2

PARTICIPATORY FARM-LEVEL
INNOVATION

Tilahun Amede, Charles Lyamchai, Girma Hailu,
Bekele Kassa, Leulseged Begashaw, Juma Wickama,
Adugna Wakjira, and Gebremedhin Woldegiorgis

Context and rationale

Improved natural resource management in the densely settled highlands of eastern Africa must begin at the farm level, as this is the basic decision unit and the locus of key inputs (land, labor, capital) which have a defining role in agricultural production systems. It is also here where concrete livelihood improvements can be made most directly, thus enhancing the likelihood that farmers will invest in natural resource management (at farm and landscape scale) with slower returns, or wider landscape governance initiatives (covered in Chapter 4) with more diffuse or uncertain returns.

Throughout the centuries, farmers have on their own initiative devised, developed, adopted, and adapted ingenious strategies and technologies for ensuring food security and economic welfare for their households (O'Neil, 1995). Farmer innovations such as the plough and the domestication of plants and animals that revolutionized agriculture date back more than 10,000 years. Farmer innovation is a process through which individuals or groups discover or develop new and better ways of managing available resources to suit their particular conditions. The resulting innovations or outcomes of the innovation process may be technical or socio-institutional (Waters-Bayer and Bayer, 2009). Innovation also occurs within a wider socio-economic and institutional context which conditions the extent and directionality of innovation processes, including the interaction among individuals or groups, policies and norms, and institutional and societal cultures. Innovative farmers are those who have tried or are trying out new, often value-adding agricultural or NRM practices, using their own knowledge and wisdom while also appropriating outsiders’ knowledge (Assefa and Fenta, 2006). For instance, improving crop varieties through careful selection of seed, harvesting rainwater from roads, and implementation soil and water conservation measures, among others, are often carried out in the absence of any outside facilitation or support. Traditional irrigation systems (e.g., of the Chagga and Sonjo in Tanzania and Qantas in Iran), local knowledge on weather forecasting, biological control technologies, production of new pesticide concoctions, use of different plants and roots for soil fertility improvement, and local cures for different animal and human ailments are some of the well-documented farmer innovations (Lyamchai et al., 2006). These innovations have clearly played a significant role in the improvement of rural livelihoods and will continue to do so in the future.

Yet while local knowledge is essential to household food security, income generation and risk management vis-à-vis market fluctuations and an uncertain climate, formal research can play an instrumental role in supporting farmer innovation. This is either because certain techniques are beyond farmers’ reach, as in the case of complex crop and livestock breeding technologies, or because researchers have access to a wider range of information and technologies that can assist farmers in capturing new opportunities or coping with a fast pace of change. As stated by O'Neil (1995: 1):

Farmers are the ultimate integrators of the information they receive to increase production, stabilize yields, use pesticides, etc. It is the farmer that “lives the problem,” gains the benefits and suffers the consequences. Therefore, a combination of farmers’ and scientific knowledge will increase the rate of success and identify new areas of effort that neither group alone would have discovered.

Characteristics of crop–livestock systems in the eastern African highlands

The highlands of eastern Africa, with an average altitude of 1,500 metres above sea level, occupy 23 percent of the total land area but support 80 percent of the population (Alumira and Owiti, 2000). Mixed crop–livestock systems predominate, where 70 percent of the total human population and approximately 80 percent of the cattle and small ruminants can be found (Thornton et al., 2002). These systems are predominantly small-scale and highly diversified, combining annual and perennial crops with livestock. The area receives relatively high annual rainfall (>1,000 mm), the soils are generally more productive than the adjacent lowlands and in some countries irrigation water is also prevalent. Given this endowment, the highlands have been a major locus of human settlement historically, as well as a source of food and nutritional security within the wider subregion. These areas also produce important export crops such as tea, coffee, khat, and other horticultural crops that contribute to hard currency earnings. Livestock is a major component of these systems and makes significant contributions to food production, income, and social security. The importance of livestock is most apparent in countries such as Ethiopia and Sudan, where crop production depends heavily on animal traction and nutrient recycling between crop and livestock components and livestock plays a critical role in expanding the area under cultivation and enhancing labor productivity. It also improves environmental processes such as the turnover of nutrients and soil carbon (van Keulen and Schiere, 2004).

In these crop–livestock systems, livestock and crops are produced within the same farm unit, whereby the by-products of one enterprise serve as a valuable resource for the other (Tarawali et al., 2004). Accordingly, the crop and the livestock components are strongly integrated through livestock feed, nutrient cycling (via manure), draught power, and input–output markets. The crop– livestock systems in the region are not homogeneous but vary from place to place depending on rainfall, soil fertility, population density, socio-economic characteristics, and access to capital and markets, among others. They are also at different levels of intensification, integration, and productivity. The production systems vary from the perennial banana–coffee gardens in the mid highlands of Uganda to the maize–beans systems of western Kenya and the more temperate barley-based systems of Ethiopia. These mixed farming systems have also evolved over time in response to changes in relative access to land, labor, and capital (van Keulen and Schiere, 2004).

These low-input systems are heavily reliant on the recycling of internal and organic nutrient resources and rainfall. Annual food production and availability in the region varies widely according to the seasonal climate, with the number of food-insecure people increasing significantly in seasons characterized by an uneven distribution and/or shortage of rainfall. Meanwhile, in good years not only food production but also national economies recover rapidly.

Yet while these highland areas are relatively rich in natural resources, livelihoods in the region are negatively affected by the following system constraints, which together with climate change limit people's coping and adaptive capacity:

•    High and growing population density, leading to small landholdings, high degrees of fragmentation of landholdings and land tenancy regimes limiting systems intensification.

•    Declining crop and livestock yields, owing in large part to declines in soil fertility, limited access to improved seeds and breeds, and increasing incidence of disease.

•    Limited access to reliable markets, thus discouraging farmers from intensifying their systems and investing in their land.

•    Limited capacity to develop water resources for multiple uses, including irrigation.

•    The erosion of local genetic diversity and management systems, leading to declining ability of crop and livestock enterprises to resist climatic and disease-related shocks.

•    Widespread poverty, with limited investment in yield- and value-enhancing technologies and practices.

•    Limited and/or top-down extension services that do not reflect socio-economic realities, and limited institutional and policy support for enhancing access to inputs, credit and markets.

The high human and livestock population densities, land shortage, and steep slopes jointly contribute to resource degradation through overgrazing, nutrient mining, soil erosion, and water depletion. The outcome of this is that in some countries such as Ethiopia, a quarter of the crop–livestock systems are seriously eroded—of which approximately 15 percent are so seriously affected that it will be difficult to make them economically productive in the near future (Amede, 2003). Also in Ethiopia, three out of the five principal farmlevel problems listed by farmers to be critically affecting farm productivity were the loss of seed and fertilizer from excess run-off, soil erosion, and increasing cost of fertilizers owing to soil fertility decline (Amede et al., 2006). Above all, nutrient depletion is a much more serious concern to food security in sub-Saharan Africa than in any other part of the world (Smaling, 1993).

Degradation of natural resources in the region is therefore partly a consequence of rural poverty, resulting from the interplay between rising population density, shrinking landholdings and livestock numbers, unreliable markets, and weak institutional and policy support in responding to emerging challenges. Livelihood strategies and assets management at farm level are also changing or need to change as families repeatedly face food deficits, livestock deaths, and degradation of the resource base.

Intensification of crop–livestock systems

Intensification is one option for fulfilling the growing demand for food, feed, and energy in the region. Intensification is a process to increase production levels, both in terms of amount, quality and fulfillment of local priorities and preferences (see also Morrison, 1994). In simple terms, it has been defined as an increase in average inputs of labor or capital on smallholdings, either on cultivated or uncultivated land, for the purpose of increasing the value of output per land area (Tiffen, 2003). Reardon et al. (1999) distinguish between capital-led and labor-led intensification. While labor-led intensification involves excessive dependence on labor as a key input to production, capital-led intensification refers to intensification based on substantial use of non-labor inputs such as chemical fertilizers and herbicides that enhance the productivity of land resources. Intensification has also been facilitated or induced by exogenous factors including population pressure, land shortage, and increasing labor availability. It can also be induced by external factors including increasing demand for livestock products and improved access to markets. In some cases, farmers are forced to shift from extensive cereal-based to intensive crop–livestock systems. The recent shift in policy towards market-led agriculture in Ethiopia, for instance, has influenced the way farmers are managing their resources. Introduction of small-scale irrigation into the cereal-based systems has induced farmers to diversify their cropping systems by growing high-value vegetables, fruits, coffee, and other crops. It has also led to the intensification of production systems through increased investment in high-yielding seeds and livestock breeds, chemical fertilizers, pesticides, and soil conservation practices. In these cases, both diversification and intensification are outcomes of farmers’ responses to market opportunities while simultaneously striving to satisfy household food demands. Similarly, in places where dairy enterprises are becoming a major source of household income, such as in central Kenya, farmers have integrated fast-growing forage grasses (e.g., Napier grass) and fodder shrubs (e.g., Calliandra) in an effort to intensify the dairy component of their maize–beans–dairy systems.

