INTRODUCTION
Conservation Agriculture (CA) is based on three complementary principles (minimum tillage, organic soil cover and crop diversification). CA represents a family of cropping systems which use ecological processes to produce while conserving natural resources, particularly soil resources (Scopel et al., Reference Scopel, Triomphe, Affholder, Macena Da Silva, Corbeels, Valadares Xavier, Lahmar, Recous, Bernoux, Blanchart, De Carvalho Mendes and De Tourdonnet2013). Its implementation involves a transformation of existing agricultural production systems, including changes in the relationship between people and nature (valuing ecosystem services). It also involves a transformation of agricultural practices (for example, no longer burning crop residues or tilling soil) and of the organization of work on a farm (readjusting crop calendars, labour input and division of tasks). Lastly, the successful implementation of CA relies on adequate changes in the configuration of stakeholder networks (Triomphe et al., Reference Triomphe, Goulet, Dreyfus, Tourdonnet, Bourrigaud and Sigaut2007).
In West Africa, CA is increasingly considered as a possible means to sustainably improve agricultural productivity and strengthen farmers’ resilience to climate variability (Tittonell et al., Reference Tittonell, Scopel, Andrieu, Posthumus, Mapfumo, Corbeels, Van Halsema, Lahmar, Lugandub, Rakotoarisoa, Mtambanengwe, Pound, Chikowo, Naudin, Triomphe and Mkomwa2012) in a context of increasingly frequent extreme weather events (Cooper et al., Reference Cooper, Dimes, Rao, Shapiro, Shiferaw and Twomlow2008) and rising demand for agricultural production. However, the principles of CA depart from prevailing farming practices based notably on tillage and single cropping. Before adoption takes place, farmers must be convinced that there is a need to switch to CA, and that CA can respond to their key objectives, such as improving food security and farm income. For this, technical references adapted to local conditions must be developed and rules governing the management of crop residues at the community level must be renewed. Crop residues are an open-access resource during the dry season, and crop farmers wishing to practice CA (and therefore to maintain soil cover) compete with livestock farmers who need forage for their animals. Farmers’ access to inputs (cover crops seeds and herbicides) and equipment required in CA must also be facilitated (Kassam et al., Reference Kassam, Friedrich, Shaxson and Pretty2009).
In the past, various approaches based on a linear conception of technology generation and transfer have had limited success on small family farms (Knowler and Bradshaw, Reference Knowler and Bradshaw2007). New approaches to innovation are needed, which allow local stakeholders to find acceptable solutions. Based on the recognition of the central role of stakeholder interactions in the success of innovation processes as described in the innovation system concept (Touzard et al., Reference Touzard, Temple, Faure and Triomphe2014), innovation platforms (IPs) have emerged gradually as a relevant means for the development and scaling of many kinds of innovations. IPs bring together stakeholders with different, and sometimes divergent, profiles and interests to facilitate conflict resolution, negotiation, social learning and collective decision making for concerted action (Nederlof et al., Reference Nederlof, Wongtschowski and Van der Lee2011).
Numerous experiments involving IPs to varying degrees have been reported around the world (Devaux et al., Reference Devaux, Horton, Velasco, Thiele, López, Bernet, Reinoso and Ordinola2009; Schut et al., Reference Schut, Klerkx, Sartas, Lamers, Campbell, Ogbonna, Kaushik, Atta-Krah and Leeuwis2015; Steins and Edward, Reference Steins and Edwards1999). For example, IPs have been used in agricultural extension to promote systemic and iterative learning processes and collective action among actors sharing the same problem (Steins and Edward, Reference Steins and Edwards1999). IPs have also been used widely in natural resource governance to prevent or manage user conflicts by helping stakeholders to realize that they depend on each other to manage a shared resource (Misiko et al., Reference Misiko, Mundy and Ericksen2013; Warner, Reference Warner2005). IPs can be implemented at the national level, for example with consortia of market chain actors and development organizations to address macro-level problems. They can also be implemented at the local level to facilitate interactions between farmers, local authorities and service providers in order to involve them in the design, implementation and assessment of solutions, and consequently empower the innovation capacity of farmers (Kilelu et al., Reference Kilelu, Klerkx and Leeuwis2013; Nederlof et al., Reference Nederlof, Wongtschowski and Van der Lee2011).
