Skip to main content

research: Science for scaling

Lennart Woltering

Lennart Woltering is helping colleagues achieve more and lasting impact with their work. By finding practical applications for complex theories on scaling, he has been able to support project teams in formulating scaling ambitions, help provide clarity on what an “enabling environment” for an innovation means, and create a realistic picture of what a project can and cannot address and where strategic collaboration is required to exert influence.

Prior to joining CIMMYT, Woltering worked for ICRISAT in Niger (2005-2010) on local adaptation of innovations. From 2010-2016 he worked for a management consulting firm designing, implementing and backstopping projects for international donors.

Scaling farming innovations: what, why and how

Participants in the scaling workshop stand for a group photo with the trainers. (Photo: CIMMYT)
Participants in the scaling workshop stand for a group photo with the trainers. (Photo: CIMMYT)

How to scale up agricultural innovation in a sustainable and responsible manner? Through a workshop from the International Maize and Wheat Improvement Center (CIMMYT), a group from Mexico’s state of Chiapas learnt the basic principles of scaling. This knowledge will allow them to design and implement strategies that live well beyond the end of a project and that take into account the systems in which they operate.

Through presentations, group exercises and discussions, 23 people associated with the MasAgro project in Chiapas — technical advisors, coordinators and collaborators —  received training on scaling sustainable and responsible agricultural innovations.

The scaling process is relevant because oftentimes the size of a problem does not match the size of its solutions. To support scaling efforts, all relevant actors and entities must subscribe to a paradigm shift to achieve a positive sustainable impact for the greatest number of people.

Executing innovations in scaling requires a paradigm shift in all relevant actors and entities involved in order to achieve positive, sustainable impact that reaches the greatest number of people.

The workshop was divided into three sections. The first section was focused on basic concepts and the definition of realistic and responsible scaling goals. In the second, the group discovered the “Scaling Scan” tool. The third section was focused on the identification of opportunities and potential actions at the project level and through strategic collaborations.

The workshop “Scaling farming innovations: the what, for what and how” took place on April 24 and 25, 2019, in Tuxtla GutiĂ©rrez, Chiapas. The event was led by Lennart Woltering, international scaling expert at CIMMYT, MarĂ­a Boa, CIMMYT scaling coordinator, and Jorge GarcĂ­a, CIMMYT Hub Manager from the MasAgro project in Chiapas.

This training was held under the project for Strengthening Market Access for Small Producers of Maize and Legumes in Oaxaca, Chiapas and Campeche, which is financed by the Walmart Foundation.

Paswel Marenya

Paswel Marenya is a Senior Scientist at the International Maize and Wheat Improvement Center (CIMMYT). He is the global coordinator of the SIMLESA project, based in Nairobi.

Marenya holds an MSc in Applied Economics and a PhD in Natural Resources Policy and Management, both from Cornell University. His research focuses on maize-based smallholder systems in eastern and southern Africa, and on the analysis of pathways and impacts of technology adoption within the framework of sustainable intensification.

He has published research on farmer policy, farmer preferences for supporting technology adoption among maize farmers in Malawi, soil quality mediated returns to fertilizer, and the macro impacts of disease tolerant varieties. He is currently working on understanding the most effective investment priorities for strengthening climate-smart agriculture in eastern and southern Africa.

Taking Maize Agronomy to Scale in Africa (TAMASA)

Taking Maize Agronomy to Scale in Africa (TAMASA) is a 4-year project seeking to improve productivity and profitability for small-scale maize farmers in Ethiopia, Nigeria and Tanzania.

The overall purpose of TAMASA is to use innovative approaches to transform agronomy that:

  • Use available geospatial and other data and analytics to map maize areas, soil constraints, and actual and yields at different scale.
  • Work with service providers (i.e. input suppliers, government and private research and extension services, agro-dealers, and others) to identify and co-develop systems and applications that transform this data and information to useable products that support their businesses or programs to reach clients more effectively
  • Build capacity in national programs to support and sustain these approaches.

The core products and services of this project include:

  • Annual assessments and digital maps of maize growing areas, actual and attainable yields in core research areas or focal areas.
  • Decision-support tools for ex-ante spatial analysis, nutrient management, fertilizer formulation and variety selection.
  • Open-access databases of agronomic data.
  • Increased capacity in national programs and partners through in-country data science and software application training and mentoring.

