Success story of implementing the self-sustaining agricultural extension system in Rwanda

Rwanda is implementing the self-sustaining extension system through Farmer Field Schools (FFS) and Farmer Promoters (FP) approaches. The objective of this paper was to find out the impact of selfsustaining extension system in order to help stakeholders to improve its current implementation. The methodology includes a desk review of reports, face to face interview with 60 participants and 5 focus group discussions between February and May 2016. It also includes the interview of 400 trained farmers and 400 non-trained farmers. It was found that 92% of the trained FFS facilitators and 62% of the farmer promoters were very active in extension services. It was also found that for beans, the highest average yield was 1.2 t/ha for non-trained farmers, 1.5 t/ha for FFS farmers, 1.3 t/ha for FP farmers and the average yield of all the farmers was worked out to be 1.4 t/ha. It was found that FFS trained farmers produce 37.5% more than non-trained farmers while farmers trained by Farmer Promoters produce 10.8% more than non-trained farmers. In general, 37.8% of farmers apply Good Agricultural Practices (GAP) among the non-trained farmers, 73% of FFS farmers use the GAP and 68.3% of the FP farmers adopt the GAPs. It was found that 20% of the FFS group activities are involved in various income generating activities compared to non-trained farmers (10%). It is concluded that the implementation of self-sustaining agricultural extension system in Rwanda has a strong impact in agricultural development through motivation and increased trainings of farmer promoters.