As the intensification of agricultural production commonly requires external inputs, small-scale farmers in Africa tend to rely on government support in the form of loans and subsidies to finance external inputs.¹ Access to credit is one of the three pre-conditions identified by Reardon et al. (1999) for sustainable agricultural intensification in Africa, along with the availability of productive labor and high returns to investment resulting from accessible input and output markets. For the majority of households intensifying their systems based on scarce endogenous resources, the process is much slower. Kelly et al. (1996) suggest that in order to obtain sustainable intensification of agriculture in Africa, agricultural policies must also consider:

•    improved access to quantity and quality seed/breeds;

•    improved strategies in restoring soil fertility;

•    functional land tenure policy; and

•    increasing rural cash income and investments to improve food security and input access.

From the social welfare-based perspective of AHI, agricultural intensification should be a means to improved livelihoods and household income of rural communities without degrading the natural resource base (water, nutrients, vegetation), irrespective of its manifestations.

This chapter summarizes AHI experiences in supporting farm-level innovation for improved livelihoods and more sustainable management of natural resources underpinning agricultural production. While its focus is on the farm level and individual households as decision units, it also describes group initiatives that supported, in one way or another, farm-level innovations. The chapter presents the methods used and the main approaches developed to support farmer-led system intensification and diversification as pathways to achieve farmers’ own livelihood and resource management goals. It displays strategies used to minimize resource degradation, and highlights the impact they have had on people's livelihoods and production systems. By providing detailed steps in each thematic section, it aims to facilitate the dissemination and adoption of the various approaches used by AHI and its partners within the wider eco-region.

Understanding systems and clients

The crop–livestock systems within this region differ in their agro-ecological and socio-economic features and in the wider policy, economic, and institutional contexts in which they are embedded. Efforts to understand the farming systems and farmers’ needs and priorities are therefore required as a first step to any innovation process.

Approach development

Approach 1—Understanding systems and clients through participatory approaches

The first approach, used by all site teams, was to conduct a participatory appraisal of system problems and possible solutions as a means of setting the participatory research process in motion. By attempting to understand a system from the perspectives of farmers themselves, the diagnostic process in effect integrates the deep knowledge farmers bring with them on the constraints they face and the pathways most likely to unlock the potential for change.

Following a preliminary reconnaissance survey for the identification of research sites, a general meeting was called by local government officials, involving farmers from the area and researchers. This meeting consisted of the following steps:

1.    An introduction to the purpose of the meeting and a clarification of expectations.

2.    The division of farmers into subgroups based on different system components (crops, livestock, soil, and socio-economics). Using participatory rural appraisal (PRA) techniques, farmers identified constraints they faced—often identifying more than 20 per group, both researchable and non-researchable.

3.    Return to plenary, where farmers ranked and prioritized the identified constraints.

4.    Key informants were then asked to group farmers into different wealth and gender groups. The important constraints identified by the general community, along with potential solutions proposed, were presented to these subgroups to solicit their reactions. Practical solutions were discussed with these groups and final decisions were made on what potential solutions they would like to pursue.

5.    Proposals were then developed collaboratively by groups of researchers and farmers, who were organized into small working groups according to sub-themes derived from the priorities set by farmers. These farmer groups critically discussed the various proposals put forward by farmers and researchers, and farmers were consulted on methodological issues and possible treatments to be employed in the design of on-farm trials. During proposal formulation, explicit attention was given to wealth and gender considerations.

The major system constraints identified in step 4 included soil erosion, soil fertility maintenance, livestock feed, lack of credit, and limited access to improved varieties of various crops (Amede et al., 2006). The aforementioned steps show how farmers were involved in problem identification, priority setting, the identification of solutions, and technology evaluation. Subsequent steps of engagement consisted of hands-on training, implementation of onfarm trials, evaluation, and replanning—with regular contact with technical assistants (hired or seconded by the project, and present on a daily basis) and researchers to discuss on-farm challenges and to evaluate performance. These steps will be discussed in greater detail in the sections that follow.

Approach 2—Understanding systems through optimization models

Models can also play a useful role in understanding systems and enhancing the capacity of farm managers and development actors to make strategic decisions. In AHI, models were employed to predict the short-term risks and long-term impacts of agricultural interventions. They also provided a mechanism for evaluating the vulnerability of livelihoods and assets to external factors, targeting potential technologies to clients, and extrapolating and synthesizing knowledge for wider use.

One modeling approach was designed to enhance household food security in the context of limited resources such as land, water, nutrients, and labor (Amede et al., 2004). The aim of the modeling process was to select the best crop combinations to produce the required amount and quality of food for rural households through the reallocation of cropland to expand the area under crops with high content of nutrients in deficit, considering resource availability and local preferences. Stated in another way, the model helps communities and households to maximize the returns from investments of fertilizer, labor, land, water, manure, and other resources by helping them to optimize investments to enhance system productivity and household nutrition.

Use of a participatory, multiple-criteria decision model was considered to be particularly suited to assessing appropriate resource allocation strategies, considering local resources, socio-economic preferences, and market options.² Participatory steps in the modeling process helped to define key production objectives of farmers and to define potential pathways to achieving these—in recognition that there are many alternative ways to maximize farm productivity. By considering the production objectives and socio-economic preferences of farmers together with the available resources, the modeling process aids in exploring what is possible and—with the help of farmers—feasible for achieving established objectives.

When employing the model to optimize land resources for food and nutritional security, the modeling process consisted of the following steps (see also Box 2.1):

1.    Identify representative households to participate in modeling, based on locally-defined social categories. A community meeting was organized to identify households keen to gain a deeper understanding of their resources and design cropping strategies for improved income, food security, and environmental protection. Based on these objectives, community members were asked to classify themselves into different social groups based on locally defined criteria. These could relate to farming system characteristics, the location and features of landholdings or resource endowments. The established groupings were then facilitated to identify their major production constraints, to discuss the causes of recurrent food insecurity and to suggest potential and practical solutions based on local resource endowments.

2.    Quantify household and farm resources. Household resource inventories were carried out during the growing season. Researchers recorded farm size, distribution of land for crop production and grazing, type of crops grown, amount of land allocated for each crop, frequency of cropping, grain yield, and crop residues for each participating household. Household data were also obtained from women related to household demographics (including household members, resident guests), amount and type of food consumed per day and other relevant data. Community leaders played an important role in cross-checking household information.

3.    Quantify the amount and distribution of resources. Once yields of the various crops and livestock were established, and annual household production levels estimated, additional household resources such as labor, nutrients, cash, and other assets were assessed.

4.    Compare household resource endowments with established or calculated norms to identify levels of vulnerability and nutrients in deficit or excess. In this case, community averages, resource holdings of a representative household or internationally established norms (e.g., for household nutrition, the recommended daily dietary nutritional allowance of the World Health Organization (WHO)) were employed to establish optimal levels of resources (e.g., amount of nutrition required per person per day). This was used to identify households vulnerable to malnutrition, famine, or drought; to identify factors involved (e.g., labor shortage); and to identify nutrients that are in excess or deficit in the system.

5.    Use optimization models to suggest an improved resource allocation strategy. Optimization and trade-off models were then used to optimize scarce resources, with provisions for placing constraints on certain parameters based on household objectives or constraints. It was used, for example, to discuss the cropping and resource (e.g., labor, nutrient) allocation implications for maximizing different household production objectives such as human nutrition or cash income.

6.    Validate model outputs with communities. Commonly, model outcomes represent optimal solutions, but do not represent realistic solutions unless they are validated by end-users and modified to match socio-economic realities. This process was used to update farmer preferences, such as a desire to maintain some enterprises and get rid of others—which were then newly incorporated as model parameters. In addition to discussions with individual households on the implications of different land and resource allocation decisions, this step should include an intensive and iterative process of community visioning. This is useful for deepening researchers’ understanding of farmers’ decision frames; for raising awareness among farmers on mechanisms to improved income and food security, reduce resource degradation, and reduce vulnerability to famine; and for articulating the agronomic management implications of desired future conditions.