In Africa, IPs have been tested in projects aiming to improve agricultural productivity focussing on participatory testing and adaptation of technologies at local level (Schut et al., Reference Schut, Klerkx, Sartas, Lamers, Campbell, Ogbonna, Kaushik, Atta-Krah and Leeuwis2015), most often in tandem with the creation or reinforcement of local actors’ access to markets (Nyemeck, Reference Nyemeck2011). A few studies, however, have examined how to implement IPs to address complex systemic innovations such as CA, although the potential of the IP approach to respond to the challenges involved appears high (Nederlof et al., Reference Nederlof, Wongtschowski and Van der Lee2011). This article assesses the process by which IPs were designed and implemented in Burkina Faso under the impetus and coordination of a research team working under the Agroecology-based aggradation conservation agriculture (ABACO) project (Tittonell et al., Reference Tittonell, Scopel, Andrieu, Posthumus, Mapfumo, Corbeels, Van Halsema, Lahmar, Lugandub, Rakotoarisoa, Mtambanengwe, Pound, Chikowo, Naudin, Triomphe and Mkomwa2012). These IPs aimed to foster community-based participatory research and learning on CA in order to explore with stakeholders whether CA was relevant for the local context of Burkina Faso to encourage the development of sustainable production systems. After presenting the design process and the structure of the IPs, we analyse several outputs and outcomes, and discuss the relevance of IPs in facilitating the participatory research and learning on CA.
MATERIALS AND METHODS
Study area
The three villages selected for this study (Table 1) are located in contrasting regions in the ABACO project area in Burkina Faso. Koumbia, located in the Sudanian zone of the cotton production basin, has a good biomass production potential combined with relative pressure from livestock rearing. The dominant farming system in the area includes the single cropping of cotton in rotation with cereals (maize and/or sorghum) and the use of mineral fertilisers. Up until 2011, Koumbia had not been exposed to any CA experiments; however, it had been a site of a research program on the participatory design of agro-pastoral innovations (Vall and Bayala, Reference Vall, Bayala, Faure, Gasselin, Triomphe, Temple and Hocdé2014). Sindri and Yilou, the two other study sites, are located in the Sahelian-Sudanian zone of the country. Compared to Koumbia, they have a lower biomass production potential and face less pressure from livestock rearing. The dominant farming system in these two sites is based on cereals (sorghum and millet) combined with cowpeas or peanuts. The first participatory experiments involving CA began in 2009 under the project, Promotion de l'agriculture de conservation dans les petites exploitations agricoles d'Afrique de l'Ouest et du Centre (SCAP), funded by the International Fund for Agricultural Development (ACT et al., 2012).
Characteristics of study villages.
Innovation platform design process
A multidisciplinary research team (two agronomists, one sociologist, two animal production scientists and one geographer) initiated the IP construction processFootnote 1 at the village level as this seemed best suited to facilitate participatory research and learning around the conditions of implementation (or not) of CA. At this stage, since CA was relatively new in the study sites, the technology transfer and adoption at farm level were not the core objectives for these IPs. Inspired by the three stages identified by Nederlof et al. (Reference Nederlof, Wongtschowski and Van der Lee2011), the process of IP building relied on three complementary steps: (1) the diagnosis of existing forms of organization, whether endogenous or exogenous; (2) the development of an IP framework; and (3) the validation by stakeholders of the IP framework and the planning of activities.
Diagnosis of existing forms of organization
The aim was to identify the main existing forms of community-based organizations involved in the management of agro-pastoral resource at the village level that could be involved in the IPs to avoid duplicating structures and creating dependence on the ABACO project. This step was carried out through semi-structured interviews with the leaders of existing farmers’ organizations (3 in Sindri, 11 in Yilou and 13 in Koumbia), one to two local government representatives per village, and traditional leaders. The interviews included questions about the identity of the organization, how it emerged, its internal structure and governance, operations, communication system, partners and its strengths and weaknesses.
Development of an innovation platform framework
Based on the outputs of the preceding step, the research team developed a proposal for the structure and functioning of the IPs coherent with the general objective of participatory research and learning on CA. Considering the requirements to be addressed for the promotion of CA, the specific objectives for these IPs were: (1) the co-design of technical references adapted to local conditions, facilitated by individual and collective learning about CA; (2) the exploration of renewed rules governing stakeholders’ access to land and crop residues at the community level to address the main challenge of competition for crop residues impeding the feasibility of CA.
Validation by stakeholders of the IP framework and the planning of activities
Four one-day discussion workshops were organized in the study area with stakeholders deemed likely to join or interact with the IPs: from each village, seven representatives of farmer groups, two to five government outreach agents, between one and six retailers selling agricultural products, between two and four local government agents and one to two traditional leaders. Researchers from the ABACO project were also involved, as were one agricultural inputs supplier in Yilou and two local manufacturers in Koumbia. These workshops aimed to allow stakeholders to define potential constraints on the application of CA, their expectations and views on the IP framework proposed, adaptations which they might propose for this framework, the role they wished or did not wish to play and the activities they wished to conduct.