INTRODUCTION
Rwanda is a land locked country in East Africa.The Government of Rwanda sees agriculture development as a key catalyst to engender long-term sustainable growth and remove thousands out of poverty.The Crop Intensification Programme (CIP) which is a flagship program implemented by the Ministry of Agriculture and *Corresponding author.E-mail: innocent.musabyimana@gmail.com.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License Animal Resources to attain the goal of increased agricultural productivity in the country, has been very successful in increasing production of staple food crops, through improvements to smallholder productivity, and helping Rwanda achieve food self-sufficiency.
Most of the cultivated land is under food crops grown twice a year (Season A and Season B) with a total annual harvesting area of almost 1,500,000 ha.The area covered by lakes and rivers is estimated at about 135,000 ha, while marshlands occupy around 170,000 ha.There are some 1,385,000 ha potentially arable land.The size of farmland available for agricultural production ranges between 0.25 and 2 ha with an average size of 0.60 ha per agricultural household.Most of the farmlands in Rwanda consist of fragmented plots of land.Approximately 80% of farms have a surface area of less than 1 ha each.
Farming in Rwanda remains largely subsistence in nature.With a rapid increase in Rwandan population from 8 million in 2002 to about 10.5 million in 2011; the pressure on ensuring food security is a constant challenge for the stakeholders.Significant progress has been made in Rwanda in the past decade with regard to over all agricultural production.However, operational efficiency and farm productivity, and, therefore, the prosperity of a very large proportion of the rural population, continue to be a concern.
As the share of service sector on national economy grows larger, the government seeks to transform farming into a productive, high value, market oriented sector by modernizing 50% of its agriculture by 2020, and thereby improve livelihoods of rural population, achieve food security and increase exports of agricultural products as reflected in the Millennium Development Goals (MDG) and New Partnership for Africa"s Development (NEPAD).The effectiveness of the agricultural extension system remains a constraint on further increases in production.In order to sustain the current rate of agricultural sector growth, Ministry of Agriculture and its implementing agency, the Rwanda Agriculture Board, need to successfully implement an extension system that is more effective and accountable to farmers.This paper present the impact of the adoption of self-sustaining extension system in crop productivity, increased revenue for Farmer Field School (FFS) trained farmers and Farmer Promoter (FP) trained farmers compared to non-trained farmers, Changes in applying Good Agricultural Practices (GAP), increased capabilities of farmers due to participation in FFS groups, group dynamics include the formation and functioning of the FP groups and FFS groups and the impact of FFS membership influencing the yield of different crops.
Brief reviews of literature pertaining to Self-Sustaining Agricultural Extension system are discussed in this section as the report of Bertus Wennink et al.(2016).Alston et al. (2000) provide an extensive review of the economic returns to investment in agricultural research and development.The analysis included over 1,128 estimated rates of return, and while 512 of these were for research and extension, only 18 were from extension only investments.The results of the analysis showed an average rate of return of 47 per cent for research and extension investments, while for extension only investments this was 80 per cent.However, as with other reviews, the methodology of the included studies is varied and few follow high quality impact evaluation methodologies.
In the 21st century, agriculture continues to be a fundamental instrument for sustainable development and poverty reduction.Agriculture remains the main source of income for around 2.5 billion people in the developing world (FAO, 2003).A range of approaches to extension delivery have been promoted over the years.Early models focusing on transfer of technology using a "topdown" linear approach were criticized due to the passive role allocated to farmers, as well as the failure to factor in the diversity of the socio-economic and institutional environments facing farmers and ultimately in generating behaviour change (Chambers and Ghildyal, 1984).
According to Anderson and Feder (2003) productivity improvements are only possible when there is a gap between actual and potential productivity.They suggest two types of "gaps" contribute to the productivity differentialthe technology gap and the management gap.Extension can contribute to the reduction of the productivity differential by increasing the speed of technology transfer and by increasing farmers" knowledge and assisting them in improving farm management practices (Birkhaeuser et al., 1991).Additionally, extension services also play an important role in improving the information flow from farmers to scientists.
A number of models have been implemented since the 1970s, combining approaches to outreach services and adult education, including the World Bank"s Training and Visit (T&V) model (Anderson et al., 2006), participatory approaches and most recently farmer field schools (FFSs) (Van den Berg and Jiggins, 2007).
Since the emergence of the Farmer Field School (FFS) approach in Indonesia in the late 1980s, this approach to extension has become increasingly widespread and has been introduced in some 78 countries (Van den Berg and Jiggins, 2007).The FFS approach draws on the participatory approach in terms of its focus on farmer experimentation and problem solving.Van den Berg (2004) provides a synthesis of 25 evaluation studies of integrated pest management (IPM) FFSs.Most studies focused on rice and measured immediate impact of the FFSs in terms of reduced pesticide use and changes in yields, reporting considerable reductions in pesticide use, with some studies also showing an increase in yields.However, in common with other reviews of extension services, the methodology of the studies is varied, highlighting the complexity of estimating impact for such interventions and the lack of an agreed conceptual framework for doing so.The review revealed that studies were either designed to be statistically rigorous, but with limited scope, or comprehensive, but with limited coverage.Van den Berg (2004) argues that by combining the results of different sources the comprehensiveness of the overall evaluation was improved.Building on the latter, Van den Berg and Jiggings (2007) review studies evaluating FFS and pest management, finding that FFSs have had additional benefits to that of IPM including facilitating collective action, leadership, organisation and improved problemsolving skills.Noting that discussions on the fiscal sustainability of FFSs should include considerations of who will pay for the externalities of pesticide use, they conclude that the evidence gathered in the review suggests that FFSs can be a cost-effective way of increasing farmers" skills and thus contributing towards escaping poverty.
Van den Berg and Jiggings (2008) stated that public policy in developing countries has failed to invest in educating farmers on how to deal with variable agroecosystems and a changing world.It presented an assessment of a participatory training approach in changing crop protection by farmers from chemically dependent, to more sustainable practices in line with the tenets of Integrated Pest Management (IPM).The evidence from the studies on an educational investment designed to capacitate farmers to apply IPM, and discussed these data in the light of an on-going policy debate concerning cost effectiveness.The results indicate substantial immediate and developmental benefits of participation in Farmer Field Schools.
Maize (Zea mays), beans (Phaseolus vulgaris) and potato (Solanum tuberosum) are the major crops of the country.Maize and beans are used as a staple food and are the major and most important cereal crop of Rwanda which well adapts to its environmental conditions.The per hectare yield of maize and beans are very low when compared to other countries in the continent like South Africa and Zimbabwe.The low yield in the country is mainly due to drought, mismanagement, small land holding and non-availability of appropriate extension system.Hence, there is a need for implementing the Self Sustaining Extension System in Rwanda.
The effectiveness of the agricultural extension system remains a constraint on further increases in production.In order to sustain the current rate of agricultural sector growth, Ministry of Agriculture and its implementing agency, the Rwanda Agriculture Board, need to successfully implement an extension system that is more effective and accountable to farmers.This paper present the impact of the adoption of self-sustaining extension system in crop productivity, increased revenue for Farmer Field School (FFS) trained farmers and Farmer Promoter (FP) trained farmers compared to non-trained farmers, changes in applying Good Agricultural Practices (GAP), increased capabilities of farmers due to participation in FFS groups, group dynamics including the formation and functioning of the FP groups and FFS groups and the impact of FFS membership influencing the yield of different crops.

