Background and objectives: Jaɓnde is a locally adapted ration formulation tool for resource-constrained dairy systems in sub-Saharan Africa. The version used in this study was the Excel-based prototype installed on a laptop. A mobile version, now available for free on the Google Play Store and online (https://jabnde.cirad.fr/), is under testing. This preliminary, supervised field validation pursued three objectives: (i) to compare the technical and economic performance of zebu and crossbred cows under dairy farmer-implemented rations; (ii) to assess the implementation fidelity and the predictive performance of Jaɓnde; and (iii) to assess dairy farmer satisfaction with the use of Jaɓnde. Methodology: The tool was tested by 40 volunteer dairy farmers in the Bobo-Dioulasso milkshed (Burkina Faso) on 82 lactating cows (65 zebu, 17 crossbred). Implementation fidelity was assessed from the mean difference (predicted ration, R2 - actual ration, R3), and predictive performance from robust linear regression, reported overall and by cow type using R², the slope, and error indicators (mean absolute error, MAE; root mean square error, RMSE; mean bias error, MBE), complemented by Bland-Altman analyses. A structured survey assessed the satisfaction of dairy farmers. Results: Dairy farmers implemented R3 ration close to R2 (mean differences -0.37 to +0.04 kg GM/cow/day across feed types). Predicted and actual values were strongly associated for most variables (R² > 0.70), though weaker for milk production among crossbred cows; bias remained limited (MBE = -0.27 L/d/cow for milk production). Agreement on milk production was the most independent performance indicator. Crossbred cows produced more milk (10.6 vs 1.2 L/d/cow), whereas the relative change was greater in zebu cows (33%vs 8% increase). Over 90% of dairy farmers reported increased production and were satisfied. Contributions: This first field assessment is encouraging: under supervised conditions, farmers implemented Jaɓnde's rations, predictive performance was promising for milk production, and satisfaction was high. Confirming its causal contribution and robustness will require controlled designs across more farms, seasons, and management systems.

In tropical regions, subject to significant fluctuations in feedstuffs, the body condition score (BCS) is a relevant indicator for monitoring body reserves status of farm animals. However, the most used BCS grids comes from temperate conditions and still not adapted for tropical species, which limits adoption and usefulness. The current work presents for the first time an original, low-cost, standardized BCS assessment system, suited to a large spectrum of tropical farm animals (zebu and/or crossbred cattle, buffaloes, camel, sheep, goats, horses and donkeys). Based on a rigorous set of uniform criteria for practical, easy-to-use on field conditions, the animal model used to calibrate and validate each grid is the adult female (except the male for horses and donkeys). A six-point BCS grid system is proposed (i.e., from very emaciated -0- to overweighed -5- body conditions), based on visual interpretation of the back and right-lateral sides of each animal. The overall BCS assessment criteria is built on three major anatomical regions (hindquarters; thorax and abdomen; shoulder and neck), and ten related anatomical landmarks, which provide a unique consensual global interpretation of the back and right views. On-field tests of the proposed BCS grids for each animal species revealed acceptable reproducibility with regard to the most conventional, established gold-standards BCS methods (r2 = 0.67–0.99). The proposed harmonized BCS is relatively simple and must facilitate the adoption of regular, accurate BCS assessment by adequately trained operators linked to tropical animal production environments.

