields observed in the 4 action sites in Sud-Kivu during the 2007 A season; the interaction between variety and site was significant at P<0.05; TSP application significantly increased yields, independent of site or variety.

49

Input application (TSP for soybean, manure for beans) generally increased yields, independent of site or variety. Figure 22 presents bush bean yields observed in Kibungo during the 2007 A and 2007 B (averaged for both sites), as affected by manure application. All varieties responded significantly to manure application; bean grain yields increased from on average 1200 kg ha-1 to 1900 kg ha-1. Some varieties performed superior in the control (e.g., AFR619, AFR708 and ZAA5/2); these are known “BILFA” varieties, which are resistant to low soil fertility and yield 60 – 80% more grains than the local variety without manure application. Other varieties perform relatively well in the control and responded strongly to manure application (e.g., CODMLB003, CNF5520, MLB49-89A and HM21/7). Several varieties performed equally well or better than the local variety.

0

500

1000

1500

2000

2500

3000

3500

AFR619

AFR708

ARA4

BRB194

CIM

9314-36

CIM

9331-1

CNF5520

CODMLB003

CODMLB007

CODMLB033

CODMLB078

ECAPAN021

GNP585

HM21/7

LSA144

Maharagi soja

Marungi

MLB49-89A

M'Sole

Rab618

Rab619

Rjb7

T8426F11-6F

UBR(92)25

VEF88(40)L1P4T6

ZAA5/2

ZKA93-10m/95

local variety

grain yield (kg ha-1)

control with manure application (a)

(b)

Figure 22: Bean grain yields observed in Kibungo, Rwanda, averaged for the two action sites and seasons A and B 2007, as affected by manure application at 5 t DM ha-1; error bars represent the SED for comparison between varieties (a) and treatments (b); the interaction between variety and manure application was not significant.

Biomass yield The biomass of the various species was sampled in farmers’ fields at the 50% podding stage, when biomass accumulation is maximal. High biomass production is an important selection criterion as it is associated with the variety’s ability to improve soil fertility. Biomass samples are currently being analysed for δ15N, which allows estimating the proportion of N fixed from the atmosphere. This is particularly relevant for soybean because of its potential rotational benefits on a subsequent crop.

Photograph 13: High biomass-yielding soybean varieties in an evaluation trial in Nyakigando, Umutara, Rwanda.

50

Several of the newly introduced soybean varieties, particularly soybean lines SB 38 – SB54, produced large quantities of biomass (up to 6 t ha-1, Figure 23). These varieties also responded strongly to TSP application. However, although these varieties are favourable for soil fertility enhancement, they are little suited for the Rwanda and Sud-Kivu mandate areas; in several action sites, the too short growth season did not permit these long-duration varieties to reach maturity and produce grains.

0

2000

4000

6000SB2

SB4

SB6

SB9

SB14

SB15

SB17

SB19

SB20

SB24

SB25

SB38

SB39

SB42

SB44

SB45

SB46

SB47

SB49

SB51

SB54

449/16/6

Bossier

Duiker

Ogden

Peka6

Soprosoy

local

biomass yield [kg ha-1]

control

with TSP application (a)

(b)

Figure 23: Soybean biomass yields observed in Kibungo, Rwanda, averaged for the two action sites and seasons A and B 2007, as affected by TSP application; error bars represent the SED for comparison between varieties (a) and treatments (b); the interaction between variety and manure application was significant at P<0.1.

Other varieties such as SB19, SB24, SB25 and Peka-6 are more suitable as they have short- to medium-duration cycles and produce at the same time a reasonable amount of biomass and yield relatively high amounts of grains.

Apart from soybean, climbing beans also produce high amounts of biomass (up to 8 t ha-1). Several of the newly introduced varieties, particularly VCB81012, produce much higher amounts of biomass than the local variety. There are indications that climbing beans can have substantial rotational benefits on a subsequent cereal crop; this is currently being tested in some of the demonstration trials with the associations in the Kabamba action site in Sud-Kivu (see section 7.A.). Micronutrient contents in bean grains Measurements are currently being conducted on grain samples collected from the various bush and climbing bean evaluation trials to assess variability in Fe and Zn contents as affected by environment (site × input) and genetic (variety) effects. This is presented in detail under the process research activities in section 4.A.

Photograph 14: High biomass-producing climbing bean varieties in Umutara, Rwanda.

51

Farmer evaluations Initially, two evaluation events were organized in each legume evaluation trial – a first evaluation between flowering and podding, and a second after harvest – to examine whether farmer-preferred varieties can already be identified at an early stage. Male and female association members were separated and defined their criteria for selection. They then visited the trial and specified positive and negative traits of each variety, and finally selected five varieties which they scored according to their criteria. The questionnaire used for farmer evaluation is presented in Annex 4.

Farmers used a wide range of criteria for evaluating the germplasm. These criteria depended on the growth stage of the plants, differed between mandate areas and crop species, and were influenced by the sex of the evaluating farmers. Presented below are the evaluation criteria for beans across the mandate areas (Figure 24).

0

20

40

60

80

100

number of

flowers/pods

early m

aturity

germ

ination

resistance to

heavy rain

drought

resistance

resistance to

poor soil fertility

disease

resitance

biomass

production

relative importance index

at flowering-podding

0

20

40

60

80

100

yield

grain size

grain colour

grain decay

grain lustre

grain density

market

preference

resistance to

weevils taste

ease of cooking

early m

aturity

resistance to

heavy rain

drought

resistance

resistance to

poor soil fertility

disease

resitance


Rwanda

Sud-Kivu

Bas-Congo

after harvest

Figure 24: Relative importance of evaluation criteria defined by farmer groups between flowering and podding (left) and after harvest (right) in the mandate areas of the TSBF-CIAT project. The index was calculated as the frequency of the criterion (%) divided by its average rank.

Between flowering and podding, farmers commonly used the number of flowers (or pods) as principle criterion for the expected yield. In addition, farmers in Bas-Congo evaluated the ‘germination capacity’ of the varieties, which relates to the swiftness at which the variety grows in the early stages and its ability to out-compete weeds. In Rwanda, farmers appreciated early maturity and drought resistance because of the short growing season with unreliable rains. In Sud-Kivu, both high rainfall tolerance as well as drought resistance were favoured traits, as climate unpredictability is a major constraint for bean production. At the evaluation events after harvest, evaluation criteria comprised yield, followed by grain traits (size, colour, lustre, decay, density,…). Some of the criteria identified during the season were retained, but had much less weight in the choice of varieties to be retained. In most cases, between one and three out of the five varieties preferred at the flowering-podding stage were retained after the harvest evaluation, and rarely the same variety was most preferred at both evaluation stages.

Men and women evaluators used somewhat different evaluation criteria, and usually had at least 2 or 3 common varieties in their top 5 of preferred varieties. After the final

Photograph 15: Farmer evaluation of bean germplasm in Zenga, Bas-Congo, DRC

52

evaluation, a discussion session was organized where the association decided in plenary on a selection of maximally five bush bean, three climbing bean and five soybean varieties to be multiplied in the following season. For soybean, research teams suggested including at least one soybean variety with high biomass production, in cases where farmers opted to multiply early-maturing varieties only. Because of the performance of the varieties varies between sites and regions, and because farmers base their selection on differing criteria, varieties retained differed between mandate areas, action sites and even individual associations. Nevertheless, a number of varieties could be identified that were widely acceptable to farmers across sites (Table 17).

Table 17: Three most frequently retained bush bean, climbing and soybean varieties by farmer groups at the evaluation event after harvest for the different mandate areas (or regions within mandate area); numbers between brackets represent proportion of farmer groups that retained the variety.

region bush bean climbing bean soybean

BC

Lola (100%) PVO14/2 (67%) ZAA5/2 (53%)

Lola (100%) Tuta (100%) MLV06-90B (78%)

Vuangi (63%) SB19 (63%) TGx814-26D (78%)

KK, Bugesera

local variety (75%) Marungi (63%) ARA4 (50%)

AND10 (50%) G59/1-2 (50%) VCB81012 (50%)

not available yet

KK, Kibungo

AFR708 (80%) CNF5520 (80%) ZAA5/2 (60%)

AND10 (83%) local variety (83%) VCB81012 (50%)

not available yet

UM

local variety (71%) Maharagi-soja (57%) M’Sole (43%)

Kiangara (80%) MLV06-90B (60%) local variety (60%)

not available yet

SK, Katana axis ZKA93-10m/95 (55%) Marungi (45%) BRB194 (45%)

VCB81012 (75%) AND10 (75%) G59/1-2 (75%)

Peka6 (100%) SB24 (79%) Soprosoy (79%)

SK, Walungu axis Marungi (69%) ZKA93-10m/95 (56%) BRB194 (50%)

Kiangara (75%) LIB1 (75%) AND10 (50%)

Peka6 (67%) SB19 (67%) Soprosoy (67%)

Mandate areas: BC: Bas-Congo; KK: Kigali-Kibungo, UM: Umutara, SK: Sud-Kivu; Bugesera comprises action sites Mayange and Musenyi, Kibungo comprises Kabare and Gatore, the Katana axis in Sud-Kivu comprises action sites Luhihi and Kabamba, the Walungu axis comprises Lurhala and Mwegerera.

53

666...BBB... LLLEEEGGGUUUMMMEEE SSSEEEEEEDDD MMMUUULLLTTTIIIPPPLLLIIICCCAAATTTIIIOOONNN

66..BB..11.. OONN--SSTTAATTIIOONN LLEEGGUUMMEE SSEEEEDD MMUULLTTIIPPLLIICCAATTIIOONN On-station seed multiplication activities started in October 2006, with the purpose of providing seeds for research and dissemination activities. Initially, seed was produced for a large number of varieties through the evaluation activities. When promising varieties were identified, these were multiplied on a larger scale. Currently, seed stocks are maintained at the national research centres to support activities within and beyond the project. Table 18 presents an overview of the on-station germplasm multiplication activities conducted at the ISAR and INERA stations.

Table 18: On-station seed multiplication of legume germplasm since the project start.

station 2007 A (Oct’06-Jan’07)

2007 B (Mar’07-Jun’07)

2008 A (Oct’07-Jan’08)

Nyagatare-ISAR - - BB, CB, SB Mulungu-INERA - BB, CB, SB, FB, GP BB, CB, SB M’vuazi-INERA BB, CB, SB, GN, CP BB, CB, SB, GN, CP BB, CB, SB, GN, CP

Species: CB: climbing bean; BB: bush bean; SB: soybean; GN: groundnut; CP: cowpea; FB: faba bean; GP: garden peas.

A strategy has been developed to accelerate diffusion of promising varieties by multiplying and disseminating seed through multiple channels, both formal and informal, starting from the seed base produced at the ISAR and INERA stations (Figure 25). While formal seed multiplication is led by the national legume programs, informal seed multiplication and dissemination is primarily facilitated by NGO partners in collaboration with the ISAR and INERA legume programs. National seed services (RADA and SENASEM) are involved in training and accreditation of farmer associations and individual seed multipliers. A detailed action plan, indicating the different sub-activities, responsibilities, timeline and reporting format is presented in Annex 5).

Currently, selected promising varieties are being fully characterized by the national legume programs, as a first step in the formal seed multiplication channel. In Rwanda, the ISAR bean and soybean program are collecting necessary information for the production of technical fiches, which are required for officially inscribing new varieties in the national catalogue. At the same time, the legume programs are producing the initial clean seed source, which will be multiplied by the Unité semencière de l’ISAR. Once the new varieties are homologated, foundation seed will be supplied to the Rwanda Agricultural Development Authority (RADA), who will formally distribute the varieties through its national network of accredited seed multipliers. In DRC, formal seed multiplication has not yet been initialised; however, SENASEM (Service National de Semence) has been actively involved in the seed multiplication through informal channels, by providing technical backstopping, training and accreditation to multiplying farmer associations.

54

Multiplicateurs individuels de semence

Associations paysannes dans les sites d’action

Instituts nationaux (SENASEM / RADA)

Membres qui se spécialisent comme multiplicateurs

Associations paysannes dans les sites satellites

Commerçants agriculteurs (associations ou individus)

Diffusion paysan ↔ paysan

INERA/ISAR semence de base

INERA/ISAR + NGOs

NGOs

INERA/ISAR + NGOs

INERA/ISAR + NGOs

INERA/ISAR

Réseau national de multiplicateurs certifiés

formation par SENASEM/RADA

formation par SENASEM/RADA

voie formelle voie informelle

Figure 25: Seed system strategy and involvement of various partners for rapid seed multiplication and dissemination through formal and informal channels.

66..BB..22.. FFAARRMMEERR AASSSSOOCCIIAATTIIOONN--LLEEDD LLEEGGUUMMEE SSEEEEDD

MMUULLTTIIPPLLIICCAATTIIOONN Legume seed multiplication through informal channels was initialised through the farmer associations in the action sites and facilitated by NGO partners. After selection of promising varieties during the farmer evaluation after harvesting the legume evaluation trials, a discussion session was organised with each farmer association to decide on practical aspects of seed multiplication. Generally, associations decided to organize seed production as a group activity, mostly in communal fields. The activity started in March 2007 with 38 farmer associations (Table 19) in action and satellite sites. In October 2007, an additional 5 associations were involved.

Photograph 16: A member of the farmer association Dutabarane multiplying an improved climbing bean variety in Gatore, Kibungo, Rwanda.

55

Table 19: Legume germplasm multiplication activities conducted by farmer associations in between March 2007 and January 2008.

mandate area

2007 B (Mar’07-Jun’07)

2008 A (Oct’07-Jan’08)

BC BB: 10 varieties, 5 associations CB: 4 varieties, 4 associations SB: 8 varieties, 21 associations GN: 1 variety, 2 associations CP: 2 varieties, 21 associations

BB: 10 varieties, 8 associations CB: 4 varieties, 6 associations SB: 8 varieties, 21 associations GN: 2 varieties, 4 associations CP: 2 varieties, 21 associations

KK BB: 11 varieties, 3 associations CB: 3 varieties, 1 association SB: 4 varieties, 1 association

UM BB: 10 varieties, 5 associations CB: 5 varieties, 5 associations

BB: 12 varieties, 5 associations CB: 4 varieties, 3 associations SB: 9 varieties, 5 associations

SK BB: 14 varieties, 8 associations CB: 4 varieties, 3 associations SB: 9 varieties, 11 associations

BB: 14 varieties, 11 associations CB: 4 varieties, 5 associations SB: 9 varieties, 11 associations

Mandate areas: BC: Bas-Congo; KK: Kigali-Kibungo, UM: Umutara, SK: Sud-Kivu. Species: CB: climbing bean; BB: bush bean; SB: soybean; GN: groundnut; CP: cowpea.

During the first season, members of the farmer associations involved were trained on technical aspects of seed production (Photograph 17), following the guidelines in the CIAT manual for small-scale seed production (David, 1998). Prior to these training sessions, CIALCA staff were acquainted with the guidelines and technical requirements of seed multiplication during a three-day training in February 2007 in Butare, Rwanda (Annex 6). In Rwanda, the ISAR legume program initially provided training and facilitation to farmer associations, but currently NGO partners (World Vision, RWARRI, RDO and

RHEPPI) are increasingly more involved in this activity. In DRC, seed multiplication activities are led by NGO partners (DIOBASS in Sud-Kivu; BDD and APRODEC in Bas-Congo), with technical backstopping by INERA and SENASEM. The first training sessions focused on agronomic practices and regulations for seed multiplication (planting in line, spacing, purging and disease management) (Annex 7). After the first season, an evaluation event was organized with each association in order to discuss successes and identify bottlenecks. Constraints were in the first place related to land availability and availing sufficient organic inputs (compost or manure) for communal activities.

Capacities for seed multiplication differed between farmer associations. Some associations were committed to continue multiplying in group, while others opted for a decentralised system, whereby association members receive a portion of seed, multiply in their own field, and return a proportion of the seed produced to the association at the end of the season. Terms and conditions are discussed within each association, and facilitated by NGO members. Performance in each individual multiplication field is being documented by a technical team appointed in the association, or by a local animateur engaged by the NGO partner (Annex 8). Some of the bottlenecks identified in the first season were

Photograph 17: Training session on seed production techniques organized by DIOBASS and SENASEM with farmer association representatives on the Walungu axis, Sud-Kivu, DRC.

56

addressed in the second season, through closer follow-up, providing technical advice and assisting in discussions to improve the organisation of the activity. In Sud-Kivu, for example, SENASEM officers visited all fields of multiplying individuals or associations, made observations and recommendations, and finally conferred official certificates to the seed producers (Photograph 18). Depending on the progress and organisational skill of the association, additional training is provided (e.g., related to post-harvest management).

Currently, about half of the associations have reached a level where seed quantities suffice to satisfy the seed needs of all members within the association (between 50 – 100 kg per variety). In the on-going season (season B 2008), activities are planned to promote and disseminate the seed produced through informal channels, primarily by farmer-to-farmer diffusion. Exchange visits will be organized between multiplying farmer associations, in preparation of farmer field days. During these events, key stakeholders (representatives from farmer associations inaction and satellite sits, local and regional NGOs, the national seed service (RADA or SENASEM), local authorities and policy makers, journalists and the local radio) will be invited to promote the improved varieties. Leaflets and farmer fiches will be distributed during these events (Figure 26). NGO partners will play a crucial role in linking the multiplying farmer associations (or individuals) with potential buyers and creating incentive for commercialisation of seed production.

AFR708 haricot nain

Les paysans ont apprécié : � le haut rendement � la précocité � l’adaptation au milieu � la grosse taille des graines � la couleur des graines (rouge-tachetée) � l’apparence des graines � la cuisson facile/rapide � la préférence au marché

Rendement potentiel : haut [2500 kg ha-1]

(au milieu paysan)

Maturité : précoce [75-85 jours]

Taille des graines : large [50g pour 100 graines]

Production de biomasse : moyenne

Croissance sur sol pauvre : bonne

Résistance aux fortes pluies : moyenne

Résistance à la sécheresse : pauvre

Résistance aux maladies : bonne

Gestion : � Semez en ligne : 40cm entre les lignes et 10-15cm sur la ligne (une graine par poquet).

� Pour un rendement optimal, appliquez 50kg de fumier ou compost par are. Cette variété donne aussi bien sans application d’intrants.

Cette variété est bio-fortifiée : elle est riche en matières minérales et bonne pour la santé.

Figure 26: An example of a farmer fiche of an improved (bio-fortified) bush bean variety, presenting technical and farmer-preferred characteristics identified during the germplasm evaluation trials.

Photograph 18: SENASEM officer making observation in a soybean multiplication field in Kabamba, Sud-Kivu.

57

777... PPPRRROOOGGGRRREEESSSSSS WWWIIITTTHHH NNNAAATTTUUURRRAAALLL RRREEESSSOOOUUURRRCCCEEE

MMMAAANNNAAAGGGEEEMMMEEENNNTTT---RRREEELLLAAATTTEEEDDD AAACCCTTTIIIVVVIIITTTIIIEEESSS

777...AAA... NNNAAATTTUUURRRAAALLL RRREEESSSOOOUUURRRCCCEEE MMMAAANNNAAAGGGEEEMMMEEENNNTTT OOOPPPTTTIIIOOONNNSSS FFFOOORRR

LLLEEEGGGUUUMMMEEE---BBBAAASSSEEEDDD SSSYYYSSSTTTEEEMMMSSS

77..AA..11.. OOVVEERRVVIIEEWW OOFF OOPPTTIIOONNSS CCUURRRREENNTTLLYY BBEEIINNGG TTEESSTTEEDD Testing of natural resource management (NRM) options started in March 2007 (season 2007 B) or in September 2007 (season 2008 A). These options were chosen to address some of the major constraints identified, and were targeted towards the specific conditions in the regions and action sites. Partners as well as farmer groups were involved in the process of selecting and developing options. In most sites, proven NRM options that were successful elsewhere (improved agronomic practices and nutrient management in legume-cereal rotation or intercropping, and cassava-legume intercropping systems) were immediately implemented as demonstration trials with farmer associations. Other technologies, requiring additional research and adjustment to the local conditions, were implemented on-station before engaging with farmer associations. These include technologies for soil erosion control, tested in Sud-Kivu (DRC), and rain water harvesting in combination with efficient nutrient management to counteract seasonal drought spells in Umutara (Rwanda). At present, a total number of 56 farmer associations are actively involved in the testing, management and evaluation of demonstration trials in the various action sites across the 4 mandate areas (Table 20). In addition, 50 households in the Nyakigando action site (Umutara, Rwanda) started implementing adaptation trials in September 2007 (season 2008 A) after having appraised the technologies demonstrated earlier in legume-cereal intercropping systems (Photograph 19).

Presented below are selected results for NRM options on erosion control, improved cassava-legume intercropping, and soil fertility amendment on the soils around Walungu (Sud-Kivu).

Photograph 19: An improved bean-maize intercropping system, demonstrated in Kabarore, Umutara, Rwanda.

58

Table 20: An overview of NRM research, demonstration and adaptation trials conducted in 2007 in the various action sites of the TSBF-CIAT project. NRM option acronym trial type location period involvement of farmer associations

Erosion control using legume hedgerows and reduced tillage as alternatives to terrace construction

“ERO-1”

on-station (research)

on-station at Mudaka (Sud-Kivu)

installed in March 2007 (currently running the 3rd season)

on-station activity but the local farmer community is involved in the management of the trial.

Rainwater harvesting and interaction with nutrient management to counteract seasonal drought spells

“WANU-1” on-station (research)

on-station at ISAR-Karama and ISAR-Nyagatare (Rwanda)

installed in March 2007 (currently running the 3rd season running)

on-station activity – the activity aims to identify promising options before engaging with farmer associations

Improved agronomy options in cassava-legume intercropping systems

“CAS-1” demonstration trials

2 locations, in Zenga and Nkamu in Bas-Congo

installed in April 2007 (on-going)

2 farmer associations are involved in the management and evaluation of the demonstrated options.

Improved agronomy options in cassava-legume intercropping systems


6 locations, in Kabamba in Sud-Kivu

installed in September 2007 (on-going)


Improved soil fertility management in cassava systems


2 locations, in Kisantu and Mbanza-Nzundu in Bas-Congo

installed in April 2007 (on-going)


Erosion control using leguminous and non-leguminous forage hedgerows as an alternative to terrace construction

“ERO-2” demonstration trials

6 locations, in action sites in Sud-Kivu

installed in March 2007 (on-going)

6 farmer associations are involved in the evaluation of the forage species. The trial has been combined with multiplication of improved cassava germplasm.

Soil fertility amendment using various organic and inorganic inputs on poor soils

“FER-1” demonstration trials

8 locations, in Mwegerera and Lurhala in Sud-Kivu

installed in March 2007 for one season (concluded)

6 farmer associations have been involved in the testing of soil fertility management options.

Improved cereal-legume intercropping options “SYS-1” demonstration trials

8 locations, in action sites in Umutara, Rwanda

installed in September 2006 (currently running the 3rd season)

8 farmer associations have been involved in the management and evaluation of the demonstrated options.

Demonstration of rotational benefits of high-biomass- yielding legume varieties and the micro-dose fertilizer technique

“SYS-2” demonstration trials

6 locations, in Luhihi in Sud-Kivu



Demonstration of rotational benefits of soybean and Mucuna on a subsequent maize crop

“SYS-3” demonstration trials

30 locations, in Lemfu in Bas-Congo

installed in October 2007 (on-going)

30 farmer associations (through the network of NGO partner BDD) are involved in the management and evaluation of the demonstrated options.

Adaptation of cereal-legume intercropping options

“ADA-1” adaptation trials

50 locations in Nyakigando, Umutara, Rwanda


50 individual households are testing and adapting the previously demonstrated options in their own farms.

59

77..AA..22.. SSOOIILL CCOONNSSEERRVVAATTIIOONN TTEECCHHNNOOLLOOGGIIEESS TTEESSTTEEDD IINN SSUUDD--KKIIVVUU ((““EERROO--11”” AANNDD ““EERROO--22)) In Sud-Kivu, high population has driven agriculture towards fragile land and soil erosion has become one of the major threats to agricultural production (Photograph 20). Legume-grown fields are predominantly located on lands with strong slopes and commonly unprotected against erosion (see section 3.C.1.). Inadequate soil conservation measures have given rise to rapid loss of topsoil and land degradation. In Rwanda, effective policies and community work arrangements are in place for large-scale terrace construction and combating soil erosion. These are however absent in Sud-Kivu, where the situation calls for alternative, less labour-intensive technologies that are adoptable by individual households.

Two trials were set up in March 2007 to evaluate alternative options for combating soil erosion. A first trial (“ERO-1”) aimed to compare the effectiveness in conserving soil of reduced tillage and planting hedgerows of a leguminous perennial, Calliandra callothyrsus, with the construction of physical terraces. This trial was set up on-station in 3 replicates, on a site with a strong slope (41 %), following a complete factorial design with factors (i) terrace construction, (ii) tillage, and (iii) Calliandra hedgerows. A second trial (“ERO-2”) was installed in 6 sites in farmers’ environment and aimed to examine the adaptability of various forage species when grown as hedgerows on representative slopes, and to obtain farmers’ feedback on the adoption potential of these forages.

On-station testing of soil conservation measures (“ERO-1”) In this trial, the soil is cropped with soybean in rotation with maize, and the short- and long-term effects of reduced tillage, planting Calliandra hedgerows, and installing physical terraces on crop production and soil conservation are assessed. The measurements conducted in this trial include: crop grain yield, changes in slope and soil loss, soil water profiles, and changes in soil fertility. A detailed description is given in the trial protocol (Annex 9). Presented below are selected results from the first and second season, grown with soybean and maize, respectively.

Extension programs recommend a vertical distance of 1.6 m between two contour lines for terrace construction or hedgerow planting. This translates into a plot width of 4 m for the given slope at the trial site. The installation of terraces by embanking the soil up-hill and hedgerow planting entail a loss of surface area available for cropping, equal to 27% and 20%, respectively

Photograph 20: Severe soil degradation in land with a high slope cropped with beans in Sud-Kivu, DRC.

Photograph 21: Soybean cropping without (top) or with (bottom) physical embankments.

60

(Photograph 21, Figure 27). This implies that grain yields per unit area need to increase by 37% and 25%, respectively, to compensate for these area losses, and to justify the use of these techniques by farmers.

In the first season, soybean grain yields obtained per plot (dimensions: length = 5m, vertical interval = 1.6 m) were significantly lower in plots with physical terraces (Figure 28). This was primarily related to the loss of surface area available for cropping. Plots without physical embankments could generally hold 5 soybean lines while after terrace construction, plots could only hold 4 soybean lines. However, terrace construction also reduced yields per unit area (630 kg ha-1 vs. 770 kg ha-1 without terraces). Most likely, the soil embankment brought unfertile, acid subsoil to the surface, which negatively affected crop performance. Reduced tillage did not affect soybean yields when terraces were installed or Calliandra hedgerows were planted, but higher yields were observed in plots without terraces and hedgerows. Planting of Calliandra hedgerows only decreased yields in plots without tillage and terrace construction.

Soil erosion was significantly reduced by terrace construction and Calliandra hedgerow planting (Figure 29); tillage management did not affect soil loss. During the first month after planting (2nd season), the soil loss amounted to almost 1 kg m-2in plots without conservation measures, which approximates a loss of 1 mm of the soil profile. This soil loss was reduced by 80% in plots with physical embankments. Calliandra hedgerows were less effective in reducing soil erosion. Calliandra initially grows slowly and the

hedgerows are at present not yet fully developed. Further measurements are required to assess the effectiveness in soil conservation in the longer term.

SED (a)

SED (b)

0

5

10

15

20

25

without Calliandra with Calliandra

terrace surface area (m

2)

without physical embankments

with physical embankments

Figure 27: Plot surface area (plot length = 5m and vertical interval = 1.6m) as affected by terrace construction and hedgerow planting; error bars represent SED, comparing (a) with and without terraces, and (b) with and without hedgerows.

SED

0.0

0.5

1.0

1.5

2.0

trad. tillage zero tillage trad. tillage

Calliandra

hedgerows

zero tillage

Calliandra

hedgerows

grain yield (kg per terrace)



Figure 28: Grain yields obtained per plot (for a plot length of 5m and a vertical interval of 1.6m) as affected by tillage, terrace construction and hedgerow planting; the error bars represents SED.

SED (a)

SED (b)

0.0

0.2

0.4

0.6

0.8

1.0

without Calliandra with Calliandra

soil loss (kg m

-2)



Figure 29: Soil loss during the first month after planting (2nd season) as affected by hedgerow planting and terrace construction; error bars represent SED, comparing (a) with and without terraces, and (b) with and without hedgerows.

61

As a preliminary conclusion, terrace construction is most effective in the short-term to reduce soil erosion, but the reduction in surface area and upturning of subsoil considerably reduces crop yields. Planting Calliandra hedgerows has less influence on crop yield, but is also less effective in reducing soil erosion. Long-term effects on crop yield and soil stabilisation need to be assessed to appraise Calliandra hedgerow planting as an alternative to terrace construction. Evaluation of hedgerow forages for erosion control in farmers’ environment (“ERO-2”)

Ten forage species (Brachiaria brizantha, Brachiaria decumbens, Brachiaria ibrido, Brachiaria ruziziensis, Calliandra calothyrsus, Leucaena diversifolia, Penisetum purpureum, Setaria sphacelata, Tithonia diversifolia and Tripsacum laxum) were established as hedgerows in 6 sites (Lurhala, Mwegerera, Luhihi, Kabamba, Cijingire and Mudaka). Measurements included survival rate and biomass accumulation, and slope and soil accumulation. A detailed trial protocol is presented in Annex 10.

Farmers evaluated the forages about 10 months after establishment (Photograph 22). The procedure used for forage evaluation was similar as for the legume germplasm evaluation (Annex 11). In each site, the male and female members of the association were separated and first defined their criteria for evaluation. They then visited the trial and specified positive and negative traits of each variety, and finally selected five forages which they scored according to their criteria.

The criteria defined by farmers primarily comprised the use of the biomass as a green manure, effective rooting (as an indicator for its capacity to contain the soil), use as a forage and production of high amounts of biomass (Figure 30). These criteria were mentioned by at least 80% of the evaluating farmer groups. Other minor criteria included the potential for using in construction (mainly as roofing for houses), being non-competitive with crops (shading) and producing poles for climbing bean cultivation.

Photograph 22: Farmers evaluating different forages in Lurhala, Sud-Kivu.

produces poles

non-shading

use for construction

drought-resistant

good rooting

green m

anure

good forage

biomass production

non-competitive

0

10

20

30

40

50

criterion importance index

Brachiaria briz.

Setaria

Calliandra

Tripsacum

Tithonia

Penisetum

Brachiaria ruzizi.

Brachiaria decum.

Leucaena

Brachiaria ibrido

0

10

20

30


Figure 30: Left: Farmer-defined criteria for evaluation of forages. Right: Forage species selected by farmer groups; the relative importance index was calculated as the frequency of the criterion or forage species (%) divided by its average rank.

62

Farmers selected and scored the forages based on these criteria. Tithonia, Tripsacum and Calliandra were the most preferred forages, selected in the top 5 by 67, 75 and 92 % of the evaluating farmer groups, respectively. Tithonia and Tripsacum were generally ranked higher (average rank = 2) than Calliandra (average rank = 3). Farmers appraised Tithonia and Setaria as the most effective for soil erosion control, followed by Calliandra, Tripsacum and Penisetum.

77..AA..33.. IIMMPPRROOVVEEDD AAGGRROONNOOMMYY AANNDD SSOOIILL FFEERRTTIILLIITTYY

MMAANNAAGGEEMMEENNTT IINN CCAASSSSAAVVAA--LLEEGGUUMMEE SSYYSSTTEEMMSS

In Bas-Congo, and to a lesser extent in Sud-Kivu, legumes are predominantly cultivated in association with cassava (see section 3.C.1.). Options for improved agronomic practices and nutrient management in these systems were discussed with partners and farmer groups. Three sets of demonstration trials were then installed, including options for improving productivity through (i) improved germplasm, (ii) alternative spacing, (iii) application

of locally available green manures and/or fertilizer, (iv) reduced tillage, (v) planting alternative legume species, and (vi) planting climbing beans during the second season. The first set of trials was installed in April 2007 with two farmer associations in Bas-Congo (two sites, 3 replicates per site) and focuses on improved agronomic practices (“CAS-1”, Annex 12). The second set likewise focuses on improved agronomic practices and was installed in September 2007 with 3 farmer associations in 6 sites in the Kabamba action site in Sud-Kivu (“CAS-2”, Annex 13). The third set of trials focuses specifically on nutrient input management using green manures and/or fertilizer to improve cassava production, and was installed in April 2007 with two farmer associations in Bas-Congo (two sites, 3 replicates per site (“CAS-3, Annex 14). Specific measurements included legume biomass and grain production, cassava tuber production and tuber quality/tradability, and detailed labour assessments. Farmers evaluated the trial at different stages: firstly at peak biomass production of the legumes (2 months after planting), secondly after harvest of the legumes (4 months after planting), and finally at the cassava harvest (12 months after planting).

Biomass production in the traditional system was generally low (for example 2 t DM ha-1 in Nkamu, Bas-Congo; Figure 31). Biomass production can be considerably increased by fertilizer application or replacing the traditional legume species (groundnut in Bas-Congo, beans in Sud-Kivu) by soybean. In Nkamu (Bas-Congo), biomass yield was three times higher for soybean than for groundnut. This has important implications for soil fertility management, as a biomass production of 6 t DM ha-1 may supply a net input of 30 – 40 kg N ha-1 (to be verified by BNF measurements), and entail significant rotational benefits for subsequent crops (to be verified in successive seasons).

Photograph 23: A demonstration trial on improved cassava-legume intercropping in Kabamba, Sud-Kivu, DRC.

63

Legume grain yields can also be substantially increased using improved agronomic practices and fertilizer application. In Kabamba (Sud-Kivu), for example, pod yields for the traditional legume (beans) were increased by 50% using an alternative intercropping spacing that favours the legume; fertilizer application doubled yields compared to the traditional system (Figure 32). Soybean generally performed very well in association with cassava, when planted at a spacing of 2 × 0.5m, allowing sufficient space for 4 lines of soybean (400,000 plants per hectare) between cassava lines (10,000 plants per hectare).

Farmers have currently evaluated the trial twice, at podding and harvest of the legume. At the podding stage, farmers primarily evaluated based on the production of biomass, the number of pods or flowers, the lustre (greenness) of the leaves, and the presence of diseases (Figure 33, Annex 15). Farmers particularly preferred the option with cassava planted at 2 × 0.5m and intercropped with soybean (and to a lesser extent with an improved bean variety) as well as the option with NPK application. These trials have attracted large interest by the farmer associations and neighbouring farming communities. At present (season 2008 B), this activity has proceeded into an adaptation phase. Individual members of farmer associations have been given access to improved legume varieties, cassava cuttings and fertilizer, and have been trained to test and adapt the demonstrated options in their own fields.

SED (b)

SED (a)

0

2000

4000

6000

groundnut groundnut soybean

biomass yield (kg D

M ha-1)

control controlNPK

Figure 31: Biomass yield obtained for groundnut (with and without NPK applied at 100 kg ha-1) and soybean grown in association with cassava in Nkamu, Bas-Congo, DRC; error bars represent SED for comparison of effects of NPK application (a) and legume species (b).

NPK application

2x0.5m spacing

improved variety

1x1m spacing

traditional

SED

0

1000

2000

3000

pod yield (kg ha-1)

Figure 32: Pod yields for common beans obtained by successively changing the spacing (1 × 1m), variety, spacing (2 × 0.5m) and applying NPK at 100 kg ha-1 in Kabamba, Sud-Kivu, DRC.

soybean intercrop

groundnut intercrop

NPK application

2x0.5m spacing

improved variety

1x1m spacing

traditional0

5

10

15

20

25

farmer preference (%)

women

men

Figure 33: Farmer preference of improved agronomic practices demonstrated in Kabamba, Sud-Kivu, as compared to the traditional cassava-legume intercropping system by successively changing the spacing (1 × 1m), variety, spacing (2 × 0.5m), applying NPK at 100 kg ha-1 and replacing the common legume (beans) for groundnut or soybean.

64

77..AA..44.. OOPPTTIIOONNSS FFOORR SSOOIILL FFEERRTTIILLIITTYY AAMMEENNDDMMEENNTT OONN TTHHEE

WWAALLUUNNGGUU AAXXIISS IINN SSUUDD--KKIIVVUU In the germplasm evaluation trials, a generally poor crop performance was observed in the action sites on the southern axis (Walungu axis) in Sud-Kivu (see section 6.A.). The region is ill-reputed for its acid and unfertile soils. During the baseline and characterisation studies, farmers expressed low soil fertility as one of the major constraints for crop production (see section 3.C.1.). Farmers are limited in their options for soil fertility restoration. Due to the limited cattle numbers, use of farm yard manure (FYM) is scarce, and chemical fertilizer is absent in the region. In addition, preliminary studies and observations in farmers’ fields suggested potential micronutrient deficiencies. A set of exploratory trials (“FER-1”) was installed with 6 farmer associations in the two action sites to investigate the potential of increasing crop yields using FYM (5 t DM ha-1), NPK (20 kg P ha-1), mavuno fertilizer (NPK enriched with micronutrients, 20 kg P ha-1), lime (4 t ha-1), and combinations of these resources. Application of Tithonia leaf residues (5 t DM ha-1) was included as an option that is relatively readily available to farmers. Climbing beans and maize were selected as test crops (Annex 16).

Grain yields were significantly increased by all inputs, except by lime application (Figure 34). However, both maize and bean yields generally remained much below the potential of the crops. Highest yields observed in one of the more fertile fields were 2.7 t ha-1 for maize and 1.8 t ha-1 for climbing beans (in treatments with combined application of FYM and fertilizer). There was a significant interaction between species, treatment effects and the soil fertility status; in the poor fields, maize failed in all treatments, while climbing beans responded significantly to FYM and fertilizer application.

Only in one out of the eight sites, a striking difference could be visually observed between the treatments with NPK application and mavuno application, which suggests a nutrient other than N, P or K was limiting crop growth (Photograph 24). In other sites, responses to NPK and mavuno were comparable. However, micronutrient deficiency may have been masked by P deficiency (see section 4.B.). Currently, specific analyses are conducted on young bean leaves and maize ear leaves to identify potential micronutrient deficiencies.

(a)

(b)

0

500

1000

1500

control

NPK

mavuno

FYM

FYM+NPK

FYM+mavuno

lime

lime+NPK

lime+mavuno

Tithonia

grain yield (kg ha-1)

maize

climbing beans

Figure 34: Grain yields for climbing beans and maize as affected by different inputs (NPK fertilizer, mavuno fertilizer, farm yard manure, lime and Tithonia leaf residues) on the Walungu axis in Sud-Kivu, DRC.

65

The results from these field trials instigated a series of pot trials on a wider range of soils from the area to rapidly investigate major nutrient deficiencies occurring on the Walungu axis. Preliminary results for these trials are presented in section 4.B.

