Exploring opportunities for scaling of sustainable intensification interventions through farmer engagement and positive deviance approach a case for Basona Worena, Ethiopia Daan Andreas August 2023 Farming Systems Ecology Group Droevendaalsesteeg 1 – 6708 PB Wageningen - The Netherlands i Exploring opportunities for scaling of sustainable intensification interventions through farmer engagement and positive deviance approach a case for Basona Worena, Ethiopia Name student: Daan Andreas Registration number: 1160788 Course code: FSE80436 Supervisor: Jeroen Groot Co-supervisor: Kindu Mekonnen (ILRI) Examiner: Lenora Ditzler Date: 15-08-2023 Chair group: Farming Systems Ecology Address: Droevendaalsesteeg 1 – 6708 PB Wageningen – The Netherlands Words: 8490 Front page image: A farmer herding his livestock on his way to the grazing fields in the Gudo Beret kebele in Basona Worena. Picture made by Daan Andreas on the 21st of December 2022. Ackwowledgements I would like to thank everyone that has helped me complete this thesis over the last year. First, the SI-MFS initiative for opening up the opportunity for me to work on the project. Jeroen Groot for his help with the conceptualization, and his guidance and patience during the process of the thesis. Kindu Mekonnen for his guidance during my time in Ethiopia and the interesting conversations about the Ethiopian agricultural context. Also Jim Hammond for providing me with the RHoMIS data that I used for the positive deviance analysis. Temesgen for his help during the fieldwork sessions and workshops. Tesfa for all his translation work, but also the friendship and fun activities that we could do together. Finally I would like to thank, all other friends and family that have helped me during the process. I will be forever grateful. i Abstract Enhancing African farming systems holistically, through interventions that introduce innovative practices, is crucial for ensuring ample food production and sustaining food security. Given the diverse nature of farming systems, validating these interventions in specific contexts becomes essential to optimize their effectiveness and adoption rates. Therefore, this research was conducted in Gudo Beret and Goshe Bado, two research kebeles situated in the Basona Worena district, Amhara, Ethiopia, which predominantly features mixed crop-livestock systems and struggles with challenges tied to climate variability, soil erosion, and land degradation. Employing a two-step methodological approach, this study examined sustainable intensification within these farming systems. In the initial phase, a positive deviance analysis was carried out based on a RHoMIS dataset containing 238 households using indicators from the RHoMIS dataset to identify exemplary farms. In the subsequent step, two parallel methodologies were employed. Firstly, the FarmDESIGN model facilitated an in- depth analysis of 16 case study farms; 6 positive deviant and two negative deviant farms per village. This offered a comprehensive assessment of their performance. Secondly, a participatory research methodology was conducted. This entailed organizing three workshops containing ten participants each. During the workshops an adapted version of the serious game ReHab was employed, engaging stakeholders to explore agile scaling opportunities for sustainable intensification interventions. The positive deviance analysis successfully identified 23 exemplary farms and 215 sub-optimal dominated systems based on the subset of indicators adopted from the Sustainable Intensification Assessment Framework. The exemplary farms on average scored higher on all indicators except gender control. Thus, these farms still demonstrated trade-offs and compromises between indicators. Furthermore, the analysis showed that the most prominently trialled interventions were related to improved varieties. It was also shown that the adoption rate of the interventions did not differ significantly between the exemplary and sub-optimal dominated systems. The farm system modelling aspect provided an all-encompassing evaluation of farm performance across productivity, economic, and environmental dimensions. The exemplary farms in this analysis scored higher on most of the indicators, but not on all. Therefore, also here it was shown that trade-offs and compromises occur in these systems. Through the participatory research methodology, employing the ReHab serious game, local stakeholders actively delved into sustainable intensification intervention possibilities, yielding valuable insights on decision-making, crop preferences, and fertilizer strategies. The workshops underscored farmers' preference towards improved varieties and the significance of organic resources, while shedding light on challenges like limited technology access and labour availability. By synergizing these methodologies, the study aimed to integrate and validate both new and existing knowledge related to sustainable intensification. This comprehensive approach offered a deeper understanding of the dynamics within the local farming systems. Ultimately, the research adds to the ongoing endeavours of enhancing sustainable agricultural practices, addressing obstacles, and facilitating the scaling of sustainable intensification interventions in the Basona Worena district and comparable regions. Key words: sustainable intensification, positive deviance approach, SI interventions, farm system modelling, FarmDESIGN, participatory research methodology, serious gaming, scaling i List of abbreviations Africa RISING Africa Research in Sustainable Intensification for the Next Generation CGIAR Consultative Group on International Agricultural Research ETB Ethiopian Birr GHG Greenhouse Gas MAE Male Adult Equivalent masl metres above sea level MIDIP Mean Indicator Distance to Ideal Point ND Negative Deviant PD Positive Deviance R4D research for development RHoMIS Rural Household Multi-Indicator Survey SI Sustainable Intensification SIAF Sustainable Intensification Assessment Framework SI-MFS Sustainable Intensification of Mixed Farming Systems SSA Sub-Saharan Africa TLU Total Livestock Units List of Figures FIG. 1: AN OVERVIEW OF THE METHODOLOGICAL FRAMEWORK ................................................................. 3 FIG. 2 A MAP OF ETHIOPIA ........................................................................................................ 4 FIG. 3: VISUALIZATION OF PARETO-OPTIMALITY. ............................................................................... 5 FIG. 4: SI INDICATOR PERFORMANCE BASED ON PD ANALYSIS ............................................................... 8 FIG. 5: TOTAL TIMES A CERTAIN NUMBER OF INTERVENTIONS WAS TRIALLED BY INDIVIDUAL FARMERS ...................11 FIG. 6: OVERVIEW OF THE NUMBER OF TIMES AN INNOVATIONS HAS BEEN TRIALLED .......................................11 FIG. 7: PROPORTIONS OF SUB-GROUP THAT HAVE TRIALLED CERTAIN INNOVATIONS. ......................................12 FIG. 8: INDICATOR SCORES FOR THE SIXTEEN CASE STUDY FARMS ...........................................................14 FIG. 9: FLOW DIAGRAMS WORKSHOP DECISION MAKING .......................................................................15 FIG. 10: PICTURES FROM THE THREE WORKSHOPS. ............................................................................16 List of tables TABLE 1: DESCRIPTIVE STATISTICS OF INDICATOR SCORES FOR FARM POPULATION ......................................... 7 TABLE 2: SIAF INDICATOR CORRELATION OVERVIEW .......................................................................... 9 TABLE 3: OVERVIEW OF THE INTERVENTIONS TRIALLED BY THE SELECTED CASE FARMS ...................................13 ii Table of content Abstract .............................................................................................................................. i List of abbreviations ............................................................................................................ ii List of Figures ..................................................................................................................... ii List of tables ...................................................................................................................... ii 1. Introduction ................................................................................................................... 1 1.1 Background ............................................................................................................... 1 1.2 Research focus and problem statement ......................................................................... 1 1.3 Objectives ................................................................................................................. 2 1.3.2 Specific objectives ................................................................................................. 2 1.3.3 Research questions ............................................................................................... 3 2. Methods ......................................................................................................................... 3 2.1 Methodological framework ........................................................................................... 3 2.2 Case study site description .......................................................................................... 4 2.3 Positive Deviance Analysis ........................................................................................... 5 2.3.1 SIAF indicators ..................................................................................................... 5 2.3.2 Pareto ranking ...................................................................................................... 5 2.3.3 Identification of exemplary and other systems .......................................................... 5 2.3.4 Farm system modelling and analysis ........................................................................ 6 2.5 Participatory Research Methodology .............................................................................. 6 3. Results .......................................................................................................................... 7 3.1 Positive Deviance Analysis ........................................................................................... 7 3.1.1 Farm population description ................................................................................... 7 3.1.2 Correlation between indicators ................................................................................ 7 3.1.3 Identification of positive deviant farms and characterization of indicator performance .... 7 3.1.4 Innovation descriptive statistics ............................................................................ 11 3.2 FarmDESIGN .............................................................................................................13 3.2.1 Case study farm interventions .............................................................................. 13 3.2.2 FarmDESIGN indicators ........................................................................................ 14 3.3 Workshop and gaming ................................................................................................15 4. Discussion .................................................................................................................... 17 5. Conclusion .................................................................................................................... 19 References ....................................................................................................................... 20 Appendix 1: Stakeholder engagement ................................................................................. 23 Appendix 2: List of indicators according to the SIAF. From Hammond et al. (2021) ................... 23 Appendix 3: All indicator score overview .............................................................................. 24 Appendix 4: Complete FarmDESIGN survey.......................................................................... 25 Appendix 5: Cumulative crop areas of the case study farms ................................................... 36 Appendix 6: Feed balance of the case study farms ................................................................ 36 Appendix 7: Workshop content ........................................................................................... 37 Appendix 8: Workshop results ............................................................................................ 