January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 0 Agroecology Initiative Technical Report January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 1 Contents List of tables ......................................................................................................... 3 List of figures ....................................................................................................... 4 1. Introduction ................................................................................................. 5 1.1. Background to the initiative.............................................................................. 5 1.2. Context of the agroecology initiative in Burkina Faso ................................. 6 1.3. Evaluation of agroecology: Measurement and methodology ................... 9 2. Methodology ............................................................................................. 11 2.1. Localization of the HOLPA tool ..................................................................... 11 2.2. Implementation of the HOLPA tool .............................................................. 12 2.3. Analysis of HOLPA data .................................................................................. 15 3. Results of the HOLPA survey .................................................................... 18 3.1. Module 1: Background ................................................................................... 18 3.2. Module 2: Agroecology ................................................................................. 25 3.3. Module 3: Performance.................................................................................. 30 4. Feedback from the workshop on the results of the HOLPA tool test ...... 35 5. Use of assessment results ......................................................................... 37 6. Lessons learned ......................................................................................... 38 7. Conclusion and next steps ........................................................................ 40 8. Acknowledgements .................................................................................. 41 9. References ................................................................................................. 42 January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 2 Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool Boko Michel OROUNLADJI1, Patrice KOUAKOU1, Ollo SIB1, Souleymane SANOGO2, Adama OUEDRAOGO3, Eric VALL1 December 2024 (1) CIRAD, (2) CIRDES, (3) INERA, Agroecology Initiative Technical Report January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 3 List of tables Table 1. List of criteria for assessing the quality of indicators ..................................................................................................................... 11 Table 2. Some variables used to compare dairy farms ................................................................................................................................ 17 Table 3. Socio-professional characteristics of respondents ........................................................................................................................ 18 Table 4. Number of individuals by age category in surveyed households and age of respondents .................................................. 19 Table 5. Dairy farmers' livestock by farm species (heads) ........................................................................................................................... 20 Table 6. Animal husbandry practices on dairy farms ................................................................................................................................... 22 Table 7. Level of erosion problems and soil fertility improvement practices on dairy farms ................................................................ 23 Table 8. Grain yields (kg/ha) of crops on dairy farms ................................................................................................................................... 25 Table 9. Some structural characteristics of dairy farmer groups ................................................................................................................ 32 Table 10. Local performance indicators by type of dairy farm ................................................................................................................... 33 January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 4 List of figures Figure 1. How the work packages of the Agroecology Initiative feed into the co-construction of an agroecological Business Model for the Bobo-Dioulasso milk chain ..................................................................................................................................................................... 5 Figure 2. Diagram of the package of agroecological technologies tested on dairy farms ..................................................................... 8 Figure 3. Agroecological transition as part of the agroecological living landscape in Burkina Faso .................................................... 9 Figure 4. HOLPA tool modules and their components ................................................................................................................................ 10 Figure 5. Priority local indicators for dairy farmers ....................................................................................................................................... 12 Figure 6. Administrative map of the Hauts-Bassins region ......................................................................................................................... 13 Figure 7. Rainfall and temperature in Bobo-Dioulasso over the last 40 years (1984 - 2023) ................................................................ 14 Figure 8. Dairy farmers' satisfaction with their living conditions ................................................................................................................ 19 Figure 9. Power and freedom of decision-making for men and women on dairy farms ....................................................................... 20 Figure 10. Origin of animals reared on dairy farms ...................................................................................................................................... 21 Figure 11. Soil fertility levels on dairy farms................................................................................................................................................... 24 Figure 12. Crop pest management practices on dairy farms ..................................................................................................................... 24 Figure 13. Rating of agroecological principles on dairy farms ................................................................................................................... 27 Figure 14. Dairy farmers' level of theoretical knowledge about agroecology ........................................................................................ 28 Figure 15. Motivations for dairy farmers to achieve a more agroecological status ................................................................................ 28 Figure 16. Agroecological farming practices on dairy farms ...................................................................................................................... 29 Figure 17. Overall performance indicators for milk-producing farms in the Bobo-Dioulasso dairy basin ......................................... 31 Figure 18. Link between the degree of agroecology and the level of performance of dairy farms .................................................... 34 January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 5 1. Introduction 1.1. Background to the initiative Agroecology aims to achieve sustainable agriculture and food systems, rooted in a set of principles (recycling, input reduction, soil health, animal health, biodiversity, synergy, economic diversification, co-creation of knowledge, social values and diets, equity, connectivity, governance of land and natural resources, and participation; Wezel et al. 2020) that emphasise the need to work with nature rather than against it. It aims to build sustainable agriculture and food systems by relying as much as possible on the co-creation of knowledge, the participation of farmers and multiple stakeholders in decision-making, while strengthening the connection between farmers and consumers. The activities of the Agroecology Initiative (AEI) are structured around five Work Packages (WP):  WP1: Transdisciplinary co-design of innovations in an Agroecological Living Landscape (ALL)  WP2: Assessments of the agroecological performance of farming systems  WP3: Development of an inclusive, agroecological economic model and financial strategies tailored to the expectations of ALL stakeholders  WP4: Strengthening policies and institutional environments conducive to the agroecological transition  WP5: Understanding and influencing partnership changes and stakeholder behaviour to promote the agroecological transition. It is in this context that the AEI, through one of its specific objectives, aims to produce scientific evidence on the performance of farming systems, and in particular their agroecological performance, in order to encourage their large-scale development in the regions. This objective cannot be achieved without collecting data and evidence on the performance of farming systems. In 2023 and 2024, in Burkina Faso, activities were carried out in all five AEI work packages (Figure 1) with the aim of co-designing an Agroecological Business Model for the local dairy value chain (Sib et al., 2024). Figure 1. How the work packages of the Agroecology Initiative feed into the co-construction of an agroecological Business Model for the Bobo-Dioulasso milk chain January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 6 1.2. Context of the agroecology initiative in Burkina Faso In order to produce locally relevant and globally comparable data on the performance of agricultural systems, the AEI's Work Package 2 has developed the HOLPA (Holistic Localized Performance Assessment for Agroecology) tool. Applying this tool in Burkina Faso required three preliminary steps: (i) drawing up a context document describing the current situation of the targeted farming systems, using technical, economic, environmental and social criteria as well as the 13 principles of agroecology; (ii) getting to grips with the HOLPA tool; and (iii) identifying local indicators specific to the milk value chain in the ALL. As a reminder, in Burkina Faso, AEI has focused on the dairy value chain, with an ALL based on the Bobo-Dioulasso multi-actor dairy innovation platform established in 2020. In 2023, the Dairy Innovation Platform (DIP in English or “Plateforme d’Innovation Lait” DIP in french) was consolidated into an ALL with the incorporation of new members and partners useful for supporting dairy industry stakeholders in the agroecological transition (Sib et al., 2023a). To implement the AEI activities, and in particular the HOLPA survey, it was necessary to involve researchers, several cooperatives of dairy farmers and processors, representatives of the public sector (Direction Régionale de l'Agriculture, des Ressources Animales et Halieutiques, Mairie), technical and financial partners, NGOs and professional organisations. All these stakeholders were contacted, interviewed and involved in drawing up the background document, identifying local indicators, familiarising themselves with the HOLPA tool, collecting quantitative and qualitative data, and finally validating the results of the HOLPA survey, which is the subject of a separate report (Orounladji et al., 2024a). It should be remembered that the geographical area of the ALL is characterised by rain-fed family farming, with the majority of farms oriented towards agriculture (cotton, cereals, pulses; Kouakou et al., 2023). Livestock farming systems in this area are predominantly agropastoral, with some farmers isolating small nuclei of female zebu cattle in order to exploit them for their milk. The milk sector is developing around mini-dairies as a result of population growth and demand for dairy products. This dynamic has led to the opening of milk collection centres. And in terms of production, some farmers are specialising in milk production by setting up mini-farms. Finally, some farmers are taking an interest in the production and sale of fodder, because livestock farmers need it more and more. In Burkina Faso, the HOLPA survey was conducted on 204 dairy farms. The survey was conducted over a reference period that more or less coincided with the first year of AEI in Burkina Faso. At that time, experimentation with agroecological technologies on dairy farms had only just begun. Consequently, these new agroecological technologies had not had time to produce their effects and induce changes and impacts on the farms. Consequently, under these conditions, the HOLPA survey made it possible to draw up a sort of baseline on the levels of agroecological performance of dairy farms prior to the implementation of changes aimed at increasing their degree of agroecology. It is very important that the reader bear this in mind. To enable respondents to better appreciate their performance when presenting the results, they have been divided into three categories for two fundamental reasons: (i) firstly, since the ALL was formalised in Burkina Faso, for all the activities (development of current and agroecological business models, cost-benefit analysis, determination of areas for initiative, etc.) carried out with the various links in the milk value chain, particularly the production link, we have always had two professional groups: mini-farms and agropastoralists; and (ii) secondly, while we know that at this stage of data collection the agroecological package could not have produced any effect, we decided to separate the volunteers involved in the experimentation of the agroecological package in 2023 (52 respondents) from the other agropastoralists (not involved in this experimentation on the farms) in view of the fairly high number of agropastoralists (196 respondents). In this way, we were able to create three categories that enabled us to compare their performance (based on local indicators). The three categories are described as follows: o Mini-dairy farms: which have set up a relatively intensive milk production unit based on the breeding of zebu dairy cows crossed with exotic breeds, reared in stalls with limited access to natural pastures; o Agropastoralists who have tried out the agroecological package proposed by the AEI: These farmers have set up a relatively extensive milk production unit based on rearing female zebu cattle, with daily grazing on natural grazing lands and the addition of fodder and supplementary feed. These farmers experimented with fodder demonstration plots and received support for optimised management of crop and livestock co-products using the CoProdScope tool. They have also received support with the Jabnde tool for rationing cows and have installed covered manure pits to improve the management of organic waste. o Other agropastoralists: their extensive production system is based on dairy cows of local breeds, better adapted to the climatic conditions of the region. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 7 During the workshop on the identification of local indicators for multi-criteria assessment of agroecology performance, which took place in Bobo-Dioulasso on 23 August 2023 in the training room of the Centre International de Recherche-Développement sur l'Elevage en zone Subhumide (CIRDES), The participants, who are stakeholders in the Initiative on Agroecology, began by familiarising themselves with the vision of the DIP, before looking at the identification, prioritisation and validation of local indicators (Orounladji et al., 2023). The vision is as follows: “ By 2028, the Bobo-Dioulasso dairy basin will be producing, collecting and processing 18,000 litres of local milk a day. The overall objective of the DIP is to increase the daily production, collection, processing and marketing of milk in the Bobo Dioulasso dairy milkshed to 18,000 litres. Specifically, this will involve :  Increase milk production on the farm and make it more regular by improving the feed and health of dairy cows;  Strengthen the intellectual and technical capacities of farmers;  Ensure that the quality and quantity of milk at collection are properly measured using appropriate tools;  Harmonised milk prices applied by collectors;  Improve the milk collection, storage and distribution system;  To market a wide range of dairy products produced by processing quality milk using appropriate equipment and techniques. In 2023, when AEI activities were launched in Burkina Faso, we checked the compatibility of the 6 objectives of the DIP with the principles of agroecology. To help achieve this vision, AEI is working with stakeholders to co-create innovations at several levels:  Improving cow feed by making more efficient use of the farm's resources,  Improved milk distribution by enabling collection centres to diversify their services to dairy farmers and processors  Diversification of products made from local milk (yoghurts flavoured with extracts of local products)  Greater equity through greater inclusion of women, young people and the elderly in the governance of the value chain. Specifically, with dairy farmers, who are key players in achieving this vision, AEI supports them in establishing fodder crops, in storing and conserving fodder, in improving the use of crop and livestock co-products for fodder and manure, in improving cow diets based on quality fodder and in the production and use of manure, as shown in Figure 2 (Sib et al., 2023b; Sib et al., 2023c). Cultivated fodder and crop co-products are used to feed animals, particularly cows, and livestock and crop co-products are used to produce manure in covered manure pits. The manure produced is in turn used to fertilise the soil. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 8 Figure 2. Diagram of the package of agroecological technologies tested on dairy farms In Burkina Faso, AEI seeks to support ALL stakeholders in an agroecological transition based on five types of changes (innovations) at different levels of the food system (Figure 3):  At farm level: to achieve a sustainable increase in milk production through the integration of farming and livestock in dairy production units (production of high-quality fodder and recycling of co-products as fodder and manure).  At collection level: support milk collection centres in diversifying their services in order to increase the quantity, quality and regularity of milk collected.  At processing level: support processors in diversifying their production to meet emerging consumer demand  At the level of dairy value chain governance: improve inclusiveness in value chain governance to support the development of the entire value chain  At the ALL level: improving the governance of the ALL January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 9 Figure 3. Agroecological transition as part of the agroecological living landscape in Burkina Faso 1.3. Evaluation of agroecology: Measurement and methodology The projects involved in the agroecological transition have applied a large number of different approaches and tools to characterize farming practices and food systems and assess their performance. According to Wiget et al (2020) and Geck et al (2023), there are at least 35 assessment tools. The existing tools have at least three limitations (Darmaun et al., 2023). Firstly, most are not adapted to local conditions and are based solely on standardized indicators, which means that they cannot be applied in a new context without risking the results being irrelevant to local stakeholders. Secondly, although many tools capture a wide range of performance criteria, some social dimensions (including subjective measures of autonomy and human well-being) and environmental dimensions (including biodiversity and climate resilience) are poorly represented. Third, tools that seek to characterise practices (e.g. in terms of agroecology) tend to use in-depth methods, limiting the time that can be devoted to collecting performance data within the same survey effort. This new HOLPA tool has been developed by AEI to address these limitations. The assessment framework of this tool comprises three modules: 1) Context Module, 2) Agroecology Module, 3) Performance Module, underpinned by a process of contextualisation of the indicators. The performances are agronomic, social, economic and environmental (Figure 4). For the Agroecology and Performance Modules, there are global indicators to which local indicators have been added to better contextualise the tool and enable it to be truly relevant for assessing the agroecological performance of farms in different ALLs. To address this and ensure that the information collected via HOLPA is meaningful locally while remaining relevant to global sustainability goals, we have developed a seven-step localisation process to be completed prior to implementation. This includes: 1) translating (if necessary) the HOLPA survey into the local language to create a context-sensitive version, 2) adjusting certain questions and response options to match the local context, including local concepts, terminologies and sensitivities, 3) responding to all parts of the survey to identify and correct errors introduced in steps 1 and 2, and test the survey with at least one local researcher or practitioner familiar with local food and farming systems, to ensure that the questions are fully understood, 4) organise a multi-stakeholder local indicator selection workshop (LISP), where a range of stakeholders, including youth groups and women, envision a sustainable future for their landscape and select key indicators to monitor progress towards this goal (see below), 5) integrate the results of steps (3) and (4) to update the version adapted to the HOLPA context, 6) test the survey with at least three local farmers and resolve any problems or gaps identified in order to produce a final version adapted to the context, 7) train the surveyors, including checking the quality of the first surveys and providing feedback to ensure correct implementation. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 10 At the end of these stages, the HOLPA tool was tested with farmers who are members of the ALL presented above in Burkina- Faso. For the specific case of the ALL in Burkina Faso, the research questions are as follows: 1) What is the agroecological status of dairy farms (of all types) in the Bobo-Dioulasso ALL? 2) What are the performance indicators for all types of dairy farm and for different types of farm (mini-dairy farms, experimental agropastoralists, other agropastoralists)? 