777...BBB... NNNAAATTTUUURRRAAALLL RRREEESSSOOOUUURRRCCCEEE MMMAAANNNAAAGGGEEEMMMEEENNNTTT OOOPPPTTTIIIOOONNNSSS FFFOOORRR

BBBAAANNNAAANNNAAA---BBBAAASSSEEEDDD SSSYYYSSSTTTEEEMMMSSS

77..BB..11.. OONN--FFAARRMM TTRRIIAALLSS The selection of technologies for on-farm testing with farmers was done through feedback workshops, which have taken place in Rwanda and the DRC. Feedback workshops for Burundi are planned for early 2008. At each site where the diagnostic survey was conducted 50 persons were invited to participate in the feedback workshops. The participants included: 30 farmers who participated in the survey at the sites as well as NGO and farmer association representatives and local leaders. Invitations aimed to achieve gender balance in the participation of the feedback workshops. The form of the feedback workshops was as follows:

• The results of the diagnostic survey were presented to the participants.

• Building consensus on these perceived constraints. In some cases more constraints were added to the list. The new list of constraints were ranked using the pair-wise ranking technique.

• Farmers suggested possible solutions to their perceived constraints. In some instances the farmers solutions coincided with interventions that researchers wanted to test with farmers. Research also made contribution to the list of solutions/interventions that could be tested.

• The solutions/interventions were then assessed for their sustainability in terms of availability of resources and capacity to manage the option over time, technical and social feasibility (technical possibility and social acceptability), Cost, effectiveness in solving the constraint and time to success (time that the technology bear fruit).

• For each action site participants of the feedback workshops selected their considered best technologies for testing.

Photograph 24: Maize growth as affected by input application in a demonstration trial in Lurhala, Sud-Kivu, DRC; from left to right: control, NPK(17:17:17) at 20 kg P ha-1 and mavuno (NPK enriched with micronutrients) at 20 kg P ha-1.

66

The following technologies were identified and adopted for on-farm trials: 1. Mulch applications – type and quantities can vary depending on local availability – the

trial will consist of treatments, namele (i) soil tillage and mulch removal, (ii) self-mulch and zero-tillage, and (iii) as the second treatment and application of external mulch at a rate of 25 t dm/ha. Farmers are allowed to continue planting of intercrop beans, but this should not interfere with the mulch and tillage practices as described in these treatments.

2. Manure application – different manure management options will be tested, this will include (i) application of fresh manure in small pits in between banana plants, versus (ii) open storage of manure besides the kraal before application. Quantities of manure applied in terms of dm/ha will be similar in both treatments.

3. Soil and water conservation measures – this includes the introduction of (i) contour bunds in combination with contour mulching, (ii) with and without the planting of leguminous intercrops on the contour bunds.

4. General plantation sanitation/husbandry – Scientist will show farmers how to improve crop husbandry in a small section of his field.

Protocols for the above mentioned trials have been developed and will be similar across all sites (Annex 17). Through the feedback workshops, farmers could indicate whether they would volunteer to host any of the above-indicated trials. In general 2-5 farmers per site per trial type will be involved, resulting in some 10-15 on-farm trials per action site. Besides our focus on banana productivity, measurements will be done on the productivity of legume intercrops if farmers have these. Hence, legume intercropping will not be imposed, but if farmers do practice legume (i.e. mostly common bean) intercropping, then measurements will be done on the legume intercrop to quantify the impact of zero-tillage and mulching on the overall productivity of the plot.

77..BB..22.. OONN--SSTTAATTIIOONN TTRRIIAALLSS

Parellel to the on-farm trials, the PhD students in Burundi and DRC have started by late 2007 to prepare the installation of on-station trials. These researcher-managed trials are located on the following sites and soil types:

Burundi: 1. Gitega – Ferralsol (FAO Acrisol) 2. Cibitoke – Vertisol (FAO Vertisol) 3. Kirundo – Ferrisol (FAO Nitisol)

DR Congo: 4. Mulungu – Brown soil – tertiary basalt (FAO Nitisol-Ferralsol) 5. Walungu – Red soil – tertiary basalt (FAO Ferralsol) Rwanda: 6. Butare-Gitarama – Ferralsol (FAO Acrisol)

7. Kibungo – Ferrisol (FAO Nitisol on schiste) 8. Ruhengeri – Brown soil (FAO Andosol)

Table 21: Treatment structure of the on-station trial to study effectsof tillage and mulching on nutrient fluxes, soil physical properties, and banana rooting and plant performance.

Treatment External mulch Removal mulch Tillage Beans

T0 No Yes Yes Yes T1 No No No Yes T2 Yes (Hyparrhenia diplandra) No No Yes T3 Yes (Tripsacum laxum) No No Yes

Each treatment will be repeated 4 times. Every plot consists of 6 x 6 plants, leaving a net plot of 4 x 4 plants. The plants will be planted at a density of 2 x 2 m. The quantity of mulch applied should be approximately 5 cm thickness, equivalent tot 25 t/ha dry matter.

67

The objective of these trials will be to study the impact of tillage and mulch application on nutrient pool fluxes and soil physical properties, and their subsequent impact on the banana rooting systems and the plant performance (Table 21). These trials will be entirely handled by the PhD students in the IITA-led project. As in the on-farm trials, bean intercropping will also be captured in these trials, in order to have a full economic an agronomic evaluation of the bean intercrop on the total productivity of these mixed cropping systems. The measurement protocols on the bean intercrops will be established in collaboration with the TSBF-led legumes project staff.

Within the framework of the PhD research of Telesphore Ndabamenye, trials have been established to assess the effect of planting density on banana (Musa spp) productivity, soil fertility dynamics, and nutrient uptake (Photograph 25). The aim of this research project is to demonstrate and quantify the effects of planting density on banana (AAA-EAHB) productivity and nutrient dynamics by establishing the relationship between soil fertility parameters, environmental factors (light, rainfall, soil moisture) and banana cultivars. A special focus is also given to quantifying the nutrient balance during the

cropping cycle. Three planting density trials were established at three contrasting agro-ecological zones (Rubona ISAR station, Kibungo, ISAR station and Ruhengeri/Kinigi). An initial soil characterization has been carried out at the 3 sites. Five different planting densities are being assessed (1,428 plants ha-1 at a spacing of 3.5x2.0 m, 2,500 plants ha-1 spaced at 2.0x2.0 m, 3,333 plants ha-1 at 1.5x2.0 m, 4,444 plants ha-1 at 1.5x1.5 m and 5,000 plants ha-1 at 1.0x2.0 m). Three AAA banana cultivars were used (two cooking cultivars: ‘Injagi’ and ‘Ingaju’, and one beer cultivar: ‘Intuntu’). Data collection of growth parameters and environmental characteristics is being conducted.

77..BB..33.. IINNIITTIIAALLIISSAATTIIOONN OOFF BBAANNAANNAA DDIISSEEAASSEE CCOONNTTRROOLL

SSTTRRAATTEEGGIIEESS Initial steps have been taken to start Xanthomonas Wilt (XW) activities in Rwanda in partnership with RADA. Mr. Frank Turyagyenda was recruited as a consultant to carry out XW work in the Central African region. The XW activities in Rwanda will focus on:

• Screening banana germplasm for Xanthomonas wilt tolerance;

• Systemicity studies after infections with contaminated garden tools (e.g. during de-leafing, de-suckering) and after an inflorescence infection in both East African highland bananas (Musa AAA-EA group) and “Pisang Awak” (Musa AABB group);

• Seasonal influence of the systemicity and speed of bacterial movements;

• Replanting trials targeting different agro-ecological zones.

Photograph 25: Banana planting density trial at the ISAR, Rubona research station.

68

888... PPPRRROOOGGGRRREEESSSSSS WWWIIITTTHHH MMMAAARRRKKKEEETTT---RRREEELLLAAATTTEEEDDD


888...AAA... BBBAAANNNAAANNNAAA VVVAAALLLUUUEEE CCCHHHAAAIIINNN AAANNNAAALLLYYYSSSIIISSS From February to March 2007, market surveys were conducted together with ISAR, ISABU, and INERA to identify opportunities and constraints in the banana value chains. A total of 400 traders and 150 transporters were interviewed. Presented below are some of the summary statistics of the respondents.

As shown in Figure 35 the rural assemblers mostly deal in beer types and so do the transporters and urban wholesalers. The same patterns were observed in Burundi. The rural retailers and the urban retailers mostly sell the cooking types. From the above it is evident that the banana value chains are as dipictued in Figure 36. Costs incurred by the traders varied with the location of the trader. Whereas the rural traders have high handling costs representing 70-90% of their total costs, urban traders have much lower handling costs (<30%) and spend most of their costs on communication (50%) and storage (20%), despite the fact that transport costs per bunch per kilometer are highest in the urban centra.

Taxes and dues are the major constraints incurred by traders while selling the scarcity of buyers and insufficient finances are also among the constraints mentioned.

Similarly taxes and tolls are mentioned as the most pressing constraints encountered by traders during purchase of bananas while the other constraints include pricing mechanisms, transport, insufficient finance are also mentioned.

0

20

40

60

80

100

120

140

160

Rural

assemblers

Rural

retailers

Transporters Urban

wholesalers

Urban

retailers

Category of traders

Average daily sales of bananas (bunches)

Beer Cooking Dessert Plantain

0

20

40

60

80

100

120

140

160

Rural

assemblers

Rural

retailers

Transporters Urban

wholesalers

Urban

retailers

Category of traders

Average daily sales of bananas (bunches)

Beer Cooking Dessert Plantain

Figure 35: Average daily sales of banana types by category of traders in Rwanda.

Farmers

Rural assemblers

Key

Major Flow

Minor Flow

Restaurants,

Hotels, Institutions

Transporters

Consumers

Rural retailer

Urban

Retailers

Processors

Urban

Wholesalers

Figure 36: The marketing channels of all banana types in the Great Lakes region.

69

The transport in the banana trade sector heavily relies on trucks for the long distances (60-100km), whereas pick-ups are still frequently used for medium distances (40-80km). In Rwanda and DRC, part of the long distance transport is also happening by boat over the Lake Kivu. Bicycle transport is used for the shorter distances (<30 km) and carrying bananas on the head is often done for relatively short distances (<15km). The costs for short distance transport by bicycle or head is also the highest (4-12ct US$ per km) compared to transport by truck (2-4ct US$ per km). Whereas transport distances for most rural and provincial markets are relatively short (15-65km), urban markets are often supplied from much further distances; e.g., the average distance for bananas supplied to Kigali was 170km.

Taxes and road tolls are among the most critical constraints highlighted by the transporters while other constraints mentioned include road drudgery, delays and police interceptions, difficulty in assembling produce and insufficient finance and difficulty in selling produce as well (Figure 38).

In the first half of 2007, banana cross-border trade studies were also conducted at the Rwandan, Burundian, DR Congo, and Ugandan border posts. These data are currently analysed. Preliminary results highlight an important flow of bananas from Uganda and DRC to Kigali, but only minor flow of produce between Burundi and DRC and between Burundi and Rwanda.

888...BBB... LLLEEEGGGUUUMMMEEE VVVAAALLLUUUEEE CCCHHHAAAIIINNN AAANNNAAALLLYYYSSSIIISSS The overall objective of this study is to identify the most important market channels of grain legumes (common beans, soybeans and groundnuts) in the Bas-Congo, Sud-Kivu Montagneux, Kigali-Kibungo, and Umutara Mandate areas from the production to the final consumption level, analyze market structure, performance, and conduct, and calculate margins at each identified market level along the chain. The specific objectives are: (i) to identify the most important market channels (one, two or three) for each commodity (interviewing the operators at earlier levels along the marketing chain helps to reveal the next level until we arrive at the level where most of the traded commodity are sold to final consumers), (ii) to identify where producers take their commodities to, those who purchase from them (the producers), (iii) to identify and analyse the different on-farm profit, expenditure, investment made over grain legume production, (iv) to assess the role of women on legume production and its impacts on household livelihoods improvement, (v) to conduct a survey at each identified major point along the channel (to ascertain all the costs associated with the marketing of the commodities and the prices at which commodities are sold), (vi) to ascertain that all costs and benefits incurred on legumes commercialization are relevant on people’s daily livelihoods, (vii) to determine market structure, performance, conduct and profit margins at each level along the marketing

4%

6%

7%

14%

29%

15%

5%

16%

2%

2%

0 5 10 15 20 25 30 35

Insufficient finances

Inadequate supply

Transport and distance

Pricing

Taxes and dues

None

Storage

Scarcity of buyers

Poor Quality

Theft

Problems encountered in selling bananas

Intensity of problem (% of respondents)

Figure 38: Problems encountered in selling bananas in Rwanda according to traders.

70

chain, and (viii) to make appropriate recommendations for improving marketing efficiency and equity in the marketing of the selected grain legumes in the study area.

Preliminary market studies were carried out in the Action sites in order to trace out the chain of the commodities to be studied (common beans, soybeans and groundnuts). Detailed information collection tools were developed for the household and market level surveys and enumerator were trained for administering these tools. Finally, a detailed sampling frame was constructed, encompassing selection of households and market operators for the different grain legumes. The obtained data will be analyzed using descriptive statistics (percentages, frequencies, graphic displays) and quantitative techniques (relationships will be determined using correlations and ANOVA).

Preliminary results show that at the household level, most farmers do not sell legumes but consume these at the household level because the production is not sufficient for marketing. For farmers who sell their produce, revenues derived from legume sales are mostly used to meet basic needs such as school fees payment, clothing, food staff, medical care. At the market level, it was found that most traders were women, indicating that the commercialization of legumes is headed by women (Table 22). An exception was Kavumu market where 59% of the traders were men. Most traders were rural wholesalers (Table 23). Another particularity of Kavumu market was that traders have the highest initial capital (1,200,000 FC or 2,400 USD) while the lowest is Kabamba (6,500 FC or 13 USD) (data not shown).

Table 22: Trader’s gender in the target markets in the Sud-Kivu mandate area.

Traders gender distribution Market name Legume Male Female

total

Groundnut 1 10 11

Common beans 0 12 12 Soybeans 0 11 11

Cabwine-Mwami

Total 1 33 34 Groundnut 3 14 17 Common beans 3 15 18 Soybeans 0 18 18

Kadutu

Total 6 47 53 Groundnut 10 7 17 Kavumu

Total 10 7 17 Soybeans 0 16 16 Mudaka Total 0 16 16 Groundnut 7 5 1 Common beans 0 6 6 Soybeans 7 11 18

Mugogo

Total 14 22 36 Total 21 125 156 Proportion (%) 20 80 100

In terms of costs and margins, there was a significant difference in purchase price and sale price between different common bean, groundnut, and soybean varieties. This could be to a certain extend because most traders sell their commodities using different measuring units and at times the scales they use are not well calibrated. The price of common beans, groundnuts and soybeans was also affected by periods of food abundance and shortages because during food abundance sales happen at the lowest prices. Most traders are informed on market prices (Table 24) and obtain such information through other traders (data not shown).

71

Table 23: Trader categories in the target markets in the Sud-Kivu mandate area.

Category of traders Market name Legume Rural

wholesalers Urban

wholesalers Urban retailers

Total

Groundnut 11 0 0 11 Common beans 12 0 0 12 Soybeans 11 0 0 11

Cabwine-Mwami

Total 34 0 0 34 Groundnut 0 12 5 17 Common beans 0 12 6 18 Soybeans 0 0 18 18

Kadutu

Total 0 24 29 53 Groundnut 17 0 0 17 Kavumu Total 17 0 0 17 Soybeans 16 0 0 16 Mudaka Total 16 0 0 16 Groundnut 12 0 0 1 Common beans 6 0 0 6 Soybeans 18 0 0 18

Mugogo

Total 36 0 0 36 Total 103 24 29 156 Proportion (%) 66 15 19 100

Table 24: Knowledge on market prices in the target markets in the Sud-Kivu mandate area.

Legumes Name of the market

Information on market pricing Groundnut Common bean Soybean

Total

Cabwine Mwami Yes 9 7 10 26 No 2 5 1 8 Total 11 12 11 34 Kadutu Yes 15 18 16 49 No 2 0 2 4 Total 17 18 18 53 Kavumu Yes 13 0 0 13 No 4 0 0 4 Total 17 0 0 17 Mudaka Yes 0 0 10 10 No 0 0 6 6 Total 0 0 16 16 Mugogo -77 0 0 1 1

Yes 8 3 12 23 No 4 3 5 12 Total 12 6 18 36

Total 57 36 63 156 Proportion (%) Yes 79 78 70 -

72

999... PPPRRROOOGGGRRREEESSSSSS WWWIIITTTHHH NNNUUUTTTRRRIIITTTIIIOOONNN---RRREEELLLAAATTTEEEDDD


999...AAA... SSSOOOYYYBBBEEEAAANNN PPPRRROOOCCCEEESSSSSSIIINNNGGG AAANNNDDD UUUTTTIIILLLIIIZZZAAATTTIIIOOONNN In the context of the legume activities, it was felt that substantial efforts need to be made to expose farmer associations to the processing and utilization of soybean since the other legumes (beans, groundnut) are well known to the farming communities. In 2007, various training tools were prepared and a detailed strategy was developed with partners, including health centers at the Action site level, to be implemented in 2008.

In terms of training tools, one training manual for trainers was developed and translated in French (Photograph 26), covering the following topics: bases de la nutrition, présentation du soja, hygiène de base, méthodes domestiques de préparation des produits de soja, contenu nutritif des produits de soja, recettes du soja, cartes postales des menus de soja, teste d’acceptabilité par le consommateur, and évaluation de la connaissance sur la transformation et l’utilisation du soja.

In terms of strategy, activities are: (i) assessment of the knowledge of soybean processing, (ii) establishment of demonstration gardens, (iii) training of trainers, and (iv) training of farmer associations. Assessment of the knowledge of soybean processing will aid organizing the curriculum of the processing activities. Such issues will be included in each training session as a pre-test evaluation. During the training, a wide range of products for different targets (markets, small scale business, household consumption, there is potential food AID

market for NGOs on USAID Title II programs advocating for locally procured food AID etc) will be produced and after acceptability studies, the project will promote what is preferred by farming communities. Near the health centers, demonstration gardens will be established, focusing on soybean production. An area of about 2500 m2 (25 are), will give about 200 kg of soybean grains. Of this 200 kg, 20 kg will be reserved for planting the next season. About 80 kg will be used for the processing training sessions and about 100 kg will be used for giving out to people who have followed the training to try on their own plot (to about 200 g per mama at 500 g per visiting female farmer). The gardens will be managed by the health center with technical supervision by ISAR or INERA. Training events will be organized by health center staff (see next paragraph on training of trainers) around the demonstration gardens with attendance of female farmers visiting the health centers. These will be given 2 documents (besides the soybean seeds): (i) a one-page folder in local language related to the production (see draft attached) (Photograph 27) and (ii) recipes that include soybean. Monitoring tools and monitoring timeframes will be developed in order to follow up if the female farmers are really planting the soybean, how

Photograph 26: Cover of the soybean processing training manual.

73

this is working, if they are able to increase soybean consumption in their families, if this has an impact on children’s health, if they would be interested in disseminating to other people, etc.

In terms of training of trainer sessions (ToT), there will be one session per action site with about 20 people attending. The ToTs will take about 3 days per session and need to be held before end of April 2008 in order to have the receipt document ready to be used in training the farmers belonging to the associations. The technical document will be multiplied in 300 copies (50 for BasCongo, 50 for Kivu, 100 for Rwanda) to satisfy the demand of the ToT sessions and have extra copies for extra demand by partners. For the trainers to advance the training through training sessions with farmers, the project will (i) facilitate trainers and (ii) develop tools to monitor if the trainers are organizing farmer training events, who they train, what is happening after training, etc. In terms of training of farmer associations, the associations that we are directly working with will be trained by the nutritionists working in the different Mandate areas. This includes about 8-12 associations with about 20-40 members or about 400 people. The training sessions will be held before the end of June 2008 (before the next growing season) but after having finalized the ToT sessions during which the final recipe book will be developed. The recipe document in local languages will be multiplied and distributed during the training. During these events, an acceptability test will be implemented for the processed products. The soybean production flyer and the commodity fact sheet will also be distributed. Again, monitoring tools and monitoring agendas will de developed for creating the necessary feedback.

Photograph 27: Soybean production flyer in French (left) and Kinyarwanda (right) to be distributed to female farmers interested in producing soybean through the health centers.

74

999...BBB... IIINNNIIITTTIIIAAALLLIIISSSAAATTTIIIOOONNN OOOFFF BBBAAANNNAAANNNAAA NNNUUUTTTRRRIIITTTIIIOOONNN---HHHEEEAAALLLTTTHHH---RRREEELLLAAATTTEEEDDD AAACCCTTTIIIVVVIIITTTIIIEEESSS

Mrs. Beatrice Nakhauka Onyango Ekesa, a nutrition-health expert, was recruited as a Dutch associate expert at Bioversity International. Beatrice will carry out nutrition-health work in the framework of CIALCA and will specifically be looking at the potential contribution of banana-based systems to the nutrition of small holder communities. She will mainly carry out her CIALCA nutrition activities around Butembo/Beni in north Kivu, DR-Congo and in Burundi. She submitted a CIALCA nutrition-health proposal to the NGO ‘HealthNet TPO’. This proposal got approved and totals 22,330 $ (with both HealthNet TPO and CIALCA financial contributions). The different CIALCA/HealthNet TPO nutrition-health activities to be carried out comprise:

• Analysis of CIALCA baseline survey datasets;

• Carry out nutrition surveys on the contribution of banana and plantain to the diet of community members and on post harvest technologies in north Kivu, DR-Congo and Burundi;

• Carry out research on the links between agriculture, nutrition and health in Burundi and north Kivu, DR-Congo;

• Train community-own resource persons on sustainable agricultural interventions for better nutrition and health in Burundi and north Kivu, DR-Congo;

• Establish comprehensive demonstration gardens (including bio-fortified beans, soybean and banana) in Burundi and north Kivu, DR-Congo.

75

111000... PPPRRROOOGGGRRREEESSSSSS WWWIIITTTHHH MMMOOONNNIIITTTOOORRRIIINNNGGG AAANNNDDD

EEEVVVAAALLLUUUAAATTTIIIOOONNN Monitoring and evaluation activities take place at different stages of project implementation and can be sub-divided in the following broad set of activities: (i) monitoring and evaluating progress with implementation of the various projects, (ii) monitoring and evaluation the various project interventions in collaboration with farmer associations and development partners, and (iii) monitoring and evaluating the impact of project interventions on farmer livelihoods.

M&E of project implementation Project planning happens at three levels (ii) across the three projects constituting CIALCA, (ii) within a single project but across the various mandate areas (Photograph 28), and (iii) within a single project, within a specific mandate area. A common agenda point of each of these meetings is review of progress, either through presentations of and discussions on progress reports or through formal logframe evaluations (Annex 1). Based on this, project activities are adjusted, if necessary, and prioritization is done for initiating new activities. During subsequent meetings, this process is repeated. This approach entails that, although the level and range of activities initiated across the different Mandate areas were similar at project inception, this is not longer the case today since progress with activities in some Mandate areas is relatively greater than progress with activities in other Mandate areas.

M&E of project interventions Various project interventions are being evaluated with farmer associations, including improved legume and banana germplasm and natural resource management options (Photograph 29). A critical stage of each of those activities is to get gender-differentiated feedback on the various positive and negative aspects of the interventions demonstrated in order to guide the demonstration of future interventions and to identify researchable issues that need to be addressed either through controlled on-station field experimentation or through more strategic greenhouse and laboratory experiments.

M&E of project impacts on rural livelihoods: During the baseline, several livelihood-related indicators were quantified and these will be evaluated again towards the end of 2008 (end of phase I of CIALCA) with an initial focus on the Action sites. At the end of a potential CIALCA-II project, the impact assessment activities will be broadened to include the various Satellite sites and eventually the total Mandate areas.

Photograph 28: The Director General of INERA, the Director of Science of ISAR and the TSBF-CIAT Director co-chairing a session during a TSBF-CIAT general planning meeting.

Photograph 29: Training session with farmers on participatory evaluation of improved legume germplasm.

76

111111... PPPRRROOOGGGRRREEESSSSSS WWWIIITTTHHH DDDEEEGGGRRREEEEEE---RRREEELLLAAATTTEEEDDD

AAACCCTTTIIIVVVIIITTTIIIEEESSS Capacity building is a crucial component of CIALCA since all countries in which CIALCA operates are recovering from civil strife which has had a detrimental impact on the capacity of various partners in the mandate areas.

MSc projects In the context of CIALCA, 13 MSc-related projects are currently supported (Table 25). Most of these are implemented in DR Congo and Rwanda since the TSBF-CIAT-led project is not directly operating in Burundi. MSc projects are also supported for students form the Belgian university partners.

Table 25: MSc projects supported by CIALCA.

Name Nationality University Topic

Julie Lunzihirwa

DR Congo Facultés Catholiques de Kinshasa, DR Congo

The impact of beans and groundnut channels on the productivity and agricultural income of households in the cataractes area’.

Rachel Zozo DR Congo Makerere University, Uganda

Assessing the socio-economic importance legumes-based on the livelihoods of farmers at Mugogo and Mudaka Markets in Ngweshe and Katana axes, Democratic Republic of Congo.

Muke Manzekele

DR Congo Université de Kinshasa, DR Congo

Techniques d’amélioration de la production agricole et de la stabilisation des sols en pente au Sud-Kivu Montagneux.

Aime Herikazi DR Congo Makerere University Kampala, Uganda

The impact of soil tillage and mulch application on soil physical properties and productivity of banana-bean intercrop systems.

Idja Sikyolo DR Congo UCG, Butembo, DR Congo

Altitude effects on plant performace in banana and plantain demonstration plots and Musa collections in North-Kivu, DR Congo

Agnes Mukandinda

Rwanda National University of Rwanda, Rwanda

Nutrient flows in banana based cropping systems.

Placide Rukundo

Rwanda Katholieke Universiteit Leuven, Belgium

Banana biotechnology.

Edouard Rurangwa

Rwanda Jomo Kenyatta University of Agriculture and Technology, Kenya

Initial survival of tissue culture bananas as affected by inoculation with arbuscular-mycorrhizal fungi.

Anaclet Nibasumba

Burundi Université Catholique de Louvain-la-neuve, Belgium

Relationship between nutrients (cations) in the soil mineral and organic pools and nutrients at the banana root surface.

Oswald Ntakirutimana

Burundi Université de Burundi

Contribution a l'etude de l'état phytosanitaire du bananier dans les Provinces de Gitega, Kirundo et Cibitoke'

77

Geoffroy Germeau

Belgium Université Catholique de Louvain-la-neuve, Belgium

Explaining banana yield differences in Rwanda through quantification of banana crop performance, soil fertility, pest and diseases and crop management practices.

Ellen Vandamme

Belgium Katholieke Universiteit Leuven, Belgium

Nutrient deficiency and unavailability in the soils of Walungu, South-Kivu, Democratic Republic of Congo.

Julie Vandamme

Belgium Université Catholique de Louvain-la-neuve, Belgium

Analysis of stakeholder perceptions of constraints and solutions in the banana sector in Rwanda.

PhD projects CIALCA is currently supporting 8 PhD students (Table 26). In addition, CIALCA is actively supporting research staff to pursue further scholarship opportunities that can build on the ongoing research. In that respect, there’s an outlook for two Belgian PhD students and one IITA-Uganda staff to do their PhD within the CIALCA project on farming systems, soil nutrient pools and recycling, and banana value chain and market analysis.

Table 26: PhD projects supported by CIALCA.


Dowiya Nzawele Benjamin

DR Congo Sokoine University, Tanzania

Characterization of Musa germplasm in Eastern DR Congo.

Tony Muliele DR Congo Université Catholique de Louvain-la-neuve, Belgium

Soil moisture and soil physical constraints in highland banana systems.

Svetlana Gaidashova

Rwanda Université Catholique de Louvain-la-neuve, Belgium

Research on banana-soil fertility-soil biology interactions, with special emphasis on the role of plant-parasitic nematodes and abuscular mycorrhizal fungi (AMF).

Telesphore Ndabamenye

Rwanda University of Pretoria, South Africa

Planting density, soil fertility, leaf nutrient status and nutrient absorption

Josaphat Rusisiro Mugabo

Rwanda Katholieke Universiteit Leuven, Belgium

Agricultural intensification under population pressure in Rwanda: An analysis of fertilizers policy and legume-based systems economic incentives.

Leon Nabahungu

Rwanda Wageningen University, the Netherlands

Competing Claims on Wetland in Eastern Rwanda: Challenges and opportunities.

Syldie Bizimana

Burundi Université Catholique de Louvain-la-neuve, Belgium

Effect of soil management on nutrient availability and nutrient recycling in highland banana cropping systems.

Célestin Niyongere

Burundi JKUAT, Nairobi, Kenya

Banana Bunchy Top Virus (BBTV) in the Great Lakes region.

78

Undergraduate projects CIALCA is currently supporting 13 Undergraduate students (Table 27).

Table 27: Undergraduate projects supported by CIALCA.


Matara Murhonyi (Memoire-Ingénieur)

DR Congo Université Catholique de Bukavu

Identifying fungal diseases affecting banana production in South Kivu.

Bahati Lukangira (Memoire-Ingénieur)


Quantifying the spread and importance of banana bunchy top virus (BBTV) in South Kivu.

Kambale Mboho (Memoire-Ingénieur)

DR Congo Université Catholique de Graben

Pest and disease problems in banana systems in Nord Kivu.

Sereka Saghasa (Memoire-Ingénieur)


Understanding soil management in banana-based farming systems in Nord Kivu.

Kakule Lukalango (Memoire-Ingénieur)


Characterizing and understanding banana germplasm diversity in Nord Kivu.

Sondirya Tsongo Michel (Memoire-Ingénieur)


Identifying socio-economic constraints in banana-based farming systems in Nord Kivu.

Janvier Bashagaluke Bigabwa (Stage-Ingénieur)


Assessment of erosion features in farmers’ fields.

Rehani Jumaine (Memoire-Ingénieur)


Demonstration of the microdosing fertilizer technique and of benefits of high biomass-yielding legumes in cereal-based rotation systems

Wivine Zirhahwakuhingwa Munyahali (Memoire-Ingénieur)


Demonstration of improved agronomic practices in cassava-legume intercropping systems

Chantal Karondo (Stage-Ingénieur)

Burundi Université de Bujumbura

Etude de la diversité génétique du germoplasme de bananier au Burundi.

Fidès Barigenera (Stage-Ingénieur)


Evaluation de l'état phytosanitaire des bananiers dans les communes les plus productrices de banane de Gitega: Giheta, Itaba et Makebuko.

Léonidas Ndikuriyo (Stage-Ingénieur)


Détermination des équivalents taxonomiques en nomenclature Américaine (Soil Tax.) et FAO-INEAC comme une façon de définition des zones potentielles de culture du bananier.

Félix Gatoto (Stage-Ingénieur)


Enquête de prospection de BBTV sur base de symptômes caractéristiques et les pertes causées par le BBTV dans la province de Cibitoke.

79

Post-doctoral positions CIALCA is supporting post-doctoral positions:

• Dr Joyce Mnyazi Jefwa, a Kenyan national, was recruited as a post doc to work on AMF in banana systems in the framework of the Bioversity project. She will assess AMF species associated banana cultivars and evaluate banana cultivars inoculated with indigenous AMF for their performance in the nursery and in farmer’s fields.

• Dr Pieter Pypers, a Belgian national, was recruited to work as a soil scientist within the TSBF-led legume project. He’s backstopping the project activities in DRC and Rwanda, and is particularly focusing on soil fertility aspects and the beneficial role of legumes within the overall productivity of cropping systems.

• Dr. Charles Murekezi, a Ugandan national, has been recruited as a project scientist with IITA. Dr. Murekezi is a specialist in the field of banana agronomy and virology and is based in Rwanda. He will spend 60% of this time to train and backstop the banana research team at ISAR. Another 40% of his time will be spent in Burundi and DR Congo to backstop other project activities.

• Dr. Emily Ouma, a Kenyan national, has been recruited as a project scientist on socio-economics with IITA. Dr Ouma has a background in agricultural economics in the CGIAR system, having conducted her PhD on cattle traits in East Africa within the framework of ILRI –led projects. She finished her PhD at the University of Kiel in Germany. Dr. Ouma will be based at IRAZ in Burundi. Her role is to lead the socio-economics research in the IITA-led project and to backstop and train partners in the national instates of Rwanda, DR Congo and Burundi. DR. Emily Ouma will join CIALCA as of February 1, 2008.

Training of national institute scientists

• Formal training events have been organized with national institute scientists and NGO partners in the context of the PRA’s, baseline surveys, legume germplasm evaluation, seed systems, banana nematology training, banana macro-propagation training, monitoring and evaluation, and final characterization studies. Such training events are aiming at cross-regional exchange of expertise.

• On-the-job training of national institute scientists and NGO partner has taken place in the context of agronomy, data collection, and sample processing.

• Various MSc and PhD projects, mentioned above, are implemented by colleagues from the national research institutes.

Training of farmer associations:

• In the context of legume activities, various farmer associations in each of the mandate areas have been trained in participatory germplasm evaluation (including training on planting, fertilizer application, weeding and harvesting techniques) and seed multiplication and storage.

• Plans are underway to initiate training events on soybean processing, market linkages, and participatory monitoring and evaluation.

80

111222... AAANNNNNNEEEXXXEEESSS

AAANNNNNNEEEXXX 111... LLLOOOGGG---FFFRRRAAAMMMEEE EEEVVVAAALLLUUUAAATTTIIIOOONNNSSS

AANNNNEEXX 11..AA.. LLOOGG--FFRRAAMMEE OOFF TTHHEE TTSSBBFF--CCIIAATT--LLEEDD PPRROOJJEECCTT

Objectively verifiable indicators/milestones Level of success

Work Package 1: Baseline information and site selection 1.1. By June 2006, all partners are involved in the planning and implementation of the project activities in all action sites. Achieved (see planning and meetings and review mission); new NGO partners engaged in all 4 mandate areas.

100%

1.2. By June 2006, all necessary information is available to select action and satellite sites for all mandate areas. Achieved for action sites (see baseline and PRA reports); activation of satellite sites is currently on-going (9 active in Bas-Congo).

AS: 100% SS: 20%

1.3. By Dec 2006, sufficient information is available to direct marketing-related activities for all mandate areas. Only preliminary market surveys were conducted through student MsC and PhD projects

33%

1.4. By Sept 2006, at least 5 action and 30 satellite sites have been identified across all mandate areas. Achieved for action sites (16 active); potential satellite sites are identified (see baseline and PRA reports) but only 10 are currently active.

AS: 100% SS: 33%

1.5. By March 2007, baseline information on livelihood status has been collected in the action sites. PRAs, baseline study and detailed characterisation studies have been conducted. Production of the reports is currently on-going

75%

1.6. By March 2007, farmer groups have been identified in all action sites. Achieved: 39 farmer associations involved in the action sites, 61 farmer associations involved in satellite sites.

100%

1.7. By March 2007, farm typologies have been constructed that will form the basis for evaluation of the appropriateness of specific technologies to specific groups. Detailed characterisation studies have been conducted at action site level; data analysis and reporting is pending.

33%

Work package 2: Participatory evaluation of best-bet options 2.1. By June 2006, a list of promising NRM options is available for initial testing, taking into consideration the overall action site characteristics. Achieved: NRM options have been identified and are currently being tested at action and satellite site level.

100%

2.2. Between Sept 2006 and the Dec 2008, the number of on-going on-farm trials increases from 50 to 1000, across all mandate areas. At the end 2007, over 200 on-farm trials have been conducted (not taking into account seed multiplication fields). These include demonstration trials (about 120) and adaptation trials (about 80). In addition, an estimated number of 350 seed multiplication fields have been installed with farmer associations.

100%

2.3. By Dec 2008, the impact of NRM options on various aspects of rural livelihoods is evaluated in all action sites. Baseline information has been collected; the activity itself is to be implemented in year 3.

N/A yr 3 activity

81


2.4. By the Dec 2008, the role of access to markets and centres de santé in improving livelihoods is evaluated for all action sites. Baseline information has been collected; the activity itself is to be implemented in year 3.

N/A yr 3 activity

2.5. By March 2007, seed multiplication is on-going in all action sites to satisfy the demand in the action and the satellite sites. Achieved: seed multiplication is currently on-going in all action and satellite sites.

100%

Work package 3: Understanding mechanisms and contributions 3.1. By Dec 2008, sufficient knowledge on mechanisms driving tolerance to drought and low soil P is available to guide breeding efforts. On-going: a selection of potential P-efficient germplasm is being identified through the legume evaluation activities; specific trials to unravel mechanisms of P tolerance will then be conducted.

25%

3.2. By Dec 2008, relationship between soil fertility status and the nutritional quality of bio-fortified crops is used by development partners to target production of these crops. On-going: micronutrient analysis of bean grains are pending – a preliminary analysis has been conducted.

50%

3.3. By Dec 2008, the potential for occurrence of positive interactions between organic and mineral inputs is evaluated for the most common cropping systems in each mandate area. On-going: one or more NRM options in each of the mandate areas include specific treatments for testing of positive interactions between organic and mineral inputs.

50%

3.4. By Dec 2008, the contribution of resilient germplasm in driving overall system resilience is understood for the conditions occurring in all mandate areas. A long term trial will be initiated in 2008, using specific information from the detailed characterisation study to investigate the contribution of improved germplasm to system resilience and soil fertility.

0%

3.5. Throughout the project life, new questions generated in the evaluation efforts of Work Package 2 are addressed and fed back to these evaluation activities. On-going: such questions include for example nutrient deficiency studies in Sud-Kivu and an investigation into the effect of nutrient inputs on cassava tuber quality.

100%

Work package 4: Trade-off analysis and impact assessment 4.1. Once each year an annual planning meetings and once each season, an action site meeting is organised. Achieved (see planning and meetings and review mission)

100%

4.2. By Sept 2006, a monitoring and evaluation framework is established and operationalised. Partly achieved: training was done; monitoring and evaluation of project activities is done informally; an external review mission by DGDC was done in July 2007. A formal farmer monitoring and evaluation framework needs to be put into operation.

60%

4.3. By Dec 2006, local and scientific indicators have been identified to measure progress with project interventions against baseline information. In progress: a baseline was conducted; indicators are currently being prioritized as part of the baseline report.

50%

4.4. By the Dec 2007, products of the trade-off analysis are guiding the introduction and evaluation of alternative NRM options, better suited to the farmer production objectives and the environment of the actions sites. In progress: discussions on-going for increased involvement of AfricaNUANCES to do trade-off analysis; necessary information is available through baseline and detailed characterisation studies. Selection of NRM options is principally guided by identification of constraints through the baseline study and discussions with farmer associations.

20%

82


4.5. By the Dec 2008, the impact of the project activities in the action sites, satellite sites, and mandate areas is quantified, against the baseline information collected. Baseline information has been collected; the activity itself is to be implemented in year 3.

