Jonathan Steinke, Ronnie Vernooy, and Jacob van Etten
In the previous module, you learned how to acquire new germplasm according to the rules and regulations that exist at international and national levels, so that it can be tested in farmers’ fields without any impediments.
In this module, you will be introduced to a number of participatory approaches to test the newly acquired germplasm. These approaches put farmers at the centre of the experimental cycle(s). Two of them, participatory variety selection and participatory plant breeding, have a considerable track record whereas the third, crowdsourcing of field trials, is a more recent technique.
At the end of this module, you will be able to:
For decades, national and international research and breeding institutes have been developing new crop varieties that have led to remarkable increases in yield potential, market value, and the capacity to adapt to climatic hazards, such as dry spells, floods, etc. Although conventional crop research has contributed to substantial yield increases, mainly in high-input areas, smallholders with limited access to inputs or credit, often relying on rain-fed farming at marginal sites and under variable climatic conditions, have enjoyed little benefit. In Africa and Latin America in particular, even farmers who have had access to improved varieties have often stopped growing them after a few years because the seeds failed to meet their needs in their production systems.
Farmers often have different priorities from scientists or breeders. To identify varieties that are of maximum use to farmers — for household consumption, sales, cultural use, animal feed, or a combination of these — locally specific characteristics need to be taken into account, beyond purely agronomic attributes like productivity. These characteristics might include cooking quality, marketability, or the quality of stubble for animal feed — features that are of high local and cultural specificity.
In addition, past experience has shown that not all farmers have the same needs, interests, and preferences. Women farmers often have different ideas than men about what traits or characteristics are important. Younger farmers may have different views from older ones, as a result, for example, of their higher education levels or more exposure to influences from outside the community.
Participatory crop improvement emerged as a response to the shortcomings of conventional approaches. It is based on the principle that farmers participate as equal partners alongside agricultural scientists, fairly sharing their knowledge, expertise, and seeds. The results of such collaboration include, not only more effective crop improvement practices, but also strengthening of farmers’ capacity to experiment, learn, and adapt.
In successful participatory variety selection (PVS), organized farmer groups (usually made up of a mix of women and men farmers, but sometimes women only or young farmers only) grow a set of promising varieties or fixed lines in experimental quantities on one or more plots that have been volunteered by one or more farmers in the group. In other words, experimentation takes place in target environments that represent real-life agro-ecological conditions. Together with a facilitator, the farmers evaluate the varieties according to the attributes that matter most to them and maintain the best materials for replication in the next season. Farmer groups may intuitively select a variety that complies with local needs and preferences and discard varieties recommended by breeding stations or governments.
PVS includes five steps:
PVS has been in use since the 1990s and has become a mainstream practice in many plant breeding and rural development programs and projects. The extent of farmer participation in selection activities may range from merely visually selecting among a few pre-release varieties at a field day to selecting and ranking during the growing cycle; participating in selection from a larger, initial pool of materials; cultivating a large number of varieties and selecting on-farm; or following seed production and marketing of promising materials.
At the lesser end of participation, preselected, improved varieties at physiological maturity are presented to farmers by researchers at a breeding station. Usually during a field day, the invited farmers select a variety they would prefer to grow and are then offered small amounts of seeds to test on-farm. In more participatory-oriented approaches, farmers make key decisions throughout the whole process, although usually in close consultation with researchers. This allows the farmers to take into account qualities that do not show up at the field day: resistance to wind, drought, or flooding; weed incidence; labour requirements, etc.
In Latin America, farmer research committees (or CIALs [Comité de investigación agrícola local], as they are called locally) repeat multi-variety trials of unreleased materials in a breeding program over several growing periods. Some CIALs begin their variety trials with 35 different lines, to see which new accession is most suitable to the environment. Spreading the process over several seasons helps them form an accurate picture of the varieties under varying climatic conditions and generates profound local learning with regard to the varieties’ advantages and disadvantages during the whole growing cycle.