Monitoring the operations, outputs and outcomes of the IPs
The IPs were established through the process described above and began to operate at the end of 2011. To monitor and guide their operations, the research team defined four general criteria and seven indicators (Table 2) based on Tenywa et al. (Reference Tenywa, Rao, Tukahirwa, Buruchara, Adekunle, Mugabe, Wanjiku, Mutabazi, Fungo, Kashaija, Pali, Mapatano, Ngaboyisonga, Farrow, Njuki and Abenakyo2011) and Nederlof et al. (Reference Nederlof, Wongtschowski and Van der Lee2011). The effects and impacts of IPs were assessed in terms of the evolution of perceptions (construction/information shaping behaviour), attitudes (predisposition to act) and practices (actions) in relation to the objectives assigned to the IPs. To assess the ʻnetworking of local stakeholdersʼ criteria, we focussed on two sub-indicators: the exchange network, covering all the people/institutions with which IP farmers interact (for seeds, fertilizer, labour force, advice and information) and the frequency of exchange, characterizing the number of interactions per farmer (Lemieux, Reference Lemieux1999).
Criteria and monitoring indicators of CA innovation platform functioning.
One of the activities of the IPs were the design and implementation of farm experiments to test the effect of CA cropping systems on yields, pH, soil organic carbon, labour, gross margin and work productivity compared to conventional cropping systems. Grain yield and aboveground biomass at harvest were measured on CA and conventional fields of 58 farmers. Four squares of 12 m2 were randomly selected in each field. Within these squares, aboveground biomass was cut, grains were separated from straw and fresh grains and straw were weighed. Subsamples were taken and oven dried (at 70 °C for 72 h) to determine dry weight. Dry:fresh weight ratios were used to estimate the overall dry biomass based on fresh samples.
Before the onset of the rainy season in 2014, three soil samples were collected with an auger at a depth of 0–20 cm along the main diagonal of each field of farmers using CA. Samples were combined to form one composite sample of 500 g soil per field. The composite samples were air-dried and ground to pass a 2 mm screen in order to carry out chemical analyses. Soil pH was measured potentiometrically in a 1:2.5 soil water suspension following the procedure outlined by MacLean (Reference McLean, Page, Miller and Keeney1982) and organic carbon (C) was determined by the Walkley and Black method (Walkley and Black, Reference Walkley and Black1934). Soil analyses were conducted by the Soil-Water-Plant Laboratory of Farako-Bâ, Agronomic National Research Center of Burkina Faso.
The labour input (based on recorded time and people involved) was recorded for the different field operations. Gross margin was the revenue after incurring the direct costs (seeds, fertilizers and labour) and work productivity was the ratio between yield and labour per hectare.
Statistical analyses (one-way ANOVA) were done by using XlSat software and means were compared by Newman and Keuls’ test at p < 0.05.
RESULTS
Design of the IPs
Step 1: Diagnosis of existing forms of organization
Diverse forms of organization co-existed in the three villages, including self-help groups (exchanging labour) and service provision groups. There were also traditional organizations grouping male household heads or lineages under the authority of traditional leaders, which notably played a role in the management of conflicts at the village territory level. These organizations were experiencing some difficulties in mobilizing members due to a rise in individualism.
The exogenous organizations identified were the product of various external dynamics. Public authorities were behind the creation of village development councils (VDC) and chambers of agriculture (CoA). These organizations had formal structures and sometimes even action plans, but their functional activities were low (VDC) to non-existent (CoA) because they did not have their own funding, members lacked training, and they were polarized by diverse power struggles. Farmer organizations were started by value chain promotion schemes. In Sindri and Yilou, many were characterized by a lack of initiative, a low level of technical equipment and low skill sets among members, and were structured around cereal, legume and vegetable crops mainly destined for home consumption rather than markets. In Koumbia, cotton farmer organizations had comparatively more equipment and technical partnerships. Village Coordination Committees (VCC) and farmer field schools (FFS) were initiated by research and development (R&D) projects, grouping farmers from diverse social and ethnic backgrounds (government extension agents and cotton extension agents were also involved in the VCC). Although they differ somewhat in terms of their history and operations, both VCC and FFS functioned as spaces or fora for interaction between farmers, the research teams which launched them and public extension services, in order to learn new technologies.