N/A yr 3 activity

Work package 5: Scaling up and out 5.1. By Sept 2006, partners in the action and satellite sites are aware of the project and ready to collaborate. Achieved (see planning and meetings and review mission); new NGO partners engaged in all 4 mandate areas.

100%

5.2. By Sept 2006, partners have identified the optimum way of communicating project-related information. Achieved: an arrangement is made to have a communication specialist to serve needs of all partners in the 3 countries.

100%

5.3. From Sept 2006 onwards, various initiatives are taken to facilitate farmer-to-farmer dissemination, including seasonal field days and farmer exchange visits between action and satellite sites. Not achieved: exchange visits were only organized in Sud-Kivu around the seed multiplication activities. Farmer-to-farmer dissemination will be strengthened in 2008 through exchange visits and field visits, led by NGO partners in action and satellite sites.

10%

5.4. By Dec 2008, sufficient information is available to advice on optimum ways to disseminate and scale up project products, taking into account the overall conditions of the mandate areas. Not achieved: this requires strengthening by involving NGO partners and an in-depth analysis of the baseline and characterisation studies on principal information distribution systems in the mandate areas.

0%

5.5. By Dec 2008, at least 10% of the farmers and 50% of local policy makers are aware of the project products in the mandate areas. The current status of activities allows an increased number of farmer adaptation trials in 2008; knowledge of the project will be increased through dissemination events organized around the various activities (including nutrition and processing activities with involvement of ‘Centres de Santé’).

N/A yr 3 activity

Work package 6: Capacity building 6.1. By Sept 2006, specific training needs for all stakeholders are identified. Partly achieved: training needs of project research, NGO partners and farmer associations are addressed ad-hoc rather than through formal need assessments.

75%

6.2. By Sept 2006, research for development teams have been identified in each action sites, comprising partners from the NARS and NGOs. Achieved (see planning and meetings and review mission); new NGO partners engaged in all 4 mandate areas.

100%

6.3. Between Sept 2006 and Dec 2008, at least 50 farmer groups and 1000 farmers across the various mandate areas have acquired the necessary skills to test, evaluate, and adapt alternative NRM options. Achieved: currently almost 100 farmer associations (with on average 15 members per association) are involved in project activities of testing and multiplying germplasm and/or testing NRM options.

100%

6.4. By Dec 2008, at least 2 input dealers in each action and satellite site have acquired sufficient knowledge to guide farmers in the most appropriate management of inputs for their respective environments. Not achieved: input dealer networks are poorly functional – alternatives using credit systems for access to inputs and potential collaboration with other projects are explored.

N/A yr 3 activity

83


6.5. By Dec 2008, at least 6 MSc projects have been submitted for defences and at least 3 PhD projects are nearly completed. On-going: currently 5 MsC and 3 PhD projects are being conducted

100%

6.6. By Dec 2008, proceedings of a final symposium are submitted for publication.

N/A yr 3 activity

AANNNNEEXX 11..BB.. LLOOGG--FFRRAAMMEE OOFF TTHHEE BBIIOOVVEERRSSIITTYY--LLEEDD PPRROOJJEECCTT


Work Package 1: Establishing Musa sector linkages within each country (INERA, ISAR, ISABU, IRAZ) 1.1. Musa sector development framework established. Musa sector strategic plans (for Burundi, DR-Congo and Rwanda) developed by a wide range of Musa stakeholders during a Bioversity-led planning meeting in Butare, Rwanda.

100%

1.2. GIS-based compilation of information on agro-climate, production zones, socio-economics of farm communities and organisations. Participatory rural appraisal [PRA], baseline and diagnostic surveys carried out across the 3 counties. GIS activities carried out in the CIALCA framework by a TSBF GIS expert.

85%

1.3. Musa production zones characterized and pilot sites selected. Participatory rural appraisal [PRA], baseline and diagnostic surveys carried out across the 3 counties. Benchmark sites selected for demo-plot establishment, PhD, MSc and BSc studies, and farmer participatory research.

85%

1.4. Strategy and resource mobilization approach formulated and implemented. Musa sector strategic plans (for Burundi, DR-Congo and Rwanda) developed by a wide range of Musa stakeholders during a Bioversity-led planning meeting in Butare, Rwanda.

70%

1.5. Final plan developed to guide future sector development and resource mobilization.

N/A yr 3 activity

1.6. Methods guide for national sector development compiled.

N/A yr 3 activity

Work package 2: Building Musa partnerships regionally (Bioversity-BARNESA, Bioversity-MUSACO) 2.1. Country perspective shared with region / GIS based compilation of information / pilot sites prioritized according to the regional perspective. Musa sector strategic plan for the CEPGL region developed by a wide range of Musa stakeholders from Burundi, DR-Congo and Rwanda during a Bioversity-led planning meeting in Butare, Rwanda. GIS activities carried out in the CIALCA framework by a TSBF GIS expert. Benchmark sites selected for demo-plot establishment, PhD, MSc and BSc studies, and farmer participatory research according to the regional needs.

80%

2.2. Regional agenda initiated. Musa sector strategic plan for the CEPGL region developed. Research topics chosen according to the regional priorities.

85%

2.3. Final results shared with regional networks and plan for future developed.

N/A yr 3 activity

Work package 3: Serving international germplasm needs (Bioversity-KUL ITC: INIBAP Transit Centre) 3.1. Germplasm collected and maintained The INIBAP Transit Centre (ITC) continues to maintain 1183 accessions and efforts are underway to introduce more germplasm into the collection from different countries, especially from the Democratic Republic of Congo (38 plantain accessions were received at the ITC in December06-January07), Tanzania, Kenya, Republic of Central Africa, Uganda, Rwanda and Burundi. This will ensure the conservation of Musa germplasm in face of the BXW epidemic.

40%

84


3.2. Collection rejuvenated and cryo-preserved Most of the rejuvenated germplasm is grown in the field, and during the next 2 years information on their trueness-to-type is expected. Every day still 5-7 accessions are distributed from the ITC worldwide and for the moment 598 accessions are already cryo-preserved.

40%

3.3. PATHOGENS treated and germplasm disseminated Activities start on June 1st 2007

40%

Work package 4: Integrating local and improved germplasm (INERA, ISAR, ISABU, IRAZ) 4.1. Plan for collection, characterisation and conservation developed with regional perspective. Musa germplasm collections are being established in north and south Kivu, DR-Congo by an INERA PhD student. Info on the accessions will be entered into the Bioversity MGIS software which will make it possible to link the DR-Congo germplasm to already established Musa collections in Rwanda, Burundi, Tanzania and Uganda.

85%

4.2. Local germplasm inventoried/characterised. A PhD student from INERA DR-Congo is collecting and characterising Musa germplasm from eastern DR-Congo, and comparing this germplasm with Musa collections in Rwanda [ISAR], Burundi [IRAZ] and Uganda [NARO/IITA] IRAZ is strongly involved in the Musa germplasm activities.

70%

4.3. In situ conservation piloted Pending the identification of local germplasm

pending

4.4. Cultivar performance data compiled. N/A yr 3 activity

4.5. Cultivars in trials. 21 Musa germplasm demo-plots (each containing over 20 genotypes) have been established in contrasting agro-ecological zones across the 3 countries

60%

4.6. Feasibility of alternate seed multiplication and dissemination systems diagnosed. A macro-propagation training course was given to regional stakeholders (including CIALCA partners) in the framework of a USAID funded project (C3P).

40%

4.7. Regional plan on cultivar introduction, evaluation and seed multiplication established.

N/A yr 3 activity

Work package 5: Understanding stress resistance (supportive and strategic research – KULeuven) 5.1. Abiotic stress resistance measured Activities start on June 1, 2007

40%

5.2. Gene tagging and transgenic lines Activities start on June 1, 2007

40%

5.3. Fungal resistance genes characterized Activities start on June 1, 2007

40%

Work package 6: Developing improved production systems (INERA, ISAR, ISABU, IRAZ with TSBF) 6.1. Technical options for soil fertility, plant nutrition and pest and disease management compiled from the region and beyond. Info available in Bioversity’s Musalit database and several Bioversity project final workshop proceedings. Bioversity, France is also developing a Musa resource knowledge centre.

50%

6.2. Plans established for on farm work with farmer research groups. Identified constraints and farmer’s needs emerging from the PRA/baseline surveys and diagnostic surveys will determine the best-bet technologies to be tested on farm. On farm trials with best-bet technologies to be established by both IITA and Bioversity during September 2007.

60%

85


6.3. Market oriented improved production systems developed. On farm trials with best-bet technologies to be established by both IITA and Bioversity during September 2007.

45%

6.4. Soil fertility and plant nutrition constraints identified. PRA, baseline surveys and diagnostic surveys carried out across the 3 counties.

85%

6.5. Biological and agronomic feasibility of soil improvement and plant health options determined in research trials. On farm trials with best-bet technologies to be established by both IITA and Bioversity during September 2007. A PhD study (ISABU) on Banana Bunchy Top Virus is ongoing. A PhD study (ISAR) on planting density, soil nutrient uptake and leaf nutrient status is ongoing.

55%

6.6. Farmer participatory research groups involved in on farm studies on soil enhancement technologies. On farm trials with best-bet technologies to be established by both IITA and Bioversity during September 2007.

55%

6.7. Methods and examples compiled in a manual for use by other extensionists and farmers.

N/A yr 3 activity

AANNNNEEXX 11..CC.. LLOOGG--FFRRAAMMEE OOFF TTHHEE IIIITTAA--LLEEDD PPRROOJJEECCTT


Work Package 1: Baseline assessment (BASELINE) 1.1. Gather and synthesize available information on banana pest and disease and soil constraints, banana markets, existing production and post-harvest technologies, and cultivar distriubtion at national and regional levels; identify NGO, farmer group, and development partners; select benchmark sites. Achieved (see various planning meeting outputs and consultancy report).

100 %

1.2. Assess demand patterns, price and income elasticities, consumer preferences, regional and international trade for fresh and processed banana products Partially achieved – (Market surveys 100%, Farm gate prices 100%, Post-harvest prod. 75%).

85%

1.3. Identify nutritional constraints to banana production in major production zones On-farm diagnostic data collected at all sites but few lab analysis left

90%

1.4. Assess economic characteristics of existing technologies at national and regional levels. Achieved– all farm economics data collected, but modelling (see 1.6) in progress

100%

1.5. Characterize in detail for 6 benchmark sites banana-based farming systems (soil, pest and disease constraints, cultivars, farm and household characteristics, and production patterns). Finished in all sites.

100%

1.6. Economically assess banana production constraints at sites All data collected. Economic modelling of constraints and technologies ongoing.

60%

Work package 2: Integrated banana systems development, evaluation, and demonstration (SYSTEMS) 2.1. Develop and select with farmers three best-bet technologies per site for further evaluation Feedback workshops and technology selection done in Rwanda, South Kivu, and ongoing in Burundi and North Kivu.

75%

86


2.2. Multiply banana germplasm in low-cost multiplication centers in 3 countries; production of starter material in tissue culture lab in Uganda Macro-propagation training was provided in 2006. Centres have been established in DR Congo and Rwanda, but Burundi is still trailing. TC material is continuously multiplied at IITA and IRAZ

75%

2.3. Demonstrate to farmers and development partners improved germplasm, IPM and soil-improving technologies (and their integration) on-station and on-farm; obtain farmer feedback Germplasm trials established. On-station mulch trials and on-farm trials started in Dec 2007

75%

2.4. Evaluate the economic and yield benefits of best-bet banana integrated practices (IPM, nutrient, and agronomic) imposed upon existing fields; discuss and obtain feedback from farmers, adapt trials based on farmer feedback

N/A yr 3 activity

2.5. Evaluate the economic and yield benefits of best-bet banana integrated practices (germplasm, IPM, nutrient, and agronomic) imposed upon newly established fields on-farm and on-station; discuss and obtain feedback from farmers, adapt trials based on farmer feedback

N/A yr 3 activity

2.6. Quantify on-farm nutrient flow dynamics with the objective of optimizing nutrient cycling PhD studies ongoing. First abstracts/papers submitted.

50%

2.7. Disseminate production packages (germplasm, IPM and soil management) to farmers within benchmark sites through trainings.

N/A yr 3 activity

Work package 3: Post-harvest (POST-HARVEST) 3.1. Evaluate potential post-harvest options (processing and value adding) from inside and outside the study areas. � Finished in Rwanda, but validation of results in Burundi, DRC needed. Also product quality tests in laboratories remain.

50%

3.2. Demonstrate to and train farmers in novel post harvest technologies to farmers; adapt technologies with farmer feedback

N/A yr 3 activity

3.3. Farmers trained in business plans for the post-harvest options identified in year 2

N/A yr 3 activity

Work package 4: Capacity-building (CAPACITY) 4.1. 2 PhD dissertation studies on topics relating to banana, soil types, rhizosphere processes, nutrient uptake and pest/disease tolerance. PhD students identified. Managed to be more cost efficient so recruited 3 PhD students for UCL.

50%

4.2. Public awareness of project goals and outputs increased throughout project. Good interaction with partners and farmers but this will remain an continuous effort in the project

60%

Work package 5: Monitoring and evaluation (M&E) 5.1. Project monitored annually; progress on milestones assessed; logframe and budgets adjusted as necessary; next year's activities planned; project staff evaluated Progressive activity; i.e. logframes and budgets are continuously adjusted when drawing new agreements with partners. Planning through meetings with partners. Project staff evaluated annually

66%

87

AAANNNNNNEEEXXX 222::: FFFIIINNNAAALLL CCCHHHAAARRRAAACCCTTTEEERRRIIIZZZAAATTTIIIOOONNN TTTOOOOOOLLL FFFOOORRR TTTHHHEEE

LLLEEEGGGUUUMMMEEE---BBBAAASSSEEEDDD SSSYYYSSSTTTEEEMMMSSS

CIALCA – TSBF-CIAT PROJET LEGUMINEUSES SYSTEMES/SOLS, ACCES AU MARCHE et NUTRITION/SANTE

CARACTERISATION FINALE

Outils nécessaires pour cette activité Chaque équipe aura besoin d’au moins:

- 1 GPS (toutes GPS doivent être calibrés avant le départ de l’activité) - 1 sonde (à 0-20cm) - 1 camera digital - papier manila (A3) + crayons couleur (pour dessiner les cartes de la ferme) - 1 × 1 m cadre pour mesurer la densité des légumineuses - fil marqué avec des nœuds chaque 20cm pour mesurer la couverture des mauvaises herbes - 1 calculateur à calculer le couverture des mauvaises herbes - des sachets à stocker des échantillons des graines et les sols - crayons pour étiqueter les échantillons - mètre à ruban (MUAC tape – mesures anthropométriques) - un pèse-personne (mesures anthropométriques) - tableau de taille (mesures anthropométriques)

Procédure La caractérisation finale est réalisée pour un sous-échantillon des ménages interviewés dans le baseline. Cette sélection est faite par hasard. Dans chacun des villages (ou 20 ménages ont été sélectionnés pour le baseline), 3 (au Rwanda) ou 4 (au Sud-Kivu/Bas-Congo) ménages sont choisis arbitrairement pour la caractérisation finale. L’objectif est de caractériser 3 ménages par jour en terme du rôle des légumineuses dans les sols/systèmes, nutrition/santé, et accès au marché+autres aspects socio-économiques. Pour les aspects de nutrition/santé, le double de ménages doit être caractérisé parmi les ménages avec au moins un enfant d’âge entre 2 et 5 ans. La caractérisation finale vise à ajouter des données quantitatives au baseline. Avant la visite, une partie du baseline est copiée au questionnaire de la caractérisation finale dans l’intention de faciliter le déroulement de l’interview. Ces cellules sont délimitées en gras. Une équipe d’enquêteurs est constituée d’au moins 1 socio-économiste, 1 nutritionniste et 2 agronomes. Chaque membre de l’équipe est responsable pour des sections spécifiques dans le questionnaire. Le questionnaire comprend une introduction (Sections A-B), qui inclue le profil du ménage, suivi par la section sur la nutrition et la santé (Section C), les sections agronomiques (Sections D-F) et les questions socio-économiques (Section G-J). La dernière section (Section K) est un ensemble des questions agronomiques spécifiques sur les parcelles comprenant des légumineuses.

Les membres d’équipe doivent optimiser le temps disponible. Les sections différentes peuvent être complétées avec des différents membres du ménage. Après avoir rempli le profile du ménage, ces membres peuvent être identifiés. Il faut faire attention de sélectionner les membres qui sont le mieux placés à répondre certaines questions. Il est par exemple possible que l’épouse connaisse mieux les caractéristiques des légumineuses que le chef du ménage. Les questions sur la nutrition et santé doivent être posées à la maman. Les questions spécifiques sur les parcelles cultivées avec des légumineuses peuvent être posées au chef, ou à un autre membre qui est impliqué dans la gestion de la ferme et les opérations de champs. Les agronomes assistent d’abord ensemble avec le dessin de la carte de la ferme (Section D). La carte est dessinée par le paysan, ou sous ses instructions. Un deuxième membre doit être présent, qui doit joindre un des agronomes dans le champ pour collecter les informations spécifiques sur les parcelles (après avoir dessiné la carte de la ferme). L’autre agronome remplit d’abord les Sections E-F avant de joindre le premier agronome au champ. Dans le cas ou il y a des champs à grande distance par rapport à la maison, les deux agronomes peuvent travailler séparément (dans ce cas, deux kits de matériels sont nécessaires). Alternativement, les deux agronomes peuvent partager les tâches. Le premier agronome fait

88

l’échantillonnage de sol et les mesures de densité de légumineuses et couverture par les mauvaises herbes. Le deuxième agronome fait les autres observations et complète l’interview avec le paysan.

____________________________________________________________ Note: ‘0’ signifie mesuré et la valeur est zéro; ‘-99’ signifie manque d’information; ‘-88’ signifie

non-applicable; ‘-77’ signifie que le répondant ne connaît pas la réponse.

Détails sur les observations 1. détermination de la pente

Une personne monte 10 m sur la pente (à mesurer à pas de pied), pendant que une personne reste en bas. Il faut alors estimer la distance verticale entre les deux personnes. La pente peut être calculé à base de cette distance verticale (cf. tableau):

2. dessiner la forme des parcelles, numérotage des coins, prise des coordonnées géographiques et altitude

Des piquets sont mis à chaque coin et dans le centre principal de la parcelle. Apres, les coordonnées (longitude et latitude) et l’altitude du centre principal sont notées, en utilisant un GPS. Les coins des parcelles sont numérotés (en utilisant des codes ‘C1’, ‘C2’, ‘C3’, etc.) et les coordonnées sont prises. La forme de la parcelle est dessinée et les positions des coins avec leur numéro sont indiquées.

3. mesurage de la densité des légumineuses

La densité des légumineuses est déterminée en plaçant un cadre de 1 × 1 m au milieu d’une diagonale qui connecte le centre avec un coin de la parcelle. Le nombre de plantes est compté. Ceci est répété pour 3 diagonales choisies par hasard. La somme pour les 3 cadres est notée. (cf. exemples pour des parcelles de forme triangulaire, rectangulaire, pentagonale et irrégulière)

4. mesurage de couverture par des mauvaises herbes Un bout de fil de 5m avec des marquages chaque 20cm par un nœud est place au milieu d’une diagonale qui connecte le centre avec un coin de la parcelle. La proportion des nœuds ou il y a des mauvaises herbes sont calculée. Ceci est répété pour 3 diagonales choisies par hasard. La moyenne pour les trois bouts est calculée et notée. (cf. exemples pour des parcelles de forme triangulaire, rectangulaire, pentagonale et irrégulière)

5. échantillonnage et étiquetage des sols

Des échantillons de sol (0-20cm) sont pris sur chaque diagonale qui connecte le centre avec un coin de la parcelle. Pour des champs avec 4 ou moins de coins, 2 échantillons sont pris sur chaque diagonale plus un échantillon dans le centre. Pour des champs avec 5 ou plus coins, un échantillon est pris sur chaque diagonale, plus un échantillon dans le centre (cf. exemples). Toutes les sondes collectées dans la parcelle sont mélangées et un sous-échantillon de 250g est gardé (une grande tasse). L’échantillon est étiqueté avec l’ID de l’exploitation (zone mandataire / site d’action / numéro de l’exploitation), la date d’échantillonnage et le code de la parcelle (‘P1’, ‘P2’, ‘P3’, etc.). Les sols sont séchés au soleil le plus vite possible et stockés.

6. échantillonnage et étiquetage des graines

L’agronome laisse des sachets étiquetés au chef de ménage (zone mandataire / site d’action / numéro de l’exploitation, le code de la parcelle (‘P1’, ‘P2’, ‘P3’, etc.), l’espèce de légumineuse et le nom de la variété). Le paysan est demandé de garder un échantillon de 50 graines de la récolte de chaque parcelle avec des légumineuses (en utilisant le dessin de la ferme). Il faut bien expliquer au paysan comment prendre cet échantillon d’une manière représentative, et de bien conserver ces 50 graines, protégés contre des ravageurs et humidité. Après collection, les échantillons sont stockés pour analyse.

distance verticale (m) pente (%) code 0 m 0 % 0 entre 0 et 0.5m entre 0 et 5% 1 entre 0.5 et 1m entre 5 et 10% 2 entre 1 et 2m entre 10 et 20% 3 entre 2 et 4m entre 20 et 40% 4 plus que 4m plus que 40% 5

89

7. échantillonnage et étiquetage de compost /fumier

L’agronome laisse des sachets papier étiquetés au chef de ménage (zone mandataire / site d’action / numéro de l’exploitation et le numéro suivant l’ordre dans le tableau en Section D.1.). Le paysan est demandé de garder un échantillon d’environ 500g en mettant un petit peu de matière dans le sachet chaque fois qu’il remplit un panier à appliquer dans le champ. Il faut mettre le sachet au soleil et sécher l’échantillon. Cet échantillon sera collecté après le début de la saison A’08.

8. prise des photos des parcelles et les systèmes de stockage de compost / fumier

Pour chaque parcelle avec des légumineuses, une photo est prise avec un membre de l’équipe ou le paysan au milieu de la parcelle, tenant une étiquette avec le numéro de l’exploitation et le code de la parcelle (‘P1’, ‘P2’, ‘P3’, etc.). Pour les systèmes de stockage de compost ou fumier, le numéro de l’exploitation et le numéro suivant l’ordre dans le tableau en Section D.1 sont indiqués.

90

1/2

1/2

densité des légumineuses – parcelle de forme triangulaire

C0

C1

C2

C3

1/2

densité des légumineuses – parcelle de forme rectangulaire

C0

C1 C2

C3C4

1/2

5m

couverture par des mauvaises herbes – parcelle de forme triangulaire

C0

C1

C2

C3

couverture par des mauvaises herbes – parcelle de forme rectangulaire

C0

C1 C2

C3C4

5m

1/3

1/3

échantillonnage de sol – parcelle de forme triangulaire

C0

C1

C2

C3

1/3

échantillonnage de sol – parcelle de forme rectangulaire

C0

C1 C2

C3 C4

1/3

1/3

1/3

Légende:

piquet à démarquer le centre (C0) et les coins (C1, C2, C2,…) de la parcelle

ligne imaginaire connectant le centre avec un des coins de la parcelle

cadre de 1m x 1m à mesurer la densité des légumineuses

ficelle de 5m avec noeuds après chaque 20cm à mesurer la

couverture par des mauvaises herbes dans la parcelle

point à prendre des échantillons de sol

91

densité des légumineuses – parcelle de forme pentagonale

C0

C1

C2

C3

C4

1/21/2

C5

densité des légumineuses – parcelle de forme irrégulière

C0

C1

C2

C3

C5

1/2

C0’

1/2

C4

C6

couverture par des mauvaises herbes – parcelle de forme pentagonale

C0

C1

C2

C3

C4

C5

5m

couverture par des mauvaises herbes – parcelle de forme irrégulière

C0

C1

C2

C3

C5

C0’

C4

C6

5m

échantillonnage de sol – parcelle de forme pentagonale

C0

C1

C2

C3

C4

C5

1/2

1/2

échantillonnage de sol – parcelle de forme irrégulière

C0

C1

C2

C3

C5

C0’

C4

C6

1/2

1/2

Légende:

piquet à démarquer le centre (C0) et les coins (C1, C2, C2,…) de la parcelle

ligne imaginaire connectant le centre avec un des coins de la parcelle

cadre de 1m x 1m à mesurer la densité des légumineuses

ficelle de 5m avec noeuds après chaque 20cm à mesurer la

couverture par des mauvaises herbes dans la parcelle

point à prendre des échantillons de sol

92

CIALCA – TSBF-CIAT PROJET LEGUMINEUSES SYSTEMES/SOLS, ACCES AU MARCHE et NUTRITION/SANTE

CARACTERISATION FINALE

SECTION A: INFORMATION GENERALE (10’) 1. Date de l’interview _______________________ 2. Nom du chef de l’équipe __________________________________________________________ 3. Noms des autres membres de l’équipe: ________________________________________________

______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________

4. ID de l’exploitation: (zone mandataire / site d’action / numéro de l’exploitation)

(même numéro que celui qui a été utilisé pendant le baseline) __________ /___________ /___________

Bas-Congo = BC Kigali-Kibungo = KK Sud-Kivu = SK Umutara = UM Kanga-Kipeti = 1 Gatore = 1 Kabamba = 1 Kabarore = 1 Lemfu = 2 Kabare = 2 Luhihi-Centre = 2 Murambi = 2 Mbanza Nzundu = 3 Mayange = 3 Lurhala-Centre = 3 Nyakigando = 3 Zenga = 4 Musenyi = 4 Mwegerera = 4 Rugarama = 4

5. Coordonnées GPS de la maison du chef de ménage (en dégrées décimales – à copier du baseline):

latitude (N/S) ______________; longitude (W/E) ______________; altitude: _____________ masl. 6. Observations sur la qualité de la maison:

Les murs (1=terre glaise; 2=bois; 3=briques; 4=autres: spécifiez) _____________________________________ Le toit (1=herbe; 2=tôles; 3=tuiles; 4=autres: spécifiez) ___________________________________________

93

SECTION B: PROFILE DU MENAGE (30’) [nom de l’enquêteur: _________________________________; temps début Section B: _______________]

scolarisation members d

e m

énag

e

nom (définition: un ménage = un groupe de gens qui vivent et mangent ensemble)

âge (ans)

sexe (1=

male; 2=

féminin)

résident (0=

non; 1=oui)

relation par rapport au chef 1=chef 2=époux/épouse 3=fils/fille 4=parent du chef 5=petit-enfant 6=belle-famille 7=relatif du chef 8=relatif de l’époux/se 9=serviteur/servante 10=autres: spécifiez

(ans) niveau plus élevé atteint 0=aucun 1=primaire 2=secondaire 3=tertiaire

Quelle activité prend le plus de temps de ce membre ? 1=au champ (propre ferme) 2=labour au champ des autres paysans 3=ménage à la maison 4=école 5=autres: spécifiez

Combien de te

mps a

u champ

par rap

port a

ux autre

s mem

bres?

(0=peu; 1=

moyen

; 2 =beau

coup)

répond aux questions? 1 = section B (profile de ménage); 2 = section C (nutrition et santé) 3 = section D (carte de la ferme); 4 = section E-F (questions agronomiques); 5 = sections G-J (questions soc.-économ.); 6 = section K (données des champs). ----------------------------------------------- $ = a participé dans le PRA; $$ = a participé dans le baseline; # = membre d’une association participante; ## = accueille un essai de démonstration.

enfant sé

lectio

nné(e) e

ntre

2 et 5

ans ? (m

arquez avex ‘×’)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

[temps fin Section B: _______________]

94

SECTION C: NUTRITION & SANTE (60’-90’) [nom de l’enquêteur: ___________________________; temps début Section C: _______________] Note: dans les ménages additionnels à caractériser pour la nutrition, il faut d’abord faire la liste des enfants en utilisant le tableau de Section B, et remplir les 6 premières colonnes (nom, âge, sexe, résidence, relation par rapport au chef et scolarisation). Etat socio-démographique du ménage 1. Qu’est-ce qui est la source principale d’eau à boire pour l’enfant sélectionné(e) du ménage à ce moment? (1=eau de robinet dans la maison; 2=eau de robinet dans la parcelle; 3=eau de conduite public; 4=puits; 5=rivière; 6=réservoir à collecter eaux pluviales; 7=autres: spécifiez) ____________________________________________ 2. Comment est-ce que vous préparez l’eau à la maison avant de boire? (1=bouillir; 2=filtrer; 3=laisser se déposer; 4=javelliser; 5=pas de préparation; 6=autres: spécifiez – plusieurs numéros possible) ______________________ 3. Comment est-ce que vous stockez l’eau à la maison avant de boire? (1=réservoir fermé; 2=réservoir ouvert; 3=non applicable (p.ex. eau de conduite); 4=autres: spécifiez) __________________________________________ 4. Quel type de toilette utilisez-vous? (1=toilette à chasse; 2=latrine puisard; 3=toilette à trou; 4=pas de toilette; 5=autres: spécifiez) ___________________________________________________________________ 5. Qu’est-ce qui est la source principale de combustible à cuire dans le ménage? (1=électricité; 2=gaz; 3=pétrole; 4=charbon de bois; 5=du bois; 6=autres: spécifiez) _________________________________________ 6. Conservez-vous les restes de nourritures après le repas? (0=non; 1=oui) ________________________ Si oui, pendant combien de temps? (0=mangé le même jour; 1=pendant 1-2 jours; 2=pendant 3-5 jours; 3=>5 jours)? _____________________________________________________________________________ Si oui, décrivez comment vous conservez les restes. (Décrivez seulement pour les nourritures principales) __________________________________________________________________________________________________________________________________________________________

Informations spécifiques sur un enfant sélectionné(e) arbitrairement d’âge entre 2 et 5 ans 7. Décrivez comment vous avez alimenté votre enfant pendant les premiers 6 mois. (1=allaitement au sein exclusif; 2=allaitement au sein + autres aliments (spécifiez); 3=lait de bétail frais; 4=lait en poudre; 5=lait soja; 6=autres aliments: spécifiez – plusieurs numéros possible) ____________________________________ ________________________________________________________________________________ 8. Décrivez comment vous avez alimenté votre enfant entre 6 et 12 mois. (1=allaitement au sein exclusif; 2=allaitement au sein + autres aliments (spécifiez); 3=lait de bétail frais; 4=lait en poudre; 5=lait soja; 6=autres aliments: spécifiez – plusieurs numéros possible) ____________________________________ ________________________________________________________________________________ 9. Est-ce que votre enfant a une carte de santé ? (0=non; 1=oui) _______________________________

Si oui, note le poids et la date à la dernière visite au centre de santé. _________________________ _____________________________________________________________________________ Si oui, note la tendance de croissance après les dernières trois visites (1=bon gagne de poids; 2=pas de gagne de poids; 3=perte de poids) ____________________________________________________________

10. Mesures anthropométriques à prendre:

-Poids actuel de l’enfant (kg) ______________________________________________________ -MUAC (circonférence du mi-haute bras gauche) (cm) ___________________________________ -Taille de l’enfant (cm) ___________________________________________________________

11. Est-ce que votre enfant a déjà été admis dans un centre de nutrition ? (0=non; 1=oui) ____________

Si oui, combien de fois? __________________________________________________________ Si oui, pour combien de temps? (spécifiez pour chaque admission) _________________________ _____________________________________________________________________________ Si oui, quel aliment est-ce que l’enfant avait reçu? ______________________________________ _____________________________________________________________________________

12. Présence de kwashiorkor/marasme après examen physique ? (0=non; 1=oui) __________________ Si oui, quels symptômes ? (1=manque de masse musculeux; 2=abdomen étendu; 3=peau ridé; 4=poils rêches et bruns;

95

5=oedème; 6=autres: spécifiez – plusieurs numéros possible) ________________________________________ _____________________________________________________________________________

13a. Votre enfant, a-t-il eu la diarrhée pendant les 3 mois passés ? (0=non; 1=oui) __________________ 13b. Votre enfant a-t-il eu du sang dans les selles pendant les 3 mois passés ? (0=non; 1=oui) _________ 13c. Votre enfant a-t-il eu difficulté à respirer pendant les 3 mois passés ? (la toux, respiration accélérée, respiration

nez/bouche bloquée, inhalation thorax difficile - 0=non; 1=oui) _____________________________ 14. Votre enfant a-t-il été admis à l’hôpital pendant les 3 mois passés? (0=non; 1=oui) ______________

Si oui, combien de fois ? _________________________________________________________ Si oui, pourquoi ? ______________________________________________________________

Nutrition de l’enfant sélectionné(e) 15. Combien de fois votre enfant mange-t-il les aliments suivants? (Spécifiez le nombre de fois) nombre de fois pendant les 7

derniers jours nombre de fois pendant les 30 derniers jours

lait frais de bétail

lait en poudre

lait soja

viande

poisson

œufs

légumes spécifiez

haricots

soja

niébé

arachide

pois cajan

manioc

banane

patate douce

mais

riz

fruits spécifiez:

aliment de bébé formulé / fortifié, spécifiez:

bouillie des céréales

autres spécifiez:

autres spécifiez:

autres spécifiez:

96

20. 24-heures rappel de nourriture de l’enfant sélectionné(e) – date (jj/mm/aa): __________ /__________ /__________ Notez les aliments; spécifiez les quantités ou possible (p.ex. 1 tasse de thé: (2): 1, 1 tasse).

boisson (1)

produits laitiers (2)

viande/poisson/ œufs (3)

de l’huile/ lipides (4)

céréales (5)

banane/plantain (6)

racines/ tubercules (7)

légumineuses (8)

légumes (9)

fruits (10)

petit-déjeuner

pause du matin

repas de midi

pause de l’après-midi

repas du soir

(1): 1=thé; 2=lait; 3=de l’eau; 4=jus de fruit; 5=sucrée; 6=autres: spécifiez. (2): 1=yaourt; 2=fromage; 3=autre: spécifiez. (3): 1=boeuf; 2=chèvre; 3=porc; 4=poulet; 5=intestins; 6=poisson; 7=oeufs; 8=autres: spécifiez. (4): 1=graisse d’animal; 2=l’huile végétale; 3=beurre; 4=margarine; 5=autres: spécifiez. (5): 1=riz; 2=mais; 3=millet; 4=sorgho; 5=sosome; 6=pain; 7=autres: spécifiez. (6): 1=banane dessert; 2=banana à cuire; 3=plantain; 4=matoke; 5=farine de banane; 6=chips de banane; 7=autres: spécifiez. (7): 1=manioc; 2=yams; 3=patates douces; 4=pommes de terre; 5=autres: spécifiez. (8): 1=haricots; 2=soja; 3=niébé; 4=arachide; 5=petits pois; 6=pois cajan; 7=autres: spécifiez. (9): 1=chou; 2=amarante; 3=tomate; 4=oignons; 5=feuilles de manioc; 6=feuilles vertes (comme les épinards); 7=autres: spécifiez. (10): 1=orange; 2=mangue; 3=papaye; 4=ananas; 5=fruits de passion; 6=avocat; 7=autres: spécifiez.

[temps fin Section C: _______________]

97

SECTION D: CARTE DE LA FERME (45’) [nom de l’enquêteur: _______________________; temps début Sections D-F: _______________] - A dessiner par le paysan ou sous son instruction sur un papier A3. Dépendamment de la situation, le dessin peut être photographié et rester chez le paysan, ou peut être amené et réduit à taille A4.

- Indiquez où se trouve la maison du chef, les étables du bétail, les pâturages et les systèmes de stockage de compost et/ou fumier.

- Indiquez toutes les parcelles de la ferme (soit louée, propriété ou commune); les parcelles qui appartiennent au chef mais qui se sont louées à un autre paysan peuvent être indiquées, mais il faut alors demander ce qui se passe avec la récolte. Indiquez si les parcelles sont contiguës (attachées) ou éloignées (détachées) de la maison du chef du ménage. Pour les parcelles détachées, indiquez la distance entre la parcelle et la maison du chef, si nécessaire calculée en termes de temps à se promener (1km = 12 minutes).

- Numérotez chaque parcelle (‘P1’, ‘P2’, ‘P3’, etc.). - Indiquez les cultures actuellement (saison B’07) dans les champs (ou jachère) en bleu, et les cultures (ou jachère) qui étaient aux champs dans la saison passée (saison A’07) en rouge.

Les informations spécifiques pour les parcelles (Section K), ne doivent être complétés que pour les parcelles actuellement cultivées par une légumineuse (soit seule ou en association), et pour les parcelles cultivées avec une légumineuse pendant la saison passée. Des parcelles qui appartiennent au ménage mais qui se sont louées à autres paysannes, dont la récolte ne revient pas au ménage interviewé, ne doivent pas être caractérisées. Les informations suivantes doivent être copiées du baseline et peuvent aider à réaliser avec le dessin de la carte de la ferme (copiez du baseline sections 4.2 et 4.5):

superficie (en ha)

distance par rapport à l’habitation

champs de case

1

2

3

4

5

6

champs de marais/bas-fond

1

2

3

4

champs de collines

1

2

3

4

5

boisement

1

2

pâturages

1

2

ordre d’importance

liste de cultures

1

2

3

4

5

6

7

8

9

10

98

SECTION E: PROFILE DE LA FERME (30’) 1. Utilisation des intrants organiques type de matière organique Est-ce que c’est composté?

(1) spécifiez la source (0=non; 1=oui) si oui, pendant combien de temps ? (nombre de mois)

quantité de matière organique appliqué sur la ferme pendant cette saison B’07 (en unités locales; spécifiez le poids d’une unité locale)

système de stockage (2)

(1): 1=fumier; 2=résidus des cultures; 3=engrais verts; 4=autres: spécifiez – indique plusieurs numéros pour des mélanges; (2): 1=tas en plein air; 2=tas sous des arbres; 3=tas sous un toit; 4=fosse en plein air; 5=fosse sous des arbres; 6=fosse sous un toit;

7 = autres: spécifiez. 2. Gestion de bétail et fumier

Le fumier est-il collecté? type nombre to

tale su

r la fe

rme

nombre qui ap

partie

nnent

au m

énage

gestion d’aliment (1)

matières utilisées comme aliment (à remplir seulement pour le bétail en stabilisation)

combien de fois par semaine? (0=non;

1=oui) si oui, combien de fois par mois?

vache locale

vache améliorée

mouton

chèvre

porc

(1): 1=pâturage individuel; 2=pâturage collectif; 3=divagation; 4=en stabulation avec aliments venant de l’extérieur; 5=en stabulation avec plantation des fourrages/herbes; 6=achat des suppléments; 7=autres: spécifiez.

99

SECTION F: CARACTERISTIQUES DES VARIETES LEGUMINEUSES CULTIVEES (60’) Note 1: toutes réponses sont des opinions du paysan, pas des mesures absolus; le paysan donne des scores relatifs par rapport aux autres variétés. Note 2: le marché local est le marché le plus proche (cf. baseline 7.4a); le marché régionale est le marché avec plus probablement le prix le plus élevé dans la région (soit urbain ou marché régional). Variations en prix

sont les prix minimum et maximum observés pendant l’année passée.