MATERIALS AND METHODS
This section explains the general approaches adopted to implement the self-sustaining extension system in Rwanda including the study area.The source of the primary and secondary data collected is from the documents of the Ministry of Agriculture and Animal Resources.The detailed methodology and the methods adopted to bring out the processed information to meet the objectives of the study are also explained.

Location of study area
The study was carried out by collecting the data from the Ministry of Agriculture and Animal Resources in Rwanda.The researcher was a senior officer in charge of organizing, executing and implementing the self-sustainable extension system in the entire country.The various data collected and reports produced under the guidance of the researcher in the ministry form the basis for the analysis of the extension system in Rwanda.

Methodology adopted
The methodology adopted to analyze the self-sustainable extension system consists of developing the institutional development of selfsustaining extension system.The self-sustaining extension system will be based on a pluralistic approach involving farmer to farmer extension model with many actors from both public and private sector playing different roles.Operating within the decentralization system, agriculture committees at village, cell, sector, district, province and national levels ensure that agricultural development agenda is prioritized in overall development agenda.Village is the entry point of self-sustaining extension system in Rwanda as shown in Figure 1 and the organization of Self Sustaining Agricultural Extension systems is shown in Figure 2.

Farmer Field School (FFS)
The FFS plot was the learning place for the members of the FFS group.The FFS Facilitator guided the FFS group members through a process of experimental learning by conducting weekly assessments of the crop growth in various comparative trials.Farmers got deep understanding of crop production in FFS plots and also learnt how to make good decisions based on observations and analysis.FFS groups at the rate of one per village were established across the country (Table 2).

RESULTS AND DISCUSSION
The self-sustaining extension system is a true "homegrown solution" that has been developed and implemented by Rwanda Agriculture Board, under the responsibility of Ministry of Agriculture and Animal Resources, in close collaboration with Districts and Sectors.It is therefore a decentralized system which  gives farmers a key role in agricultural extension.The self-sustaining extension system relies on two extension approaches: the FFS approach and the Farmer Promoter approach.
Farmer Promoters reached all farmers with basic extension messages through mobilization of farmers and demonstration plots in each village.Farmers were organized in groups to serve as extension entry points.Organization of farmers into strong groups enhances farmer to farmer knowledge transfer with a view of making the farmers truly involved in the learning process.Each village identifies one Farmer Promoter through a participatory exercise based on criteria which were developed in a participatory way.The Farmer Promoters also mobilized the farmers to consolidate land, plant in time and buy and use inputs such as improved seed and inorganic fertilizer.The Farmer Promoters supervised the village demonstration plots in which the self-sustaining groups meet three times during the planting season.
FFS Facilitators gradually reached all farmers with indepth knowledge by offering an experimental learning experience in the FFS plot.Farmers were organized in FFS Groups which are facilitated by Facilitators.Each FFS Group had its own experimental learning plot in which the group meets on a weekly basis.The FFS approach builds the skills and capacity of farmers to identify and analyze problems, and to conduct experiments aiming at developing local solutions appropriate to local specific challenges.Based on the principle "Learning by doing" farmers develop their decision-making skills which helps them to handle current and future challenges effectively and to become progressively managers of their farming activities.The self-sustaining extension system builds capacity of the Farmer Promoters to become the first line extension worker in the village while FFS Facilitators are capacitated to be competent facilitators (with strong technical and facilitation skills) to lead FFS Group members through the hands-on learning process.
At the end of 2015, the self-sustaining extension system was implemented by 2,300 FFS Facilitators and 14,200 Farmer Promoters as per details shown in Table 1.
The main role of Rwanda Agriculture Board (RAB) is to provide technical support, especially through the deployment of FFS Master Trainers, as well as other technical staff.The role of the Districts is to ensure that the agricultural extension activities are in line with the development plans of the District.Therefore, the decentralized levels play a crucial role in the planning process as well as in the day to day coordination of selfsustaining extension system activities.

Crop productivity in self-sustaining extension system
The survey was conducted in 80 villages.Six farmers were chosen from every village with the 3 categories (FFS farmers, FP farmers and Non-trained Farmers) giving 6 x 3 x 80 = 1440 farmers, who were randomly selected for studying the average yield.The production of their plots were effectively measured.Data from the Harvest Survey (season 2015B) is shown in Table 2.
Table 2 shows that the FFS participants achieve higher yields than farmer who have been trained by Farmer Promoters and non-trained farmers.It was found that for beans, the highest average yield was 1.25 t/ha for nontrained farmers, 1.52 t/ha for FFS farmers, 1.26 t/ha for FP farmers and the average yield of all the farmers was worked out to be 1.39 t/ha.Similar trend was noted for other crops like cassava, maize, rice, soya and wheat also.Hence, it is established that the FFS farmers perform better than FP farmers and non-trained farmers.Table 2 also shows that on average, FFS farmers produced 37.45% more than non-trained farmers while farmers trained by Farmer Promoters produced 10.78% more than non-trained farmers.However, these averages are strongly influenced by a few villages where very high production increases were noted.

