"Breaking through the 40% adoption ceiling: Mind the seed system gaps." A perspective on seed systems research for development in One CGIAR.

Seed systems research is central to achieving the United Nations Sustainable Development Goals. Improved varieties with promise for ending hunger, improving nutrition, and increasing livelihood security may be released, but how do they reach and benefit different types of farmers? Without widespread adoption the genetic gains achieved with improved crop varieties can never be actualized. Progress has been made toward demand responsive breeding, however the draft CGIAR 2030 Research and Innovation Strategy fails to recognize the complexity of seed systems and thus presents a narrow vision for the future of seed systems research. This points to the lack of evidence-based dialogue between seed systems researchers and breeders. This perspective paper presents findings from an interdisciplinary group of more than 50 CGIAR scientists who used a suite of seed systems tools to identify four knowledge gaps and associated insights from work on the seed systems for vegetatively propagated crops (VPCs), focusing on bananas (especially cooking bananas and plantains), cassava, potato, sweetpotato, and yam. We discuss the implications for thinking about and intervening in seed systems using a combined biophysical and socioeconomic perspective and how this can contribute to increased varietal adoption and benefits to farmers. The tools merit wider use, not only for the seed systems of VPCs, but for the seed of crops facing similar adoption challenges. We argue for deeper collaboration between seed systems researchers, breeders and national seed system stakeholders to address these and other knowledge gaps and generate the evidence and innovations needed to break through the 40% adoption ceiling for modern varieties, and ensure good quality seed once the new varieties have been adopted. Without this, the achievements of breeders may remain stuck in the seed delivery pipeline.

Understanding innovation: The development and scaling of orange-fleshed sweetpotato in major African food systems.

The development and scaling of orange-fleshed sweetpotato (OFSP) during the past 25 years is a case study of a disruptive innovation to address a pressing need - the high levels of vitamin A deficiency among children under five years of age in sub-Saharan Africa. When the innovation was introduced consumers strongly preferred white or yellow-fleshed sweetpotato, so it was necessary to create a demand to respond to that need. This was at odds with the breeding strategy of responding to consumers' demands. Additional elements of the innovation package include seed systems and nutrition education to create the awareness amongst consumers of the significant health benefits of OFSP. Complementary innovation is required in promotion and advocacy to ensure a supportive institutional environment. Four dimensions-- technical, organizational, leadership, and institutional environment-- are explored across five distinct phases of the innovation process, from the emergence of the innovative idea (1991-1996) through scaling phase in 15 countries under a major institutional innovation (2015-mid-2019), the Sweetpotato for Profit and Health Initiative (SPHI). Systematically gathering evidence of nutritional impact and ability to scale cost-effectively was requisite for obtaining support for further development and diffusion of the crop. Positive findings from a major study coincided with a major change in the institutional environment which placed agriculture and nutrition at the forefront of the development agenda, resulting in an inflection point in both research and diffusion investment. The role of committed leadership during all phases was critical for success, but particularly during the first decade of limited support in a challenging institutional environment. The most critical technical achievement underpinning scaling was moving from 2 to 13 African countries having local breeding programs. Evidence is presented that adapted, well performing varieties which consumers prefer is the foundation for successful scaling to occur. Building a cadre of within country and regional advocates was critical for getting sustained commitment and local buy-in to the concept of biofortification by regional bodies and governments, which in turn built within country ownership and the willingness of donors to invest. The SPHI united diverse organizations under a common vision with a simple metric--- the number of households reached with improved varieties of sweetpotato. Since 2009, 6.2 million households were reached by July 2019 in 15 SSA countries. Much more remains to be done. Advocacy efforts led to the integration of nutritious foods into many national and regional policies, setting the stage for further investment.

Market Intelligence and Incentive-Based Trait Ranking for Plant Breeding: A Sweetpotato Pilot in Uganda.

Crop breeding programs must accelerate crop improvement, spur widespread adoption of new varieties and increase variety turnover they are to meet the diverse needs of their clients. More comprehensive quantitative approaches are needed to better inform breeding programs about the preferred traits among farmers and other actors. However, the ability of current breeding programs to meet the demands of their clients is limited by the lack of insights about value chain actor preference for individual or packages of traits. Ranking traits based on monetary incentives, rather than subjective values, represents a more comprehensive, consistent, and quantitative approach to inform breeding programs. We conducted a large pilot in Uganda to assess the implementation of a novel approach to trait ranking, using a uniquely large sample of diverse sweetpotato value chain actors. We found meaningful differences in trait ranking and heterogeneity among different actors using this approach. We also show our approach's effectiveness at uncovering unmet demand for root quality traits and at characterizing the substantial trait demand heterogeneity among value chain players. Implementing this approach more broadly for sweetpotato and other crops would increase the effectiveness of breeding programs to improve food security in developing countries.

