trees playing a key role in food security and livelihoods ... for... · trees playing a key role in...

Trees Playing a Key Role in Food Security and Livelihoods in Rwanda- Highlights from The T4FS Project

Trees Playing a Key Role in Food Security and Livelihoods in Rwanda

Highlights from the T4FS Project2

Where the project works

The T4Fs project works in Kadahenda and Bahimba (Gishwati) and Juru, Gashora and Rweru (Bugesera) as shown in Figure 1.

Figure 1: A harmonized map of the project sites in Rwanda: Kadahenda and Bahimba (Gishwati) and Juru, Gashora and Rweru (Bugesera)

Why Rwanda?

Over the last decade, Rwanda has been facing decreasing food security and increasing land degradation, which has impacted negatively on livelihoods. Key challenges experienced in Gishwati and Bugesera necessitated implementation of the Trees for Food Security (T4FS) Project in the sites. In Gishwati, the key challenges are: low tree cover due to deforestation and tree cover loss in Gishwati forest resulting to soil erosion and soil loss and shortage of tree products such as stakes for climbing beans, fuelwood, fruits and timber. Others challenges

include: flash flooding and landslides due to cultivation in steep slopes, loss of soil fertility due to continuous cultivation without fertilizers, heavy land fragmentation, siltation of water sources and low tree diversity. In Bugesera, the key challenges are: water shortage, low tree cover leading to shortage of tree products such as fodder and fuelwood, low crop yields due to declining soil fertility and low rainfall and drought (Figure 2).


Highlights from the T4FS Project 3

Low tree density in pasture landsFlash flooding

Cultivation on steep slopes and soil erosion Land Fragmentation

Tree cutting for fuelwoodWater scarcity and low tree density

River siltation Poor tree management and design

Figure 2: Challenges to food security in Rwanda study areas



Figure 3: Maize & Grevillea, Senna, Acacia polyacantha on border/farm in lower altitude parts (Rweru) of the semi-arid zone (left); Markhamia naturally generated and protected on farm with maize/napier, along with Alnus and Eucalyptus woodlots in Bahimba, Nyundo of the sub-humid zone (right)

1.2 Understanding on-farm tree composition through tree diversity studies

A tree diversity study involving 104 households in Bugesera and 96 households in Gishwati identified 60 on-farm tree species from 193 plots selected in 20 villages. In Bugesera, 53 tree species were identified represented by 28 families and in Gishwati 20 tree species were identified represented by 15 families. In Gishwati, the most abundant species were

Eucalyptus maidenii (71.5 percent), Yushania alpina (9.90%), Erythrina_abyssinica (7.2 %) and Alnus acuminata (5.25 %). In Bugesera, Euporbia tirucalli was the most abundant (26.6%) followed by Grevillea robusta (15.6%) and Eucalyptus sp. (14.3 %) (Figure 4). The niches of inventoried trees were homestead compound, boundary planting, Woodlots, orchards around

Project Partners

The project partners include ICRAF, Rwanda Agricultural Board (RAB), CIMMYT, University of Rwanda, World Vision Rwanda and CSIRO. The project has engaged PhD and MSc students to achieve some of the project objectives.

Project Outcomes and Achievements

1. CHARACTERIZATION

1.1 Socioeconomic

Socio-economic baseline studies were on 320 households in three semiarid sites and 324 households in three sub-humid

sites to understand patterns of tree adoption. The study revealed patterns of trees on farms that present farmers managed natural regeneration (FMNR) of timber, fruit, and fuel species and planting of species for environmental services. This could be attributed to Rwanda’s conditions susceptible to soil erosions which provide strong incentives for farmers to actively plant trees with environmental services (Figure 3).



homesteads and trees mixed with annual crops.

