Developing breakthrough products that deliver at least 25% improvements in productivity in farmers' fields requires that breeding pipelines more closely align with farmers' and end-user needs and the target population of environments (TPEs). In the case of maize, climate change is accelerating climatic shifts within the TPE, potentially limiting future realised genetic gain. On-farm breeding approaches can play a valuable role in supporting breeding programs transitioning towards breakthrough products. Increased sampling of the TPE by moving selection into farmers' fields can overcome the challenges of climatic changes within the TPE and extensive farmer diversity, whilst providing information to refine managed environment trials designed to sample the TPE. This review highlights advances in inclusive host farmer sampling, sparse testing strategies and phenotyping to facilitate scaling of on-farm testing within breeding pipelines.

KEY MESSAGE:Our scalable two-step method estimates genotypic performance and genetic parameters from ranking data, producing reliable results comparable to quantitative analyses, enabling the integration of ranking data into breeding pipelines. Plant breeding research has chiefly relied on on-station experiments to evaluate varietal performance. Nevertheless, these trials often fail to represent on-farm growing conditions and farmers' preferences, potentially leading to poorly defined breeding targets. Recent work has demonstrated the potential of using on-farm verification trials combined with ranking data to support farmers in evaluating varieties while providing information that is representative of farmers' needs. Despite this potential, scalable methods for quantifying genetic differences and assessing the strength of the genetic signal in such trials remain limited. Here, we present a two-step procedure for analyzing trials based on ranking data, allowing the estimation of genetic parameters. The approach follows a common strategy in quantitative genetics, in which parameters are estimated from tables of genotypic means and their variances. In our framework, these estimates are obtained from Thurstonian and/or Plackett-Luce models, which treat rankings as observations of an underlying continuous trait associated with genotypic performance. Using simulated data, we showed that genotypic mean estimates derived from ranking analyses are linearly related to those obtained from quantitative trait analyses and that their variances adequately capture estimation uncertainty. We further demonstrated that incorporating these estimates and their variances into a second-step mixed-effects model yields accurate estimates of variance components. Analyses of groundnut, maize, and sweetpotato datasets confirmed the applicability of the approach and showed that ranking data can provide reliable estimates of genetic parameters. We argue that this framework can be scaled to obtain genotypic performance estimates from multi-trial on-farm data.

ABSTRACT High kernel yield and farmer‐ and market‐preferred traits are overriding considerations for groundnut (Arachis hypogaea L.) breeding, production and adoption. However, yield expression and cultivar selection responses in groundnuts are influenced by genotype‐by‐environment interactions (GEI) and management conditions. Therefore, it is essential to evaluate GEI to identify high‐yielding and stable groundnut genotypes preferred by farmers and markets for breeding or variety recommendations. The objectives of this study were to assess the GEI and farmers' preference traits in groundnut to facilitate the selection of superior genotypes for release and guide breeding with specific or broad adaptation while integrating farmers' and gender perceived traits. The study was conducted across 18 environments representing agroecological zones and potential groundnut production areas in Tanzania. Sixteen genotypes, including two commercial checks, were evaluated in selected locations using a randomized complete block design. Furthermore, on‐farm decentralized trials were conducted across 42 locations following the tricot approach. Significant (p < 0.05) variations were detected among the tested genotypes (G), environments (E), and GEI effects on kernel yield. Genotype ICGV‐SM 05534 had a relatively highest kernel yield of 927.232 kg ha−1. The GGE biplot identified ICGV‐SM 16645 and ICGV‐SM 10014 as the most stable genotypes across locations, with mean kernel yields of 936.39 and 877.67 kg ha−1, respectively. The triadic comparison of technologies (TRICOT) analysis identified gender differences in trait preferences among groundnut growers. Early maturity, ease of harvesting and shelling were the most preferred traits by female farmers, and haulm yield by males. Tan color and small‐medium kernel seed size were identified as the top traits selected by farmers' overall varietal preferences. Additionally, early maturity is an important trait to consider in the market segment. Farmers' preference traits are crucial during the initial stage of designing the target product profile to increase the adoption of the deployed groundnut variety. The selected genotypes, test environments, and farmer‐preferred traits are vital for breeding pipelines, targeted variety release, and production for different target population environments (TPEs) in Tanzania.