Beyond the solidarity between members, one of the main strengths of these fora resided in the desire of members to maintain a partnership with research teams to continue to test new techniques and gain access to training and agricultural inputs. However, those in charge of existing fora pointed out numerous difficulties and concerns such as a lack of interest and availability of members to participate in experimental activities without some form of material or financial compensation, failure on the part of members to respect commitments and questions regarding the legal status and sustainability of these fora.
Step 2: Definition of the innovation platform framework
The diagnosis demonstrated the strong interest of existing forms of organization in learning via the new systems (VCC, FFS), managing shared resources and related conflicts (traditional authorities) and promoting access to inputs and markets (farmer organizations), which are objectives coherent with those expected of the IPs (the co-design of technical references and the exploration of renewed rules governing stakeholders’ access to land and crop residues).
Considering the complexity of CA and the two main objectives under pursuit, the research team opted for an IP structure consisting of two bodies, a technical body and an institutional body (Supplementary Figure S1). The technical body, composed of farmers from existing R&D mechanisms, government agriculture extension services and the research team, was meant to generate CA technical references by proposing, testing and assessing cropping systems based on CA principles. The institutional body, also called the ʻforumʼ, was meant to facilitate the coming together and interactions of all stakeholders (technical body actors, private sector actors and traditional authorities) on a voluntary basis. The interactions between actors gathered within the forum aimed to identify and engage organizational changes needed to facilitate access to crop residues and land, and also to share with others insights of characteristics and performance of CA systems under development. The forum also aimed to lobby political decision makers at the village and communal level to support experimentation of CA.
The facilitators of the IPs were the elected leaders of the existing R&D mechanisms and took into account their past experience in experimenting with new cropping systems. The research team had many functions: support the facilitators and the monitoring of the process by a sociologist of CIRDES, technical support and production of scientific knowledge by the agronomists, the geographer and the animal production scientists of CIRDES and ACT, methodological support by NRI researchers and the coordination of the whole project by CIRAD and ACT researchers.
The research team proposed an informal mode of coordination within the IPs so that the IPs could function in a flexible manner (Nederlof et al., Reference Nederlof, Wongtschowski and Van der Lee2011). The decision was motivated by lessons learned from past experiences with more formal modes (within existing R&D mechanisms) which met with mixed results (Koutou et al., Reference Koutou, Vall, Chia, Andrieu, Traoré, Vall, Andrieu, Chia and Nacro2012).
Furthermore, certain activities which had proven successful in the past within R&D mechanisms were retained, such as the organization of training sessions, guided tours and inter-village exchanges focussing on innovative techniques. New tools to facilitate discussion and knowledge sharing were proposed, such as a map to identify areas where CA could be introduced and simulation models to assess the effect of different levels of CA adoption at the farm or village scale (Djamen et al., Reference Djamen Nana, Andrieu, Zerbo, Ouedraogo and Le Gal2015).
Step 3: Stakeholder validation of the IP framework and planning activities
All of the actors who had attended the workshops seemed to share a common vision of the IP as a space for coming together and exchanging information and experiences about CA. The stakeholders furthermore proposed ways to contribute to the functioning of the IP coherent with their respective conventional roles, for example, the public extension services offered to monitor the implementation of cropping system trials based on CA. However, each type of stakeholders also had his or her own, at times, opportunistic vision of the IP. For example, men saw it as an alternative to acquire inputs and agricultural equipment in a context of low soil productivity, whereas women as a way to add value to their production and commercial activities involving milk, shea butter, peanut butter and cowpeas.
The IP workshops also provided an occasion for some stakeholders to express their doubts, particularly in relation to the CA technical model, regarding the feasibility of obtaining permanent soil cover and of consequently modifying rules on access and management of crop residues. Some farmers pointed to the practice of common pasture in the dry season, and the difficulty of ensuring sufficient cover on large fields, fencing and keeping watch over their fields to prevent grazing or the theft of crop residues. Likewise, some public extension agents expressed doubts regarding the possibility of intensifying existing systems through CA. Lastly, government environment agents were reticent about the inclusion of herbicides in some of the CA systems tested.
Despite these various expectations and specific doubts, the stakeholders were able to define an action plan structured around activities involving training, field trials, study visits and trips, and work on a land charter (Table 3). This plan included the responsibilities shared between different members of the IPs. The heads of the VCC in Koumbia proposed innovative instruments such as a field trial competition which would award prizes to farmers who had best conducted their trials and who could explain the principles of CA to stimulate respect for the collectively defined technical specifications. Communication and awareness raising activities were proposed by farmers to promote better understanding of CA among other villagers (Table 3).
Action plans and activities carried out by the IPs in Koumbia, Sindri, and Yilou.