A well-established practice in PVS, “mother–baby trials” combine the benefits of trials at a research station or at one farm (with enough land for the experiment) with trials under farmer management. In “mother” trials, breeders and researchers usually assess 10–20 varieties under controlled conditions and measure yield and other variables. A smaller number of these varieties is tested by a larger group of farmers under diverse farming conditions, either in a subsequent cycle or concurrently. These are the “baby” trials. Observations may then be collected from farmers about the varieties’ performance under realistic management conditions and constraints as compared with a local variety (assessment may be simply better than/worse than/same as a local variety), in terms of days to maturity, plant height, disease incidence, taste, yield, and overall preference. The initial information from the mother trial is complemented by that from the baby trials and sometimes even corrected or discarded. This is an effective way to increase varietal diversity and relay information to farmers.
Varieties that are released after PVS often have a higher adoption rate and higher field sustainability, because they respond to the specific requirements of marginal environments, whereas conventional released varieties often do not. PVS is, therefore, especially suitable for typical staple crops of smallholders or vulnerable dryland farmers, such as legumes, maize, wheat, barley, rice, sorghum, or tef. Although PVS is considered to be a mainstream method now, there has been some criticism about the possibility of the approach leading to researcher bias, lack of sustainability, or socially unacceptable outcomes. Pitfalls include biased questioning at the time of selection or providing an incentive for “correct” answers, e.g., by letting farmers hope that answering in a certain way will increase their chances of getting further benefits from the research program. Special attention must be paid to not attaching an incentive to productivity in baby trials, as this could lead farmers to over- or underreport observations of yield. Restricting participation to final field days also brings the risk of “impulse” selection based on the snapshot impression farmers have on that day, regardless of climatic conditions, soil, stress, and other circumstances during the production cycle. When the selected variety is later grown under usual farmers’ conditions, this knowledge gap can have serious repercussions. Farmers might then observe characteristics that are perceived as not desirable. However, despite these possible shortcomings, PVS has been very successful in numerous countries.
 Misiko, M. , 2013 Dilemma in participatory selection of varieties Agricultural Systems 139, , pp. 35–42 PVS has become a standard practice, but this publication shows a number of pitfalls. [limited access] |
|
 Joshi, A., Witcombe, J.R. , 1996 Farmer participatory crop improvement. II. Participatory varietal selection, a case study in India. Experimental Agriculture 32(4), , pp. 461–477 This article describes the pioneering experience when PVS was used for chickpeas and rice in India. The results led to several important recommendations concerning the Indian variety testing and release system. |
|
 Ashby, J.A., Braun, A.R., Gracia, T., del Pilar Guerrero, M., Hernández, L.A., Quirós, C.A., Roa, J.I. , 2001 Investing in farmers as researchers: experience with local agricultural research committees in Latin America Centro internacional de Agricultura Tropical, Cali, Colombia This book is the number one resource for anyone who is interested in the CIAL method; it includes experiences, tips for practitioners, and references to practical handbooks, many of which are useful for the implementation of a PVS and/or participatory plant breeding program. |
|
 International Rice Research Institute., 2006 Module 6. Participatory approaches. Lesson 1. Participatory variety trials for rainfed rice cultivar evaluation In IRRI’s Rice Breeding Course. International Rice Research Institute, Los Baños, Philippines A handy guide on how to carry out PVS, especially mother–baby trials. A section on how to collect farmers’ observations is especially useful. |
|
Vernooy, R., Shrestha, P., Ceccarelli, S., Ríos Labrada, H., Song, Y., Humphries, S., 2009 Chapter 23. Towards new roles, responsibilities and rules: the case of participatory plant breeding In Ceccarelli, S., Guimarães, E.P., Weltzien, E. (editors). , Plant breeding and farmer participation. Food and Agriculture Organization, Rome, Italy. , pp. 613–628 This book chapter provides an overview of various context-adapted, yet successful, experiences in participatory breeding and variety selection from Asia and Latin America. It highlights how participatory crop improvement requires a new way of doing research in terms of roles, responsibilities, and rules. |
Crowdsourcing is an approach used by scientists and companies worldwide to collect data from large numbers of volunteers instead of just a few researchers. Well-established crowdsourcing projects include ones in which thousands of hobby birdwatchers contribute to regular national surveys on bird migration or citizens classify the quality of their nearby water bodies. “Crowds” can fulfill many tasks that highly specialized researchers cannot, because of their geographic spread, the accumulated time they can dedicate to a task, and the sheer number of contributors.