à remplir si le paysan connaît les prix, même s’il ne vend pas

VARIETE productio

n

(1)

beso

in de

tuteur? (2)

précocité

(3)

résiste à

la

sécheresse

(4)

résiste a

ux

forte

s pluies

(4)

résiste a

ux

peste

s et

maladies (4

)

résisite

à la

basse

fertilité de so

l (4)

couleur taille

des

graines

(5)

% qui se

ra

vendu ce

tte

saiso

n (%

)

pure ou

mélangée?

(6)

min-max prix sur le marché locale (en FR/kg)

min-max prix sur le marché régional (en FR/kg)

haricots

1. - -

2. - -

3. - -

4. - -

niébé

1. - -

2. - -

3. - -

arachide

1. 0 - -

2. 0 - -

soja

1. 0 - -

2. 0 - -

3. 0 - -

pois cajan

1. 0 - -

2. 0 - -

autres: spécifiez

- -

(1): 1=très pauvre; 2=pauvre; 3=moyen; 4=bon; 5=très bon; (2): 0=non; 1=oui; (3): 0=précoce, 1=moyenne; 2=longue durée; (4): 1=résistance très pauvre (avec beaucoup de perte de production); 2=résistance pauvre (avec perte de production significatif); 3=résistance moyenne; 4=résistance bonne (mais le paysan connaît ou cherche encore des variétés meilleures); 5=résistance très bonne (la variété est assez bonne et il n’y a pas de problème en terme de production; (5): 1=très petit; 2=petit; 3=moyen; 4=large; 5=très large; (6): 1=pure; 2=mélangé.

[temps fin Sections D-F: _______________]

100

SECTION G: APPUI INSTITUTIONEL (30’) [nom de l’enquêteur: ____________________________; temps début Sections G-J: _______________] Formation 1. Avez-vous suivi une formation au champ ? (0=non; 1=oui) ______________________________________

Si oui, combien de jours pendant les 5 dernières années? _____________________________________ 2. Combien de fois par mois écoutez-vous des programmes d’agriculture? __________________________ 3. Combien de fois avez-vous assisté à une réunion des associations paysannes / jour de champ pendant l’année

passée ? ____________________________________________________________________ Quel type de réunion ? _______________________________________________________________

Information sur le marché 4. Y a-t-il une organisation qui vous a appuyé sur la connaissance comment améliorer la commercialisation de vos

produits (0=non; 1=oui) _________________________________________________________ Si oui, quelle est le nom de cette organisation ? ____________________________________________ Si oui, sur quelle culture/élevage est-ce qu’ils vous ont aidé ? __________________________________ _________________________________________________________________________________ Si oui, quel type d’appui ont-ils fourni ? __________________________________________________ _________________________________________________________________________________ Si oui, quand était le dernier contact avec l’organisation ? _____________________________________

5. Comment obtenez-vous l’accès aux informations suivantes ? (marquez dans chaque cellule si 0=pas de source d’information, 1=source principale d’information, ou 2=source secondaire d’information)

information obtenu par information sur

radio journal visites au marché commerçants voisins ONG autres: spécifiez

engrais minéraux

fertilisants organiques

accès aux sémences

caractéristiques des variétés

qualité des graines

prix des graines

Commercialisation collective 6. Connaissez-vous des initiatives de commercialisation collective des produits agricoles? (0=non; 1=oui) ____ 7. Avez-vous déjà participé dans une de ces initiatives ? (0=non; 1=oui) _____________________________

Si oui, quand était la dernière fois que vous avez participé ? ___________________________________ Si oui, pour la commercialisation de quelle culture/élevage ? __________________________________ _________________________________________________________________________________ Si oui, qui avait organisé cette initiative de commercialisation collective ? _________________________ _________________________________________________________________________________ Si oui, quels avantages en avez-vous tirés ? Expliquez pourquoi c’était avantageux. _________________ _________________________________________________________________________________

Contrats de production 8. Connaissez-vous des initiatives de production sous contrat ? (0=non; 1=oui) __________________________ 9. Etes-vous actuellement impliqué dans une de ces initiatives ? (0=non; 1=oui) _______________________

Si oui, pour quelle culture/élevage ? _____________________________________________________ _________________________________________________________________________________ Si oui, quelle organisation achète vos produits ? ____________________________________________ Si oui, quels avantages en avez-vous tirés ? Expliquez pourquoi c’était avantageux. _________________ _________________________________________________________________________________

101

SECTION H: FONCTIONS ET CONTRAINTES DE COMMERCIALISATION (60’)

1. Avez-vous reçu une formation sur une des fonctions de commercialisation suivantes?

activité formation reçu? (0=non; 1=oui)

de quelle organisation?

combien de fois pendant l’année passée?

Ca a change la mise en oeuvre de cette activité? (0=non; 1=peu; 2=oui; 3=beaucoup)

rassemblage

triage

normalisation

mise en emballage

transport

transformation

Note: rassembler = collecter et mélanger des graines des différents parcelles/producteurs; trier = enlever les graines de mauvaise qualité; normaliser = sélectionner les graines d’une certaine taille; emballer = mettre en certaines emballages ce qui augmente les possibilités de commercialisation; transport = amener à un marché plus éloigné que le marché local; transformation = toute forme de conversion qui augmente les possibilités de commercialisation.

2. Quelles légumineuses vendez-vous ? _____________________________________________________ ____________________________________________________________________________________ Avant de vendre vos produits légumineuses, est-ce que vous faites les activités suivantes? légumineuse

rassem

blage

triage

norm

alisation

embalage

transp

ort

transform

ation prix du

produit brut (en FR/kg)

prix du produit converti (en FR/kg)

Trouvez-vous l’activité rentable par rapport à l’effort qu’elle demande ? (0=non; 1=peu; 3=oui; 4=beaucoup)

problèmes avec les activités (question ouverte)

haricot

niébé

arachide

soja

pois cajan

autre: spécifiez

Note: les prix des produits brut et converti se comparent à un moment donné par rapport à un paysan qui n’a pas fait l’activité.

3. Stockage et pertes pendant le stockage légumineuse normalement

stocké? (0=non; 1=oui)

quelle proportion? (spécifiez le %)

pendant combien de semaines?

raison pour le stockage (1)

perte pendant le stockage (2)

pourquoi cette perte ?

haricots niébé arachide soja

pois cajan

autres: spécifiez

(1): 1=impossible de vendre au moment de la récolte; 2=prix bas au marché; 3=stocké à consommer plus tard; 4=stocké comme semence; 5=autres: spécifiez (plusieurs numéros possible).

(2): 0=pas de perte; 1=peu de perte (moins que 10%); 2=entre 10 et 25% perdu; 3=entre 25 et 50% perdu; 4=plus que 50% perdu.

102

4. Quelles sont les contraintes principales pour la commercialisation de vos légumineuses? Note: ces contraintes peuvent se rencontrer dans toute la filière de la production. légumineuse 1er constrainte principale 2ieme constrainte principale 3ieme constrainte principale

haricots

niébé arachide soja pois cajan autres: spécifiez

1=manque de terrain; 2=loyer de terrain cher; 3=manque de capital/crédit; 4=manque de labour; 5=bas fertilité de sol; 6=sécheresse; 7=pestes/maladies; 9=manque des variétés améliorées/adaptées; 10=manque de semences; 11=manque des engrais minéraux; 12=manque des intrants organiques; 13=manque de marché; 14=distance au marché; 15=mauvaise route au marché; 16=tracasseries pour arriver au marché; 17=hautes taxes; 18=prix bas au marché; 19=haute fluctuations/incertitude des prix; 20=autres: spécifiez. SECTION I: INVESTISSEMENT DES BENEFICES OBTENUES PAR LA VENTE DES PRODUITS

(30’) Comment dépenserez-vous l’argent réalisé avec la vente des produits agricoles? (Rangez d’abord les dépenses, puis estime le pourcentage en utilisant 30 cailloux à partager parmi la liste de dépenses) dépense rang % des

dépenses raison pour la décision sur la dépense

loyer de terrain

labour pour les opérations de champ

semence/matérielle de plantation

engrais minéraux

intrants organiques

expérimentation et formations d’agriculture

achat des aliments

frais de scolarité des enfants

frais médicaux de la famille

dépenses sociaux (divertissement des visiteurs, mariages, enterrements,...)

autres: spécifiez

autres: spécifiez autres: spécifiez

total 100%

SECTION J: CONCLUSION (10’) Quelles sont vos 3 suggestions par ordre d’importance pour la commercialisation des légumineuses dans votre milieu ? (question ouverte) Priorité 1 ____________________________________________________________________________ _________________________________________________________________________________ Priorité 2 ____________________________________________________________________________ ____________________________________________________________________________________ Priorité 3 ____________________________________________________________________________ ____________________________________________________________________________________

[temps fin Sections G-J: _______________]

103

SECTION K: INFO PARCELLES

Propriété ? (1=propre propriété; 2=loué; 3=communal; 4=autres: spécifiez) Quand est-ce que la parcelle a été prise en cultivation pour la 1ère fois ? (donne l’année)

Appréciation du paysan sur la fertilité du sol (1=pauvre; 2=moyenne; 3=bonne) Nom local du type de sol (si connu) Contrainte principale pour la production de la légumineuse (1=érosion; 2=bas fertilité du sol; 3=mauvaises herbes; 4=pestes/maladies; 5=pierres; autres: spécifiez)

Labour de terrain ? (0=pas préparé; 1=avec hoe; 2=labouré avec vaches; autres: spécifiez) Cultures au champ pendant la saison passée (A’07) Cultures envisagées pour la saison suivante (A’08) Spécifiez le nom de la variété légumineuse Spécifiez la source de semence de la légumineuse

Engrais minéraux appliqués? (0=non; 1=oui) si oui, spécifiez le type et la dose (en unités locales; spécifiez le poids d’une unité locale) type: dose: si oui, spécifiez le temps d’application (1=à la plantation, 2=autres: spécifiez) mode d’application (1=épandu; 2=épandu et incorporé; 2=dans la ligne; 3=dans le poquet; 4=autres: spécifiez) Intrants organiques appliqués? (0=non; 1=oui) si oui, spécifiez le type et la dose (en unités locales; spécifiez le poids d’une unité locale) type: dose: si oui, spécifiez le temps d’application (1=à la plantation, 2=avant la plantation, pendant la préparation du terrain; 3=à différents moments irrégulières pendant la saison; 4=autres: spécifiez)

mode d’application (1=épandu; 2=épandu et incorporé; 2=dans la ligne; 3=dans le poquet; 4=autres: spécifiez) Insecticide ou herbicide appliqué? (0=non; 1=oui) si oui, spécifiez le type (1=local: spécifiez; 2=produit chimique: spécifiez) estime la production attendue (en unités locales; spécifiez le poids d’une unité locale en kg) estime la production attendue des espèces en association (en unités locales) Quoi sera/était fait avec la récolte de la légumineuse? (1=vendue entièrement, 2=consommée, 3=partiellement vendue, partiellement consommée, 4=autres: spécifiez)

Quoi sera/était fait avec les résidus de la légumineuse? (>1 option possible) (1=incorporés; 2=compostés; 3=aliments bétail; 4=mangés par la famille; 5=utilisé comme combustible; 6=vendus; 7=autres: spécifiez)

Le paysan a-t-il été informé de garder 3 cuillères de graines ? (0=non; 1=oui)

Code de la parcelle (numéro suivant le code (‘P1’, ‘P2’, ‘P3’, etc.) dans la carte de la ferme) Altitude du centre principal de la parcelle (‘C0’) m

Coordonnées GPS du centre principal (‘C0’): S E/O

Coordonnées GPS de chaque coin de la parcelle: coin 1 (‘C1’): S E/O coin 2 (‘C2’): S E/O coin 3 (‘C3’): S E/O coin 4 (‘C4’): S E/O coin 5 (‘C5’): S E/O

Croquis de la parcelle et numéro de chaque coin:

coin 6 (‘C6’): S E/O

Attaché / détaché de la domaine de la maison (1=attaché; 2=détaché)

Position dans le paysage (1=plateau; 2=haute-pente; 3=mi-pente; 4=basse-pente; 5=vallée) Pente (0=0%, 1=0-5%, 2=5-10%, 3=10-20%, 4=20-40%, 5=>40%) Cultures actuellement au champ (liste les cultures) S’il y a actuellement une légumineuse, spécifiez l’espèce mesure la densité des légumineuses (nombre de plantes par 3m2) S’il y a une légumineuse en association, spécifiez l’espèce en association mesure la densité de l’espèce en association (nombre de plantes par 10×10m)

Mesure et calcule la couverture par des mauvaises herbes (%) Type dominant mauvaises herbes (1=herbes, 2=feuilles larges, 3=impérata, 4=autres: spécifiez) Présence des of roches, pierres, cailloux ou gravier à la surface (échelle 1=0-5%; 2=5-25%; 3=25-50%; 4=50-75%; 5=75-95%; 6=95-100%)

Erosion visible (0=non; 1=à nappe; 2=à sillon; 3=à rigole) Techniques d’in situ collection d’eaux pluviales (0=absente; 1=zai; 2=plates-bandes; 3=plates-bandes serrées; 4=demi-lune; 5=autres: spécifiez)

Présence des structures anti-érosifs (0=absente; 1=végétatif; 2=structurel; 3=les deux) si oui, type de structures anti-érosifs (spécifiez l’espèce dominante pour les structures végétatifs; spécifiez le type pour les autres structures, p.ex. murs de pierre, fanya juu; fanya chini; terrasses; autres: spécifiez)

si oui, nombre de structures anti-érosifs (note: les structures en bas ne font pas partie de la parcelle) Echantillon de sol (0-20cm) pris? (0=non; 1=oui) Photo prise ? (0=non; 1=oui)

104

AAANNNNNNEEEXXX 333::: LLLEEEGGG---222::: LLLEEEGGGUUUMMMEEE GGGEEERRRMMMPPPLLLAAASSSMMM DDDEEEMMMOOONNNSSSTTTRRRAAATTTIIIOOONNN

TTTRRRIIIAAALLLSSS Objectives: This set of trials aims (i) to obtain detailed farmer feedback on germplasm demonstrated; (ii) to obtain a multi-locational (1 site = 1 replicate) agronomic evaluation of the germplasm demonstrated (G × E); (iii) to obtained information on BNF of all germplasm demonstrated, on low-P tolerance in selected soybean cultivars, on tolerance to low soil fertility in selected bean cultivars and on micronutrient contents in produce of selected biofortified bean cultivars; (iv) to multiply seeds for the following 2007 season. Sites: Demonstration trials are installed in a field identified by the associations. Number of sites = 4 mandate areas × 4 action sites × 2 associations = 32 sites

mandate area action site association 1 2

Musenyi

3 A KAMEGA 4

Mayange B TUZAMURANE

5 A BENISHYAKA 6

Gatore

7 A DUTERIMBERE A 8

Kigali-Ngali/Kibungo

Kabare B DUTERIMBERE B

9 A DUFATANYE 10

Nyakigando B INGANDURARUGO

11 A ISOKO Y’UBUMWE 12

Kabarore B ABAHUJUMUGAMBI

13 A TWISUNGANE 14

Rugarama B IMBARAGA

15 A DUFASHANYE 16

Umutara

Murambi B IRIBA

17 A CINAMULA 18

Lurhala B ALEMALU

19 A RUSINAME 20

Luhihi B RHUBEHAGUMA

21 A MAENDELEO 22

Kabamba B TUUNGANE

23 A APACOV 24

Sud-Kivu

Burhale B ABAGWASINYE

25 A CALDZ 26

Zenga B ADESCO

27 A ADEKO 28

Mbanza Nzundu B ACKI

29 A APDKI 30

Kanga Kimpeti B ACDPP

31 A ADERKI 32

Bas-Congo

Lemfu B APEKI

Germplasm: Soybean and bean have been identified as test crops. -soybean: selected TGx cv. from Nairobi, selected TGm cv. from Nairobi adapted to low-P conditions, 2 cv.

from Uganda, 6 cv. from ISAR adapted to high altitude. -beans: BIWADA cv. (drought-tolerant), BILFA cv. (tolerant to low soil fertility) and biofortified cv.

105

soybean bush bean climbing bean microplots local variety

cv1: SB2(TGx1831-32E) cv2: SB4(TGx1871-12E) cv3: SB6(TGx1895-4F) cv4: SB9(TGx1895-49F) cv5: SB14(TGx1878-7E) cv6: SB15(TGx1889-12F) cv7: SB17(TGx1893-10F) cv8: SB19(TGx1740-2F) cv9: SB20(TGx1448-2E) cv10: SB24(MAKSOY1a) cv11: SB25(NAMSOI4m) cv12: Duiker(ISAR) cv13: Bossier 06B(ISAR) cv14: 449/16/6 06B(ISAR) cv15: Soprosoy 06B(ISAR) cv16: Peka6 06B(ISAR) cv17: Ogden(ISAR) cv18: TGM1781(ISAR)

local variety

BIOFORTIFIED: cv1: BRB194 cv2: CODMLB003 cv3: CODMLB007 cv4: CODMLB030 cv5: CODMLB078 cv6: MAHARAGI-SOJA cv7: MARUNGI cv8: MLB49-89A cv9: M’SOLE cv10: ZKA93-10m/95

BILFA: cv11: AFR619 cv12: AFR708 (+biofort) cv13: ARA4 (+biofort) cv14: CIM9314-36 cv15: CIM9331-1 cv16: CNF5520 cv17: VEF88(40)L1P4T6 cv18: ECA-PAN021 cv19: HM21-7 (+biofort) cv20: LSA144 (+biofort) cv21: T8426F11-6F cv22: UBR(92)25 cv23: ZAA5/2

BIWADA: (only to be included in Umutara and Bas-Congo mandate areas) cv24: GNP585(BIWADA) cv25: Rab618(BIWADA) cv26: Rab619(BIWADA) cv27: Rjb7(BIWADA)

local variety

BIOFORTIFIED: cv1: AND10 cv2: G59/1-2 cv3: KIANGARA cv4: LIB1 cv5: MLV06-90B cv6: MLV59/97B cv7: VCB81012 cv8: VCB81013 cv9: VNB81010

single lines cv19: SB38(TGx1903-2F) cv20: SB39(TGx1903-3F) cv21: SB42(TGx1904-4F) cv22: SB44(TGx1908-8F) cv23: SB45(TGm1909-3F) cv24: SB46(TGM1420) cv25: SB47(TGm1511) cv26: SB49(TGm1360) cv27: SB51(TGm1196) cv28: SB54(TGm1576)

note: BIWADA cultivars only to be tested in Umutara and Bas-Congo mandate areas, where drought problems occur; if seeds available, also to be tested in Kigali-Ngali/Kibungo; note: in each trial (every association) a local variety of soybean, climbing bean and bush bean variety should be included; note: varieties of which enough seed is available are included in microplots (4 lines of 2 m), varieties of which limited seeds are available are included as single lines (1 line of 2 m).

Plant spacing -soybean: 4 rows, 2 m long (plot = 4 × 0.75 × 2 = 6 m2)

0.75 m between rows and 0.05 m between plants within a row drill seeds (2 cm depth) and thin to 5 cm between-plant distance at 3 weeks after planting

-bush beans: 4 rows, 2 m long (plot = 4 × 0.40 × 2 = 3.2 m2) 0.40 m between rows and 0.10 m between plants within a row plant 1 seed per hill

-climbing beans: 4 rows, 2 m long (plot = 4 × 0.50 × 2 = 4 m2) 0.50 m between rows and 0.10 m between plants within a row plant 1 seed per hill

note: area needed = 5.6 are without border zones; total area should be about 7 are.

106

Treatments -soybean: with and without TSP applied at a rate of 30 kg P ha-1

(band application: make trenches and apply the TSP in the trench, cover with a little soil and plant seeds at 5 cm distance between seeds) 25 kg P ha-1 = 125 kg TSP ha-1 = 12.5 g TSP m-2 = 75 g TSP per plot = 18.75 g TSP per row

-bush beans: with and without FYM (from goats) applied at a rate of 5 t fresh matter ha-1 (broadcasted over the entire plot) 10 t fresh matter ha-1 = 1 kg m-2 = 3.2 kg per plot

-climbing beans: with and without FYM (from goats) applied at a rate of 5 t fresh matter ha-1 (broadcasted over the entire plot) 10 t fresh matter ha-1 = 1 kg m-2 = 4.0 kg per plot

note: amounts of inputs per demonstration trial (association) required are: -TSP: 21 plots × 75 g TSP per plot + 10 single lines × 18.75 g TSP per row = 1.763 kg TSP per association -FYM: 30 plots × 3.2 kg per plot + 12 plots × 4.0 kg per plot = 144 kg FYM (fresh matter) per association note: the FYM needs to be well homogenized and representatively sampled. The moisture content needs to be determined and a dried sample needs to be sent to Nairobi for analysis. Trial design Split plot design: 1 association = 1 replicate; main plots = species × inputs; split plots = cultivars The various species and treatments are blocked in main plots; the trial comprises 6 blocks:

1. soybean, no inputs 2. soybean + TSP 3. bush bean, no inputs 4. bush bean + FYM 5. climbing bean, no inputs 6. climbing bean + FYM

Each block consists of two sub-blocks, a first comprising the microplots with different cultivars and a second comprising cultivars in single lines. Within the various sub-blocks, the location of the various cultivars is completely randomized. A possible design with the 6 blocks is attached. note: each block includes also two microplots where maize and sorghum are grown (alternatively, a long and a short duration maize can be grown) as reference crops for BNF assessment. For maize, 2 seeds are planted per hill and thinned to 1 plant at 3 weeks after planting. note: keep border areas between the various sub-blocks and between the various microplots. note: keep the area weed-free. note: spray insecticide as needed. Biophysical observations 1. Soil sampling Soil samples are taken at the block-level, prior to application of inputs. Per block, 9 cores (0-15 cm) are taken and mixed, air-dried and stored. At least 2 kg is required for soil analysis. Following sampling strategy is followed:

1/6 1/3 1/3 1/61/6

1/3

1/3

1/6

1/6 1/3 1/3 1/61/6

1/3

1/3

1/6

2. Weather data If possible, precipitation is recorded on a daily basis, but this is not essential as satellite data is also available.

107

3. Physiology and disease scoring Physiology and disease scorings are recorded according to ECABREN protocols.

4. Biomass and BNF assessment Aboveground biomass is sampled at 50 % podding (“50% formation des gousses”). Within the middle two rows of the microplots, a random section of 50 cm of plants (equivalent to 10 soybean plants, or 5 bean plants) is cut at the ground level. The random section should be at least 20 cm away from the border. The total fresh weight of the biomass is taken, using an accurate electronic balance. Subsequently, the biomass is separated in leaves, stems and pods, and the fresh weights of the 3 fractions are also recorded. The plant fractions are then dried (65 oC) and dry weights are recorded, using the same balance. At each time biomass assessments are made for species/cultivars that have reached the 50% podding stage (this will occur at different times during the season), 3 maize and 3 sorghum plants are taken from the reference plots. This is extremely important as these crops serve as a reference for determination of BNF in the legume cultivars.

note: since nodule sampling, counting and analysis take a large effort, and since nodule numbers relate weakly with BNF, nodulation will not be assessed.

5. Harvest At crop maturity, all pods (“gousses”) should be harvested from the net plots (the net plot is obtained by removing both outside rows and removing 20 cm – equal to 2 bean plants or 4 soybean plants – on each end of the middle rows). The total fresh weight of the pods is taken (using the 2 kg balance) and a sub-sample is taken and immediately weighed (using the 200 g balance). The sub-sample is air-dried, oven-dried (65 oC overnight), and separated into grains and husks (“fanes”). The dry grains and husks are then weighted with the same 200g balance.

note: use of grains harvested: The grains will be used for (i) taste tests at harvest, (ii) use in the next generation of demonstration trials (only some varieties), (iii) further multiplication by the farmer groups (some varieties), and (iv) further multiplication on-station to keep a minimal stock of seeds (all varieties).

Activities around the demonstrations 1 .Farmer group training At each important trial management event (planting, fertilizer application, weeding, spraying, harvest), it will be essential for the associations to be trained on the most appropriate way to implement these management practices. Members of the associations should be facilitated through the members of the ‘comités techniques’ of each association. It is thereby essential that all members of the associations are present during these training events. Information should be obtained on the members present during each such event (name, association membership, gender).

2. Participatory evaluation of the germplasm At flowering and harvest of the legumes tested, these should be evaluated by all members of the farmer associations. At flowering, traits related to general growth, biomass production, and pest/disease resistance will be evaluated, while at harvest, traits related to grain quality (taste, size, colour, cookability, etc) will be evaluated. A standard protocol, drafted by CIAT, will be used as a tool for carrying out these evaluations. Team members in all sites will be trained on how best to implement the protocol.

note: since the bush beans, climbing beans, and soybeans are going to mature after different times, this evaluation events should take place at different times during the season.

3. Linking farmers to markets Since the project puts a lot of emphasis on linking farmers to markets in order to get better output prices and produce those crops that attract considerable market interest, activities will be initiated to foster these linkages. Team members in all sites will be trained on how to initiate these activities towards November 2006.

4. Training on soybean processing Since soybean is like going to raise issues of processing and marketing in the various locations, it is envisaged that two events will be organised during the second season of 2006: (i) training on processing of soybean into products for local consumption and (ii) visits of association representatives to sites where soybean processing and marketing is on-going. Note that for the latter, local inventories of such activities are needed for the different mandate areas. These activities are planned to be implemented around November – December 2006.

108

5. Integration of partners to prepare selection of satellite sites Since we need to identify development partners that are potentially interested to disseminate products of the current project through satellite sites, managed by these partners, we need to start engaging these around our own activities at the action site level. This could happen best through the following activities: (i) Organisation of a meeting with all potential stakeholders to understand their objectives and activities and evaluate potential synergies. (ii) Institutional analysis of partner organisations that are interested in collaborating. (iii) Three-monthly meetings, preferably around the demonstration sites, maybe in the form of a field day each season. (iv) Exchange visits between the action sites.

Trial layout Split plot design 1 association = 1 replicate, main plots = species × inputs, split plots = cultivars

mandate area

action site 1

action site 3

action site 2

association Bassociation A

action site 4

soybean single lines




association A

L

soybean microplots

soybean microplots +TSP

climbing bean microplots

climbing bean microplots +FYM

bush bean microplots

bush bean microplots +FYM

association B

L

soybean microplots






association A

L

soybean microplots






association B

L

soybean microplots






association A

L

soybean microplots






association B

L

soybean microplots






association A

L

soybean microplots






association B

L

soybean microplots






mandate areamandate area

action site 1

action site 3

action site 2

association Bassociation A

action site 4





association A

L

soybean microplots






association B

L

soybean microplots






association A

L

soybean microplots






association A

L

soybean microplots






association B

L

soybean microplots






association B

L

soybean microplots






association A

L

soybean microplots






association B

L

soybean microplots






association A

L

soybean microplots






association A

L

soybean microplots






association B

L

soybean microplots






association B

L

soybean microplots






association A

L

soybean microplots






association B

L

soybean microplots






association A

L

soybean microplots






association A

L

soybean microplots






association B

L

soybean microplots






association B

L

soybean microplots






association A

L

soybean microplots






association B

L

soybean microplots






association A

L

soybean microplots






association A

L

soybean microplots






association B

L

soybean microplots






association B

L

soybean microplots






109

Detailed layout (set u

p in every asso

ciation)

soybean m

icroplots (n

o in

puts)

3m

2m

0.75mcv8

soybean single lines (no inputs)

7.5m

2m0.75m

cv20cv25cv27cv19cv24cv21cv23cv26cv22cv18

BLOCK 1

3m

2m

0.75m

cv15

3m

2m

0.75mcv63

m

2m

0.75m

cv13

3m

2m

0.75m

cv12

3m

2m

0.75mS3

m

2m

0.75mcv53

m

2m

0.75mcv2

3m

2m

0.75m

L3m

2m

0.75mcv1

3m

2m

0.75m

cv14

3m

2m

0.75mcv9

3m

2m

0.75mcv73

m

2m

0.75m

cv17

3m

2m

0.75mM

3m

2m

0.75m

cv16

3m

2m

0.75m

cv11

3m

2m

0.75mcv3

3m

2m

0.75m

cv10

3m

2m

0.75mcv4

soybean m

icroplots (n

o in

puts)

3m

2m

0.75mcv8

soybean single lines (no inputs)

7.5m

2m0.75m


BLOCK 1

3m

2m

0.75m

cv15

3m

2m

0.75mcv63

m

2m

0.75m

cv13

3m

2m

0.75m

cv12

3m

2m

0.75mS3

m

2m

0.75mcv53

m

2m

0.75mcv2

3m

2m

0.75m

L3m

2m

0.75mcv1

3m

2m

0.75m

cv14

3m

2m

0.75mcv9

3m

2m

0.75mcv73

m

2m

0.75m

cv17

3m

2m

0.75mM

3m

2m

0.75m

cv16

3m

2m

0.75m

cv11

3m

2m

0.75mcv3

3m

2m

0.75m

cv10

3m

2m

0.75mcv4

soybean m

icroplots (+

TSP)

3m

2m

0.75mcv5

soybean single lines (+TSP)7.5m

2m0.75m


BLOCK 2

3m

2m

0.75mcv3

3m

2m

0.75mM

3m

2m

0.75m

cv10

3m

2m

0.75m

cv15

3m

2m

0.75m

cv11

3m

2m

0.75mcv13

m

2m

0.75m

L

3m

2m

0.75mcv83

m

2m

0.75mcv23

m

2m

0.75mcv93

m

2m

0.75m

cv13

3m

2m

0.75m

cv17

3m

2m

0.75mS3

m

2m

0.75mcv73

m

2m

0.75mcv4

3m

2m

0.75m

cv12

3m

2m

0.75mcv6

3m

2m

0.75m

cv14

3m

2m

0.75m

cv16

soybean m

icroplots (+

TSP)

3m

2m

0.75mcv5

soybean single lines (+TSP)7.5m

2m0.75m


BLOCK 2

3m

2m

0.75mcv3

3m

2m

0.75mM

3m

2m

0.75m

cv10

3m

2m

0.75m

cv15

3m

2m

0.75m

cv11

3m

2m

0.75mcv13

m

2m

0.75m

L

3m

2m

0.75mcv83

m

2m

0.75mcv23

m

2m

0.75mcv93

m

2m

0.75m

cv13

3m

2m

0.75m

cv17

3m

2m

0.75mS3

m

2m

0.75mcv73

m

2m

0.75mcv4

3m

2m

0.75m

cv12

3m

2m

0.75mcv6

3m

2m

0.75m

cv14

3m

2m

0.75m

cv16

bush bean microplots (no inputs)

BLOCK 3

1.6m

2m

0.4m

cv7

1.6m

2m

0.4m

cv14

1.6m

2m

0.4m

cv4

1.6m

2m

0.4m

cv25

1.6m

2m

0.4m

cv18

1.6m

2m

0.4m

cv22

1.6m

2m

0.4m

cv8

1.6m

2m

0.4m

cv19

1.6m

2m

0.4m

L

1.6m

2m

0.4m

cv9

1.6m

2m

0.4m

M

1.6m

2m

0.4m

cv1

1.6m

2m

0.4m

cv15

1.6m

2m

0.4m

cv21

1.6m

2m

0.4m

cv13

1.6m

2m

0.4m

cv16

1.6m

2m

0.4m

cv5

1.6m

2m

0.4m

cv23

1.6m

2m

0.4m

cv17

1.6m

2m

0.4m

cv3

1.6m

2m

0.4m

cv27

1.6m

2m

0.4m

cv12

1.6m

2m

0.4m

cv10

1.6m

2m

0.4m

cv6

1.6m

2m

0.4m

cv24

1.6m

2m

0.4m

cv2

1.6m

2m

0.4m

cv20

1.6m

2m

0.4m

S

1.6m

2m

0.4m

cv26

1.6m

2m

0.4m

cv11

bush bean microplots (+FYM)

BLOCK 4

1.6m

2m

0.4m

S

1.6m

2m

0.4m

cv25

1.6m

2m

0.4m

cv12

1.6m

2m

0.4m

cv7

1.6m

2m

0.4m

cv17

1.6m

2m

0.4m

cv26

1.6m

2m

0.4m

cv14

1.6m

2m

0.4m

cv23

1.6m

2m

0.4m

cv3

1.6m

2m

0.4m

M

1.6m

2m

0.4m

cv1

1.6m

2m

0.4m

cv22

1.6m

2m

0.4m

cv27

1.6m

2m

0.4m

cv5

1.6m

2m

0.4m

cv9

1.6m

2m

0.4m

cv8

1.6m

2m

0.4m

cv11

1.6m

2m

0.4m

L

1.6m

2m

0.4m

cv10

1.6m

2m

0.4m

cv21

1.6m

2m

0.4m

cv19

1.6m

2m

0.4m

cv2

1.6m

2m

0.4m

cv16

1.6m

2m

0.4m

cv18

1.6m

2m

0.4m

cv4

1.6m

2m

0.4m

cv15

1.6m

2m

0.4m

cv6

1.6m

2m

0.4m

cv24

1.6m

2m

0.4m

cv20

1.6m

2m

0.4m

cv13


BLOCK 3

1.6m

2m

0.4m

cv7

1.6m

2m

0.4m

cv14

1.6m

2m

0.4m

cv4

1.6m

2m

0.4m

cv25

1.6m

2m

0.4m

cv18

1.6m

2m

0.4m

cv22

1.6m

2m

0.4m

cv8

1.6m

2m

0.4m

cv19

1.6m

2m

0.4m

L

1.6m

2m

0.4m

cv9

1.6m

2m

0.4m

M

1.6m

2m

0.4m

cv1

1.6m

2m

0.4m

cv15

1.6m

2m

0.4m

cv21

1.6m

2m

0.4m

cv13

1.6m

2m

0.4m

cv16

1.6m

2m

0.4m

cv5

1.6m

2m

0.4m

cv23

1.6m

2m

0.4m

cv17

1.6m

2m

0.4m

cv3

1.6m

2m

0.4m

cv27

1.6m

2m

0.4m

cv12

1.6m

2m

0.4m

cv10

1.6m

2m

0.4m

cv6

1.6m

2m

0.4m

cv24

1.6m

2m

0.4m

cv2

1.6m

2m

0.4m

cv20

1.6m

2m

0.4m

S

1.6m

2m

0.4m

cv26

1.6m

2m

0.4m

cv11


BLOCK 3

1.6m

2m

0.4m

cv7

1.6m

2m

0.4m

cv14

1.6m

2m

0.4m

cv4

1.6m

2m

0.4m

cv25

1.6m

2m

0.4m

cv18

1.6m

2m

0.4m

cv22

1.6m

2m

0.4m

cv8

1.6m

2m

0.4m

cv19

1.6m

2m

0.4m

L

1.6m

2m

0.4m

cv9

1.6m

2m

0.4m

M

1.6m

2m

0.4m

cv1

1.6m

2m

0.4m

cv15

1.6m

2m

0.4m

cv21

1.6m

2m

0.4m

cv13

1.6m

2m

0.4m

cv16

1.6m

2m

0.4m

cv5

1.6m

2m

0.4m

cv23

1.6m

2m

0.4m

cv17

1.6m

2m

0.4m

cv3

1.6m

2m

0.4m

cv27

1.6m

2m

0.4m

cv12

1.6m

2m

0.4m

cv10

1.6m

2m

0.4m

cv6

1.6m

2m

0.4m

cv24

1.6m

2m

0.4m

cv2

1.6m

2m

0.4m

cv20

1.6m

2m

0.4m

S

1.6m

2m

0.4m

cv26

1.6m

2m

0.4m

cv11


BLOCK 4

1.6m

2m

0.4m

S

1.6m

2m

0.4m

cv25

1.6m

2m

0.4m

cv12

1.6m

2m

0.4m

cv7

1.6m

2m

0.4m

cv17

1.6m

2m

0.4m

cv26

1.6m

2m

0.4m

cv14

1.6m

2m

0.4m

cv23

1.6m

2m

0.4m

cv3

1.6m

2m

0.4m

M

1.6m

2m

0.4m

cv1

1.6m

2m

0.4m

cv22

1.6m

2m

0.4m

cv27

1.6m

2m

0.4m

cv5

1.6m

2m

0.4m

cv9

1.6m

2m

0.4m

cv8

1.6m

2m

0.4m

cv11

1.6m

2m

0.4m

L

1.6m

2m

0.4m

cv10

1.6m

2m

0.4m

cv21

1.6m

2m

0.4m

cv19

1.6m

2m

0.4m

cv2

1.6m

2m

0.4m

cv16

1.6m

2m

0.4m

cv18

1.6m

2m

0.4m

cv4

1.6m

2m

0.4m

cv15

1.6m

2m

0.4m

cv6

1.6m

2m

0.4m

cv24

1.6m

2m

0.4m

cv20

1.6m

2m

0.4m

cv13


BLOCK 4

1.6m

2m

0.4m

S

1.6m

2m

0.4m

cv25

1.6m

2m

0.4m

cv12

1.6m

2m

0.4m

cv7

1.6m

2m

0.4m

cv17

1.6m

2m

0.4m

cv26

1.6m

2m

0.4m

cv14

1.6m

2m

0.4m

cv23

1.6m

2m

0.4m

cv3

1.6m

2m

0.4m

M

1.6m

2m

0.4m

cv1

1.6m

2m

0.4m

cv22

1.6m

2m

0.4m

cv27

1.6m

2m

0.4m

cv5

1.6m

2m

0.4m

cv9

1.6m

2m

0.4m

cv8

1.6m

2m

0.4m

cv11

1.6m

2m

0.4m

L

1.6m

2m

0.4m

cv10

1.6m

2m

0.4m

cv21

1.6m

2m

0.4m

cv19

1.6m

2m

0.4m

cv2

1.6m

2m

0.4m

cv16

1.6m

2m

0.4m

cv18

1.6m

2m

0.4m

cv4

1.6m

2m

0.4m

cv15

1.6m

2m

0.4m

cv6

1.6m

2m

0.4m

cv24

1.6m

2m

0.4m

cv20

1.6m

2m

0.4m

cv13

clim

bing bean m

icroplots (n

o in

puts)

BLOCK 5

2m

2m

0.5m

cv6

L2m

2m

0.5m

cv3

2m

2m

0.5m

cv5

2m

2m

0.5m

cv12

m

2m

0.5mS2

m

2m

0.5mL

2m

2m

0.5m

cv4

2m

2m

0.5m

cv9

2m

2m

0.5m

cv7

2m

2m

0.5mM2m

2m

0.5m

cv8

2m

2m

0.5m

cv2

clim

bing bean m

icroplots (+

FYM)

BLOCK 6

2m

2m

0.5mS

L2m

2m

0.5m

cv1

2m

2m

0.5m

cv9

2m

2m

0.5m

cv4

2m

2m

0.5mL2

m

2m

0.5m

cv5

2m

2m

0.5m

cv62

m

2m

0.5m

cv72

m

2m

0.5m

cv3

2m

2m

0.5m

cv22

m

2m

0.5m

cv82

m

2m

0.5mM

clim

bing bean m

icroplots (n

o in

puts)

BLOCK 5

2m

2m

0.5m

cv6

L2m

2m

0.5m

cv3

2m

2m

0.5m

cv5

2m

2m

0.5m

cv12

m

2m

0.5mS2

m

2m

0.5mL

2m

2m

0.5m

cv4

2m

2m

0.5m

cv9

2m

2m

0.5m

cv7

2m

2m

0.5mM2m

2m

0.5m

cv8

2m

2m

0.5m

cv2

clim

bing bean m

icroplots (n

o in

puts)

BLOCK 5

2m

2m

0.5m

cv6

L2m

2m

0.5m

cv3

2m

2m

0.5m

cv5

2m

2m

0.5m

cv12

m

2m

0.5mS2

m

2m

0.5mL

2m

2m

0.5m

cv4

2m

2m

0.5m

cv9

2m

2m

0.5m

cv7

2m

2m

0.5mM2m

2m

0.5m

cv8

2m

2m

0.5m

cv2

clim

bing bean m

icroplots (+

FYM)

BLOCK 6

2m

2m

0.5mS

L2m

2m

0.5m

cv1

2m

2m

0.5m

cv9

2m

2m

0.5m

cv4

2m

2m

0.5mL2

m

2m

0.5m

cv5

2m

2m

0.5m

cv62

m

2m

0.5m

cv72

m

2m

0.5m

cv3

2m

2m

0.5m

cv22

m

2m

0.5m

cv82

m

2m

0.5mM

note: L

= local variety; M

= maize; S =

sorgh

um

110

AAANNNNNNEEEXXX 444::: QQQUUUEEESSSTTTIIIOOONNNNNNAAAIIIRRREEE UUUSSSEEEDDD FFFOOORRR LLLEEEGGGUUUMMMEEE GGGEEERRRMMMPPPLLLAAASSSMMM

EEEVVVAAALLLUUUAAATTTIIIOOONNN BBBYYY FFFAAARRRMMMEEERRR AAASSSSSSOOOCCCIIIAAATTTIIIOOONNNSSS

EVALUATION DES VARIETES D’HARICOT ET ESSAIS EN MILIEU PAYSAN

Données générales

Nom de l’Association : _________________________ Village : ________________________________

Nombre de femmes présentes : __________ Nombre d’hommes présents : __________

Date de l’évaluation : __________ Noms et sexes des facilitateurs : __________________________________

Première étape : Discussion en groupe

FICHE 1: Quelles sont les caractéristiques que vous voudriez trouver dans une bonne variété d’haricot nain? Qu’est ce que vous considérez quand vous voulez sélectionner une bonne variété d’haricot nain?