41 iii 1. Introduction 1.1 Background The food supply of the African continent is still highly reliant on a large population of smallholder farmers (Schut and Giller, 2020). At least two-thirds of the African rural households are dependent on agricultural production (Pretty et al., 2011). Together this population is responsible for the production of 90 percent of the food consumed on the continent (Mwaniki, 2006). These farms often operate using a mixed-farming approach, in which crops are cultivated to provide food and cash, while the residues of the crops can be used as fodder for animals. The animals subsequently provide food, and additionally supply manure and draft power (CGIAR, 2021). Currently these farms are predominantly subsistence oriented, meaning that food is mostly produced for household consumption (Tittonell et al., 2010). However, it has been shown that most of these farmers are net food consumers, indicating that the households rely on external income sources to provision their dietary needs (Schut and Giller, 2020). On top of that, it is projected that the African population will more than triple at the end of this century (Roser, 2013). The increase in population will put additional pressure on the African smallholder farmers to produce ample food to sustain food security for the continent’s population (Schut and Giller, 2020). However, paradoxically smallholder farmers in Sub-Saharan Africa (SSA) are often characterized by low productivity, while the need for increased productivity is more urgent than ever (Ittersum et al., 2016). Furthermore, biophysical characteristics of land and continuous soil fertility deterioration and soil degradation due to inefficient use of agricultural land, deforestation, and extreme weather events, has limited the increase in agricultural productivity (Breman et al., 2019). Besides, it has been shown that degraded land and the associated yield gaps function as poverty traps, because the initial high investment costs to regenerate soils and increase productivity are hard to overcome (Tittonell and Giller, 2013). Increasing the productivity of these smallholder farmers over the coming decades through intensification and area expansion is therefore essential to feed the increasing African population (VanLauwe et al., 2013). Farm types are heterogeneous and context dependent (Vanlauwe et al., 2014). Environmental diversity and complex socio-economic structures limit the opportunity to develop one- size fit all solutions to increase agricultural productivity and thus food security (Breman et al., 2019). As a consequence, solutions to overcome these constraints are place-based and often require in depth local studies that allow for locally best-fit interventions (techniques and practices) to sustainably intensify the production. The concept of Sustainable Intensification (SI) can be seen as farms obtaining higher outputs, using the same amount of resources without negatively affecting the environmental, social and economic domains (Öborn et al., 2017). Thus, SI could result in equitable outcomes, contributing to an increased livelihood resilience (Struik & Kuyper, 2017). However, even though farmers possess a wide variety of knowledge and experience about the context in which they operate that could help to overcome the consequences of present challenges, such as climate variability, the rate at which these challenges are being expressed is exceeding their adaptation capacity for adequate SI (Meijer et al., 2015). The Sustainable Intensification of Mixed Farming Systems (SI-MFS) initiative of the CGIAR aims to ‘provide equitable, gender-transformative pathways for improving the livelihoods of 1.5 million female and male actors in seven prioritized MFS by 2024’ (CGIAR, 2021). As the SI-MFS initiative builds upon previous research programs there is much opportunity to integrate the extensive research findings and data sources acquired through these programs. Over the coming decades it is of importance to come up with development and adoption strategies of SI interventions. Furthermore, coming up with novel and innovative methods and tools that can support the understanding of what SI interventions work where, in what context and for whom is key to build capacity for the adoption of SI in a wider context (CGIAR, 2021). 1.2 Research focus and problem statement One of the areas where extensive research has been done to validate SI interventions are the Ethiopian Highlands, for example through the Africa RISING research for development (R4D) program (Negra et al., 2020). In the Ethiopian context it is now important to focus on scaling, which can be defined as increasing the intensity of intervention implementation by individual households, 1 and an increase of beneficiaries by reaching larger numbers of farmers through an increase in target districts and research kebeles; the smallest administrative unit in Ethiopia (Gebreyes et al., 2021). Understanding the present status of SI intervention adoption and how this has benefited farming system performance, depending on the farm resource endowment is an important step towards the scaling to other districts with similar characteristics. One of the more novel approaches that has been adopted over the past few years, is the positive deviance (PD) approach, in which according to a specific set of sustainability indicators, a multi-criteria performance analysis can be done to identify exemplary farms that perform exceptionally well, in comparison to the sub-optimal dominated farm systems, which are the remaining farms in the population (Modernel et al., 2018; Steinke et al., 2019; Adelhart-Toorop et al., 2021; Liang et al., 2022). Adelhart-Toorop et al. (2021) have shown that such an approach can be used to redesign farm systems for different farm types through whole-farm modelling using the FarmDESIGN model. For the Basona Worena district, one of the short-term research themes proposed by Seifu et al. (2022) are the community, stakeholders, conversation and discussion platforms. Using such platforms to build capacity for farmers through the engagement using simple serious games or farms system modelling, can be exceptionally valuable for locally scaling out SI intervention (Kirimbu, 2017; Michalscheck et al., 2020). A wide variety of interventions have been promoted to Ethiopian farmers (Hammond et al., 2021). However, through the Africa RISING R4D program this has happened in a somewhat unstructured manner, meaning that farmers were able to choose the interventions by themselves (Jim Hammond, personal communication, September 30 2022). As a result farm implemented different numbers of interventions in random combinations. Now the AfricaRISING project has ended, it would be interesting to see how the adoption of interventions, and in what combinations, has affected farm system performance. Basona Worana is a district characterized by low-productive mixed farming systems, that often lack diversification, of sustainability intensification interventions. Additionally, the area is characterized by low soil quality, limited adoption of crop diversity strategies, low staple crop yield, and low off-farm income. Consequently, these characteristics result in a generally low farm productivity (Hammond et al., 2021). Furthermore, Seifu et al. (2022) in their system analysis assessment of the SI-MFS initiative have come up with a set of research themes associated to the scaling of SI interventions. One example is to come up with strategies to overcome low soil fertility and the adoption of improved varieties are priority in the region. Therefore, this research will set out to come up with methodologies to explore the opportunity for out-scaling of SI intervention in different villages in the Basona Worena district. 1.3 Objectives The main aim of this study is to identify successful exemplary farm households within the diversity in farm system performance as influenced by SI intervention implementation, and to explore the factors that explain farm system performance diversity. Furthermore, it aims to seek opportunities to use new and existing knowledge related to SI interventions in the scaling processes by conduction a participatory research methodology in the Basona Worena district, Ethiopia. 1.3.2 Specific objectives The specific objectives of this study aimed to: 1. Assess farm system performance in Goshe Bado and Gudo Beret, Basona Worena district, Amhara region, Ethiopia, using a PD analysis to identify outperforming farms and the sub- optimal dominated farms in the farm population. 2. Acquire in depth understanding of farm system performance of multiple farms, using farm system modelling. 3. Explore scaling opportunities related to SI intervention implementation by conducting a participatory research methodology. 2 1.3.3 Research questions The research questions of the study are: 1. What is the SI performance of successful exemplary farms, and how does this compare to the dominated sub-optimal farm population? 2. How do the exemplary farm perform in comparison to the sub-optimal dominated systems when analysed through farm system modelling? 3. Does the utilization of serious gaming as a participatory research methodology effectively contribute to the discussion and implementation of SI interventions, resulting in concrete benefits for scaling? 2. Methods 2.1 Methodological framework This research used a two-step approach in which a PD analysis was used as entry point to identify exemplary, sub-optimal dominated, and negative deviant farms. Following the PD analysis two parallel methodological approaches were applied to both further analyse the farm systems using farm system modelling and a participatory research methodology to explore opportunities for agile SI interventions scaling methods (Fig. 1) This research, thus, aimed to combine existing methodologies to test whether such a combination of methodologies would be an efficient and effective way to integrate and validate, new and existing knowledge related to SI. Fig. 1: an overview of the methodological framework consisting of one primary step (yellow) and two subsequent parallel steps (green and blue). The grey box indicates that within the two sub-steps that knowledge related to SI interventions has been integrated within those steps. 3 Step 1: involved conducting a positive deviance (PD) assessment of farms in the Gudo Beret and Goshe Bado kebeles within the AfricaRISING R4D project in the Basona Worena district. The PD analysis employed indicators extracted from an existing RHoMIS dataset, which covered 238 mixed farms across five distinct domains as outlined in the Sustainable Intensification Assessment Framework (SIAF): productivity, economic, environmental, human condition, and social (Musumba et al., 2017). Each domain comprised multiple relevant indicators, with a comprehensive list available in Appendix 1. Initially, the farms underwent Pareto ranking based on seven indicators, aiming to identify the PD farms. Subsequently, a threshold of five out of the seven indicators was established, utilizing mean indicator scores. This approach allowed for the identification of exemplary systems that represented the best compromise across the selected indicators. Step 2A: the utilization of farm system modelling as a comprehensive tool for in-depth analysis of both exemplary and ND farms. The multi-objective optimization model FarmDESIGN was employed for this purpose (Groot et al., 2012). A total of sixteen farms were included, selected from both kebeles. This selection consisted of six exemplary farms and two negative deviant (ND) farms per kebele, with considerations given to farm size. The model was parameterized using a combination of existing research and secondary data sources. Additionally, primary data was collected through interviews conducted with the chosen exemplary and negative deviant farms. Step 2B: The participatory research methodology aimed to engage local stakeholders into farm systems thinking in relation to intervention implementation through simple serious gaming sessions. This was done in an iterative process in which the serious game ReHab (Page et al., 2016) was adapted and developed over a period of three interactive workshops so that it was able to represent the context specific challenges related to SI. The goal of the sessions was to create awareness on the impact that new farm system configurations, through SI interventions implementation, has on farm system performance related to SIAF indicator scores. 2.2 Case study site description This study was conducted in Gudo Beret and Goshe Bado, two Africa RISING research kebeles in the Basona Worena district in the Amhara region found in the Northern Shewa zone in Ethiopia (Fig. 2). The complete district covers an area of 1208 km2, with an elevation level between 2000 – 3800 metres above sea level (masl). The average daily temperature ranges from 6 to 20 °C. The mean annual precipitation of the area is between 900 – 2000 mm per year. The district predominantly falls within the mid and highland agroecological zones. The mean farm size in the region is 1.3 hectares and the dominant farming system is mixed crop-livestock. Major crops that are grow, are barley, Fig. 2 A map of Ethiopia on the left with the Basona Worena woreda in red. The North Shewa zone on the right. 