7.    Establish potential trade-offs of different farming system innovations. Tradeoff analysis could be done either as an integral component of optimization models or as ex ante analysis once optimum choices for the intended objective function are made. In the mixed crop–livestock systems of the Ethiopian highlands, farmers considered both the crop and livestock sub-systems when deciding to integrate new interventions such as crop varieties into their farms. Hence, there is generally a need to establish how changes in one or more components affect other system components (Box 2.1). In other words, there is a need to quantify how possible changes in one enterprise affect the performance of other enterprises, with all other factors remaining constant.

This modeling process was found to be a powerful tool not only for helping farmers better understand and manage their farming systems, but also for enhancing collaboration between farming households and researchers by deepening mutual understanding.

Lessons learned

The following lessons were learned from AHI efforts to understand farming systems and clients in Phase II:

•    Efforts to understand farmers’ socio-economic realities from the start of any innovation process strengthen researcher–farmer communication and, ultimately, increase the chance of technology adoption and impact.

•    Feedback mechanisms between researchers and farmers and timely responses from research are critical for identifying solutions likely to be effective in addressing the challenges faced by agricultural systems and clients.

Building rapport and farmer confidence to innovate

One of the first steps in any farmer innovation process supported by outside actors and institutions is to build rapport between farmers and service providers (research, extension, NGOs) and to boost farmers’ confidence in their ability to solve their own problems. Although research in natural resource management needs to take a holistic view as well as acknowledge the complexities and diversity of farming systems, research with farmers should also address critical problems that they have identified and prioritized (Amede et al., 2001). While a range of options were available to address the wide array of constraints identified by farmers in the diagnostic phase, it was important that those first tested would advance these aims, so as to encourage farmer engagement in more complex endeavors later on. This section therefore focuses on the use of “entry points”—defined here as an initial action that is strategically applied to enhance the likelihood of success of early innovations, and thus to build rapport between actors jointly engaged in an innovation process.

The choice of entry points has been proven to have a significant effect on whether farmers will be keen to invest in a partnership with researchers and extension agents for the purpose of experimentation, and whether farmers will continue to innovate in solving their problems without the support of external actors (Amede et al., 2001; Amede et al., 2006). Entry points can be an intervention in the form of an attractive technology or incentive. Entry points are essential to build trust between the community and outside actors, arouse their interest and keep their spirits high as the innovation process evolves—despite ever-complex challenges that may be tackled.

Approach development

Entry points utilized by AHI were commonly crop varieties, which could be identified on the basis of key constraints identified by farmers (e.g. market requirements), tested and disseminated rapidly. In some sites, varieties of high-value vegetable crops were used as entry points while in others fast growing forage grasses, such as napier, were used—based on farmer preferences. Researchers involved in AHI used “entry points” as a strategy to quickly get engaged with the farmers by providing some “best bet” technical solutions to priority problems (Wickama and Mowo, 2001).

Key steps in the process included the following:

1.    Identification of the constraints faced by different wealth and gender groups, along with potential solutions, as discussed above.

2.    Identification of criteria for selecting entry points. These often included:

a)    of high priority, addressing felt needs of intended beneficiaries;

b)    capable of bringing quick benefits (often economic in nature); and

c)    low risk (e.g., involving limited cost or having been tested and validated in similar agro-ecological zones previously).

3.    Generate a basket of options for farmers by matching specific technologies (sourced from research stations, extension agencies, sites with similar agroecologies, innovative farmers within the benchmark site) to farmers’ felt needs and goals.

4.    Facilitate the formation of “interest groups” for technology testing. Researchers facilitated the formation of thematically-based ‘interest groups’ based on technology preferences, to work as a group in carrying out a number of experiments on a particular theme.

5.    Participatory testing of a wide range of technologies by farmer interest groups. Based on group plans, group members implement experiments assigned to them. Group members periodically visit each others’ experiments, monitor performance, share information (e.g., on yield, observed characteristics or performance of technological options) and disseminate popular technologies to other group members.

6.    Facilitate cross-group sharing of popular technologies likely to be of interest to the wider community.

In some cases, as a result of proper selection and implementation of entry points, AHI and its partners managed to reach more than 75 percent of farmers with income-enhancing technologies (Box 2.2). Entry points actually adopted by farmers were also found to vary according to their social status and agro-ecologies, evidencing the need for effective targeting and participatory selection of entry points to be tested by different households. A detailed analysis of selection preferences in the Areka benchmark site, for example, showed that resource-rich farmers with fertile plots and many livestock (and ample manure) preferred high yielding crop varieties, while resource-poor farmers with degraded land and limited access to manure preferred interventions contributing to soil fertility improvement as entry points (Amede et al., 2006). Examples of successful entry points with win–win effects from AHI sites are presented in Box 2.3.

Lessons learned

The following lessons were learned from AHI's experiments with entry points:

•    Interventions that bring immediate and visible benefits to farmers and their families are essential within any INRM initiative, as they build farmers’ confidence in their ability to solve their own problems and help to build trust and rapport between farmers and support services. This is particularly true when the entry point addresses multiple concerns simultaneously (at least one of which brings quick returns).

•    Characteristics of good entry points include their high priority for farmers, their ability to address concrete problems of local concern, their ability to generate quick benefits—particularly income, their simplicity for and accessibility to a wide range of households (so that unequal benefits do not compromise the enthusiasm of large portions of the community), their high chance of success (as early successes go a long way to enhance enthusiasm and trust) and the ease with which they can be managed and multiplied.

•    To maximize impact, entry points need to be matched to household preferences and constraints, as well as to local agro-ecological and marketing conditions. For example for teff, the staple crop in Ethiopia, women's major selection criterion was color (white grain fetches more money than red, and is preferred for cooking the local bread enjera), while men considered yield and resistance to lodging as the most important criteria. Meanwhile, in Tanzania farmers preferred high yielding and firm tomato varieties with long shelf life and that can withstand transportation because better markets are more than 300km away.

•    Owing to the simplicity and low-cost, low-risk nature of entry point technologies, they can often be effective in reaching less advantaged social groups.

•    To maximize the contribution of entry points to addressing more complex system or NRM challenges, it is important to consider entry points that can enhance the subsequent adoption of other NRM technologies.

•    Where benefits, especially monetary benefits, accrued from entry point technologies, farmers are often more willing to engage in more complex and integrated technological innovations.

•    Initially unaware of the potential benefits of a lasting partnership with researchers, farmers may initially come to the innovation process with expectations of quick rewards such as fertilizers and seeds. With a lack of experience working with research and understanding its value, they may be unprepared to take risks associated with adopting complex technologies and practices. Finding means to respond to their immediate demands without creating dependency while working on more complex innovations with slower returns can go a long way in fostering interest in the latter and in moving towards more sustainable farming practices.

Supporting farmer innovation to address farm-level constraints

There is general agreement in the agricultural research and development community in the region that low agricultural productivity and resource degradation in Africa are not owing to the absence of technologies, but to the limited adoption, adaptation, and dissemination capacity of farmers and the ineffectiveness of methodologies employed to support these processes. Farmer experimentation and innovation are recognized as essential in efforts to improve productivity and reverse natural resource degradation. This innovation is not only technological in orientation, but may also encompass networking and communication, the strengthening of local institutions, planning and monitoring, or accessing resources and marketing—anything that may be considered “new ways of doing business” (Assefa and Fenta, 2006).

While diverse approaches were employed in supporting farmer experimentation in AHI, some elements were common to all. For example, different actors tended to make different types of decisions. Decisions on the location of trials, choice of crops, and harvesting time are usually made by farmers. Experimental design and implementation were carried out jointly, but researchers had a strong input into the basic research design to ensure adequate replication and controls. Researchers also participated in identifying parameters to be tested and in carrying out the analysis, but farmers were involved in managing experiments and evaluating technologies. Officers of local offices of agricultural ministries were also involved in the decision-making process, given their familiarity with wider areas over which technologies could be applicable. Farmers generally tried to address their specific problems by testing a wide range of technological options selected in response to the problems they face. In certain cases, researchers also assisted client farmers in resource mobilization, leadership, building organization skills, group management, and conflict resolution.

Approach development

AHI experimented with at least three different approaches to supporting farmer innovation: local testing and adaptation of the farmer field school (FFS) approach, approaches for inducing innovation based on local knowledge, and approaches for linking complementary technologies to achieve synergies between livelihood and natural resource improvements.