In Burkina Faso, the growing demand for milk prompts dairy farmers to pursue solutions that improve their production while maintaining profitability and sustainability. This study aimed to analyse the link between the level of agroecological transition achieved by dairy farms in the Bobo-Dioulasso milkshed area and their multidimensional performance. To address this issue, 204 dairy farms were assessed in the Bobo-Dioulasso milkshed area according to their level of progress in the agroecological (AE) transition, using the Holistic Localised Performance Assessment for Agroecology (HOLPA) tool. A typology of the dairy farms was produced according to their level of progress in the AE transition. Averaged scores for all questions in the HOLPA tool relating to the 13 principles of agroecology were used to rate each farm on a 5-point Likert scale, from 1 (Non-AE) to 5 (Very Strong AE). Performance was then compared according to these score levels. Results indicate that the dairy farms studied are mostly at the beginning (Non-AE: 7%, Weak AE: 54%, Moderate AE: 33%, and Strong AE: 6%) of their AE transition. Our findings suggest that dairy farms which have adopted more advanced AE practices demonstrate improved environmental, social and economic performance compared to conventional or less transitioned farms. The most AE dairy farms stand out on several key performance indicators. In terms of environmental performance, the most AE dairy farms achieve higher scores for crop diversity and natural resource and land management. In terms of social performance, the most AE dairy farmers share more knowledge about agroecology and natural resources, report higher levels of household satisfaction, and perceive trade associations as efficient. In terms of economic performance, the most AE dairy farms enjoy more stable and diversified income streams, which may reflect their greater resilience. With regard to agronomic performance, the most AE dairy farms perform better in terms of organic manure and fodder production. However, heavy reliance on labor and low crop productivity (e.g., for maize, sorghum, millet, rice, and cowpea) have been identified as key challenges to the AE transition and the sustainability of dairy farms. The findings of this study justify the implementation of several policy measures to support farmer training in AE on fodder and organic manure production, and natural resources management, to guaranty a better access to forages seeds and agricultural equipment to manage manure production, to increase the skills of farmers community in collective natural resources management in order to strengthen and accelerate the AE transition and enhance the resilience of the dairy sector in Burkina Faso. This study provides robust evidence to help improve public policy, guide dairy industry strategies, and support the agroecological transition in the Bobo-Dioulasso milkshed area.

Description of the subject. Agriculture-livestock integration is one of the strategies adopted by agropastoralists in the North Cameroon region to reduce the risks associated with demographic pressure and climate change, in order to improve their productivity. Objectives. The aim of the study was to characterize the systems and practices for managing and recycling animal and plant co-products on agropastoral farms in northern Cameroon, as this is one of the keys to agroecology in agropastoral systems. Method. An in-depth survey was conducted on 108 selected farms. The data were analyzed using multivariate analysis (PCA) followed by classification (CAH). Results. Characterization according to farm structure, co-product management practices and technical performance enabled us to identify four classes of farm, namely small farms mainly focused on crop production (class 1), farms specializing in livestock farming (class 2), medium-sized farms combining crop production and livestock farming (class 3) and very large agro-pastoral farms (class 4). On average, farms used 28% of plant by-products to cover 25% of their fodder requirements (straw, tops), and 23% of animal by-products to cover 35% of their organic manure requirements (animal waste). Conclusions. This study showed that the recycling of co-products in agro-pastoral farms is low but increasing.

Description of the subject. Agriculture-livestock integration is one of the strategies adopted by agropastoralists in the North Cameroon region to reduce the risks associated with demographic pressure and climate change, in order to improve their productivity. Objectives. The aim of the study was to characterize the systems and practices for managing and recycling animal and plant co-products on agropastoral farms in northern Cameroon, as this is one of the keys to agroecology in agropastoral systems. Method. An in-depth survey was conducted on 108 selected farms. The data were analyzed using multivariate analysis (PCA) followed by classification (CAH). Results. Characterization according to farm structure, co-product management practices and technical performance enabled us to identify four classes of farm, namely small farms mainly focused on crop production (class 1), farms specializing in livestock farming (class 2), medium-sized farms combining crop production and livestock farming (class 3) and very large agro-pastoral farms (class 4). On average, farms used 28% of plant by-products to cover 25% of their fodder requirements (straw, tops), and 23% of animal by-products to cover 35% of their organic manure requirements (animal waste). Conclusions. This study showed that the recycling of co-products in agro-pastoral farms is low but increasing.

In Sub-Saharan Africa, dairy value chains’ stakeholders face many challenges and have expectations for change. Step-by-step innovation design methodologies and multi-stakeholder innovation platforms are implemented to drive changes desired by stakeholders. We assumed that combining these two approaches would reinforce the potentiality of achieving the changes. To the best of our knowledge, the specific mechanisms and actions involved in such a combination are poorly documented. This study contributes to fill this gap by reporting on modalities of engagement, collaboration, and change generation with stakeholders of dairy innovation platforms deriving from a step-by-step innovation design approach that is embedded within an overall loop-structure dynamic and accounting for three levels of stakeholders’ engagement. We applied this step-by-step approach as part of the “Africa-Milk project” on ten dairy innovation platforms located in four African countries (Senegal, Burkina Faso, Kenya, and Madagascar). The approach was led by a core team and applied adaptively across the various innovation platforms, according to both their organizational context and objectives. In this paper, we captured the lessons learned along the key implementation stages of the approach (i.e., engagement, action, and assessment) and regarding the type of stakeholders involved. Our results show that the initiation of the engagement highly depends on the pre-existence of an innovation platform. The action stage proceeds then through either cascading actions or parallel actions. Finally, the outcome assessment stage enables to identify different types of changes induced by the approach (i.e., changes in practices, interactions, capacities, and opinions). Owing to its adaptability, the overall loop-structure of the approach enables practical adjustments and reflexivity to best meet the needs of innovation platform stakeholders. This study paves the way to implement co-design of innovation approaches to broader multi-stakeholder platforms involved in agri-food system transformations.