4 wheat, Faba bean, tef, and field pea. These crops are grown in barley-dairy-fattening and wheat- livestock fattening systems, respectively (Seifu et al., 2022). Currently the district is suffering from climate variability, soil erosion and land degradation (Tesfaye and Hammond, 2021). Additionally, a blanket fertilizer recommendation exists in the region, meaning that all farms are recommended to apply the same the amount of fertilizer per hectare. Because the soil fertility in the area is highly heterogeneous, the application of nutrients is often not adequately matched with the requirements of the soil and the crops. Depending on the soil type in a specific cropping system, farmers would need to apply different amount of fertilizer per hectare (Seifu et al., 2022). Over the last decade a diverse number of SI interventions has been implemented in the region. These interventions consist of improved breeds, improved seed varieties, soil water conservation interventions (Tesfaye and Hammond, 2021). 2.3 Positive Deviance Analysis A dataset from the Rural Household Multi-Indicator Survey (RHoMIS) tool that included an in depth survey from 2022 in the district has been used as primary source for the initial analysis. The dataset included a sample of 238 households that had a mixed farming system and were extensively surveyed, in the Gudo Beret and Goshe Bado research kebeles. Hammond et al. (2021) have shown that RHoMIS survey data is suitable data source to assess SI scores. Therefore, a similar approach, using a subset of indicators, has been used to identify PD farms. 2.3.1 SIAF indicators The RHoMIS survey data was used in line with the SIAF (Musamba et al., 2017). Thus, a selection of SI indicators has been made and was then used to identify the PD farms within the population, but also to determine the sub-optimal dominated farms which are the remaining non-PD farms within the population (Liang et al., 2022, see Fig. 3). Appendix 1 includes a complete list of indicators that was used to assess farm performance, and includes the Fig. 3: Visualization of Pareto-optimality in productivity, economic, environmental, human, and social which the Pareto frontier is represented by the red line on which positive deviant domains. The following seven indicators have been selected farms are situated. The black lines for the Pareto ranking process as these indicators allow for connecting the positive deviants from the a holistic assessment covering each of the SIAF domains: ideal point represent the distance to the ideal point. Furthermore, the grey dots crop productivity (kCal ha-1 yr-1), diversity (number), gross represent the dominated sub-optimal farm margin (ETB ha-1 yr-1), crop market orientation (%), soil population. The yellow dots show the quality (number), HDDS (number), gender female control extreme cases in the population where (%). strong trade-off between indicators exist. Finally, the exemplary farms show the best compromise between their goals and 2.3.2 Pareto ranking indicator outcomes. Positive deviant farms were identified through the process of Pareto ranking in which the well-performing individuals are s eparated without assigning subjective weights to multiple indicators (Groot and Rossing, 2011). Based on the set of indicators the positive deviant analysis was conducted using the ParetoRanker tool. Farms that had missing data on specific indicators were excluded from the analysis (Adelhart Toorop et al., 2021). Pareto optimal solutions lie on a frontier (in case of 2 objectives, and a surface for >2 objectives) while non-Pareto optimal solutions, the dominated sub-optimal farms, lie below the line (Adelhart Toorop et al., 2021, see Fig. 3). 2.3.3 Identification of exemplary and other systems Pareto-ranking is able to identify many possibilities that are scoring exceptionally well on one indicator, but score low on other indicators, so called extreme cases (Adelhart-Toorop et al., 2021). These cases might indicate that trade-offs exist between indicators. Therefore, to overcome the 5 inclusion of farms with unbalanced performance a threshold value for PD farms was set. As a result, only farms were included that performed well on most indicators without significant trade-offs to occur. Therefore, a farm is only defined as exemplary when it performs better than the mean indicator score on five out of seven selected indicators. This analysis has identified the exemplary farms within the population of survey participants in the Basona Worena district. Furthermore, the dominant sub-optimal population can be identified as the remaining farms that are not PD farms. 2.3.4 Farm system modelling and analysis Based on the PD analysis outcome for each kebele a selection of six exemplary farms including two small, medium, and large farms was made. Additionally, for both kebeles, two negative deviant farms were selected. These farms were visited, and the household head was interviewed using a survey that can be found in Appendix 4. The FarmDESIGN model is a tool that allows for in depth evaluation of farm system performance on productive, economic and environmental scale, through multi-objective optimization (Groot et al., 2012). A repository farm was created in the FarmDESIGN model based on previous research and secondary data. The model was set up so that the known crops, animals, and input in the region were already parameterized. By doing this a solid base for further exploration of the diversity of farms is created and the methods for modelling specific farms was more agile. In the end, the aim was to create deeper understanding of the characteristics of the existing farming landscape in the Basona Worena district using FarmDESIGN by assessing farming system performance. 2.5 Participatory Research Methodology The second parallel step (Fig. 1) of the methodological framework consisted of a participatory research methodology. Three stakeholder engagement sessions were organized to stimulate knowledge transfer and to create awareness about the possibilities of SI intervention implementation. Michalscheck et al. (2019) have shown that serious games can be a time and cost- effective tool to explore household dynamics and land-use decisions. Therefore, this research additionally aimed to create a low-tech simple serious game that allows the farmers and other stakeholders to explore possibilities in adoption of SI interventions. Eventually, the aim was to assess whether such methodologies and tools can serve as suitable options to support scaling of SI interventions using existing data and knowledge. ReHab is a game that supports experimental learning in which players have to make decisions in the field of renewable resource management (Page et al., 2016). Before organizing the workshops an adaptation of the serious game ReHab was made based on the results from the PD analysis and RHoMIS data so that it fitted the context of local farming systems and was related to SI. Players had to make decisions related to what crops to grow (local or improved varieties), how to fertilize the plots that the farmers were cultivating and subsequently what crops to grow successively over a period of three seasons. Some underlying rules, that were not explained to the participants, forced players to make decisions so that they would achieve the best outcome for themselves. During the third workshop the game was played two times. The first playthrough no communication was allowed to stimulate to development of an individual strategy. The second playthrough communication was allowed so that a collective strategy could be formed. Furthermore, communication allowed to explore the group dynamics and leadership within the participant group. After the game was played an in-depth debriefing was conducted to allow participants to discuss the game and underlying principles. A detailed explanation with the setup of the game and underlying rules can be found in Appendix 7. Three workshops were organized with local partners and farmers in January 2023. The first workshop was held with local partners in Debre Birhan, while the other two workshops were conducted with farmers in Gudo Beret and Goshe Bado. Each workshop had 10 participants. Participants were selected in consultation with the local extension officers. The participant of the first workshop was based on the availability of local partners consisting of local agricultural researchers, agricultural agents and officers. The participant selection for the second and third workshop was based on the inclusion of farms with differing resource endowment and size, and also the inclusion of female household heads. The first workshop with the local partners consisted of a 6 group discussion related to SI. Afterwards the serious game was played. The second workshop was with farmers in Gudo Beret and consisted of playing the serious game. Some adaptations were made after the first workshop to optimize the gameplay. The third workshop was with farmers in Goshe Bado and consisted of playing the serious game followed by an in-depth debriefing. The development of the game thus was an iterative process that allowed to optimize the game over a period of three workshops. 3. Results 3.1 Positive Deviance Analysis 3.1.1 Farm population description Table 1: descriptive statistics of indicator scores for In the Basona Worena district, the RHoMIS survey farm population based on RHoMIS 2022 data was conducted with 238 farmers spread across two Descriptive Statistics villages Gudo Beret and Goshe Bado. Because the Indicator Unit Mean SI-MFS project focusses on mixed farming systems Land owned hectares 1.40 the households that did not keep livestock were excluded from the analysis, resulting in a sample Land ownership ratio % 84% size of 238 households. Table 1 shows an overview Household size MAE 3.39 of the mean indicator scores for the sample. The Prop Land cultivated % 1.3 farms had an average farm size of 1.40 hectares Fertilizer Rate kg/ha 218 with an average land ownership ratio of 84%. Crop Count number 3.2 Household size on average was 3.38 Male Adult Livestock holdings TLU 6.08 Equivalent (MAE). Livestock income ETB 21156 The sample showed large variability in indicator scores as some farms scores Crop income ETB 19520 exceptionally high on certain indicators while other Tree income ETB 3936 farms exceptionally low or do not have any score Technologies adopted Count 4.1 (Fig. 4). For example, some farmers in the sample Months food shortage Months 11.76 did not have any on farm income as these farmers Months irrigated Months 0.67 do not sell any of their farm products. However, these farms were still included in the analysis. 3.1.2 Correlation between indicators Table 2 presents the correlations between all SIAF indicators that were quantified from the RHoMIS dataset. Some indicators correlated negatively while other correlated positively. This can indicate potential synergies and trade-off between indicators. Some indicators such as GHG emissions and livestock diversity (0.359) have a direct relation due to the method of calculation. As a result, these indicators have a higher positive correlation than other indicators. Additionally, the total income sources were strongly positively related to the total value of farm activity (0.485). This can indicate that the higher number of income sources can be beneficial for the total value of farm income. Next to that off farm income was negatively correlated to months food secure (-0.305), possibly explaining that higher off farm income results in higher food security due to increased income security. Finally, livestock diversity was negatively correlated to female control (-0.310). This might suggest that female controlled household keep lower number of livestock species. 