Approach 1—Farmer field schools

The FFS approach is an innovative, participatory, and interactive learning approach developed by FAO in the 1980s to address pest and disease problems of rice farmers in Southeast Asia (Pontius et al., 2002). It builds farmers’ capacity to understand their systems, identify system constraints, and to test and adopt technologies and practices matched to those constraints. With more than 1,000 FFS active in Kenya, the FFS approach is not new to the eastern Africa region. The FFS approach employs non-formal adult education methods, particularly experiential learning techniques. Farmers are selected based on their interest and willingness to follow the proposed methodology or action plan and their commitment to invest and allocate their time in the program. Typically, a group of 20 to 25 neighboring farmers meets regularly, commonly once a week, on one of the farmers’ fields during the entire experimental cycle. The school is not meant to introduce farmers to new technologies developed outside their environment, but to provide them with tools and methods that will enable them to analyze their own production practices and identify possible solutions.

The AHI team and its partners used FFS in selected cases where there was a need for intensive interaction between researchers and farmers. It was applied particularly when a farm constraint demanded a comprehensive package of knowledge, practices, and technologies, such as controlling crop diseases. It was also tested for its relevance to addressing complex natural resource management challenges and community organizing. Communities were empowered to establish FFS to organize, test, adopt, and disseminate improved technologies and practices. They were also facilitated to sustain the learning process by building the capacity of colleague farmers and communities to enable them to respond to emerging local challenges. The approach was used to build local capacity and interest in sustaining farmer experimentation on their own, even in the absence of external material and technical support.

The following steps were involved in adapting the FFS approach to achieve the program aims:

1.    Facilitate a dialogue among farmers and with outside agencies (research, extension) to enhance local awareness and refresh people's memory about key system constraints identified in the diagnostic phase (e.g., through PRAs) and their implications for food security, income, and natural resources.

2.    Plan and facilitate discussions together with local institutions on how to solve these system constraints using local solutions, skills and collective action.

3.    Identify farmers with similar farm-level problems or constraints and a shared interest to find practical solutions, and assist them in organizing themselves into thematic groups (generally consisting of a commodity and related NRM innovations) to identify and test endogenous and exogenous innovations.

4.    Organize and conduct a formal, classroom-based training program to help farmers to analyze the biological and socio-economic causes of the problem. This included the development of a detailed theoretical and practical curriculum to enable farmers to understand the causes and develop skills to solve the given constraint. In some cases, “classroom” learning was supported by laboratory experiments, field days, and other practical methods.

5.    Identify a basket of technological options based on local economic and social criteria and introduce them to farmers through cross-site visits, exposure to on-station trials, or other means.

6.    Conduct formal trainings to equip farmers with the skills needed to enable them to successfully compare options, including formal experimental methods.

7.    Organize farmers and support their efforts to test, adapt, and adopt the interventions in their own fields, assisting them to capture data on key parameters such as yield, income, and labor and to compare the performance of different enterprises and management options.

Box 2.4 describes a case where FFS were used to solve farm-level constraints.

Approach 2—Farmer experimentation using local knowledge as a starting point

Efforts to support farmer innovation to address farm-level constraints must begin with local innovation processes, a critical starting point for building partnerships of mutual respect between different actors in an innovation system. It starts with looking at what farmers are already trying to do to solve problems or grasp opportunities they have identified (Waters-bayer and Bayer, 2009).

Acknowledging this reality, the AHI research team in collaboration with NARIs and international research organizations employed several participatory techniques for integrating local knowledge into agricultural experimentation. The aims were to: 1) develop strategies to address complex NRM issues; 2) foster a change from a commodity orientation to a more holistic systems and participatory approach in the research system; and 3) develop and improve technologies and approaches that could be used by policy makers, development actors, and farmers to address identified NRM challenges. Farmers were in the forefront throughout the processes of technology development, technology dissemination, and impact assessment—a process that included the following basic steps:

1.    Participatory identification of problems and opportunities from the standpoint of farmers and researchers;

2.    Characterization of various local innovations employed by different farmers, how widespread they are and their potential benefits;

3.    Scientific validation of local knowledge or innovations to better understand their features and benefits, and to explore how to link them to scientific knowledge in addressing system constraints;

4.    Feedback of findings and discussion with the holders of local knowledge and other community members;

5.    Demonstration and experimentation of ways to link local innovations with exogenous technologies (an optional step, employed only where possible synergies are identified); and

6.    Promotion of best performing innovations as integrated packages to the wider community through training and awareness creation.

Box 2.5 presents experiences from Areka, southern Ethiopia where farmers employed local knowledge to control mole rats, a vertebrate pest causing yield reductions of up to 60 percent in root crops. Conventional methods of controlling the pest (e.g., use of poisonous substances) had proven to be ineffective in addressing the problem. In addition to the expense, the mole rats quickly learn to dodge them once they detect they are poisonous. Fortunately, a combination of local experiences and conventional techniques proved to work, and were both less costly and more environmentally friendly. At the time of writing, more than 50 percent of the farmers in the village where experiments were carried out were using the technology.

Approach 3—Promoting linked technologies

One key challenge in supporting farmer innovation is to identify and integrate technologies addressing one or more identified problems without negatively affecting other system components. For instance, during the colonial era, promotion of conservation technologies was led by conservation programmes while technologies for improving agricultural production were facilitated by agricultural research institutes. Separating conservation and production, and piecemeal promotion of technologies and management practices (using a “commodity” or “single factor” approach) did not bring real benefits to farmers; in fact it failed to create the desired impact. In response to this challenge, AHI developed and tested approaches for facilitating farmer experimentation with an explicit effort to link conservation with production-enhancing technologies.

The term “linked technologies” was coined to define technologies that when applied simultaneously at plot or farm level render multiple benefits by enhancing adoptability of discrete technologies or fostering synergies that would not exist had technologies been applied in isolation. For instance, given the steep slopes, intensive cropping, and high rainfall intensity in most of the AHI sites, soil fertility decline was very apparent. The research teams employed several participatory techniques in order to develop the capacity of farmers and researchers in integrated soil fertility management; foster partnerships among stakeholders to avail best-bet technologies; and foster a change from commodity-oriented to a more holistic and participatory approach placing farmers at the forefront of technology development and evaluation (see, for example, Stroud, 1993). The guiding hypotheses were the following (Stroud, 2003):

•    Technologies with win–win benefits (e.g., increased income, improved soil fertility) will build farmers’ confidence to test more complex NRM technologies, and strengthen the demand side in the technology innovation process.

•    Problem-solving technologies with multiple benefits will bring more food and cash income to farmers of different resource endowments by solving multiple problems simultaneously, with solutions attractive enough to “sell” to others.

The methodology for developing linked technologies starts with the methods described in the above section “Understanding systems and clients”—namely, participatory rural appraisal techniques at village level to identify constraints faced by different social groups, followed by participatory testing of a wide range of technologies by thematically based farmer groups—starting with identified entry points. The next steps were specific to the linked technology approach, as follows:

1.    Once solutions are found to the most pressing issues (addressed through the testing of entry points), researchers facilitate access to more complex technologies. These technologies often relate in one way or another (e.g., through nutrient or capital flows, or labor savings) with technologies that have already been tested. The latter could have already been adopted, or could face some constraint that a new technology can assist in alleviating. The new innovations can also be unrelated to the entry point, and build on farmers’ enthusiasm to innovate rather than on proven technologies as a means to propel interest in more complex innovations. Importantly, however, the linked technologies bring immediate benefits while also fostering farmer investments in more complex NRM technologies with slower returns. For instance, while soil and water conservation was a key intervention to minimize erosion, increase water infiltration and increase input efficiency at farm and landscape scales, farmers had a difficult time engaging in such labor-intensive practices without immediate financial returns to their labor. By combining the testing of conservation bunds with forage grasses, organic nutrient management, and multipurpose trees as linked innovations, farmers were able to generate immediate benefits such as livestock feed, improved yields (crops, milk, manure), and fuel wood from investments in soil stabilization and fertility improvements (see also Box 2.4).

2.    Gradually farmers intensify and specialize in a system such as horticulture that renders the much-needed economic as well as social benefits and sustains or expands NRM investments. As the economic returns from NRM investments begin to materialize from the high-value crop and/or livestock enterprises, farmers are often propelled to invest more or expand the area over which the innovation is carried out. Thus, complementary innovations not yet tested in the first step such as integrated pest management (IPM) or other high-value crops with complementary growing cycles (for intercropping or relay cropping) can be brought into the innovation system.

The role of research was to facilitate access to technologies, train lead farmers, support farmer experimentation, and guide and monitor what different farmer groups did to integrate the various technological options—and their perceived impacts. The gradual, iterative process of planning, testing, evaluation, and replanning in a system that becomes ever-more diversified and integrated, is portrayed in Figure 2.1.