3) What is the link between the degree of progress in agroecology and the performance levels of farms (all types taken together)? Figure 4. HOLPA tool modules and their components January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 11 2. Methodology 2.1. Localization of the HOLPA tool As part of the localization of the HOLPA tool, a workshop to identify local indicators was organised and held on 23 August 2023 at the Centre International de Recherche-Développement sur l'Elevage en zone Subhumide (CIRDES) in Bobo-Dioulasso. Further information on the workshop is contained in the report Orounladji et al. (2023). During this workshop, in order to build a relevant list of specific indicators, participants were asked to identify what they would like to see in relation to the different objectives and changes envisaged, and then to discuss how they could measure or monitor the changes. The participants were divided into four groups. The members of each group included a facilitator, one or two experts in the field, and 5 to 7 participants. The groups discussed the indicators in terms of their agronomic, social, economic and environmental dimensions. Each group was asked to reflect on the theme in the context of the project. The following guiding questions were used as a basis for identifying the indicators: (i) what changes {agricultural, social, economic, environmental} do we want to see as a result of agroecology? (ii) how will we know if we have achieved our objectives {agricultural, social, economic, environmental} with agroecology? (iii) what are the obstacles that could prevent us from achieving these objectives? how can we assess these obstacles? Using the world café facilitation method, each group took 20 minutes to discuss each dimension, first identifying the changes they would like to see and then the actions to be taken in relation to the ALL's objectives and vision, as well as the indicator to be used to evaluate these changes. After 20 minutes of discussion, the group moved on to another dimension, with the facilitator and expert remaining on hand to explain to the newcomers the elements captured with the previous groups, and the newcomers adding what they felt was missing from the inventory. Before assessing the indicators, the participants were introduced to what constitutes a good indicator. The five criteria selected for easy ranking and evaluation, summarised in Table 1, were presented to the participants. Table 1. List of criteria for assessing the quality of indicators Criteria Explanations Relevance to the sustainability of agricultural systems in the Sahel The indicator quantifies the effects of a farming system on key sustainability issues for ALL stakeholders. Scientific relevance The indicator is transparent and clearly defined (method, data source, assumptions), scientifically validated and recognised, reproducible in different contexts, accurate and robust. Feasibility The indicator is easy to enter, simple to calculate (time and cost of implementation), and tailored to the target users (availability of users at key measurement times, skills, experience). Usefulness The indicator covers the users' needs/objectives, produces results that are understandable to the target users, and can be easily communicated. Sensitivity The indicator is sensitive to change when the system is moving towards less or more sustainability (to be able to act before it's too late, or to recognise situations on the right track). Source: Report on workshop to identify local indicators Next, an indicator ranking procedure was also introduced, to prioritise the criteria and identify priority criteria with ratings such as: low = 1, medium = 2 and high = 3. The ranking was applied according to the five selected criteria and a final score was calculated by adding up the scores for each indicator. Participants were encouraged to be as rational as possible in their ranking in order to avoid many indicators obtaining high scores. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 12 The nine (09) priority indicators identified by the ALL members (ie. The DIP members), according to the four dimensions, are presented succinctly in Figure 5 Figure 5. Priority local indicators for dairy farmers 2.2. Implementation of the HOLPA tool Purpose of the HOLPA survey. The purpose of the HOLPA tool is to collect data and evidence on the current state of agroecology and its performance. The data were collected from dairy farmers located in the intervention area of the Bobo Dioulasso ALL, which is located in the Hauts-Bassins region in western Burkina Faso (Figure 6). Localities and households included in the survey. The population of the Hauts-Bassins region is young. According to the Institut National de la Statistique et de la Démographie (INSD, 2022), the 5-14 age group accounts for 27% and the 15-64 age group for 55% of the total population, estimated at 2,239,840 (including 1,094,100 men and 1,145,740 women). The gender structure of the population of the Hauts-Bassins region is similar to that of Burkina Faso as a whole. The population is made up of approximately 49% men and 51% women. The Hauts-Bassins region, with 10.9% of the total population of Burkina Faso, remains one of the most densely populated regions. Over the period 2006-2019, the Hauts-Bassins region recorded an intercensal population growth rate of 3.29%. The working-age population represents 54.7% of the total. This represents an economic and social challenge for the local authorities in terms of health, education and employment. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 13 Figure 6. Administrative map of the Hauts-Bassins region The Hauts-Bassins region lies in the southern Sudanian savannah zone, characterised by a sub-humid tropical climate with three seasons: a rainy season from May to October, a cold dry season from November to February, and a hot dry season from March to May (Figure 7). Average daily temperatures also vary seasonally. In the middle of the rainy season, they are low, with an average of 26°C. During the dry season, they are high, with an average maximum of 32 to 33°C. Evapotranspiration is generally very high. It exceeds rainfall during the period from October to June, i.e. for more than nine months, leading to a significant drop in water resources, which is detrimental to livestock farming. However, the effects of climate change are a reality in the Hauts-Bassins region. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 14 Figure 7. Rainfall and temperature in Bobo-Dioulasso over the last 40 years (1984 - 2023) Household sampling strategies. With regard to sampling, we used a non-probabilistic method to select farmer households for our data collection. Our survey covered 204 farmers associated with a milk production unit. In order to ensure an agroecological gradient within each production system, the selection of farms was based on the following inclusion criteria: o Farmers affiliated to the Bobo Dioulasso ALL who may or may not have already taken part in agroecology projects. o Farmers who intend to join the Bobo Dioulasso ALL and who express a desire to convert to sustainable practices. o Voluntary farmers implementing the agroecological package (Fodder Demo-Plot, manure pit, reasoned management of co-products (with the CoProdScope tool), reasoned dry-season rationing of dairy cows (with the Jabnde tool). Qualitative and quantitative data were collected in this study using an electronic questionnaire loaded onto KoboToolbox. The information was collected by successively applying the household and farm surveys to each farmer. After obtaining the respondent's consent for the data to be published without their personal information, the questionnaire covered some general information such as location, scale (e.g. plot, farm, landscape) and interviewer details. Four modules were covered in the household questionnaire. These were: (i) context module, (ii) agroecology integration (Ae) module, (iii) global key performance indicators module and (iv) local indicators module. The various topics covered in the context module include the collection of demographic information and involves recording respondent characteristics such as age, gender, sociolinguistic group, education, marital status, occupation, length of time living in the community, relationship with the head of household, participation in farming activities, participation in farmers' associations and participation in agricultural research or development projects. In addition, farm household characteristics such as household structure, agricultural production system, end use of agricultural products, fertiliser inputs and disease management, farm size and land tenure patterns are collected to assess socio-economic and environmental factors related to the unit of assessment. The context assessment also seeks to explore motivation and attitude towards agroecology by assessing personal perspectives on agroecology. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 15 The agroecology integration module is designed to assess the current state of practices using questions covering the 13 principles of agroecology and two additional questions to determine self-perceived adherence to these principles. The module aims to characterise the current level of adherence to agroecology or the degree of agroecological transition by assessing farming practices and the overall benefits derived from these practices. Most of the questions are multiple-choice, using a five- point Likert scale. All survey responses are scored from 1 to 5. On the basis of a median score reported for all the principles of agroecology, a composite score between 1 and 5 can be generated to characterise the overall agroecological state. The cross- cutting theme of 'self-perceived adherence' is assessed on the basis of questions designed to assess the respondent's opinion of the extent to which their field, farm or landscape is agroecological. Responses ranged from completely non-agroecological to completely agroecological. In fact, self-perceived adherence provides an alternative way of assessing the level of transition to agroecology. The global indicators module is used to assess the agroecological performance of the farming system at a selected scale using a set of survey questions and field measurements of the indicators. The sections of the survey questionnaire and field measurements cover several different elements related to the four general domains of agriculture, economy, environment and social. The data collected in this module is used to estimate agricultural, economic, environmental and social performance when formulating the following questions: 1) What are the impacts of increased adherence to agroecology? 2) What are the trade-offs between the dimensions of sustainability? The module on local indicators addressed questions on the four dimensions of farm performance assessment (Figure 5). For the farm survey, data on biodiversity, soil health and crop health are collected for use in assessing these different parameters. Soil samples are also taken for laboratory analysis to determine soil organic matter and organic carbon on these farms Soil samples were taken at three locations on the farm. The first (site 1) near the buildings, the second (site 2) in the middle of the cultivated land and the third (site 3) near the natural vegetation. At each site, we marked out a square area measuring 10 x 10 m². We then took soil samples at five points corresponding to the centre and each corner of the delimited square. A composite sample was taken for each site on each farm. A total of 612 samples were analysed, 204 for each site. Before the survey was launched, the questionnaire was tested by the interviewers on a number of farms in order to: i) finalise the survey questionnaire and make the final adjustments; check that the interviewers were familiar with using the tablets and the questionnaire (good understanding of the questions and possible answers). 