A review of varietal change in roots, tubers and bananas: consumer preferences and other drivers of adoption and implications for breeding.

This review of the literature on varietal change in sub-Saharan Africa looks in detail at adoption of new varieties of bananas in Uganda, cassava in Nigeria, potato in Kenya, sweetpotato in Uganda and yams in Côte d'Ivoire. The review explored three hypotheses about drivers of varietal change. There was a strong confirmation for the hypothesis that insufficient priority given to consumer-preferred traits by breeding programmes contributes to the limited uptake of modern varieties (MVs) and low varietal turnover. Lack of evidence meant the second hypothesis of insufficient attention to understanding and responding to gender differences in consumer preferences for quality and post-harvest traits was unresolved. The evidence on the third hypothesis about the informal seed system contributing to slow uptake of MVs was mixed. In some cases, the informal system has contributed to rapid uptake of MVs, but often it appears to be a barrier with inconsistent varietal naming a major challenge.

Factors influencing participation dynamics in research for development interventions with multi-stakeholder platforms: A metric approach to studying stakeholder participation.

Multi-stakeholder platforms have become mainstream in projects, programmes and policy interventions aiming to improve innovation and livelihoods systems, i.e. research for development interventions in low- and middle-income contexts. However, the evidence for multi-stakeholder platforms' contribution to the performance of research for development interventions and their added value is not compelling. This paper focuses on stakeholder participation as one of the channels for multi-stakeholder platforms' contribution to the performance of research for development interventions, i.e. stakeholder participation. It uses a quantitative approach and utilizes descriptive statistics and ARIMA models. It shows that, in three Ugandan multi-stakeholder platform cases studied, participation increased both in nominal and in unique terms. Moreover, participation was rather cyclical and fluctuated during the implementation of the research for development interventions. The study also shows that, in addition to locational and intervention factors such as type of the area along a rural-urban gradient targeted by the intervention and human resources provided for multi-stakeholder platform implementation, temporal elements such as phases of research for development intervention objectives and the innovation development process play significant roles in influencing participation. The study concludes that contribution of multi-stakeholder platforms to the performance of research for development projects, programs, policies and other initiatives is constrained by locational and temporal context and conditional on the participation requirements of the objectives pursued by research for development intervention.

Planning an integrated agriculture and health program and designing its evaluation: Experience from Western Kenya.

Multi-sectoral programs that involve stakeholders in agriculture, nutrition and health care are essential for responding to nutrition problems such as vitamin A deficiency among pregnant and lactating women and their infants in many poor areas of lower income countries. Yet planning such multi-sectoral programs and designing appropriate evaluations, to respond to different disciplinary cultures of evidence, remain a challenge. We describe the context, program development process, and evaluation design of the Mama SASHA project (Sweetpotato Action for Security and Health in Africa) which promoted production and consumption of a bio-fortified, orange-fleshed sweetpotato (OFSP). In planning the program we drew upon information from needs assessments, stakeholder consultations, and a first round of the implementation evaluation of a pilot project. The multi-disciplinary team worked with partner organizations to develop a program theory of change and an impact pathway which identified aspects of the program that would be monitored and established evaluation methods. Responding to the growing demand for greater rigour in impact evaluations, we carried out quasi-experimental allocation by health facility catchment area, repeat village surveys for assessment of change in intervention and control areas, and longitudinal tracking of individual mother-child pairs. Mid-course corrections in program implementation were informed by program monitoring, regular feedback from implementers and partners' meetings. To assess economic efficiency and provide evidence for scaling we collected data on resources used and project expenses. Managing the multi-sectoral program and the mixed methods evaluation involved bargaining and trade-offs that were deemed essential to respond to the array of stakeholders, program funders and disciplines involved.

Lapses, infidelities, and creative adaptations: lessons from evaluation of a participatory market development approach in the Andes.

Participatory approaches are frequently recommended for international development programs, but few have been evaluated. From 2007 to 2010 the Andean Change Alliance evaluated an agricultural research and development approach known as the "Participatory Market Chain Approach" (PMCA). Based on a study of four cases, this paper examines the fidelity of implementation, the factors that influenced implementation and results, and the PMCA change model. We identify three types of deviation from the intervention protocol (lapses, creative adaptations, and true infidelities) and five groups of variables that influenced PMCA implementation and results (attributes of the macro context, the market chain, the key actors, rules in use, and the capacity development strategy). There was insufficient information to test the validity of the PMCA change model, but results were greatest where the PMCA was implemented with highest fidelity. Our analysis suggests that the single most critical component of the PMCA is engagement of market agents - not just farmers - throughout the exercise. We present four lessons for planning and evaluating participatory approaches related to the use of action and change models, the importance of monitoring implementation fidelity, the limits of baseline survey data for outcome evaluation, and the importance of capacity development for implementers.