Scaling to new heights in agriculture

How to scale? This question frequently comes up as projects look to expand and replicate results. In order to sustain enduring impacts for projects after their lifetime, agricultural programs are turning to scaling strategies. These strategies look beyond the numbers that are reached within a project and include sustainability and transformation beyond the project context. Methods and tools exist that help anticipate realistic and responsible scaling pathways.

The Scaling team at the International Maize and Wheat Improvement Center (CIMMYT), led by Lennart Woltering, drives the initiative to incorporate scaling principles into existing and developing projects to maximize impact.

Maria Boa recently joined the team as Scaling Coordinator. Last year Boa and Woltering participated in regional meetings on scaling in Morocco, Tunisia and Vietnam, which highlighted the need for better dissemination of information on how to approach scaling, in addition to its benefits.

Participants of the Tunisia workshop collaborate on a group exercise.
Participants of the Tunisia workshop collaborate on a group exercise.

According to Boa, one of the key messages highlighted throughout these events was that in order for scaling to take hold and be integrated into projects, “
there needs to be a shift in mindset to accept that change is complex and that most projects only address a fraction of the problem.” This is essential in using scaling to effectively support long-term results.

At a workshop in Tunisia organized by ICARDA, IFAD and CIMMYT in November 2018, many participants expressed interest in scaling strategy tools, but were puzzled on how to integrate them into their specific projects. Many determined that they were stuck developing scaling strategies in an outdated framework, or one that strictly focused on using technological innovations. One participant admitted that she was skeptical of scaling perspectives because many did not lie in her field of expertise.

The November 2018 CCAFS SEA Conference on Scaling in Vietnam provided a platform for the sharing and learning of experiences in the scaling world. Some of the key messages from the event included the importance of scaling agricultural innovations taking place in complex systems of agricultural transformation, and the necessity of joint cooperation from all involved stakeholders and their openness to taking on challenges as a way to support sustainable system change.

According to Boa, scaling is a process that heavily relies on strategic collaboration for lasting impact. “Projects often don’t take into account how they’re a part of a larger chain of potential change,” she says.

Already recognized as a sustainable leader within scaling, CIMMYT is looking to strengthen scaling efforts in order to foster a more enduring impact within CIMMYT projects and beyond.

Lennart Woltering presents at the CCAFS SEA Conference in Vietnam.
Lennart Woltering presents at the CCAFS SEA Conference in Vietnam.

Currently, the Scaling team at CIMMYT is conducting research on the “science of scaling” as it continues to function as a “help desk,” providing support integrating scaling principles in proposals and projects. Its primary role is to consider a project’s scaling needs and guide the development of an informed strategy to leverage efforts and resources. Boa hopes that by integrating responsible scaling approaches early on, projects can better balance the trade-offs associated with change.

Success in scaling is measured by a project’s enduring impact. However, stakeholders need more experience and capacity to see programs through to their end and be willing to monitor them beyond that lifespan. CIMMYT is developing and collecting the tools to support stakeholders with these specific capacities.

Developing a scaling strategy can also bring additional benefits: a discussion about scaling opens the door for raising awareness and fostering actions among different stakeholders towards system change and sustainable impact.

Scaling up mechanization in Bangladesh through partnerships

Over the last two decades, a significant number of rural Bangladeshis – especially youth – have migrated to urban centers, looking for higher paying jobs and an escape from agricultural labor. Reaper-DemoConor Riggs is the Global Director of Markets and Entrepreneurship at iDE. He says smallholder farmers in Southern Bangladesh are increasingly struggling to find and afford farm labor to help harvest crops and perform a variety of other on-farm activities.

Riggs says small-scale mechanization, such as two-wheeled tractors fitted with intensification machinery and surface irrigation pumps, can help farmers make up for this labor gap and increase productivity, while boosting the local economy by supporting micro- and small enterprises.

But as Riggs discussed at the recent Scale Up Conference at Purdue University, designing the perfect machine or technology is not enough to create sustainable, far-reaching impact. On the International Day for the Eradication of Poverty, we’re following up with him to learn more about the role of markets and partnerships in bringing small-scale mechanization to rural Bangladesh.

Q: Five years ago, CIMMYT and iDE co-designed and began implementing the USAID Cereal Systems Initiative for South Asia – Mechanization and Irrigation (CSISA-MI) project. What were the goals when you began?