Monitoring and evaluation of innovation platforms
The activities carried out from the end of 2011 to the end of 2014 broadly followed the plan that had been established collectively (Table 3). Under the initiative of the research team, some proposals made in one village were reproduced in others; this was, for example, the case of the CA field trial farmer competition.
Effectiveness of innovation platforms in adapting CA principles
Prior to the beginning of CA experiments, most farmers in Yilou and Sindri and part of the farmers in Koumbia were implementing indigenous farming practices, which were in line with CA principles. However, these indigenous practices were not always implemented to achieve sustainable land management, but rather to face climatic and socio-economic constraints. For instance, direct sowing was an answer to erratic rainfall and lack of tillage equipment (Djamen et al., Reference Djamen Nana, Andrieu, Zerbo, Ouedraogo and Le Gal2015). Crop association was also widespread, reflecting both the increasing pressure on land and a strategy to diversify production systems in the face of climate risks.
The farmers played an active role in the definition of new CA cropping system adapted to local conditions. In the first year, members of the Koumbia IP technical body defined specific objectives for the cropping system to be tested: to jointly produce food and feed. They reviewed different modalities of CA principles suggested by the research team (over 10 cropping systems consisting of various combinations of main crops, associated leguminous crops and soil tillage patterns), and selected direct sowing of sorghum associated with pigeon pea, with a soil cover based on residues from other fields or brushwood. This combination was selected due to the food and fodder potential of the two crops, their resilience under climate hazards and the good soil cover in the dry season provided by sorghum stems, which are not consumed much by animals. This cropping system and a conventional system corresponding to soil tillage and single cropping of the main cereal were implemented by 40 farmers selected by the IP on plots of 0.125 ha. However, in the second year, the farmers preferred to test maize combined with cowpea because the production of sorghum had been sparse compared to maize in the first year, and the pigeon pea flowered late compared to cowpea, which rendered it necessary to keep closer watch over the crops to avoid grazing by animals.
Given the difficulty of conserving biomass during the dry season, experiments were set up to test the effect of different levels of soil cover on yields and infiltration. Farmers requested improvements to facilitate the task of sowing, which was more arduous on CA plots due to the absence of ploughing and the presence of mulch. In an area where agricultural activities were already highly mechanized, the proposal of direct manual sowing initially appeared to be a step backwards. The animal drawn direct seeder was consequently proposed by the research team, which motivated farmers who viewed this as a necessary support measure to allow them to practice CA. The farmers remained interested in this system in the third year but required more training on the use of the direct seeder. Its condition of use with three different disks was consequently analysed by the research team. Although CA plots did not result in higher yields in comparison to the control plots in this study site, improved physical and chemical soil characteristics on CA plots could, in the medium term, lead to improved yields (Table 4).
Some comparative performances of conventional and CA systems in Yilou, Sindri, and Koumbia.
Different superscript letters (a, b) per village indicate significant differences (Newman and Keuls’ test at p < 0.05) in the mean value.
In Yilou and Sindri, the initial assessment of existing cropping and farming systems led to the identification by the research team of one potential CA cropping system consisting of cereal (sorghum and millet) grown under mulch of cereal straw and in rotation with leguminous food or fodder crops. The application of herbicides was also considered as an option for weed management. It was tested by 38 farmers selected by the IPs on plots of 0.5 ha and compared to the conventional system corresponding to soil tillage and sole cropping of the main cereal without herbicides. However, participatory annual assessments and learning sessions in the IPs highlighted challenges and shortcomings of the CA cropping system tested. Iterative discussions with members of the IPs therefore led to the proposition of two similar CA cropping systems based mainly on the implementation patterns of crop diversification, the third principle of CA. The two CA cropping systems included: (i) cereal (sorghum or millet), directly sown under mulch of crop straw, intercropped preferably with leguminous food crop for farmers facing land constraints, and (ii) cereal, directly sown under mulch of crop straw, intercropped or in rotation with leguminous fodder crops for farmers with larger farms and livestock. Furthermore, several existing or new species of leguminous crops presenting specific or combined functions including food, fodder and soil fertility were identified by the research team and proposed to farmers for intercropping or rotation in CA cropping systems. The majority of the farmers opted for leguminous food crops (mainly cowpea and groundnuts that can more easily be found in the local market and ambérique – Phaseolus aureus), leaving leguminous fodder crops (mucuna and dolichos) to farmers who keep livestock, in particular cattle.