3.1.3 Identification of positive deviant farms and characterization of indicator performance Using seven indicators extracted from the RHoMIS 2022 survey, a Pareto ranking was executed that aimed to identify PD farms within the Basona population. A group of 71 farms was classified as Pareto rank 1. The remainder of the of the farms was characterized as dominated systems (sub- optimal cases). Figure 4 visualizes the indicator scores for each of the groups. No distinction was made between the two research kebeles as the indicator scores did not differ significantly between the two research sites. Pareto rank 1 farms still included some extreme cases where certain farms score high on some indicators and low on others, indicating potential trade-off between indicators. After applying the mean threshold of farms scoring above average for five out of seven indicators, a subset of 24 farms was identified as exemplary. These systems provided the best compromise based on the set 7 of indicators used for the PD analysis. Additionally, Fig. 4 shows that for most indicators the exemplary farms score higher. Interestingly, gender control is the only indicator for which the exemplary farms on average score lower than the sub-optimal dominated systems. Furthermore, for the crop productivity and gross margin indicators there are some farms that score exceptionally high, but are still defined as sub-optimal dominated systems as they score low on other indicators. This indicates that compromises are made by some farming systems. Fig. 4: SI indicator performance based on PD analysis. Exemplary (n=23) systems were separated from sub-optimal dominated systems (n=215). Boxplots with the x representing the mean of the observations and the error bars representing the standard error. 8 Production Economic Environment Human Social Table 2: SIAF indicator correlation overview. The table shows the identified indicators from the RHoMIS 2022 dataset sorted in the five SIAF domains. Crop productivity Crop diversity 0,172 Production Milk Yield 0,088 0,042 Livestock diversity 0,189 0,234 0,207 Total value of farm 0,151 0,177 0,294 0,081 activity Income sources 0,153 0,335 0,250 0,172 0,485 Economic Market Orientation -0,030 -0,064 -0,027 -0,158 0,030 0,068 Off farm income 0,028 -0,045 0,069 -0,150 0,131 0,469 0,159 PPI (asset based -0,041 -0,092 0,039 -0,156 0,255 0,122 -0,016 0,041 poverty estimate) GHG emissions 0,375 -0,069 0,117 0,359 -0,044 -0,110 -0,108 -0,061 -0,020 Irrigation 0,033 0,071 0,012 0,064 -0,020 0,127 0,199 0,073 -0,088 0,023 Environment Land conservation 0,092 -0,036 0,117 0,208 -0,128 -0,025 -0,078 -0,054 -0,129 0,020 0,185 Soil Quality -0,003 0,008 -0,052 0,014 0,073 -0,117 0,038 -0,212 0,062 0,047 -0,021 -0,076 Months food secure 0,191 0,178 0,110 0,275 0,164 0,103 -0,268 -0,305 -0,004 0,053 -0,028 0,102 0,223 Household dietary -0,001 0,108 0,091 0,084 0,088 0,064 -0,062 0,007 0,016 0,025 -0,082 -0,006 -0,002 0,256 Human diversity Education -0,098 -0,055 -0,007 0,013 0,006 0,037 0,002 -0,015 0,264 0,023 0,086 0,061 -0,103 -0,062 -0,016 Novel practices trialled -0,015 0,175 0,072 0,263 0,072 0,147 -0,046 0,066 0,071 0,037 0,280 0,200 -0,009 0,088 0,016 0,177 Female control -0,160 -0,225 -0,081 -0,310 -0,013 -0,038 0,038 0,065 -0,085 -0,190 -0,030 -0,031 -0,064 -0,143 -0,060 -0,173 -0,177 Social Group membership -0,024 -0,150 -0,034 -0,018 -0,103 -0,148 0,034 -0,214 0,072 -0,024 -0,038 0,260 0,183 0,145 0,122 0,005 0,117 -0,019 9 Crop productivity Crop diversity Milk Yield Livestock diversity Total value of farm activity Income sources Market Orientation Off farm income PPI GHG emissions Irrigation Land conservation Soil Quality Months food secure Household dietary diversity Education Novel practices trialled Female control Group membership 3.1.4 Innovation descriptive statistics Fig. 5 shows an overview of the number of interventions that individual farmers have trialled during the last three years. Most farmers have only trialled few innovations while a small portion of the farmers have trialled larger number of innovations. This indicates that some farmers have been able to acquire more access to interventions than other farmers. Fig. 6 provides an overview of the total number of times interventions have been trialled by Fig. 5: total times a certain number of interventions was trialled farmers based on the RHoMIS 2022 by individual farmers during the AfricaRISING project. data. A total of 30 of innovations has been trialled in practice over the last years of the Africa RISING R4D project. Figure 5 shows that the trials of innovations were mostly related to either improved varieties or new types of crops and forages such as Phalaris grass and tree lucerne. In contrast, technological innovations such as raised bed/ ridge and furrow, and rope and washer pumps have been trialled only a limited number of times. The tractor thresher has been trialled by some farmers, possibly because it could be tested by multiple farmers at the same time. However, most households did not have direct access to a tractor or threshing machine. Fig. 7 provides an overview of the proportion of the sub-groups of Fig.6: overview of the number of times an innovations has been trialled based on the RHoMIS 2022 survey 11 exemplary and sub-optimal dominated systems that have trialled. The interventions implementation rate does not differ significantly after conducting the two-sample z-test for proportions at α=0.05. However, fava bean (improved variety) and malt barley (improved variety) were more often used by PD farmers. Moreover, chickpea (improved variety) and lentil (improved variety) were also more often adopted by PD farmers. However, the latter has a much lower implementation rate than the former. Fig. 7: proportions of sub-group (exemplary and sub-optimal dominated systems) that have trialled certain intervention. The z-test for proportion was executed and it was concluded that non-of the proportions of adoption rate of SI interventions different significantly at α=0.05. 12 3.2 FarmDESIGN 3.2.1 Case study farm interventions Based on the PD analysis, a selection of case study farms was made that provided a representative sample for the Basona research sites. This selection process aimed to capture the range of practices and outcomes present within the local farming communities. From each of the villages within the Basona research sites, a well-balanced combination of six exemplary farms and two ND farms were chosen for closer examination. Through in-depth interviews utilizing a well-structured survey (appendix 5), these selected farms were probed on various aspects, specifically focusing on the FarmDESIGN parameters. One aspect of the study involved the analysis of the interventions that each of the case farms had implemented to sustainably intensify their practices. This analysis, presented in Table 3, provides a clear picture of the interventions that these farms had implemented. Notably, the exemplary farms show a higher level of innovation and experimentation with SI interventions. The farms delved into a broader range of interventions, demonstrating their willingness to explore diverse approaches to enhance their agricultural performance. Finally, in contrast, the ND farms from the Goshe Bado village present a more limited picture. These farms had only focussed on one intervention: growing an improved variety of bread wheat. This finding suggest that they may be missing out on the potential benefits of a more diverse and comprehensive approach to SI. Table 3: Overview of the interventions trialled by the selected case farms. A distinction is made between the two research kebeles Gudo Beret and Goshe Bado and also the exemplary and ND farms are separated. Fig. 8: Overview of the interventions trialled by the selected case farms. Farm Kebele Type # 1 5 1 1 0 0 1 1 0 0 1 0 0 0 0 0 0 2 6 1 1 0 1 0 0 0 0 1 1 1 0 0 0 0 3 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 Exemplary Gudo 4 3 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 Beret 5 6 1 1 0 1 0 0 1 0 0 1 0 1 0 0 0 6 7 1 1 0 1 0 0 0 1 1 0 0 1 1 0 0 Negative 1 3 1 0 1 0 0 0 0 0 0 0 0 0 1 0 0 Deviant 2 3 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 2 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 2 4 0 0 0 1 0 0 0 0 0 1 1 0 0 1 0 3 4 1 1 0 0 0 0 0 0 0 1 1 0 0 0 0 Exemplary Goshe 4 3 1 1 0 0 0 0 0 0 1 0 0 0 0 0 0 Bado 5 3 1 0 0 0 0 0 0 0 0 1 1 0 0 0 0 6 3 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 Negative 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Deviant 2 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 13 Interventions trialled Bread wheat (improved varieties) Faba bean (improved varieties) Food barley (improved varieties) Malt barley (improved varieties) Potato (improved varieties) Food barley (seed production) Potato (seed production) Alfalfa Oat and vetch mix Phalaris grass Tree lucerne Improved feeding troughs Apple trees Fertiliser trials (inorganic) Tractor thresher 3.2.2 FarmDESIGN indicators The inclusion of three additional indicators from the parameterized farms in the FarmDESIGN analysis enriched the understanding of the complex dynamics. Each indicator offers insights into different aspects of farm performance, providing a more comprehensive perspective for decision- making and intervention design. Farms were analyzed for dietary energy produced, which serves as an indicator to evaluate the productivity and potential food output of the farm, operating profit in (ETB ha-1 yr-1), which allows to assess the economic performance of the farms, and SOM balance (kg OM ha-1 yr-1), which helps to measure the farms ability maintain soil fertility. Looking at Fig. 8 it shows that in (A) the exemplary farm do mostly score higher on the dietary energy production but do not necessarily have a higher SOM. Furthermore, the operating profit of most exemplary farmer is higher than the ND farms. Also, (C) shows a Pareto frontier on which the exemplary farms score higher on at least one of the indicators. However, still compromises between indicators are made. Surprisingly, a high organic matter balance was found in one of the ND farms, even though the operating profit of this farm was negative. The existence of compromises between indicators highlights the inherent trade-offs that farmers face when striving to optimize multiple aspects of their farm systems, which is both occurring in exemplary and ND farms. Fig. 8: Indicator scores for the sixteen case study farms. The blue dots represent the twelve exemplary (PD) farms while the orange dots represent the four negative deviant (ND) farms. (a) denotes the dietary energy against SOM balance, (b) operating profit against SOM balance, and (c) operating profit against dietary energy. The red dotted line in figure c indicates a pareto frontier. 14 3.3 Workshop and gaming Three workshops were organized of which the first participant group was the local scaling partners, while the other two groups were farmers from Gudo Beret and Goshe Bado respectively. Communication was not allowed during the first game, while during the second game communication was allowed. As a result, the first game helped farmers to develop an individual strategy. The second game, because communication was allowed, helped farmers develop a collective strategy. This meant that some participants took the lead in communicating what types of crops and how to fertilize the crops, would result in the highest yields. Other farmers would either follow the proposed strategy or use the information provided by the other farmers to develop their own strategies. It can be seen in figure 11 that in the first diagram that more different crops were chosen, while in the second diagram mainly the improved cereal and legume varieties were chosen. During the first game farmers decision making regarding what crops to grow was somewhat random (Fig. 11). This happened because farmers did not know what the underlying rules were. During the debrief farmers mentioned that they would first mimic real-life choices. As the seasons (rounds) passed and yield were influenced by the underlying rules, farmers changed their behaviour to maximize their profit. Workshops showed that farmers preferred to cultivate improved varieties instead of local varieties as these clearly resulted in increased yields (Fig. 11). Additionally, farmers would mention that in real life organic manure would mainly be used for cereal crop such as wheat and barley while they would use inorganic fertilizers for legumes such as field pea and Faba bean. The debriefing sessions provided interesting insights in how participant perceived the game scenarios in comparison to real life situations. Some participants mentioned that the game represented real life situation. Also the importance of improved varieties was Fig. 9: Two flow diagrams that show the overall crop choices made during the two games of the third workshop. On top the first game and below the mentioned: ‘the improved varieties are second game. The left side shows the first crops chosen, the middle bar shows crucial to productivity’. This is in line with the second round, and the right bar the third round. In between are the flows that provide an overview of the rotational decisions. 