Step-change innovation in seed product design by public sector crop breeding has led to major contributions to global food security. The literature, however, provides few insights on how to identify forward-looking innovation opportunities. Inspired by discussions in the product innovation literature, this article describes our application of product concept testing in the context of hybrid maize in Uganda and Kenya. We identified the following eight maize seed product concepts based on interactions with seed companies, crop breeders, and farmers: ‘Resilience’, ‘Drought escape’, ‘Food and fodder’, ‘Home use’, ‘Green maize’, ‘Livestock feed’, ‘Intercropping’, and ‘Family nutrition’. These were described and presented to 2400 farmers using videos, where each farmer saw three concept-presentation videos. Farmers were most likely to have selected the resilience (Kenya and Uganda), drought escape (Uganda), and intercropping (Kenya) concepts. Farmers showed mixed interest in other concepts, such as home use and food and fodder, suggesting that investments in product production and promotion would be required in addition to investments in breeding. These results provide new entry points for conversations among transdisciplinary teams at regional and national levels on the current and future opportunities for crop breeding to respond to farmers’ requirements for new seed products.

Potato (Solanum tuberosum L.) is crucial for food security in Rwanda, but its production growth has slowed. Improved potato varieties are urgently needed for Rwanda potato farmers. Crop breeding can effectively support smallholder farmers when it aligns with their environmental conditions and preferences. Additionally, integrating citizen science into variety development can enhance variety adoption and suitability for smallholder farmers. We assessed the insights from a crop trial following a triadic comparison of technology options (tricot) approach, linking the results with environmental, socio-economic, and on-station trial data. Under a tricot trial, 460 farmers tested eleven potato varieties, randomly allocated in incomplete blocks of three, allowing each farmer to test and compare three varieties. Biological data, reflecting breeding and variety genotypic values, were generated from multi-environmental tests conducted during 2018-2019 to evaluate the adaptability of new varieties. This research revealed that Rwandan farmers preferred the pre-1990 varieties (Cruza and Kirundo), while Gisubizo and Kazeneza, post-2018 varieties, were also considered competitive. Farmers' preferences were influenced by diverse environmental and socio-economic conditions, with taste being crucial for home consumption and yield prioritized for market sales. Additionally, seasonal temperatures influenced the yield performance ranking of potato varieties across regions, while economic considerations and gender dynamics shaped different patterns of variety preferences. Despite challenges in aligning on-station and on-farm data, our integrated approach provides actionable insights for breeding programmes to develop potato varieties that better align with farmers' needs, as well as environmental and socio-economic conditions. This innovative method can enhance breeding efficiency, variety adoption, and potato productivity, contributing to food security and agricultural sustainability.

Societal Impact Statement Amid global challenges of food insecurity, poor nutrition, and climate change, neglected crops like amaranth are gaining renewed attention. We studied farmers' preferences for amaranth varieties across diverse geographical contexts to guide targeted breeding. Our results revealed significant variation in farmer preferences, emphasizing the need for context‐specific breeding strategies. These findings can support the development of improved amaranth varieties that meet local needs, expand economic opportunities—especially for women—promote healthier diets, and boost biodiversity. This work also offers a model for participatory research on opportunity crops, informing inclusive agricultural policies and sustainable development strategies across Africa and beyond. Summary Opportunity crops, also known as neglected and underutilized species (NUS), offer benefits to diversify food systems with nutritious and climate‐resilient foods. A major limitation to incorporating these crops in farming systems is the lack of improved varieties, which impedes farmers from accessing quality planting materials of these crops. The study explored how citizen science methods can support demand‐driven breeding and seed production of NUS using leafy amaranth – a nutritious and hardy vegetable ‐ as a case study. The study identified farmer preferences and market segments, with particular attention to gender and social differentiation. We used the tricot approach to conduct participatory on‐farm trials of 14 varieties with 2,063 farmers from Benin, Mali, and Tanzania. We then analyzed farmer traits and varietal preferences in aggregate and among segments of farmers, using cluster analysis. Farmers' overall preferences for amaranth varieties were driven principally by plant survival, yield, leaf size, taste, and marketability. Distinct farmer segments (older women generalists, young women specialists, older men generalists, and young men specialists) preferred different varieties depending on gender and business orientation. The identified farmer segments, along with their unique variety preferences, provide valuable information for breeders and seed enterprises, and support demand‐driven amaranth breeding and seed system development. The methods used and lessons learned from our citizen science exercise can be applied to enhance breeding and seed supply of other opportunity crops that are underutilized in Africa and elsewhere.