Farmers’ perceptions, attitudes and CA practices
In the three study sites, farmers’ attitudes towards each principle of CA evolved in a positive direction (Figure 1) with particularly a high percentage of farmers prone to adopt intercropping, which is illustrated by the following remarks: ‘intercropping allows for the production of two commodities on the same field’ or ‘with intercropping, we have maize as well as fodder for the animals’. Farmers’ uptake of CA practices also evolved positively, particularly practices related to the soil cover and conservation of crop residues (Figure 2). Farmers realized the added value of crop residues and banished their burning: ‘when it rains, soils now remain humid’ or ‘we did not know that the straw could be used in the compost pits to fertilize the soil’.
Willingness to adopt the principles of the CA technical model by members of the IPs in Koumbia, Sindri and Yilou at the end of the project.
Changes in practices by members of the IPs in Koumbia, Sindri and Yilou: (a) management of crop residues for soil cover, (b) soil tillage pattern and (c) cropping pattern.
In Yilou and Sindri, where the CA technical model was introduced earlier (2009), there was a favourable evolution in practices with a 16% increase in the adoption rate of mulch between 2012 and 2014 (Figure 3). The testing of CA cropping systems enabled farmers to improve their understanding of the usefulness of CA principles and improve their technical mastery, fostering adoption throughout the project life time. The evolution curb of area under CA was similar to that of mulching confirming that keeping crop residue on the field is the main challenge for the adoption of CA in this context. Furthermore, the area under direct sowing tended to stagnate and even decrease as farmers understood that the implementation of direct sowing without soil cover in semi-arid area may rather lead to a decrease in production.
Evolution of the adoption of CA and CA principles by members of the IPs in Yilou and Sindri.
Although in Sindri the IP farmers reported they would apply CA to a higher area than that noted by non-IP farmers, in Yilou they reported equivalent areas and in Koumbia they reported they would apply CA to a smaller area than that noted by non-IP farmers. In this last case, this may be because the IP farmers had gained more in-depth knowledge of the difficulties involved in implementing CA through their own experiments, and thus of the need to proceed in stages to successfully adopt CA.
Networking of stakeholders
The IPs attracted a growing number of farmers to their meetings. A comparison of the local social networks at the start and the end of the project, however, shows that the stakeholder network in Koumbia did not change much, except for the introduction of three external stakeholders (two training institutes and ACT). This can be explained by the fact that the VCC were already inclusive, fostering a strong network among diverse stakeholders (Figure S2). In Sindri and Yilou, new linkages between farmers and other stakeholders were established during the process of implementation of IPs (Figure S3). The IPs also improved the frequency of interaction of farmers with other stakeholders in all three villages (Figure 4); the strengthening of stakeholder interaction was one of the IPs’ declared objectives. The increased interaction with the research institutes CIRDES and INERA can be explained by the fact that they constituted the research team of the ABACO project. The IPs also increased interaction with other stakeholders, in particular the input suppliers, council, village chief, credit bank and artisans, reflecting the effect of the installation of an institutional body of the IP.
Frequency of interactions between farmers and other organizations and actors in the three villages.
Evolution of perceptions, attitudes and practices in relation to land and crop residue management
The IPs also resulted in changing perceptions and attitudes regarding crop residues; initially more than 70% of farmers considered crop residues to be public goods that could not be appropriated. During the fora of the institutional body, the participants discussed rules of access to crop residues with reference to difficulties noted by pilot farmers in the technical body regarding the conservation of straw and biomass produced by cover crops, a necessary condition for the success of CA trials. In Koumbia village, where a local land charter for the governance of access to the natural resources of the village territory had been established in 2012 but never implemented, discussions focussed on putting the charter into operation (Karambiri et al., Reference Karambiri, Andrieu, Chia, Laurens and Diallo2015). The discussions led to the identification of the obstacles to the effective application of the charter, such as the absence of line items in the village budget that would allow various conflict management commissions to operate, fear on the part of the town council to lose the support of voters and a lack of understanding of this instrument by some villagers. The suggestions proposed to overcome these obstacles included raising awareness and training of local stakeholders in the definition of official opening and closing dates of the common pasture, renewing land conciliation village commissions and convoking annual cooperation meetings between farmers, the town council, and livestock breeders.
In the villages of Sindri and Yilou, the discussions focussed more on developing new management rules for crop residues at the individual level inspired by the tested practices by the IP technical body to improve the level of soil cover, such as cutting down stems after the harvest. In Sindri, eight rules were developed such as the obligation to inform the owner of the field before collecting crop residues, the prohibition of collecting crop residues in degraded areas and in gravel areas, or the allowance of their collection in low-lying areas and ravines. Three of these rules specifically defined the use of the crop residues: ‘lay down stalks produced in sealed and crusted bare soils locally called zipéllé after the harvest’; ‘lay down stalks produced in low-lying areas and let animals feed on leaves’; and ‘use the stems of millet, the piliostigma pods (bagna), the stems of sesame and sorrel for the potash’. In Yilou, six rules were developed. The common pasture was not prohibited as long as the owner is notified; however, the collection of residues on degraded areas (zipéllé) was prohibited. Four of these rules addressed the use of crop residues such as the prohibition of their stealing or burning, the systematically cutting down of the stems after the harvest, or the collection of thin stems to feed to livestock.