FIGURE 2.1 Simplified model of farm level entry point and linked technologies

Integrating technologies is a function of time, space, demand, and appropriateness of the interventions under the given circumstances. It should be done through targeting clients and system niches and by providing problemsolving interventions addressing the most important household priorities. This will improve the confidence of both farmers and researchers and the rapport between them as they seek to address more complex system constraints that may need more than one technological and institutional intervention.

It should be noted that the process of developing evermore complex linkages between technologies in wider system-wide innovations is not a one-off process but rather a time-consuming, stepwise engagement whereby farmers integrate options to supplement earlier investments for increasing returns from their farms and investments. Figure 2.2 illustrates where the stepwise approach is used to foster integrated soil fertility management in the Ethiopian highlands. These innovations can then be integrated into watershed-level innovations at a later stage (see Chapter 3 for details).

The approach should give emphasis to building the capacity of the communities and R&D teams to implement a systems approach and address the needs of diverse social groups. Farmers play a key role in linking technologies. Box 2.6 presents another success case, linking soil conservation with fodder production in Ethiopia to conserve soil while enhancing livestock production.

FIGURE 2.2 Stepwise integration of various technologies and approaches to improve natural resources management in the Ethiopian highlands

Lessons learned

AHI's experience in supporting farmer innovation confirmed the need for scientists to facilitate a dialogue based on mutual respect and learning by accepting and respecting farmers’ knowledge. Scientists have important roles to play by bringing in information, methods and analyses that complement what farmers already know and can do themselves. Key lessons learned on efforts to foster farmer experimentation and innovation include the following:

•    A host of considerations and decision criteria enter into any innovation process, many of which are specific to the local setting or cultural preferences, posing a challenge to diffusion of innovations. The same constraint is not necessarily resolved in the same way in different locations, even within the same agro-ecological region. This implies the need to replicate farmer innovation as an approach to problem solving, not the solutions generated through these approaches.

•    A critical element to developing effective partnerships between farmers, researchers, and development actors in supporting farmer innovation is overcoming the widespread tendency to underestimate farmers’ knowledge and innovation capacity and treating them as equal partners. Learning to listen to farmers and take their feedback on board whenever they report challenges faced in testing and adopting interventions is essential if researchers are to play an important role in farmer-led innovation processes.

•    Research and extension practices that build on farmers’ knowledge, engage farmers’ creativity, and allow for their active involvement in outreach activities are capable of producing results that far exceed and outlast those possible through more conventional approaches.

•    The “linked technology” approach enabled farmers, development agencies, and research organizations to address poverty and natural resources degradation in a holistic manner.

•    Market opportunities are an important impetus for technology adoption and systems intensification, and efforts to identify and meet market demands within a wider innovation effort can go a long way in catalyzing change.

•    Farmer–researcher partnerships for farm-level innovation require flexibility when defining the role of research. Some interventions do not require formal experimentation, as the returns are quickly visible. In some cases, the researcher's role became one of conceptualizing a system so as to introduce new ideas rather than the design and implementation of experiments, of monitoring with the aim of understanding farmers’ innovations and evaluations, and of providing support to dissemination and scaling-up processes.

•    Interventions with win–win benefits are effective in bringing about immediate impact at household and community scales.

•    Mechanisms to involve innovative farmers as local champions of an innovation process can be an effective means of stimulating local innovation, providing technical backstopping to other farmers and facilitating dissemination—a topic to which we now turn.

Disseminating proven technologies and approaches

Conventional approaches for technology dissemination are usually top-down and commodity-oriented, with the mode of technology dissemination assumed to be “linear”—namely, from research to extension to farmers. Critical factors affecting adoption such as socio-cultural, policy and institutional conditions were not considered in this approach. Furthermore, most of the technologies were generated on station through researcher designed and managed trials. Direct feedback from farmers as well as several formal adoption studies have clearly shown that technologies developed using this conventional approach were often not appropriate to local circumstances, thus leading to low adoption (Amede et al., 2001). This was largely owing to the limited involvement of farmers in the development and dissemination of technologies, as well as to weak institutional support to facilitate the adoption capacity of target groups. Horizontal and geographical spread of technologies is limited even when facilitated by public institutions and NGOs. The challenge for AHI was therefore to identify socio-economic and biophysical incentives to facilitate the scaling up of innovations proven to work in select locations. Using more “bottom-up” approaches, AHI has seen a gradual increase in farmers’ interest and adoption of different technologies, resulting in higher incomes and food security for households (Mowo et al., 2002).

Technology dissemination beyond partner households and villages has been hindered by blanket recommendations and poor packaging. Contrasting production systems and socio-economic circumstances demands a diversity of technological innovations and approaches. The diversity of household production objectives, for example, with some households concentrating on cash crops and others focused more on achieving food self-sufficiency, requires careful targeting of technological interventions. Resource-poor farmers, especially those distant from markets, face difficult decisions over the allocation of scarce resources (e.g., land, labor, nutrients, and water). Decisions on the allocation of resources are often associated with immediate financial gains and food security, with limited assessment or appreciation of the impact of management decisions on long-term effects or other system components (e.g., soils). There is therefore a need to explore mechanisms for matching technologies to specific recommendation domains, as defined by agro-ecological conditions, cropping systems, cultural values, system niches, or socio-economic variables.

Besides technologies being poorly adapted to different agro-ecologies and socio-economic circumstances, some technologies and approaches demand collective decisions and policy support to be adopted—further limiting their spread when these factors are not taken into consideration. Farmer-to-farmer dissemination of technologies through existing social networks—be they defined by area of residence, friendship, kinship, marriage, religion, or other factors—has been found to be one successful approach (Adamo, 2001), though the reach is limited. AHI has developed approaches that have significantly increased the spread and adoption of technologies within benchmark sites where spontaneous farmer-to-farmer sharing was limited or socially biased (German et al., 2006b). Research played a critical role in helping farmers to organize themselves, access and multiply preferred technologies, and sustainably utilize these interventions and promote them within the locality and beyond.

Approach development

Various technologies and practices may demand different dissemination approaches. This is because some technologies are easy to disseminate while others are more knowledge-intensive and difficult to scale up unless accompanied by intensive mentoring, guidance and institutional support. AHI used three different approaches for scaling up: farmer research groups, externally mediated diffusion in which dissemination is governed through locally formulated by-laws, and a formal approach based on research to identify social and biophysical barriers to adoption.

Approach 1—Farmer research groups

In addition to being employed as a platform for farmer experimentation, FRGs were used as a means to scale up impacts through technology dissemination and the testing of additional innovations that might work synergistically with technologies to enhance impact (e.g., micro-credit). As mentioned above, the process of establishing FRGs for technology dissemination included the organization of farmers into thematically based groups for testing technologies on behalf of the wider community, formal training, the identification and implementation of experiments and evaluation of results. With this approach, the transition between experimentation and scaling up is a relatively seamless one—with steps in the latter a natural transition from the former. This means that the use of FRGs as a platform for scaling up is informed as much by the natural progression of farmers’ interests and experiences as it is by a set of discrete steps. However, in most cases it consisted of the following components:

1.    Members of different thematic groups presenting their findings to the wider community (including FRGs working on other themes) at different stages of experimentation.

2.    Research teams and FRG members developing a scaling-up strategy and jointly organizing field days, farm visits, posters, and demonstration trials.

3.    Farmers beyond pilot communities or groups seeking support from research to expand the FRG methodology, requiring both continued support to farmer experimentation and a proper strategy for ensuring continuity and sustainable delivery of technological options.

4.    FRG members starting their own community seed multiplication initiative as a business venture, often on their own initiative.

5.    FRG experimentation with other non-technological innovations (e.g., credit provision, marketing) to alleviate the constraints to adoption and thereby enhance technology adoption and dissemination either directly or indirectly (see Box 2.7 for the case of savings and credit associations in Tanzania).

6.    Research and development teams facilitating linkages between successful FRGs and local authorities to disseminate proven innovations beyond pilot sites.

