Maricelis Acevedo is Associate Director for Science for the Delivering Genetic Gain in Wheat Project at Cornell University. She leads the Climate-resilient Plants research team for Ceres2030, one of eight teams that has used evidence synthesis methods combined with machine learning to assess the scope of existing agricultural research — including that research not normally captured by Western databases.
“When we do an evidence synthesis of this type, we frequently look within our school of thinking, and at the authors that we know, but when you provide the machine with a set of keywords, with no bias for author, or institution, or impact – you have a much more unbiased net,” Acevedo said. “I think the machine learning, in the future, will be key in doing this type of work.”
All of the Ceres2030 teams focused on interventions that had solid evidence of beneficial impact for smallholder farmers in developing countries. For Acevedo and her team, that meant looking at crop-specific research that helped farmers cope with the challenges of climate change, such as wheat varieties that were more tolerant to high temperatures, rice that could survive more time being submerged, and maize that could handle drought conditions. The team looked at research from cutting-edge genetic research labs, and from studies on heirloom varieties.
“It may be that farmers in one region have been selecting for drought tolerance for hundreds of years. And maybe those seeds could be useful in another area that is now experiencing drought,” Acevedo said. “Farmers are very knowledgeable — if we can connect this knowledge to other farmers, that will also strengthen climate resilience.”
Acevedo’s team was interested in knowing not just which seeds were superior, but how, why, and if the farmers who could benefit from them were actually adopting and planting them. The diverse Ceres2030 research team — plant pathologists, plant breeders, geneticists, specialists in seed systems, and data managers— helped to keep that goal front and center.
“Having this diverse team was key to be able to really dig deep on the topic and see what was relevant to the smallholder farmers we want to serve,” Acevedo said.
The research team discovered that the single most important factor in a smallholder farmers’ adoption of a new variety is extension and education.
“That could be surprising to some people because it’s not something new. It’s not something that requires a new technology. But it’s something we know is extremely important. For farmers to take in the new research that has bene developed, they need help adopting it in their farm,” Acevedo said. “This intervention of providing extension, education, training, about climate change, and about potential solutions of how to adapt, is key.”
Also crucial is having a strong seed system, whether public or private, formal or informal, so farmers can access new seed varieties, she said.
“Being part of this project allowed me to be hopeful that we can communicate as scientists with donors and policymakers what the science says and how to make sense of it, so we can make real impacts,” Acevedo said. “I’ll be really happy if the next generation of investments are based on data and science and evidence.”