FICHE 2 : Quelles sont les CINQ CARACTERISTIQUES les plus importantes que vous utiliserez pour évaluer les différentes variétés ?

Critères d’évaluation Ordre d’importance

Proposez d’ajouter les critères suivants s’ils n’ont pas été mentionnés, en expliquant que ceci intéresse la recherche: 1. performance en conditions de basse fertilité des sols (sans fumier) ; 2. réponse au fumier

CIALCA-TSBF-CIAT

111

Deuxième étape : Evaluation au champ

FICHE 3. EVALUATION OUVERTE DES VARIETES D’HARICOTS NAINS

Gendre de groupe : Hommes ou Femmes …………………………………… Nombre de participants : …………………………………………

Codes pour « Réponse au fumier »: 0= Moins que le traitement sans fumier; 1= Pas de différence avec le traitement sans fumier; 2= Seulement une légère amélioration; 3= Beaucoup plus que le traitement sans fumier; 4= Plus du double que le traitement sans fumier

No Variété Aspects positifs (+) Nombre

rubans

Aspects négatifs (-) Nombre

Rubans

Reponse au

fumier

BRB194

CODMLB003

CODMLB007

CODMLB030

CODMLB078

MAHARAGI-SOJA

MARUNGI

MLB49-89A

M’SOLE

ZKA93-10m/95

AFR619

AFR708

ARA4

CIM9314-36

CIM9331-1

CNF5520

VEF88(40)L1P4T6

112

ECA-PAN021

HM21-7

LSA144

T8426F11-6F

UBR(92)25

ZAA5/2

GNP585

Rab618

Rab619

Rjb7

VARIETE LOCALE

113

Troisième étape : Analyse préférentielle

FICHE 4. ANALYSE PREFERENTIELLE DES VARIETES D’HARICOT NAIN (5-6

VARIETES SEULEMENT)

Combien de personnes classent une variété comme 1iere, 2ieme, 3ieme, 4ieme et 5ieme ?

Variétés 1er 2e 3e 4e 5e Ordre

En cas de divergence, indiquer les raisons: _______________________________________________________________________________ ______________________________________________________________________________________________________________________________________________________________ FICHE 5. MATRICE D’EVALUATION PREFERENTIELLE D’HARICOT NAIN

Caractéristiques

Variétés Production

sans fumier

Réponse au fumier

Demande au

marché

Total Rang

114

Quatrième étape : Restitution et planification pour la saison suivante

1. Restitution des résultats en groupe 1. Groupe de femmes : ______________________________________________________ 2. Groupe d’hommes : _______________________________________________________ 3. Commerçants : ___________________________________________________________

2. Demandez aux commerçants d’expliquer les filières haricot et les possibilités pour les associations de participer à cette filière ? __________________________________________________________________________________________________________________________________________________________________ 3. Inviter les paysans à poser des questions aux commerçants. Rassurez-vous que les questions ci-après sont posées. 3.1. Seriez-vous intéresser à acheter les haricots directement chez les paysans? Si oui, quels pourraient être les termes et conditions d’achat ? Conditions

Prix

Volumes Minimum Maximum

Qualité (pureté, humidité, triage, emballage,…)

Quelles sont les variétés les plus préférées ?

3.2. Que devraient faire les paysans s’ils doivent vendre mieux ? _________________________________________________________________________________ _________________________________________________________________________________ _________________________________________________________________________________ Suggestions pour la saison suivante

1. Quelles sont vos suggestions pour la saison prochaine ? _________________________________________________________________________________ _________________________________________________________________________________ _________________________________________________________________________________

2. Quelles sont les variétés que vous voudriez multiplier et réévaluer ? Variétés à multiplier Variétés à réévaluer

3. Quel sera le sort des variétés que vous n’avez pas sélectionnées ? _________________________________________________________________________________ _________________________________________________________________________________ _________________________________________________________________________________

115

AAANNNNNNEEEXXX 555::: AAACCCTTTIIIOOONNN PPPLLLAAANNN FFFOOORRR IIINNNFFFOOORRRMMMAAALLL LLLEEEGGGUUUMMMEEE SSSEEEEEEDDD SSSYYYSSSTTTEEEMMMSSS,,, 222000000777BBB–––222000000888BBB,,, TTTSSSBBBFFF---CCCIIIAAALLLCCCAAA

Activity Responsible Others to be involved Time frame Indicators (tracking changes) Tools /necessary Variety evaluation by farmers

Agronomists in charge of sites

-Socio-economists -Local leadership/extension -Local partner organizations -Traders

To be reported by 2nd week of August 2007

Number of preferred varieties identified in each site

Forms available

Farmer based variety description


Farmers and site managers End of the 2007 B season

Number of bean flyers / brochures produced

Excel sheets available

Training of farmers on post-harvest management

Phytopathology technician.

ISAR/INERA legume programs, NGO partners, local sector staff

depending on progress and need

Number of training sessions held and farmers trained

-Demonstration materials (pesticide) -Training manuals

Carry out post harvest management (PHM)

Association members

Supported by agronomists, NGO partners


-Minimized storage losses -Amount of bean dusted

-Pesticides -Training of farmers

Training/information sharing on group and social dynamism

NGO partners Local government extension staff, technicians in charge of sites


Number of people trained and training sessions held.

-Training documents (PPMR: Project de promotion pour micro-realization)

To facilitate farmers seed multiplication: -determine the amount of seeds and preferred varieties -land preparation -multiplication site visits in order to produce quality/quantity seeds

NGO partners or technicians in charge of sites

-Local leadership and extension -Local partners

before planting Area availed

To facilitate the contracting of association members in case of extra land required.

NGO partners Association leadership and members, technicians in charge of sites and local partner organizations

before planting Areas under contract availed

Crop establishment Association members

NGO partners, technicians in charge of sites

at planting -Varieties and areas planted in each association -Date of planting -Agronomic practices

Measurement tools, ropes,…

Carry out field data recording

Technical teams or animateurs


during crop growth farmer fiches (provided by Pieter)

116

Activity Responsible Others to be involved Time frame Indicators (tracking changes) Tools /necessary Training of farmers (associations) in seed production techniques / pest and disease control


NGO partners, technicians in charge of sites, legume programs, RADA/SENASEM

mid-season Number of people trained and training sessions held

Training manuals

Training of farmers on seed/grain business organization/marketing

Socio-economists NGO partners, technicians in charge of sites, legume programs, RADA/SENASEM


Number of people trained and training sessions held

Training manuals

Crop harvest Farmer associations


end of season Varieties and area under the crops, quantity of seed produced

farmer fiches (provided by Pieter)

Evaluation of season’s results


NGO partners, technicians in charge of sites, legume programs, RADA/SENASEM

after harvest evaluation sheet (provided by Pieter)

Carry out exchange visits and field days

NGO partners farmer association representatives, NGO partners, technicians in charge of sites, legume programs, RADA/SENASEM, local policy makers, local radio

Physiological maturity

Number of visites / field days and presence of stakeholders

Refreshment/transport standardized reporting format

Facilitate seed marketing and dissemination


farmer association representatives, NGO partners, technicians in charge of sites, legume programs, RADA/SENASEM, local policy makers


-Amount / varieties marketed -No of farmer seed producers selling seeds -No of farmers accessing or buying seeds (which varieties?)

Characterization of -existing seed systems / local varieties/mixtures -existing seed channels and their importance -farmer profiles


legume programs along the project time

-Existing varieties before the interventions -% of new varieties in the farmer associations along the duration of the project -Characterization of access / availability of new varieties

117

AAANNNNNNEEEXXX 666::: TTTRRRAAAIIINNNIIINNNGGG OOOFFF CCCIIIAAALLLCCCAAA SSSTTTAAAFFFFFF OOONNN LLLEEEGGGUUUMMMEEE SSSEEEEEEDDD

SSSYYYSSSTTTEEEMMMSSS AAANNNDDD SSSEEEEEEDDD PPPRRROOODDDUUUCCCTTTIIIOOONNN Date: 4-8/02/07; Venue: Butare, Rwanda Trainees: Adrien Bahizi Chifizi (PF DIOBASS/BUKAVU), Kasereka Bishikwabo (CIAT/BUKAVU), Sanginga JM (CIAT/BUKAVU), Mugiraneza Thierry (ISAR-Kibungo), Habitegeko Francis (ISAR-Umutara), Ngoga Tenge Gislain (ISAR-Karama) Subject maters covered during the training

• Elements and types of Seed Systems (both formal and decentralized) and how to link to improved bean varieties. Objectives were: 1. Understand the seed sector in DRC/Rwanda 2. Characterize seed system actors 3. Appreciate the actors’ roles in seed systems

• Strength, weakness and opportunity and threat (SWOT) analysis of existing seed systems in the different projects/countries (DRC and Rwanda) (formal and informal/decentralized) 1. Assess the strength of seed systems existing DRC/Rwanda 2. Assess their weakness regarding the accessibility of improved varieties to farmers 3. Assess the threats/opportunities they offer to accelerate improved access of varieties

• Planning of decentralized seed systems schemes (very important because the commercial seed sector is quasi existing in DRC/Rwanda) 1. Understanding elements necessary to establish a decentralised seed system 2. Assess stakeholders and their roles 3. Development of incentive systems

• Choosing the right crop and variety 1. Select the best and preferred varieties by farmers 2. Facilitate farmers’ selection 3. Link variety selection to seed systems

• Partnership for strengthening community-based seed systems scheme 1. To identify partners relevant in decentralized seed systems 2. To appreciate their roles 3. To map relationship models 4. To sustain relationships and partnerships

• Designing community seed systems schemes: To use the above-mentioned elements and design a decentralized seed scheme adapted to the participants’ conditions

Training methodology -Plenary Participatory presentations and discussions/comments -Case studies analysis conducted in groups and plenary presentation of the results Workshop Evaluation and gaps to be filled The workshop was assessed very well by the participants; however, since the workshop gave the overview on existing seed systems in DRC/Rwanda without crop-specific seed management practices, the following points/wishes were expressed by participants:

1. Soybean and bean seed multiplication sites 2. Seed business skills 3. Disease control and post harvest management 4. Crop-specific seed production (related to soil fertility) 5. Need to help participants to set up decentralised seed systems for both beans and soybeans

118

AAANNNNNNEEEXXX 777::: AAATTTEEELLLIIIEEERRR DDDEEE FFFOOORRRMMMAAATTTIIIOOONNN SSSUUURRR LLLAAA

MMMUUULLLTTTIIIPPPLLLIIICCCAAATTTIIIOOONNN DDDEEESSS SSSEEEMMMEEENNNCCCEEESSS DDDEEESSS LLLEEEGGGUUUMMMIIINNNEEEUUUSSSEEESSS

PLATE FORME DIOBASS AU KIVU

219, Avenue P.E. Lumumba/Nyawera C/o Bâtiment E.C.C/Diaconie et Développement

B.P. 1914 Bukavu/ RD Congo E-mail : [email protected]

RAPPORT

ATELIER DE FORMATION SUR LA MULTIPLICATION DES SEMENCES

Période : Du 04 au 05 avril 2007, Axe Katana Du 09 au 10 avril 2007, Axe walungu Bénéficiaires : Associations partenaires à la PF DIOBBASS Filière : Multiplication des semences autour des essais agronomiques Sigles et abréviations ALEMALU: Association pour la Lutte contre la Malnutrition à Lurhala APACOV: Actions Paysannes pour l’Amélioration des Conditions de Vie A.D.E.P.B: Association de Développement pour les éleveurs de Petits Bétails A.D.E.A: Association pour le Développement de l’Elevage et de l’Agriculture CIALCA: Consortium for Improvement Agriculture-Based Livelihoods in Central Africa CIAT: Centre International pour l’Agriculture Tropical DIOBASS: Démarche pour une Interaction entre Organisations à la Base et Autres Sources de Savoirs SENASEM: Service National de Semences ITM: Institut des Techniques Medicales FOMULAC: Fondation Médicale de l’Université Libre de Belgique en Afrique centrale I. Introduction et contexte L’introduction des nouvelles variétés de haricot et de soja, en septembre 2006 par le CIAT, a nécessité de procéder à des essais démonstrations pendant 2 saisons dans les 4 sites d’actions. Culture Nbre variétés

introduites en début saison A 2007

Quantité de semences

Variétés adoptées en fin saison A 2007

Haricots nains 28 60 graines 5 Haricots volubiles 10 60 graines 5 Soja 32 200 graines 5 A l’issu de ces essais et de l’évaluation des variétés qui s’en est suivi, les paysans ont choisis 5 bonnes variétés de haricot et de soja à la fin de la saison A 2007 comme repris dans le tableau ci- dessus. L’adoption de ces variétés a permis donc d’entrevoir la possibilité de les disponibiliser auprès des paysans pour leur multiplication dans les champs paysans avec des perspectives de marché.

119

Ia. Durée de la formation L’atelier de formation en multiplication qui s’est déroulé du 04 au 05 avril à Katana et du 09 au 10 avril 2007 à Walungu, avait pour but de former les paysans sur les aspects techniques de la multiplication en perspective de la saison B 2007. Ib.Objectifs de la formation La formation poursuivait des objectifs suivants: (i) doter les paysans membres des associations, des connaissances techniques nécessaire en matière de production des semences selon les normes du Senasem; (ii) permettre aux paysans de prendre une décision de devenir multiplicateur de semences et d’en faire une activité génératrice de revenu; (iii) outiller les paysans sur les éléments pouvant leur permettre de faire une étude de marché tenant compte des réalités de leur terroir. II. METHODES UTILISEES IIa. Choix du lieu de formation La formation en la multiplication s’est déroulée à l’ITM/FOMULAC Katana pour les paysans des localités Kabamba et Luhihi, du 04 au 05 avril. Pour les paysans des localités de Lurhala et Burhale, WALUNGU-CENTRE a servi de cadre à la formation en multiplication, du 09 au 10 avril 2007. Ces lieux de formations ont été choisis en fonction de réalités sociologiques de chaque axe, l’accessibilité au site et visualisation des essais en milieu ouvert et de la facilité de communication entre paysans de même terroir. IIb. Choix des associations participantes La formation en multiplication des semences, a concerné 12 associations dont 6 sur l’axe Katana et 6 sur l’axe Walungu. Ces associations ont été désignées pour la plupart, en fonction de leur participation active à toutes les phases des activités d’introduction, d’évaluation et de diffusion des germoplasmes de légumineuses depuis septembre 2006 IIc. Choix des participants Le choix des participants à la formation en multiplication, s’est faite en tenant compte de la représentativité dans l’association et par sexe, comme repris dans le tableau suivant :

Participants Axe Localités Association

Hommes Femmes

Total

Rhusimane 2 2 4 Luhihi Rhubehaguma 2 2 4 Tuungane 2 2 4 Maendeleo 3 1 4

Kabamba

ADEPB 1 - 1

Katana

Kavumu ADEA 1 - 1 Alemalu 4 0 4 Lurhala Cinamula 1 3 4 Apacov 3 1 4 Abagwasinye 1 3 4 Rhucihangane 1 - 1

Walungu

Burhale

Bololoke 1 - 1 La lecture de ce tableau montre que chaque fois 4 paysans ont pu être formé par association, à l’exception de ADEPB, ADEA, Rhucihangane et Bololoke qui ont disponibilisé chacun 1 membre. Cette discrimination est due au fait que les associations évoquées ci- haut ont joint la dynamique après les autres.

120

Le critère de choix de ces 4 participants a été: - la capacité de suivre la formation pendant toute sa durée - la capacité de restituer les apprentissages acquis aux autres membres de l’association - la présence d’au moins une femme dans l’équipe

La plupart des associations ayant un effectif moyen de 40 membres, il est donc possible qu’une personne puisse restituer à 10 autres ; ce qui a conduit au choix de 4 participants par association. Ainsi, le total a été de 18 participants par axe. IId. Choix de formateurs La formation en multiplication des semences a été conduite par une équipe d’Ingénieur agronome ayant une certaine expérience dans le monde associatif et de la vulgarisation du matériel végétal. Il s’agit de: 1. Bahizire Chifizi Adrien, Agronome PF DIOBASS: Chargé de l’organisation, du suivi et de

l’accompagnement des associations partenaires des sites d’actions, dans les activités CIAT-TSBF à l’Est de la RDC. Expérience également dans l’accompagnement des associations partenaires de DIOBASS sur les questions touchant le renforcement de la sécurité alimentaire et l’agriculture durable.

2. Sanginga Jean-Marie, Agronome CIAT: Chargé des implantations et du suivi des essais de démonstrations dans les sites d’action à l’Est de la RDC. A participé à des ateliers de formations sur la multiplication des semences au FHI en RDC

3. Byakombe Mazambi, Coordonnateur de SENASEM Sud-Kivu: Formateur dans le domaine de multiplication des semences.

IIe. Approche de la formation La formation des associations en multiplication entre dans un processus de capacitation des paysans membres, en outils pouvant leur permettre d’influencer positivement leur avenir. Ceci passe par la formation et l’information, 2 clés essentielles pour ouvrir les portes du progrès au milieu rural et faire reculer les murs de la misère, le tout dans une perspective de marché. La finalité de la rencontre était également de renforcer les capacités des organisations dans le but de trouver des solutions endogènes et durables aux questions que se posent les communautés. La démarche Diobass vise à :

� Améliorer la communication et les échanges entre tous les acteurs de développement : paysans, techniciens, cadres ou animateurs paysans.

� Développer une dynamique de réflexion au sein des communautés rurales, autour de thèmes qui intéressent directement ces communautés

� Promouvoir la recherche paysanne et valoriser les savoirs paysans. Cette finalité implique la formation conjointe de cadres et de paysans au cours de périodes de stages ou d’ateliers résidentiels, dans des situations concrètes de terrain, au sein d’organisation paysannes structurées IIf. Répartition des thèmes par formateur Les thèmes de formation ont été repartis comme suit entre différents intervenants :

Module 1 : Processus de la production des semences (Ir Sanginga Jean-Marie / CIAT-TSBF)

Celui-ci a tout d’abord brossé succinctement les activités CIAT dans cette partie de RDC et les partenaires locaux, régionaux et internationaux qui collaborent dans la réalisation des projets agricoles en milieu paysan et dans le cadre de la recherche. Il a montré comment, à partir 60 graines de différentes variétés de haricot (18 au total) et de soja (32 au total), introduites en septembre 2006, il a été possible d’en sélectionner 5 bonnes variétés, qui font l’objet d’un atelier de formation en ce jour.

121

Il a, en plus, attiré l’attention des participants sur l’importance alimentaire et nutritionnelle des légumineuses notamment le haricot et le soja. A partir de cette dernière culture, il est donc possible de produire, le lait, la viande et même l’huile de cuisine. Processus de production de semences Le processus de production ou chaîne de production des semences a été donné comme suit : - Semences de souches (Breeder seeds) : Au niveau des sélectionneurs - Semences de pré-base (selected seeds) : Après 3 générations - Semences de base (registred seeds) : après 1 génération - Semences certifiées : avec les agriculteurs- multiplicateurs, les associations, les privés, sous contrôle

du SENASEM

Module 2 : Elements techniques sur la production des semences de soja et de haricots (Ir Byakombe/ Senasem)

L’intervenant a donné la fiche technique du haricot et du soja :

1.. Fiche technique relative à la production des semences des variétés de Soja (Nom Scientifique : Glycine max)

1.1. Les variétés : Impérial, TGX888-49C, MUNANGA (TGX814-26D), AFIA (TGX849-249D), TGX573-209D, JUPITER, UFV, SIATA 194. Ces variétés sont celles qui sont inscrites, jusque là, au catalogue national des semences en RDC et cerfifiées par le Senasem 1.2. La qualité des semences à multiplier (Normes) Les exigences concernant la production de semence du soja sont les mêmes que pour que l’arachide. Il faut souligner que la germination du soja démunie sérieusement pendant le stockage. C’est pourquoi, avant le semis, une nouvelle analyse de la faculté germinative est absolument nécessaire pour pouvoir assurer une densité normale de la culture. La parcelle utilisée n’aura pas portée du soja pendant 2 saisons, sauf s’il s’agit de la même variété et génération ou d’une génération antérieure et que cette multiplication ait été agréée. Elle sera vierge de toute repousse accidentelle de la même espèce. 1.3. L’isolement Pour la production de semences de base, l’espace d’isolement entre différentes variétés de soja est de 30 mètres, et de 5 mètres entre mêmes variétés. Pour la production de semences certifiées, cet espace est de 5 mètres entre différentes variétés de soja et de 1 mètre entre même variété. 1.4. Préparation du terrain et fumure de base Sur les sols légers (sableux à sablo-limoneux), un labour léger est suffisant. Sur les sols plus lourds (texture limoneuse), un labour profond (15à 20 cm) suivi d’un hersage énergétique pour écraser les mottes est à recommander. La végétation grossière qui ne peut être enfouie avec le labour doit être sortie du champ semencier. Une fumure de base de 100 à 150 kg/ha d’engrais NPK (17 :17 :17) est à recommander de l’inoculum pendant les semis avec le Rhizobium 1.5.Le semis. Le semis du soja se fait soit en ligne continues soit en poquets en utilisant 40à 80 kg/ha et démarier à 1-2 plantes par poquet, 7 jours après levée.

122

• Les périodes de semis : les trois saisons (A, B et C avec irrigation d’appoint) conviennent à la culture du soja. En saison A, on sème après les premières pluies de manière à ce que la maturation et la récolte coïncide avec la petite saison sèche, afin d’éviter la pourriture des gousses.

• Les écartements : Selon les variétés et la fertilité du sol, les écartements suivants sont recommandés : Semis est mécanique (ligne continue) : 60cmx5cm ce qui correspond à 50-80kg/ha Semis manuel: 60cm x 5cm ; 40-60 cm x 20 cm avec 2 plants par poquet ce qui correspond à 40-60 kg de semences par hectare.

• La profondeur : 2 à 4 cm 1.6. Les travaux d’entretien Les travaux d’entretien consistent notamment en sarclages suivis de binages, afin d’aérer le sol et détruire les mauvaises herbes. A cet effet, au moins deux (2) sarclages sont indispensables : le premier intervient 7 jours après la levée ; le deuxième à la fructification et le troisième selon le degré d’envahissement par les mauvaises herbes. 1.7. Récolte, conditionnement et conservation

• Récolte : Elle intervient avant 4 mois après le semis selon la variété, dès la maturité qui se remarque par les signes ci-après :

• Les feuilles jaunissent complètement et tombent ;

• Les gousses se dessèchent ;

• La graine résiste à la pression des doigts et prend sa couleur caractéristique. Il faut récolter avant l’éclatement des gousses, surtout pour les variétés à gousses déhiscentes ; couper les tiges à ras du sol, et non les arracher pour ne pas priver le sol des nodosités développées sur les racines.

• Battre les tiges pour détacher les gousses, mais décortiquer les gousses à la main pour éviter de blesser l’embryon du soja qui est très sensible ; les graines sont ensuite triées en vue d’éliminer celles qui sont malformées, trop petites, blessées ou attaquées, ainsi que les corps étrangers.

• Enrober les semences, au Super Homai¨ à la dose de 10gr de produit pour 10 kg de semences ou du ALMTHIO, un mélange de Thirame et de Lindane à la dose de 250 gr pour 100kg de semences, afin de les protéger contre les insectes et les champignons ; puis les ensacher, les exposer au soleil pendant une journée, pour éliminer l’humidité due à l’enrobage et conserver dans un endroit sec, aéré et frais.

1.8. Le contrôle des cultures Un minimum de deux (2) contrôles sont indispensables : le premier à la floraison, le second avant la récolte, après la chute des feuilles. Lors des contrôles on observe les hors-types et vérifie la bonne exécution des épurations. Il est parfois nécessaire de procéder à des inspections supplémentaires en cas de problèmes particuliers. En pratique, l’inspecteur s’assurera que les plants du soja présentent bien les caractéristiques de la variété, puis examinera la bordure du champ afin de vérifier l’isolement. Il inspectera ensuite le champ dans son ensemble fera une évaluation des plantes adventices présentes et de la situation phytosanitaire. Lots de cette inspection, l’inspecteur examinera avec soin 150 plantes choisies au hasard en cinq endroits différents du champ (à raison de 30 plantes par endroit) et établira séparément le nombre des plants non conformes aux caractéristiques de la variété et le nombre de plants d’autres espèces cultivées dont les graines sont de dimension comparable. Le champ sera déclaré impropre si l’on dénombre plus de trois (3) hors-types (2%), ou plus de trois (3) plants d’autres espèces cultivées (2%).

123

Lors des contrôles, l’attention doit être sur les plants malades (bactérioses).Un champ portant plus de 1 plant malade sur 200plants contrôlés sera refusé comme champ semencier. Les champs de production de semences du soja ne porteront pas plus de 0,1 % d’adventices nuisibles. Pour les travaux du laboratoire, on prélève un échantillon de 1.000 gr pour le dénombrement et les autres analyse, dont 400gr pour l’analyse de pureté.

2. FICHE TECHNIQUE RELATIVE A LA PRODUCTION DE SEMENCES DES VARIETES DE HARICOTS (Nom scientifique : PHASEOLUS VULGARIS)

2.1. Les variétés Pour les régions des Kivu :

• Variétés VOLUBILES : ALIYA, VCB 81012 (jaune), AND 10 (blanche striée de rouge), kiangara, pigeon vert,

• Variétés NAINS : KIRUNDO, G2858, M’MAFUTALA, M’solé, D6 Bean, Sugar beans, blanket,…

Ces variétés sont celles qui sont inscrites, jusque là, au catalogue national des semences en RDC et cerfifiées par le Senasem

2.2. Le choix de terrain Pour produire des semences de haricot, le terrain doit avoir une texture moyenne, assurant un bon drainage. Les sols neutres légèrement acides, bien labourés, propres (sans mauvais herbes) sont préférables. 2.3. La qualité des semences à multiplier Les normes de qualité pour les semences de haricot et de niébé sont les suivantes :

Semences de « base » Semences « certifiées »

Pureté spécifique (minimum) Pureté spécifique (variété) (minimum) Matières inertes (maximum) Graines de mauvaises herbes Graines d’autres plantes cultivées (maximum) Germination (maximum) Humidité (maximum)

97% 98% 2% 0,005% 4%/kg 70% 12%

95% 97% 3% 0,1% 10% 70% 12%

2.4. L’antécédent cultural Les meilleures culturales précédant le haricot sont la pomme de terre, le mais et le sorgho ayant été bien entretenus par des sarclages, binages et désherbage. Il est conseillé d’attendre une campagne avant de faire revenir le haricot sur le même terrain. 2.5. L’isolement Pour la production des semences certifiées cet espace est de 5 mètres entre différentes variétés de haricots et de 1 mètre entre même variété. Pour la production des semences certifiées, cet espace est 5 mètres entre différentes variétés de haricot et de 1 mètre entre même variété. 2.6. Préparation du terrain et fumure de base Le labour à une profondeur de 20 cm est indiqué. Avant le semis, on plane le terrain avec un ou deux passages de pulvérisateurs. Avant le pulvérisation, on apporte la fumure de base: 30 à 50 unités d’azote de 30 à 40 unités de phosphore (150-200kg/ha NPK 17:17:17).

124

2.7. L’époque de semis Elle est en fonction des régions et du cycle végétatif de la variété. On cherchera toujours à arriver avec la maturation en fin de saison de pluies et début de saison sèche. 2.8. Le semis Sur les petites parcelles semencières, le semis se fait à la main, en lignes de 60 cm et en assurant 4 à 5 cm entre chaque poquets, de telle manière qu’on réalise une densité de semis de 33à 40 graines/m2 sur les superficies plus importantes, le semis est réalisé avec le semoirs de précision, en assurant la même densité. Avant le semis, un traitement des semences avec un fongicide (cryptodine, bénomy) est nécessaire pour protéger les jeunes plantules, contre les bactérioses, l’anthracnose, etc. La profondeur de semis est de 3à 4 cm, en fonction de la texture du sol. 2.9. Les travaux d’entretien Même si on utilise des herbicides, un ou deux sarclages sont nécessaires pour éliminer les mauvaises herbes et aérer le sol. Le premier sarclage sera exécuté 10 à 15 jours après la levée et le deuxième 20 à30 jours après le premier sarclage. A l’occasion de chaque sarclage, il est indiqué d’arracher toutes les mauvaises herbes .Pour les variétés volubiles, il faut faire le tuteurage au moment opportun (1ère feuille trifoliolée) La quantité de semences par hectare es t de 60 à 70 kg. 2.10. La récolte et le battage

• La récolte se réalise manuellement, en arrachant les plantes ou en les coupant avec la faucille. L’époque de la récolte d’une culture semencière doit coïncider avec la maturation physiologique. Les plantes récoltées sont ramassées en meules pour que le séchage continue pendant 2 à 3 jours. Dans les zones de haute pluviométrie, les meules ne donnent pas satisfaction. En ce qui concerne les haricots volubiles dont la récolte est échelonnée, il faut récolter les gousses au fur et à mesure de la maturité, en plusieurs passages.

• Le battage : se fait au fléau pour les petites quantités et à la batteuse, quand il s’agit de grosses quantités. Pour éviter l’écrasement des graines, la rotation du batteur ne doit pas dépasser 300 tours par minute. 2.11. Le séchage, le conditionnement et le stockage des semences L’humidité des semences ne doit pas dépasser 9 % pendant le stockage de longue durée. C’est pourquoi, les dépôts doivent être bien sec, bien aéré et pourvus d’installations de ventilation. Pour les grandes unités de production de semences, les installations de séchage sont indispensables. Pendant le stockage, une attention particulière doit être accordée à la protection contre les charançons. Pour cela, la fumigation des locaux de stockage avec tétrachlorure à la dose de 1kg/m3 de semences est indiquée, en prenant les précautions nécessaires contre les risques d’inflammation et d’intoxication. On peut utiliser également l’Actelic (2%) à la raison de 200 grammes pour une tonne de semences. Avant la commercialisation, il faut enrober les semences avec du Super Homai¨¨ (10 gr de produit pour 10 kg de graines) ou de Almithio (mélange de Thirame 250 gr et de Lindane 200gr) en raison de 250 gr pour 100kg de graines qui protégera les jeunes plantules. 2.12. Le contrôle du champ semencier Les champs de multiplication doivent être inspectés au moins deux fois : le premier contrôle doit être effectué pendant la floraison date où sont observés les hors types et où ont fait l’épuration aussi des plants malades. Le deuxième contrôle se fait à la maturation pâteuse pour éliminer les plantes qui sont soit très précoces, soit trop tardives par rapport à la majorité des plantes. Il est parfois nécessaire de procéder à des inspections supplémentaires en cas de problèmes particuliers. En pratique, l’inspecteur s’assurera que les plants de haricot présente bien les caractéristiques de la variété, plus examinera les bordures au champ afin de vérifier l’isolement. Il inspectera ensuite le champ dans son ensemble dans son ensemble et fera une évaluation des plantes adventices présentes et de la situation phytosanitaire.

125

Lors de cette inspection, l’inspecteur examinera avec soin 150 plantes choisies au hasard en cinq endroits différents du champ ( à raison de 30plantes par endroit) et établira séparément le nombre de plants d’autres espèces cultivées dont les graines sont de dimension comparable. Le champ sera déclaré impropre si l’on dénombré plus de trois hors types (2%), ou plus de trois plants d’autres espèces cultivées (2%).

Le troisième contrôle se fera à la récolte où on prélève les échantillons :

• 1.000 gr pour l’échantillon à soumettre ;

• 400 gr pour la pureté ;

• 1.000 gr pour le dénombrement et les autres analyses.

Module 3 : Compétence en gestion : Elements d’étude de marché (Ir Bahizire Chifizi/PFDiobass)

L’intervenant a d’abord tenu à préciser que la question relative au marché sera approfondie dans les prochaines activités prévues dans le projet. En revanche, il était important que les participants puissent avoir une certaine généralité sur les outils de gestion, notamment des outils de contrôle du marché.

Il s’agit des éléments suivants:

• la connaissance du client

• Fournir un bon service à son client A cet effet, il faudra faire connaître les avantages et les désavantages de chaque variété exploitée en être prêts d’en parler. Egalement, il serait utile de défendre et expliquer comment vous produisez et la différence entre vos variétés et les variétés dites locales.

• Les relations avec les concurrents

Il faut donc connaître ce que font les autres concurrents car certaines de leurs semences peuvent ressembler au votre. Ainsi donc, des essais démonstratifs sur les endroits accessibles peuvent se faire pour permettre à vos concurrents de voir et de suivre votre manière de travailler.

• Pour accroître votre demande - Pour accroître donc la demande de vos produits, il faut : - accroître votre marché en étendant le rayon de vente au delà de votre village, votre localité. - Chercher des nouveaux clients en ne vendant pas uniquement aux paysans. Etendre le rayon

de vente aux écoles, marchés locaux… - Changer fréquemment de variétés : cela peut se faire sur petite échelle puis sur grand terrain.

• Le conditionnement: Pour les haricots par exemple, utiliser des sacs translucides permettant aux clients de voir ce qu’ils achètent. Il est possible de vendre également en des petits emballages. Sur l’étiquette, on pourra lire : - Le nom et adresse du producteur - Nom de la variété - La qualité de la semence - Avertir si la semence est traitée - La classe de la semence

• Le transport: Il faut s’assurer de livrer, si les moyens le permettent, les semences à vos clients et protéger suffisamment les semences pendant leur déplacement

126

III. RESULTATS OBTENUS 1. Niveau d’assimilation des thèmes et points de vue des formés A l’issue de la formation, les participants formés ont manifestés leur contentement par rapport aux thèmes retenus et développés. Ils ont donc apprécié les méthodes de formation retenues, notamment la visite au terrain qui a marqué la dernière journée de formation sur les 2 axes. Quant aux formateurs, ils ont été captivés par l’intérêt des participants pendant les débats et les visites au terrain. Ceci a poussé le formateur sur les fiches techniques, à formuler des propositions ci-après pour les prochains ateliers : - prévoir une formation sur les fiches techniques des autres cultures - accroître la capacité organisationnelle des associations intéressées par la multiplication des

semences - faire précéder un atelier de formation sur la multiplication par des estimations chiffrées

sur la quantité de semences à semer en fonction de la quantité de semences recherchées à la récolte.

- Organiser un atelier à part sur les aspects liés au marché et à la commercialisation des produits agricoles.

- Négocier avec le Senasem de la possibilité de devenir multiplicateur des semences sur de terrains de dimensions inférieurs aux dimensions exigées par les normes.

2. Visite des champs de multiplication : réaction des agriculteurs Les visites au champ se sont effectuées dans le but de renforcer l’observation et confronter les techniques apprises aux réalités du terrain. La 1ère visite a concerné les paysans de l’axe Katana. Elle s’est effectuée au champ de multiplication d’une superficie de 1 ha, de l’association ADEA/ Kavumu qui est un site satellite. La 2eme visite s’est effectuée au champ de multiplication de haricot nain de APACOV/ Burhale, pour les paysans de l’axe Walungu. Dans ou l’autre site, les paysans étaient curieux de savoir ce qu’il faut faire pour être un multiplicateur agrée par le Senasem. En plus des préoccupations techniques liées à la multiplication des semences, les formés ont voulu également savoir la manière dont ils vont s’y prendre pour vendre à des prix compétitifs lors de la commercialisation. 3. Comment les paysans formés estiment intégrer les nouvelles connaissances dans leurs pratiques agricoles Les participants qui ont été pour la plupart des paysans, ont estimé qu’à la prochaine campagne, ils seront prêts à se lancer dans les activités de multiplication des semences. Cependant, les difficultés pour les associations d’avoir des grands terrains demeurent. A cet, les organisations voudraient solliciter au Senasem, la possibilité de présenter des terrains de dimensions réduites. Ces terrains mis ensemble pourraient constituer la superficie exigée par le Senasem. Egalement, les paysans voudraient bénéficier d’un appui en intrant par CIALCA, à la phase de multiplication des semences. IV. COMMENTAIRES DU REDACTEUR DU RAPPORT Les réalités socio-économiques du Sud- Kivu et la situation de conflit qui a prévalue dans cette partie de la RDC, n’ont pas épargné le mouvement associatif dans les sites d’action. Les structures travaillant avec la PF DIOBASS sur l’introduction des nouvelles semences, sont confrontées à des problèmes d’ordre structurel et technique autour des activités agricoles. La formation en multiplication avait justement pour but de doter les structures intéressées en technique de multiplication ainsi qu’en éléments méthodologiques d’organisation et de gestion dans la filière.