Crowdsourced data collection can be helpful in generating a broad and holistic response to a research question. Crowdsourcing field trials of new crop varieties implies that many farmers carry out small trials instead of a research station conducting one large trial. Researchers merge and analyze data from all trials. This offers the possibility of testing promising material in different climatic regions, on different soil types, under different management regimes, and, most important, under the real-life conditions of many participating farmers. However, crowdsourcing requires special preparation to motivate enough contributors and ensure data quality.
A crowdsourced project can operate over a large geographic area, but is most effective when it focuses on one crop only. A sufficient amount of promising material, such as advanced lines from breeding programs, released varieties that are not well known in the area, or promising landraces, is preselected. Within one crowdsourced project, many varieties may be assessed, but every variety must be tested various times in comparable environments to make statistical analysis possible.
Local learning is stimulated if all varieties are present in each community and if some varieties are repeated. In this way, farmers can compare varieties empirically by visiting plots within their community and making at least one observation of each variety. For example, if there are only five participants in a community, the total number of tested varieties should not exceed 10 to ensure that all varieties occur at least once in the random sample assigned to the community and some varieties are repeated.
Farmers receive and grow small, experimental quantities of three varieties that are randomly allocated and identified only with a code (A, B, C). They closely observe their growth, and report their observations and yield results to the researchers. Farmers do not know the names of the varieties until the final evaluation workshop that closes the experimental cycle. Farmers who are already organized in some way (in rural cooperatives or agricultural research committees) can be more easily approached than individual farmers for crowdsourced trials. As a first step, a trial managed by the group at large could be set up. Subsequently, individual farmers may carry out trials on their own farms to fully exploit the potential of the approach.
Selected seed is produced by researchers in a sufficient quantity to allow for a small plot, e.g., five 6-m rows per variety. In the case of common beans in Central America, for example, that means about 120 g of seed per variety per participant. A local implementing agency, such as an NGO, university, or extension bureau, then establishes a distribution pathway: are there existing farmer groups or other grassroots organizations to contact or will the trials be publicized via corner shops or agricultural fairs, for example. The implementer must also define the criteria for evaluation in consultation with farmers.
The user-friendly software package and online platform, ClimMob (www.climmob.net), is a free all-in-one tool used to design and execute crowdsourcing projects. It allows randomization of varieties for setting up the project and monitoring its progress. It carries out statistical analysis and automatically generates useful information for both the researcher/project leader and the participants. Not only are data about the performance of the varieties collected and analyzed, but differences in varietal preferences and field performance across households and environments are also explained. Varietal preferences are often influenced by the socioeconomic profile of households, while field performance of varieties depends on land characteristics, agronomic management, and weather. ClimMob can also determine whether different groups of participants have different preference profiles. The more information available about factors that might explain farmers’ evaluation, the more useful the results will be, specifically in terms of targeting variety recommendations to types of households and areas.
Some aspects of the performance of varieties, such as early vigour and disease resistance, are evaluated during the cultivation cycle, while others, such as yield and consumption characteristics, are assessed post-harvest. Local trained facilitators are responsible for keeping track of the collection and reporting to the project.
Various channels are available for entering information into ClimMob. Data can be collected on paper observation sheets by the farmers, then handed over to researchers. Information can also be collected by telephone interviews. New data can be added regularly by filling out and uploading a spreadsheet, via an online form, or using an application for Android smartphones (the Android operating system is used by more than 80% of smartphones worldwide at the time of writing). The variety of channels available for information flow enables flexible adaptation to local conditions.