DISCUSSION
To what degree are innovation platforms relevant for the adaption of CA principles?
The research team initiated the process of building and putting into operation the village IPs to address a relative lack of experience with CA and multi-stakeholder approaches in the study area, a lack of appropriate technical references for CA, and the relative weakness of local organizations. The point of departure largely influences the objectives assigned to the IPs, the type of platforms built (in this case with a focus on experiments and knowledge generation and sharing/dissemination), and their operations (Schut et al., Reference Schut, Klerkx, Sartas, Lamers, Campbell, Ogbonna, Kaushik, Atta-Krah and Leeuwis2015; Swaans et al., Reference Swaans, Cullen, van Rooyen, Adekunle, Ngwenya, Lema and Nederlof2013). Here, the main objective for the established IPs was to support a participatory research and learning around CA to address the complex technical and organizational challenges at the field and village levels, in a context where crop residues are an open-access resource during the dry season. This objective, largely promoted by the research team, joined a local demand for alternative farming systems to decrease soil depletion and low productivity. Thus far, the results obtained by the three village IPs during their first three years of existence have largely been positive, despite the low productivity of CA systems compared to conventional systems in one of the study site. As expected, the IPs have been an effective space for the joint design, testing and discussion of new cropping systems and crop residue management modes. They have also been a space for training, emulation and networking of stakeholders to address challenges related to crop residue management and land access at a higher level than the field or the farm where CA research programs generally focus on (Dugué et al., Reference Dugué, Djamen Nana, Faure and Le Gal2015).
Farmers’ perceptions of CA have evolved favourably, CA practices are better mastered and organizational changes (for example, land charters) have been initiated to facilitate its implementation. In other terms, IPs were relevant spaces to trigger learning and social processes needed for the transition to new agricultural systems. Facilitation plays a key role in the positive outcomes of IPs (Swaans et al., Reference Swaans, Cullen, van Rooyen, Adekunle, Ngwenya, Lema and Nederlof2013). Here, the facilitators were the leaders of the existing R&D mechanisms that were already known and respected in the villages. These facilitators had developed facilitation skills in previous R&D projects, and had strong and wide personal networks that allowed the PIs to gain quick legitimacy. Swanss et al. (Reference Swaans, Cullen, van Rooyen, Adekunle, Ngwenya, Lema and Nederlof2013) showed that the function of facilitator (also called innovation broker) is often fulfilled by those who instigate platform processes. Here, the close relationship between the facilitator and the scientist in charge of monitoring the process led in fact to the emergence of a binomial of innovation brokers that jointly catalysed the process: the former linking farmers to other local stakeholders and the latter linking the farmers to other members of the research team.
Some argue that when IPs achieve their objective, they can consequently be dismantled (Schut et al., Reference Schut, Klerkx, Sartas, Lamers, Campbell, Ogbonna, Kaushik, Atta-Krah and Leeuwis2015). Here, these IPs achieved their initial objectives. But to sustain positive impacts over the medium and long term, including the sustainable adoption of CA systems, it is deemed necessary to keep supporting the organizational changes that have been made and particularly the implementation of the new rules of access to land and crop residues. In this study, the research team provided most of the financial resources for the functioning of the three IPs. Supporting local leadership and the reinforcement of local stakeholders’ capacity for innovation is consequently a key feature in a new lifecycle of the IPs that means modifying the initial objectives and operations of these IPs. Other cases demonstrated that CA can be developed successfully in the context of multi-stakeholder, ʻbottom-upʼ processes in which farmers had and/or still play leading roles (Triomphe et al., Reference Triomphe, Goulet, Dreyfus, Tourdonnet, Bourrigaud and Sigaut2007). To avoid the risk of the IPs stagnating, one must identify among the ʻlocalʼ IP members those who are most likely to assume leadership (farmers organizations, public organizations – notably extension services – or NGOs with a long term local or national level presence) and are sufficiently interested in continuing the IPs. These stakeholders would need to obtain a sustainable source of resources to fund their routine operations (cost of periodic member meetings and exchange visits for experimental trials), including through self-financing activities (Nederlof et al., Reference Nederlof, Wongtschowski and Van der Lee2011).