2.3. Analysis of HOLPA data The analyses were carried out using R 4.4.1 software (R Core Team, 2024). In order to present the context and general information on the farms, data relating to the socio-demographic profiles and technical performance of the farms were subjected to descriptive statistics. Farmers' perceptions of agroecology and their well-being are presented in the form of Likert-scale graphs after subjecting the data to tests using the tidyverse (Wickham et al., 2016 and 2019) and likert (Bryer and Speerschneider, 2016) packages. The status of agroecology in the Bobo-Dioulasso ALL, taking all farms together, was assessed and the results are presented in the form of boxplots for all 13 agroecological principles as well as the overall agroecological level of the farms. The overall Key Farm Performance Indicators have been calculated and the results presented using boxplots. In the context of Burkina Faso, to enable farmers to assess their performance more accurately, a classification was established by distinguishing three categories of farmers, as described in section 1.2. To this end, data from local indicators considering the four dimensions of farm performance assessment were subjected to an analysis of variance (ANOVA) which was carried out considering the variables presented in Table 1. This was supplemented by a Student Newman-Keuls test for the comparison of means in the event of significant differences (p < 0.05) being observed between the groups of farms using the agricolae package (de Mendiburu, 2023). However, Kruskal Wallis tests were applied when certain data did not comply with the conditions for applying analysis of variance. In order to determine the link that might exist between the degree of progress towards agroecology and the performance levels of the farms, we calculated the overall score of the level of agroecology per farmer on the basis of the scores for each agroecological principle, for all 13 agroecological principles. For each agroecological principle, the score was calculated by taking the arithmetic mean of all the scores for the different indicators/questions for the principle in question. As the scores range from 1 to 5, this range has been divided into groups or levels of agroecological transition so as to have approximately the same amplitude/extent for the three levels considered January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 16 o Low: Agroecology scores between 1 and 2.33; households in the initial phase of agroecological transition o Medium: Agroecology scores between 2.34 and 3.66; households showing average progress in agroecological transition o High: Agroecology scores between 3.67 and 5; households showing high progress in agroecological transition. Farmers were classified into these agroecology groups/levels according to the overall score obtained. After cross-referencing the data on agroecology with that on performance, boxplots were produced to show the links between each performance indicator and the levels of progress towards agroecology. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 17 Table 2. Some variables used to compare dairy farms Variables Units Description Total surface area ha Total cultivated area declared by respondent Proportion of surface area with uncontested ownership % (Cultivated area for which the respondent has full ownership rights / Total area) x 100 Size of cattle herd TLU Number of cattle declared x 0.7 (conversion coefficient) No. of cows milked in hds U Number of cows milked during the cold dry season (November, December, January, February) No. of cows milked in cds U Number of cows milked during the hot dry season (March, April, May) No. of cows milked in rs U Number of cows milked during the rainy season (June, July, August, September, October) Manure pits U Number of working manure pits per farm Quantity of quality forage stored per cow kgDM/TLU ∑quantity of quality forage [Number per means of transport (U) x Conversion coefficient (kgDM)]/Total number of dairy cattle (TLU). Conversion coefficients: Boots: 3; Rickshaw: 35; Motorbike: 20; Gondola: 100; Small flatbed cart: 170; Large flatbed cart: 250; Tricycle: 150; Trailer: 530 Quantity of roughage stored per cow KgDM/TLU ∑quantity of roughage [Number per means of transport (U) x Conversion coefficient (kgDM)]/Total number of dairy cattle (TLU). Conversion coefficients: Boots: 3; Rickshaw: 40; Motorbike: 30; Gondola: 100; Small flatbed cart: 170; Large flatbed cart: 250; Tricycle: 150; Trailer: 540 Agroecological facilities (hayloft, silos, forage sheds) U Number of agroecological facilities (hayloft, silos, fodder sheds) in operation per farm Annual cost of monitoring a cow's health FCFA/cow/year Annual health monitoring costs for a cow Raising awareness of the rules for living together % Proportion of respondents who have been made aware of the rules for living together Legend: ha: hectare; TLU: Tropical Livestock Unit (1 TLU = 1 bovine of 250 kg live weight); U: unit; kgDM: kilogram of dry matter; No.: number; cds: cold dry season; hds: hot dry season; rs: rainy season; 1 USD = 605 FCFA. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 18 3. Results of the HOLPA survey 3.1. Module 1: Background 3.1.1. Socio-professional characteristics of respondents The majority of respondents were male, representing 93.63% of the sample, while women made up only 6.37% of the sample (Table 3). The low representation of women in the sample can be explained by the fact that they are very poorly represented as farm managers in this category of farmers, who are essentially dairy farmers. However, they are omnipresent in dairy activities on the farm and in the household, but they are not solicited for the surveys. Most respondents (71.08%) cannot read or write, while 22.06% can read and write. The majority of respondents (70.59%) had no education at all, followed by those with primary education, who made up 19.61% of the sample. The vast majority of respondents are cohabiting (80.88%) and 14.71% are married. The Peulh community is the most represented in the sample, with a representativeness rate of 81.37%. The vast majority of respondents (98.53%) are engaged in agricultural and/or livestock work. Table 3. Socio-professional characteristics of respondents Variables Workforce Percentage (%) Type Men 191 93,63 Woman 13 6,37 Can you read and write in any language? Cannot read or write 145 71,08 Can read and write 45 22,06 Can only read 12 5,88 Can only write 2 0,98 Level of education No 144 70,59 Primary 40 19,61 Secondary 15 7,35 University 5 2,45 Marital status Cohabitation 165 80,88 Married 30 14,71 Single 8 3,92 Divorced / Separated 1 0,49 Sociolinguistic group Peulh 166 81,37 Mossi 16 7,84 Bobo 10 4,90 Other (Bissa, Gouroussi, Samo, Sambla) 5 2,45 Dafing 5 2,45 Dioula 2 0,98 Main activity Agricultural and/or livestock work 201 98,53 Study/education/training 1 0,49 Housekeeper 1 0,49 Public administration 1 0,49 Source: HOLPA survey, Bobo-Dioulasso, 2023-2024. 3.1.2. Age of respondents and demographic status of households surveyed The average number of men or women aged between 18 and 65 in the households surveyed was 3 (Table 4). There are virtually no men or women over 65 in these households. In the households surveyed, there was an average of 3 men and an average of 3 women per household in the under-18 age group. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 19 The average age of respondents is 46. The minimum age is 20 and the maximum age is 75. This indicates a certain variability in the age ranges of respondents, with a concentration around the average age of 46. Table 4. Number of individuals by age category in surveyed households and age of respondents Variables Average Standard deviation Min Max Number of individuals by age category Men (≥18 and ≤65 years old) 3 2 0 12 Women (≥18 and ≤65 years old) 3 2 0 15 Male (>65 years) 0 0 0 2 Women (>65 years) 0 0 0 2 Male (<18 years) 3 2 0 12 Female (<18 years) 3 2 0 9 Age of respondents Age 46 13 20 75 Source: HOLPA survey, Bobo-Dioulasso, 2023-2024. 3.1.3. Satisfaction and concern among dairy farmers about their living conditions Overall, dairy farmers expressed a feeling of satisfaction with their living conditions (Figure 8), such as (i) nutritional security, (ii) their standard of living, (iii) their own life and personal situation, (iv) their personal relationships, (v) their sense of belonging to the community, (vi) their fulfilment in life, (vii) their profession, (viii) the time they have to do the things they like to do, (ix) their economic security, (x) their health, (xi) and the quality of the local environment. On the other hand, their opinions are mixed when it comes to "feeling safe". Figure 8. Dairy farmers' satisfaction with their living conditions 3.1.4. Power and freedom to make important decisions in the food system Among men, the feeling of having the power and freedom to make most of the important decisions in the life of their household was almost total 10 years ago (99%), but now stands at 98% (Figure 9). Among women, the feeling of not having the power and freedom to make most of the important decisions in the life of their household remains dominant (74%), even though the situation has improved over the past decade. Furthermore, these results may be influenced by the sample, which is predominantly made up of men (94% of respondents) rather than women (6% of respondents). January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 20 Figure 9. Power and freedom of decision-making for men and women on dairy farms Legend: Step 1: Almost no power or freedom to make decisions; Step 2: Only a small amount of power and freedom; Step 3: Power and freedom to make important life decisions; Step 4: Power and freedom to make many important life decisions; Step 5: Power and freedom to make most important life decisions. 3.1.5. Diversity of animal species on farms The different species of livestock reared on the farms and their numbers are shown in Table 5. Among the livestock, cattle come first with an average of 38 heads/farm (varying from 1 to 180 heads). For goats, the average is 14 heads/farm, with numbers varying from 4 to 50. For sheep, the average is 15 heads/farm, with figures ranging from 4 to 40. For poultry, mainly chickens, the average is significantly higher, with an average of 255 heads per farm. The values range from 0 to 6,000 heads, reflecting a wide variability in chicken numbers, with some flocks having very large numbers. Table 5. Dairy farmers' livestock by farm species (heads) Species Average Standard deviation Min Max Cattle 38 22 1 180 Goats 14 7 4 50 Ovine 15 10 4 40 Poultry 255 724 10 6000 Source: HOLPA survey, Bobo-Dioulasso, 2023-2024. The majority of dairy farmers (65%) self-produce and/or exchange with their peers or collectively manage all the farm's animal genetic resources (Figure 10). For 32% of farmers, 25% of animal genetic resources (e.g. chicks, young animals, semen) are bought on the market and the remaining 75% are self-produced or exchanged. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 21 Figure 10. Origin of animals reared on dairy farms 3.1.6. Herd management practices on dairy farms The majority of respondents rely on surface water (dams/lakes - 83.82%) and groundwater (wells, boreholes - 67.65%) as a source of watering (Table 6) In terms of grazing management practices, only 1.47% of respondents stated that they try to reduce grazing pressure on grazing areas. According to 23.4% of respondents, fodder cultivation practices are geared more towards the production of fodder legumes. This shows the extent to which the package of agroecological technologies tested on dairy farms during the AEI makes sense and will undoubtedly lead to greater adoption of fodder production on these farms. Moreover, the workshop to identify the changes brought about by the initiative revealed that dairy farmers have already started to produce more fodder (Orounladji et al., 2024b). An impact assessment in the strict sense of the term in the coming years will enable us to better assess the positive changes brought about and the real impact of the initiative on these farms. Animal genetic management practices show that 6.86% of respondents use improved breeds, in particular zebu dairy cows crossed with exotic breeds. These results show that the majority of dairy farms use local breeds that are better adapted to the region's climatic conditions. This is borne out by the number of mini-farms seeking to develop dairy farming based on zebu x exotic dairy cross-breeds (around 8) compared with the large number of zebu-breeding agropastoralists (196 respondents) involved in data collection The practices concerning the feeding, care and welfare of animals, such as providing shelter, providing constant access to adequate feed, providing clean drinking water, providing medical assistance if necessary, carrying out regular checks for injuries/diseases, providing a hygienic environment, and offering diversified diets are each declared by at least 80% of respondents to be widely adopted. Vaccination (99.51%), the use of antibiotics (92.65%) and the use of plants or veterinary pharmacopoeia (58.34%) are the disease management practices most commonly used by most respondents (>58%). January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 22 Table 6. Animal husbandry practices on dairy farms Variables Workforce Percentage (%) Watering sources Surface water (dam/lake) 171 83,82 Underground water (wells, boreholes) 138 67,65 Rainwater 59 28,92 Water from wetlands 49 24,02 Rivers 46 22,55 Regional piped water 2 0,98 Grazing management practices Reducing pressure on grazing areas 3 1,47 Forage growing practices Production of fodder legumes 47 23,04 Animal genetics management practices Maintaining improved breeds 14 6,86 Animal feeding, care and welfare practices Providing shelter 203 99,51 Provide constant access to adequate feed 202 99,02 Provide constant access to clean drinking water 199 97,55 Provide medical assistance if necessary 198 97,06 Carry out regular injury/illness checks 181 88,73 Providing a hygienic environment 170 83,33 Offering a variety of diets 165 80,88 Animal health management Vaccination 203 99,51 Antibiotics 189 92,65 Herbal remedies or veterinary pharmacopoeia 119 58,34 Quarantine 29 14,22 Genetic selection for disease resistance 15 7,35 Source: HOLPA survey, Bobo-Dioulasso, 2023-2024. 3.1.7. Soil erosion and soil fertility improvement practices on dairy farms Information on farmers' perceptions of soil erosion, the fertility level of their farmland and the practices they use to improve soil fertility is presented in Table 7. With regard to soil erosion, the majority of respondents (81.86%) considered that erosion was not a problem on their farm. However, soil analyses showed that they were relatively low in organic matter. The soil analysis results show that there is no significant difference between the three sites (site 1: soil sampling site close to dwellings/buildings; site 2: soil sampling site in the heart of the farm; site 3: soil sampling site close to natural vegetation) considered per farm in terms of either soil organic January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 23 carbon or organic matter (Figure 11). The median, for all sites combined, for organic carbon is 0.85% and for organic matter is 1.45%. Organic matter is the main source of nutrients (C, N, S, P, etc.) in highly altered tropical soils with low mineralogical reserves. If we compare the organic matter values with the norm, which is 2 to 3% (Landon, 1991; Sawadogo, 2006; Mulaji et al., 2016), the soils studied are poor in organic matter (organic carbon). These low levels of organic matter in dairy farmers' soils predispose them to acidification and rapid degradation. Low levels of organic matter have a negative impact on soil fertility, generating numerous deficiencies due to its physical, chemical and biological effects. This situation exposes these soils to degradation by water erosion during heavy rainfall in humid tropical environments, especially when slopes become steep (Muladji et al., 2016). With regard to soil fertility improvement practices, almost all respondents (96.57%) applied manure to improve soil fertility. However, a significant majority of respondents (65.69%) also used mineral fertilisers. In terms of manure management practices, manure collection (75.98%) is the most common practice on dairy farms. Table 7. Level of erosion problems and soil fertility improvement practices on dairy farms Variables Workforce Percentage (%) Level of soil erosion problems Soil erosion is not a problem on my farm 167 81,86 Soil erosion is a minor problem on my farm 36 17,65 Soil erosion is a major problem on my farm 1 0,49 Practices to improve soil fertility Application of organic fertiliser or manure 197 96,57 Application of mineral fertilisers 134 65,69 Use of environmentally-friendly practices 5 2,45 No ecological practices or chemical or organic fertilisers have been applied 4 1,96 Manure management practices Collecting manure 155 75,98 Improving the storage of manure 53 25,98 Recovering soil from parks 38 18,63 Animal enclosure (with branches or barbed wire) 38 18,63 No manure management 26 12,75 No practice 16 7,84 Source: HOLPA survey, Bobo-Dioulasso, 2023-2024. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 24 Figure 11. Soil fertility levels on dairy farms Legend: SOC: soil organic carbon; OM: organic matter; site 1: soil sampling site close to dwellings/buildings; site 2: soil sampling site in the heart of the farm; site 3: soil sampling site close to natural vegetation. 3.1.8. Pest management on farms Crop pest control practices are shown in Figure 12. The most widely adopted method is the planting of resistant varieties (64% of farms), most of which are local varieties. Cultural control (crops and fruit showing signs of disease are removed manually) is also used to a lesser extent (31%) to control crop pests. Figure 12. Crop pest management practices on dairy farms 64,22 31,37 27,45 0,49 0,49 0 10 20 30 40 50 60 70 Planting resistant varieties Cultural control Use of cover crops Planting natural repellent plants Encouraging biodiversity and spatial diversity within the agroecosystem January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 25 3.1.9. Crop diversity and yields Grain yields for crops grown on dairy farms are shown in Table 8. In terms of cereals, maize had an average yield of 1,500 kg/ha. Sorghum and millet yielded 414 kg/ha and 349 kg/ha respectively. Rice yielded an average of 1,440 kg/ha. For pulses, soya has an average yield of 600 kg/ha, while groundnuts have an average yield of 365 kg/ha. Cowpea yields average 269 kg/ha. Sesame yields 83 kg per hectare, while cashew yields 400 kg per hectare. Declared grain yields are low compared with local averages (maize between 2,000 and 2,500 kg/ha, millet and sorghum between 800 and 1,000 kg/ha, rice between 1,500 and 2,000 kg/ha, groundnuts and cowpeas between 500 and 1,000 kg/ha). Table 8. Grain yields (kg/ha) of crops on dairy farms Crops Yield (kg/ha) Standard deviation Min Max Cereals Maize 1500 847 148 6200 Sorghum 414 337 33 2000 Mil 349 312 100 1400 Rice 1440 1371 167 4800 Pulses Soya 600 - 600 600 Peanut 365 163 250 480 Cowpeas 269 291 32 1200 Other Sesame 83 - 83 83 Cashew nuts 400 - 400 400 Source: HOLPA survey, Bobo-Dioulasso, 2023-2024. 3.2. Module 2: Agroecology 3.2.1. Agroecology status of dairy farms The agroecology rating of milk-producing farms in the Bobo-Dioulasso ALL, shown in Figure 13, highlights three levels of implementation of the 13 agroecology principles at this level: Agroecology principles with a high score o Recycling o Animal health o Social values o Equity o Participation Agroecology principles with an average score o Reducing inputs o Biodiversity o Synergies January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 26 o Co-creation of knowledge o Connectivity o Governance of land and natural resources Agroecology principles with a low score o Soil health o Economic diversification In general, dairy farms in the Bobo-Dioulasso dairy basin had an average agroecological score (2.62 ± 1.06) The heterogeneity of agroecological levels observed on dairy farms reflects the diversity of farmers' practices and priorities. This variability can be explained by differences in the technical support provided by the decentralised government departments in charge of agriculture and livestock and by other DIP partners who have supported farmers in various agricultural initiatives. Also, in an effort to reduce production costs or to meet immediate economic needs, farmers themselves frequently choose practices. Dairy farmers, who are the upstream actors of the value chain, adapt their practices in line with these constraints and opportunities, resulting in uneven implementation of agroecological principles. Agroecological principles with a high implementation rating, such as recycling, animal health, social values, equity and participation, reflect a strong integration of the productive and social dimensions on dairy farms. Recycling, for example, is encouraged by the integration of crops, animals and trees, which is characteristic of the region's farms. Animal health, for its part, is prioritised because of its direct impact on milk production, which is an essential economic resource for farmers. In addition, social values, equity and participation reflect a community dynamic rooted in local traditions, involving all members of households, including women and young people. On the other hand, the principles with an average agroecological rating, such as input reduction, biodiversity, synergies, co- creation of knowledge, connectivity and governance of land and natural resources, reflect efforts that are underway but still limited. Input reduction and biodiversity are emerging practices, albeit hampered by persistent dependence on chemical inputs and economic constraints. Biodiversity is being enriched, despite being impoverished by the use of chemical inputs, particularly pesticides and herbicides, which deplete the fauna and flora. Similarly, the co-creation of knowledge and the governance of natural resources require more structured collaboration between farmers and technical partners, which can be a challenge for ALL actors for several reasons. The advance of the agricultural front and the extension of residential areas, combined with the effects of climate change, are gradually reducing grazing areas and the availability of fodder and water, which will require improved management and access to the natural resources needed for dairy farming. Finally, the principles with a low agroecological rating, such as soil health and economic diversification, highlight significant challenges. Soil health is severely challenged by difficulties in obtaining labour, which leads most farmers use herbicides and mineral fertilisers. This situation hampers the full adoption of sustainable practices such as composting, even though farmers are now making efforts to produce manure. Among dairy farmers, the low score for economic diversification can be explained by the fact that they limit themselves to selling milk from their dairy to collectors and dairies. On the other hand, if they started processing milk on the farm, it's a safe bet that the score for this principle would go up. But that's not the case at the moment. The principles with highest ratings, such as recycling and animal health, stand out for their direct link with productivity and economic resilience, while those lagging behind, such as soil health and economic diversification, require long-term resources and support. This disparity highlights the importance of prioritising the immediate needs of farmers while preparing strategic actions to integrate less developed agroecological principles. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 27 Figure 13. Rating of agroecological principles on dairy farms 3.2.2. Dairy farmers' level of theoretical knowledge about the meaning of agroecology The majority (54%) of dairy farmers have only a little theoretical knowledge about the meaning of agroecology, but have ambitions to learn more (Figure 14). Those who said they had very little knowledge represented 42% of respondents, while only 4% said they had a clear understanding of what agroecology means. These results show that, in theory, respondents have some knowledge of the term agroecology. This assessment may vary according to the language used or the description of the terminology given to the respondents. However, on these farms, a fair number of agroecological farming practices are implemented by dairy farmers. This shows in more ways than one that, in a practical sense, dairy farmers are implementing agroecological farming practices, even if, as the results show, they are unaware that agroecology is involved. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 28 Figure 14. Dairy farmers' level of theoretical knowledge about agroecology 3.2.3. Dairy farmers' motivations for the agroecological transition Of the 13 statements used to gather farmers' motivations for moving towards more agroecological practices, 10 were favourably received: (i) enjoy nature, (ii) care about nature, (iii) prefer food purchased to be produced and processed in a way that provides a fair wage and good conditions for workers, (iv) prefer to eat locally produced food, (v) prefer to eat food produced without chemical inputs, (vi) switching to agroecological farming is a wise business decision, (vii) taking care of nature and the soil on their farm, (viii) having the power and freedom to solve problems faced by farmers working with other members of the community (ix) having the power and freedom to change agricultural production practices, (x) identifying themselves as an agroecological farmer (Figure 15). On the other hand, farmers recognise that current farming systems are not working well and need to be changed. They also state that decisions about what food to buy are not made primarily on the basis of price. As for the assertion that the farmer lives in a place where most of the population takes good care of the land and nature, opinions were also unfavourable. Figure 15. Motivations for dairy farmers to achieve a more agroecological status January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 29 3.2.4. Agroecological farming practices on dairy farms Several agroecological farming practices are used on dairy farms in the Bobo-Dioulasso dairy basin (Figure 16). Of all the practices, eight (08) were implemented by at least 50% of the farmers: (i) moderate use of herbicides (88%), (ii) depositing manure in the field by parking livestock (80%), (iii) using fodder to supplement livestock feed (76%), (iv) moderate use of mineral fertilisers (75%), (v) storage and conservation of fodder (71%), (vi) use of agroecological equipment (fodder sheds, silos, cart, etc. - 65%), (vii) maintaining mulch on crop plots (64%). The following practices were implemented on less than 50% of farms: (i) crop rotation (47%), (ii) moderate use of animal feed (33%), (iii) use of animal energy (32%), (iv) use of pasture to feed livestock (30%), depositing manure on cultivated fields (23%), growing legumes (18%), and maintaining tree parks on fields (11%). Figure 16. Agroecological farming practices on dairy farms 3.2.5. Agroecological nature of crop and livestock systems on dairy farms Cropping system On dairy farms, which are more livestock-oriented, cropping is also practised. According to the respondents, the soils are moderately fertile and they do not seem to have soil erosion problems. Manure from livestock is mainly used to improve soil fertility. Cropping yields are poor, indicating that farmers have production problems. These poor yields may be linked to the low level of soil fertility, as confirmed by the results of soil organic carbon and organic matter analyses carried out in a dedicated laboratory. However, this factor alone cannot explain the low crop yields recorded. These low yields could be linked to a number of factors, such as: biases in estimating production and area per crop, failure to subtract the area of plots decimated by animals, thus making the production reported per unit area low, and pockets of drought recorded during the reference period for data collection (October 2022-September 2023). Breeding system The farms studied are mainly oriented on cattle production and most of the farmers rely on the farm's own herd for the renewal of their animals, which enables them to better control the risk of diluting the animal genetic resources existing on the farm, to conserve the performance of the best sires and to limit the diseases that can come from other animals outside the farm. These practices do have a negative effect on increasing the risk of inbreeding if reproduction is not properly controlled. The farms collect sufficient manure, which is used to improve soil fertility or sold to other farmers. This limits the use of mineral fertilisers. Most respondents say they want to ensure that their animals are fed, watered, sheltered and cared for properly. Watering is January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 30 mainly provided by surface water and groundwater, and animal health is managed more through conventional veterinary care that does not comply with agroecology principles. It should be noted, however, that there is strong interest in herbal remedies or veterinary pharmacopoeia to support conventional care in animal health management. With the right support, this trend could be reversed and a genuine start made on managing animal health in a more agroecological way. 3.3. Module 3: Performance 3.3.1. Overall performance indicators for dairy farms From an agronomic point of view, the crop and soil health indicators show high values (Figure 17). On the basis of these results, one might expect a high level of soil fertility, whereas the results of soil organic carbon analyses have shown that these soils are not very fertile (see section 3.1.7). This shows that it is more important to rely on the proven results of soil analysis in the laboratory than on statements that are, after all, subjective. As far as animal health is concerned, despite a high degree of variability, this indicator shows an average value overall. This may be linked to a number of factors, such as: (i) unsatisfactory cow feeding practices; (ii) failure to scrupulously adhere to prophylactic programmes; (iii) variations in the availability of water or fodder resources, sometimes influenced by climatic or economic conditions, and so on. In environmental terms, the indicators for energy use, tree diversity and seed variety diversity are high. On the other hand, the indicators for climate mitigation and animal diversity show relatively average performance. In the dairy farms where data was collected, there were at most four categories of animal: cattle, sheep, goats and poultry. This may justify the average value for animal diversity on dairy farms. However, performance is weak when it comes to landscape complexity and reducing water stress. From an economic point of view, household income indicators and their stability show low and moderately stable values, highlighting a notable economic fragility of dairy farms. This is due to the fact that these households' sources of income are mainly based on the sale of milk. Income stability is undoubtedly affected by seasonal fluctuations in the quantities of milk sold. During the rainy season (June to October), these households will tend to have high incomes, whereas in the dry season, incomes may plummet due to low milk production. In social terms, the indicators show high values, revealing a very good level of well-being among household members. The very good level of social well-being of households can be explained by factors such as: (i) sufficient access to basic services (education, health), (ii) a balanced distribution of tasks between household members, allowing for a better quality of life. However, dietary diversity is judged to be average, and access to land is highly variable. The average dietary diversity can be explained by the fact that food consumption is limited to a few basic products, as these farms do not diversify their food crops very much, and buy little food on the markets. This may also indicate a low capacity to access a variety of foods due to financial or logistical constraints. As far as access to land is concerned, those with sufficient financial means sometimes manage to acquire farmland. For the majority, who have modest financial means, access to land is complicated. On the other hand, the other methods of accessing land, such as gifts, inheritance and sharecropping, are not as easy because of the ever-increasing market value of land. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 31 Figure 17. Overall performance indicators for milk-producing farms in the Bobo-Dioulasso dairy basin 3.3.2. Local performance indicators for dairy farms Structural characteristics of farms Mini-farmers are dairy farmers with the largest areas of cultivated land (5.25±2.91 ha), followed by agropastoralists who experimented the AE technologies (3.50±1.75 ha) and other agropastoralists come last with the small areas held per farmer (1.96±1.58 ha). In terms of cattle herd size, agropastoralists who experimented the AR technologies had the largest herds (26.80±22.86 TLU), compared with mini-farmers (11.64±7.9 TLU) and other agropastoralists (13.28±10.60 TLU) (Table 9). There was no difference in the number of cows milked each season between these groups of dairy farmers. On average, 6 cows were milked in the cold dry season and in the rainy season per farm, whereas five were milked in the hot dry season. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 32 Table 9. Some structural characteristics of dairy farmer groups Parameters Units Average Median 8 Mini-farms (4%) 52 AP who experimented AE technologies (25%) 144 other AP (71%) p-value Total surface area ha 2.49 2.00 5.25± 2.91 a 3.50 ± 1.75b 1.96 ± 1.58 c <0,001*** Proportion of surface area with uncontested ownership % 88.6 100 94 ± 18 84 ± 34 90 ± 25 0,291 Size of cattle herd TLU 16.72 12.50 11.64±7.9 b 26.80±22.86 a 13.28±10.60 b <0,001*** No. of cows milked in cds U 6 5 5±0 5±1 6±4 0,282 No. of cows milked in hds U 5 4 4±0 4±1 5±3 0,082 No. of cows milked in rs U 6 5 5±0 5±1 6±4 0,186 Source: HOLPA survey, Bobo-Dioulasso, 2023-2024. AP: Agropastoralists; ha: hectare; TLU: Tropical Livestock Unit (1 TLU = 1 bovine of 250 kg live weight); U: unit; cds: cold dry season (November, December, January, February); hds: hot dry season (March, April, May); rs: rainy season (June, July, August, September, October). Local performance indicators for dairy farms The local indicators in Table 10 show the diversity of the three groups of dairy farmers selected. The mini-farmers group accounts for 4% of farmers. They have the highest number of manure pits per farm (on average one manure pit), the largest quantities of fodder produced (1,457 kgDM/TLU of quality fodder and 1,210 kgDM/TLU of coarse fodder), the largest quantity of manure produced (5 tonnes/ farm) and the highest number of fodder storage facilities (2 facilities on average/ farm). All the mini-farmers produce fodder and have the largest areas secured by title deeds. They spend the most on veterinary inputs (at least 3,000 FCFA/cow/year). The majority of mini-farmers (62%) have received training in innovative farming practices and farm management Agropastoralists who experimented AE technologies account for 25% of dairy farmers. Like mini-farmers, they also have the highest number of manure pits per farm (an average of one manure pit). All the farmers in this group produce fodder. The quantities of fodder produced are average (412 