Our goal was to establish a new industry for attachments in two-wheel tractors in agricultural mechanization, technologies like seeders, reapers and high-volume irrigation pumps for surface water. We wanted to help farmers access services through a fee-for-service model – small entrepreneurs buy machines and rent out those machines to farmers or directly provide that service themselves. There wasn’t really a market naturally growing for these machines in ways that included smallholders while being commercially viable, so we aimed to build it as best we could.

The results of this effort to date have been strong: 191,000 farmers can now access machinery services from a growing network of nearly 3,000 local microenterprise service providers, representing improved cultivation across 92,000 hectares in Southern Bangladesh. And we see abundant evidence that this market is scaling organically now that it’s established a model that works for both firms and farms.

Q: How did you create a market?

We incentivized several large conglomerates in the agri-business space to co-invest with us on several container-loads of these machines, which we imported from Thailand and China. We helped them find some early adopter dealers and local service providers who would actually buy them. Then we developed short-term smart subsidies to drive down the costs of supply chain development, accelerated customer adoption of the machines, and overall market growth.

An important aspect of our strategy is that we did not present these accelerating investments as typical subsidies; rather, we worked with our private partners to offer commercial discounts so that service providers and farmers Md Lalchan Shardar is a farmer and local service provider. The axial flow pump cuts the amount of irrigation time in halfwould recognize the true value of the product and the short-term opportunity to adopt the technology in its initial commercialization phase.

We first implemented this strategy with two leading firms in the market who concurrently launched a very proactive marketing campaign. Then we started pulling back those discounts overtime, year by year, as the initial partner firms found the market opportunity, and redirected this acceleration process with an additional group of interested companies that also wanted to enter the market in an inclusive manner.

Q: So the companies were benefiting from the discount?

Yes, but we created a lot of conditions. Essentially, the more project investment that was committed by the project to discount the cost of the machinery, the more we expected to see both cash and in-kind investment from those companies. In the end, about a dozen companies come into the game with about five that have really driven a lot of heavy investment.

Partnerships have been key throughout this project. What were the different strengths iDE and CIMMYT brought to the table?

iDE is a market development organization. We focus on market-based solutions, technology commercialization, last mile distribution, and market access. Fundamentally, we see our job as de-risking the market for companies to invest in lower income areas and empower the farmer and their family as both consumers and suppliers in the formal economy. To do this, we employ a lot of supply chain development, product re-design and most importantly, we develop networks of micro-entrepreneurs to serve the ‘missing middle’ between the formal and informal economies.

CIMMYT brings leading capabilities in linking science and practice, with an un-paralleled strength in understanding the agronomic rationale and the agronomic and economic combinations of the technologies as they’re applied on the ground.

CIMMYT knew what technologies were needed on the ground in Southern Bangladesh to genuinely improve productivity and efficiency in the face of changing economic circumstances, and understood how to apply them to real world conditions in alignment with market-based diffusion mechanisms. CIMMYT was also instrumental in working closely with the Bangladesh Agriculture Research Institute (BARI) and extension services, key government partners that helped us ensure market development was in alignment with public and social policy.

It can be difficult finding a synergy between two different organizations. Did you run into any challenges?  

CIMMYT and iDE have different specializations, and at the beginning, we had natural, friendly debates about how to best integrate them and achieve highly ambitious project objectives. But relatively quickly, we figured out how to learn from each other and synthesize our approaches for the best results. Both CIMMYT and iDE approached the partnership with a mission driven focus and a sense of constant, mutual respect for the value each partner brought to the table.

What do you see for the future? 

As for iDE, we’re excited to expand this successful partnership with CIMMYT to figure out how we can further replicate this success in other countries where we both work. While some of the market conditions in Bangladesh have provided us with unique opportunities for technology scaling in mechanization, we’re highly optimistic that the underlying partnership principles and management systems of CSISA-MI can be replicated in other programs and country contexts – even in ostensibly more challenging market environments.

The CSISA-MI project is funded by USAID.

Avoiding the next Aral Sea: Scaling responsibly

The Aral Sea was once the world’s fourth largest inland body of water. But in 1959, Soviet premier Nikita Khruschev unfurled a plan for industrialized agriculture across Central Asia. The government constructed irrigation canals to divert water from the Amu Syr and Amu Darya rivers, the two primary feeders for the Aral Sea, to thirsty cotton fields in Uzbekistan. Today, only about two-fifths of the sea remain. Evaporation exasperated by climate change and pesticide runoff have left the remaining body of water salty and polluted.