It is interesting to note that the dissemination phase exposed a number of weaknesses in the FRG methodology as a whole, highlighting some of the challenges that need to be managed for the successful implementation of FRG-based experimentation and dissemination efforts. With the majority of researchers initially lacking experience in applying principles and concepts of participatory methodologies, they were ill-prepared to assist moving farmers away from individualistic attitudes and to support the evolution of cohesive farmer groups. In some instances, this led to limited interest in fostering collective benefits, with FRG members seeing technologies as their own property rather than something they have been given a mandate to test on behalf of the wider community. This became apparent at the time of scaling up. In another instance where participatory approaches were tested in a site where government programmes were providing cash and inputs free of charge to farmers implementing soil and water conservation, some of the technologies used as entry points were accepted by farmers owing to these benefits rather than the technologies themselves. This hindered subsequent efforts to scale up technologies requiring significant financial or labor inputs, as the underlying motives for uptake were weak. Yet these experiences were more the exception to the rule, with most farmer groups realizing the benefits of working with researchers during participatory technology testing and dissemination.

Approach 2—Externally mediated diffusion

AHI's experience in several benchmark sites suggests that the spread of knowledge-intensive technologies is not as fast and simple as crop varieties and forage, even within a village. Moreover, scaling up “fast-moving” technologies does not mean these innovations will reach different social groups or locations without external support and facilitation by local institutions, extension departments, or research institutions. AHI and its partners therefore experimented with different approaches for mediating technology dissemination and farmer-to-farmer “spillover” to ensure equitable access to technologies being tested by different FRGs. The question of how to equitably share knowledge and technology among male and female farmers, and to reach farmers with different resource endowments, was discussed at community meetings involving community leaders, local authorities, and early adopters. This was done in part to ensure seeds continue to spill over from one FRG to the next, to counter the tendency for FRG members to take project seed as their own property and stop there. The participatory formulation of local by-laws and their subsequent endorsement by local administrative authorities was also considered as a means to guide or govern the technology dissemination process according to guidelines agreed upon by all. In short, this approach may be characterized by a very direct mediation from outside the community designed to enhance equity.

The following steps were used to facilitate externally mediated technology diffusion:

1.    Mobilize and sensitize community members on key crop and livestock issues identified through participatory diagnostic procedures such as PRAs, to explore the extent to which those who wish to access technologies are able to do so and to identify barriers to technology access and uptake.

2.    In cases where inequitable access to technologies is observed (either by gender or any other factor), hold community meetings to generate by-laws to govern modes of technology dissemination (Box 2.8).

3.    Establish technology testing sites with beneficiary farmers.

4.    Form marketing committees and higher level organizations from village to sub-county levels to facilitate market linkages, including establishment of collection centers.

5.    Establish village libraries for publicity (reading materials, pamphlets regarding NRM and other agricultural activities).

6.    Improve the capacity of farmers to multiply seeds, including phyto-sanitary measures during production (as required with Irish potato) up to post-harvest handling.

7.    Promote the products using dramas, role plays, shows, demonstrations, and other tools.

8.    Encourage farmers to keep records and use them for monitoring and tracking progress.

Key stakeholders such as local leaders and extension agents were involved in planning and implementation at all levels.

Approach 3—Targeting systems and clients for dissemination of technologies

While technological innovations are vital in solving farm-level constraints to food security and sustainable NRM, their adoption and utilization by local communities can be limited unless interventions effectively target clients and key system constraints, and contribute to overall household objectives. Although farmers are keen to learn about technologies through farmer field schools and on-farm testing, not all farmers are involved in piloting technologies and other farmers will need time to test and adapt them to their own, often sub-optimal, conditions. Yet managing such a process with each and every household is costly, and tools for predicting adoptability under different conditions can help to reduce such costs.

Predicting the likelihood of adoption of different technologies and formulating relevant recommendations are difficult owing to the variable nature of biological and social-economic systems and the trade-offs that characterize production and resource/input management decisions. Thus, the generation of decision support tools based on detailed analyses of farming systems may provide a complementary tool to more participatory techniques in identifying technologies and their potential socio-economic and biophysical niches. With the ability to consider multiple variables simultaneously, the tools described below can enable more accurate targeting of innovations and clients to foster multiple household objectives simultaneously (e.g. increased productivity of crop and livestock systems, income, and food security). Key steps in the approach include the following:

1.    Characterization of systems and clients. As systems with different characteristics will differ in their capacity to intensify and the pathways through which this occurs, and different drivers of change will influence enterprise choices and their management, the targeting of interventions should start with a characterization of the system. This includes both socio-economic and biophysical perspectives. The former includes household resource endowments by wealth, gendered perspectives on constraints and priorities, household involvement in institutions of collective action, and access to technologies and innovations. The latter includes a characterization of the production system, access to water and nutrient resources, soil fertility, and other relevant biophysical parameters.

2.    Identification of socio-economic and biophysical factors affecting adoption (Box 2.9). Farmers employ multiple criteria when deciding whether a technology in question is appropriate for their circumstances, and whether it can be productively integrated into their farming practices. This step involves identifying variables affecting the adoption of a particular technology by households with variable economic and demographic characteristics, resource endowments, and system constraints. These include the extent to which the technology is aligned or compatible with household preferences and cultural values, its actual or anticipated performance in different farm and landscape niches, the immediacy of benefits derived from its adoption, complementarities or conflicts with other system sub-components and users, and the potential of the intervention to address multiple challenges simultaneously. These factors are determined based on past experience by some or all households with the same or similar technologies.

3.    Prioritization of major socio-economic and biophysical determinants of adoption using pair-wise ranking, community validation, and case study analysis.

4.    Development of decision guides to assist development agencies, extension personnel, and farmers to target systems and clients for a specific intervention.

Lessons learned

The following lessons were learned in our efforts to test different strategies for local technology dissemination:

•    Building the skills and capacity of “elite” farmers may help to kick-start the technology innovation process; however, technologies developed by these farmers are not guaranteed to reach the broader community. This may be owing to limitations in farmer-to-farmer sharing or, in cases where the early innovators are wealthier farmers, to different resource endowments.

•    Contrary to common perception, communities are not homogeneous entities in which benefits to some households will automatically “trickle down” to all. Explicit strategies are often needed to ensure resources brought from the outside and intended for the collective benefit are well governed based on principles of equity.

•    Where select individuals step forward to test technologies on behalf of the group, the acquired technologies will often be considered to be their personal property unless the individual responsibilities to the group (e.g., subsequent sharing of information or seed) are clarified in advance.

•    Building farmers’ confidence, trust and collegial spirit will go a long way in building strong groups and enhancing farmer-to-farmer sharing of technologies. The initial trust between farmers and research and development teams was an important factor contributing to building strong local institutions.

•    Building farmers’ capacity to access loans and services and linking them to district- and national-level financial institutions will significantly contribute to agricultural productivity, rural livelihoods, and ability to invest in natural resource management.

•    Targeting potential clients and system niches can help to facilitate technology dissemination and adoption by providing a cost-effective tool for predicting adoptability of agricultural innovations. There is a need to develop these tools together with the potential users through participatory and data-based approaches. Testing and validating decision tools in diverse settings can help to expand the tool's reach beyond pilot sites.

•    Different technologies and practices may demand different dissemination approaches as some (e.g., crop varieties) are easy to disseminate, while other technologies (e.g., conservation agriculture) are knowledge-intensive and difficult to scale up unless the process is strongly facilitated by intensive mentoring, guidance and external or internal institutional support.

•    Finally, and most importantly, the establishment of strong local institutions for technology dissemination requires that the demand for such institutions comes from the grassroots, whether the community at large or historically disadvantaged groups therein.

Tracking technology spread and impacts

In addition to proactively engaging in technology testing and dissemination strategies, it is often useful to understand the actual fate of technologies following such formal interventions. This can help to identify adoption bottlenecks, whether social, economic, or biophysical. It can also highlight the spontaneous ways in which farmers adapt technologies or their management to enhance their compatibility with local farming systems or increase the benefits derived from them, so as to ensure these innovations are popularized. Finally, it can help to identify areas where complementary innovations are needed, for example to minimize negative social or biophysical impacts resulting from technology adoption.

This section describes a methodology for tracking the spontaneous “spillover,” or farmer-to-farmer sharing, of introduced technologies. Conventional adoption studies emphasize identification of factors influencing adoption and evaluation of impact in terms of the numbers of adopters and the area over which the technology is applied. The proposed methodology operates under an expanded set of objectives and research questions. Identification of pros, cons, and adoption barriers for different technologies can assist the targeting of improvements on the technology or its mode of delivery. Identification of the characteristics of adopting households and farming systems enables our understanding of who benefits from introduced technologies and can improve technology targeting for diverse social groups. Characterization of social networks through which technology flows in the absence of outside intervention can enable us to tap into existing social networks or to target strategies to overcome social biases inherent in these (i.e., gender bias within patrilineal or male-dominated societies). Identification of social and biophysical innovations made by farmers can help in our understanding of how technologies may be modified to better fit the farming system, and integrated into scaling out efforts. Finally, identification of positive and negative social and agro-ecological impacts can shed light on how to maximize positive while minimizing negative spin-offs of technological innovation (German et al., 2006a, 2006b).