When all available information has been uploaded, the researchers compile and merge data from all implementation areas, carry out analyses using ClimMob, and send the results back to the local facilitators. Because the analysis is automatized, results can be ready the day the last information is added to the system. This makes it possible to present the results in workshops to local groups of participants shortly after harvest. In these workshops, participants discuss the results and make suggestions for improvements. Seed exchange may be encouraged, and registration for the following project cycle may even take place.
ClimMob generates two final outputs: an analytical report and a set of information sheets. The analytical report sums up the results for the researcher. It shows which varieties were ranked highest for each criterion, whether there were any significant differences in the performance of varieties, and whether specific producer groups reported different evaluation results (e.g., variety X was ranked best by women, whereas variety Y was ranked best by men).
The information sheets contain personalized results for every participant. An individual sheet shows all relevant information about the trial and project results, including the names of the varieties that were included in the package for this participant, which are only revealed now. It also includes a specific variety recommendation based on the information available about the farm and the household. At the final workshop, the information sheets serve as a comparison of the experiences among small groups of farmers.
Incentives for farmers to participate are mainly the overall learning experience and the information about crop varieties that they gain through the project. Disclosure of the identity of the varieties only at the end is an incentive to complete the whole cycle. Farmers can also be encouraged to participate by giving them the opportunity to obtain a larger quantity of seed of their preferred variety for free or at a discount; this incentive is made clear at the beginning of the project. Additional appropriate motivations may include information about aspects of crop production practices. Care must be taken not to create incentives that are conditional on providing information, as this may encourage farmers to submit data even when the trial has failed or when no accurate observations were made.
Crowdsourcing is a relatively new approach in this area. Bioversity International is using it in its Seeds for Needs initiative, reaching a total of more than 25 000 farming households in 2015, including:
Crowdsourcing provides an alternative to the mother–baby trials discussed above. Each approach has advantages, but there are two main differences. First, in crowdsourcing, the varieties tested by farmers are randomly allocated, whereas in mother–baby trials, farmers each select the varieties they want to grow in their baby trials. Thus, analysis of the farmer trials is more robust under crowdsourcing. A second difference is that crowdsourcing omits mother trials altogether.
Distributing material for crowdsourcing trials is easier than organizing community plots for variety trials and organizing events during the crop cycle. Thus, crowdsourcing can reach more farmers, as costs per farmer are lower and even local organizations with limited technical capacity can help to organize the trials. Opportunities for training in crop observation, close control of experimental conditions, and detailed interaction on evaluation aspects are more difficult in crowdsourcing. However, in compensation, crowdsourcing provides many more data points representative of a range of growing conditions in an area.
The two approaches, mother–baby trials and crowdsourcing, may be used in combination: the former allowing more detailed observations and in-depth discussions with farmers in a smaller number of locations; the latter following as a scale out to reach a much larger group of farmers without losing the participatory aspect.