MSI's founder and president Larry Cooley presents at the Purdue Scale Up Conference 2018. (Photo: Rachel Cramer/CIMMYT)
MSI’s founder and president Larry Cooley presents at the Purdue Scale Up Conference 2018. (Photo: Rachel Cramer/CIMMYT)

The disappearance of the Aral Sea is a tragic story about scaling gone wrong. Larry Cooley, one of the top scaling experts in the world, describes scaling as the attempt to overcome a gap between the need for something and the extent to which that need is being met. In the case of the Aral Sea, the Soviet Union saw a need for more robust cotton production and decided to overcome the gap through large-scale irrigation.

They were successful in reaching their scaling ambition but at a high and unsustainable cost. Would Kruschev still go ahead with his development scheme if he knew it would cause irreversible ecological damage in the future? Would he still prioritize high cotton yields if he knew it would decimate the local fishing industry and leave thousands unemployed?

At the recent Scale Up Conference at Purdue University, over 200 researchers and practitioners gathered to discuss effective approaches to scaling up agricultural technologies and innovations in the developing world. The tagline read “Innovations in agriculture: Scaling up to reach millions.” Several of the presenters, however, argued development organizations should think about potential tradeoffs before trying to reach the biggest impact.

Finding the optimal scale

CIMMYT’s scaling advisor Lennart Woltering (left) and mechanization specialist Jelle van Loon led a session. (Photo: Rachel Cramer/CIMMYT)
CIMMYT’s scaling advisor Lennart Woltering (left) and mechanization specialist Jelle van Loon led a session. (Photo: Rachel Cramer/CIMMYT)

CIMMYT’s scaling advisor Lennart Woltering and mechanization specialist Jelle van Loon led a session on the opportunities and challenges to scaling two-wheeled tractors in Africa, Asia and Latin America. Van Loon explained how mechanization can decrease labor costs, improve livelihoods and help farmers stay locally and internationally competitive, but he acknowledged a few potential downsides. Small tractors of this kind require fossil fuels and maintenance, and introducing mechanization to a rural community has the potential to displace jobs and shift gender roles.

Woltering explained a new tool can help researchers and development organizations think through these tradeoffs in a systematic way. The Scaling Scan, which he developed in a collaboration with The PPPLab, guides users through a series of questions and prompts them to reflect on what scaling means, what it takes to take a project to scale and what the unintended consequences could be in a particular context.

Lennart Woltering (second from left) presents the "ingredients" of the Scaling Scan tools during one of the sessions of the Purdue Scale Up Conference. (Photo: Rachel Cramer/CIMMYT)
Lennart Woltering (second from left) presents the “ingredients” of the Scaling Scan tools during one of the sessions of the Purdue Scale Up Conference. (Photo: Rachel Cramer/CIMMYT)

The first step of the Scaling Scan is “Defining a realistic scaling ambition.” It contains a responsibility check, prompting users to consider how an intervention could affect power equity and natural resources if that scaling ambition is indeed reached. “We tried to make this check as simple as possible, but still have people anticipate what unintended consequences their scaling effort might have ten years down the line,” said Woltering.

The responsibility check includes questions like: Who are the winners and who are the losers when the innovation is adopted at a large scale? Will the scaling of the innovation affect the availability of important natural resources, such as water and land?

Woltering emphasized that development organizations should try to identify the scale that optimizes tradeoffs. “We want people to be aware that bigger is not always better,” he said.

“You might think you’re benefitting the irrigation farmers, but at the same time, the fishermen or other people might be paying the price for that,” Woltering explained. “If you’re only focused on those irrigation farmers and not the whole system, it’s easy to think, ‘Oh, we’re doing a fantastic job,’ when you’re not.”

The reasons to scale up responsibly

At the conference, Tricia Wind and Robert McLean of the International Development Research Center (IDRC) presented some of their lessons learned about responsible scaling.

“If you’re working on the problem at different scales, you need to think about the problem differently and think about the solutions differently,” said Wind. “The first principle is thinking about what scale you are starting with and what the optimal scale would be for the problems that you’re focused on solving.”