Approach development

This expanded scope is achieved through a number of variations in conventional adoption studies, which tend to follow four basic steps: 1) researcher identification of variables likely to influence adoption; 2) structured household questionnaires focusing on key variables; 3) statistical analysis to correlate key variables with technology adoption; and 4) researcher interpretation of observed patterns. The modified methodology includes these same steps, but systematically builds local perceptions into the approach. Focus group discussions with different social groups (adopting and non-adopting farmers, or by gender and wealth) during Step 1 of the methodology aid in identifying basic patterns of adoption and technology sharing, as observed by farmers. Newly identified variables from these focus group discussions are then integrated into the standard household surveys, to enable quantification of relevant variables. Focus group discussions are also utilized during Step 4 of the methodology to integrate farmers’ interpretation of observed patterns into the analysis. While researchers may believe an observed pattern may be explained in one way, farmers will often have their own explanation that differs considerably from researchers’ interpretations. Each of these steps ensures that the methodology is sensitive to patterns of adoption and social interaction specific to the local context.

Household survey methods used in Step 2 also differ in two important respects. Sampling of interviewees can be done through the standard random sampling approach or through a form of “snowball sampling” in which social networks are traced from the original “project farmers” (L0) to “level one adopter” (L1) (farmers adopting from project farmers) to “level two adopters” (L2), and so on as presented in Figure 2.4. While random sampling may be better for rigorous econometric analysis of adoption variables, snowball sampling is best for understanding social networks through which technologies spread in the absence of outside interventions and how adoption levels and technologies themselves change through successive levels of “spillover.” The latter also provides a picture of local adoption dynamics and pathways. The household survey methods employed here also differ by the integration of more in-depth qualitative interviews in a selected number of households. This aids in understanding social and biophysical innovations, livelihood and environmental impact, and the steps associated with technology adoption—generally, information requiring qualitative inquiry.

FIGURE 2.4 Levels of technology “spillover” relative to project interventions

As a whole, this methodology helps us to move from a view of technological innovation as a one-off step (introducing new technologies) to a process that proceeds from problem definition to technology targeting, testing, monitoring, troubleshooting, and dissemination or discontinuation. This is of fundamental importance in ensuring that patterns and lessons are not lost, and to minimize the risks introduced through technological innovation—such as negative agro-ecological impacts or socio-economic gap-widening resulting from biases towards wealthier or male farmers.

The methodology is applied for at least two consecutive growing seasons after technology dissemination, so that patterns of farmer-to-farmer sharing may be identified. The steps in the modified methodology may be summarized as follows:

1.    Reaching a common conceptual understanding and agreeing on technologies to be tracked.

2.    Focus group discussions to identify basic adoption patterns.

3.    Identification of networks through which technologies flowed from source farmers (“Tracking surveys”).

4.    On-farm interviews with new adopters.

5.    Data analysis to identify patterns of technology spillover.

6.    Focus group discussions to interpret emerging findings.

For a detailed description of the methodology, including research instruments and sample findings, please see German et al. (2006b).

Following application of the methodology, new technologies and dissemination approaches are targeted to overcome identified problems. These problems might include social, economic or technical barriers to technology adoption, or negative social and agro-ecological impacts of adoption. Table 2.2 presents some of the agro-ecological impacts that have been identified through application of the methodology. This table illustrates the substantial spin-offs, both positive and negative, that often accompany technological innovation. These impacts are generally obscured under conventional adoption studies, but have a profound impact on the technology's success and system sustainability.

Examples of social and economic barriers to technology adoption are summarized in Box 2.10. Different types of barriers lend themselves to different types of solutions. Negative agro-ecological impacts can be addressed by testing complementary technologies that help to minimize negative effects of innovation, or by further research (breeding or on-farm experimentation) to further improve upon the technology itself. Economic barriers to technology adoption may require coupling technology dissemination activities with credit systems, facilitating negotiations among early and late innovators prior to technology testing, and dissemination to establish rules for technology dissemination that will ensure technology access by low-income farmers. Gendered barriers to technology access can also be addressed through negotiation of rules for equitable technology access, as was done in Areka benchmark site.

TABLE 2.2 Positive and negative^a agro-ecological impacts associated with technologies introduced in Lushoto, Tanzania

Note: ^a Positive impacts, as viewed by farmers, are in bold font and negative impacts in italics.

In addition to its application as a retrospective impact study, this methodology can be applied within an iterative process of technology targeting, dissemination and monitoring. In this case, adoption barriers or negative effects of new technologies are periodically captured and addressed through further technological or methodological innovations. The methodology would need to be simplified for regular use, focusing on the most salient observations of farmers and perhaps minimizing the level of formal data collection.

Lessons learned

Lessons learned from efforts to develop methods to track technology spread and impacts are several:

•    Technological innovation often involves substantial spin-offs, both positive and negative; failure to identify and address these can reduce demand for the technology or introduce a set of problems that are propagated along with the technology. Identifying them provides an opportunity for corrective measures to be designed and tested and for “linking” multiple technologies for improved impact.

•    “Sharing biases” within rural communities can propagate inequitable access to technologies, irrespective of efforts by extension agencies to work with equal numbers of male and female farmers. Systematically tracking sharing patterns can help to identify such biases and to design and test strategies to overcome them for improved adoption and equity.

•    The new approach systematically integrates farmers’ perceptions and experience on the introduced technology into the formal methodology, broadening the scope of what is learned and integrating farmer recommendations into research and dissemination strategies designed to overcome identified problems.

•    The new approach provides an opportunity for adapting the introduced technology to address its negative effects and better fit the targeted farming system, by identifying local innovations introduced during the technology's spontaneous spread or proactively identifying adoption niches and negative impacts.

Missing links

While substantial progress has been made in identifying effective approaches for enabling livelihood improvements while also countering the degradation of resources at farm level, a few key areas of methodological innovation remain to be explored. These include the following:

1.    There is a need to consider how to tap into a wider set of opportunities and drivers in designing interventions, so as to tap into potential motivating factors (e.g., market outlets) and to move from a reactive to a proactive approach in supporting farmer innovation. The choices of interventions and innovations were often based exclusively on local preferences, without considering wider market opportunities and policy drivers and the existing institutional capacity to scale up interventions beyond pilot sites. By placing attention on wider market opportunities and policies shaping farmer behavior, it may be possible to tap into wider motivating factors and thus support more widespread adoption. By embedding innovation processes within existing institutions, it would be possible to embed the innovations—and the innovation process itself—within organizational structures capable of sustaining the innovation process within and beyond benchmark sites.

2.    Systems optimization through the use of models requires an analytical simplification of the system that may depart from real life decision processes and management principles. The more detailed the analysis of system features, community needs and preferences, market opportunities, and drivers of change, the more likely that optimization models and participatory optimization processes will be effective. More effort is needed to develop and test cost-effective methods that simultaneously enhance system understanding by farmers and researchers while targeting “best bet” facilitation processes for system change (including policy reforms).

3.    One critical gap was in the development of approaches for building on the knowledge and skills of innovative farmers (a source of learning and innovation) to bring change across a wider area, and thus achieve watershed-wide farm productivity gains. Most “early innovators” are either isolated from others, lacking the mechanisms or motives to support innovation at a wider scale, or have unique characteristics that enable them to take risks and try out new innovations—thus limiting the extent to which proven innovations will be automatically accessible to others. The development and testing of methods for linking such early innovators with the needs and capacities of a wider set of actors at village and landscape scales is needed, including mechanisms to incentivize efforts expended for the collective good rather than for personal gain.

4.    There is a need to explore how to move beyond “linked technologies” to “linked innovations.” The success of efforts to couple technological innovations with credit facilities and with social and governance innovations, and the tendency for farmers themselves to employ social innovations when adopting new technologies (German et al., 2006a), illustrate the promise of linking social and technical innovations. More effort is needed to bring social scientists and marketing specialists into efforts to support farmer innovation, so as to identify and test social and marketing innovations that work in synergy with technologies in enhancing impact at farm level. 5. More can be done to explore opportunities for enhancing impact by going further “downstream” along the farmer-to-farmer dissemination pathway. The technology spillover methodologies identified a number of factors constraining adoption and positive impacts, both social and biophysical, which could be the subject of further experimentation and innovation. This is likely to be a very fertile area of technological and social innovation, and thus impact, as it is informed by actual adoption bottlenecks. While the program engaged in a few such innovations, for example to address germplasm constraints to the spread of banana in Lushoto BMS, much work remains to be done in this regard.