The need for transformation of food and agriculture systems to be aligned with sustainable development goals is widely acknowledged. Evidence from many parts of the world shows that agroecology, which considers the social and environmental performance of agricultural systems along with economic aspects, is helping farmers transition to sustainable agricultural systems. However, there are demands from national, regional and international food and agriculture planners and funders for evidence that agroecology can work at scale. Providing that evidence requires understanding how farmers use environmental factors when selecting agricultural practices, which is poorly documented. This study contributes to filling this gap by reporting how environmental factors are important in farmers’ decisions related to adopting agroecological practices. Qualitative and quantitative data from 239 key informants, a survey of 5025 farms, 85 focus group discussions with farmers and five participatory cross-benefit analyses in eleven case studies across eight African countries were used. We show that farmers use information on and perceptions of a wide range of environmental variables and processes when assessing the usefulness of agroecological practices. In most cases, farmers cited environmental factors more frequently than economic reasons for choosing to use agroecological practices. Most of the environmental factors articulated by farmers were components of the local or farm environment that were connected to their livelihood, including aspects of soils, water, microclimate, pests and diseases, other animals, and vegetation. Intrinsic and relational environmental values such as conservation of biodiversity, long-term maintenance of land quality and aesthetics were also important. These findings show first that providing data on environmental roles of agroecology will always be challenging because of the breadth of factors that are important. Secondly, viability or usefulness of an agroecological practice is not a characteristic of the practice alone, but also of the context in which it is used. Third, impact analyses of agroecological interventions cannot be confined to consideration of a few globally comparative indicators but need to include the context specific environmental factors that farmers care about.

IntroductionBurkina Faso's dairy farms, whether extensive or semi-intensive, face major sustainability challenges. Introducing agroecological technologies offers an opportunity to optimize the use of local resources and enhance farm resilience and productivity. This study presents an innovative approach that tests and assesses, in co-design with farmers, the integration of two major agroecological technologies that underpin crop–livestock integration.MethodsTo assess the impact of these agroecological technologies, dairy farmers of the Bobo-Dioulasso milkshed implemented on-farm trials involving: (i) the introduction of dual-purpose fodder cereals (maize and sorghum) and legumes (Cowpea and mucuna) for feeding lactating cows, and (ii) the recycling of livestock and crop co-products to produce manure in covered manure pits. We employed a four-step methodological approach for co-designing agroecological innovations with 43 dairy farmers: diagnosis, co-design and experimentation, assessment, and identification of induced changes. Data from experiments and participatory assessment workshops were analyzed to compare dairy farm performance before and after the introduction of agroecological technologies. The first changes induced by these technologies were identified and characterized through participatory workshops.Results and discussionThe introduction of dual-purpose fodder crops into production systems led to improvements in on-farm milk production. Legume fodder storage increased from 356 to 518 kg DM/farm (+45%), while cereal fodder storage decreased slightly (1,388 to 1,091 kg DM/farm). The establishment of covered manure pits enhanced manure quality, with an average increase of 4,679 kg DM/farm. Milk production rose markedly, by over 80% for mixed cows (6.43 vs. 11.67 L/d/cow) and 12% for zebu cows (1.11 vs. 1.25 L/d/cow). Agroecological performance scores also improved particularly in diversity (38% vs. 94% after introduction of agroecological technologies), recycling (31% vs. 88%), and synergy (25% vs. 69%). These results demonstrate that a step-by-step integration of agroecological technologies can enhance their effectiveness and milk productivity, while strengthening the resilience of dairy farms, particularly in the face of climatic and economic challenges. Under conditions of prolonged dry seasons, irregular rainfall, and volatile prices of industrial inputs, the combination of these technologies ensures both the availability of fodder for livestock and manure for soil fertilization. Nevertheless, continued efforts are required to maintain and further improve this performance over time.