15 the fact that improved varieties are the most widely adopted SI interventions. Furthermore, it was clear that the farmers were aware of the benefits of adequately fertilizing the land and soil to productivity: ‘what the land gives you depends on what you give to the land’. The use of organic fertilizers such as manure and compost was also mentioned to be common practice in the area. However, the supply of organic resources is limited: ‘if we want to use more organic compost, there will not be sufficient raw materials’; ‘animals are source of high concentration of organic fertilizer’. Finally, a lack of labour is a major constraint for the farmers to improve their practices and productivity. While this could be improved through technological innovations such as tractor threshers, the availability of these technologies is still limited: ‘the farmers have the ability to use technology, but they do not have the suppliers and do not have access to technology’ Fig. 10: pictures from the three workshops. The top image shows a image of the debriefing session that provided a knowledge sharing platform. The image on the left show the playing field during the first workshop. Here it can be seen that some plots have more ‘yield’ tokens than others depending on the choices (variety, rotation and manure) made during the game. Finally, the bottom right image shows how the game was explained by the extension officer to the participating farmers in the Gudo Beret kebele. 16 4. Discussion The primary objective of this study was to assess the viability of SI interventions in both new and existing agricultural areas. By closely examining exemplary systems that outperform other farms in the same environment with similar resources, it was aimed to gain a deeper understanding of the exceptional performance attributed to SI intervention implementation and the underlying dynamics (Adelhart-Toorop et al., 2021; Liang et al., 2022). Through a positive deviance analysis based on a set of SIAF indicators, the research identified a group of exemplary farms in two kebeles; Gudo Beret and Goshe Bado, within the Basona Worena district (Musumba et al., 2017). While these farms outperformed the sub-optimal dominated part of the farmer population, it was evident that even the exemplary systems showed compromises in their indicator scores shown by significant variability. Additionally, farms that performed poorly; the ND farms, were selected for on-farm interviews and demonstrated low scores on all indicators compared to the exemplary systems, leading to lower SI scores. Finally, the tree workshops that were conducted during the research effectively served as a platform for knowledge sharing, shedding light on both the opportunities and limitations of SI intervention implementation (Belotti et al., 2013). These workshops facilitated fruitful discussions among participants from diverse backgrounds, helping them find common ground while playing the game and sharing insights into their decision-making processes. First the PD analysis utilized an existing RHoMIS dataset from which the set of SIAF indicators could be extracted. While the RHoMIS tool has been designed to be able to robustly collect data from rural households, there is always risk of collection unreliable data that is non-credible due to inadequate estimations (Fraval et al., 2018). The sample of farms from both kebeles showed large variability in indicator scores with some big outliers. It is suspected that this variability is caused by both the diversity in farm types in the area and the credibility of some of the data in the dataset (Fraval et al., 2018; Tittonel et al., 2010) Furthermore, in this study, certain indicators were normalized so that the values of the whole farm cumulated. For instance, crop productivity was expressed in terms of energy rather than mass (unit kCal ha-1 yr-1), which is not the conventional unit for expressing crop yields (Cassidy et al., 2013). The reason behind this unconventional approach is the presence of multiple staple crops such as malt barley, food barley, wheat, teff, and Faba bean. This diversity makes it challenging to directly compare farms that cultivate different staple crops. Therefore, we avoided the use of conventional units like dry matter or grain production (in kg ha-1 ¬yr-1). In addition, the environmental indicators employed in the PD analysis relied solely on the farmers' perception and did not incorporate precisely measured quantitative variables. Although farmers in general have the ability to adequately assess their soil quality, this approach may still introduce subjectivity and potential biases in the analysis. Furthermore, the study revealed disparities in the farms' financial status, as some farms reported no gross income from on-farm product sales, indicating a reliance on subsistence farming. On the other hand, certain farms had additional income sources from off-farm activities. Consequently, this income variability contributed to the lower average gross income observed in the sample. An approach using Pareto ranking and a mean indicator value threshold was used to select the exemplary farm systems within the two research kebeles. Pareto ranking as able to identify 71 PD farms . Then using the mean indicator threshold methodology 23 exemplary systems were identified distributed over the two research kebeles Gudo Beret and Goshe Bado. While the analysis focused on identifying exemplary systems from the two villages, it did not consider the different types of farms within the landscape. As a result, high resource endowed farms are compared on equal terms with low resource endowed farms (Tittonel et al., 2010). Thus, the mean indicator threshold that was chosen as methodology might not have been the most suitable method to separate the exemplary systems from the rest of the PD farms as mean indicators scores included values of all farm types. This was also seen in the indicator values showing large variability with large outliers. Furthermore, farms did not have to score above average for all indicators, but only five out of seven. By doing this, the inclusion of exemplary farm that showed compromises in their indicator scores had become inevitable. Comparing the research that has previously used a PD approach, other methodologies such as the mean indicator to the ideal point (MIDIP) in which farms are scaled on their relative distance to the ideal point that represents the least compromise compared over all the used indicators, would have been able to better rule out the compromising systems (Adelhart-Toorop et al., 2021; Liang et al., 2022). Nevertheless, a distinction between the 17 exemplary systems and the sub-optimal dominated systems was made that could then be used in the further steps of this research. Between these challenges, the PD analysis proved to be an agile and effective methodology for identifying exemplary systems within the farm population. The selected indicators, to some extent, adequately captured the diverse nature of the farm systems in the region. However, it is crucial to emphasize the holistic nature of the SIAF (Musumba et al., 2017). By incorporating a comprehensive set of indicators, the ranking and understanding of farming systems can be further improved, enabling more targeted and context-specific interventions. After conducting the parameterization of the farms in FarmDESIGN, several key observations emerged. The analysis revealed an unbalanced feed balance across all farms, which could be attributed to several factors. It became apparent that some farmers may not have provided accurate information regarding the sources of feed, leading to discrepancies in the model. Additionally, there were instances where the amount and timing of off-farm grazing were underestimated, further contributing to the feed imbalance (Ayele et al., 2021). Interestingly, during the data collection process, some farmers mentioned collecting off-farm fodder, which they added to the total animal feed. However, regrettably, this crucial information was not accounted for in the parameterizing of the model, potentially leading to incomplete results. Thus, it was evident that the majority of farms did not meet the necessary feed requirements to achieve optimal animal productivity (Assefa et al., 2013). This finding highlights a significant challenge in the pursuit of sustainable intensification, as livestock productivity is directly linked to the availability and quality of feed resources (Duguma and Janssens, 2021). While the feed balance in FarmDESIGN appeared unbalanced for most of the case study farms, it is essential to note that this does not necessarily imply a scarcity of livestock feed availability. Rather, it indicates that the origins of the feed are uncertain, and farmers may underestimate their residue production or misjudge the time allocated to off-farm grazing. These uncertainties contribute to the observed inconsistencies in the feed balance, further impacting livestock productivity. Another aspect revealed during the study was the varying access to irrigation among the farming communities. Some farmers in the Gudo Beret kebele reported having access to irrigation, allowing them to cultivate crops during two seasons per year. However, the availability of water sources was uneven, resulting in unequal access to irrigation facilities. In contrast, in Goshe Bado, there was no access to irrigation at all, limiting crop cultivation to a single season. As a result, irrigation was left out of the modelling process. Additionally this research was able to effectively come up with a simple serious game that was based around the existing serious game ReHab (Page et al., 2016). Through an iterative process of three workshops it was possible to develop the game in such a way that it was both engaging and educational for the participants (Daré et al., 2020). During all workshops it was mentioned that the game was able to represent real life situations and that the decisions made in the game are similar to real life decisions making (Andreotti et al., 2020). All workshops identified constraints related to availability of chemical fertilizers. Next to that the participants preferred using chemical fertilizers for the cereal crops (wheat and barley), and using organic manure for the fertilization of leguminous crops such as Faba bean. During the workshops and interviews farmers mentioned their dependence on inorganic fertilizers. Recent price increase due to uncontrollable global situations further constrain their the profitability of their enterprises and limits them to invest in other resources and interventions (Snapp et al., 2023). While farmers prefer to use organic alternatives the amount of organic material in form of manure and residues is disproportionate for the amount compost needed to fertilize the land. Thus a status quo exists of farmers being dependent on inorganic fertilizers and limited possibility to increase reliance on organic resources. Due to limited time and number of workshops the serious games was only fully played one single time during the last workshop. The first workshop, including the scaling partners, functioned mainly as a trial to playtest the game to identify potential faults in the game design. During the Goshe Bado workshop, it was observed that soil characteristics significantly influenced farmers' decisions regarding the use of chemical or organic fertilizers. This complexity highlighted the farmers' ability to understand the dynamics of soil fertility and the necessary nutrient additions based on soil characteristics (Tesfahunegn, 2019; Mesfin et al., 2023). Interestingly, during the game, farmers showed a willingness to use chemical fertilizers and improved varieties. 18 However, during the debrief, it became apparent that in reality, the availability of both chemical fertilizers and improved varieties posed major constraints for the farmers. Particularly in Goshe Bado, these resources were already extensively shared among farmers and households. Moreover, the farmers in Goshe Bado strategically planned their crop planting to optimize labour efficiency. This indicates their resourcefulness and ability to adapt their farming practices to make the best use of available resources. These findings underscore the complex and nuanced nature of decision- making in agricultural contexts. Farmers often face multiple challenges and constraints, and their actual choices may differ from what is observed in controlled scenarios like the game. Understanding these realities is crucial for designing effective and contextually appropriate interventions to support sustainable intensification and improve agricultural practices. It was expected that the PD analysis would be able to identify the exemplary farm systems that could be used as an example for other farms that have more difficulty to sustainably intensify their practices. The adoption of SI interventions has resulted in significant benefits for household livelihoods in the Basona Worena district. However, scaling these opportunities to new areas is key to stimulate further development. Serious gaming can be utilized as an quick and agile methodology to engage farmers in real life scenarios that allow them to visualize alternative practices and to create a deeper understanding for the relevant SI interventions. Next to that, these methodologies can help overcome the difficulty of integrating knowledge at the CG centres. However, it can still be questioned whether the extensive knowledge pool built by CG centres can be utilized in such a manner that it is both integrative, practical, and adaptable so that it aids scaling opportunities and possibilities. Such methodology can be utilized so that existing data and research that contain sustainable intensification indicators obtained by the CG centres, can be utilized in an integrated manner, aiding the scaling opportunities in the new research sites of the SI-MFS initiative. 