Conclusions

The crop–livestock farming systems in the highlands of eastern Africa are characterized by low-input farming, heavily reliant on the recycling of internal resources. Resource degradation is aggravated by high human and livestock population densities, which lead to overgrazing, nutrient mining, erosion, and water depletion. Intensification and diversification of these systems is one important pathway for improving rural livelihoods. This chapter sought to share AHI experiences in farm-level intensification and diversification through approaches for characterizing systems and clients (and thus potential adoption niches), supporting farmer experimentation, reaching larger numbers of farmers in benchmark sites through technology dissemination, and ex-post tracking of the spontaneous farmer-to-farmer spread of innovations. Ultimately, a combination of strategies is needed at different stages of an innovation process to effectively support farmers to generate greater returns from a limited resource base. Farmer field schools, farmer-managed experimentation, and farmer-led dissemination enhance farmers’ capacity to make informed choices and test them through an experimental learning approach, whereas researcher-led development of decision support tools and documentation of farmer-to-farmer dissemination provide a means to identify strategic interventions to enhance impact for further testing with farming communities.

Farmers’ choices of livelihood strategies substantially influence crop and livestock decisions and welfare and resource outcomes. Based on our experience to date, adoption of technological innovations often depends on a few key factors, including: 1) the type of technical and material support farmers receive from extension and research; 2) the level of familiarity of farmers with the suggested interventions; 3) the demands placed on limited resources by innovations; 4) the associated benefits that are derived, both financial and other; and 5) the time required to derive these benefits. Bottom-up processes for engaging communities are essential in integrating these and other considerations into the innovation process, and in motivating farmers to individually and collectively address production constraints and capture new livelihood opportunities.

There are two possibilities for achieving wider impact from the innovations presented in this chapter. The first is to scale up the actual technological or social innovations that were successfully employed to intensify or diversify local farming systems. The second is to scale up the approaches and tools used to generate these innovations or to target specific niches for further uptake among development agencies. The latter approach is the preferred approach for accounting for the diversity in local resources, preferences, and conditions. Each approach will be treated in greater depth in Chapter 6.

Notes

1  As observed in Areka, these subsidies also entail risks if they are not continued, given the tendency for farmers to invest less in organic nutrient management when using chemical fertilizers. If farmers have limited ability to continue purchasing these inputs, this means they will have less fertile soils to fall back on.

2  The model had three basic modeling components: 1) an objective function, which minimizes or maximizes a function of the set of activity levels; 2) a description of the activities within the system, with coefficients representing their productive responses; and 3) a set of constraints that define the operational conditions and the limits of the model and its activities.

References

Adamo, A. (2001) Participatory agricultural research processes in Ethiopia using farmers’ social networks as entry points. CIAT Occasional Papers Series No. 33. Cali, Colombia: CIAT.

Alumira, J. and A.O. Awiti (2000) The African Highlands Eco-region: Situation Analysis. Kampala, Uganda: AHI.

Amede, T. (ed.) (2003) Natural Resource Degradation and Environmental Concerns in the Amhara Regional State, Ethiopia: Impact on Food Security. Bahir Dar, Ethiopia: Ethiopian Society of Soil Sciences.

Amede, T., T. Belachew, and E. Geta (2001) Reversing the degradation of arable land in Ethiopian highlands. Managing African Soils No. 23. London: IIED.

Amede, T. and R. Kirkby (2004) Guidelines for integration of legume cover crops into the farming systems of the east African highlands. In: A. Bationo (ed.), Managing Nutrient Cycles to Sustain Soil Fertility in Sub-Saharan Africa. Nairobi, Kenya: Academy Science Publishers, pp. 43–64.

Amede, T., A. Stroud, and J. Aune (2004) Advancing human nutrition without degrading land resources through modeling cropping systems in the Ethiopian Highlands. Food and Nutrition Bulletin 25(4): 344–353.

Amede, T., R. Kirkby, and A. Stroud (2006) Intensification pathways from farmer strategies to sustainable livelihoods: AHIs’ experience. Currents 40/41: 30–37. Uppsala, Sweden: SLU.

Assefa, A. and T. Fenta (2006) Harnessing local and outsiders’ knowledge: Experiences of multi-stakeholder partnership to promote farmer innovation in Ethiopia. Prolinnova Working Paper No. 12.

German, L., J.G. Mowo, and M. Kingamkono (2006a) A methodology for tracking the “fate2 of technological innovations in agriculture. Agriculture and Human Values 23: 353–369.

German, L., J. Mowo, M. Kingamkono and J. Nuñez (2006b) Technology spillover: A methodology for understanding patterns and limits to adoption of farm-level innovations. AHI Methods Guide A1. Kampala, Uganda: AHI.

Kelly, V., B. Diagana, T. Reardon, M. Gaye, and E. Crawford (1996) Cash crop and food grain productivity in Senegal: Historical view, new survey evidence and policy implications. MSU International Development Paper No. 20.

Lyamchai, C.J., M.N. Kingamkono, J.G. Mowo, and L. German (2006) Farmers innovations in natural resource management: Lessons and challenges from Lushoto, Tanzania. In: T. Amede, L. German, S. Rao, C. Opondo and A. Stroud (eds.), Integrated Natural Resource Management in Practice: Enabling Communities to Improve Mountain Livelihoods and Landscapes. Kampala, Uganda: AHI.

Morrison, K.D. (1994) The intensification of production: Archeological approaches. Journal of Archaeological Methods and Theory 1(2): 111–159.

Mowo, J.G., S.T. Mwihomeke, J.B. Mzoo, and T.H. Msangi (2002) Managing natural resources in the West Usambara Mountains: A glimmer of hope in the Horizon. Mountains High Summit Conference for Africa, Nairobi, Kenya, May 6–10, 2002. Available at: http://www.mtnforum.org/sites/default/files/pub/213.pdf (accessed May 29, 2009).

O'Neil, R.J. (1995) Farmer Innovation and Biological Control. Midwest Biological control Newsletter 2(3): 1–2.

Pontius, J.C., R. Dilts and A. Bartlett (2002) From farmer field school to community IPM: Ten years of IPM training in Asia. RAP/2002/15, FAO Regional Office for Asia and the Pacific, Bangkok.

Reardon, T., C. Barrett, V. Kelly, and K. Svadog (1999) Policy reforms and sustainable agricultural intensification in Africa. Development Policy Review 17(4): 375–86.

Smaling, E.M.A. (1993) Soil nutrient depletion in Sub-Saharan Africa. In: H. Van Reuler and W.H. Prins (eds.), The Role of Plant Nutrients for Sustainable Food Production in Sub-Saharan Africa. Leidschendam: Dutch Association of Fertilizer Producers. Leidschendam, pp. 53–67.

Stroud, A. (1993) Conducting On-farm Experimentation—A Guide Book. Cali, Colombia: CIAT.

Stroud, A. (2003) Linked technologies for increasing adoption and impact. AHI Brief A3. Kampala: AHI.

Tarawali, S., J.D.H. Keatings, J.M. Powel, P. Hiernaux, O. Lyasse, and N. Sanginga (2004) Integrated natural resources in West African crop–livestock systems. In: T.O. Williams, S.S. Tarawali, P. Hiernaux and S. Fernandez-Rivera (eds.), Proceeding of the International Conference on Sustainable Crop–livestock Production for Improved Livelihoods and Natural Resource Management in West Africa. Ibadan, Nigeria: IITA, pp. 349–370.

Tiffen, M. (2003) Transition in sub-Saharan Africa. World Development 31(8): 1343–1366.

Thornton, P.K, R.L. Kruska, N. Henninger, P.M. Kristanson, R.S. Reed, F. Ateno, A. Odero, and T. Ndegwa (2002) Mapping Poverty and Livestock in the Developing World. Nairobi, Kenya: International Livestock Research Institute.

Van Keulen, H. and J.B. Schiere (2004) Crop–livestock systems: Old wine in new bottles. In: New Directions for a Diverse Planet: Proceedings of the 4th International Crop Science Congress. Brisbane, Australia, September 26–October 1, 2004. Available at: www.cropscience.org.au (accessed May 29, 2009). Waters-Bayer, A. and W. Bayer (2009) Enhancing local innovation to improve water productivity in crop–livestock systems. The Rangeland Journal 31: 231–35.

Wickama, J. M. and J.G. Mowo (2001) Using Local resources to improve soil fertility in Tanzania. Managing African Soils No. 21. London: IIED.