Increased revenue for FFS farmers and FP farmers due to self-sustaining extension
The increased crop productivity for FFS farmers and FP farmers compared to non-trained farmers causes and increased revenue.The computation of increased yield of FFS farmers and FP farmers over non trained farmers is shown in Table 3.It should be noted that these are gross revenues which do not integrate the costs for applying the improved technologies; for instance, the use of fertilizer which often means considerable financial costs for smallholder farmers.
Table 3 reveals the fact that the cassava crop provides additional income of 560 Rwf/ha for the FFS farmers and 330 Rwf/ha for FP farmers.It was also found that rice crop, provides additional income of 172.50 Rwf/ha for the FFS farmers and a marginal negative effect of -12.5 Rwf/ha for FP farmers.This negative trend is not attributed because of training of FP farmers.The rice yield is affected by the season, rainfall, irrigation, weeding and other crop husbandry aspects.The data collected during the harvest Survey (season 2015B) show those farmers who have been trained by either FFS Facilitators or Farmer Promoters obtained higher yields than farmers who did not receive any training from these extension agents.Thus, the self-sustaining extension system used in Rwanda caused increased gross revenues of the agricultural households.The FFS trained farmers and FP trained farmers used new trained skill, additional inputs like employing family labors and buying fertilizers etc. also were the causes of increased yield.The increased income is not only due to training efforts but also due to application of land, family labors,

Changes in applying GAP and differences in yield due to application of GAP
The data collected during the harvest survey (season 2015B) show those farmers who have been trained by either FFS Facilitators or Farmer Promoters used more Good Agricultural Practices (GAP) than the farmers who did not receive any training from these extension agents.GAP are the technologies that, when applied correctly, increase the quantity and the quality of food crop production.They often are to be used in combination with agricultural inputs in order to achieve maximum increase  4. Table 4 shows the fact there is an average of 37.8% of farmers that applied GAP among the non-trained farmers, 73% of FFS farmers used the GAP and 68.3% of the FP farmers adopted the GAPs to get higher yield from their farms.It was found that there was a difference of 4.7% between the FFS farmers and FP farmers.Hence, it was found that there was small difference between the FFS trained farmers and FP trained farmers in self-sustaining extension system used in Rwanda.
Table 4 reveals the fact that the average yield difference between the farmers not applying GAP and applying GAP was 64.63%.It is appreciably very high.It shows the importance of adopting self-sustaining extension system in increased yield.It was found that there was highest yield difference of 122.09% between adopting GAP and non-adopting GAP for the beans crop.It is followed by 70% for maize and the lowest yield difference of 37.31% is for the rice crop.

Increased capabilities of farmers (changes) due to activities initiated by FFS groups
The self-sustaining extension system adopted in Rwanda made many changes like increased capabilities of farmers due to participation in FFS groups.FFS training included the analyzes of the agro-eco system, design and implement experiments, taking decisions as a group to act and to work as group to solve the community problems.The activities initiated by FFS group and changes made are shown in Table 5.
Table 5 shows that there were increased capabilities of farmers due to the activities initiated by FFS groups.Some of the high lights of the changes made in farmers are increased knowledge from the experimental plots, decision making capacities, increased crop production capabilities, improved crop storage, handling and marketing, improved access to agricultural inputs, financial services and extension systems.It improved the relations with other stakeholders, group activities and acquiring knowledge in areas other than farming.