Improving agricultural productivity and resilience is essential to meet future food needs in sub-Saharan Africa under changing climate conditions. Achieving this will necessitate the development of high-yielding locally adapted crop varieties to mitigate the impacts of climate change. Despite advancements in crop improvement, varietal turnover in smallholder farms remains notably low. Continuous turnover of locally adapted varieties is essential, necessitating active dissemination of new varieties and withdrawal of obsolete ones across diverse target populations using participatory breeding approaches. A decentralised experimental approach, known as tricot, supported by citizen science, has proven effective in accelerating genotype selection while promoting inclusivity and diversity. However, the methodology has strongly relied on farmer-generated rankings, which provide relative performance insights but fall short in informing breeders with absolute yield data, limiting the ability to measure genetic gain or assess economic returns on breeding investments. To address this gap, we validated the accuracy of farmer-generated yield data for common bean (Phaseolus vulgaris L.), by comparing it with technician-generated volumes and researcher-generated absolute yield data. Results revealed strong correlations between farmer and technician volumes (r = 0.96, p < 0.001). The mean difference in farmer-technician log-yield was close to zero, indicating significant agreement. We further developed a predictive model to estimate absolute yields using farmer showing minimal influence from intrinsic and extrinsic factors. Our findings demonstrate that farmer-generated yield data can reliably inform breeding decisions and support the accelerated turnover of improved varieties. Integrating such data into breeding programs offers a cost-effective and scalable pathway to enhance agricultural productivity and sustainability across smallholder systems in sub-Saharan Africa.

BACKGROUND: Nigeria and Cameroon are multi-ethnic countries with diverse preferences for food characteristics. The present study aimed to inform cassava breeders on consumer-prioritized eba quality traits. Consumer testing was carried out using the triadic comparison of technologies (tricot). Diverse consumers in villages, towns and cities evaluated the overall acceptability of eba made from different cassava genotypes. Data from both countries were combined and linked to laboratory analyses of eba and the gari used tomake it. RESULTS: There is a strong preference for eba with higher cohesiveness and eba from gari with higher brightness and especially in Cameroon, with lower redness and yellowness. Relatively higher eba hardness and springiness values are preferred in the Nigerian locations, whereas lower values are preferred in Cameroon. Trends for solubility and swelling power of the gari differ between the two countries. The study also reveals that the older improved cassava genotype TMS30572 is a benchmark genotype with superior eba characteristics across different regions in Nigeria, whereas the recently released variety Game changer performs very well in Cameroon. In both locations, the recently released genotypes Obansanjo-2 and improved variety TM14F1278P0003 have good stability and overall acceptability for eba characteristics. CONCLUSION: Thewide acceptance of a single genotype across diverse geographical and cultural conditions in Nigeria, aswell as three acceptable new improved varieties in both locations, indicates that consumers' preferences are surprisingly homogeneous for eba. This would enhance breeding efforts to develop varieties with wider acceptability and expand potential target areas for released varieties. (c) 2023 The Authors. Journal of The Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

CONTEXT: Opportunity crops, also known as neglected and underutilized species (NUS), offer benefits to diversify food systems with nutritious and climate-resilient foods. A major limitation to incorporate these crops in farming systems is the lack of improved varieties impedes farmers accessing quality planting material of these crops. OBJECTIVES: The study explored how citizen science methods can support demand-driven breeding and seed production of opportunity crops using leafy amaranth, a nutritious and hardy vegetable, as a case study. The study identified farmer preferences and market segments, with particular attention to gender and social differentiation. METHODS: We used the tricot approach to conduct participatory on-farm trials of 14 varieties with 2,063 farmers from Benin, Mali, and Tanzania. We then analyzed farmer trait and varietal preferences in aggregate and among segments of farmers, generated using cluster analysis. RESULTS: Farmers overall preferences for amaranth varieties was driven principally by plant survival, yield, leaf size, taste, and marketability. Distinct farmer segments (older women generalists, young women specialists, older men generalists, and young men specialists) preferred different varieties depending on gender, business-orientation. DISCUSSION AND CONCLUSION: The farmer segments identified here, along with their unique variety preferences provide valuable information for breeders and seed enterprises, and support demand-driven amaranth breeding and seed system development. We specifically noted the need for breeding programs to understand the preferences of young amaranth specialists, both men and women, and to explore organoleptic and market-related properties of amaranth. ### Competing Interest Statement The authors have declared no competing interest.