127

Ce faisant, la durée de 2 jours impartie à l’atelier sur la multiplication des semences a paru nettement insuffisant. 4 jours auraient suffis pour aborder également d’autres modules sur d’autres cultures que le haricot et le soja. Les travaux ont suscités beaucoup d’espoirs .D’abord au niveau des capacités d’organisation des acteurs en tant que synergie de multiplicateurs de semences et enfin en tant que cadre ouvrant les perspectives de marché pour lutter contre la pauvreté endémique dans le milieu. V. CONCLUSION La formation sur la multiplication des semences a permis : De regrouper les paysans membres des associations en un seul endroit. Ce cadre a favorisé des échanges d’expériences fructueux entre les paysans des milieux différents De comprendre les difficultés d’intégration des nouvelles semences, les difficultés organisationnelles des groupes ainsi que les pistes de solutions retenues par les paysans eux-mêmes et d’autres part de comprendre les difficultés rencontrées dans l’adoption des nouvelles semences, les difficultés organisationnelles des groupes et les pistes de solutions adoptées par les paysans eux-mêmes ont été également mise en surface. C’était aussi selon l’approche Diobass un rendez-vous entre paysans et scientifique. Une occasion offerte aux paysans d’échanger avec les spécialistes de la filière notamment le SENASEM et les agronomes CIALCA Des leçons ont été tirées de part et d’autres, nous citerons : L’opportunité est offerte aux privés et aux organisations d’encadrement des paysans intéressées de devenir multiplicateur des semences en respectant certaines directives ; La semence qui ainsi produite est un produit commerçable soumis aux lois de l’offre et de la demande, à la concurrence et aux fluctuations du marché. Le multiplicateur devra se préparer en conséquence et ne pas s’improviser dans la filière. Ceci offre également une porte aux autres activités Ciat qui peuvent être menées dans le milieu d’être favorablement accueillies. La PF DIOBASS a des perspectives suivantes pour l’activité de multiplication des semences dans les sites d’action et satellites: Faire des associations intéressées par la production, la distribution et le stockage des semences, des structures dotées d’outils de gestion et de suivi pour faire de cette activité un gagne –pain du paysan; Développer une dynamique de business autour de l’activité, en vue de rehausser le revenu du paysan et faciliter la circulation de l’argent dans le milieu rural; Ouvrir les paysans multiplicateurs sur le marché régional et sous régional en livrant leur production à des prix compétitifs.

128

ANNEXE 1 Liste des participants :

1. Axe Katana

N° Noms Organisation Site d’action Tél. et e-mail 01 Matunguru Mirindi Maendeleo Kabamba 02 Dieudonné Byende Kabira Maendeleo Kabamba 0994262306 03 M’Kaseremba Marceline Maendeleo Kabamba 04 Bahati cidabagizi Jean Maendeleo Kabamba 05 Mweze kabale Tuungane Kabamba 06 Skolastika Tuungane Kabamba 07 Eugénie M’mukonda Tuungane Kabamba 09 Mwila M’nyamukamwa Rhubehaguma Luhihi 10 Mirindi Kabujege Tuungane Kabamba 11 Eugenie M’moka Rhubehaguma Luhihi 12 M’mlangota Rhubehaguma Luhihi 13 M’rungora Rhusimane Luhihi 14 Mulume Bijaci Berlin ADEPB Kabamba +2439942236060 16 Alphonse Bisusa ADEA Kavumu +243997796109 17 Olinabanji Dieudonné Marafiki wa mazingira Katana +243997740827 18 Sangara Désiré Rhusimane Luhihi 19 Bahati Rugina Rhusimane Luhihi 20 Cizungu Luhirika Rhubehaguma Luhihi 21 Fidèle Mupenda Rhubehaguma Luhihi 22 Ir Gaetan Mazombo Senasem Bukavu +243997742028 23 Ir Adrien Bahizire chifizi PF Diobass Bukavu +243 -997252039

[email protected] 24 Ir Jean Marie SANGINGA Ciat/Cialca Bukavu + 243 998666101 25 Augustin Chyoka M. PF Diobass Bukavu +243 813176372

[email protected] 26 Dr Katunga Musale Cialca Bukavu

2. Axe Walungu

N° Noms Organisation Site d’action Tél.

01 Jonas Balola Luhene Alemalu Lurhala +243810258908 02 Jean Maheshe Chyoka Cinamula Lurhala 03 Ngwasi visita Alemalu Lurhala 04 Dieudonné Kusinza Alemalu Lurhala 05 Ursula Camunani Cinamula Lurhala +243994066752 06 Bahati Venant Apacov Burhale +243997788911 07 Geneviève M’mushaba Abagwasinye Burhale 08 Odette Lushunju Cinamula lurhala 09 Fabien kolondwa Rhucihangane Burhale 10 Vincent Kiriza Bololoke Burhale 11 Masimango Claude Abagwasinye Burhale +243994304755 12 Eugenie Basimara Abagwasinye Burhale 13 Vumilia lorensi Abagwasinye Burhale 14 N’simire M’birindwa Apacov Burhale 15 Kalyo kahirwe D. Apacov Burhale 16 Buhendwa Shaba Deux Alemalu Lurhala 17 Kaboza Mutabesha Alemalu Lurhala 18 Désiré Birindwa Apacov Burhale 19 Ir Jean-marie Sanginga Ciat/Cialca Bukavu +243998666101 20 Ir Byakombe mazambi Senasem/SK Bukavu +243998669463 21 Ir Adrien Bahizire chifizi PF Diobass Bukavu +243997252039

129

AAANNNNNNEEEXXX 888::: FFFIIICCCHHHEEE FFFOOORRR DDDAAATTTAAA CCCOOOLLLLLLEEECCCTTTIIIOOONNN IIINNN LLLEEEGGGUUUMMMEEE

MMMUUULLLTTTIIIPPPLLLIIICCCAAATTTIIIOOONNN FFFIIIEEELLLDDDSSS Fiche de caractérisation et suivi des champs de multiplication

Association :

Numéro de champ : (chaque association doit compter le nombre de champs de multiplication et numérotez chaque champ)

Décrivez ou se trouve le champ :

Position du champ dans le paysage : (plateau, en pente, ou dans la vallée)

Nom local de type de sol :

Cultures cultivés la saison passée : (spécifiez les cultures, ou jachère)

Superficie du champ : (en mètres carrés, ou largeur et longueur en mètres)

Nom du propriétaire du champ : (notez le nom de l’individu ou de l’association pour les champs communs)

Appréciation de la fertilité du champ (par le propriétaire) : (pauvre, moyenne ou bonne)

Espèce : (haricot nain, haricot volubile, soja, niébé ou arachide)

Nom de la variété :

Quantité de semence reçue : (s’applique seulement si la multipl ication est géré par un individu) (en kg)

Quantité de semence semée : (en kg)

Entrants appliqués ? Si oui, spécifiez le type d’entrant (fumier, composte, résidus de culture, engrais verts, engrais chimiques) et la source. Si oui, spécifiez la dose appliquée (en kg, ou en paniers).

Date de semis :

Spécifiez les écartements : (distances entre les lignes, et entre plantes dans la ligne ; si ne pas semée en lignes, comptez le nombre de plants dans 3 cadres de 1×1m) (les écartements recommandés sont 75× 5 cm pour le soja et l’arachide, 50× 20 cm pour l ’haricot volubile, et 40×10 cm pour l’haricot nain)

Dates de sarclage : (spécifiez toutes les dates de sarclage)

Nombre de plantes arrachées pour le démariage :

Nombre de plantes arrachées de couleur jaune : (à ajouter à différents moments pendant la saison)

Nombre de plantes arrachées avec signes de maladie : (à ajouter à différents moments pendant la saison)

Date de récolte :

Rendement obtenu : (en kg)

Quantité remise à l’association : (s’applique seulement si la multipl ication est géré par un individu) (en kg)

130

AAANNNNNNEEEXXX 999::: EEERRROOO---111::: EEERRROOOSSSIIIOOONNN CCCOOONNNTTTRRROOOLLL IIINNN SSSUUUDDD---KKKIIIVVVUUU

Objectives: (i) to investigate the relative importance of mechanically formed embankments (with/without hedgerows planted for stabilization) vs. surface management (traditional tillage / zero tillage) for reducing soil erosion on sloping land in S-Kivu. (ii) to evaluate Calliandra planted as hedgerows to reduce soil erosion (in presence or not of mechanically formed embankments) and determine Calliandra biomass production; (iii) to assess agronomic performance of a cropping system planted on the terraces, and investigate effects of physical embankments, Calliandra hedgerows and surface management on crop production and erosion. Sites: Replicated trial on-station (controlled conditions) on a field with a constant (representative) slope. Treatment structure - Main plots: physical embankments (combination of “fanya juu” and “fanya chini”) vs. no physical embankments; - Sub-plots: soil tillage (traditional tillage or zero tillage) × contour hedgerows (control or Calliandra hedgerow); - 3 replicates (blocks), positioned down-hill.

no physical embankment

fanya juu

CalliandraCazero tillageZT

control0traditional tillageTT

TT-C

a

ZT-C

a

ZT-0

TT-0

ZT-C

a

ZT-0

TT-C

a

TT-0

TT-C

a

ZT-0

TT-0

ZT-C

a

TT-0

TT-C

a

ZT-0

ZT-C

a

ZT-C

a

TT-0

TT-C

a

ZT-0

ZT-C

a

ZT-0

TT-0

TT-C

a

slope dire

ctio

n

rep 1

rep 2

rep 3

grass strip

grass strip

grass strip

no physical embankment

fanya juu

CalliandraCazero tillageZT

control0traditional tillageTT

TT-C

a

ZT-C

a

ZT-0

TT-0

ZT-C

a

ZT-0

TT-C

a

TT-0

TT-C

a

ZT-0

TT-0

ZT-C

a

TT-0

TT-C

a

ZT-0

ZT-C

a

ZT-C

a

TT-0

TT-C

a

ZT-0

ZT-C

a

ZT-0

TT-0

TT-C

a

slope dire

ctio

n

rep 1

rep 2

rep 3

grass strip

grass strip

grass strip

TT-C

a

ZT-C

a

ZT-0

TT-0

ZT-C

a

ZT-0

TT-C

a

TT-0

TT-C

a

ZT-0

TT-0

ZT-C

a

TT-0

TT-C

a

ZT-0

ZT-C

a

ZT-C

a

TT-0

TT-C

a

ZT-0

ZT-C

a

ZT-0

TT-0

TT-C

a

slope dire

ctio

n

rep 1

rep 2

rep 3

grass strip

grass strip

grass strip

Figure ERO-1 trial layout Three grass strips (length = 4 m) are included. The outer 1 m (on both sides) serves as a grassed waterway to allow excess water to run down. The middle 2 m serves as a reference area where erosion is zero (or very minimal) and is sampled at later stages to determine erosion using isotopic methods.

131

1 m

(drainage canal)

1 m

(drainage canal)

2 m

(reference strip with minimal erosion)

1 m

(drainage canal)

1 m

(drainage canal)

2 m

(reference strip with minimal erosion) Calliandra is included as hedgerow species to improve soil fertility. At regular times, plants are pruned and the residues are applied to crop on the terraces. Other interesting species are: Tithonia, sugarcane, Desmodium+Paspalum, Penisetum, Setaria, Trypsacum, Panicum, Brachiaria, Tephrosia, Vetiver… Performance and biomass production of these species is tested in a 2nd trial (ERO-2). Trial management Every plot consists of 3 hedgerows (strips) and 3 alleys (terraces). The first strip is constructed as a “fanya chini” (earth moved down the slope) to avoid run-off into the plots down-slope. Water coming down the slope is first trapped in a trench and looses erosive power. The trench is sown with grass and needs to be well maintained. Sediment accumulating in the trench needs to be removed regularly, so that water is conducted laterally towards the grassed channels going down the slope. The embankment in the first strip is physically enforced by wooden boards. All 3 alleys are cropped (with a short duration bean cv.). The first alley serves as a buffer and no data are recorded. Crop performance is assessed in the 2nd alley; biomass production in the hedges is assessed in the 3rd hedgerow. A 3rd alley is included to avoid recording data from strips that borders a physically enforced “fanya chini” strip. In the treatments where Calliandra is used as a hedgerow species, the hedges are regularly pruned and the residues are exported from the trial.

terrace 2 (cropped)

(crop yield assessment)

W = 8m

H= 1.5m

A= at least 2 m

(depends on slope)

strip 1 (fanya chini with physical stabilization)

strip 2 (fanya juu)

(fanya juu) strip 3

(biomass assessment)

(fanya chini with physical stabilization) strip 4

terrace 1 (cropped)

terrace 3 (cropped)

Treatment 1: with fanyas, with hedgerows

terrace 2 (cropped)


W = 8m

H= 1.5m

A= at least 2 m

(depends on slope)


strip 2 (fanya juu)

(fanya juu) strip 3



terrace 1 (cropped)

terrace 3 (cropped)terrace 3 (cropped)

Treatment 1: with fanyas, with hedgerows

132

terrace 2 (cropped)


W = 8m

H= 1.5m

A= at least 2 m

(depends on slope)


strip 2 (fanya juu)

(fanya juu) strip 3


terrace 1 (cropped)

terrace 3 (cropped)

Treatment 2: with fanyas, without hedgerows

terrace 2 (cropped)


W = 8m

H= 1.5m

A= at least 2 m

(depends on slope)


strip 2 (fanya juu)

(fanya juu) strip 3


terrace 1 (cropped)


Treatment 2: with fanyas, without hedgerows

A + H/2

= at least 2.75 m

(depends on slope)

terrace 2 (cropped)


W = 8m


strip 2

strip 3



terrace 1 (cropped)

terrace 3 (cropped)

Treatment 3: without fanyas, with hedgerows

A + H/2

= at least 2.75 m

(depends on slope)

terrace 2 (cropped)


W = 8m


strip 2

strip 3



terrace 1 (cropped)


Treatment 3: without fanyas, with hedgerows

133

A + H/2

= at least 2.75 m

(depends on slope)

terrace 2 (cropped)


W = 8m


strip 2

strip 3


terrace 1 (cropped)

terrace 3 (cropped)

Treatment 4: without fanyas, without hedgerows

A + H/2

= at least 2.75 m

(depends on slope)

terrace 2 (cropped)


W = 8m


strip 2

strip 3


terrace 1 (cropped)


Treatment 4: without fanyas, without hedgerows

Figure Individual plot lay-out. In treatments with physical embankments, the embankments between the 1st and 2nd, and between the 2nd and 3rd alley are constructed using “fanya juu” (earth moved slope upwards). They have the advantage not to increase the slope of the land in the alleys (unlike “fanya chini”) and can gradually develop into bench terraces. However, the risk of breakage during heavy storms is larger than for “fanya chini” embankments. The 4th strip is again constructed as a “fanya chini” embankment and coincides with the 1st hedgerow of the next replicate plot down the slope. The width of each plot is 6 m (following the slope contours). The length of the plot down the slope depends on the slope and is based on a 1.6 m change in elevation. The length (down the slope) of the alley should be at least 2 m; the length of the hedgerows (down the slope) is 1.5 m. The total area reserved for the trial should thus be 64 m wide (following the slope contours) by 40 – 60 m long (down the slope). ERO-1 is intended as a long-term trial. It needs to be ensured that the site is reserved for at least 5 years.

134

0.6m

0.6m

0.6m

0.6m

1.6m

d

D

L= D / tg(αααα)

αααα

H (1.5m)

A = D / sin(αααα) - H

0.6m

0.6m

0.6m

0.6m

1.6m

d

D

L= D / tg(αααα)

αααα

H (1.5m)

A = D / sin(αααα) - H

Figure Calculation of distance between two hedgerows as affected by slope (|, [%]) and change in

elevation (D, [m]). The Calliandra hedgerows are established as two lines at 50 cm between lines and 50 cm between plants as presented in the Figure below.

0.5 m

0.5 m

0.5 m

0.5 m

0.5 m

Figure Calliandra hedgerow spacing

135

ERO-1: Soybean planting protocol Land preparation: - For treatments with traditional tillage, the land on the terraces is cleared and tilled using a hoe. - For treatments with zero tillage, the land is simply cleared using a cutlass (except for the 1st . Establishment of soybean planting lines and soil sampling: The number of soybean lines in each terrace plot will vary because of (i) the size of the plot, (ii) presence of fanyas and/or Calliandra hedgerows. Planting lines are not necessarily straight, because of the irregular shape of the terrace plot and should more or less follow the contour line. Distance between lines can vary between 50 and 75 cm, but should as much as possible be kept at 75 cm. Examples of how soybean lines are delineated in the different treatments are shown below. 1. treatment with fanya juu and Calliandra hedgerows:

ditch

ditch

60cm

50cm

50cm Calliandra hedgerow

Calliandra hedgerow

soybean planting row 1





fanya juu embankment

100cm

50-75cm

50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

6m

ditch

ditch

60cm

50cm


Calliandra hedgerow







100cm

50-75cm

50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

ditch

ditch

60cm

50cm


Calliandra hedgerow







100cm

50-75cm

50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

6m 6m 2. treatment with fanya juu and without Calliandra hedgerows:

ditch

ditch

60cm

100cm

grass fallow







50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

25cm

6m

ditch

ditch

60cm

100cm

grass fallow







50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

25cm

ditch

ditch

60cm

100cm

grass fallow







50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

25cm

6m 6m 3. treatment without fanya juu and with Calliandra hedgerows:

136

25cm


Calliandra hedgerow






100cm

50-75cm

50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

25cm

Calliandra hedgerow

Calliandra hedgerow

soybean planting row 650-75cm

50-75cm

6m

25cm


Calliandra hedgerow






100cm

50-75cm

50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

25cm

Calliandra hedgerow

Calliandra hedgerow


50-75cm

25cm


Calliandra hedgerow






100cm

50-75cm

50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

25cm

Calliandra hedgerow

Calliandra hedgerow


50-75cm

6m 6m 4. treatment without fanya juu and without Calliandra hedgerows:







50-75cm

50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

25cm



50-75cm

6m







50-75cm

50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

25cm



50-75cm







50-75cm

50-75cm

50-75cm

50-75cm

50-75cm

contour line

contour line

25cm



50-75cm

6m 6m Soybean lines in each plot are numbered starting from the lowest line in the plot (closest to the lower contour). Each plot should minimally have 3 soybean lines; control plots (without fanyas and without Calliandra hedgerows) can have up to 8 soybean lines. The soybean line is established by digging a trench (about 10 cm deep). Seeds are planted at 5 cm within-line distance. Variety PEKA-6 is used. Installation of erosion pins: Erosion pins are installed only in the measurement plots (i.e. plot numbers 5, 6, 7, 8, 17, 18, 19, 20, 29, 30, 31, 32, 41, 42, 43, 44, 53, 54, 55, 56, 65, 66, 67, 68). Per plot, 6 erosion pins are installed: 3 at the upper end of the plot and 3 at the lower end of the plot (see figure below).

137

6m

1.5m 1.5m 1.5m 1.5m

6m 6m

1.5m 1.5m 1.5m 1.5m 1.5m 1.5m 1.5m 1.5m

In treatments with Calliandra hedgerows, erosion pins are installed 25 cm away from the Calliandra planting line. In treatments with fanya juus, the upper erosion pins are installed 25 cm down the slope from the ditch; the lower erosion pin is installed 25 cm away from the Calliandra hedgerow in treatments with planted embankments and 125 cm away from the lower ditch in treatments with bare embankments. In control treatments (without fanya juu and without Calliandra hedgerows), erosion pins are installed on the contour line. This is schematically presented in the figures below.

1. with fanya juu and with Calliandra hedgerows:

25cm

25cm

25cm

25cm

3. without fanya juu and with Calliandra hedgerows:

25cm

25cm

25cm

25cm

2. with fanya juu and without Calliandra hedgerows:

25cm

125cm

25cm

125cm

4. without fanya juu and without Calliandra hedgerows:

contour line

contour line

contour line

contour line

138

Trial management and observations:

-Soil profile description.

-Soil sampling. Soil samples are taken separately in between each of the soybean line, but only in the measurement plots (i.e. plot numbers 5, 6, 7, 8, 17, 18, 19, 20, 29, 30, 31, 32, 41, 42, 43, 44, 53, 54, 55, 56, 65, 66, 67, 68). In the middle between each soybean line, a composite soil sample (4 cores per line) is taken, using a mass balance auger. Two soil depths are sampled: 0-7.5 cm and 7.5-15 cm. Samples must be well labelled, indicating date, plot number, soybean line numbers in between which the sample was taken, the soil depth, and the number of cores (this should normally be 4). Samples are air-dried, weighed and stored.

-Precipitation is measured on a daily basis, using a rain gauge.

-Weeding (when necessary).

-Clearing and reparation of fanya chini ditches is performed when necessary (not of fanya juu ditches!!).

-At 50% flowering and at 50% podding, soybean aboveground biomass is sampled, separately for every row (only in the measurement plots!). In every row, the number of plants is counted in the “net line”, i.e. the middle 5 m without the outer 50cm. Subsequently, 5 plants are cut (randomly within the row, but not from the outer 0.5 m at both ends). The plants are dried, weighed, ground and stored pending on analysis.

-During the season, the distance between each line and the upper and lower boundary of the soybean-grown area is assessed at 1.5, 3 and 4.5 m (see file with examples).

-Harvest of soybean is conducted separately for every row in the measurement plots (only the “net row”, i.e. not the outer 0.5 m at both ends). In all other plots, the entire plot is harvested at once (excluding the outer 0.5 m at both ends of the rows).

-Moisture profile measurements. Moisture tubes are installed up to a depth of ***cm, in treatments without fanya juu embankments. The first tube is installed at ***cm from the contour line or the upper Calliandra line downslope. The second tube is installed at ***cm from the contour line or the upper Calliandra line downslope. Weekly, moisture profiles are assessed using a diviner probe.

-Slope assessment at planting (after land preparation) and at harvest of every season, using a triangle.

-Erosion pin soil levels at planting and harvest of every season.

-Initially, Calliandra hedges are allowed to obtain a height of 1.5 m; subsequently, hedges are pruned to a height of 75 cm. From then onwards, Calliandra hedgerows are pruned regularly and not allowed to be taller than 1 m. At each pruning event, residues are dried, separated in wood and leaves, cut and subsampled for analysis. The wood is exported from the trial; the leaf residues are surface-applied to the terrace located upslope from the hedge [or exported as livestock feed is the most likely entry point for adoption?].

-At the end of the trial, at harvest of the final season, Calliandra hedges are removed and root depth and root distribution of Calliandra and soybean are measured.

139

ERO-1: Sediment traps protocol A number of modifications need to be made to allow installing sediment traps for assessing soil loss in the various treatments. Plot separations: Plots need to be separated by making small ridges between the plots to prevent run-off from a given plot to enter neighbouring plots. The outer plots similarly need to be separated from the grass buffer strips. Ridges need to be about 50 cm wide and 30 cm high and need to be installed without affecting the soil surface in the plots. Therefore, these need to be constructed using topsoil (0-20cm) dug out from outside the trial. Ridges are planted with a grass, and maintained and trimmed regularly.

Sediment traps: Sediment traps will be installed in the ditch of the fanya chini in each replicate treatment (24 in total). These traps do not need to be installed in the top ditch protecting the first replicate from erosion occurring upslope, and do not need to be installed in the ditches of the ERO-2 trial. Walls are constructed inside the ditch to separate treatments. These walls are 0.5m wide (occupying 25cm at both outer ends of each treatment) and reinforced using wooden boards. The figures on the right indicates the modifications to be made (example showing 4 treatments); the figure below shows the entire plot lay-out with the position and numbering of the sediment traps.

7.5m W = 7.5m 0.5m

UPPER BORDER

PLOTS

MEASUREMENT

PLOTS

LOWER BORDER

PLOTS

grass strip

grass strip

sediment trap

(D=60cm, W=7.5m, L=60cm)

ridge separating treatments

(H=30cm, W=50cm)

reinforced wall separating sediment traps

(D=90cm, W=50cm, L=60cm)

UPPER BORDER

PLOTS

MEASUREMENT

PLOTS

LOWER BORDER

PLOTS

grass strip

grass strip

sediment trap

(D=60cm, W=7.5m, L=60cm)

ridge separating treatments

(H=30cm, W=50cm)

reinforced wall separating sediment traps

(D=90cm, W=50cm, L=60cm)

140

TT-C

a

ZT-C

a

ZT-0

TT-0

ZT-C

a

ZT-0

TT-C

a

TT-0

TT-C

a

ZT-0

TT-0

ZT-C

a

TT-0

TT-C

a

ZT-0

ZT-C

a

ZT-C

a

TT-0

TT-C

a

ZT-0

ZT-C

a

ZT-0

TT-0

TT-C

a

slope dire

ctio

n

rep 1

rep 2

rep 3

grass strip

grass strip

grass strip

1 2 3 4

5 6 7 8

9 10 11 12

13 14 15 16

17 18 19 20

21 22 23 24

TT-C

a

ZT-C

a

ZT-0

TT-0

ZT-C

a

ZT-0

TT-C

a

TT-0

TT-C

a

ZT-0

TT-0

ZT-C

a

TT-0

TT-C

a

ZT-0

ZT-C

a

ZT-C

a

TT-0

TT-C

a

ZT-0

ZT-C

a

ZT-0

TT-0

TT-C

a

slope dire

ctio

n

rep 1

rep 2

rep 3

grass strip

grass strip

grass strip

1 2 3 4

5 6 7 8

9 10 11 12

13 14 15 16

17 18 19 20

21 22 23 24

A plastic sheet is placed inside the ditch to catch the run-off. The sheet is perforated at the bottom (holes of ~0.5cm diameter) to allow water to drain. The front and back wall need to be covered by the sheet in such a way that run-off can freely flow into the trap. Observations: Sediment accumulating in the traps needs to be determined every two weeks. At all times, it needs to be prevented that large amounts of sediment accumulate and restrain water infiltration; in such cases, sediment needs to be collected more frequently. At least every two weeks, the total amount of sediment in each of the traps is scooped out and weighed (the date of sampling is noted in the field book). Subsequently the sediment is thoroughly mixed and a subsample of exactly 1 kg (use an accurate balance) is collected immediately. The sediment is then air-dried and weighed again. After weighing, a subsample of about 100 grams is stored for analysis. Each sample needs to be carefully labelled with the treatment and replicate, number of the sediment trap (see picture above), and the date of sediment collection.

141

AAANNNNNNEEEXXX 111000::: EEERRROOO---222::: CCCOOOMMMPPPAAARRRIIISSSOOONNN OOOFFF VVVAAARRRIIIOOOUUUSSS FFFOOORRRAAAGGGEEE

SSSPPPEEECCCIIIEEESSS FFFOOORRR EEERRROOOSSSIIIOOONNN CCCOOONNNTTTRRROOOLLL

Objectives: (i) to compare biomass production and biomass quality of various forage species when grown as hedges on sloping land; (ii) to assess effectiveness of various forage species to stabilize the soil (assessment of soil erosion and rooting depth and distribution). Sites: 6 sites: (i) down slope from the ERO-1 trial, (ii) degraded slope at INIBAP site in Cijingiri and (iii) 4 sites with various associations in the action sites of the Sud-Kivu mandate area. Treatment structure: - 8 treatments: (i) control, (ii) Calliandra at 50 cm planting density, (iii) Calliandra at 25 cm density, (iv) Leucena diversifolia, (v) Penisetum, (vi) Brachiaria, (vii) Setaria and (viii) Tithonia; - 1 site = 1 replication (no replications on site). One of the following two layouts can be used, depending on the land area available (36 x 16 m on the left, or 66 x 6 m on the right):

viii

iiiiiv

viiiivvi

slope dire

ctio

n

grass strip

grass strip

fanya chini

fanya chini

viii

iiiiiv

viiiivvi

slope dire

ctio

n

grass strip

grass strip

fanya chini

fanya chini

viiiivviii

viiiivii

slope

grass strip

grass strip

grass strip

fanya chini fanya chini

viiiivviii

viiiivii

slope

grass strip

grass strip

grass strip

fanya chini fanya chini

Figure ERO-2 possible trial layouts Species are planted in hedges of 6 m long, as 2 parallel lines at densities recommended by local extension services. The position of the hedges is randomized. A small initial dose of NPK fertilizer can be applied to aid establishment (rate?), particularly on degraded slopes. Upslope from every hedge, a fanya chini is installed to avoid treatments higher up the slope to affect the hedge down the slope. Distances between each fanya chini and hedge are based on a vertical interval of 1.6 m. The terraces between the hedges and fanya chinis should not be cleared and kept fallow (natural vegetation). Fanya chinis canals are 60 cm deep and 60 cm wide and planted with grass. These canals need to be well maintained; sediment accumulating in the trenches needs to be removed regularly. Fanya chini embankments are reinforced by wooden boards. At the ERO-1 site the fanya chini coincides with the final fanya chini of the ERO-1 trial.

142

0.6m

0.6m

1.6m

d

D

L= D / tg(αααα)αααα

A = D / sin(αααα)0.5m

0.6m

0.6m

1.6m

d

D

L= D / tg(αααα)αααα

A = D / sin(αααα)0.5m

Figure Calculation of distance between the hedgerow and the fanya chini slope upward as affected by

slope (|, [%]) and change in elevation (D, [m]). Trial management and observations: -Precipitation assessed on a daily basis. -Initial soil description and soil fertility evaluation. -Soil accumulation and soil loss assessed by erosion pins installed before and behind the hedge. -Isotopic measurements for assessment of soil erosion -At regular times, plants are pruned, biomass production is determined and the residues are exported (not applied to the above terrace). The quality of the biomass as a green manure or forage is assessed (C:N ratio, % lignin and polyphenols, protein and carbohydrate contents, mineral nutrient contents). As this trial is established at different sites with different slopes and/or soil fertility, the relation between biomass production/quality and soil physicochemical measures can be studied. -At the end of the trial (x years), hedges are removed and root depth and root distribution is measured.

143

AAANNNNNNEEEXXX 111111::: EEEVVVAAALLLUUUAAATTTIIIOOONNN DDDEEESSS FFFOOOUUURRRRRRAAAGGGEEESSS EEENNN MMMIIILLLIIIEEEUUU

PPPAAAYYYSSSAAANNN (((EEERRROOO---222)))

Données générales

Nom de l’Association …………………………… Village …………………………………………….

Nombre de femmes présentes …………….… Nombre d’hommes présents………………

Sexe du groupe interviewé ……………………

Date de l’évaluation ………………………… Noms et sexes des facilitateurs ………………………

Première étape : Discussion en groupe

FICHE 1: Quelles sont les caractéristiques que vous voudriez trouver dans un bon fourrage? Ou Qu’est ce que vous considérez quand vous voulez sélectionner un bon fourrage?

FICHE 2 : Quelles sont les CINQ CARACTERISTIQUES les plus importantes que vous allez utiliser pour évaluer les différentes variétés ?

Criteres d’évaluation Ordre d’importance

Proposez d’ajouter les critères suivants si ils n’ont pas été mentionné, en expliquant que ceci intéresse la recherche: 1. production de biomasse ; 2. efficacité de lutter l’érosion.

144

Deuxième étape : Visite au champ

Gendre de groupe : Hommes ou Femmes …………………………………… Nombre de participants : …………………………………………

Liste des fourrages installés (à remplir basé sur les observations au champ par l’enquêteur): nom du fourrage fourrage compris dans l’essai ?

(0 = non, 1 = oui) signes de non-adaptation (maladies, pauvre biomasse,…) ? (0 = non, 1 = oui)

vol ou mangé par des betes au champ ? (0 = non, 1 = oui)

1 Brachiaria brizantha

2 Brachiaria decumbens

3 Brachiaria ibrido

4 Brachiaria ruziziensis

5 Calliandra callothyrsus (à 0.25m)

6 Calliandra callothyrsus (à 0.5m)

7 Leucaena diversifolia

8 Penisetum purpureum

9 Setaria sphacelata

10 Tithonia diversifolia

11 Tripsacum laxum

145

FICHE 3. EVALUATION OUVERTE DES FOURRAGES

Production de biomasse / efficacité contre l’érosion : 0 = très pauvre; 1 = pauvre; 3 = moyen; 4 = bon; 5 = très bon

No Fourrage Aspects positifs (+) Nombre rubans

Aspects négatifs (-) Nombre rubans

production de biomasse

efficacité contre érosion

Brachiaria brizantha

Brachiaria decumbens

Brachiaria ibrido

Brachiaria ruziziensis

Calliandra callothyrsus (à 0.25m)

Calliandra callothyrsus (à 0.5m)

Leucaena diversifolia

Penisetum purpureum

Setaria sphacelata

Tithonia diversifolia

Tripsacum laxum

146

FICHE 4. ANALYSE PREFERENTIELLE DES FOURRAGES (1)

Combien de personnes classent un fourrage comme 1er, 2e, 3e, 4e, 5e ?

Fourrage 1er 2e 3e 4e 5e Ordre

En cas de divergence, indiquez les raisons : ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ FICHE 5. ANALYSE PREFERENTIELLE DES FOURRAGES (2)

Caractéristiques

Fourrages Critère 1 Critère 2 Critère 3 Critère 4

Production

biomasse

Efficacité contre érosion

Total Rang

147

AAANNNNNNEEEXXX 111222::: CCCAAASSS---111::: IIIMMMPPPRRROOOVVVEEEDDD CCCAAASSSSSSAAAVVVAAA AAAGGGRRROOONNNOOOMMMYYY Objectives: (i) to evaluate the contributions of various legumes to cassava production; (ii) to evaluate the use of alternative agronomic practices (tilled vs no-till; varying cassava planting density; use of inputs) on cassava and legume production; (iii) to evaluate the potential to include a second legume crop in a cassava field. Treatment structure DESCRIPTION OF THE FACTORS The following factors will be considered (Table 1): - Factor ‘Cassava variety’: One treatment will be cropped using local cassava while all the other plots will have CMD-resistant varieties that are known to still have full resistance to CMD; priority should be given to varieties that delay the formation of vigorous branching and leaf production. - Factor ‘Legume species/variety’: Dual purpose soybean will replace the commonly grown groundnut in specific treatments. - Factor ‘Planting density’: Cassava will be planted at a distance of 1 x 1m or 2 x 0.5 m (2 m between the lines, 0.5 m between plants within a line). - Factor ‘Planting time’: Since the best results are usually obtained when both the legume and the cassava are planted at the same time (maximally one week difference), this factor has only one level: simultaneous planting. - Factor ‘Tillage’: Some plots will be conventionally tilled while on other plots, cassava will be grown without tillage (‘au plat’). - Factor ‘Second legume’: The treatments with a second legume will have a bush bean variety planted at the start of the second season. - Factor ‘Inputs used’: Some treatments will have no inputs while other treatments will be treated with NPK fertilizer at 2 bags per hectare, with the fertilizer equally distributed over the cassava and the legumes and applied in the planting hole. Table 1: Detailed description of the various proposed treatments. Note that improved components are written in bold.

Trt Cassava variety

Legume species/variety

Cassava planting density

Tillage Planting time

Second legume

Inputs used

1 local Gnut/improved 1x1m Tilled Simultaneous None None 2 CMD-resist. Gnut/improved 1x1m Tilled Simultaneous None None 4 CMD-resist. Gnut/improved 2x0.5m Tilled Simultaneous None None 6 CMD-resist. Gnut/improved 2x0.5m No tillage Simultaneous None None 7 CMD-resist. Gnut/improved 2x0.5m Tilled Simultaneous Cl beans None 8 CMD-resist. Gnut/improved 2x0.5m Tilled Simultaneous None Yes 5 CMD-resist. Soybean/improved 2x0.5m Tilled Simultaneous None None 9 CMD-resist. Cowpea/improved 2x0.5m Tilled Simultaneous None None 10 CMD-resist. Bush-beans/improved 2x0.5m Tilled Simultaneous None None

EXPERIMENTAL DESIGN - The design will be a completely randomized block design with 3 replicates per site. - The trial will be laid out in 2 of the 4 Action Sites. - One site should be place on a ‘savanna’ soil and one on a ‘forest soil’ - Fields with homogeneous history of management and a minimal slope should be chosen that are representative for the farmer environment (not having been under a long fallow period; not strongly eroded; not too many trees in the plots; etc).

148

Trial implementation and management PLOT SIZE AND ARRANGEMENT - Plot size will be 10 by 10 m (10 lines of cassava in the 1x1 m arrangement and 5 lines of cassava in the 2x0.5 m arrangement) (Figure 1). - The legumes will be planted at distances of 40 cm between 2 lines. The distances between plants within a legume line will be: 10 cm for soybean, 20 cm for groundnut; 20 cm for cowpea; 10 cm for bush beans and 50 cm for the second season climbing beans. Figure 1: Sketch of a cassava-legume intercrop with 1x1 m and 2x0.5 m cassava spacing. CONTROL MAIZE PLOT - Per replicate, one plot needs to be included with maize as a reference crop for the BNF measurements. MANAGEMENT - The plots will be researcher-managed. - All operations will be implemented following proper agronomic principles (e.g., timely weeding). Observations - Initial soil description and soil fertility evaluation (at least 8 cores per plot at 0-20 cm using a ‘W’ desing; bulked; air-dried; stored pending shipment and analysis). - Labour (chronometer time to perform certain operations with a special emphasis on planting, weeding, and harvesting). - Cassava biomass at harvest (roots, stems, etc) from the net plot (8 middle lines of cassava in the 1x1 m spacing or 3 middle lines in the 2x0.5 m spacing). - Legume biomass at mid-podding (biomass sampling, also for BNF measurements). - Legume grain yield at harvest.