 van Etten, J., 2011 Crowdsourcing crop improvement in sub-Saharan Africa: a proposal for a scalable and inclusive approach to food security IDS Bulletin 42(4), , pp 102–110 This article lays the foundation for a comprehensive crowdsourcing approach to crop trials and climate change adaptation, through the massive distribution of promising crop varieties. |
|
 Bioversity International, 2015 Offers training manuals, instructional videos and readings Bioversity International, Rome, Italy This website provides offers training manuals, instructional videos, and other information and resources for designing and executing successful crowdsourcing projects. |
|
Meldrum, G., Sthapit, S., Rojas, W., King, O. , 2013 Agricultural biodiversity enhances capacity to adapt to climate change New Agriculturalist 13(6), This short article highlights some examples of the multiple ways farmers worldwide use agricultural biodiversity to adapt to climate change and backs the demand for more variety availability. |
|
|
Sumberg, J., Okali, C., Reece, D. , 2003 Agricultural research in the face of diversity, local knowledge and the participation imperative: theoretical considerations Agricultural Systems 76(2), , pp 739–753 This article supports the claim that farmer participation can lead to more successful outcomes in marginal environments and advocates the supply of sufficient experimental material, such as (diverse) varieties, for farmer experimentation. |
|
|
Dickinson, J.L., Shirk, J., Bonter, D., Bonney, R., Crain, R.L., Martin, J., Phillips, T., Purcell, K. , 2012 The current state of citizen science as a tool for ecological research and public engagement Frontiers in Ecology and Environment 10(6), , pp 291–297 This article provides an inspiring overview of the different, creative ways citizen science is being used as a tool for ecological research. |
Like PVS, participatory plant breeding (PPB) was born out of the insight that small farmers in unfavourable conditions benefit very little from formal crop research. In PPB, farmers and plant breeders jointly select cultivars by segregating materials in target environments. PPB is more difficult than PVS because some degree of knowledge of heritability and genetic gain is required.
Officially released varieties are often designed for high-input farming practices used in large agro-ecological areas. In contrast, smallholder farmers practise agriculture in micro-regions where particular environmental conditions predominate. In other words, conventional breeding fails to take smallholders’ constraints, needs, and preferences into account. To produce varieties that match the specific requirements of small-scale farmers, PPB involves producers at earlier stages of the plant breeding process than PVS. In PPB, farmers set breeding goals and select parental material; they are trained in the identification and choice of parent lines, making crosses, and managing trials of a large number of varieties. Professional researchers and breeders act as facilitators in this process, guided by the paradigm that the more decision-making power the farmers have in the breeding program, the better adapted and more useful the outcome will be.
PPB relies on lasting relations between researchers and farmers and seeks out “custodian farmers” or local farmer experts who can be further trained in breeding techniques, conserve varietal diversity on their farms, and act as links between the scientific world and their fellow farmers. Experience in various parts of the world has shown that farmers are able to distinguish between very large numbers of crop lines and make choices that are at least as effective as those made by researchers. Moreover, the learning process and access to new varieties can stimulate and encourage farmers to perform further on-farm experimentation and keep adapting varieties to the environmental conditions via selection. A local or regional PPB program can include many farmers at all stages, provided that some technical and financial support is available over several years of experimentation.
There is no general roadmap for PPB. Successful experiences in various countries have taken very different paths, but they all share the acknowledgment that:
In a barley breeding project undertaken by the International Center for Agricultural Research in the Dry Areas (ICARDA), Syria, promising parent lines were identified from both research station germplasm and farmers’ fields. Farmers from various villages were involved as partners from the beginning in deciding on breeding objectives, the number of varieties, plot size, the development of a common rating system, etc. Nine “host farmers” then planted over 200 varieties in small quantities on their farms and carefully recorded their observations in a field book throughout the growing cycle. Based on various categories of interest (e.g., tillering and grain size), each variety was given a numeric score, and a final score for each variety was assessed after harvest. Using this method of farmer evaluation and selection from an early stage of the breeding program, promising lines of barley were identified. In fact, some of them would have been rejected had professional breeders carried out the selection alone. The varieties released and disseminated after this PPB success were widely adopted and yields increased in areas where plant breeding had not been successful before.
Although the Syrian example was still “breeder-led,” i.e., initiated by the researchers, farmers in Honduras took the lead and, forming CIALs, requested new maize and bean plant material and technical assistance from a breeding facility. In some cases, breeding objectives were well defined by the farmers and specific crosses were provided accordingly by professional breeders. For example, CIAL members decided to send seeds of a popular, yet disease-susceptible bean landrace to a breeding station, where it was crossed with five resistant lines. Farmers then received seed of the F3 (third) generation, cultivated it, and carried out evaluation and selection up to the F8 generation, when a line was prepared for local release by the CIAL members. Throughout the PPB process, the farmers received agronomic advice and training from an NGO-funded agronomist. The resulting bean variety became one of the most popular varieties in the impact area. The CIAL has proved to be a very successful structure for PPB, empowering farmers throughout Latin America.