Agroecological transition (AET) of agrifood systems is a multidimensional process involving diverse stakeholders. Participatory innovation and co-creation of knowledge are also at the heart of agroecology. Deliberately designed Multi-stakeholder platforms (MSPs) are collaborative spaces conducive to driving this type of collaboration and change. However, how these processes are operationalized is poorly documented. This study contributes to filling this gap by highlighting how MSPs can be important vehicles for enabling AETs. As part of the CGIAR Initiative on Agroecology, Agroecological Living Landscapes (ALLs) were set up as MSPs specifically focusing on co-creation of agroecological innovations. By examining different ALL configurations in seven countries, we analyze MSP attributes that can enable and shape AETs at different stages or "levels" of the transition pathway. We find that MSPs can advance different levels of the transition pathway simultaneously by working on multiple innovations involving various stakeholder coalitions. MSPs navigate complex environments, with diverse stakeholders playing multiple policy and non-policy roles to support AETs. The results also illustrate the intersection between MSP configuration and AET objectives, such as policy change, agroecology market support, and scaling adoption. Finally, we discuss the non-linear nature of AET pathways and political economy considerations for stakeholder engagement in MSPs.

Aims: This paper shows the contribution of participatory research to upgrade the integration of legumes into cropping systems. Place and Duration of Study: The investigation is based on survey data collected in 2008 and 2011 (two reference years) and 2021 and 2022 (current situation) in the village of Koumbia in western Burkina Faso. Methodology: Data was collected through surveys and field measurements. The surveys were conducted with a sample of fifty (50) volunteer farmers. In 2008, 2011 and 2022, the same questionnaire was administered to all 50 farmers. Field measurements with GPS (Global Positioning System) were conducted on crop area. Results: The results show that the proportion of area under legumes crops increased by 77,38% after 15 years. Other legumes crops (Glycine max and Mucuna sp.) were introduced in addition to those previously grown (groundnuts and cowpeas) by farmers. Surface cropped in sorghum (35.36%) and millet (82.79%) declined while cotton and maize remain the main crops in the zone. The results on farmers’ perception show that they prefer single cropping of legume crops as their insertion modality in the existing farming systems compared to their association with cereals. Conclusion: The study shows that the role of legumes is increasing in the cropping system, and represents an opportunity to succeed in the agro-ecological transition, even though the intercropping of cereal/legume, which is put forward by the theoretical models of the agro-ecological transition, is still little practiced. The structuration of the legume sector can be a condition for achieving it.

The objective of this work was to carry out long-term monitoring of protein banks (PB) established under the real-low-input-farm conditions of Burkina Faso, characterized by a Sudano-Sahelian climate type with wide variations in rainfall. The PBs, planted in 2016–2017 with Leucaena leucocephala (LL) and Morus alba (MA) at a high density (20,000 plants/ha), were monitored from 2016/2017 to 2021 for their biomass (fodder) production, chemical composition and nutritional value of the edible fodder, and their impacts on soil carbon content. The average annual fodder production of LL for the 6 PBs (1, 2, 3, 4, 5, and 6) was 11.4 ± 2.12 t DM/ha/year, while that of MA was 3.8 ± 1.2 t DM/ha/year for the 3 PB (1, 2, and 3). Such differences are related to higher adaptive capacity for LL in such climatic conditions. Crude protein was higher in LL (19

This study aimed to examine the correlations between the agroecological status of farms and their viability in agro-sylvo-pastoral systems located in the savannah areas of West Africa. The methodology is based on farm surveys (to assess agroecological farm status) and on farmers' focus group discussions (to assess viability at the farm level). Agroecological farm status was established by considering 15 agricultural practices: 4 on by-products recycling; 4 on soil protection; 4 on industrial inputs limitation; 2 on smart use of natural resources; 1 on cropping diversification; An agroecological (Ae) scoring system was used to establish an Ae ranking by farm type. Farm level viability was established by men and women farmers' focus group discussions: Inventory of viability enhancing and viability-weakening factors; Rating intensity of each viability factor. A scoring system was used to establish Viability ranking by farm type. Four iconic farm types were considered: Small crop-oriented (T1); Medium-size crop-oriented (T2); Large crop-livestock oriented (T3); Livestock-oriented (T4). The findings demonstrate that the viability of agro-sylvo-pastoral systems is not solely dependent on their agroecological status. The ranking of the four farm types, from the first to the last position, is T4, T2, T3, and T1 for agroecology; and T2, T3, T4 and T1 for viability. Our study reveals that T1 farms in Koumbia are not highly agroecological or highly viable. Similarly, farms with the largest cultivated areas (T3) and herds (T4) are not the most agroecological or viable. The best balance between agroecology and viability is struck by T2 farms. We recommend promoting the development of T2 farms that demonstrate the best congruence between agroecology and viability as a model for T1, T3, and T4 farmers. Depending on the agroecological status, one farm could be more or less viable.