Group dynamics of FFS group and FP groups
Group dynamics include the formation and functioning of the FP groups and FFS groups.The functioning of the groups refers to activities undertaken collectively by group members.Rwanda has two major agricultural seasons.They were season A and season B. Data was collected for season A and B for the years 2015 and 2016.They were named as season 2015A and 2015B for the year 2015 and season 2016A and 2016B for the year 2016.Data on group dynamics are provided through the season 2015B and season 2016A.Table 6 shows the group dynamics of FFS groups and FP groups during the seasons 2015B and 2016A.Table 6 shows that during season 2015B, there was 3.4 average numbers of groups per village and it was 4.2 for 2016A season.It was found that there were 72 farmer members in FP group per village during 2015B season whereas there were 80 farmer members in FP group per village during 2016A season.It was found that there were 22 farmers per FFS Group in season 2015B and 20 farmers per FP group in season 2015B.It was found that there was 52% of women membership in FFS groups.It was found that 68% of the FFS group activities were towards saving and credit development.It was also found that 20% of the FFS group activities were towards various income generating activities.It was found out that both FFS groups and FP groups undertake collective activities pertaining to group savings and credits schemes stand out.Both FFS groups and FP groups undertook collective procurement of agricultural inputs and marketing of products.The most important benefits of being a FFS group member are having more food on the table and members helping each other.The most important benefits of being a FFS group work is that they are able to sell more agricultural produce to the market.Older groups have more group income generating activities and

Membership of FFS Groups and differences in yields
The period of FFS members has influence on the yield of different crops.The FFS members that started from 2009 to 2014 were taken as one entity and the FFS members that started during 2015 was taken as another entity in this study.The average yield difference of FFS members compared to non-trained farmers were worked out in % and are shown in Table 7.
Table 7 shows the fact that the % yield difference of principal food crop of Rwanda is high for the FFS members started since 2009 to 2014 compared to the recently joined FFS member.The recent FFS members since 2015 recorded lesser yields because they acquired lesser knowledge, understanding and practices of GAP.The table shows that the soya crop yield for older FFS members was 133.6% whereas it is 56.9% for members that joined in 2015.The differences in yield between older FFS members and recent FFS members may be due to the facts like education levels and experience between older FFS members and recent FFS members.Similarly for maize crop also, there was wider gap between the yield of older FFS members (114.7%)compared to recent FFS member (26.2%) due to differences in knowledge and experience of the members in older and recent FFS members.

Access to extension services
A household survey was conducted during 2012 and 2015.Data was collected about the access to extension services by different randomly selected households.It included FFS group members, FP group members and not trained farmers households.The result of the household survey for accessing the benefit of extension services is shown in Figure 3.
Figures 3 and 4 show the household survey 2012 and 31.7% of the households are accessing the benefits of self-sustaining extension system where as in 2015, 31.0% of the households are accessing the benefits of extension services.There is no appreciable difference between 2012 and 2015.It is also found that 68.30% of households during 2012 and 69% of households during 2015 are not accessing the extension services.Hence, there is a need to increase the FFS and FP members for better extension.The number of trainings, quality of trainings, field visit and sharing the experiences of demonstrations farms has to be included for increased benefits to households by accessing the self-sustaining extension system..These efforts were less between the survey periods of 2012 and 2015.Hence, there is no appreciable difference between the periods of 2012 and 2015 in accessing the benefits of self-sustaining extension system.