5. Conclusion In conclusion, this study aimed to assess the viability of SI interventions in Basona Worena by identifying exemplary farm systems, analysing indicator performance, and utilizing serious gaming for knowledge dissemination and decision-making. The research employed PD analysis to identify exemplary farms, based on a set of SIAF indicators. The results highlighted the diversity and variability in indicator scores, leading to the selection of farms that exhibited exceptional performance across the indicators. The FarmDESIGN analysis provided insights into the interventions implemented by case study farms to sustainably intensify their practices. This analysis revealed challenges such as unbalanced feed resources and varying access to irrigation, influencing the productivity of livestock and crops. It also emphasized the importance of accurate data collection and interpretation to capture the complexities of farm systems accurately. Serious gaming workshops served as an effective tool to engage farmers, facilitating discussions and providing an interactive platform to explore decision-making processes. The game allowed participants to experiment with different interventions and scenarios, leading to valuable insights into real-world challenges and opportunities related to SI interventions. However, the findings also highlighted the discrepancies between game choices and actual constraints faced by farmers in terms of resource availability and access. The study's methodology demonstrated the potential of combining quantitative analyses like PD and FarmDESIGN with participatory approaches like serious gaming to gain a comprehensive understanding of sustainable intensification. It also highlighted the need to carefully consider data quality, context-specific factors, and the complexities of decision-making when designing interventions and disseminating knowledge to support agricultural development. In conclusion, this research provides valuable insights into the complexities of implementing SI interventions, the challenges farmers face, and the potential of innovative methodologies to engage stakeholders and promote sustainable agricultural practices. Future efforts should focus on refining these methodologies, integrating local knowledge, and tailoring interventions to specific contexts to achieve meaningful and scalable impacts on agricultural systems. 19 References Adelhart Toorop, R., Ceccarelli, V., Bijarniya, D., Jat, M. L., Jat, R. K., Lopez-Ridaura, S., & Groot, J. C. J. (2020). Using a positive deviance approach to inform farming systems redesign: A case study from Bihar, India. Agricultural Systems, 185(August), 102942. https://doi.org/10.1016/j.agsy.2020.102942 Andreotti, F., Speelman, E. N., Van den Meersche, K., & Allinne, C. (2020). Combining participatory games and backcasting to support collective scenario evaluation: an action research approach for sustainable agroforestry landscape management. Sustainability Science, 15(5), 1383-1399. https://doi.org/10.1007/s11625-020-00829-3 Assefa, D., Nurfeta, A., & Banerjee, S. (2013). Assessment of feed resource availability and livestock production constraints in selected Kebeles of Adami Tullu Jiddo Kombolcha District, Ethiopia. African Journal of Agricultural Research, 8(29), 4067-4073. https://doi.org/10.5897/AJAR2013.7096 Ayele, J., Tolemariam, T., Beyene, A., Tadese, D. A., & Tamiru, M. (2021). Assessment of livestock feed supply and demand concerning livestock productivity in Lalo Kile district of Kellem Wollega Zone, Western Ethiopia. Heliyon, 7(10). https://doi.org/10.1016/j.heliyon.2021.e08177 Barrett, P., 2005. Euclidean Distance raw, normalized, and double-scaled coefficients [WWW Document]. URL. http://www.pbarrett.net/techpapers/euclid.pdf (accessed 1.11.22). Bellotti, F., Kapralos, B., Lee, K., Moreno-Ger, P., & Berta, R. (2013). Assessment in and of serious games: An overview. Advances in human-computer interaction, 2013, 1-1. https://doi.org/10.1155/2013/136864 Breman, H., Schut, A. G., & Seligman, N. A. G. (2019). From fed by the world to food security: Accelerating agricultural development in Africa. Plant Production Systems Wageningen University. https://doi.org/10.18174/498300 Cassidy, E. S., West, P. C., Gerber, J. S., & Foley, J. A. (2013). Redefining agricultural yields: from tonnes to people nourished per hectare. Environmental Research Letters, 8(3), 034015. http://dx.doi.org/10.1088/1748-9326/8/3/034015 CGIAR (2021). Sustainable Intensification of Mixed Farming Systems ( SI-MFS ). 1–78. November 23, 2021. https://www.cgiar.org/news-events/news/new-cgiar-initiative-on-sustainable intensification-of-mixed-farming-systems-launched/ Daré William's, Hassenforder Emeline, Dray Anne. (2020). Observation manual for collective serious games. Montpellier : CIRAD-ComMod Association, 68 p. ISBN 978-2-87614-765-2 https://doi.org/10.19182/agritrop/00144 Duguma, B., & Janssens, G. P. (2021). Assessment of livestock feed resources and coping strategies with dry season feed scarcity in mixed crop–livestock farming systems around the gilgel gibe catchment, Southwest Ethiopia. Sustainability, 13(19), 10713. https://doi.org/10.3390/su131910713 Fraval, S., Hammond, J., Wichern, J., Oosting, S. J., De Boer, I. J., Teufel, N., ... & Van Wijk, M. T. (2019). Making the most of imperfect data: a critical evaluation of standard information collected in farm household surveys. Experimental Agriculture, 55(2), 230-250. https://doi.org/10.1017/S0014479718000388 Gebreyes, M., Mekonnen, K., Thorne, P., Derseh, M., Adie, A., Mulema, A., Kemal, S. A., Tamene, L., Amede, T., Haileslassie, A., Gebrekirstos, A., Mupangwa, W. T., Ebrahim, M., Alene, T., Asfaw, A., Dubale, W., & Yasabu, S. (2021). Overcoming constraints of scaling: Critical and empirical perspectives on agricultural innovation scaling. PLoS ONE, 16(5 May), 1 19. https://doi.org/10.1371/journal.pone.0251958 Groot, J.C.J., Rossing, W.A., 2011. Model-aided learning for adaptive management of natural resources: an evolutionary design perspective. Methods Ecol. Evol. 2, 643–650. https://doi.org/10.1111/j.2041-210X.2011.00114.x. Groot, J.C.J., Oomen, G.J., Rossing, W.A., 2012. Multi-objective optimization and design of farming systems. Agric. Syst. 110, 63–77. https://doi.org/10.1016/j.agsy.2012.03. 012. 20 Hammond, J., van Wijk, M., Teufel, N., Mekonnen, K., & Thorne, P. (2021). Assessing smallholder sustainable intensification in the Ethiopian highlands. Agricultural Systems, 194(April), 103266. https://doi.org/10.1016/j.agsy.2021.103266 Ittersum, M. K. Van, Bussel, L. G. J. Van, Wolf, J., Grassini, P., Wart, J. Van, & Guilpart, N. (2016). Can sub-Saharan Africa feed itself ? 113(52), 14964–14969. https://doi.org/10.1073/pnas.1610359113 Kirimbo, G.E. (2017). Engaging smallholder-farmers in model-based identification of alternative options, trade-offs and synergies for sustainable intensification in North-Tanzania. MSc thesis. Wageningen, The Netherlands: Wageningen University and Research Centre. Liang, Z., van der Werf, W., Xu, Z., Cheng, J., Wang, C., Cong, W. F., ... & Groot, J. C. (2022). Identifying exemplary sustainable cropping systems using a positive deviance approach: Wheat-maize double cropping in the North China Plain. Agricultural Systems, 201, 103471. https://doi.org/10.1016/j.agsy.2022.103471 Tobias Lunt, Jim Ellis-Jones, Kindu Mekonnen, Steffen Schulz, Peter Thorne, Elmar Schulte Geldermann & Kalpana Sharma (2018) Participatory community analysis: identifying and addressing challenges to Ethiopian smallholder livelihoods, Development in Practice, 28:2, 208-226, https://doi.org/10.1080/09614524.2018.1417354 Massop, E. 2021. Farm performance evaluation: Holistic impact assessment of project promoted sustainable intensification innovations at farm-level in Tanzania. MSc thesis in Farming Systems Ecology. Wageningen, The Netherlands: Wageningen University and Research Centre. Meijer, S. S., Catacutan, D., Ajayi, O. C., Sileshi, G. W., & Nieuwenhuis, M. (2015). The role of knowledge, attitudes and perceptions in the uptake of agricultural and agroforestry innovations among smallholder farmers in sub-Saharan Africa. International Journal of Agricultural Sustainability, 13(1), 40–54. https://doi.org/10.1080/14735903.2014.912493 Mesfin, D., Assefa, E., & Simane, B. (2023). Farmers and scientific-based assessment of soil quality in the Choke agroecosystem, Upper Blue Nile Basin, Ethiopia. Environmental and Sustainability Indicators, 100279. https://doi.org/10.1016/j.indic.2023.100279 Michalscheck, M., Groot, J. C. J., Fischer, G., & Tittonell, P. (2020). Land Use Policy Land use decisions : By whom and to whose benefit ? A serious game to uncover dynamics in farm land allocation at household level in Northern Ghana. Land Use Policy, 91(December 2019), 104325. https://doi.org/10.1016/j.landusepol.2019.104325 Modernel, P., Dogliotti, S., Alvarez, S., Corbeels, M., Picasso, V., & Tittonell, P. (2018). Identification of beef production farms in the Pampas and Campos area that stand out in economic and environmental performance. Ecological Indicators, 89(May 2017), 755–770. https://doi.org/10.1016/j.ecolind.2018.01.038 Musumba, M., and Grabowski, P., Palm, C. & Snapp, S., Guide for the Sustainable Intensification Assessment Framework (August 24, 2017). Available at SSRN: https://ssrn.com/abstract=3906994 or http://dx.doi.org/10.2139/ssrn.3906994 Mwaniki, A. (2006). Achieving food security in Africa: Challenges and issues. UN Office of the Special Advisor on Africa (OSAA). Negra, C., Powell, M., & McCarthy, N. (2020). Performance evaluation of the Africa Research in Sustainable Intensification for the Next Generation (Africa RISING) program. April. https://cgspace.cgiar.org/handle/10568/108031 Öborn, I., Vanlauwe, B., Atta-Krah, K., Thomas, R. J., Phillips, M., & Schut, M. (2017). Integrated systems research for sustainable intensification of smallholder agriculture. In I. Oborn, B. Vanlauwe, M. Phillips, R. Thomas, K. Atta-Krah, & W. Brooijmans (Eds.), Sustainable Intensification in Smallholder Agriculture: An Integrated Systems Research Approach (pp. 1-15). Taylor and Francis Inc.. Pretty, J., Toulmin, C., & Williams, S. (2011). Sustainable intensification in African agriculture. International journal of agricultural sustainability, 9(1), 5-24. https://doi.org/10.3763/ijas.2010.0583 21 Roser, M., Ritchie, H., Ortiz-Ospina, E., & Rodés-Guirao, L. (2013) - "World Population Growth". Published online at OurWorldInData.org. Retrieved from: ‘https://ourworldindata.org/world-population-growth’ [Online Resource] Seifu, H., Gebreyes, M., Mekonnen, K., & Whitbread, A. (2022). Sustainable Intensification of Mixed Farming Systems (SI-MFS) Initiative: Report on the Ethiopia Team System Analysis meeting. Schut, A. G. T., & Giller, K. E. (2020). Sustainable intensi fi cation of agriculture in Africa. 