These two examples are meant to illustrate that PPB is feasible in very different environments and conditions. Numerous other examples exist (see recommended reading). Creating a system adapted to local conditions and capacities requires motivation and creativity, but can lead to rewarding outcomes.
 Ceccarelli, S., Grando, S., Tutwiler, R., Baha, J., Martini, A.M., Salahieh, H., Goodchild, A., Michael, M. , 2000 A methodological study on participatory barley breeding. I. Selection phase. Euphytica 111(2), , pp 91–104 One very successful PPB project, ICARDA’s participatory barley breeding project, showed that farmers can handle large numbers of unreleased varieties in their fields. Farmers select for different criteria than researchers, and their selections are at least as high yielding as those of breeders, especially off-station. |
|
 Witcombe, J.R., Joshi, A., Joshi, K.D., Sthapit, B.R. , 1996 Farmer participatory crop improvement. I. Varietal selection and breeding methods and their impact on biodiversity Experimental Agriculture 32(4), , pp 445–460 This pioneering article reviews various forms of PVS and PPB and compares their key features. It makes the case for PPB as an approach that can lead to more acceptable products in particular in marginal environments. |
|
 Sthapit, B.R., Joshi, K.D., Witcombe, J.R., 1996 Farmer participatory crop improvement. III. Participatory plant breeding, a case study for rice in Nepal. Experimental Agriculture 32(4), , pp 479–496 This article describes the initial results of a rice PPB program carried out in high mountain areas of Nepal. In addition to identifying better performing varieties, the research also increased crop diversity in the area. |
|
 Vernooy, R. , 2003 Seeds that give: participatory plant breeding. International Development Research Centre, Ottawa, Canada This concise booklet, which is also available in French, Spanish, Chinese, Arabic, Vietnamese, and Nepali, offers an overview of key learning and recommendations for practitioners based on a decade of experience in PPB. |
|
Witcombe, J.R., Joshi, K.D., Virk, D.S., Sthapit, B.R., 2011 Impact of introduction of modern variety on crop diversity In Lenné, J.M.,Wood, D. (editors)., Agrobiodiversity management for food security: a critical review. CABI International, London, UK, pp. 87–98 These authors discuss whether and how PPB changes crop diversity over a long period based on three case studies from India and Nepal for which good data are available. The study concludes that diversity has not decreased due to PPB. PPB has been successful in improving crop production, although sustaining PPB through institutionalization has not been achieved. |
|
|
Halewood, M., Deupmann, P., Sthapit, B.R., Vernooy, R., Ceccarelli, S. , 2007 Participatory plant breeding to promote farmers’ rights Bioversity Internationa, Rome, Italy This brief seeks to raise the profile of PPB as an additional, complementary, and powerful strategy for advancing the rights and interests of farmers. |
|
|
Ruiz, M., Vernooy, R. (editors), 2012 The custodians of biodiversity: sharing access and benefits of genetic resources Earthscan, London, UK, and International Development Research Centre, Ottawa, Canada This book illustrates a number of successful examples of collaboration between breeders and farmers though the lens of access to and benefit-sharing of plant genetic resources. Case studies include PPB experiences in China, Cuba, Honduras, Jordan, Nepal, and Syria. |
Here is a quiz that will help you test your newly acquired knowledge. Once you have covered the content sections and completed the assigned readings, please answer the Field-testing quiz.
In this module, you were introduced to a number of participatory approaches to test newly acquired germplasm. These approaches put farmers at the centre of the experimental cycle(s). Now your task is to select the most appropriate approach given the context in which you are working and to develop a plan for field-level implementation. Please, document the following:
The next module in our research resource box is Germplasm conservation. Let us begin!