kgDM/TLU of quality fodder and 728 kgDM/TLU of roughage). They spend an average of 2,500 FCFA/cow per year on health monitoring. The majority (67%) have received training in innovative farming practices and farm management. The other agropastoralists make up the majority group (71% of respondents). They have the lowest quantities of quality fodder and roughage distributed per TLU (respectively 89 an 382 kgDM/TLU/yr) They have the best milk yields (3.7 L/d/cow in the rainy season compared with 2.44 L/d/cow in the mini-farmers). This result is surprising because these agropastoralists mainly rear zebus, unlike the mini-farmers, who mainly rear cross-bred cows (zebus x exotic dairy breeds). On the question of milk yields, the opinions of participants at the HOLPA survey results workshop were divided. For some, the low quantities observed seem consistent and can be explained by several factors: poor adaptation of the cows to local climatic conditions, feeding difficulties, or inadequate management of the animals. On the other hand, others believe that these results could be biased due to errors in reporting the amount of milk produced per cow or problems with data entry. These agropastoralists produce the smallest quantities of manure, but they lead the way in terms of soil fertility. This can be explained by the fact that the animals rotate around their plots in the dry season. Finally, this group of farmers spends the least on animal health care (1,500 FCFA/cow). January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 33 Table 10. Local performance indicators by type of dairy farm Indicators Units Average Median 8 Mini-farms (4%) 53 AP who experimented AE technologies (25%) 144 Other AP (71%) p-value Agronomic dimension Number of manure pits per farm U 0 0 1±1a 1±1a 0±0b <0,001*** Quantity of quality forage kgDM/TLU 225,15 200,00 1457 ±2644a 412±330b 89±177c <0,001*** Quantity of roughage kgDM/TLU 502,82 447,50 1210±1000a 728±457b 382±514c <0,001*** Storage equipment (hayloft, silos, forage sheds) U 1 1 2±1a 1±0b 1±1b <0,001*** Quantity of milk produced per cow in cds L/cow/d 2,6 2 2,31±0,88 1,97±0,25 2,78±3,07 0,154 Quantity of milk produced per cow in hds L/cow/d 1,69 1,5 1,81±0,88 1,47±0,12 1,76±1,00 0,109 Quantity of milk produced per cow in rs L/cow/d 3,22 2 2,44±1,24b 2,02±0,14b 3,70±2,24a <0,001*** Proportion of forage farmers % 54 100 100±00a 100± 00a 42±39b <0,001*** Quantity of OM produced kg 3372,57 3000 5125±834a 3961±1804b 2778±1689c <0,001*** Quantity of OM spread per ha kg/ha 1651,1 1600 1837±644 1721±670 1720±842 0,92 Proportion of soil organic carbon % 0,89 0,85 0,45±0,21c 0,74±0,28b 0,96±0,36a <0,001*** Proportion of soil organic matter % 1,53 1,45 0,78±0,36c 1,28±0,48b 1,66±0,61a <0,001*** Environmental dimension Areas secured through title documents ha 2,25 2 4,62±2,20a 3,11±2,00b 1,82±1,55c <0,001*** Percentage of users of biodegradable packaging % 1 0 12±35a 2±14b 0±0b 0,001** Economic dimension Fodder production costs per farm FCFA/ha/year 65700 37500 84375±59300 56700±42600 57200±42600 0,239 Annual cost of monitoring a cow's health FCFA/cow/year ~1500 ~1500 3000a 2500b 1500c <0,001*** Social dimension Training on innovative farming practices and farm management % 38 0 62±25a 67±19a 26±12b <0,001*** Raising awareness of the rules for living together % 81 100 100±0 75±44 83±38 0,187 Source: HOLPA survey, Bobo-Dioulasso, 2023-2024. Legend: AP: Agropastoralists; ; ha: hectare; TLU: Tropical Livestock Unit (1 TLU = 1 bovine of 250 kg live weight); U: unit; L/cow/d: litre per cow per day; kgMS: kilogramme of dry matter; 1 USD = 605 FCFA; cds: cold dry season (November, December, January, February); hds: hot dry season (March, April, May); rs: rainy season (June, July, August, September, October). 3.3.3. Correlation between degree of agroecology and level of farm performance On the basis of the criteria established to categorise dairy farms according to their level of agroecological transition, we ended up with two categories: farms showing a medium level of agroecological transition and those showing a high level of agroecological transition. The correlation between these two levels of agroecological transition and performance was established by considering local indicators. Local indicators were chosen because upstream stakeholders in the dairy value chain January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 34 consider them to be the most important indicators. This was confirmed at the HOLPA survey results workshop when all the indicators (global and local) were ranked in order of importance (see Orounladji et al., 2024a). For indicators such as: production of manure, proportion of land with undisputed title, quantity of roughage, area receiving manure, farms showing a high level of agroecological transition also show the best performance (Figure 18). On the other hand, for indicators of soil fertility levels (organic matter and organic carbon). Farms with an average level of agroecological transition show the best performance. Finally, there was no visible difference between the two levels of agroecological transition for the following performance indicators: quantity of milk produced during each season, quantity of manure applied per hectare, annual cost of fodder production, level of training in farming practices and farm management, quantity of quality fodder distributed and number of storage facilities. On the basis of these results, there is no strict correlation between the degree of agroecology on farms in the Bobo-Dioulasso dairy basin and their level of performance. Figure 18. Link between the degree of agroecology and the level of performance of dairy farms Legend: Medium: households showing average progress in agroecological transition with agroecology scores between 2.34 and 3.66; High: households showing high progress in agroecological transition with agroecology scores between 3.67 and 5. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 35 4. Feedback from the workshop on the results of the HOLPA tool test The attendants to the HOLPA result presentation workshop (Orounladji et al., 2024a) generally appreciated the results presented, which reflect their context, the status of agroecology on dairy farms and the performance indicators for these farms. The participants also became aware of the importance of closely monitoring the size of their herds, knowing their numbers by age category and by type of animal species, as well as their productivity. This approach has enabled them to better assess the resources available and adapt their practices. Several directions have been identified for the future. Firstly, a change of behaviour in farming and livestock rearing practices is needed to promote more agroecological production. This includes phasing out the use of pesticides and herbicides in favour of environmentally-friendly methods. Participants also stressed the importance of maintaining and intensifying the production of fodder and manure, which are essential for improving the fertility of impoverished soils. Awareness of the use of manure has motivated farmers to adopt more sustainable practices to preserve and restore their farmland. Diversifying sources of income is also a priority. In addition to selling milk, participants identified economic opportunities in the production and marketing of fodder seeds and manure. This diversification could strengthen the economic resilience of farms while meeting local needs. Finally, it is essential to continue raising awareness and building the capacity of farmers to maintain these positive dynamics and further improve farm performance. Stakeholders have also proposed strategies for moving more quickly towards achieving the DIP vision. These strategies, presented below, are grouped into three categories according to the stakeholders most concerned by their implementation: Farmers  Develop a network of seed-producing farmers to increase seed availability.  Fencing off production sites.  Composting in piles.  Strengthening solidarity between farmers to facilitate access to seeds.  Promoting the best farmers (champions) to encourage emulation.  Ensuring that animals are kept on plots to improve soil fertility.  Better water management for fodder production. Ministry of Agriculture, Animal Resources and Fisheries  Improve seed distribution to farmers.  Promoting the milk sector through the agropastoral offensive.  Supporting DIP in fodder production.  Supporting farmers to improve fodder production.  Encourage widespread adoption of forage production technologies.  Rewarding the best farmers to motivate everyone involved. Other partners  Subsidise the purchase of equipment, particularly shredders.  Training farmers in innovative farming practices, farm management and equipment use.  Install boreholes to secure fodder production in the dry season.  Train trainers to support farmers in the use of manure pits and heap composting.  Working with farmers to improve fertilisation practices.  Build the capacity of farmers not involved in the project to increase fodder production. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 36  Strengthen farmers' capacity to use the heap composting technique, which is considered more accessible and better suited to their needs. On the fringes of this workshop, a number of suggestions were put forward to improve practices and consolidate achievements. These included :  Make the results of soil analyses available to the farmers surveyed via the DIP. This would give them precise information about the fertility of their soils and enable them to adjust their practices accordingly.  In a few years' time, carry out a new evaluation using the HOLPA tool with the same farmers. This will enable us to measure the impact of the agroecological packages introduced on certain farms and assess the progress made. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 37 5. Use of assessment results The results obtained from this evaluation of the agroecological performance of dairy farms established into the Bobo-Dioulasso dairy basin are very important and will help to guide decision-making at national and local level. They are invaluable in more ways than one for the stakeholders of the DIP, specifically through the results of the local indicators, which will serve as a compass to determine the current state of agroecology on the dairy farms of Bobo-Dioulasso and the decisions that can be taken to support the achievement of the DIP's vision. It should be remembered that these local indicators were identified and prioritised by the stakeholders before the data was collected. The most significant results that DIP stakeholders can focus on are indicators relating to fodder production, fodder storage, milk production, manure production and soil fertility. The results on fodder production and storage, manure production and soil fertility will be better used by farmers, livestock breeders, agricultural and livestock technical services and researchers. The results on milk production are very useful for all the actors of the dairy value chain (farmers, collectors, processors, consumers), not forgetting the technical services for agriculture and livestock, researchers, local authorities and the Burkina Faso government. The revitalisation of the DIP in recent years has given the local milk sector greater credibility, with the government of Burkina Faso including the sector in its new policy known as the "agropastoral offensive". This shows the extent to which milk production indicators are highly beneficial in new policy decisions concerning the milk sector. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 38 6. Lessons learned Data was collected from households and milk-producing farms to assess their performance in general and their agroecological performance in particular. The results provide comprehensive information on the three modules (context, agroecology and performance) of the HOLPA tool. Integrating local indicators into the data collected has enabled the results to be better contextualised so that they are more meaningful to local stakeholders. However, the HOLPA tool deserves to be further improved in terms of agroecological practices for agropastoral systems. In a recent publication (Vall et al., 2023), we showed that in agropastoral systems in western Burkina Faso, crop-livestock integration and the recycling of crop and livestock by-products into fodder and manure make a major contribution to the agroecological characteristics and performance of these farming systems. The practices concerned are as follows:  Storage of crop by-products (straw, tops) for fodder purposes  The production of manure and compost from livestock and crop by-products in night pens and manure pits  Night-time stabling of herds in fields for fertilisation purposes  Rational management of organic fertilisation of fields  The use of animal power for tillage and transport However, it has to be said that these practices are not taken into account in the overall indicators of the HOLPA tool. The absence of these indicators in the HOLPA tool will have the effect of obscuring a very large proportion of the agroecological characteristics of the farming systems studied, not only in Burkina Faso but also in countries where agropastoral systems are very common, which is the case in many regions of Africa, Asia and tropical Latin America, both dry and humid. Also, the tool does not address the doses of organic and mineral fertilisers applied per crop, and this lack of information means that it is not possible to quantify fertiliser applications per crop and per unit area. It would be important to integrate this information into the consolidation of the HOLPA tool. At the current stage, the HOLPA tool is being used to determine soil fertility levels using two approaches. The first approach is based on respondents' declarations in the household questionnaire, where they are asked to choose a category (not very fertile, moderately fertile or very fertile) according to their assessment of the level of fertility of their soil. The second, and more objective, approach involves taking composite soil samples from farms for laboratory analysis. This two-pronged methodology sometimes leads to contradictions in the results obtained. This was also the case with the results for dairy farms, where using the first approach (based on declarations), the majority of respondents (89.22%) stated that the soils were moderately fertile (compared with 10.78% who said they were not very fertile). In contrast, the second approach (laboratory analysis) showed that the soils are not very fertile. For this reason, we propose that the HOLPA tool should be limited to the results of laboratory analyses, to the detriment of declarations, if such analyses carried out at all The method used to assess agroecology indicators and performance is qualitative and does not provide sufficiently tangible results for decision-making purposes. Although we can say that proportions/percentages have been used to assess the agroecology scores, these proportions/percentages do not allow an objective comparison to be made between two ALLs, whether they are in the same country or in two different countries. This state of affairs shows that a farm can be at a high level of agroecological transition in one ALL and, by superimposing it on another ALL, find itself at another level (low or medium) of agroecological transition. For example, at the same percentage of manure production on a farm (25%), for 75% of purchased manure, this may well represent a production of 500 kg on one farm and 5,000 kg on another, which in both cases does not represent the same effort in manure production (10 times greater in the case of the second farm which produces and purchases much more manure). In terms of proportions/percentages, there is no difference, whereas in terms of quantity, the difference is clear. To this end, we propose to reformulate these types of modalities in the assessment questionnaire by using intervals for numbers, quantities, volumes, etc. instead of proportions/percentages. With regard to the definition of agroecological levels, the HOLPA tool should clearly describe how the classification will be carried out so that any user knows how to rank farms according to their degree of agroecological transition. To establish the link between levels of agroecology and levels of farm performance, in our case we used a method of classifying farms into three categories, detailed in the data analysis section, which could be adopted for HOLPA. Once the tool has been improved, HOLPA's digital platforms (mobile application & web interface) should be created and made available with a language tab so that the assessor can choose the language he or she wants. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 39 In terms of lessons learned, at the workshop held to present the results of the HOLPA survey (Orounladji et al., 2024a), the stakeholders acknowledged the effectiveness of the HOLPA tool in assessing the agroecological performance of farms. Although the survey was time-consuming according to the respondents, it yielded relevant results, particularly on the state of soil fertility, the level of fodder production (proportion of farmers, quantity produced, types of fodder species, etc.), and the agroecological status of dairy farms in the Hauts-Bassins region. The presentation of the results was seen as a highly beneficial initiative. It enabled the participants to gain a better understanding of the agroecological status and performance of the farms, as well as the impact of certain practices, such as the use of pesticides and herbicides, which were considered to be non-agroecological. This awareness changed their initial perception, some of whom had previously regarded the survey as a mere constraint. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 40 7. Conclusion and next steps Most of the respondents to the HOLPA survey conducted on 204 dairy farms located in the AEI intervention area in Burkina Faso (an area corresponding to the Bobo-Dioulasso milkshed and the Agroecological Living Landscape intervention area) were men (93.63%), illiterate (71.08%), of Peulh origin (81.37%) and living with a partner (80.88%). On the whole, they said they were satisfied with their living conditions, with the exception of concerns about insecurity. The men expressed a feeling of freedom in their decision-making, unlike the women. A majority of respondents (54%) said they had very little theoretical knowledge of agroecology. For them, the main motivations for adopting agroecological principles are as follows: (i) enjoying nature, (ii) caring about nature, (iii) I would prefer the food I buy to be produced and processed in a way that provides a fair wage and good conditions for workers, (iv) preferring to eat locally produced food, (v) preferring to eat food produced without chemical inputs, (vi) switching to agroecological farming is a wise business decision and (vii) caring for nature and the soil on their farm. The farms are mainly oriented on cattle production, with the renewal of livestock mainly provided by the farm's own herd. Most respondents said they wanted to ensure that their animals were fed, watered, sheltered and well cared for. Watering is mainly provided by surface water and groundwater. In terms of feed, the practice of fodder cultivation remains limited, but in the context of growing insecurity and reduced grazing space, dairy farmers are showing more and more interest in this practice which would allow them to better feed their animals in safety. This is indicative of the strong pressure exerted on pastures despite the poor security situation. There is also strong interest in manure collection. However, on these livestock-oriented farms, where agriculture is also practised, there are major soil fertility problems. Pests are managed mainly by using resistant varieties, most of which are local varieties, although these are not known for their high yields. Crop yields in 2023 were poor. Given that most farmers are livestock farmers, field sizes are not very large, but herd sizes are high. Normally, we would expect better crop yields because the herd dairy farmers produce large quantity of manure available to improve soil fertility. However, this was not the case, and the low yields can be explained by a number of factors, such as: biases in estimating yields, failure to subtract the area of plots decimated by animals, making the reported production per unit area low, pockets of drought recorded during the reference period for data collection, manure given or sales out of the farms. When we look at the agroecological farming practices (according to the HOLPA terminology) implemented on the farms, we find that at least 60% of farmers implement the following practices on their farms: (i) moderate use of herbicides; (ii) depositing manure in the field by parking livestock; (iii) using fodder to supplement livestock feed; (iv) moderate use of mineral fertilisers; (v) storing and conserving fodder; (vi) using agroecological equipment (fodder sheds, silos, carts, dumpers, etc.). Looking at the agroecological status of dairy farms in the Bobo-Dioulasso milkshed, the study showed that these farms are at an average level of progress in the agroecological transition. The exercise of establishing a correlation between the degree of agroecology and the level of performance of these farms showed that a strict link does not exist, or at least not at the current stage of testing the HOLPA tool. The results obtained from this study constitute a baseline for the agroecological performance of Bobo-Dioulasso milkshed dairy farms. They will be used to guide policy decisions relating to the dairy sector in Burkina Faso, and provide a benchmark against which results can be compared in future years as part of an impact assessment. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 41 8. Acknowledgements The authors would like to express their deep gratitude to Hati Konaté, President of the Dairy Innovation Platform, and to all the stakeholders of the Bobo-Dioulasso Agroecological Living Landscape for their collaboration and commitment throughout this study. They would also like to warmly thank Raoul Zoundi and Amadou Barry for their exemplary dedication to data collection. Special thanks go to Sarah Jones and Andrea Cecilia Sanchez of the Bioversity International Alliance & CIAT for their valuable contribution to the data analysis. Finally, the authors would like to express their gratitude to Maryline Darmaun for her guidance and support throughout the preparation of this report. January 25|Agroecological Performance Assessment of Dairy Farms in the Bobo-Dioulasso Milkshed using the HOLPA tool 42 9. References Bryer J, Speerschneider K, 2016. Likert: Analysis and Visualization Likert Items. R package version 1.3.5; https://CRAN.R- project.org/package=likert Darmaun M, Chevallier T, Hossard L, Lairez J, Scopel E, Chotte J L, Lambert-Derkimba A, de Tourdonnet S, 2023. Multidimensional and multiscale assessment of agroecological transitions. A review. International Journal of Agricultural Sustainability, 21(1). https://doi.org/10.1080/14735903.2023.2193028 de Mendiburu F (2023). agricolae: Statistical Procedures for Agricultural Research. R package version 1.3-7. https://CRAN.R- project.org/package=agricolae Geck M S, Crossland M, Lamanna C, 2023. Measuring agroecology and its performance: An overview and critical discussion of existing tools and approaches. Outlook Agric. 52, 349-359. https://doi.org/10.1177/00307270231196309 INSD, 2022. 5th RGPH. Fifth General Cen