1m

1m

0.4m 0.3m0.3m

Cassava-groundnut intercrop [1 x 1 m]

Cassava hill

Legume line

2m

0.5m

0.4m

Cassava-groundnut intercrop [2 x 0.5 m]

0.4m 0.4m0.4m0.4m

149

AAANNNNNNEEEXXX 111333::: CCCAAASSS---222::: IIIMMMPPPRRROOOVVVEEEDDD CCCAAASSSSSSAAAVVVAAA AAAGGGRRROOONNNOOOMMMYYY Objectives: (i) to evaluate the contributions of various legumes to cassava production; (ii) to evaluate the use of alternative agronomic practices (varying cassava planting density; use of additional NPK input) on cassava and legume production; (iii) to evaluate the potential to include a second legume crop in a cassava field. Sites The trial will be established with the associations in Kabamba (TUUNGANE and MAENDELEO), at 2 replicates per association. If possible, neighbouring associations can be involved and install additional replicates (maximally 8 fields in total). Sites should be chosen carefully and be as homogeneous as possible, have the same position in the landscape (not on sloping land), a similar cropping & management history and a comparable fertility status. Treatment structure DESCRIPTION OF THE FACTORS The following factors will be considered (Table 1): - Factor ‘Cassava variety’: two treatments will be cropped using local cassava while all the other plots will have CMD-resistant varieties that are known to still have full resistance to CMD; priority should be given to varieties that delay the formation of vigorous branching and leaf production. - Factor ‘Legume species/variety’: In the 1st two treatments, the traditional cassava and bush bean varieties are used. In the other treatments, improved varieties (e.g. Sawasawa for cassava and ZKA93-10m/95 or Marungi for bush bean) are used. Dual purpose soybean (SB24 or SB25) or groundnut will replace the commonly grown bush beans in specific treatments. - Factor ‘Planting density’: In the 1st treatment, farmers use their traditional spacing (broadcasted seeds, random placement of cuttings). In all the other treatments, exact spacing is used, with the beans planted in lines. Cassava will be planted at a distance of 1x1m or 2x0.5 m (2 m between lines, 0.5 m between plants within a line). - Factor ‘Second legume’: The treatments with climbing beans as a second legume, planted at the start of the second season and climbing the maturing cassava crop. - Factor ‘Inputs used’: All treatments receive a blanket of FYM at 5 tonnes fresh matter ha-1, as farmers commonly apply organic inputs in their cassava systems. A treatment is included where additional NPK is applied at 2 bags per hectare, with the fertilizer equally distributed over the cassava and the legumes and applied in the planting hole. Table 1: Detailed description of the various treatments. Tr Cassava

variety Legume species/variety

Cassava planting density

Second legume Inputs used

1 Local Bean/local Traditional None FYM 2 Local Bean/local 1x1m None FYM 3 CMD-resist. Bean/improved 1x1m None FYM 4 CMD-resist. Bean/improved 2x0.5m None FYM 5 CMD-resist. Bean/improved 2x0.5m Climbing beans FYM 6 CMD-resist. Bean/improved 2x0.5m None FYM+NPK 7 CMD-resist. groundnut/improved 2x0.5m None FYM 8 CMD-resist. soybean/improved 2x0.5m None FYM

150

Trial implementation and management PLOT SIZE AND ARRANGEMENT - Plot size will be 10 by 6 m (10 lines of cassava in the 1x1m arrangement and 5 lines of cassava in the 2x0.5m arrangement) (Figure 1). - The legumes will be planted at distances of 40 cm between 2 lines. The distances between plants within a legume line will be: 10 cm for soybean, 20 cm for groundnut, 10 cm for bush beans and 50 cm for the second season climbing beans. Figure 1: Sketch of a cassava-legume intercrop with 1x1 m and 2x0.5 m cassava spacing (see additional details in protocol CAS-2 annex). FYM AND NPK APPLICATION FYM will be broadcasted and incorporated at 5 tonnes ha-1 at trial establishment (per plot of 10x6 m, 30 kg of FYM is incorporated). NPK will be applied at two bags per hectare, more or less equally distributed over the cassava and legume species (i.e. one bag for each species). In each planting hole of cassava, 2 bottle caps of NPK are applied (2 × ~3.65 g per planting hole). Per 1.5 m of legume, 1 bottle cap (~3.65 g) is applied. As such, per planting line of 6 m, 4 bottle caps are applied, and as much as possible concentrated at the seed planting spots. CONTROL MAIZE PLOT Per site, two small plots (2x3m) need to be included with a maize reference crop for BNF measurements, with and without NPK application. The maize is replanted in the 2nd season at the same time as the climbing beans. In both plots, the maize receives a basal FYM application at the same rate as the cassava (5 tonnes ha-1 = 3 kg per plot of 2x3m). In one of the two reference plots, additional NPK is added at one bag per hectare (1 g per planting hole). FIELD LAYOUT A detailed field layout for both spacings, with the delineation of the net plot, location of the cassava cuttings, legume planting lines and the planting beds, are presented in following two figures:

1m

1m

0.4m 0.3m0.3m

Cassava-groundnut intercrop [1 x 1 m]

Cassava hill

Legume line

2m

0.5m

0.4m

Cassava-groundnut intercrop [2 x 0.5 m]

0.4m 0.4m0.4m0.4m

151

parcelle utile

(6m x 4m)

bouture manioc ligne legumineuse

1m 0.4m 0.6m

1m

10m

6m

echantillonnage de

biomasse (1m x 0.5m)

parcelle utile

(6m x 4m)

bouture manioc ligne legumineuse

1m 0.4m 0.6m

1m

10m

6m

echantillonnage de

biomasse (1m x 0.5m) « Ecartement 1 × 1 m » « Ecartement 2 × 0.5 m »

parcelle utile

(6m x 4m)

bouture manioc lignes legumineuses

0.4m 0.8m 2m 0.4m

0.5m

10m

6m

echantillonnage de biomasse

(2 fois 0.5m x 0.5m)

parcelle utile

(6m x 4m)

bouture manioc lignes legumineuses

0.4m 0.8m 2m 0.4m

0.5m

10m

6m

echantillonnage de biomasse

(2 fois 0.5m x 0.5m) OBSERVATIONS - Initial soil description and soil fertility evaluation (at least 8 cores per plot at 0-20 cm using a ‘W’ design; bulked; air-dried; stored pending shipment and analysis). - Economic analysis (chronometer time to perform certain operations). The number of man-days required for overall land preparation (for the entire trial area) needs to be estimated. All other operations need to be timed using a chronometer on a per plot basis, particularly: (1) planting bed preparation; (2) fertilizer application; (3) cassava and legume planting; (4) weeding (1st, 2nd, 3rd and 4th weeding); (5) pesticide spraying (if conducted);

(6) legume harvesting (including pod collection and shucking), possibly conducted in several subsequent pod collections;

(7) cassava tuber harvesting;

152

note: operations performed for scientific purposes (biomass sampling, BNF assessment,…) should be differentiated from essential agronomic operations and not be included in the economic analysis. In addition, the market price of all inputs (seeds, cuttings, fertilizer, pesticide,…) should be collected. Market price information on all produce (groundnuts, cowpea, soybean, beans, maize and cassava) should be collected at a nearby market on a monthly basis, during a one year period. Also price information on side-products (e.g. cassava leaves and stems,…) should be gathered. - Legume biomass sampling at 50% podding (or 75% flowering for groundnut). Two random 0.50 m strips within two neighbouring legume lines in the net plot area are cut; the number of plants cut is counted and recorded. In treatments with 1 × 1 cassava spacing, one biomass sample is collected. In treatments with 2 × 0.5 cassava spacing, two biomass samples are collected: a first in a legume line bordering a cassava line (within the net plot area), and a second in a central legume line, bordered by two other legume lines (within the net plot area). At each biomass sampling event, 3 random maize plants are cut from the maize reference plots (3 replicates). As legume species are likely to attain the 50 % podding at different times, repeated maize reference samples may need to be taken. All biomass samples need to be well labelled (with the date of sampling clearly marked), and sun-dried. After oven-drying and recording the dry weight, the biomass samples are ground and stored pending analysis. - Legume grain yield at harvest. Similarly as for the biomass yield assessment, the yield in rows bordering cassava and rows bordered by two other legume rows are collected and sub-sampled separately in treatments with 2 × 0.5 cassava spacing. - Cassava harvesting at 12 months after planting. All cassava within the net plot are harvested and separated in tubers, stems and leaves. The number of stems and the number of tubers are recorded. The total weight of stems, leaves and tubers within the net plot is determined. Subsequently, the stems are cut in pieces of 20cm and 30 random pieces are subsampled. A subsample of 16 tubers and a subsample of about 1 kg of fresh leaves are collected. The weights of the subsamples are recorded at the same time as total yield measurements. Subsamples are taken to the station and dried. After recording the oven-dry weight, the samples are ground and stored pending analysis.

153

AAANNNNNNEEEXXX 111444::: CCCAAASSS---333::: IIIMMMPPPRRROOOVVVEEEDDD FFFEEERRRTTTIIILLLIIITTTYYY

MMMAAANNNAAAGGGEEEMMMEEENNNTTT IIINNN CCCAAASSSSSSAAAVVVAAA SSSYYYSSSTTTEEEMMMSSS Objectives: (i) to evaluate potential of green manure application to improve cassava production, as compared to the traditional slash and burn systems; (ii) to evaluate additional and interactive effects of fertilizer application and green manure application on cassava production; (ii) to assess nutrient uptake during growth, as affected by nutrient availability; Sites The trial will be established as a researcher-led on-farm trial with an association at Kisantu (CIALCA-managed) and one association at Mbanza Nzundu (VLIR-managed). Sites should be chosen carefully and be as homogeneous as possible, neither on strong-sloping land (avoid erosion/run-off problems) nor in low valleys (avoid flooding), and have a moderate soil fertility status (where a good response to inputs can be expected). The trial will be established in a completely randomized block design, with 3 replicates set up as separate blocks. Note: An improved, CMD-resistant cassava variety must be used (same variety for both sites), e.g. Butama or Disanka. Treatment structure DESCRIPTION OF THE FACTORS The following factors will be considered (Table 1): - Factor ‘natural vegetation’: the natural vegetation is either cut and removed, or applied under the planting hills. Only in the 3rd treatment, the natural vegetation is cut and burnt before being buried under the ridges. - Factor ‘OM application’: Tithonia and Chromolaena are applied at 2.5 t DM ha-1, supplying approximately 75 – 100 kg K ha-1. DM contents need to be calculated and application rates need to be corrected for moisture content. Moisture contents for Tithonia and Calliandra are 81% and 72%, respectively (determined on a subsample taken by W. Biponda). - Factor ‘fertilizer application’: NPK is applied at different rates (40, 120 or 200 kg K ha-1). Higher rates are split-applied (an initial 40 kg K ha-1 and additional application of 80 kg K ha-1 after 2 (and 4) months, depending on the treatment). Table 1: Detailed description of the various treatments Nr. natural vegetation OM application fertilizer application 1 cut & removed - - 2 cut & placed under hill - - 3 cut, burnt & placed under

hill - -

4 cut & removed Tithonia, 2.5 t DM ha-1 - 5 cut & removed Chromolaena, 2.5 t DM ha-1 - 6 cut & removed - 40 kg K ha-1

7 cut & placed under hill - 40 kg K ha-1 8 cut & removed Tithonia, 2.5 t DM ha-1 40 kg K ha-1

9 cut & removed Chromolaena, 2.5 t DM ha-1 40 kg K ha-1 10 cut & removed - 120 kg K ha-1

11 cut & removed Tithonia, 2.5 t DM ha-1 120 kg K ha-1 12 cut & removed Chromolaena, 2.5 t DM ha-1 120 kg K ha-1 13 cut & removed - 200 kg K ha-1

154

Trial implementation and management PLOT SIZE AND ARRANGEMENT Plot size will be 12 by 8 m (6 lines of cassava on ridges of 12 m long and 1 m wide, 40cm between ridges; Figure 1). On the ridge, the distance between cassava plants is 1 m (within the row). Care should be taken that the planting beds are installed perpendicular to the slope direction.

Figure 1: Layout of a cassava plot with 1x1.4 m cassava spacing and 2 net plots to

be harvested at 6 months (net plot I) and at 11-12 months (net plot II) after planting. The selection of the net plots can follow either the outline in (a) or in (b) (randomly selected).

Note: Each trial consists of 3 blocks (replicates) of 13 plots each. Total land area required per site equals 42 are (including border zones between plots).

MANAGEMENT OF NATURAL VEGETATION In all plots, the natural vegetation is slashed and exported (not burned!). In a separate area of 864m2 (24m × 36m), however, the vegetation is slashed separately and kept isolated. This will provide the organic matter for treatments 2, 3 and 7. The total amount of vegetation slashed in

(a)

(b)

8m

12m

1m

(ridge)

40cm

(furro

w)

1m

1.4m

I

slope dire

ctio

nII

8m

12m

1m

(ridge)

40cm

(furro

w)

1m

1.4m

I

slope dire

ctio

nII

8m

12m

1m

(ridge)

40cm

(furro

w)

1m

1.4m

I

slope dire

ctio

n

II

8m

12m

1m

(ridge)

40cm

(furro

w)

1m

1.4m

I

slope dire

ctio

n

II

155

this area is weighed, cut in smaller pieces (<20cm), thoroughly homogenized and divided in 9 equal quantities (3 treatments × 3 reps). A subsample is taken by mixing a handful out of each of the 9 quantities of residues. The fresh weight of the subsample is recorded; the subsample is then taken to the station for drying, determination of the dry weight, grinding and storage pending analysis. In each of the 9 plots where residues of the natural vegetation are retained, a quantity of residues is spread equally in 6 strips of 8 m long and 1 m wide, where the ridges will be established. These are then either burnt and buried (treatment 3), or immediately buried (treatments 2 and 7) by creating the ridges. These preparations should be performed about two weeks before planting, to allow the residues to decompose. ORGANIC MATTER APPLICATION Organic matter (either Tithonia or Chromolaena residues, chopped < 20cm) is collected from nearby sites. Only younger plant parts should be gathered; woody stems must be avoided. The material collected is thoroughly homogenized, and a representative subsample is taken. After recording the fresh weight, the subsamples are taken to the station for drying, determination of the dry weight, grinding and storage pending analyses. The residues are applied similarly as the natural vegetation residues, by spreading the residues equally in 6 strips of 8 m long and 1 m wide, where the ridges will be established. Fresh matter is applied, but rates need to be corrected for moisture content. The application rate of 2.5 t ha-1 corresponds to 4 kg DM per strip of 8 m (24 kg DM per plot = 6 strips). This corresponds to 21 kg fresh matter per strip for Tithonia and 14 kg fresh matter per strip for Chromolaena. The residues are immediately buried by creating the ridges. These preparations should be performed about 2 weeks before planting, to allow the residues to decompose.

Note: Per site, 216 kg Tithonia and 216 kg Chromolaena dry residues are required (3 treatments × 3 reps × 24 kg per plot), corresponding to 1137 kg fresh Tithonia residues and 758 kg fresh Chromolaena residues.

Note: Per site, the organic residues (Tithonia and Chromolaena) and natural vegetation (both the slashed material and the ashes after burning) should be representatively subsampled. The subsample is weighed fresh, oven-dried, weighed after drying to determine the MC, and subsequently ground and stored pending analysis for nutrient contents.

FERTILIZER APPLICATION AND PLANTING NPK (0.17:0.17:0.17) is applied at different rates (40, 120 and 200 kg K ha-1). Fertilizer is applied at planting, in the “planting hole” at an initial rate of 40 kg K ha-1 (23.53 g NPK per planting hole). With the hoe, a small pit is dug in which the fertilizer is applied, after mixing thoroughly with some soil to avoid chemical burning. The pit is subsequently covered with soil and a cassava cutting is planted vertically on the pocket. In treatments with NPK application at 120 kg K ha-1, additional fertilizer (80 kg K ha-1) is applied at 2 months after planting, by digging a small trench next to each plant, and applying the NPK (47.06 g NPK per planting hole) in the trench. In treatments with NPK application at 200 kg K ha-1, two additional doses of fertilizer (each 80 kg K ha-1) are applied at 2 and 4 months after planting respectively, by digging a small trench next to each plant, 20 cm away from the plant, and applying the NPK (47.06 g NPK per planting hole) in the trench.

Note: The quantities of 23.53 and 47.06 g of NPK do not need to be weighed with a balance. Application can be facilitated by finding a cap that approximately contains 23.53 g or 47.06 g of NPK.

Observations - Initial soil description and soil fertility evaluation: per replicate block, a composite topsoil (0-20 cm) sample is collected by mixing two cores from each plot within the replicate block. The soil is air-dried and stored pending shipment and analysis. Soils should be sampled before trial establishment, i.e. after slashing but before the application of organic matter and ridging.

156

- Economic analysis (chronometer time to perform certain operations). The number of man-days required for overall land preparation (for the entire trial area) needs to be estimated. All other operations need to be timed using a chronometer on a per plot basis, particularly: (1) organic matter collection, transport, cutting and application; (2) burning (treatment 3); (3) planting bed preparation; (4) fertilizer application; (5) cassava planting; (6) weeding (1st, 2nd, 3rd and 4th weeding);

(7) cassava tuber harvesting; note: operations performed for scientific purposes (biomass sampling,…) should be differentiated from essential agronomic operations and not be included in the economic analysis. In addition, the market price of all inputs (cuttings, fertilizer,…) should be collected. Market price information on produce (cassava tubers) should be collected at a nearby market on a monthly basis, during a one year period. Also price information on side-products (cassava leaves and stems,…) should be gathered. - Three cassava biomass samplings and a final yield assessment are included at 6 months after planting and at harvest (11-12 months after planting). At each event, one of the 2 net plots (see Figure 1) is randomly harvested. The biomass is separated in tubers, stems and leaves. The height of the stems and number of stems and the number of tubers are recorded. The total weight of stems, leaves and tubers within the net plot is determined. Subsequently, the stems are cut in pieces of 20cm and 30 random pieces are subsampled. A subsample of 12 tubers and a subsample of about 1 kg of fresh leaves are collected. The weights of the subsamples are recorded at the same time as total yield measurements. Subsamples are taken to the station and dried. After recording the oven-dry weight, the samples are ground and stored pending analysis. Note: In treatments with OM application or buried natural vegetation, OM decomposition is

assessed qualitatively (or quantitatively if possible, by digging up, drying and weighing the remaining residues).

157

AAANNNNNNEEEXXX 111555::: PPPRRREEEFFFEEERRREEENNNCCCEEESSS DDDEEESSS EEESSSSSSAAAIIISSS

DDD’’’AAAMMMEEELLLIIIOOORRRAAATTTIIIOOONNN DDDEEESSS SSSYYYSSSTTTEEEMMMEEESSS AAAGGGRRRIIICCCOOOLLLEEESSS BBBAAASSSEEESSS SSSUUURRR

LLLEEE MMMAAANNNIIIOOOCCC PPPAAARRR LLLEEESSS AAAGGGRRRIIICCCUUULLLTTTEEEUUURRRSSS DDDEEE KKKAAABBBAAAMMMBBBAAA

Consortium for Improving Agriculture-based Livelihoods in Central Africa = CIALCA /Sud-Kivu =

DGDC Legume project of TSBF-CIAT

Préférences des essais d’amélioration des systèmes agricoles basés sur le manioc par les agriculteurs de Kabamba

Résultats de l’évaluation à la floraison des légumineuses associées avec le manioc Bukavu, janvier 2008

INTRODUCTION L’évaluation à la floraison des essais CAS-2 s’est déroulée du 18 au 19 décembre 2008. L’évaluation a été assurée par Kasereka Bishikwabo et Adrien Chifizi respectivement socio-économiste au CIAT/Bukavu et responsable des activités avec les associations au sein de la plateforme DIOBASS. L’objectif a été d’évaluer les préférences par les agriculteurs des essais de gestion des systèmes agricoles avec manioc. Ce rapport préliminaire conçue pour permettre de décider des activités en saison B2008 présente les méthodes et matériel utilisées, les résultats de l’évaluation et quelques commentaires ainsi que des suggestions pour préparer l’acceptation des technologies qui sont préférées. 1. METHODES ET MATERIEL 1.1 Description de l’essai L’essai d’association manioc-légumineuses comportait cinq traitements dans huit parcelles (Tableau 1). Tableau 1 : Traitements et contenu des parcelles dans les essais CAS-2 No Traitement Numéro parcelle et contenu Numéro parcelle et contenu

1 Effet de la technique de semis

Parcelle no 1 : Traditionnel (en vrac) Variétés locales de manioc en semis traditionnel (en vrac) et de haricot nain.

Parcelle no 2 : Ligne Variétés locales de manioc à écartement 1mx1m et de haricot nain.

2 Effet de la variété

Parcelle no 2 : Locales Variétés locales de manioc à écartement 1mx1m et de haricot nain

Parcelle no 3 : Améliorées Variétés améliorées de manioc à écartement 1mx1m et de haricot nain

3 Effet de l’écartement

Parcelle no3 : 1mx1m Variétés améliorées de manioc à écartement 1mx1m et haricot nain

Parcelle no 4 : 2mx0,5m Variétés améliorées de manioc à écartement 2mx0,5m et de haricot nain.

4 Effet du NPK Parcelle no 5 : Sans NPK Variétés améliorées de manioc à écartement 2mx0,5m et de haricot nain sans NPK.

Parcelle no 6 : Avec NPK Variétés améliorées de manioc à écartement 2mx0,5m et de haricot nain avec NPK.

5 Effet d’autres espèces de légumineuses

Parcelle no 7 : Variétés améliorées de manioc à écartement 2mx0,5m et arachide.

Parcelle no 8 : Variétés améliorées de manioc à écartement 2mx0,5m et soja.

158

Ce protocole a été respecté. Une erreur a été remarquée au niveau de l’association Maendeleo où il y a eu deux variétés différentes de haricot nain, l’une dans la parcelle no 1 et l’autre dans la parcelle no 2. Dans la parcelle no 2 la variété de manioc était améliorée au lieu d’être locale. 1.2 Les agriculteurs évaluateurs Trente et un agriculteurs évaluateurs membres de trois associations ont évalué les essais de gestion CAS-2 installés dans le site d’action de Kabamba (Tableau 2). Il y avait parmi eux 16 femmes et 15 hommes, tous membres des associations partenaires Maendeleo, Tuungane et ADEPB. Tableau 2 : Nombre d’agriculteurs évaluateurs des essais CAS-2 par sexe et par association

Association Femme Homme Total

Maendeleo 3 4 7 Tuungane 8 6 14 ADEPB 5 5 10 Total 16 15 31

1.3 Procédure d’évaluation Une réunion a été tenue avec les membres évaluateurs au niveau de chaque association. Au cours de la réunion, les participants ont décrit leur système traditionnel d’association manioc - légumineuse. Ils l’ont par la suite comparé au système amélioré. Les objectifs des essais ont été discutés dans l’intérêt que les agriculteurs comprennent mieux l’activité qu’ils sont en train de réaliser. Par la suite, deux groupes par association ont été constitués : l’un composé uniquement de femmes et l’autre d’hommes. Chaque groupe a généré et rangé selon l’ordre d’importance ses critères d’appréciation du haricot nain, du manioc, de l’arachide et du soja. Dans l’essai, chaque groupe a apprécié les différentes parcelles l’une après l’autre en donnant les observations négatives et positives sur chaque parcelle d’essai. Ensuite il y a eu une comparaison des parcelles par traitement. Chaque femme a reçu 10 billes de couleur verte et chaque homme en reçu pareil mais de couleur bleue. Un sachet noir était suspendu sur l’étiquette dans chaque parcelle et chaque agriculteur évaluateur était invité à partager ses billes entre les parcelles préférées en prenant soin de mettre plus de billes dans le sachet à la parcelle la plus préférée. A la fin de l’opération, les membres des deux groupes sont passés ensemble parcelle par parcelle pour compter le nombre de billes bleues et celui des billes vertes. Une discussion s’en suivait chaque fois pour justifier les écarts. (N.B. : dans les résultats, bille=cailloux). Toutes les données récoltées ont été notées sur des fiches préalablement préparées par Mr Pieters. Pour élaborer ce rapport préliminaire, quelques données synthèses sur chaque fiche ont été saisies dans Excel et analysées dans SPSS12. Les graphiques ont été produites en utilisant le logiciel Excel. 2. RESULTATS DE L’APPRECIATION DES ESSAIS CAS-2 2.1 Les critères d’appréciation Dans l’ensemble, les paysans ont généré 15 critères (Figure 1). Adaptation est la résistance de la culture aux intemperies (eau abondante et sécheresse). Biomasse implique l’abondance des feuilles et/ou tige ainsi que leur vigueur. Cynophoresol égal cynophore de la plante bien fixé dans le sol. Feuilleaspect est l’apparence de la feuille ; quand les agriculteurs évoquent ce critère, c’est souvent pour vanter la couleur vert-foncée mais aussi des feuilles larges. Fleursgousseab est l’abondance des fleurs et/ou des gousses. Germinationbon veut dire que la plupart des plants avaient levé. Maladiepas est l’absence de maladies ou d’autres signes inquiétant sur la plante. Eauresiste est la résistance de la plante contre une abondance d’eau. Planthaut est la hauteur de la plante ; une plante d’une hauteur élevée est mieux appréciée sauf pour l’arachide où la faible hauteur ‘taillereduite’ est la qualité recherchée. Floraison2pas a été cité pour l’appréciation des arachides. D’après les évaluateurs paysans, quand les arachides fleurissent deux fois on ne s’attend plus à une bonne production. Ravageurspas est que la plante n’a pas été affectée par les ravageurs. Meristemeeleve est le méristème élevé qui est préféré par les agriculteurs. Tigegrosse est la grosse tige qui est préférée tandis que Tigetachepas est une tige sans tâche.

159

Les critères Feuilleaspect et Maladiepas ont été utilisés pour apprécier toutes les quatre cultures dans l’essai : manioc, arachide, haricot et soja. Tigegrosse a été appliqué sur toutes les cultures sauf l’arachide. Le haricot nain et le soja sont plus préférés pour l’abondance de leurs gousses et/ou fleurs. Le critère Maladiepas a plus servi pour l’appréciation du manioc et de l’arachide. Ces deux cultures sont plus malades à Kabamba comparativement au haricot et au soja. Le manioc a depuis plus de quatre ans été attaqué par la mosaïque. Les variétés locales d’arachide sont très vulnérables aux maladies à tel enseigne que certains paysans se découragent à pratiquer cette culture qui avec le manioc sont perçus comme les cultures qui rapportent le plus d’argent au paysan de Kabamba. Certains critères s’appliquent uniquement à l’appréciation de l’arachide : c’est le cas de taillereduite, meristemeeleve, Cynophoresol et Floraison2pas.

0

1

2

3

4

5

6

7

8

9

Adaptation

Biomasse

Cynophoresol

Feuilleaspect

Fleursgousseab

Floraison2pas

Germinationbon

maladiepas

eauresiste

Planthaut

ravageurspas

meristemeeleve

Tigegrosse

Tigetachepas

taillereduite

arachide

haricot

manioc

soja

Figure 1 : Critères générés et utilisés par les agriculteurs pour l’appréciation des cultures d’arachide, haricot, manioc et soja dans les essais CAS-2 Le critère Feuilleaspect est très important dans l’appréciation des cultures. Il a été relevé 11 fois comme premier critère sur les 14 fois qu’il a été cité (Figure 2).

1

0

2

11

2

1

3

9

3

0

3

5

4

0 0

3

0

1

0

1

0

1

0

1

0

2

4

6

8

10

12

Critere 1 Critere 2 Critere 3

Adaptation

Biomasse

Cynophoresol

Feuilleaspect

Fleursgousseab

Floraison2pas

Germinationbon

maladiepas

eauresiste

Planthaut

ravageurspas

meristemeeleve

Tigegrosse

Tigetachepas

Figure 2 : Fréquence de critères d’appréciation par rang (critère 1 = premier critère, plus important).

160

Les agriculteurs disent que la qualité de feuilles est le premier indicateur pour espérer ou non une bonne production. Aussi est-il qu’à part le soja et l’arachide, les feuilles de manioc et des haricots constituent un aliment consommé à Kabamba. Le critère Fleurgousseab est cité trois fois comme premier, neuf fois comme deuxième et trois fois comme troisième. C’est un critère important d’autant plus que les agriculteurs, acculés par le besoin de manger d’abord, voudraient avoir des technologies qui entraînent une plus grande production. Un plant ou une variété qui porte beaucoup de gousses et de fleurs est perçu comme devant avoir un haut rendement. Ce critère s’applique aux légumineuses. Le critère maladiepas est cité trois fois au rang premier, cinq fois au rang deuxième et quatre fois au troisième rang. La maladie des cultures est perçue comme l’obstacle à la production. Tigegrosse, cinq fois cité au deuxième rang et deux fois au troisième rang est un critère important pour apprécier le manioc surtout. 2.2 Préférences des essais de gestion : association manioc-légumineuses Les traitements évalués étaient les suivants : VARLOECART : variété locale écartement locale ; VARLO1X1 : variété locale écartement 1 m x 1m ; VARAM 1x 1 : variété améliorée écartement 1mx1m ; VARAM 2x 05 : variété améliorée écartement 2mx0,5m ; VARAM2L2x05 : variété améliorée parcelle prévue pour la deuxième légumineuse et NPK non appliqué ; VARAM2L2x05N : variété améliorée parcelle prévue pour la deuxième légumineuse et NPK appliqué ; VARAMAR2X05 : variété améliorée arachide comme légumineuse ; VARAMSO2X05 : variété améliorée soja comme légumineuse. En considérant la moyenne totale du nombre de cailloux pour les associations Maendeleo, Tuungane et ADEPB et celle des rangs totaux pour les différents traitements, les traitements les plus appréciés à Kabamba lors de l’évaluation à la floraison sont : variété améliorée soja comme légumineuse (VARAMSO2X05) avec une moyenne de 11 cailloux et le rang moyen 2 , variété améliorée parcelle prévue pour la deuxième légumineuse et NPK appliqué (ARAM2L2x05N) avec une moyenne de 10 cailloux et le rang moyen 3 et variété améliorée écartement 2mx0,5m (VARAM 2x 05) avec une moyenne de 9 cailloux et le rang moyen 3 (Figure 3). Le traitement le moins préféré est la variété améliorée parcelle prévue pour la deuxième légumineuse et NPK non appliqué (VARAM2L2x05).

0

2

4

6

8

10

12

MCaillouxTO

MRangTO

MCaillouxTO 5 6 5 9 3 10 6 11

MRangTO 6 6 6 3 7 3 5 2

VARLOE

CARL

VARLO1

X1

VARAM

1X1

VARAM

2X05

VARAM

2L2X05

VARAM

2L2X05N

VARAM

AR2X05

VARAM

SO2X05

Figure 3 : Fréquences totales des moyennes des cailloux (MCailluxTO) et des rangs (MRangTO) par parcelle d’essai La moyenne des cailloux (MCailluxTO) est e relation inverse de celle des rangs (MRangTO). Plus une parcelle avait de cailloux, moins son rang était élevé ; ce qui veut dire qu’elle était plus préférée. Pour l’appréciation de trois traitements préférés (en gras) et du seul traitement moins préféré (en italique), 48 appréciations positives et 19 appréciations négatives ont été données (Tableau 3). La proportion du nombre d’avis négatifs sur l’ensemble des avis émis est très élevée pour la parcelle moins préférée VARAM2L2X05 (6/16) ; elle est faible dans la parcelle la plus préférée VARAMSO2X05 (4/13).

161

Tableau 3 : Avis positifs et négatifs émis par critère pour chacune des quatre parcelles.

VARAM2X05 VARAM2L2X05 VARAM2L2X05N VARAMSO2X05 Avis positifs

Avis négatifs

Avis positifs

Avis négatifs

Avis positifs

Avis négatifs

Avis positifs

Avis négatifs

Fleursgousseab 4 5 5 3 Biomasse 1 1 1 1 2 2 Feuilleaaspect 2 1 2 4 Hauteur manioc 2 1 2 1 2 2 3 Eaurésiste 2 3 3 Maladiepas 3 1 1 1 Tigegrosse 1 1 1 Germinationbon 1 Précocité 1 Ravageurspas 1 Total 13 4 10 6 12 5 13 4

De quatre parcelles reprises dans le Tableau 3, les légumineuses (haricot et soja) sont préférées pour l’abondance de leurs fleurs et gousses. Mais la variété de haricot utilisée CODMLB003 n’a pas résisté à l’eau. La plupart des plants ont des feuilles ‘brûlées’ par l’eau, surtout dans la parcelle non préférée VARAM2L2X05. La hauteur des plants de manioc est faible dans les parcelles où la biomasse des légumineuses est abondante : c’est le cas de la parcelle avec soja où la variété SB24 ombrage le manioc ainsi que la parcelle avec NPK où la biomasse des haricots nains avec feuilles toujours vertes est abondante. En examinant l’effet des traitements, le haricot dans la parcelle avec NPK est plus préféré que le haricot dans la parcelle sans NPK qui est cité sept fois comme étant beaucoup inférieur et seulement une fois comme étant beaucoup supérieur (Tableau 4). La parcelle avec NPK a plus de fleurs et gousses que la parcelle sans NPK. La préférence est inverse pour le manioc. Le manioc dans la parcelle avec NPK est moins préféré que le manioc dans la parcelle sans NPK qui est quatre fois supérieur et trois fois beaucoup supérieur. L’aspect du manioc n’est pas bon sous l’ombrage d’une légumineuse ayant une biomasse élevée. Tableau 4 : Résultats de la comparaison des parcelles sans et avec NPK ; VARAM2X05≤ < = > ≥ VARAM2L2X05NPK. Critère Culture Nbre ≤ Nbre < Nbre = Nbre > Nbre ≥

Haricot 2 0 1 1 0 Feuilleaspect Manioc 0 1 1 2 2 Haricot 4 0 1 1 0 Fleursgousseab Manioc 0 0 0 0 0 Haricot 1 0 0 0 0 Tigegrosse Manioc 0 1 1 1 0 Haricot 0 0 1 1 0 Maladiepas Manioc 0 0 3 1 0 Haricot 0 0 0 1 1 Eauresiste Manioc 0 1 0 0 0 Haricot 0 0 0 0 0 Planthaut Manioc 0 0 0 0 1 Haricot 7 0 3 4 1 Total Manioc 0 3 5 4 3

N.B. : ≤ : beaucoup inférieur que, < : inférieur que, = pas de différence, > : supérieur que ≥ : beaucoup supérieur que. En comparant les parcelles avec écartements de 1mx1m et 2mx0,5m la différence n’est pas évidente entre les plants de manioc sur les deux parcelles, sauf au niveau du critère aspect de feuilles où la parcelle à 1mx1m est cité trois fois comme inférieure à la parcelle à écartement de

162

2mx0,5m. Le haricot sur la parcelle à écartement de 1mx1m est cité sept fois comme étant inférieur à celui dans la parcelle à écartement de 2mx0,5m contre trois fois supérieur. La différence liée au sexe n’est pas évidente entre les hommes (McaillouHO) et les femmes (McaillouFE) pour la moyenne de cailloux utilisés lors de l’évaluation (Figure 4 ).

0

5

10

15

McaillouFE

McaillouHO

McaillouFE 5 6 5 9 3 10 6 9

McaillouHO 4 3 5 9 3 10 5 12

VARLOE

CARL

VARLO1

X1

VARAM1

X1

VARAM2

X05

VARAM2

L2X05

VARAM2

L2X05NP

VARAM

AR2X05

VARAMS

O2X05

Figure 4 : Effet du sexe sur la préférence des parcelles d’essais Il en va de même pour la moyenne de rangs entre les deux sexes. 3. QUELQUES COMMENTAIRES, OBSERVATIONS FAITES SUR LE TERRAIN ET SUGGESTIONS Les résultats de cette évaluation montrent que les parcelles avec technologies améliorées (engrais NPK et écartement 2mx0,5m) ont été plus préférées que celle n’ayant pas ces technologies (parcelles sans engrais et écartement de 1mx1m). Cette observation s’applique surtout à la culture de haricot dont la tendance de rendement était déjà perceptible à partir des gousses. Pour le manioc, il est trop tôt de juger de la pertinence de la préférence. Bien que les parcelles préférées soient celles avec écartement de 2mx0,5m pour le manioc, les agriculteurs ont dit qu’ils reconnaissent la facilité d’entretien et l’économie des semences dans un champ semé en ligne. Mais le semis en ligne prend plus de temps de semis par rapport au semis en vrac : ce qui fait qu’il peut faire rater la saison à un agriculteur n’ayant pas suffisamment de main d’oeuvre. Lors du semis à ADEPB, à presque même durée, huit personnes auraient semé une parcelle en ligne contre deux femmes qui ont semé en vrac la même dimension de parcelle. Parfois les parcelles avec technologies améliorées ont un reçu un score d’appréciation inférieur à celui des parcelles sans ces technologies. Les cultures sur une parcelle avec technologie non améliorée ont dans certains cas eu la même préférence ou ont été mieux préférées (selon les évaluateurs paysans) que celles sur une parcelle avec technologie améliorée. Ce résultat inattendu ne semble pas lié au traitement appliqué mais plutôt aux conditions de terrain. Au niveau de l’association ADEPB, la pente de la parcelle avec engrais était plus prononcée que celle de la parcelle correspondante sans engrais. Cette dernière où l’érosion n’est pas manifeste était sur un terrain plat, elle accumule aussi des sols des couches supérieures érodés des pentes situées en amont dans le même essai. Au niveau de l’association Maendeleo, les paysans se sont demandés si l’ombrage des bananiers n’a pas déterminé un mauvais comportement de haricot comparativement à une parcelle qui était totalement ensoleillée. Claude Rubyogo s’était aussi posé la même question mais en l’absence des agriculteurs évaluateurs. S’agissant du NPK, au niveau de l’association Maendeleo, les femmes n’ont pas préféré la parcelle avec NPK qui était la plus préférée par les hommes. Les femmes ont reconnu que cette appréciation était liée à leurs observations faites dans un autre champ où les cultures sur la parcelle avec engrais n’ont pas poussé.

163

Eu égard à ce qui précède, nous suggérons : - La variété améliorée de haricot nain COODMLB003 appliquée dans les essais CAS-2

n’a pas résisté à l’eau. Elle n’était pas une des meilleures variétés. Pour les prochains essais, il serait mieux d’utiliser une des variétés naines qui a été le plus appréciées et qui continue à l’être à Kabamba ;

- Il existe des outils de technologie appropriée pour semer rapidement en ligne, par exemple le rayonneur. Il pourrait être opportun d’inventorier ces outils et d’utiliser un pour le semis en saison B-2008 ;

- Dès qu’il y a des assurances que l’engrais est efficace dans l’amélioration des rendements des cultures à Kabamba, une campagne de sensibilisation pourrait être amorcée dans le but de favoriser l’adoption ;

- Ce n’est qu’à l’évaluation des récoltes qu’on peut se faire une opinion sur les préférences des technologies essayées.

- Nous n’avons pas enregistré de reproche lié à l’essai CAS-2 lors de l’évaluation. Cependant lors de l’évaluation de l’essai SYS-2 à Luhihi, l’association Rhusimane a reproché l’utilisation des monocultures dans les essais : la monoculture, ont dit les membres, constitue une perte car après la récolte des légumineuses et des céréales il ne reste plus rien dans le champ alors que si ces cultures étaient associées au manioc, celui-ci leur permettrait de gagner plus ;

- Les cultures semées en association avec le manioc au champ en saison B à Kabamba sont surtout le haricot et les céréales (maïs et/ou sorgho).