7(4), 371 375. https://doi.org/10.15302/J-FASE-2020357 Snapp, S., Sapkota, T., Chamberlin, J., Cox, C., Gameda, S., Jat, M., ... & Govaerts, B. (2023). Rebalancing global nitrogen management in response to a fertilizer and food security crisis. https://doi.org/10.21203/rs.3.rs-2318855/v1 Speelman, E. N., García-Barrios, L. E., Groot, J. C. J., & Tittonell, P. (2014). Gaming for smallholder participation in the design of more sustainable agricultural landscapes. Agricultural Systems 126, 62-75. https://doi.org/10.1016/j.agsy.2013.09.002 Steinke, J., Mgimiloko, M. G., Graef, F., Hammond, J., van Wijk, M. T., & van Etten, J. (2019). Prioritizing options for multi-objective agricultural development through the Positive Deviance approach. PLoS ONE, 14(2), 1–20. https://doi.org/10.1371/journal.pone.0212926 Tesfahunegn, G. B. (2019). Farmers’ perception on land degradation in northern Ethiopia: Implication for developing sustainable land management. The Social Science Journal, 56(2), 268-287. https://doi.org/10.1016/j.soscij.2018.07.004 Tesfaye A, Hammond J, Radeny M, Recha JW, Nigussie A, Ambaw G, van Wijk MT, Tamene L, Abera W, Solomon D. 2021. The impacts of climate smart agricultural practices on household income and food security: evidence from Doyogena and Basona climate-smart landscapes in Ethiopia. CCAFS Technical Report. Wageningen, the Netherlands: CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS). Tittonell, P., Muriuki, A., Shepherd, K. D., Mugendi, D., Kaizzi, K. C., Okeyo, J., Verchot, L., Coe, R., & Vanlauwe, B. (2010). The diversity of rural livelihoods and their influence on soil fertility in agricultural systems of East Africa – A typology of smallholder farms. Agricultural Systems, 103(2), 83–97. https://doi.org/10.1016/j.agsy.2009.10.001 Tittonell, P., & Giller, K. E. (2013). Field Crops Research When yield gaps are poverty traps : The paradigm of ecological intensification in African smallholder agriculture. Field Crops Research, 143, 76–90. https://doi.org/10.1016/j.fcr.2012.10.007 Vanlauwe, B., Coyne, D., Gockowski, J., Hauser, S., Huising, J., Masso, C., Nziguheba, G., Schut, M., & Van Asten, P. (2014). Sustainable intensification and the African smallholder farmer. Current Opinion in Environmental Sustainability, 8, 15–22. https://doi.org/10.1016/j.cosust.2014.06.001 22 Appendix 1: Stakeholder engagement A project managed through the CGIAR in cooperation with many of the CGIAR institutes using multiple participatory methodologies requires a significant amount of stakeholder engagement. Both formal institutions and local stakeholders are involved in the process. For this thesis main contact organization are ILRI and ICARDA. Furthermore, local farmers will be visited with help and under supervision of local experts. For the farm redesign focus group sessions it is preferred to not only involve farmers, but also other local institutions. By creating a diverse set of stakeholders the opportunity for fruitful discussion sessions is increased. In the end, the aim is to hopefully create understanding for the approaches used in this thesis so that these methods and tools can potentially be adopted on a larger more regional scale. The complete list of stakeholders is shown below. Name Description Wageningen University and Research Supervision and examination of the thesis project CGIAR Coordination of SI-MFS initiative ILRI Supervision during fieldwork trip, facilitation of logistics ICARDA Supervision of SI-MFS initiative in Ethiopia Debre Birhan University Local experts and researchers included in focus groups Local experts in agriculture that can be involved in focus Debre Birhan Agricultural Research Centre groups both to moderate and the learn from the sessions Basona Worena district Office of Agriculture Local authority related to agriculture Farmers Local household members involved in agriculture Appendix 2: List of indicators according to the SIAF. From Hammond et al. (2021) Domain Indicator Unit Staple crop yield t ha-1yr-1 Crop productivity kCal ha-1 yr-1 Crop diversity species grown Milk Yield Litres cow-1 day-1 Production Livestock productivity kCal TLU-1 yr-1 Livestock diversity Number of species kept Total value of activity USD person-1 day-1 Income sources Count Market Orientation % produce sold Economic Off farm income USD household-1 yr-1 PPI (asset based poverty estimate) % chance out of poverty GHG emissions tCO2eq pers-1 yr-1 Irrigation Months yr-1 Environment Land conservation Counts of practices Soil Quality Farmer perception (0-3) HFIAS (hunger) Severe to no experience of hunger (1-4) Months food secure Months yr-1 Human Household dietary diversity Food groups consumed weekly (0-10) Education Ordinal (1-7) Novel practices trialled Count over last 5 years Female assets owned Assets owned or co-owned by females (0-4) Female control % of livelihood controlled by females Dependency ratio Ratio of non-worker: workers Social Group membership Group membership yr-1 Extension services Frequency and quality (1-18) Skill sharing Count of peers receiving skills Count of substantial items given or received Gifts and exchange yr-1 23 Appendix 3: All indicator score overview Sub-optimal Indicator Unit All farms (n=238) Exemplary (n=23) dominated (n=215) Mean SD Mean SD Mean SD MIDIP - 0,4 0,2 0,4 0,2 0,4 0,2 HH Size Number 4,3 1,5 4,3 1,3 4,3 1,5 HH Size MAE 3,4 1,2 3,6 1,2 3,4 1,3 Education Ordinal (1-7) 3,7 2,3 3,6 2,3 3,7 2,3 Land Ownership Ratio - 0,8 0,5 0,8 0,4 0,8 0,5 Land Owned ha 1,4 1,1 1,3 1,1 1,4 1,1 Land Cultivated ha 1,8 1,3 1,7 1,0 1,9 1,4 Prop land cultivated - 1,3 1,1 1,3 0,7 1,3 1,1 Livestock Holdings TLU 6,1 3,3 8,6 2,7 5,8 3,3 Total energy kCal yr-1 7415101 6590366 12610511 8257821 6859313 6171075 Crop Productivity kCal/ha/yr 4683779 3552193 8151122 4813198 4312854 3197059 Crop Diversity Number 3,2 0,8 3,7 0,7 3,2 0,8 Livestock Diversity Number 4,7 1,4 5,7 0,8 4,6 1,4 Food Self Sufficiency # HHMAE/yr 2,7 2,4 4,5 3,2 2,5 2,2 Diversity Number 7,9 1,8 9,1 1,8 7,7 1,7 Gross income ETB/hhmember/day 31,8 42,1 60,9 50,4 28,7 40,2 Income per hectare ETB/ha/yr 27868 30591 53179 32130 25160 29299 Total Income Sources Number 3,1 1,7 4,0 2,0 3,0 1,6 Crop Market Orientation % 22% 21% 27% 25% 21% 20% Off farm Income Sources Number 0,5 0,7 0,5 0,5 0,4 0,7 PPI Score Number 55,3 11,1 54,2 13,2 55,4 11,0 GHG Emissions CO2eq/ha/yr 3229 2278 4845 2675 3056 2173 Irrigation Months/yr 0,7 1,8 1,0 2,0 0,6 1,8 Land Conservation Practices Number 2,1 1,0 2,3 1,0 2,1 1,0 Fertilizer Rate kg/ha 218 252 301 184 209 258 Fert Input kg 348 434 443 309 338 445 Soil Quality Number 2,5 0,8 2,9 0,4 2,5 0,8 Months Food Secure months/yr 11,8 0,8 11,9 0,4 11,7 0,8 HDDS score Number 4,3 1,4 5,0 1,6 4,2 1,4 Education Ordinal (1-7) 3,7 2,3 3,6 2,3 3,7 2,3 Novel Practices Trialled Number 4,1 2,6 5,0 2,5 4,0 2,6 Gender Female Control % 34% 25% 25% 18% 35% 26% Membership Count Number 1,6 1,0 1,8 1,0 1,5 1,0 Total Income ETB/yr 44611 52597 85125 74007 40277 48150 Value Farm Produce ETB/yr 40676 51955 78821 74075 36595 47590 Forage income ETB/yr 3936 10750 6304 10438 3682 10801 Livestock Prod Sales ETB/yr 21156 31119 40560 38890 19080 29609 Crop Sales ETB/yr 19520 34585 38261 50474 17515 32040 24 Appendix 4: Complete FarmDESIGN survey General Information: No Question Question Type Options 1 Does the household consent to provide information? Acknowledge 2 Name of head of household Text 3 Name of respondent (if not head of household) Text 4 Gender of respondent Select one Male Female Other 5 Contact/ Mobile Number Number 6 Country Name Text 7 District/ Sub-location Name Text 8 Village Name Text 9 Location of farm GPS Point 10 Local currency unit (LCU) T e x t (Hint: Indian rupee, Bangladeshi taka etc.) 11. Household Information: Household Position in HH (in relation to Gender Age Notes members head) HH1 Head Male Husband/wife Female Son/daughter Other Son/daughter in law Father/mother Grandchild Sibling Permanently employed worker Other HH2 HH3 HH4 HH5 HH6 25 12. Household labour Labour hours per season per Season Off-farm work [Select one] Notes HH member On-farm Off-farm HH1 Agricultural (off-farm) Non-agricultural labour (off-farm) Government employment Private formal employment Informal employment Self-employed (no employees) Business (with employees) Student Other HH2 HH3 HH4 HH5 HH6 Land Information: 13 Local land unit (LLU) followed by farmer Text Timad 14 LU conversion rate, 1 hectare = Specify LLU value Decimal 1 timad = 0.25 hectares Hint: 1 hectare = 2.47 acre. So if LLU is acre then enter 2.47 15. Indicate duration of seasons Planting month Harvesting month Notes Season 1 Season 2 26 16. Basic farm structure [Question Matrix] Area (in hectare) Season Notes How much land do you cultivate? How much land do you own (including leased out)? How much land do you lease? How much land do you lease-out? Number of plots (same location and same crop pattern over the year) How many crops did you grow in the past year? Hint: (Include all crops including cereals, cash crops, feed crops, vegetables and fruits) 1. Yes 17 Do you have a kitchen garden? Select One 2. No (Hint: garden close to house where vegetables are grown) 18 How big is your kitchen garden? [If Q 18 is answered yes] Decimal (length x width) 19 What crops do you grow in the kitchen garden? [If Q 18 is answered yes] Text 20. Scattered Trees: Tree Tree Type Number of Utilized for Quantity % home % sold at Value of sales Notes [Text] trees [Select produced consumption market [Decimal] [Integer] Many] (include unit) 1. Food 1 2. Feed 3. Timber 4. Fuel 5. Other 2 27 Cropping System Information: 21 Plot Number Integer 22 Area (timed) Decimal 1. Owned 23 Ownership of plot Select 2. Rented One 3. Shared 1. Season 1 24 When are crops grown Select 2. Season 2 in this plot? Many 3. Season 3 4. Perennial 25 Crops grown in season Text [If Q 28 selects Season1] 26 Perennial crops [If Q Text 28 selects Perennial] 27 Does this plot include Select intercropping in any one season? 28 Is this plot irrigated? Select One 29 Explain [If ‘yes’ is Text selected for Q34] 30 Do you receive any Select specific crop subsidies? one 31 For which crop (s)? Text 32 Value of additional crop Text subsidies year [Only displayed if Q 36 is Yes] 28 33. Crop production and use per plot [Question Matrix]: Note: Indicate crop products produced from this particular plot. If two crop products are produced from the same crop [ ex. Banana leaf and banana fruits] then include them as separate lines. If residue is sold then include it here. Otherwise include residue management details in next question. Product Crop Crop Yield Yield Unit Percent home Percent Percent Total value Notes (Include any No. [Text] Product [Decimal] [Text] consumption used as sold at of crop other use for product [Text] (%) feed (%) market sales (LCU) ex. Kept as seed etc.) (%) 1 2 3 4 34. Residue Management [Question Matrix]: Residue Crop Yield (If Type of residue % used % used % used % burnt % used as Notes (including Management [Text] known. [Select one] as as feed as fuel in field bedding for any other uses) Include unit) green livestock [Text] manure 1. Straw 1 2. Stover 3. Leaves 4. Stalks 5. Vines 6. Whole plant 7. Other 2 3 4 5 6 29 35. Cropping activities consuming labour and inputs: Costs for HH Hired Input(s) Labour Type [Select inputs Number Crop Activity [Select one] Labour labour (hired or Notes Many] (hrs) (hrs) purchased) (fertilizers, pesticides) 1. Land preparation 1. Household member 1 2. Intercultural activities 2. Permanent employed 3. Fertilizer/ chemical worker application 3. Unpaid labour 4. Harvesting 4. Hired labour (temp) 5. Post-harvest 5. Contract worker 6. Other 2 3 36. Fertilizer and chemical application Number Crop/ Name of Quantity Price per kg (in ETB) Fertilizer Quantity Price of Notes Season Organic Applied (Use 0 if produced on Name/ Applied fertilizer amendment (kg/ha) farm) composition (kg/ha) per kg 1 2 3 37. Biocide [Question Matrix] Number Crop Pesticide Name/ composition Quantity Applied (kg) Price per kg Notes 1 2 3 30 Other Cultivation costs 38 Other cultivation costs (not covered elsewhere) Decimal 39 Hired labour (for general farm activities) in hrs Decimal Hint: 40 Average cost of casual labour per hour Decimal Hint: Usually unskilled labour 41 Average cost of regular labour per hour Decimal Hint: Usually skilled labour 42 Any further notes on crop production? Text 43 Livestock species [If Cow Adult Cow Adult Cow Young Ox Sheep Goat Donkey Horse Chicken Bees Q 50 is yes] (F) (M) 44 Sex [If Q 51 selects ruminants or non- ruminants] 45 Breed(s) [If Q 51 selects ruminants] 46 Breed Type [If Q 50 selects ruminants] 47 Number being milked right now 48 Average milk production (L/day) 49 Average weight of animal 50 Weight of bedding in kg 51 Animal used for ploughing 31 52 Sources of feed 53 Other sources of feed [If Q 50 is yes] 54 Details of feed and season [If Q 50 is yes] 55 Quantity fed (kg/day) [If Q 50 is yes] 56 Days fed in a week [If Q 50 is yes] 57 Quantity purchased (kg) [If Q 50 is yes] 58 Purchase price of feed(s) [If Q 50 is yes] 59. Livestock Products [Question Matrix]: [If Q 50 is yes] Product No. Livestock Total quantity Unit per time (ltr Percent home Percent sold at Total value of Notes (Include any other produced per day) [Text] consumption (%) market (%) sales (LCU) use for product ex. Feed Product [Decimal] for calves, gift) [Text] 1 2 3 4 5 32 60. Livestock activities and inputs [Question Matrix]: Total HH labour Hired labour Labour Type Cost of Number Activity [Select one] required required Input [Text] Notes [Select Many] input per (hrs/animal) (hrs/animal) year 1. Herding/ grazing 1. Household 1 2. Feed collection/ prep member 3. Feeding 2. Permanent 4. Watering employed 5. Housing maintenance worker 6. Milking 3. Unpaid labour 7. Insemination 4. Hired labour 8. Calving (temp) 9. Health 5. Contract 10. Other worker 2 3 4 5 Other livestock questions 61 Other herd costs (not covered elsewhere) [If Q 50 is yes] Decimal Hint: 62 Hired labour (for general herd activities) in hrs [If Q 50 is yes] Decimal 63 Average cost of casual labour per hour [If Q 50 is yes] Decimal Hint: Usually unskilled labour 64 Average cost of regular labour per hour [If Q 50 is yes] Decimal Hint: Usually skilled labour 65 Any further notes on livestock? [If Q 50 is yes] Text 33 66. Livestock movement: Animal Stable/ barn Own fields or Yard (days; Off-farm Notes (days; hr/day) pasture (days; hr/day) (days; hr/day) hr/day) 67. Manure Utilization: % % used as Total value % used as % used as % Season returned house of sales Notes (include any other use; columns fuel bio-gas sold to soil maintenance (LCU) should be 100%) 68. Manure storage Manure storage Other storage [Text] Duration (weeks) Any added material? Notes [Select one] [Decimal] (kind and amount) [Text] 1. Applied fresh without Season storage 2. Open heap 3. Under roof 4. Sealed 5. Other 34 Income and Expenditure: 69 Total value of off-farm income per year Decimal 70 Total value of income from land per year Decimal 71 Do you receive any additional income (from pensions, Select one government schemes, family remittances etc.)