The objective of this work was to carry out a long-term monitoring of shrub fodder banks (SFB), established under the real-low-input-farm conditions of Burkina Faso, characterised by a Sudano-Sahelian climate type with wide variations in rainfall. The SFBs, which were planted in 2016-2017 with Leucaena leucocephala (LL) and Morus alba (MA) at a high-density (20,000 plants/ha), were monitored from 2016/2017 to 2021 for their biomass (fodder) production, chemical composition and nutritive value of the edible forage, and their impacts on soil carbon content. The average annual forage production of LL for the 6 SFBs (1, 2, 3, 4, 5 and 6) was 11.4 ± 2.12 t DM/ha/year while that of MA was 3.8 ± 1.2 t DM/ha/year for the 3 SFB (1, 2, and 3). Such differences are likely related to higher adaptive capacity for LL in such climatic conditions. The high nutritive value observed in both fodder resources agreed with available literature worldwide. Crude protein was higher in LL (19% vs. 17%), but dry matter digestibility (65% vs 77%) were superior in MA. Soil carbon increased during the five monitored years, which is an interesting result for these poor, low fertility soil conditions. Our results demonstrate the technical and environmental feasibility for establishing this low-input agroforestry technology in these conditions, as an agroecological alternative to overcoming feed shortages for local livestock farming. However, the successful management of a SFB relies on the selection of the right species, the protection provided, and the management during the establishment and fodder production periods, with suitable cutting frequency and appropriate post-harvest management.

Description of the subject. Considering the increasing price of animal feeds and synthetic fertilizers, agro-pastoral farmers of Western Africa, in Burkina Faso in particular, try to increase the use of Crops Co-Products (CCP: straw, tops, stalks) and Livestocks Co-Products (LCP: faeces) to feed animals and fertilize cultivated fields, however face challenges to do so. Objectives. Our objectives were to characterize the management and recycling practices of CCP and LCP at the farm level and to assess their contribution to the coverage of forage and manure needs. Method. Data collected by survey on 60 agro-pastoral farms in six villages in western Burkina Faso. On-farm management practices of CCP and LCP characterized by eight variables. Flows of CCP and LCP characterized at three levels: farm, crop and livestock units, and type of co-product management system. Results. Out of the ~21 tons of CCP available per farm, 23% are recycled as forage, and 77% are not or scarcely recovered by farms. Out of the ~24 tons of LCP available per farm, 40% are recovered for use as manure, 60% are scattered off-farm during grazing and mobility. Co-products recycling covers only 16% of farm needs in forage and manure. Conclusions. Co-products are poorly recycled, and their level of recovery depends on farm equipment, manpower and herd mobility. Facing this challenge, we are developing a farm-level co-products recycling/advisory tool to support farmers in the agroecological transition.

In Western Africa, agro-sylvo-pastoral systems are dominant and food demand is booming. To meet this demand, many farmers intensify the production with industrial inputs (mineral fertilizers, feeds, pesticides, herbicides). However, the price of these inputs is rocketing. To face this issue, some farmers reconsider crops, livestock, and tree synergies and by-product recycling to increase their production sustainably at a lower cost. The study aimed to characterize the diversity of Koumbia’s farming systems and to assess farming systems’ technical performance in an agroecological perspective. We surveyed 391 farms in the county of Koumbia (Burkina Faso). Considering 15 agricultural practices (4 on by-products recycling, 4 on soil protection, 4 on industrial inputs limitation, 2 on smart use of natural resources, and 1 on cropping diversification), a multivariate analysis (PCA+HAC) combined to an agroecology (Ae) scoring system (−15 to +15) based on 15 specific indicators (one/practice), we highlighted 3 agroecological farming systems clusters. These clusters are distributed along a gradient of agroecology intensity (Ae+: high degree of Ae, Ae+/−: medium degree, and Ae−: poor degree). Ae+ farms (Ae score: +3.0) group 17% of the farms, Ae+/− (Ae score: −4.5) group 58% of the farms, and Ae− (Ae score: −10.5) group 25% of the farms. Ae+ raise more livestock and recycle a higher rate of crop-livestock by-products in fodder, organic manure, and mulch. These recycling practices are facilitated by better levels of equipment for transportation and storage and soil water and crop residue conservation measures, including maintenance of the wooded park on the cultivated fields. This set of practices, which close better the agricultural system, produces a systemic effect which has a positive impact on yields and on the whole ecosystem. Our findings outline for the first time that crop-livestock synergies and by-product recycling are major factors of agroecological transition in agro-sylvo-pastoral systems.