Figure 4: Top ten abundant on-farm trees in Bugesera, Rwanda

1.3 Drivers of tree cover loss- local knowledge perspective

Local knowledge undertaken using the AKT5 tool on 30 farmers in Gishwati and 70 farmers in Bugesera helped to identify and characterize the bio-physical and socio-economic context of the study sites. Local knowledge revealed evidence of declining tree cover and tree species diversity especially native species, which has led to loss of initial regulating and livelihood functions of trees in farming systems. Tree cover loss was triggered by post-genocide deforestation (1994/1995) to meet needs for cultivated land,

settlements and tree products from an increasing population. Although farmers had detailed knowledge about a diverse set of trees, in practice, they concentrated on few species to secure the supply of high value products, regulating and cultural ecosystem services namely the home compound, along boundaries, soil conservation structures and woodlots. Knowledge gaps limiting agroforestry adoption include: limited knowledge on tree management, ecological functions and ecological suitability of tree species (Figure 5).

Figure 5: Farmers in Gishwati undertaking Village Resource Mapping



1.5 Mapping Farmers’ Dependency on Ecosystem Services

A study conducted in Gishwati to map ecosystem services that farmers’ livelihood depend on over time revealed that farmers have become more dependent on sourcing ecosystem services away from/ outside their own farms. However, the study also revealed that local knowledge of ecosystem services is strongest at the farm level and becomes weaker with an increase in distance away from the farm. In addition, a farmer defines his/ her own livelihood boundary, not based on ecological ecosystem services or socio-political boundaries but on the source of tree products (Figure 7). Figure 7: A farmer in Gakoma, Kadahenda in Gishwati staking

her climbing beans using Alnus and Bamboo stakes harvested from her farm

1.4 Understanding the extent of land degradation through land health surveillance

The aim of this study was to generate relevant information on land health to assist planning and targeting site-specific management interventions. We applied the Land Degradation Surveillance Framework (LDSF) to characterize two sentinel sites (each 10 km × 10 km) in semi- arid (Bugesera) and sub-humid (Gishwati) agro-ecologies in Rwanda. Gishwati site had significantly higher carbon content and better soil quality for

crop production and environmental services than Bugesera site. Soil organic carbon increases with increasing longitude at Bugesera and increasing latitude for the Gishwati site which follows the trend in increasing soil moisture in the two sites. Distribution of soil organic carbon in the two sentinel sites is indicated in figure 6 below.

Figure 6: Spatial distribution of Soil Organic Carbon (SOC) in Bugesera and Gishwati sentinel sites.



1.6 Suitable tree species selection tool

A suitable tree species selection tool has also been produced that aims at matching agroforestry options to the sites’ biophysical context and farmers’ circumstances for both agro-ecologies. A total of 115 (61 exotic and 54 native) tree species have been identified in both Gishwati and Bugesera. This has been achieved through collating data on tree species identified from local knowledge, baseline studies and tree

diversity studies; and complemented with scientific data. The tool entails three outputs namely: the tree utility table indicating products and ecological functions of trees (including niches where trees are located), the bio-physical requirements for each species and also the links to other data sources (Figure 8). This tool is key in informing scaling-up interventions.

2. EXPERIMENTATION

2.1 Long term trials

Two long term trial experiments were established in Karama RAB station in Bugesera and RAB station, in Rubavu district (Gishwati). Data collection on tree performance and crop

productivity is on course. Details of the experimental design and results on these trials are found in the 2016 Rwanda Report.