? 72 Value of additional income per year [Only displayed if Q 84 is Decimal Yes] 73 Do you receive any additional income from general farm Select one subsidies? 74 Value of additional income per year [Only displayed if Q 86 is Decimal Yes] 75 Any additional information on income? Text 76. Other expenses [Question Matrix]: Type Per year Notes General farm costs (not covered elsewhere) Cost of land (per ha) Salaries paid to permanently employed workers Hint: This includes household workers and NOT farm workers Housing costs Food costs (per week) Health costs Personal costs 35 Appendix 5: Cumulative crop areas of the case study farms Cumulative crop areas from the case study farms Wheat (bread) Fava Bean Barley (food) Teff Field Pea Chickpea Lentil Vetch Sorghum Potato Kidney bean Linseed 0 2 4 6 8 10 12 14 GudoPD1 GudoPD2 GudoPD3 GudoPD4 GudoPD5 GudoPD6 GudoND1 GudoND2 GoshePD1 GoshePD2 GoshePD3 GoshePD4 Appendix 6: Feed balance of the case study farms One of the indicators that was analysed through FarmDESIGN and could not be extracted from the RHoMIS survey data, was the feed balance. This includes the total requirements for all animals on the farm and whether these requirements are met by the supply of Feed balance deviation overview 100,00% 80,00% 60,00% 40,00% 20,00% 0,00% -20,00% -40,00% -60,00% -80,00% -100,00% Saturation Energy Protein 36 Gudo PD1 Gudo PD2 Gudo PD3 Gudo PD4 Gudo PD5 Gudo PD6 Gudo ND1 Gudo ND2 Goshe PD1 Goshe PD2 Goshe PD3 Goshe PD4 Goshe PD5 Goshe PD6 Goshe ND1 Goshe ND2 feed. Figure 9 provides an overview of the feed balance deviations based on the FarmDESIGN data acquired through farmer interviews. It shows that based on the given information most farms are not able to provide ample feed for their livestock to maintain sufficient productivity. Furthermore, farms that were able to supply sufficient feed such as Goshe PD 4 and Goshe PD5 had to buy significant amounts of external feed. These feed sources include concentrates such as oil cake and wheat bran. Consequently, based on the interviews only few farms were able to sustain adequate levels of livestock productivity. Appendix 7: Workshop content Introduction: - The purpose of the workshop is to stimulate knowledge transfer and to create awareness about the possibilities of farm redesign in relation to adoption of specific sets of SI interventions. o What do participant value when it comes to farm system redesign? Based of SIAF indicators - Sustainable intensification has been adopted in the region through projects such as AfricaRISING. However, scaling of the practices related to SI is important to allow for larger adoption rate. Thus, more farmers will benefit from these innovations - Experience sharing of farmers key to create understanding for benefits - Feedback of farm system redesign Program 1. Short game related to decision making in farm system redesign 2. Final discussion and wrap up session Part 1: short game related to farm system functioning and decision making - Soil fertility consequences related to compost manure application and crop yield - Market orientation: some market orientation can be beneficial to provide opportunity ReHab adaptation SI (Page et al., 2016) - 4 x 5 playing field - Soil fertility test - Players play in five round - Each round players determine what they plant and how they want to harvest - Manure can be added to increase fertility Description: The game represents a landscape consisting of 20 plots of 1 tamad (total 5 hectares). Each of the player has to manage the plot for four growing seasons. Players individually have to make decisions regarding what crop they grow each season, whether these crops are native or improved varieties, and how they utilize the manure produced on their farm. Growing improved varieties result in an increased yield which is beneficial for the farms income and food security. Growing native varieties helps conserve the genetic variety in the region. Manure can either be used to make compost which is added to the soil to increase soil fertility or it can be used for fuel to generate energy. The game represents a simplification of a real life farm situation, but simulates real life decision making. Players play two games in total. The first game communication between players is forbidden to allow for individual strategy development. After the first game, a second game is played in which communication is allowed. This stimulated team strategy building and allows to analyse the difference in strategy development to the game without communication. Also group dynamics can be analysed. 37 The goal is to create an understanding of soil fertility management related to compost manure application and crop rotation. Also showing the effect of growing improved variety or native variety is included in the game. In the end it is up to the players to pick a strategy that they see fit to achieve their own goals. Materials required - 20 land ownership tokens (2 x 10 colours) - 50 soil fertility tokens (small rocks) - 100 crop tokens for each crop o Barley (native/improved) o Wheat (native/improved) o Teff (native/improved) o Faba bean (native/improved) o Field pea (native/improved) - 1 playing field (a net consisting of 20 plots) o Can be created by making a drawing on a large paper sheet o Can be created in the field by drawing on the ground with a stick - 10 pencils for the players - 20 player cards (1 per player per game) - 2 moderator cards (1 per game) Game preparation - The game is played using a playing field consisting of 5 x 4 grid that represents 20 plots of 1 tamad (0.25 hectares). - The game is played with a number of players ranging from 5 to 10. - Each player receives two tokens that are used to indicate the plots of the particular farmer. Farmers one after another place one token on a plot of their choice. After two rounds all the tokens are divided over the plots. It is not allowed to have multiple tokens on one plot. - Each farmer receives two barley, wheat, teff, faba bean, and field pea tokens to start with. Additionally the farmers every round receive two manure tokens. Game play - First an overview of the game is given to the players. The underlying rules are not mentioned. The moderator addresses the fact that communication is forbidden during the first game. - The first round starts with the farmers determining what crops they would like to grow in what plot. Farmers therefore have to select two crops, one per plot. - The second phase of the round is to determine the destination of the manure. Farmers can make compost and put it on the soil or use it as fuel. o Farmers can decide to put all compost on one plot or to divide between the plots o It is allowed to use one token as fuel and one token for compost o It is not possible to divide one token of compost over the two plot - The compost decision is also written on the card - All cards of the farmers are collected and the results processed on the moderator card - The moderator then manages the yield of the farmers and places the correct tokens on the plot of the players - After placing the crop tokens the moderator will place the soil fertility tokens on the plots indicating the current level of soil fertility - Player can then collect their yield and start off with two (new) manure tokens - This process is repeated three times until three growing seasons have been played - At the end farmers show their total harvest which is written down by the moderator - The game is then repeated and play similarly to the first game, however, communication is allowed during the second game. 38 Explanation of underlying rules: The game consists of few rules that are know the players before playing the game. However, there are underlying dynamics that are unknown to the players. - It is assumed that improved varieties result in higher yields than native varieties - Manure application will results in more nutrient availability also affecting the yield positively - Crop rotation will also affect yield in subsequent season. Growing legumes previous to cereals, increases cereals’ yield in the next year. The following tables show the effect of underlying rules: Manure Rotational effect Barley -1 -1 0 0 Compost ++ Compost + Compost - Wheat -1 -1 0 0 Improved 2 1 0 Season 1 Crop Faba bean 1 1 1 1 Local 2 1 0 Field pea 1 1 1 1 Barley Wheat Faba Field pea Season 2 Strategies (still to observe during the workshop) - Native species conservationist - Big harvester - Manure fuel dependent - Big farmer Indicators to score the game - Total amount harvested per crop type - Soil fertility number per farm and cumulative - make a distinction between with and without communication Part 2: Debrief and closing of session (Page et al., 2016) Immediate emotional response of the game - How did you feel while playing the game? - How do you feel now? Visualization of indicators - What do you think about the collective results? - What do think about your own individual result? Understanding of the game - What did you understand? - What were the hidden rules? - What did you learn? Collective analysis of group dynamics - Did you reach agreements? - How were conflicts resolved? - Was there someone that was able to act as a leader? How did that person do? 39 Bridging gaming context and reality of tragedy of the commons - What are the differences and similarities between the game and reality? - Why is it more beneficial to apply manure to the field then to use it as fuel? - How would you approach market orientation in such situations? Example of the game moderating sheet Session Season Season Season Result Game 1 2 3 Input Input Input Season 1 Season 2 Season 3 Crop Plot Crop Manure Plot Crop Manure Rot. Plot Crop Manure Rot. Plot # Yield Plot # Yield Plot # Yield result 1 wi 1 2 1 1 1 3 1 0 1 0 2 0 2 2 2 0 2 0 2 0 3 0 3 3 3 0 3 0 3 0 4 0 4 4 4 0 4 0 4 0 5 0 5 5 5 0 5 0 5 0 6 0 6 6 6 0 6 0 6 0 7 0 7 7 7 0 7 0 7 0 8 0 8 8 8 0 8 0 8 0 9 0 9 9 9 0 9 0 9 0 10 0 10 10 10 0 10 0 10 0 11 0 11 11 11 0 11 0 11 0 12 0 12 12 12 0 12 0 12 0 13 0 13 13 13 0 13 0 13 0 14 0 14 14 14 0 14 0 14 0 15 0 15 15 15 0 15 0 15 0 16 0 16 16 16 0 16 0 16 0 17 0 17 17 17 0 17 0 17 0 18 0 18 18 18 0 18 0 18 0 19 0 19 19 19 0 19 0 19 0 20 0 20 20 20 0 20 0 20 0 40 Appendix 8: Workshop results Individual gaming results Game 1 - Individual result 10 9 8 7 6 5 4 3 2 1 0 Player 1 Player 2 Player 3 Player 4 Player 5 Player 6 Player 7 Player 8 Player 9 Player 10 Round 1 Round 2 Round 3 Game 2 - Individual result 14 12 10 8 6 4 2 0 Player Player 2 Player 3 Player Player 5 Player 6 Player 7 Player 8 Player 9 Player 10 1 4 Round 1 Round 2 Round 3 41 Wheat improved Wheat improved Faba bean improved Faba bean improved Wheat improved Wheat improved Faba bean improved Barley improved Field pea Faba bean improved Wheat improved Wheat improved Barley improved Wheat local Faba bean improved Faba bean local Barley improved Field pea Faba bean improved Wheat improved Wheat improved Faba bean improved Faba bean improved Faba bean local Wheat improved Wheat improved Barley improved Barley improved Faba bean improved Faba bean improved Faba bean local Wheat improved Field pea Wheat improved Barley improved Barley improved Faba bean improved Faba bean improved Wheat improved Faba bean local Barley local Wheat improved Wheat local Faba bean improved Barley improved Faba bean local Faba bean improved Wheat improved Faba bean local Barley improved Wheat improved Barley improved Faba bean improved Field pea Wheat improved Wheat improved Barley improved Barley improved Faba bean improved Faba bean local Wheat improved Wheat improved Wheat local Faba bean improved Barley local Faba bean local Faba bean improved