Abstract Enhancing food security for smallholder farmers amidst challenging climatic conditions requires accurate quantification of agricultural production. However, there are increasing debates on the weaknesses of commonly used yield estimation approaches in smallholder production systems. There is also a limited understanding of the implications of methodological choice of different yield estimation approaches at varying scales in diverse production systems. As such, smallholder farming systems in sub-Saharan Africa (SSA) have been characterized with substantial crop yield variability which is persistent even within the same agro-ecological zones. Moreover, broad comparisons for regional and national agricultural productivity have often formed the basis of certain misconceptions about crop productivity across a wide diversity of production systems, each with its own peculiarities and oddities. Additionally, the key factors influencing accuracy and reliability of yield data are still not well understood and are often overlooked when estimating yield in both small and large-scale surveys. This review provides an in-depth comparative analysis of trade-offs and sources of error in commonly applied yield estimation approaches in SSA production systems. Also, here for the first time, we collectively discuss the key factors that impact accuracy and reliability of yield data in diverse production systems in sub-Saharan Africa. Our review provides useful insights for standardizing on-farm yield measurement approaches and benchmarking crop production in smallholder production systems, a prerequisite for recommendations and decision-making in agricultural research. Improving the measurement of yield data will increase the understanding of diverse smallholder production systems and consequentially improve the targeting of productivity-enhancing interventions.

Experimental citizen science offers new ways to organize on-farm testing of crop varieties and other agronomic options. Its implementation at scale requires software that streamlines the process of experimental design, data collection and analysis, so that different organizations can support trials. This article considers ClimMob software developed to facilitate implementing experimental citizen science in agriculture. We describe the software design process, including our initial design choices, the architecture and functionality of ClimMob, and the methodology used for incorporating user feedback. Initial design choices were guided by the need to shape a workflow that is feasible for farmers and relevant for farmers, breeders and other decision-makers. Workflow and software concepts were developed concurrently. The resulting approach supported by ClimMob is triadic comparisons of technology options (tricot), which allows farmers to make simple comparisons between crop varieties or other agricultural technologies tested on farms. The software was built using Component-Based Software Engineering (CBSE), to allow for a flexible, modular design of software that is easy to maintain. Source is open-source and built on existing components that generally have a broad user community, to ensure their continuity in the future. Key components include Open Data Kit, ODK Tools, PyUtilib Component Architecture. The design of experiments and data analysis is done through R packages, which are all available on CRAN. Constant user feedback and short communication lines between the development teams and users was crucial in the development process. Development will continue to further improve user experience, expand data collection methods and media channels, ensure integration with other systems, and to further improve the support for data-driven decision-making.

Matching crop varieties to their target use context and user preferences is a challenge faced by many plant breeding programs serving smallholder agriculture. Numerous participatory approaches proposed by CGIAR and other research teams over the last four decades have attempted to capture farmers’ priorities/preferences and crop variety field performance in representative growing environments through experimental trials with higher external validity. Yet none have overcome the challenges of scalability, data validity and reliability, and difficulties in capturing socio-economic and environmental heterogeneity. Building on the strengths of these attempts, we developed a new data-generation approach, called triadic comparison of technology options (tricot). Tricot is a decentralized experimental approach supported by crowdsourced citizen science. In this article, we review the development, validation, and evolution of the tricot approach, through our own research results and reviewing the literature in which tricot approaches have been successfully applied. The first results indicated that tricot-aggregated farmer-led assessments contained information with adequate validity and that reliability could be achieved with a large sample. Costs were lower than current participatory approaches. Scaling the tricot approach into a large on-farm testing network successfully registered specific climatic effects of crop variety performance in representative growing environments. Tricot’s recent application in plant breeding networks in relation to decision-making has (i) advanced plant breeding lines recognizing socio-economic heterogeneity, and (ii) identified consumers’ preferences and market demands, generating alternative breeding design priorities. We review lessons learned from tricot applications that have enabled a large scaling effort, which should lead to stronger decision-making in crop improvement and increased use of improved varieties in smallholder agriculture.