Market orientation of IPs
Throughout the ABACO project, an important question lingered in the background: Can the IPs be sustainably organized without establishing a strong market link? To attract more farmers and private-sector stakeholders, IPs are commonly designed around a value chain to improve market functions and relationships, with expected short-term benefits such as improved production and income through technical and organizational changes, improved networking among stakeholders and platform autonomy (Devaux et al., Reference Devaux, Horton, Velasco, Thiele, López, Bernet, Reinoso and Ordinola2009; Tenywa et al., Reference Tenywa, Rao, Tukahirwa, Buruchara, Adekunle, Mugabe, Wanjiku, Mutabazi, Fungo, Kashaija, Pali, Mapatano, Ngaboyisonga, Farrow, Njuki and Abenakyo2011). In our case, market considerations, while not entirely absent, remain for the moment marginal in relation to the IPs’ operations. This does not seem to be a problem as the IPs have succeeded in attracting a growing number of producers without financial incentives. The kind of systemic change represented by CA (e.g., the management of natural resources during the grazing season) does not only involve market factors, but rather a shift in communal management of local natural resources which can be initiated and supported by a local IP.
Other objectives of IPs can be, for example, the testing of new cropping systems or the development of land charters. It is likely that farmers will soon want the IPs to progressively give greater consideration to market access to sustain their continued participation. This may include placing greater emphasis in the near future on supplying inputs needed for the implementation of CA (purchase of herbicides, cover crop seeds and seeders), providing access to certain services (such as agricultural advice or contract work), or marketing of expected cereal production surpluses. The fact that the producers’ organizations with links to value chains are IP members should facilitate such evolutions in objectives and activities. It is critical to also identify and address constraints within the ‘enabling environment’ (Hounkonnou et al., Reference Hounkonnou, Kossou, Kuyper, Leeuwis, Nederlof, Röling, Sakyi-Dawson, Traoré and van Huis2012) as agricultural policies often focus on single commodity crops and are therefore not always conducive for the implementation of complex farming systems such as CA. A medium-term strategy would thus be to develop one or several provincial or national level platforms, as a complement to local platforms, to influence institutional and policy changes which could promote CA. These may include, for example, training for extension services to accompany the CA transition, setting up subsidies or financial incentives for good practices, changes in rules of land access for migrants and women or, more generally, innovation policies favourable to family farming (Devaux et al., Reference Devaux, Horton, Velasco, Thiele, López, Bernet, Reinoso and Ordinola2009; Kilelu et al., Reference Kilelu, Klerkx and Leeuwis2013).
CONCLUSION
This article analysed the design and implementation process of local multi-stakeholder innovation platforms piloted by a research team, aiming to promote the co-design of local technical references for CA and revise the rules of interaction between stakeholders in three villages in Burkina Faso. These IPs were structured into two components: a technical body to test new CA cropping systems and an institutional body to gather local actors into a network to address the challenges posed by CA, such as access to crop residues and land. The IPs were engaged in a wide range of activities (negotiation of protocols, experimenting, annual assessments, meetings and competitions for the best pilot farmers), whereas the research team steering the process provided appropriate methods and facilitation tools.
Three main lessons can be learned from this process: (1) build on existing inclusive organizations already involved in activities coherent with those expected of the IPs helped to quickly gain legitimacy and an audience to promote dialogue around new CA agricultural systems; (2) the mode of functioning of IPs based on technical and institutional bodies gave stakeholders the opportunity to address not only technical issues as in conventional research on CA but also to enable new linkages between local stakeholders for the identification of solutions to govern the access to crop residues, which had up until then been considered to be a common good; (3) IPs are consequently appropriate instruments to promote changes in farmers’ perception, attitudes and management practices of crop residues. However, various improvements should be made in the near future if these platforms are to outlive the framework and funding of the project through which they were created. In particular, this will involve reducing the platforms’ dependence on the research team. This may involve a closer link with economic objectives and more clearly reinforcing the capacity of local actors to innovate, notably the extension services, which are the real ‘innovation facilitators’. This will also allow the platforms to address innovations other than CA. Another challenge will be to eventually complement the village-level mechanism with the development of provincial and national platforms, which will allow key institutional and economic questions to be addressed.
Acknowledgements
This research received financial support from the ABACO project (Grant agreement no. DCI-Food 2010/230-178) funded by the European Commission. We thank the members of the innovation platforms in Koumbia, Sindri and Yilou, and also thank Grace Delobel who did the language editing of this manuscript.
SUPPLEMENTARY MATERIALS
For supplementary material for this article, please visit http://dx.doi.org/10.1017/S0014479716000636