Conclusion
The self-sustaining agricultural extension system implemented in Rwanda has two pillars FFS and FP groups.These groups are spreading the improved agricultural technologies from one to another through supply of inputs and field demonstrations.The summary of the present research is given below: 1) It was found that at the end of 2015, the self-sustaining extension system was implemented by 2,300 FFS  Facilitators and 14,200 Farmer Promoters.FFS facilitators comprised of 72% male and 28% female.It was found that 92% of the trained FFS facilitators were active.It was found that 95% of the FFS facilitators are members of cooperatives.The FP groups comprised of 80% male and 20% female.It was found that 62% of the 2015 in Rwanda.It shows the fact that during 2012, Farmer promoters were active in 2013.It was found that 25% of FP is active in 2014 and it is decreased to 13% active during 2015.Hence, the activeness of the Farmer Promoters is decreasing year after year of starting the FP groups.
2) It was found that the FFS participants achieve higher yields than other farmers.It was found that for beans, the highest average yield was 1.25 t/ha for non-tr.52t/ha for FFS farmers, 1.26 t/ha for FP farmers and the average yield of all the farmers was worked out to be 1.39t/ha.Similar trend is noted for other crops like cassava, maize, rice, soya and wheat also.Hence, it is concluded that the FFS farmers perform better than FP farmers and nontrained farmers.It was found that on an average, FFS farmers produced 37.45% more than non-trained farmers while farmers trained by Farmer Promoters produced 10.78% more than non-trained farmers.
3) It was found that the cassava crop provides additional income of 560 Rwf/ha for the FFS farmers and 330 Rwf/ha for FP farmers.It was also found that rice crop, provideD additional income of 172.50 Rwf/ha for the FFS farmers and a marginal negative effect of -12.5 Rwf/ha for FP farmers.This negative was not because of training of FP farmers.The rice yield was affected by the season, rainfall, irrigation, weeding and other crop husbandry aspects.Thus, it was concluded that the selfsustaining extension system used in Rwanda caused the increased gross revenues of the agricultural households.The increased income is not only due to training efforts but also due to application of land, family labors, livestock, financial capital used to purchase of inputs like fertilizers and pesticides applied.It was found that an average of 37.8% of farmers applied Good Agricultural Practices (GAP) among the non-trained farmers, 73% of FFS farmers used the GAP and 68.3% of the FP farmers adopted the GAPs to get higher yield from their farms.It was found that there was a difference of 4.7% between the FFS farmers and FP farmers.Hence, it is concluded that there is small difference between the FFS trained farmers and FP trained farmers in self-sustaining extension system used in Rwanda.
4) It was found that there were increased capabilities of farmers due to the activities initiated by FFS groups.Some of the high lights of the changes made in farmers are increased knowledge from the experimental plots, decision making capacities, increased crop production capabilities, improved crop storage, handling and marketing, improved access to agricultural inputs, financial services and extension systems.The selfsustaining extension system improved the relations with other stakeholders, group activities and acquiring knowledge in areas other than farming.
5) It was found that there is 52% of women membership in FFS groups.It was found that 68% of the FFS group activities were towards saving and credit development.It was also found that 20% of the FFS group activities were towards various income generating activities.It was found out that both FFS groups and FP groups undertook collective activities pertaining to group savings and credits schemes stand out.Both FFS groups and FP groups undertake collective procurement of agricultural inputs and marketing of products.The most important benefits of being a FFS group member are having more food on the table and members helping each other.The most important benefit of being a FFS group work is that members were able to sell more agricultural produce to the market.Older groups have more group income generating activities and often they have become a formal cooperative.
6) It was found that the percentage yield difference of principal food crop of Rwanda is high for the FFS members started since 2009 to 2014 compared to the recently joined FFS member.The recent FFS members since 2015 recorded lesser yields because they acquired lesser knowledge, understanding and practices of GAP.It was found that the soya crop yield for older FFS members was 133.6% whereas it was 56.9% for members that joined in 2015.Similarly for maize crop also, there was wider gap between the yields of older FFS members (114.7%)compared to recent FFS member (26.2%).The differences in yield between older FFS members and recent FFS members may be due to the facts like education levels and experience between older FFS members and recent FFS members.It indicates that the self-sustaining extension system works well in Rwanda.
7) It was found from the household survey conducted during 2012 and 2015 in Rwanda, that 31.7% and 31.0% of the households accessed the benefits of selfsustaining extension system respectively.There is no appreciable difference of accessing benefits of extension system between 2012 and 2015.It was also found that 68.30% of households during 2012 and 69% of households during 2015 did not access the extension services.Hence, there is a need to increase the FFS and FP members for better extension.There is a need to increase the number of trainings, quality of trainings, field visit and share the experiences of demonstrations farms for increased benefits to households by accessing the self-sustaining extension system.This research concludes that there was appreciable improvements in the spheres of crop productivity, increased revenue, applying GAP, differences in yield due to application of GAP, increased capabilities of farmers due to improved activities, group dynamics, membership of groups and access to extension services because of implementation of self-sustaining extension system.The most important implication studied from the study are 1) to continue the self-sustaining extension system in Rwanda for increased crop production, 2) to increase the knowledge base and capacity of farmers in Musabyimana et al. 185 crop productivity and 3) better organization of farming communities to meet the climate change and market dynamics etc.

Figure 1 .
Figure 1.Village as the entry point of Self Sustaining Extension System.

Figure 2 .
Figure 2. Organization of Self Sustaining Agricultural Extension systems.

Table 1 .
Numbers and characteristics of FFS Facilitators and Farmer Promoters.

Table 2 .
Average yields (t/ha) for FFS farmers, FP farmers and non-trained farmers.

Table 3 .
Increased revenue for FFS farmers and FP farmers compared to non-trained farmers.

Table 4 .
Percentage of farmers applying GAP and differences in yield between farmers applying and not applying GAP.

Table 5 .
Activities initiated by FFS group and changes made.

Table 6 .
Group dynamics of FFS groups and FP groups during the seasons 2015B and 2016A.

Table 7 .
Membership of FFS Groups and differences (%) in yields.