CONTEXT: Digital innovations can enhance the participation of often-marginalized social groups - including women and resource-poor farmers in low- and middle-income countries - in sustainable, profitable food systems. But digital interventions can also reinforce existing inequities by further increasing the competitive advantage of user groups privileged with literacy, access to smartphones, or high investment capacity. To ensure that the digital transformation in the Global South leaves no one behind, therefore, deliberate efforts are needed to promote the inclusivity of emerging digital innovations. To date, however, there is a lack of practical guidelines and tools to critically assess, demonstrate, and enhance the inclusivity of digital food systems interventions. Too often, inclusivity remains a blurry concept and distant objective. In result, digital development researchers and practitioners have limited incentives for investing time and effort into safeguarding inclusivity. OBJECTIVE: With this short communication, we intend to contribute to future, practice-oriented discussions about social inclusivity in development-oriented digital interventions for sustainable food systems. We provide a critical reflection on the current discourse around digital inclusion in development context and outline challenges and opportunities for considering inclusivity in the design and deployment of digital food system innovations. METHODS: Drawing on literature as well as the authors' own experiences with the design and implementation of digital innovations within research-for-development, we highlight 'blind spots' in the current discourse around digital inclusion in lowand middle-income country context. We then develop practical suggestions for overcoming these limitations. RESULTS AND CONCLUSIONS: We propose a concrete agenda for enabling researchers and other innovation stakeholders, including donors, to contribute to more inclusive digital food system innovation in lowand middle-income countries. First, a standard concept and procedure is required for transparently assessing the inclusivity of digital services. Second, as many digital development stakeholders work under resource constraints, simple design tools can help them effectively consider social inclusion criteria during the design of digital solutions. Lastly, a stronger emphasis on inclusivity is required throughout the research-for-development system, ensuring that design processes themselves are inclusive, rather than considering only the final digital products. SIGNIFICANCE: As the importance of digital innovation keeps growing within the wider agricultural development discourse, this article helps researchers and practitioners gain conceptual clarity on the goal of digital inclusion. Through concrete suggestions on how inclusivity could be considered in practice, the article promotes a more equitable, inclusive digital transformation of food systems.

Participatory approaches for crop variety testing can help breeding teams to incorporate traditional knowledge and consider site-specific sociocultural complexities. However, traditional participatory approaches have drawbacks and are seldom streamlined or scaled. Decentralized on-farm testing supported by citizen science addresses some of these challenges. In this study, we compare a citizen science on-farm testing approach - triadic comparisons of technology options (tricot-PVS) - with the benchmark state-of-the-art group-based participatory variety testing approach (group-PVS) over a set of socioeconomic outcomes. We focus on on-farm testing of common bean (Phaseolus vulgaris L.) in the Trifinio area of Central America. We measure the impact of these two approaches on bean growers in terms of on-farm diversification and food security. We use data from 1978 smallholder farmers from 140 villages, which were randomly assigned to tricot-PVS, group-PVS or control. Utilizing a difference-in-difference model with inverse probability weighting and an instrumental variable approach, we observe that farmers involved in group-PVS, and tricot-PVS had comparable levels of on-farm varietal diversification with respect to control farmers. Nonetheless, group-PVS appears to be significantly more effective in boosting household food security, which can be attributed to improved agronomic management of the crops. This study contributes to the next generation of innovations in exploring trait preferences to produce more inclusive, demand-driven varietal design that democratize participatory varietal selection programs.