The second principle is the justification for scaling. “So stepping back from the how and thinking about the why,” she explained. “What difference would this make?” Similar to the responsibility check in the Scaling Scan, the second principle explores the issue of equity. Who would be reached by this solution, and who would be left out or even negatively affected by it?

The third principle is about coordination. McLean said, “This is about accepting that all scaling happens in a system. Are the alternative solutions? How do you displace solutions that might already exist if you try to scale something? What about the cultural norms and the institutions that exist in the area where you’re scaling, and how do you coordinate to scale responsibly?”

The fourth principle is dynamic evaluation. Maclean said an organization should learn as it scales. “It’s never going to be a 1-2-3 step process that’s going to get you from innovation to impact at scale,” he explained. “Scaling itself is also an intervention. So you have your intervention you’re trying to scale, and as you scale, systems change.”

Participants and panelists of the Scale Up Conference pose for a group photograph. (Photo: Courtesy of Purdue University)
Participants and panelists of the Scale Up Conference pose for a group photograph. (Photo: Courtesy of Purdue University)

Johannes Linn, Nonresident Senior Fellow with the Brookings Institute and another one of the world’s top scaling experts, emphasized, “Scaling is not a linear process. It is iterative with feedback loops to learn and adapt.”

During the opening reception, Woltering and van Loon congratulated Seerp Wigboldus, a senior advisor and researcher with Wageningen University, on his recently completed PhD thesis, published as a book: To scale, or not to scale – that is not the only question.

Someone asked, “What do you do if 40 people are going to be harmed by an intervention while 50 people benefit?” Wigboldus replied, “Well, unfortunately, there’s no formula for this kind of thing. There will always be tradeoffs, but hopefully we can get people to slow down a bit. We need to be transparent and justify our decisions.”

Nearly all of humanity’s greatest challenges originate from the scaling of innovations. The depletion of the Aral Sea in order to scale cotton production is just one example. Climate change and industrialization is another. By adopting a responsible scaling approach, the agricultural development sector can minimize negative impacts and side effects and seek optimal solutions.

The full version of the Scaling Scan contains detailed practical information on how and when to use this tool. A condensed, two-page version is also available. We also recommend the companion Excel sheet, which generates average scores and results automatically.

This work is supported by the German Development Cooperation (GIZ) and led by the International Maize and Wheat Improvement Center (CIMMYT).

Scaling Scan: A simple tool for big impact

Eleven years ago this week, Apple Inc. released the iPhone. While it was not the first smartphone on the market, industry experts often credit the iPhone’s groundbreaking design with the launch of the mobile revolution. The device, its competitors and the apps that emerged with them have changed how over two billion people interact with the world on a daily basis.

The success of this revolution, however, goes far beyond the actual technology. At the International Maize and Wheat Improvement Center (CIMMYT) outside Mexico City, scaling expert Lennart Woltering points to a smartphone lying on his desk.

“We have to remember that this phone is just hardware. It is useless if you don’t have a network connection or an outlet in your house with electricity,” he says.

Woltering joined CIMMYT last year as part of the German Development Cooperation’s effort to aid the scaling-up of agricultural innovations. New, improved seeds, small-scale machinery and conservation practices can all play a role in achieving several of the Sustainable Development Goals, but Woltering says many other non-technological factors, such as markets and policies, can prevent these innovations from having significant impact.

Roadside vendor sells roasted maize cobs in Kenya. (Photo: P.Lowe/CIMMYT)
Roadside vendor sells roasted maize cobs in Kenya. (Photo: P.Lowe/CIMMYT)

“Many research institutes and nongovernmental organizations tend to focus on technology as the solution for everything,” he says. “But we find that 9 out of 10 cases, limiting factors have more to do with financing not being available to people, or poor policies that are hampering the adoption of technology.”

For example, CIMMYT has many initiatives in South Asia to promote conservation agriculture. Adopting no-till practices can help reduce erosion and improve soil health for better yields, but farmers who make this transition often need access to a different kind of machinery, such as the Happy Seeder, to plant their seeds. If government subsidies exist for conventional rototillers but not for the Happy Seeder, it is difficult to persuade farmers to make that economic sacrifice.

“It is a completely different ballgame in the real world, and you have to be honest about whatever fake reality you created in your project,” says Woltering.