Figure 8: Alnus located along bench terraces and Eucalyptus woodlot in Gihira, Kadahenda

Figure 9: Google earth map of the site in Rubavu |(left) and the trial in Karama (right)



Treatments where Croton was planted alone increased maize grain yields significantly (4.5 t ha-1) compared to Alnus alone, control ( maize planted alone) and mixture of Croton and Alnus treatment. The treatment where maize was intercropped with Alnus showed lower maize grain yield ( 2.8 t ha-1) implying that Croton planted increased maize grain by 40.6% while Alnus treatment planted alone reduced maize by 12.5%

On farm experiments

Pruning G. robusta reduces tree water uptake and increases maize yields in dry Bugesera

A study set up to assess tree crop water interactions in managed and unmanaged Grevillea robusta trees revealed

that pruned Grevillea takes up much less water than the unprunned ones.. The daily sap flow volumes range from 70- 88 l day-1 compared to 7- 34 l day-1 (Figure 10) The average maize yields across twelve farms under pruned G. robusta trees was higher (4, 724 Kg ha-1) compared to that under the unpruned trees which was 2,787 Kg ha-1 translating to nearly 70% increase in yields.

Figure 10. Daily Sap Volumes for un-pruned and pruned G. robusta trees in Bugesera between Aug-Sept 2014

More soil moisture and less solar radiation for crops under selected tree species

In Bugesera, a study assessing the direct effects of trees on annual food crops revealed that Senna and Markhamia trees showed increased soil moisture under their canopy while it was not the case for Grevillea and Alnus. Alnus reduced total

radiation by 35% while Senna reduced it by 75%. Maize yields were generally reduced under trees with some exceptions during good seasons with enough rains (for both sites) and radiation in Gishwati.



Hours

05:00 07:00 09:00 11:00 13:00 15:00 17:00 19:00

Sola

r ra

dia

tion (

watt m

-2)

0

200

400

600

800

1000Solar Rad_Under AlnusSolar Rad_Open field

Hours

05:00 07:00 09:00 11:00 13:00 15:00 17:00 19:00

Solar Rad_Open fieldSolar Rad_Under Senna

Figure 11. Solar radiation under (1m from tree in the North) and away (20m from tree in the North) from Alnus in Gishwati (A) and from Senna in Bugesera (B). Data are averages of 3 days measurements on weather stations logging every 5 minutes.



Participatory Trials- farmers trying out different agroforestry technologies on their farms

Increased crop yields with green tree biomass incorporation

Effect of location on maize yield

The biomass incorporated were green manure from Vernonia amygdalina, Gliricidia sepium, Calliandra calothyrsus, Leuceana diversifolia applied alone, combined with inorganic fertilizers (DAP +Urea) compared to inorganic fertilizers applied alone and farmyard (Farmer practices). The biomass equivalent to 5 t ha-1 (dry matter basis) involved 130 farmers in Juru (62), Rweru (20) and Murama (48). The effect of the green manure varied with site, with Rweru recording systematically higher maize grain yield in all the treatments (3.1. to 6.9 t ha-1) than Juru (1.9 t0 6.7 t ha-1). All green manure showed higher maize grain yield than the control (Farmyard manure). The combination of Biomass and inorganic fertilizer was the best treatment on maize yield increases. In all treatments, Gliricidia sepium had the highest impact on maize yields amongst the four shrubs.

The case of Alnus Acuminata green manure in

Gishwati

In Gishwati, where the effect of tree biomass was tested on different crops, a combination of Alnus acuminata leaf biomass and inorganic fertilizers (DAP +Urea) significantly increased crops yields.

Climbing beans yields increased by 38.9%, maize by 33% while Irish potato yields increased by 16.7%, compared to the previous farmer practices where only inorganic fertilizer was used.

Improvement of food security through provision of alternative species for Stakes for climbing beans in Gishwati

The introduction of climbing beans, which produce thrice more than bush beans has increased food security and improved livelihoods of the Rwandan population. Climbing beans require stakes for optimum production and any shortage of them greatly affect yields. Traditionally, Napier (Pennisetum purperum) has been used to stake beans. However, it is weak and can only be used for one planting season and it is very competitive with adjacent crops. The T4FS project introduced tree species like Acacia augusitissima, Alnus acuminata and Vernonia amygdalina as solution to stakes for climbing beans in Rubavu district. Fifty-two thousand (52,000) stakes are required per hectare and one stake cost 35 Rwanda Francs ($US1.04) on average for alnus, acacia, etc. per stake. These trees are fast growing and also contributes to the improvement of soil fertility through biomass incorporation, rich in nitrogen, control soil erosion and reduce distance for famers collecting stakes in the bushes.