Les éleveurs laitiers ouest-africains connaissent des contraintes d’alimentation des vaches en saison sèche conduisant à une chute de l’offre de lait local. Pour y faire face, ils achètent cher et utilisent sporadiquement des aliments concentrés. Certains semblent s’en sortir en combinant le pâturage, les fourrages et l’aliment concentré. Cependant, ces stratégies restent méconnues et leur efficacité peu optimisée. La présente étude vise à caractériser les pratiques d’alimentation des vaches traites en saison sèche chez les éleveurs laitiers extensifs en zone sud-soudanienne du Burkina Faso pour repérer et promouvoir des stratégies d’alimentation innovantes économiquement viables. Une enquête ménage a été conduite auprès de 134 éleveurs qui approvisionnent les laiteries des villes de Banfora et Bobo-Dioulasso, à l’ouest du pays. La typologie des systèmes d’alimentation proposée a mis en évidence un système prometteur, techniquement innovant et économiquement plus performant, qui n’est cependant pratiqué que par une minorité d’éleveurs. Dans ce système, les vaches traites nourries au pâturage (9 h/j) reçoivent en saison sèche, en complément, des apports de fourrage de bonne qualité et en quantité suffisante (3360 ± 1424 kg de matière sèche/vache/an) et des quantités modérées de concentrés alimentaires (253 ± 244 kg matière brute/vache/an). Cette pratique permet de maintenir la production laitière et d’assurer des revenus réguliers à l’éleveur. Il apparaît donc opportun de soutenir la co-conception de systèmes d’alimentation des vaches intégrant les cultures fourragères et une utilisation modérée des concentrés. L’effort d’amélioration du système d’alimentation des vaches par les éleveurs semble avoir comme motivation l’existence du dispositif de collecte de lait facilitant l’écoulement et les incitations pour la production. En vue d’amplifier les dynamiques d’intensification agroécologique de la production laitière en cours, les opérateurs de l’amont de la filière lait local pourraient donc développer un modèle économique qui crée davantage d’incitation pour les éleveurs.

West African dairy farmers are faced with constraints in feeding their cows during the dry season when rangelands are not rich enough in soft grass. This leads to a drop in the supply of local milk. To address this problem, farmers purchase expensive feedstuffs that they use sporadically. Some seem to cope by combining grazing, fodder and cattle feed. However, these strategies remain poorly known. This study aimed at characterizing dry season feeding practices of dairy cows in extensive farms in the South Sudanian zone of Burkina Faso in order to identify and promote innovative and economically viable feeding strategies. A household survey was conducted among 134 farmers supplying dairy processors in the cities of Banfora and Bobo-Dioulasso, in western Burkina Faso. The interviews aimed at characterizing the feeding practices of dairy cows during the dry season. The study presents a typology of dairy cows feeding systems, highlighting a technically and economically promising feeding system, however practiced only by few farmers. In this system, the dairy cows receive good quality fodder in sufficient quantity (3360 +/- 1424 kg dry matter/cow/year) and moderate quantities of feed concentrates (253 +/- 244 kg gross weight/cow/year) during the dry season, in addition to a reasonable duration of grazing (9 h/d). This practice maintains milk production and provides a steady income to the farmer. It appears therefore appropriate to support the co-design of feeding systems integrating forage crops, grazing and moderate use of concentrate feed. The improvement of dairy cow diet appears to be mainly motivated by the organization of the collection and the incentives offered to farmers. In order to amplify the dynamics of agro-ecological intensification of milk production, and in particular this type of dairy cows feeding system, upstream operators of the local dairy value chain could develop an economic model that creates more incentives for the farmers.