164

AAANNNNNNEEEXXX 111666::: FFFEEERRR---111::: IIIDDDEEENNNTTTIIIFFFIIICCCAAATTTIIIOOONNN OOOFFF IIINNNPPPUUUTTTSSS

RRREEEQQQUUUIIIRRREEEDDD FFFOOORRR SSSOOOIIILLL FFFEEERRRTTTIIILLLIIITTTYYY AAAMMMEEENNNDDDMMMEEENNNTTT

Objectives: (i) to investigate yield limiting soil constraints for bean and maize and examine yield improvements by amendments with combinations of lime, manure, NPK and mavuno; (ii) to explore variability in soil fertility status within the study area;

Sites: 8 sites: 2 fields (one of medium and one of poor soil fertility status) at each of the 2 associations at Lurhala (CINAMULA and ALEMALU) and Burhale (APACOV and ABAGWASINYE).

Treatment structure Each trial consists of 10 plots (treatments), in 2 separate blocks (blocked per species): - main plots: soil amendments 1. control; 2. NPK only; 3. mavuno (NPK + micronutrients); 4. FYM only; 5. FYM + NPK; 6. FYM + mavuno; 7. lime only; 8. lime + NPK; 9. lime + mavuno; 10. Tithonia residues (take subsample per trial, determine FW and DW and store sample).

- blocks: species (maize, climbing bean);

Trial management Plant varieties and spacing: Maize: Variety: Katsumani cv.;

Plant spacing: 0.75m between rows, 0.25m between plants within the row; Plot size: 6 rows of 3m = 4.5m × 3m = 13.5m2.

Climbing bean: Variety: AND10; Plant spacing: 0.50m between rows, 0.10m between plants within the row; Plot size: 6 rows of 3m = 3m × 3m = 9m2.

Area required per trial = 10 × (9 + 13.5) = 225 m2 ≈ 3 are.

Soil amendments and mode of application: Lime: 4 tonnes ha-1, broadcasted and incorporated; (=5.4 kg per maize plot of 13.5 m2 and 3.6 kg per climbing bean plot of 9 m2) FYM (from goats): 10 tonnes fresh matter ha-1, broadcasted and incorporated; (=13.5 kg per maize plot of 13.5 m2 and 9 kg per climbing bean plot of 9 m2) NPK: 20 kg P ha-1, banded application, in the planting line (but deep enough not to hamper

germination). (=26.5 g per maize row of 3 m and 17.7 g per climbing bean row of 3 m)

Mavuno: 20 kg P ha-1, banded application, in the planting line (but deep enough not to hamper germination).

(=40.9 g per maize row of 3 m and 27.3 g per climbing bean row of 3 m) For maize, additional urea (top-dressing) is applied 6 weeks after planting at 60 kg N ha-1 (for

treatments with NPK or mavuno only); Tithonia: 5 tonnes DM ha-1, (≈ 25 tonnes fresh matter ha-1), broadcasted and incorporated. (=33.75 kg per maize plot of 13.5 m2 and 22.5 kg per climbing bean plot of 9 m2)

0.75m 0.5m

3m

Maize Climbing bean

0.75m 0.5m

3m

Maize Climbing bean

165

Observations DAY 0 - Initial soil description and soil fertility evaluation. Per trial, at least 8 soil samples are taken at 0-15cm prior to trial installation and application of soil amendments, using a ‘W’ design. Soil samples are bulked, air-dried and stored pending shipment and analysis (in total, 8 soil samples must be collected: 2 action sites × 2 associations × 2 trials per association). DAY 28 - 32 - Soil sampling at 4 weeks after planting. Per treatment and species, 6 soil cores (0-15cm) are collected diagonally across the plot from between the planting rows and are bulked and air-dried. A subsample of 200 g is send to Nairobi for analysis; the remaining soil is stored at INERA. - Maize leaf sampling at 4 weeks after planting. Cut off (using a thoroughly cleaned knife) the youngest fully-developed leaf (i.e. the leaf collar emerged) from its collar for 1 plant in the 3rd and 1 plant in the 4th line (at least 0.5 m away from the plot ends). Choose representative plants of average height (not the largest or the smallest plants). Avoid plants with diseased or damaged leaves, or plants having leaves stained with soil. Place the leaves flat into large brown paper bags. Do not fold the leaves – if the leaves are too long then cut the leaves in half, using a thoroughly cleaned knife. Label the bags with the trial name, date, mandate area, action site, association name, field identification or farmer name, “MAIZE LEAVES at 4WAP” and treatment. The bags with samples need to be laid out in the sun and dried as quickly as possible. Leave samples should not be placed in plastic bags as this may cause “respiration”. Neither should the leaf samples be laid bare to dry. These leave samples will be analysed for micronutrient concentration, so all contamination should be avoided. If sun-drying is not possible, leaf samples can be directly dried in an oven (65ºC) – care should however be taken and not too many samples should be dried at once, to allow fluid air circulation and swift drying. At all times the samples need to remain in the closed paper bags and should be kept in a dry, well-aired place, protected from dust. DAY 55 - 70 - Maize biomass and ear leave sampling at tasseling. At tasseling, once the ear leaf is completely visible (silking to early milk stage, commonly between 55 and 70 days after planting), 3 maize plants are harvested randomly within the 3rd row or 4th row (at least 0.5 m away from the plot ends). Avoid plants sampled for leaves at 4 weeks after planting, and damaged, diseased or abnormally developed plants. The plants are separated in leaves (cut off at the collar), stem + tassel, ear leaves and the immature kernel + silk + cob. Use a thoroughly cleaned knife for cutting up, to avoid contamination. Do not fold up any plant parts, but cut up in pieces (keep pieces as large as possible) to fit the paper bags. Label the bags with the trial name (FER-1), date, mandate area (SK), action site (Lurhala or Burhale), association name, field identification or farmer name, plant part (“MAIZE LEAVES at TASSELING”, STEM + TASSEL”, “EAR LEAVES” or “KERNEL + SILK + COB”) and treatment. The bags with samples need to be laid out in the sun and dried as quickly as possible. Samples should not be placed in plastic bags as this may cause “respiration”. Neither should any samples be laid bare to dry. These samples will be analysed for micronutrient concentration, so all contamination should be avoided. If sun-drying is not possible, leaf samples can be directly dried in an oven (65ºC) – care should however be taken and not too many samples should be dried at once, to allow fluid air circulation and swift drying. At all times the samples need to remain in closed paper bags and should be kept in a dry, well-aired place, protected from dust. BEANS 50% FLOWERING - Bean biomass and leaf sampling at 50 % flowering. At 50 % flowering, 4 bean plants are harvested: 2 plants in the 3rd row and 2 plants in the 4th row (at least 0.5 m away from the plot ends). Plants are cut at least 5 cm above the soil surface to avoid contamination by soil particles. For each of the plants cut, the growing tips are identified

166

(several per plant) and starting from each growing tip, the youngest fully developed leaves (tri-foliates) are separated and stored in a separate paper bag. The leave sample should not be folded, but placed neatly inside the bag. Bags are labelled with the trial name (FER-1), date, mandate area (SK), action site (Lurhala or Burhale), association name, field identification or farmer name, plant part (“BEAN LEAVE SAMPLE at 50% FLOWERING” or “BEAN BIOMASS at 50% FLOWERING”) and treatment. The bags with samples need to be laid out in the sun and dried as quickly as possible. Samples should not be placed in plastic bags as this may cause “respiration”. Neither should any samples be laid bare to dry. These samples will be analysed for micronutrient concentration, so all contamination should be avoided. If sun-drying is not possible, leaf samples can be directly dried in an oven (65ºC) – care should however be taken and not too many samples should be dried at once, to allow fluid air circulation and swift drying. At all times the samples need to remain in the closed paper bags and should be kept in a dry, well-aired place, protected from dust. BEANS 50% PODDING - Climbing bean biomass sampling at 50 % podding. At 50 % podding, plants within a 50 cm section in the net plot are cut (note the number of plants cut). Sections missing plants due to sampling at 50 % flowering need to be avoided. At the same time, 3 maize plants per treatment are cut as reference for BNF assessment (in the 2nd, 3rd, 4th of 5th row, at least 50 cm away from the plot ends; avoid plants sampled for leaves at 4 weeks after planting). If this coincides (less than 5 days difference) with the maize biomass sampling at tasseling, harvesting of maize reference plants for BNF assessment can be omitted. Biomass samples are air-dried, weighed, ground and stored pending shipment and analysis. HARVEST - Maize grain and biomass yield at harvest. Bean grain yield at harvest. Samples are air-dried, weighed, ground and stored pending shipment and analysis.

167

AAANNNNNNEEEXXX 111777::: PPPRRROOOTTTOOOCCCOOOLLL FFFOOORRR OOONNN---SSSTTTAAATTTIIIOOONNN AAANNNDDD OOONNN---FFFAAARRRMMM

MMMUUULLLCCCHHH TTTRRRIIIAAALLLSSS Introduction Deux types de dispositifs expérimentaux sont retenus : - essais menés en milieu contrôlé, soit en station expérimentale, soit en milieu agricole,

mais dans les deux cas gérés par les chercheurs (étudiants doctorants) : ces essais sont destinés à comprendre les processus (process oriented research) et sont assortis de mesures et déterminations choisies et formatées dans ce but ;

- essais menés milieu paysan, dans des parcelles situées en milieu réel, dans un triple but d’extrapolation, de validation et démonstration.

Ces deux dispositifs sont assortis de matériaux partagés, de suivis communs et de suivis distincts : - Matériaux communs : sites (1), protocoles (2) - Suivi commun : paramètres agronomiques (3)

- Suivis distincts : - cycle des éléments minéraux (Syldie Bizimana) - propriétés physiques (Tony Muliele)

1. Sites 1.1. Essais gérés par les chercheurs Burundi Gitega Ferralsol (FAO Acrisol / « latérite ») Cibitoke Vertisol (FAO Vertisol / alluvions) Kirundo Ferrisol (FAO Nitisol sur schiste) Congo Mulungu “sol brun s/ basalte tertiaire” (FAO Nitisol-Ferralsol / basalte) Walungu “sol rouge s/ basalte tertiaire” (FAO Ferralsol / basalte) Rwanda Butare-Gitarama Ferralsol (FAO Acrisol sur granite) Kibungo Ferrisol (FAO Nitisol sur schiste) Ruhengeri sol brun (FAO Andosol sur cendres volcaniques) 1.2. Essais en milieu paysan Pour chaque site/région, nous sélectionnerons des exploitations dont le système local est dominant, afin de réduire la variabilité du « témoin » (systèmes similaires). Par système local, nous sélectionnerons au moins trois exploitations. Le nombre minimum d’essais en milieu paysan s’élèvera donc à 24. 2. Protocoles 2.1. Essais gérés par les chercheurs Traitements : témoin + 3 traitements

Pailles importées Pailles exportées Labour Haricot T0 Non Oui Oui Oui T1 Non Non1 Non Oui T2 Oui3 Non2 Non Oui T3 Oui4 Non Non Oui

1 autopaillage 2 autopaillage + importation 3 paillage graminéen (Hyparrhenia diplandraà 4 paillage graminéen (Guatemala grass : Tripsacum laxum)

168

Répétitions : 4 Disposition : en fonction du terrain (topographie, …) Carré 6x6 (36 bananiers), ‘net plot’ 4x4 plants (16 bananiers) Rectangle 7x5 (35 bananiers), ‘net plot’ 3x5 plants (15 bananiers) Densité 2x2 m Epaisseur du paillage : ~ 5cm (~25T/ha –matière sèche–) 2.2. Essais en milieu paysan Traitements Pailles importées Labour

Ta Non Oui Tb Non Non Tc Oui1 Non

1 paillage graminée

Le protocole dépend du système local appliqué par l’agriculteur, l’objectif étant de valider, en milieu paysan, les résultats obtenus par les essais gérés par les chercheurs. Dans ces conditions : - le témoin varie en fonction du système local ; - l’exportation du paillage n’est pas imposée, mais sera enregistrée ; - si Ta/Tb intègre le haricot, cette plante doit aussi être cultivée dans Tb/Tc ; - la densité sera enregistrée, en tentant de retenir des systèmes locaux adoptant des

densités proches de 2500 p/ha (2x2m). Répétitions : aucune au niveau de l’exploitation ; par système local, nous sélectionnerons au

minimum trois exploitations. Disposition : ‘net plot’ 3x3 plants (9 bananiers), ‘total plot’ 5x5 plants (25 bananiers) 3. Matériaux et caractérisation des sites Matériel de plantation

rejets baïonnettes à sélectionner sur pieds ; parage très soigneux ; désinfection à l’eau bouillante (<30 sec) ?

Caractérisation des sols Description morphologique des profils (00-160 cm) Analyses :

- granulométrie, densité apparente - propriétés physico-chimiques (CEC, pH, cations échangeables, C, N, P)

- analyses totales - analyses minéralogiques - analyses spécifiques

4. Suivi de la croissance et des paramètres agronomiques Tous les bananiers mesurés seront soigneusement numérotés. A chaque bananier enregistré, correspondra un « carnet de croissance » (fichier) reprenant toutes les mesures liées à la croissance et aux paramètres agronomiques. 4.1. Essais gérés par les chercheurs De la plantation à la récolte (rythme mensuel) : - circonférence (base du plant, 100cm dès que possible) ;

- nombre de feuilles mortes (retirées après mesure) et vivantes, feuille morte <50% de limbe vert ;

- hauteur du plant.

169

Floraison : - circonférence (base du plant, 100cm) ; - nombre de mains à fleurs femelles

- nombre de doigts de la rangée inférieure de l’avant-dernière main (complète) ; - hauteur du plant.

Récolte : - poids du régime.

Appréciation des contraintes biotiques (selon protocoles standards existants) : - dégâts foliaires dus aux maladies fongiques ;

- dégâts au rhizome et à la base du pseudo-tronc (charançon) ; - dégâts racinaires (nématodes, …).

4.2. Essais en milieu paysans Avant la floraison, rythme bimensuel à trimestriel (de la plantation à la récolte) : - circonférence (base du plant, 100cm dès que possible) ;

- nombre de vivantes, feuille vivante >50% de limbe vert ; - hauteur du plant.

Entre floraison et récolte : - circonférence (base du plant, 100cm) ; - nombre de mains à fleurs femelles ; - nombre de doigts de la rangée inférieure de l’avant-dernière main (complète) ; - hauteur du plant.

5. Paramètres relatifs aux propriétés physiques des sols Note : les déterminations réalisées dans les essais gérés par les chercheurs et en milieu paysan sont affectées par

les symboles respectifs ec et mp.

5.1. Caractérisation de base (ec) : - granulométrie ; - courbe teneur en eau succion ; - densité apparente (en fonction du volume de la motte) ; - constitution minéralogique et stock organique.

5.2. Suivi (plantation–récolte) selon un rythme saisonnier : - densité racinaire par ‘core sampling’ : ec + mp

- résistance mécanique à la pénétration : ec + mp - densité apparente : ec + mp

- teneur en eau (EM probes in PVC tubes) : ec - température (sondes) : ec - infiltration (double anneau) : ec - rainfall simulation (bananiers de bordure (1ère récolte), net plot (2ème récolte)) : ec - pluviométrie : ec + mp

5.3. Caractérisation en fin d’essai (près de la récolte du 2ème cycle) Ces caractérisations seront réservées aux observations et mesures destructives près de la récolte du 2ème cycle. Des profils racinaires1 seront ouverts au pied des bananiers significatifs :

- comptages racinaires réalisés à l’aide d’une grille : ec + mp

- résistance mécanique à la pénétration : ec + mp - densité apparente: ec + mp - infiltrométrie à disque par horizon : ec + mp

En fonction des possibilités et uniquement en scénario ec, des profils racinaires pourraient être étudiés à la récolte du 1er cycle au pied des bananiers de bordures.

1 Delvaux et Guyot (1989) Caractérisation de l'enracinement du bananier au champ. Incidences sur les relations sol-plante dans les bananeraies intensives de Martinique. Fruits, 44 (12): 633-647.

170

6. Suivi des paramètres relatifs au cycle des éléments minéraux 6.1. Caractérisation de base des sols (ec) :

- granulométrie ;

- densité apparente ; - constitution minéralogique et stock organique ; - analyses chimiques totales et stock des éléments minéraux ; - stocks minéraux et organiques des éléments nutritifs ; - stocks des éléments échangeables et non échangeables ;

- N, P minéralisables. 6.2. Suivi mensuel des paramètres pédologiques (ec, échantillon composite) :

- C, N totaux ; C oxydable, N minéralisable ;

- P total et P disponible ; - pH, cations échangeables.

6.3. Suivi cyclique des paramètres pédologiques (mp, échantillon composite, 1er+2ème cycle à la floraison cycle saisonnier à préciser) : - C, N totaux ; C oxydable ; - P disponible ; - pH, cations échangeables.

6.4. Diagnostic foliaire à la floraison (ec + mp, 1er+2ème cycle, échantillon composite) : - éléments majeurs N, P, K, Ca, Mg ; - éléments mineurs.

6.5. Minéralomasse des bananiers (ec, 1er+2ème cycle, stade récolte) :

- éléments majeurs N, P, K, Ca, Mg ; - éléments mineurs ;

- échantillonnage et analyse des parties aériennes des bananiers (cfr essais sol-plante, Martin-Prével et al.) : pseudo-tronc, feuilles, régime ;

- détermination de la minéralomasse et des exportations minérales. 6.6. Minéralomasse des pailles importées :

- éléments majeurs N, P, K, Ca, Mg ; - éléments mineurs

171

AAANNNNNNEEEXXX 111888::: RRREEELLLEEEVVVAAANNNTTT RRREEEPPPOOORRRTTTSSS,,, PPPRRREEESSSEEENNNTTTAAATTTIIIOOONNNSSS,,, AAANNNDDD PPPUUUBBBLLLIIICCCAAATTTIIIOOONNNSSS

Articles in international refereed journals Bairagi, R., and Ahsan, R.I., 1998. Inconsistencies in the findings of child nutrition surveys in

Bangladesh. American Journal of Clinical Nutrition 68:1267-1271. Carpentier S., Witters E., Laukens K., Deckers P., Swennen R. and Panis B., 2005. Preparation of

protein extracts from recalcitrant plant tissues: An evaluation of different methods for two-dimensional gel electrophoresis analysis. Proteomics 5 (10):2497-2507.

Carpentier S., Witters E., Laukens K., Panis B. and Swennen R., 2005. Proteome analysis: a successful approach for functional research in recalcitrant non-model crops. Proceedings 11th symposium on Applied Biological Sciences. Leuven, Belgium, 6 October 2005. Communications in Agricultural and Applied Biological Science 70 (2):3-4.

Carpentier S., Witters E., Laukens K., Van Onckelen H. , Swennen R. and Panis B., (in press). Banana (Musa spp.) as a model to study the meristem proteome: acclimation to osmotic stress. Proteomics.

Chun X. Xu, Panis B., Strosse H., Hua P. Li, Huo G. Xiao, Huai Z. Fan and Swennen R., 2005. Establisment of embryogenic cell suspension and plant regeneration of the dessert banana "Williams" (Musa AAA group). Journal of Horticultural Science & Biotechnology 80 (5):523-528.

Coemans B., Matsumura H., Terauchi R., Remy S., Swennen R. and Sági L., 2005. SuperSAGE combined with PCR walking allows global gene expression profiling of banana (Musa acuminata), a non-model organism. Theoretical and Applied Genetics 111 (6):1118-1126.

Coemans B., Takahashi Y., Berberich T., Ito A., Kanzaki H., Matsumura H., Saitoh H., Tsuda S., Kamoun S., Sági L., Swennen R. and Terauchi R., (submitted). High-throughput in planta expression screening identifies an ADP-ribosylation factor (ARF1) that is involved in non-host resistance and R gene mediated resistance. Plant Physiology.

Dolezelova M., Dolezel J., Van den houwe I., Roux N. and Swennen R., 2005. Ploidy levels revealed. Infomusa 14 (1):34-36.

Dussert S., Davey M. W., Laffargue A., Doulbeau S., Swennen R. and Etienne H., 2006. Oxidative stress, phospholipid loss and lipid hydrolysis during drying and storage of intermediate seeds. Physiologia Plantarum 127:192-204.

Fermont, A.M., P.J.A. van Asten, K.E. Giller. Increasing land pressure in farming systems in the mid-altitude zone of East Africa: The changing role of cassava and consequences for system sustainability. Submitted to Agriculture, Ecosystems and Environment.

Geuns J., Ledo Orriach M., Swennen R., Zhu G. Y., Panis B., Compernolle F. and Van der Auweraer M., 2006. Simultaneous Liquid Chromatography Determination of Polyamines and Arylalkyl Monoamines. Analytical Biochemistry 354: 127-131.

Henriet C., Draye X., Oppitz I., Swennen R. and Delvaux B., 2006. Effects, distribution and uptake of silicon in banana (Musa spp.) under controlled conditions. Plant and Soil 287:359-374.

Mtambanengwe F, Mapfumo P, Vanlauwe B 2006 Comparative short-term effects of different quality organic resources on maize productivity under two different environments in Zimbabwe. Nutrient Cycling in Agroecosystems, In Press.

Ohl T. J., Cullis M. A., Kunert K., Engelborghs I., Swennen R. and Cullis. C.A., (submitted). Genomic changes associated with somaclonal variation in banana (Musa spp.). Physiologia Plantarum.

Panis B., Helliot B., Strosse H., Remy S., Lepoivre P. and Swennen R., 2005. Germplasm conservation, virus eradication and safe storage of transformation competent cultures in banana: The importance of cryopreservation. Chang W.-C., Drew R. (ed.). Proceedings IInd IS on Biotech. of Trop. & Subtrop. Species. Acta Horticulturae 692. ISHS, 51-59.

Panis B., Piette B. and Swennen R., 2005. Droplet vitrification of apical meristems: a cryopreservation protocol applicable to all Musaceae. Plant Science 168:45-55.

172

Pérez Hernández J. B., Galán Saúco V., Swennen R. and Sági L., (submitted). Agrobacterium-mediated transformation is an efficient tool to generate transgenic plants from embryogenic cell suspension cultures of banana (Musa spp.), a tropical monocot species.

Pérez Hernández J. B., Swennen R. and Sági L., 2006. Number and accuracy of T-DNA insertions in transgenic banana (Musa spp.) plants characterized by an improved anchored PCR technique. Transgenic Research 15:139-150.

Pypers, P, J Delrue, J Diels, E Smolders, R Merckx 2006 Phosphorus intensity determines short-term P uptake by pigeon pea (Cajanus cajan L.) grown in soils with differing P buffering capacity Plant Soil 284:217-227.

Pypers, P, L Van Loon, J Diels, R Abaidoo, E Smolders & R Merckx 2006 Plant-available P for maize and cowpea in P-deficient soils from the Nigerian Northern Guinea savanna - Comparison of E- and L-values Plant and Soil 283:251-264.

Pypers, P, M Huybrighs, J Diels, R Abaidoo, E Smolders & R Merckx 2006 Does the enhanced P acquisition by maize following legumes in a rotation result from improved soil P availability? Soil Biology and Biochemistry, submitted.

Pypers, P., S. Verstraete, Cong Phan Thi, R. Merckx 2005 Changes in mineral nitrogen, phosphorus availability and salt-extractable aluminium following the application of green manure residues in two weathered soils of South Vietnam. Soil Biology & Biochemistry 37, 163-172.

Remy S., Thiry E., Coemans B., Windelinckx S., Swennen R. and Sági L., 2005. Improved T-DNA vector for tagging plant promoters via high-throughput luciferase screening. BioTechniques 38:763-770.

Samyn B., Sergeant K., Carpentier S., Debyser G., Panis B., Swennen R. and Van Beeumen J., (in press). Homology-based functional proteome analysis: a successful approach for the non-model plant Musa spp. Jounal of Proteome Research.

Shakir, A., and Morley, D., 1974. Measuring malnutrition. Lancet 1(7860):758-759. Strosse H., Schoofs H., Panis B., André E., Reyniers K. and Swennen R., 2006. Development of

embryogenic cell suspensions from shoot meristematic tissue in bananas and plantains (Musa spp.). Plant Science 170 (1):104-112.

Talwana H., Speijer P. R., Gold C. S., Swennen R. and De Waele D., 2006. Effect of nematode infection and damage on the root system and plant growth of three Musa cultivars commonly grown in Uganda. Nematology 8(2):177-189.

Tittonell P, Leffelaar PA, Vanlauwe B, van Wijk MT and Giller KE 2006 Exploring diversity of crop and soil management within smallholder African farms: a dynamic model for simulation of nutrient (N) balances and use efficiencies at field scale. Agricultural Systems, 91, 71-101.

van Asten, P.J.A., S. Kaaria, A.M. Fermont, R.J. Delve. Challenges of conducting rigorous farmer participatory research in Africa. Submitted to Experimental Agriculture.

Vanlauwe, B, P Tittonell and J Mukalama 2006 Within-farm soil fertility gradients affect response of maize to fertilizer application in western Kenya. Nutrient Cycling in Agroecosystems, In Press.

Zhu G. Y., Geuns J., Dussert S., Swennen R. and Panis B., 2006Change in sugar, sterol and fatty acid composition in banana meristems caused by sucrose-induced acclimation and its effects on cryopreservation . Physiologia Plantarum 128:80-94.

Papers presented at international congresses and symposia Aert R., Hribova E., Delezel J., Swennen R. and Sági L., (in press). Cot filtration of the banana

(Musa acuminata) genome. Plant Genomics European Meeting V. Venice, Italy, 11-14 October 2006. Poster Abstract.

Arinaitwe G., Remy S., Thiry E., Sági L. and Swennen R., 2005. Integration of rice chitinase genes in banana. International Consortium on Agricultural Biotechnology Research (ICABR). Ravello, Italy, 6-10 July 2005. Poster abstract.

Bationo A, J Kihara, B Vanlauwe, J Kimetu and K L Sahrawat. 2006 Integrated nutrient management - Concepts and experience from SSA (in Press).

Bationo A, Vanlauwe B, Kimetu J, Kihara J, Abdoulaye MS, Adamou A, Tabo R, Koala S 2005 Increasing land sustainability and productivity through soil fertility management in the West African Sudano-Sahelian zone. In: Nutrient and Water Management Practices for Increasing

173

Crop Production in Rainfed Arid/Semi-arid Areas. International Atomic Energy Agency, TECDOC 1468, pp. 53-75, IAEA, Vienna, Austria.

Birabwa, R., P.J.A. van Asten, I.N. Alou, G. Taulya. Got matooke (Musa AAA-EA) for Christmas. Accepted for African banana conference, Mombasa, October 2008.

Carpentier S., Swennen R. and Panis B., 2006. The use of uni- and multivariate statistics in data analysis: a case study of the banana meristem proteome. Proteom'Lux 2006: Bridging the gap between gene expression and biological function. Luxembourg, Luxembourg, 11-14 October 2006. Abstract.

Carpentier S., Witters E., Laukens K., Swennen R. and Panis B., 2005. Two-dimensional gel electrophoresis and subsequent protein identification via MALDI-MS/MS: a successful approach to unravel the abiotic stress responses in a non-model organism (Musa spp.). HUPO 4th Annual world congress: "From defining the proteome to understanding function. Technical University Munich, Germany, 28 August - 1 September 2005. Molecular & Cellular Proteomics 4 (8): S257.

Carpentier S., Witters E., Laukens K., Swennen R. and Panis B., 2006. Proteome research of banana meristems to study cryoprotection. 43rd Meeting of the Society for Cryobiology in association with the Society for Low Temperature Biology . Hamburg, Germany, 24-27 July 2006. Poster abstract.

Coemans B., Takahashi Y., Berberich T., Saitoh H., Matsumura H., Kanzaki H., Ito A., Remy S., Swennen R. , Sági L. and Terauchi R., 2006. High-throughput in planta expression screening identifies an ADP-ribosylation factor (ARF1) that is involved in plant cell death. 3rd EPSO conference. Plant dynamics: from molecules to ecosystems. Viségrad, Hungary, 28 May - 1 June 2006.

Gaidashova, S.V., G. Germeau, P. van Asten, B. Delvaux. 2007. Identification of constraints to banana production in three eco-regions of Rwanda. Proceedings of ISAR Conference, March 2007, Rwanda.

Gaidashova, S., M. Bagabe, A. Nsabimana, P. van Asten. Fusarium wilt disease in Rwanda: A real threat to apple banana market. Accepted for African banana conference, Mombasa, October 2008.

Hauser, S., P. van Asten. Methodological considerations on banana and plantain yield determinations. Accepted for African banana conference, Mombasa, October 2008.

Jagwe, J., S. Abele, L. Niyangabo. Banana Marketing in Burundi. Accepted for African banana conference, Mombasa, October 2008.

Jefwa, J.M., B. Vanlauwe, N. Sanginga, A. Elsen, E. Kahangi, L. Ruto, P. van Asten, T. Losentge, M. Mwajita, J. Mungatu.. Abuscular Mycorrhizal Fungi (AMF) and root nematode damage in banana plantains of maragua district in Central Kenya. Accepted for African banana conference, Mombasa, October 2008.

Kahangi, E., B. Vanlauwe, N Sanginga, P van Asten, Mbugua, L Ruto, Kimenju, J Jefwa. Characteristics of banana farms and causes of decline in production in maragua district in Central Kenya. Accepted for African banana conference, Mombasa, October 2008.

Muchunguzi, P, M Batte, M Nyine, M Pillay, C Kiwanuka, P van Asten, L Wairegi, J Lorezenzen. Innovations for producing clean banana planting materials by macropropagation in Uganda. Accepted for African banana conference, Mombasa, October 2008.

Murekezi, C., P.J.A. van Asten. Farm banana constraints in Rwanda: A farmer’s perspective. Accepted for African banana conference, Mombasa, October 2008.

Mwangi, M., Vigheri, N., Fiaboe, K., Bandyopadhyay, R., 2006. Emergence and Spread of Banana Xanthomonas Wilt in East D.R. Congo and Strategies to Halt its Spread Towards Central and West Africa. Presentation given by P. van Asten at the MUSACO meeting, 18-22 September 2006, Limbé, Cameroon.

Nyombi, K., van Asten, P.J.A., Taulya, G., K. Giller, 2006. The effect of soil quality parameters on growth rates of Kisansa TC banana plants in Senge, Uganda. Paper presented at the Soil Science Society of East Africa Conference, Masaka, November 2006.

Nyombi, K., P.J.A. van Asten, P.A. Leffelaar, M. Corbeels, C. K. Kaizzi and K.E. Giller. Developing a simulation model to understand and improve growth of East African highland banana (AAA-EAHB, cv. Kisansa). Accepted for African banana conference, Mombasa, October 2008.

174

Ochieno, D.M.W., T. Dubois, D. Coyne, M. Dicke, A. van Huis, P. van Asten. Benefits of non pathogenic Fusarium oxysporum on bananas as influenced by nutrients. Accepted for African banana conference, Mombasa, October 2008.

Panis B. and Lambardi M., 2005. Status of cryopreservation technologies in plants (crops and forest trees). International Workshop on "The role of biotechnology for the characterisation and conservation of crop, forestry, animal and fishery genetic resources". Turin, Italy, 5-7 March 2005. 43-54.

Panis B. and Swennen R., 2005. Applications of cryopreservation in banana. COST 843 final conference / COST 843 and COST 851 joint meeting. Stara Lesna, Slovakije, 28 June - 2 July 2005. 17-19. Abstract.

Sági L., Remy S., Coemans B., Thiry E., Santos E., Matsumura H., Terauchi R. and Swennen R., 2005. Functional Analysis of the Banana Genome by Gene Tagging and Sage. Plant & Animal Genomes XIII Conference. San Diego, CA, USA, 15-19 January 2005.

Santos E., Remy S., Thiry E., Windelinckx S., Swennen R. and Sági L., (submitted). Tagging Novel Promoters in Banana Using the Luciferase Reporter Gene. 27th International Horticultural Congress on "Global Horticulture: Diversity and Harmony". Seoul, Korea, 13-19 Augustus 2006. Acta Horticulturae.

Santos E., Remy S., Thiry E., Windelinckx S., Swennen R. and Sági L., 2006. Tagging Low-Temperature Responsive Promoters in Banana Using the Luciferase Reporter Gene. 27th International Horticultural Congress on "Global Horticulture: Diversity and Harmony". Seoul, Korea, 13-19 Augustus 2006. 218-219. Abstract.

Taulya, G., van Asten, P.J.A., Okech, S.H.O, Gold, C.S., 2006. Effect of drought on the yields of an East African highland banana (cv. Kisansa) in Uganda. Paper presented at the Soil Science Society of East Africa Conference, Masaka, November 2006.

van Asten, P.J.A., A. Twagirayezu, S. Gaidashova. 2008. Effect of Guatamala grass (Tripsacum laxum) mulch applications on soil moisture conservation and soil fertility status. Proceedings of ISAR Conference, March 2007, Rwanda

van Asten, P.J.A., L.W.I. Wairegi, F. Bagamba, and C. Drew. Factors driving fertilizer adoption in banana systems in Uganda. Accepted for African banana conference, Mombasa, October 2008.

van Asten, P.J.A., D. Florent, M.S. Apio. Opportunities and constraints for dried dessert banana export in Uganda. Accepted for African banana conference, Mombasa, October 2008.

van Asten, P.J.A., D. Mukasa, N.O. Uringi. Farmers earn more money when banana and coffee are intercropped. Accepted for African banana conference, Mombasa, October 2008.

van Asten, P.J.A., C. Murekezi, L. Wairegi, S. Gaidashova, T. Muliele, S. Hakizamana, S. Bizimana, N. Vigheri, J. Walangululu, B. Delvaux. Commonly perceived production levels and constraints in East African highland banana systems do not correspond to field realities. Accepted for African banana conference, Mombasa, October 2008.

van Asten, P.J.A., I.N. Alou, B. Vanlauwe, D. Mubiru, T.K.W Mugaaga. Resource availability and yield differences between farms in a banana producing valley in Southwest Uganda. Accepted for African banana conference, Mombasa, October 2008.

Van Asten, P.J.A., 2006. Soil quality constraints in EA highland bananas. Poster presented at the Farmer Field School Experiences'Workshop on Land and Water Management in Africa, Jinja April 24-29 2006.

Van Asten, P.J.A., Abele, Blomme, G., S. Sanginga, P., Vanlauwe, B., Vigheri, N., Lunze, L., Walangululu J., Farrow, A., 2006. The role of bananas in Eastern DRCongo. Presentation at the MUSACO meeting, 18-22 September 2006, Limbé, Cameroon.

Van Asten, P.J.A., Fermont, A.M. 2006. Is farmer knowledge guiding us in the right direction? Paper presented at the Farmer Field School Experiences'Workshop on Land and Water Management in Africa, Jinja April 24-29 2006. 11 pp.

Van den houwe I., Panis B., Arnaud E., Markham R. and Swennen R., 2005. Banana (Musa spp.) genetic resources maanged in the INIBAP-IPGRI gene bank: conservation and documentation status. 3rd GBIF Science Symposium "Tropical Biodiversity: Science, Data, Conservation". Brussels, Belgium, 18-19 April 2005. Abstract.

Van den houwe I., Panis B., Arnaud E., Markham R. and Swennen R., 2006. The management of banana (Musa spp.) genetic resources at the IPGRI/INIBAP-IPGRI gene bank: the

175

conservation and documentation status. Segers H., Desmet P. , Baus E. (ed.). Tropical Biodiversity: Science, Data, Conservation. Proceedings of the 3rd GBIF Science Symposium. Brussels, Belgium, 18-19 April 2005. 143-152.

Vanlauwe, B., P. van Asten, E. Kahangi, N. Sanginga, T. Losenge. J.M. Jefwa J.M., Ruto. Banana fertilizer application and their effects on growth, flowering and production in an on-station trial in Maragua district of Central Kenya. Accepted for African banana conference, Mombasa, October 2008.

Vigheri, L.N., L. Lukalango, K.Sikyolo, N. Dowiya, P.J.A. van Asten, G. Blomme. Etude du germoplasme des bananiers et bananiers plantains en territoire de Beni et de Lubero. Accepted for African banana conference, Mombasa, October 2008.

Vigheri, L.N., L.K. Mbogho, K.Sikyolo, G.Blomme, P.J.A. van Asten. Contribution à l’etude du maladies et ravageurs du bananiers et bananier plantain en territoire de Beni et de Lubero. Accepted for African banana conference, Mombasa, October 2008.

Vigheri, L.N., S. Saghasa, K.Sikyolo, G. Blomme, P..J.A. van Asten. La place du bananier et bananier plantain dans les systèmes de cultures en territoire de Beni et Lubero, Province du Nord Kivu, R.D. Congo. Accepted for African banana conference, Mombasa, October 2008.

Wairegi, L.W.I., P.J.A. van Asten, C. Kiwanuka, M. Tenywa, M. Bekunda. 2007. Assessment of soil management practices in East African highland cooking banana (Musa spp. AAA-EA) systems in Uganda. Paper presented in ‘International Symposium on Innovations for a Green Revolution in Africa: Exploring the Scientific Facts’, 17-21 September, 2007.

Wopereis, M, K E Giller, A Maatman, B Vanlauwe, A Mando, A Bationo 2006 Innovations for increasing productivity through improved nutrient use in Africa. Proceedings of the World Congress of Soil Science, July 2006, Philadelphia, USA.

Articles in books Panis B. and Lambardi M., 2006. Status of cryopreservation technologies in plants (crops and

forest trees). Chapter 6. Ruane J., Sonnino A. (ed.). The role of biotechnology in exploring and protecting agricultural genetic resources. Food and Agriculture Organization of the United Nations, Rome, Italy: 61-78.

Pérez Hernández J. B., Remy S., Swennen R. and Sági L., 2006. Banana (Musa sp.). Wang K. (ed.). Methods in Molecular Biology, vol. 344: Agrobacterium Protocols 2/e, volume 2. Humana Press Inc., Totowa, NJ:167-175.

Roux N., Strosse H., Toloza A., Panis B. and Dolezel J., 2005. Potential of flow cytometry for monitoring genetic stability of banana embryogenic cell suspension cultures. Chapter 25. Hvoslef-Eide A. K., Preil W. (ed.). Liquid culture Systems for in vitro Plant Propagation. Springer, 337-344.

Theses and other reports David, S., 1998. La production de semences de haricot: Manuels pour les producteurs de

semence de haricot sur une petite échelle. Manuel 1. Réseau de la Recherche sur le Haricot en Afrique, Séries de publications occasionnelles, N°29. CIAT, Kampala, Ouganda.

Germeau, G., 2006. Identification des contraintes en culture bananière traditionnelle dans trois régions du Rwanda par enquête diagnostic. MSc thesis, Université Catholique de Louvain, 113pp.

Nibasumba, A. 2007. Garniture Cationique des sols et des racines dan des systèmes de culture bananière du Burundi et du Rwanda. MSc thesis, UCL, Belgium.

Rishirumuhirwa, T., 2006. The role and management of bananas in Burundian farming systems. Consultancy report for the DGDC-funded IITA-led project on bananas in the Great Lakes region. 52pp.

Report-DGDC Feb 2008 - V to print - Betuco Annual Report 2006-2007.pdf · livelihoods in central...

Documents

Transcript of Report-DGDC Feb 2008 - V to print - Betuco Annual Report 2006-2007.pdf · livelihoods in central...