Figure 13 shows the earlier measurements of different species. A. angustissima reach more 1.4 meters 90 days after planting.

Figure 12. Farmers incorporating Alnus acuminata biomass into the soil in Kadahenda, Karago sector



Figure 13. Agroforestry trees alternative as stakes for climbing

Case Study

Agroforestry for watershed protection- Busoro Gishwati

The landscape of Busoro and Matyazo was affected by landslides which reduce the size of arable land, destroyed tea plantations downhill and sediments were transported into Nyamukongori River, which carries it to Lake Karago. The size of Lake Karago was reduced due to the siltation. The area was mapped and characterized (Structure and texture) and the intervention designed. Alnus acuminata was planted at 50cm within the lines of trees while the distance between the lines depend on the slopes. After 4 years, the landslides were reduced significantly and the land for tea plantation was recovered. Capacity building of farmers was conducted for maintenance of the planted trees.Figure 14. Mr Idelphonse and his wife Nirere Beatrice removing

harvested Acacia angustissima seeds from pods

According to Mr Mbahingana Ildephose, a farmer in Bahimba, Nyundo sector, bean production has increased significantly. A. angustissima is also fast growing and he has already harvested bean stakes twice and he is very happy to replace the short-life P. purpureum stakes with the Acacia. He also noted that since climbing beans are

strongly anchored, they grow with more vigour, produce more, and are less susceptible to yellowing of leaves and diseases. His family can now comfortably produce enough beans for consumption and income; and he plans to plant more trees from the harvested seeds.



Fig. 15. Watershed management of Karago

3. SCALING UP AND ENABLING ENVIRONMENT

Since its inception in Rwanda, the T4FS has reached 7,347 farmers through various approaches as shown in the table below.

Scaling up approach Numbers of community members reached

Capacity development and technical training 171

World Vision Rwanda engagement 3438

Farmer trainings in Bugesera and Gishwati 129

Participatory trials and tree planting 3112

Project surveys 947

Rural Resource Centres in Karago and Karama 1743

Farmers reached through Umuganda 1894

Total 11434

3.1 Tree seed and seedling system

A survey carried out to map tree seed and seedling supply among nursery workers, operators, and farmers revealed that seed and seedling system in Rwanda uses a top-down approach and is largely government-controlled, with seeds

distributed by Tree Seed Center of Rwanda Agricultural Board (RAB) and seedlings raised mostly through cooperatives. Most farmers and nursery operators therefore lacked the technical capacity for seedling production and nursery management, access to improved/quality germplasm, and had low diversity of tree species to choose from. However, the results showed that self-collected seeds have a rapid and higher germination rate (89% to 95%) than seeds sourced from RAB (73% to 97%). Nursery operators involvement into tree nursery business rely on previous experiences obtained either from Government or projects nurseries, while farmers rely on observation or own knowledge.

Figure 16: Tree nursery in Karago

3.2 Policy Dialogues

The overall objective of the step-wise policy dialogue process was to establish a policy framework suitable for scaling up the use of trees within farming systems and then scale out success to relevant agro-ecological zones. Results indicate that constraining policy factors that may inhibit the flow of benefits from trees from local to the national level were: socio-economic and cultural factors (lack of seeds; shortage of land; lack of skills on tree management; lack of markets, weak extension services, theft of seedlings/ tree products, conflicts along farm boundaries). Environmental constraints include: trees take long to harvest; some trees compete for nutrients with crops; deforestation, drought; termites; and agro-ecological inadaptability of some species. Priority



action plan issues in Rwanda are: capacity building for identification of suitable tree species, government driven incentive, programs for positive tree management, support of management practices such as FMNR and the availability and diversity of tree species distributed through government programs.