Projects are designed in a very controlled way. They have a fixed budget and a fixed end date, and they are often shielded from the social and economic complexities that can propel or hinder an innovation from scaling.

“So if a donor says, ‘We want two million people to be reached,’ well, how are you going to do that? That’s where the Scaling Scan can help,” says Woltering.

Extension agents in Mexico use the Scaling Scan. (Photo: L. Woltering/CIMMYT)
Extension agents in Mexico use the Scaling Scan. (Photo: L. Woltering/CIMMYT)

The Scaling Scan helps an individual analyze, reflect on, and sharpen one’s scaling ambition and approach through a series of questions and prompts. It focuses on ten scaling ‘ingredients’ that need to be considered (e.g. knowledge and skills, public sector governance, awareness and demand) to reach the desired outcome.

“The Scaling Scan helps you figure out what exactly is required, what is possible, and what bottlenecks exist that you need to address in your strategy,” Woltering says.

Woltering collaborated with The PPPLab, a consortium of four Dutch institutes, to release the first version of the Scaling Scan last year. They tested it with project teams in the Netherlands, Mexico, India, Nepal and Kenya, and based on the feedback, they are now releasing a second version, which is available here.

In the trials with the first Scaling Scan, some teams realized the results they wanted to achieve were too ambitious given the circumstances. For other teams, it helped them clarify exactly what they wanted to achieve.

“Having a project objective is not enough to internalize the main goal,” says Woltering. “It also changes over time, especially if it’s a long-term project. The scaling scan can be good for an annual checkup.”

Woltering emphasizes that successful scaling requires multidisciplinary collaboration.

“If you only have a team of agronomists, you will not reach a scale of millions you want to achieve. If you only have a team of policy experts, you will not succeed,” he says. “There are professionals that can really help and add value to what we are doing.”

“It’s hard to get an agronomist and an economist in the same room together, but we’re not going to change the world if we don’t work together with others who have their specific specialty or expertise,” he says.

The Scaling Scan also includes a responsibility check through some very simple but strategic questions.

“Every system has its pros and cons – some people benefit, some do not. Some have power, some do not,” says Woltering. “So what does it mean if your innovation goes to scale? Maybe there’s a whole new power dimension.”

Successfully scaling something may have unintended consequences. There are always tradeoffs and resistance to change. Woltering says the responsibility check can help actors in the development sector to think through these questions and consider what the possible outcomes could be.

For more explanation on how and when to use the tool, we invite you to download the Scaling Scan (also available in Spanish) which contains detailed practical information. We recommend the Excel sheet (also available in Spanish) to have the average scores and results generated automatically. A condensed, two-page PDF is also available.

This work is supported by the German Development Cooperation (GIZ) and led by the International Maize and Wheat Improvement Center (CIMMYT).

Overcoming gender gaps in rural mechanization

A new publication suggests strategies to improve rural women's access to agricultural machinery. Photo: CIMMYT/ Martin Ranak
A new publication suggests strategies to improve rural women’s access to agricultural machinery. Photo: CIMMYT/ Martin Ranak

A new research note published for International Women’s Day, details current gender gaps in rural mechanization in Bangladesh, and outlines plans to overcome these challenges.

Using simple technologies, such as multi-crop reaper-harvesters can reduce the time farmers spend harvesting by up to 80 percent and can reduce the costs of hiring field labor by up to 60 percent. The problem is that women may face cultural constraints to working in the field, running machinery service provision businesses, and do not have equal access to financing, which is a huge barrier, as the technologies can cost $500-2000 up front.

The authors suggest a number of gender-balanced approaches to scaling-out technologies such as use of targeted, selective and smart subsidies and access to finance to women-headed households, methods to spread investment risks, and prioritizing joint learning, with husbands and wives attending field courses together and jointly developing business plans.

View the new research note here.

The research note is a result of joint efforts between the USAID/Washington and Bill and Melinda Gates Foundation supported Cereal Systems Initiative for South Asia (CSISA), the USAID/Bangladesh CSISA – Mechanization and Irrigation Project, and the the USAID/Washington funded USAID funded Gender, Climate Change, and Nutrition Integration Initiative (GCAN) project, all of which involve collaborations between the International Maize and Wheat Improvement Center, the International Food Policy Research Institute, International Development Enterprises, the International Rice Research Institute and the CGIAR Research Program on Climate Change, Agriculture and Food Security.