The triadic comparison of technology options (tricot) approach to participatory varietal selection has been demonstrated to scale out the on-farm testing of elite candidate crop varieties. In this study, we evaluated elite clones of cassava (Manihot esculenta Crantz) using the tricot approach. We sought to (1) synthesize on-farm trial data from two cycles evaluating different sets of elite cassava clones; (2) assess the agronomic performance of elite cassava clones considering sociodemographic and climatic factors; and (3) assess the performance of elite cassava clones for both agronomic and food quality traits. The study involved 10 districts in Uganda, two cycles/seasons of evaluation, 20 elite cassava clones, one check variety, and 294 men and 320 women farmers. Our results indicate that the agronomic performance of elite cassava clones was influenced more by geographic than sociodemographic factors. Our analysis identified the number of days with rainfall higher than 20 mm as the most influencing climatic factor over agronomic performance. Further, the study identified superior elite cassava clones UG110164, UGC14170, and UG120193 as promising candidate varieties for release, targeting food products. Overall, our study emphasizes the important contribution of end-users to crop improvement and provides insights into use of tricot on-farm testing methodology to evaluate elite cassava clones during cassava variety development in Uganda, which can be used to support decision making for variety release. We applied the tricot (triadic comparison of technology options) approach to evaluate elite cassava clones on-farm, exploring how the linked data can be used to advance clones for breeding. We used a rank-aggregation approach to assess the on-farm performance of two different sets of elite cassava clones. We assessed the effect of sociodemographic and geographic factors on the on-farm agronomic performance of elite cassava clones. We assessed the performance of elite cassava clones for beneficial agronomic and food quality traits for consideration in advancement and varietal release.

To cope with interannual climate variability, many farmers in tropical and sub-tropical regions choose crop varieties that fit seasonal climate conditions. Therefore, seed demand for different varieties, such as early- or late-maturing cultivars, varies between years. Resulting mismatches between relatively constant supply and variable demand create losses for both seed suppliers and farmers. Because demand for seed of different varieties is influenced by seasonal climate, however, probabilistic seasonal rainfall forecasts could help seed suppliers better anticipate upcoming seed demand. To explore this idea, we engaged decision-makers from seed supply organizations in Zimbabwe and Ethiopia. Through a participatory design process, we identified opportunities and challenges for using seasonal rainfall forecasts to inform seed supply decisions. In a case study of maize seed sales in Zimbabwe, we tested our assumptions and iteratively devised a systematic procedure for forecast-based planning in seed supply, relying on free online data sources and expert deliberations. We found that currently accessible rainfall forecasts could indeed be useful for prioritizing likely high-demand varieties during the stages of seed treatment, packaging, and logistics. In practice, though, more flexible and adaptive management of seed supply pipelines might be required to make use of seed demand forecasts. In the future, targeting farmers with climate forecasts along with recommended variety portfolios may strengthen the association between seasonal climate and farmers’ variety demand, increasing the accuracy of demand anticipation. This study highlights opportunities for increased case-specific collaboration between climate scientists and the seed sector to make seasonal forecast information operational.

Synthesis of crop trial data can generate insights that are not available from the analysis of individual studies, but such synthesis is often constrained by the heterogeneity of data among studies. Rank-based data synthesis provides the flexibility to combine data of heterogeneous types and from different sources. We demonstrate the application of rank-based data synthesis of heterogeneous trial data to assess the effect of climatic factors on the reaction of several Musa genotypes to black leaf streak disease (BLSD; caused by Pseudocercospora fijiensis [sexual morph: Mycosphaerella fijiensis]). We aggregated data from the main public repositories of Musa trial data. We applied model-based recursive partitioning with the Plackett-Luce model, using climatic data as covariates. The model identified the maximum length of the dry spell as the main variable influencing differences in genotypic response to BLSD, dividing the aggregated trial dataset into humid and dry environments. We found differences in the reaction of genotypes to BLSD between these environments. In humid environments, NARITA 8 was found to be the most resistant genotype, while in dry environments FHIA-01 was the best performing improved genotype. We also assessed reliability, which is the probability of outperforming the reference genotype (Calcutta 4). In humid environments, NARITA 2, NARITA 8, and FHIA-01 had the highest reliability, while in dry environments only the landrace Saba surpassed 50% reliability. The information generated by our data synthesis approach supports selecting Musa genotypes for further evaluations at new locations.

Agrometeorological data is important in agricultural research, especially in agronomy and crop science, for investigating genotype by environment interactions. The AgERA5 dataset from the Copernicus Climate Data Store provides free and public access to global gridded daily agrometeorological data, from 1979 to present, with ready to use variables tailored for agricultural and agro-ecological studies. We developed the R package ag5Tools, which provides a simplified interface for downloading and extracting AgERA5 data. The package facilitates extracting time-series data for sets of geographic points in a format that can be conveniently used in statistical models applied in agricultural research. The use of the package is demonstrated with a synthetic dataset of multi-location trials in Arusha, Tanzania.