Figure 18b: .A farmer promoter/lead farmer explaining about the different technologies on his farm

3.4 Scaling out activities through Umuganda and WV led activities

Through ‘Umuganda’ community service, farmers are being sensitized on agroforestry best practices. We have found this to be one of the most effective ways of scaling up and out, as many farmers including their local leaders are reached within a short period of time. Through this approach, we have reached over 2,000 farmers in both Gishwati and Bugesera since 2015. We are also scaling up tree planting activities during umuganda sensitization meetings, where farmers and their leaders collectively identify a suitable site and they are provided with trees to plant, especially for provision of ecological services such as soil erosion control at the landscape scale. Through the umuganda approach, we are also scaling out agroforestry technologies beyond the T4FS project areas such as Kanama sector in Gishwati, which was identified as having high level of land degradation and lacking soil and water conservation interventions.

Figure 17: Participants in a group activity for work plan development in Rwanda

3.3 Survey of the agricultural extension system

This study which involved key informants interviews revealed that RAB, an institution under MINAGRI (Ministry of Agriculture) is the highest level of authority at the national level which heads extension, while the lowest level of extension is at the village (Umudugudu) level, mainly coordinated by farmer promoters. The government has adopted a pluralistic model of extension with many extension providers that include: farmer organizations, civil society, NGOs, places of worship, private sector, institutions of learning and financial institutions. The government also builds on existing initiatives such as Umuganda, Imihigo, Ubudehe, Integrated Development Program, Girinka, and Agasozi Ndatwa. The main extension methods used are both individual and group extension; lead farmers/farmer promoters, farmer field schools, demonstrations on individual farms, individual farm visits, field days and mass media e.g. print outs and brochures and community mobilization campaigns. The most common and effective extension method is farmer promoters who host demonstration plots. The main obstacle of extension common in all areas is lack of resources due to constrained budget, including lack of transport facilities and extension materials.

Figure 18a:.Farmers in a demonstration plot



Figure 19: Farmer sensitization through umuganda in Bahimba, Nyundo sector, Gishwati

3.5 Prioritization and Rapid Market Appraisal for tree products in Rwanda

Rapid Market Appraisal (RMA) tools (key informant interviews, observation, FGDs) were used to conduct market analysis for potential products in both agro-ecologies. Results indicate that the preferred tree species in the sub-humid zones include avocado, tamarillo, passion fruits, papaya and eucalyptus, while mango and guava were important in the semi-arid zones in addition to those preferred in the sub humid zones. Producers usually sell products to middlemen who then sell to traders. No business services exist in terms of extension, credit and market information for tree products. The main

form of governance of the chain is spot market. Demand for tree products preferred by farmers generally exceeds supply indicating availability of markets. Products are usually sold with the only value addition being grading and sorting. No processing of the fruit takes place at the local level. There are no group activities in production and marketing. Lack of grafted planting materials and germplasm for most of the fruit trees such as avocado, mangos, oranges, plum and tamarillo was a major constraint in the two study areas.

Figure 20: Tamarillo in Gishwati, and Traders selling fruits in Mahoko market, Rubavu, Gishwati



3.6 Rural Resource Centers (RRC)

Two RRCs were successfully initiated in Karago (Gishwati) and Karama (Bugesera) sector to accelerate farmer learnings through trainings and demonstrations. This has been done in collaboration with farmer cooperatives, government extension services and non-governmental organizations such as World Vision Rwanda. Successes, challenges and future plans of the RRCs have already been documented but some key highlights include:

• Setting up of seedlings production facilities and demo sites in the two important agricultural areas

• Production of at least 30,000 (value US $ 26,400) fruit seedlings for avocado, tamarillo and mango

• Production and distribution of over 400,000 (value US $ 220,000) seedlings of A. acuminata, A. angustissima C.

megalocarpus, Dombeya torrida, G. sepium, G. robusta and C. calothyrsus

• Trainings for 3 farmer cooperatives on good nursery practices and fruit grafting such as mango

• Establishment of 3 satellite nurseries for trained farmers

• Demonstrations on tree biomass utilization for improved soil fertility management

Key challenge includes financial resources to support required satellite nurseries and hiring technical staff to support activity implementation in the two sites. A more sustainable approach would be to build the capacity of SACCO members and nursery operators.

Figure 21: Acacia angustissima and tamarillo seedlings production at the Karago RRC

4. MONITORING AND IMPACT ASSESSMENT

4.1 Farmers’ perceptions about participatory trials

A survey conducted in 2014 to determine farmers’ perception on the agroforestry participatory trials established in Bugesera and Gishwati indicated that the major challenges facing farmers in implementation of the agroforestry technologies include prolonged drought periods, unavailability of tree germplasm and browsing of trees by livestock. In the Alnus

biomass accumulation trials, 100% of the interviewed farmers preferred a combination of Alnus biomass and inorganic fertilizers. Despite the laborious tasks of sourcing and applying biomass on fields, farmers highlighted that crop yields and soil fertility had significantly improved. Furthermore, use of Alnus biomass maintained soil fertility over several cropping



seasons. In the Nyundo sector trials, farmers preferred Acacia. angustissima trees to Alnus because of their adaptability to the dry conditions and due to their multiple functions. In the Bugesera trials, farmers prefer Calliandra calothyrsus and Leucaena trichandra to Gliricidia sepium mainly due to provision of better fodder. Gliricidia sepium was perceived to be most effective for soil fertility improvement. In all trials, farmers were willing to expand and maintain the participatory trials even without external interventions.

Figure 22: An enumerator interviweing a farmer in Gishwati

4.2 Tracking the Spread of Agroforestry Technologies

Farmers involved in the participatory trials reported that there has been an increase in the number of farmers enquiring about the new technologies, mainly: immediate neighbours, casual labourers working on participatory trial farms, people from within their village or neighbouring villages, relatives or strangers/ passers-by who are keen to know about the well performing technologies. The exchange has been mainly through sharing of knowledge and germplasm. We have translated the ‘tracking spread of agroforestry technologies’ record sheet into Kinyarwanda language and trained farmers who can read and write on how to record this information. For

Figure 23: RAB Technician (left) and Farmer promoter (right) training farmers on how to record information using the ‘Tracking spread of technologies sheet’ in Karandaryi, Kadahenda

the significant number of farmers with no formal education, we are working with farmer promoters in each village who are visiting these farmers regularly to record the information.



LOOKING AHEAD…

Learning from the past successes and impact the project has made in Rwanda, in order to accelerate the adoption of these locally adapted technologies beyond the current project areas, we will employ several approaches that have been deemed effective. These include:

• Using individual and group extension approaches to reach as many farmers as possible such as: RRCs, farmer exchange visits, umuganda, demonstrations and field days

• Addressing bottlenecks within the tree product value chain

• Training of agronomists and other extension workers in agroforestry best-practices

• Undertake further assessments and simulations of tree-crop-water-soil nutrient interactions for more tree species and differing agro-ecological conditions and

• Monitoring and evaluating the impacts of trees on farm

Contributors

Athanase Mukuralinda, Tenge G. Ngonga, Bernard S. Musana, Catherine Muthuri, Amini Mutaganda, Alain Ndoli, Miyuki Iiyama, Safari Dieudonne, Anne Kuria, Ermias Betemariam, Evelyn Kiptot, Edmundo Barrios, Ruth Kinuthia, Albert Mwangi, Sammy Carsan, Joseph Tanui, Amos Gyau, Jeremias Mowo and Fergus Sinclair

April 2016

trees playing a key role in food security and livelihoods ... for... · trees playing a key role in...

Documents